The Golgi Apparatus: State of the art 110 years after Camillo Golgi's discovery

SpringerWienNewYork

Alexander A. Mironov and Margit Pavelka (eds.)

The Golgi Apparatus State of the art 110 years after Camillo Golgi’s discovery

SpringerWienNewYork

Prof. Dr. Alexander A. Mironov Consorzio Mario Negri Sud Laboratory of Intracellular Traffic Department of Cell Biology and Oncology, S. Maria Imbaro (Chieti), Italy

Prof. Dr. Margit Pavelka Department of Cell Biology and Ultrastructure Research Institute of Histology and Embryology Center for Anatomy and Cell Biology Medical University of Vienna Vienna, Austria € tzung des Bundesministeriums fu € r Wissenschaft und Forschung Gedruckt mit Unterstu in Wien

This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically those of translation, reprinting, re-use of illustrations, broadcasting, reproduction by photocopying machines or similar means, and storage in data banks. Product Liability: The publisher can give no guarantee for all the information contained in this book. This does also refer to information about drug dosage and application thereof. In every individual case the respective user must check its accuracy by consulting other pharmaceutical literature. The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use.  2008 Springer-Verlag/Wien Printed in Austria SpringerWienNewYork is a part of Springer Science þ Business Media springer.at Typesetting: Thomson Press (India) Ltd., Chennai Printing: Holzhausen Druck & Medien, 1140 Wien Printed on acid-free and chlorine-free bleached paper SPIN: 12054541 With 106 partly coloured figures Library of Congress Control Number: 2008932528 ISBN 978-3-211-76309-4 SpringerWienNewYork

Preface

*

V

Preface In 2008, we celebrate the 110th anniversary of the first description of the complex apparato reticolare interno by Camillo Golgi in the Bollettino della  Medico-Chirurgica di Pavia. The biography of Camillo Golgi, and the Societa Golgi apparatus history have impressively been described by Paolo Mazzarello in his book “The Hidden Structure. A Scientific Biography of Camillo Golgi” (Oxford University Press, 1999). During the 20th century, Camillo Golgi’s discovery had a changing up and down and up history, timely being assessed as an artefact, and then again coming into the centre stage of cell biologic research. Today, it is well established that the Golgi apparatus constitutes a main crossroads in the intracellular transport routes of the biosynthetic, endocytic, and recycling systems. During the past decades, multiple new discoveries contributed to the understanding of the organization and the functions of the complex organelle (see Chapter 1.1). In 1997, the excellent book about the Golgi apparatus edited by J. Roth and E. Berger provided a comprehensive summarizing presentation of the state of research 100 years after the first description of the organelle. Now, after further 10 years, it is necessary to summarize again what it is known about the complex organization of Camillo Golgi’s apparatus. Our book is an attempt to bring together multiple new results obtained by different techniques, and addressing different aspects of the Golgi apparatus and intracellular transport. We are hopeful that the presentation of the state of the art 110 years after Camillo Golgi’s discovery of the complex apparato reticolare interno will lead to an improved understanding, novel insights, and new perspectives for future research.

Acknowledgements The editors cordially thank all authors of the Chapters, all our colleagues involved in the works presented in the Chapters of this book written by us and by others. We thank the Springer Company for the possibility to publish this book and in particular we thank Mag. Franziska Brugger, Mag. Angelika Heller, and Mrs. Ursula Szorger for a huge help in our work. A.M. is especially thankful to Dr. A. Fusella and D. Gaindomenico for the technical help and to Chris Berrie and Raman Parashuraman for critical reading of the manuscripts, in which A.M. is a co-author. A. A. Mironov and M. Pavelka S. Maria Imbaro/Vienna, May 2008

Movies can be viewed online at: www.springer.com/springerwiennewyork/lifeþsciences/book/978-3-211-76309-4

Contents

*

VII

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1

1. General considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1.1. The Golgi apparatus and main discoveries in the field of intracellular transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 1.2. The Golgi apparatus as a crossroads in intracellular traffic . . . 16 2. Main machineries operating at the Golgi apparatus . . . . . . . . . . . . . 2.1. SNAREs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Rabs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. COPII . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4. COPI: mechanisms and transport roles . . . . . . . . . . . . . . . . . . . 2.5. Arfs and Arls: models for Arf family members in membrane traffic at the Golgi . . . . . . . . . . . . . . . . . . . . . . . . . 2.6. COG complex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.7. The TRAPP complex. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.8. The role of Ca2þ in the regulation of intracellular transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.9. Golgi glycosylation enzymes . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.10. Nucleotide sugar transporters of the Golgi apparatus. . . . . . . 2.11. Luminal lectins. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.12. The Golgi ribbon and the function of the golgins . . . . . . . . . . 2.13. Functional cross talk between membrane trafficking and cell signalling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.14. The role of the cytoskeleton in the structure and function of the Golgi apparatus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.15. The dynamin–cortactin complex as a mediator of vesicle formation at the trans-Golgi network . . . . . . . . . . . . . . . . . . . 2.16. The geometry of organelles of the secretory pathway . . . . . . 3. Main transport steps . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. ER-to-Golgi transport . . . . . . . . . . . . . . . . . . . . . 3.2. Intra-Golgi transport . . . . . . . . . . . . . . . . . . . . . . 3.3. Structure and domain organization of the trans-Golgi network . . . . . . . . . . . . . . . . . . . . . . 3.4. Golgi-to-PM transport . . . . . . . . . . . . . . . . . . . . . 3.5. Protein transport from the trans-Golgi network to endosomes . . . . . . . . . . . . . . . . . . . . . . . . . . .

41 43 66 78 87 106 120 130 143 161 190 207 223 247 270 301 314

. . . . . . . . . . . 331 . . . . . . . . . . . 333 . . . . . . . . . . . 342 . . . . . . . . . . . 358 . . . . . . . . . . . 375 . . . . . . . . . . . 388

VIII

*

Contents

3.6. 3.7. 3.8. 3.9. 3.10. 3.11. 3.12. 3.13. 3.14. 3.15.

The transport of soluble lysosomal hydrolases from the Golgi complex to lysosomes . . . . . . . . . . . . . . . . . . . . . . . . . . . Transport of lysosomal membrane proteins from the Golgi complex to lysosomes . . . . . . . . . . . . . . . . . . . . . . . . Retrograde endosome-to-TGN transport . . . . . . . . . . . . . . . . . Retrograde plasma membrane-to-Golgi apparatus transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interactions between endocytosis and secretory transport . . Origins of the regulated secretory pathway . . . . . . . . . . . . . . Secretion and endocytosis in endothelial cells . . . . . . . . . . . . . Formation of mucin granules . . . . . . . . . . . . . . . . . . . . . . . . . . Golgi apparatus and epithelial cell polarity . . . . . . . . . . . . . . . Golgi apparatus inheritance . . . . . . . . . . . . . . . . . . . . . . . . . . .

4. Peculiarities of intracellular transport in different organisms . . 4.1. Features of the plant Golgi apparatus . . . . . . . . . . . . . . . 4.2. Yeast Golgi apparatus . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3. Morphodynamics of the yeast Golgi apparatus . . . . . . . . 4.4. Structure and function of the Golgi organelle in parasitic protists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5. Evolution of the Golgi complex . . . . . . . . . . . . . . . . . . . .

.... .... .... ....

402 414 425 459 475 485 520 535 563 580 609 611 623 630

. . . . 647 . . . . 675

5. General conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 693 Contributors in alphabetical order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 697 Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 706

Introduction

*

1

Introduction Alexander A. Mironov and Margit Pavelka

All cells secrete a diversity of macromolecules to modify their environment or to protect themselves. On the other hand, there is the necessity to replace membrane proteins and lipids that are being constantly degraded in compartments of the secretory and endocytic pathways. Therefore, eukaryotic cells synthesize proteins for either export (secretion) or delivery to the secretory and endocytic compartments and to the plasma membrane (PM) for replacement of degraded proteins and lipids. The synthesis is carried out by ribosomes attached to the cytosolic surface of the endoplasmic reticulum (ER). The synthesis of most of cellular lipids and all fatty acids also occurs in the smooth ER. During or after the synthesis, the polypeptide chains containing transmembrane domains composed of hydrophobic amino acids are inserted into the ER membrane, whereas soluble proteins are transferred into the ER lumen. After the cleavage of their leading hydrophobic signal peptides and after protein folding both the groups of proteins are transported along the secretory pathway. The Golgi apparatus (GA) is the central station along this pathway. While passing through the GA, proteins and lipids undergo posttranslational modifications (mainly glycosylation by Golgi glycosidases and transferases) and sorting. The transport of proteins and lipids from the ER to their destinations may be carried out in several ways: via the dissociation mechanism, the progression mechanism, and/or the lateral diffusion mechanism (see Chapter 1.2). Over the past 30 years, the field of intracellular traffic has seen tremendous advances towards the identification of the relevant molecular machineries. Substantial progress in the isolation, cloning, and characterization of proteins involved in intracellular transport and its regulation, as well as in deciphering corresponding molecular events was achieved. Proteins involved in budding, fission, fusion, and sorting have been discovered, and, in some cases, a picture of how such proteins are assembled and work has been glimpsed. In contrast, perhaps surprisingly, a satisfactory understanding of how transport occurs in vivo at the organelle level has not been achieved, and the general picture of this process remains obscure. As a consequence, our present view of the overall mode of intracellular traffic (the “physiology” of traffic) in living cells is rather poorly developed. Although in vitro reconstitution experiments have been crucial in establishing the minimum number of components required for carrier budding using biochemical techniques, a more critical evaluation of intracellular transport could be complemented by in vivo experiments. The reasons for this lag are both technological and conceptual. Technological, because intracellular traffic is an essentially dynamic event in time and

2

*

A. A. Mironov and M. Pavelka

space. Yet, methods to study its dynamic features have been lacking until very recently. The recent advent of green fluorescent protein (GFP) technology has now partially filled this gap. Conceptual, because the field has been dominated until very recently by the model of intracellular traffic by anterograde vesicles. The vesicular scheme has been very successful in providing a framework to integrate an enormous amount of biochemical and genetic evidence collected over the past two decades into a coherent picture. In the past years, the progress in development of new scientific methods accelerated. New microscopic and biochemical methods, new scientific instruments together with the development of molecular tools has given the possibility to study biochemical reactions and molecules in single cells with a resolution impossible to achieve before. Significant progress was made in the development of new synchronization protocols suitable for study of transport of several cargoes. On the other hand, the complete deciphering of the full genetic code of yeast, humans, plants and several other species together with the development of proteomics provided a significant number of new proteins and details on protein–protein and protein–lipid interactions involved into the traffic. In the literature, there are thousands of details regarding protein interactions, their sorting signals, and the effects of their inhibition and deletion. As such, it seems that the field of intracellular transport appears to be facing a serious crisis. There are now so many proteins and inter-protein interactions involved that it has become almost impossible to follow and understand the meaning of all of these details. The main idea of this book is thus to collect the full range of expertise and to examine the problems from different points of view and with different approaches. This book is devoted to the molecular mechanisms of morpho-functional organization of the GA and summarizes most of new data related to the GA. The book is a collection of chapters written by different groups and therefore expresses different views especially on the mechanisms of traffic. The possibility to follow the evaluation of the intracellular transport from different points of view and on the basis of different expertises could help to resolve this contradiction in the field. There are several levels in the book. The main is the description of cell physiology with emphasis on the physiology of intracellular transport. The second level is the presentation of the morphology in wide term including light microscopy, analysis of live cells and so on. Finally, there are several chapters devoted to the molecular mechanisms involved in different physiological processes related to intracellular transport. In this book, we could not exclude completely some degree of overlapping. First, because the authors have different opinions about models of transport, and the second, it was necessary to illustrate some ideas from different points of view. We apologize to colleagues, whose relevant work has not been mentioned because of space limitations and focusing on work published most recently.

Introduction

*

3

In the first part of the book there are two chapters. One is an overview of the history of this field of cell biology. The second one is a morpho-functional overview of the main steps of intracellular transport including functional organization and architecture of the GA. In the second part of the book, main machineries operating along the secretory pathway are presented. These include SNAREs, Rabs, COPII, COPI, ARFs, ARFGEFs and ARLs, COG, TRAPP, dynamin and cortactin, Golgi enzymes and sugar transporters, ERGIC-53, and other luminal lectins, Golgins and Golgi matrix proteins. Lipids and lipid signalling, as well as common signalling mechanisms will be assessed. Additionally, the role of calcium and other ions in the regulation of intracellular transport, and the role of the cytoskeleton in Golgi function will be described. In the third part of the book, different transport steps of intracellular transport will be evaluated, namely, ER-to-Golgi transport, intra-Golgi transport, Golgi-to-PM transport, Golgi-to-endosome transport, transport of lysosomal enzymes, retrograde endosome-to-TGN transport, and retrograde PM-to-Golgi transport. Mechanisms of regulated secretion will be explained in a separate chapter and in particular, mechanisms involved in formation of mucin granules are described. Interactions between endocytosis and secretory transport, the relationship between the GA and cell polarity, as well as structure and domain organization of the trans-Golgi network, and the questions of GA formation and inheritance will be discussed in separate chapters. Finally, in the last part of the book the reader will find some aspects about the peculiarities of intracellular transport in different organisms: plants, yeast, and protists. A separate chapter is devoted to the endomembrane ultrastructure and dynamics in yeasts. Finally, the models of Golgi evolution will be discussed in the final chapter of this part. This book is intended for cell biologists and histologists, who work with students, and also for scientists working in other fields of biology as well as for students per se. The most important item for teaching is the understanding of not a single or several mechanisms but the comprehensive view upon the full drama of development of scientific models. The readers should understand the logic of model replacements. Therefore, the aim of the book is to make the field of intracellular transport more understandable by keeping it as simple as possible, but also as full as possible, and at minimal cost. Finally, we would like to mention the question of terminology. Since the original term was “Golgi apparatus”, this term is used in most of the chapters. However, both terms, Golgi apparatus and Golgi complex, are customary today.

General considerations

The Golgi apparatus and main discoveries

*

7

The Golgi apparatus and main discoveries in the field of intracellular transport Alexander A. Mironov and Margit Pavelka

In this chapter, we summarize important findings in the field of intracellular transport, which have considerably contributed to the understanding of the function and organization of the Golgi apparatus (GA). It is not possible to mention all authors in this huge field. We apologize for gaps and incompleteness, and are thankful for suggestions and corrections. The GA is named after its discoverer Camillo Golgi, who first described the complex apparato reticolare interno in 1898 (Golgi 1898a,b; reviewed € scher 1998). Although Camillo Golgi had presented his by Berger 1997; Dro discovery convincingly, for a long time his data have been considered as an artifact of cell staining (Farquhar and Palade 1981). Only after the electron microscopic confirmation of the existence of the GA in cells by Dalton in 1951, scientists started to believe in its reality. Therefore, we will not list the discoveries within the area of intracellular transport made in the time, before the existence of the GA was confirmed electron microscopically. However, the names of A. Negri, H. Fuch, A. Perroncito, S. Ramon y Cajal, D.N. Nassonov, R.H. Bowen, G.S. Carter, H.W. Beams and R.L. King, V.M. Emmel, H.W. Deane and E.M. Dempsey, W.C. Schneider et al. should be mentioned, because they have considerably contributed to the understanding of the Golgi function (reviewed in Berger 1997). Here, we want to address most important discoveries within the area of intracellular transport after 1951 (Table 1). Additionally, we would like to mention further important contributions to this field. The hypothesis of lipid rafts was proposed and developed by van Meer and Simons. The Lodish group made the invention of the synchronization of the transport of cargoes. The role of lectins in ER-to-Golgi transport was discovered by H.-P. Hauri. The most important contribution to the characterization of Rab machinery (although in the endocytic pathway) was made by M. Zerial. W. Hong, R. Sheller and R. Jahn made important contributions to the understanding of the function of the SNARE machinery. R. Schekman and W. Balch deciphered the functions of the COPII coat. A. Rambourg, Y. Clermont, G. Griffiths, A. Staehelin and K. Howell made significant contributions to the 3D-analysis of the GA in different cell types. J. Slot and H. Geuze provided new insight into the morphology of the endocytic system and its interaction with exocytosis. The important contribution into the analysis of the kinetics of the plant GA was made by C. Hawes. The characterization of the 3D-structure of several proteins important for intracellular transport, and protein coat complexes in their crystal state is linked with W. Balch and J. Goldberg’s names. We apologise again for possible

8

*

A. A. Mironov and M. Pavelka

Table 1. The Golgi apparatus and main discoveries in the field of physiology of intracellular transport 1898 1951 1961 1964 1964 1964 1964 1966

1966 1967 1969 1976 1977 1980 1981 1982

Discovery of the GA Confirmation of the presence of the GA (Dalton 1951) The regional distribution of the thiamine-pyrophosphatase activity within the GA (Novikoff and Goldfischer 1961) The trans ER (Novikoff 1964; Novikoff et al. 1964) GERL concept (Novikoff 1964)  and Mollenhauer 1964) Isolation of Golgi membranes from cells (Morre The process of sulphation in the GA (Godman and Lane 1964) The sugar–nucleotide transport from the cytosol to the Golgi lumen across the Golgi membranes, the role of the GA in glycosylation (Neutra and Leblond 1966) The origin of lysosomes and the function of clathrin-coated vesicles during protein absorption (Bainton and Farquhar 1966; Friend and Farquhar 1967) The intracellular transport (Jamieson and Palade 1967a,b)  et al. 1969) Galactosyltransferase as a Golgi marker (Whur et al. 1969; Morre Isolation of clathrin-coated vesicles (Pearse 1976) The PM-to-Golgi transport of the endogenously added marker (Herzog and Farquhar 1977) M6P-mediated sorting of Golgi enzymes at the GA (Tabas and Kornfeld 1980) Clathrin-coated buds in the trans side of the GA (Griffiths et al. 1981) Immunocytochemical localization of galactosyltransferase (Roth and Berger 1982) Topology of N-glycosylation (Dunphy and Rothman 1983) Reconstitution of intra-Golgi transport in vitro (Balch et al. 1984) The 15 C temperature block (Saraste and Kuismanen 1984) Clathrin-independent endocytosis (Moya et al. 1985; Sandvig et al. 1985)

1983 1984 1984 1985 1985– 1987 The mitotic form of the GA and mechanisms of mitotic Golgi transformation in animal cells (Featherstone et al. 1985; Lucocq et al. 1987) 1986 The COPI-coated vesicles and characterization of molecular mechanisms involved into the function of COPI coat (Orci et al. 1986; Serafini et al. 1991) 1986 The structure and function of the TGN and the 20 C temperature block (Griffiths and Simons 1986) 1987 KDEL-signal for the retention of luminally located proteins (Munro and Pelham 1987) 1989 BFA was applied for the study of intra-Golgi transport (Doms et al. 1989; Lippincott-Schwartz et al. 1989) 1990 SNAREs (Newman et al. 1990) 1990 The main genes involved in intracellular transport, the genetic evidence in favour of the vesicular model of the transport in yeast (Kaiser and Schekman 1990) 1991 A Golgi retention signal in the membrane-spanning domain (Swift and Machamer 1991) 1993 The role of oligomerization for the retention of Golgi enzymes (Weisz et al. 1993) 1993 The role of PM-derived signalling for intra-Golgi transport (De Matteis et al. 1993) 1994 Golgi matrix (Slusarewicz et al. 1994) and cis-Golgin, GM130 (Nakamura et al. 1995) 1994 COPI-dependent retrieval sorting signals (Cosson and Letourneur 1994)

The Golgi apparatus and main discoveries

*

9

Table 1. (Continued)

1994 1996 1996 1997 1997 1997 1998 1998 1998 1999 2001 2003 2003 2004 2006 2006 2007

COPII coat. Isolation of COPII-dependent small vesicles in cell-free system (Barlowe et al. 1994) Application of GFP-technology for the study of the GA in living cells (Cole et al. 1996) Characterization of the ER exit sites (Bannykh et al. 1996) The AP3 and AP4 coats (Dell’Angelica et al. 1997, 1999) Characterization of ER-to-Golgi transport carriers in living cells (Presley et al. 1997; Scales et al. 1997; Mironov et al. 2003) Characterization of post-Golgi transport carriers in living cells (Wacker et al. 1997; Hirschberg et al. 1998; Polishchuk et al. 2000) Intra-Golgi transport of large cargo aggregates (Bonfanti et al. 1998) The role of endocytic TGN in the formation of the most-trans Golgi cisterna (Pavelka et al. 1998) Discovery of R- and Q-SNAREs (Fasshauer et al. 1998) Tomographic reconstruction of the GA (Ladinsky et al. 1999) The concentration of regulatory secretory proteins within the Golgi cisternae (Oprins et al. 2001) The understanding of the evolution of small GTPases had changed the model kely 2003) of the Golgi evolution (Je Characterization of Golgi-to-apical PM transport carriers in living cells (Kreitzer et al. 2003) Intercisternal connections in transporting GA (Marsh et al. 2004; Trucco et al. 2004) Characterization of the Golgi-to-endosome carriers in living cells (Polishchuk et al. 2006) The role of GM130 in the maintenance of the Golgi ribbon (Puthenveedu et al. 2006) The role of ER-to-Golgi transport in the maintenance of the Golgi ribbon (Marra et al. 2007)

gaps (all authors quoted in the consecutive chapters deserve to be listed here). The list is open for suggestions. The development of the research in the field of intracellular transport has been comprehensively discussed in 1998 at the conference in Pavia devoted to the 100th anniversary of the Golgi discovery.

History of models of intracellular transport Historically, the first mechanism that had been proposed for intracellular transport was the progression. The origin of the progression model (or the concept of cis-to-trans flow) links to Grasse’s name (1957) who proposed that the continuous formation of cis Golgi cisternae balances the conversion of trans one into secretory granules. However, the first experimental data in favour of the progression concept were obtained in 1971 (Franke et al. 1971). In 1967, it has been demonstrated that proteins newly synthesized in the ER appeared, after a few minutes, not only over Golgi stacks but also over

10

*

A. A. Mironov and M. Pavelka

round profiles surrounding the GA and the conclusion that secretory proteins bypass the GA was made (Jamieson and Palade 1967a,b, 1968a,b). Then, in 1981, the vesicular model replaced the progression model because the main support for the progression model, the cis–trans movement of scales in algae has been considered to be a rare formula connected with unusual geometry and size of the product (Farquhar and Palade 1981). Ironically, the major supporting data for the vesicular model at that time was based on the isolation of Golgi-derived clathrin-coated vesicles (Rothman et al. 1980). However, after the discovery of coat protein I (COPI) (Orci et al. 1986), the vesicular model was changed, and instead of clathrin-dependent vesicles, COPI-dependent vesicles were proposed to serve as anterograde carriers. The strongest support for the vesicular model appeared from the experiments in yeast with the temperature sensitive Sec genes (Kaiser and Schekman 1990). The in vitro isolation of functional (containing VSVG and able to fuse with acceptor Golgi membranes) COPI-coated vesicles (Osterman et al. 1993) was interpreted as the second proof for the role of COPI-coated vesicles in the anterograde intra-Golgi transport. Importantly, however, that the first author of this paper later stressed, that actually, these data support the cisterna maturation model (Ostermann 2001). On the other hand, it has also been demonstrated that 20 min after fusion of two (or more) cells (one cell is VSV-infected, another is a non-infected cell) and formation of a heterokaryon, VSVG seems to move from the GA derived from the infected cell to the GA derived from non-infected cells (Rothman et al. 1984). These results were interpreted as confirmation of the ability of vesicular carriers to diffuse through the cytosol of the heterokaryon from one GA to another. However, later, the Rothman group (Orci et al. 1998) laid less emphasis on the heterokaryon experiments, suggesting that those observations appeared as a result of the treatment of cells with an acidic medium. Instead, the “string theory” was proposed, according to which a proteinaceous-like string links vesicles to cisternal elements and prevents budded vesicles from diffusing away, while still allowing them to diffuse laterally. With time, due to accumulation of contradictions, the current vesicular paradigm became less and less effective in the explanation of growing body of observations (Mironov et al. 1997). As a result, the original version of the vesicular paradigm began to be modified not only by the opponents of the vesicular model but also by its authors and proponents (Orci et al. 1998). In order to resolve accumulated contradictions within the field, almost simultaneously several groups (Bannykh and Balch 1997; Mironov et al. 1997; Glick et al. 1997; Schekman and Mellman 1997) have published the cisterna maturation-progression model based on the COPI vesicles-mediated Golgi enzyme recycling. The first experimental confirmation that large aggregated cargo, such as procollagen I, can be transported through the GA by maturation mechanism came in 1998 (Bonfanti et al. 1998). Previous stereological observations in

The Golgi apparatus and main discoveries

*

11

P. scheffelii suggesting that their scales being much too large to be packaged into vesicles are transported by the progression of Golgi cisternae towards the plasmalemma were published not in an original paper but in a review (Becker et al. 1995) and were not confirmed later because glycoprotein and polysaccharide synthesis are uncoupled during flagella regeneration (Perasso et al. 2000). Next, it has been demonstrated (Mironov et al. 2001) that both diffusible and non-diffusible cargoes are transported in the same carriers through the Golgi stacks. It has been proved that vesicles are not transport carriers for cargo in the intra-Golgi transport not only in situ, but also in vitro, in cell-free assay (Happe and Weidman 1998). After these publications, there was a short period when the cisterna maturation model became dominant. With time new contradictions not compatible with the cisterna maturation-progression model have accumulated (Mironov et al. 2005). The attempts to use transport models based on combination of basic principles were not successful (see Chapter 3.2). Therefore now, there is no consensus on the models of intra-Golgi transport. The existence of the maturation mechanism is almost finally established for the secretion of large polymeric structures incompatible in size with COPI-dependent vesicles in many types of cells and under the infection of some viruses.

References Bainton DF, Farquhar MG (1966) Origin of granules in polymorphonuclear leukocytes. Two types derived from opposite faces of the Golgi complex in developing granulocytes. J Cell Biol 28(2): 277–301 Balch WE, Dunphy WG, Braell WA, Rothman JE (1984) Reconstitution of the transport of protein between successive compartments of the Golgi measured by the coupled incorporation of N-acetylglucosamine. Cell 39: 405–416 Bannykh SI, Rowe T, Balch WE (1996) The organization of endoplasmic reticulum export complexes. J Cell Biol 135: 19–35 Bannykh SI, Balch WE (1997) Membrane dynamics at the endoplasmic reticulum–Golgi interface. J Cell Biol 138: 1–4 Barlowe C, Orci L, Yeung T, Hosobuchi M, Hamamoto S, Salama N, Rexach MF, Ravazzola M, Amherdt M, Schekman R (1994) COPII: a membrane coat formed by Sec proteins that drive vesicle budding from the endoplasmic reticulum. Cell 77: 895–907 Becker B, Bolinger B, Melkonian M (1995) Anterograde transport of algal scales through the Golgi complex is not mediated by vesicles. Trends Cell Biol 5: 305–307 Berger EG (1997) The Golgi apparatus: from discovery to contemporary studies. In: Berger EG, Roth J (eds) The Golgi apparatus. Basel et al., Birkhauser Verlag, pp 1–35  rguez JA, Martella O, Fusella A, Baldassarre M, Bonfanti L, Mironov AA Jr, Martínez-Mena Buccione R, Geuze HJ, Mironov AA, Luini A (1998) Procollagen traverses the Golgi stack without leaving the lumen of cisternae: evidence for cisternal maturation. Cell 95(7): 993–1003 Cole NB, Smith CL, Sciaky N, Terasaki M, Edidin M, Lippincott-Schwartz J (1996) Diffusional mobility of Golgi proteins in membranes of living cells. Science 273 (5276): 797–801

12

*

A. A. Mironov and M. Pavelka

Cosson P, Letourneur F (1994) Coatomer interaction with di-lysine endoplasmic reticulum retention motif. Science 263: 1629–1631 Dalton AJ (1951) Observations of the Golgi substance with the electron microscope. Nature 168(4267): 244–245 De Matteis MA, Santini G, Kahn RA, Di Tullio G, Luini A (1993) Receptor and protein kinase C-mediated regulation of ARF binding to the Golgi complex. Nature 364: 818–821 Dell’Angelica EC, Ohno H, Ooi CE, Rabinovich E, Roche KW, Bonifacino JS (1997) AP-3: an adaptor-like protein complex with ubiquitous expression. EMBO J 16(5): 917–928 Dell’Angelica EC, Mullins C, Bonifacino JS (1999) AP-4, a novel protein complex related to clathrin adaptors. J Biol Chem 274: 7278–7285 Doms RW, Russ G, Yewdell JW (1989) Brefeldin A redistributes resident and itinerant Golgi proteins to the endoplasmic reticulum. J Cell Biol 109: 61–72 € scher A (1998) Camillo Golgi and the discovery of the Golgi apparatus. Histochem Dro Cell Biol 109: 425–430 Dunphy WG, Rothman JE (1983) Compartmentation of asparagine-linked oligosaccharide processing in the Golgi apparatus. J Cell Biol 97(1): 270–275 Farquhar MG, Palade GE (1981) The Golgi apparatus (complex)-(1954–1981)-from artifact to center stage. J Cell Biol 91(3 Pt 2): 77s–103s Fasshauer D, Sutton RB, Brunger AT, John R (1998) Conserved structural features of the synaptic fusion complex: SNARE proteins reclassified as Q- and R-SNAREs. Proc Natl Acad Sci USA 95(26): 15781–15786 Featherstone C, Griffiths G, Warren G (1985) Newly synthesized G protein of vesicular stomatitis virus is not transported to the Golgi complex in mitotic cells. J Cell Biol 101(6): 2036–2046 Franke WW, Morre DJ, Deumling B, Cheetham RD, Kartenbeck J, Jarasch E-D, Zengtraf HW (1971) Synthesis and turnover of membrane proteins in rat liver: an examination of the membrane flow hypothesis. Z Naturforsch 26b: 1031–1039 Friend DS, Farquhar MG (1967) Functions of coated vesicles during protein absorption in the rat vas deferens. J Cell Biol 35(2): 357–376 Glick BS, Elston T, Oster G (1997) A cisternal maturation mechanism can explain the asymmetry of the Golgi stack. FEBS Lett 414: 177–181 Godman GC, Lane N (1964) On the site of sulfation in the chondrocyte. J Cell Biol 21: 353–366 Golgi C (1898a) Intorno alla struttura della cellula nervosa. Boll Soc Med Chir Pavia 13: 1–14 Golgi C (1898b) Sur la structure des cellules nerveuses des ganglions spinaux. Arch Ital Biol 30: 60–71 Grasse PP (1957) Ultrastructure, polarity and reproduction of Golgi apparatus. C R Hebd Seances Acad Sci 245(16): 1278–1281 Griffiths G, Warren G, Stuhlfauth I, Jockusch BM (1981) The role of clathrin-coated vesicles in acrosome formation. Eur J Cell Biol 26(1): 52–60 Griffiths G, Simons K (1986) The trans Golgi network: sorting at the exit site of the Golgi complex. Science 34: 438–443 Happe S, Weidman P (1998) Cell-free transport to distinct Golgi cisternae is compartment specific and ARF independent. J Cell Biol 140(3): 511–523 Herzog V, Farquhar MG (1977) Luminal membrane retrieved after exocytosis reaches most Golgi cisternae in secretory cells. Proc Natl Acad Sci USA 74(11): 5073–5077 Hirschberg K, Miller CM, Ellenberg J, Presley JF, Siggia ED, Phair RB, LippincottSchwartz J (1998) Kinetic analysis of secretory protein traffic and characterization of Golgi to plasma membrane transport in living cells. J Cell Biol 143: 1485–1503

The Golgi apparatus and main discoveries

*

13

Jamieson JD, Palade GE (1967a) Intracellular transport of secretory proteins in the pancreatic exocrine cell. I. Role of the peripheral elements of the Golgi complex. J Cell Biol 34(2): 577–596 Jamieson JD, Palade GE (1967b) Intracellular transport of secretory proteins in the pancreatic exocrine cell. II. Transport to condensing vacuoles and zymogen granules. J Cell Biol 34(2): 597–615 Jamieson JD, Palade GE (1968a) Intracellular transport of secretory proteins in the pancreatic exocrine cell. 3. Dissociation of intracellular transport from protein synthesis. J Cell Biol 39(3): 580–588 Jamieson JD, Palade GE (1968b) Intracellular transport of secretory proteins in the pancreatic exocrine cell. IV. Metabolic requirements. J Cell Biol 39(3): 589–603 kely G (2003) Small GTPases and the evolution of the eukaryotic cell. Bioessays 25(11): Je 1129–1138 Kaiser CA, Schekman R (1990) Distinct sets of SEC genes govern transport vesicle formation and fusion early in the secretory pathway. Cell 61(4): 723–733 Kreitzer G, Schmoranzer J, Low SH, Li X, Gan Y, Weimbs T, Simon SM, Rodriguez-Boulan E (2003) Three-dimensional analysis of post-Golgi carrier exocytosis in epithelial cells. Nat Cell Biol 5(2): 126–136 Ladinsky MS, Mastronarde DN, McIntosh JR, Howell KE, Staehelin LA (1999) Golgi structure in three dimensions: functional insights from the normal rat kidney cell. J Cell Biol 144: 1135–1149 Lippincott-Schwartz J, Yuan LC, Bonifacino JS, Klausner RD (1989) Rapid redistribution of Golgi proteins into the ER in cells treated with Brefeldin A: evidence for membrane cycling from the Golgi to ER. Cell 56: 801–813 Lucocq JM, Pryde JG, Berger EG, Warren G (1987) A mitotic form of the Golgi apparatus in Hela cells. J Cell Biol 104: 865–874 Marra P, Salvatore L, Mironov A Jr, Di Campli A, Di Tullio G, Trucco A, Beznoussenko G, Mironov A, De Matteis MA (2007) The biogenesis of the Golgi ribbon: the roles of membrane input from the ER and of GM130. Mol Biol Cell 18(5): 1595–1608 Marsh BJ, Volkmann N, McIntosh JR, Howell KE (2004) Direct continuities between cisternae at different levels of the Golgi complex in glucose-stimulated mouse islet beta cells. Proc Natl Acad Sci USA 101(15): 5565–5570 Mironov AA, Weidman P, Luini A (1997) Variations on the intracellular transport theme: maturing cisternae and trafficking tubules. J Cell Biol 138: 481–484 Mironov AA, Beznoussenko GV, Nicoziani P, Martella O, Trucco A, Kweon HS, Di Giandomenico D, Polishchuk RS, Fusella A, Lupetti P, Berger EG, Geerts WJ, Koster AJ, Burger KN, Luini A (2001) Small cargo proteins and large aggregates can traverse the Golgi by a common mechanism without leaving the lumen of cisternae. J Cell Biol 155: 1225–1238 Mironov AA, Mironov AA Jr, Beznoussenko GV, Trucco A, Lupetti P, Smith JD, Geerts WJ, Koster AJ, Burger KN, Martone ME, Deerinck TJ, Ellisman MH, Luini A (2003) ER-toGolgi carriers arise through direct en bloc protrusion and multistage maturation of specialized ER exit domains. Dev Cell 5: 583–594 Mironov AA, Beznoussenko GV, Polishchuk RS, Trucco A (2005) Intra-Golgi transport. A way to a new paradigm? BBA Mol Cell Res 1744: 340–350  DJ, Mollenhauer HH (1964) Isolation of Golgi apparatus from plant cells. J Cell Morre Biol 23: 295–305  DJ, Merlin L, Keenan T (1969) Localization of glycosyl transferase activities in a Morre Golgi apparatus-rich fraction isolated from rat liver. Biochem Biophys Res Commun 37(5): 813–819 Moya M, Dautry-Varsat A, Goud B, Louvard D, Boquet P (1985) Inhibition of coated pit formation in Hep2 cells blocks the cytotoxicity of diphtheria toxin but not that of ricin. J Cell Biol 101: 548–559

14

*

A. A. Mironov and M. Pavelka

Munro S, Pelham HRB (1987) A C-terminal signal prevents secretion of luminal ER proteins. Cell 48: 899–907 Nakamura N, Rabouille C, Watson R, Nilsson T, Hui N, Slusarewicz P, Kreis TS, Warren G (1995) Characterization of a cis-Golgi matrix protein, GM130. J Cell Biol 131: 1715–1726 Neutra M, Leblond CP (1966) Radioautographic comparison of the uptake of galactoseH and glucose-H3 in the Golgi region of various cells secreting glycoproteins or mucopolysaccharides. J Cell Biol 30: 137–150 Newman AP, Shim J, Ferro-Novick S (1990) BET1, BOS1, and SEC22 are members of a group of interacting yeast genes required for transport from the endoplasmic reticulum to the Golgi complex. Mol Cell Biol 10(7): 3405–3414 Novikoff A, Goldfischer S (1961) Nucleosidediphosphatase activity in the Golgi apparatus and its usefulness for cytological studies. Proc Natl Acad Sci USA 47: 802–810 Novikoff AB (1964) GERL, its form and function in neurons of rat spinal ganglia. Biol Bull 127: 358 Novikoff AV, Essner E, Quintana N (1964) Golgi apparatus and lysosomes. Fed Proc 23: 1010–1022 Oprins A, Rabouille C, Posthuma G, Klumperman J, Geuze HJ, Slot JW (2001) The ER to Golgi interface is the major concentration site of secretory proteins in the exocrine pancreatic cell. Traffic 2: 831–838 Orci L, Glick BS, Rothman JE (1986) A new type of coated vesicular carrier that appears not to contain clathrin: its possible role in protein transport within the Golgi stack. Cell 46: 171–184 Orci L, Perrelet A, Rothman JE (1998) Vesicles on strings: morphological evidence for processive transport within the Golgi stack. Proc Natl Acad Sci USA 95(5): 2279–2283 Ostermann J, Orci L, Tani K, Amherdt M, Ravazzola M, Elazar Z, Rothman JE (1993) Stepwise assembly of functionally active transport vesicles. Cell 75(5): 1015–1025 Ostermann J (2001) Stoichiometry and kinetics of transport vesicle fusion with Golgi membranes. EMBO Rep 2(4): 324–329 Pavelka M, Ellinger A, Debbage P, Loewe C, Vetterlein M, Roth J (1998) Endocytic routes to the Golgi apparatus. Histochem Cell Biol 109: 555–570 Pearse BM (1976) Clathrin: a unique protein associated with intracellular transfer of membrane by coated vesicles. Proc Natl Acad Sci USA 73(4): 1255–1259 € linger B, Melkonian M, Becker B (2000) The Golgi € ntrup IM, Bo Perasso L, Grunow A, Bru apparatus of the scaly green flagellate Scherffelia dubia: uncoupling of glycoprotein and polysaccharide synthesis during flagellar regeneration. Planta 210(4): 551–562 Polishchuk RS, Polishchuk EV, Marra P, Alberti S, Buccione R, Luini A, Mironov AA (2000) Correlative light-electron microscopy reveals the tubular–saccular ultrastructure of carriers operating between Golgi apparatus and plasma membrane. J Cell Biol 148 (1): 45–58 Polishchuk RS, San Pietro E, Di Pentima A, Tete S, Bonifacino JS (2006) Ultrastructure of long-range transport carriers moving from the trans Golgi network to peripheral endosomes. Traffic 7: 1092–1103 Presley JF, Cole NB, Schroer TA, Hirschberg K, Zaal KJ, Lippincott-Schwartz J (1997) ER-toGolgi transport visualized in living cells. Nature 389: 81–85 Puthenveedu MA, Bachert C, Puri S, Lanni F, Linstedt AD (2006) GM130 and GRASP65dependent lateral cisternal fusion allows uniform Golgi-enzyme distribution. Nat Cell Biol 8: 238–248 Roth J, Berger EG (1982) Immunocytochemical localization of galactosyltransferase in HeLa cells: codistribution with thiamine pyrophosphatase in trans-Golgi cisternae. J Cell Biol 93(1): 223–229 Roth J, Berger EG (eds) (1997) The Golgi apparatus. Basel. Birkhauser

The Golgi apparatus and main discoveries

*

15

Rothman JE, Bursztyn-Pettegrew H, Fine RE (1980) Transport of the membrane glycoprotein of vesicular stomatitis virus to the cell surface in two stages by clathrincoated vesicles. J Cell Biol 86(1): 162–171 Rothman JE, Miller RL, Urbani LJ (1984) Intercompartmental transport in the Golgi complex is a dissociative process: facile transfer of membrane protein between two Golgi populations. J Cell Biol 99: 260–271 Sandvig K, Sundan A, Olsnes S (1985) Effect of potassium depletion of cells on their sensitivity to diphtheria toxin and pseudomonas toxin. J Cell Physiol 124: 54–56 Saraste J, Kuismanen E (1984) Pre- and post-Golgi vacuoles operate in the transport of Semliki Forest virus membrane glycoproteins to the cell surface. Cell 38(2): 535–549 Scales SJ, Pepperkok R, Kreis TE (1997) Visualization of ER-to-Golgi transport in living cells reveals a sequential mode of action for COPII and COPI. Cell 90: 1137–1148 Schekman R, Mellman I (1997) Does COPI go both ways? Cell 90: 197–200 Serafini T, Orci L, Amherdt M, Brunner M, Kahn RA, Rothman JE (1991) ADP-ribosylation factor is a subunit of the coat of Golgi-derived COP-coated vesicles: a novel role for a GTP-binding protein. Cell 67(2): 239–253 Slusarewicz P, Nilsson T, Hui N, Watson R, Warren G (1994) Isolation of a matrix that binds medial Golgi enzymes. J Cell Biol 124(4): 405–413 Swift AM, Machamer CE (1991) A Golgi retention signal in a membrane-spanning domain of coronavirus E1 protein. J Cell Biol 115(1): 19–30 Tabas I, Kornfeld S (1980) Biosynthetic intermediates of beta-glucuronidase contain high mannose oligosaccharides with blocked phosphate residues. J Biol Chem 255 (14): 6633–6639 Trucco A, Polishchuk RS, Martella O, Di Pentima A, Fusella A, Di Giandomenico D, San Pietro E, Beznoussenko GV, Polishchuk EV, Baldassarre M, Buccione R, Geerts WJ, Koster AJ, Burger KN, Mironov AA, Luini A (2004) Secretory traffic triggers the formation of tubular continuities across Golgi sub-compartments. Nat Cell Biol 6 (11): 1071–1081 Varki A, Cummings R, Esko J, Freeze H, Hart G, Marth J (1999) Essentials of glycobiology. Cold Spring Harbor Laboratory Press, Cold Spring Harbor Wacker I, Kaether C, Kromer A, Migala A, Almers W, Gerdes HH (1997) Microtubuledependent transport of secretory vesicles visualized in real time with a GFP-tagged secretory protein. J Cell Sci 110: 1453–1463 Weisz OA, Swift AM, Machamer CE (1993) Oligomerization of a membrane protein correlates with its retention in the Golgi complex. J Cell Biol 122(6): 1185–1196 Whur P, Herscovics A, Leblond CP (1969) Radioautographic visualization of the incorporation of galactose-3H and mannose-3H by rat thyroids in vitro in relation to the stages of thyroglobulin synthesis. J Cell Biol 43: 289–311

16

*

A. A. Mironov and M. Pavelka

The Golgi apparatus as a crossroads in intracellular traffic Alexander A. Mironov and Margit Pavelka

In this chapter, we will briefly describe the structure and list main functions of different compartments along the secretory pathway.

ER–Golgi interface After its synthesis, folding and quality control a cargo exits from the endoplasmic reticulum (ER) and moves to the Golgi apparatus (GA) through the ER–Golgi interface. Here, the most peculiar structures are vesicular–tubular clusters (VTC) defined as clusters of a few small vesicles and tubular–saccular elements associated with the region of the rough ER and even with the nuclear envelope containing COPII-coated buds (Bannykh et al. 1996; Mironov et al. 2003). On routine EM sections, the size of VTCs varies from 200 to 1,000 nm. They are positive for glucose-6-phosphatase (Thorne-Tjomsland et al. 1991) and at immunofluorescence level, appear as COPII-positive distinct sites (0.5–1 mm in size) associated with the ER (Stephens et al. 1997). In interphase cells at steady state, the average number of VTC remains constant varying from several dozens to several hundreds of VTC per one cell (Bannykh et al. 1996; Aridor et al. 1999; Hammond and Glick 2000). Typically, COPII-coated buds are described as elevations on the surface of the ER with a width of 65–85 nm, extruded from the membrane by at least 50% of their diameter. Buds are covered with an 8–10 nm thick electron-dense COPII coat. On grazing sections, the buds possess a lattice-like appearance due to semi-regular array of 4–5 nm elongated particles arranged in a semi-regular pattern with mostly tetrahedron organization (Bannykh et al. 1996). Separated COPII-coated vesicles do exist (Zeuschner et al. 2006) although they are few and mostly are devoid of secretory proteins (Mironov et al. 2003). Some elongated profiles (which might represent cross-sections of saccules or tubules) within VTCs have a dense COPI-like coat at their tips and on their central parts (Martinez-Menarguez et al. 1999). VTCs are not carriers, which undergo centralization and deliver cargo to the GA (Stephens et al. 2000; Mironov et al. 2003; see details in Chapter 3.1). For this purpose, ER-to-Golgi carriers (EGCs) are used. These appear mostly as saccular containers filled with either the large supramolecular cargo (i.e. procollagen) or the small diffusible cargo proteins. They arise through cargo concentration and direct en bloc protrusion of specialized ER domains in the vicinity of VTC (Mironov et al. 2003).

Golgi apparatus as a crossroads

*

17

ER–Golgi connections Direct membrane continuity between the ER and the GA has been described in many papers over the years in variety of tissues and cell types under different functional conditions (Flickinger 1969, 1973; Maul and Brinkley 1970; Claude 1970; Bracker et al. 1971; Holzman 1971; Morre et al. 1974; Franke and Kartenbeck 1976; Novikoff and Yam 1978; Uchiyama 1982; Broadwell and Cataldo 1983; Sasaki et al. 1984; Williams and Lafontane 1985; Lindsey and Ellisman 1985a,b; Tanaka etal. 1986; Krijnse-Locker et al. 1994; Sesso et al. 1994; Stinchcombe et al. 1995; Trucco et al. 2004) and using different methods of analysis, including three-dimensional (3D) observation in high voltage electron microscope (Lindsey and Ellisman 1985a,b), scanning electron microscopy (Tanaka et al. 1986), reconstruction of serial sections (Sesso et al. 1994) and even functional analysis of transport (Krijnse-Locker et al. 1994). Connections were described between the ER and EGC (Stinchcombe et al. 1995; Mironov et al. 2003). After 3D tomographic reconstruction (Ladinsky et al. 1999), a connection between the ER and the membrane disk integrated between Golgi cisternae has been found. However, the nature of this disk is not established. The author interpreted this disk as the specialized domain of the ER. Recently, two reports about ER-to-Golgi connections have been published. In one (Koga and Ushiki 2006) the existence of connections between the ER and the GA was not confirmed. However, the method used is not completely free from artefacts. ER-to-Golgi connections cannot be a result of fixation, because fixative usually disrupt pre-existing tubules rather than induce their formation (McIntosh 2001). In the report by Vivero-Salmeron et al. (2008), the existence of ER-to-Golgi connections was confirmed. The relatively low frequency of these observations might simply be due to the fact that thin sections are technically unsuitable for revealing a convoluted and transient (Vivero-Salmeron et al. 2008) structure extending through a large three-dimensional space or due to temporality of the connections (Fig. 1, 2c–e).

The function of compartments within the ER-to-Golgi interface Within the ER-to-Golgi interface, several important posttranslational functions are performed: COPII-mediated concentration of defined membrane and soluble cargoes (to improve the efficiency of transport), delivery of cargo to the GA (see Chapter 3.1), O-glycosylation (Tooze et al. 1988), acylation (Rizzolo et al. 1985), generation of mannose-6-phosphate signal for lysosomal protein targeting (Pelham et al. 1988), protein palmitoylation (Bonatti et al. 1989), retrieval of misfolded proteins (Hammond and Helenius 1994), and segregation of secretory cargoes, namely, regulatory secretory

18

*

A. A. Mironov and M. Pavelka

Figure 1. Scheme of structures within the ER–Golgi interface. The ER is shown as the structure filled with a grey content and covered with ribosomes (dark dots). The ER has a specialized domain that is attached to the Golgi stack (trans-ER). At the cis-side of the stack, the ER contains the ER exit site (ERES, or early intermediate compartment [IC] or vesicular tubular cluster; arrow). The ERES can include COPII-dependent vesicles (arrow). Near the ERES, there is the forming precursor of an ER-to-Golgi carrier (EGC, arrow). The cis Golgi network is composed of several domains. One represents the highly perforated disk similar in shape with Golgi cisternae. It is attached to the Golgi stacks (G) and is named as attached CGN (AC). There is a small part of the total CGN that appears as the three-dimensional tubular network or a cage (indicated as CGN) near the Golgi stacks. This part produces tubules moving towards the ERES (in the center of the image). Another part of the CGN is connected with the AC by tubules and has similar shape with the CGN. It localizes out of the Golgi stack and appears as the late IC that could be rather stable compartment. At the trans-side of the Golgi stack (G), there is the trans-Golgi network (TGN), parts of which reside apart of the stack, and are connected with the other part, the attached TGN (AT). The TGN contains clathrin-coated buds (double arrow). Precursors of Golgi-to-PM carriers (GPC) could form from the last two COPI-positive (with COPI-buds) cisternae, or from the entire TGN. COPI-vesicles are present near the rims of Golgi cisternae.

proteins, constitutive secretory proteins, proteins destined for the apical plasma membrane (PM) and basolateral PM, endosomal and lysosomal proteins by elimination of the mechanisms responsible to their retention within the ER. The role of microtubules in centralization of EGCs is described in Chapter 2.14. However, this function is absent in plant and yeast cells (Nebenfuhr and Staehelin 2001) and even in several cell types in mammals (such as oocytes, Motta et al. 1995) and, thus, might be not directly related to ER–Golgi transport per se. Transient ER–Golgi connections could serve for the diffusion of cargo proteins and recycling of resident proteins.

Golgi apparatus as a crossroads

*

19

Morphology of the Golgi apparatus The structural organization of the GA varies among species (see Chapters 4.1–4.4). In yeast (Chapter 4.3), and some protists (Chapter 4.4), the GA is composed of tubular networks and isolated disks. In animal and plant cells, the GA appears as a series of closely associated flattened membrane ‘sacs’ aligned in parallel to form a stack (Polishchuk and Mironov 2004; Chapter 4.1). In plants, S. cerevisiae, protists and some insect cells, stacks remain separated from each other whereas in mammalian cells stacks form the single ribbon (Chapter 2.12). Here, we describe mainly the mammalian GA. The GA is embedded into a (so-called) “zone of exclusion”, a polymer-based derivative of the cytosol that is especially evident in plant. This zone does not contain ribosomes but cytoskeleton elements can pass through it (Mollenhauer and Morre 1978). In epithelial cells of epididymis, goblet cells in the jejunum, gonadotrophs in pituitary glands and dorsal root ganglion cells, an “empty” zone of exclusion also surrounds the GA with the thickness of about 200 nm (Koga and Ushiki 2006). The GA represents 2% of hepatocellular membrane (Blouin 1983) or 20% of that of the ER system (Griffiths et al. 1989). The GA is capable to undergo rapid and reversible reorganization in response to a variety of experimental manipulations (Polishchuk and Mironov 2004). According to Griffiths et al. (1995), the GA begins from the place where mannosidase I is localized. Man I exhibits a highly polarized staining at the cis-pole of Golgi stacks usually composed of the first, sometimes of the first two cisternae (Marra et al. 2001). The canonical GA consists of a series flattened cisternal membranes closely associated, aligned in parallel and forming a stacked structure, abundant tubular–reticular networks and vesicles. In the perinuclear area, dozens or hundreds of Golgi stacks are linked together to form an interconnected, ribbon-like structure as a single organelle with alternating compact (stacked cisternae) and non-compact (tubular–reticular) zones (Mogelsvang et al. 2004; see also Chapter 2.12). The GA can be viewed schematically as being composed of three main compartments: the cis-, medial- and trans-Golgi. Both the cis- and the transmost Golgi elements are largely tubular. The medial-Golgi stacked cisternal compartment resides between these two networks (Rambourg and Clermont 1997; Polishchuk and Mironov 2004). Several gradients exist within a Golgi stack: (i) a gradient in the cisterna fenestration; (ii) a gradient in the cisterna thickness; (iii) a gradient in the localization of the Golgi enzymes; (iv) a gradient in the lipid bilayer thickness; (v) a gradient in the pH; (vi) a gradient in concentration of cholesterol. The concentration of cholesterol is higher at the trans-side of a Golgi stack. Especially high concentration of cholesterol is found in endosomes (Orci et al. € bius et al. 2003). 1981; Cluett et al. 1997; Mo

20

*

A. A. Mironov and M. Pavelka

The cis-to-trans changes in cisterna fenestration and thickness mean that the size of the fenestrae and the wells initially become smaller towards the medial cisternae, while the thickness of the cisternae decreases from the cis to the trans-pole. Then, still in a trans-wise direction, the cisternae become more perforated again (Ladinsky et al. 1999). The enzymes involved in the early stages of glycosylation are located mostly at the cis-side, whereas the late and terminal glycosylation enzymes are situated at the trans-side of the stacks. The trans-compartments are enriched in terminal processing enzymes involved in sialylation. However, a significant difference in the distribution of the Golgi enzymes can be detected only between the cis-most and trans-most cisternae of the central

Golgi apparatus as a crossroads

*

21

Golgi domain, and not between adjacent cisternae of the same stack (Rabouille et al. 1995). Thus, a single cisterna does not necessarily represent a separated Golgi compartment from either the biochemical or the structural point of view, potentially representing instead a part of a larger compartment.

Cis-Golgi network (CGN) The cis-Golgi network is composed of several domains. One represents the highly perforated disk similar in shape with Golgi cisternae. It is attached to the Golgi stacks and is named as the attached CGN or the Cis-perforated cisternae of the intermediate compartment or CGN (CISCIC). CISCIC appears as a disk with 30 nm perforations. This is especially evident in epithelial cells of epididymis, goblet cells in the jejunum, gonadotrophs in pituitary glands and dorsal root ganglion cells (Koga and Ushiki 2006). It is highly labelled for COPI (Oprins et al. 1993; Kweon et al. 2004). The second part of the CGN is a small tubular part of the total CGN that appears as the three-dimensional tubular network or a cage near the Golgi stacks. Let’s name it as the free CGN. This part produces tubules moving towards the ER-export sites (ERES; Marra et al. 2001; Mironov et al. 2003). Another part of the CGN is connected with the attached CGN by tubules and has similar shape with the CGN. It localizes out of the GA and appears as the late intermediate compartment (Marra et al. 2001) that could be rather a stable compartment (Ben-Tekaya et al. 2005). In spermatides, the cis-most cisterna appears as a regular network of anastomotic membranous tubules, and the medial saccules usually have fewer but larger irregular fenestrations in them (Ho et al. 1999). There, the cis-elements of Golgi stacks were slightly reactive for G6Pase. Labelling can be seen in some cis-Golgi cisternae. The CGN contains also early processing enzymes such as alpha-mannosidases that trim high mannose N-linked

"

Figure 2. Tomographic 3D reconstruction of different compartments of the secretory pathway. A,B Comparison of the resting (A) and transporting ministacks (B) in fibroblast-like cells. Ministacks are formed after the complete depolymerization of MTs (3 h of treatment with nocodazole). The resting stack (the stack before the release of the block of intra-Golgi transport) contains more vesicles (white spheres) and less cisternae. In resting stack (A), the cis-most and the trans-most cisternae are almost absent whereas in the transporting stack (B) both the transcisterna (white arrow) and the cis-cisterna (red arrow) are visible. In the resting stack (A), the cismost cisterna is replaced by the tubular network (white arrow). C–E Structure of moving (C,D) and stationary (E) ER-to-Golgi carriers. The ER is pictured in green. The uncoated domains of the EGCs are pictured into brownish. The domains coated with COPII-like coat are pictured into yellow whereas the buds coated with COPI-like coat are indicated by light-blue colour. Samples were prepared according to the correlative light-electron microscopy (Mironov et al. 2003) 20 min after the restoration of ER-to-Golgi transport of tsVSVG. F,G Structure of the Golgi exit site at the moment of the formation of Golgi-to-PM carriers filled with tsVSVG (12 min after the release of the ER exit block according to the small pulse-chaise protocol (Mironov et al. 2001). Connections (arrow) between the COPI-positive Golgi cisterna and the TGN (yellow network) are shown from different points of view. Red arrows show the Golgi stack. The ER is pictured in green. Models A and B were made by A. Trucco, Models C–G were made by G. V. Beznoussenko. Bar: 120 nm.

22

*

A. A. Mironov and M. Pavelka

oligosaccharides added to the nascent chain in the ER (Thorne-Tjomsland et al. 1991). The CGN receives newly synthesized or recycled polypeptides from the endoplasmic reticulum, which are then posttranslationally modified by glycosylation, sulphation, phosphorylation, palmitoylation, myristoylation or methylation (De Graffenried and Bertozzi 2004).

Cisternal shape Cisternae are oriented along microtubules. Although the number of cisternae in the Golgi stacks varies from one cell type to another (three to eight cisternae, in the majority of cases), however within the same cell line, the number of cisternae could be constant representing a specific characteristic of cell type (Ladinsky et al. 1999) or varies depending on the functional state. At least, this number is reproduced in cells washed out after brefeldin A and nocodazole treatment. The number of not perforated COPI-positive cisternae is almost not changed after the arrival of cargo (Trucco et al. 2004; see Chapter 2.16). All Golgi cisternae have roughly the same surface area, although they can differ in volume by as much as 50% (Ladinsky et al. 1999). The length of all of the cisternae within the stack is equal. When the GA becomes fragmented, this feature becomes particularly evident even when cargo is being transported through the GA. Even after arrival of cargo to the cis-side of the GA, the length of all Golgi cisternae rapidly became equal (Trucco et al. 2004). Both the cis-most perforated cisterna and all medial cisternae contain COPI-coated buds, whereas the trans-most perforated cisterna(e) contains only clathrin-coated buds and usually have no COPI-coated buds (Ladinsky et al. 1999). All Golgi cisternae are fenestrated; the existence of Golgi cisternae without these fenestrae has yet to be demonstrated, at least in mammalian cells. The large openings in cisternae can often form “wells” (Ladinsky et al. 1999). These fenestrae are necessary for movement of secretory granules (SGs; Rambourg and Clermont 1997). The lumen of a Golgi cisterna is usually quite narrow (10–20 nm). There is a systematic decrease in luminal diameter in the trans-direction in quick-frozen NRK cells (Ladinsky et al. 1999). Mechanisms responsible for Golgi cisterna stacking and maintenance of the cisterna shape remain mostly unknown. Attempts to explain stacking by the presence of so-called Golgi matrix proteins or Golgins (see Chapter 2.12) were not successful (Seemann et al. 2000) because these proteins are not present between all Golgi cisternae. Low affinity antiparallel dimerization of cytosolic domains of sugar transporters might be responsible for attachment of Golgi cisternae to each other. At least overexpression of GDP-mannose transporter in the yeast Saccharomyces cerevisiae induces formation of the stacked GA (Hashimoto et al. 2002). Stacks are also formed in S. cerevisiae after deletion of function of Sec7

Golgi apparatus as a crossroads

*

23

(Rambourg et al. 1993). In any case, additional analysis is necessary to resolve this problem.

Golgi vesicles There are several known coats along the secretory pathway participating in the formation of coat-dependent vesicles. The most important of them are the following: COPII (see above), COPI, clathrin. Each of them contains several variants of the coat. COPII-based coat has two main forms (see Chapter 2.3), COPI-based coat also has two forms (see Chapter 2.4), and clathrin coat has two or four variants interacting with different adaptors, namely, AP1, AP2, AP3 and AP4 (reviewed by Traub 2005). Membrane budding with the help of these coats could induce generation of corresponding coat-dependent vesicles. So far, only four types of coat-dependent small vesicles have been found in cells, namely, clathrin-coated vesicles that could be formed from clathrin/AP1coated buds present on the TGN attached to COPI-positive stacked Golgi cisternae, secretory granules, endosomes and clathrin/AP2-coated buds found on the PM. COPII-dependent vesicles are formed near the ER exit sites. COPI-dependent vesicles are formed at the level of the ER-to-Golgi carriers, the intermediate compartment, the cis-Golgi and medial Golgi (reviewed by McMahon and Mills 2004). One of the most important features of the GA is the presence of small 52 nm COPI-dependent vesicles surrounding each Golgi cisterna (Ladinsky et al. 1999; Marsh et al. 2001). COPI vesicles do not appear to be really “free” because most of them are unmistakably tethered to neighbouring vesicles and/or to the Golgi membranes (Orci et al. 1998). This explains why vesicles do not diffuse towards the cell periphery. In contrast, most of the clathrindependent or irregularly shaped “vesicles” are clustered away from the GA (Ladinsky et al. 1999). A dense COPI coat observed on Golgi buds/vesicles on thin sections appears as a lace-like cytoplasmic structure closely attached to the lipid bilayer and composed of a series obtuse spikes separated by an average of 20 nm centerto-center. These spikes do not have the “bristle” or “spiny” appearance of clathrin subunits in basketworks (Orci et al. 1986). The thickness of COPI coat is about 10 nm whereas the thickness of clathrin coat is 18 nm (Oprins et al. 1993). The number of vesicles is the result of the equilibrium between two processes – the activity of COPI machinery and the activity of SNARE/Ca machinery. Inhibition of COPI activity causes reduction of the number of COPI-dependent vesicles. Similar effects are observed, when the COPI/ARF machinery is inhibited with brefeldin A In contrast, when the SNARE machinery is inhibited, the number of vesicles increases (Kweon et al. 2004). Most of data suggest that these vesicles are formed from COPI-coated buds abundant along the cisternal rims. COPI-dependent vesicles could derive

24

*

A. A. Mironov and M. Pavelka

from varicosities within tangential tubules along the rims of Golgi cisternae (Weidman et al. 1993). There could be following functions for COPI vesicles. 1. Formation of COPI vesicles could be specific mechanisms controlling the geometry of the Golgi elements (see Chapter 2.16). 2. COPI vesicles could control the fusion between adjacent Golgi cisternae extracting Qb SNAREs from there. COPI-dependent vesicles are two-fold enriched in Qb-SNAREs of the same type: membrin and GOS28 (Trucco et al. 2004; our unpublished observations) and depleted of syntaxin 5 (Orci et al. 2000a,b). Extracting GOS28 and membrin from the Golgi membrane, COPI vesicles prevent fusion between the cisternae (Trucco et al. 2004). 3. Formation of coat-dependent vesicles could be the way for fast uncoating. The Role of COPI vesicles as anterograde or retrograde carriers will be analyzed in Chapter 3.2.

Intercisternal connections One of the important questions of Golgi morphology is the issue of intercisternal heterogeneous connections (Tanaka et al. 1986; Rambourg and Clermont 1990, 1997; Sesso et al. 1994). In some cells, the Golgi forms even a single continuous membranous system (Tanaka et al. 1986; Inoue 1992; Rambourg and Clermont 1990, 1997). At steady state and in unstimulated cells, the connections are rare (Marsh et al. 2004). Intercisternal connections are augmented after arrival of cargo to the GA (Trucco et al. 2004) or when the islet beta cells have been stimulated for 1 h with 11 mM glucose (Marsh et al. 2004). Recently, the existence of such connections has been confirmed at steady state (Beznoussenko et al. 2006; Vivero-Salmeron et al. 2008). The connections between cisternae at different levels of the GA are of three types. The first type is observed at points, where the Golgi ribbon branches (Rambourg and Clermont 1997). Cisternae could be connected at both equivalent and non-equivalent levels. This “Y” configuration of cisternae at the branch in the Golgi ribbon also means that there is direct continuity between cisternae at different levels around the periphery of the upper stack of Golgi membranes. The second type of connection occurs when one cisterna projects through an opening (fenestration) in an adjacent cisterna to form a continuous lumen with its next-nearest neighbour. In the third type of connection, membrane tubules connecting non-equivalent cisternae bypass interceding cisternae at the periphery of the stack in Golgi regions where the ribbon is unbranched (Marsh et al. 2004; Trucco et al. 2004). Intercisternal connections were detected rarely because on conventional EM sections luminal continuity between consecutive cisternae is almost undetectable (Marsh 2005).

Golgi apparatus as a crossroads

*

25

The tubules connecting different stacks have been proposed to participate in intra-Golgi traffic (Mellman and Simons 1992; Weidman 1995). These connections could be used for the transport of soluble cargo, like albumin, or serve for the movement of proton from the trans- to the cis-compartments, or for Golgi enzyme diffusion (Trucco et al. 2004). The speed diffusion of ions, lipids and even transmembrane proteins along Golgi lumen and membranes is high. For instance, the speed of cholesterol diffusion in clear lipid bilayer is about 2 mm/s. The speed of ceramide diffusion is about 0.4–0.5 mm/s (Cooper et al. 1990). If connections are constant, the ionic, lipid and protein gradients that are known to exist between the Golgi poles (reviewed in Mironov et al. 1998, 2005) have to be expected to dissipate through these continuities. However, since this is not the case the connections have to be transient and highly regulated as within the framework of kiss-and-run models of transport (see Chapter 3.2).

Function of the Golgi apparatus In mammalian cells, occupying a central position, the GA plays a central role in the classical secretory pathway, as well as in endocytic pathways, and in multiple recycling routes. 1. The GA exchanges membrane components with several other subcellular organelles, including endosomes, caveosomes, autophagosomes, and lipid droplets participating in sorting (Mironov et al. 2005). 2. The GA is the main station of cellular glycosylation. During movement along the Golgi membranes, cargoes undergo glycosylation. The Golgi stack is composed of a series of compartments containing oligosaccharide processing and other enzymes that are generally arranged in a cis-totrans orientation (Rabouille et al. 1995). 3. Assembly of triglycerides with apoB and other apolipoproteins occurs in the GA. During this process, apoB undergoes conformational changes, and the expanding lipoproteins recruit more apoE (Gusarova et al. 2007). 4. The GA is involved in various other cellular processes such as transcription, apoptosis, and mitosis via signalling pathways mediated by Ras proteins, protein kinases, and G proteins (Helms et al. 1998; DeBose-Boyd et al. 1999; Lane et al. 2002; Sutterlin et al. 2002; Bivona et al. 2003; Nardini et al. 2003; Preisinger et al. 2004). 5. The GA provides a connection between exocytosis and endocytosis.

Structure of the Golgi exit site After its passage through the GA cargo exits from it, and moves to the sites of its destinations. In the literature, this Golgi exit site (GES) is usually called as the trans-Golgi network (TGN, Griffiths and Simons 1986; Ladinsky et al. 1994; Clermont et al. 1995) suggesting that the TGN is composed of distinct tubules

26

*

A. A. Mironov and M. Pavelka

with little indication of anastomosis. However, this name is misleading because the term TGN reflects mostly the structural organization of only one compartment, the whole TGN, within GES (Fig. 2f, g). The identity of the GES could be defined by membrane components delivered to the TGN from both anterograde and retrograde transport, together with the recruitment from the cytosol of coat proteins, regulatory GTPases, fusion and Golgi matrix and motor proteins. Among the most important proteins localized within the TGN, one could observe clathrin, M6PR, TGN38, AP1, furin, GGA, EEA1, clathrin, Golgin-97, and other transGolgins (reviewed in Robinson and Bonifacino 2001). Two to four Golgi cisternae are stained with ceramide (Pagano et al. 1989) and TPPase (Novikoff et al. 1971). The GESs are composed of the two last (trans) COPI-positive cisternae of the Golgi stack (Ladinsky et al. 2002), the trans-most perforated cisterna(e) with clathrin-coated buds, and the network of tubules surrounding the stack near its trans-pole and sometimes being continuous with either multiple trans-cisternae (Rambourg et al. 1979) or with only the trans-most cisterna in the stack (Griffiths and Simons 1985; Griffiths et al. 1989). “Peeling off” configurations of the last Golgi cisterna with clathrin-coated buds are frequently described in the past (Ladinsky et al. 1994; Clermont et al. 1995). This trans-most cisterna contains exclusively clathrin-coated buds, whereas the other cisternae have COPI-coated buds only (Ladinsky et al. 1999, 2002). Only 12% of the total TGN surface area is attributable to the flattened cisternal part of the TGN which is labelled by the presence of TPPase and which is morphologically indistinguishable from the other cisternae of the Golgi stack. In many cases most of the tubules located within the TGN area are devoid of the reaction for TPPase (Griffiths et al. 1989). The trans-most cisterna containing clathrin-coated buds actually represents a highly perforated disk containing clathrin-coated buds along its rims (Ladinsky et al. 1999). This perforated cisterna is accessible for WGAHRP added from outside (Pavelka et al. 1998). Thus, it represents a part of the endocytic TGN connected with endosomes. As such, this trans cisterna of the endocytic trans-Golgi network could be named as the TRANSCET or the attached TGN. In epithelial cells of the epididymis, goblet cells in the jejunum, gonadotrophs in pituitary glands and dorsal root ganglion cells, the TRANSCET appears as perforated sheet, varicose tubules or small plates connected with each other two-dimensionally by tubules (Koga and Ushiki 2006). Structures within GES are highly dynamic and continuously undergo renewal (Clermont et al. 1995). Tubules continuously emanate from the Golgi cisternae going towards the cell periphery (Cooper et al. 1990). The structure of the GES depends on the cell type. In cells, where secretory granules (SGs) are not seen being associated with the GA, the non-attached TGN appears as a tubular network connected with last two Golgi cisternae. However, the TGN does not form the continuous ribbon along the Golgi

Golgi apparatus as a crossroads

*

27

ribbon. In the second group of the cells, where the SGs are seen exclusively within the trans-most cisterna of the stack or within the TGN, the TGN here is small. The SGs are formed within the TGN. In the third group of cells, immature secretory granules (ISGs) are formed as distensions of all Golgi cisternae with the exception of the cis-most perforated cisternae. The TGN here is almost invisible (Clermont et al. 1995). In spermatides, the TGN is also composed of irregular saccules fenestrated near the edge, and long strings of connected vesicles. However, only the edge portions of the trans-Golgi cisternae but not the whole cisternae are fenestrated (Ho et al. 1999). In the cells not possessing secretory granules, where MTs are depolymerized, the non-attached TGN still exhibits tubular organization but its size is reduced (Trucco et al. 2004). The ER cisternae attached to the TRANSCET or to one of the two of the last medial cisternae (these last medial cisternae usually exhibit the presence of TPPase activity; Paavola 1978a,b,c) are one of the most fundamental features of the Golgi exit site. This trans-ER described in multiple cell types has ribosomes bound to their surface, while their other side associates over considerable distances with Golgi cisternae (Novikoff et al. 1964; Pavelka and Ellinger 1983, 1986; Hermo and Smith 1998; Ladinsky et al. 1999; Marsh et al. 2001). The contact between the trans-ER and the trans-GA could serve for direct lipid transfer by, i.e. the ceramide-transfer protein, the oxysterolbinding protein and others (De Matteis et al. 2007). At the level of the GES, two morphologically distinct coats have been identified, namely, clathrin-based and lace-like (Ladinsky et al. 1994). Here, clathrin coat could form three different variants using adaptors AP1, AP3, AP4 and the GGAs. It has been postulated (but never proved) that there exists an additional coat composed of p62 and p200. The p200 is BFA-sensitive protein localized on Golgi membranes (Narula et al. 1992; Narula and Stow 1995). TGN tubules have varicosities coated by either clathrin or lace-like coat. Each individual tubule is covered by only one type of coat (Ladinsky et al. 1994). Within GES, there are only few clathrin-coated vesicles (Ladinsky et al. 1994, 1999). Many vesicles seen in the routine sections through TGN are actually continuous strings of vesicles (Ho et al. 1999).

Function of the Golgi exit site The first main functions of the GES are protein and lipid posttranslational and postsynthetic modifications. The compartments of GES contain a number of resident enzymes involved in the processing of cargo molecules, such as glycosyltransferases involved in the addition of terminal sugars (Rabouille et al. 1995), several pro-protein convertases including furin (Thomas 2002), and tyrosine sulphation enzymes. For instance, glycosamino glycan chains are synthesized and sulphated in the trans-Golgi/TGN, when cells are incubated with a membrane permanent xyloside (Farquhar 1985; Velasco et al. 1988).

28

*

A. A. Mironov and M. Pavelka

On the other hand, GES is at the crossroads of many transport pathways (Pavelka et al. 1998; Medigeshi and Schu 2003; Shewan et al. 2003; Sannerud et al. 2003; Young et al. 2005; Bonifacino and Rojas 2006). Therefore, the precise sorting is one of the main functions of the GES. This occurs when proteins undergo packaging into their appropriate membrane carriers for delivery not only to the cell surface but also to a number of other compartments (Griffiths and Simons 1986; Traub and Kornfield 1997; Keller and Simons 1998). Signals used for the sorting of proteins at GES for the apical PM (O-glycosylation and N-glycosylation) and endosomal/lysosomal deliveries are discussed in Chapters 3.5–3.7.

Role of cargoes in the structure of the Golgi apparatus Cargo proteins modulate structure and function of the GA. In yeast, a 3 h treatment with cyclohexamide leads to the complete disappearance of the GA (Morin-Ganet et al. 2000). The size of the GA is changed under different conditions. In mammalian cells, the GA tends to become larger when transport intensifies, while it declines in the absence of cargo input (Rambourg et al. 1993; Mironov and Mironov 1998; Trucco et al. 2004). The volume of the TGN depends on the cargo transported (Griffiths et al. 1989). When transport through the GA was inhibited, tubular–reticular membranes in the trans-Golgi area were not detected in tomograms (Trucco et al. 2004). In contrast, when more lysosomal enzymes are produced, there is also proliferation of the TGN (Decker 1974; Novikoff and Novikoff 1977; Paavola 1978a,b,c; Blest et al. 1978; Morre et al. 1979). There are many descriptions of transporting and not transporting stacks, although these are not systematic. Here, we list the main features of transporting and non-transporting stacks (Fig. 2a, b). 1. In the absence of intra-Golgi transport, the Golgi ribbon is fragmented (Marra et al. 2007; see also Chapter 2.12). 2. In the resting stacks, the volume of the GA is smaller. For instance, after accumulation of VSVG at the no permissive temperature the surface area of Golgi membrane reduces 1.6-fold (Aridor et al. 1999). During the starvation of cells, the volume of the GA declines (Mironov and Mironov 1998). In random sections, cisternae of resting stacks are shorter (Marra et al. 2007). In mammalian prolactin cells, blockage of prolactin secretion leads to a reduction of size of the GA (Rambourg et al. 1993). 3. Blockade of protein synthesis with cyclohexamide for 4 h reduces the size of the GA and leads to the formation of very thin cisternae in onion-like Golgi stacks (Taylor et al. 1997). However, in mammals, even 6 h after the blockage of protein synthesis the GA is still present (Beznoussenko et al. submitted). During the stimulation and inhibition of prolactin transport in lactating rats, the GA tends to decline in the absence of cargo input and

Golgi apparatus as a crossroads

4.

5.

6. 7.

8.

9.

*

29

becomes larger when transport intensifies and when the medial Golgi cisternae increase in size (Rambourg et al. 1993). In contrast, when cargo influx increases sharply (e.g. 5 min after ERtransport-block release), the GA immediately grows in size (Aridor et al. 1999). For instance, synchronous release of the pre-Golgi transport block could induce a 3-fold augmentation of the Golgi surface (Trucco et al. 2004). Amplification of the GA is observed in cells secreting human growth hormone (Rudick et al. 1993). In transporting stacks, there are more COPI-dependent 52 nm vesicles than in resting stacks (Rambourg et al. 1993; Trucco et al. 2004). In starved cells, the volume of COPI-dependent vesicle is 20% higher (Mironov and Mironov 1998). In resting stacks, there are no intercisternal connections (Trucco et al. 2004). In resting stacks, there is no cis-most highly perforated cisterna positive for GM130. For instance, at 40 C, the no permissive temperature for tsVSVG, when VSVG is accumulated in the ER after 3 h of the temperature block these perforated cisternae were not visible (Marra et al. 2007; our unpublished observations). Two hours after the blockage of hormonal stimulation of prolactin producing cells the cis Golgi network disappears from the stacks. Lindsey and Ellisman (1985b) demonstrated that in neurons in the very same cell the cis side of the GA could be in five different stages forming the circle of function. In resting stacks, there is a reduction of the TRANSCET (Rambourg et al. 1993; Beznoussenko et al. submitted). The TGN augments when cells are stimulated to secrete (Rambourg et al. 1993, Fig. 3) or when more lysosomal enzymes are produced (Novikoff and Novikoff 1977; Paavola 1997, 1978a,b,c; Blest et al. 1978). In resting stacks, the trans-ER is attached not to the TRANSCET but to the last COPI-positive cisterna, or not attached at all.

Usually, at steady state, the features of the GA are more similar with transporting stacks than to the resting ones.

General principles of intracellular transport Currently, there are three basic principles for the intracellular and in particular intra-Golgi transport, namely, progression, dissociation and diffusion (Fig. 4). Their modifications and combinations of two or more principles in one model give the whole list of existing models of intra-Golgi transport. We think the same situation is present not only for intra-Golgi transport but also for all main steps of intracellular transport, such as ER-to-Golgi and post-Golgi transport. The essence the of progression model could be expressed as the following. At the ER exit sites, there is a formation of large containers for

30

*

A. A. Mironov and M. Pavelka

Figure 3. Tomographic 3D reconstruction of the Golgi exit site during the formation of the precursors of GPCs containing procollagen I (PCI). Samples are prepared at 12 min after the release of the ER exit block for PCI, according to the small pulse-chase protocol (Mironov et al. 2001). A Virtual tomographic section. B–D The 3D model. Each PCI distension (empty cavities coloured in yellow) is covered by the pieces of the trans-most cisterna (green) at least from one side. The single exception is the distension indicated with white arrows in B and D, where the distension appears between the trans-most cisterna and the additional cisternae (pictured in orange). The late endosome (red arrows) contains internal vesicles (blue). The TGN (green) has the connection (blue arrow in B) or a very close association with the plasma membrane (indicated by red lines). During maturation of GPCs these are surrounded by the trans-most cisterna. Models were made by G. V. Beznoussenko. Bars: 150 nm.

transport of cargo and these containers move through the entire pathway of intracellular transport without change of their composition and size. The compartments along the secretory pathway are not stable and each of them contains all proteins necessary for posttranslational modification of secretory proteins. In contrast, the dissociation models implies that the specialized small (much smaller than the compartments) transport carriers that are formed on the proximal (closer to the ER along the secretory pathway) compartment, detach from that compartment, and are delivered (by simple diffusion through the cytosol or with the participation of cytoskeleton motors) to the consecutive but more distal (closer to the PM) compartment, and then dock and fuse with it. Importantly, the fission of TCs occurs before the fusion. Therefore consecutive compartments do not have a common lumen at any time. The single variant of the dissociation model is the vesicular model (VM). One of the less known version of the vesicular model implied that tubules serve as retro-

Golgi apparatus as a crossroads

*

31

Figure 4. Scheme showing the main principles of intracellular transport. A The dissociation (mainly vesicular) model of intracellular transport. Initially, membrane buds are formed on the first (proximal) compartment with the help of protein coat and then, after fission and subsequent uncoating, coat-dependent vesicles move to the second (distal) compartment and are captured by tethering system. Then using SNAREs, the vesicle fuses with the second compartment. B The progression model of intracellular transport. Initially the large membrane protrusion is formed from the first (proximal) compartment with the help of not-coat mechanisms. This protrusion concentrates cargoes or contains them at the same concentration as in the first compartment. The large carrier is captured by the tethering system and with the help of SNAREs undergoes fusion with the second (distal) compartment. C The lateral diffusion model of intracellular transport.

32

*

A. A. Mironov and M. Pavelka

grade carriers (Lippincott-Schwartz 1993). The VM has been proposed by Palade (1975) and further developed by Rothman (1994). Within the framework of VM, membrane budding is the function of the coat proteins (Antonny and Schekman 2001; Bonifacino and Glick 2004; see also Chapter 2.3). Now, several coat complexes functioning within the secretory pathway are known: COPII (ER-to-CISCIC), COPI (CISCIC-to-Golgi; Golgi-to-ER; intra-Golgi: anterograde and retrograde transport), clathrin-AP1 (trans-Golgi networkto-lysosomes). At the TGN level, the existence an additional coat composed of p62 and p200 has been postulated (Zehavi-Feferman et al. 1995; Narula and Stow 1995; Ikonen et al. 1997; as well as clathrin-AP-3 and AP-4 (see Chapters 3.5–3.9). According to the VM, the compartments are stable, isolated entities. Secretory proteins move through the secretory pathway from each compartment to the next in discrete membrane-bound small coat-dependent vesicles. The process of transport at each transport step includes formation of coated buds, concentration of cargoes inside of them, detachment of coated buds with the formation of small 50–100 nm spheres and then their uncoating and fusion with the consecutive compartment. The diffusion mechanism is based on the presence of membrane and luminal connections along the secretory pathway and transport occurs by simple diffusion from the proximal to the distal compartment. In the following chapters, mechanisms of transport will be specified for each step of exocytosis.

Conclusion Structure and functions of the compartments along the secretory pathway are extremely complicated and their morphology depends on their functional state. Analysis of their functional roles within the framework of the transport models together with the precise mapping of known protein machines will be presented in the following chapters.

Abbreviations CISCIC CGN EGC EM ER ERES GA GES PM TRANSCET VM VTC

cis-cisterna of the CGN cis-Golgi network ER-to-Golgi carrier electron microscopy endoplasmic reticulum ER exit site Golgi apparatus Golgi exit site plasma membrane trans-cisterna of the endosomal TGN vesicular model vesicular–tubular cluster

Golgi apparatus as a crossroads

*

33

References Antonny B, Schekman R (2001) ER export: public transportation by the COPII coach. Curr Opin Cell Biol 13: 438–443 Aridor M, Bannykh SI, Rowe T, Balch WE (1999) Cargo can modulate COPII vesicle formation from the endoplasmic reticulum. J Biol Chem 274: 4389–4399 Bannykh SI, Rowe T, Balch WE (1996) The organization of endoplasmic reticulum export complexes. J Cell Biol 135: 19–35 Bannykh SI, Plutner H, Matteson J, Balch WE (2005) The role of ARF1 and rab GTPases in polarization of the Golgi stack. Traffic 6(9): 803–819 Ben-Tekaya H, Miura K, Pepperkok R, Hauri HP (2005) Live imaging of bidirectional traffic from the ERGIC. J Cell Sci 118(Pt2): 357–367 Beznoussenko GV, Sesorova IS, Banin VV (2006) Electron-tomographic analysis of the Golgi complex structure in cultured cells. Morfologiia 129(3): 41–44 Bivona TG, Perez De Castro I, Ahearn IM, Grana TM, Chiu VK, Lockyer PJ, Cullen PJ, Pellicer A, Cox AD, Philips MR (2003) Phospholipase Cgamma activates Ras on the Golgi apparatus by means of RasGRP1. Nature 424: 694–698 Blest AD, Powell K, Kao L (1978) Photoreceptor membrane breakdown in the spider Dinopis: GERL differentiation in the receptors. Cell Tissue Res 195(2): 277–297 Blouin A (1983) Anatomy, ultrastructure and morphometry of the liver. In: Glaumann H, Peters T Jr, Redman C (eds) Plasma protein secretion by the liver. Academic, New York, pp 31–53 Bonatti S, Migliaccio G, Simons K (1989) Palmitylation of viral membrane glycoproteins takes place after exit from the endoplasmic reticulum. J Biol Chem 264(21): 12590–12595 Bonifacino JS, Glick BS (2004) The mechanisms of vesicle budding and fusion. Cell 116(2): 153–166 Bonifacino JS, Rojas R (2006) Retorgrade transport from endosomes to the trans-Golgi network. Nature Rev Mol Cell Biol 7: 568–579 Bracker C, Grove SN, Heintz CE, Morre DJ (1971) Continuity between endomembrane components in hyphae of Pythium spp. Cytobiologie 4: 1–8 Broadwell RD, Cataldo AM (1983) The neuronal endoplasmic reticulum: its cytochemistry and contribution to the endomembrane system. I. Cell bodies and dendrities. J Histochem Cytochem 31: 1077–1088 Claude A (1970) Growth and differentiation of cytoplasmic membranes in the course of lipoprotein granule synthesis in the hepatic cell. I. Elaboration of elements of the Golgi complex. J Cell Biol 47: 745–766 Clermont Y, Rambourg A, Hermo L (1995) Trans-Golgi network (TGN) of different cell types: three-dimensional structural characteristics and variability. Anat Rec 242(3): 289–301 Cluett EB, Kuismanen E, Machamer CE (1997) Heterogeneous distribution of the unusual phospholipid semilysobisphosphatidic acid through the Golgi complex. Mol Biol Cell 8: 2233–2240 Cooper MS, Cornell-Bell AH, Chernjavsky A, Dani JW, Smith SJ (1990) Tubulovesicular processes emerge from trans-Golgi cisternae, extend along microtubules, and interlink adjacent trans-Golgi elements into a reticulum. Cell 61(1): 135–145 De Graffenried CL, Bertozzi CR (2004) The roles of enzyme localisation and complex formation in glycan assembly within the Golgi apparatus. Curr Opin Cell Biol 16: 356–363 De Matteis MA, Di Campli A, D’Angelo G (2007) Lipid-transfer proteins in membrane trafficking at the Golgi complex. Biochim Biophys Acta 1771(6): 761–768 De Matteis MA, Luini A (2008) Exiting the Golgi complex. Nat Rev Mol Cell Biol 9(4): 273–284

34

*

A. A. Mironov and M. Pavelka

DeBose-Boyd RA, Brown MS, Li WP, Nohturfft A, Goldstein JL, Espenshade PJ (1999) Transport-dependent proteolysis of SREBP: relocation of site-1 protease from Golgi to ER obviates the need for SREBP transport to Golgi. Cell 99: 703–712 Decker RS (1974) Lysosomal packaging in differentiating and degenerating anuran lateral motor column neurons. J Cell Biol 61(3): 599–612 Farquhar MG (1985) Progress in unraveling pathways of Golgi traffic. Annu Rev Cell Biol 1: 447–488 Flickinger CJ (1969) The development of Golgi complexes and their dependence upon the nucleus in amoebae. J Cell Biol 49: 250–262 Flickinger CL (1973) Maintenance and regeneration of cytoplasmic organelles in hybrid amoebae formed by nuclear transplantation. Exp Cell Res 80: 31–46 Franke WW, Kartenbeck J (1976) Some principles of membrane differentiation. In: Mueller-Berat N (ed) Progress in differentiation research. Elsevier-North Holland Publishing Co., Amsterdam, pp 213–243 Griffiths G, Simons K (1986) The trans Golgi network: sorting at the exit site of the Golgi complex. Science 234(4775): 438–443 Griffiths G, Fuller SD, Back R, Hollinshead M, Pfeiffer S, Simons K (1989) The dynamic nature of the Golgi complex. J Cell Biol 108: 277–297 Griffiths G, Pepperkok R, Locker JK, Kreis TE (1995) Immunocytochemical localization of beta-COP to the ER–Golgi boundary and the TGN. J Cell Sci 108(Pt 8): 2839–2856 Gusarova V, Seo J, Sullivan ML, Watkins SC, Brodsky JL, Fisher EA (2007) Golgi-associated maturation of very low density lipoproteins involves conformational changes in apolipoprotein B, but is not dependent on apolipoprotein E. J Biol Chem 282(27): 19453–19462 Hammond C, Helenius A (1994) Quality control in the secretory pathway: retention of a misfolded viral membrane glycoprotein involves cycling between the ER, intermediate compartment, and Golgi apparatus. J Cell Biol 126: 41–52 Hammond AT, Glick BS (2000) Dynamics of transitional endoplasmic reticulum sites in vertebrate cells. Mol Biol Cell 11(9): 3013–3030 Hashimoto H, Abe M, Hirata A, Noda Y, Adachi H, Yoda K (2002) Progression of the stacked Golgi compartments in the yeast Saccharomyces cerevisiae by overproduction of GDP-mannose transporter. Yeast 19: 1413–1424 Helms JB, HelmsBrons D, Brugger B, Gkantiragas I, Eberle H, Nickel W, Nurnberg B, Gerdes HH, Wieland FT (1998) A putative heterotrimeric G protein inhibits the fusion of COPI-coated vesicles – segregation of heterotrimeric G proteins from COPI-coated vesicles. J Biol Chem 273: 15203–15208 Hermo L, Smith CE (1998) The structure of the Golgi apparatus: a sperm’s eye view in principal epithelial cells of the rat epididymis. Histochem Cell Biol 109(5–6): 431–447 Ho HC, Tang CY, Suarez SS (1999) Three-dimensional structure of the Golgi apparatus in mouse spermatids: a scanning electron microscopic study. Anat Rec 256(2): 189–194 Holzman E (1971) Cytochemical studies of protein transport in the nervous system. Philos Trans R Soc Lond (Biol) 261: 407–421 Ikonen E, De Almeid JB, Fath KR, Burgess DR, Ashman K, Simons K, Stow JL (1997) Myosin II is associated with Golgi membranes: identification of p200 as nonmuscle myosin II on Golgi-derived vesicles. J Cell Sci 110(Pt18): 2155–2164 Inoue T (1992) Complementary scanning electron microscopy: technical notes and applications. Arch Histol Cytol 55(Suppl): 45–51 Keller P, Simons K (1998) Cholesterol is required for surface transport of influenza virus hemagglutinin. J Cell Biol 140(6): 1357–1367 Koga D, Ushiki T (2006) Three-dimensional ultrastructure of the Golgi apparatus in different cells: high-resolution scanning electron microscopy of osmium-macerated tissues. Arch Histol Cytol 69(5): 357–374

Golgi apparatus as a crossroads

*

35

Krijnse-Locker J, Ericsson M, Rottier PJ, Griffiths G (1994) Characterization of the budding compartment of mouse hepatitis virus: evidence that transport from the RER to the Golgi complex requires only one vesicular transport step. J Cell Biol 124 (1–2): 55–70 Kweon HS, Beznoussenko GV, Micaroni M, Polishchuk RS, Trucco A, Martella O, Di Giandomenico D, Marra P, Fusella A, Di Pentima A, Berger EG, Geerts WJ, Koster AJ, Burger KN, Luini A, Mironov AA (2004) Golgi enzymes are enriched in perforated zones of Golgi cisternae but are depleted in COPI vesicles. Mol Biol Cell 15: 4710–4724 Ladinsky MS, Kremer JR, Furcinitti PS, McIntosh JR, Howell KE (1994) HVEM tomography of the trans-Golgi network: structural insights and identification of a lace-like vesicle coat. J Cell Biol 127(1): 29–38 Ladinsky MS, Mastronarde DN, McIntosh JR, Howell KE, Staehelin LA (1999) Golgi structure in three dimensions: functional insights from the normal rat kidney cell. J Cell Biol 144: 1135–1149 Ladinsky MS, Wu CC, McIntosh S, McIntosh JR, Howell K (2002) Structure of the Golgi and distribution of reporter molecules at 20 C reveals the complexity of the exit compartments. Mol Biol Cell 13: 2810–2825 Lane JD, Lucocq J, Pryde J, Barr FA, Woodman PG, Allan VJ, Lowe M (2002) Caspasemediated cleavage of the stacking protein GRASP65 is required for Golgi fragmentation during apoptosis. J Cell Biol 156: 495–509 Lindsey JD, Ellisman MH (1985a) The neuronal endomembrane system. I. Direct links between rough endoplasmic reticulum and the cis-element of the Golgi apparatus. J Neurosci 5: 3111–3123 Lindsey JD, Ellisman MH (1985b) The neuronal endomembrane system. II. The multiple forms of the Golgi apparatus cis element. J Neurosci 5(12): 3124–3134 Lippincott-Schwartz J (1993) Bidirectional membrane traffic between the endoplasmic reticulum and Golgi apparatus. Trends Cell Biol 3(3): 81–88 Marra P, Maffucci T, Daniele T, Tullio GD, Ikehara Y, Chan EK, Luini A, Beznoussenko G, Mironov A, De Matteis MA (2001) The GM130 and GRASP65 Golgi proteins cycle through and define a subdomain of the intermediate compartment. Nat Cell Biol 3 (12): 1101–1013 Marra P, Salvatore L, Mironov A Jr, Di Campli A, Di Tullio G, Trucco A, Beznoussenko G, Mironov A, De Matteis MA (2007) The biogenesis of the Golgi ribbon: the roles of membrane input from the ER and of GM130. Mol Biol Cell 18(5): 1595–1608 Marsh BJ (2005) Lessons from tomographic studies of the mammalian Golgi. Biochim Biophys Acta 1744(3): 273–292 Marsh BJ, Mastronarde DN, Buttle KF, Howell KE, McIntosh JR (2001) Organellar relationships in the Golgi region of pancreatic beta cell line, HIT-T15, visualized by high resolution electron tomography. Proc Nat Acad Sci USA 98: 2399–2406 Marsh BJ, Volkmann N, McIntosh JR, Howell KE (2004) Direct continuities between cisternae at different levels of the Golgi complex in glucose-stimulated mouse islet beta cells. Proc Natl Acad Sci USA 101(15): 5565–5570  rguez JA, Geuze HJ, Slot JW, Klumperman J (1999) Vesicular tubular Martínez-Mena clusters (VTCs) between ER and Golgi mediate concentration of soluble secretory proteins by exclusion from COPI-coated vesicles. Cell 98: 81–90 Maul GG, Brinkley BR (1970) Golgi apparatus during mitosis in human melanoma cells in vitro. Cancer Res 30: 2326–2332 McMahon HT, Mills IG (2004) COP and clathrin-coated vesicle budding: different pathways, common approaches. Curr Opin Cell Biol 16: 379–391 McIntosh JR (2001) Electron microscopy of cells: a new beginning for a new century. J Cell Biol 153: F25–F32 Medigeshi GR, Schu P (2003) Characterization of the in vitro retrograde transport of MPR46. Traffic 4: 802–811

36

*

A. A. Mironov and M. Pavelka

Mellman L, Simons KS (1992) The Golgi complex: in vitro veriras? Cell 68: 829–840 Mironov A Jr, Luini A, Mironov A (1998) A synthetic model of intra-Golgi traffic. FASEB J 12: 249–252 Mironov AA Jr, Mironov AA (1998) Estimation of subcellular organelle volume from ultrathin sections through centrioles with a discretized version of vertical rotator. J Microsc 192: 29–36 Mironov AA, Beznoussenko GV, Nicoziani P, Martella O, Trucco A, Kweon HS, Di Giandomenico D, Polishchuk RS, Fusella A, Lupetti P, Berger EG, Geerts WJ, Koster AJ, Burger KN, Luini A (2001) Small cargo proteins and large aggregates can traverse the Golgi by a common mechanism without leaving the lumen of cisternae. J Cell Biol 155: 1225–1238 Mironov AA, Mironov AA Jr, Beznoussenko GV, Trucco A, Lupetti P, Smith JD, Geerts WJ, Koster AJ, Burger KN, Martone ME, Deerinck TJ, Ellisman MH, Luini A (2003) ER-toGolgi carriers arise through direct en bloc protrusion and multistage maturation of specialized ER exit domains. Dev Cell 5: 583–594 Mironov AA, Beznoussenko GV, Polishchuk RS, Trucco A (2005) Intra-Golgi transport. A way to a new paradigm? BBA Mol Cell Res 1744: 340–350 € bius W, Ohno-Iwashita Y, Van Donselaar EG, Oorschot VM, Shimada Y, Fujimoto T, Mo Heijnen HF, Geuze HJ, Slot JW (2002) Immunoelectron microscopic localization of cholesterol using biotinylated and non-cytolytic perfringolysin O. J Histochem Cytochem 50(1): 43–55 € bius W, Van Donselaar E, Ohno-Iwashita Y, Shimada Y, Heijnen HF, Slot JW, Geuze Mo HJ (2003) Recycling compartments and the internal vesicles of multivesicular bodies harbor most of the cholesterol found in the endocytic pathway. Traffic 4(4): 222–231 Mogelsvang S, Marsh BJ, Ladinsky MS, Howell KE (2004) Predicting function from structure: 3D structure studies of the mammalian Golgi complex. Traffic 5: 338–345  DJ (1978) Structural differences contrast higher plant and Mollenhauer HH, Morre animal Golgi apparatus. J Cell Sci 32: 357–362 Morin-Ganet MN, Rambourg A, Deitz SB, Franzusoff A, Kepes F (2000) Morphogenesis and dynamics of the yeast Golgi apparatus. Traffic 1: 56–68  DJ, Keenan TW, Huang CM (1974) Membrane flow and differentiation: origin of Morre Golgi apparatus membranes from endoplasmic reticulum. Adv Cytopharmacol 2: 107–125  DJ, Kartenbeck J, Franke WW (1979) Membrane flow and intercoversions among Morre endomembranes. Biochim Biophys Acta 559(1): 71–152 Motta PM, Nottola SA, Familiari G, Macchiarelli G, Vizza E, Correr S (1995) Ultrastructure of human reproduction from folliculogenesis to early embryo development. A review. Ital J Anat Embryol 100(4): 9–72 Nardini M, Spano S, Cericola C, Pesce A, Massaro A, Millo E, Luini A, Corda D, Bolognesi M (2003) CtBP/BARS: a dual-function protein involved in transcription co-repression and Golgi membrane fission. EMBO J 22: 3122–3130 Narula N, McMorrow I, Plopper G, Doherty J, Matlin KS, Burke B, Stow JL (1992) Identification of a 200-kD, brefeldin-sensitive protein on Golgi membranes. J Cell Biol 117(1): 27–38 Narula N, Stow JL (1995) Distinct coated vesicles labeled for p200 bud from trans-Golgi network membranes. Proc Natl Acad Sci USA 92(7): 2874–2878 € hr A, Staehelin LA (2001) Mobile factories: Golgi dynamics in plant cells. Trends Nebenfu Plant Sci 6(4): 160–167 Novikoff AB (1964) GERL, its form and function in neurons of rat spinal ganglia. Biol Bull 127: 358 Novikoff PM, Novikoff AB, Quintana N, Hauw JJ (1971) Golgi apparatus, GERL, and lysosomes of neurons in rat dorsal root ganglia, studied by thick section and thin section cytochemistry. J Cell Biol 50(3): 859–886

Golgi apparatus as a crossroads

*

37

Novikoff AB, Novikoff PM (1977) Cytochemical contributions to differentiating GERL from the Golgi apparatus. Histochem J 9(5): 525–551 Novikoff PM, Yam A (1978) Sites of lipoprotein particles in normal rat hepatocytes. J Cell Biol 76(1): 1–11 Oprins A, Duden R, Kreis TE, Geuze HJ, Slot JW (1993) b-COP localizes mainly to the cis-Golgi side in exocrine pancreas. J Cell Biol 121: 49–59 Orci L, Montesano R, Meda P, Malaisse-Lagae F, Brown D, Perrelet A, Vassalli P (1981) Heterogeneous distribution of filipin–cholesterol complexes across the cisternae of the Golgi apparatus. Proc Natl Acad Sci USA 78(1): 293–297 Orci L, Glick BS, Rothman JE (1986) A new type of coated vesicular carrier that appears not to contain clathrin: its possible role in protein transport within the Golgi stack. Cell 46: 171–184 Orci L, Perrelet A, Rothman JE (1998) Vesicles on strings: morphological evidence for processive transport within the Golgi stack. Proc Natl Acad Sci USA 95(5): 2279–2283 Orci L, Amherdt M, Ravazzola M, Perrelet A, Rothman JE (2000a) Exclusion of Golgi residents from transport vesicles budding from Golgi cisternae in intact cells. J Cell Biol 150: 1263–1270 Orci L, Ravazzola M, Volchuk A, Engel T, Gmachl M, Amherdt M, Perrelet A, Sollner TH, Rothman JE (2000b) Anterograde flow of cargo across the Golgi stack potentially mediated via bidirectional “percolating” COPI vesicles. Proc Natl Acad Sci USA 97 (19): 10400–10405 Palade G (1975) Intracellular aspects of the process of protein synthesis. Science 189: 347–359 Paavola LG (1978a) The corpus luteum of the guinea pig. Fine structure at the time of maximum progesterone secretion and during regression. Am J Anat 150(4): 565–603 Paavola LG (1978b) The corpus luteum of the guinea pig. II. Cytochemical studies on the Golgi complex, GERL, and lysosomes in luteal cells during maximal progesterone secretion. J Cell Biol 79(1): 45–58 Paavola LG (1978c) The corpus luteum of the guinea pig. III. Cytochemical studies on the Golgi complex and GERL during normal postpartum regression of luteal cells, emphasizing the origin of lysosomes and autophagic vacuoles. J Cell Biol 79(1): 59–73 Pagano RE, Sepanski MA, Martin OC (1989) Molecular trapping of a fluorescent ceramide analogue at the Golgi apparatus of fixed cells: interaction with endogenous lipids provides a trans-Golgi marker for both light and electron microscopy. J Cell Biol 109(5): 2067–2079 Papp S, Dziak E, Michalak M, Opas M (2003) Is all of the endoplasmic reticulum created equal? The effects of the heterogeneous distribution of endoplasmic reticulum Ca2þ -handling proteins. J Cell Biol 160: 475–479 Pavelka M, Ellinger A (1983) The trans Golgi face in rat small intestinal absorptive cells. Eur J Cell Biol 29: 253–261 Pavelka M, Ellinger A (1986) The Golgi apparatus in the acinar cells of the developing embryonic pancreas. I. Morphology and enzyme cytochemistry. Am J Anat 178: 215–223 Pavelka M, Ellinger A, Debbage P, Loewe C, Vetterlein M, Roth J (1998) Endocytic routes to the Golgi apparatus. Histochem Cell Biol 109: 555–570 Pelham HR, Hardwick KG, Lewis MJ (1988) Sorting of soluble ER proteins in yeast. EMBO J 7(6): 1757–1762 Polishchuk R, Di Pentima A, Lippincott-Schwartz J (2004) Delivery of raft-associated, GPI-anchored proteins to the apical surface of polarized MDCK cells by a transcytotic pathway. Nat Cell Biol 6(4): 297–307 Polishchuk RS, Mironov AA (2004) Structural aspects of Golgi function. Cell Mol Life Sci 61(2): 146–158

38

*

A. A. Mironov and M. Pavelka

Preisinger C, Short B, De Corte V, Bruyneel E, Haas A, Kopajtich R, Gettemans J, Barr FA (2004) YSK1 is activated by the Golgi matrix protein GM130 and plays a role in cell migration through its substrate 14-3-3{zeta}. J Cell Biol 164: 1009–1020 Rabouille C, Hui N, Hunte F, Kieckbusch R, Berger EG, Warren G, Nilsson T (1995) Mapping the distribution of Golgi enzymes involved in the construction of complex oligosaccharides. J Cell Sci 108(Pt. 4): 1617–1627 Rambourg A, Clermont Y, Hermo L (1979) Three-dimensional architecture of the Golgi apparatus in Sertoli cells of the rat. Am J Anat 154: 455–476 Rambourg A, Clermont Y (1990) Three-dimensional electron microscopy: structure of the Golgi apparatus. Eur J Cell Biol 51(2): 189–200 Rambourg A, Clermont Y (1997) Three-dimensional structure of the Golgi apparatus in €user, Basel, mammalian cells. In: Roth J, Berger EG (eds) The Golgi apparatus. Birkha pp 37–61  tien M, Olivier L (1993) Modulation of the Golgi Rambourg A, Clermont Y, Chre apparatus in stimulated and nonstimulated prolactin cells of female rats. Anat Rec 235(3): 353–362 Rizzolo LJ, Finidori J, Gonzalez A, Arpin M, Ivanov IE, Adesnik M, Sabatini DD (1985) Biosynthesis and intracellular sorting of growth hormone-viral envelope glycoprotein hybrids. J Cell Biol 101(4): 1351–1362 Robinson MS, Bonifacino JS (2001) Adaptor-related proteins. Curr Opin Cell Biol 13(4): 444–453 Rothman JE (1994) Mechanisms of intracellular protein transport. Nature 372: 55–63 Rudick VL, Rudick MJ, Brun-Zinkernagel AM (1993) Amplification of the Golgi complex in MDCK cells secreting human growth hormone. J Cell Sci 104 (Pt. 2): 509–520 Sannerud R, Saraste J, Goud B (2003) Retrograde traffic in the biosynthetic-secretory route: pathways and machinery. Curr Opin Cell Biol 15: 438–445 Sasaki T, Motegi N, Higashi S (1984) Morphological analysis of the Golgi apparatus in rat amelogenesis as revealed by the Ur–Pb–Cu block staining method and freeze fracture replication. J Electron Microsc 33: 19–33 Seemann J, Jokitalo E, Pypaert M, Warren G (2000) Matrix proteins can generate the higher order architecture of the Golgi apparatus. Nature 407: 1022–1026 Sesso A, De Faria FP, Iwamura ESM, Correa H (1994) A three-dimensional reconstruction study of the rough ER–Golgi interface in serial thin sections of the pancreatic acinar cell of the rat. J Cell Sci 107: 517–528 Shewan AM, Van Dam EM, Martin S, Luen TB, Hong W, Bryant NJ, James DE (2003) GLUT4 recycles via a trans-Golgi network (TGN) subdomain enriched in Syntaxins 6 and 16 but not TGN38: involvement of an acidic targeting motif. Mol Biol Cell 14: 973–986 Stephens DJ, Crump CM, Clarke AR, Banting G (1997) Serine 331 and Tyrosine 333 are both involved in the interaction between the cytosolic domain of TGN38 and the g2 subunit of the AP2 clathrin adaptor complex. J Biol Chem 272: 14104–14109 Stephens DJ, Lin-Marq N, Pagano A, Pepperkok R, Paccaud JP (2000) COPI-coated ER-toGolgi transport complexes segregate from COPII in close proximity to ER exit sites. J Cell Sci 113: 2177–2185 Stinchcombe JC, Nomoto H, Cutler DF, Hopkins CR (1995) Anterograde and retrograde traffic between the rough endoplasmic reticulum and the Golgi complex. J Cell Biol 131(6 Pt 1): 1387–1401 Sutterlin C, Hsu P, Mallabiabarrena A, Malhotra V (2002) Fragmentation and dispersal of the pericentriolar Golgi complex is required for entry into mitosis in mammalian cells. Cell 109: 359–369 Tanaka K, Mitsushima A, Fukudome H, Kashima Y (1986) Three-dimensional architecture of the Golgi complex observed by high resolution scanning electron microscopy. J Submicrosc Cytol 18(1): 1–9

Golgi apparatus as a crossroads

*

39

Taylor RS, Jones SM, Dahl RH, Nordeen MH, Howell KE (1997) Characterization of the Golgi complex cleared of proteins in transit and examination of calcium uptake activities. Mol Biol Cell 8: 1911–1931 Thomas G (2002) Furin at the cutting edge: from protein traffic to embryogenesis and disease. Nat Rev Mol Cell Biol 3: 753–766 Thorne-Tjomsland G, Clermont Y, Tang X (1991) Glucose-6-phosphatase activity of endoplasmic reticulum and Golgi apparatus in spermatocytes and spermatids of the rat: an electron microscopic cytochemical study. Biol Cell 71: 33–41 Tooze SA, Tooze J, Warren G (1988) Site of addition of N-acetyl-galactosamine to the E1 glycoprotein of mouse hepatitis virus-A59. J Cell Biol 106(5): 1475–1487 Tooze SA (1998) Biogenesis of secretory granules in the trans-Golgi network of neuroendocrine and endocrine cells. Biochim Biophys Acta 1404: 231–244 Traub LM, Kornfeld S (1997) The trans-Golgi network: a late secretory sorting station. Curr Opin Cell Biol 9: 527–533 Traub LM (2005) Common principles in clathrin-mediated sorting at the Golgi and the plasma membrane. Biochim Biophys Acta 1744(3): 415–437 Trucco A, Polishchuk RS, Martella O, Di Pentima A, Fusella A, Di Giandomenico D, San Pietro E, Beznoussenko GV, Polishchuk EV, Baldassarre M, Buccione R, Geerts WJ, Koster AJ, Burger KN, Mironov AA, Luini A (2004) Secretory traffic triggers the formation of tubular continuities across Golgi sub-compartments. Nat Cell Biol 6(11): 1071–1081 Uchiyama Y (1982) The membrane association among the rough- and smooth- surfaced endoplasmic reticulum and Golgi apparatus in rat hepatocytes at certain circadian stages. Tohoku J Exp Med 136: 299–302  rez-Vilar J, García-Herdugo G, Navas P (1988) Detection of Velasco A, Hidalgo J, Pe glycosaminoglycans in the Golgi complex of chondrocytes. Eur J Cell Biol 47(2): 241–250 Vivero-Salmeron G, Ballesta J, Martinez-Menarguez JA (2008) Heterotypic tubular connections at the endoplasmatic reticulum–Golgi interface. Histochem Cell Biol (In press) Weidman P, Roth R, Heuser J (1993) Golgi membrane dynamics imaged by freeze-etch electron microscopy: views of different membrane coatings involved in tubulation versus vesiculation. Cell 75(1): 123–133 Weidman P (1995) Anterograde transport through the Golgi complex: do Golgi tubules hold the key? Trends Cell Biol 5: 302–305 Williams LM, Lafontane JC (1985) An electron microscope thick section study of endomembrane organization in the Myx-amoebae of Physarum polycephalum. Protoplasma 124: 42–51 Young J, Stauber T, Del Nery E, Vernos I, Pepperkok R, Nilsson T (2005) Regulation of microtubule-dependent recycling at the trans-Golgi network by Rab6A and Rab6A0 . Mol Biol Cell 16: 162–177 Zehavi-Feferman R, Burgess JW, Stanley KK (1995) Control of p62 binding to TGN38/41 by phosphorylation. FEBS Lett 368(1): 122–124 Zeuschner D, Geerts WJ, Van Donselaar E, Humbel BM, Slot JW, Koster AJ, Klumperman J (2006) Immuno-electron tomography of ER exit sites reveals the existence of free COPII-coated transport carriers. Nat Cell Biol 8: 377–383

Main machineries operating at the Golgi apparatus

SNAREs

*

43

SNAREs David K. Banfield and Wanjin Hong

Introduction Eukaryotic cells contain multiple membrane-bound compartments between which proteins and lipid molecules are continually shuttled via membranebound vesicular carriers. Despite the constant flux of proteins and lipid through these compartments their functional and composition integrity is maintained. While the molecular machinery involved in vesicle recognition and fusion can often be transport-step/fusion-event specific, one group of proteins – the SNAREs (soluble N-ethylmaleimide-sensitive factor attachment protein receptors) play a common and central role in this process. Transport-step-specific combinations of SNARE proteins, localized to the vesicle and the target organelle, form complexes that facilitate the final step leading to the fusion of vesicles with their cognate target organelles. In general, the role of SNAREs appears to be conserved irrespective of their location of function in the cell, and much of what has been established for SNAREs in a particular trafficking pathway or organelle, is broadly applicable to SNAREs that function in the Golgi. Here we review Golgi SNAREs and the role they play in membrane and protein trafficking in the Golgi apparatus with, a particular emphasis on their functions in yeast and human cells.

General features of Golgi SNAREs The majority of SNARE proteins that function in the Golgi are type II integral membrane proteins anchored in the lipid bilayer by virtue of their single C-terminal transmembrane domain (TMD) see Fig. 1. The TMDs of SNAREs are presumably crucial for the stable association of SNARE proteins with membranes, but also play a role in establishing the steady-state distribution of SNAREs in the Golgi (Banfield et al. 1994; Rayner and Pelham 1997; Watson and Pessin 2001). In addition, in vitro fusion assays have established that the transmembrane domains of v-SNAREs (Xu et al. 2005) and of Qa-SNAREs (Han et al. 2004) are important for the formation of the hemi-fusion intermediates that precede membrane fusion and vesicle content mixing with the target compartment. Adjacent to the TMD is a short stretch of amimo acids (10 in length) referred to as the membrane proximal region (MPR). The amino acid sequence and length of this region is not evolutionarily conserved. The MPR

44

*

D. K. Banfield and W. Hong

Figure 1. General features of Golgi SNAREs. Based on their structural and functional features Golgi SNAREs are grouped into four categories. Category I is comprised of the Qa-SNAREs and some Qb- and Qc-SNAREs. Category II is mainly comprised of the Qc-SNAREs. Category III the R-SNARE Sec22p/Sec22b and category IV the R-SNARE Ykt6. The filled rectangles denote the location of the membrane proximal region (MPR) see Table 1 and the text for further details.

serves to separate the TMD from the SNARE-motif, which precedes it. The length of the MPR appears to be important for the function of some SNAREs, at least in vitro (McNew et al. 1999, 2000; Melia et al. 2002) however, whether these observations extend to Golgi-localized SNAREs is presently not known. The SNARE-motif is comprised of a number of heptad-repeats, typically 7–8, which are responsible for the formation of the amphipathic helical bundles characteristic of SNARE complexes. An evolutionarily conserved amino acid residue that occupies a central position in the SNARE-motif, and which contributes to the zero ionic layer of SNARE complexes, is the basis of a SNARE protein family classification scheme (see below). In addition to the so-called SNARE-motif or core domain, SNARE proteins also contain N-terminal extensions (N-terminal domain (NTD)) of varying length and folds (see Figs. 1, 2 and Table 1). Golgi Qa- and Qb-SNAREs contain a domain which adopts a three-helix fold, termed an Habc domain, whereas Golgi R-SNAREs, with the exception of VAMP4, contain a longin fold. Golgi Qc-SNAREs typically contain short N-terminal regions that are predicted to be unstructured, although the NTDs of the Qc-SNAREs Tlg1p and Syntaxin 6, likely adopt an Habc fold. The longin fold, found in Golgi R-SNAREs, is also present in several sub-units of the Golgi-localized vesicle tethering complexes TRAPPI and TRAPPII (Kim et al. 2006) and is predicted to be present in two sub-units of the Golgi vesicle coat complex – coatomer, although the significance of this is not presently understood (Schlenker et al. 2006). In some cases the N-terminal domains of Golgi SNAREs are capable of binding to their respective SNARE-motif, in which case the SNARE is said to adopt a closed or folded-back conformation. Folded-back conformations are known to occur for the R-SNAREs Ykt6p (Tochio et al. 2001) and Sec22p

SNAREs

*

45

Figure 2. The N-terminal domain folds of Golgi SNAREs. (A) Cartoon representation of the crystral structure of the human Vti1b Habc domain (Miller et al. 2007; pdb accession number 2qyw) viewed from the side. The three helices of the domain are labelled from N – to C  a, b and c. (B) The same structure is in (A) but viewed from the N-terminus down the three helix bundle. (C) The NMR-derived solution structure of the longin domain of yeast Ykt6p (Tochio et al. 2001; pdb accession number 1h8m). The cartoons represent 180 rotations of one another.

(Mancias and Goldberg 2007). For, Ykt6p this conformation appears to be important for the protein’s stability and likely plays a key role in the targeting of this protein by regulating the association of the cytoplasmic prenylated form of Ykt6 with membranes (Tochio et al. 2001; Fukasawa et al. 2004;

46

*

D. K. Banfield and W. Hong

Table 1. Yeast and human Golgi-resident SNAREs Type

Human

Yeast homolog

TMD

N-terminal extention

N-terminal fold

Qa

Syntaxin 5

Sed5p

Yes

Yes

Habc

Syntaxin 16

Tlg2p

Yes

Yes

Habc (predicted)

Syntaxin 10

–

Yes

Yes

Habc (predicted)

Syntaxin 11

–

No

Yes

Habc (predicted)

GS27 (membrin, GOS-27)

Bos1p

Yes

Yes

Habc (predicted)

Vti1a (Vti1-rp2)

Vti1p

Yes

Yes

Habc

GS28 (GOS-28)

Gos1p

Yes

Yes

Habc (predicted)

Syntaxin 6

Tlg1p

Yes

Yes

Habc (predicted)

Bet1

Bet1p

Yes

No

Random coil

GS15

Sft1p

Yes

No

Random coil

Sec22b (ERS-24)

Sec22p

Yes

Yes

Longin

Ykt6

Ykt6p

No (prenyl)

Yes

Longin

VAMP4

–

Yes

Yes

Unstructured

SNAP-29 (GS32)

–

No

No

–

Qb

Qc

R

Qb þ Qc

Hasegawa et al. 2004). For Sec22p, a folded-back conformation appears to be a prerequisite for this SNARE’s efficient incorporation into COPII-coated vesicles (Liu et al. 2004; Mancias and Goldberg 2007). The N-terminal domain of the Golgi syntaxins Sed5p (yeast)/Syn5p (mammals) are known to bind to the Sec1–Munc18 (SM) family member protein Sly1 (Yamaguchi et al. 2002; Dulubova et al. 2003; Arac et al. 2005). The association of Sly1p with Sed5p is important for the specificity of Golgi SNARE complex assembly (Peng and Gallwitz 2002) whereas the association of Sly1 with Syntaxin 5 is important for ER–Golgi transport (Williams et al. 2004). A folded-back conformation of Sed5p may be involved in the efficient packaging of this SNARE into COPIIcoated vesicles (Mossessova et al. 2003). Apparently, COPII preferentially binds Sed5p when the protein is part of the Sed5p–Bos1p–Sec22p SNARE

SNAREs

*

47

Figure 3. The topological arrangements of v- and t-SNAREs.

complex because t-SNARE assembly presumably removes the auto-inhibitory contacts of the closed conformation of the protein, exposing its COPII sorting signal (Mossessova et al. 2003).

SNARE protein classification and nomenclature Functionally, SNAREs can be classified as either v-SNAREs or t-SNAREs. v-SNAREs are localized to the transport vesicle, whereas t-SNAREs are predominantly localized to the vesicle’s target compartment. Currently, the generally accepted view is that a single membrane-anchored v-SNARE forms a SNARE-complex in trans with a heterotrimeric t-SNARE. See Fig. 3 and Table 1 for a description of yeast and human Golgi-resident SNARE proteins. SNARE proteins can be further sub-divided based on their amino acid sequence similarities and the position their homologs occupy in SNARE complex macromolecular structures (Fasshauer et al. 1998; Bock et al. 2001). The macromolecular structures of the exocytic and endocytic SNARE complexes revealed that they are parallel four-helical bundles (Sutton et al. 1998; Antonin et al. 2002). In the case of the endocytic SNARE complex, four different SNARE proteins contribute a single helix each to the complex (Fig. 4), this arrangement is also very likely to be the case for Golgi SNARE-complexes. Thus, the syntaxin sub-family has been termed the Qa-SNAREs whereas SNAREs that share the greatest degree of amino acid similarity with the Nterminal SNARE-motif of SNAP-25 (SNAP-25N) are referred to as Qb-SNAREs. Similarly, SNAREs that are most similar to the C-terminal SNARE-motif of SNAP-25 (SNAP-25C) are referred to as Qc-SNAREs. Members of the so-called VAMP family are collectively referred to as R-SNAREs. The Qa-, Qb-, Qc- and R-SNARE nomenclature refers to the presence of a highly evolutionarily conserved amino acid residue at the so-called zero ionic layer of the four-helical bundle – a glutamine for the Q-SNAREs and an arginine for the R-SNAREs (Fig. 4). The general expectation is that members of each family will occupy the equivalent position in their respective SNARE complexes as the corresponding SNARE in the exocytic and endocytic SNARE complexes – adopting a Qa:Qb:Qc:R stoichiometry often referred to as the 3Q:1R rule.

48

*

D. K. Banfield and W. Hong

Figure 4. The SNARE complex is a four-helical bundle. (A) An elongated side view cartoon representation of the macromolecular structure of the endocytic SNARE complex (Zwilling et al. 2007). The position of the zero ionic layer is indicated by the arrow. Syntaxin 6 is represented by the yellow helix, whereas Syntaxin 13, Vti1a and VAMP4 are represented by the blue, magenta and green helices, respectively. (B) An enlarged and skewed side view of the SNARE complex cartoon. The colour scheme used is as in (A). (C) A view down the helical bundle of a cartoon representation of the SNARE complex in which the amino acid residue side-chains defining the zero ionic layer are indicated. The colour scheme used is as in (A). (D) The zero ionic layer residues of the endocytic SNARE complex. Note that while Vti1a is classified as a Q-SNARE, it contributes an aspartic acid, rather than glutamine to the layer. Cartoons where generated using MacPyMOL and the pdb file 2nps.

SNAREs

*

49

Although the 3Q:1R rule is crucial for the formation of properly functioning SNARE complexes, amino acid substitution experiments have shown that it is not necessary, for example, that arginine be contributed by an R-SNARE per se (Katz and Brennwald 2000; Graf et al. 2005). However, exceptions to these general rules exist. For example, Sft1p and Bet1p, two yeast Golgi resident Qc-SNAREs, contain an aspartic acid and a serine (respectively) at the zero layer position, however the biological significance of this variability is presently unknown.

The general mode of SNARE protein function It is now generally accepted that the predominate function of SNARE proteins is to act as facilitators of intra-cellular membrane fusion events within the endomembrane system, through the formation of complexes between SNAREs on vesicles and SNAREs on organellar membranes. This association of SNAREs in trans is thought to be important in bringing the vesicle and organellar membranes close enough together to facilitate membrane fusion. In general, the SNARE complex that forms conforms to the 3Q:1R rule (Katz and Brennwald 2000). How are the individual SNARE proteins contributed to the SNARE complex? In vitro fusion assays with yeast Golgi SNAREs revealed that the heterotrimeric t-SNARE is comprised of a heavy chain – a Qa-SNARE (a syntaxin such as Sed5p or Syn5) and two different SNAREs which comprise the two t-SNARE light chains. Thus the t-SNARE consists of one Qa-SNARE together with either a Qb þ Qc, Qb þ R or Qc þ R pair defining the t-SNARE light chains. Employing this scheme the v-SNARE would be contributed by the remaining SNARE, i.e., either a Qb-, Qc- or R-SNARE, depending on the composition of the t-SNARE complex (Fig. 2). The v-SNARE is often an RSNARE, but this may not be so for SNAREs in the Golgi, as liposome fusion assays have established that the Qc-SNAREs, Bet1p and Sft1p, function as vSNAREs in this context (McNew et al. 2000; Parlati et al. 2002). However, an in vitro transport employing yeast Golgi SNAREs revealed that, in addition to Bet1p, the Qb-SNARE Bos1p and the R-SNARE Sec22p may also function as vSNAREs in transport between the ER and Golgi (Spang and Schekman 1998). Although the composition of the t-SNARE complex and it’s respective vSNARE appears to be quite rigid in vitro (Parlati et al. 2000, 2002) it seems likely that greater compositional flexibility exists in cells (Tsui and Banfield 2000; Tsui et al. 2001; Banfield 2001). In yeast, for example, some Golgi SNAREs interact with Qa-SNAREs other than Sed5p and in so doing participate in multiple transport steps (e.g. Vti1p, Ykt6p and Tlg1p). In addition, several yeast Golgi SNAREs are not essential for yeast cell growth. Given the importance of SNAREs in membrane fusion these observations have been viewed as being consistent with a functionally redundant role of SNAREs and reflexing a lack of selectivity in the composition of SNARE complexes. Despite the apparent flexibility in SNARE pairing interactions in cells, adherence to

50

*

D. K. Banfield and W. Hong

Figure 5. The Golgi SNARE cycle.

the 3Q:1R rule remains important. This requirement has been successfully exploited as a means to identify functionally interacting SNARE complexes (Graf et al. 2005). While the importance of residues in the immediate vicinity of the zero ionic layer have been documented (Stone et al. 1997; Graf et al. 2005) a comprehensive examination of the relative importance of other regions of the SNARE-motif in Golgi SNARE function is lacking. What directs the specificity of SNARE complex formation? Prior to SNARE complex formation, the v-SNARE and t-SNAREs encounter each other in cis (Fig. 5) and SNARE complex formation appears to proceed from the N- to Cterminus (Sorensen et al. 2006; Pobbati et al. 2006). The close opposition of the v- and t-SNAREs is mediated by a variety of factors, so-called tethering factors, which presumably function to ensure that only the correct SNAREs form biologically meaningful trans-complexes. The formation of cognate, fusogenic SNARE complexes between opposing membranes drives fusion. Although cartoons, such as the one depicted in Fig. 5, often show the formation of trans-SNARE complexes comprised of 1–2 complexes (for the sake of simplicity) the average number of complexes participating in one fusion reaction, based on studies on the neuronal exocytic SNARE complex, is likely to be on the order of 3–8 (Han et al. 2004; Rickman et al. 2005; Montecucco et al. 2005). The rosette-like structures that are observed to form from the association of multiple SNARE complexes may be important for mediating membrane fusion, perhaps via the transmembrane domain of SNAREs (Han et al. 2004). Following fusion of the vesicle with the Golgi, SNAREs remain bound to one another in trans. Trans-SNARE complexes are dissociated through the combined actions of a-SNAP/Sec17p and NSF/Sec18p after which, SNAREs are free to be recycled and reused in another round of vesicular transport and membrane fusion. Several Golgi SNAREs have been shown to cycle between

SNAREs

*

51

the Golgi and the ER (Ballensiefen et al. 1998; Wooding and Pelham 1998; Ossipov et al. 1999; Cosson et al. 2004), reflecting the requirement for SNAREs in anterograde as well as retrograde vesicle-mediated transport (Spang and Schekman 1998). In addition, apart from the requirement that cells reuse SNARE proteins in successive rounds of transport, it seems likely that this recycling process is intimately linked to the establishment and dynamic nature of the Golgi apparatus itself (Cosson et al. 2004).

The specificity of SNARE complex formation Table 2 lists SNARE complexes known to function in transport to the Golgi in yeast and mammalian cells. The complexes that mediate such traffic in mammalian cells have predominantly been identified by co-immune precipitation experiments. In contrast, in budding yeast this information has been obtained from a variety of approaches, including co-immune precipitation, genetic studies and in vitro mixing and fusion assays. An observation that has

Table 2. SNARE complexes known to function in transport to the Golgi Mammals

Yeast

Complex

Transport step (s)

Complex

Transport step (s)

Syntaxin 5 (Qa) GS28 (Qb) GS15 (Qc) Ykt6 (R)

Recyling endosome–TGN

Sed5p Gos1p Sft1p Ykt6p

Intra-Golgi

Syntaxin 5 GS28 Bet1 Ykt6

ERGIC – Golgi

Sed5p Bos1p Bet1p Sec22p

ER–Golgi

Syntaxin 5 GS27 Bet1 Sec22p

ER–ERGIC

Sed5p Bos1p Bet1p Ykt6pa

ER–Golgi

Syntaxin 16 Vti1a Syn6 VAMP4

Early endosome–TGN

Sed5p Gos1p Bet1pa Ykt6p

Intra-Golgi

Syntaxin 16 Vti1a Syntaxin 10 VAMP3

Late endosome–TGN

a

Assumed on the basis of over-expression experiments in sec22D (Liu and Barlowe 2004) and sft1D cells (Tsui et al. 2001). SNAREs in bold, italicized font are encoded by non-essential genes. ERGIC (ER-Golgi intermediate compartment).

52

*

D. K. Banfield and W. Hong

dogged the role of SNAREs in the specificity of membrane fusion events has been the apparent lack of specificity among many SNARE–SNARE associations. This lack of specificity is particularly apparent in in vitro mixing experiments using bacterially expressed mammalian as well as yeast SNARE proteins (Yang et al. 1999; Fasshauser et al. 1999; Tsui and Banfield 2000). Recent evidence suggests that non-cognate SNARE complexes form in cells, but that cells have some, as yet to be identified mechanism, which selects only the physiologically relevant complexes for use in membrane fusion reactions (Bethani et al. 2007). Thus, the identification of SNARE–SNARE interactions by co-immune precipitation may not be sufficient to assign particular SNAREs to a functional complex (Bethani et al. 2007). In in vitro fusion assays using the theoretical maximum tetrameirc combinations of SNAREs encoded by the yeast genome, only 9/275 were found to be fusogenic (McNew et al. 2000; Parlati et al. 2000, 2002; Paumet et al. 2004). Two of the nine fusogenic complexes contained the Golgi Qa-SNARE Sed5p: Sed5p/Bos1p/Sec22p (t-SNARE) þ Bet1p (v-SNARE) and Sed5p/Gos1p/Ykt6p (t-SNARE) þ Sft1p (v-SNARE), complexes which mediate fusion of vesicles with the cis- and trans-Golgi, respectively. These two complexes correspond to the mammalian Syntaxin 5-containing complexes: Syntaxin 5/membrin/ Sec22b (t-SNARE) þ Bet1 (v-SNARE) (Hay et al. 1998), although some studies suggest that the v-SNARE may be Sec22b (Xu et al. 2000; Joglekar et al. 2003) and to Syntaxin 5/GS28/Ykt6 (t-SNARE) þ GS15 (v-SNARE) (Xu et al. 2002). These in vitro fusion assay data suggest that SNARE proteins encode the necessary information to direct the formation of fusogenic SNARE complexes. In yeast, Sed5p is the only syntaxin required for transport through the Golgi, however, Sec22p and Gos1p are encoded by non-essential genes. Thus in cells lacking either the SEC22 or GOS1 genes (presumably) only a single Sed5p-containing SNARE complex would remain. Additional Sed5pcontaining fusogenic SNARE complexes have been proposed to form in cells on the basis of biochemical as well as genetic studies (Tsui and Banfield 2000; Liu and Barlowe 2002) see Table 2. In cells, Ykt6p appears to be able to substitute for Sec22p (Liu and Barlowe 2002). Similarly, under conditions when the Qc-SNARE Bet1p is ectopically over-expressed, cells can survive without the Qc-SNARE Sft1p (Tsui and Banfield 2000). Thus, with the exception of the QaSNARE, yeast Golgi Qb-, Qc- and R-SNAREs display varying degrees of presumptive functional redundancy. Whether these additional complexes constitute functionally overlapping SNARE complexes, redundant complexes or complexes that form as a result of the absence of the cognate SNARE, requires further investigation. The observation that some Sed5p interacting SNAREs also form complexes with other Qa-SNAREs functioning on other organelles suggests that a single SNARE is likely to be insufficient to direct complex specificity. For example, Vti1p (Lupashin et al. 1997; Von Mollard et al. 1997) and Ykt6p (Kweon et al. 2003) bind to multiple Qa-SNAREs and function in multiple transport pathways. Combinatorial binding interactions may therefore influence the specificity of

SNAREs

*

53

SNARE complex formation (Banfield 2001) in vivo. Finally, the extent to which regions of the SNARE-motif, other than the zero ionic layer, contribute to the specificity of Golgi SNARE complex assembly is an important issue which remains to be addressed.

i-SNAREs In vitro mixing studies with the soluble forms of Sed5p and its Golgi SNARE binding partners revealed far more ternary complexes than were identified on the basis of SNARE-mediated liposome fusion assays (Tsui et al. 2001; McNew et al. 2000; Parlati et al. 2002). The presence of more than two fusion competent Sed5p-containing SNARE complexes would help to reconcile conceptual problems arising from the fact that one SNARE from each of these complexes is non-essential – Sec22p and Gos1p, respectively (see Table 2) (McNew et al. 2000; Parlati et al. 2002). However, another explanation has been proposed to account for the additional Sed5p-containing SNARE complexes observed in in vitro mixing studies with soluble SNAREs. Using their well established SNARE-mediated liposome fusion assay Varlamov et al. (2004) established that certain sub-units of the cis-Golgi SNARE complex could inhibit fusion mediated by the trans-Golgi SNARE complex and vice versa (Varlamov et al. 2004) – the authors termed these SNAREs as i-SNAREs. While the opposing distribution of cis- and trans-Golgi SNAREs (Volchuk et al. 2004) could in principle account for the distribution of the fusogenic SNARE complexes, the authors argue that i-SNAREs would enhance this phenomena – essentially fine-tuning the specificity of membrane fusion events in the Golgi. While this is a particularly attractive notion, the concept of i-SNAREs functioning in the Golgi still awaits in vivo validation.

Localization of SNAREs to the Golgi In general SNAREs are predominantly localized to the vesicles and compartments on which they function and are absent from those on which they do not. How are SNAREs localized to the Golgi? Accumulating evidence suggests that both the transmembrane domain of SNAREs as well as signals in their cytoplasmic domains accounts for their steady-state distributions. A requirement of the transmembrane domain in the localization of Golgi enzymes is well established in mammalian cells. Such studies have led to the proposals that (1) the length of the transmembrane is an important factor in Golgi localization and that sorting/or localization is the result of Golgi membrane bilayer thickness (Bretcher and Munro 1993) or (2) that the TMDs of Golgi residents oligomerize and are prevented from exiting the Golgi (Nilsson et al. 1993). The transmembrane domains of the yeast SNAREs Sed5p and Sft1p contribute to their Golgi localization (Banfield et al. 1994; Rayner and Pelham 1997) and TMD length has been shown to be important for the Golgi localization of Syn5 in mammalian cells (Watson and Pessin 2001). However

54

*

D. K. Banfield and W. Hong

in yeast, TMD length alone is not sufficient to ensure exclusive Golgi localization of SNARE protein chimeras (Rayner and Pelham 1997). The steady-state distribution of SNAREs proteins in the Golgi appears be dynamic – relying on the active retrieval of SNAREs from distal cisternae and their retrieval to earlier sub-compartments or to the ER, from where they return to Golgi. In yeast, Sed5p (Wooding and Pelham 1998), Sec22p (Ballensiefen et al. 1998) and Bos1p (Ossipov et al. 1999) have been shown to cycle between the Golgi and the ER. Although yeast Bet1p does not undergo such cycling (Ossipov et al. 1999), in mammalian cells Bet1 and Sec22b have been shown to continually cycle between the Golgi and the ER (Hay et al. 1998). The recycling of Bet1 does not appear to require interactions with its cognate SNAREs (Joglekar et al. 2003). In the case of yeast Sec22p and Bos1p, the retrieval of these SNAREs from the Golgi requires a functional COPI coat (Ballensiefen et al. 1998; Ossipov et al. 1999). An analysis of the lateral distribution and vesicle incorporation of SNAREs in the mammalian Golgi using electron microscopy is also consistent with a dynamic localization mechanism (Cosson et al. 2005). The Golgi SNARE Ykt6 does not contain a TMD, but rather is dually lipid modified at its C-terminus (Fukasawa et al. 2004). The farnesylated form of Ykt6 resides in the cytoplasm whereas farnesylated, palmitoylated Ykt6 is found predominantly on Golgi membranes in non-neuronal mammalian cells (Fukasawa et al. 2004; Hasegawa et al. 2004). How Ykt6 is targeted to Golgi membranes remains to be determined.

SNAREs and COPI interactions Several SNAREs have been shown cycle within or from the Golgi in a COPIdependent manner, data that implies an interaction between these SNAREs and the coat protein complex. The COPI coat is comprised of the heptameric complex termed coatomer, together with the GTPase Arf1. Arf1 cycles on and off Golgi membranes as a function of its nucleotide-bound state. GTP-bound Arf localizes to membranes whereas GDP-bound Arf is found in the cytoplasm. The nucleotide status on Arf1 is controlled through the action of its exchange factor (Arf GEF) and its activating protein (Arf GAP). In vitro studies using yeast Golgi SNAREs revealed that the Arf1p GAPs, Glo3p and Gsc1p, act catalytically on Golgi SNAREs promoting a conformational change that facilitates stoichiometric recruitment of Arf1p to SNAREs (Rein et al. 2002). In agreement with these findings, studies in mammalian cells have identified a motif on Arf that is required for the recruitment of Arf to Golgi membranes by the Qb-SNARE membrin (Honda et al. 2005). Schindler and Spang (2007) have shown that Gcs1p accelerates the formation of SNARE complexes in vitro and suggested that Arf GAPs may function as folding chaperones for SNAREs. Such mechanisms may function to couple SNARE recruitment to vesicle formation in cells, thus ensuring that each vesicle carries sufficient SNAREs capable of forming cognate SNARE complexes at its target compartment.

SNAREs

*

55

The role of SNAREs in the morphological and functional organization of the Golgi Presumably recycling Golgi SNAREs is important for ensuring efficient vesiclemediated transport, as this would allow these proteins to be employed in successive rounds of trafficking. Mathematical modeling using a minimal system, in which the variables were restricted to cytoplasmic coat protein complexes and SNAREs, was sufficient to generate stable non-identical compartments (Heinrich and Rapoport 2005). A requirement of Heinrich and Rapoport’s (2005) model was that each vesicle generating coat complex preferentially bound and packaged a characteristic set of SNAREs. The lateral distribution of Golgi SNAREs observed by Cosson et al. (2005) may similarly reflect differential affinity of Golgi SNAREs for the COPI coat in vivo. Thus, the affinity of vesicle coats, or their cargo sorting affiliated partners, may function to promote and maintain the compositional integrity of Golgi cisternia through their intrinsic ability to bind different SNAREs with varying affinities.

Regulators of Golgi SNARE function The activity of SNAREs is regulated at various stages of their action including the assembly post-translational of the t-SNARE and the assembly of the trans-SNARE complex (Fig. 5). A variety of proteins have been identified that modulate the activity of SNAREs. In addition, post-translational modifications such prenylation, palmitoylation and phosphorylation also influence the activity and or localization of SNAREs.

NSF/Sec18p and a-SNAP/Sec17p NSF and a-SNAP represent two co-operating core regulators of SNARE protein activity. These proteins are responsible for the disassembly of cis-SNARE complexes (Fig. 5), an activity that frees-up SNAREs to be used in successive rounds of vesicle fusion. Three molecules of a-SNAP link the cis-SNARE complex with a hexamer of the ATPase NSF/Sec18p and together this complex is referred to as the 20 S complex (Hohl et al. 1998; Wimmer et al. 2001; Furst et al. 2003; Brunger and DeLaBarre 2003). NSF contains two ATPase domains termed D1 and D2. The D2 ATPase domain mediates the formation of the NSF hexamer whereas the D1 ATPase domain effects the dissociation of the cisSNARE complex. The association of NSF with a-SNAP into the 20 S complex stimulates the ATPase activity of NSF (Marz et al. 2003).

The Sec1/Munc-18 like (SM) proteins Sec1/Munc-18 (SM) proteins bind directly to SNAREs and act downstream of vesicle tethering events. Sly1p, the yeast SM protein which binds to Sed5p, was identified because a mutant of this protein (sly1-20p) could suppress the loss of the essential Rab/Ypt GTPase Ypt1p (Dascher et al. 1991). The association of Sly1p with Sed5p has been shown to enhance Sed5p-containing transSNARE complexes (Kosodo et al. 2002), and to be important for the specificity

56

*

D. K. Banfield and W. Hong

of Golgi SNARE complex assembly (Peng and Gallwitz 2002) whereas the assocation of Sly1 with Syntaxin 5 has been demonstrated to be important for ER–Golgi transport (Williams et al. 2004). However, the interaction between Sed5p and Sly1p is dispensible for transport (Peng and Gallwitz 2004). Unlike the exocytic and neuronal syntaxins and their requisite SM protein interactions, the interaction between Sly1p and Sed5p is mediated by a short stretch of amino acids at the N-terminus of the protein which does not involve either the Habc or SNARE-motif domains (Yamaguchi et al. 2002; Bracher and Weissenhorn 2002; Dulubova et al. 2003) and thus this association does not promote a folded-back conformation for Sed5p. A similar mode of binding is also evident between the SM protein Vps45/Vps45p and the Qa-SNARE Syntaxin 16/Tlg2p (Dulubova et al. 2002). It is now apparent that Sly1p is capable of binding to non-syntaxin SNAREs as well as to SNARE complexes and that this SNARE binding property of Sly1p is important in the specificity of cognate SNARE complex formation (Peng and Gallwitz 2002, 2004; Li et al. 2005).

The Golgins The Golgins are a class of large coiled-coil containing Golgi localized proteins with roles in tethering vesicles to the Golgi. Some golgins contain a single Cterminal transmembrane domain whereas other members of the family are peripherally associated with Golgi membranes. The peripheral membrane protein p115 has been shown to bind directly to SNAREs involved in ERGolgi intermediate compartment (ERGIC) as well as ERGIC-Golgi transport (Allan et al. 2000; Shorter et al. 2002). A SNARE-related coiled-coil region of p115 interacts with many Golgi SNAREs and such interactions likely promote the formation of trans-SNARE complexes (Shorter et al. 2002). The functional consequences of such interactions may be to ensure a direct connection between the tethering machinery and SNAREs as well as to facilitate the recruitment of p115 to membrane sites where unassembled SNAREs are located (Brandon et al. 2006; Bentley et al. 2006). Mutational analysis of p115 suggests that the SNARE-modulating activity of the protein is more important than its tethering activity in maintaining the structure and function of the Golgi (Puthenveedu and Lindstedt 2004). Uso1p is the yeast homologue of p115 (Sapperstein et al. 1995, 1996; Cao et al. 1998). Other Golgins have also been shown to bind to Golgi SNAREs, including GM130 which interacts directly with Syntaxin 5 (Diao et al. 2007) and GCC185, which binds directly to Syntaxin 16 (Ganley et al. 2008). The emerging picture of the role of Golgins is one in which these proteins sequester Rab GTPases and Qa-SNAREs/syntaxins, and in so doing keep these two key proteins in close proximity to the tether.

Conserved oligomeric Golgi (COG) The COG complex is a member of the oligomeric vesicle tethering factor family which comprise a structurally diverse group of peripheral membrane

SNAREs

*

57

protein complexes involved in vesicle-organellar tethering events prior to SNARE complex assembly (Oka and Krieger 2005; Stzul and Lupashin 2006). The COG complex is involved in retrograde trafficking of Golgi-resident proteins and extensive genetic interactions have been documented between the COG complex and SNAREs. In addition the yeast and mammalian COG complexes co-immune precipitate with Golgi SNAREs (Suvorova et al. 2002; Zolov and Lupashin 2005) and the localization and stability of Golgi SNAREs is altered in cells with defective COG complex components (Oka et al. 2004; Fotso et al. 2005; Zolov and Luphasin 2005; Shestakova et al. 2007). More recently, the yeast COG complex has been shown to interact with the SNAREmotif of Sed5p and to preferentially bind to Sed5p-containing SNARE complexes, leading the authors to propose that one function of the COG complex is to stabilize intra-Golgi SNARE complexes (Shestakova et al. 2007).

GATE-16 GATE-16, a member of the ubiquitin-fold (UF) protein family, is localized to the Golgi and interacts with NSF as well as GS28 (Sagiv et al. 2000). NSF/a-SNAP facilitates the interaction of GATE-16 with GS28 in a manner that requires ATP-binding but not ATP hydrolysis. Interestingly, GATE-16 binding prevents GS28 from interacting with Syntaxin 5 and in so doing prevents the formation of a functional t-SNARE (Muller et al. 2002). In addition, the yeast GATE-16 homologue, Aut7p, interacts with Bet1p, a Qc-SNARE involved in ER-to-Golgi transport and shows genetic interactions with BET1 and the ER-Golgi RSNARE SEC22 (Legesse-Miller et al. 2000).

FIG FIG (also known as CAL, PIST and GOPC) localizes to the TGN where it interacts with the Qc-SNARE Syntaxin 6 (Charest et al. 2001). FIG contains two coiledcoil regions and a single PDZ domain and the protein’s interaction with Syntaxin 6 is mediated via the second coiled-coil region and its C-terminal flanking region. Although the biological significance of this interaction remains to be determined, knock-out of the FIG gene in mice results in selective ablation of acrosome formation during spermatogenesis (Yao et al. 2002). The acrosome is believed to form from the Golgi apparatus and the absence of FIG leads to fragmented acrosomal vesicles suggestive of a role for FIG in the fusion of vesicles into the acrosome. Curiously, FIG also interacts with Golgin-160 (Hicks and Machamer 2005).

Phosphorylation Many SNAREs and their regulatory proteins are known to be phosphorylated by a variety of kinases (Gerst 2003; Snyder et al. 2006). The yeast Golgi Qa-SNARE Sed5p is a phosphoprotein and Weinberger et al. (2005) have shown that amino acid substitutions to an evolutionarily conserved protein kinase A phosphorylation site adjacent to the transmembrane domain of the protein has dramatic effects on Golgi morphology. While expression of the

58

*

D. K. Banfield and W. Hong

pseudophosphorylated form of Sed5p (Ser317Asp) results in the accumulation of ER membranes and vesicles, expression of the non-phosphorylatable form of the protein (Ser317Ala) results in the accumulation of Golgi membranes reminiscent of the mammalian cell Golgi (Weinberger et al. 2005). The Ser317Ala mutant also shows an increased affinity for the COPI coat, suggesting that phoshorylation status of Sed5p in cells may play a role in regulating Golgi morphology.

Palmitoylation Several SNAREs are known to be palmitoylated. Some of these SNAREs lack transmembrane domains and are anchored to the membrane by their palmitate moieties, such is the case for SNAP-25, SNAP-23 and Syntaxin 11 (Vogel and Roche 1999; Veit 2000; Prekeris et al. 2000). In contrast, the Golgi SNARE Ykt6p is anchored by a combination of prenylation and palmitoylation (Fukasawa et al. 2004). In addition, it is now apparent that several SNAREs bearing TMDs are also palmitoylated (Valdez-Taubas and Pelham 2005) including the TGN/ endosomal Qc-SNARE, Tlg1p. The yeast DHCC-CDR family member Swf1p is required for palmitoylation of Tlg1p and prevention of Tlg1p palmitoylation results in its ubiquitination and transportation, via the multivesicular body, to the vacuole for degradation. While palymitoyation of TMD-anchored SNAREs does not appear to be essential for their function, this modification may play a role in the membrane partitioning of these SNAREs and or in dissociation of these modified proteins from other SNAREs, following fusion (Valdez-Taubas and Pelham 2005). Based on amino acid sequence similarities with Tlg1p, the mammalian Golgi resident SNAREs Syntaxin 6, Syntaxin 10 and VAMP4 may also be substrates for palymitoylation (Valdez-Taubas and Pelham 2005). Unlike the TMD-anchored SNAREs, which are modified via DHHC-CDR palmitoyltransferases, Ykt6 appears to be capable of mediating its own palymitoylation (Veit 2004) via its longin fold (Dietrich et al. 2004). Ykt6 is found in two pools in cells–a cytoplasmic pool and a membrane associated pool. Ykt6 lacks a proteinaceous membrane anchor but contains a prenylation consensus sequence (a so-called CAAX box) at its C-terminus. The cytoplasmic pool of Ykt6 has been shown to farnesylated and the farnesylation of Ykt6 is prerequiste for the subsequent palymitoylation and membrane association of the protein (Fukasawa et al. 2004). Fukasawa et al. (2004) propose a cycle of membrane association of Ykt6 in which the farnesylated fold-back conformation of Ykt6p (mediated by an interaction between the longin domain and SNARE-motif, Tochio et al. 2001) is targeted to membranes, whereupon the protein is palymitoylated. This dual lipid modification may be required for stable membrane association of Ykt6 (Fukasawa et al. 2004).

Golgi SNAREs and apoptosis During programmed cells death (apoptosis) the Golgi loses its cisternal organization and is fragmented into clusters of tubulovesicular elements

SNAREs

*

59

(Lane et al. 2001) and the early secretory pathway is blocked (Lane et al. 2002). While this phenotype is associated with the proteolytic cleavage of members of the Golgin family (Mancini et al. 2000; Chiu et al. 2002; Lane et al. 2002) Lowe et al. (2004) have shown that Syntaxin 5 is cleaved by caspase during apoptosis. Caspase-3 cleaves Syntaxin 5 at Asp 263, separating the SNARE-motif from the N-terminal Habc domain. Syntaxin 5 participates in several SNARE complexes in the Golgi, including partnerships with Bet1 (Qc-), membrin/GS27 (Qb-) and Sec22b (R-); Bet1 (Qc-), GS28 (Qb-) and Ykt6 (R-) as well as with GS15 (Qc-), GS28 (Qb-) and Ykt6 (R-) (Hay et al. 1998; Zhang et al. 2001; Xu et al. 2002). Thus, cleavage of Syntaxin 5 by caspase-3 is likely to affect several trafficking steps to and within the Golgi.

Future perspectives While much has been learned about the role of SNAREs in the Golgi many important questions remain to be addressed. These include identification of the sorting signals/motifs on SNAREs as well as the macromolecular details governing interactions between SNAREs and the COPI vesicle generating machinery. Establishing, whether like SNAREs and the ER vesicle coat COPII (Morsomme et al. 2003), Golgi SNAREs influence the incorporation of particular cargo proteins into COPI-coated vesicles. Finally, a more detailed understanding of the mechanisms governing the steady-state localization of SNAREs to the Golgi will make important contributions to our understanding of how Golgi SNARE trafficking impacts the morphological and functional organization of this fascinating organelle.

References Allan BB, Moyer BD, Balch WE (2000) Rab1 recruitment of p115 into a cis-SNARE complex: programming budding COPII vesicles for fusion. Science 289: 444–448 Antonin W, Fasshauer D, Becker S, Jahn R, Schneider TR (2002) Crystal structure of the endosomal SNARE complex reveals common structural principles of all SNAREs. Nat Struct Biol 9: 107–111 Arac D, Dulubova I, Pei J, Huryeva I, Grishin NV, Rizo J (2005) Three-dimensional structure of the rSly1 N-terminal domain reveals a conformational change induced by binding to syntaxin 5. J Mol Biol 346: 589–601 Ballensiefen W, Ossipov D, Schmitt HD (1998) Recycling of the yeast v-SNARE Sec22p involves COPI-proteins and the ER transmembrane proteins Ufe1p and Sec20p. J Cell Sci 111(Pt 11): 1507–1520 Banfield DK (2001) SNARE complexes – is there sufficient complexity for vesicle targeting specificity? Trends Biochem Sci 26: 67–68 Banfield DK, Lewis MJ, Rabouille C, Warren G, Pelham HR (1994) Localization of Sed5, a putative vesicle targeting molecule, to the cis-Golgi network involves both its transmembrane and cytoplasmic domains. J Cell Biol 127: 357–371 Bentley M, Liang Y, Mullen K, Xu D, Sztul E, Hay JC (2006) SNARE status regulates tether recruitment and function in homotypic COPII vesicle fusion. J Biol Chem 281: 38825–38833

60

*

D. K. Banfield and W. Hong

Bethani I, Lang T, Geumann U, Sieber JJ, Jahn R, Rizzoli SO (2007) The specificity of SNARE pairing in biological membranes is mediated by both proof-reading and spatial segregation. EMBO J 26: 3981–3992 Bock JB, Matern HT, Peden AA, Scheller RH (2001) A genomic perspective on membrane compartment organization. Nature 409: 839–841 Bracher A, Weissenhorn W (2002) Structural basis for the Golgi membrane recruitment of Sly1p by Sed5p. EMBO J 21: 6114–6124 Brandon E, Szul T, Alvarez C, Grabski R, Benjamin R, Kawai R, Sztul E (2006) On and off membrane dynamics of the endoplasmic reticulum-Golgi tethering factor p115 in vivo. Mol Biol Cell 17: 2996–3008 Bretscher MS, Munro S (1993) Cholesterol and the Golgi apparatus. Science 261: 1280–1281 Brunger AT, DeLaBarre B (2003) NSF and p97/VCP: similar at first, different at last. FEBS Lett 555: 126–133 Cao X, Ballew N, Barlowe C (1998) Initial docking of ER-derived vesicles requires Uso1p and Ypt1p but is independent of SNARE proteins. EMBO J 17: 2156–2165 Charest A, Lane K, McMahon K, Housman DE (2001) Association of a novel PDZ domaincontaining peripheral Golgi protein with the Q-SNARE (Q-soluble N-ethylmaleimide-sensitive fusion protein (NSF) attachment protein receptor) protein syntaxin 6. J Biol Chem 276: 29456–29465 Chiu R, Novikov L, Mukherjee S, Shields D (2002) A caspase cleavage fragment of p115 induces fragmentation of the Golgi apparatus and apoptosis. J Cell Biol 159: 637–648 Cosson P, Ravazzola M, Varlamov O, Sollner TH, Di Liberto M, Volchuk A, Rothman JE, Orci L (2005) Dynamic transport of SNARE proteins in the Golgi apparatus. Proc Natl Acad Sci USA 102: 14647–14652 Diao A, Frost L, Morohashi Y, Lowe M (2007) Coordination of golgin tethering and SNARE assembly: GM130 binds syntaxin 5 in a p115-regulated manner. J Biol Chem 283: 6957–6967 Dietrich LE, Gurezka R, Veit M, Ungermann C (2004) The SNARE Ykt6 mediates protein palmitoylation during an early stage of homotypic vacuole fusion. EMBO J 23: 45–53 Dulubova I, Yamaguchi T, Arac D, Li H, Huryeva I, Min SW, Rizo J, Sudhof TC (2003) Convergence and divergence in the mechanism of SNARE binding by Sec1/Munc18like proteins. Proc Natl Acad Sci USA 100: 32–37 Dulubova I, Yamaguchi T, Gao Y, Min SW, Huryeva I, Sudhof TC, Rizo J (2002) How Tlg2p/ syntaxin 16 ‘snares’ Vps45. EMBO J 21: 3620–3631 Fasshauer D, Antonin W, Margittai M, Pabst S, Jahn R (1999) Mixed and non-cognate SNARE complexes. Characterization of assembly and biophysical properties. J Biol Chem 274: 15440–15446 Fasshauer D, Sutton RB, Brunger AT, Jahn R (1998) Conserved structural features of the synaptic fusion complex: SNARE proteins reclassified as Q- and R-SNAREs. Proc Natl Acad Sci USA 95: 15781–15786 Fotso P, Koryakina Y, Pavliv O, Tsiomenko AB, Lupashin VV (2005) Cog1p plays a central role in the organization of the yeast conserved oligomeric Golgi complex. J Biol Chem 280: 27613–27623 Fukasawa M, Varlamov O, Eng WS, Sollner TH, Rothman JE (2004) Localization and activity of the SNARE Ykt6 determined by its regulatory domain and palmitoylation. Proc Natl Acad Sci USA 101: 4815–4820 Furst J, Sutton RB, Chen J, Brunger AT, Grigorieff N (2003) Electron cryomicroscopy structure of N-ethyl maleimide sensitive factor at 11 A resolution. EMBO J 22: 4365–4374 Ganley IG, Espinosa E, Pfeffer SR (2008) A syntaxin 10-SNARE complex distinguishes two distinct transport routes from endosomes to the trans-Golgi in human cells. J Cell Biol 180: 159–172

SNAREs

*

61

Gerst JE (2003) SNARE regulators: matchmakers and matchbreakers. Biochim Biophys Acta 1641: 99–110 Graf CT, Riedel D, Schmitt HD, Jahn R (2005) Identification of functionally interacting SNAREs by using complementary substitutions in the conserved ‘0’ layer. Mol Biol Cell 16: 2263–2274 Han X, Wang CT, Bai J, Chapman ER, Jackson MB (2004) Transmembrane segments of syntaxin line the fusion pore of Ca2 þ -triggered exocytosis. Science 304: 289–292 Hasegawa H, Yang Z, Oltedal L, Davanger S, Hay JC (2004) Intramolecular protein– protein and protein–lipid interactions control the conformation and subcellular targeting of neuronal Ykt6. J Cell Sci 117: 4495–4508 Hay JC, Klumperman J, Oorschot V, Steegmaier M, Kuo CS, Scheller RH (1998) Localization, dynamics, and protein interactions reveal distinct roles for ER and Golgi SNAREs. J Cell Biol 141: 1489–1502 Heinrich R, Rapoport TA (2005) Generation of nonidentical compartments in vesicular transport systems. J Cell Biol 168: 271–280 Hicks SW, Machamer CE (2005) Isoform-specific interaction of golgin-160 with the Golgi-associated protein PIST. J Biol Chem 280: 28944–28951 Hohl TM, Parlati F, Wimmer C, Rothman JE, Sollner TH, Engelhardt H (1998) Arrangement of subunits in 20 S particles consisting of NSF, SNAPs, and SNARE complexes. Mol Cell 2: 539–548 Honda A, Al-Awar OS, Hay JC, Donaldson JG (2005) Targeting of Arf-1 to the early Golgi by membrin, an ER-Golgi SNARE. J Cell Biol 168: 1039–1051 Joglekar AP, Xu D, Rigotti DJ, Fairman R, Hay JC (2003) The SNARE motif contributes to rbet1 intracellular targeting and dynamics independently of SNARE interactions. J Biol Chem 278: 14121–14133 Katz L, Brennwald P (2000) Testing the 3Q:1R “rule”: mutational analysis of the ionic “zero” layer in the yeast exocytic SNARE complex reveals no requirement for arginine. Mol Biol Cell 11: 3849–3858 Kim YG, Raunser S, Munger C, Wagner J, Song YL, Cygler M, Walz T, Oh BH, Sacher M (2006) The architecture of the multisubunit TRAPP I complex suggests a model for vesicle tethering. Cell 127: 817–830 Kosodo Y, Noda Y, Adachi H, Yoda K (2002) Binding of Sly1 to Sed5 enhances formation of the yeast early Golgi SNARE complex. J Cell Sci 115: 3683–3691 Kweon Y, Rothe A, Conibear E, Stevens TH (2003) Ykt6p is a multifunctional yeast R-SNARE that is required for multiple membrane transport pathways to the vacuole. Mol Biol Cell 14: 1868–1881 Lane JD, Lucocq J, Pryde J, Barr FA, Woodman PG, Allan VJ, Lowe M (2002) Caspasemediated cleavage of the stacking protein GRASP65 is required for Golgi fragmentation during apoptosis. J Cell Biol 156: 495–509 Lane JD, Vergnolle MA, Woodman PG, Allan VJ (2001) Apoptotic cleavage of cytoplasmic dynein intermediate chain and p150(Glued) stops dynein-dependent membrane motility. J Cell Biol 153: 1415–1426 Legesse-Miller A, Sagiv Y, Glozman R, Elazar Z (2000) Aut7p, a soluble autophagic factor, participates in multiple membrane trafficking processes. J Biol Chem 275: 32966–32973 Li Y, Gallwitz D, Peng R (2005) Structure-based functional analysis reveals a role for the SM protein Sly1p in retrograde transport to the endoplasmic reticulum. Mol Biol Cell 16: 3951–3962 Liu Y, Barlowe C (2002) Analysis of Sec22p in endoplasmic reticulum/Golgi transport reveals cellular redundancy in SNARE protein function. Mol Biol Cell 13: 3314–3324 Liu Y, Flanagan JJ, Barlowe C (2004) Sec22p export from the endoplasmic reticulum is independent of SNARE pairing. J Biol Chem 279: 27225–27232

62

*

D. K. Banfield and W. Hong

Lowe M, Lane JD, Woodman PG, Allan VJ (2004) Caspase-mediated cleavage of Syntaxin 5 and giantin accompanies inhibition of secretory traffic during apoptosis. J Cell Sci 117: 1139–1150 Lupashin VV, Pokrovskaya ID, McNew JA, Waters MG (1997) Characterization of a novel yeast SNARE protein implicated in Golgi retrograde traffic. Mol Biol Cell 8: 2659– 2676 Mancias JD, Goldberg J (2007) The transport signal on Sec22 for packaging into COPIIcoated vesicles is a conformational epitope. Mol Cell 26: 403–414 Mancini M, Machamer CE, Roy S, Nicholson DW, Thornberry NA, Casciola-Rosen LA, Rosen A (2000) Caspase-2 is localized at the Golgi complex and cleaves golgin-160 during apoptosis. J Cell Biol 149: 603–612 Marz KE, Lauer JM, Hanson PI (2003) Defining the SNARE complex binding surface of alpha-SNAP: implications for SNARE complex disassembly. J Biol Chem 278: 27000–27008 McNew JA, Weber T, Engelman DM, Sollner TH, Rothman JE (1999) The length of the flexible SNAREpin juxtamembrane region is a critical determinant of SNAREdependent fusion. Mol Cell 4: 415–421 McNew JA, Weber T, Parlati F, Johnston RJ, Melia TJ, Sollner TH, Rothman JE (2000) Close is not enough: SNARE-dependent membrane fusion requires an active mechanism that transduces force to membrane anchors. J Cell Biol 150: 105–117 Melia TJ, Weber T, McNew JA, Fisher LE, Johnston RJ, Parlati F, Mahal LK, Sollner TH, Rothman JE (2002) Regulation of membrane fusion by the membrane-proximal coil of the t-SNARE during zippering of SNAREpins. J Cell Biol 158: 929–940 Montecucco C, Schiavo G, Pantano S (2005) SNARE complexes and neuroexocytosis: how many, how close? Trends Biochem Sci 30: 367–372 Morsomme P, Prescianotto-Baschong C, Riezman H (2003) The ER v-SNAREs are required for GPI-anchored protein sorting from other secretory proteins upon exit from the ER. J Cell Biol 162: 403–412 Mossessova E, Bickford LC, Goldberg J (2003) SNARE selectivity of the COPII coat. Cell 114: 483–495 Muller JM, Shorter J, Newman R, Deinhardt K, Sagiv Y, Elazar Z, Warren G, Shima DT (2002) Sequential SNARE disassembly and GATE-16-GOS-28 complex assembly mediated by distinct NSF activities drives Golgi membrane fusion. J Cell Biol 157: 1161–1173 Nilsson T, Slusarewicz P, Hoe MH, Warren G (1993) Kin recognition. A model for the retention of Golgi enzymes. FEBS Lett 330: 1–4 Oka T, Krieger M (2005) Multi-component protein complexes and Golgi membrane trafficking. J Biochem (Tokyo) 137: 109–114 Oka T, Ungar D, Hughson FM, Krieger M (2004) The COG and COPI complexes interact to control the abundance of GEARs, a subset of Golgi integral membrane proteins. Mol Biol Cell 15: 2423–2435 Ossipov D, Schroder-Kohne S, Schmitt HD (1999) Yeast ER-Golgi v-SNAREs Bos1p and Bet1p differ in steady-state localization and targeting. J Cell Sci 112(Pt 22): 4135–4142 Parlati F, McNew JA, Fukuda R, Miller R, Sollner TH, Rothman JE (2000) Topological restriction of SNARE-dependent membrane fusion. Nature 407: 194–198 Parlati F, Varlamov O, Paz K, McNew JA, Hurtado D, Sollner TH, Rothman JE (2002) Distinct SNARE complexes mediating membrane fusion in Golgi transport based on combinatorial specificity. Proc Natl Acad Sci USA 99: 5424–5429 Paumet F, Rahimian V, Rothman JE (2004) The specificity of SNARE-dependent fusion is encoded in the SNARE motif. Proc Natl Acad Sci USA 101: 3376–3380 Peng R, Gallwitz D (2004) Multiple SNARE interactions of an SM protein: Sed5p/Sly1p binding is dispensable for transport. EMBO J 23: 3939–3949

SNAREs

*

63

Peng R, Gallwitz D (2002) Sly1 protein bound to Golgi syntaxin Sed5p allows assembly and contributes to specificity of SNARE fusion complexes. J Cell Biol 157: 645–655 Pobbati AV, Stein A, Fasshauer D (2006) N- to C-terminal SNARE complex assembly promotes rapid membrane fusion. Science 313: 673–676 Prekeris R, Klumperman J, Scheller RH (2000) Syntaxin 11 is an atypical SNARE abundant in the immune system. Eur J Cell Biol 79: 771–780 Puthenveedu MA, Linstedt AD (2004) Gene replacement reveals that p115/SNARE interactions are essential for Golgi biogenesis. Proc Natl Acad Sci USA 101: 1253–1256 Rayner JC, Pelham HR (1997) Transmembrane domain-dependent sorting of proteins to the ER and plasma membrane in yeast. EMBO J 16: 1832–1841 Rein U, Andag U, Duden R, Schmitt HD, Spang A (2002) ARF-GAP-mediated interaction between the ER-Golgi v-SNAREs and the COPI coat. J Cell Biol 157: 395–404 Rickman C, Hu K, Carroll J, Davletov B (2005) Self-assembly of SNARE fusion proteins into star-shaped oligomers. Biochem J 388: 75–79 Sagiv Y, Legesse-Miller A, Porat A, Elazar Z (2000) GATE-16, a membrane transport modulator, interacts with NSF and the Golgi v-SNARE GOS-28. EMBO J 19: 1494–1504 Sapperstein SK, Lupashin VV, Schmitt HD, Waters MG (1996) Assembly of the ER to Golgi SNARE complex requires Uso1p. J Cell Biol 132: 755–767 Sapperstein SK, Walter DM, Grosvenor AR, Heuser JE, Waters MG (1995) p115 is a general vesicular transport factor related to the yeast endoplasmic reticulum to Golgi transport factor Uso1p. Proc Natl Acad Sci USA 92: 522–526 Schindler C, Spang A (2007) Interaction of SNAREs with ArfGAPs precedes recruitment of Sec18p/NSF. Mol Biol Cell 18: 2852–2863 Schlenker O, Hendricks A, Sinning I, Wild K (2006) The structure of the mammalian signal recognition particle (SRP) receptor as prototype for the interaction of small GTPases with Longin domains. J Biol Chem 281: 8898–8906 Shestakova A, Suvorova E, Pavliv O, Khaidakova G, Lupashin V (2007) Interaction of the conserved oligomeric Golgi complex with t-SNARE Syntaxin5a/Sed5 enhances intraGolgi SNARE complex stability. J Cell Biol 179: 1179–1192 Shorter J, Beard MB, Seemann J, Dirac-Svejstrup AB, Warren G (2002) Sequential tethering of Golgins and catalysis of SNAREpin assembly by the vesicle-tethering protein p115. J Cell Biol 157: 45–62 Siddiqi SA, Siddiqi S, Mahan J, Peggs K, Gorelick FS, Mansbach CM II (2006) The identification of a novel endoplasmic reticulum to Golgi SNARE complex used by the prechylomicron transport vesicle. J Biol Chem 281: 20974–20982 Snyder DA, Kelly ML, Woodbury DJ (2006) SNARE complex regulation by phosphorylation. Cell Biochem Biophys 45: 111–123 Sorensen JB, Wiederhold K, Muller EM, Milosevic I, Nagy G, De Groot BL, Grubmuller H, Fasshauer D (2006) Sequential N- to C-terminal SNARE complex assembly drives priming and fusion of secretory vesicles. EMBO J 25: 955–966 Spang A, Schekman R (1998) Reconstitution of retrograde transport from the Golgi to the ER in vitro. J Cell Biol 143: 589–599 Stone S, Sacher M, Mao Y, Carr C, Lyons P, Quinn AM, Ferro-Novick S (1997) Bet1p activates the v-SNARE Bos1p. Mol Biol Cell 8: 1175–1181 Sutton RB, Fasshauer D, Jahn R, Brunger AT (1998) Crystal structure of a SNARE complex involved in synaptic exocytosis at 2.4 A resolution. Nature 395: 347–353 Suvorova ES, Duden R, Lupashin VV (2002) The Sec34/Sec35p complex, a Ypt1p effector required for retrograde intra-Golgi trafficking, interacts with Golgi SNAREs and COPI vesicle coat proteins. J Cell Biol 157: 631–643 Sztul E, Lupashin V (2006) Role of tethering factors in secretory membrane traffic. Am J Physiol Cell Physiol 290, C11–C26

64

*

D. K. Banfield and W. Hong

Tochio H, Tsui MM, Banfield DK, Zhang M (2001) An autoinhibitory mechanism for nonsyntaxin SNARE proteins revealed by the structure of Ykt6p. Science 293: 698–702 Tsui MM, Banfield DK (2000) Yeast Golgi SNARE interactions are promiscuous. J Cell Sci 113(Pt 1): 145–152 Tsui MM, Tai WC, Banfield DK (2001) Selective formation of Sed5p-containing SNARE complexes is mediated by combinatorial binding interactions. Mol Biol Cell 12: 521–538 Valdez-Taubas J, Pelham H (2005) Swf1-dependent palmitoylation of the SNARE Tlg1 prevents its ubiquitination and degradation. EMBO J 24: 2524–2532 Varlamov O, Volchuk A, Rahimian V, Doege CA, Paumet F, Eng WS, Arango N, Parlati F, Ravazzola M, Orci L, Sollner TH, Rothman JE (2004) i-SNAREs: inhibitory SNAREs that fine-tune the specificity of membrane fusion. J Cell Biol 164: 79–88 Veit M (2004) The human SNARE protein Ykt6 mediates its own palmitoylation at C-terminal cysteine residues. Biochem J 384: 233–237 Veit M (2000) Palmitoylation of the 25-kDa synaptosomal protein (SNAP-25) in vitro occurs in the absence of an enzyme, but is stimulated by binding to syntaxin. Biochem J 345(Pt 1): 145–151 Vogel K, Roche PA (1999) SNAP-23 and SNAP-25 are palmitoylated in vivo. Biochem Biophys Res Commun 258: 407–410 Volchuk A, Ravazzola M, Perrelet A, Eng WS, Di Liberto M, Varlamov O, Fukasawa M, Engel T, Sollner TH, Rothman JE, Orci L (2004) Countercurrent distribution of two distinct SNARE complexes mediating transport within the Golgi stack. Mol Biol Cell 15: 1506–1518 Von Mollard GF, Nothwehr SF, Stevens TH (1997) The yeast v-SNARE Vti1p mediates two vesicle transport pathways through interactions with the t-SNAREs Sed5p and Pep12p. J Cell Biol 137: 1511–1524 Watson RT, Pessin JE (2001) Transmembrane domain length determines intracellular membrane compartment localization of syntaxins 3, 4, and 5. Am J Physiol Cell Physiol 281: C215–C223 Weinberger A, Kamena F, Kama R, Spang A, Gerst JE (2005) Control of Golgi morphology and function by Sed5 t-SNARE phosphorylation. Mol Biol Cell 16: 4918–4930 Williams AL, Ehm S, Jacobson NC, Xu D, Hay JC (2004) rsly1 binding to Syntaxin 5 is required for endoplasmic reticulum-to-Golgi transport but does not promote SNARE motif accessibility. Mol Biol Cell 15: 162–175 Wimmer C, Hohl TM, Hughes CA, Muller SA, Sollner TH, Engel A, Rothman JE (2001) Molecular mass, stoichiometry, and assembly of 20 S particles. J Biol Chem 276: 29091–29097 Wooding S, Pelham HR (1998) The dynamics of Golgi protein traffic visualized in living yeast cells. Mol Biol Cell 9: 2667–2680 Xu D, Joglekar AP, Williams AL, Hay JC (2000) Subunit structure of a mammalian ER/ Golgi SNARE complex. J Biol Chem 275: 39631–39639 Xu Y, Martin S, James DE, Hong W (2002) GS15 forms a SNARE complex with syntaxin 5, GS28, and Ykt6 and is implicated in traffic in the early cisternae of the Golgi apparatus. Mol Biol Cell 13: 3493–3507 Xu Y, Zhang F, Su Z, McNew JA, Shin YK (2005) Hemifusion in SNARE-mediated membrane fusion. Nat Struct Mol Biol 12: 417–422 Yamaguchi T, Dulubova I, Min SW, Chen X, Rizo J, Sudhof TC (2002) Sly1 binds to Golgi and ER syntaxins via a conserved N-terminal peptide motif. Dev Cell 2: 295–305 Yang B, Gonzalez LJr, Prekeris R, Steegmaier M, Advani RJ, Scheller RH (1999) SNARE interactions are not selective. Implications for membrane fusion specificity. J Biol Chem 274: 5649–5653

SNAREs

*

65

Yao R, Ito C, Natsume Y, Sugitani Y, Yamanaka H, Kuretake S, Yanagida K, Sato A, Toshimori K, Noda T (2002) Lack of acrosome formation in mice lacking a Golgi protein, GOPC. Proc Natl Acad Sci USA 99: 11211–11216 Zhang T, Hong W (2001) Ykt6 forms a SNARE complex with syntaxin 5, GS28, and Bet1 and participates in a late stage in endoplasmic reticulum-Golgi transport. J Biol Chem 276: 27480–27487 Zolov SN, Lupashin VV (2005) Cog3p depletion blocks vesicle-mediated Golgi retrograde trafficking in HeLa cells. J Cell Biol 168: 747–759

66

*

Ch. Claas et al.

Rabs Christoph Claas, Alexander A. Mironov and Vytaute Starkuviene

In this chapter we will describe the Rab family of proteins, one of the most important molecular machines operating within the secretory pathway. Rab proteins (Ras-related proteins in brain) are evolutionary conserved regulators of membrane traffic. They are of key importance for proper functioning of the exocytic secretory pathway and endocytosis. In fact, Rabs seem to play a role in all events of intracellular transport, i.e. they take part in cargo selection and budding of vesicular (or tubular) structures in donor compartments, they regulate the transport of these membrane-bound structures along cytoskeletal tracks, they control their docking to the target membrane, and they are involved in the final fusion of donor and target membrane. Moreover, they guarantee that organelles keep their protein and lipid composition as well as their right position within the cell (Seabra et al. 2002; Grosshans et al. 2006). Whereas 11 Rabs have been described in Saccharomyces cerevisiae, more than 70 Rabs and Rab-like proteins exist in mammals (Pereira-Leal and Seabra 2001; Gurkan et al. 2005). Consequently, together with the SNAREs the Rabs are the largest of the protein families involved in membrane trafficking.

Membrane attachment of Rabs Rabs are synthesized as soluble molecules, but their function depends on association with the cytoplasmic leaflet of cellular membranes – only after post-translational addition of two isoprenoid molecules (prenylation) at their C-terminus they are able to become membrane-associated (Pereira-Leal et al. 2001; An et al. 2003). This post-translational modification is catalyzed by socalled Rab escort protein (REP). REP binds to the newly synthesized Rab in GDPbound form and simultaneously to Rab geranylgeranyl transferase (RGGT or protein prenyl transferase, Pylypenko et al. 2003). After modification is accomplished, RGGT dissociates from REP–Rab complex, and REP delivers the Rab to the membrane (Thoma et al. 2001). Then, REP recycles back to bind new molecules of freshly synthesized Rab in the cytoplasm. Since there are only two isoforms of REP in mammalian cells, REP is able to bind many distinct Rab proteins via conserved residues (Alory and Balch 2001; Pfeffer 2005). Membrane-associated Rabs can either be converted into the active GTPbound form (see below) or they are retrieved from the membrane as Rab–GDP by cytoplasmic proteins called GDP dissociation inhibitor (GDI, Wu et al. 2007). GDI is able to bind to prenylated Rabs in the GDP-bound form (Wilson et al. 1996). Similar to the Rab-binding subunit of REP, GDIs hide

Rabs

*

67

prenyl groups attached to Rabs within a hydrophobic groove (Pylypenko et al. 2006), thereby keeping Rabs hydrophilic and enabling them to reside in the cytoplasm. Upon the action of GDI displacement factor (GDF), it is proposed that the complex of Rab–GDP and GDI dissociates, prenyl anchors can be inserted into the membrane, and as a result the Rab molecule becomes associated with the membrane (Dirac-Svejstrup et al. 1997). Little is known about the detailed mechanism leading to removal of GDI in the course of Rab membrane attachment. In fact, to date only one protein with GDF activity has been identified in yeast and mammals. This protein, Yip3 (PRA1), belongs to the Yip (Ypt-interacting proteins) family of proteins and is able to dissociate the Rab9–GDI complex and to recruit Rab9 to endosomal membranes (Silvar et al. 2003). At least five GDI isoforms are expressed in mammalian cells (Alory and Balch 2003), and structural analysis of some of them highlighted the mechanisms of their association to membranes and Rabs (Schalk et al. 1996; Luan et al. 2000). Human aGDI is highly enriched in brain tissue with low abundance in other cells, whereas bGDI is the main, “housekeeping” isoform being expressed ubiquitously (Nishimura et al. 1994). Therefore, their function could be different: aGDI is involved in sorting of highly specialised vesicles in brain such as neurosecretory vesicles, in difference to bGDI, which plays a general role in vesicular trafficking in diverse types of cells. Consequently, GDI isoforms are not specific and recognize a broad range of Rab species (Ullrich et al. 1993). Curiously, REP and GDI proteins are sharing common Rab-binding properties and high sequence similarity, so that they can be grouped into one evolutionary conserved family of REP/GDI proteins (Alory and Balch 2001). Still, the functions of both proteins are not interchangeable: GDI cannot assist in the prenylation of newly synthesized Rab proteins and REP cannot retrieve Rab proteins from membranes.

Rabs as molecular switches Rabs function as GTPases that cycle between a GTP- and a GDP-bound state (Dumas et al. 1999; Pasqualato et al. 2004). Only the GTP-bound form is functionally active and can recruit diverse proteins, so-called effectors (see later), which, in turn, participate in different steps of vesicular traffic (Grosshans et al. 2006). The inactive GDP-bound form can be either cytoplasmic or membrane-associated (Wittmann and Rudolph 2004). Because of the low rates of nucleotide exchange and hydrolysis, the activity switch function of Rabs is under the control of GDP–GTP exchange factors (GEFs) and GTPase activating proteins (GAPs) (Bos et al. 2007). Hydrolysis of GTP leads to inactivation of the Rab and the GDP-bound form can then be extracted from the membrane by GDI (Ullrich et al. 1993). GDI then periodically delivers cytosolic Rab to the appropriate organelle (Pfeffer and Aivazian 2004). When Rab–GDP becomes membrane-bound, a Rab GEF can promote exchange of GDP with GTP, thus converting Rab into the active GTP-bound

68

*

Ch. Claas et al.

form (Bos et al. 2007). Standard models of Rab dynamics suggest a continuous cycle of organelle association and dissociation (Segev 2001; Pfeffer and Aivazian 2004; Bos et al. 2007). In contrast to general regulators of Rab function such as REP and GDI proteins, GEFs and GAPs are specific for one or a few Rab family members, tissues or even subcellular localization (Bernards 2003; Bos et al. 2007). For instance, over-expression of Rab GAPs can be used to specifically inactivate the endogenous pool of a Rab and, thus, to interfere with the process this Rab is involved in as it was shown by dissecting the internalisation pathways of Shiga toxin and epidermal growth factor EGF (Fuchs et al. 2007). The human genome potentially encodes 39 Rab GAPs; therefore one Rab GAP should be acting upon several Rabs (Bernards 2003). Still, the identification of specific Rab–Rab GAP pairs is not a trivial task (Fuchs et al. 2007). Rab-specific GEFs comprise a diverse group of proteins; therefore their prediction presents a daunting task. Most Rab GEFs are Vps9-, Sec2- or Mss4like proteins (Bos et al. 2007). Little data is available about specific pairs of Rabs and GEFs. By inducing conformational changes to the Switch regions of an associating Rab protein, GEFs are decreasing the affinity for binding GTP or GDP As GEFs and Rabs have no preference for a particular nucleotide, it is the higher concentration of GTP within the cell that results in increased amounts of Rab–GTP upon interaction with GEF (Bos et al. 2007). Finally, since Rabs are regulators of all steps in vesicular traffic, a timely coordinated exchange in the Rabs being “in charge” of a given step has to be ensured (Grosshans et al. 2006). This so-called Rab conversion can be achieved by organizing Rab recruitment and activation into so-called “Rab cascades”. This concept postulates that the GEF of a downstream Rab GTPase will serve at the same time as an effector of an upstream Rab protein (Rink et al. 2005). For instance, Sec2 serves as GEF for the Rab Sec4, which is a regulator of Golgi to plasma membrane (PM) transport. Sec2 is recruited to membranes by the upstream acting Rabs Ypt31/32, thus ensuring timely and spatially coordinated regulation of secretory granule transport to the PM (Ortiz et al. 2002).

Structure of Rabs The structure of no other protein family was so extensively analysed as that of Rabs, including active and non-active states, complexes with REP, GDI and numerous effectors. The interest was raised due to the universality of Rab function, but their high specificity with respect to their target compartments and functional effectors. The versatility of Rabs contrasts with their high sequence homology and similarity in the three-dimensional fold (Pfeffer 2005). As members of the Ras superfamily of proteins the overall structure of Rabs is very similar to other small GTPases (Wennerberg et al. 2005). They share a set of G-box GDP/GTP-binding motifs at their N-terminus, and two switch regions (Switch I and Switch II), which change their conformation depending on which nucleotide is bound (GDP or GTP). Within Rab molecules,

Rabs

*

69

Switch regions are the only structural elements changing their conformation upon binding of GDP or GTP (Stroupe and Brunger 2000). Despite functional similarities, Switch domains of Rab GTPases have a different orientation than those of Ras proteins, serving as family discriminants. Switch domains of various Rabs crystallized to date overlap significantly in terms of their overall lengths and boundaries (Pfeffer 2005). Therefore, considerable effort was made to find specific structural features that could provide the basis for functional heterogeneity of Rabs. Taking into consideration that Rab effectors are preferentially interacting with GTP-bound proteins, they have to be able to “sense” whether GTP or GDP is bound by the Rab. Consequently, they should interact with sequences close to or within the Switch regions. By this, Switch regions will act not only as providers of information about the activity state of Rabs, but also as recognition domains for effectors. By comparative sequence analysis Pereira-Leal and Seabra (2000) identified five so-called Rab family sequences, F1–F5, that are conserved among Rabs but not Ras or Rho GTPases. Indeed some of them, namely F1, F3 and F5, are located within Switch domains. Moreover, four Rab subfamily-conserved (RabSF) regions were identified by sequence comparison, and have been used to define 10 subfamilies of Rab GTPases (Pereira-Leal and Seabra 2000). Molecular determinants of such specificity were described for some Rab proteins (Constantinescu et al. 2002). Similar to RabF motifs, some of the RabSFs are located within Switch regions (Stroupe and Brunger 2000). Likely, specificity of interactions between Rabs and their effectors are achieved by co-operative binding to RabF and RabSF regions (Pereira-Leal and Seabra 2000). However, regardless of the possibility to cluster Rab molecules according to their primary sequence features, function–structure predictions of uncharacterized Rabs might be not accurate enough as several additional specificity determinants were described (Schwartz et al. 2007). The variety of Rab interactions was attributed not only to differences in the primary sequence, but also to conformational heterogeneity. For instance, structural studies on Rab5C and Rab3A demonstrated that an invariant hydrophobic triad at the Switch region interface might be positioned by diverse angles, thereby creating very distinct surfaces of even closely related Rabs (Merithew et al. 2001). Variability between Rabs is also extended towards the C-terminus. C-termini of Rabs have different lengths and composition and, being the most distinct structural elements of Rabs (Chavrier et al. 1991), contribute to generation of distinct Rab protein surfaces and diversification of their function. Usually, they are called “hyper variable domains”; it was shown that Rab protein localization depends on these domains (Chavrier et al. 1991). However, whereas effectors interacting with multiple Rabs hardly can use these regions for recognition of their target molecules, these regions contain highly conserved residues important for the interaction with GDI and REP proteins (Pylypenko et al. 2003, 2006).

70

*

Ch. Claas et al.

Compartmentalisation of Rabs Rabs are widely distributed along the secretory pathway, but each member of the Rab family has its particular function inside the cell (Schwartz et al. 2007). Each Rab protein is localized at the membrane of specific intracellular compartments and is highly specific for a particular transport step and its effectors. For instance, Rabs 1 and 2 are important for ER-to-Golgi traffic (Tisdale et al. 1992; Saraste et al. 1995), Rab6 for intra-Golgi transport (Jiang and Storrie 2005), and Rab8 and Rab10 act in the late secretory pathway (Peranen et al. 1996; Babbey et al. 2006). Different Rabs are involved in various types of secretory processes, e.g., Rab3 and Rab27 are playing a role in regulated secretion (Oberhauser et al. 1992; Stinchcombe et al. 2001), Rab5, Rab7 and Rab11 in endocytosis (Ullrich et al. 1996; Pelkmans et al. 2004; Rink et al. 2005). Although some Rabs are tissue-specific (Gurkan et al. 2005), many are ubiquitous in their expression. Often multiple Rabs play a role in the same chain of trafficking events, so their functions need to be coordinated. Rab5, for example, is found on the plasma membrane and early endosomes. Early endosomes also carry Rab4, which acts downstream of Rab5. Rab4 and Rab11 then are localized together on recycling endosomes (Sonnichsen et al. 2000; Zerial and McBride 2001). Similarly, late endosomes carry both Rab7 and Rab9 (Barbero et al. 2002). How is correct delivery of Rabs to destination compartments achieved? Because geranylgeranylation is a common feature of essentially all Rabs and serves for membrane anchoring, it cannot account for their organelle-specific targeting. Specific targeting of Rab proteins via the hyper variable region (Chavrier et al. 1991; Stenmark et al. 1994) is a widely accepted model. It was shown, that replacing the C-terminus of Rab6 with the equivalent C-terminal region of Rab5 resulted in re-localization of the hybrid Rab to Rab5-positive structures, like early endosomes (Stenmark et al. 1994). Similar observations were reported for Rab5 and Rab7 proteins (Ali et al. 2004). However, it has been shown recently that the hyper variable region does not contain a general Rab targeting signal. Reciprocal exchanges of the hyper variable domains of Rab1a, Rab2a, Rab5a, Rab7 and Rab27a failed to re-direct the hybrid proteins away from their original compartment to the new compartment designated by the hyper variable region (Chavrier et al. 1991; Ali et al. 2004). Other regions have been demonstrated to be required within the RabF and RabSF motifs for specific targeting of Rab27a to secretory granules or melanosomes, and Rab5a to endosomes (Ali et al. 2004). Mutations in these targeting-determining regions frequently induced localization to the ER Indeed, the ER and the Golgi apparatus (GA) are preferential sites of membrane translocation for Rabs that have lost targeting information. Interestingly, the localization of the Rab27a was shown to be dependent on the presence of luminally expressed proteins, thus suggesting that cargo molecules might influence their environment including recruitment of specific Rabs (Hannah et al. 2003). In addition, there is evidence that certain receptors

Rabs

*

71

for Rabs are present in membranes and that those receptors show some preference for the interaction with certain Rab isoforms. Potentially, Yip proteins aid in specific Rab targeting as they can bind to Rab proteins that are in complex with GDI (Pfeffer and Aivazian 2004). However, binding to Yip cannot completely explain highly specific localization of Rabs within a cell as 16 Yip family members in humans (Pfeffer and Aivazian 2004) can most probably not account for proper delivery of roughly 70 Rabs. Many researchers, therefore, assume that Rabs are initially delivered nonspecifically. Only if the Rab protein does not encounter activating proteins and effector proteins already present at this localization or does not recruit these molecules itself to establish a proper Rab-defined complex, GDI will eventually extract Rab–GDP from the membrane and initiate a new cycle of membrane delivery. If the former events take place, positive feedback loops between the function of Rabs and of Rab effectors might lead to stabilization of the Rab localization resulting in a creation of a transient scaffold, a so-called Rab microdomain (Zerial and McBride 2001; Pfeffer and Aivazian 2004; Cai et al. 2007).

Rab effectors Effectors of Rab proteins are defined as proteins that are able to interact with the active, GTP-bound form of a specific Rab, and they exert at least one specific function downstream of this Rab. In contrast to some GDI, GAP, and Rabs themselves, effector proteins display a high degree of structural diversity and belong to a multitude of different protein superfamilies. Numerous interactions of Rabs with effectors have been described and the list of effector molecules is ever growing (Grosshans et al. 2006). Examples for such associations, therefore, have to be limited to a few examples covering each step in membrane trafficking. The first step in trafficking encompasses the selection and concentration of cargo molecules in a subdomain of a donor membrane and the formation of a vesicle. It has been demonstrated that the association of Rab9–GTP with tail-interacting protein of 47 kDa (TIP47) is important for the recycling of mannose6-phosphate receptors from late endosomes to the GA Mannose6phosphate receptor binds to hydrolases bearing mannose6-phosphate in the trans-Golgi network and transports them to late endosomes. Recycling back to the GA enables it to initiate a new cycle of transport. Rab9 on late endosomes increases the affinity of associated TIP47 for binding to the cytoplasmic tail of mannose6-phosphate receptor, thus regulating cargo selection for recycling vesicles (Carroll et al. 2001). Moreover, interaction of Rab9 with its effector TIP47 is required for the correct localization of TIP47 on late endosomes (Carroll et al. 2001). Vice versa, the presence of TIP47 on late endosomes stabilizes the steady-state localization of Rab9 on these organelles (Aivazian et al. 2006). This interplay of a Rab and its effector demonstrates the importance of these interactions for the correct targeting of the trafficking machinery.

72

*

Ch. Claas et al.

After budding from a donor membrane, newly formed carriers have to become associated with motor proteins moving them along actin filaments or along the microtubule network. Association can be established by direct interaction with a motor protein or indirectly via an adaptor protein. A wellstudied example for the latter case is the association of Rab27a expressed by melanocytes with myosin Va. The effector melanophilin links Rab27a on melanosomes to the actin motor myosin-Va (Fukuda et al. 2002; Nagashima et al. 2002; Strom et al. 2002). Without these interactions melanosomes loose their localization at the periphery of melanocytes and cannot be transported to neighboring keratinocytes (Wu et al. 1998). In retinal cells the same Rab27a was demonstrated to interact with another myosin motor, myosin VII (Gibbs et al. 2004). This interaction depends on the Rab effector MyRIP (Desnos et al. 2003; El-Amraoui et al. 2002), showing that Rab-effector complexes can be cell-type specific. Carriers on the way to their destination compartment have to recognize and to bind to structures on the target membrane. This is achieved by tethering factors, whose function is also regulated by Rabs. In general, tethering factors are either extended proteins with a coiled-coil structure or they are multi-subunit complexes. An example of the former is p115, which is present on ER-derived COPII-coated vesicles. P115 binds to the GM130/ GRASP65 complex present at the cis-GA, and both proteins have been shown to be effectors of Rab1. An intricate net of interactions between Golgi matrix proteins, tethering factors and Rab1 was shown to be essential for material flow through the early secretory pathway (Allan et al. 2002; Moyer et al. 2001; Beard et al. 2005). An example for a Rab-regulated multisubunit tether is represented by the exocyst. The exocyst is an octameric complex that tethers secretory granules to the PM. The yeast Rab Sec4p in its active state is able to interact with the exocyst subunit Sec15p (Guo et al. 1999). Biochemical analyses revealed that another component, Sec8p, also associates with Sec4p, suggesting that Sec4p serves as central regulator of exocyst function (Toikkanen et al. 2003). Later studies indicate that this interaction of Rab and exocyst is conserved in mammals (Zhang et al. 2004). Finally, Rabs play a role in membrane fusion, the last step in trafficking. An influence of Rabs on SNARE (soluble N-ethylmaleimide-sensitive factor attachment protein receptor, see Chapter 2.1) function has been shown by many investigators. The effect seems to be indirect, however, and is exerted by effector molecules directly interacting with Rabs (Grosshans et al. 2006). SNARE proteins induce fusion by bringing the bilayers of opposing vesicular and target membranes into close proximity (Pfeffer 2007; see Chapter 2.1). Shorter et al. (2002) describe that in the process of intra-Golgi transport p115 catalyzes the assembly of a trans-SNARE complex by linking the v-SNARE GOS28 to the t-SNARE Syntaxin5. Another role of Rabs in vesicle fusion could be the prevention of premature complex formation from monomeric SNAREs situated in the same compartment.

Rabs

*

73

Figure 1. Scheme of the Rab cycle. After its synthesis the Rab molecule (1) is prenylated by RGGT (A). Alternatively, initially Rab binds GDP (2) and then undergoes prenylation. After prenylation Rab binds REP (4) that accompanies Rab to the membrane (tree-layered arc filled with grey color). After detachment of REP (4), Rab (1) bound to GDP (2) binds RabGEF (5) that exchanges GDP for GTP (3). RabGTP interacts with SNARE preventing formation of a SNARE complex with the membrane plane. After fusion, the SNARE detaches from the Rab and the Rab binds to Rab GAP (6). This leads to hydrolysis of GTP and liberation of phosphate (P). After detachment of RabGAP, RabGDP can detach from the membrane and bind to Rab GDI (7). Rab GDI delivers the Rab back to a membrane and a new cycle of Rab activation can start.

Conclusion A plethora of different proteins is involved in the trafficking of membranebound structures within a cell. Besides Rabs, coat proteins, SNAREs, and motor proteins are of key importance. The number of Rabs during evolution, however, has increased much more than those of other proteins in this context, matching the needs of increasing complexity of membrane systems from yeast to mammals. This reflects the central role that Rabs play in the coordination of diverse steps during trafficking – some authors conceptualize Rabs as the central organizers or central hubs of membrane trafficking (Gurkan et al. 2005). Certainly, their function as binary switches oscillating between an active and an inactive form alone cannot explain how these molecules can contribute to complex spatial and timely regulation which is needed for proper trafficking. Presumably, positive feedback loops between Rab, Rab function modulating proteins and effector molecules stabilize the activity and localization of Rabs, and contribute to the establishment of Rabdefined microdomains within the membrane (Zerial and McBride 2001). Moreover, protein machineries in such microdomains have to be coordinated

74

*

Ch. Claas et al.

in terms of assembly, disassembly and interplay with other machineries to ensure proper regulation of successive steps during vesicle transport and, consequently, to ensure cellular homeostasis. The concept of organizing different Rabs into Rab cascades as described above may explain at least in part how this task is achieved (Rink et al. 2005). Still, studies on so far uncharacterized Rabs, the identification of more Rab-interacting proteins and, most importantly, knowledge about detailed mechanisms of the interplay of all these components will be required to fully understand the surprising versatility of Rab–GTPases.

Abbreviations GA GAP GDF GDI GEF PM REP RGGT Yip

Golgi apparatus GTPase-activating protein GDI displacement factor GDP dissociation inhibitor GDP–GTP exchange factor plasma membrane Rab escort protein Rab geranylgeranyl transferase Ypt-interacting protein

References Aivazian D, Serrano RL, Pfeffer S (2006) TIP47 is a key effector for Rab9 localization. J Cell Biol 173(6): 917–926 Ali BR, Wasmeier C, Lamoreux L, Strom M, Seabra MC (2004) Multiple regions contribute to membrane targeting of Rab GTPases. J Cell Sci 117(Pt 26): 6401–6412 Allan BB, Moyer BD, Balch WE (2002) Rab1 recruitment of p115 into a cis-SNARE complex: programming budding COPII vesicles for fusion. Science 289(5478): 444–448 Alory C, Balch WE (2001) Organization of the Rab–GDI/CHM superfamily: the functional basis for choroideremia disease. Traffic 2(8): 532–543 Alory C, Balch WE (2003) Molecular evolution of the Rab-escort-protein/guaninenucleotide-dissociation-inhibitor superfamily. Mol Biol Cell 14: 3857–3867 An Y, Shao Y, Alory C, Matteson J, Sakisaka T, Chen W, Gibbs RA, Wilson IA, Balch WE (2003) Geranylgeranyl switching regulates GDI–Rab GTPase recycling. Structure 11: 347–357 Babbey CM, Ahktar N, Wang E, Chen CC, Grant BD, Dunn KW (2006) Rab10 regulates membrane transport through early endosomes of polarized Madin-Darby canine kidney cells. Mol Biol Cell 17(7): 3156–3175 Barbero P, Bittova L, Pfeffer SR (2002) Visualization of Rab9-mediated vesicle transport from endosomes to the trans-Golgi in living cells. J Cell Biol 156(3): 511–518 Bernards A (2003) GAPs galore! A survey of putative Ras superfamily GTPase activating proteins in man and Drosophila. Biochim Biophys Acta 1603(2): 47–82 Beard M, Satoh A, Shorter J, Warren GA (2005) Cryptic Rab1-binding site in the p115 tethering protein. J Biol Chem 280(27): 25840–25848 Bos JL, Rehmann H, Wittinghofer A (2007) GEFs and GAPs: critical elements in the control of small G proteins. Cell 129(5): 865–877

Rabs

*

75

Cai H, Reinisch K, Ferro-Novick S (2007) Coats tethers, Rabs and SNAREs work together to mediate the intracellular destination of a transport vesicle. Dev Cell 12(5): 671–682 Carroll KS, Hanna J, Simon I, Krise J, Barbero P, Pfeffer SR (2001) Role of Rab9 GTPase in facilitating receptor recruitment by TIP47. Science 292(5520): 1373–1376 Chavrier P, Gorvel JP, Stelzer E, Simons K, Gruenberg J, Zerial M (1991) Hypervariable C-terminal domain of rab proteins acts as a targeting signal. Nature 353(6346): 769–772 Constantinescu AT, Rak A, Alexandrov K, Esters H, Goody RS, Scheidig AJ (2002) Rabsubfamily-specific regions of Ypt7p are structurally different from other RabGTPases. Structure 10(4): 569–579 Desnos C, Schonn JS, Huet S, Tran VS, El-Amraoui A, Raposo G, Fanget I, Chapuis C, nasche  G, De Saint Basile G, Petit C, Cribier S, Henry JP, Darchen F (2003) Rab27A Me and its effector MyRIP link secretory granules to F-actin and control their motion towards release sites. J Cell Biol 163(3): 559–570 Dirac-Svejstrup AB, Sumizawa T, Pfeffer SR (1997) Identification of a GDI displacement factor that releases endosomal Rab GTPases from Rab–GDI EMBO J 16(3): 465–472 Dumas JJ, Zhu Z, Connolly JL, Lambright DG (1999) Structural basis of activation and GTP hydrolysis in Rab proteins. Structure 15: 413–423 € ssel-Andermann P, Blanchard S, Desnos C, Henry JP, El-Amraoui A, Schonn JS, Ku Wolfrum U, Darchen F, Petit C (2002) MyRIP, a novel Rab effector, enables myosin VIIa recruitment to retinal melanosomes. EMBO Rep 3(5): 463–470 Fuchs E, Haas AK, Spooner RA, Yoshimura S, Lord JM, Barr FA (2007) Specific Rab GTPaseactivating proteins define the Shiga toxin and epidermal growth factor uptake pathways. J Cell Biol 177: 1133–1143 Fukuda M, Kuroda TS, Mikoshiba K (2002) Slac2-a/melanophilin, the missing link between Rab27 and myosin Va: implications of a tripartite protein complex for melanosome transport. J Biol Chem 277(14): 12432–12436 Gibbs D, Azarian SM, Lillo C, Kitamoto J, Klomp AE, Steel KP, Libby RT, Williams DS (2004) Role of myosin VIIa and Rab27a in the motility and localization of RPE melanosomes. J Cell Sci 117(Pt26): 6473–6483 Grosshans BL, Ortiz D, Novick P (2006) Rabs and their effectors: achieving specificity in membrane traffic. Proc Natl Acad Sci USA 103(32): 11821–11827 Guo W, Roth D, Walch-Solimena C, Novick P (1999) The exocyst is an effector for Sec4p, targeting secretory vesicles to sites of exocytosis. EMBO J 18(4): 1071–1080 Gurkan C, Lapp H, Alory C, Su AI, Hogenesch J, Balch WE (2005) Large-scale profiling of Rab GTPase trafficking networks: the membrome. Mol Biol Cell 16: 3847–3864 Hannah MJ, Hume AN, Arribas M, Williams R, Hewlett LJ, Seabra MC and Cutler DF (2003) Weibel-Palade bodies recruit Rab27 by a content-driven, maturation-dependent mechanism that is independent of cell type. J Cell Sci 116: 3939–3948 Jiang S, Storrie B (2005) Cisternal rab proteins regulate Golgi apparatus redistribution in response to hypotonic stress. Mol Biol Cell 16(5): 2586–2596 Luan P, Heine A, Zeng K, Moyer B, Greasely SE, Kuhn P, Balch WE, Wilson IA (2000) A new functional domain of guanine nucleotide dissociation inhibitor (alpha-GDI) involved in Rab recycling. Traffic 1(3): 270–281 Merithew E, Hatherly S, Dumas JJ, Lawe DC, Heller-Harrison R, Lambright DG (2001) Structural plasticity of an invariant hydrophobic triad in the switch regions of Rab GTPases is a determinant of effector recognition. J Biol Chem 276(17): 13982– 13988 Moyer BD, Allan BB, Balch WE (2001) Rab1 interaction with a GM130 effector complex regulates COPII vesicle cis-Golgi tethering. Traffic 2(4): 268–276 Nagashima K, Torii S, Yi Z, Igarashi M, Okamoto K, Takeuchi T, Izumi T (2002) Melanophilin directly links Rab27a and myosin Va through its distinct coiled-coil regions. FEBS Lett 517(1–3): 233–238

76

*

Ch. Claas et al.

Nishimura N, Nakamura H, Takai Y, Sano K (1994) Molecular cloning and characterization of two rab GDI species from rat brain: brain-specific and ubiquitous types. J Biol Chem 269: 14191–14198 Oberhauser AF, Monck JR, Balch WE, Fernandez JM (1992) Exocytotic fusion is activated by Rab3a peptides. Nature 360(6401): 270–273 Ortiz D, Medkova M, Walch-Solimena C, Novick P (2002) Ypt32 recruits the Sec4p guanine nucleotide exchange factor, Sec2p, to secretory vesicles; evidence for a Rab cascade in yeast. J Cell Biol 157: 1005–1015 Pasqualato S, Senic-Matuglia F, Renault L, Goud B, Salamero J, Cherfils J (2004) The structural GDP/GTP cycle of Rab11 reveals a novel interface involved in the dynamics of recycling endosomes. J Biol Chem 279(12): 11480–11488 € rli T, Zerial M, Helenius A (2004) Caveolin-stabilized membrane domains Pelkmans L, Bu as multifunctional transport and sorting devices in endocytic membrane traffic. Cell 118(6): 767–780 € nen J, Auvinen P, Virta H, Wepf R, Simons K (1996) Rab8 promotes polarized Pera membrane transport through reorganization of actin and microtubules in fibroblasts. J Cell Biol 135(1): 153–167 Pereira-Leal JB, Hume AN, Seabra MC (2001) Prenylation of Rab GTPases: molecular mechanisms and involvement in genetic disease. FEBS Lett 498: 197–200 Pereira-Leal J, Seabra MC (2000) The mammalian Rab family of small GTPases: definition of family and subfamily sequence motifs suggests a mechanism for functional specificity in the Ras superfamily. J Mol Biol 301: 1077–1087 Pereira-Leal JB, Seabra MC (2001) Evolution of the Rab family of small GTP–binding proteins. J Mol Biol 313(4): 889–901 Pfeffer SR (2005) Structural clues to Rab GTPase functional diversity. J Biol Chem 280(16): 15485–15488 Pfeffer SR (2007) Unsolved mysteries in membrane traffic. Annu Rev Biochem 76: 629–645 Pfeffer S, Aivazian D (2004) Targeting Rab GTPases to distinct membrane compartments. Nat Rev Mol Cell Biol 5(11): 886–896 Pylypenko O, Rak A, Reents R, Niculae A, Sidorovitch V, Cioaca MD, Bessolitsyna E, € NH, Waldmann H, Schlichting I, Goody RS, Alexandrov K (2003) Structure of Thoma Rab escort protein-1 in complex with Rab geranylgeranyltransferase. Mol Cell 11(2): 483–494 Pylypenko O, Rak A, Durek T, Kushnir S, Dursina BE, Thomae NH, Constantinescu AT, Brunsveld L, Watzke A, Waldmann H, Goody RS, Alexandrov K (2006) Structure of doubly prenylated Ypt1: GDI complex and the mechanism of GDI-mediated Rab recycling. EMBO J 11: 13–23 Rink J, Ghigo E, Kalaidzidis Y, Zerial M (2005) Rab conversion as a mechanism of progression from early to late endosomes. Cell 122(5): 735–749 Saraste J, Latineen U, Goud B (1995) Localization of the small GTP-binding protein rab1p to early compartments of the secretory pathway. J Cell Sci 108(Pt4): 1541–1552 Schalk I, Zeng K, Wu SK, Stura EA, Matteson J, Huang M, Tandon A, Wilson IA, Balch WE (1996) Structure and mutational analysis of Rab GDP-dissociation inhibitor. Nature 381: 42–48 Schwartz S, Cao C, Pylypenko O, Rak A, Wandinger-Ness A (2007) Rab GTPases at a glance. J Cell Sci 120(Pt 22): 3905–3910 Seabra MC, Mules EH, Hume AN (2002) Rab GTPases intracellular traffic and disease. Trends Mol Med 8(1): 23–30 Segev N (2001) Ypt and Rab GTPases: insight into functions through novel interactions. Curr Opin Cell Biol 13(4): 500–511 Shorter J, Beard MB, Seemann J, Dirac-Svejstrup AB, Warren G (2002) Sequential tethering of Golgins and catalysis of SNAREpin assembly by the vesicle-tethering protein p115. J Cell Biol 157(1): 45–62

Rabs

*

77

Sivars U, Aivazian D, Pfeffer SR (2003) Yip3 catalyses the dissociation of endosomal Rab–GDI complexes. Nature 425: 856–859 € nnichsen B, De Renzis S, Nielsen E, Rietdorf J, Zerial M (2000) Distinct membrane So domains on endosomes in the recycling pathway visualized by multicolor imaging of Rab4, Rab5, and Rab11. J Cell Biol 149(4): 901–914 Stenmark H, Valencia A, Martinez O, Ullrich O, Goud B, Zerial M (1994) Distinct structural elements of rab5 define its functional specificity. EMBO J 13: 575–583 Stinchcombe JC, Barral DC, Mules EH, Booth S, Hume AN, Machesky LM, Seabra MC, Griffiths GM (2001) Rab27a is required for regulated secretion in cytotoxic T lymphocytes. J Cell Biol 152(4): 825–834 Strom M, Hume AN, Tarafder AK, Barkagianni E, Seabra MC (2002) A family of Rab27binding proteins. Melanophilin links Rab27a and myosin Va function in melanosome transport. J Biol Chem 277(28): 25423–24530 Stroupe C, Brunger AT (2000) Crystal structures of a Rab protein in its inactive and active conformations. J Mol Biol 304(4): 585–598 Thoma NH, Iakovenko A, Kalinin A, Waldmann H, Goody RS, Alexandrov K (2001) Allosteric regulation of substrate binding and product release in geranylgeranyl transferase type II Biochemistry 40: 268–274 Tisdale EJ, Bourne JR, Khosravi-Far R, Der CJ, Balch WE (1992) GTP-binding mutants of Rab1 and Rab2 are potent inhibitors of vesicular transport form the endoplasmic reticulum to the Golgi complex. J Cell Biol 119: 749–761 € derlund H, Ja € ntti J, Kera €nen S (2003) The beta subunit of the Toikkanen JH, Miller KJ, So Sec61p endoplasmic reticulum translocon interacts with the exocyst complex in Saccharomyces cerevisiae. J Biol Chem 278(23): 20946–20953 Ullrich O, Stenmark H, Alexandrov K, Huber LA, Kaibuchi K, Sasaki T, Takai Y, Zerial M (1993) Rab GDP dissociation inhiband itor as a general regulator for the membrane association of rab proteins. J Biol Chem 268(24): 18143–18150 , S, Zerial M, Parton RG (1996) Rab11 regulates recycling Ullrich O, Reinsch S, Urbe through the pericentriolar recycling endosome. J Cell Biol 135(4): 913–924 Wennerberg K, Rossman KL, Der CJ (2005) The Ras superfamily at a glance. J Cell Sci 118 (Pt 5): 843–846 Wilson AL, Erdman RA, Maltese WA (1996) Association of Rab1B with GDP-dissociation inhibitor (GDI) is required for recycling but not initial membrane targeting of the Rab protein. J Biol Chem 271: 10932–10940 Wittmann JG, Rudolph MG (2004) Crystal structure of Rab9 complexed to GDP reveals a dimer with an active conformation of switch II. FEBS Lett 568(1–3): 23–29 Wu X, Bowers B, Rao K, Wei Q, Hammer JA III (1998) Visualization of melanosome dynamics within wild-type and dilute melanocytes suggests a paradigm for myosin V function In vivo. J Cell Biol Dec 28;143(7): 1899–1918 Wu YW, Tan KT, Waldmann H, Goody RS, Alexandrov K (2007) Interaction analysis of prenylated Rab GTPase with Rab escort protein and GDP dissociation inhibitor explains the need for both regulators. Proc Natl Acad Sci USA 104(30): 12294–12299 Zerial M, McBride H (2001) Rab proteins as membrane organizers. Nat Rev Mol Cell Biol 2(2): 107–117 Zhang XM, Ellis S, Sriratana A, Mitchell CA, Rowe T (2004) Sec15 is an effector for the Rab11 GTPase in mammalian cells. J Biol Chem 279(41): 43027–43034

78

*

K. Sato and A. Nakano

COPII Ken Sato and Akihiko Nakano

Introduction Vesicular traffic provides a dynamic and elaborate communication network between the subcellular compartments that define the structure and identity of membrane-bound organelles (Bonifacino and Glick 2004). The molecular and structural mechanisms that direct lipid and protein cargo flow between discontinuous subcellular organelles involve specialized multiprotein machineries that are defined by the molecular and structural properties of cytosolic coat protein complexes (Bonifacino and LippincottSchwartz 2003). The Golgi apparatus is certainly involved in this flow. The endoplasmic reticulum (ER) is responsible for the synthesis of the proteins of most of the cellular organelles. Newly synthesized secretory proteins are translated at the rough ER and translocated into the ER lumen or ER membrane through the translocation channel, where they undergo folding, assembly and post-translational modifications with the aid of a variety of ER chaperones. Correctly folded and assembled secretory proteins are then segregated from ER resident proteins and transported to the Golgi apparatus for further processing and secretion. The ER-to-Golgi transport step is thought to occur via membrane-bound vesicles or carrier intermediates, which are formed by the assembly of the coat protein complex II (COPII) on the ER membranes. COPII is the name given to a cytosolic protein complex required for direct capture of cargo molecules and for the physical deformation of the ER membrane that drives the formation of the so-called “COPII vesicles” or carrier intermediates in anterograde transport from the ER to the Golgi. Protein export by COPII vesicle from the ER is the default ER-toGolgi route that has been proposed in yeast and mammals. Cargo proteins that are destined for delivery to the Golgi apparatus need not only refer to newly synthesized biosynthetic cargo molecules, but also a variety of other machinery proteins that constantly cycle between the ER and the Golgi are included. The molecular mechanistic details of COPII vesicle formation and following cargo delivery to the Golgi have been defined with greater precision through yeast genetics and in vitro reconstitution. Orthologues of almost all of these yeast genes have now been shown to play an equally important role in higher eukaryotes COPII function. Although none of the COPII components are predominantly localized to the Golgi, the COPII-mediated ER-to-Golgi transport is essential for maintaining the function and identity of the Golgi. In this

Mechanisms of COPII vesicle formation

*

79

section, we focus on the individual contributions of the COPII components to the ER-to-Golgi trafficking.

COPII coat recruitment and assembly Pioneering studies by Schekman and colleagues first identified the components of the COPII coat, Sar1 (Nakano and Muramatsu 1989), Sec23/24 (Hicke et al. 1992), Sec13/31 (Salama et al. 1993), and Sar1 regulator Sec12 (Nakano et al. 1988) in the yeast Saccharomyces cerevisiae as proteins required for ER exit in a genetic screen. Their physical interaction and ability to generate COPII vesicles in vitro from isolated ER membranes was shown a few years after the original discovery of these proteins (Barlowe et al. 1994). These COPII components generate COPII vesicles through a sequence of events (Fig. 1). A wealth of genetic and biochemical experiments has led to consensus that in the first step of COPII coat assembly is initiated by activation of the small Ras-like GTPase Sar1. Similar to other small GTPases, conversion of Sar1-GDP to Sar1-GTP is mediated by guanine nucleotide exchange factor (GEF). Sec12 is an ERanchored transmembrane GEF for Sar1 (Barlowe and Schekman 1993). Since Sec12 is strictly regulated to localize to the ER by static retention (Sato et al. 1996), Sar1 activation is restricted to the ER. The GTP binding triggers the exposure of the N-terminal amphipathic a-helix element of Sar1 that inserts into the ER membrane (Huang et al. 2001; Bi et al. 2002). Membrane insertion of the N-terminal helix is a prerequisite for GTP exchange and thus the GTP loading to Sar1 proceeds only in the presence of a membrane surface. Sar1GTP recruits the Sec23–Sec24 heterodimer by binding to the Sec23 portion to form a so-called “prebudding complex” (Kuehn et al. 1998). The Sec23 subunit of Sec23/24 complex is the GTPase-activating protein (GAP) for Sar1 (Yoshihisa et al. 1993) and therefore stimulates Sar1 GTP hydrolysis upon binding to Sar1, which leads to the loss from the membrane of both Sar1 and Sec23/24 complex (Antonny et al. 2001). However, kinetically stable Sec23/24–Sar1 complexes are maintained on the membrane by the presence of the GEF Sec12, which counteracts the GTPase-stimulating activity of the Sec23 by continually recharging Sar1 with GTP (Futai et al. 2004). Subsequently, the prebudding complex recruits Sec13–Sec31 heterotetramer onto the prebudding complex, providing the outer layer of the coat (Lederkremer et al. 2001). Since the Sec31 subunit interacts with both the Sec23 and Sec24 (Shaywitz et al. 1997), but not with Sar1 or directly with the membrane, it is likely that Sec13/31 complex cross-links the preassembled prebudding complexes and drive membrane deformation to form COPII vesicles (60–70 nm in diameter). A recent cryoelectron-microscopy study has proposed that Sec13/31 complex forms a cuboctahedral lattice whose faces are squares and triangles (Stagg et al. 2006). COPII vesicles could be generated from synthetic liposomes incubated with only the above-defined proteins Sec23/24 complex, Sec13/31 complex and GTP-locked Sar1 with non-hydrolyzable GTP analog (Matsuoka et al. 1998). They are therefore “core” components together with the Sec12 GEF.

80

*

K. Sato and A. Nakano

Figure 1. COPII vesicle formation and the selective uptake of cargo proteins. The COPII vesicle formation is initiated by GDP–GTP exchange on Sar1 catalyzed by Sec12. Activated Sar1–GTP binds to the ER membrane and recruits the Sec23/24 subcomplex. The cytoplasmically exposed signal of transmembrane cargo is captured by direct contact with Sec24, forming the prebudding complex. These prebudding complexes are clustered by the Sec13/31 subcomplex, generating COPII-coated vesicles. The assembled cargo has a high affinity for the Sec23/24 because of the combined export signals. This cargo-Sec23/24 association persists during the GDP–GTP exchange of Sar1 catalyzed by Sec12 (upper panel). In contrast, the Sar1 GTP hydrolysis dissociates the weak association between the coat and unassembled cargo or lipid before polymerizing into COPII coats (lower panel). Thus, the prebudding complex stabilities are biased towards the complex including assembled cargo, ensuring that the fully assembled cargo is preferentially incorporated into COPII vesicles.

In addition to the core COPII components, given the likely specialization of trafficking components to accommodate an extraordinary variety of cargo proteins with different structures, sizes and functions with COPII coat assembly, many other proteins are seem to be required to export cargo from the ER. The evolutionarily conserved Sec16 is a large peripheral protein that associates with the ER membrane, and shown to be an essential gene in S. cerevisiae and Pichia pastoris (Espenshade et al. 1995; Connerly et al. 2005). In vitro COPII vesicle budding reaction performed with isolated ER membranes stripped of their Sec16 reveals that vesicle formation is significantly reduced. Sec16 contains domains that make direct contact with the

Mechanisms of COPII vesicle formation

*

81

COPII coat components and may act as a scaffold for assembly of the coat (Gimeno et al. 1996; Supek et al. 2002). Another related protein isolated as a dosage-dependent suppressor of temperature-sensitive SEC16 mutations is Sed4. Sed4 is an integral membrane protein located at the ER membrane, and deletion of the SED4 gene from wild-type cells retards transport from the ER to the Golgi (Gimeno et al. 1995). Although the cytoplasmic domain of Sed4 shares significant homology with that of Sec12, no GEF activity has been found (Saito-Nakano and Nakano 2000). This cytoplasmic domain interacts directly with Sec16 at the ER membrane (Gimeno et al. 1995) and these factors are likely to function together, though their roles are less understood. Yip1p is a member of a conserved family of integral membrane proteins that interact with Rab/Ypt GTPases (Yang et al. 1998). Yip1 forms a complex with at least two other proteins Yif1 and Yos1, and Yip1/Yif1/Yos1 complex cycles between the ER and the Golgi. This complex is required for COPII vesicle biogenesis as Yip1 antibodies inhibit cargo export from the ER and yeast yip1 mutants are defective in COPII vesicle generation (Heidtman et al. 2003, 2005). The mammalian Yip1 is found to interact with the Sec23/24 complex (Tang et al. 2001). A role for Yip1/Yif1/Yos1 complex in COPII assembly is not still elucidated but obviously required for COPII-mediated cargo exit. COPII does not assemble randomly throughout the ER membranes in vivo but instead is concentrated at specialized regions termed transitional ER (tER) or ER exit sites (ERES) (Orci et al. 1991; Bannykh et al. 1996). Immunoelectron and fluorescence microscopy of COPII in coated buds shows a restricted localization to these sites, which represent domains of the ER responsible for the generation of COPII vesicles. At least, Sec16 is shown to be required for normal tER organization (Connerly et al. 2005), but it is not certain how these distinct zones are maintained and the proteins that build these sites are not fully identified. Moreover, the functional consequence of these specialized budding zones is not known, although the extent of organization of these ER sites might influence the morphology of newly forming Golgi elements (Rossanese et al. 1999).

Cargo selection by COPII components It is now widely accepted that the majority of cargo proteins are actively sorted into COPII vesicles. The formation of the prebudding complex consisting of Sec23/24–Sar1 bound to cargo protein is the cargo recognition step prior to polymerization by Sec13/31. The Sec24 subunit in the prebudding complex is generally responsible for interactions with cargo molecules (Mossessova et al. 2003; Miller et al. 2003). The selective capture is basically driven by export signals within the amino acid sequence of each transmembrane cargo protein contained on their cytoplasmically exposed regions, but some transmembrane and most soluble cargo proteins require

82

*

K. Sato and A. Nakano

specific transmembrane cargo receptors/adaptors to mediate the interaction with the COPII coats. These signals are quite diverse, and several classes of ER export signals have so far been identified on the cytoplasmic regions of transmembrane cargo proteins and transmembrane receptors/ adaptors from various organisms (Barlowe 2003). For example, di-acidic ((D/E)x(D/E), with x representing any amino acid residue) and di-hydrophobic (FF, YY, LL, or FY) motifs are well characterized. Interestingly, multiple distinct Sec24 family members are able to pair with Sec23 and bind to export signals different from those recognized by Sec24 (Pagano et al. 1999; Miller et al. 2003), expanding the cargo multiplicity captured by COPII coat. The ER contains a certain amount of newly synthesized unfolded or unassembled cargo proteins, which should be segregated from secretory proteins to be exported. To ensure efficient incorporation of fully folded and assembled cargo proteins into COPII vesicles, many exported proteins are required multiple signals in a specific display that is achieved only through proper assembly. In vitro experiments with cargo-reconstituted proteoliposomes have demonstrated that the Sec23/24 can remain transiently associated with properly assembled cargo even after Sar1 GTP hydrolysis, while the interaction between Sec23/24 and unassembled cargo or lipid on membranes is disrupted immediately upon Sar1 GTP hydrolysis (Sato and Nakano 2005). This is probably due to combined export signals displayed on assembled complex might increase the affinity to Sec23/24. In the presence of Sec12, stable binding of Sec23/24 with assembled cargo proteins has been observed due to continuous reactivation of Sar1 to its GTP-bound form before Sec23/24 release. In contrast, continuous Sec23/24 binding and release occurs on unassembled cargo and lipids accompanied by Sar1 GTPase cycles (Fig. 1). Thus, the prebudding complex stability is biased toward the complex containing properly assembled cargo proteins during Sar1 GTPase cycles. Thus, Sar1 selectively promotes exclusion of unassembled cargo proteins from emerging COPII vesicles by virtue of its GTP hydrolysis, explaining how only proper cargo molecules can be efficiently concentrated into COPII vesicles.

COPII vesicle budding and pinch-off Sar1 provides not only as a spatial landmark that recruits Sec23/24 and Sec13/31 but also as a mediator of membrane deformation and vesicle scission. There is evidence that Sar1-GTP can deform liposomes into tubules about 26 nm in diameter (Lee et al. 2005). This was observed only with activated Sar1-GTP, suggesting an involvement of the N-terminal helix into the outer leaflet of the ER membrane would act to displace the lipid headgroups, and this asymmetric expansion may promote curvature toward the cytoplasmic region (Farsad and De Camilli 2003). Sar1-GTP then stimulates the binding of an inner shell of Sec23/24 to form prebudding complex. It has

Mechanisms of COPII vesicle formation

*

83

been shown in structural studies that the Sec23/24 complex has a concave surface that matches the size of a 60-nm vesicle and is thus thought to facilitate membrane bending (Bi et al. 2002). Such prebudding complexes only give a random distribution of local membrane deformations because they have no ability to interact with each other. Sec13/31 alone is shown to self-assemble in solution into COPII cage-like structures (Stagg et al. 2006), and hence the spherical shape of the COPII coat seems to be formed by prebudding complex clustering by Sec13/31. Importantly, although Sec23/24 and Sec13/31 complexes can form budded vesicles, the buds rarely closed off to generate free vesicles when Sar1 is anchored in the nickel-conjugated lipids containing liposomes by a polyhistidine tag in place of N-terminal helix (Lee et al. 2005). So, the neck of a COPII coated bud is not likely to break spontaneously and the N-terminal helix of Sar1 may have an active role in membrane fission. Other experiments suggest that fission is more efficient when Sar1 hydrolyses GTP (Bielli et al. 2005), suggesting that the Sar1 GTP hydrolysis may play some roles in vesicle fission. Further analyses are required to determine how Sar1 N-terminal helix initiates and completes the fission of a COPII vesicle.

Building a Golgi with COPII vesicles COPII vesicles shed their coats before fusion with Golgi membrane and this uncoating reaction is thought to be achieved by the Sar1 GTP hydrolysis (Oka and Nakano 1994). In mammalian cells COPII vesicles derived from the tER do not fuse directly to the Golgi membrane, instead, they appear to tether and fuse to each other (homotypic fusion) to form carrier intermediates that lie adjacent to the tER (Xu and Hay 2004). COPII vesicles continue to fuse with the intermediates, which becomes larger and eventually thought to fuse with the Golgi. Numerous names have been given to the discontinuous carriers that move from the ER compartment to Golgi such as ER–Golgi intermediate compartment (ERGIC), pre-Golgi intermediate and vesicular tubular complex (VTC). Although machineries of the ERto-Golgi trafficking are highly conserved from yeast to human, no equivalent to a pre-Golgi compartment has been identified in the yeast, and COPII vesicles are thought to fuse directly to the Golgi membranes (heterotypic fusion). However, the high degree of homology between yeast and mammalian components required for the late stages of ER-to-Golgi traffic raises the possibility that yeast COPII vesicles may undergo homotypic tethering and fusion before they fuse with the Golgi. The other way round, the recent observation that the pre-Golgi compartment is stable (Ben-Tekaya et al. 2005; Sannerud et al. 2006) has raised the possibility that homotypic fusion is not exclusive and mammalian COPII vesicles may also heterotypically fuse with stable ER–Golgi intermediate compartments. Much is yet to be determined on how the pre-Golgi structure is formed from COPII vesicles.

84

*

K. Sato and A. Nakano

References Antonny B, Madden D, Hamamoto S, Orci L, Schekman R (2001) Dynamics of the COPII coat with GTP and stable analogues. Nat Cell Biol 3: 531–537 Bannykh SI, Rowe T, Balch WE (1996) The organization of endoplasmic reticulum export complexes. J Cell Biol 135: 19–35 Barlowe C (2003) Signals for COPII-dependent export from the ER: what’s the ticket out? Trends Cell Biol 13: 295–300 Barlowe C, Orci L, Yeung T, Hosobuchi M, Hamamoto S, Salama N, Rexach MF, Ravazzola M, Amherdt M, Schekman R (1994) COPII: a membrane coat formed by Sec proteins that drive vesicle budding from the endoplasmic reticulum. Cell 77: 895–907 Barlowe C, Schekman R (1993) SEC12 encodes a guanine-nucleotide-exchange factor essential for transport vesicle budding from the ER. Nature 365: 347–349 Ben-Tekaya H, Miura K, Pepperkok R, Hauri HP (2005) Live imaging of bidirectional traffic from the ERGIC. J Cell Sci 118: 357–367 Bi X, Corpina RA, Goldberg J (2002) Structure of the Sec23/24-Sar1 pre-budding complex of the COPII vesicle coat. Nature 419: 271–277 Bielli A, Haney CJ, Gabreski G, Watkins SC, Bannykh SI, Aridor M (2005) Regulation of Sar1 NH2 terminus by GTP binding and hydrolysis promotes membrane deformation to control COPII vesicle fission. J Cell Biol 171: 919–924 Bonifacino JS, Glick BS (2004) The mechanisms of vesicle budding and fusion. Cell 116: 153–166 Bonifacino JS, Lippincott-Schwartz J (2003) Coat proteins: shaping membrane transport. Nat Rev Mol Cell Biol 4: 409–414 Connerly PL, Esaki M, Montegna EA, Strongin DE, Levi S, Soderholm J, Glick BS (2005) Sec16 is a determinant of transitional ER organization. Curr Biol 15: 1439–1447 Espenshade P, Gimeno RE, Holzmacher E, Teung P, Kaiser CA (1995) Yeast SEC16 gene encodes a multidomain vesicle coat protein that interacts with Sec23p. J Cell Biol 131: 311–324 Farsad K, De Camilli P (2003) Mechanisms of membrane deformation. Curr Opin Cell Biol 15: 372–381 Futai E, Hamamoto S, Orci L, Schekman R (2004) GTP/GDP exchange by Sec12p enables COPII vesicle bud formation on synthetic liposomes. Embo J 23: 4146–4155 Gimeno RE, Espenshade P, Kaiser CA (1995) SED4 encodes a yeast endoplasmic reticulum protein that binds Sec16p and participates in vesicle formation. J Cell Biol 131: 325–338 Gimeno RE, Espenshade P, Kaiser CA (1996) COPII coat subunit interactions: Sec24p and Sec23p bind to adjacent regions of Sec16p. Mol Biol Cell 7: 1815–1823 Heidtman M, Chen CZ, Collins RN, Barlowe C (2003) A role for Yip1p in COPII vesicle biogenesis. J Cell Biol 163: 57–69 Heidtman M, Chen CZ, Collins RN, Barlowe C (2005) Yos1p is a novel subunit of the Yip1p–Yif1p complex and is required for transport between the endoplasmic reticulum and the Golgi complex. Mol Biol Cell 16: 1673–1683 Hicke L, Yoshihisa T, Schekman R (1992) Sec23p and a novel 105-kDa protein function as a multimeric complex to promote vesicle budding and protein transport from the endoplasmic reticulum. Mol Biol Cell 3: 667–676 Huang M, Weissman JT, Beraud-Dufour S, Luan P, Wang C, Chen W, Aridor M, Wilson IA, Balch WE (2001) Crystal structure of Sar1-GDP at 1.7 A resolution and the role of the NH2 terminus in ER export. J Cell Biol 155: 937–948 Kuehn MJ, Herrmann JM, Schekman R (1998) COPII-cargo interactions direct protein sorting into ER-derived transport vesicles. Nature 391: 187–190

Mechanisms of COPII vesicle formation

*

85

Lederkremer GZ, Cheng Y, Petre BM, Vogan E, Springer S, Schekman R, Walz T, Kirchhausen T (2001) Structure of the Sec23p/24p and Sec13p/31p complexes of COPII. Proc Natl Acad Sci USA 98: 10704–10709 Lee MC, Orci L, Hamamoto S, Futai E, Ravazzola M, Schekman R (2005) Sar1p N-terminal helix initiates membrane curvature and completes the fission of a COPII vesicle. Cell 122: 605–617 Matsuoka K, Orci L, Amherdt M, Bednarek SY, Hamamoto S, Schekman R, Yeung T (1998) COPII-coated vesicle formation reconstituted with purified coat proteins and chemically defined liposomes. Cell 93: 263–275 Miller EA, Beilharz TH, Malkus PN, Lee MC, Hamamoto S, Orci L, Schekman R (2003) Multiple cargo binding sites on the COPII subunit Sec24p ensure capture of diverse membrane proteins into transport vesicles. Cell 114: 497–509 Mossessova E, Bickford LC, Goldberg J (2003) SNARE selectivity of the COPII coat. Cell 114: 483–495 Nakano A, Brada D, Schekman R (1988) A membrane glycoprotein, Sec12p, required for protein transport from the endoplasmic reticulum to the Golgi apparatus in yeast. J Cell Biol 107: 851–863 Nakano A, Muramatsu M (1989) A novel GTP-binding protein, Sar1p, is involved in transport from the endoplasmic reticulum to the Golgi apparatus. J Cell Biol 109: 2677–2691 Oka T, Nakano A (1994) Inhibition of GTP hydrolysis by Sar1p causes accumulation of vesicles that are a functional intermediate of the ER-to-Golgi transport in yeast. J Cell Biol 124: 425–434 Orci L, Ravazzola M, Meda P, Holcomb C, Moore HP, Hicke L, Schekman R (1991) Mammalian Sec23p homologue is restricted to the endoplasmic reticulum transitional cytoplasm. Proc Natl Acad Sci USA 88: 8611–8615 Pagano A, Letourneur F, Garcia-Estefania D, Carpentier JL, Orci L, Paccaud JP (1999) Sec24 proteins and sorting at the endoplasmic reticulum. J Biol Chem 274: 7833–7840 Rossanese OW, Soderholm J, Bevis BJ, Sears IB, O’Connor J, Williamson EK, Glick BS (1999) Golgi structure correlates with transitional endoplasmic reticulum organization in Pichia pastoris and Saccharomyces cerevisiae. J Cell Biol 145: 69–81 Saito-Nakano Y, Nakano A (2000) Sed4p functions as a positive regulator of Sar1p probably through inhibition of the GTPase activation by Sec23p. Genes Cells 5: 1039–1048 Salama NR, Yeung T, Schekman RW (1993) The Sec13p complex and reconstitution of vesicle budding from the ER with purified cytosolic proteins. EMBO J 12: 4073–4082 Sannerud R, Marie M, Nizak C, Dale HA, Pernet-Gallay K, Perez F, Goud B, Saraste J (2006) Rab1 defines a novel pathway connecting the pre-Golgi intermediate compartment with the cell periphery. Mol Biol Cell 17: 1514–1526 Sato K, Nakano A (2005) Dissection of COPII subunit-cargo assembly and disassembly kinetics during Sar1p–GTP hydrolysis. Nat Struct Mol Biol 12: 167–174 Sato M, Sato K, Nakano A (1996) Endoplasmic reticulum localization of Sec12p is achieved by two mechanisms: Rer1p-dependent retrieval that requires the transmembrane domain and Rer1p-independent retention that involves the cytoplasmic domain. J Cell Biol 134: 279–293 Shaywitz DA, Espenshade PJ, Gimeno RE, Kaiser CA (1997) COPII subunit interactions in the assembly of the vesicle coat. J Biol Chem 272: 25413–25416 Stagg SM, Gurkan C, Fowler DM, LaPointe P, Foss TR, Potter CS, Carragher B, Balch WE (2006) Structure of the Sec13/31 COPII coat cage. Nature 439: 234–238 Supek F, Madden DT, Hamamoto S, Orci L, Schekman R (2002) Sec16p potentiates the action of COPII proteins to bud transport vesicles. J Cell Biol 158: 1029–1038 Tang BL, Ong YS, Huang B, Wei S, Wong ET, Qi R, Horstmann H, Hong W (2001) A membrane protein enriched in endoplasmic reticulum exit sites interacts with COPII. J Biol Chem 276: 40008–40017

86

*

K. Sato and A. Nakano

Xu D, Hay JC (2004) Reconstitution of COPII vesicle fusion to generate a pre-Golgi intermediate compartment. J Cell Biol 167: 997–1003 Yang X, Matern HT, Gallwitz D (1998) Specific binding to a novel and essential Golgi membrane protein (Yip1p) functionally links the transport GTPases Ypt1p and Ypt31p. Embo J 17: 4954–4963 Yoshihisa T, Barlowe C, Schekman R (1993) Requirement for a GTPase-activating protein in vesicle budding from the endoplasmic reticulum. Science 259: 1466–1468

COPI mechanisms and roles

*

87

COPI: mechanisms and transport roles Victor W. Hsu, Jia-Shu Yang, and Stella Y. Lee

Coat protein I (COPI) is considered one of the best characterized coat complexes, which represent the core machinery by which vesicle formation and cargo sorting are coupled to initiate vesicular transport (Bonifacino and Lippincott-Schwartz 2003; McMahon and Mills 2004). Our understanding of the molecular mechanisms by which COPI acts and the transport pathways in which it operates has evolved significantly over the years, and with considerable accompanying controversy. These aspects of COPI research will be reviewed. See also Fig. 1 for a timeline that summarizes its key discoveries.

Historic background The origin of COPI could be traced to two lines of investigations, which initially seemed distinct. One avenue of research had sought to reconstitute transport among the Golgi cisternae using a cell-free system (Balch et al. 1984). This reconstitution took advantage of a mutant cell line that was defective in a specific medial Golgi glycosylation reaction. These cells were infected with the vesicular stomatitis virus (VSV). A Golgi-enriched fraction was then collected and incubated with a similarly enriched fraction from uninfected wild-type cells. Transport between the two Golgi fractions was marked by the transfer of the major G protein expressed by the VSV infection (and thus known as VSVG), which was monitored by its glycosylation pattern upon arrival to the acceptor fraction (Balch et al. 1984). Subsequent characterization of this reconstitution system revealed that it could be blocked by the addition of a nonhydrolyzable analog of guanosine triphosphate (GTP), known as GTPgS. Electron microscopy (EM) revealed that this block induced the accumulation of coated vesicles (Orci et al. 1986). As this coating was distinct from the only other coat protein known at the time, clathrin with the AP2 adaptor, the Golgi-derived vesicles were proposed to be formed by a novel, non-clathrin coat complex (Orci et al. 1986). Later, key components of this coating were identified to be part of a multimeric complex, known as coatomer (Waters et al. 1991). Moreover, coated vesicles formed in the presence of GTPgS contained coatomer in stoichiometric level to the small GTPase ADP Ribosylation factor 1 (ARF1) (Serafini et al. 1991a). Thus, the novel coat was considered to consist of ARF1 and coatomer (Orci et al. 1993; Serafini et al. 1991a), and given the name coat protein (COP) (Serafini et al. 1991a; Waters et al. 1991). Subsequent-

88

*

V. W. Hsu et al.

Figure 1. A timeline summary of key insights in COPI research. Key mechanistic insights or events are highlighted above the arrow line, while key components are highlighted below this line.

ly, this name was changed to COPI when COPII was identified (Barlowe et al. 1994). Another line of research had sought to elucidate how a pharmacologic agent, brefeldin-A (BFA), blocked cellular secretion. Initially, BFA was found to induce the redistribution of the Golgi complex to the endoplasmic reticulum (ER), resulting in a mixed organellar system that was incapable of supporting transport to the plasma membrane (Lippincott-Schwartz et al. 1989). To understand how BFA induced this redistribution of the Golgi complex, an early insight was that certain Golgi-localized proteins were observed to become released to the cytosol upon treatment by BFA. The redistribution of one such protein, dubbed the 110 kDa protein, was notable due to its additional regulation by GTPgS, such that pre-incubation with GTPgS prevented BFA from releasing this protein on Golgi membrane (Donaldson et al. 1991). Notably, this line of investigation merged with the ongoing studies on reconstituted intra-Golgi transport, when the 110 kDa protein was found to be identical to a subunit of coatomer, named b-COP (Duden et al. 1991; Serafini et al. 1991b). Subsequent studies revealed that activation of ARF1 by its binding of GTP recruited coatomer from the cytosol onto Golgi membrane to initiate COPI vesicle formation (Donaldson et al. 1992a; Orci et al. 1991). Moreover, BFA inhibited ARF1 activation by blocking the guanine nucleotide exchange activity that catalyzed this activation (Donaldson et al. 1992b; Helms and Rothman 1992). These findings in COPI research have provided seminal contributions to our understanding of how the ARF family of small GTPases regulates coat proteins to initiate vesicle formation. However, subsequent studies on this regulation have revealed surprising mechanistic complexities. Moreover, whereas the role of COPI vesicles was originally envisioned to mediate anterograde (forward) transport through the Golgi complex, this role has also undergone dramatic revisions subsequently. Below, we will first discuss the key components discovered for COPI vesicle formation, and then review the intracellular pathways in which COPI has been implicated to act.

COPI mechanisms and roles

*

89

Key components in COPI vesicle formation Much of our current mechanistic understanding of how the different factors act in COPI vesicle formation has come largely from a vesicle reconstitution system that evolved from the original intra-Golgi transport assay. Experimentally, this involves the incubation of Golgi membrane with purified protein components, which initially suggested how the GTPase cycle of ARF1 regulates coatomer to initiate COPI vesicle formation (Orci et al. 1991; Serafini et al. 1991a; Tanigawa et al. 1993). Later, in its more refined form, this reconstitution system has been instrumental in identifying additional key factors (Lee et al. 2005; Ostermann et al. 1993; Yang et al. 2002, 2005, 2006). Summarizing how these factors are currently thought to act, Fig. 2 proposes a model.

Coatomer Coatomer is a multimeric complex (composed of a, b, b0 , g, d, e, and z subunits) (Harrison-Lavoie et al. 1993; Stenbeck et al. 1993; Waters et al. 1991). Its role as the principal coating on COPI vesicles was initially suggested through the COPI vesicle reconstitution system (Orci et al. 1993; Serafini et al. 1991a). Examining a minimal system using liposomal membrane rather than Golgi membrane, later studies further suggested that coatomer and ARF1 were sufficient to form vesicular-like structures (Bremser et al. 1999; Spang et al. 1998).

Figure 2. A model proposed for how key components act in COPI vesicle formation. Besides coatomer, ARFGAP1 is now realized to be another coat component, while BARS/endophilin B have been identified to have mechanistically interchangeable roles in the fission step.

90

*

V. W. Hsu et al.

However, more recent vesicle reconstitution studies that have re-visited the use of Golgi membrane have revealed additional factors needed, not only for COPI vesicle formation but also for its proper cargo sorting (see subsections below that discuss such factors in more detail). Sub-complexes of coatomer have been delineated based on physical interactions among its subunits (Eugster et al. 2000; Lowe and Kreis 1995; Pavel et al. 1998). One sub-complex consists of a, b0 , and e subunits, while the other consists of the remaining subunits. With their sequencing, the b, g, and z subunits of coatomer have been revealed to have detectable levels of sequence similarity to b, m, and s subunits of the clathrin AP2 adaptors, respectively (Cosson et al. 1996; Serafini et al. 1991b). Supporting this parallel, structural analysis also indicates that g-COP shares folding similarity to the appendage domains of a and b subunits of the clathrin AP2 adaptor (Hoffman et al. 2003). Altogether, these findings have led to the suggestion that coatomer subunits are organized in structurally similar ways as the clathrin coat complex (McMahon and Mills 2004). For cargo sorting, coatomer recognizes two distinct types of di-basic sequences on cargo proteins, known as sorting signals. Di-lysine signals are located near the carboxyl terminus within the cytoplasmic domains of COPI cargo proteins. In contrast, di-arginine signals occur near the amino terminus within the cytoplasmic domain of cargo proteins. Moreover, whereas the dilysine motif shows more strict spacing requirement with respect to the end of the carboxy terminus, the di-arginine motif exhibits more flexibility within the amino terminal region of cargo proteins (Teasdale and Jackson 1996; Zerangue et al. 2001). These two sorting signals are also recognized by distinct components of coatomer, with the di-lysine motif interacting with the a, b0 , and g subunits (Cosson and Letourneur 1994; Harter et al. 1996), and the diarginine motif interacting with b and d subunits (Michelsen et al. 2007). Recently, further complexity in COPI cargo sorting has been revealed by studies on ion channels. Phosphorylation near some di-basic sequences in potassium channels was found to inhibit the binding of coatomer to these motifs. As a mechanistic basis for this inhibition, isoforms of 14-3-3 were identified to recognize such phosphorylated sites, and thereby providing steric hindrance to prevent binding by coatomer (O’Kelly et al. 2002; Yuan et al. 2003). As 14-3-3 is a large protein family that participate in diverse biological events (Aitken, 1996), an intriguing aspect of this finding is that regulation of cargo sorting by the COPI complex may be coordinated with other cellular events through the actions of 14-3-3.

ARF1 ARF1 is the founding member of the ARF family of small GTPases, which are now well known to act in vesicular transport, actin rearrangement and signaling (D’Souza-Schorey and Chavrier 2006). ARF1 was originally discovered as a co-factor needed for the ADP-ribosylation of cholera toxin (Kahn and Gilman 1984), hence the origin of its name as ADP-ribosylation factor 1.

COPI mechanisms and roles

*

91

However, in contrast to this role for which the physiologic meaning still remains relatively obscure, its subsequently elucidated role in regulating COPI transport has considerably advanced our understanding of how vesicle formation is regulated to initiate vesicular transport (D’Souza-Schorey and Chavrier 2006). Like all small GTPases, guanine nucleotide exchange factors (GEFs) are needed to catalyze the activation of ARF1 (Casanova 2007), while GTPaseactivating proteins (GAPs) are required to catalyze its deactivation (Inoue and Randazzo 2007). As another mode of regulation, ARF1 is myristoylated at its amino terminus, which is critical for its stabilization on target membrane upon its binding to GTP (Franco et al. 1996; Randazzo et al. 1995). ARF1 has been shown to interact with multiple effectors, including coatomer, specific cargo proteins, and multiple lipid-modifying enzymes (D’Souza-Schorey and Chavrier 2006). Moreover, a recently appreciated effector has been its GAP, which has been shown to act as a coat component in COPI vesicle formation (Lee et al. 2005; Yang et al. 2002). Altogether, these findings suggest that, rather than one specific interaction, the coordination of multiple interactions of activated ARF1 results in the stable recruitment of the coat complex onto membrane for COPI vesicle formation.

GEF The first GEFs identified for ARF small GTPases were Gea1p and Gea2p, two homologous yeast proteins (Peyroche et al. 1996). While neither was essential for yeast viability, deletion of both led to lethality (Peyroche et al. 1996). Moreover, similar to perturbation of coatomer subunits, perturbing these GEFs also led to secretion abnormalities and perturbation of the Golgi complex (Peyroche et al. 2001). Thus, these two yeast GEFs are currently thought to act redundantly in activating ARF1 for COPI transport. The first mammalian GEF identified for ARF small GTPases was ARNO, based on its sequence similarity to the yeast Gea1p and Gea2p, particularly within the region responsible for catalysis, known as the Sec7 domain (Chardin et al. 1996). However, ARNO turned out to be insensitive to BFA (Chardin et al. 1996). Moreover, it had a significant distribution on the plasma membrane, where it acted on ARF6 for the regulation of actin rearrangement (Frank et al. 1998a, b; Santy and Casanova 2001). GBF1 was subsequently identified as the more likely GEF that acts on ARF1 in COPI transport. Overexpression of GBF1 conferred resistance to cells treated with BFA (Kawamoto et al. 2002), while expression of a catalytic dead mutant mimicked the effect of BFA in redistributing coatomer from the Golgi (Garcia-Mata et al. 2003). Moreover, a viral protein expressed by picornavirus has recently been shown to inhibit GBF1, resulting in trafficking defects likely attributable to perturbations in COPI transport (Wessels et al. 2006). However, in contrast to all other key components currently known for COPI vesicle formation, the mechanistic details of how GBF1 participates in COPI transport have not been studied using the COPI vesicle reconstitution system.

92

*

V. W. Hsu et al.

GAP The first GAP identified to act on an ARF small GTPase is ARFGAP1 (Cukierman et al. 1995). However, the manner by which this GAP participates in COPI transport has turned out to be surprisingly complex. In early studies, when Golgi membrane was incubated with purified ARF1 and coatomer, introduction of either an ARF1 mutant with impaired GTPase activity (ARF1-Q71L) or the addition of GTPgS to inhibit this activity, led to the formation of coated vesicles that could not undergo uncoating (Tanigawa et al. 1993). Moreover, later COPI vesicle reconstitution studies that involved the incubation of liposomal membrane with purified protein components also concluded that the GAP activity acted in COPI vesicle uncoating (Bigay et al. 2003; Reinhard et al. 2003). Altogether, these findings suggested that the GAP would act simply in destabilizing coatomer on membrane, leading to its mechanistic assignment in COPI vesicle uncoating. A need to revise this view was initially suggested by the finding that the GAP activity promoted COPI cargo sorting (Lanoix et al. 1999; Nickel et al. 1998; Pepperkok et al. 2000). As cargo sorting is tightly coupled to vesicle formation (Springer et al. 1999), early attempts to reconcile the seemingly disparate roles of the GAP activity (in promoting cargo sorting and yet also in inhibiting vesicle formation) focused on how it may be regulated. One notable example was based on the principle of kinetic proofreading, for which distinct members of the p24 family of cargo proteins were shown to exhibit differential regulatory effects on the GAP activity (Goldberg 2000; Lanoix et al. 2001). Moreover, another mechanism was suggested by the observation that the GAP activity was regulated by membrane curvature (Bigay et al. 2003). Recently however, the prevailing view that the GAP activity antagonizes COPI vesicle formation has been challenged by examining the behavior of ARFGAP1 in a refined COPI vesicle reconstitution system that used Golgi membrane rather than the simpler liposomal membrane. Remarkably, reconstituted COPI vesicles were found to be coated with coatomer and ARFGAP1 in stoichiometric levels (Yang et al. 2002). Moreover, ARFGAP1 was found to play a direct role in recruiting coatomer to bind cargo proteins (Lee et al. 2005). Also, its catalytic activity was found to be required for vesicle formation (Lee et al. 2005). Altogether, these findings have led to a new conclusion that, rather than simply acting as a negative regulator of the ARF1 GTPase cycle, the GAP also acts as an effector by being a component of the COPI complex.

Fission factors Besides GAP, the COPI vesicle reconstitution system that uses Golgi membrane has been further refined recently, leading to the identification of brefeldin-A ADP-ribosylated substrate (BARS) and endophilin B to act mechanistically interchangeable for a late step of COPI vesicle formation, known as the fission step.

COPI mechanisms and roles

*

93

BARS BARS has a complex history. Its role in membrane traffic was originally suggested by the discovery that BFA activated an endogenous ADP-ribosylation reaction with BARS identified as one target (Di Girolamo et al. 1995; Spano et al. 1999). With its sequencing, BARS was realized to be a splice variant of C-terminal binding protein 1 (CtBP1) (Spano et al. 1999), which belonged to a protein family previously characterized to act as transcription co-repressors (Chinnadurai 2002). Thus, BARS is also known as CtBP3. Initial mechanistic insight on BARS suggested that it possessed an acyltransferase activity, which was associated with its ability to induce the fission of Golgi tubules into vesicles (Weigert et al. 1999). Subsequently however, this acyltransferase activity was shown not to be intrinsic to BARS (Gallop et al. 2005). In the meantime, when Golgi membrane was washed more stringently in the refined COPI vesicle reconstitution system, the existing known factors in COPI vesicle formation no longer became sufficient, implying additional factor(s) needed. By examining cytosol for a fraction that could complement the existing factors, a recent study has identified BARS (Yang et al. 2005). Characterization of its role revealed that BARS played a critical role during the fission process, which is late stage of vesicle formation when the neck of coated buds underwent constriction for their eventual release as vesicles from compartmental membrane (Yang et al. 2005). Also, a minimal domain of BARS was shown to be sufficient for this role, and this domain did not possess acyltransferase activity (Yang et al. 2005). Thus, even though these recent findings in COPI transport has confirmed a role for BARS in membrane fission, how this action is achieved remains to be determined. How BARS is regulated to act in COPI vesicle formation has also been revealed to be surprisingly complex. For transcription, CtBP members required NAD as a co-factor (Chinnadurai 2002). However, binding to this cofactor inhibited the role of BARS in COPI vesicle formation by preventing its association with ARFGAP1 (Yang et al. 2005). Instead, binding to p-coA by BARS as an alternate co-factor allowed its association with ARFGAP1 for COPI vesicle fission (Yang et al. 2005). Intriguingly, these observations have raised the possibility that transcription and transport, two intracellular events that had not been previously appreciated to have much in common, could be coregulated through the action of BARS.

Endophilin B Endophilin B belongs to a protein family for which endophilin A had been the best characterized member. Endophilin A has been shown to act in the fission of clathrin coated vesicles (with the AP2 adaptor) from the plasma membrane (Ringstad et al. 1999; Schmidt et al. 1999; Simpson et al. 1999). In contrast, endophilin B was initially found to localize at the Golgi but its function had remained unknown (Farsad et al. 2001). A role for endophilin B in COPI transport was recently discovered in an unexpected manner. As background, three forms of CtBP have been found: CtBP1, CtBP2, and CtBP3 (also known as

94

*

V. W. Hsu et al.

BARS). CtBP2 is encoded by one gene, while the other two members are expressed through another gene. Deletion of both genes that encode for all CtBP members in the mouse, while being lethal with respect to the generation of the whole animal, led to viable embryonic cells (Hildebrand and Soriano 2002). This result seemingly contradicted the finding that COPI transport is critical for cell viability (Guo et al. 1994; Hosobuchi et al. 1992). Providing an explanation, a recent study examined embryonic cells derived from mice that had all CtBP members deleted, and found that these cells used endophilin B instead of BARS for COPI vesicle formation (Yang et al. 2006). Remarkably, endophilin B was found also to act in COPI vesicle fission in different adult mouse cell types, in a manner mutually exclusive to that of BARS (Yang et al. 2006). This latter finding is surprising, because all previously identified critical factors in COPI vesicle formation had been shown to play inflexibly roles from yeast to mammals.

COPI transport pathways Since its original discovery, COPI vesicles are now implicated to act in both intra-Golgi transport and also transport from the Golgi to the ER. While the discovery of the latter role was initially considered surprising, it has become more widely accepted in recent years. In contrast, the precise role of COPI vesicles in intra-Golgi transport remains debated. Notably, COPI has also been suggested recently to have roles other than in forming transport vesicles, such as potentially in organizing membrane domains on organellar compartments (Bonifacino and Lippincott-Schwartz 2003). However, as this proposed role has been relatively uncharacterized, we will focus instead on its intensely investigated roles in intracellular transport pathways.

Intra-Golgi transport The intra-Golgi transport assay originally designed to monitor the transfer of VSVG between Golgi cisternae (Balch et al. 1984). Thus, as VSVG is transported anterograde through the secretory system to reach the plasma membrane, the simplest explanation at the time was that COPI vesicles functioned in anterograde transport. However, a series of subsequent findings have led to major changes in this view. Initially, coatomer was found to bind to the cytoplasmic domain of cargo proteins that contained a di-lysine-based motif (Cosson and Letourneur 1994), with functional evidence suggesting that this binding led to the retrograde transport of di-lysine-containing cargo proteins from the Golgi to the ER (Letourneur et al. 1994). Other yeast studies at this time also revealed that perturbation of coatomer subunits perturbed anterograde transport between the ER and the Golgi by affecting more directly the retrograde arm of bidirectional transport that connected these two organelles (Gaynor and Emr 1997; Lewis and Pelham 1996). A role for COPI in retrograde transport was further suggested subsequently by the realization that cisternal maturation played a predominant role in

COPI mechanisms and roles

*

95

anterograde intra-Golgi transport (Bonfanti et al. 1998; Mironov et al. 2001). In this model, rather than being static membrane compartments, which was the prevailing view at the time, Golgi cisternae dynamically transformed themselves by earlier stacks becoming later stacks. This dynamic transformation provided the basis by which anterograde intra-Golgi transport was accomplished. A key implication of this model was that retrograde transport must also exist, as markers of Golgi cisternae (which were typically transmembrane Golgi glycosylation enzymes) were noted to remain in their distinct distribution in the face of anterograde transport by cisternal maturation. Thus, COPI vesicles were scrutinized for a potential role in retrograde intraGolgi transport. For these studies, two complementary approaches were taken: (i) a biochemical approach that examined the content of COPI vesicles reconstituted from Golgi membrane, and (ii) a morphologic approach that visualized COPI vesicles and their content by immunogold electron microscopy. The biochemical approach initially found that Golgi enzymes were concentrated in COPI vesicles, and thus, concluding that these vesicles played a role in retrograde intra-Golgi transport. Specifically, in an initial re-visit of the intra-Golgi transport assay, Golgi enzymes were found to be transferred among Golgi stacks in a COPI-dependent manner (Love et al. 1998). Moreover, COPI vesicles reconstituted in the presence of GTPgS that blocked ARF1 deactivation were found to have reduced cargo sorting (Lanoix et al. 1999; Nickel et al. 1998; Pepperkok et al. 2000). When vesicles were generated instead by incubating Golgi membrane with cytosol in the presence of GTP, Golgi enzymes, but not examples of anterograde cargo protein, were detected to be concentrated in the reconstituted vesicles (Lanoix et al. 1999). However, these biochemical studies could be criticized that their conclusion rested on indirect evidence. As cytosol contained factor(s) that prevented the stabilization of COPI on vesicular membrane, these studies examined vesicles that did not have COPI coating. Instead, these vesicles were suggested to have been coated by COPI and then having undergone uncoating, because their generation was inhibited when coatomer was depleted from cytosol (Lanoix et al. 1999). In the other main approach, immunogold EM studies were undertaken. However, these studies led to conflicting results, with some studies concluding that Golgi enzymes were not transported by COPI vesicles (Cosson et al. 2002; Orci et al. 2000a), while others concluding the opposite (Martinez-Menarguez et al. 2001). A limitation of the traditional EM approach that involves thinsectioning has been the ability to distinguish vesicles from buds and tubules with certainty. To overcome this hurdle, EM tomography has been used more recently. These studies concluded that Golgi enzymes were not significantly concentrated in COPI vesicles (Kweon et al. 2004). Instead, an intriguing possibility was raised that Golgi enzymes could be transported retrograde through tubular connections among the Golgi stacks (Trucco et al. 2004). A recent review has summarized the conflicting conclusions regarding the role

96

*

V. W. Hsu et al.

of COPI in retrograde intra-Golgi transport, suggesting potential methodologic and interpretational differences (Rabouille and Klumperman 2005). As it currently stands however, whether COPI vesicles plays a significant role in this transport remains disputed. Even more uncertain has been whether COPI vesicles participate in anterograde intra-Golgi transport, its originally proposed role. As alluded to above, early evidence for this role came from the cell-free reconstitution of intra-Golgi transport that monitored the transfer VSVG, which is an anterograde cargo protein (Balch et al. 1984). However, later studies suggested experimental caveats for this observation. First, because the relative fraction of VSVG detected in reconstituted COPI vesicles was low as compared to host proteins, a suggestion was made that the detection of VSVG in reconstituted COPI vesicles represented their mis-sorting due to viral-mediated overexpression of this cargo protein (Love et al. 1998). Second, GTPgS that was used to reconstitute COPI vesicles in the original studies was subsequently revealed to prevent the proper sorting of cargo proteins (Lanoix et al. 1999; Nickel et al. 1998; Pepperkok et al. 2000). When vesicles that were reconstituted in the presence of GTP and shown to be dependent on COPI, anterograde cargo proteins were not enriched in these vesicles (Lanoix et al. 1999; Love et al. 1998). Finally, EM studies have led to conflicting conclusions regarding whether anterograde cargo proteins are concentrated in COPI vesicles (Martinez-Menarguez et al. 2001; Mironov et al. 2001; Orci et al. 2000b, 1997). Recent studies have focused on the possibility that subpopulations of COPI vesicles may exist, for which one potential role would be in anterograde intraGolgi transport. Variants of different subunits of coatomer have been identified (Wegmann et al. 2004), and shown to have similar but non-overlapping distributions at the Golgi (Moelleken et al. 2007). These findings raise the specter that they may form distinct subpopulations of COPI vesicles. Moreover, another recent study has found that COPI vesicles could be distinguished based on their association with different tethering complexes and also in cargo content (Malsam et al. 2005). However, COPI vesicles reconstituted in this study did not include ARFGAP1 as a purified component. Thus, as GAP activity plays an important role in proper cargo sorting (Lanoix et al. 1999; Nickel et al. 1998; Pepperkok et al. 2000), whether subpopulations of COPI vesicles exist remains an interesting possibility that will require additional supporting evidence in the future.

Retrograde Golgi-to-ER transport Early studies on some chaperones that functioned in protein folding and assembly in the ER suggested that a fraction of these proteins leaked from the ER, and was then retrieved by retrograde transport from the Golgi (Munro and Pelham 1987). Subsequent studies revealed the molecular basis of this retrieval, which identified a receptor that recognized a common KDEL motif on the leaked ER proteins. Thus, the receptor was named the KDEL receptor (KDELR) (Lewis and Pelham 1990; Lewis et al. 1990; Semenza et al. 1990).

COPI mechanisms and roles

*

97

Subsequent studies on the KDELR revealed that it was retrieved from the Golgi to the ER through COPI transport (Girod et al. 1999; Lewis and Pelham 1996; White et al. 1999). Moreover, the KDELR had a more complex role than simply being a passive passenger in COPI vesicles, as it was also found to play a critical role in recruiting a GAP that acts on ARF1 to Golgi membrane in regulating COPI transport (Aoe et al. 1997, 1998). Altogether, these findings not only showed that the KDELR was a COPI cargo protein in retrograde transport from the Golgi to the ER, but also revealed that it was a special cargo protein that regulated its pathway of transport. Another cargo protein that has been well-established to be transported retrograde from the Golgi to the ER by COPI is ERGIC-53. It was originally identified as a marker of the intermediate compartment (Saraste et al. 1987; Schweizer et al. 1988), and hence the basis for its name as ER–Golgiintermediate compartment—53 kDa. Early studies on this cargo protein using temperature blocks revealed its cycling between the ER and the Golgi (Lippincott-Schwartz et al. 1990). Subsequently, the luminal domain of ERGIC-53 was revealed to act as a lectin that recognized a subset of glycosylated soluble luminal proteins, which was critical for these soluble proteins to be transported through the early secretory system to reach peripheral lysosome-like organelles (Appenzeller et al. 1999). Moreover, the physiologic relevance of this binding was revealed by mutations in ERGIC-53 causing human coagulopathies (Nichols et al. 1998). The cycling of ERGIC-53 between the ER and the Golgi has been shown to involve COPII recognizing a diphenylalanine motif in the cytoplasmic domain of ERGIC-53 for its anterograde transport (Kappeler et al. 1997), and COPI recognizing di-lysine motif for its retrograde transport (Kappeler et al. 1997; Tisdale et al. 1997). Like ERGIC-53, the p24 family of transmembrane proteins also cycle between the ER and the Golgi complex. Moreover, a phenylalanine-based motif in their cytoplasmic domain is recognized by COPII and a di-lysine-based motif is recognized by COPI (Dominguez et al. 1998). However, their precise cellular roles remain debated. The p24 family was originally identified based on their abundance in purified COPI vesicles (Sohn et al. 1996; Stamnes et al. 1995). As such, they were proposed to be critical for the binding of COPI to membrane for its role in forming transport vesicles. Studies on one member in particular, known as p23, supported this contention (Bremser et al. 1999; Reinhard et al. 1999). Remarkably however, when all p24 family members were deleted in the yeast, a surprisingly mild phenotype was observed (Springer et al. 2000). Thus, as COPI is essential for cell viability from yeast to mammals (Guo et al. 1994; Hosobuchi et al. 1992), an essential role for p24 family members in COPI vesicle formation has been questioned. Instead, other evidence suggests that the luminal domain of p24 members binds to different ligands to mediate their exit from the ER (Muniz et al. 2000). In this light, p24 is essentially acting as a receptor that facilitates the exit of select cargo proteins from the ER, which is fundamentally similar to the role currently attributed to ERGIC-53.

98

*

V. W. Hsu et al.

Studies on SNAREs (soluble N-ethylmaleimide-sensitive factor attachment protein receptor) have revealed yet another class of cargo proteins for which retrograde transport from the Golgi to the ER by COPI plays a critical role. SNAREs mediate the fusion of transport vesicles with their compartment of destination (Jahn and Scheller 2006). Studies on Ufe1p, a yeast syntaxin, revealed that its distribution at the ER depended on COPI (Lewis and Pelham, 1996). Moreover, yeasts with mutations in either Ufe1p or subunits of coatomer had defective anterograde transport between the ER and the Golgi, which was deduced to be the result of a more direct defect in retrograde COPI transport (Lewis and Pelham 1996).

Perspective Studies on COPI transport have made fundamental contributions to our understanding of how coat proteins are regulated by the ARF family of small GTPases during vesicle formation. However, methodologic issues have led to controversies regarding the precise function of COPI vesicles and how they are formed. Notably, recent mechanistic insights on COPI vesicle formation have revealed that the GAP which deactivates ARF1 has surprisingly complex roles. Rather than simply acting as an upstream regulator of the GTPase cycle, the GAP also acts as a key downstream effector, by being a component of the COPI complex. Thus, it is now appreciated to be intimately involved in vesicle formation, rather than in vesicle uncoating as originally postulated (Fig. 3)

Figure 3. A revised model for the role of the GAP that acts on ARF1 in COPI transport. Rather than simply acting an upstream regulator of the GTPase cycle, ARFGAP1 also acts as a key downstream effector, by being a component of the COPI complex. A key implication of this new model is that COPI vesicle uncoating will involve factor(s) yet to be identified.

COPI mechanisms and roles

*

99

This shift in our mechanistic understanding of COPI vesicle formation has three key implications. First, it has allowed further mechanistic insights into COPI vesicle formation, including the identification of novel critical factors. Second, it predicts that, rather than the GAP that acts on ARF1, COPI vesicle uncoating will be more directly regulated by other factor(s). Third, it simplifies models to explain how COPI vesicle formation and cargo sorting are coupled. In this last case, an intriguing prospect is that further studies on mechanisms of cargo sorting may eventually result in a more precise understanding of transport roles by COPI, for which some are still currently debated. Acknowledgements. We thank Jian Li for helpful discussions. This work is funded by a grant from the NIH to VWH. We apologize to our colleagues for not having cited all their work related to COPI due to space constraint.

References Aiken A (1996) 14-3-3 and its possible role in co-ordinating multiple signalling pathways. Trends Cell Biol 6: 341–347 Aoe T, Cukierman E, Lee A, Cassel D, Peters PJ, Hsu VW (1997) The KDEL receptor, ERD2, regulates intracellular traffic by recruiting a GTPase-activating protein for ARF1. EMBO J 16: 7305–7316 Aoe T, Lee AJ, Van Donselaar E, Peters PJ, Hsu VW (1998) Modulation of intracellular transport by transported proteins: insight from regulation of COPI-mediated transport. Proc Natl Acad Sci USA 95: 1624–1629 Appenzeller C, Andersson H, Kappeler F, Hauri HP (1999) The lectin ERGIC-53 is a cargo transport receptor for glycoproteins. Nat Cell Biol 1: 330–334 Balch WE, Dunphy WG, Braell WA, Rothman JE (1984) Reconstitution of the transport of protein between successive compartments of the Golgi measured by the coupled incorporation of N-acetylglucosamine. Cell 39: 405–416 Barlowe C, Orci L, Yeung T, Hosobuchi M, Hamamoto S, Salama N, Rexach MF, Ravazzola M, Amherdt M, Schekman R (1994) COPII: a membrane coat formed by Sec proteins that drive vesicle budding from the endoplasmic reticulum. Cell 77: 895–907 Bigay J, Gounon P, Robineau S, Antonny B (2003) Lipid packing sensed by ArfGAP1 couples COPI coat disassembly to membrane bilayer curvature. Nature 426: 563–566 Bonfanti L, Mironov AA Jr, Martinez-Menarguez JA, Martella O, Fusella A, Baldassarre M, Buccione R, Geuze HJ, Mironov AA, Luini A (1998) Procollagen traverses the Golgi stack without leaving the lumen of cisternae: evidence for cisternal maturation. Cell 95: 993–1003 Bonifacino JS, Lippincott-Schwartz J (2003) Coat proteins: shaping membrane transport. Nat Rev Mol Cell Biol 4: 409–414 Bremser M, Nickel W, Schweikert M, Ravazzola M, Amherdt M, Hughes CA, Sollner TH, Rothman JE, Wieland FT (1999) Coupling of coat assembly and vesicle budding to packaging of putative cargo receptors. Cell 96: 495–506 Casanova JE (2007) Regulation of arf activation: the Sec7 family of guanine nucleotide exchange factors. Traffic 8(11): 1476–1485 Chardin P, Paris S, Antonny B, Robineau S, Beraud-Dufour S, Jackson CL, Chabre M (1996) A human exchange factor for ARF contains Sec7- and pleckstrin-homology domains. Nature 384: 481–484 Chinnadurai G (2002) CtBP, an unconventional transcriptional corepressor in development and oncogenesis. Mol Cell 9: 213–224

100

*

V. W. Hsu et al.

Cosson P, Amherdt M, Rothman JE, Orci L (2002) A resident Golgi protein is excluded from peri-Golgi vesicles in NRK cells. Proc Natl Acad Sci USA 99: 12831–12834 Cosson P, Demolliere C, Hennecke S, Duden R, Letourneur F (1996) Delta- and zeta-COP, two coatomer subunits homologous to clathrin-associated proteins, are involved in ER retrieval. EMBO J 15: 1792–1798 Cosson P, Letourneur F (1994) Coatomer interaction with di-lysine endoplasmic reticulum retention motifs. Science 263: 1629–1631 Cukierman E, Huber I, Rotman M, Cassel D (1995) The ARF1-GTPase-Activating Protein: zinc finger motif and Golgi complex localization. Science 270: 1999–2002 D’Souza-Schorey C, Chavrier P (2006) ARF proteins: roles in membrane traffic and beyond. Nat Rev Mol Cell Biol 7: 347–358 Di Girolamo M, Silletta MG, De Matteis MA, Braca A, Colanzi A, Pawlak D, Rasenick MM, Luini A, Corda D (1995) Evidence that the 50-kDa substrate of brefeldin A-dependent ADP-ribosylation binds GTP and is modulated by the G-protein beta gamma subunit complex. Proc Natl Acad Sci USA 92: 7065–7069 Dominguez M, Dejgaard K, Fullekrug J, Dahan S, Fazel A, Paccaud JP, Thomas DY, Bergeron JJ, Nilsson T (1998) gp25L/emp24/p24 protein family members of the cis-Golgi network bind both COP I and II coatomer. J Cell Biol 140: 751–765 Donaldson JG, Cassel D, Kahn RA, Klausner RD (1992a) ADP-ribosylation factor, a small GTP-binding protein, is required for binding of the coatomer protein beta-COP to Golgi membranes. Proc Natl Acad Sci USA 89: 6408–6412 Donaldson JG, Finazzi D, Klausner RD (1992b) Brefeldin A inhibits Golgi membranecatalysed exchange of guanine nucleotide onto ARF protein. Nature 360: 350–352 Donaldson JG, Lippincott-Schwartz J, Klausner RD (1991) Guanine nucleotides modulate the effects of brefeldin A in semipermeable cells: regulation of the association of a 110-kDa peripheral membrane protein with the Golgi apparatus. Journal of Cell Biology 112: 579–588 Duden R, Griffiths G, Frank R, Argos P, Kreis TE (1991) Beta-COP, a 110 kDa protein associated with non-clathrin-coated vesicles and the Golgi complex, shows homology to beta-adaptin. Cell 64: 649–665 Eugster A, Frigerio G, Dale M, Duden R (2000) COP I domains required for coatomer integrity, and novel interactions with ARF and ARF–GAP. EMBO J 19: 3905–3917 Farsad K, Ringstad N, Takei K, Floyd SR, Rose K, De Camilli P (2001) Generation of high curvature membranes mediated by direct endophilin bilayer interactions. J Cell Biol 155: 193–200 Franco M, Chardin P, Chabre M, Paris S (1996) Myristoylation-facilitated binding of the G protein ARF1-GDP to membrane phospholipids is required for its activation by a soluble nucleotide exchange factor. J Biol Chem 271: 1573–1578 Frank S, Upender S, Hansen SH, Casanova JE (1998a) ARNO is a guanine nucleotide exchange factor for ADP-ribosylation factor 6. J Biol Chem 273: 23–27 Frank SR, Hatfield JC, Casanova JE (1998b) Remodeling of the actin cytoskeleton is coordinately regulated by protein kinase C and the ADP-ribosylation factor nucleotide exchange factor ARNO. Mol Biol Cell 9: 3133-3146 Gallop JL, Butler PJ, McMahon HT (2005) Endophilin and CtBP/BARS are not acyl transferases in endocytosis or Golgi fission. Nature 438: 675–678 Garcia-Mata R, Szul T, Alvarez C, Sztul E (2003) ADP-ribosylation factor/COPI-dependent events at the endoplasmic reticulum–Golgi interface are regulated by the guanine nucleotide exchange factor GBF1. Mol Biol Cell 14: 2250–2261 Gaynor EC, Emr SD (1997) COPI-independent anterograde transport: cargo-selective ER to Golgi protein transport in yeast COPI mutants. J Cell Biol 136: 789–802 Girod A, Storrie B, Simpson JC, Johannes L, Goud B, Roberts LM, Lord JM, Nilsson T, Pepperkok R (1999) Evidence for a COP-I-independent transport route from the Golgi complex to the endoplasmic reticulum. Nat Cell Biol 1: 423–430

COPI mechanisms and roles

*

101

Goldberg J (2000) Decoding of sorting signals by coatomer through a GTPase switch in the COPI coat complex. Cell 100: 671–679 Guo Q, Vasile E, Krieger M (1994) Disruptions in Golgi structure and membrane traffic in a conditional lethal mammalian cell mutant are corrected by epsilon-COP. J Cell Biol 125: 1213–1224 Harrison-Lavoie KJ, Lewis VA, Hynes GM, Collison KS, Nutland E, Willison KR (1993) A 102 kDa subunit of a Golgi-associated particle has homology to beta subunits of trimeric G proteins. EMBO J 12: 2847–2853 Harter C, Pavel J, Coccia F, Draken E, Wegehingel S, Tschochner H, Wieland F (1996) Nonclathrin coat protein gamma, a subunit of coatomer, binds to the cytoplasmic dilysine motif of membrane proteins of the early secretory pathway. Proc Natl Acad Sci USA 93: 1902–1906 Helms JB, Rothman JE (1992) Inhibition by brefeldin A of a Golgi membrane enzyme that catalyses exchange of guanine nucleotide bound to ARF. Nature 360: 352–354 Hildebrand JD, Soriano P (2002) Overlapping and unique roles for C-terminal binding protein 1 (CtBP1) and CtBP2 during mouse development. Mol Cell Biol 22: 5296–5307 Hoffman GR, Rahl PB, Collins RN, Cerione RA (2003) Conserved structural motifs in intracellular trafficking pathways: structure of the gammaCOP appendage domain. Mol Cell 12: 615–625 Hosobuchi M, Kreis T, Schekman R (1992) SEC21 is a gene required for ER to Golgi protein transport that encodes a subunit of a yeast coatomer. Nature 360: 603–605 Inoue H, Randazzo PA (2007) Arf GAPs and their interacting proteins. Traffic 8: 1465–1475 Jahn R, Scheller RH (2006) SNAREs—engines for membrane fusion. Nat Rev Mol Cell Biol 7: 631–643 Kahn RA, Gilman AG (1984) Purification of a protein cofactor required for ADPribosylation of the stimulatory regulatory component of adenylate cyclase by cholera toxin. J Biol Chem 259: 6228–6234 Kappeler F, Klopfenstein DR, Foguet M, Paccaud JP, Hauri HP (1997) The recycling of ERGIC-53 in the early secretory pathway. ERGIC-53 carries a cytosolic endoplasmic reticulum-exit determinant interacting with COPII. J Biol Chem 272: 31801–31808 Kawamoto K, Yoshida Y, Tamaki H, Torii S, Shinotsuka C, Yamashina S, Nakayama K (2002) GBF1, a guanine nucleotide exchange factor for ADP-ribosylation factors, is localized to the cis-Golgi and involved in membrane association of the COPI Coat. Traffic 3: 483–495 Kweon HS, Beznoussenko GV, Micaroni M, Polishchuk RS, Trucco A, Martella O, Di Giandomenico D, Marra P, Fusella A, Di Pentima A, et al (2004) Golgi enzymes are enriched in perforated zones of Golgi cisternae but are depleted in COPI vesicles. Mol Biol Cell 15: 4710–4724 Lanoix J, Ouwendijk J, Lin CC, Stark A, Love HD, Ostermann J, Nilsson T (1999) GTP hydrolysis by arf-1 mediates sorting and concentration of Golgi resident enzymes into functional COP I vesicles. EMBO J 18: 4935–4948 Lanoix J, Ouwendijk J, Stark A, Szafer E, Cassel D, Dejgaard K, Weiss M, Nilsson T (2001) Sorting of Golgi resident proteins into different subpopulations of COPI vesicles: a role for ArfGAP1. J Cell Biol 155: 1199–1212 Lee SY, Yang JS, Hong W, Premont RT, Hsu VW (2005) ARFGAP1 plays a central role in coupling COPI cargo sorting with vesicle formation. J Cell Biol 168: 281–290 Letourneur F, Gaynor EC, Hennecke S, Demolliere C, Duden R, Emr SD, Riezman H, Cosson P (1994) Coatomer is essential for retrieval of dilysine-tagged proteins to the endoplasmic reticulum. Cell 79: 1199–1207 Lewis MJ, Pelham HR (1990) A human homologue of the yeast HDEL receptor. Nature 348: 162–163

102

*

V. W. Hsu et al.

Lewis MJ, Pelham HR (1996) SNARE-mediated retrograde traffic from the Golgi complex to the endoplasmic reticulum. Cell 85: 205–215 Lewis MJ, Sweet DJ, Pelham HR (1990) The ERD2 gene determines the specificity of the luminal ER protein retention system. Cell 61: 1359–1363 Lippincott-Schwartz J, Donaldson JG, Schweizer A, Berger EG, Hauri HP, Yuan LC, Klausner RD (1990) Microtubule-dependent retrograde transport of proteins into the ER in the presence of brefeldin A suggests an ER recycling pathway. Cell 60: 821–836 Lippincott-Schwartz J, Yuan LC, Bonifacino JS, Klausner RD (1989) Rapid redistribution of Golgi proteins into the ER in cells treated with brefeldin A: evidence for membrane cycling from Golgi to ER. Cell 56: 801–813 Love HD, Lin CC, Short CS, Ostermann J (1998) Isolation of functional Golgi-derived vesicles with a possible role in retrograde transport. J Cell Biol 140: 541–551 Lowe M, Kreis TE (1995) In vitro assembly and disassembly of coatomer. J Biol Chem 270: 31364–31371 Malsam J, Satoh A, Pelletier L, Warren G (2005) Golgin tethers define subpopulations of COPI vesicles. Science 307: 1095–1098 Martinez-Menarguez JA, Prekeris R, Oorschot VM, Scheller R, Slot JW, Geuze HJ, Klumperman J (2001) Peri-Golgi vesicles contain retrograde but not anterograde proteins consistent with the cisternal progression model of intra-Golgi transport. J Cell Biol 155: 1213–1224 McMahon HT, Mills IG (2004) COP and clathrin-coated vesicle budding: different pathways, common approaches. Curr Opin Cell Biol 16: 379–391 Michelsen K, Schmid V, Metz J, Heusser K, Liebel U, Schwede T, Spang A, Schwappach B (2007) Novel cargo-binding site in the {beta} and {delta} subunits of coatomer. J Cell Biol 179: 209–217 Mironov AA, Beznoussenko GV, Nicoziani P, Martella O, Trucco A, Kweon HS, Di Giandomenico D, Polishchuk RS, Fusella A, Lupetti P, et al (2001) Small cargo proteins and large aggregates can traverse the Golgi by a common mechanism without leaving the lumen of cisternae. J Cell Biol 155: 1225–1238 Moelleken J, Malsam J, Betts MJ, Movafeghi A, Reckmann I, Meissner I, Hellwig A, Russell RB, Sollner T, Brugger B, Wieland FT (2007) Differential localization of coatomer complex isoforms within the Golgi apparatus. Proc Natl Acad Sci USA 104: 4425–4430 Muniz M, Nuoffer C, Hauri HP, Riezman H (2000) The Emp24 complex recruits a specific cargo molecule into endoplasmic reticulum-derived vesicles. J Cell Biol 148: 925–930 Munro S, Pelham HR (1987) A C-terminal signal prevents secretion of luminal ER proteins. Cell 48: 899–907 Nichols WC, Seligsohn U, Zivelin A, Terry VH, Hertel CE, Wheatley MA, Moussalli MJ, Hauri HP, Ciavarella N, Kaufman RJ, Ginsburg D (1998) Mutations in the ER-Golgi intermediate compartment protein ERGIC-53 cause combined deficiency of coagulation factors V and VIII. Cell 93: 61–70 Nickel W, Malsam J, Gorgas K, Ravazzola M, Jenne N, Helms JB, Wieland FT (1998) Uptake by COPI-coated vesicles of both anterograde and retrograde cargo is inhibited by GTPgammaS in vitro. J Cell Sci 111: 3081–3090 O’Kelly I, Butler MH, Zilberberg N, Goldstein SA (2002) Forward transport. 14-3-3 binding overcomes retention in endoplasmic reticulum by dibasic signals. Cell 111: 577–588 Orci L, Amherdt M, Ravazzola M, Perrelet A, Rothman JE (2000a) Exclusion of Golgi residents from transport vesicles budding from Golgi cisternae in intact cells. J Cell Biol 150: 1263–1270 Orci L, Glick BS, Rothman JE (1986) A new type of coated vesicular carrier that appears not to contain clathrin: its possible role in protein transport within the Golgi stack. Cell 46: 171–184

COPI mechanisms and roles

*

103

Orci L, Palmer DJ, Amherdt M, Rothman JE (1993) Coated vesicle assembly in the Golgi requires only coatomer ARF proteins from the cytosol. Nature 364: 732–734 Orci L, Ravazzola M, Volchuk A, Engel T, Gmachl M, Amherdt M, Perrelet A, Sollner TH, Rothman JE (2000b) Anterograde flow of cargo across the Golgi stack potentially mediated via bidirectional “percolating” COPI vesicles. Proc Natl Acad Sci USA 97: 10400–10405 Orci L, Stamnes M, Ravazzola M, Amherdt M, Perrelet A, Sollner TH, Rothman JE (1997) Bidirectional transport by distinct populations of COPI-coated vesicles. Cell 90: 335–349 Orci L, Tagaya M, Amherdt M, Perrelet A, Donaldson JG, Lippincott-Schwartz J, Klausner RD, Rothman JE (1991) Brefeldin A, a drug that blocks secretion, prevents the assembly of non-clathrin-coated buds on Golgi cisternae. Cell 64: 1183–1195 Ostermann J, Orci L, Tani K, Amherdt M, Ravazzola M, Elazar Z, Rothman JE (1993) Stepwise assembly of functionally active transport vesicles. Cell 75: 1015–1025 Pavel J, Harter C, Wieland FT (1998) Reversible dissociation of coatomer: functional characterization of a beta/delta-coat protein subcomplex. Proc Natl Acad Sci USA 95: 2140–2145 Pepperkok R, Whitney JA, Gomez M, Kreis TE (2000) COPI vesicles accumulating in the presence of a GTP restricted arf1 mutant are depleted of anterograde and retrograde cargo. J Cell Sci 113 (Pt 1): 135–144 Peyroche A, Courbeyrette R, Rambourg A, Jackson CL (2001) The ARF exchange factors Gea1p and Gea2p regulate Golgi structure and function in yeast. J Cell Sci 114: 2241–2253 Peyroche A, Paris S, Jackson CL (1996) Nucleotide exchange on ARF mediated by yeast Gea1 protein. Nature 384: 479–481 Rabouille C, Klumperman J (2005) Opinion: the maturing role of COPI vesicles in intraGolgi transport. Nat Rev Mol Cell Biol 6: 812–817 Randazzo PA, Terui T, Sturch S, Fales HM, Ferrige AG, Kahn RA (1995) The myristoylated amino terminus of ADP-ribosylation factor 1 is a phospholipid- and GTP-sensitive switch. J Biol Chem 270: 14809–14815 Reinhard C, Harter C, Bremser M, Brugger B, Sohn K, Helms JB, Wieland F (1999) Receptor-induced polymerization of coatomer. Proc Natl Acad Sci USA 96: 1224–1228 Reinhard C, Schweikert M, Wieland FT, Nickel W (2003) Functional reconstitution of COPI coat assembly and disassembly using chemically defined components. Proc Natl Acad Sci USA 100: 8253–8257 Ringstad N, Gad H, Low P, Di Paolo G, Brodin L, Shupliakov O, De Camilli P (1999) Endophilin/SH3p4 is required for the transition from early to late stages in clathrinmediated synaptic vesicle endocytosis. Neuron 24: 143–154 Santy LC, Casanova JE (2001) Activation of ARF6 by ARNO stimulates epithelial cell migration through downstream activation of both Rac1 and phospholipase D. J Cell Biol 154: 599–610 Saraste J, Palade GE, Farquhar MG (1987) Antibodies to rat pancreas Golgi subfractions: identification of a 58-kD cis-Golgi protein. J Cell Biol 105: 2021–2029 Schmidt A, Wolde M, Thiele C, Fest W, Kratzin H, Podtelejnikov AV, Witke W, Huttner WB, Soling HD (1999) Endophilin I mediates synaptic vesicle formation by transfer of arachidonate to lysophosphatidic acid. Nature 401: 133–141 Schweizer A, Fransen JA, Bachi T, Ginsel L, Hauri HP (1988) Identification, by a monoclonal antibody, of a 53-kD protein associated with a tubulo-vesicular compartment at the cis-side of the Golgi apparatus. J Cell Biol 107: 1643–1653. Semenza JC, Hardwick KG, Dean N, Pelham HR (1990) ERD2, a yeast gene required for the receptor-mediated retrieval of luminal ER proteins from the secretory pathway. Cell 61: 1349–1357.

104

*

V. W. Hsu et al.

Serafini T, Orci L, Amherdt M, Brunner M, Kahn RA, Rothman JE (1991a) ADP-ribosylation factor is a subunit of the coat of Golgi-derived COP-coated vesicles: a novel role for a GTP-binding protein. Cell 67: 239–253 Serafini T, Stenbeck G, Brecht A, Lottspeich F, Orci L, Rothman JE, Wieland FT (1991b) A coat subunit of Golgi-derived non-clathrin-coated vesicles with homology to the clathrin-coated vesicle coat protein beta-adaptin. Nature 349: 215–220 Simpson F, Hussain NK, Qualmann B, Kelly RB, Kay BK, McPherson PS, Schmid SL (1999) SH3-domain-containing proteins function at distinct steps in clathrin-coated vesicle formation. Nat Cell Biol 1: 119–124 Sohn K, Orci L, Ravazzola M, Amherdt M, Bremser M, Lottspeich F, Fiedler K, Helms JB, Wieland FT (1996) A major transmembrane protein of Golgi-derived COPI-coated vesicles involved in coatomer binding. J Cell Biol 135: 1239–1248 Spang A, Matsuoka K, Hamamoto S, Schekman R, Orci L (1998) Coatomer, Arf1p, and nucleotide are required to bud coat protein complex I-coated vesicles from large synthetic liposomes. Proc Natl Acad Sci USA 95: 11199-11204 Spano S, Silletta MG, Colanzi A, Alberti S, Fiucci G, Valente C, Fusella A, Salmona M, Mironov A, Luini A, Corda D (1999) Molecular cloning and functional characterization of brefeldin A-ADP-ribosylated substrate. A novel protein involved in the maintenance of the Golgi structure. J Biol Chem 274: 17705–17710 Springer S, Chen E, Duden R, Marzioch M, Rowley A, Hamamoto S, Merchant S, Schekman R (2000) The p24 proteins are not essential for vesicular transport in Saccharomyces cerevisiae. Proc Natl Acad Sci USA 97: 4034–4039 Springer S, Spang A, Schekman R (1999) A primer on vesicle budding. Cell 97: 145–148 Stamnes MA, Craighead MW, Hoe MH, Lampen N, Geromanos S, Tempst P, Rothman JE (1995) An integral membrane component of coatomer-coated transport vesicles defines a family of proteins involved in budding [published erratum appears in Proc Natl Acad Sci USA 1995 Nov 7; 92(23): 10816]. Proc Natl Acad Sci USA 92: 8011–8015 Stenbeck G, Harter C, Brecht A, Herrmann D, Lottspeich F, Oric L, Wieland FT (1993) Beta prime COP, a novel subunit of coatomer. EMBO J 12: 2841–2845 Tanigawa G, Orci L, Amherdt M, Ravazzola M, Helms JB, Rothman JE (1993) Hydrolysis of bound GTP by ARF protein triggers uncoating of Golgi-derived COP-coated vesicles. J Cell Biol 123: 1365–1371 Teasdale RD, Jackson MR (1996) Signal-mediated sorting of membrane proteins between the endoplasmic reticulum and the Golgi apparatus. Annu Rev Cell Dev Biol 12: 27–54 Tisdale EJ, Plutner H, Matteson J, Balch WE (1997) p53/58 binds COPI and is required for selective transport through the early secretory pathway. J Cell Biol 137: 581–593 Trucco A, Polishchuk RS, Martella O, Di Pentima A, Fusella A, Di Giandomenico D, San Pietro E, Beznoussenko GV, Polishchuk EV, Baldassarre M, et al. (2004) Secretory traffic triggers the formation of tubular continuities across Golgi sub-compartments. Nat Cell Biol 6: 1071–1081 Waters MG, Serafini T, Rothman JE (1991) ‘Coatomer’: a cytosolic protein complex containing subunits of non-clathrin-coated Golgi transport vesicles. Nature 349: 248–251 Wegmann D, Hess P, Baier C, Wieland FT, Reinhard C (2004) Novel isotypic gamma/ zeta subunits reveal three coatomer complexes in mammals. Mol Cell Biol 24: 1070–1080 Weigert R, Silletta MG, Spano S, Turacchio G, Cericola C, Colanzi A, Senatore S, Mancini R, Polishchuk EV, Salmona M, et al. (1999) CtBP/BARS induces fission of Golgi membranes by acylating lysophosphatidic acid. Nature 402: 429–433 Wessels E, Duijsings D, Niu TK, Neumann S, Oorschot VM, De Lange F, Lanke KH, Klumperman J, Henke A, Jackson CL, et al. (2006) A viral protein that blocks Arf1-

COPI mechanisms and roles

*

105

mediated COP-I assembly by inhibiting the guanine nucleotide exchange factor GBF1. Dev Cell 11: 191–201 White J, Johannes L, Mallard F, Girod A, Grill S, Reinsch S, Keller P, Tzschaschel B, Echard A, Goud B, Stelzer EH (1999) Rab6 coordinates a novel Golgi to ER retrograde transport pathway in live cells. J Cell Biol 147: 743–760 Yang J, Lee S, Spanò S, Gad H, Zhang L, Nie Z, Bonazzi M, Corda D, Luini A, Hsu V (2005) A role for BARS at the fission step of COPI vesicle formation from Golgi membrane. EMBO J 24: 4133–4143 Yang JS, Lee SY, Gao M, Bourgoin S, Randazzo PA, Premont RT, Hsu VW (2002) ARFGAP1 promotes the formation of COPI vesicles, suggesting function as a component of the coat. J Cell Biol 159: 69–78 Yang JS, Zhang L, Lee SY, Gad H, Luini A, Hsu VW (2006) Key components of the fission machinery are interchangeable. Nat Cell Biol 8: 1376–1382 Yuan H, Michelsen K, Schwappach B (2003) 14-3-3 dimers probe the assembly status of multimeric membrane proteins. Curr Biol 13: 638–646 Zerangue N, Malan MJ, Fried SR, Dazin PF, Jan YN, Jan LY, Schwappach B (2001) Analysis of endoplasmic reticulum trafficking signals by combinatorial screening in mammalian cells. Proc Natl Acad Sci USA 98: 2431–2436

106

*

R. A. Kahn

Arfs and Arls: models for Arf family members in membrane traffic at the Golgi Richard A. Kahn

Introduction The ADP-ribosylation factor (Arf) family GTPases are highly conserved 20 kDa GTP-binding proteins that play a number of roles in the regulation of cellular physiology, most relevant here is that they act to recruit soluble proteins to a membrane surface and coordinate the assembly of multi-protein complexes that are required for the biogenesis of nascent carriers of membrane traffic. The Arfs can also recruit and directly activate lipid-modifying enzymes, providing important functional links between localized changes in lipid composition and protein assemblies. The Arf family GTPases and their interactions have been the subject of a recent book (Kahn 2004) and reviews (Gillingham and Munro 2007; Inoue and Randazzo 2007). Molecular aspects of Arf family members acting at the Golgi and models for their actions are summarized in this chapter.

The Arf family of regulatory GTPases and their actions at the Golgi The Arf family includes both the Arf sub-family, with six members in mammals (Arf1–6) that share >60% primary sequence identity, and the more divergent Arf-like (Arl; typically 40–60% identity to Arfs or each other), and Sar (20–30% identical to Arfs) sub-families (Kahn et al. 2006; Li et al. 2004; Logsdon and Kahn 2004). This is an ancient gene family with already six members in the earliest eukaryotes, prior to the emergence of Ras or heterotrimeric G proteins, and likely arising in prokaryotes (Dong et al. 2007). Members of each of these sub-families play important roles in membrane traffic, while several Arls have distinct roles in the regulation of microtubule dynamics (Arl2 and Arl8 (Antoshechkin and Han 2002; Bhamidipati et al. 2000; Hoyt et al. 1990; McElver et al. 2000; Okai et al. 2004; Zhou et al. 2006)), ciliogenesis (Arl3, Arl6, and Arl13b (Caspary et al. 2007; Chiang et al. 2004; Fan et al. 2004; Sahin et al. 2004; Schrick et al. 2006; Zhou et al. 2006)), and cytokinesis (Arl3, Arf1, and Arf6 (Altan-Bonnet et al. 2003; Schweitzer and D’Souza-Schorey 2005; Zhou et al. 2006)). Sar1 (two genes/proteins in mammals) is the most divergent member of the Arf family yet acts to regulate carrier biogenesis at ER exit sites in a fashion that is very analogous to the actions of other family members at the Golgi (Barlowe et al. 1994; Kuge et al. 1994; Lee et al. 2005; Nakano and

Arfs and Arls: models for Arf family members at the Golgi

*

107

Muramatsu 1989; Oka et al. 1991). In many ways the mechanism of action of Sar1 in regulating exit of cargo from the ER and recruitment of the COPII complex is a paradigm for the actions of Arfs at the Golgi, though the latter have additional complexities as a result of the increased (a) diversity in highly homologous proteins, (b) number of coat complexes involved, and (c) role of lipid/membrane interactions, as summarized above. Ten of the thirty members of the mammalian Arf family are implicated as regulators of membrane traffic at the Golgi; Arf1–5, Arl1, and Arl5A–C, and Arfrp1. Arf6 is active predominantly in endocytosis at the plasma membrane and also impacts the actin cytoskeleton so is not included in this list. Our current models for Arf and Arl actions at the Golgi come from studies of Arf1–5, (though overwhelmingly only Arf1) and Arl1 so will form the focus of the rest of this chapter. Arf2 has been lost in humans and no specific functional studies have been reported for this isoform so will not be discussed further. Arl5s (three genes/proteins in mammals) localize to the Golgi and either knockdowns by siRNA or expression of dominant mutants have only quite subtle effects on Golgi morphology (Yawei Li and R. A. Kahn; unpublished observations), but no further information is available on their actions. Arfrp1 is discussed with Arl1 as they are genetically and functionally linked. With 10 proteins all active at the Golgi, where is the specificity in Arfregulated membrane traffic found? With clear differences in genetic and physical interactions it was relatively easy to distinguish functions of Arfs from those of Arl1 or Arfrp1, but this has not been true for Arf1–5. The high degree of conservation of primary sequence of the Arfs throughout eukaryotic evolution (human and S. cerevisiae Arf1 are 74% identical) is accompanied by functional conservation. For example, each of the five human Arfs (Arf1, Arf3–6) or the single Arf gene from Giardia lamblia can complement the lethality resulting from deletion of the two yeast Arf genes, ARF1 and ARF2 (Kahn et al. 1991; Lee et al. 1992), but neither the S. cerevisiae or other ARL1 genes can complement the deletion of the ARFs. In addition, in several in vitro assays of Arf activities the specific activities of the different human Arf proteins are indistinguishable. We still lack reagents that can distinguish between the Arf proteins in cells because the isoform specific antisera generated to date do not work for immunofluorescence and epitope tagging is fraught with artifacts resulting from (often poorly characterized) alterations in protein localization and affinities for protein- and lipid-binding partners. Thus, truly specific roles for any Arf1–5 have not been described, while the actions of Arfs and Arl1 are clearly distinct and will be discussed separately. In stark contrast to our failure to generate useful antibodies for immunocytochemistry and the difficulties in interpretation from the use of dominant mutants of such a highly conserved protein family, the use of siRNA to specifically deplete cells of each Arf, alone or in combination, has provided an opportunity to address the question of specificities within the Arf proteins. This approach has led to the identification of a number of distinctive roles for

108

*

R. A. Kahn

Arf1–5 in ER–Golgi traffic, exit from the Golgi, and in endosome recycling (Volpicelli-Daley et al. 2005). Surprisingly, the most profound phenotypes were only observed with different combinations of depletion of two Arfs. While these data may be interpreted in different ways, one intriguing model to arise from them is that Arfs act in pairs to regulate carrier biogenesis at a donor membrane. Indeed, a growing number of the proteins involved in membrane traffic are being found to dimerize on the membrane surface and it is predicted that this will be a central theme in molecular models in the future (e.g., see Burguete et al. (2008)).

Arfs bind lipids and lipid-modifying enzymes Arfs act to control aspects of carrier biogenesis at the Golgi, trans-Golgi network (TGN), ER–Golgi intermediate compartment (ERGIC), and likely endosomes. As regulatory GTPases, Arfs cycle between GDP- and GTP-bound conformations in a cycle at rates that are determined by binding to guanine nucleotide exchange factors (GEFs), GTPase activating proteins (GAPs), and effectors. In addition to these functionally critical protein interactions, Arfs are unusual among regulatory GTPases in that their affinity for biological membranes is directly impacted by the nucleotide bound. Specifically, Arfs are N-myristoylated and this hydrophobic anchor functions as a “myristoyl switch” to dock and orient the protein on the surface of the membrane when the Arf is GTP-bound, but the protein binds the myristate and shields it from aqueous interactions to promote its solubility when “inactive”, i.e., when GDP is bound (Franco et al. 1996; Kahn et al. 1988, 1992; Ames et al. 1997; Goldberg 1998). Thus, activation of Arf (i.e., GTP-binding) promotes a more stable membrane attachment as well as increasing its affinity for effectors. This myristoylation-dependent membrane translocation is predominantly nonspecific as far as lipid interactions. In contrast, some specific phospholipids are capable of altering the guanine nucleotide-binding properties of Arfs and some Arf GAPs (Brown et al. 1998; Kam et al. 2000; Randazzo 1997; Randazzo and Kahn 1994; Terui et al. 1994; Zheng et al. 1996). In a related fashion, some Arf GAPs appear capable of sensing membrane curvature and become more active as the curvature increases, providing a link between carrier maturation and Arf inactivation (Bigay et al. 2005, 2003). Together these data highlight the intimate interplay between Arfs and the membranes at which they act to regulate membrane traffic. In addition to Arfs and Arf GAPs being capable of sensing localized changes in the lipid environment, Arfs also have roles in both recruitment to a membrane and direct activation of a number of lipid-modifying enzyme. Activated Arfs increase the recruitment to membranes of PI 4-kinase and PI(4P) 5-kinase (De Matteis et al. 2005; Godi et al. 1999; Honda et al. 1999; Jones et al. 2000) and increase their activity. Arfs are also potent and direct activators of phospholipase D (Brown et al. 1993; Cockcroft et al. 1994; Ktistakis et al. 1995, 1996). The intimate co-dependence of lipid

Arfs and Arls: models for Arf family members at the Golgi

*

109

modifications and protein recruitment to Arf action and membrane traffic at the Golgi is evident from studies of the FAPP and related proteins, that have both Arf-binding and PI4P-binding domains, each of which are required for their action in export from the TGN (Godi et al. 2004; Vieira et al. 2005). Molecular details of how the lipid-modifying activities of Arfs and adaptor recruitment are integrated in the cell and the respective contributions of each activity at different sites remain controversial. But there is no disagreement among researchers that our understanding of molecular membrane traffic will require detailed understanding of the localized changes in the lipid bilayer, membrane curvature, protein recruitment, and enzyme activities; thus assuring that Arfs are key players in this complex and essential aspect of cell biology.

Activated Arfs recruit protein adaptors to nascent carriers of membrane traffic – models for Arf action at the Golgi/TGN Three observations from the late 1980’s and early 1990’s were central to the current models of the actions of Arfs at the Golgi. First was that the Arfs are GTPases that act at the Golgi to regulate aspects of membrane traffic (Kahn and Gilman 1986; Kahn et al. 1992; Stearns et al. 1990). The GTPase field that the Arfs arose from was defined by the plasma membrane localized, heterotrimeric G proteins, and thus the models involved initiation of GTPase signaling in response to agonist binding to a specific receptor, that has GEF activity. No such agonist has been defined for Arfs and thus the initiator of Arf signaling remains uncertain. While Arfs clearly operate through the canonical GTP/GDP cycle for regulatory GTPases, it is not necessarily true that agoniststimulated G protein signaling is the best molecular model for Arfs as regulators of membrane traffic. Second was the observation that Arfs were required to recruit the COPI coat to Golgi membranes in an in vitro assay of intra-Golgi transport (Serafini et al. 1991). Although several controversies grew out of this early work, some of which remain unresolved, it has become clear that a central role for Arfs as regulators of membrane traffic is the recruitment to membranes of protein adaptors or adaptor complexes. This is true not only for the heptameric COPI complex, but also the tetrameric adaptin complexes (AP-1 and AP-4 at the TGN (Dell’Angelica et al. 1999; Hirst et al. 1999; Stamnes and Rothman 1993; Traub et al. 1993), AP-2 at the plasma membrane and endosomes (Krauss et al. 2003; Paleotti et al. 2005; West et al. 1997), and AP-3 at endosomes (Drake et al. 2000; Faundez et al. 1998; Faundez and Kelly 2000; Ooi et al. 1998)), and monomeric GGA (Boman et al. 2000; Dell’Angelica et al. 2000; Hirst et al. 2001; Puertollano et al. 2001a, b; Takatsu et al. 2001) and Mint (Hill et al. 2003; Shrivastava-Ranjan et al. 2008) families (which each contain three different genes/proteins).

110

*

R. A. Kahn

The third landmark observation was that the macrolide antibiotic brefeldin A (BFA) was found to be a rapidly acting, membrane permeant, specific inhibitor of a subset of Arf GEFs that act at the Golgi/TGN (Donaldson et al. 1991, 1992; Fujiwara et al. 1988; Lippincott-Schwartz et al. 1989; Misumi et al. 1986; Robinson and Kreis 1992). Arf GEFs and their sensitivity to BFA have been the subject of an excellent recent review so will not be discussed in detail here (Casanova 2007). Use of BFA has allowed live cell imaging of Arfdependent processes and the emergence of one imporant criterion for an Arf-dependent adaptor at the Golgi – rapid release (<2 min typically) from membranes in response to the drug. For example, this feature has been confirmed for each of the Arf-dependent adaptors listed above. The importance of BFA to our understanding of Arf actions in membrane traffic and the BFA-sensitive Arf GEFs (GBF1, BIG1, and BIG2) is hard to overstate but its extensive use has also led to a severe underrepresentation of studies of the BFA-insensitive Arf GEFs. Together, these results support a model (see Fig. 1) in which activated Arf is the initiating event in the formation of nascent carriers from the Golgi/TGN. Activation of Arf promotes its stable assocation with membranes (step 1), which in turn supports recruitment of Arf-dependent adaptors (step 2). These adaptors also bind specific transmembrane protein cargos, through short sorting motifs in their cytoplasmic tails (Hill et al. 2003; Ohno et al. 1998; Puertollano et al. 2001a; Zhu et al. 2001) so that the presence of Arf and adaptors on the cytoplasmic surface of a membrane leads to selective recruitment of cargo (step 3) and repeated rounds of recruitment and stabilization of these and accessory proteins (step 4) that result in a protein coat on the nascent carrier and ultimately on the mature carrier (step 5). Such a model captures the essential role of Arfs in carrier biogenesis and is useful in that it summarizes a large number of studies and provides a number of testable predictions. This is similar to earlier models of cargo recruitment by adaptors (e.g., see Kirchhausen et al. (1997)) but includes the specific role for Arfs, non-clathrin coated carriers, and is more generalized to include the large number of Arf-dependent adaptors. An extension of this minimal three-component model has been proposed (Goldberg 2000) to describe a role for an ArfGAP in the process, specifically to provide a proofreading function. When cargo with an incorrect sorting signal is present, GAP activity is stimulated by the adaptor (e.g., COPI in the Goldberg studies); leading to GTP hydrolysis on ARF, release of COPI, and prevention of coat assembly and bud maturation (Fig. 1; step 2b). If the sorting motif on the cargo is appropriate to the adaptor, GTP hydrolysis is not stimulated, and vesicle coating proceeds (steps 2a, 3–5). Such a model is appealing in that it confers upon the Arf GTPases the role of kinetic proofreading that couples cargo selection with carrier maturation. Appealing as it is, this aspect of the model requires additional testing, particularly in cell-based assays and with different adaptors.

Arfs and Arls: models for Arf family members at the Golgi

*

111

Figure 1. Model for the actions of Arf GTPases to recruit adaptors to the site of carrier biogenesis (“budding”) and correctly pair adaptors with transmembrane cargoes. The translocation of Arf onto the cytosolic surface of a membrane is shown in step 1, with the insertion of the Nmyristate into the bilayer to assist in the docking process. Correct pairing of cargo with adaptors is shown in step 2a and 3, using an adaptin (e.g., AP-1) in this example. The subsequent recruitment of unspecified accessory proteins is depicted in step 4, with generation of the mature adaptor-coated and cargo-enriched carrier shown in step 5. Note that the Arf is lost during the carrier maturation process, in response to GAP activity that is stimulated by increasing curvature of the nascent bud. Step 2b shows that when a cargo and adaptor are incorrectly paired, there is activation of the GAP at an earlier stage, with consequent loss of Arf and adaptors from the membrane surface. Three different sorting motifs are shown on the cytosolic tail of the “correct” cargo, though each is found on different proteins where each acts autonomously.

How then is GTP hydrolysis and release of Arf from a nascent bud controlled when the appropriate Arf–adaptor–cargo complex has been generated? Results from the laboratory of Bruno Antonny have shown that the Arf GAP activity of ArfGAP1 is progressively increased with the curvature of the membrane on which it is bound (Bigay et al. 2003, 2005). This curvature sensor is found in a domain termed the ArfGAP1 lipid packing sensor (ALPS) and thus couples the increasing curvature of nascent carriers with the hydrolysis of GTP on the Arf, and subsequent dissocation of the GTPase from the bud. This model leads to the prediction that Arfs are absent from the mature carriers, which has been the case so far. Two different Arf-dependent carriers have been purified and characterized and in neither case were Arfs found as part of the mature coated carriers (Salazar et al. 2005; ShrivastavaRanjan et al. 2008).

112

*

R. A. Kahn

An important remaining question from the model shown in Fig. 1 is how is Arf activated? By analogy to classical G protein signaling, there should be a ligand-binding event that initiates the GTPase cycle. But no such ligand is evident and membrane traffic is a constitutive process that is shut off only during mitosis. This does not mean that Arf GEFs are not important. On the contrary, results from the use of BFA have convincingly demonstrated the required role for at least the BFA-sensitive Arf GEFs in adaptor recruitment and re-cycling. But it is much more likely that Arf activation is a stochastic process that involves localized changes in cargo concentration, phospholipid composition of the membrane microenvironment, and the concentrations of Arf effectors, Arf GEFs, and Arf GAPs. One strong indication that Arfs do not act solely in a linear, canonical, G protein fashion is that Arf effectors have been shown to be capable of increasing the activity of the Arf, through an increase in its affinity for GTP (Zhu et al. 2000). In a related fashion, a recent report out of the Kornfeld laboratory has revealed a role for the sorting sequences, derived from the cytoplasmic tails of cargo, in binding to adaptors (in this case AP-1) to increase their affinity for activated Arf and recruit the Arf to a liposome (Lee et al. 2008). Thus, the trigger to Arf signaling and the initiation of carrier budding is likely to lie not in a single ligand-binding event but in the aggregate of many factors including several that may be highly localized.

Arl1 recruits Golgins to the Golgi Arfs and Arl1 are each essential genes in worms (Li et al. 2004) and flies (Tamkun et al. 1991) and are predicted by this to each possess required and distinctive functions. Arl1 orthologs were localized in early studies to the Golgi (Lee et al. 1997; Lowe et al. 1996), and later the trans-Golgi, where they were implicated as a required component in Golgi traffic and morphology (Lu et al. 2001; Van Valkenburgh et al. 2001). Early biochemical and cell biological studies suggested close functional relatedness between Arf and Arl1 actions, as they are N-myristoylated, localize to the Golgi, alter Golgi morphology upon expression of a dominant active point mutant, and share an overlapping set of effectors and even an Arf GAP (Lee et al. 1997; Liu et al. 2005; Lu et al. 2001; Van Valkenburgh et al. 2001). Among the Arl1 effectors not shared by Arfs are proteins that share a common fold and domain structure, termed the GRIP domain; these include only Imh1p in S. cerevisiae but four proteins in humans, Golgin-97, Golgin-245, GCC88, and GCC185. These effectors have recently come to dominate models of Arl1 actions as a growing body of evidence supports their importance in endosome–Golgi traffic, and we still lack cellular data to support the functional significance of the specific binding of Arl1GTP to other effectors, e.g., Arfaptin2/POR1, MKLP1, RanBP2a, and SCOCO (Derby et al. 2004; Lu and Hong 2003; Lu et al. 2004; Panic et al. 2003a, b; Van Valkenburgh et al. 2001; Wu et al. 2004). Analogous to the recruitment of COPII by Sar1 and Arf-dependent adaptors by Arfs, the primary actions of Arl1 are currently thought to include the

Arfs and Arls: models for Arf family members at the Golgi

*

113

direct binding and recruitment to membranes of GRIP domain proteins that are required for endosome to Golgi traffic (Burd et al. 2004; Derby et al. 2007; Lu et al. 2004; Munro 2005). A recent structural study has shed new light into Arl1 actions as it revealed that Rab6 binds to a different region of GCC185 to increase its affinity for Arl1GTP and the two GTPases coordinately recruit this Golgin to assist in the docking of endosome derived carriers at the Golgi (Burguete et al. 2008). Both this study and an earlier one of the structure of Arl1 binding to the GRIP domain of Golgin-245 (Panic et al. 2003a) reveal that GRIP domains homodimerize and bind two Arl1s per homodimer (Wu et al. 2004). While direct binding of Arl1 to each GRIP domain protein is predicted and supported by yeast two-hybrid data (Lu and Hong 2003) they do so with different apparent affinities. It is likely that Arl1GTP binds Golgin-97 and Golgin-245 with sufficient avidity that other interactions, like Rab6 with GCC185, are not required for membrane recruitment but may play a supportive role. It is not yet clear whether GCC88 recruitment to membranes will involve a Rab or other cooperating regulator (Burguete et al. 2008). Additional insights into the actions of Arl1 at the Golgi came from two studies in yeast that each found that another member of the Arf family, Arfrp1 acts upstream and is required for the binding of Arl1 at the Golgi (Panic et al. 2003b; Setty et al. 2003). These studies were each performed in S. cerevisiae, in which the ortholog of Arfrp1 is named Arl3, so it is important to clarify that in mammals Arl1 acts downstream of Arfrp1 (Zahn et al. 2006) and not the mammalian Arl3, which has completely distinct actions in cells (Zhou et al. 2006). Thus, an ordered process of Arfrp1 activation leading to activation of Arl1, and consequently recruitment of Golgins is consistent with data from several labs and is conserved between yeast and mammals. Arfrp1 is an essential gene that is required during gastrulation in mice (Mueller et al. 2002) and its absence in yeast or mammalian cells leads to the loss of Arl1 and Golgin recruitment to the Golgi (Panic et al. 2003b; Setty et al. 2003; Shin et al. 2005; Zahn et al. 2006). In contrast to Arfs and Arl1, Arfrp1 is N-acetylated and this modification is required for its biological activity as well as for binding to Sys1p (Behnia et al. 2004). But what activates Arfrp1, how does Arfrp1GTP lead to activation of Arl1, and what benefit to cells or regulatory signaling is achieved by the sequential actions of these two GTPases in this fashion? We don’t have answers to these questions yet, but they highlight the fact that our models are still at a rather primitive stage in explaining the cellular actions of Arl1. In addition to the observations that support a model in which Arl1 acts in endosome–Golgi traffic through Golgin recruitment, other data suggest that Arl1 acts in anterograde traffic from the Golgi or in regulated secretion (IcardLiepkalns et al. 1997; Liu et al. 2006; Lock et al. 2005; Rosenwald et al. 2002; Shin et al. 2005). It is possible that the traffic of proteins from endosomes to Golgi is required to supply one or more shared components of anterograde Golgi traffic and thus these effects of Arl1 may be secondary to its role

114

*

R. A. Kahn

in endosome to Golgi traffic, or vice versa. But it seems more likely and consistent with precedence from Arf studies that Arl1 has more than one cellular role. Thus, while Arfs and Arls have specific roles in cell regulation at the Golgi their larger importance to cell biology may lie in their having multiple functions at more than one location and in this way allow coordination of metabolism and signaling throughout the cell.

References Altan-Bonnet N, Phair RD, Polishchuk RS, Weigert R, Lippincott-Schwartz J (2003) A role for Arf1 in mitotic Golgi disassembly, chromosome segregation, and cytokinesis. Proc Natl Acad Sci USA 100: 13314–13319 Ames JB, Ishima R, Tanaka T, Gordon JI, Stryer L, Ikura M (1997) Molecular mechanics of calcium–myristoyl switches. Nature 389: 198–202 Antoshechkin I, Han M (2002) The C. elegans evl-20 gene is a homolog of the small GTPase ARL2 and regulates cytoskeleton dynamics during cytokinesis and morphogenesis. Dev Cell 2: 579–591 Barlowe C, Orci L, Yeung T, Hosobuchi M, Hamamoto S, Salama N, Rexach MF, Ravazzola M, Amherdt M, Schekman R (1994) COPII: a membrane coat formed by Sec proteins that drive vesicle budding from the endoplasmic reticulum. Cell 77: 895–907 Behnia R, Panic B, Whyte JR, Munro S (2004) Targeting of the Arf-like GTPase Arl3p to the Golgi requires N-terminal acetylation and the membrane protein Sys1p. Nat Cell Biol 6: 405–413 Bhamidipati A, Lewis SA, Cowan NJ (2000) ADP ribosylation factor-like protein 2 (Arl2) regulates the interaction of tubulin-folding cofactor D with native tubulin. J Cell Biol 149: 1087–1096 Bigay J, Casella JF, Drin G, Mesmin B, Antonny B (2005) ArfGAP1 responds to membrane curvature through the folding of a lipid packing sensor motif. EMBO J 24: 2244–2253 Bigay J, Gounon P, Robineau S, Antonny B (2003) Lipid packing sensed by ArfGAP1 couples COPI coat disassembly to membrane bilayer curvature. Nature 426: 563–566 Boman AL, Zhang C, Zhu X, Kahn RA (2000) A family of ADP-ribosylation factor effectors that can alter membrane transport through the trans-Golgi. Mol Biol Cell 11: 1241–1255 Brown HA, Gutowski S, Moomaw CR, Slaughter C, Sternweis PC (1993) ADP-ribosylation factor, a small GTP-dependent regulatory protein, stimulates phospholipase D activity [see comments]. Cell 75: 1137–1144 Brown MT, Andrade J, Radhakrishna H, Donaldson JG, Cooper JA, Randazzo PA (1998) ASAP1, a phospholipid-dependent arf GTPase-activating protein that associates with and is phosphorylated by Src. Mol Cell Biol 18: 7038–7051 Burd CG, Strochlic TI, Gangi Setty SR (2004) Arf-like GTPases: not so Arf-like after all. Trends Cell Biol 14: 687–694 Burguete AS, Fenn TD, Brunger AT, Pfeffer SR (2008) Rab and Arl GTPase family members cooperate in the localization of the golgin GCC185. Cell 132: 286–298 Casanova JE (2007) Regulation of Arf activation: the Sec7 family of guanine nucleotide exchange factors. Traffic 8: 1476–1485 Caspary T, Larkins CE, Anderson KV (2007) The graded response to Sonic Hedgehog depends on cilia architecture. Dev Cell 12: 767–778 Chiang AP, Nishimura D, Searby C, Elbedour K, Carmi R, Ferguson AL, Secrist J, Braun T, Casavant T, Stone EM, Sheffield VC (2004) Comparative genomic analysis identifies an ADP-ribosylation factor-like gene as the cause of Bardet–Biedl syndrome (BBS3). Am J Hum Genet 75: 475–484

Arfs and Arls: models for Arf family members at the Golgi

*

115

Cockcroft S, Thomas GM, Fensome A, Geny B, Cunningham E, Gout I, Hiles I, Totty NF, Truong O, Hsuan JJ (1994) Phospholipase D: a downstream effector of ARF in granulocytes. Science 263: 523–526 De Matteis MA, Di Campli A, Godi A (2005) The role of the phosphoinositides at the Golgi complex. Biochim Biophys Acta 1744: 396–405 Dell’Angelica EC, Mullins C, Bonifacino JS (1999) AP-4, a novel protein complex related to clathrin adaptors. J Biol Chem 274: 7278–7285 Dell’Angelica EC, Puertollano R, Mullins C, Aguilar RC, Vargas JD, Hartnell LM, Bonifacino JS (2000) GGAs: a family of ADP ribosylation factor-binding proteins related to adaptors and associated with the Golgi complex. J Cell Biol 149: 81–94 Derby MC, Lieu ZZ, Brown D, Stow JL, Goud B, Gleeson PA (2007) The trans-Golgi network golgin, GCC185, is required for endosome-to-Golgi transport and maintenance of Golgi structure. Traffic 8: 758–773 Derby MC, Van Vliet C, Brown D, Luke MR, Lu L, Hong W, Stow JL, Gleeson PA (2004) Mammalian GRIP domain proteins differ in their membrane binding properties and are recruited to distinct domains of the TGN. J Cell Sci 117: 5865–5874 Donaldson JG, Finazzi D, Klausner RD (1992) Brefeldin A inhibits Golgi membranecatalysed exchange of guanine nucleotide onto ARF protein. Nature 360: 350–352 Donaldson JG, Kahn RA, Lippincott-Schwartz J, Klausner RD (1991) Binding of ARF and beta-COP to Golgi membranes: possible regulation by a trimeric G protein. Science 254: 1197–1199 Dong JH, Wen JF, Tian HF (2007) Homologs of eukaryotic Ras superfamily proteins in prokaryotes and their novel phylogenetic correlation with their eukaryotic analogs. Gene 396: 116–124 Drake MT, Zhu Y, Kornfeld S (2000) The assembly of AP-3 adaptor complex-containing clathrin-coated vesicles on synthetic liposomes. Mol Biol Cell 11: 3723–3736 Fan Y, Esmail MA, Ansley SJ, Blacque OE, Boroevich K, Ross AJ, Moore SJ, Badano JL, MaySimera H, Compton DS, Green JS, Lewis RA, Van Haelst MM, Parfrey PS, Baillie DL, Beales PL, Katsanis N, Davidson WS, Leroux MR (2004) Mutations in a member of the Ras superfamily of small GTP-binding proteins causes Bardet–Biedl syndrome. Nat Genet 36: 989–993 Faundez V, Horng JT, Kelly RB (1998) A function for the AP3 coat complex in synaptic vesicle formation from endosomes. Cell 93: 423–432 Faundez VV, Kelly RB (2000) The AP-3 complex required for endosomal synaptic vesicle biogenesis is associated with a casein kinase ialpha-like isoform [In Process Citation]. Mol Biol Cell 11: 2591–2604 Franco M, Chardin P, Chabre M, Paris S (1996) Myristoylation-facilitated binding of the G protein ARF1GDP to membrane phospholipids is required for its activation by a soluble nucleotide exchange factor. J Biol Chem 271: 1573–1578 Fujiwara T, Oda K, Yokota S, Takatsuki A, Ikehara Y (1988) Brefeldin A causes disassembly of the Golgi complex and accumulation of secretory proteins in the endoplasmic reticulum. J Biol Chem 263: 18545–18552 Gillingham AK, Munro S (2007) The small G proteins of the Arf family and their regulators. Annu Rev Cell Dev Biol 23: 579–611 Godi A, Di Campli A, Konstantakopoulos A, Di Tullio G, Alessi DR, Kular GS, Daniele T, Marra P, Lucocq JM, De Matteis MA (2004) FAPPs control Golgi-to-cell-surface membrane traffic by binding to ARF and PtdIns(4)P. Nat Cell Biol 6: 393–404 Godi A, Pertile P, Meyers R, Marra P, Di Tullio G, Iurisci C, Luini A, Corda D, De Matteis MA (1999) ARF mediates recruitment of PtdIns-4-OH kinase-beta and stimulates synthesis of PtdIns(4,5)P2 on the Golgi complex [see comments]. Nat Cell Biol 1: 280–287 Goldberg J (1998) Structural basis for activation of ARF GTPase: mechanisms of guanine nucleotide exchange and GTP-myristoyl switching. Cell 95: 237–248 Goldberg J (2000) Decoding of sorting signals by coatomer through a GTPase switch in the COPI coat complex. Cell 100: 671–679

116

*

R. A. Kahn

Hill K, Li Y, Bennett M, McKay M, Zhu X, Shern J, Torre E, Lah JJ, Levey AI, Kahn RA (2003) Munc 18 interacting proteins: ADP-ribosylation factor-dependent coat proteins that regulate the traffic of beta-Alzheimer’s precursor protein. J Biol Chem 278: 36032–36040 Hirst J, Bright NA, Rous B, Robinson MS (1999) Characterization of a fourth adaptorrelated protein complex [In Process Citation]. Mol Biol Cell 10: 2787–2802 Hirst J, Lindsay MR, Robinson MS (2001) GGAs: roles of the different domains and comparison with AP-1 and clathrin. Mol Biol Cell 12: 3573–3588 Honda A, Nogami M, Yokozeki T, Yamazaki M, Nakamura H, Watanabe H, Kawamoto K, Nakayama K, Morris AJ, Frohman MA, Kanaho Y (1999) Phosphatidylinositol 4phosphate 5-kinase alpha is a downstream effector of the small G protein ARF6 in membrane ruffle formation. Cell 99: 521–532 Hoyt MA, Stearns T, Botstein D (1990) Chromosome instability mutants of Saccharomyces cerevisiae that are defective in microtubule-mediated processes. Mol Cell Biol 10: 223–234 Icard-Liepkalns C, Ravassard P, Liepkalns VA, Chatail F, Mallet J (1997) An ADP-ribosylation-factor(ARF)-like protein involved in regulated secretion. Eur J Biochem 246: 388–393 Inoue H, Randazzo PA (2007) Arf GAPs and their interacting proteins. Traffic 8: 1465–1475 Jones DH, Morris JB, Morgan CP, Kondo H, Irvine RF, Cockcroft S (2000) Type I phosphatidylinositol 4-phosphate 5-kinase directly interacts with ADP-ribosylation factor 1 and is responsible for phosphatidylinositol 4,5-bisphosphate synthesis in the Golgi compartment. J Biol Chem 275: 13962–13966 Kahn RA (2004) ARF family GTPases. Kluwer Academic Publishers, Dordrecht Kahn RA, Cherfils J, Elias M, Lovering RC, Munro S, Schurmann A (2006) Nomenclature for the human Arf family of GTP-binding proteins: ARF, ARL, and SAR proteins. J Cell Biol 172: 645–650 Kahn RA, Gilman AG (1986) The protein cofactor necessary for ADP-ribosylation of Gs by cholera toxin is itself a GTP binding protein. J Biol Chem 261: 7906–7911 Kahn RA, Goddard C, Newkirk M (1988) Chemical and immunological characterization of the 21-kDa ADP-ribosylation factor of adenylate cyclase. J Biol Chem 263: 8282–8287 Kahn RA, Kern FG, Clark J, Gelmann EP, Rulka C (1991) Human ADP-ribosylation factors. A functionally conserved family of GTP-binding proteins. J Biol Chem 266: 2606–2614 Kahn RA, Randazzo P, Serafini T, Weiss O, Rulka C, Clark J, Amherdt M, Roller P, Orci L, Rothman JE (1992) The amino terminus of ADP-ribosylation factor (ARF) is a critical determinant of ARF activities and is a potent and specific inhibitor of protein transport. J Biol Chem 267: 13039–13046 Kam JL, Miura K, Jackson TR, Gruschus J, Roller P, Stauffer S, Clark J, Aneja R, Randazzo PA (2000) Phosphoinositide-dependent activation of the ADP-ribosylation factor GTPase-activating protein ASAP1. Evidence for the pleckstrin homology domain functioning as an allosteric site. J Biol Chem 275: 9653–9663 Kirchhausen T, Bonifacino JS, Riezman H (1997) Linking cargo to vesicle formation: receptor tail interactions with coat proteins. Curr Opin Cell Biol 9: 488–495 Krauss M, Kinuta M, Wenk MR, De Camilli P, Takei K, Haucke V (2003) ARF6 stimulates clathrin/AP-2 recruitment to synaptic membranes by activating phosphatidylinositol phosphate kinase type I {gamma}. J Cell Biol 162: 113–124 Ktistakis NT, Brown HA, Sternweis PC, Roth MG (1995) Phospholipase D is present on Golgi-enriched membranes and its activation by ADP ribosylation factor is sensitive to brefeldin A. Proc Natl Acad Sci USA 92: 4952–4956 Ktistakis NT, Brown HA, Waters MG, Sternweis PC, Roth MG (1996) Evidence that phospholipase D mediates ADP ribosylation factor-dependent formation of Golgi coated vesicles. J Cell Biol 134: 295–306

Arfs and Arls: models for Arf family members at the Golgi

*

117

Kuge O, Dascher C, Orci L, Rowe T, Amherdt M, Plutner H, Ravazzola M, Tanigawa G, Rothman JE, Balch WE (1994) Sar1 promotes vesicle budding from the endoplasmic reticulum but not Golgi compartments. J Cell Biol 125: 51–65 Lee FJ, Huang CF, Yu WL, Buu LM, Lin CY, Huang MC, Moss J, Vaughan M (1997) Characterization of an ADP-ribosylation factor-like 1 protein in Saccharomyces cerevisiae. J Biol Chem 272: 30998–31005 Lee FJ, Moss J, Vaughan M (1992) Human and Giardia ADP-ribosylation factors (ARFs) complement ARF function in Saccharomyces cerevisiae. J Biol Chem 267: 24441–24445 Lee I, Doray B, Govero J, Kornfeld S (2008) Binding of cargo sorting signals to AP-1 enhances its association with ADP ribosylation factor 1-GTP. J Cell Biol: jcb.200709037 Lee MC, Orci L, Hamamoto S, Futai E, Ravazzola M, Schekman R (2005) Sar1p N-terminal helix initiates membrane curvature and completes the fission of a COPII vesicle. Cell 122: 605–617 Li Y, Kelly WG, Logsdon JM Jr, Schurko AM, Harfe BD, Hill-Harfe KL, Kahn RA (2004) Functional genomic analysis of the ADP-ribosylation factor family of GTPases: phylogeny among diverse eukaryotes and function in C. elegans. FASEB J 18: 1834–1850 Lippincott-Schwartz J, Yuan LC, Bonifacino JS, Klausner RD (1989) Rapid redistribution of Golgi proteins into the ER in cells treated with brefeldin A: evidence for membrane cycling from Golgi to ER. Cell 56: 801–813 Liu YW, Huang CF, Huang KB, Lee FJ (2005) Role for Gcs1p in regulation of Arl1p at transGolgi compartments. Mol Biol Cell 16: 4024–4033 Liu YW, Lee SW, Lee FJ (2006) Arl1p is involved in transport of the GPI-anchored protein Gas1p from the late Golgi to the plasma membrane. J Cell Sci 119: 3845–3855 Lock JG, Hammond LA, Houghton F, Gleeson PA, Stow JL (2005) E-cadherin transport from the trans-Golgi network in tubulovesicular carriers is selectively regulated by golgin-97. Traffic 6: 1142–1156 Logsdon JJM, Kahn RA (2004) The Arf family tree. In: Kahn RA (ed) Arf family GTPases. Kluwer Academic Publishers, Dordrecht, pp 1–22 Lowe SL, Wong SH, Hong W (1996) The mammalian ARF-like protein 1 (Arl1) is associated with the Golgi complex. J Cell Sci 109: 209–220 Lu L, Hong W (2003) Interaction of Arl1-GTP with GRIP domains recruits autoantigens Golgin-97 and Golgin-245/p230 onto the Golgi. Mol Biol Cell 14: 3767–3781 Lu L, Horstmann H, Ng C, Hong W (2001) Regulation of Golgi structure and function by ARF-like protein 1 (Arl1). J Cell Sci 114: 4543–4555 Lu L, Tai G, Hong W (2004) Autoantigen Golgin-97, an effector of Arl1 GTPase, participates in traffic from the endosome to the trans-Golgi network. Mol Biol Cell 15: 4426–4443 McElver J, Patton D, Rumbaugh M, Liu C, Yang LJ, Meinke D (2000) The TITAN5 gene of Arabidopsis encodes a protein related to the ADP ribosylation factor family of GTP binding proteins. Plant Cell 12: 1379–1392 Misumi Y, Miki K, Takatsuki A, Tamura G, Ikehara Y (1986) Novel blockade by brefeldin A of intracellular transport of secretory proteins in cultured rat hepatocytes. J Biol Chem 261: 11398–11403 Mueller AG, Moser M, Kluge R, Leder S, Blum M, Buttner R, Joost HG, Schurmann A (2002) Embryonic lethality caused by apoptosis during gastrulation in mice lacking the gene of the ADP-ribosylation factor-related protein 1. Mol Cell Biol 22: 1488–1494 Munro S (2005) The Arf-like GTPase Arl1 and its role in membrane traffic. Biochem Soc Trans 33: 601–605 Nakano A, Muramatsu M (1989) A novel GTP-binding protein, Sar1p, is involved in transport from the endoplasmic reticulum to the Golgi apparatus. J Cell Biol 109: 2677–2691

118

*

R. A. Kahn

Ohno H, Aguilar RC, Yeh D, Taura D, Saito T, Bonifacino JS (1998) The medium subunits of adaptor complexes recognize distinct but overlapping sets of tyrosine-based sorting signals. J Biol Chem 273: 25915–25921 Oka T, Nishikawa S, Nakano A (1991) Reconstitution of GTP-binding Sar1 protein function in ER to Golgi transport. J Cell Biol 114: 671–679 Okai T, Araki Y, Tada M, Tateno T, Kontani K, Katada T (2004) Novel small GTPase subfamily capable of associating with tubulin is required for chromosome segregation. J Cell Sci 117: 4705–4715 Ooi CE, Dell’Angelica EC, Bonifacino JS (1998) ADP-Ribosylation factor 1 (ARF1) regulates recruitment of the AP-3 adaptor complex to membranes. J Cell Biol 142: 391–402 Paleotti O, Macia E, Luton F, Klein S, Partisani M, Chardin P, Kirchhausen T, Franco M (2005) The small G-protein Arf6GTP recruits the AP-2 adaptor complex to membranes. J Biol Chem 280: 21661–21666 Panic B, Perisic O, Veprintsev DB, Williams RL, Munro S (2003a) Structural basis for Arl1dependent targeting of homodimeric GRIP domains to the Golgi apparatus. Mol Cell 12: 863–874 Panic B, Whyte JR, Munro S (2003b) The ARF-like GTPases Arl1p and Arl3p act in a pathway that interacts with vesicle-tethering factors at the Golgi apparatus. Curr Biol 13: 405–410 Puertollano R, Aguilar RC, Gorshkova I, Crouch RJ, Bonifacino JS (2001a) Sorting of mannose 6-phosphate receptors mediated by the GGAs. Science 292: 1712–1716 Puertollano R, Randazzo PA, Presley JF, Hartnell LM, Bonifacino JS (2001b) The GGAs promote Arf-dependent recruitment of clathrin to the TGN. Cell 105: 93–102 Randazzo PA (1997) Functional interaction of ADP-ribosylation factor 1 with phosphatidylinositol 4,5-bisphosphate. J Biol Chem 272: 7688–7692 Randazzo PA, Kahn RA (1994) GTP hydrolysis by ADP-ribosylation factor is dependent on both an ADP-ribosylation factor GTPase-activating protein and acid phospholipids [published erratum appears in J Biol Chem 1994 Jun 10; 269(23), 16519]. J Biol Chem 269: 10758–10763 Robinson MS, Kreis TE (1992) Recruitment of coat proteins onto Golgi membranes in intact and permeabilized cells: effects of brefeldin A and G protein activators. Cell 69: 129–138 Rosenwald AG, Rhodes MA, Van Valkenburgh H, Palanivel V, Chapman G, Boman A, Zhang CJ, Kahn RA (2002) ARL1 and membrane traffic in Saccharomyces cerevisiae. Yeast 19: 1039–1056 Sahin A, Lemercier G, Tetaud E, Espiau B, Myler P, Stuart K, Bakalara N, Merlin G (2004) Trypanosomatid flagellum biogenesis: ARL-3A is involved in several species. Exp Parasitol 108: 126–133 Salazar G, Craige B, Wainer BH, Guo J, De Camilli P, Faundez V (2005) Phosphatidylinositol-4-kinase type II alpha is a component of adaptor protein-3-derived vesicles. Mol Biol Cell 16: 3692–3704 Schrick JJ, Vogel P, Abuin A, Hampton B, Rice DS (2006) ADP-ribosylation factor-like 3 is involved in kidney and photoreceptor development. Am J Pathol 168: 1288–1298 Schweitzer JK, D’Souza-Schorey C (2005) A requirement for ARF6 during the completion of cytokinesis. Exp Cell Res 311: 74–83 Serafini T, Orci L, Amherdt M, Brunner M, Kahn RA, Rothman JE (1991) ADP-ribosylation factor is a subunit of the coat of Golgi-derived COP-coated vesicles: a novel role for a GTP-binding protein. Cell 67: 239–253 Setty SR, Shin ME, Yoshino A, Marks MS, Burd CG (2003) Golgi recruitment of GRIP domain proteins by Arf-like GTPase 1 is regulated by Arf-like GTPase 3. Curr Biol 13: 401–404 Shin HW, Kobayashi H, Kitamura M, Waguri S, Suganuma T, Uchiyama Y, Nakayama K (2005) Roles of ARFRP1 (ADP-ribosylation factor-related protein 1) in post-Golgi membrane trafficking. J Cell Sci 118: 4039–4048

Arfs and Arls: models for Arf family members at the Golgi

*

119

Shrivastava-Ranjan P, Faundez V, Fang G, Rees H, Lah JJ, Levey AI, Kahn RA (2008) Mint3/ X11{gamma} is an ADP-ribosylation factor-dependent adaptor that regulates the traffic of the Alzheimer’s precursor protein from the trans-Golgi network. Mol Biol Cell 19: 51–64 Stamnes MA, Rothman JE (1993) The binding of AP-1 clathrin adaptor particles to Golgi membranes requires ADP-ribosylation factor, a small GTP-binding protein. Cell 73: 999–1005 Stearns T, Willingham MC, Botstein D, Kahn RA (1990) ADP-ribosylation factor is functionally and physically associated with the Golgi complex. Proc Natl Acad Sci USA 87: 1238–1242 Takatsu H, Katoh Y, Shiba Y, Nakayama K (2001) Golgi-localizing, gamma-Adaptin ear homology domain, ADP-ribosylation factor-binding (GGA) proteins interact with acidic dileucine sequences within the cytoplasmic domains of sorting receptors through their Vps27p/Hrs/STAM (VHS) domains. J Biol Chem 276: 28541–28545 Tamkun JW, Kahn RA, Kissinger M, Brizuela BJ, Rulka C, Scott MP, Kennison JA (1991) The arflike gene encodes an essential GTP-binding protein in Drosophila. Proc Natl Acad Sci USA 88: 3120–3124 Terui T, Kahn RA, Randazzo PA (1994) Effects of acid phospholipids on nucleotide exchange properties of ADP-ribosylation factor 1. Evidence for specific interaction with phosphatidylinositol 4,5-bisphosphate. J Biol Chem 269: 28130–28135 Traub LM, Ostrom JA, Kornfeld S (1993) Biochemical dissection of AP-1 recruitment onto Golgi membranes. J Cell Biol 123: 561–573 Van Valkenburgh H, Shern JF, Sharer JD, Zhu X, Kahn RA (2001) ADP-ribosylation factors (ARFs) and ARF-like 1 (ARL1) have both specific and shared effectors: characterizing ARL1-binding proteins. J Biol Chem 276: 22826–22837 Vieira OV, Verkade P, Manninen A, Simons K (2005) FAPP2 is involved in the transport of apical cargo in polarized MDCK cells. J Cell Biol 170: 521–526 Volpicelli-Daley LA, Li Y, Zhang CJ, Kahn RA (2005) Isoform-selective effects of the depletion of ADP-ribosylation factors 1–5 on membrane traffic. Mol Biol Cell 16: 4495–4508 West MA, Bright NA, Robinson MS (1997) The role of ADP-ribosylation factor and phospholipase D in adaptor recruitment. J Cell Biol 138: 1239–1254 Wu M, Lu L, Hong W, Song H (2004) Structural basis for recruitment of GRIP domain golgin-245 by small GTPase Arl1. Nat Struct Mol Biol 11: 86–94 Zahn C, Hommel A, Lu L, Hong W, Walther DJ, Florian S, Joost HG, Schurmann A (2006) Knockout of Arfrp1 leads to disruption of ARF-like1 (ARL1) targeting to the transGolgi in mouse embryos and HeLa cells. Mol Membr Biol 23: 475–485 Zheng Y, Glaven JA, Wu WJ, Cerione RA (1996) Phosphatidylinositol 4,5-bisphosphate provides an alternative to guanine nucleotide exchange factors by stimulating the dissociation of GDP from Cdc42Hs. J Biol Chem 271: 23815–23819 Zhou C, Cunningham L, Marcus AI, Li Y, Kahn RA (2006) Arl2 and Arl3 regulate different microtubule-dependent processes. Mol Biol Cell 17: 2476–2487 Zhu X, Boman AL, Kuai J, Cieplak W, Kahn RA (2000) Effectors increase the affinity of ADP-ribosylation factor for GTP to increase binding. J Biol Chem 275: 13465–13475 Zhu Y, Doray B, Poussu A, Lehto VP, Kornfeld S (2001) Binding of GGA2 to the lysosomal enzyme sorting motif of the mannose 6-phosphate receptor. Science 292: 1716–1718

120

*

V. Lupashin and D. Ungar

COG complex Vladimir Lupashin and Daniel Ungar

COG structure The conserved oligomeric Golgi (COG) protein complex consists of eight subunits named Cog1 through Cog8 (Kingsley et al. 1986; Ram et al. 2002; Suvorova et al. 2001, 2002; Ungar et al. 2002, Whyte and Munro 2001). Based on yeast genetic studies, biochemical observations and electron micrographs of the complex (Ungar et al. 2002) the COG subunits have been grouped into two lobes consisting of Cog1 to Cog4 and Cog5 to Cog8, respectively (Fig. 1). The two lobes appear to be interconnected by thin rods and/or globules. Cog1p is likely a bridging subunit between the two COG lobes in yeast (Fotso et al. 2005). The bridge that joins lobe A and lobe B in the mammalian COG complex is similarly composed of Cog1 and Cog8 (Oka et al. 2005; Ungar et al. 2005). The COG complex is evolutionarily conserved; homologues of its most conserved subunits, Cog3p, Cog4p, Cog6p and Cog8p can be found in most of the sequenced eukaryotic genomes, including yeast, plants, insects, worms and mammals (Koumandou et al. 2007). In addition of its conservation between species, COG also appears to be related to tethering complexes acting at different vesicular transport steps in the cell (Whyte and Munro 2001). Iterative database searches using N-terminal domains of several COG components revealed similarities in the components of the exocyst (for a recent review see Hsu et al. (2004)) and the GARP (Liewen et al. 2005) complex. Further support for the relationship between COG and the exocyst comes from the recently reported solution structure of a large portion of yeast Cog2p (Cavanaugh et al. 2007). The structure reveals a six-helix bundle with few conserved surface features but a general resemblance to crystal structures of several exocyst subunit-domains. Subcellular fractionation and immunofluorescence microscopy both show that COG is localized to Golgi membranes (Suvorova et al. 2002; Ungar et al. 2002; Walter et al. 1998). Moreover, immunogold electron microscopy using functional, hemagglutinin epitope-labeled Cog1 demonstrated that COG localizes primarily on or in close proximity to the tips and rims of the Golgi cisternae and their associated vesicles (Vasile et al. 2006). However, none of the COG subunits has either a transmembrane or a lipid-binding domain and thus the complex most likely attaches to the membrane via specific and highly-regulated protein–protein interactions. Data from our and other labs suggest that both Golgi localization and membrane attachment of the

COG complex

*

121

Figure 1. COG structure. (a) Rotary shadowed cryoelectron micrographs of purified bovine brain COG complex either after a brief glutaraldehyde fixation (bottom row), or untreated (top row). Modified from (Ungar et al. 2002). [Reproduced from the Journal of Cell Biology, 2002, 157: 405–415.  2002 The Rockefeller University Press.] (b) Subunit interaction map of mammalian and yeast COG complex. Modified from: (Ungar et al. 2005; Fotso et al. 2005). Insert depicts a ribbon diagram of a large portion of yeast Cog2p (Cavanaugh et al. 2007).

COG complex is probably multifaceted and may involve several COG subunits. We have found that the absence of one subunit of the yeast COG complex does not significantly affect membrane association of others (Fotso et al. 2005). Both Golgi-localized lipid-modified Rab proteins and transmembrane SNARE proteins could be implicated as factors that facilitate COG-membrane binding. Since these interactions are likely all transient, it is possible that the complex also has a more permanent membrane receptor, but experimental evidence for such is still lacking.

The interactome of COG Not much is known about the molecular details of vesicle tethering, but it is likely a series of protein interactions and rearrangements involving a tethering complex such as COG and members of other protein families including Rab GTPases, vesicle coats and SNAREs. Therefore it is worth considering what is known about the protein interactions of the COG complex. Rab proteins act as molecular switches that shuttle on/off membranes while attracting a diverse group of effectors, to control protein trafficking in

122

*

V. Lupashin and D. Ungar

eukaryotes (Zerial and McBride 2001). There are more than 20 Rabs associated with the mammalian Golgi apparatus (Gilchrist et al. 2006), and tethering factors are known to interact with them (Sztul and Lupashin 2006). In yeast cells the COG complex interacts with two Golgi Rabs, Ypt1 and Ypt6 (Ram et al. 2002; Suvorova et al. 2002; Whyte and Munro 2001). As a functional consequence, overexpression of YPT1 suppresses growth defects associated with the loss of COG function (VanRheenen et al. 1998, 1999). Conversely, the simultaneous loss of a non-essential COG subunit and Ypt6, produces synthetic lethality (Tong et al. 2004). Mammalian COG has recently been shown to interact with several Golgi Rabs in a GTP-dependent manner as well (D. Ungar, unpublished data). Epistatic depletion of mammalian Rab6 inhibited the Golgi disruptive effects of COG complex inactivation by siRNA or antibodies, indicating that COG operates downstream of Rab6 (Sun et al. 2007). A number of oligomeric tethering complexes co-localize and directly interact with SNARE molecules (Fridmann-Sirkis et al. 2006; Price et al. 2000; Suvorova et al. 2002), but the mechanism of this tether-SNARE interaction is not well understood. Besides numerous genetic interactions (Kim et al. 1999; Ram et al. 2002; VanRheenen et al. 1998, 1999), the yeast COG complex also co-immunoprecipitates (co-IP) with the Golgi SNAREs Sed5p, Gos1p, Ykt6p and Sec22p (Suvorova and Lupashin 2002). Moreover, mammalian COG co-IPs with GS28, the mammalian homologue of Gos1p (Zolov and Lupashin 2005), and FRET reveals an in vivo interaction between Syntaxin5a and the COG complex (Shestakova et al. 2007). Data from several labs also indicate that the localization and stability of several Golgi SNAREs are altered in cells expressing a defective COG complex (Fotso et al. 2005; Oka et al. 2004; Shestakova et al. 2006; Zolov and Lupashin 2005). A direct functional link between COG and SNAREs comes from an experiment using RNAi knockdown (KD) of the mammalian COG complex. Here, Golgi SNARE mobility was decreased and free Syntaxin5 accumulated suggesting a direct involvement of COG in promoting SNARE complex formation and dynamics (Shestakova et al. 2007). Other protein families to which the COG complex has been linked are the COPI coat (Oka et al. 2004; Ram et al. 2002; Suvorova et al. 2002), and golgin tethers (Shestakova et al. 2007; Sohda et al. 2007). COPI coats are found on trafficking vesicles originating within the Golgi, and have an established role in retrograde trafficking (Letourneur et al. 1994). Oka et al. (2004) have employed short interfering RNA (siRNA)-mediated z-COP depletion and established that COG and z-COP contribute to similar functions necessary for normal Golgi organization. This data is in a good agreement with the proposal that in yeast cells COG complex mediates the tethering of COPI coated trafficking intermediates (Suvorova et al. 2002). The interaction found between the golgin p115 and Cog2 (Sohda et al. 2007) has similar implications, and suggests that the interactions between other golgins and COG should be further explored in the future. Thus, the COG complex “interactome” includes SNAREs, Rabs and COPI coat proteins, all proteins with well-established roles in vesicular transport

COG complex

*

123

Table 1. COG interactome COG subunits

Species

Interacting protein

Comments

COG complex

Budding yeasts

Ypt1-GTP

In vitro binding (Suvorova et al. 2002)

Ypt6

In vitro binding (Suvorova et al. 2002)

Sed5p

In vitro binding (Suvorova et al. 2002)

Sed5p

Co-IP (Suvorova et al. 2002)

Cog1

Budding yeasts

COPI

Co-IP (Suvorova et al. 2002)

Cog2

Mammals

p115

Y2H, in vitro binding, co-IP (Sohda et al. 2007)

Budding yeasts

g-COPI

Y2H (Suvorova et al. 2002)

Sed5p

Co-IP (Suvorova et al. 2002)

GS28

Co-IP (Zolov and Lupashin 2005)

Cog3

Mammals

Budding yeasts

Cog4

Budding yeasts Mammals

Syntaxin5a

Co-IP (Shestakova et al. 2007)

b-COPI

Co-IP (Zolov and Lupashin 2005)

p115

Co-IP (Shestakova et al. 2007; Sohda et al. 2007)

GM130

Co-IP (Shestakova et al. 2007; Sohda et al. 2007)

Giantin

Co-IP (Sohda et al. 2007)

Sed5p

Co-IP, in vitro binding (Suvorova et al. 2002)

Ykt6p

Co-IP (Suvorova et al. 2002)

Gos1p

Co-IP (Suvorova et al. 2002)

Sec22p

Co-IP (Suvorova et al. 2002)

COPI

Co-IP (Suvorova et al. 2002)

Sed5p

Co-IP (Suvorova et al. 2002)

COPI

Co-IP (Suvorova et al. 2002)

Syntaxin5a

Y2H, Co-IP (Shestakova et al. 2007)

Cog5

Mammals

Syntaxin5a

Y2H (Shestakova et al. 2007)

Cog6

Budding yeasts

Sed5p

Co-IP (Suvorova et al. 2002)

COPI

Co-IP (Suvorova et al. 2002)

Mammals

Syntaxin5a

Y2H, Co-IP (Shestakova et al. 2007)

Mammals

Syntaxin5a

Y2H (Shestakova et al. 2007)

Cog8

124

*

V. Lupashin and D. Ungar

(Table 1). Further analysis of interactions between the COG complex subunits and other Golgi-localized components of vesicle trafficking machinery will help in understanding the COG complex cellular function.

The COG complex and Golgi-glycosylation Mutation or deletion of individual COG subunits give rise to different phenotypes, suggesting that each member protein plays a distinct role within the complex. This also means that a number of different phenotypes can be associated with COG defects, from lethality in the yeast S. cerevisiae (deletion of COG1, 2, 3 or 4 (Kim et al. 1999, 2001; Ram et al. 2002; Suvorova et al. 2002; VanRheenen et al. 1998, 1999)) to male sterility in Drosophila (Farkas et al. 2003) and to disruptions in Golgi structure in mammals (Zolov and Lupashin 2005; Ungar et al. 2002). The overarching theme of mutant phenotypes is, however, a defect in one of the core Golgi-functions: glycosylation. The following paragraphs summarize some of the glycosylation-related phenotypic alterations associated with COG defects in different organisms. As opposed to mammalian cells cultured ex vivo, Golgi-glycosylation is essential for the proliferation of the yeast S. cerevisiae. Therefore it is not surprising, that deletion of any of the Cog1–4 subunits is lethal, since experiments with temperature-sensitive mutant strains or with suppressor genes show that these subunits are necessary for proper Golgi glycosylation. In contrast, deletion of any of the Cog5–8 subunits does not cause any observable growth defect, and indeed the defects in glycosylation are minimal (Suvorova et al. 2002; Whyte and Munro 2001). In C. elegans, the glycoprotein Mig17 is an ADAM metalloprotease that plays a role in distal tip cell migration, which is necessary for gonad morphogenesis. The deletion of either C. elegans Cog1 or Cog3 leads to underglycosylation of Mig17, causing a misdirection of the distal tip cells, similar to a Mig17 deletion (Kubota et al. 2006). Monty Krieger’s lab at MIT created several mutant Chinese hamster ovary (CHO) cell lines that display severe defects in low density lipoprotein (LDL) receptor function (Kingsley and Krieger 1984). It was subsequently shown that the ldlB (COG1 knockout) and ldlC (COG2 knockout) cells are deficient in glycosylation of both N-linked and O-linked glycans (Kingsley et al. 1986), and that altered glycosylation of LDL receptor results in reduced steady state levels and therefore decreased activity of the receptor. An elegant demonstration of the glycosylation defects in these cells is their differential sensitivity to toxic lectins. These plant toxins are endocytosed after binding to specific glycan sequences on cell surface glycoproteins, and then inhibit vital cell functions such as protein synthesis. For ldlB and ldlC cells the LD10 was lower than wild type for the lectins Concanavalin A and ricin (which bind mannose, GalNAc, or terminal galactose) but higher than wild type for WGA, PHA, and LCA (binding sialic acid, GlcNAc, branched-chain galactose, or mannose in fucosylated chains) (Kingsley et al. 1986).

COG complex

*

125

COG-associated glycosylation defects are not restricted to model organisms. Several different COG mutations have been identified in patients suffering from congenital glycosylation disorders (also referred to as CDGs) (Freeze 2007). CDGs are recessive multisystem genetic disorders in which the glycosylation machinery is defective. In type I CDGs the initial steps of Nglycosylation in the ER are deficient, whereas type II CDGs are defective in Golgi-associated O-glycosylation or N-glycosylation (Foulquier et al. 2006). COG defects are all type II CDGs, but whereas most other CDG IIs are defective in a particular enzyme, and thus present either N- or O-glycan defects, COG patients, similar to the above-discussed ldlB and ldlC cells have both (Foulquier et al. 2006). The first case of COG-associated CDG, a defect in COG7, was described in 2004 (Wu et al. 2004). The three siblings described in this study showed a reduction in total serum sialic acid and died shortly after birth. Since then six other mutations in different COG subunits were identified in patients (summarized in Table 2). All of the known COG deletions in humans cause microcephaly (“small brain”), hypotonia (low muscle tone due to defect in Table 2. Mutations in COG complex subunits causing congenital glycosylation disorders Subunit

Mutation

Glycosylation defects

Cog1(ldlB)

1. Cog1-2660iC, this one base insertion replaces the C-terminal 94 amino acids with 13 unrelated ones (Foulquier et al. 2006). The connection between the two COG lobes is lost.

1. The mutation causes partial sialylation and galactosylation defects, a decrease in sugar-transporters and glycosyltransferases (Foulquier et al. 2006).

Cog7

The first splice site in the Cog7 gene is mutated. The absence of splicing causes premature termination of translation (Morava et al. 2007; Wu et al. 2004). The absence of Cog7 causes an almost complete loss of the lobe B subunits.

Sialylation and galactosylation defects. Decrease in sugar-transporters and glycosyltransferases (Morava et al. 2007; Ng et al. 2007; Steet and Kornfeld 2006; Wu et al. 2004).

Cog8

1. Cog8-Tyr537STOP (Foulquier et al. 2007). The connection between the two COG lobes is lost. 2. In one Cog8 allele the 3rd intron is not spliced causing premature termination, in the other allele the 1687–1688 del TT deletion causes truncation of the C-terminal 47 amino acids (Kranz et al. 2007).

1. Partial sialylation and galactosylation defects and a decrease in b-1,4-galactosyltransferase levels (Foulquier et al. 2007). 2. Mild sialylation defect (Kranz et al. 2007).

126

*

V. Lupashin and D. Ungar

cerebellar growth), and failure to thrive (Wu et al. 2004). It is clear, however, that these mutations only affect part of COG’s function, since none of them shows the severe glycosylation defects observed in the ldlB and ldlC cells. A complete loss of COG function would likely be embryonic lethal, and therefore would not make it into a clinical setting.

COG complex function in vesicle trafficking We have recently demonstrated that the acute knock down of COG3 in HeLa cells is accompanied by reduction in Cog1, 2 and 4 protein levels and results in the accumulation of COG complex-dependent (CCD) vesicles carrying Golgi vSNARE molecules (Zolov and Lupashin 2005). Vesicle accumulation as a response for the COG complex dysfunction is evolutionarily conserved, since previous electron microscopy experiments revealed that yeast cog2 and cog3 temperature-sensitive mutants accumulated 50 nm vesicles (Wuestehube et al. 1996). COG complex-dependent docking of isolated CCD vesicles was reconstituted in vitro supporting their role as functional trafficking intermediates (Shestakova et al. 2006), and providing crucial evidence for the role of COG as a tethering factor. A prolonged block in CCD vesicle tethering is accompanied by substantial fragmentation of the Golgi ribbon followed by defects in glycan processing. Fragmented Golgi membranes maintain their juxtanuclear localization, cisternal organization and competence for anterograde protein trafficking to the plasma membrane. These findings led to the hypothesis that the COG complex acts as a tether which connects COPI vesicles with cis-Golgi membranes during retrograde intra-Golgi traffic (Suvorova et al. 2002; Ungar et al. 2002). Additional evidence that COG plays a role in the retrograde vesicular transport of Golgi proteins, including the glycosylation enzymes required for normal Golgi function, came from surveying the steady state levels of Golgi proteins in wild type and COG-deficient mammalian cells (Oka et al. 2004). Seven Golgi membrane proteins (called “GEAR”s, which is not an acronym), including the processing enzyme a-1,3-1,6-mannosidase II (Man II), were found to exhibit reduced steady state levels in both Cog1- and Cog2-deficient CHO cells. Detailed analysis of cellular phenotypes at different stages of COG3 KD was very instrumental to uncover the molecular link between COG function and glycosylation disorders. It was found that in COG-depleted cells two medial-Golgi enzymes, GlcNAc-TI and ManII, are transiently relocated into CCD vesicles. As a result, Golgi modifications of both plasma membrane (CD44) and lysosomal (Lamp2) glycoproteins are distorted. Final intracellular localization of these glycoproteins was not altered indicating that the COG complex is not required for either anterograde trafficking or postGolgi sorting. COG7 KD and double COG3/COG7 KD caused similar defects with respect to both Golgi traffic and glycosylation suggesting that the entire COG complex orchestrates recycling of medial-Golgi resident proteins (Shestakova et al. 2006). Similar defects in localization of cis-Golgi

COG complex

*

127

mannosyltransferase Och1p were found in yeast cog3-temperature-sensitive mutant cells (Bruinsma et al. 2004) supporting the hypothesis that COG complex-dependent recycling of Golgi glycosylation enzymes is evolutionarily conserved. Altogether, the data suggest that constantly cycling medialGolgi enzymes are transported from distal compartments to the cis/medial Golgi in CCD vesicles. Dysfunction of the COG complex leads to separation of glycosyltransferases from anterograde cargo molecules passing along the secretory pathway thus affecting normal protein glycosylation. Acknowledgments. Supported by grant from the NSF (MCB-0645163) for VL, and a Marie Curie reintegration grant from the EU (No. 201098) for DU.

References Bruinsma P, Spelbrink RG, Nothwehr SF (2004) Retrograde transport of the mannosyltransferase Och1p to the early Golgi requires a component of the COG transport complex. J Biol Chem 279: 39814–39823 Cavanaugh LF, Chen X, Richardson BC, Ungar D, Pelczer I, Rizo J, Hughson FM (2007) Structural analysis of conserved oligomeric Golgi complex subunit 2. J Biol Chem 282: 23418–23426 Farkas RM, Giansanti MG, Gatti M, Fuller MT (2003) The Drosophila Cog5 homologue is required for cytokinesis, cell elongation, and assembly of specialized Golgi architecture during spermatogenesis. Mol Biol Cell 14: 190–200 Fotso P, Koryakina Y, Pavliv O, Tsiomenko AB, Lupashin VV (2005) Cog1p plays a central role in the organization of the yeast conserved oligomeric Golgi complex. J Biol Chem 280: 27613–27623 Foulquier F, Ungar D, Reynders E, Zeevaert R, Mills P, Garcia-Silva MT, Briones P, Winchester B, Morelle W, Krieger M, Annaert W, Matthijs G (2007) A new inborn error of glycosylation due to a Cog8 deficiency reveals a critical role for the Cog1–Cog8 interaction in COG complex formation. Hum Mol Genet 16(7): 717–730 Foulquier F, Vasile E, Schollen E, Callewaert N, Raemaekers T, Quelhas D, Jaeken J, Mills P, Winchester B, Krieger M, Annaert W, Matthijs G (2006) Conserved oligomeric Golgi complex subunit 1 deficiency reveals a previously uncharacterized congenital disorder of glycosylation type II. Proc Natl Acad Sci USA 103: 3764–3769 Freeze HH (2007) Congenital Disorders of Glycosylation: CDG-I, CDG-II, and beyond. Curr Mol Med 7(4): 389–396 Fridmann-Sirkis Y, Kent HM, Lewis MJ, Evans PR, Pelham HR (2006) Structural analysis of the interaction between the SNARE Tlg1 and Vps51. Traffic 7: 182–190 Gilchrist A, Au CE, Hiding J, Bell AW, Fernandez-Rodriguez J, Lesimple S, Nagaya H, Roy L, Gosline SJ, Hallett M, Paiement J, Kearney RE, Nilsson T, Bergeron JJ (2006) Quantitative proteomics analysis of the secretory pathway. Cell 127: 1265–1281 Hsu SC, TerBush D, Abraham M, Guo W (2004) The exocyst complex in polarized exocytosis. Int Rev Cytol 233: 243–265 Kim DW, Massey T, Sacher M, Pypaert M, Ferro-Novick S (2001) Sgf1p, a new component of the Sec34p/Sec35p complex. Traffic 2: 820–830 Kim DW, Sacher M, Scarpa A, Quinn AM, Ferro-Novick S (1999) High-copy suppressor analysis reveals a physical interaction between Sec34p and Sec35p, a protein implicated in vesicle docking. Mol Biol Cell 10: 3317–3329 Kingsley DM, Kozarsky KF, Segal M, Krieger M (1986) Three types of low density lipoprotein receptor-deficient mutant have pleiotropic defects in the synthesis

128

*

V. Lupashin and D. Ungar

of N-linked, O-linked, and lipid-linked carbohydrate chains. J Cell Biol 102: 1576–1585 Kingsley DM, Krieger M (1984) Receptor-mediated endocytosis of low density lipoprotein: somatic cell mutants define multiple genes required for expression of surfacereceptor activity. Proc Natl Acad Sci USA 81: 5454–5458 Koumandou VL, Dacks JB, Coulson RM, Field MC (2007) Control systems for membrane fusion in the ancestral eukaryote; evolution of tethering complexes and SM proteins. BMC Evol Biol 7: 29 Kranz C, Ng BG, Sun L, Sharma V, Eklund EA, Miura Y, Ungar D, Lupashin V, Winkel RD, Cipollo JF, Costello CE, Loh E, Hong W, Freeze HH (2007) COG8 deficiency causes new congenital disorder of glycosylation type IIh. Hum Mol Genet 16: 731–741 Kubota Y, Sano M, Goda S, Suzuki N, Nishiwaki K (2006) The conserved oligomeric Golgi complex acts in organ morphogenesis via glycosylation of an ADAM protease in C. elegans. Development 133: 263–273 mollie re C, Duden R, Emr SD, Riezman H, Letourneur F, Gaynor EC, Hennecke S, De Cosson P (1994) Coatomer is essential for retrieval of dilysine-tagged proteins to the endoplasmic reticulum. Cell 79: 1199–1207 Liewen H, Meinhold-Heerlein I, Oliveira V, Schwarzenbacher R, Luo G, Wadle A, Jung M, Pfreundschuh M, Stenner-Liewen F (2005) Characterization of the human GARP (Golgi associated retrograde protein) complex. Exp Cell Res 306: 24–34 Morava E, Zeevaert R, Korsch E, Huijben K, Wopereis S, Matthijs G, Keymolen K, Lefeber DJ, De Meirleir L, Wevers RA (2007) A common mutation in the COG7 gene with a consistent phenotype including microcephaly, adducted thumbs, growth retardation, VSD and episodes of hyperthermia. Eur J Hum Genet 15: 638–645 Ng BG, Kranz C, Hagebeuk EE, Duran M, Abeling NG, Wuyts B, Ungar D, Lupashin V, Hartdorff CM, Poll-The BT, Freeze HH (2007) Molecular and clinical characterization of a Moroccan Cog7 deficient patient. Mol Genet Metab 91: 201–204 Oka T, Ungar D, Hughson FM, Krieger M (2004) The COG and COPI complexes interact to control the abundance of GEARs, a subset of Golgi integral membrane proteins. Mol Biol Cell 15: 2423–2435 Oka T, Vasile E, Penman M, Novina CD, Dykxhoorn DM, Ungar D, Hughson FM, Krieger M (2005) Genetic analysis of the subunit organization and function of the conserved oligomeric Golgi (COG) complex: studies of COG5- and COG7-deficient mammalian cells. J Biol Chem 280: 32736–32745 Price A, Seals D, Wickner W, Ungermann C (2000) The docking stage of yeast vacuole fusion requires the transfer of proteins from a cis-SNARE complex to a Rab/Ypt protein. J Cell Biol 148: 1231–1238 Ram RJ, Li B, Kaiser CA (2002) Identification of sec36p, sec37p, and sec38p: components of yeast complex that contains sec34p and sec35p. Mol Biol Cell 13: 1484–1500 Shestakova A, Suvorova E, Pavliv O, Khaidakova G, Lupashin V (2007) Interaction of the conserved oligomeric Golgi complex with t-SNARE Syntaxin5a/Sed5 enhances intra-Golgi SNARE complex stability. J Cell Biol 179(6): 1179–1192 Shestakova A, Zolov S, Lupashin V (2006) COG complex-mediated recycling of Golgi glycosyltransferases is essential for normal protein glycosylation. Traffic 7: 191–204 Sohda M, Misumi Y, Yoshimura S, Nakamura N, Fusano T, Ogata S, Sakisaka S, Ikehara Y (2007) The interaction of two tethering factors, p115 and COG complex, is required for Golgi integrity. Traffic 8: 270–284 Steet R, Kornfeld S (2006) COG-7-deficient human fibroblasts exhibit altered recycling of Golgi proteins. Mol Biol Cell 17: 2312–2321 Sun Y, Shestakova A, Hunt L, Sehgal S, Lupashin V, Storrie B (2007) Rab6 Regulates Both ZW10/RINT-1 and COG complex dependent Golgi trafficking and homeostasis. Mol Biol Cell 18(10):\ 4129–4142 Suvorova E, Lupashin VV (2002) COG complex interacts with the components of the Golgi tethering machinery. Mol Biol Cell 13: 266A–266A

COG complex

*

129

Suvorova ES, Duden R, Lupashin VV (2002) The Sec34/Sec35p complex, a Ypt1p effector required for retrograde intra-Golgi trafficking, interacts with Golgi SNAREs and COPI vesicle coat proteins. J Cell Biol 157: 631–643 Suvorova ES, Kurten RC, Lupashin VV (2001) Identification of a human orthologue of Sec34p as a component of the cis-Golgi vesicle tethering machinery. J Biol Chem 276: 22810–22818 Sztul E, Lupashin V (2006) Role of tethering factors in secretory membrane traffic. Am J Phys Cell Physiol 290: C11–C26 Tong AH, Lesage G, Bader GD, Ding H, Xu H, Xin X, Young J, Berriz GF, Brost RL, Chang M, Chen Y, Cheng X, Chua G, Friesen H, Goldberg DS, Haynes J, Humphries C, He G, Hussein S, Ke L, Krogan N, Li Z, Levinson JN, Lu H, Menard P, Munyana C, Parsons AB, Ryan O, Tonikian R, Roberts T, Sdicu AM, Shapiro J, Sheikh B, Suter B, Wong SL, Zhang LV, Zhu H, Burd CG, Munro S, Sander C, Rine J, Greenblatt J, Peter M, Bretscher A, Bell G, Roth FP, Brown GW, Andrews B, Bussey H, Boone C (2004) Global mapping of the yeast genetic interaction network. Science 303: 808–813 Ungar D, Oka T, Brittle EE, Vasile E, Lupashin VV, Chatterton JE, Heuser JE, Krieger M, Waters MG (2002) Characterization of a mammalian Golgi-localized protein complex, COG, that is required for normal Golgi morphology and function. J Cell Biol 157: 405–415 Ungar D, Oka T, Vasile E, Krieger M, Hughson FM (2005) Subunit architecture of the conserved oligomeric Golgi complex. J Biol Chem 280: 32729–32735 VanRheenen SM, Cao X, Lupashin VV, Barlowe C, Waters MG (1998) Sec35p, a novel peripheral membrane protein, is required for ER to Golgi vesicle docking. J Cell Biol 141: 1107–1119 VanRheenen SM, Cao X, Sapperstein SK, Chiang EC, Lupashin VV, Barlowe C, Waters MG (1999) Sec34p, a protein required for vesicle tethering to the yeast Golgi apparatus, is in a complex with Sec35p. J Cell Biol 147: 729–742 Vasile E, Oka T, Ericsson M, Nakamura N, Krieger M (2006) IntraGolgi distribution of the Conserved Oligomeric Golgi (COG) complex. Exp Cell Res 312: 3132–3141 Walter DM, Paul KS, Waters MG (1998) Purification and characterization of a novel 13 S hetero-oligomeric protein complex that stimulates in vitro Golgi transport. J Biol Chem 273: 29565–29576 Whyte JR, Munro S (2001) The Sec34/35 Golgi transport complex is related to the exocyst, defining a family of complexes involved in multiple steps of membrane traffic. Dev Cell 1: 527–537 Wu X, Steet RA, Bohorov O, Bakker J, Newell J, Krieger M, Spaapen L, Kornfeld S, Freeze HH (2004) Mutation of the COG complex subunit gene COG7 causes a lethal congenital disorder. Nat Med 10: 518–523 Wuestehube LJ, Duden R, Eun A, Hamamoto S, Korn P, Ram R, Schekman R (1996) New mutants of Saccharomyces cerevisiae affected in the transport of proteins from the endoplasmic reticulum to the Golgi complex. Genetics 142: 393–406 Zerial M, McBride H (2001) Rab proteins as membrane organizers. Nat Rev Mol Cell Biol 2: 107–117 Zolov SN, Lupashin VV (2005) Cog3p depletion blocks vesicle-mediated Golgi retrograde trafficking in HeLa cells. J Cell Biol 168: 747–759

130

*

T. Scanu and C. Wilson

The TRAPP complex Tiziana Scanu and Cathal Wilson

Introduction Intracellular transport of biosynthetic cargo from the endoplasmic reticulum to the plasma membrane occurs by membrane-bound vesicular or tubulovesicular carriers that dissociate from a donor compartment and fuse with an acceptor compartment. The arrival of a transport carrier to the correct destination along the exocytic pathway is important for the appropriate spatio-temporal processing and delivery of the cargo molecules. Membrane fusion requires the formation of a trans SNARE complex (SNAREpin) of SNARE proteins contributed by the donor and acceptor membranes that is thought to overcome the energy barrier that would prevent two membranes from fusing. Prior to this event, however, proteins called tethering factors appear to act as physical links between membrane compartments (Whyte and Munro 2002). In addition to acting as physical links the tethering factors may contribute to the specificity of compartmental fusion by their interaction with various molecules on the donor and acceptor membranes. Two broad classes of tethering factors are represented by proteins containing extensive coiled-coil domains (such as p115/Uso1 in mammals/yeast) or by a number of large multiprotein complexes that mediate membrane traffic between various compartments within the cell (Whyte and Munro 2002). A common feature of the tethering factors seems to be their interaction with small GTPases of the Rab/Ypt family and with SNAREs that appears to contribute to the specificity of membrane fusion.

Transport protein particle (TRAPP) TRAPP is an evolutionarily conserved multisubunit complex that has been proposed to act as a tethering factor. Two TRAPP complexes have been identified in yeast cells: TRAPP I is 300 kDa in size and contains seven subunits while TRAPP II is 1,000 kDa and contains all of the TRAPP I subunits plus an additional three subunits (Table 1). All of the subunits except Trs33, Trs65 and Trs85 are required for yeast viability (Sacher et al. 2000). Similar analyses of human cell lysates identified a single complex of 670 kDa (Sacher et al. 2000). Phylogenetic analyses (Cox et al. 2007) suggest that the human TRAPP I complex contains only six subunits (Table 1). It is still unsure whether a TRAPP II complex exists in mammals: putative orthologues of Trs120 (NBIP) and Trs130 (TMEM1/EHOC1) have been identified based on sequence

The TRAPP complex

*

131

Table 1. Subunits of the TRAPP complex in yeast and mammals Yeast TRAPP I/II Bet3 Bet5 Trs20 Trs23 Trs31 Trs33 Trs85/Gsg1 TRAPP II Trs65/Kre11 Trs120 Trs130

MW (Da) 22,129 18,434 19,700 24,863 31,721 30,749 80,494 63,367 147,660 128,129

Mammals

MW (Da)

TRAPPC3/hBet3 TRAPPC1/MUM2 TRAPPC2/SEDL TRAPPC4/Synbindin TRAPPC5 TRAPPC6A/6B –

20,274 16,831 16,444 24,340 20,783 18,932/17,982 –

– NIBP TMEM1/EHOC1

– 139,422 142,189

comparisons (Cox et al. 2007) and EHOC1 was found as part of a Bet3 complex after tandem-affinity purification (Gavin et al. 2002). For simplicity, the yeast nomenclature is used throughout and readers should refer to Table 1 for the corresponding mammalian homologues.

The architecture of the TRAPP I complex The architecture of the TRAPP I complex has been characterized using X-ray crystallography, single-particle electron microscopy (EM) and analysis of the molecular interactions of six subunits of the TRAPP I complex (Kim et al. 2006). Two TRAPP I subcomplexes were identified in mammals, one a heterotetramer of Bet3–Trs33–Bet5–Trs23 and the other a trimer of Bet3–Trs31–Trs20. By contrast, all of the six yeast subunits (Trs85 was not included) could form a stable complex with a stoichiometry of 2:1:1:1:1:1 (Bet3:Trs33:Bet5:Trs23: Trs31:Trs20) with an apparent size of 170 kDa, in good agreement with the size of the yeast TRAPP I complex previously determined by size exclusion chromatography if the Trs85 subunit is included (see also Table 1). This enabled the determination of the morphology of the yeast TRAPP I complex by single-particle EM, which appeared as a bi-lobed elongated structure. The superimposition of the crystal structures of the mammalian subcomplexes into the density map of the recombinant yeast TRAPP I suggested that the two subcomplexes might interact with each other through contacts between Bet3 and Trs31 of the trimeric subcomplex with Trs23 of the tetrameric subcomplex, although the mammalian subcomplexes could not be shown to interact in vitro. The surface of the TRAPP I complex that is expected to be membraneproximal is rather flat, being formed by the positively charged flat surfaces of two Bet3 subunits and those of Trs31 and Trs33 (Kim et al. 2006; Fig. 1). Trs33 has been proposed to facilitate the recruitment and/or binding of Bet3 into the assembling TRAPP complex where a Bet3/Trs33 heterodimer rather than a

132

*

T. Scanu and C. Wilson

Figure 1. The TRAPP I complex appears as an elongated flat structure lying flat on the membrane. Membrane binding is facilitated by positively charged surfaces of Bet3–Trs33 and Bet3–Trs31. Additional as yet unidentified proteins might act as anchors for binding TRAPP to membranes.

Bet3 homodimer is the functional unit of the TRAPP complex (Kim et al. 2005). The formation of the Bet3–Trs33 complex is followed by the recruitment of the Bet5 subunit to this same complex (Kim et al. 2005; Kummel et al. 2006).

TRAPP localization The localization of TRAPP I was determined by analysis of yeast lysate fractions on sucrose density gradients that were capable of resolving subcompartments of the Golgi complex. TRAPP I is stably anchored to the Golgi complex through the Bet3 subunit, which essentially lies flat on these membranes. Mutagenesis of two lysine residues that form part of the positively charged flat surface of Bet3 that is expected to interact with membranes led to the dissociation of the TRAPP complex from membranes (Sacher et al. 2000). TRAPP I co-fractionates with markers of early Golgi compartments (Sacher et al. 2000). The TRAPP complex appears to be stably associated with Golgi membranes, even when anterograde transport is blocked which leads to disassembly of the Golgi structure (Sacher et al. 2000). A small amount of Bet3 is also present in the cytosol of yeast cells as the addition of Bet3-containing cytosol reconstituted transport activity in Bet3-depleted vesicle fusion assays (Barrowman et al. 2000). In mammalian cells, by contrast, significant free pools of monomeric Bet3 and Sedlin (Trs20) are present in addition to the 670-kDa TRAPP complex, and the majority of Bet3 is cytosolic (Loh et al. 2005). In addition to the cytosolic pool of Bet3, a perinuclear localization was also found in several mammalian cells lines (Yu et al. 2006) which largely co-localized with the tER marker Sec31. This tER-localization of Bet3 was confirmed by immuno-EM (Yu et al. 2006). In both mammalian and yeast cells the TRAPP complex is resistant to extraction by Triton X-100, suggesting that it is anchored to a Triton X-100resistant matrix in the cell (Sacher et al. 2000). Kummel et al. (2006) have shown that Bet3 has a strong self-palmitoylating activity but this is not

The TRAPP complex

*

133

required for membrane localization of Bet3 or TRAPP, nor for yeast cell viability. So far the physiological role of this modification remains unclear, although it appears to be required for protein stability and interaction with Trs33 (Kummel et al. 2006).

TRAPP I is involved in ER to Golgi transport A role for the TRAPP complex in transport from the ER to the Golgi emerged from the observation that a temperature-sensitive mutant of yeast Bet3 (bet3-1) failed to transport the soluble cargo proteins invertase, a-factor, and carboxypeptidase Y (CPY) from the ER to the Golgi complex (Rossi et al. 1995). However, the block of ER-to-Golgi transport in this mutant was not complete, since small amounts of p10 CPY (the early Golgi form) and of the Golgi form of invertase accumulated inside the cells (Rossi et al. 1995; Sacher et al. 2001), indicating that the cargo proceeded past an initial block between the ER and the Golgi complex and was then transported to the Golgi. In mammalian cells, antibodies against Bet3 inhibited transport of the viral glycoprotein VSV-G in a dose-dependent manner in a semi-intact cell system that reconstitutes ER-to-Golgi transport. Moreover, a similar defect in VSV-G transport caused by cytosol that was immuno-depleted of Bet3 could be rescued by the addition of recombinant GST-Bet3 (Loh et al. 2005). Inactivation of Bet3 in vivo, by microinjecting an anti-Bet3 antibody into cells, resulted in a 70% decrease in trafficking of YFP-tagged VSV-G to the cell surface (Yu et al. 2006), although this assay did not address at which stage of the secretory pathway transport had been blocked. Transfection of the plasma membrane protein CD8 into BSC-1 cells that had been microinjected with an anti-Bet3 antibody resulted in the co-localization of the CD8 protein with the COP II component Sec31, but adjacent to the Golgi membrane marker Golgin-84 (in control cells CD8 was found at the cell surface as well as co-localizing with Golgin-84), suggesting that a block occurred at ER exit sites. Moreover, the knockdown of Bet3 using Bet3-siRNA resulted in the dispersion of the tER sites and disruption of the Golgi architecture (Yu et al. 2006). Additionofananti-Bet3antibodytopermeabilizedcellsinvitroincreasedthe release of vesicles, indicating that Bet3 is not required for vesicle budding (Yu et al. 2006). Using a two-stage transport assay with combinations of cytosol depleted or not of various essential components known to act in ER to Golgi transport together with Bet3-depleted cytosol indicated that Bet3 acts after COPII but before Rab1 and a-SNAP in ER-to-Golgi transport (Loh et al. 2005). Bet3 depletion did not affect budding efficiency in yeast cells (Morsomme and Riezman 2002; Sacher et al. 1998) but was required for the fusion of vesicles with the Golgi. When Bet3-depleted Golgi membranes were incubated with vesicles in the presence of Bet3-depleted cytosol, the vesicles failed to fuse with the Golgi (Barrowman et al. 2000). Addition of Bet3-containing cytosol to this reaction, or incubation of Bet3-depleted cytosol with Bet3-containing Golgi membranes, resulted in vesicle fusion. Together the data point to a role

134

*

T. Scanu and C. Wilson

for the TRAPP I complex in a transport step after exit of vesicles from the ER and during or before fusion with the acceptor compartment.

TRAPP I acts as a tethering factor The TRAPP I complex could function in tethering membranes before fusion or be required during the fusion step itself. The observation that vesicles could bind to, but not fuse with, Golgi membranes in the presence of an anti-Bos1 (SNARE) antibody while Bet3 depletion prevented both binding and fusion indicated that the TRAPP complex acts as a tethering factor (Barrowman et al. 2000). Purified TRAPP I was found to bind ER-derived COP II vesicles in vitro (Sacher et al. 2001). Binding was drastically reduced by the addition of GTPgS, a non-hydrolyzable analogue of GTP that is known to stabilize the COP II coat, and no interaction of COP II components with TRAPP I was found leading to the suggestion that COP II vesicles must uncoat before TRAPP I can bind (Sacher et al. 2001). Subsequently, however, the same group showed that yeast TRAPP I tethers COP II vesicles via an interaction with Sec23p, a component of the COP II coat, suggesting that vesicle tethering precedes vesicle uncoating (Cai et al. 2007). Furthermore, they found that TRAPP I binds efficiently to vesicles formed in the presence of GTPgS, and TRAPP I failed to bind vesicles when they were striped of their coat by sedimentation through a sorbitol cushion, although they were still fusion competent. The addition of excess GST-Sec23 blocked vesicle tethering in mammalian homotypic and yeast ER-to-Golgi COP II tethering assays, and Sec23 and the COP II cage component Sec31 were detected on the co-isolated mammalian vesicles (Cai et al. 2007). In addition, immuno-EM showed that Bet3 and Sec31 reside on the same vesicles (Yu et al. 2006). These new data suggest that TRAPP I acts as a tethering factor. It had been thought that the COP II coat is removed soon after vesicle formation (Bonifacino and Glick 2004) but it now appears that the COP II coat is longer-lived than previously supposed since ER-derived vesicles should recognize and bind to their target membrane before uncoating. To what extent TRAPP I contributes to uncoating remains to be clarified.

The nucleotide exchange activity of the yeast TRAPP I complex The Ypt/Rab small GTP-binding proteins cycle between their GDP-bound and GTP-bound forms through nucleotide exchange and GTP hydrolysis, respectively. The yeast Ypt1 protein (a homologue of mammalian Rab1) regulates ER-to-Golgi membrane trafficking and interacts genetically with the Bet3 and Bet5 genes, and its over-expression suppresses the bet3-1 mutation (Jiang et al. 1998; Rossi et al. 1995). The yeast TRAPP complex can stimulate guanine nucleotide exchange on Ypt1p suggesting a role for TRAPP as a guanine nucleotide-exchange factor (GEF) for Ypt1 (Wang et al. 2000). Biochemical analysis confirmed that purified TRAPP can bind to the nucleotide-free form

The TRAPP complex

*

135

of Ypt1, and can stimulate GDP release from Ypt1–GDP (Wang et al. 2000). The minimal TRAPP subunit composition that is needed for GEF activity is composed of Bet5–Trs23–Bet3–Trs31 (Kim et al. 2006). Although Ypt1 is found in COP II vesicles, it may be the Golgi-located Ypt1 that is important in the fusion of vesicles with the acceptor compartment (Cao and Barlowe 2000). This would be in keeping with the Golgi localization of the yeast TRAPP complex and place activation of Ypt1 after vesicle tethering by TRAPP. No such GEF activity has been demonstrated so far for the mammalian TRAPP complex. Ypt1 appears to function in the same process as the tethering factor Uso1 and may regulate Uso1 binding to membranes (Cao et al. 1998; Sapperstein et al. 1996). Similarly, Rab1 appears to function prior to p115 (the mammalian homologue of Uso1) and activated Rab1 recruits p115 to transport vesicles (Allan et al. 2000; Alvarez et al. 1999) which may also be regulated by unassembled SNAREs (Brandon et al. 2006). In addition, activated Rab1 can recruit other tethering/docking factors such as GRASP65 and GM130 to COP II vesicles independently of p115 (Moyer et al. 2001). Uso1, in turn, is required for the assembly of the SNAREpin (Sapperstein et al. 1996). Therefore, activation of a small GTPase on vesicles (in mammalian cells) or on the Golgi (in yeast) appears to be necessary for the recruitment of tethers that are required for SNAREpin assembly, leading to the formation of the VTC in mammals or fusion with the Golgi in yeast. Based on the above data, a model of how TRAPP I might function in ER to Golgi transport in yeast and mammals is presented in Fig. 2. TRAPP I binds to COP II-coated vesicles, activates Ypt1/Rab1 which results in Uso1/p115 recruitment from the cytosol followed by SNAREpin formation and finally membrane fusion.

The TRAPP II complex The yeast TRAPP II complex co-fractionates with late Golgi/early endosomal markers on sucrose gradients (Cai et al. 2005). Mutants in Trs120 were reported to mislocalize COPI subunits and a specific, albeit weak, interaction between TRAPP (purified using the Bet3-tagged subunits that are common to TRAPP I and TRAPP II) and COPI was demonstrated in vitro (Cai et al. 2005). However, a mutation in Trs120 blocked neither invertase secretion nor trafficking of CPY, but showed a defect in the recycling of the R-SNARE Snc1 (which resides on the plasma membrane and moves through the early endosomes before reaching the late Golgi) and of the Chs3 enzyme (which has a recycling pathway similar to Snc1p). Thus, Trs120 appeared to be involved in trafficking from early endosomes to the late Golgi (Cai et al. 2005). The trafficking of GFP–Snc1 and Chs3–GFP was also defective in trs130 mutants, where, in contrast to trs120 mutants, invertase and CPY transport were also affected. Therefore, Trs130 may be required at multiple steps on the secretory and endocytic pathways whereas Trs120 appears to be required only in the endocytic pathway. Consistent with TRAPP II having a function at the

136

*

T. Scanu and C. Wilson

Figure 2. A model for TRAPP function in ER-Golgi transport. Cargo exits the ER in COP II-coated vesicles. The COP II coat consists of an inner layer of Sec23/Sec24 surrounded by a cage formed by the Sec13/Sec31 proteins. Bet3 interacts directly with Sec23, leading to the tethering of vesicles with the cis-Golgi membrane (in yeast) or to the tethering of homotypic vesicles (in mammals). Following binding of TRAPP to Sec23 the COP II coat is released from the vesicles. Activation of the small GTPase Rab1 (Ypt1 in yeast) then occurs through the guanine nucleotideexchange factor activity of TRAPP. Rab1 recruits the docking protein p115 (Uso1 in yeast). p115/ Uso1 is required for the recruitment of other tethering/docking factors such as GRASP65 and GM130 (not shown) and for SNAREpin formation. The SNAREpin, which is required for membrane fusion, is then resolved by NSF (Sec18 in yeast).

late Golgi are the genetic and functional interactions of Trs130, Trs33, and Trs65 with the small GTPase Ypt31 and the phosphatidylinositol-4-kinase Pik1 (Sciorra et al. 2005; Yamamoto and Jigami 2002; Zhang et al. 2002), both of which regulate exocytic/endocytic trafficking at the trans-Golgi. More recently, Morozova et al. (2006) showed that the two essential TRAPP IIspecific subunits Trs120 and Trs130 were required for switching the GEF specificity of TRAPP from Ypt1 to Ypt31. Trs130 is required for GEF activity on Ypt31 but not Ypt1. Indeed, a temperature-sensitive mutant or the deletion of Trs130 resulted in a higher GEF activity of TRAPP towards Ypt1 (although the opposite effect was observed by Wang et al. 2000). This suggested a model where TRAPP I acts as a GEF for Ypt1 while attachment ofTRAPP II-specific subunits toTRAPP I at the trans-Golgi switch its GEF specificity from Ypt1 to Ypt31/32 thus regulating entry into and exit from the Golgi complex. The Trs65 subunit, which is only found in fungi, might contribute to the assembly and/or stability of TRAPP II, and thereby to the GEF activity of this complex (Liang et al. 2007).

The TRAPP complex

*

137

Other functions of TRAPP complex subunits Much of the data described above have been obtained through the analysis of the function of the Bet3 subunit. As Bet3 is considered to be a central component of the TRAPP complex, these findings have been extrapolated to predict the function of the TRAPP complex as a whole. However there is evidence that functions associated with the TRAPP complex or some of its subunits may have roles other than tethering transport vesicles. Spondyloepiphyseal dysplasia tarda (SEDT) is a late onset skeletal disorder, which manifests in childhood. It is caused by mutations in the sedlin gene (SEDL) located on Xp22.12–p22.31 (Gedeon et al. 1999), which encodes a subunit of the mammalian TRAPP complex (TRAPPC2, a homologue of yeast Trs20). The human sedlin gene can complement the lethality caused by cz et al. deletion of the Trs20 gene, illustrating conservation of function (Ge 2003). Most of the disease-associated mutations that have been identified throughout the sedlin gene are splice alterations, insertions and deletions that can lead to premature termination signals and truncated proteins (Gedeon et al. 2001), while only four missense mutations have been described. Although the different SEDL mutations display the same clinical phenotype, expression in yeast of three of the four missense mutations that are associated with the development of SEDT have shown that only one of the mutations could not complement the yeast deletion, while the other two led to very subtle effects on yeast growth (Gecz et al. 2003). Sedlin occupies a peripheral position in the TRAPP complex and its yeast homologue Trs20 is not required for the GEF activity of the TRAPP complex (Kim et al. 2006). The crystal structure of sedlin revealed that, despite having no sequence homology, it has a folded structure similar to the N-terminal regulatory domain of two SNAREs, Ykt6 and Sec22, suggesting a possible interaction between sedlin and SNAREs (Jang et al. 2002). In addition, Trs20 is substoichiometric with respect to the other subunits in the TRAPP I complex but is present in equimolar amounts in TRAPP II (Sacher et al. 2001). However, despite these structural insights into the sedlin protein, its function remains completely unknown. Although Trs20 has been reported to be involved in ER to Golgi transport, as assayed by repression of galactose-driven Trs20 gene expression (Sacher et al. 2001), similar strong shut-down of sedlin or Trs20  expression in yeast cells led to no observable phenotype (Belgareh-Touze et al. 2003; Gecz et al. 2003). An isoform of SEDL, called SEDLP1 or MIP-2A, was isolated via yeast twohybrid analysis. MIP-2A interacted with the c-myc promoter-binding protein 1 (MBP-1) and relieved the activity of MBP-1 as a transcriptional repressor (Ghosh et al. 2001). MBP-1 negatively regulates both human and mouse c-myc promoter activity and its ectopic expression induces cell death, reduction of cmyc expression and regression of tumour growth. In addition, it has been reported that MIP-2A can regulate transcription by inhibiting the transactivation of the luteinizing hormone-b promoter that is mediated by the SF-1

138

*

T. Scanu and C. Wilson

nuclear receptor and by the pituitary homeobox 1 protein (Ghosh et al. 2003). Thus, the interaction with MBP-1 raises the intriguing possibility of a dual function for sedlin, both in membrane trafficking and in the regulation of gene expression. The function of the sedlin/Trs20 protein and how it operates in membrane trafficking, whether it has other functions, and how these result in the development of SEDT remain open questions. The syndecan-binding protein synbindin (TRAPPC4), the human homologue of yeast Trs23, is a neuronal cytoplasmic protein that was identified by two-hybrid screening using the syndecan-2 cytoplasmic domain as a bait (Ethell et al. 2000). Syndecan-2 is a transmembrane heparan sulfate proteoglycan that is mainly found in fibroblasts and appears to be involved in the control of ECM assembly. Immuno-EM has demonstrated that synbindin is present on the Golgi cisternae, and also in vesicles within the soma and dendrites of cortical neurons. Syndecan-2 might induce spine formation by recruiting intracellular vesicles through its interaction with synbindin. Trs23 is one of the subunits of the tetramer that exhibits GEF activity, and mutation of Trs23 abolished this activity. Thus, the recruitment of synbindin by syndecan-2 might direct trafficking to maturing spines and postsynaptic structures (Ethell et al. 2000). It cannot be excluded that a role of synbindin in neuronal spine formation via its interaction with syndecan-2 represent a specific function that occurs outside of the TRAPP complex. The non-essential TRAPP subunit Trs85 is, like Trs20, substoichiometric with respect to the other subunits in the TRAPP I complex. Although deletion of the Trs85 gene was reported to lead to an ER-to-Golgi transport defect (Sacher et al. 2001), subsequent analyses showed that deletion of Trs85 does not affect CPY transport to the vacuole and that Trs85 might play a role in the biogenesis of Cvt vesicles (cytoplasm to vacuole targeting) during autophagy (Meiling-Wesse et al. 2005; Nazarko et al. 2005).

Concluding remarks and perspectives The TRAPP complex has been highly conserved throughout evolution (Koumandou et al. 2007) and has emerged as a key player in the trafficking of cargo between the ER and the Golgi. Binding of TRAPP to the COP II coat would ensure vesicle recognition and specificity of targeting to the correct acceptor compartment membrane. The GEF activity of TRAPP could co-ordinate the activation of Rab1/Ypt1 with the recruitment of the coiled-coil tether p115/Uso1p that is required for subsequent SNAREpin formation and thus membrane fusion. However, many questions remain about the mechanisms and functions of the TRAPP complex. COP II vesicles are surrounded by an inner layer composed of Sec23, Sec24 and Sar1 which in turn are encased in a cage of Sec13 and Sec31 (Fath et al. 2007). The vesicles isolated from mammalian cells that were used in vesicle tethering assays were shown to contain two of these components, Sec23 and Sec31 (Cai et al. 2007). It has been proposed that the elongated structure of

The TRAPP complex

*

139

the TRAPP I lies flat on the membrane (Kim et al. 2006). If this is the case, it is difficult to envisage how Bet3 could interact with the inner layer of a COP IIcoated vesicle. If the TRAPP complex is attached to an acceptor compartment membrane only on one side via one of the Bet3 molecules, could the other Bet3 molecule of the complex then insert into the open cage to interact with Sec23, or is partial uncoating required before interaction? How would the membrane binding of TRAPP fit in with such a scenario in tethering COP II vesicles via Sec23 in mammalian cells? Furthermore, it has been proposed that vesicular–tubular carriers exit from the ER and transport does not involve homotypic vesicle fusion (Mironov et al. 2003). Such tubular carriers, although dependent on COP II, may exit without a surrounding COP II coat and eventually give rise to the VTC. What is the function of TRAPP, and Bet3 binding to Sec23, in the formation and/or fusion of these transport carriers? There may be differences in the function of the TRAPP complex, or its subunits, in yeast and mammalian cells. Somewhat surprisingly, human Bet3 cannot complement deletion of the Bet3 gene in yeast (Turnbull et al. 2005), despite being the most highly conserved of the TRAPP subunits whereas the less conserved sedlin gene complements deletion of Trs20 (Gecz et al. 2003). The majority of the TRAPP complex is cytosolic in mammalian cells that also contain large pools of monomeric Bet3 and sedlin (Trs20), in contrast to yeast, but the functional meaning of this is still unclear. The minimal TRAPP subunit complex with nucleotide exchange activity for Ypt1p in yeast could not be reconstituted with the mammalian paralogues. The role of Bet3 binding to Sec23 is still unknown. Whether it acts simply as a physical tether or can also function in uncoating of the vesicle/carrier, and how these functions are integrated into Rab/Ypt activation and SNARE pairing are open questions. Finally, future consideration should be given to the possibility that some TRAPP subunits, such as Trs85, synbindin (Trs23), and sedlin (Trs20), might have functions outside those of membrane trafficking, and how the presence of additional isoforms of Bet3, Trs33 and sedlin might contribute to the function of the TRAPP complex. Acknowledgements. Work in the author’s laboratory was supported by Marie Curie Action “Transfer of Knowledge” project number 14505 and the Telethon grant number GGP07075. We thank Maria Antonietta De Matteis for critical reading of the manuscript and Elena Fontana for the artwork.

References Allan BB, Moyer BD, Balch WE (2000) Rab1 recruitment of p115 into a cisSNARE complex: programming budding COPII vesicles for fusion. Science 289: 444–448 Alvarez C, Fujita H, Hubbard A, Sztul E (1999) ER to Golgi transport: requirement for p115 at a pre-Golgi VTC stage. J Cell Biol 147: 1205–1222 Barrowman J, Sacher M, Ferro-Novick S (2000) TRAPP stably associates with the Golgi and is required for vesicle docking. EMBO J 19: 862–869

140

*

T. Scanu and C. Wilson

 N, Corral-Debrinski M, Launhardt H, Galan JM, Munder T, Le Panse S, Belgareh-Touze Haguenauer-Tsapis R (2003) Yeast functional analysis: identification of two essential genes involved in ER to Golgi trafficking. Traffic 4: 607–617 Bonifacino JS, Glick BS (2004) The mechanisms of vesicle budding and fusion. Cell 116: 153–166 Brandon E, Szul T, Alvarez C, Grabski R, Benjamin R, Kawai R, Sztul E (2006) On and off membrane dynamics of the endoplasmic reticulum–Golgi tethering factor p115 in vivo. Mol Biol Cell 17: 2996–3008 Cai H, Yu S, Menon S, Cai Y, Lazarova D, Fu C, Reinisch K, Hay JC, Ferro-Novick S (2007) TRAPPI tethers COPII vesicles by binding the coat subunit Sec23. Nature 445: 941–944 Cai H, Zhang Y, Pypaert M, Walker L, Ferro-Novick S (2005) Mutants in trs120 disrupt traffic from the early endosome to the late Golgi. J Cell Biol 171: 823–833 Cao X, Ballew N, Barlowe C (1998) Initial docking of ER-derived vesicles requires Uso1p and Ypt1p but is independent of SNARE proteins. EMBO J 17: 2156–2165 Cao X, Barlowe C (2000) Asymmetric requirements for a Rab GTPase and SNARE proteins in fusion of COPII vesicles with acceptor membranes. J Cell Biol 149: 55–66 Cox R, Chen SH, Yoo E, Segev N (2007) Conservation of the TRAPPII-specific subunits of a Ypt/Rab exchanger complex. BMC Evol Biol 7: 12 Ethell IM, Hagihara K, Miura Y, Irie F, Yamaguchi Y (2000) Synbindin, a novel syndecan-2binding protein in neuronal dendritic spines. J Cell Biol 151: 53–68 Fath S, Mancias JD, Bi X, Goldberg J (2007) Structure and organization of coat proteins in the COPII cage. Cell 129: 1325–1336 € sche M, Krause R, Grandi P, Marzioch M, Bauer A, Schultz J, Rick JM, Gavin AC, Bo € fert C, Schelder M, Brajenovic M, Ruffner H, Michon AM, Cruciat CM, Remor M, Ho Merino A, Klein K, Hudak M, Dickson D, Rudi T, Gnau V, Bauch A, Bastuck S, Huhse B, Leutwein C, Heurtier MA, Copley RR, Edelmann A, Querfurth E, Rybin V, Drewes G, Raida M, Bouwmeester T, Bork P, Seraphin B, Kuster B, Neubauer G, Superti-Furga G (2002) Functional organization of the yeast proteome by systematic analysis of protein complexes. Nature 415: 141–147 Gecz J, Shaw MA, Bellon JR, De Barros Lopes M (2003) Human wild-type SEDL protein functionally complements yeast Trs20p but some naturally occurring SEDL mutants do not. Gene 320: 137–144 Gedeon AK, Colley A, Jamieson R, Thompson EM, Rogers J, Sillence D, Tiller GE, Mulley JC, Gecz J (1999) Identification of the gene (SEDL) causing X-linked spondyloepiphyseal dysplasia tarda. Nat Genet 22: 400–404 Gedeon AK, Tiller GE, Le Merrer M, Heuertz S, Tranebjaerg L, Chitayat D, Robertson S, Glass IA, Savarirayan R, Cole WG, Rimoin DL, Kousseff BG, Ohashi H, Zabel B, Munnich A, Gecz J, Mulley JC (2001) The molecular basis of X-linked spondyloepiphyseal dysplasia tarda. Am J Hum Genet 68: 1386–1397 Ghosh AK, Majumder M, Steele R, White RA, Ray RB (2001) A novel 16-kilodalton cellular protein physically interacts with and antagonizes the functional activity of c-myc promoter-binding protein 1. Mol Cell Biol 21: 655–662 Ghosh AK, Steele R, Ray RB (2003) Modulation of human luteinizing hormone beta gene transcription by MIP-2A. J Biol Chem 278: 24033–24038 Jang SB, Kim YG, Cho YS, Suh PG, Kim KH, Oh BH (2002) Crystal structure of SEDL and its implications for a genetic disease spondyloepiphyseal dysplasia tarda. J Biol Chem 277: 49863–49869 Jiang Y, Scarpa A, Zhang L, Stone S, Feliciano E, Ferro-Novick S (1998) A high copy suppressor screen reveals genetic interactions between BET3 and a new gene. Evidence for a novel complex in ER-to-Golgi transport. Genetics 149: 833–841 Kim YG, Raunser S, Munger C, Wagner J, Song YL, Cygler M, Walz T, Oh BH, Sacher M (2006) The architecture of the multisubunit TRAPP I complex suggests a model for vesicle tethering. Cell 127: 817–830

The TRAPP complex

*

141

Kim YG, Sohn EJ, Seo J, Lee KJ, Lee HS, Hwang I, Whiteway M, Sacher M, Oh BH (2005) Crystal structure of bet3 reveals a novel mechanism for Golgi localization of tethering factor TRAPP. Nat Struct Mol Biol 12: 38–45 Koumandou VL, Dacks JB, Coulson RM, Field MC (2007) Control systems for membrane fusion in the ancestral eukaryote; evolution of tethering complexes and SM proteins. BMC Evol Biol 7: 29 Kummel D, Muller JJ, Roske Y, Henke N, Heinemann U (2006) Structure of the Bet3–Tpc6B core of TRAPP: two Tpc6 paralogs form trimeric complexes with Bet3 and Mum2. J Mol Biol 361: 22–32 Liang Y, Morozova N, Tokarev AA, Mulholland JW, Segev N (2007) The role of Trs65 in the Ypt/Rab guanine nucleotide exchange factor function of the TRAPP II complex. Mol Biol Cell 18: 2533–2541 Loh E, Peter F, Subramaniam VN, Hong W (2005) Mammalian Bet3 functions as a cytosolic factor participating in transport from the ER to the Golgi apparatus. J Cell Sci 118: 1209–1222 Meiling-Wesse K, Epple UD, Krick R, Barth H, Appelles A, Voss C, Eskelinen EL, Thumm M (2005) Trs85 (Gsg1), a component of the TRAPP complexes, is required for the organization of the preautophagosomal structure during selective autophagy via the Cvt pathway. J Biol Chem 280: 33669–33678 Mironov AA, Mironov AA Jr, Beznoussenko GV, Trucco A, Lupetti P, Smith JD, Geerts WJ, Koster AJ, Burger KN, Martone ME, Deerinck TJ, Ellisman MH, Luini A (2003) ER-toGolgi carriers arise through direct en bloc protrusion and multistage maturation of specialized ER exit domains. Dev Cell 5: 583–594 Morozova N, Liang Y, Tokarev AA, Chen SH, Cox R, Andrejic J, Lipatova Z, Sciorra VA, Emr SD, Segev N (2006) TRAPPII subunits are required for the specificity switch of a Ypt–Rab GEF. Nat Cell Biol 8: 1263–1269 Morsomme P, Riezman H (2002) The Rab GTPase Ypt1p and tethering factors couple protein sorting at the ER to vesicle targeting to the Golgi apparatus. Dev Cell 2: 307–317 Moyer BD, Allan BB, Balch WE (2001) Rab1 interaction with a GM130 effector complex regulates COPII vesicle cis-Golgi tethering. Traffic 2: 268–276 Nazarko TY, Huang J, Nicaud JM, Klionsky DJ, Sibirny AA (2005) Trs85 is required for macroautophagy, pexophagy and cytoplasm to vacuole targeting in Yarrowia lipolytica and Saccharomyces cerevisiae. Autophagy 1: 37–45 Rossi G, Kolstad K, Stone S, Palluault F, Ferro-Novick S (1995) BET3 encodes a novel hydrophilic protein that acts in conjunction with yeast SNAREs. Mol Biol Cell 6:1769–1780 Sacher M, Barrowman J, Schieltz D, Yates JR III, Ferro-Novick S (2000) Identification and characterization of five new subunits of TRAPP. Eur J Cell Biol 79: 71–80 Sacher M, Barrowman J, Wang W, Horecka J, Zhang Y, Pypaert M, Ferro-Novick S (2001) TRAPP I implicated in the specificity of tethering in ER-to-Golgi transport. Mol Cell 7: 433–442 Sacher M, Jiang Y, Barrowman J, Scarpa A, Burston J, Zhang L, Schieltz D, Yates JR III, Abeliovich H, Ferro-Novick S (1998) TRAPP, a highly conserved novel complex on the cis-Golgi that mediates vesicle docking and fusion. EMBO J 17: 2494–2503 Sapperstein SK, Lupashin VV, Schmitt HD, Waters MG (1996) Assembly of the ER to Golgi SNARE complex requires Uso1p. J Cell Biol 132: 755–767 Sciorra VA, Audhya A, Parsons AB, Segev N, Boone C, Emr SD (2005) Synthetic genetic array analysis of the PtdIns 4-kinase Pik1p identifies components in a Golgi-specific Ypt31/rab-GTPase signaling pathway. Mol Biol Cell 16: 776–793 € mmel D, Prinz B, Holz C, Schultchen J, Lang C, Niesen FH, Hofmann KP, Turnbull AP, Ku € ck H, Behlke J, Mu € ller EC, Jarosch E, Sommer T, Heinemann U (2005) Structure Delbru

142

*

T. Scanu and C. Wilson

of palmitoylated BET3: insights into TRAPP complex assembly and membrane localization. EMBO J 24: 875–884 Wang W, Sacher M, Ferro-Novick S (2000) TRAPP stimulates guanine nucleotide exchange on Ypt1p. J Cell Biol 151: 289–296 Whyte JR, Munro S (2002) Vesicle tethering complexes in membrane traffic. J Cell Sci 115: 2627–2637 Yamamoto K, Jigami Y (2002) Mutation of TRS130, which encodes a component of the TRAPP II complex, activates transcription of OCH1 in Saccharomyces cerevisiae. Curr Genet 42: 85–93 Yu S, Satoh A, Pypaert M, Mullen K, Hay JC, Ferro-Novick S (2006) mBet3p is required for homotypic COPII vesicle tethering in mammalian cells. J Cell Biol 174:359–368 Zhang CJ, Bowzard JB, Greene M, Anido A, Stearns K, Kahn RA (2002) Genetic interactions link ARF1, YPT31/32 and TRS130. Yeast 19: 1075–1086

Ca2þ in intracellular transport regulation

*

143

The role of Ca2þ in the regulation of intracellular transport Massimo Micaroni, Alexander A. Mironov and Rosario Rizzuto

It is known that cytosolic Ca2þ is essential for signalling, cell-cycle function, cell growth, cell death and other functions (Gill et al. 1996; Orrenius et al. 2003). Recent evidence has also highlighted the functional importance of Ca2þ for intracellular trafficking, and in particular for the fusion of endomembranes, with further functions inside the lumen of Golgi cisternae and other endomembranes. In this chapter we will discuss these latter roles of Ca2þ mechanisms in the regulation of intracellular trafficking (see also Chapter 2.1).

Ca2þ and membrane fusion What is the role of Ca2þ in intracellular transport? One of the possible mechanisms is the participation of Ca2þ in the final stages of membrane fusion (see Chapter 2.1) during cargo transport along the secretory (Porat and Elazar 2000; Sorrentino and Rizzuto 2001; Chen et al. 2002) and endocytic (Lauvrak et al. 2002) pathways. For instance, chelation of extracellular Ca2þ with EGTA, which is not membrane permeant, blocks transport at a very early stage: during endoplasmic reticulum (ER)-to-Golgi transport when carriers arrive at the Golgi complex (GC; Pind et al. 1994). Ca2þ is required at a late step preceding the fusion of vesicular–tubular clusters (VTCs) to the Golgi stack, but not for vesicle budding from the ER (Pind et al. 1994; Aridor et al. 1995). Chelation of extracellular Ca2þ with EGTA inhibited VSV-G transport to the GC when added together with a calcium ionophore (Chen et al. 2002). It has been established that Ca2þ is necessary for the final step of SNAREdriven fusion. Ca2þ acts downstream of the SNAREs, and in the absence of Ca2þ, fusion between membranes is very slow (Coorssen et al. 1998; Cho et al. 2002; Jeremic et al. 2004). Thus, to facilitate membrane fusion, the cell cytosol should contain basal-free Ca2þ concentrations ([Ca2þ]). Indeed, Ca2þ is one of the most abundant cations in vertebrates: in the extracellular space, the [Ca2þ] is about 2 mM. However, eukaryotic cells sequester Ca2þ efficiently, taking it up into intracellular stores or pumping it outside of the cell, and thus reducing the basal cytosolic (intracellular) [Ca2þ] ([Ca2þ]i) to 50–100 nM (Brostrom and Brostrom 1990; Pozzan et al. 1994; Michelangeli et al. 2005; Dolman and Tepikin 2006). However, it appears that at steady-state, even this low [Ca2þ]i is sufficient to promote many constitutive membrane fusion reactions along the secretory pathway (Beckers and Balch 1989; Baker et al. 1990; Ivessa et al. 1995; Porat and Elazar

144

*

M. Micaroni et al.

2000; Ahluwalia et al. 2001; Stojilkovic 2005). For instance, COPI-dependent vesicles fuse with the Golgi cisternae at steady-state without apparent changes in the [Ca2þ]i (Kweon et al. 2004). However, several reports show the need for transient increases in [Ca2þ]i for the execution of several other transport steps (Sullivan et al. 1993; Peters and Mayer 1998; Chen et al. 1999; Pryor et al. 2000; Reddy et al. 2001; Rettig and Neher 2002). For instance, intra-Golgi transport needs rapid elevations in [Ca2þ]i (Chen et al. 2002). Elevation of [Ca2þ]i over basal levels is also required for regulated exocytosis, as an increase in the [Ca2þ]i triggers the exocytosis of storage granules and of synaptic vesicles (Burgoyne and Morgan 1995, 1998; see Chapter 3.11). Transient, but significant, [Ca2þ] gradients across the GC have also been described (Dolman et al. 2005; Dolman and Tepikin 2006). However, there have also been contrasting reports. For instance, there is no obvious accumulation of VSV-G in the GC or in a pre-Golgi compartment upon treatment with the acetoxymethyl ester of EGTA (EGTA-AM). This compound can enter the cell and thus chelate cytosolic Ca2þ. Thus VSV-G was seen to be endo-H-resistant regardless of when EGTA-AM was added in relation to the switch to the permissive temperature for transport (Chen et al. 2002). There is also evidence that suggests that bilayer fusion can proceed normally in the presence of EGTA and BAPTA (Flanagan and Barlowe 2006). Retrograde membrane trafficking from the GC to the ER under the action of the fungal toxin brefeldin A (BFA) critically depends on luminal Ca2þ (Ivessa et al. 1995); this study suggested that in the absence of luminal Ca2þ, there is no fusion between the COPI-positive Golgi cisternae and the trans-Golgi network (TGN). Retrograde PM-to-ER transport of the Shiga toxin-B (STB) subunit is blocked when Ca2þ is chelated. However, internalization of STB into an endosomal compartment is not affected by the calcium chelation (Chen et al. 2002). When BAPTA-AM (the permeable chelator of Ca2þ) was added prior to STB, the toxin was internalized but appeared to accumulate in punctate endosomal compartments. Although the GC had a normal morphology under these conditions, co-localization of STB with Golgi markers was never seen. STB remained non-glycosylated here, suggesting inhibition of endosome-to-Golgi transport in the presence of the Ca2þ chelator BAPTA-AM (Chen et al. 2002). The simple incubation of cells in a Ca2þ-free medium at 37 C often results in a substantial modification of the structure of the GC, with a fragmentation of Golgi cisternae that is possibly due to a block of membrane fusion. However, at 4 C cells can maintain an intact Golgi morphology when they have been treated with Ca2þ-depletion protocols. However, when Ca2þdepleted cells at 4 C were warmed to 37 C and Ca2þ (or Sr2þ) was re-added to the medium, the appearance of the GC was not modified for at least the first 30 min (Pinton et al. 1998). Thus, only lipid membranes in a fluid state participate in Ca2þ-driven fusion.

Ca2þ in intracellular transport regulation

*

145

In the cytosol, there are various Ca2þ-binding proteins that help in the sequestering of Ca2þ (Maki et al. 2002; Shibata et al. 2007). One of these, calmodulin (CaM), has an important role in fusion between yeast vacuoles (Peters and Mayer 1998), and in endosome fusion in animal cells (Colombo et al. 1997). Strong inhibition of intra-Golgi transport by a CaM antagonist suggested that CaM can regulate various intracellular fusion processes in a Ca2þ-dependent manner (Porat and Elazar 2000). As indicated above, eukaryotic cells maintain a low basal [Ca2þ]i (50–100 nM). This basal [Ca2þ]i is sufficient for some transport reactions, but not for others. Thus the basal [Ca2þ]i is enough for fusion of COPI-dependent vesicles and ER-to-Golgi carriers with the GC, while local changes in [Ca2þ]i could regulate the fusion of organelles within the exocytic and endocytic pathways, ensuring the directionality of these processes (Fig. 1). Cytosolic Ca2þ-binding proteins would thus regulate this process.

Ca2þ in the lumen of the ER and Golgi cisternae If membrane fusion within some specific stages of intracellular transport needs an elevation of [Ca2þ]i, the question is what are the sources of this Ca2þ? One source is the extracellular space. However, the GC and most of the endosomes are situated far away from the plasma membrane (PM). Thus, increases in [Ca2þ]i should arise from internal sources that are suitable for the fast release of Ca2þ and are situated near to a place of fusion. In eukaryotic cells, there are several stores of Ca2þ that are suitable for this task. In mammalian cells, the best-characterized Ca2þ store is the ER. The free [Ca2þ] in the ER ([Ca2þ]ER)has been estimated to be as high as 0.7 mM (Meldolesi and Pozzan 1998). In most cells, in addition to the ER, the GC has been shown to be a Ca2þ store that can sequester high levels of Ca2þ (Chandra et al. 1994; Pezzati et al. 1997) through the Ca2þ binding to proteins in the Golgi lumen (Lin et al. 1999). Indeed, it has been estimated that the GC can store up to 5% of the total intracellular Ca2þ (Chandra et al. 1991). In different reports, the total [Ca2þ] in the GC ([Ca2þ]GC) has varied from 1–2 mM (Chandra et al. 1991) to >10 mM (Pezzati et al. 1997). However, targeting aequorin to the GC to measure the [Ca2þ]GC of the Golgi cisternae has demonstrated that in the cisternal lumen of unstimulated cells the free [Ca2þ] is around 0.3 mM; therefore, a gradient in [Ca2þ] exists between the lumen of the Golgi and the cytosol (Pinton et al. 1998). These storage compartments are also equipped with Ca2þ-release channels: the inositol 1,4,5-trisphosphate (IP3) receptor and/or the ryanodine receptor (RyR, Dolman and Tepikin 2006). Ca2þ release from GC and subsequent reuptake are faster than in the ER (Missiaen et al. 2004a,b). Thus, cells can actively maintain a [Ca2þ]i that is some 40,000-fold lower than that outside of the cell, and some 14,000-fold lower than that in the lumen of endomembranes. To maintain such low [Ca2þ]i will also expend a lot of energy.

146

*

M. Micaroni et al.

Figure 1. Scheme for the role of Ca2þ at different transport steps. The arrows indicate the direction of the delivery of membrane carriers. The Ca2þ indications show where Ca2þ is important in the processes involved in the fusion of these carriers with their target compartments. The bold Ca2þ indicate the need for an increased [Ca2þ] for membrane fusion at the different stages of intracellular membrane transport. These include the following fusion steps: between the cargo domain in the GC and the TGN (arrow 2 and double-directed arrow near the Golgi-to-PM carrier [GPC]); between the Golgi-to-basolateral PM and Golgi-to-apical PM carriers and basolateral PM (arrow 3) and apical PM (arrow 6), correspondingly; between the Golgi-to-endosome carrier (GEC) and endosomes (E) (arrow 4); between the GPC ferrying albumin and other soluble cargoes to the basolateral PM (arrow 5); between apically directed mature secretory granules (AMSG) and apical PM (arrow 7), between different endosomes (double directed arrow near E) and between endosomes and lysosomes (not shown). The indications of grey Ca2þ show where even basal [Ca2þ]i in quiescent cells (50–100 nM) are sufficient for membrane fusion: between the ER-to-Golgi carrier and the Golgi cisternae (arrow 1), and between COPI-dependent vesicles and the Golgi cisternae (double directed arrow near “cis”). The role of Ca2þ for the fusion of the putative clathrin/AP-2-dependent vesicles and endosomes is not known (arrow 8, indicated as no Ca2þ). Dark squares show the border between the apical PM and the basolateral PM. Further abbreviations: AISG, apically directed immature secretory granules; AGPC, precursor of apically directed Golgi-to-PM carrier; APM, apical PM; AT, attached TGN; BLPM, basolateral PM; CCB, clathrin-coated buds; ER, endoplasmic reticulum; ERES, ER exit site; Nu, nucleus; PM, plasma membrane; TGN, trans-Golgi network.

Ca2þ in intracellular transport regulation

*

147

Ca2þ pumps Ca2þ ATPase pumps use a lot of energy to maintain the gradients of [Ca2þ] between the cytosol and the lumen of the ER and the GC. Golgi membranes contain two types of Ca2þ ATPase pumps that contribute to Ca2þ uptake in the GC. One of these belongs to the thapsigargin-sensitive, sarcoendoplasmic reticulum calcium ATPase (SERCA) family, that sequester Ca2þ mainly in the ER, and to a lesser extent in Golgi compartments. The other ATPase pump belongs to the thapsigargin-insensitive, secretory pathway calcium ATPase (SPCA) family, which are present on the medial-trans region of the GC (Behne et al. 2003). All of the Ca2þ uptake by the ER is mediated by the SERCA Ca2þ pumps (Vanoevelen et al. 2004). Various relative contributions of the SPCA and SERCA Ca2þ pumps to the total Ca2þ uptake in the GC have been reported, which could be due to celltype dependency. For instance, according to Rojas et al. (2000), Ca2þ uptake in a Golgi-enriched fraction of rat liver depended totally on a SERCA Ca2þ pump, since it was almost completely inhibited by thapsigargin. In contrast, Taylor et al. (1997) showed only 50% inhibition by thapsigargin of Ca2þ uptake into a stacked Golgi fraction from rat liver. A thapsigargin-independent Ca2þ uptake has also been ascribed to PMCA Ca2þ pumps in transit through the GC to the PM. This thapsigargin-resistant Ca2þ uptake disappears when SPCA1 expression is disrupted using RNA interference (Van Baelen et al. 2003). Using aequorin to measure Ca2þ uptake, it has been demonstrated that SERCA Ca2þ pumps are responsible for 50–85% of Ca2þ uptake in the Golgi compartment of HeLa cells (Pinton et al. 1998; Van Baelen et al. 2003). In HeLa and CHO cells overexpressing the Ca2þ-binding protein CALNUC, about 70% of the Ca2þ uptake by the GC depends on the SERCA pumps (Lin et al. 1999). In contrast, it has been reported that the SPCA1 Ca2þ pump is mainly used (67%) to load the GC with Ca2þ in human keratinocytes (Callewaert et al. 2003). Whereas the SERCA pumps are expressed in both the ER and the GC, the SPCAs appear to be more specifically confined to the latter compartments of the secretory pathway, i.e. the Golgi stack, the TGN, and maybe secretory granules. The cis-Golgi region appears to express SERCA and IP3 receptors (Missiaen et al. 2004a,b; Vanoevelen et al. 2004), while the trans-Golgi region appears to contain SPCA1 and to lack IP3 receptors (Missiaen et al. 2004a,b). Thus, SPCA1 is responsible for the uptake of Ca2þ into the trans area of the GC (Michelangeli et al. 2005). Two isoforms of the SPCA Ca2þ pump, known as SPCA1 and SPCA2, have so far been identified, although only SPCA1 has been shown to be active. Unlike SERCA, the SPCA1 Ca2þ pump can transport Mn2þ in addition to Ca2þ (Van Baelen et al. 2001; Missiaen et al. 2004a,b). SPCA2 is present in mammalian cells, although its level of expression and distribution in tissues remains controversial (Vanoevelen et al. 2005; Xiang et al. 2005) and it is not yet well characterized. However, it is known to have similar functions to SPCA1, which

148

*

M. Micaroni et al.

include of the same capacity for transporting Mn2þ and Ca2þ in a thapsigargin-insensitive way, and the same GC localization. Over-expression of SPCA can cause significant alterations in several of the Ca2þ transport molecules in the cell and dramatically increased the celldivision rate (Reinhardt et al. 2004). In Darier Disease, where SERCA2 function is impaired, the overexpression of SPCA1 might compensate for this SERCA2 dysfunction, suggesting a role in changes of resting [Ca2þ]i (Foggia et al. 2005). Thus, the SPCAs have different crucial roles: (a) they contribute in a significant manner to luminal Ca2þ uptake; (b) they are responsible for the cytosolic regulation and storage in the GC of Mn2þ, promoting the correct functioning of the few Golgi enzymes that require Mn2þ as a cofactor. The ATP2C1 gene, encoding for the SPCA1 in mammals cells, have an homolog in yeast and Caenorhabditis elegans. In fact, the PMR1 gene product was shown to be a Ca2þ-ATPase that is located in the GC (Sorin et al. 1997; Van Baelen et al. 2001). For the latter, these are seen as type IIA (ER-type) Ca2þ-ATPases, and type IIB (PM-type) Ca2þ-ATPases (Sze et al. 2000). The mechanisms by which Ca2þ is transported into the GC and maintained at high levels has not been completely defined yet. Ca2þ accumulation in isolated Golgi membranes varies according to both Ca2þ and Mg-ATP concentrations, and it can be inhibited by thapsigargin, but not stimulated by calmodulin (Rojas et al. 2000). Thapsigargin-independent Ca2þ accumulation was not affected by pre-treatment with agents such as NH4Cl or chloroquine, which collapse the H+ gradient across cell membranes (Pinton et al. 1998). Thus, there are two main Ca2þ pumps (SERCA and SPCA) that are responsible for the accumulation of Ca2þ in the ER and the GC.

Protein buffers for Ca2þ in the lumen of the GC To save on the energy that is needed to maintain such high gradients of Ca2þ between these intracellular Ca2þ stores and the cytosol, cells express specific Ca2þ-binding proteins, which can sequester most of the free Ca2þ in these stores. In the ER, there are several Ca2þ-binding proteins, including calreticulin (Michalak et al. 1999), calnexin (Ohsako et al. 1994), reticulocalbin, calumenin and ERC-55 (Vorum et al. 1999). Some of these, although not calumenin, have also been found in the lumen of the Golgi cisternae (Vorum et al. 1999). However, the Ca2þ sequestered in the Golgi lumen is mainly buffered by CALNUC (nucleobindin; Lin et al. 1998, 1999; Kawano et al. 2000), Cab45, the first resident protein that was described for the Golgi lumen (Scherer et al.  and Vorum 2000), and P54/NEFA (Morel-Haux et al. 2002). 1996; Honore CALNUC has also been localized to both the ER and the Golgi lumen (Lin et al. 1998, 1999).

Ca2þ in intracellular transport regulation

*

149

When the influx of Ca2þ into the cytosol is increased chronically, cells can express more of the Golgi calcium-binding protein CALNUC in order to prevent calcium cytotoxicity. Additionally, the overexpression of SPCA1 increases CALNUC levels (Rejnhardt et al. 2000). On the other hand, over expression of CALNUC leads to an enhancement of agonist-evoked calcium release (Lin et al. 1999). Thus, cells have developed specific molecular mechanisms to maintain these Ca2þ gradients between the Ca2þ stores and the cell cytosol.

Ca2þ release What is the role of these Ca2þ gradients and what are the mechanisms regulating the local [Ca2þ]i, during membrane transport? These mechanisms primarily involve Ca2þ channels. It appears that during the synchronous passage of cargo through the GC, the local [Ca2þ]i increases due to the activation of Ca2þ channels. Ca2þ is released from either the Golgi cisternae per se or from the trans ER (see Chapter 1), which is closely attached to the GC (Micaroni et al. submitted). Both the ER and the SERCA-expressing part of the GC are involved in the setting up of [Ca2þ]i signals. The release of Ca2þ from the ER can be triggered not only by the activation of IP3 and ryanodine receptors (Pinton et al. 1998) but also increase in [Ca2þ]i can themselves promote further Ca2þ release from these intracellular Ca2þ stores via the direct binding of Ca2þ to IP3 and RyR channels (Roderick et al. 2003). The GC has been shown to act in concert with the ER, albeit with different kinetics, in the elevation of [Ca2þ]i in response to agonist stimulation (Missiaen et al. 2004a,b). IP3 receptors have been immunolocalized in the GC (Lin et al. 1999), and agonist stimulation of the production of IP3 can activate these channels on the Golgi membranes, resulting in Ca2þ release from the GC. All of the 0.3 mM of free Ca2þ inside the Golgi lumen can be released upon stimulation with agonists coupled to IP3 production; for example, addition of histamine to cells, which is an agonist coupled to IP3 production, results in a rapid and extensive drop in the free Ca2þ in the GC. However, this drop in [Ca2þ]GC caused by histamine has been shown to be slightly smaller and slower than that observed in the ER (Pinton et al. 1998; Vanoevelen et al. 2004). Ca2þ release from the GC is also inactivated faster than that from the ER (Michelangeli et al. 2005), and it has been shown that the SPCA1-based part of the GC does not contribute to these changes in [Ca2þ]i (Vanoevelen et al. 2004; Missiaen et al. 2004a,b). This release of Ca2þ from the GC does not depend on the functions of the ARF/COPI machinery. Indeed, in digitonin-permeabilized cells, the release and uptake of Ca2þ from the GC was not affected by GTPgS, with neither the loading of Ca2þ into the GC nor the rapid emptying of the GC in thapsigargintreated cells (see above) is affected (Pinton et al. 1998).

150

*

M. Micaroni et al.

Thus, both the ER and the GC participate in local increases in [Ca2þ]i, which also facilitates vectorial membrane transport. Upon stimulation of a receptor coupled to IP3 production, not only does the GC contribute in part to the increased [Ca2þ]i, but there are also major changes in [Ca2þ]GC (Pinton et al. 1998). During intra-Golgi transport, the [Ca2þ]i local to the GC increases, with this increase not occurring immediately after the arrival of cargo at the GC, but a little time (5–7 min) later (Micaroni et al. submitted), which corresponds to the need for fusion between the cargo domain and the TGN (Mironov et al. 2005). Increases in the local [Ca2þ]i at the trans side of the GC can thus facilitate the transfer of a cargo domain from the cis to the trans side of the GC. A possible scheme of this process is shown in Fig. 2.

The positioning of the Ca2þ source The position of these Ca2þ stores is indeed very important for the regulation of intracellular transport (Dolman et al. 2005; Dolman and Tepikin 2006). For instance, in polarized cells, the positioning of the GC is such that both mitochondria and GC are segregated from the lateral regions of the PM, the nucleus and the basal part of the cytoplasm. Here, the ER and nucleus are located in the basolateral part of the cell, whereas the secretory granules are located at the apical pole (Gerasimenko et al. 2002). The GC is therefore positioned between the main Ca2þ release sites in the apical region of the cell and the important Ca2þ sink formed by the perigranular mitochondria. During acetylcholine-induced [Ca2þ]i signalling in the apical region, large Ca2þ gradients can form over the GC because the GC is sandwiched between the Ca2þ source (release sites in the apical region) and the Ca2þ sink (mitochondrial uniporters), Ca2þ gradients are formed with higher [Ca2þ]i over the trans-Golgi than over the cis-Golgi (Dolman et al. 2005). When low doses of acetylcholine were given apically, the Ca2þ gradient had almost completely dissipated at a distance of 2 mm from the GC, which was the region of the cell occupied by the perigranular mitochondrial belt (Dolman and Tepikin 2006). These [Ca2þ]i gradients can also reach hundreds of nanomoles per micrometer when measured along a line drawn from the apical to the basal part of an acinar cell (Gerasimenko et al. 1996). The well-established cell-free assay for intra-Golgi transport that measures glycosylation of the VSV-G protein is inhibited by BAPTA (IC50 0.8 mM) but not by EGTA. This indicates that luminal Ca2þ is required for transport (Porat and Elazar 2000) and suggesting a role for Ca2þ release from the luminal stores in the membrane fusion that accompanies intracellular transport. As the local release of Ca2þ occurs from the luminal stores of the organelles involved in fusion, can this explain the differential effects of BAPTA and EGTA? The on-rate of Ca2þ binding to BAPTA (Naraghi 1997) is similar to that for calmodulin (Falke et al. 1994), so that this Ca2þ chelator

Ca2þ in intracellular transport regulation

*

151

Figure 2. Behaviour of the GC and Ca2þ during the passage of cargo through the GC. A Exit of cargo (black lines) from the ER in ER-to-Golgi carriers. B,C Arrival of cargo at the GC triggers Ca2þ exit from the GC (grey shadow) that shifts the equilibrium between formation and consumption of COPI-dependent vesicles (blue circles) towards consumption, and that induces the SNAREdependent disappearance of vesicles and the formation of intracisternal connections. Ca2þ channels open, leading to release of Ca2þ (black dots in C), or there is a temporary reduction in SPCA activity. Increased [Ca2þ]i leads to recruitment of cPLA2 to GC membranes. D Shift of cargo domain to the trans side of the GC. E Reuptake of Ca2þ on the GC region due to closure of Ca2þ channels and maturation of the cargo domain into a post-Golgi carrier. F Enhanced activity of SPCA pumps restore basal [Ca2þ]i. Exit of post-Golgi carrier and restoration of vesicle equilibrium at the Golgi level.

can compete with calmodulin in its initial binding of Ca2þ. In contrast, EGTA has a considerably slower on-rate (Naraghi 1997). These differences in onrate would only be relevant for Ca2þ gradients lasting for less than 1 ms. It has been calculated that buffering by 10 mM BAPTA would reach equilibri-

152

*

M. Micaroni et al.

um in around 3 s, whereas that due to EGTA would take around 1.2 s (Adler et al. 1991). From the known diffusion rate for Ca2þ in the cell cytosol, it is clear that over 1 ms such a [Ca2þ]i gradient would dissipate within a distance of around 20 nm. This is a possible situation for a fast calmodulin-dependent fusion process that requires [Ca2þ]i increases in the micromolar range (or higher, as in yeast vacuole–vacuole fusion; Peters and Mayer 1998). If the basal [Ca2þ]i was sufficient, a differential effect of these two chelators on a Ca2þ/calmodulindependent fusion process would not be expected. Thus these differential effects indicate that the requirement must be for a local increase in [Ca2þ]i due to short-lived pulses of Ca2þ release from very close to the site of membrane fusion. This Ca2þ release would, therefore, have to be tightly coupled to the fusion, both spatially and temporally. One way in which this can be achieved for synaptic exocytosis is through the direct interactions of neuronal SNARE proteins with voltage-gated Ca2þ channels (Sheng et al. 1996). Perhaps similar interactions can occur with intracellular Ca2þ release channels. Thus, the special organization and localization of the Ca2þ stores can regulate the directionality of vectorial membrane transport.

Other functions of Ca2þ in trafficking Transient gradients in [Ca2þ]i also regulate the assembly and disassembly of the coat proteins that are responsible for vesicular trafficking between Golgi stacks and beyond the TGN. For instance, experiments with BAPTA and EGTA have demonstrated that some of their effects can be explained by an important role for Ca2þ in the stabilizing of the coat of the forming coated buds (Ahluwalia et al. 2001). As with Ca2þ-sensitive membrane fusion, COPI coat assembly is more sensitive to BAPTA than EGTA (Pryor et al. 2000). Also, after 90 min treatment of NRK cells with BAPTA-AM, while the GC was of similar size and retained a stacked structure, the number of cisternae within each stack was smaller. At the same time, accumulation of COPI-dependent vesicles was not seen because low Ca2þ blocks the formation of the COPI coat; indeed, the coatomer dissociates from the GC after treatment with BAPTAAM (Chen et al. 2002). Similar effect of Ca2þ on COPII coat have also been described. Sec31A (a subunit of COPII)-positive spots have been shown to increase in number and to be concentrated in a juxtanuclear region in response to Ca2þ mobilization. In contrast, Ca2þ chelation by BAPTA-AM decreases the number of punctate dots associated with Sec31A. The distribution pattern of the Sec23 (another subunit of COPII)-interacting protein p125 is not affected by treatment with either a Ca2þ ionophore or a Ca2þ chelator (Shibata et al. 2007). Additionally, at least at the level of the fusion with the PM, Ca2þ channels have been seen to associate physically and functionally with the Q-SNAREs, syntaxin and SNAP-25 (Yoshida et al. 1992; Sheng et al. 1994; Mochida et al.

Ca2þ in intracellular transport regulation

*

153

1996; Wiser et al. 1996; Rettig et al. 1997). Trans interactions between SNAREs on opposite membranes have been proposed to facilitate or trigger [Ca2þ]i signals in response to docking (Bezprozvanny et al. 1995; Scheckman 1998). Local increases in [Ca2þ]i near the GC lead to the recruitment of the cytosolic, Ca2þ-sensitive phospholipase A2 isoform cPLA2a to Golgi membranes (Ghosh et al. 2006; Polishchuk R, personal communication). The role of this event is not clear. Thus, normal and transiently increased [Ca2þ]i appears to be important not only for membrane fusion, but also for the formation of specialized membrane coats.

Roles of luminal Ca2þ in the Golgi apparatus Why it is necessary to have high concentrations of Ca2þ in the lumen of the ER and the GC? Fairly constant and rather high [Ca2þ]GC suggest that Ca2þ may be needed for the correct execution of luminal functions. However, little is known about the role of Ca2þ of this compartment, and its dynamic changes under physiological conditions. [Ca2þ]GC controls a variety of important functions, including protein and lipid synthesis, chaperone-dependent processing, glycosylation, sorting and eventual breakdown of newly formed proteins (Carnell and Moore 1994; Ivessa et al. 1995; Austin and Shields 1996; Duncan and Burgoyne 1996; Meldolesi and Pozzan 1998), as well as transport of proteins, cargo condensation, and precursor processing (Chanat and Huttner 1991; Carnell and Moore 1994; Austin and Shields 1996; Duncan and Burgoyne 1996; Corbett and Michalak 2000; Wuytack et al. 2003). Of interest, a sufficient supply of Mn2þ is also an absolute requirement for correct glycosylation of secretory proteins in the GC (Durr et al. 1998). Indeed, the activities of most enzymes are Ca2þ and Mn2þ dependent (Sharma et al. 1974; Parodi 1979). In mammalian cells, the endoproteolytic proprotein convertases (Davidson et al. 1988; Schmidt and Moore 1995; Steiner 1998) and the secretases (LaFerla 2002) in the GC and secretory vesicles are Ca2þ dependent. Ca2þ dyshomeostasis in these compartments could contribute to various amyloidoses, including Alzheimer’s disease. The role of luminal Ca2þ for human pathology can be illustrated by the intracellular pathogenic bacteria, such as Mycobacteria and Salmonellae, that have developed means to control fusion reactions in their host cells. They persist in phagosomes, the fusion of which with lysosomes is actively suppressed to ensure the bacteria survival inside host cells (Michelangeli et al. 2005). However, the maintenance of the high [Ca2þ]ER and [Ca2þ]GC appears not to be vital. Indeed, depletion of calcium from the ER and possibly from the GC, either by treatment with A23187 or thapsigargin, has no effects on the folding or secretion of newly synthesized albumin (Lodish et al. 1992). Moreover, cells can adapt their growth when the pumping of Ca2þ from the

154

*

M. Micaroni et al.

cytosol is blocked by thapsigargin, by expressing large amounts of thapsigargin-resistant Ca2þ-ATPase pumps (Waldron et al. 1995). Since under these conditions the Ca2þ pumps in the PM are functional, Ca2þ levels in the ER and the cytosol are expected to equilibrate at a low level. Thus, the maintenance of a high [Ca2þ] in the lumen of the ER and of the GC facilitates some important functions. These are necessary for enzymatic activity in these organelles and for the prevention of the early exit of unfolded proteins from the ER. The trans ER represents the source of the rapid increases in local [Ca2þ]i near the trans side of the GC. However, these high [Ca2þ] in the ER and GC lumen are not vital per se.

The role of Ca2þ in the TGN and endosomes The [Ca2þ] in the TGN and post-Golgi carriers is lower than in the GC (Pezzati et al. 1997). However, it has been shown that the selective aggregation of regulated secretory proteins in the TGN depends on luminal [Ca2þ] (Chanat and Huttner 1991; see also Chapter 3.11). On the other hand, it is known that in cells with regulatory secretion, most of the constitutively secreted proteins pass through endosomes (see Chapter 3.10). It is important to stress that these endosomal clathrin-coated buds and clathrin-independent tubules contain millimolar levels of Ca2þ when they pinch off from the PM, because clathrin-coated buds are in continuity with the external fluid that contains 1–2 mM Ca2þ. The fusion of recently uncoated clathrin-dependent buds or vesicles with endosome could provide a flux of Ca2þ from the extracellular space towards an endosome along the very thin tube that connects a bud with the PM, destabilizing the internal lipid leaflet. Ca2þ-sensitive probes that undergo endocytosis have demonstrated that the luminal [Ca2þ] is rapidly reduced from 1–2 mM to around 3 mM, within 20 min of endocytosis (Gerasimenko et al. 1998). In phagosomes, the [Ca2þ] can be as high as 400–600 mM (Christensen et al. 2002). Thus, the [Ca2þ] in the lumen of post-Golgi compartments and within the endosomal pathway is lower than in the [Ca2þ]ER and the [Ca2þ]GC. However, even this lower concentration is important for the execution of some functions that occur there. One of these functions could be the uptake of Ca2þ from outside.

Conclusion The cellular [Ca2þ] has an important role not only in signalling, but also in intracellular transport, and in particular in the directionality of consecutive fusion events along the secretory pathway. The GC contains a significant amount of Ca2þ that is involved not only in the regulation of intraluminal functions, but that is also released into the cytosol as part of the regulation of local [Ca2þ]i that is needed to maintain vectorial membrane transport during exocytosis.

Ca2þ in intracellular transport regulation

*

155

Abbreviations [Ca2þ] ER GC PM RyR STB TGN

Ca2þ concentration endoplasmic reticulum Golgi complex plasma membrane ryanodine receptor Shiga toxin-B trans-Golgi network

References Adler EM, Augustine GJ, Duffy SN, Charlton MP (1991) Alien intracellular calcium chelators attenuate neurotransmitter release at the squid giant synapse. J Neurosci 11(6): 1496–1507 Ahluwalia JP, Topp JD, Weirather K, Zimmerman M, Stamnes M (2001) A role for calcium in stabilizing transport vesicle coats. J Biol Chem 276(36): 34148–34155 Aridor M, Bannykh SI, Rowe T, Balch WE (1995) Sequential coupling between COPII and COPI vesicle coats in endoplasmic reticulum to Golgi transport. J Cell Biol 131(4): 875–893 Austin CD, Shields D (1996) Prosomatostatin processing in permeabilized cells. Calcium is required for prohormone cleavage but not formation of nascent secretory vesicles. J Biol Chem 271(2): 1194–1199 Baker D, Wuestehube L, Schekman R, Botstein D, Segev N (1990) GTP-binding Ypt1 protein and Ca2þ function indipendently in a cell-free protein transport reaction. Proc Natl Acad Sci 87(1): 355–359 Beckers CJ, Balch WE (1989) Calcium and GTP: essential components in vesicular trafficking between the endoplasmic reticulum and Golgi apparaturs. J Cell Biol 108(4): 1245–1256 Behne MJ, Tu CL, Aronchik I, Epstein E, Bench G, Bikle DD, Pozzan T, Mauro TM (2003) Human keratinocyte ATP2C1 localizes to the Golgi and controls Golgi Ca2þ stores. J Invest Dermatol 121(4): 688–694 Bezprozvanny I, Scheller RH, Tsien RW (1995) Functional impact of syntaxin on gating N-type and Q-type calcium channels. Nature 378(6557): 623–626 Brostrom CO, Brostrom MA (1990) Calcium-dependent regulation of protein synthesis in intact mammalian cells. Annu Rev Physiol 52: 577–590 Burgoyne RD, Morgan A (1995) Ca2þ and secretory-vesicle dynamics. Trends Neurosci 18(4): 191–196 Burgoyne RD, Morgan A (1998) Calcium sensors in regulated exocytosis. Cell Calcium 24(5–6): 367–376 Callewaert G, Parys JB, De Smedt H, Raeymaekers L, Wuytack F, Vanoevelen J, Van Baelen K, Simoni A, Rizzuto R, Missiaen L (2003) Similar Ca2þ-signaling properties in keratinocytes and in COS-1 cells overexpressing the secretory-pathway Ca2þ-ATPase SPCA1. Cell Calcium 34(2): 157–162 Carnell L, Moore HP (1994) Transport via the regulated secretory pathway in semi-intact PC12 cells: role of intra-cisternal calcium and pH in the transport and sorting of secretogranin II. J Cell Biol 127(3): 693–705 Chanat E, Huttner WB (1991) Milieu-induced, selective aggregation of regulated secretory proteins in the trans-Golgi network. J Cell Biol 115(6): 1505–1519

156

*

M. Micaroni et al.

Chandra S, Kable EP, Morrison GH, Webb WW (1991) Calcium sequestration in the Golgi apparatus of cultured mammalian cells revealed by laser scanning confocal microscopy and ion microscopy. J Cell Sci 100(4): 747–752 Chandra S, Fewtrell C, Millard PJ, Sandison DR, Webb WW, Morrison GH (1994) Imaging of total intracellular calcium and calcium influx and efflux in individual resting and stimulated tumor mast cells using ion microscopy. J Biol Chem 269(21): 15186–15194 Chen JL, Ahluwalia JP, Stamnes M (2002) Selective effects of calcium chelators on anterograde and retrograde protein transport in the cell. J Biol Chem 277(38): 35682–35687 Chen YA, Scales SJ, Patel SM, Doung YC, Scheller RH (1999) SNARE complex formation is triggered by Ca2þ and drives membrane fusion. Cell 97(2): 165–174 Cho SJ, Kelly M, Rognlien KT, Cho JA, Horber JK, Jena BP (2002) SNAREs in opposing bilayers interact in a circular array to form conducting pores. Biophys J 83(5): 2522–2527 Christensen KA, Myers JT, Swanson JA (2002) pH-dependent regulation of lysosomal calcium in macrophages. J Cell Sci 115(3): 599–607 Colombo MI, Beron W, Stahl PD (1997) Calmodulin regulates endosome fusion. J Biol Chem 272(12): 7707–7712 Coorssen JR, Blank PS, Tahara M, Zimmerberg J (1998) Biochemical and functional studies of cortical vesicle fusion: the SNARE complex and Ca2þ sensitivity. J Cell Biol 143(7): 1845–1857 Corbett EF, Michalak M (2000) Calcium, a signalling molecule in the endoplasmic reticulum? Trends Biochem Sci 25(7): 307–311 Davidson HW, Rhodes CJ, Hutton JC (1988) Intraorganellar calcium and pH control proinsulin cleavage in the pancreatic beta cell via two distinct site-specific endopeptidases. Nature 333(6168): 93–96 Dolman NJ, Gerasimenko JV, Gerasimenko OV, Voronina SG, Petersen OH, Tepikin AV (2005) Stable Golgi–mitochondria complexes and formation of Golgi Ca2þ gradients in pancreatic acinar cells. J Biol Chem 280(16): 15794–15799 Dolman NJ, Tepikin AV (2006) Calcium gradients and the Golgi. Cell Calcium 40(5–6): 505–512 Duncan JS, Burgoyne RD (1996) Characterization of the effects of Ca2þ depletion on the synthesis, phosphorylation and secretion of caseins in lactating mammary epithelial cells. Biochem J 317(2): 487–493 Durr G, Strayle J, Plemper R, Elbs S, Klee SK, Catty P, Wolf DH, Rudolph HK (1998) The medial-Golgi ion pump Pmr1 supplies the yeast secretory pathway with Ca2þ and Mn2þ required for glycosylation, sorting, and endoplasmic reticulum-associated protein degradation. Mol Biol Cell 9(5): 1149–1162 Falke JJ, Drake SK, Hazard AL, Peersen OB (1994) Molecular tuning of ion binding to calcium signaling proteins. Q Rev Biophys 27(3): 219–290 Flanagan JJ, Barlowe C (2006) Cysteine-disulfide cross-linking to monitor SNARE complex assembly during endoplasmic reticulum–Golgi transport. J Biol Chem 281(4): 2281–2288 Foggia L, Aronchik I, Aberg K, Brown B, Hovnanian A, Mauro TM (2005) Activity of the hSPCA1 Golgi Ca2þ pump is essential for Ca2þ-mediated Ca2þ response and cell viability in Darier disease. J Cell Sci 119(4): 671–679 Gerasimenko JV, Tepikin AV, Petersen OH, Gerasimenko OV (1998) Calcium uptake via endocytosis with rapid release from acidifying endosomes. Curr Biol 8(24): 1335–1338 Gerasimenko OV, Gerasimenko JV, Petersen OH, Tepikin AV (1996) Short pulses of acetylcholine stimulation induce cytosolic Ca2þ signals that are excluded from the nuclear region in pancreatic acinar cells. Pflugers Arch 432(6): 1055–1061 Gerasimenko OV, Gerasimenko JV, Rizzuto RR, Treiman M, Tepikin AV, Petersen OH (2002) The distribution of the endoplasmic reticulum in living pancreatic acinar cells. Cell Calcium 32(5–6): 261–268

Ca2þ in intracellular transport regulation

*

157

Ghosh M, Tucker DE, Burchett SA, Leslie CC (2006) Properties of the Group IV phospholipase A2 family. Prog Lipid Res 45(6): 487–510 Gill DL, Waldron RT, Rys-Sikora KE, Ufret-Vincenty CA, Graber MN, Favre CJ, Alfonso A (1996) Calcium pools, calcium entry, and cell growth. Biosci Rep 16(2): 139–157  B, Vorum H (2000) The CREC family, a novel family of multiple EF-hand, lowHonore affinity Ca2þ-binding proteins localised to the secretory pathway of mammalian cells. FEBS Lett 466(1): 11–18 Ivessa NE, De Lemos-Chiarandini C, Gravotta D, Sabatini DD, Kreibich G (1995) The brefeldin A-induced retrograde transport from the Golgi apparatus to the endoplasmic reticulum depends on calcium sequestered to intracellular stores. J Biol Chem 270(43): 25960–25967 Jeremic A, Kelly M, Cho JA, Cho SJ, Horber JK, Jena BP (2004) Calcium drives fusion of SNARE-apposed bilayers. Cell Biol Int 28(1): 19–31 Kawano J, Kotani T, Ogata Y, Ohtaki S, Takechi S, Nakayama T, Sawaguchi A, Nagaike R, Oinuma T, Suganuma T (2000) CALNUC (nucleobindin) is localized in the Golgi apparatus in insect cells. Eur J Cell Biol 79(3): 208–217 Kweon HS, Beznoussenko GV, Micaroni M, Polishchuk RS, Trucco A, Martella O, Di Giandomenico D, Marra P, Fusella A, Di Pentima A, Berger EG, Geerts WJ, Koster AJ, Burger KN, Luini A, Mironov AA (2004) Golgi enzymes are enriched in perforated zones of Golgi cisternae but are depleted in COPI vesicles. Mol Biol Cell 15(10): 4710–4724 LaFerla FM (2002) Calcium dyshomeostasis and intracellular signalling in Alzheimer’s disease. Nat Rev Neurosci 3(11): 862–872 Lauvrak SU, Llorente A, Iversen TG, Sandvig K (2002) Selective regulation of the Rab9independent transport of ricin to the Golgi apparatus by calcium. J Cell Sci 115(17): 3449–3456 Lin P, Le-Niculescu H, Hofmeister R, McCaffery JM, Jin M, Hennemann H, McQuistan T, De Vries L, Farquhar MG (1998) The mammalian calcium-binding protein, nucleobindin (CALNUC), is a Golgi resident protein. J Cell Biol 141(7): 1515–1527 Lin P, Yao Y, Hofmeister R, Tsien RY, Farquhar MG (1999) Overexpression of CALNUC (nucleobindin) increases agonist and thapsigargin releasable Ca2þ storage in the Golgi. J Cell Biol 145(2): 279–289 Lodish HF, Kong N, Wikstrom L (1992) Calcium is required for folding of newly made subunits of the asialoglycoprotein receptor within the endoplasmic reticulum. J Biol Chem 267(18): 12753–12760 Maki M, Kitaura Y, Satoh H, Ohkouchi S, Shibata H (2002) Structures, functions and molecular evolution of the penta-EF-hand Ca2þ-binding proteins. Biochim Biophys Acta 1600(1–2): 51–60 Meldolesi J, Pozzan T (1998) The endoplasmic reticulum Ca2þ store: a view from the lumen. Trends Biochem Sci 23(1): 10–14 Michalak M, Corbett EF, Mesaeli N, Nakamura K, Opas M (1999) Calreticulin: one protein, one gene, many functions. Biochem J 344(2): 281–292 Michelangeli F, Ogunbayo OA, Wootton LL (2005) A plethora of interacting organellar Ca2þ stores. Curr Opin Cell Biol 17(2): 135–140 Mironov AA, Beznoussenko GV, Polishchuk RS, Trucco A (2005) Intra-Golgi transport. A way to a new paradigm? BBA Mol Cell Res 1744(3): 340–350 Missiaen L, Raeymaekers L, Dode L, Vanoevelen J, VanBaelen K, Parys JB, Callewaert G, DeSmedt H, Segaert S, Wuytack F (2004a) SPCA1 pumps and Hailey–Hailey disease. Biochem Biophys Res Commun 322(4): 1204–1213 Missiaen L, Van Acker K, Van Baelen K, Raeymaekers L, Wuytack F, Parys JB, De Smedt H, Vanoevelen J, Dode L, Rizzuto R, Callewaert G (2004b) Calcium release from the Golgi apparatus and the endoplasmic reticulum in HeLa cells stably expressing targeted aequorin to these compartments. Cell Calcium 36(6): 479–487

158

*

M. Micaroni et al.

Mochida S, Sheng ZH, Baker C, Kobayashi H, Catterall WA (1996) Inhibition of neurotransmission by peptides containing the synaptic protein interaction site N-type Ca2þ channels. Neuron 17(4): 781–788 Morel-Huaux VM, Pypaert M, Wouters S, Tartakoff AM, Jurgan U, Gevaert K, Courtoy PJ (2002) The calcium-binding protein p54/NEFA is a novel luminal resident of medial Golgi cisternae that traffics independently of mannosidase II. Eur J Cell Biol 81(2): 87–100 Naraghi M (1997) T-jump study of calcium binding kinetics of calcium chelators. Cell Calcium 22(4): 255–268 Ohsako S, Hayashi Y, Bunick D (1994) Molecular cloning and sequencing of calnexin-t. An abundant male germ cell-specific calcium-binding protein of the endoplasmic reticulum. J Biol Chem 269(19): 14140–14148 Ordenes VR, Reyes FC, Wolff D, Orellana A (2002) A thapsigargin-sensitive Ca2þ pump is present in the pea Golgi apparatus membrane. Plant Physiol 129(4): 1820–1828 Orrenius S, Zhivotovsky B, Nicotera P (2003) Regulation of cell death: the calciumapoptosis link. Nat Rev Mol Cell Biol 4(7): 552–565 Parodi AJ (1979) Biosynthesis of yeast mannoproteins. Synthesis of mannan outer chain and of dolichol derivatives. J Biol Chem 254(17): 8343–8352 Peters C, Mayer A (1998) Ca2þ/calmodulin signals the completion of docking and triggers a late step of vacuole fusion. Nature 396(6711): 575–580 Pezzati R, Bossi M, Podini P, Meldolesi J, Grohovaz F (1997) High-resolution calcium mapping of the endoplasmic reticulum–Golgi-exocytic membrane system. Electron energy loss imaging analysis of quick frozen-freeze dried PC12 cells. Mol Biol Cell 8(8): 1501–1512 Pind SN, Nuoffer C, McCaffery JM, Plutner H, Davidson HW, Farquhar MG, Balch WE (1994) Rab1 and Ca2þ are required for the fusion of carrier vesicles mediating endoplasmic reticulum to Golgi transport. J Cell Biol 125(2): 239–252 Pinton P, Pozzan T, Rizzuto R (1998) The Golgi apparatus is an inositol 1,4,5-trisphosphate-sensitive Ca2þ store, with functional properties distinct from those of the endoplasmic reticulum. EMBO J 17(18): 5298–5308 Porat A, Elazar Z (2000) Regulation of intra-Golgi membrane transport by calcium. J Biol Chem 275(38): 29233–29237 Pozzan T, Rizzuto R, Volpe P, Meldolesi J (1994) Molecular and cellular physiology of intracellular calcium stores. Physiol Rev 74(3): 595–636 Pryor PR, Mullock BM, Bright NA, Gray SR, Luzio JP (2000) The role of intraorganellar Ca2þ in late endosome–lysosome heterotypic fusion and in the reformation of lysosomes from hybrid organelles. J Cell Biol 149(5): 1053–1062 Reddy A, Caler EV, Andrews NW (2001) Plasma membrane repair is mediated by Ca2þregulated exocytosis of lysosomes. Cell 106(2): 157–169 Reinhardt TA, Filoteo AG, Penniston JT, Horst RL (2000) Ca2þ-ATPase protein expression in mammary tissue. Am J Physiol Cell Physiol 279(5): C1595–C1602 Reinhardt TA, Horst RL, Waters WR (2004) Characterization of Cos-7 cells overexpressing the rat secretory pathway Ca2þ-ATPase. Am J Physiol Cell Physiol 286(1): C164–C169 Rettig J, Heinemann C, Ashery U, Sheng ZH, Yokoyama CT, Catterall WA, Neher E (1997) Alteration of Ca2þ dependance of neurotransmitter release by disruption of Ca2þ channel/syntaxin interaction. J Neurosci 17(17): 6647–6656 Rettig J, Neher E (2002) Emerging roles of presynaptic proteins in Ca2þ-triggered exocytosis. Science 298(5594): 781–785 Roderick HL, Berridge MJ, Bootman MD (2003) Calcium-induced calcium release. Curr Biol 13(11): R425 Rojas P, Surroca A, Orellana A, Wolff D (2000) Kinetic characterization of calcium uptake by the rat liver Golgi apparatus. Cell Biol Int 24(4): 229–233 Scheckman R (1998) Membrane fusion. Ready . . . aim . . . fire!. Nature 396(6711): 514–515

Ca2þ in intracellular transport regulation

*

159

Scherer PE, Lederkremer GZ, Williams S, Fogliano M, Baldini G, Lodish HF (1996) Cab45, a novel (Ca2þ)-binding protein localized to the Golgi lumen. J Cell Biol 133(2): 257–268 Schmidt WK, Moore HP (1995) Ionic milieu controls the compartment-specific activation of pro-opiomelanocortin processing in AtT-20 cells. Mol Biol Cell 6(10): 1271–1285 Sharma CB, Babczinski P, Lehle L, Tanner W (1974) The role of dolicholmonophosphate in glycoprotein biosynthesis in Saccharomyces cerevisiae. Eur J Biochem 46(1): 35–41 Sheng ZH, Rettig J, Takahashi M, Catterall WA (1994) Identification of a syntaxinbinding site on N-type calcium channels. Neuron 13(6): 1303–1313 Sheng ZH, Rettig J, Cook T, Catterall WA (1996) Calcium-dependent interaction of N-type calcium channels with the synaptic core complex. Nature 379(6564): 451–454 Shibata H, Suzuki H, Yoshida H, Maki M (2007) ALG-2 directly binds Sec31A and localizes at endoplasmic reticulum exit sites in a Ca2þ-dependent manner. Biochem Biophys Res Commun 353(3): 756–763 Sorin A, Rosas G, Rao R (1997) PMR1, a Ca2þ-ATPase in yeast Golgi, has properties distinct from sarco/endoplasmic reticulum and plasma membrane calcium pumps. J Biol Chem 272(15): 9895–9901 Sorrentino V, Rizzuto R (2001) Molecular genetics of Ca2þ stores and intracellular Ca2þ signaling. Trends Pharm Sci 22(9): 459–464 Steiner DF (1998) The proprotein convertases. Curr Opin Chem Biol 2(1): 31–39 Stojilkovic SS (2005) Ca2þ-regulated exocytosis and SNARE function. Trends Endocrinol Metab 16(3): 81–83 Sullivan KM, Busa WB, Wilson KL (1993) Calcium mobilization is required for nuclear vesicle fusion in vitro: implications for membrane traffic and IP3 receptor function. Cell 73(7): 1411–1422 Sze H, Liang F, Hwang I, Curran AC, Harper JF (2000) Diversity and regulation of plant Ca2þ pumps: insights from expression in yeast. Annu Rev Plant Physiol Plant Mol Biol 51: 433–462 Taylor RS, Jones SM, Dahl RH, Nordeen MH, Howell KE (1997) Characterization of the Golgi complex cleared of proteins in transit and examination of calcium uptake activities. Mol Biol Cell 8(10): 1911–1931 Van Baelen K, Vanoevelen J, Missiaen L, Raeymaekers L, Wuytack F (2001) The Golgi PMR1 P-type ATPase of Caenorhabditis elegans. Identification of the gene and demonstration of calcium and manganese transport. J Biol Chem 276(14): 10683–10691 Van Baelen K, Vanoevelen J, Callewaert G, Parys JB, DeSmedt H, Raeymaekers L, Rizzuto R, Missiaen L, Wuytack F (2003) The contribution of the SPCA1 Ca2þ pump to the Ca2þ accumulation in the Golgi apparatus of HeLa cells assessed via RNA-mediated interference. Biochem Biophys Res Commun 306(2): 430–436 Van Baelen K, Dode L, Vanoevelen J, Callewaert G, De Smedt H, Missiaen L, Parys JB, Raeymaekers L, Wuytack F (2004) The Ca2þ/Mn2þ pumps in the Golgi apparatus. Biochim Biophys Acta 1742(1–3): 103–112 Vanoevelen J, Raeymaeker L, Parys JB, DeSmedt H, Van Baelen K, Callewaert G, Wuytack F, Missiaen L (2004) Inositol-trisphosphate producing agonists do not mobilize the thapsigargin-insensitive part of the endoplasmic reticulum and Golgi Ca2þ store. Cell Calcium 35(2): 115–121 Vanoevelen J, Dode L, Van Baelen K, Fairclough RJ, Missiaen L, Raeymaekers L, Wuytack F (2005) The secretory pathway Ca2þ/Mn2þ-ATPase 2 is a Golgi localized pump with high affinity for Ca2þ ions. J Biol Chem 280(24): 22800–22808  B (1999) Human calumenin Vorum H, Hager H, Christensen BM, Nielsen S, Honore localizes to the secretory pathway and is secreted to the medium. Exp Cell Res 248(2): 473–481

160

*

M. Micaroni et al.

Waldron RT, Short AD, Gill DL (1995) Thapsigargin-resistant intracellular calcium pumps. Role in calcium pool function and growth of thapsigargin-resistant cells. J Biol Chem 272(20): 11955–11961 Wiser O, Bennett MK, Atlas D (1996) Functional interaction of syntaxin and SNAP-25 with voltage-sensitive L- and N-type Ca2þ channels. EMBO J 15(16): 4100–4110 Wuytack F, Raeymaekers L, Missiaen L (2003) PMR1/SPCA Ca2þ pumps and the role of the Golgi apparatus as a calcium store. Pflugers’ Arch 446(2): 148–153 Xiang M, Mohamalawari D, Rao R (2005) A novel isoform of the secretory pathway Ca2þ/Mn2þ ATPase, hSPCA2, has unusual properties and is expressed in the brain. J Biol Chem 280(12): 11608–11614 Yamasaki A, Tani K, Yamamoto A, Kitamura N, Komada M (2006) The Ca2þ-binding protein ALG-2 is recruited to endoplasmic reticulum exit sites by Sec31A and stabilizes the localization of Sec31A. Mol Biol Cell 17(11): 4876–4887 Yoshida A, Oho C, Omori A, Ruwahara R, Ito T, Takahashi M (1992) HPC-1 is associated with synaptotagmin and omega-conotoxin receptor. J Biol Chem 267(35): 24925–24928

Golgi glycosylation enzymes

*

161

Golgi glycosylation enzymes Eric G. Berger and Jack Rohrer

Historical perspective: the Golgi apparatus as the main site of glycosylation From time to time the question is posed by colleagues and research pupils about an object blackened by the classical Golgi techniques [black reaction or reazione nera developed by C. Golgi to identify the Golgi Apparatus], “Is this a Golgi body?” I suggest the proper answer would be: “I do not think that your question has a meaning. It is framed in terms of an improbable hypothesis” (Baker 1953). This telling citation coincides with the end of the long-lasting Golgi controversy about its mere existence; it was finally resolved by the clear definition of the GA1 as an ultrastructural entity (Dalton and Felix 1954). In fact, at these times the believers already recognized that the GA is likely to contain mono- and polysaccharides by virtue of specific histochemical staining (discussed by Bensley (1951)). The breakthrough to recognize the GA as the main cellular site of glycosylation can be traced back to the metabolic incorporation of glucose into cellular components shown by autoradiography to occur in the GA (Neutra and Leblond 1966). The procedure applied by these authors was inspired by Palade’s pioneering work on the secretory pathway (Caro and Palade 1964). The next milestone in associating glycosylation mechanisms with the GA was the advent of fractionation techniques combined with identification of subcellular fractions by marker enzymes such as galactosyltransferase [EC 2.4.1.22]. In the late 1960s a number of groups introduced a corresponding enzyme to identify Golgi fractions which were morphologically assigned to the GA (for review see Farquhar and Palade (1981)). The circle was then closed by the first immunocytochemical staining of the GA using antibodies to this enzyme (Berger et al. 1981), as shown in Fig. 1. Hence biochemical characterization of Golgi fractions was rendered feasible leading to a wealth of data allowing a coherent view on the stepwise assembly of glycans, mainly those of the N-glycosylation pathway of glycoproteins 1 The abbreviations are: AA, amino acids; BFA, brefeldin A; CT, cytoplasmic tail; CFP, cyan fluorescent protein; CDG, congenital disorder of glycosylation; cer, ceramide; fuc, fucose; fuc-T, fucosyltransferase, GA, Golgi Apparatus; GAG, glycosaminoglycan; gal, galactose; GalNAc-T, N-acetylgalactosaminyltranasferase; GalNA, N-acetylgalactosamine; GFP, green fluorescent protein; Glc, glucose; GlcA, glucuronic acid; GlcNAc, N-acetylglucosamine; Gn-T, N-acetylglucosaminyltransferase; GT, glycosyltransferase; hyl, hydroxylysine; man, mannose; PM, plasma membrane; PI, phosphatidyl-inositol; sia-T, sialyltransferase; TGN, trans Golgi network; TMD, transmembrane domain; wt, wild type; xyl, xylose

162

*

E. G. Berger and J. Rohrer

Figure 1. Immunocytochemical staining of b4gal-T, the classical marker enzyme for the Golgi apparatus using polyclonal antibodies to a soluble form of this enzyme secreted in human milk (Berger et al. 1981).

(Kornfeld and Kornfeld 1985). This review is the milestone of the biochemical era. Remarkably, the pathway outlined therein is still valid, but has since been refined and complemented in cell-type-, species-, and development-dependent aspects. Introduction of antibodies to specific Golgi components initiated the cell biological era (Berger et al. 1981; Louvard et al. 1982) soon followed by cloning of the first genuine Golgi enzymes (Narimatsu et al. 1986; Shaper et al. 1986; Weinstein et al. 1987) introducing the molecular era. Both these new approaches have been fully accounted for in the monograph edited for the centenary of its discovery by C. Golgi (Berger and Roth 1997). Where do we stand now? In the past 10 years, new concepts of trafficking, biogenesis and organellar proteomics (Au et al. 2007) have emerged which all will push our current level of ignorance a bit farther away.

The basics Golgi glycosylation enzymes comprise two classes: processing glycosidases and glycosyltransferases. Glycosidases cleave and glycosyltransferases (GTS) create glycosidic linkages. Both are class II membrane proteins with their catalytic portion oriented to the lumen of the GA. Bioinformatics concerning these enzymes are found at http://www.cazy.org (Coutinho and Henrissat 1999). Among the glycosidases, only class I and class II mannosidases are associated with the GA (Herscovics 1999; Mast and Moremen 2006), the remaining glycosylation enzymes being glycosyltransferases. Their general function is depicted on Fig. 2. Catalytic properties are described in the legend of Fig. 2. A hallmark of these enzymes is their high substrate specificity. This property gave rise to the

Golgi glycosylation enzymes

*

163

Figure 2. Glycosyltransferases catalyze the transfer of a sugar residue from a nucleotide-sugar (those acting in the GA are listed on the left side; the abbreviations are given in footnote 1). Catalysis of sugar transfer is classified into retaining or inverting mode depending on the anomeric configuration of the sugar before and after transfer. Usually, glycosyltransferases are activated by bivalent cations (Mn2+ or Ca2+) and follow an ordered compulsory mechanism by first binding the donor, then the acceptor substrate. Michaelis constants for the donor substrates are within the micromolar range whereas those for the acceptors are difficult to estimate; in vitro they are between micro- and millimolar. Optimum pH is around 7. The products are glycosylated acceptor and the nucleotide (NDP) which is immediately cleaved by a luminally acting nucleoside phosphatase to prevent kinetic inhibition. The resulting nucleoside monophosphate (NMP) is exchanged by a specific antiporter with the NDP-sugar synthesized in the cytoplasmic compartment (Caffaro and Hirschberg 2006). The fate of luminally produced phosphate is not known.

one enzyme-one linkage paradigm (Hagopian et al. 1968) predicting a specific glycosyltransferase, thus a gene encoding it, for the formation of each glycosidic bond. While this prediction was a fruitful basis for the search of distinct activities, molecular cloning has now revealed the existence of gene families for many glycosyltransferases with overlapping specificities. These are compiled in the Cazy database. An overview of glycosidic linkages found in mammals has been published (Ohtsubo and Marth 2006). The sequences of human Golgi glycosyltransferases forming the cytoplasmic, transmembrane and stem regions have recently been compared (Patel and Balaji 2007). Their respective topogenetic functions are discussed below. Like all genuine membrane proteins of the secretory pathway with type II topology, glycosyltransferases are synthesized in the ER with an internal signal sequence. They usually contain N- and/or O-linked glycans and follow the membrane flow of the secretory pathway. As discussed in ‘Trafficking of Golgi glycosyltransferases’ section, at some specific stage

164

*

E. G. Berger and J. Rohrer

Figure 3. Golgi GTs are type II membrane proteins composed of four domains: a short cytoplasmic tail (CT), a transmembrane (TMD), a stem and a catalytic domain. They may be proteolytically cleaved in a post Golgi compartment and released as a soluble enzyme. PDB ID: 1O0R (Qasba et al. 2005).

within the Golgi cisternal stack GTs accumulate by mechanisms which are still poorly understood and which may involve complex formation (retention) and/or recycling by vesicular transport (retrieval) (see ‘Trafficking of Golgi glycosyltransferases’ section). Finally, they can move beyond the GA to undergo processing by furin-type proteases and to be released into the extracellular space as catalytically active, soluble enzymes devoid of their membrane anchor and part of the stem region (Fig. 3). Clearly, the mechanisms governing assembly and sequential arrangement of Golgi GTs are at the center stage of current efforts to understand the molecular mechanisms of Golgi-associated glycosylation.

Structural aspects Domain structure, topology Elucidation of the domain structure of Golgi GTs (see Fig. 3) was an important milestone (Paulson and Colley 1989). In a bioinformatics approach, the sequences of the domains of human GTs have recently been characterized (Patel and Balaji 2007): the cytoplasmic

Golgi glycosylation enzymes

*

165

domain usually comprises 6–10 AAs but varies between a few and 80. They may contain motifs important for topogenesis (‘Summary and outlook’ section). The transmembrane domain contains 17–33 residues which would correspond to a length of up to 4.2 nm in an a-helical conformation. It is flanked by more positively charged residues (Arg, Lys) on the N-terminal than on the C-terminal side as predicted for type II membrane proteins according to the “positive-inside” rule (Higy et al. 2004). Moreover, in several cases intramembrane Cys residues possibly involved in dimer formation have been found (Qian et al. 2001; Sousa et al. 2003). The stem domain tethers the catalytic portion and enables molecular encounters of enzyme and acceptor, both anchored to the membrane; thus, the stem containing 70% of disorderpromoting AAs is believed to be flexible. The length varies between 10 and 165 AA and probably reflects the distance needed to bind the corresponding acceptor substrate; in accordance with this idea, glycolipid-specific GT appear to have shorter stem domains than those elongating N- or O-glycans.

3D structures of glycosyltransferases 3D structures are available for the catalytic portions of a number of Golgi GTs all belonging to the GT-A class as reviewed by Qasba et al. (2005) and Breton et al. (2006). A synopsis of the hitherto available structures is available at the following website: http://www.cermav.cnrs.fr/glyco3d/. These structures reveal common binding motifs (DXD or EXD) for the metal ion, the nucleotideactivated sugar followed by the acceptor substrate in an ordered sequential reaction. They also infer on the molecular mechanisms of the two catalytic processes, i.e. inversion and the less well understood retention of anomericity of the donor sugar. Moreover, substrate-induced conformational changes of a loop which acts as lid covering the donor substrate could be visualized; this in turn opens the space to bind the acceptor substrate (Qasba et al. 2005). Moreover, they form a rationale for designing GTs with defined specificities (Hancock et al. 2006).

Glycosylation pathways The GA is an impressive biosynthetic machine of glycans which accounts for the bulk of biomass in the living world. Notwithstanding the fact that an important glycosaminoglycan (hyaluronic acid) is synthesized at the plasma membrane in animal cells (Rilla et al. 2005) most glycans are assembled in the GA. These include peripheral sugars of protein-bound N-glycans and Oglycans, glycolipids and proteoglycans. A widely accepted view on this process predicts an assembly-line in which sequentially ordered GTs would elongate a glycan chain on a protein or lipid substrate along its passage through the GA. Several predictions can be made from this assumption: (1) glycosylation enzymes are expected to build-up in the GA in a steady state (i.e. inflow from endoplasmic reticulum (ER)-associated biosynthesis into the GA matches their outflow); (2) glycosylation enzymes would not be uniformly distributed

166

*

E. G. Berger and J. Rohrer

along the Golgi cisternal stack; instead they would be sequentially arranged along the cis–trans axis of the GA; (3) co-localization of enzymes specific for the same acceptor substrate would result in competition. The first postulate has amply been confirmed and can be used as a litmus test for the specificity of antibodies to Golgi glycosylation enzymes: they have to provide the characteristic picture of the usually compact juxtanuclear localization of the GA at the level of the light microscope (cf. Fig. 1). There are cell-type specific variations but a close topographic relationship to the MTOC is invariably seen (Tassin et al. 1985). The best evidence for the second postulate relates to Golgi glycosylation enzymes involved in chain elongation and termination of N-glycans shown to be sequentially arranged (Rabouille et al. 1995). The third postulate has been documented in many instances; a recent striking example addresses competition of overexpressed fuc-T IV (EC 2.4.1.-) with a-gal-T (EC 2.4.1.87) for the terminal LacNAc structure leading to downregulation of the a-gal epitope (Hansen et al. 2005). The study of the topography of glycosylation enzymes with respect to the Golgi cisternal organization relies on different methods, each having intrinsic limitations as reviewed by Varki (1998): these included in the early days biochemical purification of GTs (Beyer et al. 1981) or their heterologous expression, immunocytochemical localization of the enzymes (Roth and

Figure 4. Glycan classes. The scheme depicts the sugars attached to the aglycone. Relevant to Golgi associated reactions are the elongation and termination of N-glycans, initiation of glycans, polymerization of glycosaminoglycans (GAG) and biosynthesis of glycolipids. The symbols designate: &GlcNAc; *Gal; Man; 4 Fuc; & GalNAc; Glc; rXyl.



Golgi glycosylation enzymes

*

167

Taatjes 1998; Roth 1998), fractionation of the GA (Bretz et al. 1980), use of Golgi disturbing agents (Dinter and Berger 1998), intercompartmental transport assays (Pfeffer and Rothman 1987). Newer approaches include studies by life microscopy of GFP-tagged GTs (Lorenz et al. 2006). The concerted actions of Golgi GTs result in specific glycosylation pathways which can be classified according to (i) type of aglycone (lipid or protein), (ii) the first sugar linked to it and (iii) the species. An overview of all known start points of glycosylation pathways is given on Fig. 4: out of them, only those which are initiated or elongated in a mammalian GA are described in more detail. These include chain extension and termination of N-glycans and Oglycans as paradigms for further work aimed at understanding the molecular organization of these enzymes.

N-glycosylation N-glycosylation of proteins is initiated in the endoplasmic reticulum by en bloc transfer of the preformed oligosaccharide Glc3Man9GlcNAc2 from its dolichol carrier to the nascent polypeptide chain by the multimeric enzyme oligosaccharyltransferase. ER-associated glycans then fulfill a variety of intracellular functions such as quality control, sorting, degradation and secretion (for recent review see Helenius and Aebi (2004)). Their detailed description is beyond the scope of this chapter. Briefly, following transfer of the oligosaccharide to the peptide chain, one man residue is cleaved in the ER by mannosidase I (for review see Spiro (2004)). Glycoproteins then move to the GA (as outlined in ‘Domain structure, topology’ section) to arrive to the cis cisterna. The basic scheme of Golgi-associated trimming, chain elongation and termination reactions as proposed 20 years ago by (Kornfeld and Kornfeld 1985) is still basically valid and is depicted in Fig. 5. This pathway can be viewed as a theme with many species and cell-type specific variations. The final products are glycans exerting their functions (i) outside of the cell as part of secreted or shed glycoproteins, (ii) on the plasma membrane exposing the glycans on the cell surface, (iii) in post Golgi recycling compartments and (iv) in lysosomes. The enzymes modifying the N-glycans act sequentially from cis to trans as part of the multiglycosyltransferase system (Roseman 1970) characteristic for a given cell-type. Thus, the product of one becomes the substrate for the next enzyme. Annotations of these enzymes are available at the database CAZy (http://www.cazy.org/). Another useful website addressing GTs is found at http://www.functionalglycomics.org/glycomics/molecule/jsp/ glycoEnzyme/geMolecule.jsp. In the following, the main steps restricted to a human GA will be reviewed and recent findings mentioned. 1. 1a. The first Golgi-associated step is the conversion of the Man8GlcNAc2 N-glycan by mannosidases IA and IB to Man5GlcNAc2 (EC 3.2.1.113) as reviewed by Herscovics (Herscovics 1999). Recently, an additional type designated mannosidase IC has been cloned from a human fetal brain

168

*

E. G. Berger and J. Rohrer

Figure 5. General N-glycosylation pathway in the Golgi apparatus. N-glycosylated proteins carrying a Man8GlcNAc2 glycan move by vesicular transport to the cis-Golgi where further processing (arrow 1a) takes place. Alternatively, lysosomal enzymes are substituted by one to two GlcNAc residues linked by a phosphodiester bond (arrow 1b). They are not further modified until removal of the GlcNAc by an uncovering enzyme in the TGN to expose the mannose-6 phosphate recognition marker. All other N-glycans may be elongated and terminated in the medial and trans-Golgi cisternae as depicted. Many alternative or additional substitutions are possible depending on the cell-type. These include core fucosylation (step 5), branching up to pentaantennary N-glycans and a number of different terminal structures. All symbols are explained in the legend to Fig. 4. ¤ designates sialic acid. The numbered arrows refer to the steps explained in the text. Detailed glycosidic linkages are shown on Fig. 6.

cDNA library and shown to have a tissular expression pattern different from IA and IB and a slightly different fine specificity in the order of cleavage leading to different intermediate compounds (for review see Herscovics (2001)). The final product of the class I mannosidases is the Man5GlcNAc2 species. 1b. An alternative glycosylation reaction substitutes lysosomal enzymes with a GlcNAc residues linked by a phosphodiester bond. The corresponding enzyme, the UDPGlcNAc: lysosomal enzyme-1-phosphotransferase (EC 2.7.8.15), a hexameric enzyme, has been purified (Bao et al. 1996), cloned (Kudo et al. 2005) and analyzed for pathogenic mutations (Tiede et al. 2005). The a/b subunits may carry loss of catalytic function mutations leading to mucolipidosis II (OMIM 252500) whereas defects of the g subunit which specifically recognizes lysosomal enzymes cause mucolipidosis III (OMIM 252600).

Golgi glycosylation enzymes

*

169

2. Man5GlcNAc2-R then serves as a substrate for Gn-T I (EC 2.4.1.101) thereby initiating synthesis of hybrid or complex type N-glycans. This enzyme is of critical importance in the development of higher eukaryotes since mice in which the gene was ablated died around embryonic day 10 (Campbell et al. 1995). In fact, this enzyme has been extensively investigated in C. elegans leading to the surprising observation that worms have very little complex or hybrid N-glycans (Schachter 2004). A homologous enzyme encoded by a different gene has been described and designated Gn-T I.2 which may also be involved in extension of O-linked mannose (Zhang et al. 2002). 3. The ensuing hybrid-type glycan GlcNac1Man5GlcNAc2 becomes a substrate for processing mannosidase II (EC 3.2.1.114), a class II mannosidase (Moremen 2002). The alternative homologous enzyme mannosidase X (EC no entry; for recent discussion see: Akama et al. (2006)) may substitute for the former in case of its ablation. Persistence of the two remaining mannose residues by mannosidase II leads to the formation of hybrid glycans with different biological properties. Hence the development of inhibitors to this enzyme other than the classical compound swainsonine has been a subject of considerable interest in recent years (Kawatkar et al. 2006). 4. Following removal of mannose residues, Gn-T II (EC 2.4.1.143) adds another GlcNAc residue to the 1!6 branch to form GlcNAc2Man3GlcNAc2-R. Transcriptional regulation of this enzyme has been investigated and shown to depend on the Ets transcription factors but not src or neu (Zhang et al. 2000). The biosynthetic importance of this enzyme is also underlined by its defect in CDG IIa characterized by frequent postnatal lethality with multiple defects (OMIM 212066). 5. The formation of GlcNAc2Man3GlcNAc2-R permits a6fuc-T8 (EC 2.4.1.68) to transfer a fucose residue to the innermost GlcNAc, a process called core fucosylation. Also triantennary N-glycans with terminal GlcNAc residues can be core fucosylated whereas bisected glycans (see below) are not substrates. Expression of this enzyme is upregulated in a variety of cancer cells (Miyoshi et al. 1999). In addition, the crystal structure of the human enzyme has recently been reported (Ihara et al. 2007).

Figure 6. Branching of N-glycans. Specific N-acetylglucosaminyltransferases act in a cellspecific manner to form up to five branches (pentaantennary N-glycan). The depicted hexaantennary structure has not been observed in vivo.

170

*

E. G. Berger and J. Rohrer

6. At this critical stage, either further branching or chain elongation to biantennary complex N-glycans may take place. A detailed insight into the numerous possibilities of branching by Gn-Ts is provided in the classical GO-NOGO lecture by Schachter (1986). One of the six branching reactions (Fig. 6) is catalyzed by “bisecting” Gn-T III (not shown on Fig. 5; EC 2.4.1.143), which received a great deal of attention in recent years (for review see Ikeda and Taniguchi (2001)). Remarkably, this branch is never extended and prevents further processing by mannosidase II, branching by core fuc-T, or importantly, Gn-T II, Gn-T IV (EC 2.4.1.145) and Gn-T V (EC 2.4.1.155) (Zhao et al. 2006). However, the enzyme can act on any bi-, triand tetraantennary form of N-glycans provided that they contain no galactose. Expression of Gn-T III appears to exert a pivotal role as it impairs branching by Gn-T V, a branch conferring metastatic potential to melanoma or NIH3T3 cells (for review see Gu and Taniguchi (2004)). 7. All the intermediate species are substrates for Gn-T IV which adds a b4GlcNAc to the 3-branch of core mannose leading to a triantennary species, or tetraantennary species if Gn-T V has already acted on the 6branch of core mannose. A second isoform exists with an ubiquitous tissue distribution (Yoshida et al. 1998). An intriguing increase of its expression has been observed in choriocarcinoma where overexpression of this enzyme forms so-called abnormal biantennary glycans, i.e. glycans lacking the product of Gn-T II (Takamatsu et al. 1999). 8. The a6 branch of core mannose can be substituted by Gn-T V (not shown on Fig. 5) which is also represented by a genetically distinct isoform designated Gn-T VB (Kaneko et al. 2003) or Gn-T IX with predominant expression in the brain (Inamori et al. 2006). Most interestingly, Gn-T VA has been implicated in a number cancer-associated changes of glycosylation as this enzyme was found to be responsible for increased branching of N-glycans in highly metastatic cancer cells (Dennis 1988); moreover this branch may carry polylacNAc structures which also contribute to the increased size of the N-glycans, the “Warren phenomenon” known since the seventies (Warren et al. 1978). Gn-T VA has a complex promoter set-up with several binding sites for transcription factors (Saito et al. 1995) and, remarkably, PEA3/Ets binding sites which can by activated by RAS-RAFMAPK signalling (Buckhaults et al. 1997). An entirely new connection of the b1,6 branch introduced by Gn-T V has been revealed in Mgat5(  /  ) mice: they appeared to be less sensible to anabolic cytokines since their receptors have reduced surface expression in absence of this branch (Cheung et al. 2007). While pentaantennary structures are commonly found, hexaantennary structures, although theoretically possible, have not been reported. 9. N-glycans substituted with GlcNAc residues then move by mechanisms addressed in ‘Trafficking of Golgi glycosyltransferases’ section to the transGolgi compartment, the site of chain elongation and termination. TransGolgi cisternae harbour gal-Ts (Roth and Berger 1982), sia-Ts (Roth et al.

Golgi glycosylation enzymes

*

171

1985) and probably a number of terminating GTs which have not been mapped yet by immunoelectron microscopy. The first reaction occurring in trans-Golgi cisternae is galactosylation of GlcNAc residues. Most commonly, this reaction is catalyzed by the ubiquitous (though downregulated in neural tissue) b4gal-T1 (EC 2.4.1.22), probably the most thoroughly investigated mammalian GT (Berger and Rohrer 2003). Interest in this enzyme dates back to an intriguing finding of a high concentration of soluble gal-T in embryonic chicken brain later implicated in specific intercellular adhesion (Den et al. 1970). This line of research pertains to ectopic localization of Golgi GTs reviewed by Berger (2002) and is not addressed in this review. Galactosylation takes place by two classes of enzymes, the b1!4 and b1!3gal-T (for reviews see Amado et al. (1999) and Hennet (2002)). While the glycosidic linkage formed by these families is the same among their members, the acceptor substrates are different and define the biological role. Among all gal-Ts acting along this pathway, only gal-T1 has been deleted in mice (Asano et al. 1997; Lu et al. 1997) leading to postnatal lethality, endocrine defects and skin abnormalities. This enzyme also served as a unique paradigm for a protein modifier function: its binding of a-lactalbumin, a protein uniquely expressed in lactating mammary gland shifts recognition of GlcNAc-R as acceptor to glucose thereby forming lactose (Hill and Brew 1975). At present, interest in this enzyme focuses on trafficking, complex formation, shedding and possible regulation by phosphorylation. 10. 10a. Biosynthesis of N-glycan chains very often is terminated by sialylation which appears to immediately follow galactosylation. Although co-localization of endogenously expressed gal-Ts and siaTs has been confirmed at the level of light microscopy (Taatjes et al. 1987), evidence for co-localization at the ultrastructural level was only possible with transfected sia-T (Kweon et al. 2004). Out of the numerous different sia-Ts reviewed by Harduin-Lepers et al. (2005), ST6Gal1 has been investigated with respect to its targeting in some detail (see ‘Trafficking of Golgi glycosyltransferases’ section). Another paradigmatic feature of Golgi GTs was the identification of sequence motifs involved in donor or acceptor substrate binding, respectively. These were called sialyl motifs (Datta and Paulson 1995; Datta et al. 1998) and shown to be present also in polysiaTs (see below). Chain termination may involve other GTs as schematically shown on Fig. 7. An alternative to 2!6 is 2!3 sialylation (EC 2.4.99.6) depending on the relative expression of 2,3 versus 2,6 sia-Ts (Lee et al. 1989). 10b. Synthesis of the blood groups is catalyzed by allozymes encoded on the ABO locus; differences of four AAs between the blood group B (a3gal-Ts, EC 2.4.1.37) and blood group A forming enzyme (a-galNAc-T, EC 2.4.1.40) are known since their cloning (Yamamoto et al. 1990); however, the switch of their donor substrate specificity

172

*

E. G. Berger and J. Rohrer

Figure 7. Biosynthesis of ABO(H) histo blood groups. Symbols are defined in the legend to Fig. 4

appears to rely on a single nucleotide exchange only leading to a P234S mutation. Thus, affinity changes from UDPGal to UDPGalNAc (Marcus et al. 2003). This type of inference on the basis of substrate specificity of a GT also lays the foundation for enzyme engineering projects as outlined in Hancock et al. (2006). The 0 enzyme most frequently is truncated at position 117 (instead of the full length 354 AAs) but a full length variant devoid of catalytic activity but Golgilocalized has been described (Amado et al. 2000). This raises the intriguing question whether the phenomenon of a correctly expressed yet inactive enzyme is an incidentally recognized tip of an iceberg. Blood group specifying GTs act only on fucosylated terminal galactoses as shown on Fig. 7. Thus, corresponding fuc-Ts precede their action to form the H epitope (more commonly designated 0 blood group) or Se (EC 2.4.1.69) on both type 1 and type 2 LacNAc’s. These two types of fuc-Ts are encoded by the FUT1 or the FUT2 gene, respectively. Together with fuc-Ts involved in biosynthesis of the Lewis structures (thoroughly reviewed by Lowe (1995)) a sizable diversity of structures can be made in the distal Golgi compartments as recently reviewed by Ma et al. (2006). A common theme of these structures is their involvement in cellular adhesion and recognition

Golgi glycosylation enzymes

*

173

events both important in inflammation (Rosen 2004) and cancer metastasis (Borsig 2004). 10c. A variety of other chain termination reactions have been described such as polysialic acid, a polymer of a2!8 linked sialic acid residues. This structure extends a2,3 sialic acids of N-glycans preferably attached to the neural cell adhesion molecule (Bonfanti 2006) and bears important properties during embryonic neurogenesis. It is synthesized by two synergistic polysia-Ts designated ST8Sia II and ST8Sia IV (EC 2.4.), respectively (reviewed by Angata and Fukuda (2003)). Considering the conventional topology of these enzymes as luminally oriented catalytic sites, polysialic acid most likely occurs on the luminal side of trans-Golgi cisternae although a cytoplasmic orientation has been proposed (Bonfanti 2006). 10d. An important terminal biosynthetic reaction is sulfation catalyzed by a variety of sulfo-Ts with similar domain structure and topology as the GTs. The donor substrate is 30 -phosphoadenosine 50 -phosphosulfate (PAPS) synthesized in the cytoplasm and transported across the Golgi membrane alike the sugar nucleotides (Abeijon et al. 1997). Sulfo-Ts occur in the cytoplasm as well as in the GA both enzyme species belonging to a single gene superfamily with some common features of their enzymatic mechanisms (Negishi et al. 2001); those directed to the secretory pathway carry an internal signal sequence specifying their import into the endoplasmic reticulum; however, little is known concerning their further topogenesis within the Golgi cisternal stack. Golgi-associated sulfo-Ts either decorate in a highly specific manner glycosaminoglycans on their repetitive carbohydrate backbone or they add sulfate to terminal glycans or tyrosines on specific proteins as schematically shown on Fig. 8 (reviewed by Chapman et al. (2004). 11. The sulfated products exert highly specialized functions as also highlighted by an emerging group of genetic defects known as sulfation defects, in particular those caused by deficient 6-O-sulfotransferase-1 entailing spondyloepiphyseal dysplasia (OMIM 603799 (Thiele et al. 2004) and corneal GlcNAc-6-sulfo-T (C-GlcNAc6ST) causing macular corneal dystrophy (OMIM 217800) (Akama et al. 2000). An intriguing structure is the HNK-1 epitope almost exclusively found in neural tissue involved in neural development (Schachner et al. 1995); the role of the terminal sulfate, however, is not clear, since knockout-mice for the corresponding sulfo-T did not reveal any phenotype (Chou et al. 2002). The precursor structure, e.g. a glucuronic acid residue attached in b1!3 to LacNAc is formed by two specific glcA-Ts, GlcAT-P and GlcAT-S (EC not specified) of which GlcAT-P in fact seems to be involved in spatial memory formation (Yamamoto et al. 2002) and whose catalytic domain has recently been crystallized (Kakuda et al. 2004).

174

*

E. G. Berger and J. Rohrer

Figure 8. Terminal glycans (including sulfates) expressed on a restricted set of protein carriers reflecting cell specificity of terminal GT and sulfotransferase expression.

O-glycosylation On Fig. 4 the currently known O-glycans are listed. Only the most abundant O-glycans initiated by GalNAc-Ts (EC 2.4.1.41) are exclusively synthesized in the GA whereas attachment of O-linked mannose and O-linked fucose as well as collagen glycosylation is ER-associated. The former may be extended by Golgi-associated GTs in analogy to O-linked GalNAc. O-linked GlcNAc is a special case as it is reversibly transferred to cytoplasmic and nuclear protein acceptors by a cytosolic Gn-T and appears not to be elongated by further carbohydrates (Hart et al. 2007). 1. The first step involves transfer of GalNAc to a Ser or Thr residue of a mucin-type domain of a glycoprotein catalyzed by one of the 18 different UDPGalNAc:polypeptide GalNAc-Ts (EC 2.4.1.41) hitherto listed on the CAZy database, enzyme family 27 (http://www.cazy.org/fam/GT27.html) (Ten Hagen et al. 2003). A general scheme of O-glycan biosynthesis is given on Fig. 9, again as a theme with many variations dependent on celltype, developmental stage and cellular differentiation. A few hallmarks distinguish the O- from the N-glycan biosynthetic pathway: (i) O-glycan biosynthesis is confined to the GA; (ii) the high diversity of genetic isoforms of polypeptide GalNAc-Ts reflects their restricted specificity to the different polypeptide backbones; (iii) the topography of O-glycosylation

Golgi glycosylation enzymes

*

175

Figure 9. Simplified scheme of O-glycosylation initiated by GalNAc-Ts. GalNAc-Ts seem to be located throughout the Golgi stack (see text). Thus, truncated glycans may be expressed on the cell surface such as the TF (Thomsen-Friedenreich), Tn and sialosyl-Tn antigens. In many cells, only the glycophorin type of the O-linked tetrasaccharide is formed while in others (e.g. activated lymphocytes) a polylacNAc extension including a sialyl-Lex epitope is synthesized on a core 2 structure.

reactions is largely unknown, as the three different GalNAc-Ts localized by immuno electron microscopy did not reveal any cisternal preference for GalNAc-T1, whereas GalNAcT2 and T3 were preferentially labeled over the trans cisternae (Rottger et al. 1998); these results suggested that initiation of O-glycan biosynthesis may occur at any stage along cisternal progression and may partially explain the variable size of O-glycans on defined proteins; (iv) an interesting kinetic phenomenon has been observed in the case of GalNAc-T2 and -T4 in that a lectin domain of the enzyme specific for O-linked GalNAc-peptide promotes further transfer of GalNAc residues thereby increasing the density of O-glycans (Wandall et al. 2007). 2. The second step in O-Glycan biosynthesis may involve formation of the tumor-associated antigen sialyl-Tn by ST6GalNAc II (EC 2.4.99.3) (Sewell et al. 2006). A more common alternative is chain extension by b3gal-T also designated T-transferase (C1GalT1 or EC 2.4.1.122) as this enzyme is involved in the formation of the Thomsen–Friedenreich antigen and whose deficiency may cause the Tn-syndrome (Berger 1999). Cloning of this enzyme by Cummings and associates revealed a hitherto unique feature in GT enzymology as the activity in vivo depends on the co-

176

*

E. G. Berger and J. Rohrer

expression of an X-linked chaperone designated COSMC or C1GalT2 (Ju and Cummings 2002) which may be mutated in the Tn-syndrome (Ju and Cummings 2005). Defects of both C1GalT1 or C1GalT2 have been associated also with IgA nephropathy (Barratt et al. 2007). 3. The structure galb1!4GalNAc-R (TF-antigen) is usually sialylated to form the glycophorin type tetrasaccharide depicted on Fig. 9. by ST3Gal I (EC 2.4.99.4) and ST6GalNAc IV (EC 2.4.99.7). 4. Formation of core 2 structures as depicted on Fig. 9 depends on expression of core 2 Gn-T (EC 2.4.1.102), which is up-regulated in T lymphocytes upon stimulation as initially shown by Piller et al. (1988). These may then be extended by repeating LacNAc’s and terminated by ABO(H) or sialylLewis blood groups. In fact, seven different core structures have been described which all may be elongated. A detailed description is given by Brockhausen (1995).

Glycolipids A huge diversity of glycolipids is synthesized in the Golgi stack essentially along a similar assembly-line like the mechanism described above for protein glycan chain elongation and termination. A few aspects specific to glycolipid biosynthesis are as follows: (i) the first glycosylation reaction, i.e. the formation of glucosylceramide, occurs on the cytoplasmic face (Coste et al. 1986) on the cis side of the GA where it can be bound by FAPP2 for transport to more distal sites of the GA (D’ Angelo et al. 2007). Very recent results also suggest that part of cytoplasmically oriented glucosylceramide is flipped to the luminal side in the ER whence elongation and termination of the glycolipids occurs along the transit through the GA (Halter et al. 2007). A detailed chart of biosynthetic pathways is available (Ichikawa and Hirabayashi 1998).

Proteoglycans The four main classes of proteoglycans comprise sulfated forms of keratan, chondroitin, dermatan and heparan. Biosynthesis of proteoglycans includes ER-associated steps such as translation and import of the core protein and glycan chain initiation by transfer of xylose. Further steps include core glycan elongation by two galactoses followed then by chain extension with alternating sugars. These occur in the GA. A common hallmark is their assembly in repeating disaccharides of variable length and the modifications imparted to the sugar chains in parallel to their assembly. An interesting and somewhat unique feature in glycan biosynthesis are the two established cases of families of bifunctional enzymes participating in the assembly of heparan sulfate: these are known as copolymerase EXT1 bearing two catalytic sites in tandem, one specific for the transfer of a4GlcNAc, the other for b4GlcA (EC 2.4.1.223). Concerted action of both catalytic sites produces repeating units of GlcNAca4GlcAb4. These structures

Golgi glycosylation enzymes

*

177

are modified by N-deacetylase N-sulfotransferase (NDST1) (EC 2.8.2.8.) leading to sulfation of specific sites of GlcNAc. Little is known about the molecular interactions of the alternating GTs and their topography along the cisternal stack. An interesting view on their putative supramolecular structure is its designation as a “gagosome” (Esko and Selleck 2002): this would be a physical complex of enzymes involved in polymerizing the glycosaminoglycan backbone along with their modifications. Such a complex could explain the speed of the enzymatic reaction by substrate channelling and the formation of bifunctional enzymes during the evolution. It will be a challenging task to investigate how this concept fits to the notion of cisternal progression.

Comparative and evolutionary aspects Details of the glycosylation pathways described above refer to the human species; however, it is clear that all eukaryotes comprising as diverse species as yeasts, plants, worms, insects and mammals all synthesize their glycome by virtue of a GA. These aspects are dealt with in chapter 4.5. The basic principles of glycan assembly are common to all eukaryotes. The changes in evolution concern expression and specificities of the GT repertoire in a given organism (Varki 2006). This evolutionary change occurs over time by proneness to infections as a result of specific interactions between the host glycans and infectious agents. Alternatively, also sexual selection can lead to shifts in the glycan make-up as the example of the ABH blood groups show. Shifts of specificities of GTs by random genetic drift may always implicate pleiotropic changes with metabolic and morphogenetic consequences. This is exemplified by the genetic defects designated congenital disorders of glycosylation (Freeze 2006) which, in fact, cover the entire spectrum from embryonic lethality in case of complete enzymatic knock-out (for example b3galT, Xia et al. (2004)) to no apparent phenotypic changes in the case of absence of blood groups A or B. In between, GT polymorphisms can lead to all kinds of subtle phenotypic differences.

Trafficking of Golgi glycosyltransferases Sorting machineries required for the sorting of glycosyltransferases Following synthesis in the ER (see ‘The basics’ section), GTs are concentrated at ER exit sites and are packaged in transport vesicles that, in mammalian cells, seem to undergo homotypic fusion to form the ER–Golgi intermediate compartment before being delivered to the cis cisterna of the Golgi stack. Upon arrival in the GA the cell faces the difficult task to sort the individual GTs to specific cisternae of the GA (cis-, medial- or trans-Golgi) where at steady state the majority of a particular enzyme accumulates to perform its function within pathways outlined in ‘Glycosylation pathways’ section of this chapter. To determine how the cell can maintain this asymmetrical distribution of GTs in the GA while at the same time ensuring efficient passage of newly synthesized proteins through the secretory pathway is a major challenge for

178

*

E. G. Berger and J. Rohrer

cell biologists. Chapters 2.6 and 3.2 address current models and the molecular machineries of intra Golgi trafficking. In the context of GTs, the conserved oligomeric Golgi (COG) complex appears on center stage. It consists of two lobes each comprising of four proteins. The precise role of the COG complex is not known so far but current evidence suggests that the COG complex is required for the retrograde trafficking of GTs (Chatterton et al. 1999; Kingsley et al. 1986; Ungar et al. 2002; Wu et al. 2004; Zolov and Lupashin 2005). Somatic mutations and siRNA experiments in mammalian cells as well as subtypes of CDG (Wu et al. 2004) showed global defects in protein glycosylation which are at least in part due to the mislocalization or instability of GTs. Despite intense research over the past decades it is still under debate whether the mechanism for the correct localization within the GA of different GTs is determined by the cytoplasmic-, transmembrane- and/or luminal domain (reviewed originally by Colley (1997)). Most likely this has to be assessed for each GT individually and possibly there are multiple signals/ mechanisms to localize the enzymes to their correct place within the GA.

Trafficking of galactosyltransferases At steady state the b4gal-T1 (EC 2.4.1.22) was found to be localized to the trans cisterna of the GA as determined by immuno electron microscopy (Roth and Berger 1982). Initial studies identified the transmembrane domain of b4gal-T1 as the important feature for the correct Golgi localization (Aoki et al. 1992; Nilsson et al. 1991; Teasdale et al. 1992). However, further data imply additional domains like the cytoplasmic domain to be required for efficient localization (Evans et al. 1993; Nilsson et al. 1991; Russo et al. 1992). Using a GFP-tagged form of the b4gal-T1 or the endogenous form, it was shown that the enzyme cycles between the trans-Golgi and the ER (Rhee et al. 2005; Zaal et al. 1999). Most recently (Schaub et al. 2006) could demonstrate that the cytoplasmic domain contains a specific signal that is required for the transport of b4gal-T1 from the trans-Golgi to the TGN. In contrast to the vesicular transport model, this result is unexpected in the view of a pure cisternal maturation model which would not require signals for proteins to be included in anterograde transport. In fact, mutating such a signal prevented delivery of b4gal-T1 to the TGN (Schaub et al. 2006, unpublished results). Either b4gal-T1 represents an exceptional case as it is also cycling back to the ER (Zaal et al. 1999) and that there are competing transport events which might be regulated by special signals or the cisternal maturation model has some unexpected requirements for the transport of Golgi enzymes between the trans-Golgi and the TGN. In any case, a leakage at the level of the recycling from the TGN back to the trans-Golgi could also explain the presence of a small fraction of b4gal-T1 at the surface in some cells (reviewed by Shur et al. (1998)). A soluble form of the enzyme was originally observed in bovine milk (Babad and Hassid 1966) and later purified from human milk (Gerber et al. 1979). However, it is not known if the soluble form is created by clipping the enzyme in the TGN with subsequent exocytosis (Strous and Berger 1982) or at the cell surface or both.

Golgi glycosylation enzymes

*

179

For a3gal-T (EC 2.4.1.87) it was shown that overexpression of a soluble form of the cytoplasmic domain led to a mislocalization of the enzyme and that replacement of the TMD with a TMD from a PM protein did not affect the Golgi localization indicating that the cytoplasmic domain is a critical feature for its correct localization (Milland et al. 2002).

Trafficking of sialyltransferase ST6Gal1 (EC 2.4.99.1) seems to have a complicated mechanism or rather multiple mechanisms for its localization in the trans-Golgi cisterna as several reports using different experimental systems implicate different domains that are required for correct localization of the enzyme. Early on, Colley et al. (1989) found that sequences required for Golgi localization are within the CT, TMD and/or stem region. Later it was suggested that the TMD and flanking regions are sufficient for the Golgi localization with complete retention in COS and CHO cells only if CT and stem were present (Dahdal and Colley 1993; Munro 1991). Gradually increasing the length of the TMD resulted in the mislocalization of an ST6Gal1-lysozyme chimera to the plasma membrane (Munro 1995) but replacing the TMD of ST6Gal1 by the longer TMD of neuraminidase did not affect the localization (Dahdal and Colley 1993). Furthermore, the formation of disulfide bonded dimers of ST6Gal1 was assumed to promote retention of the enzyme in the GA as only monomers were found to be secreted (Chen et al. 2000; El-Battari et al. 2003) but Qian et al. (2001) identified Cys24 within the TMD as crucial for dimer formation explaining the absence of secreted dimers in the media. Recently, a very elegant study by Fenteany and Colley (2005) found that on one hand the CT is required for Golgi localization and on the other hand that oligomerization of the enzyme, mediated by luminal sequences, is required as well. This leads to a model where ST6Gal1 is concentrated in the Golgi due to its TMD and cytoplasmic domain which subsequently leads to the oligomerization mediated by the luminal sequences (Cys123 in particular).

Trafficking of N-acetylglucosaminyltransferase Early experiments using chimeric proteins between type II surface proteins (dipeptidyltransferase and transferrin receptor) and of Gn-T1 (EC 2.4.1.101) revealed that the TMD of Gn-T1 is required for localization of the enzyme in the medial/trans-Golgi (Tang et al. 1992). This was initially confirmed by Burke et al. (1992) using the TMD and immediate flanking regions to retain ovalbumin in the GA. However, a more detailed analysis later demonstrated that all three domains of the Gn-T1 contribute to the efficient localization within the GA (Burke et al. 1994). Further dissecting the role of the luminal domain of the GnT3 (EC 2.4.1.144) with a comprehensive examination of the three N-glycosylation sites of the enzyme revealed that not only the enzyme activity but also Golgi localization decreased by reducing the number of glycosylation sites (Nagai et al. 1997). Moreover, additional work from Taniguchi’s group showed that part of the stem region of the Gn-T5 (EC 2.4.1.155) participates in Golgi

180

*

E. G. Berger and J. Rohrer

retention through disulfide bond mediated homo-oligomer formation without affecting the enzymatic activity (Sasai et al. 2001). In contrast, analysis of constructs containing domains of the human O-glycan core 2 Gn-T (C2Gn-T, EC 2.4.1.102) fused to GFP clearly identified the CT and TMD but not the stem region as the necessary and sufficient parts of the enzyme for Golgi localization (Zerfaoui et al. 2002). The authors could also show that the CT and TMD of C2Gn-T fused to the luminal domain of fuc-T7, (EC 2.4.1.-) displaced the chimera from a trans-Golgi cisterna to cis-/medial-cisterna as analyzed by changes of the cell surface glycosylation pattern. Taken all results into account it becomes obvious that the different Gn-T’s either require all three domains for efficient Golgi localization but each of the domains might mediate a different function within the scheme or that the individual Gn-T’s might use different mechanisms to achieve Golgi localization and therefore require just one or two domains for this task.

Trafficking of fucosyltransferase Early studies demonstrated that fuc-T6 (EC 2.4.1.65) co-localizes with b4gal-T1 in the GA and monensin-induced swollen vesicles (Borsig et al. 1999) indicating a cycling of the enzyme through the TGN (see section on b4gal-T1). Milland et al. (2001) analyzed the significance of the CT for the localization of the fuc-T1 (EC 2.4.1.69) and found that upon deletion of the CT more of the mutant fuc-T1 retained a perinuclear staining pattern after addition of BFA compared to the wt fuc-T1, indicating a mislocalization of the tailless enzyme to the TGN probably due to a missing retrieval/recycling signal. Mutations of individual AAs within the CT (MWVPSRRH) further revealed a role of residues 3–7 for correct localization with a particular role for Ser5, suggesting a potential role of phosphorylation for this process (Milland et al. 2001). In contrast to this, Sousa et al. (2004) reported data based on immunofluorescence that deletion of the CT of fuc-T3 (EC 2.4.1.65) led to a mislocalization of the enzyme to an earlier cisterna of the Golgi than the wt protein. This apparent contradiction between the two studies can hardly be explained by a different experimental set-up but seems rather due to different localization requirements of fuc-T1 and fuc-T3. Analyzing the TMD of fuc-T3 it was shown that mutation of four non-hydrophobic residues (C16, Q23, C29 and Y33) to leucines led to its surface accumulation (Sousa et al. 2003). Furthermore, C16 and C29 were shown to be required for dimer formation of fuc-T3 and all four residues seem to be important for the incorporation of the enzyme into COPI vesicles indicating that the TMD is also required for recycling of the enzyme from later compartments (Sousa et al. 2003).

Trafficking of N-acetylgalactosaminyltransferase Initial experiments with the b4GalNAc-T (EC 2.4.1.92) demonstrated that homodimers are formed as a result of intermolecular disulfide bonds creation in the ER which might be an important feature for the localization (see above) (Zhu et al. 1997). Recently, a study using chimeric molecules between GalNAc-

Golgi glycosylation enzymes

*

181

T and Sia-T2, CT and TMD fused to fluorescent protein tags, revealed that the CTof the GalNAc-T is sufficient to cause accumulation of the chimera mostly in the TGN as indicated by resistance to BFA (Uliana et al. 2006). In these experiments the TMDs did not seem to have an effect for the correct localization of the enzymes.

General remarks on trafficking of glycosyltransferases Taken together, it becomes obvious that the individual GTs do not have one common signal or feature that determines the localization and trafficking pathways for all of them. The enzymes have to be analyzed individually and what has been discovered for one does not necessarily apply to another. It seems that the localization for many GTs is mediated by a mix of oligomerization which requires the TMD and sometimes the stem region, and specific transport signals localized in the CT. The individual requirements and pathways for the correct localization of the GTs within the GA are especially important if the enzymes are used as markers for the GA in general because experiments using one particular enzyme could yield different results if performed using another enzyme.

Summary and outlook There is considerable advance in knowledge on Golgi glycosylation enzymes since publication of the centennial book (Berger and Roth 1997), mainly with respect to 3D-structures, catalytic mechanisms and defects leading to type II of CDG. In some cases new genetic isoforms have been added to the list of known glycosyltransferases complementing our knowledge on cell specificity of glycosylation pathways. An emerging field addresses the still poorly understood topogenetic mechanisms, the definition of interacting partners and activity regulation by complex formation and/or phosphorylation. At present, however, we are still far away from a unifying concept regarding the fine distribution of glycosyltransferases along the Golgi cisternal stack.

References Abeijon C, Mandon EC, Hirschberg CB (1997) Transporters of nucleotide sugars, nucleotide sulfate and ATP in the Golgi apparatus. Trends Biochem Sci 22: 203–207 Akama TO, Nakagawa H, Wong NK, Sutton-Smith M, Dell A, Morris HR, Nakayama J, Nishimura S, Pai A, Moremen KW, Marth JD, Fukuda MN (2006) Essential and mutually compensatory roles of {alpha}-mannosidase II and {alpha}-mannosidase IIx in N-glycan processing in vivo in mice. Proc Natl Acad Sci USA 103: 8983–8988 Akama TO, Nishida K, Nakayama J, Watanabe H, Ozaki K, Nakamura T, Dota A, Kawasaki S, Inoue Y, Maeda N, Yamamoto S, Fujiwara T, Thonar EJMA, Shimomura Y, Kinoshita S, Tanigami A, Fukuda MN (2000) Macular corneal dystrophy type I and type II are caused by distinct mutations in a new sulphotransferase gene. Nature Genetics 26: 237–241

182

*

E. G. Berger and J. Rohrer

Amado M, Almeida R, Schwientek T, Clausen H (1999) Identification and characterization of large galactosyltransferase gene families: galactosyltransferases for all functions. Biochim Biophys Acta 1473: 35–53 Amado M, Bennett EP, Carneiro F, Clausen H (2000) Characterization of the histo-blood group O(2) gene and its protein product. Vox Sang 79: 219–226 Angata K, Fukuda M (2003) Polysialyltransferases: major players in polysialic acid synthesis on the neural cell adhesion molecule. Biochimie 85: 195–206 Aoki D, Lee N, Yamaguchi N, Dubois C, Fukuda MN (1992) Golgi retention of a trans-Golgi membrane protein, galactosyltransferase, requires cysteine and histidine residues within the membrane-anchoring domain. Proc Natl Acad Sci USA 89: 4319–4323 Asano M, Furukawa K, Kido M, Matsumoto S, Umesaki Y, Kochibe N, Iwakura Y (1997) Growth retardation and early death of beta-1,4-galactosyltransferase knockout mice with augmented proliferation and abnormal differentiation of epithelial cells. EMBO J 16: 1850–1857 Au CE, Bell AW, Gilchrist A, Hiding J, Nilsson T, Bergeron JJ (2007) Organellar proteomics to create the cell map. Curr Opin Cell Biol 19: 376–385 Babad H, Hassid WZ (1966) Soluble uridine diphosphate D-galactose: D-glucose beta-4-Dgalactosyltransferase from bovine milk. J Biol Chem 241: 2672–2678 Baker JR (1953) The expressions “Golgi Apparatus”, “Golgi body” and “Golgi substance”. Nature 172: 617–618 Bao M, Elmendorf BJ, Booth JL, Drake RR, Canfield WM (1996) Bovine UDP-N-acetylglucosamine: lysosomal-enzyme N-acetylglucosamine-1-phosphotransferase. II. Enzymatic characterization and identification of the catalytic subunit. J Biol Chem 271: 31446–31451 Barratt J, Smith AC, Feehally J (2007) The pathogenic role of IgA1 O-linked glycosylation in the pathogenesis of IgA nephropathy. Nephrology (Carlton) 12: 275–284 Bensley RR (1951) Facts versus artefacts in cytology: the Golgi apparatus. Exp Cell Res 2: 1–9 Berger EG (1999) Tn-syndrome. Biochim Biophys Acta 1455: 255–268 Berger EG (2002) Ectopic localizations of Golgi glycosyltransferases. Glycobiology 12: 29R–36R Berger EG, Mandel T, Schilt U (1981) Immunohistochemical localization of galactosyltransferase in human fibroblasts and HeLa cells. J Histochem Cytochem 29: 364–370 Berger EG, Rohrer J (2003) Galactosyltransferase—still up and running. Biochimie 85: 261–274 €user, Basel Berger EG, Roth J (1997) The Golgi Apparatus. Birkha Beyer TA, Sadler JE, Rearick JI, Paulson JC, Hill RL (1981) Glycosyltransferases and their use in assessing oligosaccharide structure and structure–function relationships. Adv Enzymol Relat Areas Mol Biol 52: 23–175 Bonfanti L (2006) PSA-NCAM in mammalian structural plasticity and neurogenesis. Prog Neurobiol 80: 129–164 Borsig L (2004) Selectins facilitate carcinoma metastasis and heparin can prevent them. News Physiol Sci 19: 16–21 € chli M, Berger EG (1999) Alpha1,3Fucosyltransferase VI is Borsig L, Imbach T, Ho expressed in HepG2 cells and codistributed with beta1,4galactosyltransferase I in the Golgi apparatus and monensin-induced swollen vesicles. Glycobiology 9: 1273–1280 Breton C, Snajdrova L, Jeanneau C, Koca J, Imberty A (2006) Structures and mechanisms of glycosyltransferases. Glycobiology 16: 29R–37R Bretz R, Bretz H, Palade GE (1980) Distribution of terminal glycosyltransferases in hepatic Golgi fractions. J Cell Biol 84: 87–101 Brockhausen I (1995) Biosynthesis of O-Glycans of the N-acetylgalactosamine-alpha-Ser/ Thr linkage type. In: Montreuil J, Vliegenthart JFG, Schachter H (eds) Elsevier, Amsterdam, pp 201–250

Golgi glycosylation enzymes

*

183

Buckhaults P, Chen L, Fregien N, Pierce M (1997) Transcriptional regulation of N-acetylglucosaminyltransferase V by the src oncogene. J Biol Chem272: 19575–19581 Burke J, Pettitt JM, Humphris D, Gleeson PA (1994) Medial-Golgi retention of N-acetylglucosaminyltransferase I. Contribution from all domains of the enzyme. J Biol Chem 269: 12049–12059 Burke J, Pettitt JM, Schachter H, Sarkar M, Gleeson PA (1992) The transmembrane and flanking sequences of beta 1,2-N-acetylglucosaminyltransferase I specify medialGolgi localization. J Biol Chem 267: 24433–24440 Caffaro CE, Hirschberg CB (2006) Nucleotide sugar transporters of the Golgi apparatus: from basic science to diseases. Acc Chem Res 39: 805–812 Campbell RM, Metzler M, Granovsky M, Dennis JW, Marth JD (1995) Complex asparagine-linked oligosaccharides in Mgat1-null embryos. Glycobiology 5: 535–543 Caro LG, Palade GE (1964) Protein synthesis, storage, and discharge in the pancreatic exocrine cell. An autoradiographic study. J Cell Biol 20: 473–495 Chapman E, Best MD, Hanson SR, Wong CH (2004) Sulfotransferases: structure, mechanism, biological activity, inhibition, and synthetic utility. Angew Chem Int Ed 43: 3526–3548 Chatterton JE, Hirsch D, Schwartz JJ, Bickel PE, Rosenberg RD, Lodish HF, Krieger M (1999) Expression cloning of LDLB, a gene essential for normal Golgi function and assembly of the 1d1Cp complex. Proc Natl Acad Sci USA 96: 915–920 Chen C, Ma J, Lazic A, Backovic M, Colley KJ (2000) Formation of insoluble oligomers correlates with ST6Gal I stable localization in the Golgi. J Biol Chem 275: 13819–13826 Cheung P, Pawling J, Partridge EA, Sukhu B, Grynpas M, Dennis JW (2007) Metabolic homeostasis and tissue renewal are dependent on beta1,6GlcNAc-branched N-glycans. Glycobiology 17: 828–837 Chou DKH, Schachner M, Jungalwala FB (2002) HNK-1 sulfotransferase null mice express glucuronyl glycoconjugates and show normal cerebellar granule neuron migration in vivo and in vitro. J Neurochem 82: 1239–1251 Colley KJ (1997) Golgi localization of glycosyltransferases: more questions than answers. Glycobiology 7: 1–13 Colley KJ, Lee EU, Adler B, Browne JK, Paulson JC (1989) Conversion of a Golgi apparatus sialyltransferase to a secretory protein by replacement of the NH2-terminal signal anchor with a signal peptide. J Biol Chem 264: 17619–17622 Coste H, Martel MB, Got R (1986) Topology of glucosylceramide synthesis in Golgi membranes from porcine submaxillary glands. Biochim Biophys Acta 858: 6–12 Coutinho PM, Henrissat B (1999) Life with no sugars? J Mol Microbiol Biotechnol 1: 307–308 D’Angelo G, Polishchuk E, Di Tullio G, Santoro M, Di Campli A, Godi A, West G, Bielawski J, Chuang CC, Van der Spoel AC, Platt FM, Hannun YA, Polishchuk R, Mattjus P, De Matteis MA (2007) Glycosphingolipid synthesis requires FAPP2 transfer of glucosylceramide. Nature 449: 62–67 Dahdal RY, Colley KJ (1993) Specific sequences in the signal anchor of the betagalactoside alpha-2,6-sialyltransferase are not essential for Golgi localization. Membrane flanking sequences may specify Golgi retention. J Biol Chem 268: 26310–26319 Dalton AJ, Felix MD (1954) Cytologic and cytochemical characteristics of the Golgi substance of epithelial cells of the epididymis in situ, in homogenates and after isolation. Am J Anat 94: 171–207 Datta AK, Paulson JC (1995) The sialyltransferase “sialylmotif” participates in binding the donor substrate CMP-NeuAc. J Biol Chem 270: 1497–1500 Datta AK, Sinha A, Paulson JC (1998) Mutation of the sialyltransferase S-sialylmotif alters the kinetics of the donor and acceptor substrates. J Biol Chem 273: 9608–9614

184

*

E. G. Berger and J. Rohrer

Den H, Kaufman B, Roseman S (1970) Properties of some glycosyltransferases in embryonic chicken brain. J Biol Chem 245: 6607–6615 Dennis JW (1988) Asn-linked oligosaccharide processing and malignant potential. Cancer Surv 7: 573–595 Dinter A, Berger EG (1998) Golgi-disturbing agents. Histochem Cell Biol 109: 571–590 El-Battari A, Prorok M, Angata K, Mathieu S, Zerfaoui M, Ong E, Suzuki M, Lombardo D, Fukuda M (2003) Different glycosyltransferases are differentially processed for secretion, dimerization, and autoglycosylation. Glycobiology 13: 941–953 Esko JD, Selleck SB (2002) Order out of chaos: assembly of ligand binding sites in heparan sulfate. Annu Rev Biochem 71: 435–471 Evans SC, Lopez LC, Shur BD (1993) Dominant negative mutation in cell surface beta 1,4galactosyltransferase inhibits cell–cell and cell–matrix interactions. J Cell Biol 120: 1045–1057 Farquhar MG, Palade GE (1981) The Golgi apparatus (complex)-(1954–1981)-from artifact to center stage. J Cell Biol 91: 77s–103s Fenteany FH, Colley KJ (2005) Multiple signals are required for alpha2,6-sialyltransferase (ST6Gal I) oligomerization and Golgi localization. J Biol Chem 280: 5423–5429 Freeze HH (2006) Genetic defects in the human glycome. Nat Rev Genet 7: 537–551 Gerber AC, Kozdrowski I, Wyss SR, Berger EG (1979) The charge heterogeneity of soluble human galactosyltransferases isolated from milk, amniotic fluid and malignant ascites. Eur J Biochem 93: 453–460 Gu J, Taniguchi N (2004) Regulation of integrin functions by N-glycans. Glycoconj J 21: 9–15 Hagopian A, Bosmann HB, Eylar EH (1968) Glycoprotein biosynthesis: the localization of polypeptidyl: N-acetylgalactosaminyl, collagen: glucosyl, and glycoprotein: galactosyl transferases in HeLa cell membrane fractions. Arch Biochem Biophys 128: 387–396 Halter D, Neumann S, Van Dijk SM, Wolthoorn J, De Maziere AM, Vieira OV, Mattjus P, Klumperman J, Van Meer G, Sprong H (2007) Pre- and post-Golgi translocation of glucosylceramide in glycosphingolipid synthesis. J Cell Biol 179: 101–115 Hancock SM, Vaughan MD, Withers SG (2006) Engineering of glycosidases and glycosyltransferases. Curr Opin Chem Biol 10: 509–519 Hansen W, Grabenhorst E, Nimtz M, Muller K, Conradt HS, Wirth M (2005) Generation of serum-stabilized retroviruses: reduction of alpha1,3gal-epitope synthesis in a murine NIH3T3-derived packaging cell line by expression of chimeric glycosyltransferases. Metab Eng 7: 221–228 Harduin-Lepers A, Mollicone R, Delannoy P, Oriol R (2005) The animal sialyltransferases and sialyltransferase-related genes: a phylogenetic approach. Glycobiology 15: 805–817 Hart GW, Housley MP, Slawson C (2007) Cycling of O-linked beta-N-acetylglucosamine on nucleocytoplasmic proteins. Nature 446: 1017–1022 Hennet T (2002) The galactosyltransferase family. Cell Mol Life Sci 59: 1081–1095 Herscovics A (1999) Importance of glycosidases in mammalian glycoprotein biosynthesis. Biochim Biophys Acta 1473: 96–107 Herscovics A (2001) Structure and function of Class I alpha 1,2-mannosidases involved in glycoprotein synthesis and endoplasmic reticulum quality control. Biochimie 83: 757–762 Higy M, Junne T, Spiess M (2004) Topogenesis of membrane proteins at the endoplasmic reticulum. Biochemistry 43: 12716–12722 Hill RL, Brew K (1975) Lactose synthetase. Adv Enzymol Relat Areas Mol Biol 43: 411–490 Ichikawa S, Hirabayashi Y (1998) Glucosylceramide synthase and glycosphingolipid synthesis. Trends Cell Biol 8: 198–202

Golgi glycosylation enzymes

*

185

Ihara H, Ikeda Y, Toma S, Wang X, Suzuki T, Gu J, Miyoshi E, Tsukihara T, Honke K, Matsumoto A, Nakagawa A, Taniguchi N (2007) Crystal structure of mammalian alpha1,6-fucosyltransferase, FUT8. Glycobiology 17: 455–466 Ikeda Y, Taniguchi N (2001) Enzymatic properties and biological functions of beta 1,4-Nacetylglucosaminyltransferase III. Trends Glycosci Glycotech 13: 167–176 Inamori K, Mita S, Gu J, Mizuno-Horikawa Y, Miyoshi E, Dennis JW, Taniguchi N (2006) Demonstration of the expression and the enzymatic activity of N-acetylglucosaminyltransferase IX in the mouse brain. Biochim Biophys Acta 1760: 678–684 Ju T, Cummings RD (2002) A unique molecular chaperone Cosmc required for activity of the mammalian core 1 beta 3-galactosyltransferase. Proc Natl Acad Sci USA 99: 16613–16618 Ju T, Cummings RD (2005) Protein glycosylation: chaperone mutation in Tn syndrome. Nature 437: 1252 Kakuda S, Shiba T, Ishiguro M, Tagawa H, Oka S, Kajihara Y, Kawasaki T, Wakatsuki S, Kato R (2004) Structural basis for acceptor substrate recognition of a human glucuronyltransferase, GlcAT-P, an enzyme critical in the biosynthesis of the carbohydrate epitope HNK-1. J Biol Chem 279: 22693–22703 Kaneko M, Alvarez-Manilla G, Kamar M, Lee I, Lee JK, Troupe K, Zhang W, Osawa M, Pierce M (2003) A novel beta(1,6)-N-acetylglucosaminyltransferase V (GnT-VB)(1). FEBS Lett 554: 515–519 Kawatkar SP, Kuntz DA, Woods RJ, Rose DR, Boons GJ (2006) Structural basis of the inhibition of Golgi alpha-mannosidase II by mannostatin A and the role of the thiomethyl moiety in ligand-protein interactions. J Am Chem Soc 128: 8310–8319 Kingsley DM, Kozarsky KF, Hobbie L, Krieger M (1986) Reversible defects in O-linked glycosylation and LDL receptor expression in a UDP-Gal/UDP-GalNAc 4-epimerase deficient mutant. Cell 44: 749–759 Kornfeld R, Kornfeld S (1985) Assembly of asparagine-linked oligosaccharides. Annu Rev Biochem 54: 631–664 Kudo M, Bao M, D’Souza A, Ying F, Pan H, Roe BA, Canfield WM (2005) The alpha- and beta-subunits of the human UDP-N-acetylglucosamine: lysosomal enzyme N-acetylglucosamine-1-phosphotransferase [corrected] are encoded by a single cDNA. J Biol Chem 280: 36141–36149 Kweon H-S, Beznoussenko GV, Micaroni M, Polishchuk RS, Trucco A, Martella O, Di Giandomenico D, Marra P, Fusella A, Di Pentima A, Berger EG, Geerts WJ C, Koster AJ, Burger KN J, Luini A, Mironov AA (2004) Golgi enzymes are enriched in perforated zones of Golgi cisternae but are depleted in COPI vesicles. Mol Biol Cell 15: 4710–4724 Lee EU, Roth J, Paulson JC (1989) Alteration of terminal glycosylation sequences on Nlinked oligosaccharides of Chinese hamster ovary cells by expression of betagalactoside alpha 2,6-sialyltransferase. J Biol Chem 264: 13848–13855 Lorenz H, Hailey DW, Lippincott-Schwartz J (2006) Fluorescence protease protection of GFP chimeras to reveal protein topology and subcellular localization. Nat Methods 3: 205–210 Louvard D, Reggio H, Warren G (1982) Antibodies to the Golgi complex and the rough endoplasmic reticulum. J Cell Biol 92: 92–107 Lowe JB (1995) Molecular Basis of Human Blood Group Antigens. Plenum Press, New York Lu Q, Hasty P, Shur BD (1997) Targeted mutation in beta1,4-galactosyltransferase leads to pituitary insufficiency and neonatal lethality. Dev Biol 181: 257–267 Ma B, Simala-Grant JL, Taylor DE (2006) Fucosylation in prokaryotes and eukaryotes. Glycobiology 16: 158R–184R Ma J, Qian R, Rausa FM, III, Colley KJ (1997) Two naturally occurring alpha2,6-sialyltransferase forms with a single AA change in the catalytic domain differ in their catalytic activity and proteolytic processing. J Biol Chem 272: 672–679 Marcus SL, Polakowski R, Seto NO, Leinala E, Borisova S, Blancher A, Roubinet F, Evans SV, Palcic MM (2003) A single point mutation reverses the donor specificity of

186

*

E. G. Berger and J. Rohrer

human blood group B-synthesizing galactosyltransferase. J Biol Chem 278: 12403–12405 Mast SW, Moremen KW (2006) Family 47 alpha-mannosidases in N-glycan processing. Methods Enzymol 415: 31–46 Milland J, Russell SM, Dodson HC, McKenzie IF, Sandrin MS (2002) The cytoplasmic tail of alpha 1,3-galactosyltransferase inhibits Golgi localization of the full-length enzyme. J Biol Chem 277: 10374–10378 Milland J, Taylor SG, Dodson HC, McKenzie IF, Sandrin MS (2001) The cytoplasmic tail of alpha 1,2-fucosyltransferase contains a sequence for Golgi localization. J Biol Chem 276: 12012–12018 Miyoshi E, Noda K, Yamaguchi Y, Inoue S, Ikeda Y, Wang W, Ko JH, Uozumi N, Li W, Taniguchi N (1999) The alpha1-6-fucosyltransferase gene and its biological significance. Biochim Biophys Acta 1473: 9–20 Moremen KW (2002) Golgi alpha-mannosidase II deficiency in vertebrate systems: implications for asparagine-linked oligosaccharide processing in mammals. Biochim Biophys Acta 1573: 225–235 Munro S (1991) Sequences within and adjacent to the transmembrane segment of alpha-2,6-sialyltransferase specify Golgi retention. EMBO J 10: 3577–3588 Munro S (1995) An investigation of the role of transmembrane domains in Golgi protein retention. EMBO J 14: 4695–4704 Nagai K, Ihara Y, Wada Y, Taniguchi N (1997) N-glycosylation is requisite for the enzyme activity and Golgi retention of N-acetylglucosaminyltransferase III. Glycobiology 7: 769–776 Narimatsu H, Sinha S, Brew K, Okayama H, Qasba PK (1986) Cloning and sequencing of cDNA of bovine N-acetylglucosamine (beta 1-4)galactosyltransferase. Proc Natl Acad Sci USA 83: 4720–4740 Negishi M, Pedersen LG, Petrotchenko E, Shevtsov S, Gorokhov A, Kakuta Y, Pedersen LC (2001) Structure and function of sulfotransferases. Arch Biochem Biophys 390: 149–157 Neutra M, Leblond CP (1966) Radioautographic comparison of the uptake of galactoseH and glucose-H3 in the Golgi region of various cells secreting glycoproteins or mucopolysaccharides. J Cell Biol 30: 137–150 Nilsson T, Lucocq JM, Mackay D, Warren G (1991) The membrane spanning domain of beta-1,4-galactosyltransferase specifies trans Golgi localization. EMBO J 10: 3567–3575 Ohtsubo K, Marth JD (2006) Glycosylation in cellular mechanisms of health and disease. Cell 126: 855–867 Patel RY, Balaji PV (2007) Length and composition analysis of the cytoplasmic, transmembrane and stem regions of human Golgi glycosyltransferases. Protein Pept Lett 14: 601–609 Paulson J, Colley K (1989) Glycosyltransferases. Structure, localization, and control of cell type-specific glycosylation. J Biol Chem 264: 17615–17618 Pfeffer SR, Rothman JE (1987) Biosynthetic protein transport and sorting by the endoplasmic reticulum and Golgi. Annu Rev Biochem 56: 829–852 Piller F, Piller V, Fox RI, Fukuda M (1988) Human T-lymphocyte activation is associated with changes in O-glycan biosynthesis. J Biol Chem 263: 15146–15150 Qasba PK, Ramakrishnan B, Boeggeman E (2005) Substrate-induced conformational changes in glycosyltransferases. Trends Biochem Sci 30: 53–62 Qian R, Chen C, Colley KJ (2001) Location and mechanism of alpha 2,6-sialyltransferase dimer formation. Role of cysteine residues in enzyme dimerization, localization, activity, and processing. J Biol Chem 276: 28641–28649 Rabouille C, Hui N, Hunte F, Kieckbusch R, Berger EG, Warren G, Nilsson T (1995) Mapping the distribution of Golgi enzymes involved in the construction of complex oligosaccharides. J Cell Sci 108: 1617–1627

Golgi glycosylation enzymes

*

187

Rhee SW, Starr T, Forsten-Williams K, Storrie B (2005) The steady-state distribution of glycosyltransferases between the Golgi Apparatus and the endoplasmic reticulum is approximately 90:10. Traffic 6: 978–990 Rilla K, Siiskonen H, Spicer AP, Hyttinen JM, Tammi MI, Tammi RH (2005) Plasma membrane residence of hyaluronan synthase is coupled to its enzymatic activity. J Biol Chem 280: 31890–31897 Roseman S (1970) The synthesis of complex carbohydrates by multiglycosyltransferase systems and their potential function in intercellular adhesion. Chem Phys Lipids 5: 270–297 Rosen SD (2004) Ligands for L-selectin: homing, inflammation, and beyond. Ann Rev Immunol 22: 129–156 Roth J (1998) Topology of glycosylation in the Golgi Apparatus. In: Berger EG, Roth J €user, Basel Boston Berlin, pp 131–161 (ed.) The Golgi Apparatus. Birkha Roth J, Berger EG (1982) Immunocytochemical localization of galactosyltransferase in HeLa cells: codistribution with thiamine pyrophosphatase in trans-Golgi cisternae. J Cell Biol 93: 223–229 Roth J, Taatjes DJ (1998) Tubules of the trans Golgi apparatus visualized by immunoelectron microscopy. Histochem Cell Biol 109: 545–553 Roth J, Taatjes DJ, Lucocq JM, Weinstein J, Paulson JC (1985) Demonstration of an extensive trans-tubular network continuous with the Golgi apparatus stack that may function in glycosylation. Cell 43: 287–295 Rottger S, White J, Wandall HH, Olivo JC, Stark A, Bennett EP, Whitehouse C, Berger EG, Clausen H, Nilsson T (1998) Localization of three human polypeptide GalNActransferases in HeLa cells suggests initiation of O-linked glycosylation throughout the Golgi apparatus. J Cell Sci 111: 45–60 Russo RN, Shaper NL, Taatjes DJ, Shaper JH (1992) Beta 1,4-galactosyltransferase: a short NH2-terminal fragment that includes the cytoplasmic and transmembrane domain is sufficient for Golgi retention. J Biol Chem 267: 9241–9247 Saito H, Gu J, Nishikawa A, Ihara Y, Fujii J, Kohgo Y, Taniguchi N (1995) Organization of the human N-acetylglucosaminyltransferase V gene. Eur J Biochem 233: 18–26 Sasai K, Ikeda Y, Tsuda T, Ihara H, Korekane H, Shiota K, Taniguchi N (2001) The critical role of the stem region as a functional domain responsible for the oligomerization and Golgi localization of N-acetylglucosaminyltransferase V. The involvement of a domain homophilic interaction. J Biol Chem 276: 759–765 Schachner M, Martini R, Hall H, Orberger G (1995) Functions of the L2/HNK-1 carbohydrate in the nervous system. Prog Brain Res 105: 183–188 Schachter H (1986) Biosynthetic controls that determine the branching and microheterogeneity of protein-bound oligosaccharides. Biochem Cell Biol 64: 163–181 Schachter H (2004) Protein glycosylation lessons from Caenorhabditis elegans. Curr Opin Struct Biol 14: 607–616 Schaub BE, Berger B, Berger EG, Rohrer J (2006) Transition of galactosyltransferase 1 from trans-Golgi cisterna to the trans-Golgi network is signal mediated. Mol Biol Cell 17(12): 5153–5162 Sewell R, Backstrom M, Dalziel M, Gschmeissner S, Karlsson H, Noll T, Gatgens J, Clausen H, Hansson GC, Burchell J, Taylor-Papadimitriou J (2006) The ST6GalNAc-I sialyltransferase localizes throughout the Golgi and is responsible for the synthesis of the tumor-associated sialyl-Tn O-glycan in human breast cancer. J Biol Chem 281: 3586–3594 Shaper NL, Shaper JH, Meuth JL, Fox JL, Chang H, Kirsch IR, Hollis GF (1986) Bovine galactosyltransferase: identification of a clone by direct immunological screening of a cDNA expression library. Proc Natl Acad Sci USA 83: 1573–1577 Shur BD, Evans S, Lu Q (1998) Cell surface galactosyltransferase: current issues. Glycoconj J 15: 537–548

188

*

E. G. Berger and J. Rohrer

Sousa VL, Brito C, Costa J (2004) Deletion of the cytoplasmic domain of human alpha3/4 fucosyltransferase III causes the shift of the enzyme to early Golgi compartments. Biochim Biophys Acta 1675: 95–104 Sousa VL, Brito C, Costa T, Lanoix J, Nilsson T, Costa J (2003) Importance of Cys, Gln, and Tyr from the transmembrane domain of human alpha 3/4 fucosyltransferase III for its localization and sorting in the Golgi of baby hamster kidney cells. J Biol Chem 278: 7624–7629 Spiro RG (2004) Role of N-linked polymannose oligosaccharides in targeting glycoproteins for endoplasmic reticulum-associated degradation. Cell Mol Life Sci 61: 1025–1041 Strous GJ, Berger EG (1982) Biosynthesis, intracellular transport, and release of the Golgi enzyme galactosyltransferase (lactose synthetase A protein) in HeLa cells. J Biol Chem 257: 7623–7628 Taatjes DJ, Roth J, Weinstein J, Paulson JC, Shaper NL, Shaper JH (1987) Codistribution of galactosyl- and sialyltransferase: reorganization of trans Golgi apparatus elements in hepatocytes in intact liver and cell culture. Eur J Cell Biol 44: 187–194 Takamatsu S, Oguri S, Minowa MT, Yoshida A, Nakamura K, Takeuchi M, Kobata A (1999) Unusually high expression of N-acetylglucosaminyltransferase-IVa in human choriocarcinoma cell lines: a possible enzymatic basis of the formation of abnormal biantennary sugar chain. Cancer Res 59: 3949–3953 Tang BL, Wong SH, Low SH, Hong W (1992) The transmembrane domain of N-glucosaminyltransferase I contains a Golgi retention signal. J Biol Chem 267: 10122–10126 Tassin AM, Paintrand M, Berger EG, Bornens M (1985) The Golgi apparatus remains associated with microtubule organizing centers during myogenesis. J Cell Biol 101: 630–638 Teasdale RD, D’Agostaro G, Gleeson PA (1992) The signal for Golgi retention of bovine beta 1,4-galactosyltransferase is in the transmembrane domain. J Biol Chem 267: 13113 Ten Hagen KG, Fritz TA, Tabak LA (2003) All in the family: the UDP-GalNAc : polypeptide N-acetylgalactosaminyltransferases. Glycobiology 13: 1R–16R Thiele H, Sakano M, Kitagawa H, Sugahara K, Rajab A, Hohne W, Ritter H, Leschik G, Nurnberg P, Mundlos S (2004) Loss of chondroitin 6-O-sulfotransferase-1 function results in severe human chondrodysplasia with progressive spinal involvement. Proc Natl Acad Sci USA 101: 10155–10160 Tiede S, Storch S, Lubke T, Henrissat B, Bargal R, Raas-Rothschild A, Braulke T (2005) Mucolipidosis II is caused by mutations in GNPTA encoding the alpha/beta GlcNAc-1phosphotransferase. Nat Med 11: 1109–1112 Uliana AS, Giraudo CG, Maccioni HJ (2006) Cytoplasmic tails of SialT2 and GalNAcT impose their respective proximal and distal Golgi localization. Traffic 7: 604–612 Ungar D, Oka T, Brittle EE, Vasile E, Lupashin VV, Chatterton JE, Heuser JE, Krieger M, Waters MG (2002) Characterization of a mammalian Golgi-localized protein complex, COG, that is required for normal Golgi morphology and function. J Cell Biol 157: 405–415 Varki A (1998) Factors controlling the glycosylation potential of the Golgi apparatus. Trends Cell Biol 8: 34–40 Varki A (2006) Nothing in glycobiology makes sense, except in the light of evolution. Cell 126: 841–845 Wandall HH, Irazoqui F, Tarp MA, Bennett EP, Mandel U, Takeuchi H, Kato K, Irimura T, Suryanarayanan G, Hollingsworth MA, Clausen H (2007) The lectin domains of polypeptide GalNAc-transferases exhibit carbohydrate-binding specificity for GalNAc: lectin binding to GalNAc-glycopeptide substrates is required for high density GalNAc-O-glycosylation. Glycobiology 17: 374–387

Golgi glycosylation enzymes

*

189

Warren L, Buck CA, Tuszynski GP (1978) Glycopeptide changes and malignant transformation. A possible role for carbohydrate in malignant behavior. Biochim Biophys Acta 516: 97–127 Weinstein J, Lee EU, McEntee K, Lai PH, Paulson JC (1987) Primary structure of betagalactoside alpha 2,6-sialyltransferase. Conversion of membrane-bound enzyme to soluble forms by cleavage of the NH2-terminal signal anchor. J Biol Chem 262: 17735–17743 Wu X, Steet RA, Bohorov O, Bakker J, Newell J, Krieger M, Spaapen L, Kornfeld S, Freeze HH (2004) Mutation of the COG complex subunit gene COG7 causes a lethal congenital disorder. Nat Med 10: 518–523 Xia L, Ju T, Westmuckett A, An G, Ivanciu L, McDaniel JM, Lupu F, Cummings RD, McEver RP (2004) Defective angiogenesis and fatal embryonic hemorrhage in mice lacking core 1-derived O-glycans. J Cell Biol 164: 451–459 Yamamoto F, Clausen H, White T, Marken J, Hakomori S (1990) Molecular genetic basis of the histo-blood group ABO system. Nature 345: 229–233 Yamamoto S, Oka S, Inoue M, Shimuta M, Manabe T, Takahashi M, Miyamoto M, Asano M, Sakagami J, Sudo K, Iwakura Y, Ono K, Kawasaki T (2002) Mice deficient in nervous system-specific carbohydrate epitope HNK-1 exhibit impaired synaptic plasticity and spatial learning. J Biol Chem 277: 27227–27231 Yoshida A, Minowa MT, Takamatsu S, Hara T, Ikenaga H, Takeuchi M (1998) A novel second isoenzyme of the human UDP-N-acetylglucosamine:alpha1,3-D-mannoside beta1,4-N-acetylglucosaminyltransferase family: cDNA cloning, expression, and chromosomal assignment. Glycoconj J 15: 1115–1123 Zaal KJ, Smith CL, Polishchuk RS, Altan N, Cole NB, Ellenberg J, Hirschberg K, Presley JF, Roberts TH, Siggia E, Phair RD, Lippincott-Schwartz J (1999) Golgi membranes are absorbed into and reemerge from the ER during mitosis. Cell 99: 589–601 Zerfaoui M, Fukuda M, Langlet C, Mathieu S, Suzuki M, Lombardo D, El-Battari A (2002) The cytosolic and transmembrane domains of the beta 1,6 N-acetylglucosaminyltransferase (C2GnT) function as a cis to medial/Golgi-targeting determinant. Glycobiology 12: 15–24 Zhang W, Betel D, Schachter H (2002) Cloning and expression of a novel UDP-GlcNAc: alpha-d-mannoside beta1,2-N-acetylglucosaminyltransferase homologous to UDPGlcNAc:alpha-3-d-mannoside beta1,2-N-acetylglucosaminyltransferase I. Biochem J 361: 153–162 Zhang W, Revers L, Pierce M, Schachter H (2000) Regulation of expression of the human beta-1,2-N-acetylglucosaminyltransferase II gene (MGAT2) by Ets transcription factors. Biochem J 347: 511–518 Zhao Y, Nakagawa T, Itoh S, Inamori K, Isaji T, Kariya Y, Kondo A, Miyoshi E, Miyazaki K, Kawasaki N, Taniguchi N, Gu J (2006) N-acetylglucosaminyltransferase III antagonizes the effect of N-acetylglucosaminyltransferase V on alpha3beta1 integrinmediated cell migration. J Biol Chem 281: 32122–32130 Zhu G, Jaskiewicz E, Bassi R, Darling DS, Young WW, Jr (1997) Beta 1,4 N-acetylgalactosaminyltransferase (GM2/GD2/GA2 synthase) forms homodimers in the endoplasmic reticulum: a strategy to test for dimerization of Golgi membrane proteins. Glycobiology 7: 987–996 Zolov SN, Lupashin VV (2005) Cog3p depletion blocks vesicle-mediated Golgi retrograde trafficking in HeLa cells. J Cell Biol 168: 747–759

190

*

W. Zhao and K. J. Colley

Nucleotide sugar transporters of the Golgi apparatus Weihan Zhao and Karen J. Colley

Introduction The Golgi apparatus is the major site of protein, lipid and proteoglycan glycosylation. The glycosylation enzymes, as well as kinases and sulfatases that catalyze phosphorylation and sulfation, are localized within the Golgi cisternae in characteristic distributions that frequently reflect their order in a particular pathway (Kornfeld and Kornfeld 1985; Colley 1997). The glycosyltransferases, sulfotransferases and kinases are “transferases” that require activated donor molecules for the reactions they catalyze. For eukaryotic, fungal and protozoan glycosyltransferases these are the nucleotide sugars UDP-N-acetylglucosamine (UDP-GlcNAc), UDP-galactose (UDP-Gal), GDPfucose (GDP-Fuc), CMP-sialic acid (CMP-Sia), UDP-glucuronic acid (UDP-GlcA), GDP-mannose (GDP-Man), and UDP-xylose (UDP-Xyl) (Hirschberg et al. 1998). For the kinases, ATP functions as the donor, while for the sulfotransferases, adenosine 30 -phosphate 50 -phosphate (PAPS) acts as the donor (Hirschberg et al. 1998). The active sites of all these enzymes are oriented towards the lumen of the Golgi cisternae. This necessitates the translocation of their donors from the cytosol into the lumenal Golgi compartments. In this chapter we will focus on the structure, function and localization of the Golgi nucleotide sugar transporters (NSTs), and highlight the diseases and developmental defects associated with defective transporters. We direct the reader to several excellent reviews on Golgi transporters for additional details and references (Hirschberg et al. 1998; Berninsone and Hirschberg 2000; Gerardy-Schahn et al. 2001; Handford et al. 2006; Caffaro and Hirschberg 2006).

The identification of NSTs and the diseases and defects in development caused by mutant transporters Abundant evidence now exists for the importance of glycoconjugates in both development and in fundamental processes in adult organisms (Varki 1993; Haltiwanger and Lowe 2004). The critical role of NSTs and the maintenance of nucleotide sugar levels in the glycosylation of proteins, lipids and proteoglycans has been highlighted by a number of transporter mutants that lead to developmental defects in model organisms such as C. elegans and Drosophila, to decreased virulence of parasites such as Leishmania, and to severe diseases in humans and cattle.

Nucleotide sugar transporters

*

191

In early studies, mutants exhibiting altered glycosylation in both mammalian cell lines and other organisms, such as yeast and the protozoan parasite Leishmania, were isolated and found defective in nucleotide sugar transport by biochemical analysis (Ballou et al. 1991; Descoteaux et al. 1995; Herman and Horvitz 1999; Patnaik and Stanley 2006). In these analyses, investigators quantified the transport of radiolabeled sugar nucleotides into sealed vesicles under different conditions using filtration or centrifugation to separate the vesicles from the assay medium (reviewed in Hirschberg et al. 1998). For example, Deutscher et al. (1984) demonstrated that Lec2 CHO cells possessed only 2% of the CMP-Sia transport activity of wild type CHO cells, while others demonstrated similar decreases in UDP-GlcNAc transport activity in the Kluyveromyces lactis mnn2-2 mutant (Abeijon et al. 1996a), in UDP-Gal transport activity in the MDCKII-RCAr mutant (Brandli et al. 1988), and in GDP-Man transport activity in both the L. donovani C3PO mutant and in the Saccharomyces cerevisiae vrg4 mutant (Ma et al. 1997; Dean et al. 1997). These studies demonstrated that nucleotide sugar transport was absolutely required for glycosylation in mammalian, yeast and protozoan cells and provided investigators with a way to clone the defective transporters by complementation. The cloning of NST coding sequences opened the way to further characterization of the structure and function of the transporters by expression in heterologous systems and reconstitution into proteoliposomes. This also allowed investigators to identify inactivating NST mutations leading to developmental defects and human disease.

Defects in the GDP-fucose and CMP-Sia transporters lead to two congenital disorders of glycosylation Leukocyte adhesion deficiency syndrome type II (LAD II), also called congenital disorder of glycosylation (CDG) IIc, is a rare autosomal recessive human syndrome characterized by a general reduction of fucose in glycoconjugates due to a deficiency in GDP-Fuc transport (reviewed in Hirschberg 2001; Becker and Lowe 1999). Patients have an abnormal facial appearance and exhibit severe psychomotor and growth retardation, recurrent infections, and periodontitis. Using patient fibroblasts and screening for recovery of glycocon€ bke et al. (2001) and Lu € hn et al. jugate fucosylation in transformants, both Lu (2001) cloned the human and C. elegans GDP-fucose transporters, respectively. These investigators and Helmus et al. (2006) have identified several specific mutations in LADII/CDG IIc patients that lead to disease. The murine CMP-Sia transporter was cloned by complementation of sialylation deficient Lec2 CHO cells and its activity verified by heterologous expression in S. cerevisiae (Eckhardt et al. 1996; Berninsone et al. 1997). Recently, inactivating mutations in this transporter were found to result in a new CDG type II that was diagnosed in a 4-month-old boy (Willig et al. 2001; Martinez-Duncker et al. 2005). Decreased sialylation in the patient led to macrothrombocytopenia, neutropenia, and complete lack of the sialyl Lewis X antigen on polymorphonuclear cells. The patient experienced progressive

192

*

W. Zhao and K. J. Colley

hemorrhaging, respiratory distress syndrome, and opportunistic infections. Ultimately, complications including pulmonary viral infection, massive pulmonary hemorrhage, and respiratory failure led to death at the age of 37 months (Willig et al. 2001).

Drosophila Fringe connection and C. elegans SQV7, two multi-substrate NSTs required for signaling and cell interactions during development The Drosophila Fringe connection transporter was cloned by Selva et al. (2001) and Goto et al. (2001) who identified mutants in Fringe connection in screens for segment polarity and limb defects. These mutants had defects in Fringe-dependent Notch signaling, and in the Wingless/Wnt, Hedgehog, and fibroblast growth factor signaling pathways that require heparan sulfate expression. Fringe is a GlcNAc transferase that modifies O-linked fucose residues on the Notch receptor’s epidermal growth factor repeats, and this modification differentially modulates the binding of Notch to receptors and its signaling pathways (Haltiwanger and Lowe 2004). Accordingly, Fringe connection is a Golgi localized multi-substrate nucleotide sugar transporter that transports UDP-GlcNAc, UDP-GlcA, and UDP-Xyl (Selva et al. 2001). Other groups cloned putative human orthologs of Fringe connection (Suda et al. 2004; Ishida et al. 2005), and over expression of one of these proteins in mammalian cells increased surface levels of heparan sulfate, consistent with the activity of Drosophila Fringe connection (Suda et al. 2004). C. elegans sqv mutants exhibit a squashed vulval phenotype and a reduction in hermaphrodite fertility (Herman et al. 1999). All eight of these mutant genes encode proteins involved in different aspects of proteoglycan biosynthesis (Herman and Horvitz 1999; Hwang and Horvitz 2002). The sqv7 gene encodes an NST that transports UDP-GlcA, UDP-GlcNAc, and UDP-Gal in a competitive and non-cooperative fashion (Berninsone et al. 2001). Surprisingly, two other C. elegans NSTs, the SRF-3 and CO3H5.2 proteins, are redundant with SQV-7 and each other, and they exhibit a dramatically different noncompetitive and simultaneous mechanism. The implications of this will be discussed below.

A defective UDP-GlcNAc transporter leads to complex vertebral malformation in cattle The yeast and canine UDP-GlcNAc transporters were cloned by complementation of the transporter defect the K. lactis mnn2-2 mutant (Abeijon et al. 1996b; Guillen et al. 1998). Interestingly, the sequence similarity of these two functionally equivalent transporters from different species is very low (22%), but not uncommon among transporters with the same specificity from different species. This and the high sequence similarities observed between transporters with different specificities, highlights the importance of biochemically verifying the true substrates of recombinant NSTs (Caffaro and Hirschberg 2006). Recently, a recessively inherited disease in cattle, complex vertebral malformation, was found to be the result of a missense mutation in

Nucleotide sugar transporters

*

193

the bovine UDP-GlcNAc transporter (Thomsen et al. 2006). The disease, which results in severe malformation of the vertebrae, abortion of fetuses, and perinatal death, has been reported in cattle all over the world. In fact, it was reported that approximately 30% of the elite sires in Japan and Denmark are carriers for this disease (Thomsen et al. 2006).

The GDP-mannose transporter is required for the virulence of parasites and is an essential protein in yeast Protozoans and yeast differ from vertebrates in that they require the translocation of GDP-Man into the Golgi lumen for extensive mannosylation of their glycoconjugates. GDP-Man transporters were cloned by complementation of the S. cerevisiae vrg4 mutant and the L. donvani C3PO mutant (Descoteaux et al. 1995; Ma et al. 1997; Poster and Dean 1996). The importance of this transporter in both organisms is underscored by the fact that vrg4 is an essential gene in yeast, and by the requirement for GDP-Man transport and the biosynthesis of mannose-containing surface glycoconjugates for Leishmania virulence. The latter observation identifies the Leishmania LPG2 GDP-Man transporter as a possible drug target for the treatment of Leishmaniasis.

Nucleotide sugar transporter specificity and mechanism In early studies, biochemical assays employing topologically correct membrane vesicles were used to identify and characterize NST activities. These initial studies showed that transport in most cases is organelle specific, is temperature sensitive, saturable (Kms of 1–10 mM), concentrates nucleotide sugars 50- to 100-fold in Golgi vesicles/liposomes relative to their concentration in the assay medium, does not require ATP, is not altered by ionophores, and that NSTs are antiporters (Hirschberg et al. 1998). These initial observations were verified by expression of recombinant NSTs in heterologous systems and the reconstitution of purified, recombinant transporters into proteoliposomes. Importantly, these more recent studies demonstrated that single transporter proteins were sufficient for transport activity and revealed surprising multi-substrate specificities and unique mechanisms for some NSTs.

Antiporter mechanism The ability of NSTs to function as antiporters, where the nucleotide sugar is stoichiometrically exchanged for the corresponding nucleoside monophosphate (see Fig. 1), was supported by early studies which showed that preloading nucleoside monophosphates into Golgi membrane vesicles, or proteoliposomes containing transporters, stimulated the transport of their respective nucleotide sugars into the lumen of these vesicles (Hirschberg et al. 1998). More recently both the recombinant L. donovani LPG2 GDP-Man transporter and the murine CMP-Sia transporter were reconstituted into

194

*

W. Zhao and K. J. Colley

A

B

Figure 1. The nucleotide sugar transport/antiport cycle in the Golgi apparatus. (A) GDP-Man, which is synthesized in the cytoplasm, is transported into the Golgi lumen by the GDP-Man transporter. In the Golgi lumen, GDP-Man is a substrate for mannosyltransferases (triangle), which transfer Man to glycoconjugate substrates (rectangle). GDP, the other product of the transfer reaction, is converted by a lumenal nucleoside diphosphatase (oval) to GMP. The export of GMP to the cytosol is coupled to the import of GDP-Man. (B) This type of antiport mechanism occurs for the exchange of CMP-Sia and CMP using a distinct CMP-Sia transporter. The major difference is that no diphosphatase activity is needed because CMP is released following the sialyltransferase reaction.

Nucleotide sugar transporters

*

195

phosphatidylcholine liposomes and their specificity and mechanism reevaluated (Segawa et al. 2005; Tiralongo et al. 2006). In both cases investigators observed that preloading the liposomes with the corresponding nucleotide monophosphate yielded a 3-fold higher initial rate of transport relative to liposomes that were not preloaded, again supporting an antiporter mechanism. Other work by Tiralongo et al. (2006) using the reconstituted murine CMP-Sia transporter showed that the rate of CMP-Sia transport was also stimulated under “equilibrium exchange conditions” and suggested that this transporter is a simple mobile carrier with a binding site that alternates between both sides of the membrane. Most biochemical evidence suggests that the nucleotide monophosphate is exchanged for the corresponding nucleotide sugar (reviewed in Hirschberg et al. 1998). This requires that the nucleoside diphosphates released after transfer of the sugar to the glycoconjugate substrate are converted to nucleoside monophosphates by diphosphatases (Fig. 1A). Of course, CMP-Sia is the one exception because CMP is directly released following sugar transfer (Fig. 1B). Additional evidence for the importance of disphosphatase activities in the nucleotide sugar antiporter mechanism came from the finding that deletion of the S. cerevisiae Golgi guanosine diphosphatase (Gda1) decreased GDP-Man transport into membrane vesicles and led to a partial defect in the addition of Man to both glycoproteins and glycolipids (Abeijon et al. 1993; Berninsone et al. 1994). Other work by D’Alessio et al. (2005) demonstrated that GDP-Man dependent glycosylation is reduced but not eliminated in nucleoside diphosphatase mutants in yeast, and suggested that other mechanisms may lead to nucleoside monophosphate translocation. Recent studies by Muraoka et al. (2007) in which the transporter mechanism was evaluated for the endoplasmic reticulum (ER) localized human UGTrel7 transporter, that is capable of transporting UDP-Gal, UDP-GlcA and UDPGlcNAc, and the Golgi localized Drosophila Fringe connection transporter, suggested that the former transporter may be a UDP-sugar/UDP-sugar antiporter, while the latter may use UDP as efficiently as UMP as an antiport substrate.

Multi-specificity and redundancy of NSTs Several NSTs have been identified as multi-substrate transporters including the C. elegans SQV-7 transporter and the Drosophila Fringe connection transporter described above. Other multi-substrate and redundant NSTs emerged as investigators searched the human, Drosophila and yeast genomes for putative nucleotide sugar transporters, and subjected these newly cloned putative transporters to extensive analyses using multiple nucleotide sugars (Segawa et al. 2002; Muraoka et al. 2001; Ashikov et al. 2005). For example, the UDP-Gal transporter was first cloned by complementation of mammalian and yeast mutants (Miura et al. 1996; Tabuchi et al. 1997). Later, Segawa et al. (2002) cloned the Drosophila UDP-Gal transporter and found that both the human UDP-Gal transporter 1 (hUGT1) and the Drosophila transporter were

196

*

W. Zhao and K. J. Colley

specific for both UDP-Gal and UDP-GalNAc. More recently, Capul et al. (2007) identified two genes (LPG5A and LPG5B) in Leishmania major that encode UDP-Gal transporters with partially overlapping activity. The multi-specificity and redundancy of NSTs raised new questions concerning mechanism and potential differences in the roles of NSTs with similar substrate specificities. Investigators have now begun to address some of these questions. The C. elegans genome encodes 18 putative NSTs based on sequence homology with transporters from other species, but the transport of only seven nucleotide sugars is required for glycosylation, and a similar situation exists in humans (Caffaro et al. 2007; Martinez-Duncker et al. 2003). Three redundant, multi-substrate specific C. elegans NSTs, SQV7, SRF-3, and CO3H5.2, have been characterized. As described above, the SQV-7 protein is a multi-substrate transporter that is specific for UDP-GlcA, UDP-GlcNAc, and UDP-Gal and transports these substrates in a competitive and non-cooperative fashion (Berninsone et al. 2001). In contrast to the sqv mutants, the C. elegans srf mutants exhibit no obvious behavioral or morphological changes, but do have defects in cell surface molecules that alter their binding to antibodies and lectins and block infection/colonization by parasites (reviewed in Hoflich et al. 2004). The srf-3 gene encodes a transporter that is specific for both UDP-Gal and UDP-GlcNAc (Hoflich et al. 2004), while the CO3H5.2 gene encodes a transporter specific for UDP-GlcNAc and UDPGalNAc (Caffaro et al. 2006). Surprisingly, the CO3H5.2 and SRF-3 transporters use a simultaneous and non-competitive substrate transport mechanism that differs from the competitive mechanism used by the SQV7 transporter (Caffaro et al. 2006, 2007). A deletion of 16 amino acids in the loop between transmembrane (TM) helices 2 and 3 of the CO3H5.2 protein preferentially decreased UDP-GalNAc transport by 85–90%, but did not impact UDP-GlcNAc transport, suggesting two independent translocation sites for these nucleotide sugars (Caffaro et al. 2006). The existence of these three C. elegans transporters that exhibit partially overlapping substrate specificity (UDP-GlcNAc) and expression patterns, led Caffaro et al. (2007) to investigate this redundancy. They used RNAi technology to knock down the CO3H5.2 gene in srf-3 mutants and found that a defect in both transporters led to developmental and morphological changes not observed in the srf-3 mutant alone. This strongly suggested that these two transporters are at least partially redundant and begged the question why redundancy was needed. One possibility, suggested by the investigators, is that certain nucleotide sugars need to be maintained at high levels so that processes requiring these molecules can proceed with high efficiency under a variety of circumstances. In addition, if nucleotide sugar levels drop, different glycosylation pathways can be differentially affected depending upon the affinity of the associated glycosyltransferases for the particular nucleotide sugar. For example, in the MDCKII-RCAr cell mutant where availability of UDP-Gal is limited, a decreased galactosylation is observed for glycoproteins, glycolipids and keratan sulfate proteoglycans,

Nucleotide sugar transporters

*

197

while the amounts of chondroitin and heparan sulfate proteoglycans remain relatively normal (Toma et al. 1996). Similarly, in LADII/CDG IIc patients who have a defect in the GDP-Fuc transporter, low dose oral Fuc supplementation therapy partially restores P-selectin-mediated, but not E-selectin-mediated binding of neutrophils (Marquardt et al. 1999a, b). Since the synthesis of P- and E-selectin glycan ligands depends upon different Fuc transferase activities (Huang et al. 2000), it is likely that the enzymes have different affinities for GDP-Fuc and that this leads to differences in glycoconjugate expression under limiting GDP-Fuc levels. Another possible reason that there may be genetic pressure to maintain the expression of rendundant multi-substrate transporters is so that the non-overlapping functions of these transporters can be maintained (Caffaro et al. 2006). A third possibility, is that different transporters with similar specificity function in conjunction with specific glycosylation pathways. This idea is supported by the work of Capul et al. (2007) who cloned two UDP-Gal transporters from L. major (LPG5A and LPG5B) and demonstrated that deficiencies in these transporters impacted the biosynthesis of the two predominant Leishmania surface glycoconjugates differently.

NST structure and sequence requirements for function, trafficking, and localization Topology and oligomerization Hydrophobicity plots and topology prediction algorithms based on the deduced amino acid sequences of NSTs suggest that these proteins are multispanning membrane proteins containing six to ten TM regions with both their amino- and carboxy-termini in the cytosol (Hirschberg et al. 1998) (Fig. 2). Initial studies to define the topology of the K. lactis UDP-GlcNAc transporter in Golgi vesicles have been performed and suggest either a six or eight TM helix topology (Berninsone and Hirschberg 2000). In contrast, Eckhardt et al. (1999) evaluated the membrane topology of the murine CMP-Sia transporter using immunofluorescence microscopy following epitope insertion and selective membrane permeablization, and obtained data supporting a 10 TM helix model for this transporter (Eckhardt et al. 1999). Further studies are needed to define the topological arrangement of other NSTs. Most NSTs are thought to exist as homodimers. For example, the rat liver UDP-GalNAc and GDP-Fuc transporters both migrate with molecular masses of approximately 40 kDa upon denaturing gel electrophoresis, but exhibit molecular masses of 80–90 kDa in native glycerol gradients (Puglielli et al. 1999; Puglielli and Hirschberg 1999). VRG4, the S. cerevisiae GDP-Man transporter has also been found to be a homodimer (Gao and Dean 2000). A carboxy-terminal region of this protein, which includes the last TM helix, is necessary for dimer formation, and truncated proteins lacking this sequence are unstable and rapidly degraded. Interestingly, overexpression of an aminoterminal truncated VRG4 protein in yeast causes a dominant negative growth

198

*

W. Zhao and K. J. Colley

Figure 2. Topology and functional regions of the Golgi NSTs. (A) Topology and TM helices required for transport activity. (i) CMP-Sia transporter: TM helix 7 (TM7) is required for the specificity of this transporter for CMP-Sia, whereas the TM helices 2 and 3 enhance the efficiency of CMP-Sia transport. (ii) UDP-Gal transporter: TM helices 1 (TM1) and 8 (TM8) are necessary but not sufficient for UDP-Gal transport, and other helices in different combinations (2, 3 and 7 OR 9 and 10) must be included with TM1 and TM8 for transport activity. (B) Sequences required for dimerization and ER export. (i) In the GDP-mannose transporter, a carboxyterminal sequence containing the last TM helix is involved in dimerization, whereas the amino-terminal 44 amino acids include an ER export signal. (ii) In the CMP-Sia transporter, a di-isoleucine motif and a terminal valine residue (boxed) at the very carboxy-terminus mediate its ER export.

phenotype. This is believed to be a consequence of the formation of inactive heterodimers of the truncated protein and the endogenous full length protein, and supports the notion that homodimerization of VRG4 is crucial for its function. In contrast LPG2, the Leishmania GDP-Man transporter, which migrates as a hexamer in native glycerol gradients and upon pore-limited native gel electrophoresis (Hong et al. 2000). Further functional studies will be needed to determine whether the hexamer form of this protein is the active state in the membrane.

Nucleotide sugar transporters

*

199

Sequence and structural requirements for substrate recognition and transport The identification of mutants in NSTs first gave hints to what regions and amino acid residues are critical for their function. Gao et al. (2001) identified a conserved region in the yeast Vrg4 GDP-Man transporter (amino acids 280–291) as the transporter’s GDP-Man binding site, because mutations in this sequence reduced binding to a photoaffinity substrate analog and led to decreased GDP-Man transport. Gerardy-Schahn and colleagues (Eckhardt et al. 1998; Oelmann et al. 2001) identified the various mutations in the Lec2 and Lec8 complementation groups that lead to defects in the CMP-Sia transporter and UDP-Gal transporter, respectively. While many of the identified mutants of the CMP-Sia transporter were deletions that led to mislocalization and low expression, the Gly189Glu mutant was localized in the Golgi and well-expressed, suggesting that this amino acid was critical for transporter activity per se (Eckhardt et al. 1998). Likewise, the DSer213 and Gly281Asp mutants of the UDP-Gal transporter were localized properly and expressed well, but still inactive. Interestingly, introducing these changes into the CMP-Sia transporter also led to its inactivation, suggesting that these conserved residues are important for general transporter mechanism (Oelmann et al. 2001). Aoki et al. (2001, 2003) identified critical TM helices in both the CMP-Sia and UDP-Gal transporters. These transporters are 43% identical and yet are absolutely specific for their respective substrates. The investigators created chimeras containing TM helices from both transporters and found that CMPSia transporter TM helix 7 was necessary and sufficient for transport of CMPSia when inserted into a UDP-Gal transporter background, while the inclusion of TM helices 2 and 3 enhanced efficiency of transport. In contrast, TM helices 1 and 8 of the UDP-Gal transporter were necessary but not sufficient for UDPGal transport in the context of the CMP-Sia transporter. Only the inclusion of either TM helices 9 and 10, or TM helices 2, 3 and 7 from the UDP-Gal transporter, in addition to TM helix 1 and 8, could generate a chimeric transporter competent to transport UDP-Gal (Aoki et al. 2003).

Sequence requirements for NST ER export and retrieval In eukaryotic secretory pathway, exit of secretory proteins from the ER relies on their sorting into ER-derived COPII-coated vesicles. Much of this sorting is mediated by specific, cytoplasmically exposed signals that can be recognized by subunits of the COPII coat (Barlowe 2003). ER export signals have been found in several NSTs. The amino-terminal 44 amino acids of VRG4, the yeast Golgi GDP-Man transporter, are likely to include an ER export signal because deletion of this region leads to ER accumulation, and fusion of these sequences to related ER proteins promotes their transport to the Golgi apparatus (Gao and Dean 2000). A di-isoleucine motif and a terminal valine in the last four amino acids of the carboxy-terminal cytoplasmic tail of the murine CMP-Sia transporter mediate the ER export of the protein (Zhao et al.

200

*

W. Zhao and K. J. Colley

2006). These signals are independent and both need to be deleted or replaced to abolish ER export. Signals predicted to allow the COPI-coated vesicle-mediated ER retrieval have been found in both ER NSTs and some Golgi NSTs (Martinez-Duncker et al. 2003). The cloning of the gene for a second functional isoform of the human UDP-Gal transporter (hUGT2) predicted that this protein is truncated at its carboxy-terminus (Ishida et al. 1996). Comparison of the localization of the two isoforms revealed that hUGT1 is localized in the Golgi, while hUGT2 is localized in the ER and Golgi (Kabuß et al. 2005). Kabuß et al. (2005) demonstrated that a dilysine motif (LysValLysGlySer) found in the carboxyterminal cytoplasmic tail of hUGT2 was responsible for its ER retrieval and its dual localization. Fusion of this motif is sufficient to redistribute the Golgi CMP-Sia transporter to the ER. A similar dilysine ER retrieval motif also is found in the K. lactis Golgi UDP-GlcNAc transporter, however it is unclear whether this motif actually functions as an ER retrieval signal (Abeijon et al. 1996b). Interestingly, work done by Abe et al. (2004) suggests that a COPImediated retrieval of the GDP-Man transporter to the ER is a critical step in the Golgi localization of this transporter and that lysine residues in its carboxylterminal cytoplasmic tail are necessary for COPI coat interaction and retrieval.

Golgi targeting of NSTs and the organization of glycosylation machinery Many studies on the signals and mechanisms of Golgi glycosylation enzyme localization have revealed that sequences mediating Golgi localization are complex and may reside in different domains of these type II membrane proteins (Colley 1997). Likewise, several redundant mechanisms including those involving lipid partitioning, oligomerization, and retrieval, may be used to maintain Golgi enzymes in their resident cisternae (Colley 1997; Mironov et al. 2005). When considering the organization of glycosylation pathways in the Golgi, it is tempting to speculate that NSTs are co-compartmentalized with the glycosyltransferases that use their nucleotide sugar substrates as donors and that they may even form functional complexes. For this reason we were surprised to find that the CMP-Sia transporter showed a more restricted medial–trans Golgi localization than might be indicated by the rather broad Golgi distribution of sialyltransferases involved in both the sialylation of glycoproteins and glycolipids (Zhao et al. 2006). This finding, together with others (D’Alessio et al. 2003; Kabuß et al. 2005), suggests that CMP-Sia as well as other nucleotide sugars move freely in the lumen of the Golgi apparatus. This is consistent with recent evidence that the Golgi cisternae are more interconnected than once believed (Mironov et al. 2005). The identity of the sequences that mediate the Golgi localization of the NSTs has not been widely investigated, however the sequence requirements for the Golgi localization of some viral multi-spanning membrane proteins have been determined. The first TM helix of the avian coronavirus E1 protein is sufficient to localize two cell surface proteins to the Golgi and is likely to

Nucleotide sugar transporters

*

201

play a role in mediating E1 Golgi localization (Machamer and Rose 1987). Later studies showed that uncharged polar residues that line one face of this TM helix are important for Golgi localization (Swift and Machamer 1991). Work by Locker et al. (1994) suggested that the cytoplasmic tail of the mouse hepatitis virus M protein plays a role in its Golgi localization. We have recently found that, while the CMP-Sia transporter cytoplasmic sequences have no direct role in Golgi localization, fusing the first TM helix of the transporter to the lumenal sequences of an inefficiently retained Golgi sialyltransferase can reduce the level of its Golgi exit (W. Zhao and K. J. Colley, unpublished data). This suggests that, like the Golgi localization of the infectious bronchitis virus E1 protein, the first TM helix of the CMP-Sia transporter may be involved in its Golgi localization. NST Golgi localization could also be mediated by interactions with glycosyltransferases. Complex formation between a transporter and its corresponding glycosyltransferase would presumably enhance the efficiency of the glycosylation reaction by facilitating the transfer of the nucleotide sugar to the glycosyltransferase. Along these lines, Sprong et al. (2003) found that a portion of the Golgi UDP-Gal transporter can associate with the ER localized UDP-galactose:ceramide galactosyltransferase to allow UDP-Gal import into this compartment. However, evidence for other functional NST-glycosyltransferase complexes is lacking. Radiation inactivation studies suggest that the galactosyl- and sialyltransferases are not in functional complexes with the corresponding transporters (Fleischer et al. 1993). Moreover, redistribution studies show that there is no complex formation between the CMP-Sia transporter and the corresponding sialyltransferases (Zhao et al. 2006). Although weak interactions between glycosyltransferases and their respective NSTs are still possible, these results suggest that NSTs may not rely on glycosyltransferases for their Golgi localization.

Conclusions Investigators have made great progress in identifying NSTs and defining their substrate specificity. However, many questions remain concerning the mechanism of nucleotide sugar transport, the roles and expression of multisubstrate and redundant NSTs, the potential connection of redundant transporters with specific glycosylation pathways, the mechanisms of NST Golgi localization, and how NSTs and glycosyltransferases are organized within the Golgi apparatus. For example, how is the developmental, cellular and tissue expression of redundant and multi-substrate NSTs controlled? Do differences in sub-Golgi localization of NSTs with similar substrate specificity allow these transporters to be compartmentalized with glycosylation enzymes in different pathways? How does the simultaneous transport of two substrates occur in one transporter? What are the precise interactions mediating nucleotide sugar recognition and transport? How are NSTs localized in the Golgi, and how are they organized vis a vis their respective glycosyltransferases to

202

*

W. Zhao and K. J. Colley

promote efficient glycosylation? The realization that mutant NSTs lead to human disease and developmental defects has and will continue to generate interest in these proteins and will hopefully stimulate additional research to answer these many questions.

References Abe M, Noda Y, Adachi H, Yoda K (2004) Localization of GDP-mannose transporter in the Golgi requires retrieval to the endoplasmic reticulum depending on its cytoplasmic tail and coatomer. J Cell Sci 117: 5687–5696 € usler A, Moremen K, Hirschberg CB (1993) Abeijon C, Yanagisawa K, Mandon EC, Ha Guanosine diphosphatase is required for protein and sphingolipid glycosylation in the Golgi lumen of Saccharomyces cerevisiae. J Cell Biol 122: 307–323 Abeijon C, Mandon EC, Robbins PW, Hirschberg CB (1996a) A mutant yeast deficient in Golgi transport of uridine diphosphate N-acetylglucosamine. J Biol Chem 271: 8851–8854 Abeijon C, Robbins PW, Hirschberg CB (1996b) Molecular cloning of the Golgi apparatus uridine diphosphate-N-acetylglucosamine transporter from Kluyveromyces lactis. Proc Natl Acad Sci USA 93: 5963–5968 Aoki K, Ishida N, Kawakita M (2001) Substrate recognition by UDP-galactose and CMP-sialic acid transporters. Different sets of transmembrane helices are utilized for the specific recognition of UDP-galactose and CMP-sialic acid. J Biol Chem 276: 21555–21561 Aoki K, Ishida N, Kawakita M (2003) Substrate recognition by nucleotide sugar transporters. Further characterization of substrate recognition regions by analyses of UDP-galactose/CMP-sialic acid transporter chimeras and biochemical analyses of the substrate specificity of parental and chimeric transporters. J Biol Chem 278: 22887–22893 Ashikov A, Routier F, Fuhlott J, Helmus Y, Wild M, Gerardy-Schahn R, Bakker H (2005) The human solute carrier gene SLC35B4 encodes a bifunctional nucleotide sugar transporter with specificity for UDP-xylose and UDP-N-acetylglucosamine. J Biol Chem 280: 27230–27235 Ballou L, Hitzeman RA, Lewis MS, Ballou CE (1991) Proc Natl Acad Sci USA 88: 3209–3212 Barlowe C (2003) Signals for COPII-dependent export from the ER: what’s the ticket out? Trends Cell Biol 13: 295–300 Becker DJ, Lowe JB (1999) Leukocyte adhesion deficiency type II. Biochim Biophys Acta 1455: 193–204 Berninsone P, Miret JJ, Hirschberg CB (1994) The Golgi guanosine diphosphatase is required for transport of GDP-mannose into the lumen of Saccharomyces cerevisiae Golgi vesicles. J Biol Chem 269: 207–211 Berninsone P, Eckhardt M, Gerardy-Schahn R, Hirschberg CB (1997) Functional expression of the murine Golgi CMP-sialic acid transporter in Saccharomyces cerevisiae. J Biol Chem 272: 12616–12619 Berninsone PM, Hirschberg CB (2000) Nucleotide sugar transporters of the Golgi apparatus. Curr Opin Struct Biol 10: 542–547 Berninsone P, Hwang H-Y, Zemtseva I, Horvitz HR, Hirschberg CB (2001) SQV-7, a protein involved in Caenorhabditis elegans epithelial invagination and early embryogenesis, transports UDP-glucuronic acid, UDP-N-acetylgalactosamine, and UDP-galactose. Proc Natl Acad Sci USA 98: 3738–3743 € ndli AW, Hansson GC, Rodriguez-Boulan E, and Simons K (1988) A polarized Bra epithelial cell mutant deficient in translocation of UDP-galactose into the Golgi complex. J Biol Chem 263: 16283–16290

Nucleotide sugar transporters

*

203

Caffaro CE, Hirschberg CB (2006) Nucleotide sugar transporters of the Golgi apparatus: from basic science to diseases. Acc Chem Res 39: 805–812 Caffaro CE, Hirschberg CB, Berninsone PM (2007) Functional redundancy between two Caenorhabditis elegans nucleotide sugar transporters with a novel transport mechanism. J Biol Chem 282: 27970–27975 Caffaro CE, Hirschberg CB, Berninsone PM (2006) Independent and simultaneous translocation of two substrates by a nucleotide sugar transporter. Proc Natl Acad Sci USA 103: 16176–16181 Capul AA, Barron T, Dobson DE, Turco SJ, Beverley SM (2007) Two functionally divergent UDP-Gal nucleotide sugar transporters participate in phosphoglycan synthesis in Leishmania major. J Biol Chem 282: 14006–14017 Colley KJ (1997) Golgi localization of glycosyltransferases: more questions than answers. Glycobiology 7: 1–13 D’Alessio C, Caramelo JJ, Parodi AJ (2005) Absence of nucleoside diphosphatase activities in the yeast secretory pathway does not abolish nucleotide sugar-dependent protein glycosylation. J Biol Chem 280: 40417–40427 Dean N, Zhang YB, Poster JB (1997) The VRG4 gene is required for GDP-mannose transport into the lumen of the Golgi in the yeast, Saccharomyces cerevisiae. J Biol Chem 272: 31908–31914 Descoteaux A, Luo Y, Turco SJ, Beverley SM (1995) A specialized pathway affecting virulence glycoconjugates of Leishmania. Science 269: 1869–1872 Deutscher SL, Nuwayhid N, Stanley P, Briles EIB, Hirschberg CB (1984) Translocation across Golgi vesicle membranes: a CHO glycosylation mutant deficient in CMP-sialic acid transport. Cell 39: 295–299 € hlenhoff M, Bethe A, Gerardy-Schahn R (1996) Expression cloning of the Eckhardt M, Mu Golgi CMP-sialic acid transporter. Proc Natl Acad Sci USA 93: 7572–7576 Eckhardt M, Gotza B, Gerardy-Schahn R (1998) Mutants of the CMP-sialic acid transporter causing the Lec2 phenotype. J Biol Chem 273: 20189–20195 Eckhardt M, Gotza B, Gerardy-Schahn R (1999) Membrane topology of the mammalian CMP-sialic acid transporter. J Biol Chem 274: 8779–8787 Fleischer B, McIntyre JO, Kempner ES (1993) Target sizes of galactosyltransferase, sialyltransferase, and uridine diphosphatase in Golgi apparatus of rat liver. Biochemistry 32: 2076–2081 Gao X-D, Dean N (2000) Distinct protein domains of the yeast Golgi GDP-mannose transporter mediate oligomer assembly and export from the endoplasmic reticulum. J Biol Chem 275: 17718–17727 Gao X-D, Nishikawa A, Dean N (2001) Identification of a conserved motif in the yeast Golgi GDP-mannose transporter required for binding to nucleotide sugar. J Biol Chem 276: 4424–4432 Gerardy-Schahn R, Oelmann S, Bakker H (2001) Nucleotide sugar transporters: biological and functional aspects. Biochimie 83: 775–782 Goto S, Taniguchi M, Muraoka M, Toyoda H, Sado Y, Kawakita M, Hayashi S (2001) UDPsugar transporter implicated in glycosylation and processing of Notch. Nature Cell Biol 3: 816–822 Guillen E, Abeijon C, Hirschberg CB (1998) Mammalian Golgi apparatus UDP-N-acetylglucosamine transporter: molecular cloning by phenotypic correction of a yeast mutant. Proc Natl Acad Sci USA 95: 7888–7892 Haltiwanger RS, Lowe JB (2004) Role of glycosylation in development. Annu Rev Biochem 73: 491–537 Handford M, Rodriguez-Furlan C, Orellana A (2006) Nucleotide-sugar transporters: structure, function and roles in vivo. Braz J Med Biol Res 39: 1149–1158 € hn K, Watson DL, McGrogan PJ, Helmus Y, Denecke J, Yaubenia S, Robinson P, Lu Vestweber D, Marquardt T, Wild MK (2006) Leukocyte adhesion deficiency II patients with a dual defect of the GDP-fucose transporter. Blood 107: 3959–3966

204

*

W. Zhao and K. J. Colley

Herman T, Horvitz HR (1999) Three proteins involved in Caenorhabditis elegans invagination are similar to components of a glycosylation pathway. Proc Natl Acad Sci 96: 974–979 Herman T, Hartweig E, Horvitz HR (1999) sqv mutants of Caenorhabditis elegans are defective in vulval epithelial invagination. Proc Natl Acad Sci USA 96: 968–973 Hirshberg CB, Robbins PW, Abeijon C (1998) Transporters of nucleotide sugars, ATP, and nucleotide sulfate in the endoplasmic reticulum and Golgi apparatus. Annu Rev Biochem 67: 49–69 Hirschberg CB (2001) Golgi nucleotide sugar transport and leukocyte adhesion deficiency II. J Clin Invest 108: 3–6 Hoflich J, Berninsone P, Gobel C, Gravato-Nobre MJ, Libby BJ, Darby C, Politz SM, Hodgkin J, Hirschberg CB, Baumeister R (2004) Loss of srf-3-encoded nucleotide sugar transporter activity in Caenorhabditis elegans alters surface antigenicity and prevents bacterial adherence. J Biol Chem 279: 30440–30448 Hong K, Ma D, Beverley SM, Turco SJ (2000) The Leishmania GDP-mannose transporter is an autonomous, multi-specific, hexameric complex of LPG2 subunits. Biochemistry 39: 2013–2022 Huang MC, Zollner O, Moll T, Maly P, Thall AD, Lowe JB, Vestweber D (2000) P-selectin glycoprotein ligand-1 and E-selectin ligand-1 are differentially modified by fucosyltransferases Fuc-TIV and Fuc-TVII in mouse neutrophils. J Biol Chem 275: 31353–31360 Hwang H-Y, Horvitz HR (2002) The SQV-1 UDP-glucuronic acid decarboxylase and the SQV-7 nucleotide-sugar transporter may act in the Golgi apparatus to affect Caenohabditis elegans vulval morphogenesis and embryonic development. Proc Natl Acad Sci USA 99: 14218–14223 Ishida N, Miura N, Yoshioka S, Kawakita M (1996) Molecular cloning and characterization of a novel isoform of the human UDP-galactose transporter, and of related complementary DNAs belonging ot the nucleotide-sugar transporter gene family. J Biochem 120: 1074–1078 Ishida N, Kuba T, Aoki K, Miyatake S, Kawakita M, Sanai Y (2005) Identification and characterization of human Golgi nucleotide sugar transporter SLC35D2, a novel member of the SLC35 mucleotide sugar transporter family. Genomics 85: 106–116 Kabuß R, Ashikov A, Oelmann S, Gerardy-Schahn R, Bakker H (2005) Endoplasmic reticulum retention of the large splice variant of the UDP-galactose transporter is caused by a dilysine motif. Glycobiology 15: 905–911 Kornfeld S, Kornfeld R (1985) The assembly of asparagine-linked oligosaccharides. Annu Rev Biochem 54: 631–664 Locker JK, Klumperman J, Oorschot V, Horzinek MC, Geuze HJ, Rottier PJ (1994) The cytoplasmic tail of mouse hepatitis virus M protein is essential but not sufficient for its retention in the Golgi complex. J Biol Chem 269: 28263–28269 € rner C (1999) A new type of carbohydrate€ bke T, Marquardt T, Von Figura K, Ko Lu deficient glycoprotein syndrome due to a decreased import of GDP-fucose into the Golgi. J Biol Chem 274: 25986–25989 € rner C (2001) Comple€ bke T, Marquardt T, Etzoni A, Hartmann E, Von Figura K, Ko Lu mentation cloning identifies CDG-IIc, a new type of congenital disorders of glycosylation, as a GDP-fucose transporter deficiency. Nat Genet 28: 73–76 € hn K, Wild MK, Eckhardt M, Gerardy-Schahn R, Vestweber D (2001) The gene Lu defective in leukocyte adhesion deficiency II encodes a putative GDP-fucose transporter. Nat Genet 28: 69–72 Ma D, Russell DG, Beverley SM, Turco SJ (1997) Golgi GDP-mannose uptake requires Leishmania LPG2. J Biol Chem 272: 3799–3805 Machamer CE, Rose JK (1987) A specific transmembrane domain of a coronavirus E1 glycoprotein is required for its retention in the Golgi region. J Cell Biol 105: 1205–1214

Nucleotide sugar transporters

*

205

€ rner C, Fabritz L, Van der Werft N, € hn K, Zimmer K-P, Ko Marquardt T, Brune T, Lu Vormoor J, Freeze HH, Louwen F, Biermann B, Harms E, Von Figura K, Vestweber D, Koch HG (1999a) Leukocyte adhesion deficiency II syndrome, a generalized defect in fucose metabolism. J Pediatr 134: 681–688 Marquardt T, Luhn K, Srikrishna G, Freeze HH, Harms E, Vestweber D (1999b) Correction of leukocyte adhesion deficiency type II with oral fucose. Blood 94: 3976–3985 Martinez-Duncker I, Mollicone R, Codogno P, Oriol R (2003) The nucleotide-sugar transporter family: a phylogenetic approach. Biochimie 85: 245–260  T, Piller V, Piller F, Candelier J-J, Trichet C, Tchernia G, Oriol R, Martinez-Duncker I, Dupre Mollicone R (2005) Genetic complementation reveals a novel human congenital disorder of glycosylation of type II, due to inactivation of the Golgi CMP-sialic acid transporter. Blood 105: 2671–2676 Mironov AA, Beznoussenko GV, Polishchuk RS, Trucco A (2005) Intra-Golgi transport: a way to a new paradigm? Biochim Biophys Acta 1744: 340–350 Miura N, Ishida N, Hoshino M, Yamauchi M, Hara T, Ayusawa D, Kawakita M (1996) Human UDP-galactose transporter: molecular cloning of a complementary DNA that complements the genetic defect of a mutant cell line deficient in UDPgalactose translocator. J Biochem 120: 236–241 Muraoka M, Kawakita M, Ishida N (2001) Molecular characterization of human UDPglucuronic acid/UDP-N-acetylgalactosamine transporter, a novel nucleotide sugar transporter with dual substrate specificity. FEBS Lett 495: 87–93 Muraoka M, Miki T, Ishida N, Hara T, Kawakita M (2007) Variety of nucleotide sugar transporters with respect to the interaction with nucleoside mono- and disphosphates. J Biol Chem 282: 24615–24622 Oelmann S, Stanley P, Gerardy-Schahn R (2001) Point mutations identified in Lec8 Chinese hamster ovary glycosylation mutants that inactivate both the UDP-galactose and CMP-sialic acid transporters. J Biol Chem 276: 26291–26300 Patnaik SK, Stanley P (2006) Lectin-resistant CHO glycosylation mutants. Meth Enz 416: 159–182 Poster JB, Dean N (1996) The yeast VRG4 gene is required for normal Golgi functions and defines a new family of related genes. J Biol Chem 271: 3837–3845 Puglielli L, Hirschberg CB (1999) Reconstitution, identification, and purification of the rat liver Golgi membrane GDP-fucose transporter. J Biol Chem 274: 35596–35600 Puglielli L, Mandon EC, Rancour DM, Menon AK, Hirschberg CB (1999) Identification and purification of the rat liver Golgi membrane UDP-N-acetylgalactosamine transporter. J Biol Chem 274: 4474–4479 Segawa H, Kawakita M, Ishida N (2002) Human and Drosophila UDP-galactose transporters transport UDP-N-acetylgalactosamine in addition to UDP-galactose. Eur J Biochem 269: 128–138 Segawa H, Soares RP, Kawakita M, Beverley SM, Turco SJ (2005) Reconstitution of GDPmannose transport activity with purified Leishmania LPG2 protein into liposomes. J Biol Chem 280: 2028–2035 € cker U (2001) Dual Selva EM, Hong K, Baeg G-H, Beverley SM, Turco SJ, Perrimon N, Ha role of the fringe connection gene in both heparan sulphate and fringe-dependent signaling events. Nat Cell Biol 3: 809–815 Sprong H, Degroote S, Nilsson T, Kawakita M, Ishida N, Van der Sluijs P, Van Meer G (2003) Association of the Golgi UDP-galactose transporter with UDP-galactose: ceramide galactosyltransferase allows UDP-galactose import in the endoplasmic reticulum. Mol Biol Cell 14: 3482–3493 Suda T, Kamiyama S, Suzuki M, Kikuchi N, Nakayama K, Narimatsu H, Jigami Y, Aoki T, Nishihara S (2004) Molecular cloning and characterization of a human multisubstrate specific nucleotide-sugar transporter homologous to Drosophila fringe connection. J Biol Chem 279: 26469–26474

206

*

W. Zhao and K. J. Colley

Swift AM, Machamer CE (1991) A Golgi retention signal in a membrane-spanning domain of coronavirus E1 protein. J Cell Biol 115: 19–30 Tabuchi M, Tanaka N, Iwahara S, Takegawa K (1997) The Schizosaccharomyces pombe gms1 þ gene encodes an UDP-galactose transporter homologue required for protein galactosylation. Biochem Biophys Res Commun 232: 121–125 Thomsen B, Horn P, Panitz F, Bendixen E, Petersen AH, Holm L-E, Nielsen VH, Agerholm JS, Arnbjerg J, Bendixen C (2006) A missense mutation in the bovine SLC35A3 gene, encoding a UDP-N-acetylglucosamine transporter, causes complex vertebral malformation. Genome Res 16: 97–105 Tiralongo J, Ashikov A, Routier F, Eckhardt M, Bakker H, Gerardy-Schahn R, Von Itzstein M (2006) Functional expression of the CMP-sialic acid transporter in Escherichia coli and its identification as a simple mobile carrier. Glycobiology 16: 73–81 Toma L, Pinhal MA, Dietrich CP, Nader HB, Hirschberg, CB (1996) Transport of UDPgalactose into the Golgi lumen regulates the biosynthesis of proteoglycans. J Biol Chem 271: 3897–3901 Varki A (1993) Biological roles of oligosaccharides: all of the theories are correct. Glycobiology 3: 97–130 Willig T-N, Breton-Gorius J, Elbim C, Mignotte V, Kaplan C, Mollicone R, Pasquier C, lot F, Cartron J-P, Gougerot-Pocidalo M-A, Debili N, Guichard, Filipe A, Mie Dommergues J-P, Mohandas N, Tchernia G (2001) Macrothrombocytopenia with abnormal demarcation membranes in megakaryocytes and neutropenia with a complete lack of sialyl-Lewis-X antigen in leukocytes – a new syndrome? Blood 97: 826–828 Zhao W, Chen TL, Vertel BM, Colley KJ (2006) The CMP-sialic acid transporter is localized in the medial–trans Golgi and possesses two specific endoplasmic reticulum export motifs in its carboxyl-terminal cytoplasmic tail. J Biol Chem 281: 31106–31118

Luminal lectins

*

207

Luminal lectins Beat Nyfeler, Eva Koegler, Veronika Reiterer and Hans-Peter Hauri

Introduction Asparagine-linked glycosylation (N-glycosylation) is a major post-translational modification of secretory and membrane proteins and influences important physical protein properties such as conformation, stability and solubility (Helenius and Aebi 2001). The majority of proteins that enter the secretory pathway receives multiple N-linked glycans. N-glycosylation is initiated cotranslationally in the lumen of the endoplasmic reticulum (ER) by oligosaccharyltransferase. This multisubunit protein complex scans nascent proteins for N-glycosylation consensus sequences (Asn–X–Ser/Thr) and catalyzes the transfer of a 14-saccharide core glycan to the asparagine residue (Fig. 1). About two-thirds of all consensus sites are glycosylated. After conjugation to the protein, the 14-saccharide core is trimmed in ER and Golgi by glycosidases and extended in the Golgi by glycosyltransferases (Kornfeld and Kornfeld 1985). ER glucosidases I and II remove the three glucose (Glc) residues, whereas ER a1,2 mannosidase I and Golgi a1,2 mannosidases 1A, 1B and 1C trim the a1,2-linked mannoses (Man). In the Golgi, two additional Man residues are cleaved and the N-glycans undergo complex glycosylation by the addition of N-acetylglucosamine (GlcNAc), fucose, galactose and sialic acid residues. After traversing the Golgi, glycoproteins carry various N-linked glycans differing in composition and structure. This heterogeneity allows mature glycoproteins to fulfill a plethora of functions including the presentation of interaction sites for other molecules (Varki 1993). In contrast, nascent glycoproteins in the early secretory pathway, termed high-mannose glycoproteins, display only few but distinct oligosaccharide structures which function as recognition tags for different sugar-binding proteins (lectins). With their carbohydrate recognition domain (CRD), the lectins bind newly synthesized glycoproteins and control their folding, degradation, transport and sorting. Here, we provide an overview of the different animal lectins localized to the lumen of the secretory pathway (Table 1) and describe them grouped according to their proposed function.

Functions of luminal lectins Protein folding and quality control in the ER: calnexin, calreticulin The lumen of the ER features a specialized cellular environment for protein folding and modification. Multiprotein networks of general chaperones,

208

*

B. Nyfeler et al.

Figure 1. The N-linked core glycan. The 14-saccharide core glycan contains two N-acetylglucosamine (GlcNAc, boxes), nine mannose (Man, circles) and three glucose (Glc, triangle) residues. The nomenclature of the different branches and the type of glycosidic linkage is indicated.

peptidyl prolyl isomerases and thiol oxidoreductases help newly synthesized proteins to gain their native conformation. Important factors in the folding process of glycoproteins are calnexin (CNX, Wada et al. 1991) and calreticulin (CRT, Fliegel et al. 1989), two homologous calcium-binding lectins which localize to the ER and function as molecular chaperones (Ou et al. 1993). CNX is a type I transmembrane protein of 90 kDa, whereas CRT is a 60-kDa soluble luminal protein that is retained in the ER by its C-terminal KDEL retention signal. Both proteins are monomeric and display a similar luminal fold consisting of a globular and an extended domain. The globular domain contains the CRD and has a b-sandwich fold related to plant legume lectins. The extended domain is proline-rich, hence termed P domain, and comprises two b-strands forming a long hairpin. Bound substrate glycoproteins are believed to localize to the space between the extended P domain and the globular lectin domain. How are glycoproteins bound? Although the contribution of protein–protein interactions to substrate capture is still a matter of debate (Williams 2006), binding of carbohydrate moieties by the CRDs of CNX and CRT is well documented and accepted. The in vitro established carbohydrate specificity for a single a1,3-linked Glc and three Man residues in the A branch is in line with the preferential binding of monoglucosylated glycoproteins observed in vivo. Hence, CNX and CRT bind glycoproteins immediately after addition of the N-linked core glycan and removal of the two outermost Glc residues by ER glycosidases I and II. The majority, if not all, glycoproteins bind to either CNX, CRT or both (Helenius et al. 1997). If the association with the two lectins is inhibited, for instance by the ER glucosidase inhibitor castanospermine, glycoprotein folding is more rapid but also more error-prone which results in more misfolded and degraded protein species. Therefore, CNX and CRT increase the efficiency of protein folding by (i) slowing down the folding reaction, (ii) preventing off-pathway folding reactions and aggregation, and (iii) favoring native disulfide-bridge forma-

Luminal lectins

*

209

Table 1. Overview of luminal animal lectins Lectin

Intracellular localization

Carbohydrate specificity

Proposed function

Calnexin

ER

Calreticulin

ER

ER a1,2mannosidasea

ER

Protein folding Protein folding Protein degradation

EDEM1

ER

EDEM2

ER

EDEM3

ER

Yos9p(yeast)a

ER

Monoglucosylated A branch Monoglucosylated A branch Man9GlcNAc2: Man trimming to Man5–6GlcNAc2 High-mannose (Man8GlcNAc2) High-mannose (Man8GlcNAc2) High-mannose (Man8GlcNAc2) Mannose?

ERGIC-53

ERGIC

ERGLa VIP36

ER ERGIC, Golgi

VIPL

ER

CI-M6PR

Trans-Golgi, endosomes, plasma membrane

CD-M6PR

Trans-Golgi, endosomes, plasma membrane

a

High-mannose, broad specificity (Man9GlcNAc2 to (Man6GlcNAc2) ? De-glucosylated A branch De-glucosylated A branch Mannose 6-phosphatemonoester, Mannose 6-phosphatediester

Mannose 6-phosphatemonoester

Protein degradation Protein degradation Protein degradation Protein degradation ER-ERGIC transport

Protein sorting in ER? Golgi–ER transport? intra-Golgi transport? Protein sorting in ER? Transport of lysosomal proteins from trans-Golgi and plasma membrane to endosomes, IGF-II receptor Transport of lysosomal proteins from trans-Golgi to endosomes.

Lectin activity uncertain

tion. The latter is achieved by exposing the substrate glycoproteins to the oxidoreductase ERp57 which binds to the P domain of CNX and CRT and catalyzes the formation and isomerization of disulfide bonds. The association of glycoproteins with CNX and CRT is transient and terminated by the action of ER glucosidase II that removes the remaining single Glc residue of the A branch. The resulting deglucosylated high-mannose oligosaccharide structure shows low affinity toward CNX and CRT leading to the release of glycoproteins into the lumen of the ER. Correctly folded glycoproteins can

210

*

B. Nyfeler et al.

now exit the ER in coat protein II (COPII)-coated vesicles. While a single association with CNX and CRT suffices for some glycoproteins to gain their native conformation, others are still misfolded. These misfolded proteins are recognized by UDP-glucose:glycoprotein glucosyltransferase (UGT1), a folding sensor which re-glucosylates nearly native glycoproteins. Terminally misfolded proteins are not recognized and are ultimately eliminated by ER-associated degradation (ERAD). Hence, UGT1 is the ER quality control component that monitors the conformational status of folding glycoproteins. Re-glucosylation of non-native glycoproteins allows multiple rounds of association with CNX and CRT, a process termed calnexin/calreticulin cycle (Parodi 2000). Having two lectins with different topologies broadens the scope of glycoproteins substrates that can be assisted in this cycle. In addition to their involvement in protein folding, CNX and CRT seem to have additional functions in development. Although CNX- and CRT-deficient cell lines can survive in culture, knockout mice show severe phenotypes. A knockout of CNX results in motor disorders associated with a dramatic loss of nerve fibers and in reduced survival. A CRT knockout is embryonically lethal.

ER-associated protein degradation: ER mannosidase, EDEM, Yos9p Terminally misfolded proteins need to be removed from the ER as their accumulation may compromise the folding and secretory capacity of the cell since misfolded proteins have a tendency to aggregate. If refolding by the CNX/CRT cycle is unsuccessful, terminally misfolded glycoproteins are retrotranslocated into the cytosol and degraded by the proteasome. This process, known as ERAD, is determined by mannose trimming as indicated by the fact that kifunensin, an inhibitor of ER mannosidase I (ERManI), or deoxymannojirimycin inhibit ERAD. Mannose trimming exposes sugars that do not allow reglucosylation by UGT1 and association with CNX/CRT. In yeast trimming of a single mannose residue from branch B suffices to divert a glycoprotein to the ERAD pathway. In contrast, mammalian glycoproteins require more extensive mannose trimming down to Man5–6 (Molinari 2007). In particular, removal of the terminal mannose of the A branch prevents reglucosylation by UGT1 and binding to CNX/CRT with additional contribution of mannose trimming from branches B and C. Overexpression and knockdown studies indicate that mannose trimming down to Man5–6 is catalyzed by ERManI (Avezov et al. 2008). Overexpressed ERManI localizes to a percentriolar ER-derived structure that may constitute a specific ER quality compartment but confirmation of this notion requires localization of endogenous ERManI. Moreover, it is presently unclear if ERManI also operates as a mannose lectin routing glycoproteins to ERAD. Additional factors in ERAD are the recently discovered putative mannosebinding proteins EDEM1, EDEM2 and EDEM3 which are homologous to ERManI (Olivari and Molinari 2007). EDEMs (for ER degradation-enhancing a-mannosidase-like proteins) have different tissue distributions and it is currently unclear if they are functionally redundant. Endogenous EDEM1 is concentrated in ER buds that give rise to 150 nm vesicles lacking COPII and

Luminal lectins

*

211

ERGIC-53 but contain misfolded alpha1-antitrypsin and occasionally Derlin-2 that may constitute part of the retrotranslocation channel (Zuber et al. 2007). Whether these structures relate to the ERManI compartment, remains to be shown. Up- or downregulation of EDEMs modulates degradation of foldingdefective glycoproteins. For instance, EDEM1 enhances removal of the terminal branch A mannose (Ermonval et al. 2001; Olivari et al. 2006), but the mechanism of action is not fully understood. EDEMs may act as mannosidases that determine the rate of ERAD substrate demannosylation. Alternatively, they may act as classical chaperones by preventing aberrant oligomerization and aggregate formation (Hosokawa et al. 2006). Yet another possibility is a lectin function in bridging misfolded proteins to the retrotranslocation pore in view of EDEMs association with derlin 2 and 3. A second class of lectin-like ERAD factors is Yos9p in yeast (Kanehara et al. 2007). Although the name Yos9p is derived from the mammalian protein OS9, Yos9p's mammalian orthologue has not been characterized. Yos9 shares a lectin-like domain with the mannose 6-phosphate receptors (Whyte and Munro 2001) and is required for efficient ERAD. In yeast Yos9p binds to misfolded glyproteins, and misfolded glycoproteins degrade very poorly in the absence of Yos9p. Although the degradation of non-glycosylated variants shows no dependence on Yos9p, the role of glycans in this process has not been clarified. Like proposed for EDEMs, Yos9p may operate in guiding terminally misfolded glycoproteins to the retrotranslocation pore.

Protein transport: L-type lectins Upon correct folding, native proteins are exported from the ER in COPIIcoated vesicles that mediate transport to the ER Golgi intermediate compartment (ERGIC). In many cases protein export from the ER is selective. In this process transmembrane proteins can directly interact with the cytosolic COPII coat and thereby convey their selective incorporation into transport vesicles. In contrast, for soluble proteins, selective recruitment into COPII vesicles requires the assistance of transmembrane receptors, termed cargo receptors (Appenzeller et al. 1999; Baines and Zhang 2007). In mammalian cells, the best characterized cargo receptor is ERGIC-53 (Hauri et al. 2000b) that belongs to a family of four related lectins also comprising ERGL, VIP36 and VIPL. These four proteins are all type I membrane proteins, share a conserved luminal CRD that is homologous to leguminous plant lectins (L-type lectins), localize to the ER/ Golgi interface, and are believed to assist glycoprotein sorting and transport in the early secretory pathway.

ERGIC-53 ERGIC-53 (ER Golgi intermediate compartment protein of 53 kDa, gene name: LMAN1) is a dynamic protein which cycles between ER and ERGIC with a minor cycling route via the cis-Golgi (Appenzeller-Herzog and Hauri 2006; Schweizer et al. 1988). At steady state ERGIC-53 is mainly localized to the ERGIC. The

212

*

B. Nyfeler et al.

cycling properties of ERGIC-53 are determined by sorting motifs in its carboxyl-terminus, which interact with cytosolic coat components. A di-phenylalanine motif interacts with the Sec24 subunit of the COPII coat and mediates ER export, and a di-lysine motif binds COPI and retrieves ERGIC-53 back from ERGIC and cis-Golgi. The luminal domain of ERGIC-53 contains the CRD and a coiled-coil stalk domain that allows oligomerization into homo-dimers and hexameres. The crystal structure of the CRD of ERGIC-53 uncovered two Ca2+binding sites and an overall b-sandwich structure composed of one concave and one convex b-sheet with the ligand-binding site located in a negatively charged cleft. ERGIC-53 binds high-mannose glycoproteins and its CRD shows in vitro a broad specificity but low affinity towards various high-mannose carbohydrate structures. What is the function of ERGIC-53’s lectin domain? By its CRD, ERGIC-53 can capture glycoproteins in the lumen of the ER and through its cytosolic di-phenylalanine motif bind to COPII, thereby recruiting soluble cargo glycoproteins into COPII vesicles. Hence, ERGIC-53 acts as a cargo transport receptor by mediating ER-to-ERGIC transport of some glycoproteins. In addition, ERGIC-53 facilitates the assembly of immunoglobulin M polymers in the lumen of the ER. The identified cargo glycoproteins of ERGIC-53 include blood coagulation factors V (FV) and VIII (FVIII), cathepsin C, and cathepsin Z. Loss-of-function mutations in human ERGIC-53 result in the clinical manifestation of combined FV and FVIII deficiency. Patients lacking ERGIC-53 show a bleeding phenotype due to reduced secretion of FV and FVIII into the blood plasma. Interestingly, efficient secretion of FV and FVIII requires an additional factor known as MCFD2 (Zhang et al. 2003). MCFD2 is a small soluble protein that interacts with ERGIC-53 in a calcium-dependent but lectin activity-independent manner and is believed to recruit FV and FVIII to ERGIC-53. In contrast, MCFD2 is dispensable for the interaction of ERGIC-53 and cathepsin Z. In this interaction ERGIC-53 recognizes a high-mannose N-glycan and a folded b-hairpin peptide structure in cathepsin Z. The requirement of a combined oligosaccharide/ peptide structure may limit the repertoire of possible ERGIC-53 cargo proteins and has led to the suggestion that ERGIC-53 might function in secondary quality control by capturing only native cargo proteins. Cargo protein release from ERGIC-53 occurs in the ERGIC probably due to a drop in pH which results in the protonation of His178, the loss of a calcium ion, and lowered affinity for glycans.

VIP36 The ERGIC-53-related lectin VIP36 (vesicular integral-membrane protein of 36 kDa, gene name: LMAN2, Fiedler et al. 1994) localizes to the early secretory pathway and cycles between ER, ERGIC, and cis-Golgi (Fullekrug et al. 1999). If highly overexpressed, VIP36 can be found in the Golgi, apical and basolateral vesicles, and the plasma membrane. In its cytoplasmic domain VIP36 carries a putative phenylalanine/tyrosine ER export motif and a di-basic motif (KR) that

Luminal lectins

*

213

may operate in retrieval although it does not entirely conform to the established di-lysine ER retention/retrieval consensus signal. Less efficient retrieval than observed for ERGIC-53 may explain why VIP36 has access to the transGolgi and why there is some controversy as to whether VIP36 also operates in the late secretory pathway. In contrast to ERGIC-53, VIP36 is N-glycosylated but lacks the luminal coiled-coil domain required for oligomerization. Hence, VIP36 likely functions as a monomer. The luminal CRD was crystallized recently (Satoh et al. 2007). It exhibits a b-sandwich fold composed of two antiparallel b-sheets. Although not supported by all studies, VIP36 seems to bind carbohydrates in a calcium-dependent manner and the crystal structure identified Asp131, Asn166 and His190 as calcium-binding residues. It was postulated that VIP36 may function in post-ER quality control of glycoproteins in the Golgi by recycling inadequately trimmed glycoproteins (Hauri et al. 2000a). Recent biochemical data lend support to this hypothesis. VIP36 shows highest affinity for the deglucosylated high-mannose A branch as assessed by affinity chromatography (Kamiya et al. 2005). The binding strength changes with pH and shows a bell-shaped dependence with an optimum at pH 6.5. This optimum corresponds to the pH of the medial/trans-Golgi. Thus, glycoproteins with an untrimmed A branch that have inadvertently escaped the ER would be captured by VIP36 in the Golgi and retrieved back to the cis-Golgi or even the ER for another round of mannose trimming. The functional characterization of VIP36 suffers from the lack of characterized cargo glycoproteins and from the fact that overexpression of the protein considerably shifts its localization from the early to the late secretory pathway.

ERGL Human ERGL (ERGIC-53-like protein, gene name: LMAN1L), identified in a prostate-specific EST cluster, is homologous to ERGIC-53 (Yerushalmi et al. 2001). SLAMP (sublingual acinar membrane protein) likely represents the corresponding rat orthologue (Sakulsak et al. 2005). In contrast to the ubiquitously expressed ERGIC-53, ERGL mRNA is found only in a few tissues, including prostate, spleen, salivary gland, cardiac atrium and distinct cells of the central nervous system. The reason for this tissue-specific expression is currently unclear. The amino acid sequence of ERGL reveals a luminal coiledcoil domain which may allow oligomerization. ERGL has a longer C-terminal tail than the other animal L-type lectins and lacks typical transport motifs. No cell line has been found that expresses endogenous ERGL protein but transfection studies with human ERGL have provided some interesting information (Liang, L. and Hauri H.P., unpublished). ERGL is confined to the ER and has a short half life of about 30 min as opposed to ERGIC-53 with a half life of days. Surprisingly, overexpression of ERGL selectively retains ERGIC-53 in the ER by disulfide-bond-mediated interaction, and this interaction reduces mannosebinding activity of ERGIC-53 in an in vitro assay. Together these data suggest that ERGL may function as an ERGIC-53-regulating protein in the ER.

214

*

B. Nyfeler et al.

VIPL VIPL (VIP36-like protein, gene name: LMAN2L) exhibits 43% and 68% sequence similarity to ERGIC-53 and VIP36, respectively. Like VIP36, VIPL is a monomeric and glycosylated membrane protein (Nufer et al. 2003). However, in contrast to VIP36 that has access to the Golgi and is complex glycosylated, VIPL is an ER resident protein that displays high-mannose-type glycans and remains endoglycosidase-H sensitive. The ER localization of VIPL is determined by a di-arginine-based ER retention motif in its cytosolic tail (RKR). In addition, VIPL contains a potential ER export motif (FY) and, if the ER retention motif is mutated, can leave the ER. The luminal CRD of VIPL shows the highest affinity for the three Mana1–2Mana1–2Man residues of the A branch (Kamiya et al. 2008). Glucosylation strongly reduces binding. Sugar-binding of VIPL is calciumdependent and most efficient at pH 7.5–8, which corresponds to the pH of the ER. This pH optimum, together with the carbohydrate specificity, suggests that VIPL binds glycoproteins in the ER. One scenario is that VIPL delivers highmannose glycoproteins, released from the CNX/CRT cycle, to ERGIC-53 for transport. Binding to VIPL might protect high-mannose glycans on native glycoproteins from extensive mannose trimming which would otherwise divert the glycoprotein to the ERAD pathway. Furthermore, VIPL may control the intracellular distribution of other lectins since its overexpression relocalizes ERGIC-53 to the ER. VIPL’s short half-life of 30 min is in line with a potential regulatory function. So far no direct cargo glycoproteins have been identified for VIPL, but a siRNA-based knockdown of VIPL results in reduced secretion of two glycoproteins of unknown identity with a molecular mass of 35 kDa and 250 kDa (Neve et al. 2003).

Lysosomal protein sorting in the Golgi: P-type lectins After export from the ER and transport through the ERGIC folded proteins reach the cis-Golgi. In the cis-Golgi newly synthesized proteins destined for lysosomes are modified by N-acetylglucosamine 1-phosphate at the 6th position of selected mannose residues by the action of GlcNAc-phosphotransferase (Kornfeld and Mellman 1989). From the resulting phosphodiester intermediate the N-acetylglucosamine residue is removed by uncovering enzyme in the TGN resulting in a phosphomonoester (Kornfeld 1987). The secondary or tertiary structure of the protein seems to be crucial for recognition by the N-glycan phosphorylation machinery. Obtaining this modification at one or two mannose residues, lysosomal enzymes can now be distinguished from other proteins traversing the Golgi apparatus. In the TGN the mannose 6-phosphate receptors (MPRs) bind about 50 different acid hydrolases, transport them to and release them in endosomes upon a drop in pH. Further transport to lysosomes occurs in a MPR-independent manner. For completeness it is worth mentioning that some lysosomal hydrolases are targeted to lysosomes in a MPR-independent manner.

Luminal lectins

*

215

MPRs are the only members of the P-type lectin family and come in two flavours: cation-dependent MPR (CD-MPR; Mr 46 kDa) and cation-independent MPR (CI-MPR; Mr 300 kDa, also termed IGF-II/MPR since it also acts as insulin-like growth factor II receptor). Both MPRs cycle constitutively between TGN, early endosomes, recycling endosomes, late endosomes and plasma membrane, but they are absent from lysosomes. Trafficking is controlled by post-translational modifications such as phosphorylation, palmitoylation, and by transport signals present in the cytoplasmic tail of the MPRs. Export from the trans-Golgi is mediated by binding to the clathrin-adapter AP-1. GGA1, 2 and 3 bind the acidic cluster-dileucine in the cytoplasmic tail and mediate binding of the receptors to AP1, which then nucleates clathrin-coated vesicles (Puertollano et al. 2001; Zhu et al. 2001). After reaching endosomes, a phenylalanine–tryptophan motif ensures retention in late endosomes. Recycling back to the trans-Golgi network is mediated by a clathrin-independent pathway involving TIP47, Rab9, PACS-1, and AP-1. MPRs can reach the plasma membrane either by recycling from early or late endosomes or by missorting directly from the trans-Golgi. Subsequent internalization at the plasma membrane is accomplished by the YSKV motif of the CI-MPR whereas the CD-MPR has three separate internalization sequences (phenylalaninecontaining sequence, tyrosine-based motif and a dileucine motif). The luminal domain of the CD-MPR binds lysosomal enzymes. It consists of a single a helix near the N-terminus and nine primarily antiparallel b strands that form two b sheets (Olson et al. 2002). When ligand is bound, the binding site encompasses the phosphate group and the terminal three mannose rings of the cargo protein. Two amino acid side chains (Q66 and R111) constitute the binding specificity for mannose 6-phosphate rather than glucose 6phosphate. Additionally, the phosphate group itself is relevant for highaffinity ligand recognition by the receptor. A Mn2+ ion and water establish additional contacts with the phosphate. Structural analysis revealed that the carbohydrate recognition domain lies relatively deep inside the protein. This explains the high binding affinity of the CD-MPR. In comparison to many other lectins, where water fills the ligand-free cleft, the binding pocket of the CD-MPR is occupied by a loop in the absence of ligand. The role of the CD-MPR at the plasma membrane remains mysterious. At neutral pH (pH 7.4) the CD-MPR does not bind ligands efficiently, but it may regulate secretion of mannose 6-phosphate containing ligands into the extracellular milieu. The CI-MPR has a large extracytoplasmic domain composed of 15 repeating units. CI-MPR possesses two distinct carbohydrate recognition sites and a single IGF-II binding site. Structure-based sequence alignments with the carbohydrate recognition domain of the CD-MPR and mutagenesis experiments predict an arginine residue and several other amino acids in two repeating units to be important for ligand recognition. The two CRDs of the CI-MPR are structurally and mechanistically similar to each other, but there are also differences. First, the amino-terminal binding site shows efficient binding at higher pH and second, it is able to bind to larger M6P-OGlcNAc

216

*

B. Nyfeler et al.

phosphodiesters, M6P-OCH3 phosphodiester and mannose 6-sulfate with only slightly lower affinities than M6P (Distler et al. 1991). CI-MPR is posttranslationally modified with N-glycans in the extracytoplasmic domain, although they are not required for ligand-binding. However, the formation of disulfide bonds is crucial for proper folding of the receptor. Furthermore, the receptor is palmitoylated but neither the attachment site nor the function

Figure 2. Localization and function of luminal lectins. Secretory and membrane proteins enter the ER co-translationally through the Sec61 translocation pore and get N-glycosylated by the oligosaccharyltransferase complex (OST). Upon trimming by ER glucosidase I and II, monoglucosylated proteins are bound by CNX and CRT which increases the efficiency of protein folding (2). After release from CNX/CRT by ER glucosidase II (3), glycoproteins can follow three different fates depending on their conformation. Nearly native glycoproteins will be re-glucosylated by UGT1 and re-enter the CNX/CRT cycle (4). Terminally misfolded glycoproteins will be subjected to extensive Man trimming by ER mannosidase I and re-translocated into the cytosol for proteasomal degradation (6). The EDEMs may act as mannosidases or as lectin chaperones. Native glycoproteins will be captured and packaged into COPII vesicles for ER exit. VIPL might bind native glycoproteins released from CNX/CRT (7) and pass the proteins on to ERGIC-53 (8). ERGIC-53 is a cargo receptor for a subset of glycoproteins and facilitates ER to ERGIC transport (9). In the ERGIC, cargo is released upon a drop in pH and perhaps calcium. The ERGIC-53-related lectin VIP36 functions at the Golgi interface and is proposed to capture proteins which have escaped the ER quality control machinery and recycle them back to the ER (10) or cis-Golgi (11). For sorting to lysosomes, GlcNAc 1-phosphate is transferred to high-mannose glycans of lysosomal enzymes in the cis-Golgi (12), uncovered in the trans-Golgi (13). Man6 phosphorylated proteins are captured in the TGN by MPRs and transported to endosomes (14).

Luminal lectins

*

217

is known. CI-MPR can also endocytose various ligands including Man6Pmodified lysosomal enzymes from the plasma membrane.

Conclusions and perspectives Animal cells control the folding, degradation, sorting and transport of newly synthesized glycoprotein by expressing several luminal lectins which differ in their intracellular localization and function as illustrated in Fig. 2. Based on the characterization of the carbohydrate-specificities of the different lectins, the A branch of the core-glycan emerges as important determinant for the fate of a glycoprotein. While mono-glucosylation of the A branch retains a glycoprotein in the CNX/CRT cycle, extensive trimming of a1,2-linked Man residues in the A branch seems to mark terminally misfolded glycoproteins for ERAD. Moreover, the de-glucosylated A branch is the preferential binding substrate of VIPL, which might protect native glycoproteins from extensive Man trimming in the ER and degradation. The increasing number of CRD crystal structures allows us to understand the molecular determinants of carbohydrate-specificity. A case in point is the change of VIP36’s sugarspecificity based on a rational amino acid substitution in its CRD (Kamiya et al. 2008). Apart from binding to N-glycans of substrate glycoproteins, at least some of the lectins seem to recognize also protein determinants. While the notion of such a bi-partite interaction is still controversial for CNX and CRT, ERGIC-53 definitively recognizes a combined oligosaccharide/peptide structure (Appenzeller-Herzog et al. 2005). Future work is likely to identify the contribution of protein determinants to substrate binding and the binding specificity of the other lectins. Further unresolved questions concern the functions of the less-characterized lectins such as the EDEMs, VIP36, VIPL or ERGL. Are the EDEMs solely lectins or do the proteins possess mannosidase activity? Do the three EDEMs fulfill redundant functions? What are the cargo proteins of VIP36, VIPL and ERGL? Do the lectins relay substrate proteins as proposed for VIPL passing cargo on to ERGIC-53?

References Appenzeller C, Andersson H, Kappeler F, Hauri HP (1999) The lectin ERGIC-53 is a cargo transport receptor for glycoproteins. Nat Cell Biol 1: 330–334 Appenzeller-Herzog C, Hauri HP (2006) The ER–Golgi intermediate compartment (ERGIC): in search of its identity and function. J Cell Sci 119: 2173–2183 Appenzeller-Herzog C, Nyfeler B, Burkhard P, Santamaria I, Lopez-Otin C, Hauri HP (2005) Carbohydrate- and conformation-dependent cargo capture for ER-exit. Mol Biol Cell 16: 1258–1267 Avezov E, Frenkel Z, Ehrlich M, Herscovics A, Lederkremer GZ (2008) ER Mannosidase I Is compartmentalized and required for N-glycan trimming to Man5 6GlcNAc2 in glycoprotein ER-associated degradation. Mol Biol Cell 19: 216–225 Baines AC, Zhang B (2007) Receptor-mediated protein transport in the early secretory pathway. Trends Biochem Sci 32, 381–388 Distler JJ, Guo JF, Jourdian GW, Srivastava OP, Hindsgaul O (1991) The binding specificity of high and low molecular weight phosphomannosyl receptors from bovine testes.

218

*

B. Nyfeler et al.

Inhibition studies with chemically synthesized 6-O-phosphorylated oligomannosides. J Biol Chem 266: 21687–21692 Ermonval M, Kitzmuller C, Mir AM, Cacan R, Ivessa NE (2001) N-glycan structure of a short-lived variant of ribophorin I expressed in the MadIA214 glycosylation-defective cell line reveals the role of a mannosidase that is not ER mannosidase I in the process of glycoprotein degradation. Glycobiology 11: 565–576 Fiedler K, Parton RG, Kellner R, Etzold T, Simons K (1994) VIP36, a novel component of glycolipid rafts and exocytic carrier vesicles in epithelial cells. EMBO J 13: 1729–1740 Fliegel L, Burns K, MacLennan DH, Reithmeier RA, Michalak M (1989) Molecular cloning of the high affinity calcium-binding protein (calreticulin) of skeletal muscle sarcoplasmic reticulum. J Biol Chem 264: 21522–21528 Fullekrug J, Scheiffele P, Simons K (1999) VIP36 localisation to the early secretory pathway. J Cell Sci 112(Pt 17): 2813–2821 Hauri H, Appenzeller C, Kuhn F, Nufer O (2000a) Lectins and traffic in the secretory pathway. FEBS Lett 476: 32–37 Hauri HP, Kappeler F, Andersson H, Appenzeller C (2000b) ERGIC-53 and traffic in the secretory pathway. J Cell Sci 113(Pt 4): 587–596 Helenius A, Aebi M (2001) Intracellular functions of N-linked glycans. Science 291: 2364–2369 Helenius A, Trombetta ES, Hebert DN, Simons JF (1997) Calnexin, calreticulin and the folding of glycoproteins. Trends Cell Biol 7: 193–200 Hosokawa N, Wada I, Natsuka Y, Nagata K (2006) EDEM accelerates ERAD by preventing aberrant dimer formation of misfolded alpha1-antitrypsin. Genes Cells 11: 465–476 Kamiya Y, Kamiya D, Yamamoto K, Nyfeler B, Hauri HP, Kato K (2008) Molecular basis of sugar recognition by the human L-type lectins ERGIC-53, VIPL and VIP36. J Biol Chem 283: 1857–1861 Kamiya Y, Yamaguchi Y, Takahashi N, Arata Y, Kasai K, Ihara Y, Matsuo I, Ito Y, Yamamoto K, Kato K (2005) Sugar-binding properties of VIP36, an intracellular animal lectin operating as a cargo receptor. J Biol Chem 280: 37178–37182 Kanehara K, Kawaguchi S, Ng DT (2007) The EDEM and Yos9p families of lectin-like ERAD factors. Semin Cell Dev Biol 18: 743–750 Kornfeld R, Kornfeld S (1985) Assembly of asparagine-linked oligosaccharides. Annu Rev Biochem 54: 631–664 Kornfeld S (1987) Trafficking of lysosomal enzymes. FASEB J 1: 462–468 Kornfeld S, Mellman I (1989) The biogenesis of lysosomes. Annu Rev Cell Biol 5: 483–525 Molinari M (2007) N-glycan structure dictates extension of protein folding or onset of disposal. Nat Chem Biol 3: 313–320 Neve EP, Svensson K, Fuxe J, Pettersson RF (2003) VIPL, a VIP36-like membrane protein with a putative function in the export of glycoproteins from the endoplasmic reticulum. Exp Cell Res 288: 70–83 Nufer O, Mitrovic S, Hauri HP (2003) Profile-based data base scanning for animal L-type lectins and characterization of VIPL, a novel VIP36-like endoplasmic reticulum protein. J Biol Chem 278: 15886–15896 Olivari S, Cali T, Salo KE, Paganetti P, Ruddock LW, Molinari M (2006) EDEM1 regulates ER-associated degradation by accelerating de-mannosylation of folding-defective polypeptides and by inhibiting their covalent aggregation. Biochem Biophys Res Commun 349: 1278–1284 Olivari S, Molinari M (2007) Glycoprotein folding and the role of EDEM1, EDEM2 and EDEM3 in degradation of folding-defective glycoproteins. FEBS Lett 581: 3658–3664 Olson LJ, Zhang J, Dahms NM, Kim JJ (2002) Twists and turns of the cation-dependent mannose 6-phosphate receptor. Ligand-bound versus ligand-free receptor. J Biol Chem 277: 10156–10161 Ou WJ, Cameron PH, Thomas DY, Bergeron JJ (1993) Association of folding intermediates of glycoproteins with calnexin during protein maturation. Nature 364: 771–776

Luminal lectins

*

219

Parodi AJ (2000) Protein glucosylation and its role in protein folding. Annu Rev Biochem 69: 69–93 Puertollano R, Aguilar RC, Gorshkova I, Crouch RJ, Bonifacino JS (2001) Sorting of mannose 6-phosphate receptors mediated by the GGAs. Science 292: 1712–1716 Sakulsak N, Wakayama T, Hipkaeo W, Yamamoto M, Iseki S (2005) Cloning and characterization of a novel animal lectin expressed in the rat sublingual gland. J Histochem Cytochem 53: 1335–1343 Satoh T, Cowieson NP, Hakamata W, Ideo H, Fukushima K, Kurihara M, Kato R, Yamashita K, Wakatsuki S (2007) Structural basis for recognition of high mannose type glycoproteins by mammalian transport lectin VIP36. J Biol Chem 282: 28246–28255 Schweizer A, Fransen JA, Bachi T, Ginsel L, Hauri HP (1988) Identification, by a monoclonal antibody, of a 53-kD protein associated with a tubulo-vesicular compartment at the cis-side of the Golgi apparatus. J Cell Biol 107: 1643–1653 Varki A (1993) Biological roles of oligosaccharides: all of the theories are correct. Glycobiology 3: 97–130 Wada I, Rindress D, Cameron PH, Ou WJ, Doherty JJ II, Louvard D, Bell AW, Dignard D, Thomas DY, Bergeron JJ (1991) SSR alpha and associated calnexin are major calcium binding proteins of the endoplasmic reticulum membrane. J Biol Chem 266: 19599–19610 Whyte JR, Munro S (2001) A yeast homolog of the mammalian mannose 6-phosphate receptors contributes to the sorting of vacuolar hydrolases. Curr Biol 11, 1074–1078 Williams DB (2006) Beyond lectins: the calnexin/calreticulin chaperone system of the endoplasmic reticulum. J Cell Sci 119: 615–623 Yerushalmi N, Keppler-Hafkemeyer A, Vasmatzis G, Liu XF, Olsson P, Bera TK, Duray P, Lee B, Pastan I (2001) ERGL, a novel gene related to ERGIC-53 that is highly expressed in normal and neoplastic prostate and several other tissues. Gene 265: 55–60 Zhang B, Cunningham MA, Nichols WC, Bernat JA, Seligsohn U, Pipe SW, McVey JH, Schulte-Overberg U, De Bosch NB, Ruiz-Saez A, White GC, Tuddenham EG, Kaufman RJ, Ginsburg D (2003) Bleeding due to disruption of a cargo-specific ER-to-Golgi transport complex. Nat Genet 34: 220–225 Zhu Y, Doray B, Poussu A, Lehto VP, Kornfeld S (2001) Binding of GGA2 to the lysosomal enzyme sorting motif of the mannose 6-phosphate receptor. Science 292: 1716–1718 Zuber C, Cormier JH, Guhl B, Santimaria R, Hebert DN, Roth J (2007) EDEM1 reveals a quality control vesicular transport pathway out of the endoplasmic reticulum not involving the COPII exit sites. Proc Natl Acad Sci USA 104: 4407–4412

Supplementary references Protein folding and quality control in the ER Denzel A, Molinari M, Trigueros C, Martin JE, Velmurgan S, Brown S, Stamp G, Owen MJ (2002) Early postnatal death and motor disorders in mice congenitally deficient in calnexin expression. Mol Cell Biol 22: 7398–7404 Ellgaard L, Riek R, Herrmann T, Guntert P, Braun D, Helenius A, Wuthrich K (2001) NMR structure of the calreticulin P-domain. Proc Natl Acad Sci USA 98: 3133–3138 Hammond C, Braakman I, Helenius A (1994) Role of N-linked oligosaccharide recognition, glucose trimming, and calnexin in glycoprotein folding and quality control. Proc Natl Acad Sci USA 91: 913–917 Jackson MR, Cohen-Doyle MF, Peterson PA, Williams DB (1994) Regulation of MHC class I transport by the molecular chaperone, calnexin (p88, IP90). Science 263: 384–387 Michalak M, Lynch J, Groenendyk J, Guo L, Robert Parker JM, Opas M (2002) Calreticulin in cardiac development and pathology. Biochim Biophys Acta 1600: 32–37

220

*

B. Nyfeler et al.

Moore SE, Spiro RG (1993) Inhibition of glucose trimming by castanospermine results in rapid degradation of unassembled major histocompatibility complex class I molecules. J Biol Chem 268: 3809–3812 Schrag JD, Bergeron JJ, Li Y, Borisova S, Hahn M, Thomas DY, Cygler M (2001) The Structure of calnexin, an ER chaperone involved in quality control of protein folding. Mol Cell 8: 633–644 Solda T, Galli C, Kaufman RJ, Molinari M (2007) Substrate-specific requirements for UGT1-dependent release from calnexin. Mol Cell 27: 238–249 Vassilakos A, Michalak M, Lehrman MA, Williams DB (1998) Oligosaccharide binding characteristics of the molecular chaperones calnexin and calreticulin. Biochemistry 37: 3480–3490

ER-associated protein degradation Buschhorn BA, Kostova Z, Medicherla B, Wolf DH (2004) A genome-wide screen identifies Yos9p as essential for ER-associated degradation of glycoproteins. FEBS Lett 577: 422–426 Eriksson KK, Vago R, Calanca V, Galli C, Paganetti P, Molinari M (2004) EDEM contributes to maintenance of protein folding efficiency and secretory capacity. J Biol Chem 279: 44600–44605 Hirao K, Natsuka Y, Tamura T, Wada I, Morito D, Natsuka S, Romero P, Sleno B, Tremblay LO, Herscovics A, Nagata K, Hosokawa N (2006) EDEM3, a soluble EDEM homolog, enhances glycoprotein endoplasmic reticulum-associated degradation and mannose trimming. J Biol Chem 281: 9650–9658 Hosokawa N, Tremblay LO, You Z, Herscovics A, Wada I, Nagata K (2003) Enhancement of endoplasmic reticulum (ER) degradation of misfolded Null Hong Kong alpha1antitrypsin by human ER mannosidase I. J Biol Chem 278: 26287–26294 Hosokawa N, Wada I, Hasegawa K, Yorihuzi T, Tremblay LO, Herscovics A, Nagata K (2001) A novel ER alpha-mannosidase-like protein accelerates ER-associated degradation. EMBO Rep 2: 415–422 Mast SW, Diekman K, Karaveg K, Davis A, Sifers RN, Moremen KW (2005) Human EDEM2, a novel homolog of family 47 glycosidases, is involved in ER-associated degradation of glycoproteins. Glycobiology 15: 421–436 Molinari M, Calanca V, Galli C, Lucca P, Paganetti P (2003) Role of EDEM in the release of misfolded glycoproteins from the calnexin cycle. Science 299: 1397–1400 Oda Y, Hosokawa N, Wada I, Nagata K (2003) EDEM as an acceptor of terminally misfolded glycoproteins released from calnexin. Science 299: 1394–1397 Olivari S, Galli C, Alanen H, Ruddock L, Molinari M (2005) A novel stress-induced EDEM variant regulating endoplasmic reticulum-associated glycoprotein degradation. J Biol Chem 280: 2424–2428 Szathmary R, Bielmann R, Nita-Lazar M, Burda P, Jakob CA (2005) Yos9 protein is essential for degradation of misfolded glycoproteins and may function as lectin in ERAD. Mol Cell 19: 765–775

Protein transport Anelli T, Ceppi S, Bergamelli L, Cortini M, Masciarelli S, Valetti C, Sitia R (2007) Sequential steps and checkpoints in the early exocytic compartment during secretory IgM biogenesis. EMBO J 26: 4177–4188 Appenzeller-Herzog C, Roche AC, Nufer O, Hauri HP (2004) pH-induced conversion of the transport lectin ERGIC-53 triggers glycoprotein release. J Biol Chem 279: 12943–12950

Luminal lectins

*

221

Dahm T, White J, Grill S, Fullekrug J, Stelzer EH (2001) Quantitative ER-Golgi transport kinetics and protein separation upon Golgi exit revealed by vesicular integral membrane protein 36 dynamics in live cells. Mol Biol Cell 12: 1481–1498 Hara-Kuge S, Ohkura T, Ideo H, Shimada O, Atsumi S, Yamashita K (2002) Involvement of VIP36 in intracellular transport and secretion of glycoproteins in polarized MadinDarby canine kidney (MDCK) cells. J Biol Chem 277: 16332–16339 Itin C, Roche AC, Monsigny M, Hauri HP (1996) ERGIC-53 is a functional mannoseselective and calcium-dependent human homologue of leguminous lectins. Mol Biol Cell 7: 483–493 Kappeler F, Klopfenstein DR, Foguet M, Paccaud JP, Hauri HP (1997) The recycling of ERGIC-53 in the early secretory pathway. ERGIC-53 carries a cytosolic endoplasmic reticulum-exit determinant interacting with COPII. J Biol Chem 272: 31801–31808 Klumperman J, Schweizer A, Clausen H, Tang BL, Hong W, Oorschot V, Hauri HP (1998) The recycling pathway of protein ERGIC-53 and dynamics of the ER-Golgi intermediate compartment. J Cell Sci 111(Pt 22): 3411–3425 Nichols WC, Seligsohn U, Zivelin A, Terry VH, Hertel CE, Wheatley MA, Moussalli MJ, Hauri HP, Ciavarella N, Kaufman RJ, Ginsburg D (1998) Mutations in the ER-Golgi intermediate compartment protein ERGIC-53 cause combined deficiency of coagulation factors V and VIII. Cell 93: 61–70 Nufer O, Kappeler F, Guldbrandsen S, Hauri HP (2003) ER export of ERGIC-53 is controlled by cooperation of targeting determinants in all three of its domains. J Cell Sci 116: 4429–4440 Nyfeler B, Michnick SW, Hauri HP (2005) Capturing protein interactions in the secretory pathway of living cells. Proc Natl Acad Sci USA 102: 6350–6355 Nyfeler B, Zhang B, Ginsburg D, Kaufman RJ, Hauri HP (2006) Cargo selectivity of the ERGIC-53/MCFD2 transport receptor complex. Traffic 7: 1473–1481 Teasdale RD, Jackson MR (1996) Signal-mediated sorting of membrane proteins between the endoplasmic reticulum and the Golgi apparatus. Annu Rev Cell Dev Biol 12: 27–54 Velloso LM, Svensson K, Pettersson RF, Lindqvist Y (2003) The crystal structure of the carbohydrate-recognition domain of the glycoprotein sorting receptor p58/ERGIC53 reveals an unpredicted metal-binding site and conformational changes associated with calcium ion binding. J Mol Biol 334: 845–851 Vollenweider F, Kappeler F, Itin C, Hauri HP (1998) Mistargeting of the lectin ERGIC-53 to the endoplasmic reticulum of HeLa cells impairs the secretion of a lysosomal enzyme. J Cell Biol 142: 377–389 Wendeler MW, Paccaud JP, Hauri HP (2007) Role of Sec24 isoforms in selective export of membrane proteins from the endoplasmic reticulum. EMBO Rep 8: 258–264 Yamaguchi D, Kawasaki N, Matsuo I, Totani K, Tozawa H, Matsumoto N, Ito Y, Yamamoto K (2007) VIPL has sugar-binding activity specific for high-mannose-type N-glycans, and glucosylation of the {alpha}1,2 mannotriosyl branch blocks its binding. Glycobiology 17: 1061–1069

Lysosomal protein sorting in the Golgi Breuer P, Korner C, Boker C, Herzog A, Pohlmann R, Braulke T (1997) Serine phosphorylation site of the 46-kDa mannose 6-phosphate receptor is required for transport to the plasma membrane in Madin-Darby canine kidney and mouse fibroblast cells. Mol Biol Cell 8: 567–576 Chao HH, Waheed A, Pohlmann R, Hille A, Von Figura K (1990) Mannose 6-phosphate receptor dependent secretion of lysosomal enzymes. EMBO J 9: 3507–3513 Dahms NM, Hancock MK (2002) P-type lectins. Biochim Biophys Acta 1572: 317–340

222

*

B. Nyfeler et al.

Dahms NM, Rose PA, Molkentin JD, Zhang Y, Brzycki MA (1993) The bovine mannose 6phosphate/insulin-like growth factor II receptor. The role of arginine residues in mannose 6-phosphate binding. J Biol Chem 268: 5457–5463 Diaz E, Pfeffer SR (1998) TIP47: a cargo selection device for mannose 6-phosphate receptor trafficking. Cell 93: 433–443 Duncan JR, Kornfeld S (1988) Intracellular movement of two mannose 6-phosphate receptors: return to the Golgi apparatus. J Cell Biol 106: 617–628 Koster A, Von Figura K, Pohlmann R (1994) Mistargeting of lysosomal enzymes in M(r) 46,000 mannose 6-phosphate receptor-deficient mice is compensated by carbohydrate-specific endocytotic receptors. Eur J Biochem 224: 685–689 Meresse S, Hoflack B (1993) Phosphorylation of the cation-independent mannose 6phosphate receptor is closely associated with its exit from the trans-Golgi network. J Cell Biol 120: 67–75 Morgan DO, Edman JC, Standring DN, Fried VA, Smith MC, Roth RA, Rutter WJ (1987) Insulin-like growth factor II receptor as a multifunctional binding protein. Nature 329: 301–307 Olson LJ, Zhang J, Lee YC, Dahms NM, Kim JJ (1999) Structural basis for recognition of phosphorylated high mannose oligosaccharides by the cation-dependent mannose 6-phosphate receptor. J Biol Chem 274: 29889–29896 Reczek D, Schwake M, Schroder J, Hughes H, Blanz J, Jin X, Brondyk W, Van Patten S, Edmunds T, Saftig P (2007) LIMP-2 is a receptor for lysosomal mannose-6-phosphate-independent targeting of beta-glucocerebrosidase. Cell 131: 770–783 Riederer MA, Soldati T, Shapiro AD, Lin J, Pfeffer SR (1994) Lysosome biogenesis requires Rab9 function and receptor recycling from endosomes to the trans-Golgi network. J Cell Biol 125: 573–582 Roberts DL, Weix DJ, Dahms NM, Kim JJ (1998) Molecular basis of lysosomal enzyme recognition: three-dimensional structure of the cation-dependent mannose 6phosphate receptor. Cell 93: 639–648 Rohrer J, Kornfeld R (2001) Lysosomal hydrolase mannose 6-phosphate uncovering enzyme resides in the trans-Golgi network. Mol Biol Cell 12: 1623–1631 Sahagian GG, Distler J, Jourdian GW (1981) Characterization of a membrane-associated receptor from bovine liver that binds phosphomannosyl residues of bovine testicular beta-galactosidase. Proc Natl Acad Sci USA 78: 4289–4293 Schweizer A, Kornfeld S, Rohrer J (1996) Cysteine34 of the cytoplasmic tail of the cationdependent mannose 6-phosphate receptor is reversibly palmitoylated and required for normal trafficking and lysosomal enzyme sorting. J Cell Biol 132: 577–584 Schweizer A, Kornfeld S, Rohrer J (1997) Proper sorting of the cation-dependent mannose 6-phosphate receptor in endosomes depends on a pair of aromatic amino acids in its cytoplasmic tail. Proc Natl Acad Sci USA 94: 14471–14476 Steet R, Lee WS, Kornfeld S (2005) Identification of the minimal lysosomal enzyme recognition domain in cathepsin D. J Biol Chem 280: 33318–33323 Tong PY, Tollefsen SE, Kornfeld S (1988) The cation-independent mannose 6-phosphate receptor binds insulin-like growth factor II. J Biol Chem 263: 2585–2588 Wan L, Molloy SS, Thomas L, Liu G, Xiang Y, Rybak SL, Thomas G (1998) PACS-1 defines a novel gene family of cytosolic sorting proteins required for trans-Golgi network localization. Cell 94: 205–216

Golgi ribbon and the function of the golgins

*

223

The Golgi ribbon and the function of the golgins Maria A. De Matteis, Alexander A. Mironov and Galina V. Beznoussenko

Introduction The Golgi apparatus (GA) is present in different organisms in very different forms. When visualized by immunofluorescence in most mammalian cells, the Golgi ribbon appears as a lacy structure that occupies a volume of 5–7 mm in length, 1–2 mm in width, and 3–5 mm in depth (Storrie and Kreis 1996), and that surrounds the centrosome (or microtubule-organizing centre; MTOC) (see Chapter 2.14). The positions of the MTOC and the GA depend on cell polarity. In many polarized epithelial cells, the centrosome is positioned near the apical portion of the cell surface (Ojakian et al. 1997), where the GA also resides. From dozens to hundreds of Golgi stacks that act as a single organelle are linked together to form an interconnected, ribbon-like structure in the perinuclear area (Hidalgo Carcedo et al. 2004; Polishchuk and Mironov 2004; Mogelsvang et al. 2004). Completely isolated stacks are rare (Cole et al. 1996a,b). The connectivity between individual stacks is appreciable in living cells, where photobleaching of a fraction of the GA containing chimeras between resident Golgi proteins (i.e. enzymes) and green fluorescent protein (GFP) quickly induces a disappearance/decrease of fluorescence from other parts of the GA, suggesting rapid diffusional exchange of the Golgi enzymes (such as ManII and GalT, but not ManI, Marra et al. 2001) between the Golgi stacks (Cole et al. 1996a,b). Continuity of the stacks within the GA has been confirmed using scanning electron microscopy (SEM). The SEM complementary observation of the fractured samples of the GA has demonstrated that while almost all of the adjacent stacks appear to be separated from one another, they are not actually separated, but remain continuous with each other (Inoue 1992). The Golgi ribbon is, however, just one of the possible shapes of this organelle, as it is found in different organisms in very different forms: from scattered tubular networks (Saccharomyces cerevisiae) and isolated multiple stacks of cisternae (Pichia pastoris, Drosophila, plants), to the above-described continuous pericentrosomal ribbon of many mammalian cells. The overall structural organization of the GA can vary also within the same species between different cell types: in mammals, for instance, it is present as isolated stacks in oocytes and skeletal muscle cells, while it forms a continuous ribbon in the majority of other cell types. Finally, the Golgi ribbon can change its organization in the very same cell through the cell cycle (since it undergoes

224

*

M. A. De Matteis et al.

cycles of fragmentation and reformation during the mitotic process) and according to the state of activity of an organelle. Despite its very “conserved” functions, the reasons for such a variegated architecture of the GA in different organisms, or even in different cells of the same organism, are not yet really understood. A number of possibilities have been put forward, with the one that is more frequently considered being related to the GA and cell “motility”: directing the movement of a continuous organelle (such as the Golgi ribbon) within the cell would be easier and more effective than coordinately moving multiple unlinked organelles (isolated Golgi stacks). This appears to be required during cell migration, when the GA is reoriented towards the leading edge of the cell, to provide membranes for the advancing cell front. However, two levels of consideration challenge this assumption: the isolated Golgi stacks in plants are highly motile along actin filaments (see Chapter 4.1), and cells deprived of a central GA migrate apparently normally (Kondylis and Rabouille 2003). Perhaps a more likely explanation is that the integrity of the Golgi ribbon may represent a sort of “signal”, e.g. for a Golgibased mitotic checkpoint (Colanzi et al. 2003; Chapter 3.15). Finally, one should consider the “easiest” explanation, that the ribbon represents just one of the possible configurations of the organelle and it is linked to its functional state (i.e. transport activity, see below). In spite of our uncertainties regarding the reasoning behind a continuous organelle instead of isolated stacks or cisternae, we know a lot about the molecular mechanisms that are responsible for the building up and maintenance of the Golgi ribbon: these are the focus of this Chapter.

The role of microtubules A common feature of the cells that do not have a continuous Golgi ribbon is the absence of the radial organization of MTs around the MTOC. This feature is seen for mammalian oocytes and myotubes, which do not have a Golgi ribbon (Polishchuk et al. 1999; Trucco et al. 2004), in insect cells (Kondylis and Rabouille 2003), and in plants, yeast, and some protists. Protist cells that have a MTOC usually also have their GA organized into a ribbon, reinforcing the correlation that exists between the existence of the MTOC and the presence of a Golgi ribbon. Thus, a first prerequisite for the formation of the Golgi ribbon is the presence of a radial MT array centred on the MTOC. The second requirement for the formation of the Golgi ribbon is the presence of MTs per se. Depolymerization of MTs with specific agents (e.g. nocodazole) (Cole et al. 1996a,b; Thyberg and Moskalewski 1999; Polishchuk et al. 1999; Trucco et al. 2004) induces the fragmentation of the Golgi ribbon into many peripheral pieces. These Golgi fragments have a typical stacked organization (Polishchuk et al. 1999). In favourable EM sections, the Golgi ministacks formed after depletion of cells from MTs can be seen in close association with endoplasmic reticulum (ER) exit sites (Polishchuk et al. 1999; Storrie et al. 1998).

Golgi ribbon and the function of the golgins

*

225

Although the mechanisms underlying this Golgi fragmentation are beyond the scope of this Chapter (see some details in Chapter 2.14), mention should however be made that Golgi fragmentation occurs with the participation of the diffusion of Golgi enzymes through the ER (Cole et al. 1996a,b; Storrie et al. 1998), with this mechanism having been questioned more recently (Pecot and Malhotra 2006). Here it is important to note that these Golgi ministacks can reassemble into a continuous ribbon upon removal of the MT-disrupting drug and restoration of the MTOC and MT organization (Ho et al. 1989). The third requirement is that MTs maintain their dynamic instability. Indeed, taxol, which stabilizes MTs and induces the polymerization of MTs not connected with the MTOC and of MT bundles localized in the cell periphery (Wehland et al. 1983), also induces the fragmentation of the Golgi ribbon. Even when MT dynamics are lowered by addition of very low concentrations of nocodazole or of taxol, the Golgi ribbon undergoes fragmentation (Minin 1997). Finally, connection between the Golgi ribbon and MT organization emerges also from the observation that the Golgi ribbon is severed into isolated stacks during the cell cycle, in G2, when reorganization of the MT star could contribute to this peripheral Golgi fragmentation (Thyberg and Moskalewski 1999). G2-blocked cells do not show major differences either in the number of cisternae that comprise a single stack, or in the average diameter of the stacks; however, when compared with control cells, they show stacks that are isolated (i.e. not interconnected by membrane tubules) in most cases and are not longitudinally aligned. The stacks in G2 cells are either isolated or connected in small groups of two to four stacks; this Golgi fragmentation in G2 is mirrored by the much lower recovery rate of GalT–GFP in G2 cells compared to interphase cells in FRAP experiments (Hidaldo Carcedo et al. 2004). The requirement for MTs and the MTOC underlies the involvement of MT-based motors, and in particular of dynein, in building up and maintaining the Golgi ribbon. Dynein is a multisubunit minus-end-directed microtubule motor that is known to transport a variety of cargoes in animal cells (Vallee et al. 2004; Vale 2003). Cytoplasmic dynein cycles constitutively between the ER and the GA. It co-localizes partially with the intermediate compartment (Roghi and Allan 1999). Dynein function requires a series of “accessory proteins”, those that activate dynein, such as dynactin, and those involved in the recruitment of dynein–dynactin to membranes, such as Bicaudal-D (Matanis et al. 2002), Rab6 (Short et al. 2002), beta III spectrin (Holleran et al. 2001), and CLIPR-59 (Perez et al. 2002). Dynactin is concentrated at MT plus ends and can transiently capture Golgi membranes that are then transported towards the minus ends by dynein (Vaughan et al. 2002). These mechanisms are discussed in more detail in Chapter 2.14. Several studies have suggested that dynein can associate with the mammalian GA through Golgi-associated spectrin and dynactin, a protein complex

226

*

M. A. De Matteis et al.

that is required for the movement of dynein cargo (Kumar et al. 2004; Schroer sy-Theulaz et al. 1992; Harada et al. 1998). 2004; Burkhardt et al. 1997; Corthe Cells lacking dynein cannot concentrate organelles in the perinuclear region, such as the GA and lysosomes (Narada et al. 1998), while the over-expression of p50/dynamitin results in the dissociation of dynein heavy chain from the membrane of peripheral Golgi elements (Roghi and Allan 1999) and induces the dispersal of the GA from the perinuclear region (Burkhardt et al. 1997). The situation is different in polarized epithelial cells, where MTs are primarily oriented with their plus ends basally, near the GA, and their minus ends in the apical cytoplasm. Here, the distribution of MT motors is also different and a selected kinesin isoform (KIFC3) has been shown to have roles in MT-dependent centralization and positioning of the GA in some polarized epithelial cells (Xu et al. 2002). The balance between the activity of the minus-end motor dynein, and the plus-end motor kinesin will determine the final positioning of the GA. However, the directionality of membrane trafficking and cycling between the ER and the GA is unlikely to be achieved through the control of motor–membrane interactions; rather, the motors, and especially dynein, probably remain bound throughout the whole anterograde–retrograde cycle, with their activity (and not their membrane association) being modulated accordingly (Fritzler et al. 1993). This is suggested also by the finding that dynein heavy chain is found on brefeldin-A-induced tubules that are moving retrogradely towards microtubule plus ends (Roghi and Allan 1999). Although the Golgi ribbon is centred on the MTOC, it does not contact it, with most of the Golgi elements situated at a distance of 1–3 mm from the centrosome. The reason for this is not clear, although it is likely that the MTs growing from the centrosome at a high density “push” the Golgi membranes out, forcing them to stay at some distance from the cell centre (Polishchuk and Mironov 2004). Kinesin, the plus-end MT-based motor that serves for centrifugal movement of post-Golgi carriers and the ER might have a role in maintaining this MTOC–Golgi distance, as its inactivation results in the collapse of the Golgi around the centrosome. In cells without kinesin, the ER is also located near to the centrosome (Feiguin et al. 1994). Another mechanism preventing overlap of the GA with the MTOC might involve GMAP-210, a protein involved in the interaction of Golgi membranes with MTs (Chabin-Brion et al. 2001). GMAP-210 captures short MT seeds that are formed at the centrosome by their minus ends, and together with the Golgilocalized pool of the CLASPs that attach to and stabilize the MT plus ends, GMAP-210 generates a meshwork of short MTs that are associated with adjacent Golgi stacks and links them up to form a ribbon (Rios et al. 2004). Another possible mechanism might involve the actin/myosin machinery: inducing actin depolymerization induces a “collapse” of the GA on the MTOC (Valderrama et al. 1998). The reason for this is also not completely clear (see also Chapter 2.14). Isolated Golgi membranes from intestinal epithelial cells are enriched in myosin-I, dynein, and its in vitro motility activator dynactin

Golgi ribbon and the function of the golgins

*

227

(p150/Glued). Myosin-I is present on all membranes in the Golgi fraction; dynein is present only on a small membrane fraction (Fath et al. 1994).

The role of membrane input from the endoplasmic reticulum The presence of an intact MT system is not sufficient, per se, to guarantee the formation of the Golgi ribbon: the continuous input of ER-derived membranes moving centripetally along MTs towards the Golgi area is also a primary requirement, as it represents the source of the membranes that promote the formation of connections between stacks. This absolute requirement for membrane input is testified by the observation that the Golgi ribbon disconnects into isolated stacks as soon as the arrival of membranes from the ER is slowed down or interrupted (Marra et al. 2007). Indeed, Rambourg et al. (1993) also showed that the Golgi ribbon of prolactin cells is fragmented when the function of the GA is blocked (compare figs. 1 and 7 in Rambourg et al. 1993). The ribbon configuration of the GA in mammalian cells thus reflects an “active” state of the organelle, as it receives and absorbs membranous carriers from the ER (the ER-to-Golgi carriers; EGCs) (Marra et al. 2007). For ER-derived membranes to sustain the formation of the Golgi ribbon, they need to be completely integrated into the cisterna stacks; this implies that the highly pleiomorphic membranes coming out from the ER (i.e. the EGCs) need to be re-shaped, and transformed into flat discs (i.e. the Golgi cisternae), and integrated into the Golgi ribbon. Indeed, following their generation at the ER, EGCs undergo a series of modification steps with regard to their molecular composition, ultrastructure and dynamics (Marra et al. 2001), which prime them for entering the GA. Among these, a key step is the acquisition of a class of Golgi-associated proteins that are collectively known as the Golgi tethering factors, the golgins, due to their property of joining membranes together.

The golgins The golgins were originally identified as a family of Golgi-localized autoantigens using antibodies derived from the sera of patients with a variety of autoimmune disorders (Chan and Fritzler 1998). Most of these golgins were then cloned by screening of expression libraries with these autoantibodies, such as Golgin-160 (Fritzler et al. 1993), giantin (Seelig et al. 1994), GMAP-210 (Rios et al. 1994), Golgin-245/p230 (Fritzler et al. 1995), Golgin-97 (Griffith et al. 1997) and Golgin-67 (Eystathioy et al. 2000). Among the different golgins, GM130, giantin, and p115 have been the most studied to date. Most of the golgins have extensive coiled-coil regions throughout their entire polypeptide chain, a common protein motif that is known to form an extended rod-like structure (Kjer-Nielsen et al. 1999; Burkhard et al. 2001). Coiled-coil proteins have two a-helices that wrap around each other with a slight left-handed superhelical twist, forming the rod-like structure

228

*

M. A. De Matteis et al.

(Gillingham et al. 2003). These proteins have elongated shapes that can reach several dozens of nm in length. Most of the golgins are peripheral membrane proteins and they often have short non-coiled-coil regions at either end that mediate their targeting and other interactions. Thus, the C-terminus of GM130 binds to GRASP65, a lipid-anchored protein on the cis-Golgi network (Barr et al. 1998). Similarly, four of the golgins share a C-terminal GRIP domain that is sufficient to target them to the trans-side of the GA (see Chapter 3.3). In contrast, other golgins, such as giantin, Golgin-84 and CASP, are anchored in the bilayer via a C-terminal transmembrane domain (Fridmann-Sirkis et al. 2004; see below). Although the golgins from different species share some structural features, their sequences are not well conserved (Renna et al. 2005). In addition, the golgins are not uniformly distributed among all species. Some organisms and cells contain significant levels of the golgins, while others lack most of them. Indeed, only a few of the golgins are present in yeast: Coy1p (similar to CASP), and Grp1p (similar to Golgin-160/GCP170) (Barr and Short 2003; Gillingham and Munro 2003), and the single GRIP protein, Imh1p (Siniossoglou et al. 2000; Tsukada et al. 1999; Short et al. 2005; Luke et al. 2003). Apart from Golgi localization and the presence of coiled-coil motifs, the only other common feature so far identified for the golgin family is that many members interact with small GTPases. Depending on their localization (Table 1), the golgins can be divided into two subfamilies: the cis- and trans-golgins. The cis-golgins include p115, GM130/Golgin-95, GRASP65, GRASP55, CASP, Coy1p (similar to CASP), Golgin-45, Golgin-67 and Golgin-84. Some of these were previously considered to be tethering factors. The trans-golgins include Golgin-97, GCC88, Golgin-160/ MEA-2/GCP170, Golgin-245/p230/tGolgin-1, GCC185, GMAP-210, and Grp1p (similar to Golgin-160/GCP170) in yeast (Barr and Short 2003; Gillingham and Munro 2003), and the related group of proteins – possibly splice variants – of GCP372 and GCP364 (Barr 1999). Giantin/macrogolgin is not restricted to either the cis or the trans pole of the GA. The golgins are recruited to Golgi membranes in several ways. Some of them are targeted by the Rabs (p115 by Rab1; Bicaudal-D1 and Bicaudal-D2 by Rab6) in a nucleotide-dependent manner. The GRASP proteins themselves are targeted via an N-terminal myristoyl group. GM130 associates with the membrane through an interaction with the GRASPs. Golgin-245, Golgin97, GCC88, and GCC185 are targeted to the trans-Golgi network (TGN) membranes by their C-terminal GRIP domain, in a G-protein-dependent process involving Arl1. In turn, the Rabs are targeted to membranes via Cterminal geranylgeranyl modifications, whereas most of the Arls are targeted to membranes by N-terminal myristoylation (Short et al. 2002; Matanis et al. 2002). Giantin, Golgin-84, and CASP contain long N-terminal coiled-coil regions that protrude into the cytosol and C-terminal transmembrane domains with aminoacid residues that are highly conserved across species. Their transmembrane domains also show a high degree of sequence similarity.

Cytosolic; rod like; 54 nm in length; coiled-coil; homodimer GM130; giantin, syn5, sly1, membrin, Ykt6, Rab1 (GTP), GOS28, ARFGEF; IRAP; No binding to MTs VTC, late IC; CGN, ciscisterna

Structure

Unknown To the IC, then ER

To the IC (?)

Role in mitotic Golgi fragmentation Redistribution under the action of BFA

Attachment of the CGN to the medial Golgi Through phosphorylation

Role in stacking

Small

IC-Golgi; intra-Golgi in vitro; ManI (but not ManII) ! IC at 15 C in Lec1

Role in transport

By cytoplasmic domain of about 100 residues adjacent to the TMD Small

By Rab1 and other binding partners

Non-compact zones

C-terminus of p115 by N-terminus of giantin

Type II protein; rod like; coiled-coil; 400 kDa

Giantin

Targeting

Localization

Main binding partners

p115

Characteristics

Table 1. Characteristics of golgins

Minimal; W/o GM130 – BFA resistance; in Drosophila w/o GRASP65 no inhibition Attachment of the CGN to the medial GA Through phosphorylation To the IC

CGN and cisternal rims of the 2nd and the 3rd cisternae; co-localization with cargo By GRASP and other binding partners

C-terminus of p115 by N-terminus of GM130; Rab1 (GTP); GRASP65

Cytosolic; rod like; coiled-coil

GM130

To the IC

(Continued)

Small; deletion of GM130 does not induce transport inhibition in Drosophila cell. Attachment of the CGN to the medial GA Significant

By myristoylation

IC/CGN

GM130. p24

Cytosolic with myristoylation

GRASP65

Golgi ribbon and the function of the golgins *

229

Role in mitosis Redistribution under the action of BFA Result of protein depletion or inhibition Result of protein over expression

Role in transport Role in stacking

Localization Targeting

IC/CGN By TMD

Unknown

Type II protein; rod like; coiled-coil

Unknown

Unknown

Cytosolic with myristoylation GM130; p24; Golgin-45; TGF a Cis; VTS; rims; tubules By myristoylation

Structure

CASP

Unknown

Central Golgi fragmentation Unknown

GRASP55

Characteristics

Unknown

Central Golgi fragmentation

Unknown Regulation of formation of the attached CGN Unknown To the ER

Peripheral Golgi fragmentation

Result of protein over expression

Central Golgi fragmentation

GM130

Unknown Attachment of the CGN to the medial GA Significant Unknown

Peripheral Golgi fragmentation

Result of protein depletion or inhibition

Giantin

Peripheral Golgi fragmentation

Unknown IC

CGN Cytoplasmic domain adjacent to the TMD Minimal Attachment of the CGN to medial Golgi

Rod; coiled-coil; type II protein (N-cytosolic) Rab1 (GTP), not matrix proteins

Golgin-84

Central Golgi fragmentation with augmentation of the cisterna length and decrease of number of cisternae Unknown

GRASP65

*

Main binding partners

p115

Characteristics

Table 1. (Continued) 230 M. A. De Matteis et al.

Rab2. GRASP55 CGN Unknown

ERES Attachment of the CGN to medial Golgi Unknown Unknown Redistribution to the ER (?)

TGN (?) Unknown

Unknown Unknown

Unknown Unknown

Unknown

Main binding partners Localization Targeting

Role in transport Role in stacking

Role in mitosis Redistribution by BFA Results of inhibition of function

Rod; coiled-coil; cytosolic

Rod; Coiled-coil; cytosolic; similar to GM130; TMD unclear Unknown

Structure

Golgin-45

Golgin-67

Characteristics

ES-TGN Attachment of the TGN to the medial Golgi Unknown Cytosolic and endosomal Central fragmentation; increased cisternal length

Trans/TGN Homodimers bind ARL1

ARL1

Rod; coiled-coil; cytosolic

Golgin-97

Peripheral Golgi fragmentation

Unknown Cytosolic

Trans/TGN Homodimers bind ARL1; MACF-1; no binding to dynactin/ dynein ES-TGN Unknown

ARL1

Rod; coiled-coil; cytosolic

Golgin-245

Unknown

Unknown Cytosolic (?)

Unknown Unknown

TGN Rab6 (?)

Rab6. Dynactin

Rod; coiled-coil; cytosolic

Bicaudal-D1/D2

Golgi ribbon and the function of the golgins *

231

232

*

M. A. De Matteis et al.

In particular, key tyrosine and histidine residues are completely conserved. When such a conserved tyrosine residue is mutated to leucine in CASP, this protein no longer localizes to the GA, and accumulates in the ER instead (Gillingham et al. 2002). Both the cis- and trans-golgins are required to maintain the Golgi ribbon. Indeed, inhibition of the function of p115 (by different means) induces Golgi fragmentation into peripheral stacked cisternae and vesicle-like structures (Schroer et al. 2004). Both the over-expression and depletion of Golgin-84 results in the peripheral fragmentation of the Golgi ribbon (Diao et al. 2003), while inhibition of GM130/GRASP65 function in any way induces the fragmentation of the central GA (Puthenveedu et al. 2006; Marra et al. 2007). HeLa cells depleted of GRASP55 show a fragmented GA (Feinstein and Linstedt 2008). Similarly, microinjection of anti-giantin antibodies induces the fragmentation of the central Golgi ribbon (Beznoussenko et al. submitted), and depletion of TMF/ARA160 by RNA interference (RNAi) in NRK cells results in a modest dispersal of the Golgi membranes (Fridmann-Sirkis et al. 2004). RNAi of Golgin-97 or microinjection of anti-Golgin-97 antibodies also induce the fragmentation of the Golgi ribbon, although the Golgi fragments are larger than those induced by the block of GM130, GRASP65, and giantin (Beznoussenko et al. submitted). Golgin-245 depletion leads to replacement of a centralized, ribbon-like pattern with a tiny spotty pattern throughout the cell body (Yoshino et al. 2005). The peripheral fragmentation of the GA induced by deletion of Golgin-245 results in the dispersal of the GA to peripheral ministacks that are well preserved ultrastructurally (Yoshino et al. 2005). Thus, impairment of the function of most of the Golgins can induce either peripheral (p115, Golgin-84, Golgin-245) or central (GM130, giantin, GRASP65, Golgin-97) fragmentation of the Golgi ribbon (Fig. 1).

Other participants in the formation of the Golgi ribbon Other factors that are required for the maintenance of the Golgi ribbon include the different members of the p23-protein family (their over-expression leads to fragmentation of the Golgi ribbon; Rojo et al. 2000) and the retromer components (Seaman 2004). Depletion of the Golgi-associated conserved oligomeric complex also leads to fragmentation of the mammalian GA (Shestakova et al. 2006). Interfering with Golgi-associated enzymes can also induce Golgi fragmentation. This occurs with inhibition of PLA2 activity or expression (R. Polishchuk, personal communication) or upon PKD over-expression (Diaz Anel and Malhotra 2005). PLA2 has been shown to be required to generate and maintain tubular membranes at the GA (Brown et al. 2003), thus including also those that connect neighbouring stacks. For PKD, this is a component of one of the fission machineries that operates at the GA (Diaz Anel and

Golgi ribbon and the function of the golgins

*

233

Figure 1. GM130 depletion induces fragmentation of the Golgi ribbon. CHO (upper panel) and ldlG cells not expressing GM130 (lower panel) were analyzed by EM tomography. The 3D models of Golgi stacks were superimposed over the corresponding virtual EM images. The 3D models contain cis-most (red) and trans-most (magenta) cisternae. Vesicles (blue, 50–60 nm; white, 80–90 nm) are represented by software-generated spheres, centred on the centre of the vesicles. Models were made by A. A. Mironov, Jr and E. Fontana.

Malhotra 2005). In this respect, the other fissioning factor at the GA, the protein CTBP1/BARS (Weigert et al. 1999; Yang et al. 2005, 2006), would be different since it induces fragmentation of the GA only in the presence of mitotic cytosol (previously depleted of BARS) and not of interphase cytosol (Hidalgo Carcedo et al. 2004). Fragmentation of the GA with the redistribution of various Golgi markers, including Mann II and GM130, has been seen after inhibition of the Golgiresident GPI-anchored protein (GREG) (Xueyi et al. 2007). Expression of 23TMGREG, a fusion protein composed of GREG and the transmembrane domain of p23, also generates a scattered Golgi structure (Xueyi et al. 2007). Similarly, fragmentation of the GA is also observed in PIG-L CHO cells (Abrami et al. 2001) that are deficient in the biosynthesis of the GPI-anchor (Xueyi et al. 2007). In these PIG-L cells labelled with NAGTI-GFP, many circular or ring-like

234

*

M. A. De Matteis et al.

structures are seen, which are reminiscent of those induced by expression of GPI-GREG (Xueyi et al. 2007). RNAi of RINT-1, a protein linker between ZW10 and the ER soluble Nethylmaleimide-sensitive factor attachment protein receptor, syntaxin 18, induces Golgi fragmentation (Sun et al. 2007). Depletion of ZW10, a mitotic checkpoint protein implicated in Golgi/ER trafficking/transport, results in a central, disconnected cluster of Golgi elements and inhibition of ERGIC53 and Golgi enzyme recycling to the ER (Sun et al. 2007). Finally, there are several observations in favour of the necessity for the fusion machinery for the transformation of centrally located Golgi stacks into a ribbon-like structure. In cells microinjected with an SNAP mutant, we have seen significant inhibition of the centralization of EGCs (Mironov et al. 2003). However, this microinjection of the SNAP mutant also induces not only central Golgi fragmentation, but also strong vesiculation of Golgi stacks (Kweon et al. 2004). It seems that the over-expression of the SNARE GS15 lacking its transmembrane domain also induces central fragmentation of ManII-positive Golgi structures (i.e. see fig. 8n in Xu et al. 2002), while siRNA-mediated knockdown of GS15 transforms central GalT into a diffuse patterns that is similar to the ER (Xu et al. 2002). An antibody against the R-SNARE Ykt6 (see Chapter 2.1) induces the central fragmentation of the Golgi ribbon within the Golgi area (Zhang and Hong 2001).

The functional role of the Golgi ribbon What is the main purpose for such a complex array of molecular machineries for the building up and maintaining of a ribbon-like GA? Surprisingly, the basic functions of the GA, i.e. membrane trafficking and glycosylation, are not significantly different if they are carried out by a Golgi ribbon or by isolated Golgi stacks. Intra-Golgi transport does not depend on the ribbonlike organization of the GA (Trucco et al. 2004) and its stacked structure. The transport functions of the GA can continue at the same level if the ribbonlike GA is fragmented by depolymerization of MTs (Trucco et al. 2004). Thus, the presence of these “ministacks” in vivo (Trucco et al. 2004; Ward and Brandizzi 2004) and in cell-free assays (Pullikith and Wiedman 2002) allows the GA to perform all of its functions. Indeed, for glycosylation and protein sorting, even a single cisterna is sufficient (Varki 1998). Studies of glycosylation in the GA have shown that sugar nucleotide transporters and many glycosyltransferases are located within a single cisterna (Opat et al. 2001; Young 2004). This suggests that either a stack or even a single cisterna could be considered as the minimal Golgi nano-unit. Indeed, in insects, a double depletion of dGRASP and dGM130 leads to the quantitative conversion of Golgi stacks into clusters of vesicles and tubules, often featuring single cisternae; at the same time, the transport function is not severely impaired (Kondylis et al. 2005).

Golgi ribbon and the function of the golgins

*

235

What is then the reason for having a Golgi ribbon? In theory, the ribbon could facilitate the even distribution of Golgi enzymes among the stacks, and consequently a more efficient glycosylation of secretory proteins (Puthenveedu et al. 2006). However, when specifically tested on selective secretory proteins, no differences in glycosylation specifically attributable to the presence of a Golgi ribbon can be detected (Marra et al. 2007). As noted above, the presence of a Golgi ribbon might serve more specialized functions, such as the polarized delivery of membranes (that is indeed lost in nocodazole-treated cells). Consistent with this possibility, it has been recently shown that the fragmentation of the Golgi ribbon induced by overexpression of GRASP65 impairs polarized dendrite outgrowth in hippocampal neurons (Horton et al. 2005).

The role of the golgins as tethering factors The functions of the golgins should be in agreement with the models of intracellular transport. Within the framework of the vesicular model, the most obvious function of the golgins would be their role as the tethering factors. The presence of extensive coiled-coil regions in the golgins suggests that they can also adopt long rod-like structures. Therefore, in the framework of the vesicular model, a tethering factor is considered as tethering ER- or Golgi-derived vesicles to the GA (Allan et al. 2000; Moyer et al. 2001; Sonnichsen et al. 1998). Thus, the golgins are believed to be involved in the tethering of vesicles. In the framework of the vesicular model of intracellular transport, these proteins were considered as the factor helping docking of transport-coat-dependent vesicles acting before the SNAREs. Tethering is defined as a “formation of physical links, between two membranes that are due to fuse, before the engagement of the SNAREs” (Whyte and Munro 2002). This process might represent the earliest stage at which specificity is conferred on a fusion reaction and it may involve multiple interactions. Two broad classes of molecules are proposed to have roles in tethering: a group of coiled-coil proteins (GM130 and p115, see below), and several large, multisubunit complexes (such as the TRAPP and COG complexes; see Chapters 2.7 and 2.6) (Whyte and Munro 2002). These long rod-like molecules, the golgins, are thus attractive candidates as factors that link the Golgi cisternae or capture the transport vesicular carriers in the proximity of the cisternae, prior to fusion (Gillingham and Munro 2003). The function of some of the golgins has been interpreted as an interaction between the cis-golgins, such as p115, and GM130, which tethers an incoming COPI or COPII vesicle to the cis-Golgi membrane, and that leads to the docking and subsequent membrane fusion mediated by the SNAREs (Shorter et al. 2002). P115 and molecules like giantin, which cooperate with it in tethering at the GA, have an extended conformation, a property that might enable them to initially bridge relatively large distances between membranes (Lesa et al. 2000; Seemann et al. 2000a,b; Sonnichsen et al. 1998).

236

*

M. A. De Matteis et al.

Two additional models are presently available for the role of GM130: one in which GM130 mediates homotypic tethering of neighbouring cisternae (Puthenveedu et al. 2006), and one in which GM130 maintains the ribbon structure through its mediation of the incorporation of EGCs (Marra et al. 2007), promoting homotypic fusion of the cisternal rims (Puthenveedu et al. 2006). However, these hypotheses do not explain the function of the other golgins, and especially those associated with the trans-side of the GA. If the vesicular model of transport is wrong, there is the need to explain golgin function within the framework of other transport models. Due to the problems that we have in the explanation of experimental data based on the vesicular model as the main model of intracellular transport (see Chapters 1.2 and 3.2), we need to consider other possibilities and to find the function of the golgins in the framework of other models of intra-Golgi transport. There is here some evidence in favour of the cisterna maturation model. Indeed, COPI vesicles dependent on the golgin CASP contain more concentrated Golgi enzymes than other COPI-dependent vesicles (Malsam et al. 2005).

The role of golgins in stacking GM130 and p115 are considered to form part of the Golgi matrix, which maintains the stacked cisterna architecture of the GA (Nakamura et al. 1995; Seeman et al. 2000a,b). GRASP65, and its related GRASP55, also have roles in Golgi stacking independent of GM130 (Barr et al. 1998; Shorter et al. 1999). P115, giantin, GM130 and GRASP65 have been shown to be required for both cisterna re-growth and cisterna stacking (two sub-reactions during in vitro Golgi reassembly after mitotic fragmentation) (Shorter and Warren 1999). The golgins could also provide the physical link between the Golgi cisternae, facilitating their stacking (Gillingham and Munro 2003). GM130/GRASP, giantin and p115 could thus be involved into the stacking of Golgi cisternae when the GA reforms following mitosis, a case of membrane tethering without subsequent fusion (Shorter and Warren 1999). Several lines of evidence suggest that the golgins could be important for stacking. In vitro, reconstruction of Golgi cisternae from mitotic fragments requires some tethering factors (Rabouille et al. 1995). Moreover, EM has identified proteinaceous bridges linking adjacent cisternae together (Franke et al. 1976; Cluett and Brown 1992). Finally, it has been shown that the HR2 domain of p115 has an essential role in the elongation of the Golgi cisternae (Sohda et al. 2007). For instance, p115 is required for the stacking of reassembling Golgi cisternae at an early stage in shack formation, before GRASP65 (Shorter and Warren 2002). Indeed, p115 is only required for cisterna stacking (Shorter and Warren 2002). Further, the acidic domain of p115 is not sufficient for the reassembly of Golgi cisternae, with p115 dimerization also reported to be needed (Dirac-Svejstrup et al. 2000). Some matrix proteins, such as the GRASPs, are considered as a “glue” to attach cisternae to each other.

Golgi ribbon and the function of the golgins

*

237

As originally proposed, this model has several weaknesses. First, most of the golgins are excluded from the space between the medial cisternae (Beznoussenko et al. submitted), which makes a role in stacking of the medial GA cisternae problematic. Then GRASP65 is not essential for € tterlin et al. 2005), as the GA shows a stacked organicisterna stacking (Su zation in GRASP65-depleted HeLa cells, where the number of cisternae per stack is reduced from a mean of 6 to a mean of 3 per stack. Precise analysis € tterlin et al. (2005) gives the impression that the cisternae of fig. 2B of Su became longer. However, as the Golgi localization of GM130 is not affected by GRASP65 depletion, its binding to GRASP65 cannot be the sole mechanism for its localization to Golgi membranes. In these GRASP65-depleted HeLa cells, VSV-G was transported to the cell surface with similar kinetics to the control, and after washout of nocodazole, Golgi membranes reassembled in the pericentriolar region of the cells, just as seen in the control cells. BFA-induced Golgi dynamics are also normal in the absence of GRASP65, and GRASP65 is important for bipolar spindle formation € tterlin et al. 2005). (Su

The role of the golgins in promoting attachment of CGN and TGN Thus, so far, three main factors have to co-exist to generate the Golgi ribbon: an intact MT star system, the input of membrane from the ER, and the activity of the different golgins that have important roles for the integration of the ER-derived membranes into the cisternae of neighbouring stacks, thus ‘glueing’ the separated stacks into a ribbon. The SNARE/Rab machineries are also important here. If these main preconditions are fulfilled, the concentration of the Golgi stacks within a narrow space (near the centrosome) inevitably leads to the generation of the ribbon. As such, the function of golgins might only be important for the generation of the Golgi ribbon and not for transport. However, the golgins are present not only in cells that can form a Golgi ribbon, but also in plant, yeast and insect cells, where the ribbon is absent. Similarly, p115 is found in all eukaryotes, and the GRASPs are found in all eukaryotes except plants, with Golgin-45 present only in vertebrates and GM130 present only in mammals. As with p115, Rab1 is found in all eukaryotes, while Rab2, as with its partner Golgin-45, is also only present in vertebrates (Short et al. 2005). Thus, in cells where the Golgi ribbon is not formed, some of the golgins are absent (plants, yeast). However, in general this machinery is present in all species. Importantly, cells for which the fragmentation of the ribbon during mitosis is more important have a more developed golgin machinery. The obvious question that arises is thus whether the function of the golgins is just for the generation of the ribbon? Detailed analysis of this aspect has revealed that in the absence of the golgins, and even when the Golgi ribbon is artificially broken by

238

*

M. A. De Matteis et al.

depolymerization of MTs, intracellular transport is impaired. Elimination of the function of p115 induces severe alterations in transport (see Table 1). Even GM130/GRASP65 depletion leads to a delay of transport and to less precise protein glycosylation (Marra et al. 2007). In turn, inhibition of Golgin-245 affects the function of the lysosomal system. After siRNAdepletion of Golgin-245, lysosomes show a more central pattern and there is accumulation of aberrant multivesicular structures (Yoshino et al. 2005). Finally, inhibition of the function of Golgin-97 slows down intra-Golgi transport, making protein sorting at the level of the trans-Golgi/TGN less precise (Beznoussenko et al. submitted) and affecting endosome-to-TGN transport (Lu et al. 2004). As depletion of each of the golgins has substantial functional effects (Tables 1 and 2), this suggests that the formation of the ribbon may not be the main function of the golgins, but actually represent a lateral effect of golgin function that appears when the Golgi stacks are centralized. In this case, the function of the golgins could be for the building of the functional Golgi stacks. Indeed, in cells lacking GM130 and devoid of MTs, the most cis cisterna of the stacks is absent both before and after the release of the transport block, whereas in control cells this cisterna is absent in the resting stacks but is visible in transporting stacks (Marra et al. 2007). Careful € tterlin et al. (2005) gives the impression that without analysis of fig. 2Bc in Su GRASP65, the attachment of the cis-most cisternae to the Golgi cisternae is affected. Moreover, examination of fig. 7 in Colanzi et al. (2000) suggests that Golgi fragments formed during incubation of permeabilized interphase cells with mitotic cytosol are not covered by the cis-most cisterna from the cis side and by the trans-most cisterna from the trans side. Instead, from the trans side they are often covered by the trans ER. Thus, within the framework of the non-vesicular models of intra-Golgi transport, and taking into account the hypothesis that the cis and trans golgins are responsible for the attachment of the CGN and the TGN (see Chapter 1.2), respectively, this means that in the absence of transport the medial GA is not covered by the attached CGN from the cis side and by the attached TGN from the trans side. The functional role of this Golgi configuration could be the regulation of the fusion–fission events that are important for protein and lipid sorting within, for example, the ‘kissand-run’ models of intra-Golgi transport. At least this hypothesis is in agreement with the finding that in the absence of GM130 the attached CGN is not formed even in transporting stacks (Marra et al. 2007), whereas after inhibition of Golgin-97 the attached TGN cannot replace the trans ER attached to the medial Golgi cisternae in resting stacks after restoration of intra-Golgi transport (Beznoussenko et al. submitted). This inability could affect protein sorting. Golgins could also be important for the correct sorting of post-Golgi carrier trafficking to the newly formed plasma membrane. For instance, the insect

Golgi ribbon and the function of the golgins

*

239

golgin Lava lamp (Lva) is required for cellularization, the rapid division of a large oocyte into small cells (Papoulas et al. 2005). During mitosis some golgins are phosphorylated. This phosphorylation can interfere with golgin function, and thus, during mitosis, the Golgi ribbon fragments and then more severe dispersion occurs due to loss of function of golgins, such as p115, GM130 and GRASPs (see Chapter 3.15), and maybe other golgins. One could imagine the following models for the function of golgins. At the ER exit sites, p115 and Golgin-84 regulate the maturation of the EGCs through their interactions with Rab1, SNAREs or other factors. The main function of p115 is its activity that is related to the centralization of EGCs. Golgin-245 has a similar function. Uso1/p115 can directly promote the formation of SNARE complexes, and thus of membrane fusion (Shorter et al. 2002). Additionally, p115 enhances the lipid activation of cytidylyltransferase (Feldman and Weinhold 1998), which also facilitates the maturation of EGCs. If the function of p115, Golgin-84 and Golgin-245 is normal and the carriers arrive at the central Golgi area, this event forces the cis-Golgi network to form the attached CGN, which facilitates the integration of the EGCs into the stack. Inhibition of golgin function would affect this maturation process, leading to a block in carrier centralization and peripheral fragmentation of the GA. CGN attachment is GM130 and GRASP65 dependent. GM130-containing EGCs develop the ability to undergo homotypic coalescence (Marra et al. 2001) and fusion, and as a possible consequence, they become larger and acquire a complex disc-like shape. The acquisition of GM130 and the development of this tendency to undergo homotypic fusion can be viewed as a sort of maturation process through which EGCs gradually acquire the properties and composition of the next compartment, i.e. the Golgi cisternae. Indeed, the arrival of a synchronized wave of EGCs into the stacks coincides with an immediate increase in the number of cisternae per stack and in the surface area of the cisternae (concomitant with a decrease in the surface area of the EGCs), thus indicating that the EGCs are integrally incorporated into the stacks. GM130 is a key player in this process (Marra et al. 2007). However, even for most studies of GM130, its ultimate mechanism of action remains to be defined (Puthenveedu et al. 2006; Marra et al. 2007). Further, an attachment of the cis-most cisterna of the CGN per se is not sufficient for the formation of the ribbon. To complete ribbon formation, it is necessary to have the attachment of the trans-most cisterna to the stack. The cisternal part of the TGN attaches to the membrane domain that contains mostly cargo. Thus, only the concentrated action of all of the golgins is sufficient for the generation of the Golgi ribbon and for the normal function of the Golgi stacks. The function of giantin is the most mysterious and needs additional study. Thus, the hypothesis about the role of golgins in the regulation of the structure of the cis and trans sides of the GA appear rather probable.

240

*

M. A. De Matteis et al.

Figure 2. Scheme showing the role of ER-to-Golgi transport, cis-Golgins and trans-Golgins for the formation of the Golgi ribbon. Two Golgi stacks are shown in the top panel, each comprising three cisternae and two ER exit sites (ERES). The second panel demonstrates the exit of cargo from the ER. The ER exit sites became larger. The third panel shows formation EGCs located between the ER exit sites and the stacks. The fourth panel shows the fusion of two EGCs. The bottom panels show the resulting structure of the GA when ER-to-Golgi transport is blocked (left-most panel), when ER-to-Golgi transport occurs in normal cells (central-left panel), when the ER-to-Golgi transport occurs in cells devoid of GM130 (central-right panel), and when ER-toGolgi transport occurs in cells where the function of Golgin-97 is inhibited (right-most panel). As a result, the Golgi ribbon is normal in the transporting cells expressing GM130, and fragmented in cells without ER-to-Golgi transport (small fragments), in transporting cells without GM130 (small fragments), and in transporting cells without Golgin-97 (larger and more variable fragments).

Conclusions The data presented here show that the organization of the Golgi ribbon depends on many factors, such as the ability of cells to induce the concentration of Golgi stacks within a restricted space, the normal functioning of the golgins and the SNARE/Rab machineries, and several other less studied factors (a possible scheme of the interactions of these factors is shown in Fig. 2). The main function of golgins seems to be the regulation of the Golgi ribbon. Characteristics of golgins are summarized in Table 1. To date though, experimental observations can, and should, still be interpreted within the framework of several models of intracellular transport.

Abbreviations GA GFP

Golgi apparatus green fluorescent protein

Golgi ribbon and the function of the golgins

GREG MT MTOC SEM TGN

*

241

Golgi-resident GPI-anchored protein microtubule microtubule-organizing centre scanning electron microscopy trans-Golgi network

References Abrami L, Fivaz M, Kobayashi T, Kinoshita T, Parton RG, Van der Goot FG (2001) Crosstalk between caveolae and glycosylphosphatidylinositol-rich domains. J Biol Chem 276: 30729–30736 Allan BB, Moyer BD, Balch WE (2000) Rab1 recruitment of p115 into a cisSNARE complex: programming budding COPII vesicles for fusion. Science 289: 444–448 Barr FA, Nakamura N, Warren G (1998) Mapping the interaction between GRASP65 and GM130, components of a protein complex involved in the stacking of Golgi cisternae. EMBO J 17: 3258–3268 Barr FA (1999) A novel Rab6-interacting domain defines a family of Golgi-targeted coiled-coil proteins. Curr Biol 9: 381–384 Barr FA, Short B (2003) Golgins in the structure and dynamics of the Golgi apparatus. Curr Opin Cell Biol 15: 405–413 Brown WJ, Chambers K, Doody A (2003) Phospholipase A2 (PLA2) enzymes in membrane trafficking: mediators of membrane shape and function. Traffic 4: 214–221 Burkhardt JK, Echeverri CJ, Nilsson T, Vallee RB (1997) Over expression of the dynamitin (p50) subunit of the dynactin complex disrupts dynein-dependent maintenance of membrane organelle distribution. J Cell Biol 139: 469–484 Burkhard P, Stetefeld J, Strelkov SV (2001) Coiled coils: a highly versatile protein folding motif. Trends Cell Biol 11: 82–88 Chan EKL, Fritzler MJ (1998) Golgins: coiled-coil proteins associated with the Golgi complex, Electron. J Biotechnol 1: 1–10 Colanzi A, Deerinck TJ, Ellisman MH, Malhotra V (2000) A specific activation of mitogenactivated protein kinase kinase 1 (MEK1) is required for Golgi fragmentation during mitosis. J Cell Biol 149: 331–339 Colanzi A, Suetterlin C, Malhotra V (2003) Cell-cycle-specific Golgi fragmentation: how and why? Curr Opin Cell Biol 15: 462–467 ColeNB, Smith CL, Sciaky N, Terasaki M,Edidin M,Lippincott-SchwartzJ (1996a) Diffusional mobility of Golgi proteins in membranes of living cells. Science 273: 797–801 Cole NB, Sciaky N, Marotta A, Song J, Lippincott-Schwartz J (1996b) Golgi dispersal during microtubule disruption: regeneration of Golgi stacks at peripheral endoplasmic reticulum exit sites. Mol Biol Cell 7: 631–650 sy-Theulaz I, Pauloin A, Pfeffer SR (1992) Cytoplasmic dynein participates in the Corthe centrosomal localization of the Golgi complex. J Cell Biol 118: 1333–1345 Diao A, Rahmann D, Pappin DJ, Lucocq J, Lowe M (2003) The coiled-coil membrane protein Golgin-84 is a novel Rab effector required for Golgi ribbon formation, J Cell Biol 160: 201–212 Diaz Anel AM, Malhotra V (2005) PKCeta is required for beta1gamma2/beta3gamma2and PKD-mediated transport to the cell surface and the organization of the Golgi apparatus. J Cell Biol 169: 83–91 Dirac-Svejstrup AB, Shorter J, Waters MG, Warren G (2000) Phosphorylation of the vesicle-tethering protein p115 by a casein kinase II-like enzyme is required for Golgi reassembly from isolated mitotic fragments. J Cell Biol 150: 475–488

242

*

M. A. De Matteis et al.

Eystathioy T, Jakymiw A, Fujita DJ, Fritzler MJ, Chan EK (2000) Human autoantibodies to a novel Golgi protein Golgin-67, high similarity with golgin-95/GM130 autoantigen. J Autoimmun 14: 179–187 Feinstein TN, Linstedt AD (2008) GRASP55 regulates Golgi ribbon formation. Mol Biol Cell Published April 23 Feldman DA, Weinhold PA (1998) Cytidylyltransferase-binding protein is identical to transcytosis-associated protein (TAP/p115) and enhances the lipid activation of cytidylyltransferase. J Biol Chem 273: 102–109 Franke WW, Kartenbeck J (1976) Some principles of membrane differentiation. In: Mueller-Berat N (ed) Progress in differentiation research. Elsevier-North Holland Publishing Co., Amsterdam, pp 213–243 Fridmann-Sirkis Y, Siniossoglou S, Pelham HR (2004) TMF is a golgin that binds Rab6 and influences Golgi morphology. BMC Cell Biol 5: 18 Fritzler MJ, Hamel JC, Ochs RL, Chan EK (1993) Molecular characterization of two human autoantigens: unique cDNAs encoding 95- and160-kD proteins of a putative family in the Golgi complex. J Exp Med 178: 49–62 Fritzler MJ, Lung CC, Hamel JC, Griffith KJ, Chan EK (1995) Molecular characterization of Golgin-245, a novel Golgi complex protein containing a granin signature. J Biol Chem 270: 31262–31268 Gillingham AK, Pfeifer AC, Munro S (2002) CASP, the alternatively spliced product of the gene encoding the CCAAT-displacement protein transcription factor, is a Golgi membrane protein related to giantin. Mol Biol Cell 13: 3761–3774 Gillingham AK, Munro S (2003) Long coiled-coil proteins and membrane traffic. Biochim Biophys Acta 1641: 71–85 Griffith KJ, Chan EK, Lung CC, Hamel JC, Guo X, Miyachi K, Fritzler MJ (1997) Molecular cloning of a novel 97-kDa Golgi complex autoantigen associated with Sjogren’s syndrome. Arthritis Rheum 40: 1693–1702 Harada A, Takei Y, Kanai Y, Tanaka Y, Nonaka S, Hirokawa N (1998) Golgi vesiculation and lysosome dispersion in cells lacking cytoplasmic dynein. J Cell Biol 141: 51–59 Hidalgo Carcedo C, Bonazzi M, Spano S, Turacchio G, Colanzi A, Luini A, Corda D (2004) Mitotic Golgi partitioning is driven by the membrane-fissioning protein CtBP3/ BARS. Science 305: 93–96 Ho WC, Allan VJ, Van Meer G, Berger EG, Kreis TE (1989) Reclustering of scattered Golgi elements occurs along microtubules. Eur J Cell Biol 48: 250–263 Holleran EA, Ligon LA, Tokito M, Stankewich MC, Morrow JS, Holzbaur ELF (2001) Beta III spectrin binds to the Arp1 subunitof dynactin. J Biol Chem 276: 36598–36605 Horton AC, Racz B, Monson EE, Lin AL, Weinberg RJ, Ehlers MD (2005) Polarized secretory trafficking directs cargo for asymmetric dendrite growth and morphogenesis. Neuron 48: 757–771 Inoue T (1992) Complementary scanning electron microscopy: technical notes and applications. Arch Histol Cytol 55(Suppl): 45–51 Kjer-Nielsen L, Teasdale RD, Van Vliet C, Gleeson PA (1999) A novel Golgi-localisation domain shared by a class of coiled-coil peripheral membrane proteins. Curr Biol 9: 385–388 Kondylis V, Rabouille C (2003) A novel role for dp115 in the organization of tER sites in Drosophila. J Cell Biol 162: 185–198 Kondylis V, Spoorendonk KM, Rabouille C (2005) dGRASP localization and function in the early exocytic pathway in Drosophila S2 cells. Mol Biol Cell 16: 4061–4072 Kumar S, Lee IH, Plamann M (2004) Cytoplasmic dynein ATPase activity is regulated by dynactin-dependent phosphorylation. J Biol Chem 275: 31798–31804 Kweon HS, Beznoussenko GV, Micaroni M, Polishchuk RS, Trucco A, Martella O, Di Giandomenico D, Marra P, Fusella A, Di Pentima A, Berger EG, Geerts WJ, Koster AJ, Burger KN, Luini A, Mironov AA (2004) Golgi enzymes are enriched in perforated

Golgi ribbon and the function of the golgins

*

243

zones of Golgi cisternae but are depleted in COPI vesicles. Mol Biol Cell 15: 4710–4724 Lesa GM, Seemann J, Shorter J, Vandekerckhove J, Warren G (2000) The amino-terminal domain of the Golgi protein giantin interacts directly with the vesicle-tethering p115. J Biol Chem 275: 2831–2836 Lu L, Tai G, Hong W (2004) Autoantigen golgin-97, an effector of Arl1 GTPase, participates in traffic from the endosome to the TGN. Mol Biol Cell 15: 4426–4443 Luke MR, Kjer-Nielsen L, Brown DL, Stow JL, Gleeson PA (2003) GRIP domain-mediated targeting of two new coiled-coil proteins, GCC88 and GCC185, to subcompartments of the trans-Golgi network. J Biol Chem 278: 4216–4226 Malsam J, Satoh A, Pelletier L, Warren G (2005) Golgin tethers define subpopulations of COPI vesicles. Science 307: 1095–1098 Marra P, Maffucci T, Daniele T, Tullio GD, Ikehara Y, Chan EK, Luini A, Beznoussenko G, Mironov A, De Matteis MA (2001) The GM130 and GRASP65 Golgi proteins cycle through and define a subdomain of the intermediate compartment. Nat Cell Biol 3: 1101–1013 Marra P, Salvatore L, Mironov A Jr, Di Campli A, Di Tullio G, Trucco A, Beznoussenko G, Mironov A, De Matteis MA (2007) The biogenesis of the Golgi ribbon: the roles of membrane input from the ER and of GM130. Mol Biol Cell 18: 1595–1608 Matanis T, Akhmanova A, Wulf P, Del Nery E, Weide T, Stepanova T, Galjart N, Grosveld F, Goud B, De Zeeuw CI, Barnekow A, Hoogenraad CC (2002) Bicaudal-D regulates COPI-independent Golgi–ER transport by recruiting the dynein dynactin motor complex. Nat Cell Biol 4: 986–992 Minin AA (1997) Dispersal of Golgi apparatus in nocodazole-treated fibroblasts is a kinesin-driven process. J Cell Sci 110: 2495–2505 Mironov AA, Mironov AA Jr, Beznoussenko GV, Trucco A, Lupetti P, Smith JD, Geerts WJ, Koster AJ, Burger KN, Martone ME, Deerinck TJ, Ellisman MH, Luini A (2003) ER-toGolgi carriers arise through direct en-bloc protrusion and multistage maturation of specialized ER exit domains. Dev Cell 5: 583–594 Mogelsvang S, Marsh BJ, Ladinsky MS, Howell KE (2004) Predicting function from structure: 3D structure studies of the mammalian Golgi complex. Traffic 5: 338–345 Moyer BD, Allan BB, Balch WE (2001) Rab1 interaction with a GM130 effector complex regulates COPII vesicle cis-Golgi tethering. Traffic 2: 268–276 Nakamura N, Rabouille C, Watson R, Nilsson T, Hui N, Slusarewicz P, Kreis TE, Warren G (1995) Characterization of a cis-Golgi matrix protein, GM130. J Cell Biol 131: 1715–1726 Ojakian GK, Nelson WJ, Beck KA (1997) Mechanisms for de novo biogenesis of an apical membrane compartment in groups of simple epithelial cells surrounded by extracellular matrix. J Cell Sci 110: 2781–2794 Opat AS, Van Vliet C, Gleeson PA (2001) Trafficking and localisation of resident Golgi glycosylation enzymes. Biochimie 83: 763–773 Papoulas O, Hays TS, Sisson JC (2005) The golgin Lava lamp mediates dynein-based Golgi movements during Drosophila cellularization. Nat Cell Biol 7: 612–618 Pecot MY, Malhotra V (2006) The Golgi apparatus maintains its organization independent of the endoplasmic reticulum. Mol Biol Cell 17: 5372–5380 Perez F, Pernet-Gallay K, Nizak C, Goodson HV, Kreis TE, Goud B (2002) CLIPR-59, a new trans-Golgi/TGN cytoplasmic linker protein belonging to the CLIP-170 family. J Cell Biol 156: 631–642 Polishchuk RS, Polishchuk EV, Mironov AA (1999) Stack coalescence in MT-deprived cells with fragmented Golgi. Eur J Cell Biol 78: 170–185 Polishchuk RS, Mironov AA (2004) Structural aspects of Golgi function. Cell Mol Life Sci 61: 146–158 Pullikuth AK, Weidman PJ (2002) In-vitro transport on cis and trans sides of the Golgi involves two distinct types of coatomer and ADP-ribosylation factor-independent transport intermediates. J Biol Chem 277: 50355–50364

244

*

M. A. De Matteis et al.

Puthenveedu MA, Bachert C, Puri S, Lanni F, Linstedt AD (2006) GM130 and GRASP65dependent lateral cisternal fusion allows uniform Golgi-enzyme distribution. Nat Cell Biol 8: 238–248 Rabouille C, Levine TP, Peters J-M, Warren G (1995) An NSF-like ATPase, p97, and NSF mediate costernal regrowth from mitotic Golgi fragments. Cell 82: 905–914  tien M, Olivier L (1993) Modulation of the Golgi Rambourg A, Clermont Y, Chre apparatus in stimulated and non-stimulated prolactin cells of female rats. Anat Rec 235: 353–362 Renna L, Hanton SL, Stefano G, Bortolotti L, Misra V, Brandizzi F (2005) Identification and characterization of AtCASP, a plant transmembrane Golgi matrix protein. Plant Mol Biol 58: 109–122 Rios RM, Tassin AM, Celati C, Antony C, Boissier MC, Homberg JC, Bornens M (1994) A peripheral protein associated with the cis-Golgi network redistributes in the intermediate compartment upon brefeldin A treatment. J Cell Biol 125: 997–1013 Rios RM, Sanchis A, Tassin AM, Fedriani C, Bornens M (2004) GMAP-210 recruits gammatubulin complexes to cis-Golgi membranes and is required for Golgi ribbon formation. Cell 118: 323–335 Roghi C, Allan VJ (1999) Dynamic association of cytoplasmic dynein heavy chain 1a with the Golgi apparatus and intermediate compartment. J Cell Sci 112: 4673–4685 €ki V, McDowall AW, Parton RG, Gruenberg J (2000) The Rojo M, Emery G, Marjoma transmembrane protein p23 contributes to the organization of the Golgi apparatus. J Cell Sci 113: 1043–1057 Schroer TA (2004) Dynactin. Annu Rev Cell Dev Biol 20: 759–779 Seaman MN (2004) Cargo-selective endosomal sorting for retrieval to the Golgi requires retromer. J Cell Biol 165: 111–122 Seelig HP, Schranz P, Schroter H, Wiemann C, Griffiths G, Renz M (1994) Molecular genetic analyses of a 376-kilodalton Golgi complex membrane protein (giantin). Mol Cell Biol 14: 2564–2576 Seemann J, Jokitalo EJ, Pypaert M, Warren G (2000a) Matrix proteins can generate the higher order architecture of the Golgi apparatus. Nature 407: 1022–1026 Seemann J, Jokitalo EJ, Warren G (2000b) The role of the tethering proteins p115 and GM130 in transport through the Golgi apparatus in vivo. Mol Biol Cell 11: 635–645 Shestakova A, Zolov S, Lupashin V (2006) COG complex-mediated recycling of Golgi glycosyltransferases is essential for normal protein glycosylation. Traffic 7: 191–204 Short B, Preisinger C, Schaletzky J, Kopajtich R, Barr FA (2002) The Rab6 GTPase regulates recruitment of the dynactin complex to Golgi membranes. Curr Biol 12: 1792–1795 Short B, Haas A, Barr FA (2005) Golgins and GTPases, giving identity and structure to the Golgi apparatus. Biochim Biophys Acta 1744: 383–395 Shorter J, Watson R, Giannakou M-E, Clarke M, Warren G, Barr FA (1999) GRASP55, a second mammalian GRASP protein involved in the stacking of Golgi cisternae in a cell-free system. EMBO J 18: 4949–4960 Shorter J, Warren G (1999) A role for the vesicle tethering protein, p115, in the postmitotic stacking of reassembling Golgi cisternae in a cell-free system. J Cell Biol 146: 57–70 Shorter J, Beard MB, Seemann J, Dirac-Svejstrup AB, Warren G (2002) Sequential tethering of golgins and catalysis of SNAREpin assembly by the vesicle-tethering protein p115. J Cell Biol 157: 45–62 Siniossoglou S, Peak-Chew SY, Pelham HR (2000) Ric1p and Rgp1p form a complex that catalyses nucleotide exchange on Ypt6p. EMBO J 19: 4885–4894 Sohda M, Misumi Y, Yoshimura S, Nakamura N, Fusano T, Ogata S, Sakisaka S, Ikehara Y (2007) The interaction of two tethering factors, p115 and COG complex, is required for Golgi integrity. Traffic 8: 270–284

Golgi ribbon and the function of the golgins

*

245

Sonnichsen B, Lowe M, Levine T, Jamsa E, Dirac-Svejstrup AB, Warren G (1998) A role for giantin in docking COPI vesicles to Golgi membranes. J Cell Biol 140: 1013–1021 Storrie B, Kreis TE (1996) Probing the mobility of membrane proteins inside the cell. Trends Cell Biol 6: 321–324 Storrie B, White J, Rottger S, Stelzer EH, Suganuma T, Nilsson T (1998) Recycling of Golgiresident glycosyltransferases through the ER reveals a novel pathway and provides an explanation for nocodazole-induced Golgi scattering. J Cell Biol 143: 1505–1521 Sun Y, Shestakova A, Hunt L, Sehgal S, Lupashin V, Storrie B (2007) Rab6 regulates both ZW10/RINT-1 and conserved oligomeric Golgi complex-dependent Golgi trafficking and homeostasis. Mol Biol Cell 18: 4129–4142 € tterlin C, Polishchuk R, Pecot M, Malhotra V (2005) The Golgi-associated protein Su GRASP65 regulates spindle dynamics and is essential for cell division. Mol Biol Cell 16: 3211–3222 Thyberg J, Moskalewski S (1999) Role of microtubules in the organization of the Golgi complex. Exp Cell Res 246: 263–279 Trucco A, Polishchuk RS, Martella O, Di Pentima A, Fusella A, Di Giandomenico D, San Pietro E, Beznoussenko GV, Polishchuk EV, Baldassarre M, Buccione R, Geerts WJ, Koster AJ, Burger KN, Mironov AA, Luini A (2004) Secretory traffic triggers the formation of tubular continuities across Golgi sub-compartments. Nat Cell Biol 6: 1071–1081 Tsukada M, Will E, Gallwitz D (1999) Structural and functional analysis of a novel coiledcoil protein involved in Ypt6 GTPase-regulated protein transport in yeast. Mol Biol Cell 10: 63–75  T, Ayala I, Kok JW, Renau-Piqueras J, Egea G (1998) Actin microValderrama F, Babia filaments are essential for the cytological positioning and morphology of the Golgi complex. Eur J Cell Biol 76: 9–17 Vale RD (2003) The molecular motor toolbox for intracellular transport. Cell 112: 467–480 Vallee RB, Williams JC, Varma D, Barnhart LE (2004) Dynein: an ancient motor protein involved in multiple modes of transport. J Neurobiol 58: 189–200 Varki A (1998) Factors controlling the glycosylation potential of the Golgi apparatus. Trends Cell Biol 8: 34–40 Vaughan PS, Miura P, Henderson M, Byrne B, Vaughan KT (2002) A role for regulated binding of p150Glued to microtubule plus ends in organelle transport. J Cell Biol 158: 305–319 Ward TH, Brandizzi F (2004) Dynamics of proteins in Golgi membranes: comparisons between mammalian and plant cells highlighted by photobleaching techniques. Cell Mol Life Sci 61: 172–185 Wehland J, Henkart M, Klausner R, Sandoval IV (1983) Role of microtubules in the distribution of the Golgi apparatus: effect of taxol and microinjected anti-alphatubulin antibodies. Proc Natl Acad Sci USA 80: 4286–4290 Weigert R, Silletta MG, Spanò S, Turacchio G, Cericola C, Colanzi A, Mancini R, Polishchuk EV, Salmona M, Facchiano F, Burger KNJ, Mironov A, Luini A, Corda D (1999) CtBP/ BARS induces fission of Golgi membranes by acylating lysophosphatidic acid. Nature 402: 429–433 Whyte JR, Munro S (2002) Vesicle tethering complexes in membrane traffic. J Cell Sci 115: 2627–2637 Xu Y, Takeda S, Nakata T, Noda Y, Tanaka Y, Hirokawa N (2002) Role of KIFC3 motor protein in Golgi positioning and integration. J Cell Biol 158: 293–303 Xu Y, Martin S, James DE, Hong W (2002) GS15 forms a SNARE complex with syntaxin 5, GS28, and Ykt6 and is implicated in traffic in the early cisternae of the Golgi apparatus. Mol Biol Cell 13: 3493–3507 Xueyi Li, Kaloyanova D, Van Eijk M, Eerland R, Van der Goot G, Oorschot V, Klumperman J, Lottspeich F, Starkuviene V, Wieland FT, Helms JB (2007) Involvement of a Golgi-

246

*

M. A. De Matteis et al.

resident GPI-anchored protein in maintenance of the Golgi structure. Mol Biol Cell 18: 1261–1271 Yang JS, Lee SY, Spanò S, Gad H, Zhang L, Nie Z, Bonazzi M, Corda D, Luini A, Hsu VW (2005) A role for BARS at the fission step of COPI vesicle formation from Golgi membrane. EMBO J 24: 4133–4143 Yang JS, Zhang L, Lee SY, Gad H, Luini A, Hsu VW (2006) Key components of the fission machinery are interchangeable. Nat Cell Biol 8: 1376–1382 Yoshino A, Setty SR, Poynton C, Whiteman EL, Saint-Pol A, Burd CG, Johannes L, Holzbaur EL, Koval M, McCaffery JM, Marks MS (2005) tGolgin-1 (p230, golgin245) modulates Shiga-toxin transport to the Golgi and Golgi motility towards the microtubule-organizing centre. J Cell Sci 118: 2279–2293 Young WW Jr (2004) Organization of Golgi glycosyltransferases in membranes: complexity via complexes. J Membrane Biol 198: 1–13 Zhang T, Hong W (2001) Ykt6 forms a SNARE complex with syntaxin 5, GS28, and Bet1 and participates in a late stage in endoplasmic reticulum–Golgi transport. J Biol Chem 276: 27480–27487

Functional cross talk between membrane trafficking and cell signalling

*

247

Functional cross talk between membrane trafficking and cell signalling The Golgi complex as a signalling platform Michele Sallese

Introduction Eukaryotic cells are complex systems that are capable of fine-tuning their functioning following internal and external perturbations. Their plasma membranes bearseveraldifferentreceptorproteinsthatcansensetheexternal milieu and transduce information across the plasma membrane to cytosolic proteins. These, in turn, transmit the messages to other partners along signalling cascades that lead to the required cellularresponse. Such events are known as ‘cell signalling’, and these cascades have critical effects on the behaviour of a cell, since they can affect cell motility and growth, and indeed, apoptosis. According to some estimations (Venter 2001; Imanishi et al. 2004), approximately 10–12% of the genes in the human genome encode for signal transduction proteins, thus highlighting the relevance of these functions to the cell. Signal transduction is a highly dynamic and regulated process. The overstimulation of a cell is limited by multiple control mechanisms, which range from desensitization to endocytosis, and which can eventually lead to the down-regulation of a receptor. The desensitization process involves several molecular events that leads to the uncoupling of a receptor from its downstream effectors. In recent years, the extension of our understanding of signal transduction mechanisms has been accompanied by the emergence of an important role of endomembranes in these signalling processes (Sorkin and Von Zastrow 2002; Di Fiore and De Camilli 2001). Receptor internalization leads to a decrease in plasma-membrane signalling (as part of the desensitization process), although the endocytosed receptor can continue to engage in signalling pathways that are different from those activated when the same receptor was present at the plasma membrane. It is now clear that the endosomal membranes are not only a sorting compartment where plasma-membrane receptors are either committed towards the degradative lysosomal pathway or dephosphorylated and recycled back to the plasma membrane for a further round of signalling; but also provide a signalling membrane from where specific signals originate (Di Fiore and De Camilli 2001; Sorkin and Von Zastrow 2002). As well as the endosomal membranes, several studies have reported the presence of signalling proteins on different organelles, including the

248

*

M. Sallese

endoplasmic reticulum (ER), the Golgi complex, the mitochondria, and the nuclear membrane, although our knowledge of their roles in these locations remains vague (Sallese et al. 2006). The aim of this chapter is thus to bring together the available information relating to the presence and roles of signalling proteins on the membranes of the Golgi complex. The Golgi membranes host a variety of classical signalling molecules that are generally present and functional at the plasma membrane (Sallese et al. 2006). These range from kinases (Birkeli et al. 2003; Colanzi et al. 2003) to phosphatases (Lavoie et al. 2000), to phospholipases (Freyberg et al. 2003), heterotrimeric G-proteins (Wilson et al. 1994) and phosphodiesterases (Asirvatham et al. 2004), to name but a few (Donaldson and LippincottSchwartz 2000). The easiest explanation for their presence on the Golgi complex is that they can be visualized on endomembranes because they are in transit towards their final destination, the plasma membrane (Michaelson et al. 2002; Sallese et al. 2006). Indeed, ER and Golgi membranes function as a platform for the assembly of multiprotein complexes (e.g. heterotrimeric G proteins). In addition, these membranes host the palmitoyl transferases (Ohno et al. 2006), the function of which is an obligatory step in the acylation of many proteins before their transfer to their functional sites of action; again, generally the plasma membrane (Marrari et al. 2007). However, a growing body of evidence shows that these signalling proteins that are localized on endomembranes can be in their active state, thus suggesting that they can have a functional role in these locations (Bivona and Philips 2003). We envisage four different scenarios for the roles of these signalling proteins on the Golgi complex. First, the Golgi localized signalling proteins could participate into Golgi disassembly during the mitosis downstream to growth factor receptor activation on the plasma-membrane (Fig. 1). Second plasmamembrane receptors could signal to the Golgi, regulating its activity and its secretory functions (Buccione et al. 1996), (Fig. 1). Third, plasmamembrane receptors could use the Golgi membranes, exploiting them as a central location, or as a signalling platform or relay station for the integration of messages in their transfer to their final destinations, which could include the cytoskeleton, mitochondria, etc. (Bivona and Philips 2003) (Fig. 1). Fourth, the traffic itself could trigger a signalling cascade that is involved in the coordination of the various secretory compartments (Pulvirenti et al. 2008) (Fig. 1) or possibly could affect other organelles or other cellular functions besides membrane trafficking (Fig. 1). Of these four scenarios, to data, the last two have only been support by a few indications. We believe that they represent an important area of investigation that it will be worthwhile to explore in the future.

The phospholipases The lipid derivatives include a large array of molecules for which an involvement in key cell signalling functions is well established; however, more

Functional cross talk between membrane trafficking and cell signalling

*

249

Figure 1. The Golgi is a signalling hub. This cartoon schematizes four possible signalling pathways involving the Golgi complex. First, the Golgi localized signalling proteins could participate into Golgi disassembly during the mitosis downstream to growth factor receptor activation on the plasma membrane (arrow 1). Second, plasma-membrane receptors could signal to the Golgi, regulating its activity and its secretory functions (arrow 2). Third, plasmamembrane receptors could use the Golgi membranes, exploiting them as a central location, or as a signalling platform or relay station for the integration of messages in their transfer to their final destinations, which could include the cytoskeleton, mitochondria, etc. (arrow 3). Fourth, the traffic itself could trigger a signalling cascade that is involved in the coordination of the various secretory compartments, or possibly could affect other organelles or other cellular functions besides membrane trafficking (arrow 4). G Golgi complex; IC intermediate compartment; ER endoplasmic reticulum; N nucleus.

recently, they have also emerged as regulators of membrane trafficking (De Matteis et al. 2005). It is still not clear if their roles in membrane trafficking impinge on their signalling properties or rely on changes in the physical properties of the membranes, and indeed the available data suggest that there are contributions from both sides. To accomplish these important tasks, their levels are tightly regulated by a wealth of lipid-modifying enzymes, including kinases, phosphatases and lipases. The phospholipases are lipid-hydrolysing enzymes that can use the phosphatidylinositols (PIs) and phosphatidylcholine (PtdCho) in cell membranes as substrates. They can be divided into three families based on the phospholipid bond that they hydrolyse: phospholipase A1 (PLA1) and PLA2 hydrolyse fatty acids esterified on the sn-1 and sn-2 positions, respectively generating free fatty acid and lysophospholipid (Fig. 2). PLC hydrolyses the bond between the

250

*

M. Sallese

Figure 2. The diacylglycerol is the centrepiece of a functional lipid network. The scheme represents the metabolic pathways of the main signalling lipids possibly involved in the regulation of the Golgi functions. The arrows indicate the transformation of one compound into another in virtue of the enzyme written on the arrow itself. The flash indicates activation of PKD. AA, arachidonic acid; LPA, lysophosphatidic acid; PA, phosphatidic acid, DAG, diacylglycerol; AcylCoa, acyl-coenzyme A; PtdIns(4,5)P2, phosphatidylinositol (4,5)-bisphosphate; IP3, inositol (1,4,5)-trisphosphate; PC, phosphatidylcholine; SM, sphingomyelin. PLA2, phospholipase A2, PLD, phospholipase D; PI-PLC, phosphatidylinositol-specific phospholipase C; PAP, phosphatidic acid phosphatase; DAGK, diacyglycerol kinase; PC-PLC, phosphatidylcholinespecific phospholipase C; CerS, ceramide synthase; CDase, Ceramidase; CERT, ceramide transporter; SMS, sphingomyelin synthase; BARS, brefeldin A ADP-ribosylated substrate; PKD, protein kinase D. Pyrrophenone, PLA2 inhibitor; D609, PC-PLC and SMS inhibitor; FB1, (fumonisin B1) CerS inhibitor; U73122, PI-PLC inhibitor.

glycerol and the phosphate, generating diacylglycerol and phosphoinositols (Fig. 2), whereas PLD hydrolyses the bond between the phosphate and the inositol moiety generating phosphatidic acid and inositols (Fig. 2). The activity of PLA1 is mainly concentrated in the lumen of the lysosomes and on extracellular membranes, and therefore they are outside the scope of this review.

Phospholipase A2 Four main subfamilies of PLA2 have been identified: secretory PLA2 (sPLA2); cytosolic, Ca2þ -dependent PLA2 (cPLA2); intracellular, Ca2þ -independent PLA2 (iPLA2); and platelet-activating factor acetylhydrolases (PAF-AHs) (Brown et al. 2003). cPLA2 contains a Ca2þ -dependent membrane-binding domain (C2 domain). The activation of plasma-membrane receptors coupled to intracellular Ca2þ stimulation promotes the association of cPLA2 with the ER and with Golgi membranes (Evans and Leslie 2004). The recruitment of

Functional cross talk between membrane trafficking and cell signalling

*

251

cPLA2 to these membranes could simply suggest that they act as relay stations along a plasma-membrane-initiated signalling cascade. However, several lines of evidence indicate the involvement of cPLA2 in the regulation of membrane trafficking (Brown et al. 2003). Membrane trafficking involves the formation of tubular-shaped membranes that can act as carriers or as structural elements in specific compartments, including in the cis- and trans-Golgi networks (CGN and TGN) (Lippincott-Schwartz et al. 2001). Brown and co-workers (2000) have reported that when the enzymatic activity of PLA2 is inhibited, the BFA-induced tubules that emanate from the Golgi complex are impaired, together with the redistribution of Golgi proteins to the ER. Along the same lines, in vitro experiments performed on purified Golgi membranes have indicated that activators of PLA2 promote tubule formation (Polizotto et al. 1999). The possible mechanisms for these PLA2-mediated effects rely on the local production of inverted cone-shaped lysophospholipids that drive the formation of positive curvature, a process that is involved in tubule formation (de Figueiredo et al. 1998); however, a role for arachidonic acid as a signalling mechanism cannot be ruled out. It appears that PLA2 also has a role in membrane fusion, but unfortunately many reports are based on in vitro evidence, which makes it hard to assess the real contribution of PLA2 to fusion steps in membrane trafficking. Specifically, snake venom PLA2 increases the fusion of liposomes and isolated secretory granules (Blackwood et al. 1996). Moreover, the use of PLA2 inhibitors impairs endosomal fusion, a phenomenon that can be overcame by the addition of arachidonic acid (Mayorga et al. 1993). PLA2 inhibitors are also able to block endosomal fusion in vivo (de Figueiredo et al. 2000). Similar to that hypothesized for tubule formation, PLA2 could participate in membrane fusion by changing the local membrane composition through the production of lysolipids and free fatty acids, and/or via signalling cascades. From a physiology standpoint, PLA2 regulates the retrograde transport of proteins between the Golgi complex and the ER. The use of PLA2 inhibitors has shown an impairment of retrograde transport of a chimeric construct between the KDEL receptor and VSVG, the temperature-sensitive variant of the vesicular stomatitis virus G protein (de Figueiredo et al. 2000). Similar approaches have shown that PLA2 activity is important for the recycling of the transferrin receptor from the recycling endosomes towards the plasma membrane (de Figueiredo et al. 2001), and for the maintenance of the Golgi ribbon (de Figueiredo et al. 1999), with PLA2 inhibitors promoting the formation of separate Golgi stacks that remain in the perinuclear area (de Figueiredo et al. 1999).

Phospholipase D The human genome contains two phospholipase D (PLD) genes known as PLD1 and PLD2 (Freyberg et al. 2003). Both of these preferentially hydrolyze

252

*

M. Sallese

the PtdCho into phosphatidic acid (PA) and choline. The PLD1 protein is distributed throughout the cell, including on endomembranes, such as the Golgi complex, the ER, and the endosomes, with only a minor fraction at the plasma membrane (Freyberg et al. 2001, 2003). In contrast, PLD2 is mainly present on the plasma membrane, although one study has reported about 20% of PLD2 at the Golgi cisterna rims (Freyberg et al. 2003). Consistent with its intracellular localization, PLD is part of signalling pathways initiated by external stimuli via growth factors, cytokines, and G-protein-coupled receptors (GPCRs). PLD also participates in membrane trafficking events and maintenance of the Golgi structure, although it is not known whether the signalling and trafficking functions of PLD are linked; e.g., receptors activated at the plasma membrane could affect vesicular transport via the PLD on the Golgi complex. A few studies that have used isolated Golgi membranes and permeabilized cells have shown the involvement of PLD in the release of nascent secretory vesicles from the TGN (Ktistakis et al. 1996). PLD also shows transphosphatidylation activity, such that in the presence of a primary alcohol, the phosphatidyl group is transferred to the alcohol, forming a phosphatidyl-alcohol instead of PA (Freyberg et al. 2003). Indeed, the exploitation of this property of PLD has provided the main tool to interfere with PLD activity, and thus to improve our understanding of the role of PLD in membrane transport. Shields and co-workers (2000) used 1-butanol to demonstrate that the PLD product PA is required for ER-to-Golgi transport of VSVG, as well as for TGN-to-plasma-membrane transport. In addition, they showed that 1-butanol heavily alters the structure of the Golgi complex, raising doubts as to the real cause of this membrane transport inhibition. A recent report has also linked the transport of the cystic fibrosis transmembrane conductance regulator (CFTR) to PLD activity (Hashimoto et al. 2008). Specifically, treatment with 1-butanol impairs the exit of CFTR from the ER, a phenotype that can be rescued by exogenous addition of PA. CFTR transport defects have also been reported in cells knocked down for PLD1 using siRNAs (Hashimoto et al. 2008). In addition, the presence of high levels of PA impairs Golgi-to-plasma-membrane transport of CFTR. Altogether, these data indicate that PLD/PA is an important regulator of CFTR transport, and it would thus be worth exploring whether the pharmacological modulation of PA levels could help in the treatment of cystic fibrosis. PLD1 appears to be normally inactive, with its activity increasing through direct interactions with protein kinase C (PKC), and Rho and Arf family GTPases (Hammond et al. 1997). Phosphatidylinositol 4,5-bisphosphate (PtdIns45P2) can also promote PLD1 activity by interacting with its pleckstrin homology (PH) domain and KR motif (Sciorra et al. 1999; Brown et al. 1993). In contrast, PLD2 has a high basal activity (at least in vitro), and it also requires PtdIns45P2 although it responds poorly to PKC, Rho and Arf (Powner and Wakelam 2002; Freyberg et al. 2003).

Functional cross talk between membrane trafficking and cell signalling

*

253

In molecular terms, PLD1-dependent membrane trafficking relies on Arf1 stimulation, and in turn, the PA produced promotes the membrane recruitment of the coatomer I (COPI) complex (by direct interaction with the b subunit), NSF, the kinesins, and Arf1 itself in a sort of positive feedback cycle aimed at the formation of membrane carriers (Ktistakis et al. 1996, 2003). The PLD product PA can also affect membrane trafficking by acting at the fusion step. Indeed, a recent study carried out with the yeast proteins indicated that PA has a dual role during membrane fusion: it is required for the correct localization of the SNAP25 family proteins, and for the stimulation of the fusion process, at least in vitro (Liu et al. 2007). PA is also a potent activator of PtdIns4P 5-kinase (PI4P5K), the final enzyme of the PtdIns45P2 synthetic pathway (Jenkins et al. 1994). Indeed PLD can stimulate PtdIns45P2 synthesis on isolated Golgi membranes in an Arf1dependent manner (Siddhanta et al. 2000). However, the presence of PtdIns45P2 on the Golgi complex remains a matter of dispute, despite some reports that have shown that Golgi PtdIns45P2 contributes to the recruitment of bIII spectrin and ankyrin, and that PI 5-phosphatase, the enzyme that catalyses the initial dephosphorylation of PtdIns45P2, is expressed on the Golgi membranes (Freyberg et al. 2003). On the whole, this evidence supports the presence of PtdIns45P2 on the Golgi complex. Along with its signalling role, PLD1 could participate in the structure and function of the Golgi complex by substituting PtdCho with PA inside the biological membranes. Thus the two leaflets of cell membranes lose their equilibrium through the change of the cylindrical lipid PtdCho into the conical lipid PA (PA has a smaller polar head), and so to cope with this, the system can establish a new equilibrium by membrane bending (Kooijman et al. 2005). In line with this hypothesis, there is the localized presence of PLD2 on the rims of the Golgi cisterna, where strong curvature is needed (Freyberg et al. 2002). Alternatively, the PA produced on the rims could be part of the recruitment process for proteins involved in membrane trafficking (see above), since the cisterna rims also represent a hot spot for post-Golgi carrier formation.

Phospholipase C The phospholipases C (PLCs) are a family of at least 13 enzymes that can be divided into six classes: PLC b, g, d, e, z, h (Rhee and Choi 1992). These preferentially hydrolyse PtdIns45P2 to generate two important second messengers: diacylglycerol (DAG) and inositol 1,4,5-trisphosphate (Ins145P3) (Rhee and Bae 1997). These PLCs are generally part of the signalling pathways initiated by GPCRs and growth-factor receptors on the plasma membrane (Exton 1996). Intracellularly, the PLCs are localized in the cytosol, on the plasma membrane (where the major pool of PtdIns45P2 is located), in the nucleus and on endomembranes, including the endosomal compartment and the Golgi complex (Mazzoni et al. 1992; Bertagnolo et al. 1995; Blayney et al. 1998). The PLC product DAG is required for membrane recruitment of

254

*

M. Sallese

several signalling proteins, including protein kinases (e.g., PKC) and small Gprotein exchange factors (e.g., Ras–GRP1). Ins145P3 binds to and opens ERand Golgi-located Ca2þ channels, which leads to an increase in cytosolic Ca2þ levels and the consequent signalling promoted by this important second messenger. Interestingly, Bikle and co-workers (2007) have reported the presence of PLCg1 in a complex with the Ins145P3 receptor, and of secretory pathway Ca2þ -ATPase1 (SPCA1) on the trans-Golgi of keratinocytes. They thus hypothesized a role for PLCg1 downstream of the Ca2þ -sensor receptor (CaR) in the regulation of the Ca2þ stores of the Golgi complex (see Ca2þ section below). PLCb2 and PLCb3 are also present on the Golgi complex (Diaz Anel 2007). In contrast, the cytosolic PLCe translocates to Golgi membranes in response to Rap1 activation, as a downstream effector of GPCRs and growthfactor receptors (Jin et al. 2001). Rap1 is a small G protein of the Ras superfamily that is mainly localized on the Golgi complex. Originally it was discovered as an inhibitor of Ras functions (cell growth, adhesion; see below), as it competes with Ras for binding to Raf-1, but does not activate Raf-1. However, later it was shown that Rap1 can activate the ERK pathway via B-Raf, and it can even synergise with Ras. The PLCe recruited to Rap1-containing membranes (e.g., the Golgi complex) also acts as a Rap1 guanine nucleotide exchange factor (GEF) through its Cdc25-homology domain, generating a positive feedback loop that is responsible for the sustained activation of Rap1 at the Golgi complex (at least 20 min), as a downstream effect of EGF-receptor stimulation (Jin et al. 2001). We would also expect that a PLC contributing to DAG levels at the Golgi complex could regulate membrane trafficking. Indeed, reduction of DAG levels in the Golgi by down-regulation of the PItransfer protein Nir2 impairs TGN-to-plasma-membrane transport of VSVG while leaving unaltered the bidirectional ER-to-Golgi and intra-Golgi transport steps (Litvak et al. 2005). The importance of DAG in membrane trafficking has been emphasized by a series of studies that have shown that the TGN exit of carriers is dependent on the fissioning protein PKD (Baron and Malhotra 2002). This is also one of the most clear signalling cascades that has been identified on the Golgi complex to date. Briefly, the bg subunit of the heterotrimeric G proteins can activate PLCb3, which by generating DAG recruits PKCh and PKD to Golgi membranes. PKCh also phosphorylates and activates PKD, which in turn, phosphorylates and activates the PI 4-kinases (PI4Ks) (Diaz Anel and Malhotra 2005; Diaz Anel 2007). The PtdIns4P generated by this signalling cascade could lead to the recruitment of several adaptor proteins (four-phosphate-adaptor proteins 1 and 2, FAPP1 and 2; ceramide transfer protein, CERT; adaptor protein complex 1, AP-1,) that are necessary for the formation/fission of carriers at the TGN that are directed towards the plasma membrane (De Matteis et al. 2005). Moreover, in a possible negative feedback loop, PKD can also directly phosphorylate CERT (Fugmann et al. 2007). Ceramide belongs to the pathway that contributes to the maintenance of DAG levels on the Golgi complex via sphingomyelin synthase (Fig. 2). Phosphorylating CERT decreases

Functional cross talk between membrane trafficking and cell signalling

*

255

the transport of ceramide to the Golgi complex, and as a consequence, the amount of DAG (Fig. 2). Finally, this CERT phosphorylation leads to the quenching of the PKD activating pathway. A recent study by Egea and co-workers (2007) has further emphasized the important role of DAG in membrane trafficking. Here they showed a specific role for DAG in the fissioning of COPI-coated buds during retrograde, Golgi-toER, transport. Furthermore, in investigating the potential pathways for DAG generation on the Golgi complex using chemical inhibitors, they showed that at least three DAG-generating pathways participate in the DAG levels at the Golgi complex: the classical PLC pathway; the PA phosphatase (PAP) pathway downstream of PLD; and the sphingomyelin synthase pathway, as a by-product of sphingomyelin (Fig. 2) (Fernandez-Ulibarri et al. 2007). However, the PAP substrate PA can also be derived from acylation of the PLA2 catabolite lysoPA (Fig. 2). This finding suggests that all of the phospholipases taking part in Golgi functioning show cross talk by converging on DAG (Fig. 2).

Calcium and Ras at the Golgi complex Ca2þ is perhaps the most ubiquitous of the second messengers in vertebrates, and even slight variations in its levels can greatly affect cell behaviour. Extracellular Ca2þ concentrations are usually in the low millimolar range, while resting free cytosolic Ca2þ concentrations are in the order of 100 nM. A large amount of intracellular Ca2þ is stored inside the ER and the Golgi complex; both of these organelles can accumulate millimolar levels of luminal Ca2þ (Montero et al. 1995; Pinton et al. 1998). These steep Ca2þ gradients between the organelle lumens and the cytosol, and between the cytosol and the extracellular space, are maintained by a series of ATP-dependent pumps and channels that fine-tune this homeostasis. According to the classical paradigm, receptor activation at the plasma membrane stimulates the release of Ca2þ from the ER stores into the cell cytosol, which then activates downstream signalling proteins. More recently, the Golgi complex has also been considered as part of the intracellular Ca2þ response that can be triggered by extracellular stimuli (Pinton et al. 1998). The ER and Golgi complex release Ca2þ upon activation of the Ins145P3 receptors in their membranes (Berridge 2002; Pinton et al. 1998), while Ca2þ uptake involves two classes of Ca2þ -ATPase pumps, the sarcoplasmic and ER Ca2þ ATPase (SERCA) in the ER and Golgi membranes, and the Golgi-specific Ca2þ ATPase SPCA1 (Missiaen et al. 2007). As well as pumping Ca2þ , SPCA1 also supplies Mn2þ as a cofactor for the Golgi complex glycosyltransferases. The importance of the Golgi complex in cellular Ca2þ homeostasis is highlighted by the skin disorder Hailey–Hailey disease, a keratinocyte disorder that is characterized by cell–cell adhesion and differentiation defects, and that is caused by an inactivating mutation in the SPCA1 Ca2þ -ATPase gene (Hu et al. 2000). The contributions of SERCA and SPCA1 to the homeostasis of Ca2þ in the Golgi complex is also cell-type dependent. Indeed, as keratino-

256

*

M. Sallese

cytes rely almost exclusively on SPCA1, this indicates why Hailey–Hailey disease is cell specific (Callewaert et al. 2003). Fluctuations in the Ca2þ concentrations in the extracellular environment are also monitored by the Ca2þ receptor (CaR). This is a low affinity, seventransmembrane-domain GPCR that is coupled to various heterotrimeric G proteins of the Gi, Gq and G12/13 classes (Ward 2004). In keratinocytes, CaR coupling to the PLC signalling pathway can trigger the release of Ca2þ from intracellular stores, and hence promote cell differentiation. Recent studies have shown that the CaR is also on the TGN, whereby it can sense the luminal Ca2þ concentrations and regulate the Ca2þ uptake into the Golgi complex accordingly, through acting in concert with PLCg1 and SPCA1 (Tu et al. 2007). If this finding can be confirmed in future studies, it represents the first evidence of a Golgi-initiated signalling circuit. Changes in Ca2þ homeostasis within the Golgi lumen or in the cytosol proximal to this organelle can affect the functions of the Golgi complex and cell signalling. Indeed, there is evidence that variations in the levels of cytoplasmic Ca2þ are involved in different transport steps (Chen et al. 2002). Specifically, a role for Ca2þ in endosomal fusion has been reported, as well as in homotypic vacuolar fusion in yeast (Peters and Mayer 1998; Pryor et al. 2000; Mayorga et al. 1993). In addition, Balch and Beckers (1989) identified a Ca2þ -dependent ER-to-Golgi transport step, while Elazar and Porat (2000) used a reconstituted intra-Golgi transport assay to demonstrate that Ca2þ released from the Golgi complex is required for intraGolgi transport of VSVG. These data would thus suggest that Golgi transport is controlled by a micro-signalling circuit that is triggered by the arrival of cargo at the Golgi complex and that leads to the regulation of transport flow (see Chapter 2.8). There is also clear evidence for a role for Ca2þ in constitutive exocytosis and endocytosis in whole cells, as provided by Stamnes and co-workers (2002), in agreement to previous studies performed with purified organelle membranes and semi-intact cell systems. Here they showed that in NRK cells, VSVG transport is impaired by the Ca2þ chelator BAPTA, for the intermediate compartment-to-Golgi and the Golgi-to-plasma-membrane transport steps (Chen et al. 2002). When probed with the Shiga toxin b fragment, the functioning of the endocytic/retrograde pathway was blocked by BAPTA treatment at the endosomal-to-Golgi and Golgi-to-ER interfaces. Ca2þ chelators also promoted the detachment of the COPI protein coat from membranes, providing the first mechanistic explanation of how Ca2þ regulates transport throughout the constitutive exocytic pathway (Chen et al. 2002). As well as these effects on the COPI machinery, Ca2þ can affect membrane trafficking by various means. In the first instance, an impairment of membrane trafficking could result from changes in the intraluminal Ca2þ concentrations. This hypothesis is supported by the presence of a number of Ca2þ -binding proteins (CALNUC, P54/NEFA and Cab45) that appear to be

Functional cross talk between membrane trafficking and cell signalling

*

257

devoted to the control of the luminal Ca2þ concentrations of the Golgi complex (Morel-Huaux et al. 2002; Scherer et al. 1996; Lin et al. 1999; Kawano et al. 2000). Secondly, changes in peri-Golgi Ca2þ concentrations can affect the activities of Golgi-localized Ca2þ -dependent proteins and/or result in the recruitment of cytosolic signalling proteins that are involved in the initiation of signalling cascades similar to those seen at the plasma membrane. For example, Ca2þ increases in the proximity of the Golgi membranes can activate L-CaBP1, a Ca2þ -binding protein that negatively modulates Ca2þ release from the Golgi complex via the Ins145P3 receptor (Haynes et al. 2004). Recently, interactions have also been seen between L-CaBP1 and the AP-1 adaptor, suggesting a specific role of L-CaBP1 in the regulation of transport (Haynes et al. 2006). Furthermore, Ca2þ activates neuronal calcium sensor-1 (NCS-1), a protein that can stimulate PI4KIIIb at the TGN, leading to an increase in constitutive and stimulated transport from the TGN to the plasma membrane (Haynes et al. 2005). NCS-1 binding to Arf1 can also compete with Arf1 in the activation of PI4KIIIb, suggesting the presence of two mutually exclusive pathways acting upstream of PI4KIIIb in the regulation of exit from the TGN (Haynes et al. 2005). A peri-Golgi Ca2þ increase can also activate the cysteine protease calpain, which is involved in many different cellular processes, including cell adhesion and migration. However, a recent study showed that Golgi calpain proteolyses the b-subunit of the COPI coatomer (Hata et al. 2006). This suggests that the activation of calpain by Ca2þ , which could be released from the Golgi complex during the transport of cargo, modulates its own transport by promoting the disassembly of the COPI coatomer. There is a second class of proteins that although not generally localized on the Golgi membranes, they are recruited to the Golgi in response to a cytosolic Ca2þ increase. These proteins include other neuronal calcium-sensor family proteins (hippocalcin, Vilip-1 and neurocalcin-d), cPLA2, K-Ras and Ras-GRP (see below for further details relating to Ras-GRP) (O’Callaghan et al. 2002; Bivona et al. 2003; Evans and Leslie 2004; Lopez-Alcala et al. 2008). Hippocalcin has been recently reported to be involved in the activation of Rasmediated Raf1-activation along the MAPK pathway initiated by N-methyl-Daspartate (NMDA) and KCl (Noguchi et al. 2007). Since the Golgi complex hosts the Ras-dependent MAPK activation pathway, we can hypothesize that the translocation of hippocalcin to the Golgi complex upon increased cytosolic Ca2þ levels allows it to participate in this signalling cascade from this Golgi location. This scenario opens the possibility of cross talk between Ca2þ released from the Golgi complex during membrane trafficking (if any) and signalling coming from the plasma membrane.

Ras The Ras proteins are the founder members of a large family of small GTPbinding proteins, which now includes Arf, Rab and Rho. Ras itself comprises

258

*

M. Sallese

three proteins, N-Ras, H-Ras and K-Ras, which are involved in the regulation of cell proliferation and cell death promoted by plasma-membrane receptors. The Ras family members are involved in cell signalling via three main pathways: Raf1/Erk, PI3K/AKT and RalGDS (Rodriguez-Viciana and McCormick 2005; Rodriguez-Viciana et al. 2004). Mutations in these Ras proteins that impair their GTPase activities are responsible for several human diseases, including cancers. Recent studies have reported a substantial bidirectional trafficking of H-Ras and N-Ras between the Golgi complex and the plasma membrane (Quatela and Philips 2006). Indeed, using genetically encoded fluorescent probes that can sense Ras activation, Philips and co-workers (2002) have demonstrated that growthfactor stimulation transiently activates Ras on the plasma membrane, while its activation is sustained on the Golgi membranes, from where it can promote cell growth or differentiation. Ras activation on the Golgi complex could be due to the retrograde transport of Ras from the plasma membrane, or it could rely on a diffusible mediator that transduces the message from the plasma membrane to the Golgi complex. Although both hypotheses might be valid, the most convincing experimental data indicate that Ca2þ is the diffusible signal involved in the activation of Ras that is already on Golgi membranes. Specifically, an intracellular Ca2þ increase recruits the Ca2þ -dependent Ras exchange factor Ras-GRP1 to the Golgi complex and the RAS GTPase-activating protein (GAP) CAPRI to the plasma membrane, leading to Ras activation at the Golgi complex and its inactivation at the plasma membrane (Bivona and Philips 2003). Of particular importance here, further studies have also shown that the outcome of Ras activation (cell growth versus cell differentiation) depends on its subcellular location (Quatela and Philips 2006; Mor and Philips 2006; Mor et al. 2007a). In T lymphocytes, activation of the T-cell receptor (TCR) promotes cell growth, while that of the lymphocyte-function-associated antigen 1 (LFA-1) receptor regulates cell adhesion, with both pathways involving Ras signalling (Mor et al. 2007b). Thus TCR engagement induces activation of Ras specifically at the Golgi complex, while the co-stimulation of the TCR and the LFA-1 receptor leads to the activation of Ras at the Golgi complex and the plasma membrane (Mor et al. 2007b). Interestingly, both of these signals rely on Ras–GRP1, but the DAG that is produced via the TCR is provided by PLC, while the PLD2 and PAP pathway produces DAG downstream of the LFA-1 receptor. It is not clear whether this PLC activation that occurs downstream of the TCR is localized to the Golgi complex or the plasma membrane. From these examples, it is evident that Ca2þ released from the Golgi complex or the ER can activate the Ras–GRP1/Ras signalling pathway. It would therefore be of importance to investigate whether the Ca2þ present in the peri-Golgi area during membrane trafficking, could do this as well. In addition, understanding whether Ras activation can affect membrane trafficking is also an important aspect that needs further investigation.

Functional cross talk between membrane trafficking and cell signalling

*

259

Additional Golgi signalling Other classes of signalling proteins are known to be present on the membranes of the Golgi complex. These are considered here only very briefly for the sake of space and because they have been recently reviewed by our group (Sallese et al. 2006). The kinases form the most important families of signalling protein, and they are also present on Golgi membranes, from where they can modulate membrane trafficking and cell growth (Sallese et al. 2006).

The cyclic adenosine monophosphate (cAMP)–PKA pathway Protein kinase A (PKA) is by far the most important cAMP effector. PKA is a tetrameric kinase that comprises two regulatory and two catalytic subunits. cAMP induces the dissociation of the inhibitory, regulatory, subunits, thus freeing the active catalytic subunits. cAMP signalling is also believed to rely on functional complexes in restricted areas, so as to limit and optimize their actions in space (Zaccolo et al. 2002). The cAMP ‘signalosome’ is comprised of a GPCR that can promote GTP loading on the alpha stimulatory subunit (Gas) of heterotrimeric G proteins via a conformational change, which in turn activates the enzyme adenylyl cylase for the production of cAMP. The cAMP formed stimulates the phosphorylating activity of PKA as the final effect. As well as the desensitization mechanisms that work at the level of the activated receptor, this pathway is counterbalanced by the action of the phosphodiesterases (PDEs), which catabolise the cAMP that is formed. This signalosome complex is maintained and organized through the actions of a scaffold protein known as AKAP (Beene and Scott 2007). Remarkably, all of these cAMP pathway modules are located on Golgi membranes, strongly suggesting a specific function at this location (Cheng and Farquhar 1976a,b; Maier et al. 1995; Pooley et al. 1997; Martin et al. 1999; Li et al. 2003). From the functional standpoint, cAMP accelerates ER-to-Golgi and Golgi-to-plasma-membrane transport, while PKA inhibitors block Golgito-plasma-membrane transport with minor effects, if any, on ER-to-Golgi transport (Muniz et al. 1996, 1997). Increases in cAMP also alter Golgi morphology, with the generation of tubular networks among the Golgi cisternae (Muniz et al. 1996). Finally, PKA activity is required for retrograde transport, at the endosome–Golgi and Golgi–ER interfaces (Birkeli et al. 2003; Cabrera et al. 2003). At the molecular level, PKA can affect multiple transport steps, since its activity promotes the recruitment of Arf1 to Golgi membranes, which can then be followed by the pleiotropic effects that are known to be downstream of this master regulator of membrane trafficking (Martin et al. 2000). PKA also phosphorylates the KDEL receptor, uncovering its COPIbinding motif, and some of the SNAREs, modulating their ability to support membrane fusion (Cabrera et al. 2003; Hong 2005). Clearly the cAMP–PKA system is involved in multiple trafficking stages. The question thus arises: is the cAMP generated specifically at its site of action, or as cAMP can freely diffuse for hundreds of microns, does it derive from a

260

*

M. Sallese

different cellular compartment? The experimental information available cannot at present distinguish between these two possibilities. However, it would appear more effective to control membrane trafficking in restricted sub-Golgi areas (e.g., CGN, CGN) using sudden local elevations of cAMP generation, in contrast to the need to establish a cAMP gradient that arises elsewhere and that would activate the whole perinuclear area.

Protein kinase C The PKCs are a class of second-messenger-dependent kinases that comprises 10 members that can be grouped into: the classical PKCs (PKCa, b and g) that are DAG and Ca2þ dependent; the novel PKCs (PKC d, e, h, m and q) that are DAG dependent and Ca2þ independent; and the atypical PKCs (PKCz and i/l) that are DAG and Ca2þ independent. Most of these are recruited to Golgi membranes through to their C1 domain, which can bind to DAG, ceramides and arachidonic acid (Schultz et al. 2004). Interestingly, exogenous addition of ceramides, or their generation through stimulation of the IFNg receptor, induces cell apoptosis via the translocation of PKCd and PKCe to the Golgi complex (Schultz et al. 2003; Kajimoto et al. 2001, 2004). The role of PKC in trafficking is, however, still not completely understood. Small chemical PKC inhibitors/activators affect intra-Golgi and TGN-to-plasma-membrane transport, as revealed by their effects on the prototypical cargoes VSVG and the glucosaminoglycans (GAGs) (De Matteis et al. 1993; Fabbri et al. 1994; Buccione et al. 1996). Similar effects have been seen by stimulating the plasma-membrane IgE receptor (Buccione et al. 1996). This work represents an example of the plasma-membrane receptors regulating the Golgi functions, as described in the introduction (second scenario). The regulation of membrane trafficking by PKC activation involves control of the recruitment of the Arf1–COPI machinery to Golgi membranes, although an alternative interpretation has proposed a catalytically independent action of PKC. (De Matteis et al. 1993; Fabbri et al. 1994). More recent work from Tisdale and co-workers has shown important roles for PKCi/l in the sorting of retrograde-directed material from the ER–Golgi compartment, through actions on COPI and GAPDH (Tisdale and Artalejo 2006; Tisdale 2000; Tisdale et al. 2004). In conclusion, the PKCs at the Golgi complex serve the dual roles of amembraneraffickingregulatorandanapoptosismediator.Nodataareavailable to support traffic-dependent activation of the PKCs.

Heterotrimeric G proteins: key regulators or simple passengers? The heterotrimeric G proteins are the family of GTP-binding proteins that transduce the downstream signalling from the seven transmembrane domain receptors (the GPCRs). They are formed from three polypeptides, the Gbg dimer and the Ga subunit, whereby the characteristics of this last define the

Functional cross talk between membrane trafficking and cell signalling

*

261

group classification of the whole trimer, in terms of its specific functional subfamily. Members of the Gai, Gaq, Gas and Gaz subfamilies have been clearly demonstrated to be on the Golgi complex (Denker et al. 1996; Helms et al. 1998). Some studies have proposed that these G proteins can be visualized on the Golgi complex since the trimer, and probably the whole signalosome that also includes a GPCR, is assembled and post-translationally modified on the Golgi membranes (Marrari et al. 2007). In contrast to this view, manipulation of the expression levels or the activity states of the heterotrimeric G-proteins influences the functionality of the secretory system. Thus, addition of the Gai activator mastoparan or of an excess of Gbg (which will titrate out the Gai subunit) in permeabilized cells results in a block of VSVG exit from the ER, which would suggest that Gai is required for this transport step (Schwaninger et al. 1992). In polarized cells, membrane trafficking from the TGN to the plasma membrane follows two separate routes, which are directed toward the apical and the basolateral membranes (Lipschutz et al. 2001). It appears that Gai is specifically required only for the basolateral-directed cargoes, since pertussis toxin treatment (a Gai-inactivating toxin) blocks only this pathway (Pimplikar and Simons 1993a,b). In contrast, Gas is important for apical, and not basolateral, transport (Pimplikar and Simons 1993a,b). Gai2 and Gaz appear to be involved in the maintenance of the structure of the Golgi complex, since their over-expression can counteract the actions of the Golgi disrupting agent, nordihydroguaiaretic acid; in addition, Gaz inactivation induces the disassembly of the Golgi complex (Yamaguchi et al. 2000). The regulators of G-protein signalling (RGSs) are a relatively recently discovered family of multi-domain proteins that have been implicated in various signalling pathways (Willars 2006). They all share an RGS domain, which can stimulate the GTPase activities of the G proteins, and thus their inactivation. RGS proteins are present on the Golgi complex (RGS–GAIP) and on Golgi-derived carriers (RGS4) (Sullivan et al. 2000; Wylie et al. 1999, 2003). The specific function of GAIP remains to be defined, although RGS4 is known to directly interact with COPI to inhibit the transport of aquaporin and alkaline phosphatase, probably by sequestering COPI from the Golgi membranes (Sullivan et al. 2000). In conclusion, there are several lines of evidence that support roles for heterotrimeric G-protein signalling in secretory transport and maintenance of Golgi morphology, although a potential role of the Golgi complex in their transport and assembly cannot be ignored. What appears to be lacking, instead, is a coherent picture of the stimulus–receptor–signalling–effector pathways in which the heterotrimeric G proteins are involved.

Conclusions The relatively small number of signalling proteins that have been considered above represent only the proof of concept that there exist on the Golgi

262

*

M. Sallese

complex classical signalling pathways that can regulate its functioning. Numerous other signalling proteins have also been reported to be on Golgi membranes, although, as yet, without any apparent roles. In addition, it has been shown that alterations in the expression/function of signalling proteins can affect the behaviour of the Golgi complex, although these might still be the results of indirect effects exerted in another cellular compartment. All of these data firmly point to roles of the classical ‘plasma-membrane’ signalling pathways superimposed on the basic trafficking machinery. We can conclude that experimental studies strongly support the concept that signals originating from the plasma membrane converge on the Golgi and can have effects on membrane trafficking. However, these signals also use the Golgi membranes as a platform to affect functions other than secretion. In contrast, there remain very few indications for traffic-triggered signalling cascades. With many signalling proteins that have been shown to be present on the Golgi complex and to influence membrane trafficking, it is difficult to envisage their primary activation at the plasma membrane; their functioning would indeed fit better with a local (Golgi complex) activatory loop. The first clear evidence of this was provided by the unfolded protein response (UPR), whereby the presence of unfolded proteins inside the ER can stimulate three sensor receptors that are localized in the ER membranes, thus initiating the signalling pathways that lead to the removal of the problem. We therefore await the evidence for similar pathways that operate on the Golgi complex. We are confident that in the near future we will understand a lot more about the information flow and involvement of these Golgi-located signalling proteins. Thus, it would be important to understand what are the signals that trigger Golgi-based cascades? What is the sensor receptors involved? What are the molecular mechanisms in the amplification/integration of these cascades? What are the effectors of the primary signals to produce a specific phenotype?

References Asirvatham AL, Galligan SG, Schillace RV, Davey MP, Vasta V, Beavo JA, Carr DW (2004) A-kinase anchoring proteins interact with phosphodiesterases in T lymphocyte cell lines. J Immunol 173: 4806–4814 Baron CL, Malhotra V (2002) Role of diacylglycerol in PKD recruitment to the TGN protein transport to the plasma membrane. Science 295: 325–328 Beckers CJ, Balch WE (1989) Calcium GTP: essential components in vesicular trafficking between the endoplasmic reticulum Golgi apparatus. J Cell Biol 108: 1245–1256 Beene DL, Scott JD (2007) A-kinase anchoring proteins take shape. Curr Opin Cell Biol 19: 192–198 Berridge MJ (2002) The endoplasmic reticulum: a multifunctional signaling organelle. Cell Calcium 32: 235–249 Bertagnolo V, Mazzoni M, Ricci D, Carini C, Neri LM, Previati M, Capitani S (1995) Identification of PI-PLC beta 1, gamma 1, and delta 1 in rat liver: subcellular distribution relationship to inositol lipid nuclear signalling. Cell Signal 7: 669–678

Functional cross talk between membrane trafficking and cell signalling

*

263

Birkeli KA, Llorente A, Torgersen ML, Keryer G, Tasken K, Sandvig K (2003) Endosometo-Golgi transport is regulated by protein kinase A type II alpha. J Biol Chem 278: 1991–1997 Bivona TG, Perez-De Castro I, Ahearn IM, Grana TM, Chiu VK, Lockyer PJ, Cullen PJ, Pellicer A, Cox AD, Philips MR (2003) Phospholipase Cgamma activates Ras on the Golgi apparatus by means of RasGRP1. Nature 424: 694–698 Bivona TG, Philips MR (2003) Ras pathway signaling on endomembranes. Curr Opin Cell Biol 15: 136–142 Blackwood RA, Transue AT, Harsh DM, Brower RC, Zacharek SJ, Smolen JE, Hessler RJ (1996) PLA2 promotes fusion between PMN-specific granules complex liposomes. J Leukocyte Biol 59: 663–670 Blayney L, Gapper P, Rix C (1998) Identification of phospholipase C beta isoforms their location in cultured vascular smooth muscle cells of pig, human and rat. Cardiovasc Res 40: 564–572 Brown HA, Gutowski S, Moomaw CR, Slaughter C, Sternweis PC (1993) ADP-ribosylation factor, a small GTP-dependent regulatory protein, stimulates phospholipase D activity Cell 75: 1137–1144 Brown WJ, Chambers K, Doody A (2003) Phospholipase A2 (PLA2) enzymes in membrane trafficking: mediators of membrane shape function. Traffic 4: 214–221 Buccione R, Bannykh S, Santone I, Baldassarre M, Facchiano F, Bozzi Y, Di Tullio G, Mironov A, Luini A, De Matteis MA (1996) Regulation of constitutive exocytic transport by membrane receptors. A biochemical and morphometric study. J Biol Chem 271: 3523–3533 Cabrera M, Muniz M, Hidalgo J, Vega L, Martin ME, Velasco A (2003) The retrieval function of the KDEL receptor requires PKA phosphorylation of its C-terminus. Mol Biol Cell 14: 4114–4125 Callewaert G, Parys JB, De SH, Raeymaekers L, Wuytack F, Vanoevelen J, Van BK, Simoni A, Rizzuto R, Missiaen L (2003) Similar Ca(2þ)-signaling properties in keratinocytes and in COS-1 cells overexpressing the secretory-pathway Ca(2þ)-ATPase SPCA1. Cell Calcium 34: 157–162 Chen JL, Ahluwalia JP, Stamnes M (2002) Selective effects of calcium chelators on anterograde retrograde protein transport in the cell. J Biol Chem 277: 35682–35687 Cheng H, Farquhar MG (1976a) Presence of adenylate cyclase activity in Golgi and other fractions from rat liver. I Biochemical determination. J Cell Biol 70: 660–670 Cheng H, Farquhar MG (1976b) Presence of adenylate cyclase activity in Golgi and other fractions from rat liver. II Cytochemical localization within Golgi and ER membranes. J Cell Biol 70: 671–684 Chiu VK, Bivona T, Hach A, Sajous JB, Silletti J, Wiener H, Johnson RL, Cox AD, Philips MR, (2002) Ras signalling on the endoplasmic reticulum and the Golgi. Nat Cell Biol 4: 343–350 Colanzi A, Sutterlin C, Malhotra V (2003) RAF1-activated MEK1 is found on the Golgi apparatus in late prophase and is required for Golgi complex fragmentation in mitosis. J Cell Biol 161: 27–32 De Figueiredo P, Doody A, Polizotto RS, Drecktrah D, Wood S, Banta M, Strang MS, Brown WJ, (2001) Inhibition of transferrin recycling and endosome tubulation by phospholipase A2 antagonists. J Biol Chem 276: 47361–47370 De Figueiredo P, Drecktrah D, Katzenellenbogen JA, Strang M, Brown WJ (1998) Evidence that phospholipase A2 activity is required for Golgi complex and trans Golgi network membrane tubulation. Proc Natl Acad Sci USA 95: 8642–8647 De Figueiredo P, Drecktrah D, Polizotto RS, Cole NB, Lippincott-Schwartz J, Brown WJ, (2000) Phospholipase A2 antagonists inhibit constitutive retrograde membrane traffic to the endoplasmic reticulum. Traffic 1: 504–511

264

*

M. Sallese

De Figueiredo P, Polizotto RS, Drecktrah D, Brown WJ, (1999) Membrane tubulemediated reassembly and maintenance of the Golgi complex is disrupted by phospholipase A2 antagonists. Mol Biol Cell 10: 1763–1782 De Matteis MA, Di CA, Godi A (2005) The role of the phosphoinositides at the Golgi complex. Biochim Biophys Acta 1744: 396–405 De Matteis MA, Santini G, Kahn RA, Di Tullio G, Luini A (1993) Receptor and protein kinase C-mediated regulation of ARF binding to the Golgi complex. Nature 364: 818–821 Denker SP, McCaffery JM, Palade GE, Insel PA, Farquhar MG (1996) Differential distribution of alpha subunits beta gamma subunits of heterotrimeric G proteins on Golgi membranes of the exocrine pancreas. J Cell Biol 133: 1027–1040 Di Fiore PP, De Camilli P (2001) Endocytosis signaling. An inseparable partnership. Cell 106: 1–4 Diaz Anel AM (2007) Phospholipase C beta3 is a key component in the Gbetagamma/ PKCeta/PKD-mediated regulation of trans-Golgi network to plasma membrane transport. Biochem J 406: 157–165 Diaz Anel AM, Malhotra V (2005) PKCeta is required for beta1gamma2/beta3gamma2and PKD-mediated transport to the cell surface the organization of the Golgi apparatus. J Cell Biol 169: 83–91 Donaldson JG, Lippincott-Schwartz J (2000) Sorting signaling at the Golgi complex. Cell 101: 693–696 Evans JH, Leslie CC (2004) The cytosolic phospholipase A2 catalytic domain modulates association and residence time at Golgi membranes. J Biol Chem 279: 6005–6016 Exton JH (1996) Regulation of phosphoinositide phospholipases by hormones, neurotransmitters, and other agonists linked to G proteins. Annu Rev Pharmacol Toxicol 36: 481–509 Fabbri M, Bannykh S, Balch WE (1994) Export of protein from the endoplasmic reticulum is regulated by a diacylglycerol/phorbol ester binding protein. J Biol Chem 269: 26848–26857 Fernandez-Ulibarri I, Vilella M, Lazaro-Dieguez F, Sarri E, Martinez SE, Jimenez N, Claro E, Merida I, Burger KN, Egea G (2007) Diacylglycerol is required for the formation of COPI vesicles in the Golgi-to-ER transport pathway. Mol Biol Cell 18: 3250–3263 Freyberg Z, Bourgoin S, Shields D (2002) Phospholipase D2 is localized to the rims of the Golgi apparatus in mammalian cells. Mol Biol Cell 13: 3930–3942 Freyberg Z, Siddhanta A, Shields D (2003) “Slip, sliding away”: phospholipase D and the Golgi apparatus. Trends Cell Biol 13: 540–546 Freyberg Z, Sweeney D, Siddhanta A, Bourgoin S, Frohman M, Shields D (2001) Intracellular localization of phospholipase D1 in mammalian cells. Mol Biol Cell 12: 943–955 Fugmann T, Hausser A, Schoffler P, Schmid S, Pfizenmaier K, Olayioye MA (2007) Regulation of secretory transport by protein kinase D-mediated phosphorylation of the ceramide transfer protein. J Cell Biol 178: 15–22 Hammond SM, Jenco JM, Nakashima S, Cadwallader K, Gu Q, Cook S, Nozawa Y, Prestwich GD, Frohman MA, Morris AJ (1997) Characterization of two alternately spliced forms of phospholipase D1. Activation of the purified enzymes by phosphatidylinositol 4,5-bisphosphate, ADP-ribosylation factor, and Rho family monomeric GTP-binding and proteins protein kinase C-alpha. J Biol Chem 272: 3860–3868 Hashimoto Y, Okiyoneda T, Harada K, Ueno K, Sugahara T, Yamashita A, Shuto T, Suico MA, Kai H (2008) Phosphatidic acid metabolism regulates the intracellular trafficking and retrotranslocation of CFTR. Biochim Biophys Acta 1783: 153–162 Hata S, Koyama S, Kawahara H, Doi N, Maeda T, Toyama-Sorimachi N, Abe K, Suzuki K, Sorimachi H (2006) Stomach-specific calpain, nCL-2, localizes in mucus cells and

Functional cross talk between membrane trafficking and cell signalling

*

265

proteolyzes the beta-subunit of coatomer complex, beta-COP. J Biol Chem 281: 11214–11224 Haynes LP, Fitzgerald DJ, Wareing B, O’Callaghan DW, Morgan A, Burgoyne RD (2006) Analysis of the interacting partners of the neuronal calcium-binding proteins L-CaBP1, hippocalcin, NCS-1 and neurocalcin delta. Proteomics 6: 1822–1832 Haynes LP, Tepikin AV, Burgoyne RD (2004) Calcium-binding protein 1 is an inhibitor of agonist-evoked, inositol 1,4,5-trisphosphate-mediated calcium signaling. J Biol Chem 279: 547–555 Haynes LP, Thomas GM, Burgoyne RD (2005) Interaction of neuronal calcium sensor-1 and ADP-ribosylation factor 1 allows bidirectional control of phosphatidylinositol 4kinase beta trans-Golgi network-plasma membrane traffic. J Biol Chem 280: 6047–6054 Helms JB, Helms-Brons D, Brugger B, Gkantiragas I, Eberle H, Nickel W, Nurnberg B, Gerdes HH, Wiel FT (1998) A putative heterotrimeric G protein inhibits the fusion of COPI-coated vesicles. Segregation of heterotrimeric G proteins from COPI-coated vesicles. J Biol Chem 273: 15203–15208 Hong W (2005) SNAREs traffic. Biochim Biophys Acta 1744: 493–517 Hu Z, Bonifas JM, Beech J, Bench G, Shigihara T, Ogawa H, Ikeda S, Mauro T, Epstein EH Jr (2000) Mutations in ATP2C1, encoding a calcium pump, cause Hailey–Hailey disease. Nat Genet 24: 61–65 Imanishi T, Itoh T, Suzuki Y, O’Donovan C, Fukuchi S, Koyanagi KO, Barrero RA, Tamura T, Yamaguchi-Kabata Y, Tanino M, Yura K, Miyazaki S, Ikeo K, Homma K, Kasprzyk A, Nishikawa T, Hirakawa M, Thierry-Mieg J, Thierry-Mieg D, Ashurst J, Jia L, Nakao M, Thomas MA, Mulder N, Karavidopoulou Y, Jin L, Kim S, Yasuda T, Lenhard B, Eveno E, Suzuki Y, Yamasaki C, Takeda J, Gough C, Hilton P, Fujii Y, Sakai H, Tanaka S, Amid C, Bellgard M, Bonaldo MF, Bono H, Bromberg SK, Brookes AJ, Bruford E, Carninci P, Chelala C, Couillault C, De Souza SJ, Debily MA, Devignes MD, Dubchak I, Endo T, Estreicher A, Eyras E, Fukami-Kobayashi K, Gopinath GR, Graudens E, Hahn Y, Han M, Han ZG, Hanada K, Hanaoka H, Harada E, Hashimoto K, Hinz U, Hirai M, Hishiki T, Hopkinson I, Imbeaud S, Inoko H, Kanapin A, Kaneko Y, Kasukawa T, Kelso J, Kersey P, Kikuno R, Kimura K, Korn B, Kuryshev V, Makalowska I, Makino T, Mano S, Mariage-Samson R, Mashima J, Matsuda H, Mewes HW, Minoshima S, Nagai K, Nagasaki H, Nagata N, Nigam R, Ogasawara O, Ohara O, Ohtsubo M, Okada N, Okido T, Oota S, Ota M, Ota T, Otsuki T, Piatier-Tonneau D, Poustka A, Ren SX, Saitou N, Sakai K, Sakamoto S, Sakate R, Schupp I, Servant F, Sherry S, Shiba R, Shimizu N, Shimoyama M, Simpson AJ, Soares B, Steward C, Suwa M, Suzuki M, Takahashi A, Tamiya G, Tanaka H, Taylor T, Terwilliger JD, Unneberg P, Veeramachaneni V, Watanabe S, Wilming L, Yasuda N, Yoo HS, Stodolsky M, Makalowski W, Go M, Nakai K, Takagi T, Kanehisa M, Sakaki Y, Quackenbush J, Okazaki Y, Hayashizaki Y, Hide W, Chakraborty R, Nishikawa K, Sugawara H, Tateno Y, Chen Z, Oishi M, Tonellato P, Apweiler R, Okubo K, Wagner L, Wiemann S, Strausberg RL, Isogai T, Auffray C, Nomura N, Gojobori T, Sugano S (2004) Integrative annotation of 21,037 human genes validated by full-length cDNA clones. PLoS Biol 2: 856–875 Jenkins GH, Fisette PL, Anderson RA (1994) Type I phosphatidylinositol 4-phosphate 5kinase isoforms are specifically stimulated by phosphatidic acid. J Biol Chem 269: 11547–11554 Jin TG, Satoh T, Liao Y, Song C, Gao X, Kariya K, Hu CD, Kataoka T (2001) Role of the CDC25 homology domain of phospholipase Cepsilon in amplification of Rap1dependent signaling. J Biol Chem 276: 30301–30307 Kajimoto T, Ohmori S, Shirai Y, Sakai N, Saito N (2001) Subtype-specific translocation of the delta subtype of protein kinase C and its activation by tyrosine phosphorylation induced by ceramide in HeLa cells. Mol Cell Biol 21: 1769–1783

266

*

M. Sallese

Kajimoto T, Shirai Y, Sakai N, Yamamoto T, Matsuzaki H, Kikkawa U, Saito N (2004) Ceramide-induced apoptosis by translocation, phosphorylation, and activation of protein kinase Cdelta in the Golgi complex. J Biol Chem 279: 12668–12676 Kawano J, Kotani T, Ogata Y, Ohtaki S, Takechi S, Nakayama T, Sawaguchi A, Nagaike R, Oinuma T, Suganuma T (2000) CALNUC (nucleobindin) is localized in the Golgi apparatus in insect cells. Eur J Cell Biol 79: 208–217 Kooijman EE, Chupin V, Fuller NL, Kozlov MM, De KB, Burger KN, Rand PR (2005) Spontaneous curvature of phosphatidic acid and lysophosphatidic acid. Biochemistry 44: 2097–2102 Ktistakis NT, Brown HA, Waters MG, Sternweis PC, Roth MG (1996) Evidence that phospholipase D mediates ADP ribosylation factor-dependent formation of Golgi coated vesicles. J Cell Biol 134: 295–306 Ktistakis NT, Delon C, Manifava M, Wood E, Ganley I, Sugars JM (2003) Phospholipase D1 and potential targets of its hydrolysis product, phosphatidic acid. Biochem Soc Trans 31: 94–97 Lavoie C, Chevet E, Roy L, Tonks NK, Fazel A, Posner BI, Paiement J, Bergeron JJ (2000) Tyrosine phosphorylation of p97 regulates transitional endoplasmic reticulum assembly in vitro. Proc Natl Acad Sci USA 97: 13637–13642 Li H, Adamik R, Pacheco-Rodriguez G, Moss J, Vaughan M (2003) Protein kinase Aanchoring (AKAP) domains in brefeldin A-inhibited guanine nucleotide-exchange protein 2 (BIG2) Proc Natl Acad Sci USA 100: 1627–1632 Lin P, Yao Y, Hofmeister R, Tsien RY, Farquhar MG (1999) Overexpression of CALNUC (nucleobindin) increases agonist and thapsigargin releasable Ca2þ storage in the Golgi. J Cell Biol 145: 279–289 Lippincott-Schwartz J, Snapp E, Kenworthy A (2001) Studying protein dynamics in living cells. Nat Rev Mol Cell Biol 2: 444–456 Lipschutz JH, O’Brien LE, Altschuler Y, Avrahami D, Nguyen Y, Tang K, Mostov KE (2001) Analysis of membrane traffic in polarized epithelial cells. Curr Protoc Cell Biol Chapter 15, Unit 15.5 Litvak V, Dahan N, Ramach ran,S, Sabanay H, Lev S (2005) Maintenance of the diacylglycerol level in the Golgi apparatus by the Nir2 protein is critical for Golgi secretory function. Nat Cell Biol 7: 225–234 Liu S, Wilson KA, Rice-Stitt T, Neiman AM, McNew JA (2007) In vitro fusion catalyzed by the sporulation-specific t-SNARE light-chain Spo20p is stimulated by phosphatidic acid. Traffic 8: 1630–1643 Lopez-Alcala C, varez-Moya B, Villalonga P, Calvo M, Bachs O, Agell N (2008) Identification of essential interacting elements in K-Ras/Calmodulin binding its role in K-Ras localization. J Biol Chem 283: 10621–10631 Maier O, Ehmsen E, Westermann P (1995) Trimeric G protein alpha subunits of the Gs Gi families localized at the Golgi membrane. Biochem Biophys Res Commun 208: 135–143 Marrari Y, Crouthamel M, Irannejad R, Wedegaertner PB (2007) Assembly and trafficking of heterotrimeric G proteins. Biochemistry 46: 7665–7677 Martin ME, Hidalgo J, Rosa JL, Crottet P, Velasco A (2000) Effect of protein kinase A activity on the association of ADP-ribosylation factor 1 to Golgi membranes. J Biol Chem 275: 19050–19059 Martin ME, Hidalgo J, Vega FM, Velasco A (1999) Trimeric G proteins modulate the dynamic interaction of PKAII with the Golgi complex. J Cell Sci 112 ( Pt 22): 3869–3878 Mayorga LS, Colombo MI, Lennartz M, Brown EJ, Rahman KH, Weiss R, Lennon PJ, Stahl PD (1993) Inhibition of endosome fusion by phospholipase A2 (PLA2) inhibitors points to a role for PLA2 in endocytosis. Proc Natl Acad Sci USA 90: 10255–10259 Mazzoni M, Bertagnolo V, Neri LM, Carini C, Marchisio M, Milani D, Manzoli FA, Capitani S (1992) Discrete subcellular localization of phosphoinositidase C beta,

Functional cross talk between membrane trafficking and cell signalling

*

267

gamma delta in PC12 rat pheochromocytoma cells. Biochem Biophys Res Commun 187: 114–120 Michaelson D, Ahearn I, Bergo M, Young S, Philips M (2002) Membrane trafficking of heterotrimeric G proteins via the endoplasmic reticulum Golgi. Mol Biol Cell 13: 3294–3302 Missiaen L, Dode L, Vanoevelen J, Raeymaekers L, Wuytack F (2007) Calcium in the Golgi apparatus. Cell Calcium 41: 405–416 Montero M, Brini M, Marsault R, Alvarez J, Sitia R, Pozzan T, Rizzuto R (1995) Monitoring dynamic changes in free Ca2þ concentration in the endoplasmic reticulum of intact cells. EMBO J 14: 5467–5475 Mor A, Campi G, Du G, Zheng Y, Foster DA, Dustin ML, Philips MR (2007a) The lymphocyte function-associated antigen-1 receptor costimulates plasma membrane Ras via phospholipase D2. Nat Cell Biol 9: 713–719 Mor A, Dustin ML, Philips MR (2007b) Small GTPases LFA-1 reciprocally modulate adhesion and signaling. Immunol Rev 218: 114–125 Mor A, Philips MR (2006) Compartmentalized Ras/MAPK signaling. Annu Rev Immunol 24: 771–800 Morel-Huaux VM, Pypaert M, Wouters S, Tartakoff AM, Jurgan U, Gevaert K, Courtoy PJ (2002) The calcium-binding protein p54/NEFA is a novel luminal resident of medial Golgi cisternae that traffics independently of mannosidase II. Eur J Cell Biol 81: 87–100 Muniz M, Alonso M, Hidalgo J, Velasco A (1996) A regulatory role for cAMP-dependent protein kinase in protein traffic along the exocytic route. J Biol Chem 271: 30935–30941 Muniz M, Martin ME, Hidalgo J, Velasco A (1997) Protein kinase A activity is required for the budding of constitutive transport vesicles from the trans-Golgi network. Proc Natl Acad Sci USA 94: 14461–14466 Noguchi H, Kobayashi M, Miwa N, Takamatsu K (2007) Lack of hippocalcin causes impairment in Ras/extracellular signal-regulated kinase cascade via a Raf-mediated activation process. J Neurosci Res 85: 837–844 O’Callaghan DW, Ivings L, Weiss JL, Ashby MC, Tepikin AV, Burgoyne RD (2002) Differential use of myristoyl groups on neuronal calcium sensor proteins as a determinant of spatio-temporal aspects of Ca2þ signal transduction. J Biol Chem 277: 14227–14237 Ohno Y, Kihara A, Sano T, Igarashi Y (2006) Intracellular localization tissue-specific distribution of human and yeast DHHC cysteine-rich domain-containing proteins. Biochim Biophys Acta 1761: 474–483 Peters C, Mayer A (1998) Ca2þ calmodulin signals the completion of docking and triggers a late step of vacuole fusion. Nature 396: 575–580 Pimplikar SW, Simons K (1993a) Regulation of apical transport in epithelial cells by a Gs class of heterotrimeric G protein. Nature 362: 456–458 Pimplikar SW, Simons K (1993b) Role of heterotrimeric G proteins in polarized membrane transport. J Cell Sci Suppl 17: 27–32 Pinton P, Pozzan T, Rizzuto R (1998) The Golgi apparatus is an inositol 1,4,5-trisphosphate-sensitive Ca2þ store, with functional properties distinct from those of the endoplasmic reticulum. EMBO J 17: 5298–5308 Polizotto RS, De FP, Brown WJ (1999) Stimulation of Golgi membrane tubulation and retrograde trafficking to the ER by phospholipase A(2) activating protein (PLAP) peptide. J Cell Biochem 74: 670–683 Pooley L, Shakur Y, Rena G, Houslay MD (1997) Intracellular localization of the PDE4A cAMP-specific phosphodiesterase splice variant RD1 (RNPDE4A1A) in stably transfected human thyroid carcinoma FTC cell lines. Biochem J 321( Pt 1): 177–185 Porat A, Elazar Z (2000) Regulation of intra-Golgi membrane transport by calcium. J Biol Chem 275: 29233–29237

268

*

M. Sallese

Powner DJ, Wakelam MJ (2002) The regulation of phospholipase D by inositol phospholipids and small GTPases. FEBS Lett 531: 62–64 Pryor PR, Mullock BM, Bright NA, Gray SR, Luzio JP (2000) The role of intraorganellar Ca (2þ) in late endosome–lysosome heterotypic fusion in the reformation of lysosomes from hybrid organelles. J Cell Biol 149: 1053–1062 Pulvirenti T, Giannotta M, Capestrano M, Capitani M, Pisanu A, Polishchuk RS, San Pictro E, Beznoussenko GV, Mironov AA, Turacchio G, Hsu VW, Sallese M, Luini A (2008) A traffic-activated Golgi-based signalling circuit coordinates the secretary pathway. Nat Cell Biol (In press) Quatela SE, Philips MR (2006) Ras signaling on the Golgi. Curr Opin Cell Biol 18: 162–167 Rhee SG, Bae YS (1997) Regulation of phosphoinositide-specific phospholipase C isozymes. J Biol Chem 272: 15045–15048 Rhee SG, Choi KD (1992) Multiple forms of phospholipase C isozymes and their activation mechanisms. Adv Second Messenger Phosphoprotein Res 26: 35–61 Rodriguez-Viciana P, McCormick F (2005) RalGDS comes of age. Cancer Cell 7: 205–206 Rodriguez-Viciana P, Sabatier C, McCormick F (2004) Signaling specificity by Ras family GTPases is determined by the full spectrum of effectors they regulate. Mol Cell Biol 24: 4943–4954 Sallese M, Pulvirenti T, Luini A (2006) The physiology of membrane transport and endomembrane-based signalling. EMBO J 25: 2663–2673 Scherer PE, Lederkremer GZ, Williams S, Fogliano M, Baldini G, Lodish HF (1996) Cab45, a novel (Ca2þ)-binding protein localized to the Golgi lumen. J Cell Biol 133: 257–268 Schultz A, Jonsson JI, Larsson C (2003) The regulatory domain of protein kinase Ctheta localises to the Golgi complex and induces apoptosis in neuroblastoma and Jurkat cells. Cell Death Differ 10: 662–675 Schultz A, Ling M, Larsson C (2004) Identification of an amino acid residue in the protein kinase C C1b domain crucial for its localization to the Golgi network. J Biol Chem 279: 31750–31760 Schwaninger R, Plutner H, Bokoch GM, Balch WE (1992) Multiple GTP-binding proteins regulate vesicular transport from the ER to Golgi membranes. J Cell Biol 119: 1077–1096 Sciorra VA, Rudge SA, Prestwich GD, Frohman MA, Engebrecht J, Morris AJ (1999) Identification of a phosphoinositide binding motif that mediates activation of mammalian and yeast phospholipase D isoenzymes. EMBO J 18: 5911–5921 Siddhanta A, Backer JM, Shields D (2000) Inhibition of phosphatidic acid synthesis alters the structure of the Golgi apparatus and inhibits secretion in endocrine cells. J Biol Chem 275: 12023–12031 Sorkin A, Von Zastrow M (2002) Signal transduction and endocytosis: close encounters of many kinds. Nat Rev Mol Cell Biol 3: 600–614 Sullivan BM, Harrison-Lavoie KJ, Marshansky V, Lin HY, Kehrl JH, Ausiello DA, Brown D, Druey KM (2000) RGS4 and RGS2 bind coatomer and inhibit COPI association with Golgi membranes intracellular transport. Mol Biol Cell 11: 3155–3168 Tisdale EJ (2000) Rab2 requires PKC iota/lambda to recruit beta-COP for vesicle formation. Traffic 1: 702–712 Tisdale EJ, Artalejo CR (2006) Src-dependent aPKCiota/lambda tyrosine phosphorylation is required for aPKCiota/lambda association with Rab2 glyceraldehyde3-phosphate dehydrogenase on pre-golgi intermediates. J Biol Chem 281: 8436–8442 Tisdale EJ, Kelly C, Artalejo CR (2004) Glyceraldehyde-3-phosphate dehydrogenase interacts with Rab2 and plays an essential role in endoplasmic reticulum to Golgi transport exclusive of its glycolytic activity. J Biol Chem 279: 54046–54052 Tu CL, Chang W, Bikle DD (2007) The role of the calcium sensing receptor in regulating intracellular calcium handling in human epidermal keratinocytes. J Invest Dermatol 127: 1074–1083

Functional cross talk between membrane trafficking and cell signalling

*

269

Ward DT (2004) Calcium receptor-mediated intracellular signalling. Cell Calcium 35: 217–228 Willars GB (2006) Mammalian RGS proteins: multifunctional regulators of cellular signalling. Semin Cell Dev Biol 17: 363–376 Wilson BS, Komuro M, Farquhar MG (1994) Cellular variations in heterotrimeric G protein localization expression in rat pituitary. Endocrinology 134: 233–244 Wylie F, Heimann K, Le TL, Brown D, Rabnott G, Stow JL (1999) GAIP, a Galphai-3-binding protein, is associated with Golgi-derived vesicles and protein trafficking. Am J Physiol 276: C497–C506 Wylie FG, Lock JG, Jamriska L, Khromykh T, D LB, Stow JL (2003) GAIP participates in budding of membrane carriers at the trans-Golgi network. Traffic 4: 175–189 Yamaguchi T, Nagahama M, Itoh H, Hatsuzawa K, Tani K, Tagaya M (2000) Regulation of the Golgi structure by the alpha subunits of heterotrimeric G proteins. FEBS Lett 470: 25–28 Zaccolo M, Magalhaes P, Pozzan T (2002) Compartmentalisation of cAMP and Ca(2þ) signals. Curr Opin Cell Biol 14: 160–166

270

*

G. Egea and R. M. Ríos

The role of the cytoskeleton in the structure and function of the Golgi apparatus Gustavo Egea and Rosa M. Ríos

Introduction Cellular organelles in mammalian cells are individualized membrane entities that often become spherical. The endoplasmic reticulum (ER) and the Golgi apparatus (GA) are exceptions to this rule, as they are respectively made up of a continuous tubular network and a pile of flat disks. Their unique shapes are regulated by molecular elements (Kepes et al. 2005; Levine and Rabouille 2005), including the cytoskeleton. All cytoskeletal elements, together with cytoskeleton-associated motors and non-motor proteins, have a role in the subcellular positioning, biogenesis and function of most organelles, being particularly relevant in the GA. The GA is the central organelle of the eukaryotic secretory pathway. While its basic function is highly conserved, the GA varies greatly in shape and number from one organism to another. In the simplest organisms like budding yeast Saccharomyces cerevisiae, the organelle takes the form of dispersed cisternae or isolated tubular networks (Preuss et al. 1992; Rambourg et al. 2001). Unicellular green alga (Henderson et al. 2007) and many protozoa like Toxoplasma gondii (Pelletier et al. 2002) and Trypanosoma brucei (He 2007; He et al. 2004) contain a single pile of flattened cisternae aligned in parallel. The organization of the GA in this manner is referred to as a Golgi stack, which usually contains two regions: one central and poorly fenestrated (compact) and other lateral and highly fenestrated (non-compact) (Kepes et al. 2005) (see the 3D modelling of a GA stack of a control NRK cell in Fig. 2). In fungi (Mogelsvang et al. 2003; Rossanese et al. 1999), plants (daSilva et al. 2004; Hawes and Satiat-Jeunemaitre 2005) or Drosophila (Kondylis and Rabouille 2003) many separate Golgi stacks are dispersed throughout the cytoplasm. In all these cases, each Golgi stack is associated with a single ER exit site (ERES), forming a secretory unit. In contrast, in most mammalian cells, the GA is a single-copy organelle shaped like a ribbon, containing numerous stacks joined by a tubular network and located near the nucleus (Ladinsky et al. 1999; Rambourg and Clermont 1986). In mammals, Golgi stacks are segregated from the ERES, and the Golgi ribbon is closely associated with the centrosome, the main organizing centre for cytoplasmic microtubules (MTOC) (Rios and Bornens 2003; Saraste and Goud 2007). The cytoskeleton imposes the localization of the GA. Depending on the cellular model, either microtubules (MTs) or actin filaments (AFs) have the

The role of the cytoskeleton in the structure and function of GA

*

271

greater influence (for instance, in mammalian and plant cells, respectively). Historically, the first cytoskeleton element to be linked to the GA and associated membrane trafficking was the MT (Thyberg and Moskalewski 1999). Some time later, both AFs and actin-associated proteins were clearly implicated (for recent reviews see Egea et al. (2006); Lanzetti (2007); Ridley (2006); Smythe and Ayscough (2006); Soldati and Schliwa (2006)), and more recently intermediate filaments (IFs) also appear to interact with Golgi membranes and participate in protein trafficking (Gao and Sztul 2001; Styers et al. 2005; Toivola et al. 2005). In this chapter, we provide a general overview of the structural and functional consequences of the coupling between the cytoskeleton and the GA in various cell models.

Microtubules and the structure and dynamics of the Golgi apparatus Microtubules and the structural integrity of the Golgi apparatus In non-polarized mammalian cells, the GA is closely associated with the centrosome, which is usually located near the nucleus at the cell centre (Rios and Bornens 2003) (Fig. 1). The spatial proximity of the GA and the centrosome has been known since Camilo Golgi’s time, but it has been confirmed by immunofluorescence studies in the last 20 years. The close association with the centrosome is maintained even under conditions where cellular architecture is undergoing major remodelling, which occurs during cell migration (Kupfer et al. 1982), fusion of myoblasts to form myotubes (Ralston 1993; Tassin et al. 1985b), the delivery of lytic granules at the immunological synapse (Stinchcombe et al. 2006), phagocytosis (Eng et al. 2007; Stinchcombe et al. 2006) or neuronal polarization (de Anda et al. 2005). However, this association is broken when MT dynamics is perturbed by drugs like nocodazole (NZ) or taxol (TX) (Sandoval et al. 1984; Wehland et al. 1983), which suggests that the main factor governing Golgi ribbon integrity and localization is the microtubular network (Thyberg and Moskalewski 1985, 1999). Pioneer studies of the role of MTs in the structural organization of the mammalian GA used drugs that favour MT disassembly (Robbins and Gonatas 1964). In the absence of MTs, the Golgi ribbon fragments, giving rise to discrete Golgi elements or mini-stacks, which are dispersed throughout the cell (Fig. 1). However, the GA does not need to be either intact or near the nucleus for protein transport or glycosylation (Rogalski et al. 1984). NZinduced Golgi mini-stacks localize at the peripheral ERES, thus enabling the cell to maintain secretory transport from the ER (Cole et al. 1996; Trucco et al. 2004). Therefore, the GA of mammalian cells lacking MTs resembles the normal state of affairs in plant cells and fungi, where Golgi architecture and function occur without centralization. Recent developments in microscope technology have advanced our understanding of the dynamics of MTs and GA interaction. For example, 3D electron microscope studies have allowed individual MTs to be modelled

272

*

G. Egea and R. M. Ríos

Figure 1. Microtubule–Golgi interaction. Immunofluorescence images of human retinal pigment epithelial cells fixed with methanol and double labelled for tubulin (green) and the Golgi protein GMAP210 (red). Cells were treated either with nocodazole (þNZ; 10 mM/3 h) to depolymerize MTs or with taxol (þTX; 10 mM/5 h) to induce the complete polymerization and stabilization of tubulin into MT bundles. Control, NZ- and TX-treated cells were also processed for electron microscopy analysis. Control cells show the characteristic ribbon-like arrangement of numerous adjacent Golgi stacks localised around centrioles (coloured in red). After NZ or TX treatments, numerous discrete Golgi elements appeared which maintained the characteristic stacked morphology (mini-stacks). Note that no significant differences in the ultrastructure of mini-stacks are seen between the two treatments. In TX-treated cells, Golgi mini-stacks are mostly localised to the cell periphery, whereas those in NZ-treated cells are uniformly distributed throughout the cytoplasm. Bars: 5 mm (immunofluorescence microscopy images) and 200 nm (electron microscopy images).

The role of the cytoskeleton in the structure and function of GA

*

273

and the relationships with cisternae to be analyzed in situ. MTs specifically associate with the first cis-cisterna over long distances. MTs also cross Golgi stacks at multiple points via non-compact regions and cisternal openings (Marsh et al. 2001). Time-lapse microscopy studies have revealed that the overall 3D arrangement of the GA near the centrosome is relatively stable (Presley et al. 1997; Scales et al. 1997; Sciaky et al. 1997) although thin tubules are constantly formed and detached from the lateral portions of the GA. After extending from the GA, the tubules break off and move along MTs to the cell periphery. Some move directly to the ERES and, once attached, they collapse into them delivering Golgi proteins to the ER (Mardones et al. 2006). The motion of membrane elements from the ER to the GA is also critically dependent on MTs. If MTs are depolymerized, Golgi proteins that have recycled back to the ER are exported into pre-Golgi intermediates, which then fail to move to the pericentrosomal region and consequently, remain stationary in the vicinity of ERES. Over long periods of time, these de novo structures acquire a normal Golgi stack morphology and become completely functional for secretion (Trucco et al. 2004). These studies indicate that the ability to form a Golgi stack is an intrinsic property of ER-derived membranes that does not require MTs. However, they are required to link stacks into a single organelle and to ensure its central location around the centrosome. Another interesting aspect is the relationship between the GA and stable MTs. These MTs are characterized by the presence of detyrosinated and/or acetylated tubulin. They have a longer half-life and are more resistant to NZinduced depolymerization (Schulze et al. 1987). Most stable MTs, which often appear short and convoluted under the fluorescence microscope, concentrate around the centrosome and colocalize with the GA (Burgess et al. 1991; Skoufias et al. 1990; Thyberg and Moskalewski 1993). Immunoelectron microscopy further demonstrated a close connection between detyrosinated MTs and vesicles transporting newly synthesized proteins from the ER to the Golgi (Mizuno and Singer 1994). Therefore, it has been proposed that there is a reciprocal relationship between MT stabilization and Golgi membrane dynamics. More than 10 years later, the molecular mechanisms mediating this relationship are now beginning to be unveiled (see below).

Microtubule-motor proteins In the current view, it is difficult to understand how MTs contribute to the Golgi structure without considering how MTs mediate in Golgi-associated transport functions. In non-polarized cells, MTs are organized in a characteristic radial pattern with minus-ends anchored at the centrosome and plusends extending toward the cell periphery. Since the GA localizes around the centrosome, the predominant-associated motor activity involved should be minus-end directed. In eukaryotic cells, the primary molecular motor for minus-end-directed movements is cytoplasmic dynein 1 (Hook and Vallee 2006). Movement of transport carriers from peripheral ERES to the GA in the cell centre along MTs is mediated by dynein (Corthesy-Theulaz et al. 1992;

274

*

G. Egea and R. M. Ríos

Harada et al. 1998). In contrast, movement of membranes from the GA back to the ER is plus-end-directed and mediated by kinesin-2 (Stauber et al. 2006). Kinesins are a large protein superfamily, most of whose members have plusend-directed activity (Miki et al. 2005). Dynein and kinesin are structurally similar, consisting of two functional parts: a motor domain that reversibly binds to the cytoskeleton and converts chemical energy into motion, and a tail that interacts with cargo either directly or through accessory chains (Caviston and Holzbaur 2006). The mechanisms by which motors interact with a diversity of cargoes and subcellular targeting sites are not completely understood. One candidate factor proposed to link dynein to endomembranes is the multiprotein complex dynactin, an essential activator for most cytoplasmic dynein functions. Dynactin contains 11 subunits organized into an elaborate structure. The best characterized subunits are the Arp1 filament, p150Glued and p50 dynamitin. p150Glued is a dimer that forms a coiled-coil and binds to dynein, the ARP1 filament and MTs. Dynamitin is required for maintaining the integrity of the complex (Schroer 2004). Expression of a dominant negative form of p150Glued or overexpression of dynamitin induces the fragmentation of the GA into multiple dispersed elements (Burkhardt et al. 1997; Quintyne et al. 1999). The movement of transport carriers from the ER to the GA along MT tracks is blocked under these conditions, as occurs in NZ-treated cells (Presley et al. 1997). Most likely, this blockade occurs at the earliest phases of ER protein export, since the Sec23p component the COPII complex interacts directly with dynactin. This interaction would facilitate the formation of transport carriers and their motion to the GA (Watson et al. 2005). It is therefore widely accepted that the dynein/dynactin motor is primarily responsible for ER-toGolgi transport and the localization of the GA in the cell centre. An alternative model postulates that dynactin plays a role in coordinating the activity of opposing MT-motors and in regulating their processivity (Berezuk and Schroer 2007; Deacon et al. 2003; Haghnia et al. 2007; Ross et al. 2006). Supporting this view, dynein and kinesin colocalize in the same membrane structures (Welte 2004). Even more relevant for Golgi dynamics, kinesin-2 interacts with dynactin (Deacon et al. 2003). This interaction appears to involve primarily the non-motor subunit of kinesin-2 (KAP3) and the p150Glued subunit of dynactin. Knockdown of KAP3 blocked the Golgi-to-ER pathway and disorganized Golgi membranes (Stauber et al. 2006). The observation that kinesin-2 binds the same dynactin subunit as dynein raises the possibility that it could act as a molecular switch that coordinates bidirectional trafficking. Recent data suggest that the scenario could actually be more complicated. Thus, dynein associates with Golgi membranes through several distinct mechanisms and dynactin, via p150Glued, also binds Golgi-associated nonmotor proteins. Dynein intermediate chain directly interacts with huntingtin, which has an important role in vesicle transport. Huntingtin silencing disrupts the GA in HeLa cells (Caviston et al. 2007). ZW10, a mitotic checkpoint protein

The role of the cytoskeleton in the structure and function of GA

*

275

that anchors dynein to kinetochores, also performs important functions in membrane traffic. These include dynein targeting to the GA and other membranes, but also SNARE-mediated ER–Golgi trafficking. ZW10 depletion provokes Golgi dispersal and decreases the frequency of minus-end-directed movements (Arasaki et al. 2006; Hirose et al. 2004; Vallee et al. 2006; Varma et al. 2006). Interestingly, the small GTPases Cdc42 and Rab6 also play a role in regulating motor recruitment to membranes (Chen et al. 2005a; Matanis et al. 2002). Coatomer-bound Cdc42 prevents dynein binding to COPI vesicles, and expression of constitutively active Cdc42 blocks translocation toward the cell centre of NZ-induced stacks and ER-to-Golgi carriers (Chen et al. 2005a). Rab6 family members are involved in some MT-dependent transport steps from the trans-Golgi network (TGN). When Rab6 is activated, BicaudalD1/2 is recruited to the TGN, which in turn recruits dynein–dynactin complexes (Hoogenraad et al. 2001; Matanis et al. 2002). It has been proposed that these complexes could participate in a recycling pathway that begins at the TGN and leads directly to the ER (Young et al. 2005). However, a later study showed that the major target for the fusion of Rab6-containing vesicles is the plasma membrane and that Rab6 regulates the transport and targeting of constitutive secretion vesicles. This study also reports an additional interaction of BicaudalD1/2 with kinesin-1. Therefore, the interactions between Rab6, BicaudalD1/2, kinesin-1 and dynein–dynactin complexes may contribute to the regulation of motor recruitment and co-ordination of their activities to specify directionality (Grigoriev et al. 2007; Saraste and Goud 2007). Defining the relative contributions of all of these mechanisms to the overall regulation of Golgi dynamics will require further investigation.

Role of the Golgi apparatus in microtubule dynamics A new concept is emerging concerning the role of the GA in MT dynamics: the GA acting as a secondary MTOC (Luders and Stearns 2007). The ability of Golgi membranes to assemble and stabilize MTs was first noticed in hepatic cells after NZ treatment (Chabin-Brion et al. 2001). During NZ recovery, short MTs were invariably seen to associate with Golgi mini-stacks. In addition, purified Golgi membranes were shown to contain a-, b- and g-tubulin and to support MT nucleation. However, this study did not resolve whether the MT nucleation was primarily carried out by the centrosome (the MTs then being released and anchored to Golgi membranes) or directly by the GA. Experimental support for the latter hypothesis came from the analysis of the mechanisms regulating the centering of a radial array of MTs in cells lacking centrosomes (Malikov et al. 2005). In stimulated cytoplasmic fragments of melanophores, pigment granules form a central aggregate that becomes the focal point from which MTs radiate. Radial MT arrays also form and become centralized in centrosome-free cytoplasts obtained from non-pigment cells. Strikingly, the GA appeared to be located in the centre of the cytoplasts, close to the MT aster (Malikov et al. 2004). Recently, the GA has been unambigu-

276

*

G. Egea and R. M. Ríos

ously identified as an MTOC by laser ablation of the centrosome (Efimov et al. 2007). MT-nucleation at the GA was shown to require g-tubulin complexes. To date, two g-tubulin-interacting proteins have been associated with the GA: GMAP210 and AKAP450/CG-NAP. The former is a cis-Golgi-associated protein that copurifies with MTs (Infante et al. 1999; Kim et al. 2007) and recruits g-tubulin complexes to the GA (Rios et al. 2004). GMAP210 depletion fragments the Golgi ribbon into elements that remain near the centrosome (Rios et al. 2004). AKAP450 localizes at both the centrosome and the GA (Keryer et al. 2003a, b; Larocca et al. 2004; Takahashi et al. 1999). It contains two MT-binding domains and interacts with g-tubulin complexes, thus providing MT-nucleating sites to the centrosome and, probably, to the GA as well (Kim et al. 2007; Takahashi et al. 2002). Interestingly, AKAP450 also interacts with p150Glued and the expression of a mutant that disrupts this interaction causes GA fragmentation and dispersion in a similar manner to that observed with the overexpression of dynamitin. In RPE-1 cells, many MTs are generated from the TGN, where microtubule plus-end-binding proteins CLASPs localize. These proteins stabilize pre-existing MTseeds by coating them, thus preventing their disassembly. In this regard, the centrosomal protein CAP350, which was originally believed to participate in MT-anchoring at the centrosome (Yan et al. 2006) actually stabilizes MTs enriched in the Golgi complex, and thus helps to maintain the integrity of the GA in the vicinity of the centrosome (Hoppeler-Lebel et al. 2007).

Relationship between the Golgi apparatus and microtubules in different cellular systems So far, we have focused on what we have learned in cells displaying a radial MT array with plus-ends facing toward the cell cortex and minus-ends anchored at the centrosome. Direct observation of MTs reveals that cell lines with well-defined radial MT arrays are really a minority. In contrast, most cell lines display a loosely organized MT array. These differences are due to variations in the number of centrosomal anchored MTs, which ranges from their totality (lymphocytes) to practically none (epithelia) (Bornens 2002). In parallel, the morphology of the GA varies from a fully compacted shape around the centrosome to a highly extended one (Rios and Bornens 2003). The differentiation of specialized cells types in multicellular organisms frequently leads to the generation of non-radial MT arrays, which better serve the specialized functions of these cells (Dammermann et al. 2003; Musch 2004). Two of the most representative examples of non-radial MT arrays are the polarized epithelial cell and skeletal muscle fiber. In polarized epithelial cells, MTs form an apico-basal array with their minus-ends concentrated near the apical surface and their plus-ends facing the basal domain (Mogensen et al. 2000). This array determines the membrane trafficking, which is central to the function of epithelia. In addition to vertically arranged MTs, cell lines derived from columnar epithelia (MDCK or Caco-2) also show networks of horizontal MTs both at the cell apex and the

The role of the cytoskeleton in the structure and function of GA

*

277

cell base. In hepatocytes, MTs converge underneath the bile-canalicular lumen at the apical surface (Dammermann et al. 2003; Musch 2004). The GA shows compact morphology and typically lies just apical to the nucleus, wellseparated from the microtubule minus-ends. In MDCK cells, the GA extends upwards from the nucleus to the apical portion of the cell. At the same time, in epithelial cells, membrane proteins are segregated into functionally and structurally different apical and basolateral domains. Whereas MTs seem to be important in the organization of apical exocytosis, the actin cytoskeleton seems to be the main organizer for basolateral secretion. However, neither the motors that participate in the exit of various classes of proteins from the TGN, nor the molecular interactions that allow MTs and actin filaments to modulate luminal and basolateral polarity are fully understood (RodriguezBoulan et al. 2005). Skeletal muscle fibers are multinucleate cells resulting from the fusion of mononucleate myoblasts in myogenesis. During this process, both pericentriolar proteins and MT nucleation sites redistribute from the centrosome to the nuclear periphery (Dammermann et al. 2003; Tassin et al. 1985a). In the same way, the GA is redistributed into smaller perinuclear elements that are formed by stacked cisternae (Tassin et al. 1985b). As in non-polarized cells, these perinuclear Golgi elements appear localized near the ERES and associated with stable MTs (Lu et al. 2001; Percival and Froehner 2007; Ralston et al. 1999, 2001).

The actin-based cytoskeleton and the Golgi apparatus The actin-based cytoskeleton and the structural organization of the Golgi apparatus in mammalian cells The first experimental evidence that AFs and the GA are linked was the observation that the GA invariably becomes compacted when a large variety of clonal cell lines are treated with actin toxins that either depolymerize (mainly cytochalasins and latrunculins) or stabilize and nucleate AFs (jasplakinolide) (di Campli et al. 1999; Lazaro-Dieguez et al. 2006; Valderrama et al. 1998, 2000, 2001) (Fig. 2). However, at the ultrastructural level, Golgi stacks from cells treated with actin-depolymerizing or -stabilizing toxins appeared different. Thus, the former mainly show swelled cisternae, while the latter have perforated/fragmented cisternae, which remain totally flat (Fig. 2). Supporting these observations, cisternae in NZtreated cells remain completely flat (Thyberg and Moskalewski 1999; Trucco et al. 2004), and the GA in cells treated with NZ plus actin toxins display the same ultrastructural alterations as those seen in cells treated with actin toxins alone (Lazaro-Dieguez et al. 2006). This indicates that there is no synergic cooperation between MTs and AFs controlling the cisterna morphology. Therefore, as a general rule, MTs determine the pericentriolar localization of the Golgi ribbon, whereas AFs maintain the shape and membrane integrity of cisternae.

278

*

G. Egea and R. M. Ríos

Figure 2. Actin filaments–Golgi interaction. NRK cells treated with the actin-depolymerizing toxin latrunculin B (þLtB; 500 nM/45 min) or the actin-stabilizing toxin jasplakinolide (þJpk; 500 nM/45 min) show a similar compaction of the Golgi complex when viewed under the fluorescence microscope. In the respective panels, we also show the resulting F- and G-actin pools. At ultrastructural level, LtB treatment mainly produced swelling of cisternae. In contrast, Jpk treatment only induced fragmentation of cisternae. We also display 3D reconstructions obtained from electron tomograms of the GA from control and actin toxin-treated cells. In control cells, both the central compact (c) and the lateral non-compact (nc) regions are seen. In LtB-treated cells, the predominant alteration is the swelling of cisternae (asterisks), but some cisternal perforation/fragmentation is also seen (arrowheads). Jpk treatment leave cisternae flat, but they show numerous perforations (arrows) and perforations/fragmentations (arrowheads). A perforation/fragmentation of a Golgi stack is indicated by the dashed red lines. We also indicate, in the respective 3D Golgi modelling panels, the intra-Golgi pH values obtained for each experimental condition (asterisk indicates a statistical significance of p  0.01 according to the Student’s t-test).

The role of the cytoskeleton in the structure and function of GA

*

279

Cisternae are always flat, despite the huge amount of cargo that is continuously crossing the Golgi stack. When the amount of cargo to be transported is much higher, then membrane continuities appear between cisternae (Trucco et al. 2004). The aforementioned ultrastructural changes induced by actin toxins indicate that AFs provide the necessary mechanical stability to cisternae to prevent their expected spontaneous swelling as a consequence of the hyperosmotic protein content in transit through the Golgi stack. By analogy with red blood cells, the unique flat morphology of cisternae could result from the structural organization of the spectrin-actinbased cytoskeleton present in the GA (Bennett and Baines 2001; De Matteis and Morrow 2000). In this respect, spectrin and ankyrin isoforms, actin, and an anion exchanger (AE2) are all present in Golgi membranes (Beck et al. 1997; Devarajan et al. 1996, 1997; Godi et al. 1998; Heimann et al. 1999; Holappa et al. 2001, 2004; Stankewich et al. 1998; Valderrama et al. 2000). Together with these, ion regulatory molecules such as vacuolar Hþ-ATPases (Moriyama and Nelson 1989) and cation (NHEs) exchangers (Nakamura et al. 2005) resident in the Golgi or in transit to the plasma membrane could contribute to this postulated actin/spectrin-based cisternal mechanical stability by providing the appropriate intra-Golgi ion and pH homeostasis. This is essential, on the one hand for Golgi-associated post-translational protein and lipid modifications (Axelsson et al. 2001) and on the other hand to keep the cisternae flat. Thus, bafilomycin A1, an inhibitor of vacuolar ATPases, both induces cisterna swelling and slows the Golgi-to-ER protein transport (Palokangas et al. 1998). Curiously, the cisterna swelling after AFs depolymerization is accompanied by an increase in the intra-Golgi pH (Fig. 2). The restitution of normal actin cytoskeleton organization after the removal of actin-depolymerizing toxins is followed by the normalization of cisterna morphology and the intra-Golgi pH (Lazaro-Dieguez et al. 2006). This correlation strongly suggests that AFs could interact and modulate the activity of (some) ionic regulatory proteins (vacuolar ATPases, anion and cation exchangers, ionic channels, pumps, and/ or transporters) present in Golgi membranes. This interaction would be highly similar to that observed for some of these proteins present at the plasma membrane. Therefore, we postulate that the equilibrium of osmotically active ions would maintain the flatness of Golgi cisternae in concert with the actin assembly state (Lazaro-Dieguez et al. 2006). At the same time, the Golgi-associated spectrin-actin cytoskeleton system could organize the secretory molecular machinery controlling the lateral distribution of the main Golgi membrane components (De Matteis and Morrow 2000). Thus, a physical barrier could be formed in the compact region of cisternae by the conjunction of the spectrin-actin-based cytoskeleton together with particular lipids (for example, cylindrical-shaped ones) and proteins (for example, glycosyltransferases). This would result in a permanent inhibitory membrane area for the biogenesis of transport carriers. Future research in the Golgi spectrin and ankyrin isoforms should provide significant insights into Golgi architecture.

280

*

G. Egea and R. M. Ríos

The actin-based cytoskeleton and the biogenesis and motion of Golgi-derived transport carriers The presence in Golgi membranes of molecular components that trigger actin polymerization (Arp2/3, Cdc42, cortactin, N-WASP, syndapin) and those that determine vesicular budding (coatomer and clathrin coats) and fission (dynamin) suggests their physiological coupling, as occurs at the plasma membrane during endocytosis. Actin assembly provides the necessary structural support that facilitates the formation of transport carriers in the lateral portions of Golgi membranes. This can be achieved by generating force through de novo actin polymerization, which in turn can be accompanied by the mechanical activity of actin motors (myosins). In respect to the former possibility, the actin nucleators Arp2/3 and Spir1 are present in the Golgi (Carreno et al. 2004; Chen et al. 2004; Kerkhoff et al. 2001; Matas et al. 2004). Their respective upstream regulators can be diverse. For the Arp2/3 complex, the more consistent are Cdc42-N-WASP and dynamin2-cortactin (Cao et al. 2005; Chen et al. 2004; Luna et al. 2002; Matas et al. 2004). At the trans-Golgi network (TGN), there is experimental evidence of the functional coupling between dynamin-mediated membrane fission and Arp2/3-mediated actinbased mechanisms (Cao et al. 2005; Carreno et al. 2004; Kerkhoff et al. 2001; Kessels and Qualmann 2004; Praefcke and McMahon 2004; Rozelle et al. 2000). Thus, the interference with dynamin2/cortactin or syndapin2/dynamin2 protein interactions blocks post-Golgi protein transport (Cao et al. 2005; Kessels et al. 2006). Fewer data are available on early Golgi compartments, but an interesting functional connection between actin polymerization governed by Cdc42, coatomer (COPI)-mediated transport carrier formation, and microtubule motor-mediated motion has been described (Chen et al. 2005a). Under the activation of the ADP-ribosylation factor 1 (ARF1), actin, coatomer and the Cdc42 are all recruited to Golgi membranes (Stamnes 2002). Cdc42 interacts with g-COP subunit of the COPI-coated transport carrier in a cargo receptor p23-sensitive manner such that coatomer cannot simultaneously bind to Cdc42 and p23 (Chen et al. 2005a, b). Interestingly, the activation of Cdc42 (Cdc42-GTP) inhibits the recruitment of dynein to COPI-coated transport carriers. In contrast, the prevention of the COPI–Cdc42 interaction by p23 stimulates dynein recruitment on Golgi-derived transport carriers, and hence their MT-based transport. Overall, this could provide a safe control mechanism by which a COPI-mediated transport carrier cannot be moved (through microtubule motors) until it is completely assembled (when Cdc42 does not bind to coatomer) (Hehnly and Stamnes 2007). Supporting this idea, the disruption of actin filaments as well as the activation of the Cdc42-N-WASP-Arp2/3 signaling pathway by the expression of the constitutively activated mutant of Cdc42 (GTP-bound) block the COPI-mediated Golgi-to-ER protein transport (Luna et al. 2002; Valderrama et al. 2001). Therefore, the local fine regulation of the actin dynamics state on the transport carrier assembly could represent an early step that precedes the scission of the transport carrier in the lateral portions of

The role of the cytoskeleton in the structure and function of GA

*

281

cisternae for its subsequent switching to MT tracks and motility (Egea et al. 2006). The coupling between actin polymerization and transport carrier biogenesis occurs both at the TGN and in early Golgi compartments, but some of the molecular mediators that regulate both processes are unevenly distributed in the GA. Thus, regarding actin polymerization, Cdc42 and Arp2/3 are present to varying degrees in the Golgi stack (cis/middle/trans-cisternae) and at the TGN, but N-WASP is absent from the trans/TGN (Matas et al. 2004). Cortactin, which like N-WASP also recruits Arp2/3, is visualized at the tips and buds at both the cis- and trans-cisternae (Cao et al. 2005). It is then reasonable to postulate that, at the TGN, cortactin substitutes N-WASP in the recruitment of the Arp2/3 that mediate in the post-Golgi protein transport. If so, the cis-totrans/TGN segregation of some fundamental components involved in actin polymerization would facilitate the targeting and assembly of the specific molecular machinery that participates in the membrane budding and fission occurring in Golgi compartments. Thus, the sequential protein interactions syndapin2- or dynamin2-cortactin-Arp2/3 and Cdc42-N-WASP-Arp2/3 are respectively restricted to the TGN and to early Golgi compartments, participating in this manner in the post-Golgi and in the ER/Golgi interface protein transport. A key aspect in the structure of polarized cells (epithelial and neuronal) is the maintenance of polarized organization based on highly specific sorting machinery for cargo destined to the apical or basolateral membrane domain at the exit of the TGN (Rodriguez-Boulan et al. 2005). In accordance with the localization of Cdc42 in the trans/TGN (Matas et al. 2004), the expression of constitutively active (GTP-bound) or inactive (GDP-bound) Cdc42 mutants slows the exit of basolateral protein markers and accelerates apical ones (Cohen et al. 2001; Kroschewski et al. 1999; Musch et al. 2001). The downstream effectors involved in the regulatory protein sorting induced by Cdc42 at the TGN are unknown. The integrity of AFs is necessary for efficient delivery of some (but not all) proteins to the apical domain. For example, in MDCK cells the apical delivery of sucrase-isomaltase, but not that of lactase-phlorizin hydrolase or gp80, occurs along AF tracks (Delacour and Jacob 2006; Jacob et al. 2003). Interference with actin dynamics using actin toxins variably affects the exit of apical- and basolateral-targeted cargo from the TGN. Moreover, actin does not participate in the TGN egress of lipid raft-associated GPI-anchored cargo (Lazaro-Dieguez et al. 2007). Actin nucleation/polymerization activity on Golgi membranes can also give rise to the formation of actin comet tails, which consist of filamentous actin and various actin-binding proteins that focally assemble and grow on a membrane surface (Welch and Mullins 2002). Actin tails have been observed in raft-enriched TGN-derived vesicles in certain experimental conditions (Rozelle et al. 2000), but this does not seem to be the most efficient mechanism to specify directionality to transport carriers. In this respect, MT and AF tracks are more suitable. However, analogously to what happens at the plasma

282

*

G. Egea and R. M. Ríos

membrane (Merrifield 2004; Merrifield et al. 2005; Perrais and Merrifield 2005), Golgi-associated Arp2/3-mediated actin polymerization generates a force. Depending on whether this acts on the lateral portion of the cisterna or on the transport carrier membrane, it could respectively facilitate the membrane elongation that precedes membrane scission or propel the transport carrier away from the cisterna. The presence of components of the Arp2/3 complex in cisternae and periGolgi transport carriers (Chen et al. 2005a; Matas et al. 2004) endorses both possibilities, but direct experimental evidence of either phenomenon is still lacking. However, an interesting in vitro approach has recently been reported (Heuvingh et al. 2007). These authors observed actin polymerization around liposomes composed of a specific lipid that facilitates the recruitment of the activated form of ARF1. This actin polymerization was dependent on Cdc42 and N-WASP present in HeLa cell extracts, and resulted in the formation of actin comets, which pushed the ARF1 liposomes forward. Tight control of the coupling between Golgi-associated actin polymerization and membrane elongation and fission reactions prevents the structural and functional collapse of the GA. Part of this control can be achieved by regulating the activation state of Cdc42 in Golgi membranes. The Cdc42 GAP (GTPase-activating protein) ARHGAP10 and GEFs (guanine nucleotide exchange factor) Fgd1 and Dbs are present in Golgi membranes (Dubois et al. 2005; Estrada et al. 2001; Kostenko et al. 2005). In addition, the low levels of phosphatidylinositol 4,5-biphosphate (PIP2) present in the Golgi (De Matteis et al. 2005) could also facilitate this control. Note that PIP2 synergizes with Cdc42-N-WASP and cortactin in Arp2/3-triggered actin nucleation/polymerization (Rohatgi et al. 2000; Schafer et al. 2002). A variety of independent experimental approaches show that Cdc42 is the only Rho GTPase that functions in the Golgi complex in mammalian cell lines (Fucini et al. 2000; Matas et al. 2005; Prigozhina and Waterman-Storer 2004; Valderrama et al. 2000). However, neurons seem to be an exception. Citron-N, a RhoA-binding protein and ROCKinase-II are both seen in the neuronal GA (Camera et al. 2003). Likewise, LIMK1, a kinase that specifically phosphorylates ADF-cofilin, localizes to Golgi membranes (Rosso et al. 2004). For a review of the role of actin and actin-binding/regulatory proteins in the GA of neuronal cells see (Bornens 2002). In addition to actin polymerization, myosins also generate a force, which can promote the formation of transport carriers and/or their movement away from Golgi membranes along AF tracks. Non-muscle myosin II mediates both Golgi-to-ER and post-Golgi protein transport (DePina et al. 2007; Duran et al. 2003; Musch et al. 1997; Stow et al. 1998). This myosin is a non-processive motor that directly interacts with Golgi membranes in vitro (Fath 2005). It is postulated that this motor is tethered to the cisterna by its tail and to actin filaments by its head. Its subsequent motion along actin filaments could provide the force needed to extend Golgi-derived membranes away from the cisterna. This would be similar to what happens to tubules originating

The role of the cytoskeleton in the structure and function of GA

*

283

from liposomes incubated with microtubular motors (kinesin) moving along microtubules (Roux et al. 2002). This cisterna-derived membrane extension could facilitate the subsequent functional coupling of membrane scission protein(s), leading to the complete release of the transport carrier. The binding of a tropomyosin isoform to Golgi-associated short actin filaments (Percival et al. 2004) could facilitate non-muscle myosin II recruitment and its interaction with them. Myosin VI is another myosin motor located in the GA (Buss et al. 2004; Warner et al. 2003). It differs from other processive myosins (for example, myosin V and X) as it only moves transport carriers towards the fast-depolymerizing minus-end pole of the microfilament. Therefore, myosin VI could provide the force and directionality for the transport carrier movement away from cisternae according to the expected fast-growing plus-end polarization of the actin filaments originated from Golgi membranes (Chen et al. 2004). The interaction between myosin VI and optineurin, a partner of Rab8 (Sahlender et al. 2005) extends to the Golgi–actin cytoskeleton interaction the known role of some Rab proteins as linkers of endomembrane systems to cytoskeletal motors (Jordens et al. 2005). Myosin VI, together with optineurin and Rab 8, operates in protein sorting and transport at the TGN in polarized cells (Au et al. 2007). The myosin VI–optineurin complex is required in the basolateral protein sorting pathway mediated by the Golgi-associated clathrin adaptor protein AP-1B, which in turn is specifically regulated by Rab 8 (Ang et al. 2003). The inhibition of myosin VI results in the incorporation of basolateral membrane proteins into apical transport carriers and their delivery to the apical plasma membrane domain. Sorting of other basolateral or apical cargo does not involve myosin VI. This result suggests that myosin motors could selectively couple protein sorting and transport carrier biogenesis and motility. Finally, class I myosins are also reported to associate with Golgi membranes and on apical Golgi-derived vesicles from polarized cells (Fath and Burgess 1993; Jacob et al. 2003; Montes de Oca et al. 1997). In myosin Ia knock-out mice, apical markers sucrase–isomaltase and galectin-4 are mislocalized to the basolateral surface in intestinal epithelial cells (Tyska et al. 2005). The sorting ability of myosin I could be linked to its capacity to interact with lipid raft-associated cargo as this monomeric, non-processive motor binds to phospholipid vesicles (Hayden et al. 1990).

The Golgi apparatus–actin interaction in other cellular models Yeast The use of a large number of mutants that produce alterations in intracellular traffic in the budding yeast Saccharomyces cerevisie has led to the identification of proteins involved in both membrane trafficking and actin organization (Kaksonen et al. 2006; Mulholland et al. 1997). Most components of the secretory pathway and many of the actin-based cytoskeleton are conserved between yeast and mammalian cells. The actin cytoskeleton in yeast consists primarily of cortical patches and cables (Adams and Pringle 1984; Kilmartin and

284

*

G. Egea and R. M. Ríos

Adams 1984; Moseley and Goode 2006). AFs, and not MTs, polarizes growth in yeast (Novick and Botstein 1985). In Saccharomyces, MTs have not been implicated in the dynamics of any organelle except the nucleus (Rossanese et al. 2001). Many actin mutants accumulate large secretory vesicles and exhibit phenotypes consistent with defects in polarized growth (Pruyne et al. 2004). This, together with the polarized organization of actin cytoskeleton, has suggested a role for actin in the polarized transport of late secretory vesicles to the plasma membrane (Finger and Novick 2000; Mulholland et al. 1997). Thus, a mutation of GRD20, a protein involved in sorting in the TGN/endosomal system, showed aberrant secretion of the vacuolar hydrolase carboxypeptidase Y, but not other TGN membrane proteins, as well as defects in the polarization of the actin cytoskeleton (Spelbrink and Nothwehr 1999). Recently, depletion of Av19p in a strain that also lacks Vps1 (dynamin) and Apl2 (adaptor–protein complex 1) proteins results in secretory defects, accumulation of Golgi-like membranes, and a non-polarized actin cytoskeleton organization (Harsay and Schekman 2007). Finally, concentration of late (but not early) Golgi elements in the sites of polarized growth (the bud) depends on actin, which is transported along actin cables by type V myosin Myo2p (Rossanese et al. 2001). With regard to the early secretory pathway, AFs depolymerization with actin toxins does not affect ER-to-Golgi (Brazer et al. 2000) or Golgi-to-ER (M. Muñiz, personal communication) protein transport. Taken together, these results demonstrate that in yeast, actin organization directly participates in post-Golgi vesicular transport and in the Golgi inheritance.

Drosophila In adherent S2 cells derived from mixed Drosophila melanogaster embryonic tissues, it has recently been reported that Golgi inheritance occurs by duplication to form a paired structure. This process requires an intact actin cytoskeleton and depends on Abi/Scar but not WASP (Kondylis et al. 2007). In another recent study, the analysis of a genome-wide RNA-mediated interference screen in these cells showed that the depletion of the tsr gene (which codifies for destrin, also known as ADF/cofilin) induces Golgi membranes to aggregate and swell, resulting in inhibition of the HRP secretion (Bard et al. 2006). Coronin proteins dpdo1 and coro regulate the actin cytoskeleton, but also govern biosynthetic and endocytic vesicular trafficking, as indicated by mutant phenotypes that show severe developmental defects ranging from abnormal cell division to aberrant formation of morphogen gradients (Rybakin and Clemen 2005).

Dictyostelium The slime mould Dictyostelium discoideum (like Drosophila) is widely studied in developmental and cell biology. Cells of this protist are easy to manipulate by genetic and biochemical means. They contain various types of vacuole, ER and very small Golgi stacks (Becker and Melkonian 1996). Comitin (p24) is a dimeric Dictyostelium actin-binding protein present in the GA and in vesicle

The role of the cytoskeleton in the structure and function of GA

*

285

membranes that contains sequence motifs homologous to lectins. It is postulated that this protein may bind Golgi-derived vesicles to the actin filaments via the cytoplasmically exposed mannosylated glycans (Jung et al. 1996; Weiner et al. 1993). Villidin is another actin-binding protein present in this organism that seems to be associated with secretory vesicular and Golgi membranes (Gloss et al. 2003). LIS1 (DdLIS1) is a centrosomal protein required for the link between MTs, the nucleus and the centrosome that also controls the GA morphology. Mutants of this protein lead to MT disruption, Golgi fragmentation and actin depolymerization (Rehberg et al. 2005).

Caenorhabditis elegans Very little is known about the GA and actin cytoskeleton interaction in this organism, but consistent with a possible role of coronin 7 in Golgi trafficking (Rybakin et al. 2004), depletion of POD-1 gene (a Coronin 7 homolog) using RNA interference leads to aberrant accumulation of vesicles in cells of the early embryo (Rappleye et al. 1999). Moreover, CRP-1, a Cdc42-related protein, localizes at the TGN and recycling endosomes. Alteration of CRP-1 expression in epithelial-like cells affected the apical but not the basolateral trafficking (Jenna et al. 2005).

Plant cells The structural organization of the GA in plants has many points in common with animal cells but there are important differences, which are largely dependent on the different cytoskeleton organization of plant cells. Thus, interphase higher plant cells (angiosperms and some gymnosperms without flagellate sperm) lack doublet and triplet MTs and a single MTOC. Instead, numerous MTOCs are aligned in the cortex, which assemble and form the transverse bands referred to as cortical MTs. These MTs are essential for the transport of Golgi-derived vesicles formed during metaphase (Segui-Simarro et al. 2004), which subsequently fuse to form the phragmoplast (Jurgens 2005), the equivalent to the contractile ring in animal cells. In contrast to MTs, stationary AFs are most prominent in plant cells (known as actin bundles) where they are all oriented with the same polarity and aligned along the plant cell. Attached to the actin bundles are the ER, vesicles and numerous discrete or a few clustered Golgi stack-TGN units (also named Golgi bodies or dictyosomes). Importantly, Golgi units are highly variable in number (a few tens to hundreds) depending on the plant type, cell type and the developmental stage of the cell (Hawes and Satiat-Jeunemaitre 2005; Kepes et al. 2005). In polarized root hairs and pollen tubes, the TGN is a vesicular-like non-tubular compartment morphologically segregated from Golgi stacks. It localizes to growing tips of these cells, where together with actin, plant Rho/Rac members (ROPs and Rac1, respectively), Rab (Rab4a and Rab11) and ARF (ARF1) small GTPases regulate vesicular secretory and endocytic trafficking (Samaj et al. 2006). In plants, most of the endomembrane compartments are in constant movement together with the cytoplasmic streaming whereby cellular metab-

286

*

G. Egea and R. M. Ríos

olites are distributed all over the cell (Shimmen and Yokota 2004). ER vesicles and the Golgi units show actin-dependent dispersal and spatial organization and are propelled by the plant-specific myosin XI (Boutte et al. 2007). Discrete Golgi units contains a fine fibrillar material enriched in actin, spectrin and myosin-like proteins (especially the former) (Mollenhauer and Morre 1976; Satiat-Jeunemaitre et al. 1996). The depolymerization of AFs with actin toxins uncouples the association between specific regions of cortical ER with individual Golgi bodies (Boevink et al. 1998; Brandizzi et al. 2003). Thus, cytochalasin or latrunculin treatments induce the aggregation of Golgi bodies and variably alter the Golgi morphology. However, the latter depends on the cell type examined and the period of treatment (Chen et al. 2006; Satiat-Jeunemaitre et al. 1996). Actin toxins also perturb the coordinated movement of Golgi bodies and ER tubules (daSilva et al. 2004; Yang et al. 2005). Actin does not participate in the ER/Golgi interface protein transport (Saint-Jore et al. 2002), but it does it in post-Golgi trafficking to the plasma membrane and the vacuole. Thus, in the tip of growing cells like pollen tubes, AFs are the tracks on which Golgi-derived secretory vesicles are transported (Picton and Steer 1981; Vidali et al. 2001). Cargoes containing polysaccharides and the enzymes necessary for cell-wall morphogenesis also require an intact actin–myosin system (Blancaflor 2002; Hu et al. 2003; Miller et al. 1995; Nebenfuhr et al. 1999). Therefore, post-Golgi trafficking and the organization of vacuoles in plant cells require an intact actin cytoskeleton (Uemura et al. 2002).

The Golgi apparatus-intermediate filaments interaction IFs are found in nearly all animal cells. They are classified according to their distribution in specific tissues. In contrast to MTs and AFs, IFs do not exhibit polarity or bind nucleotides, and they are considered a more stable structure. IFs are of intermediate size (8–12 nm) in comparison to MTs (23–25 nm) and to AFs (6–8 nm). IFs maintain cell and tissue integrity thanks to their mechanical properties, cellular distribution and, as far as we know, from disease-associated IFs phenotypes. IFs participate in the regulation of key signaling pathways that control cell survival and growth, and also in protein targeting and membrane trafficking (Coulombe and Wong 2004; Kim and Coulombe 2007; Omary et al. 2004; Oriolo et al. 2007; Styers et al. 2005; Toivola et al. 2005). IFs extend from the plasma membrane to the nucleus in close vicinity to some organelles such as mitochondria, endocytic compartments, and the GA (Fig. 3). The first Golgi–IF interaction was reported for vimentin filaments at ultrastructural level (Katsumoto et al. 1991), and later confirmed biochemically (Gao and Sztul 2001). The interaction is mediated by the Golgi membrane-associated protein formiminotransferase cyclodeaminase (FTCD), a metabolic enzyme involved in conversion of histidine to glutamic acid (Gao et al. 1998). Overexpression of FTCD resulted in the formation of extensive FTCD-containing fibres originating from the GA and inducing its fragmentation, and whose fragments remained tethered to these fibers (Gao and Sztul,

The role of the cytoskeleton in the structure and function of GA

*

287

Figure 3. Intermediate filaments–Golgi interaction. The upper panels show an NRK cell double stained to reveal the association of the GA (using anti-mannosidase II antibodies; in red) with a network of vimentin intermediate filaments (in green) seen by immunofluorescence. Bar: 2 mm. The bottom panels illustrate the severe Golgi fragmentation that occurred in transfected Huh7 cells expressing the GFP-cytokeratin18 R89C mutant. The GA was revealed with anti-galactosyltransferase (GalT) antibodies. Neighbouring non-transfected cells show a normal GA (Kumemura et al. 2004) (images used with permission of the authors and the publisher). Bar, 4 mm.

2001). However, the GA appears to be normal in vimentin–null cells (Gao et al. 2002; Styers et al. 2004). Thus, it is postulated that the vimentin–FTCD interaction at the GA is essential for FTCD functionality, but not linked to the maintenance of Golgi organization. Oxysterol-binding protein (OSBP) regulates lipid and cholesterol metabolism and interacts with the GA in the presence of oxysterol (Ridgway et al. 1992). A splice variant form of the OSBPrelated protein 4 in which the PH domain and part of the oxysterol-binding domain are deleted, colocalizes with vimentin IF in the Golgi region and inhibits the intracellular cholesterol transport pathway mediated by vimentin (Wang et al. 2002). Epithelial cells express cytokeratins, whose mutations are also associated with epidermal, oral and ocular diseases (Uitto et al. 2007). Arginine 89 of cytokeratin18 plays an important role in IF assembly. The expression of this mutant cytokeratin-induced aggregations, loss of the cytokeratin cytoskele-

288

*

G. Egea and R. M. Ríos

ton and the fragmentation of the GA (Fig. 3). Moreover, the Golgi reassembly occurring after NZ or brefeldin A treatments was perturbed only in cells expressing both cytokeratins and vimentin IFs (Huh7 and OUMS29 cells), but curiously not in cells that only possess vimentin IFs (HEK293 cells) (Kumemura et al. 2004). Whether cytokeratin or vimentin IFs are involved in Golgi-associated protein sorting, vesicular formation and/or transport remains to be established. However, recent evidence indicates that IFs could regulate some membrane protein targeting events, which take place at the GA level. Thus vimentin directly binds AP-3 and thus regulates protein sorting in endo-lysosomes (Styers et al. 2004). Polarized enterocytes and hepatocytes depleted of some keratins by antisense strategies and in cytokeratin 8-null mouse cells showed altered apical protein transport (Ameen et al. 2001; Rodriguez et al. 1994; Salas et al. 1997). Maturation of glycosphingolipids is also impaired in vimentin-deficient cells, but the defect seems to be localized to the Golgi/endosomal interface transport (Gillard et al. 1994, 1998). Since MT-motors kinesin and dynein also control the dynamics of IFs (Helfand et al. 2002, 2003; Prahlad et al. 1998), the reported Golgi/endo-lysosomal membrane trafficking mediated by IFs may be under the control of the dynamic IF–MT interaction, which can also be applied to AFs and motors. In this respect, the actin-based motor myosin Va has been identified as a neurofilament-associated protein (Rao et al. 2002). In summary, the structural and functional interaction of the GA with IFs is not yet firmly established. However, since IFs are well-integrated with both actin and microtubule cytoskeletons and their motors (Chang and Goldman 2004), the organization of the GA may also be influenced by the organization and dynamics of IFs.

Conclusion Both MTs and AFs are necessary for correct Golgi positioning, architecture and trafficking. Strong evidence in favour of this view now indicates that the GA functions as a microtubule and as an actin-nucleating organelle. In general terms, the relationships between each cytoskeleton network and Golgi dynamics are complementary. Thus, in animals cells, the actin-dependent cytoskeleton (AFs and actin-binding/regulatory proteins) plays an important role in early events of vesicular transport (sorting and/or membrane fission), and in the maintenance of the flattened morphology of cisternae. Furthermore, MTs and associated motors are directly involved in the motion of Golgiderived transport carriers to their final destinations and in the positioning and organization of the Golgi ribbon. In contrast, in plant cells, endomembrane organization and trafficking are almost exclusively mediated by AFs. Both ER and individual Golgi stacks are directly anchored to actin bundles, and transport between the ER and the GA is cytoskeleton-independent. In yeast, much less is known, but both cytoskeleton elements participate in post-Golgi protein transport and Golgi inheritance.

The role of the cytoskeleton in the structure and function of GA

*

289

At the apex of complexity, we envision the GA of mammalian cells to be assembled in three stages, in which the cytoskeleton participates in a variable extent: the first flattens cisternae (which is actin-dependent); the second maintains the discrete stack structure in the tight parallel arrangement of a variable number of cisternae (which at the moment seems to be independent of cytoskeleton proteins), and the third maintains stacks together to produce the classical single ribbon-like Golgi structure (which is fully dependent on MTs). Variations in the relative contribution of each of these three steps could generate the diversity of GA arrangements observed in different biological systems.  s Antón and members of our Acknowledgements. We thank Michel Bornens, Ine zarorespective labs for their comments, as well as Sabrina Rivero and Francisco La  guez for help with figures. The work carried out in our laboratories has been Die supported by grants from Ministerio de Educación y Ciencia (G.E. and R.M.R.), Junta de Andalucía (R.M.R.), and Distinció Award from the Generalitat de Catalunya (G.E).

References Adams AE, Pringle JR (1984) Relationship of actin and tubulin distribution to bud growth in wild-type and morphogenetic-mutant Saccharomyces cerevisiae. J Cell Biol 98: 934–945 Ameen NA, Figueroa Y, Salas PJ (2001) Anomalous apical plasma membrane phenotype in CK8-deficient mice indicates a novel role for intermediate filaments in the polarization of simple epithelia. J Cell Sci 114: 563–575 Ang AL, Folsch H, Koivisto UM, Pypaert M, Mellman I (2003) The Rab8 GTPase selectively regulates AP-1B-dependent basolateral transport in polarized Madin-Darby canine kidney cells. J Cell Biol 163: 339–350 Arasaki K, Taniguchi M, Tani K, Tagaya M (2006) RINT-1 regulates the localization and entry of ZW10 to the syntaxin 18 complex. Mol Biol Cell 17: 2780–2788 Au JS, Puri C, Ihrke G, Kendrick-Jones J, Buss F (2007) Myosin VI is required for sorting of AP-1B-dependent cargo to the basolateral domain in polarized MDCK cells. J Cell Biol 177: 103–114 Axelsson MA, Karlsson NG, Steel DM, Ouwendijk J, Nilsson T, Hansson GC (2001) Neutralization of pH in the Golgi apparatus causes redistribution of glycosyltransferases and changes in the O-glycosylation of mucins. Glycobiology 11: 633–644 Bard F, Casano L, Mallabiabarrena A, Wallace E, Saito K, Kitayama H, Guizzunti G, Hu Y, Wendler F, Dasgupta R, Perrimon N, Malhotra V (2006) Functional genomics reveals genes involved in protein secretion and Golgi organization. Nature 439: 604–607 Beck KA, Buchanan JA, Nelson WJ (1997) Golgi membrane skeleton: identification, localization and oligomerization of a 195 kDa ankyrin isoform associated with the Golgi complex. J Cell Sci 110 ( Pt 10), 1239–1249 Becker B, Melkonian M (1996) The secretory pathway of protists: spatial and functional organization and evolution. Microbiol Rev 60: 697–721 Bennett V, Baines AJ (2001) Spectrin and ankyrin-based pathways: metazoan inventions for integrating cells into tissues. Physiol Rev 81: 1353–1392 Berezuk MA, Schroer TA (2007) Dynactin enhances the processivity of kinesin-2. Traffic 8: 124–129 Blancaflor EB (2002) The cytoskeleton and gravitropism in higher plants. J Plant Growth Regul 21: 120–136

290

*

G. Egea and R. M. Ríos

Boevink P, Oparka K, Santa Cruz S, Martin B, Betteridge A, Hawes C (1998) Stacks on tracks: the plant Golgi apparatus traffics on an actin/ER network. Plant J 15: 441–447 Bornens M (2002) Centrosome composition and microtubule anchoring mechanisms. Curr Opin Cell Biol 14: 25–34 Boutte Y, Vernhettes S, Satiat-Jeunemaitre B (2007) Involvement of the cytoskeleton in the secretory pathway and plasma membrane organisation of higher plant cells. Cell Biol Int 31: 649–654 Brandizzi F, Saint-Jore C, Moore I, Hawes C (2003) The relationship between endomembranes and the plant cytoskeleton. Cell Biol Int 27: 177–179 Brazer SC, Williams HP, Chappell TG, Cande WZ (2000) A fission yeast kinesin affects Golgi membrane recycling. Yeast 16: 149–166 Burgess TL, Skoufias DA, Wilson L (1991) Disruption of the Golgi apparatus with brefeldin A does not destabilize the associated detyrosinated microtubule network. Cell Motil Cytoskeleton 20: 289–300 Burkhardt JK, Echeverri CJ, Nilsson T, Vallee RB (1997) Overexpression of the dynamitin (p50) subunit of the dynactin complex disrupts dynein-dependent maintenance of membrane organelle distribution. J Cell Biol 139: 469–484 Buss F, Spudich G, Kendrick-Jones J (2004) Myosin VI: cellular functions and motor properties. Annu Rev Cell Dev Biol 20: 649–676 Camera P, da Silva JS, Griffiths G, Giuffrida MG, Ferrara L, Schubert V, Imarisio S, Silengo L, Dotti CG, Di Cunto F (2003) Citron-N is a neuronal Rho-associated protein involved in Golgi organization through actin cytoskeleton regulation. Nat Cell Biol 5: 1071–1078 Cao H, Weller S, Orth JD, Chen J, Huang B, Chen JL, Stamnes M, McNiven MA (2005) Actin and Arf1-dependent recruitment of a cortactin–dynamin complex to the Golgi regulates post-Golgi transport. Nat Cell Biol 7: 483–492 Carreno S, Engqvist-Goldstein AE, Zhang CX, McDonald KL, Drubin DG (2004) Actin dynamics coupled to clathrin-coated vesicle formation at the trans-Golgi network. J Cell Biol 165: 781–788 Caviston JP, Holzbaur EL (2006) Microtubule motors at the intersection of trafficking and transport. Trends Cell Biol 16: 530–537 Caviston JP, Ross JL, Antony SM, Tokito M, Holzbaur EL (2007) Huntingtin facilitates dynein/dynactin-mediated vesicle transport. Proc Natl Acad Sci USA 104: 10045–10050 Cohen D, Musch A, Rodriguez-Boulan E (2001) Selective control of basolateral membrane protein polarity by cdc42. Traffic 2: 556–564 Cole NB, Sciaky N, Marotta A, Song J, Lippincott-Schwartz J (1996) Golgi dispersal during microtubule disruption: regeneration of Golgi stacks at peripheral endoplasmic reticulum exit sites. Mol Biol Cell 7: 631–650 Corthesy-Theulaz I, Pauloin A, Pfeffer SR (1992) Cytoplasmic dynein participates in the centrosomal localization of the Golgi complex. J Cell Biol 118: 1333–1345 Coulombe PA, Wong P (2004) Cytoplasmic intermediate filaments revealed as dynamic and multipurpose scaffolds. Nat Cell Biol 6: 699–706 Chabin-Brion K, Marceiller J, Perez F, Settegrana C, Drechou A, Durand G, Pous C (2001) The Golgi complex is a microtubule-organizing organelle. Mol Biol Cell 12: 2047–2060 Chang L, Goldman RD (2004) Intermediate filaments mediate cytoskeletal crosstalk. Nat Rev Mol Cell Biol 5: 601–613 Chen JL, Fucini RV, Lacomis L, Erdjument-Bromage H, Tempst P, Stamnes M (2005a) Coatomer-bound Cdc42 regulates dynein recruitment to COPI vesicles. J Cell Biol 169: 383–389 Chen JL, Lacomis L, Erdjument-Bromage H, Tempst P, Stamnes M (2004) Cytosol-derived proteins are sufficient for Arp2/3 recruitment and ARF/coatomer-dependent actin polymerization on Golgi membranes. FEBS Lett 566: 281–286

The role of the cytoskeleton in the structure and function of GA

*

291

Chen JL, Xu W, Stamnes M (2005b) In vitro reconstitution of ARF-regulated cytoskeletal dynamics on Golgi membranes. Methods Enzymol 404: 345–358 Chen Y, Chen T, Shen S, Zheng M, Guo Y, Lin J, Baluska F, Samaj J (2006) Differential display proteomic analysis of Picea meyeri pollen germination and pollen-tube growth after inhibition of actin polymerization by latrunculin B. Plant J 47: 174–195 Dammermann A, Desai A, Oegema K (2003) The minus end in sight. Curr Biol 13: R614–R624 DaSilva LL, Snapp EL, Denecke J, Lippincott-Schwartz J, Hawes C, Brandizzi F (2004) Endoplasmic reticulum export sites and Golgi bodies behave as single mobile secretory units in plant cells. Plant Cell 16: 1753–1771 De Anda FC, Pollarolo G, Da Silva JS, Camoletto PG, Feiguin F, Dotti CG (2005) Centrosome localization determines neuronal polarity. Nature 436: 704–708 De Matteis MA, Di Campli A, Godi A (2005) The role of the phosphoinositides at the Golgi complex. Biochim Biophys Acta 1744: 396–405 De Matteis MA, Morrow JS (2000) Spectrin tethers and mesh in the biosynthetic pathway. J Cell Sci 113 ( Pt 13): 2331–2343 Deacon SW, Serpinskaya AS, Vaughan PS, Lopez Fanarraga M, Vernos I, Vaughan KT, Gelfand VI (2003) Dynactin is required for bidirectional organelle transport. J Cell Biol 160: 297–301 Delacour D, Jacob R (2006) Apical protein transport. Cell Mol Life Sci 63: 2491–2505 DePina AS, Wollert T, Langford GM (2007) Membrane associated nonmuscle myosin II functions as a motor for actin-based vesicle transport in clam oocyte extracts. Cell Motil Cytoskeleton 64: 739–755 Devarajan P, Stabach PR, De Matteis MA, Morrow JS (1997) Na,K-ATPase transport from endoplasmic reticulum to Golgi requires the Golgi spectrin–ankyrin G119 skeleton in Madin Darby canine kidney cells. Proc Natl Acad Sci USA 94: 10711–10716 Devarajan P, Stabach PR, Mann AS, Ardito T, Kashgarian M, Morrow JS (1996) Identification of a small cytoplasmic ankyrin (AnkG119) in the kidney and muscle that binds beta I sigma spectrin and associates with the Golgi apparatus. J Cell Biol 133: 819–830 Di Campli A, Valderrama F, Babia T, De Matteis MA, Luini A, Egea G (1999) Morphological changes in the Golgi complex correlate with actin cytoskeleton rearrangements. Cell Motil Cytoskeleton 43: 334–348 Dubois T, Paleotti O, Mironov AA, Fraisier V, Stradal TE, De Matteis MA, Franco M, Chavrier P (2005) Golgi-localized GAP for Cdc42 functions downstream of ARF1 to control Arp2/3 complex and F-actin dynamics. Nat Cell Biol 7: 353–364 Duran JM, Valderrama F, Castel S, Magdalena J, Tomas M, Hosoya H, Renau-Piqueras J, Malhotra V, Egea G (2003) Myosin motors and not actin comets are mediators of the actin-based Golgi-to-endoplasmic reticulum protein transport. Mol Biol Cell 14: 445–459 Efimov A, Kharitonov A, Efimova N, Loncarek J, Miller PM, Andreyeva N, Gleeson P, Galjart N, Maia AR, McLeod IX, Yates JR III, Maiato H, Khodjakov A, Akhmanova A, Kaverina I (2007) Asymmetric CLASP-dependent nucleation of noncentrosomal microtubules at the trans-Golgi network. Dev Cell 12: 917–930 Egea G, Lazaro-Dieguez F, Vilella M (2006) Actin dynamics at the Golgi complex in mammalian cells. Curr Opin Cell Biol 18: 168–178 Eng EW, Bettio A, Ibrahim J, Harrison RE (2007) MTOC reorientation occurs during FcgammaR-mediated phagocytosis in macrophages. Mol Biol Cell 18: 2389–2399 Estrada L, Caron E, Gorski JL (2001) Fgd1, the Cdc42 guanine nucleotide exchange factor responsible for faciogenital dysplasia, is localized to the subcortical actin cytoskeleton and Golgi membrane. Hum Mol Genet 10: 485–495 Fath KR (2005) Characterization of myosin-II binding to Golgi stacks in vitro. Cell Motil Cytoskeleton 60: 222–235

292

*

G. Egea and R. M. Ríos

Fath KR, Burgess DR (1993) Golgi-derived vesicles from developing epithelial cells bind actin filaments and possess myosin-I as a cytoplasmically oriented peripheral membrane protein. J Cell Biol 120: 117–127 Finger FP, Novick P (2000) Synthetic interactions of the post-Golgi sec mutations of Saccharomyces cerevisiae. Genetics 156: 943–951 Fucini RV, Navarrete A, Vadakkan C, Lacomis L, Erdjument-Bromage H, Tempst P, Stamnes M (2000) Activated ADP-ribosylation factor assembles distinct pools of actin on Golgi membranes. J Biol Chem 275: 18824–18829 Gao Y, Sztul E (2001) A novel interaction of the Golgi complex with the vimentin intermediate filament cytoskeleton. J Cell Biol 152: 877–894 Gao YS, Vrielink A, MacKenzie R, Sztul E (2002) A novel type of regulation of the vimentin intermediate filament cytoskeleton by a Golgi protein. Eur J Cell Biol 81: 391–401 Gillard BK, Clement R, Colucci-Guyon E, Babinet C, Schwarzmann G, Taki T, Kasama T, Marcus DM (1998) Decreased synthesis of glycosphingolipids in cells lacking vimentin intermediate filaments. Exp Cell Res 242: 561–572 Gillard BK, Thurmon LT, Harrell RG, Capetanaki Y, Saito M, Yu RK, Marcus DM (1994) Biosynthesis of glycosphingolipids is reduced in the absence of a vimentin intermediate filament network. J Cell Sci 107 ( Pt 12): 3545–3555 Gloss A, Rivero F, Khaire N, Muller R, Loomis WF, Schleicher M, Noegel AA (2003) Villidin, a novel WD-repeat and villin-related protein from Dictyostelium, is associated with membranes and the cytoskeleton. Mol Biol Cell 14: 2716–2727 Godi A, Santone I, Pertile P, Devarajan P, Stabach PR, Morrow JS, Di Tullio G, Polishchuk R, Petrucci TC, Luini A, De Matteis MA (1998) ADP ribosylation factor regulates spectrin binding to the Golgi complex. Proc Natl Acad Sci USA 95: 8607–8612 Grigoriev I, Splinter D, Keijzer N, Wulf PS, Demmers J, Ohtsuka T, Modesti M, Maly IV, Grosveld F, Hoogenraad CC, Akhmanova A (2007) Rab6 regulates transport and targeting of exocytotic carriers. Dev Cell 13: 305–314 Haghnia M, Cavalli V, Shah SB, Schimmelpfeng K, Brusch R, Yang G, Herrera C, Pilling A, Goldstein LS (2007) Dynactin is required for coordinated bidirectional motility, but not for dynein membrane attachment. Mol Biol Cell 18: 2081–2089 Harada A, Takei Y, Kanai Y, Tanaka Y, Nonaka S, Hirokawa N (1998) Golgi vesiculation and lysosome dispersion in cells lacking cytoplasmic dynein. J Cell Biol 141: 51–59 Harsay E, Schekman R (2007) Avl9p, a member of a novel protein superfamily, functions in the late secretory pathway. Mol Biol Cell 18: 1203–1219 Hawes C, Satiat-Jeunemaitre B (2005) The plant Golgi apparatus—going with the flow. Biochim Biophys Acta 1744: 93–107 Hayden SM, Wolenski JS, Mooseker MS (1990) Binding of brush border myosin I to phospholipid vesicles. J Cell Biol 111: 443–451 He CY (2007) Golgi biogenesis in simple eukaryotes. Cell Microbiol 9: 566–572 He CY, Ho HH, Malsam J, Chalouni C, West CM, Ullu E, Toomre D, Warren G (2004) Golgi duplication in Trypanosoma brucei. J Cell Biol 165: 313–321 Hehnly H, Stamnes M (2007) Regulating cytoskeleton-based vesicle motility. FEBS Lett 581: 2112–2118 Heimann K, Percival JM, Weinberger R, Gunning P, Stow JL (1999) Specific isoforms of actin-binding proteins on distinct populations of Golgi-derived vesicles. J Biol Chem 274: 10743–10750 Helfand BT, Chang L, Goldman RD (2003) The dynamic and motile properties of intermediate filaments. Annu Rev Cell Dev Biol 19: 445–467 Helfand BT, Mikami A, Vallee RB, Goldman RD (2002) A requirement for cytoplasmic dynein and dynactin in intermediate filament network assembly and organization. J Cell Biol 157: 795–806

The role of the cytoskeleton in the structure and function of GA

*

293

Henderson GP, Gan L, Jensen GJ (2007) 3-D ultrastructure of O. tauri: electron cryotomography of an entire eukaryotic cell. PLoS ONE 2: e749 Heuvingh J, Franco M, Chavrier P, Sykes C (2007) ARF1-mediated actin polymerization produces movement of artificial vesicles. Proc Natl Acad Sci USA 104: 16928–16933 Hirose H, Arasaki K, Dohmae N, Takio K, Hatsuzawa K, Nagahama M, Tani K, Yamamoto A, Tohyama M, Tagaya M (2004) Implication of ZW10 in membrane trafficking between the endoplasmic reticulum and Golgi. EMBO J 23: 1267–1278 Holappa K, Munoz MT, Egea G, Kellokumpu S (2004) The AE2 anion exchanger is necessary for the structural integrity of the Golgi apparatus in mammalian cells. FEBS Lett 564: 97–103 Holappa K, Suokas M, Soininen P, Kellokumpu S (2001) Identification of the full-length AE2 (AE2a) isoform as the Golgi-associated anion exchanger in fibroblasts. J Histochem Cytochem 49: 259–269 Hoogenraad CC, Akhmanova A, Howell SA, Dortland BR, De Zeeuw CI, Willemsen R, Visser P, Grosveld F, Galjart N (2001) Mammalian Golgi-associated Bicaudal-D2 functions in the dynein–dynactin pathway by interacting with these complexes. EMBO J 20: 4041–4054 Hook P, Vallee RB (2006) The dynein family at a glance. J Cell Sci 119: 4369–4371 Hoppeler-Lebel A, Celati C, Bellett G, Mogensen MM, Klein-Hitpass L, Bornens M, Tassin AM (2007) Centrosomal CAP350 protein stabilises microtubules associated with the Golgi complex. J Cell Sci 120: 3299–3308 Hu Y, Zhong R, Morrison WH III, Ye ZH (2003) The Arabidopsis RHD3 gene is required for cell wall biosynthesis and actin organization. Planta 217: 912–921 Infante C, Ramos-Morales F, Fedriani C, Bornens M, Rios RM (1999) GMAP-210, A cis-Golgi network-associated protein, is a minus end microtubule-binding protein. J Cell Biol 145: 83–98 Jacob R, Heine M, Alfalah M, Naim HY (2003) Distinct cytoskeletal tracks direct individual vesicle populations to the apical membrane of epithelial cells. Curr Biol 13: 607–612 Jenna S, Caruso ME, Emadali A, Nguyen DT, Dominguez M, Li S, Roy R, Reboul J, Vidal M, Tzimas GN, Bosse R, Chevet E (2005) Regulation of membrane trafficking by a novel Cdc42-related protein in Caenorhabditis elegans epithelial cells. Mol Biol Cell 16: 1629–1639 Jordens I, Marsman M, Kuijl C, Neefjes J (2005) Rab proteins, connecting transport and vesicle fusion. Traffic 6: 1070–1077 Jung E, Fucini P, Stewart M, Noegel AA, Schleicher M (1996) Linking microfilaments to intracellular membranes: the actin-binding and vesicle-associated protein comitin exhibits a mannose-specific lectin activity. EMBO J 15: 1238–1246 Jurgens G (2005) Cytokinesis in higher plants. Annu Rev Plant Biol 56: 281–299 Kaksonen M, Toret CP, Drubin DG (2006) Harnessing actin dynamics for clathrinmediated endocytosis. Nat Rev Mol Cell Biol 7: 404–414 Katsumoto T, Inoue M, Naguro T, Kurimura T (1991) Association of cytoskeletons with the Golgi apparatus: three-dimensional observation and computer-graphic reconstruction. J Electron Microsc (Tokyo) 40: 24–28 Kepes F, Rambourg A, Satiat-Jeunemaitre B (2005) Morphodynamics of the secretory pathway. Int Rev Cytol 242: 55–120 Kerkhoff E, Simpson JC, Leberfinger CB, Otto IM, Doerks T, Bork P, Rapp UR, Raabe T, Pepperkok R (2001) The Spir actin organizers are involved in vesicle transport processes. Curr Biol 11: 1963–1968 Keryer G, Di Fiore B, Celati C, Lechtreck KF, Mogensen M, Delouvee A, Lavia P, Bornens M, Tassin AM (2003a) Part of Ran is associated with AKAP450 at the centrosome: involvement in microtubule-organizing activity. Mol Biol Cell 14: 4260–4271 Keryer G, Witczak O, Delouvee A, Kemmner WA, Rouillard D, Tasken K, Bornens M (2003b) Dissociating the centrosomal matrix protein AKAP450 from centrioles impairs centriole duplication and cell cycle progression. Mol Biol Cell 14: 2436–2446

294

*

G. Egea and R. M. Ríos

Kessels MM, Dong J, Leibig W, Westermann P, Qualmann B (2006) Complexes of syndapin II with dynamin II promote vesicle formation at the trans-Golgi network. J Cell Sci 119: 1504–1516 Kessels MM, Qualmann B (2004) The syndapin protein family: linking membrane trafficking with the cytoskeleton. J Cell Sci 117: 3077–3086 Kilmartin JV, Adams AE (1984) Structural rearrangements of tubulin and actin during the cell cycle of the yeast Saccharomyces. J Cell Biol 98: 922–933 Kim HS, Takahashi M, Matsuo K, Ono Y (2007) Recruitment of CG-NAP to the Golgi apparatus through interaction with dynein–dynactin complex. Genes Cells 12: 421–434 Kim S, Coulombe PA (2007) Intermediate filament scaffolds fulfill mechanical, organizational, and signaling functions in the cytoplasm. Genes Dev 21: 1581–1597 Kondylis V, Rabouille C (2003) A novel role for dp115 in the organization of tER sites in Drosophila. J Cell Biol 162: 185–198 Kondylis V, Van Nispen tot Pannerden HE, Herpers B, Friggi-Grelin F, Rabouille C (2007) The Golgi comprises a paired stack that is separated at G2 by modulation of the actin cytoskeleton through Abi and Scar/WAVE. Dev Cell 12: 901–915 Kostenko EV, Mahon GM, Cheng L, Whitehead IP (2005) The Sec14 homology domain regulates the cellular distribution and transforming activity of the Rho-specific guanine nucleotide exchange factor Dbs. J Biol Chem 280: 2807–2817 Kroschewski R, Hall A, Mellman I (1999) Cdc42 controls secretory and endocytic transport to the basolateral plasma membrane of MDCK cells. Nat Cell Biol 1: 8–13 Kumemura H, Harada M, Omary MB, Sakisaka S, Suganuma T, Namba M, Sata M (2004) Aggregation and loss of cytokeratin filament networks inhibit Golgi organization in liver-derived epithelial cell lines. Cell Motil Cytoskeleton 57: 37–52 Kupfer A, Louvard D, Singer SJ (1982) Polarization of the Golgi apparatus and the microtubule-organizing center in cultured fibroblasts at the edge of an experimental wound. Proc Natl Acad Sci USA 79: 2603–2607 Ladinsky MS, Mastronarde DN, McIntosh JR, Howell KE, Staehelin LA (1999) Golgi structure in three dimensions: functional insights from the normal rat kidney cell. J Cell Biol 144: 1135–1149 Lanzetti L (2007) Actin in membrane trafficking. Curr Opin Cell Biol 19: 453–458 Larocca MC, Shanks RA, Tian L, Nelson DL, Stewart DM, Goldenring JR (2004) AKAP350 interaction with cdc42 interacting protein 4 at the Golgi apparatus. Mol Biol Cell 15: 2771–2781 Lazaro-Dieguez F, Colonna C, Cortegano M, Calvo M, Martinez SE, Egea G (2007) Variable actin dynamics requirement for the exit of different cargo from the trans-Golgi network. FEBS Lett 581: 3875–3881 Lazaro-Dieguez F, Jimenez N, Barth H, Koster AJ, Renau-Piqueras J, Llopis JL, Burger KN, Egea G (2006) Actin filaments are involved in the maintenance of Golgi cisternae morphology and intra-Golgi pH. Cell Motil Cytoskeleton 63: 778–791 Levine T, Rabouille C (2005) Endoplasmic reticulum: one continuous network compartmentalized by extrinsic cues. Curr Opin Cell Biol 17: 362–368 Lu Z, Joseph D, Bugnard E, Zaal KJ, Ralston E (2001) Golgi complex reorganization during muscle differentiation: visualization in living cells and mechanism. Mol Biol Cell 12: 795–808 Luders J, Stearns T (2007) Microtubule-organizing centres: a re-evaluation. Nat Rev Mol Cell Biol 8: 161–167 Luna A, Matas OB, Martinez-Menarguez JA, Mato E, Duran JM, Ballesta J, Way M, Egea G (2002) Regulation of protein transport from the Golgi complex to the endoplasmic reticulum by CDC42 and N-WASP. Mol Biol Cell 13: 866–879 Malikov V, Cytrynbaum EN, Kashina A, Mogilner A, Rodionov V (2005) Centering of a radial microtubule array by translocation along microtubules spontaneously nucleated in the cytoplasm. Nat Cell Biol 7: 1213–1218

The role of the cytoskeleton in the structure and function of GA

*

295

Malikov V, Kashina A, Rodionov V (2004) Cytoplasmic dynein nucleates microtubules to organize them into radial arrays in vivo. Mol Biol Cell 15: 2742–2749 Mardones GA, Snyder CM, Howell KE (2006) Cis-Golgi matrix proteins move directly to endoplasmic reticulum exit sites by association with tubules. Mol Biol Cell 17: 525–538 Marsh BJ, Mastronarde DN, Buttle KF, Howell KE, McIntosh JR (2001) Organellar relationships in the Golgi region of the pancreatic beta cell line, HIT-T15, visualized by high resolution electron tomography. Proc Natl Acad Sci USA 98: 2399–2406 Matanis T, Akhmanova A, Wulf P, Del Nery E, Weide T, Stepanova T, Galjart N, Grosveld F, Goud B, De Zeeuw CI, Barnekow A, Hoogenraad CC (2002) Bicaudal-D regulates COPI-independent Golgi–ER transport by recruiting the dynein–dynactin motor complex. Nat Cell Biol 4: 986–992 Matas OB, Martinez-Menarguez JA, Egea G (2004) Association of Cdc42/N-WASP/Arp2/3 signaling pathway with Golgi membranes. Traffic 5: 838–846 Matas OB, Fritz S, Luna A, Egea G (2005) Membrane trafficking at the ER/Golgi interface: functional implications of RhoA and Rac1. Eur J Cell Biol 84: 699–707 Merrifield CJ (2004) Seeing is believing: imaging actin dynamics at single sites of endocytosis. Trends Cell Biol 14: 352–358 Merrifield CJ, Perrais D, Zenisek D (2005) Coupling between clathrin-coated-pit invagination, cortactin recruitment, and membrane scission observed in live cells. Cell 121: 593–606 Miki H, Okada Y, Hirokawa N (2005) Analysis of the kinesin superfamily: insights into structure and function. Trends Cell Biol 15: 467–476 Miller DD, Scordilis SP, Hepler PK (1995) Identification and localization of three classes of myosins in pollen tubes of Lilium longiflorum and Nicotiana alata. J Cell Sci 108 (Pt 7): 2549–2563 Mizuno M, Singer SJ (1994) A possible role for stable microtubules in intracellular transport from the endoplasmic reticulum to the Golgi apparatus. J Cell Sci 107 (Pt 5): 1321–1331 Mogelsvang S, Gomez-Ospina N, Soderholm J, Glick BS, Staehelin LA (2003) Tomographic evidence for continuous turnover of Golgi cisternae in Pichia pastoris. Mol Biol Cell 14: 2277–2291 Mogensen MM, Malik A, Piel M, Bouckson-Castaing V, Bornens M (2000) Microtubule minus-end anchorage at centrosomal and non-centrosomal sites: the role of ninein. J Cell Sci 113 (Pt 17): 3013–3023 Mollenhauer HH, Morre DJ (1976) Cytochalasin B, but not colchicine, inhibits migration of secretory vesicles in root tips of maize. Protoplasma 87: 39–48 Montes de Oca G, Lezama RA, Mondragon R, Castillo AM, Meza I (1997) Myosin I interactions with actin filaments and trans-Golgi-derived vesicles in MDCK cell monolayers. Arch Med Res 28: 321–328 Moriyama Y, Nelson N (1989) Hþ-translocating ATPase in Golgi apparatus. Characterization as vacuolar Hþ-ATPase and its subunit structures. J Biol Chem 264: 18445–18450 Moseley JB, Goode BL (2006) The yeast actin cytoskeleton: from cellular function to biochemical mechanism. Microbiol Mol Biol Rev 70: 605–645 Mulholland J, Wesp A, Riezman H, Botstein D (1997) Yeast actin cytoskeleton mutants accumulate a new class of Golgi-derived secretary vesicle. Mol Biol Cell 8: 1481–1499 Musch A (2004) Microtubule organization and function in epithelial cells. Traffic 5: 1–9 Musch A, Cohen D, Kreitzer G, Rodriguez-Boulan E (2001) cdc42 regulates the exit of apical and basolateral proteins from the trans-Golgi network. EMBO J 20: 2171–2179 Musch A, Cohen D, Rodriguez-Boulan E (1997) Myosin II is involved in the production of constitutive transport vesicles from the TGN. J Cell Biol 138: 291–306 Nakamura N, Tanaka S, Teko Y, Mitsui K, Kanazawa H (2005) Four Naþ/Hþ exchanger isoforms are distributed to Golgi and post-Golgi compartments and are involved in organelle pH regulation. J Biol Chem 280: 1561–1572

296

*

G. Egea and R. M. Ríos

Nebenfuhr A, Gallagher LA, Dunahay TG, Frohlick JA, Mazurkiewicz AM, Meehl JB, Staehelin LA (1999) Stop-and-go movements of plant Golgi stacks are mediated by the acto-myosin system. Plant Physiol 121: 1127–1142 Novick P, Botstein D (1985) Phenotypic analysis of temperature-sensitive yeast actin mutants. Cell 40: 405–416 Omary MB, Coulombe PA, McLean WH (2004) Intermediate filament proteins and their associated diseases. N Engl J Med 351: 2087–2100 Oriolo AS, Wald FA, Ramsauer VP, Salas PJ (2007) Intermediate filaments: a role in epithelial polarity. Exp Cell Res 313: 2255–2264 Palokangas H, Ying M, Vaananen K, Saraste J (1998) Retrograde transport from the preGolgi intermediate compartment and the Golgi complex is affected by the vacuolar Hþ-ATPase inhibitor bafilomycin A1. Mol Biol Cell 9: 3561–3578 Pelletier L, Stern CA, Pypaert M, Sheff D, Ngo HM, Roper N, He CY, Hu K, Toomre D, Coppens I, Roos DS, Joiner KA, Warren G (2002) Golgi biogenesis in Toxoplasma gondii. Nature 418: 548–552 Percival JM, Froehner SC (2007) Golgi complex organization in skeletal muscle: a role for Golgi-mediated glycosylation in muscular dystrophies? Traffic 8: 184–194 Percival JM, Hughes JA, Brown DL, Schevzov G, Heimann K, Vrhovski B, Bryce N, Stow JL, Gunning PW (2004) Targeting of a tropomyosin isoform to short microfilaments associated with the Golgi complex. Mol Biol Cell 15: 268–280 Perrais D, Merrifield CJ (2005) Dynamics of endocytic vesicle creation. Dev Cell 9: 581–592 Picton JM, Steer MW (1981) Determination of secretory vesicle production rates by dictyosomes in pollen tubes of Tradescantia using cytochalasin D. J Cell Sci 49: 261–272 Praefcke GJ, McMahon HT (2004) The dynamin superfamily: universal membrane tubulation and fission molecules? Nat Rev Mol Cell Biol 5: 133–147 Prahlad V, Yoon M, Moir RD, Vale RD, Goldman RD (1998) Rapid movements of vimentin on microtubule tracks: kinesin-dependent assembly of intermediate filament networks. J Cell Biol 143: 159–170 Presley JF, Cole NB, Schroer TA, Hirschberg K, Zaal KJ, Lippincott-Schwartz J (1997) ER-toGolgi transport visualized in living cells. Nature 389: 81–85 Preuss D, Mulholland J, Franzusoff A, Segev N, Botstein D (1992) Characterization of the Saccharomyces Golgi complex through the cell cycle by immunoelectron microscopy. Mol Biol Cell 3: 789–803 Prigozhina NL, Waterman-Storer CM (2004) Protein kinase D-mediated anterograde membrane trafficking is required for fibroblast motility. Curr Biol 14: 88–98 Pruyne D, Legesse-Miller A, Gao L, Dong Y, Bretscher A (2004) Mechanisms of polarized growth and organelle segregation in yeast. Annu Rev Cell Dev Biol 20: 559–591 Quintyne NJ, Gill SR, Eckley DM, Crego CL, Compton DA, Schroer TA (1999) Dynactin is required for microtubule anchoring at centrosomes. J Cell Biol 147: 321–334 Ralston E (1993) Changes in architecture of the Golgi complex and other subcellular organelles during myogenesis. J Cell Biol 120: 399–409 Ralston E, Lu Z, Ploug T (1999) The organization of the Golgi complex and microtubules in skeletal muscle is fiber type-dependent. J Neurosci 19: 10694–10705 Ralston E, Ploug T, Kalhovde J, Lomo T (2001) Golgi complex, endoplasmic reticulum exit sites, and microtubules in skeletal muscle fibers are organized by patterned activity. J Neurosci 21: 875–883 Rambourg A, Clermont Y (1986) Tridimensional structure of the Golgi apparatus in type A ganglion cells of the rat. Am J Anat 176: 393–409 Rambourg A, Jackson CL, Clermont Y (2001) Three dimensional configuration of the secretory pathway and segregation of secretion granules in the yeast Saccharomyces cerevisiae. J Cell Sci 114: 2231–2239

The role of the cytoskeleton in the structure and function of GA

*

297

Rao MV, Engle LJ, Mohan PS, Yuan A, Qiu D, Cataldo A, Hassinger L, Jacobsen S, Lee VM, Andreadis A, Julien JP, Bridgman PC, Nixon RA (2002) Myosin Va binding to neurofilaments is essential for correct myosin Va distribution and transport and neurofilament density. J Cell Biol 159: 279–290 Rappleye CA, Paredez AR, Smith CW, McDonald KL, Aroian RV (1999) The coronin-like protein POD-1 is required for anterior–posterior axis formation and cellular architecture in the nematode Caenorhabditis elegans. Genes Dev 13: 2838–2851 Rehberg M, Kleylein-Sohn J, Faix J, Ho TH, Schulz I, Graf R (2005) Dictyostelium LIS1 is a centrosomal protein required for microtubule/cell cortex interactions, nucleus/ centrosome linkage, and actin dynamics. Mol Biol Cell 16: 2759–2771 Ridgway ND, Dawson PA, Ho YK, Brown MS, Goldstein JL (1992) Translocation of oxysterol binding protein to Golgi apparatus triggered by ligand binding. J Cell Biol 116: 307–319 Ridley AJ (2006) Rho GTPases and actin dynamics in membrane protrusions and vesicle trafficking. Trends Cell Biol 16: 522–529 Rios RM, Bornens M (2003) The Golgi apparatus at the cell centre. Curr Opin Cell Biol 15: 60–66 Rios RM, Sanchis A, Tassin AM, Fedriani C, Bornens M (2004) GMAP-210 recruits gammatubulin complexes to cis-Golgi membranes and is required for Golgi ribbon formation. Cell 118: 323–335 Robbins E, Gonatas NK (1964) Histochemical and ultrastructural studies on HeLa cell cultures exposed to spindle inhibitors with special reference to the interphase cell. J Histochem Cytochem 12: 704–711 Rodriguez ML, Brignoni M, Salas PJ (1994) A specifically apical sub-membrane intermediate filament cytoskeleton in non-brush-border epithelial cells. J Cell Sci 107 (Pt 11): 3145–3151 Rodriguez-Boulan E, Kreitzer G, Musch A (2005) Organization of vesicular trafficking in epithelia. Nat Rev Mol Cell Biol 6: 233–247 Rogalski AA, Bergmann JE, Singer SJ (1984) Effect of microtubule assembly status on the intracellular processing and surface expression of an integral protein of the plasma membrane. J Cell Biol 99: 1101–1109 Rohatgi R, Ho HY, Kirschner MW (2000) Mechanism of N-WASP activation by CDC42 and phosphatidylinositol 4,5-bisphosphate. J Cell Biol 150: 1299–1310 Ross JL, Wallace K, Shuman H, Goldman YE, Holzbaur EL (2006) Processive bidirectional motion of dynein–dynactin complexes in vitro. Nat Cell Biol 8: 562–570 Rossanese OW, Reinke CA, Bevis BJ, Hammond AT, Sears IB, O’Connor J, Glick BS (2001) A role for actin, Cdc1p, and Myo2p in the inheritance of late Golgi elements in Saccharomyces cerevisiae. J Cell Biol 153: 47–62 Rossanese OW, Soderholm J, Bevis BJ, Sears IB, O’Connor J, Williamson EK, Glick BS (1999) Golgi structure correlates with transitional endoplasmic reticulum organization in Pichia pastoris and Saccharomyces cerevisiae. J Cell Biol 145: 69–81 Rosso S, Bollati F, Bisbal M, Peretti D, Sumi T, Nakamura T, Quiroga S, Ferreira A, Caceres A (2004) LIMK1 regulates Golgi dynamics, traffic of Golgi-derived vesicles, and process extension in primary cultured neurons. Mol Biol Cell 15: 3433–3449 Roux A, Cappello G, Cartaud J, Prost J, Goud B, Bassereau P (2002) A minimal system allowing tubulation with molecular motors pulling on giant liposomes. Proc Natl Acad Sci USA 99: 5394–5399 Rozelle AL, Machesky LM, Yamamoto M, Driessens MH, Insall RH, Roth MG, Luby-Phelps K, Marriott G, Hall A, Yin HL (2000) Phosphatidylinositol 4,5-bisphosphate induces actin-based movement of raft-enriched vesicles through WASP-Arp2/3. Curr Biol 10: 311–320 Rybakin V, Clemen CS (2005) Coronin proteins as multifunctional regulators of the cytoskeleton and membrane trafficking. Bioessays 27: 625–632

298

*

G. Egea and R. M. Ríos

Rybakin V, Stumpf M, Schulze A, Majoul IV, Noegel AA, Hasse A (2004) Coronin 7, the mammalian POD-1 homologue, localizes to the Golgi apparatus. FEBS Lett 573: 161–167 Sahlender DA, Roberts RC, Arden SD, Spudich G, Taylor MJ, Luzio JP, Kendrick-Jones J, Buss F (2005) Optineurin links myosin VI to the Golgi complex and is involved in Golgi organization and exocytosis. J Cell Biol 169: 285–295 Saint-Jore CM, Evins J, Batoko H, Brandizzi F, Moore I, Hawes C (2002) Redistribution of membrane proteins between the Golgi apparatus and endoplasmic reticulum in plants is reversible and not dependent on cytoskeletal networks. Plant J 29: 661–678 Salas PJ, Rodriguez ML, Viciana AL, Vega-Salas DE, Hauri HP (1997) The apical submembrane cytoskeleton participates in the organization of the apical pole in epithelial cells. J Cell Biol 137: 359–375 Samaj J, Muller J, Beck M, Bohm N, Menzel D (2006) Vesicular trafficking, cytoskeleton and signalling in root hairs and pollen tubes. Trends Plant Sci 11: 594–600 Sandoval IV, Bonifacino JS, Klausner RD, Henkart M, Wehland J (1984) Role of microtubules in the organization and localization of the Golgi apparatus. J Cell Biol 99: 113s–118s Saraste J, Goud B (2007) Functional symmetry of endomembranes. Mol Biol Cell 18: 1430–1436 Satiat-Jeunemaitre B, Cole L, Bourett T, Howard R, Hawes C (1996) Brefeldin A effects in plant and fungal cells: something new about vesicle trafficking? J Microsc 181: 162–177 Scales SJ, Pepperkok R, Kreis TE (1997) Visualization of ER-to-Golgi transport in living cells reveals a sequential mode of action for COPII and COPI. Cell 90: 1137–1148 Sciaky N, Presley J, Smith C, Zaal KJ, Cole N, Moreira JE, Terasaki M, Siggia E, LippincottSchwartz J (1997) Golgi tubule traffic and the effects of brefeldin A visualized in living cells. J Cell Biol 139: 1137–1155 Schafer DA, Weed SA, Binns D, Karginov AV, Parsons JT, Cooper JA (2002) Dynamin2 and cortactin regulate actin assembly and filament organization. Curr Biol 12: 1852–1857 Schroer TA (2004) Dynactin. Annu Rev Cell Dev Biol 20: 759–779 Schulze E, Asai DJ, Bulinski JC, Kirschner M (1987) Posttranslational modification and microtubule stability. J Cell Biol 105: 2167–2177 Segui-Simarro JM, Austin JR II, White EA, Staehelin LA (2004) Electron tomographic analysis of somatic cell plate formation in meristematic cells of Arabidopsis preserved by high-pressure freezing. Plant Cell 16: 836–856 Shimmen T, Yokota E (2004) Cytoplasmic streaming in plants. Curr Opin Cell Biol 16: 68–72 Skoufias DA, Burgess TL, Wilson L (1990) Spatial and temporal colocalization of the Golgi apparatus and microtubules rich in detyrosinated tubulin. J Cell Biol 111: 1929–1937 Smythe E, Ayscough KR (2006) Actin regulation in endocytosis. J Cell Sci 119: 4589–4598 Soldati T, Schliwa M (2006) Powering membrane traffic in endocytosis and recycling. Nat Rev Mol Cell Biol 7: 897–908 Spelbrink RG, Nothwehr SF (1999) The yeast GRD20 gene is required for protein sorting in the trans-Golgi network/endosomal system and for polarization of the actin cytoskeleton. Mol Biol Cell 10: 4263–4281 Stamnes M (2002) Regulating the actin cytoskeleton during vesicular transport. Curr Opin Cell Biol 14: 428–433 Stankewich MC, Tse WT, Peters LL, Ch’ng Y, John KM, Stabach PR, Devarajan P, Morrow JS, Lux SE (1998) A widely expressed betaIII spectrin associated with Golgi and cytoplasmic vesicles. Proc Natl Acad Sci USA 95: 14158–14163 Stauber T, Simpson JC, Pepperkok R, Vernos I (2006) A role for kinesin-2 in COPIdependent recycling between the ER and the Golgi complex. Curr Biol 16: 2245–2251

The role of the cytoskeleton in the structure and function of GA

*

299

Stinchcombe JC, Majorovits E, Bossi G, Fuller S, Griffiths GM (2006) Centrosome polarization delivers secretory granules to the immunological synapse. Nature 443: 462–465 Stow JL, Fath KR, Burgess DR (1998) Budding roles for myosin II on the Golgi. Trends Cell Biol 8: 138–141 Styers ML, Kowalczyk AP, Faundez V (2005) Intermediate filaments and vesicular membrane traffic: the odd couple’s first dance? Traffic 6: 359–365 Styers ML, Salazar G, Love R, Peden AA, Kowalczyk AP, Faundez V (2004) The endolysosomal sorting machinery interacts with the intermediate filament cytoskeleton. Mol Biol Cell 15: 5369–5382 Takahashi M, Shibata H, Shimakawa M, Miyamoto M, Mukai H, Ono Y (1999) Characterization of a novel giant scaffolding protein, CG-NAP, that anchors multiple signaling enzymes to centrosome and the Golgi apparatus. J Biol Chem 274: 17267–17274 Takahashi M, Yamagiwa A, Nishimura T, Mukai H, Ono Y (2002) Centrosomal proteins CG-NAP and kendrin provide microtubule nucleation sites by anchoring gammatubulin ring complex. Mol Biol Cell 13: 3235–3245 Tassin AM, Maro B, Bornens M (1985a) Fate of microtubule-organizing centers during myogenesis in vitro. J Cell Biol 100: 35–46 Tassin AM, Paintrand M, Berger EG, Bornens M (1985b) The Golgi apparatus remains associated with microtubule organizing centers during myogenesis. J Cell Biol 101: 630–638 Thyberg J, Moskalewski S (1985) Microtubules and the organization of the Golgi complex. Exp Cell Res 159: 1–16 Thyberg J, Moskalewski S (1993) Relationship between the Golgi complex and microtubules enriched in detyrosinated or acetylated alpha-tubulin: studies on cells recovering from nocodazole and cells in the terminal phase of cytokinesis. Cell Tissue Res 273: 457–466 Thyberg J, Moskalewski S (1999) Role of microtubules in the organization of the Golgi complex. Exp Cell Res 246: 263–279 Toivola DM, Tao GZ, Habtezion A, Liao J, Omary MB (2005) Cellular integrity plus: organelle-related and protein-targeting functions of intermediate filaments. Trends Cell Biol 15: 608–617 Trucco A, Polishchuk RS, Martella O, Di Pentima A, Fusella A, Di Giandomenico D, San Pietro E, Beznoussenko GV, Polishchuk EV, Baldassarre M, Buccione R, Geerts WJ, Koster AJ, Burger KN, Mironov AA, Luini A (2004) Secretory traffic triggers the formation of tubular continuities across Golgi sub-compartments. Nat Cell Biol 6: 1071–1081 Tyska MJ, Mackey AT, Huang JD, Copeland NG, Jenkins NA, Mooseker MS (2005) Myosin1a is critical for normal brush border structure and composition. Mol Biol Cell 16: 2443–2457 Uemura T, Yoshimura SH, Takeyasu K, Sato MH (2002) Vacuolar membrane dynamics revealed by GFP-AtVam3 fusion protein. Genes Cells 7: 743–753 Uitto J, Richard G, McGrath JA (2007) Diseases of epidermal keratins and their linker proteins. Exp Cell Res 313: 1995–2009 Valderrama F, Babia T, Ayala I, Kok JW, Renau-Piqueras J, Egea G (1998) Actin microfilaments are essential for the cytological positioning and morphology of the Golgi complex. Eur J Cell Biol 76: 9–17 Valderrama F, Duran JM, Babia T, Barth H, Renau-Piqueras J, Egea G (2001) Actin microfilaments facilitate the retrograde transport from the Golgi complex to the endoplasmic reticulum in mammalian cells. Traffic 2: 717–726 Valderrama F, Luna A, Babia T, Martinez-Menarguez JA, Ballesta J, Barth H, Chaponnier C, Renau-Piqueras J, Egea G (2000) The Golgi-associated COPI-coated buds and vesicles contain beta/gamma-actin. Proc Natl Acad Sci USA 97: 1560–1565

300

*

G. Egea and R. M. Ríos

Vallee RB, Varma D, Dujardin DL (2006) ZW10 function in mitotic checkpoint control, dynein targeting and membrane trafficking: is dynein the unifying theme? Cell Cycle 5: 2447–2451 Varma D, Dujardin DL, Stehman SA, Vallee RB (2006) Role of the kinetochore/cell cycle checkpoint protein ZW10 in interphase cytoplasmic dynein function. J Cell Biol 172: 655–662 Vidali L, McKenna ST, Hepler PK (2001) Actin polymerization is essential for pollen tube growth. Mol Biol Cell 12: 2534–2545 Wang C, JeBailey L, Ridgway ND (2002) Oxysterol-binding-protein (OSBP)-related protein 4 binds 25-hydroxycholesterol and interacts with vimentin intermediate filaments. Biochem J 361: 461–472 Warner CL, Stewart A, Luzio JP, Steel KP, Libby RT, Kendrick-Jones J, Buss F (2003) Loss of myosin VI reduces secretion and the size of the Golgi in fibroblasts from Snell’s waltzer mice. EMBO J 22: 569–579 Watson P, Forster R, Palmer KJ, Pepperkok R, Stephens DJ (2005) Coupling of ER exit to microtubules through direct interaction of COPII with dynactin. Nat Cell Biol 7: 48–55 Wehland J, Henkart M, Klausner R, Sandoval IV (1983) Role of microtubules in the distribution of the Golgi apparatus: effect of taxol and microinjected anti-alphatubulin antibodies. Proc Natl Acad Sci USA 80: 4286–4290 Weiner OH, Murphy J, Griffiths G, Schleicher M, Noegel AA (1993) The actin-binding protein comitin (p24) is a component of the Golgi apparatus. J Cell Biol 123: 23–34 Welch MD, Mullins RD (2002) Cellular control of actin nucleation. Annu Rev Cell Dev Biol 18: 247–288 Welte MA (2004) Bidirectional transport along microtubules. Curr Biol 14: R525–R537 Yan X, Habedanck R, Nigg EA (2006) A complex of two centrosomal proteins, CAP350 and FOP, cooperates with EB1 in microtubule anchoring. Mol Biol Cell 17: 634–644 Yang YD, Elamawi R, Bubeck J, Pepperkok R, Ritzenthaler C, Robinson DG (2005) Dynamics of COPII vesicles and the Golgi apparatus in cultured Nicotiana tabacum BY-2 cells provides evidence for transient association of Golgi stacks with endoplasmic reticulum exit sites. Plant Cell 17: 1513–1531 Young J, Stauber T, del Nery E, Vernos I, Pepperkok R, Nilsson T (2005) Regulation of microtubule-dependent recycling at the trans-Golgi network by Rab6A and Rab6A0 . Mol Biol Cell 16: 162–177

The dynamin–cortactin complex as a mediator of vesicle formation

*

301

The dynamin–cortactin complex as a mediator of vesicle formation at the trans-Golgi network Shaun Weller, Hong Cao and Mark A. McNiven

Dynamin function The conventional dynamins represent a family of large GTPases that are encoded by at least three distinct genes in mammalian tissue and contain four conserved domains; an N-terminal highly conserved tripartite GTP-binding domain located within the first 300 amino acids, a pleckstrin homology (PH) domain of 100 amino acids, a coiled-coil (CC) region and a modestly conserved proline-rich domain (PRD) at the C terminus. Several in vitro and in vivo studies have demonstrated convincingly that dynamin binds to phosphoinositides via its PH domain (Salim et al. 1996; Zheng et al. 1996; Achiriloaie et al. 1999; Lee et al. 1999; Vallis et al. 1999), facilitating a direct interaction of dynamin with membranes. The CC domain has been characterized as a GTPase-effector domain (GED) (Sever et al. 1999), whereas the PRD has been shown to bind to multiple effector molecules (for a review of these see McNiven et al. (2000a). Substantial evidence supports the concept that dynamin is a mechanoenzyme with the ability to compress membranes into tubules and subsequently constrict these tubules into vesicles. Seminal in vitro studies have demonstrated that dynamin self-assembles to form helical ring structures (Hinshaw and Schmid 1995). Under low-salt conditions (<50 mM NaCl), dynamin polymerizes into stacks of interconnected rings resembling the necks of invaginated coated pits, suggesting a role for dynamin in the constriction and scission of budding vesicles. In accompanying studies (Takei et al. 1995), striking images of stacked collars around membrane tubules capped by clathrin-coated buds were revealed when an isolated synaptosomal fraction was treated with the non-hydrolysable GTP analog GTPgS. These collars were shown to contain dynamin, suggesting that the GTP-bound form of dynamin assembles at the neck of budding membrane surfaces, likely facilitating the restriction and tubulation of membranes to achieve liberation of endocytic vesicles. Subsequent studies have demonstrated that purified dynamin is able to tubulate and vesiculate synthetic liposomes (Sweitzer and Hinshaw 1998; Stowell et al. 1999; Takei et al. 1999). Sweitzer and Hinshaw (1998) found that, in the absence of guanine nucleotide, the addition of purified dynamin turned spherical liposomes into elongated tubules. These membranous tubules, potential fission intermediates in vesicle formation, were constricted in diameter and subsequently fragmented into vesicles following the addition of GTP. The use of a purified GTPase-defective dynamin (containing a K44A

302

*

S. Weller et al.

mutation) or the addition of GTPgS induced tubulation of liposomes, but not constriction and vesiculation, indicating that GTP hydrolysis by dynamin is the sole requirement for membrane fission. Recent studies (Roux et al. 2006; Mears et al. 2007) have demonstrated that the coiled, helical pattern of assembled dynamin protein complexes on membrane tubules undergo a GTP dependent twisting rotation. This GTP dependent spiraling and twisting action is able to generate enough longitudinal force to sever these membrane tubules into discrete vesicles. Further, this rotational spiraling of oligomerized dynamin on lengthening membrane tubules helps reconcile the GTP hydrolysis dependent change in pitch and spacing of the dynamin rings observed at the ultra-structural level (Stowell et al. 1999). Such elegant studies and experimental findings give a more clear picture of how the mechanoenzyme dynamin tubulates and severs membranes to generate vesicular carriers within the cell.

Dynamins and the trans-Golgi network Over a decade ago it was predicted that the dynamins could provide a vesicle budding/scission function at the TGN-based on the high degree of conservation with the vesiculation machinery found at the plasma membrane (PM). Indeed, as recently reviewed (McNiven and Thompson 2006) both organelles possess related classes of molecular motor enzymes, coat proteins, adaptors, and lipid modifying enzymes so it was presumed likely that Dyn2 would also function at each of these sites. This prediction was supported by multiple morphological-based “sightings” of dynamin at the Golgi of mammalian cells (Henley and McNiven 1996; Maier et al. 1996) and functional evidence for a role of Vps1p, one of two dynamin homologues in yeast, in Golgi-based transport (Rothman et al. 1990; Vater et al. 1992; Wilsbach and Payne 1993). These initial findings included the immunofluorescense and immunoelectron microscopic localization of dynamin to the Golgi apparatus using multiple dynamin antibodies (Henley and McNiven 1996; Maier et al. 1996) (Fig. 1a, c, e) and the immunoisolation of rat liver Golgi using dynamin antibody-coated beads (Henley and McNiven 1996). Further, Dynamin 2 (Dyn2) coupled to green fluorescent protein (GFP, specifically Dyn2(aa)-GFP) was observed to target to clathrin-coated pits at both the plasma membrane and the Golgi in cultured cells (Cao et al. 1998; Jones et al. 1998). U Figure 1. Localization of Dyn2 and cortactin to the Golgi apparatus. (a) Immunofluorescence staining of a Clone 9 cells with a widely used monoclonal antibody (HUDY 1) that recognizes the C-terminal tail of the dynamin proteins and a polyclonal antibody that specifically recognizes the Proline/tyrosine rich region of cortactin (b). Both antibodies, and several other reagents localize these proteins to the TGN. Immuno-electron microscopy of Dyn2 and cortactin colocalized to Golgi buds and peripheral vesicles at the trans- and cis-face. (c–e) Double immunogold labeling of Clone 9 cells with both the HUDY1 dynamin antibody (small 5 nm gold) and the C-Tyr cortactin antibody (large 10 nm gold). Both sized gold particles can be seen in close association at the tips of cisternae (arrowheads) and peripheral Golgi vesicles (arrows). Only modest labeling of the stacks (GS) is observed. Bars = 100 nm.

The dynamin–cortactin complex as a mediator of vesicle formation

*

303

304

*

S. Weller et al.

Subsequently, a variety of functional studies from multiple laboratories have demonstrated a requirement for Dyn2 in the formation of TGN-derived secretory vesicles. This includes a well characterized in vitro, cell free assay (Jones et al. 1998) in which immuno-depletion of dynamin inhibited the budding of cargo from isolated Golgi membranes. Further, reconstitution of dynamin within the reaction mixture restored cargo vesiculation from these enriched Golgi membranes, indicating a direct role for dynamin in the budding and scission events necessary for Golgi carrier formation. As

The dynamin–cortactin complex as a mediator of vesicle formation

*

305

a follow up to the in vitro findings Cao et al. (2000) utilized a mutant Dyn2 containing a point mutation in the first GTP-binding element (K44A) of Dyn2. They found that epithelial cells expressing this mutant dynamin protein display either compacted, vesiculated, or tubulated Golgi structures (Fig. 2a–h) that are defective in the transport of GFP-tagged vesicular stomatitus virus protein (VSV-G) to the plasma membrane (Fig. 3a–h). Similar findings were reported for the Golgi to plasma membrane trafficking of the apical protein p75-GFP (Kreitzer et al. 2000). As with the basolateral membrane-targeted VSV-G cargo, p75-GFP transport out of the TGN was blocked following expression of Dyn2 protein (K44E) incapable of binding GTP. Consistent with a Dyn2-K44(A/E) mediated block in scission of post-Golgi carriers from the TGN were the presence of numerous anastomosing p75-GFP tubules that extended and retracted from this compartment (Fig. 3n). The combined observations described above strongly suggest that Dyn2 associates with the Golgi apparatus in mammalian cells where it may release newly forming secretory vesicles from the TGN. There are some implications of an alternate, dynamin independent, fission mechanism for cargo exiting the TGN. Some specific cell models exhibit varying dependence on dynamin in promoting carrier liberation from the Golgi, respective of the specific cargo in tow (Damke et al.1994; Kasai et al. 1999; Bonazzi et al. 2005). A clear alternative fission mechanism within the Golgi compartment involves carboxy-terminal-binding protein 3/brefeldin Aribosylated substrate (CtBP3/BARS). The BARS protein was originally identified from a search for cytosolic factors that control Golgi tubulation (Spano et al. 1999), and has subsequently been found to regulate Golgi membrane scission (Weigert et al. 1999) albeit via an undefined mechanism (Gallop et al. 2005). Interestingly, there was a cargo-specific dependence found for dynamin and the BARS protein in driving fission events necessary for cargo carrier emergence from the TGN within polarized cell models. Delivery of U Figure 2. Cells expressing mutant Dyn2 show altered Golgi morphologies. Fluorescence images of Clone 9 cells expressing the Golgi resident marker protein TGN38-GFP. (a) Control cells expressing TGN38-GFP show normal Golgi structures situated to one side of the nucleus (N). (b–e) Cells expressing the Dyn2(K44A) mutant display a variety of altered Golgi structures. These include a tightly compressed mass of Golgi cisternae (b), an enlarged vesiculated Golgi (c), and a highly tubulated Golgi (d) in which TGN38-GFP positive tubules extend out from the Golgi cisternae to encircle the nucleus. (e) A higher magnification image of the boxed region in d shows that these tubules vary greatly in length but are of a consistent diameter. Bars,10 mm. (f–h) Clone 9 cells expressing mutant Dyn2 possess many Golgi clusters comprised of tubulated cisternae and an extensive number of associated coated vesicle buds. (f) Electron micrographs of a cell expressing Dyn2(K44A) display prominent clusters of Golgi stacks that are comprised of classic cisternae and an extensive array of associated tubules (arrowheads) and vesicles (arrows). (g, h) Higher magnification images of this cell (boxed regions) clearly show the tubulation of the cisternae. A remarkable number of densely coated vesicle buds with elongated necks (arrows) are intimately associated with the membrane tubules. Bars: (f) 0.5 mm; (g, h) 0.1 mm. (i) BHK-21 cells expressing CortDSH3 accumulate large amounts of nascent VSV-G protein and show large, distended Golgi cisternae (arrowheads), suggesting a massive retention of secretory cargo. Scale bar, 200 nm.

306

*

S. Weller et al.

Figure 3. Nascent VSV-G-GFP accumulates in the Golgi region of cells expressing mutant Dyn2 and cortactin. Fluorescence micrographs of cultured BHK cells co-transfected with plasmids encoding the secretory marker protein VSVG-ts045-GFP and either wild-type Dyn2 (a–d), mutant Dyn2K44A (e–h) or a deletion mutant of cortactin(DSH3, i–l). Cells were transfected, then incubated for 16 h at 40  C during which cells accumulated VSVG-ts045-GFP in the ER. Cells were then treated with 100 mg/ml of cycloheximide for 30 min to stop protein synthesis and then shifted to 32  C for 15, 60, or 120 min prior to fixation to allow transport of the nascent VSVG through the secretory pathway. With increased incubation times at the permissive temperature (60, 120 min) cells expressing wild-type Dyn2 (c, d) transported most, if not all, of the nascent VSVG-ts045-GFP out of the perinuclear region to the cell surface. During this same time period, cells expressing the mutant dynamin (g, h) or cortactinDSH3 (k, l) retained the nascent viral protein in a perinuclear region even after 120 min (h, l). Bar, 10 mm.

apical cargo to the PM in these polarized systems was found to depend on functional dynamin and not BARS protein. The converse was found for basolateral membrane-targeted cargo, wherein these cells relied upon BARS protein to drive carrier liberation from the TGN and not dynamin. Making this even more complex is the finding that dynamin and not BARS protein

The dynamin–cortactin complex as a mediator of vesicle formation

*

307

is required for effective Golgi to PM transport in non-polarized systems. It will be important to define the determinants that specify dynamin and BARS utilization at the TGN to drive cargo-specific vesicle scission in polarized cells.

A dynamic dynamin–actin matrix at the TGN Dynamin is known to interact with a variety of actin-binding proteins (for review see Orth and McNiven (2003), Kruchten and McNiven (2006)) and thought to serve as a ‘polymeric contractile scaffold’ at the interface between biological membranes and filamentous actin. One of the actin regulatory proteins known to interact directly with dynamin is cortactin (McNiven et al. 2000b), a rod-shaped protein with the ability to bind filamentous actin as well as interact with Arp2/3 machinery to promote dendritic actin assembly (Olazabal and Machesky 2001). Cortactin has an essential role in the formation of clathrin-coated vesicles at the plasma membrane during receptormediated endocytosis of a variety of receptor–ligand complexes (Cao et al. 2003). There is a striking localization of cortactin both at the plasma membrane and at a perinuclear compartment reminiscent of the Golgi complex. Confirmation of cortactin's association with the Golgi was achieved through standard morphological methods using a variety of immunological and molecular reagents (Cao et al. 2005) (Figs. 1b, 4c). Higher resolution analysis using Immunoelectron microscopy revealed a localized pool of cortactin protein at the tips and buds of individual Golgi cisternae as well as within a tubular–vesicular network, likely representative of the trans-face of the Golgi compartment (Fig. 1c–e). Interestingly, dynamin was also found at these same regions of the Golgi complex, often coincident with cortactin labeling on the same structures (Fig. 1c, e). Dynamin and cortactin are known to interact through the direct binding of their carboxy-terminal PRD and SH3 domains (McNiven et al. 2000). Functional studies have demonstrated that ablation of the PRD–SH3 domain interactions of dynamin and cortactin leads to a defect in the internalization of ligand stimulated endocytosis at the cell surface (Cao et al. 2003). Extending this concept to the Golgi compartment, disruption of the cortactin/dynamin interaction through the use of deletion mutants was found to compromise the Golgi to PM transport of VSV-G (Cao et al. 2005) (Fig. 3i–m). The retention of VSV-G cargo observed following expression of an SH3-domain compromised cortactin protein was coincident with the appearance of swollen and distended Golgi tubules (Fig. 2i). As the VSV-G transport assay represents a large bolus of secretory cargo entering and exiting the Golgi within a discrete timeframe, the swollen and distended Golgi membranes seen following disruption of the cortactin/dynamin interaction are consistent with a compromised fission of carriers from the Golgi. It is currently unclear whether cortactin participates in specific or multiple transport mechanisms of the Golgi. Transport of VSV-G carriers from the TGN are mediated by a class of poorly characterized, non-clathrin-coated vesicles

308

*

S. Weller et al.

The dynamin–cortactin complex as a mediator of vesicle formation

*

309

and tubules. As cortactin is in intimate association with endocytic-based clathrin-coated pits at the cell surface (Cao et al. 2003), one might predict a similar clathrin-based function at the TGN. Indeed, mannose-6-phosphate receptor transport from the Golgi to the late endosome was also affected by cortactin mutant expression, a process known to involve clathrin machinery. These findings support the concept that cortactin, together with dynamin, are important functional components of the Golgi, facilitating the fission of transport carriers. Actin cytoskeletal networks are a structural component of the Golgi compartment. Disruption of actin dynamics or filament assembly are known to affect the overall architecture of the Golgi in addition to the functional trafficking of cargo within this compartment (Egea et al. 2006). Activation of ADP-ribosylation factor-1 (Arf-1) through GTP loading leads to a dramatic assembly of actin onto isolated Golgi membranes (Fucini et al. 2000). This Arf1 sensitive pool of assembled actin is able to recruit cortactin and Dyn2 protein onto these Golgi membranes (Cao et al. 2005) (Fig. 4a and b). Treatment of these membranes with Brefeldin-A, a toxin that prevents guanine nucleotide exchange of Arf-1, prevents assembly of this actin pool and the subsequent recruitment of the cortactin/dynamin protein complex. Inclusion of the actin depolymerizing drug latrunculin-A in this in vitro reaction prevented the actin assembly on Golgi membranes as well as the cortactin/dynamin recruitment. Correlative studies were also performed in live cells where treatment with the actin depolymerizing drug cytochalasin-D, as well as the brefeldin compound, resulted in a loss of Golgi associated cortactin and Dyn2 (Fig. 4c–h). Together, these data implicate the activation of Arf1 in the U Figure 4. Binding of cortactin and dynamin to isolated Golgi membranes is dependent upon Arf1-activated actin recruitment. (a) Western blot analysis of an in vitro Golgi-binding assay was used to determine what factors are required for recruitment of cortactin and dynamin proteins to Golgi membranes. With GTP-gS in the assay, there was a marked increase in Golgi membranebound cortactin and dynamin. However, inclusion of either BrfA, to inhibit Arf1 activity, or the actin-filament-disrupting drug latrunculin A (LatA; 3 mM) resulted in a marked inhibition of actin, cortactin and dynamin recruitment to these membranes. The reassembly of Golgi membranes after the assay period was assessed by blotting for the Golgi resident protein MannII. (b) To test further the role of Arf1 in the GTP-dependent recruitment, the same assay was performed with, or without, added Arf1 protein ‘preloaded’ with GTP and in the presence or absence of BFA. As in a, there was a substantial recruitment of the complex to Golgi membranes, and the inclusion of BFA prevented this recruitment. Whereas the inclusion of preloaded Arf1-GTP (b; lane 3) to the assay had the same positive effect on complex recruitment as the addition of GTP-gS (a; lane 2), the action of preloading the Arf1 before assay addition negated the inhibitory effects of BFA on recruitment (compare a, lane 3 with b, lanes 3 and 4). For cellular studies, rat fibroblasts were treated with 0.5 mM of cytochalasin D for 30 min, fixed and double stained for cortactin (c–e) and TGN38 (c0 –e0 ). Whereas cytochalasin treatment induced some modest disorganization of the Golgi apparatus in cells, cortactin localization to the Golgi (arrows) was altered and appeared as dispersed, peripheral puncta (d, e). BFA treatment (20 mM) of Clone 9 cells induced a rapid loss of both cortactin and Dyn2 from the Golgi after just 1–2 min despite the fact that Golgi structure appeared normal by TGN38 staining; compare control cells (c, c0 ; arrowheads indicate colocalization) with BFA cells treated for 1–2 min (f, g). (h) By 15 min post-treatment the Golgi had completely dispersed, leaving a diffuse perinuclear region of cortactin. Scale bar, 10 mm.

310

*

S. Weller et al.

recruitment of cortactin/Dyn2 protein complexes to the Golgi–actin scaffold. Through the use of a variety of different approaches including the expression of truncated proteins, and microinjection of peptides or cortactin antibodies, the recruitment of Dyn2, but not cortactin, to the Golgi was altered markedly. A consistent observation was noted wherein alterations of the Dyn2 protein had no effect on cortactin localization to the Golgi, truncations of the Dyn2interacting SH3 domain of cortactin markedly reduced the levels of Golgiassociated Dyn2. Equally striking was the observation that Dyn2-PRD alone was targeted to the Golgi (Cao et al. 2005). Notably, expression of the interaction-deficient truncated Dyn2 and cortactin proteins affected the kinetics and efficacy of nascent VSV-G protein transit from the TGN to the cell surface (Fig. 3m). The findings presented here implicate cortactin as an essential component of a complex and dynamic Golgi matrix that is composed of a variety of actin-associated proteins such as specific spectrins, Abp1, the small GTPases Arf1 and Cdc42, Arp2/3-N-WASP complexes, force-generating enzymes such as myosins as well as cytoplasmic dynein and dynamin (Egea et al. 2006). Recently another dynamin interacting partner, syndapin, has been implicated in the fission events necessary for post-Golgi carrier liberation at the TGN (Kessels et al. 2006). Syndapin, in a similar manner to cortactin, binds to the PRD of Dyn2 by virtue of an SH3 domain and is also thought to interface with the Arp2/3-NWASP actin regulatory machinery. As with cortactin, the syndapin-SH3-domain-based recruitment of Dyn2 was found as crucial in the effective Golgi to PM transport of VSV-G cargo. The relative contributions of syndapin and cortactin to the Golgi localized recruitment of Dyn2 are unclear at this time, but do indicate a common recruitment strategy for the Dyn2-based fission machinery necessary in the generation of postGolgi carrier formation at the TGN.

Dynamin as a contractile scaffold mediating vesicle formation from the TGN In addition to the actin-binding proteins such as cortactin and syndapin, dynamin is known to interact with a series of proteins bearing the recently identified BAR domain (McNiven and Thompson 2006). This domain is present in many proteins with roles in membrane dynamics, including membrane tubulation and ruffling. In its simplest form, the BAR domain functions as a membrane curvature-sensing module. However, some proteins contain an N-BAR domain, where an unstructured amphipathic helix is also present N-terminal to the BAR domain. The presence of this amphipathic helix in addition to the BAR domain seems to allow these proteins to both sense and induce membrane curvature, presumably toward vesicle formation and scission. Amphiphysin, which contains an N-BAR domain, is able to bind and tubulate membranes both in vitro and in vivo and is involved in clathrin-coated vesicle formation during endocytosis (Fig. 5a). Other proteins containing BAR domains have been identified that also play a role in endocytosis, including

The dynamin–cortactin complex as a mediator of vesicle formation

*

311

Figure 5. (a) Common membrane-tubulating and cytoskeletal proteins at the PM and TGN. Cartoons showing common players utilized in vesicle formation and scission at either the plasma membrane during endocytosis or the TGN during the packaging and liberation of secretory cargo. Illustrations focus on of conserved lipid/membrane-deforming proteins (a) or additional cytoskeletal components that interact with the central dynamin contractile scaffold (b).

endophilin and sorting nexin 9. At the TGN, variants of amphiphysin and endophilin are present and exhibit functions independent of endocytosis. Sorting nexin 9, in contrast, has been demonstrated to bind to both the subunits of AP-2 and AP-1 (McNiven and Thompson 2006). Thus, a series of related ENTH/ANTH and BAR domain-containing proteins with membranedeforming properties has been superimposed on the clathrin-adaptor sorting machinery to initiate the tubulation and vesiculation of sequestered cargo from both the PM and the TGN (Fig. 5b).

Regulation of a dynamin Golgi scaffold The interaction of Dyn2 with cortactin is known to be modified by Src phosphorylation (Zhu et al. 2007). Interestingly, ligand stimulated receptor-mediated endocytosis is associated with the phosphorylation of cortactin, subsequently promoting the assembly of cortactin–Dyn2 complexes necessary for their internalization. Whether activation of these signaling cascades from the cell surface could alter Golgi structural integrity and trafficking functions is a new concept that is just being addressed. Some observations of ligand induced Golgi fragmentation have been made in pancreatic acinar cells. Dahan et al. (2005), reported a secretory agonist-based stimulation of vesiculation in ceramide enriched sub-domains of the TGN. This agonisttriggered vesiculation of Golgi membranes was found to be dependent on dynamin. Such findings suggest that the Dyn2–actin Golgi scaffold is regulated by external stimuli to control the formation of nascent secretory vesicles for transport to the cell surface. A host of important signaling molecules such as ras (Chiu et al. 2002), PIP, src kinase (Bard et al. 2003) have recently been shown to reside at the Golgi and are likely to play an important role in the activation and attenuation of the Dyn2 scaffold and subsequently protein secretion. How these cascades affect the assembly and function of this scaffold in regulated and even “constitutive” secretory cells is an important challenge to the field.

312

*

S. Weller et al.

References Achiriloaie M, Barylko B, Albanesi JP (1999) Essential role of the dynamin pleckstrin homology domain in receptor-mediated endocytosis. Mol Cell Biol 19(2): 1410–1415 Bard F, et al (2003) Src regulates Golgi structure and KDEL receptor-dependent retrograde transport to the endoplasmic reticulum. J Biol Chem 278(47): 46601–46606 Bonazzi M, et al (2005) CtBP3/BARS drives membrane fission in dynamin-independent transport pathways. Nat Cell Biol 7(6): 570–580 Cao H, Garcia F, McNiven MA (1998) Differential distribution of dynamin isoforms in mammalian cells. Mol Biol Cell 9(9): 2595–2609 Cao H, et al (2000) Disruption of Golgi structure and function in mammalian cells expressing a mutant dynamin. J Cell Sci 113(Pt 11): 1993–2002 Cao H, et al (2003) Cortactin is a component of clathrin-coated pits and participates in receptor-mediated endocytosis. Mol Cell Biol 23(6): 2162–2170 Cao H, et al (2005) Actin and Arf1-dependent recruitment of a cortactin–dynamin complex to the Golgi regulates post-Golgi transport. Nat Cell Biol 7(5): 483–492 Chiu VK, et al (2002) Ras signalling on the endoplasmic reticulum and the Golgi. Nat Cell Biol 4(5): 343–350 Dahan S, et al (2005) Agonist-induced vesiculation of the Golgi apparatus in pancreatic acinar cells. Gastroenterology 129(6): 2032–2046 Damke H, et al (1994) Induction of mutant dynamin specifically blocks endocytic coated vesicle formation. J Cell Biol 127(4): 915–934 Egea G, Lazaro-Dieguez F, Vilella M (2006) Actin dynamics at the Golgi complex in mammalian cells. Curr Opin Cell Biol 18(2): 168–178 Fucini RV, et al (2000) Activated ADP-ribosylation factor assembles distinct pools of actin on Golgi membranes. J Biol Chem 275(25): 18824–18829 Gallop JL, Butler PJ, McMahon HT (2005) Endophilin and CtBP/BARS are not acyl transferases in endocytosis or Golgi fission. Nature 438(7068): 675–678 Henley JR, McNiven MA (1996) Association of a dynamin-like protein with the Golgi apparatus in mammalian cells. J Cell Biol 133(4): 761–775 Hinshaw JE, Schmid SL (1995) Dynamin self-assembles into rings suggesting a mechanism for coated vesicle budding. Nature 374(6518): 190–192 Jones SM, et al (1998) Role of dynamin in the formation of transport vesicles from the trans-Golgi network. Science 279(5350): 573–577 Kasai K, et al (1999) Dynamin II is involved in endocytosis but not in the formation of transport vesicles from the trans-Golgi network. J Biochem 125(4): 780–789 Kessels MM, et al (2006) Complexes of syndapin II with dynamin II promote vesicle formation at the trans-Golgi network. J Cell Sci 119(Pt 8): 1504–1516 Kreitzer G, et al (2000) Kinesin and dynamin are required for post-Golgi transport of a plasma-membrane protein. Nat Cell Biol 2(2): 125–127 Kruchten AE, McNiven MA (2006) Dynamin as a mover and pincher during cell migration and invasion. J Cell Sci 119(Pt 9): 1683–1690 Lee A, et al (1999) Dominant-negative inhibition of receptor-mediated endocytosis by a dynamin-1 mutant with a defective pleckstrin homology domain. Curr Biol 9(5): 261–264 Maier O, Knoblich M,Westermann P (1996) Dynamin II binds to the trans-Golgi network. Biochem Biophys Res Commun 223(2): 229–233 Marks B, et al (2001) GTPase activity of dynamin and resulting conformation change are essential for endocytosis. Nature 410(6825): 231–235 McNiven MA, et al (2000a) The dynamin family of mechanoenzymes: pinching in new places. Trends Biochem Sci 25(3): 115–120 McNiven MA, et al (2000b) Regulated interactions between dynamin and the actinbinding protein cortactin modulate cell shape. J Cell Biol 151(1): 187–198

The dynamin–cortactin complex as a mediator of vesicle formation

*

313

McNiven MA, Thompson HM (2006) Vesicle formation at the plasma membrane and trans-Golgi network: the same but different. Science 313(5793): 1591–1594 Mears JA, Ray P, Hinshaw JE (2007) A corkscrew model for dynamin constriction. Structure 15(10): 1190–1202 Olazabal IM, Machesky LM (2001) Abp1p and cortactin, new “hand-holds” for actin. J Cell Biol 154(4): 679–682 Orth JD, McNiven MA (2003) Dynamin at the actin-membrane interface. Curr Opin Cell Biol 15(1): 31–39 Rothman JH, et al (1990) A putative GTP binding protein homologous to interferoninducible Mx proteins performs an essential function in yeast protein sorting. Cell 61(6): 1063–1074 Roux A, et al (2006) GTP-dependent twisting of dynamin implicates constriction and tension in membrane fission. Nature 441(7092): 528–531 Salim K, et al (1996) Distinct specificity in the recognition of phosphoinositides by the pleckstrin homology domains of dynamin and Bruton’s tyrosine kinase. EMBO 15(22): 6241–6250 Sever S, Muhlberg AB, Schmid SL (1999) Impairment of dynamin’s GAP domain stimulates receptor-mediated endocytosis. Nature 398(6727): 481–486 Shpetner HS, Vallee RB (1989) Identification of dynamin, a novel mechanochemical enzyme that mediates interactions between microtubules. Cell 59(3): 421–432 Spano S, et al (1999) Molecular cloning and functional characterization of brefeldin A-ADP-ribosylated substrate. A novel protein involved in the maintenance of the Golgi structure. J Biol Chem 274(25): 17705–17710 Stowell MH, et al (1999) Nucleotide-dependent conformational changes in dynamin: evidence for a mechanochemical molecular spring. Nat Cell Biol 1(1): 27–32 Sweitzer SM, Hinshaw JE (1998) Dynamin undergoes a GTP-dependent conformational change causing vesiculation. Cell 93(6): 1021–1029 Takei K, et al (1995) Tubular membrane invaginations coated by dynamin rings are induced by GTP-gamma S in nerve terminals. Nature 374(6518): 186–190 Takei K, et al (1999) Functional partnership between amphiphysin and dynamin in clathrin-mediated endocytosis. Nat Cell Biol 1(1): 33–39 Vallis Y, et al (1999) Importance of the pleckstrin homology domain of dynamin in clathrin-mediated endocytosis. Curr Biol 9(5): 257–260 Vater CA, et al (1992) The VPS1 protein, a homolog of dynamin required for vacuolar protein sorting in Saccharomyces cerevisiae, is a GTPase with two functionally separable domains. J Cell Biol 119(4): 773–786 Weigert R, et al (1999) CtBP/BARS induces fission of Golgi membranes by acylating lysophosphatidic acid. Nature 402(6760): 429–433 Wilsbach K, Payne GS (1993) Vps1p, a member of the dynamin GTPase family, is necessary for Golgi membrane protein retention in Saccharomyces cerevisiae. EMBO J 12(8): 3049–3059 Zheng J, et al (1996) Identification of the binding site for acidic phospholipids on the pH domain of dynamin: implications for stimulation of GTPase activity. J Mol Biol 255 (1): 14–21 Zhu J, et al (2007) Receptor-mediated endocytosis involves tyrosine phosphorylation of cortactin. J Biol Chem 282(22): 16086–16094

314

*

J. Derganc et al.

The geometry of organelles of the secretory pathway Jure Derganc, Alexander A. Mironov and Saša Svetina

Introduction The organelles of the secretory pathway have distinctive geometries, and yet in fact they are all composed of similar basic shapes. Thus, the membrane geometries encountered in the secretory organelles can be classified into spheres, cylinders and flattened cisternae. Typically, the organelles are complex combinations of these basic figures and they may also possess perforations. For example, the endoplasmic reticulum (ER) is a complex system of interconnected cylinders and discs, which are often flattened and perforated. The Golgi complex appears as a membrane structure that is composed of stacks of flattened cisternae. These cisternae contain perforations, often carry tubular extensions, and are surrounded by a number of small spherical vesicles. Post-Golgi carriers are composed of tubular–saccular structures (also comprising ovoid geometry) and, finally, the plasma membrane (PM) corresponds to a planar undulated bilayer. Clearly, the shape of all these forms depends on the interactions between different physicochemical entities and is also closely related to the function of the organelles. In order to recognize a general pattern in the observed richness of organelle geometries, we shall follow a bottom-up approach by first revealing the most basic features that define the organelle geometry and then attempting to analyze how the general properties are reflected in the complex reality of secretory organelles. As the structural and functional backbone of these organelles is their lipid membrane, this chapter will thus focus to a large extent on the physical properties of the lipid bilayer. For complementary aspects of the organelle geometry the reader is directed to a number of recent reviews on the topic (Malhotra and Yaffe 2005; McMahon and Gallop 2005; Zimmerberg and Kozlov 2006; Voeltz and Prinz 2007). Historically, the main picture of membrane physical properties was established in the 1970s and 1980s, and was consequently supported by extensive experimental data on model lipid membranes – lipid vesicles – that was acquired during the 1980s and 1990s. The boom in molecular biology which began in the 1990s is also providing an immense amount of information on the relation between the molecular basis of secretory organelles and their function. Together with advancements in optical microscopy and highresolution EM tomography, which provides high-resolution 3D images of secretory organelles (Ladinsky et al. 1999), it is now possible to start building a

The geometry of organelles

*

315

complete picture connecting the function of organelles with their macroscopic geometry and microscopic structure. In this chapter we will first provide an overview of the macroscopic physical properties of lipid membranes, with special attention to the properties that are important for the membrane geometry. These macroscopic properties will then be linked to a diversity of molecular mechanisms which affect the membrane geometry, and some examples of these mechanisms in the secretory pathway will be given (a more detailed overview of the molecular aspects will be covered in other chapters of this book). Finally, the current understanding of the geometry of the Golgi complex will be discussed.

Physical basis of membrane shape (macroscopic picture) Basic concepts The lipid membrane is composed of lipid molecules arranged in a bilayer with the hydrophobic lipid tails hidden in the inside of the membrane and their hydrophilic heads oriented towards the aqueous solution. In this arrangement, the structural integrity of the bilayer is provided by hydrophobic forces, even in the absence of strong interactions between the lipid molecules. This distinctive structural feature is the basis for the unique physical properties of the membrane. For example, the unconstrained membrane constituents are freely mobile in the lateral direction. Thus, biological membranes can be regarded as a two-dimensional fluid consisting of membrane proteins that are floating in a fluid lipid bilayer, a fact recognized in the fluid mosaic model (Singer and Nicolson 1972). Likewise, the lipid bilayer is practically incompressible and on the other hand, it possesses a very small bending rigidity, and typically exhibits thermal bending fluctuations (the so-called flickering or thermal undulations). Interactions between the membrane constituents may lead to various degrees of non-random lateral membrane organization (Engelman 2005). For example, certain constituents of biological membranes may in fact segregate and form different membrane phases (domains) (Simons and Ikonen 1997). In addition, lateral organization in biological membranes can be provided by external factors such as the protein cytoskeletons (McMahon and Gallop 2005). Accordingly, the diffusion coefficients of membrane constituents in biological membranes can be much smaller than those in model lipid membranes exhibiting a truly fluid behaviour.

Membrane curvature and area difference between the membrane layers As stated above, the main deformational mode of the lipid membranes is their bending. The extent of membrane bending at a given point on the membrane surface is described in terms of membrane curvature. Since a membrane can be independently bent in two perpendicular directions, two independent curvatures can be defined at each point, denoted as principal

316

*

J. Derganc et al.

curvatures C1 and C2. The curvature in a given direction (i ¼ 1, 2) is equal to the reciprocal of the radius of the circle osculating the surface in that direction, Ci ¼ 1=Ri . By definition the curvature of a convexly curved membrane (with respect to the vesicle interior) is positive and the curvature of a concavely curved membrane is negative. Often, it is convenient to express the membrane curvature in terms of two combinations of the principal curvatures: the total membrane curvature, which is the sum of the principal curvatures C1 þ C2, and the Gaussian curvature, defined as the product of the principal curvatures C1C2. In the secretory organelles, the typical maximal principal curvatures are of the order of 1/(25 nm), with 25 nm being approximately the radius of curvature of various relevant membrane geometries, e.g., the COPI vesicles, various tubular structures, and the rims of flattened Golgi cisternae. Since the total and the Gaussian curvatures depend on the signs of the principal curvatures they can vary substantially between different membrane geometries (Fig. 1 and Table 1). For example, in vesicles both principal curvatures are positive and thus, small vesicles have a large total curvature and a large positive Gaussian curvature. In cylinders, one of the principal curvatures is zero and typical cylinders thus have twice as small total curvature as COPI vesicles, and zero Gaussian curvature. The rims of large flattened compartments (e.g., Golgi cisternae) can be regarded as bent half-cylinders, cut in the axial direction. Thus, one principal curvature of the cisternal rims is as large as the one in the COPI vesicles, while the other is small, but still larger than zero. Membranes in the narrow necks of buds and in the rims of cisternal perforations are bent in two opposite directions and thus, these saddle geometries have a small total curvature and a strongly negative Gaussian curvature. Finally, the curvatures in flat membranes (e.g., flat parts of Golgi cisternae) are close to zero.

Figure 1. Typical membrane geometries in the secretory organelles (for the corresponding membrane curvatures see Table 1). (A) Spherical membrane (e.g., a vesicle) has a positive total membrane curvature. (B) Cylindrical membrane (e.g., a membrane tubule). Here one of the principal curvatures is positive, and the other is zero. (C) Flattened perforated membrane compartment (e.g., a Golgi cisterna) possesses membrane of three different curvatures. Both curvatures in the compartment’s outer rim (1) are positive (one is larger than the other). In the flat sides, both curvatures are zero (3). In the rim of the cisternal perforation (2), one curvature is positive and the other negative. (D) Membrane neck (e.g., in a budding vesicle). Here the two principal curvatures have opposite signs, resulting in a vanishing total membrane curvature, and a negative Gaussian curvature.

The geometry of organelles

*

317

Table 1. Typical membrane curvatures in secretory organelles Total curvature

Gaussian curvature

TAA = (11 DA A)

Typical TAA (R 5 25 nm, h 5 4 nm)

Vesicle (radius R)

1 R

þ R1

1 R2

1 þ 2h R

1.32

Invaginated vesicle (radius R)

 R1  R1

1 R2

1  2h R

0.68

Cylinder (radius R)

1 R

þ0

0

1 þ hR

1.16

Rim of a discoid cisterna (disc radius Rd )

1 R

þ R1d

1 1 R Rd

1 þ hR þ Rhd

1.16

Flat cisternal side

0

0

1

1

Rim of a perforation (perforation radius Rp )

1 R



 R1 R1p

1þ 

Neck

1 R

 R1 ¼ 0

 R12

1

1 Rp

h R

h Rp

1
All values are given in terms of the typical maximal radius of curvature R encountered in the secretory organelles ðR ¼ 25 nmÞ. The value of the distance h between the neutral planes of the two membrane leaflets is approximated by the distance between the planes of the polar lipid heads in the bilayer. TAA: transbilayer area asymmetry.

Since the lipid membrane is a bilayer, its bending is closely related to another property. Namely, when the bilayer is bent, its outer layer undergoes stretching while the inner layer undergoes compression. Vice versa, if for a certain reason the surface area of one of the layers changes relative to the other one, the membrane will bend accordingly. Locally, the difference in the surface areas of the membrane layers is directly related to the total membrane curvature, (C1 þ C2)h, where h is the distance between the neutral planes of the two membrane layers. The neutral plane of a layer is defined as the plane (in the layer) that does not stretch when the layer is subject to a pure bending deformation. Clearly, the distance between the neutral planes of the membrane leaflets is related to the membrane thickness; for example, h can be taken to be the distance between the planes of the polar lipid heads. The total area difference in a closed membrane (DA) can be obtained by integrating the total curvature over the whole membrane surface R ðAÞ; DA ¼ hðC1 þ C2 ÞdA. Thus, the area difference DA of a closed membrane compartment is directly related to its shape. The area difference can be also expressed as the transbilayer area asymmetry (TAA), which describes the ratio between the surface areas of the outer and the inner monolayers, TAA ¼ 1 þ DA A . As TAA is proportional to the total area difference, it is largest in vesicles and smallest in flat membranes and in saddle-like membrane in the rims of the perforations. For example, in COPI vesicles, the surface area of the

318

*

J. Derganc et al.

cytosolic membrane leaflet is more than 30% larger than the surface area of the inner one. The smaller the vesicle, the larger is its curvature and TAA.

Membrane elastic energy There are three independent terms in the total membrane elastic energy. Two are related to the two independent membrane curvatures (Helfrich 1973), and the third to the membrane TAA (Bozic et al. 1992): (1)

(2)

(3)

The first energy term is often denoted simply as the membrane bending energy (Helfrich 1973). It describes the energy that is needed to bend a membrane from its natural state. Locally, the membrane bending energy is proportional to Wb / kc ðC1 þ C2  C0 Þ2 , where kc is the membrane bending modulus and C0 the membrane spontaneous curvature which describes the curvature that a free unconstrained membrane patch would assume. The value of the spontaneous curvature is related to the bilayer asymmetry and will be discussed further in the next section. Importantly, in closed membrane compartments, there are in general geometric constraints which prevent the membrane from fully assuming its spontaneous curvature. The value of the bending modulus of the Golgi membrane was measured to be of the order of kc 3 1019 J (Upadhyaya and Sheetz 2004). The second term of the membrane elastic energy is proportional to the Gaussian membrane curvature (Helfrich 1973). Locally the Gaussian bending energy is proportional to WG / kc C1 C2 , where kc is the Gaussian bending constant. The Gaussian modulus kc of a membrane monolayer is supposedly negative, however measurements of its value in biological membranes are rare (Siegel and Kozlov 2004). An important property of this term is that, according to the Gauss–Bonnet theorem, the total Gaussian bending energy of homogeneous closed membrane compartments depends only on the topology of the compartment and not on the actual membrane shape: WG ¼ 4pkc ð1  gÞ, where g is the compartment’s topological genus describing how many “holes” (perforations) a compartment has. For example, spheres have no perforations and genus g ¼ 0. Compartments with one perforation, i.e., the topology of a torus, have genus g ¼ 1, compartments with two perforations have genus g ¼ 2 and so on. Clearly, this term should be considered as very important for the secretory organelles because their topology changes constantly due to continuous membrane fusion, fission and formation of perforations. The third important term of the membrane elastic energy is closely related to the area difference and the possible relative stretching of the layers (since the layers can freely slide one over the other, they equilibrate their lateral stresses independently and, at equilibrium, one of the layers may remain stretched and the other compressed). This term is denoted as the non-local bending energy (Bo zic et al. 1992) or as the area

The geometry of organelles

*

319

difference elasticity (ADE) term (Miao et al. 1994), and is proportional to WADE / kr ðDA  DA0 Þ2 . Here, kr is the non-local bending constant and the preferred membrane area difference DA0 describes the area difference in an unstressed membrane patch. The preferred membrane area difference DA0 is closely related to the number of molecules in each leaflet. For example, if the number of molecules in the membrane outer leaflet increases, DA0 increases, and vice versa. The non-local bending constant kr is approximately three times larger than the membrane bending constant, kr 3 kc (Waugh et al. 1992; Hwang and Waugh 1997; Svetina et al. 1998).

Equilibrium membrane shapes First, it is instructive to focus on the shapes of homogeneous closed membrane compartments in the absence of external forces. This problem has been studied extensively, both theoretically and experimentally, in the case of lipid vesicles. In closed membrane compartments the membrane is constrained geometrically and, in addition, its shape does not depend on the Gaussian bending energy. Thus, the membrane will take the shape in which the sum of its bending and ADE energies is minimal. It was shown that the shapes of closed compartments can fall into several different shape classes, depending on their area to volume ratio and their membrane area difference (Fig. 2)

Figure 2. Schematic representation of the typical shapes of closed homogeneous membrane compartments. In general, the shape of a compartment is determined by its area difference DA and its relative volume v. The relative volume is a convenient way to describe the area to volume ratio in a dimensionless form; the larger the area to volume ratio (A/V), the smaller is the relative pffiffiffiffi volume (v ¼ 6V p=A3/2). The dashed vertical lines represent axes of symmetry. Note that the diagram is not in scale (see, Svetina and Zekš (2002)). Increased membrane spontaneous curvature C0 or preferred area difference DA0 drives the shapes towards larger values of DA and vice versa. The limiting shapes at large DA consist of evaginated spheres, whereas the limiting shapes at small DA consist of spheres with invaginations. The spheres of the shapes with extreme DA may only have two different radii (Svetina and Zekš 1989). The presented non-axisymmetric compartments have the flattened starfish-like shape, which can be regarded as a combination of an oblate (i.e., flattened) and several prolate (i.e., elongated) structures.

320

*

J. Derganc et al.

 (Svetina and Zekš 1989, 2002; Seifert et al. 1991). Interestingly, the equilibrium shapes of membrane compartments do not depend on the absolute values of the membrane elastic moduli but, rather, the value of the ratio kr/kc influences the transitions between different shape classes. The membrane spontaneous curvature and its preferred area difference affect the membrane shape in a similar way: a large increase in the spontaneous curvature or in the preferred area difference leads to membrane evaginations, while a large decrease of one of these parameters leads to membrane invaginations (in fact, these two parameters can be combined into a single one, i.e., an effective area difference DA0 ¼ DA0 þ kkcr hAC0 ). It is noteworthy that, at large area to volume ratios, the flattened discoid shapes, such as a single Golgi cisterna, do not belong to a stable class. Here, depending on the preferred area difference, the two flat sides of a cisterna stick together on the inside, or the whole compartment bends and closes into a cup-like shape, thus reducing the surface area of the highly curved rims (Majhenc et al. 2004). Also, at large area to volume ratios and sufficiently large preferred area differences, various non-axisymmetric shapes emerge as stable, such as boomerangs, rackets and starfish, which can be understood as being composed of combinations of oblate (i.e., flattened) and prolate (i.e., elongated) parts (Ziherl and Svetina 2005), and simple prolate shapes like worms and necklaces. External forces may significantly change the equilibrium membrane shape, and often an elaborate analysis is needed in order to completely describe these shapes. Because of the two-dimensional fluid character of the lipid membrane, its geometry is affected most by the forces acting perpendicular to the membrane surface. When a small point force is applied perpendicular to the membrane, the membrane deforms locally and, if the force exceeds a certain critical threshold, a thin membrane tubule (a tether) is pulled out of the membrane (Bozic et al. 1997; Heinrich et al. 1999). Typically, pulling a membrane tubule requires a force of several 10 pN, which is more than forces exerted by single molecular motors (Koster et al. 2003). When many point-like forces are applied to a given membrane surface, e.g., by a protein skeleton with a number of attachment points, the forces act more as a surface force (i.e., force per unit area) and can thus affect the membrane geometry on a larger scale without pulling tubes out of the membrane. For example, adhesion between two adjacent membranes can result from a number of adhesion molecules connecting the two membranes. In general, the forces between two membranes may be attractive or repulsive, resulting from Van der Waals, electrostatic and even entropic interactions. A typical force acting on the membrane surface is also the osmotic transmembrane pressure. If the membrane of a closed compartment is not homogeneous, its shape becomes dependent on the lateral distribution of the membrane constituents. In general, the two properties are closely intertwined: the membrane shape influences the lateral distribution and the lateral distribution has a

The geometry of organelles

*

321

certain effect on the membrane shape. Although several theoretical approaches have been proposed to deal with the shapes of laterally inhomogeneous membranes (Markin 1981; Seifert 1993; Lipowsky 1992; Bo zic et al. 2006), the complexity of this problem still renders the application of any general results to biological membranes a challenging task.

Molecular factors important for membrane shape (microscopic picture) Although the general physical properties of vesicular objects can be attributed primarily to the universal lipid bilayer structure of their membranes, the details that fine-tune the biological function clearly arise from the differences in the microscopic structure. As it is impossible to cover all the different molecular aspects in one short section, only the most prominent features that define the membrane geometry will be discussed here.

Membrane composition and intermolecular interactions Since the secretory organelles play a central role in lipid and protein production and sorting, it is not surprising that the composition of their membranes varies substantially. To begin at the cellular level, there are several lines of evidence indicating a gradient of membrane composition between ER and PM. In general, the level of unsaturated lipids decreases and the level of sphingolipids increases from ER to PM. For example, the amounts of sphingomyelin (SM) and phosphatidylserine (PS) are 2.2- and 1.5-fold larger in the trans-Golgi network (TGN) than in cis-Golgi network (CGN), respectively. For cholesterol this factor is 1.7 (Cluett and Machamer 1996). At a lower level, evidence for the variation of membrane composition is much harder to obtain, yet there are strong indications of significant variations, even within a single membrane compartment, which are a consequence of aggregation and lateral separation of certain membrane constituents. Notably, it is now established that cholesterol, glycolipids, sphingolipids and certain proteins may segregate and form membrane microdomains (the socalled membrane rafts) (Simons and Ikonen 1997), while further lateral organization is believed to be produced by external protein scaffolds or influenced by membrane curvature (McMahon and Gallop 2005; Zimmerberg and Kozlov 2006). For example, in artificial membranes composed of the right proportions of cholesterol, dioleoyl-phosphatidylcholine (DOPC) and SM, the liquid-disordered phase (enriched by DOPC) readily separates from the liquidordered phase (enriched by SM) (Dietrich et al. 2001). In addition to the lateral membrane organization, there can be an asymmetry in composition between the cytoplasmic and exoplasmic membrane leaflets. In PM, for example, the choline-containing PC and SM are highly enriched on the exoplasmic leaflet of the membrane bilayer while the aminophospholipids, such as PS and phosphatidylethanolamine (PE), are enriched on the cytoplasmic leaflet (Hanada and Pagano 1995).

322

*

J. Derganc et al.

Lipid composition and related intermolecular interactions directly affect the macroscopic physical properties that are important for the membrane geometry. For example, the thickness of the membranes of the ER, the Golgi, and apical PM, purified from rat hepatocytes, were determined to be    37.50.4 A, 39.50.4 A, and 42.50.3 A, respectively (the thickness of the  basolateral membrane was found to be 35.60.6 A) (Mitra et al. 2004). However, the same study showed that membrane thickness is not a direct consequence of ordering of the acyl chains of phospholipids due to increased cholesterol content, but that protein-related mechanisms may be involved, indicating that the relation between the membrane composition and its thickness is far from simple.

Factors affecting local membrane properties Clearly, all the membrane elastic moduli (kc, kc and kr), its thickness and its spontaneous curvature C0 are material constants and thus depend on the local membrane composition and intermolecular interactions (of note, the preferred area difference DA0 is an integral property of the closed membrane). It was for example experimentally verified that disordered membranes with short polyunsaturated chains (e.g., in ER) are more flexible than ordered membranes with long acyl chains (e.g., in PM) (Rawicz et al. 2000). The bending modulus of the liquid ordered phase was measured to be approximately two times larger than in the liquid disordered phase (Roux et al. 2005). One of the most direct effects on membrane spontaneous curvature comes from the intrinsic shape of its constituents (lipids as well as proteins). For example, cylindrical molecules form flat monolayers, while monolayers composed of conical molecules will have a tendency to bend (Fig. 3A and B). The role of molecular shape on macroscopic membrane curvature can be described in terms of the preferred molecular curvature, defined as the curvature assumed spontaneously by an unconstrained single-composition membrane monolayer (Zimmerberg and Kozlov 2006). Because the two monolayers in the membrane have opposing orientations, their curvatures have opposite signs. The spontaneous curvature of the bilayer, which is the sum of the spontaneous curvatures of the two monolayers, is therefore zero if the two monolayers have identical compositions (Fig. 3C). For lipids, the structural feature that predominantly influences their shape is the relative size of the polar and the non-polar molecular parts (Israelachvili et al. 1980). In particular, the shape is affected by the complexity of their hydrophilic heads, the number of hydrophobic tails, their length and their saturation. For example, lysolipids only have one hydrophobic tail, and therefore their shape resembles an inverted cone, so they prefer strongly positive curvatures. On the other hand, diacylglycerol (DAG), with its small head and two tails (of which one is unsaturated), prefers large negative curvatures. Importantly, the effective shape of lipid molecules in a bilayer also depends directly on the intermolecular interactions, e.g., the electro-

The geometry of organelles

*

323

Figure 3. Schematic representation of the role of the intrinsic molecular properties in the membrane spontaneous curvature. (A) Depending on the relative sizes of their hydrophobic and hydrophilic parts, the membrane constituents can have a cylindrical (e.g., PC), a conical shape (e.g., PE) or a shape of an inverted cone (e.g., lysolipids). (B) Cylindrical constituents form flat monolayers and conical constituents form bent monolayers. When a monolayer is composed of different constituents its spontaneous curvature is an average value between the values of the single-composition monolayers. (C) The spontaneous curvature of a symmetric bilayer is zero, regardless of its composition. A non-vanishing spontaneous curvature appears as a consequence of an asymmetry between the two leaflets in the bilayer.

static interactions between the charged lipid heads or lateral stresses in the membrane, and is reflected for example in the membrane fluidity. Note that for certain classes of lipids there is a close relationship between their shape and the type of lipid phase they form spontaneously in the solution. For instance, lysolipids form micelles, while cylindrically shaped PCs form bilayers. The lipids in secretory organelles are subject to extensive lipid metabolism, much of which alters their shape and thus has a direct influence on the membrane shape. For example, enzymes lipases cleave off various parts of phospholipids. Phospholipase A2 (PLA2) cleaves phospholipids into a fatty acid and wedge-shaped lysophospholipid, thus increasing local lysophospholipid concentrations (de Figueiredo et al. 1998). Phospholipase C (PLC) induces conversion of PIP2 into DAG (strongly conical shape). Phospholipase D (PLD) catalyzes the hydrolysis of PC to form phosphatidic acid (PA). PA can be further hydrolyzed to DAG. PA can also be formed by acylation of lysophosphatidic acid (LPA) with acyl-coenzyme A (acyl-CoA). This process is catalyzed by DtBP/BARS, a protein that affects the dynamics of Golgi tubules (Weigert et al. 1999). Inositol lipid kinases may also modulate the structure of the lipid bilayer (De Matteis et al. 2002).

324

*

J. Derganc et al.

Proteins may influence the local membrane properties, both as intrinsic factors via their transmembrane domain (TMD) or as external factors binding to the membrane surface. In general, the thickness of the membrane is related to the length of the protein TMD, and the “roughness” of the TMD correlates with the state of membrane order. For example, since the TMD of Sec22p is very short (15 amino acids in length), it localizes at the ER exit sites. Alternatively, shortening of TMD in VSV-G, a PM resident protein, induces its mislocalization (Cole et al. 1998). It has been noted that Golgi resident proteins contain more phenylalanine than those from the PM (Bretscher and Munro 1993). This is in a good agreement with the higher concentration of cholesterol in the PM, which makes the bilayer more ordered and thus renders less unfavourable the insertion of proteins containing disordering motifs such as phenylalanine. Any differences between the length of protein TMD and the bilayer thickness lead to the so-called hydrophobic mismatch, which may in turn affect both the protein and the lipid bilayer (Killian 1998). Proteins may also affect the local membrane properties as external factors. For example, curved proteins may bind to the membrane and impose their curvature on the membrane (Zimmerberg and Kozlov 2006). Indeed, it has recently been reported that the binding of crescent-shaped BAR domains to the membrane plays a role in the formation of membrane tubules, and that the resulting spontaneous membrane curvature might depend on the fraction of the membrane surface area covered by BAR (Blood and Voth 2006).

Factors affecting the preferred membrane area difference The equilibrium difference between the surface areas of the membrane layers is related to the surface area of individual molecules (i.e., the molecular shape; see previous section) and the number of molecules in each layer. There are several ways of changing the number of molecules in the layers. The first is the translocation of lipids from one membrane leaflet to another (flip-flop). In artificial membranes, the spontaneous flip-flop of lipids with large or charged head groups, i.e., glycolipids and phospholipids, is very slow (t1/2 of many hours or days) whereas molecules like fatty acids, cholesterol, DAG and ceramide flip in seconds (Van Helvoort and Van Meer 1995; Hamilton 2003). In biological membranes the lipid translocation is accelerated to a biologically relevant rate by protein catalysts denoted by the general term flippases (Pomorski and Menon 2006; Daleke 2007). Most of these transporters are still not identified. Often they are called flippases if they move the lipids to the cytoplasmic side, floppases if they move the lipids in the opposite direction, and scramblases if they transport the lipids in either direction. The activity of the transporters may or may not depend on ATP. For example, in ER membranes extracted from rat liver a fast (t1/2 of the order of several seconds), ATP-independent redistribution of PC and PE has been observed (Marx et al. 2000; Buton et al. 2002). The second way of affecting the TAA of membranes is by addition of lipid molecules from the cytosol via the aqueous phase or by intercalation of protein domains in only one leaflet (e.g., the hydrophobic part of the BAR

The geometry of organelles

*

325

domain). Lipid molecules may be added to the cytosolic leaflet through the aqueous phase with the help of specific (i.e., phospholipid transfer protein) or non-specific (i.e., albumin and non-specific lipid transfer proteins) lipid transporters (Fang et al. 1998; Wirtz 2006) or simply by solubilization of some molecules like fatty acids (Cullis et al. 1996). For example, phosphatidylinositol-transfer protein (Sec14p) catalyses the export of PI from the trans-Golgi membrane (Kearns et al. 1997; Wirtz 2006). Finally, factors that markedly affect the membrane area difference in secretory organelles are the fusion and the fission of membranes. For example, since there is a large excess of TAA in small vesicles, the fusion of other membrane compartments with vesicles necessarily leads to an increase in the compartments’ TAA. And vice versa, any vesiculation from a compartment decreases its TAA.

Protein coats and external forces When proteins bind to a membrane and remain laterally mobile (e.g., by BAR domain), their influence can be considered in terms of altered intrinsic membrane properties, i.e., an altered spontaneous curvature and preferred area difference. However, if the bound proteins form a rigid scaffold, their role can be described as a source of external forces exerted on the membrane. In the secretory pathway, there are many protein coats that may influence the membrane in both ways, namely COPI, COPII, dynamin, clathrin/AP1, clathrin/ AP2, AP3, AP4. For example, bound clathrin monomers initially diffuse freely in the membrane and then form a rigid cage which acts as an external scaffold for vesiculation. Secretory organelles are known to be associated with cellular protein skeletons, such as microtubules, actin filaments (Egea et al. 2006) or spectrin (Beck et al. 1994), but the role of these cellular skeletons in providing the scaffold for the membrane of secretory organelles is not understood.

The shape of the Golgi complex The most prominent geometrical feature of the Golgi complex is a stack of flattened cisternae with a large area to volume ratio. The cisternae are surrounded by a number of small vesicles and can possess perforations. The tubular extensions may connect cisternae within a single stack or cisternae of adjacent stacks. In the Golgi, the regions of relatively flat membrane (in the flat cisternal sides) coexist with the regions of highly curved membrane in the (possibly undulated) cisternal rims, the perforation rims, vesicles and tubular extensions. In general the Golgi structure varies to some extent between different cell types, but its size and the number of cisternae appear to be relatively stable parameters within a given cell type. Although the complete picture of the relations between the Golgi shape and its function remains unknown (e.g., the function of the cisternal perforations or tubular extensions), recent systematic studies have revealed some important features.

326

*

J. Derganc et al.

When analyzing the geometry of the stacked Golgi cisternae the first observation is that the stability of a single flattened cisterna with the coexisting flat and highly curved membrane cannot be explained in terms of the properties of free homogeneous membrane compartments (see above). Therefore, it was recently suggested that the stabilization may be provided by external forces due to intercisternal adhesion (Derganc et al. 2006). Although the nature of the adhesion is not completely understood, it can be linked to a number of adhesion molecules (Cluett and Brown 1992; Barr et al. 1997; Seemann et al. 2000). The adhesion strength needed for the stabilization of a typical Golgi stack was estimated to be of the order of 1.6 1016 J/mm2 (Derganc et al. 2006). Note that the microtubule cytoskeleton does not appear to be an essential scaffold for the Golgi structure because the stacked Golgi geometry is conserved even after its disruption with nocodazole (Cole et al. 1996). Another possible factor in stabilizing the flattened Golgi geometry is the lateral segregation of membrane constituents according to their preferred curvature. The segregation of the curvature-preferring membrane constituents into the cisternal rims would reduce the mechanical stress in the rims, lower the overall membrane bending energy, and thus stabilize the flattened cisternal shape. Also, such a segregation could play a role in the lateral sorting of membrane constituents. However, it was estimated that, in a typical case, the decrease in the bending energy due to the segregation of the curvaturepreferring membrane constituents into the cisternal rims is not large enough to drive an extensive segregation (Derganc 2007). This implies that the lateral segregation of membrane constituents in the Golgi involves other factors as well, e.g., phase separation or external protein scaffolds. An even more striking feature of the Golgi geometry is that it is preserved despite the fusion and fission of various saccular–vesicular membranes which possess a large excess of TAA. For example, fusion of the Golgi with the Golgian vesicles induces a rapid and large increase in the Golgi TAA, which destabilizes the Golgi geometry and cannot easily be compensated by flipase activity. Indeed, it has been observed that the arrival of the highly curved membrane from the intermediate compartment to isolated Golgi stacks increases the Golgi TAA and that this increase in TAA is compensated by formation of an additional cisterna and tubular connections between adjacent cisternae (Trucco et al. 2004). Recognition of the impact of membrane TAA on the Golgi geometry has led to the emergence of a novel function for the COPI vesicles. It has been proposed that the COPI vesicles act as reservoirs for membrane curvature and thus control the Golgi morphology (Beznoussenko et al. submitted). The role of the COPI vesicles as reservoirs of membrane curvature was validated by a controlled and independent manipulation of the SNARE and ARF/COPI machineries, which mediate the fusion and formation of vesicles, respectively. It was found that (i) inhibition of the SNARE machinery alone will reduce the TAA of the Golgi cisternae and induce a narrowing of the cisternal perforations,

The geometry of organelles

*

327

followed by invagination of cisternal membranes; (ii) inhibition of the ARF/ COPI machinery alone will increase the TAA of the Golgi cisternae and induce a widening of the cisternal perforations followed by Golgi tubulation; and (iii) inhibition of both machineries will not induce any changes in Golgi shape. To conclude, the main aspects of the geometry of cellular membrane compartments can be summarized. First, one has to bear in mind that many of the membrane geometrical features observed in vivo can be described in terms of the properties of laterally homogeneous membranes in the absence of external forces. Here, the parameters that define the shape of a membrane compartment are its area to volume ratio, the membrane spontaneous curvature C0 and the preferred membrane area difference DA0. For example, increasing C0 or DA0 generally leads to membrane evaginations while a decrease of C0 or DA0 leads to membrane invaginations. On the next level, membrane shape can be influenced by the lateral organization of membrane constituents, leading to laterally non-homogeneous membranes with variable thickness, elastic moduli or spontaneous curvature. The lateral organization may be provided either by purely intrinsic factors, e.g., intermolecular interactions, or by external factors such as protein scaffolds. Finally, the membrane shape may be affected by a variety of external forces.

List of abbreviations ADE CGN DAG DOPC ER PA PE PI PIP2 PM PS SM TAA TGN TMD VSV-G

area difference elasticity cis-Golgi network diacylglycerol dioleoyl-phosphatidylcholine endoplasmic reticulum phosphatidic acid phosphatidylethanolamine phosphatidylinositol phosphatidylinositol bisphosphate plasma membrane phosphatidylserine sphingomyelin transbilayer area asymmetry trans-Golgi network transmembrane domain vesicular stomatitis virus glycoprotein

References Barr FA, Puype M, Vandekerckhove J, Warren G (1997) GRASP65, a protein involved in the stacking of Golgi cisternae. Cell 91: 253–262 Beck KA, Buchanan JA, Malhotra V, Nelson WJ (1994) Golgi spectrin: identification of an erythroid beta-spectrin homolog associated with the Golgi complex. J Cell Biol 127: 707–723

328

*

J. Derganc et al.

Blood PD, Voth GA (2006) Direct observation of Bin/amphiphysin/Rvs (BAR) domaininduced membrane curvature by means of molecular dynamics simulations. Proc Natl Acad Sci USA 103: 15068–15072  Bo zic B, Svetina S, Zekš B, Waugh RE (1992) Role of lamellar membrane-structure in tether formation from bilayer vesicles. Biophys J 61: 963–973  Bo zic B, Svetina S, Zekš B (1997) Theoretical analysis of the formation of membrane microtubes on axially strained vesicles. Phys Rev E 55: 5834–5842 Bo zic B, Kralj-Iglic V, Svetina S (2006) Coupling between vesicle shape and lateral distribution of mobile membrane inclusions. Phys Rev E 73: 041915 Bretscher MS, Munro S (1993). Cholesterol and the Golgi apparatus. Science 261: 1280–1281 Buton X, Herve P, Kubelt J, Tannert A, Burger KN, Fellmann P, Muller P, Herrmann A, Seigneuret M, Devaux PF (2002) Transbilayer movement of monohexosylsphingolipids in endoplasmic reticulum and Golgi membranes. Biochemistry 41: 13106–13115 Cluett EB, Brown WJ (1992) Adhesion of Golgi cisternae by proteinaceous interactions – intercisternal bridges as putative adhesive structures. J Cell Sci 103: 773–784 Cluett EB, Machamer CE (1996) The envelope of vaccinia virus reveals an unusual phospholipid in Golgi complex membranes. J Cell Sci 109: 2121–2131 Cole NB, Sciaky N, Marotta A, Song J, Lippincott-Schwartz J (1996) Golgi dispersal during microtubule disruption: regeneration of Golgi stacks at peripheral endoplasmic reticulum exit sites. Mol Biol Cell 7: 631–650 Cole NB, Ellenberg J, Song J, DiEuliis D, Lippincott-Schwartz J (1998) Retrograde transport of Golgi-localized proteins to the ER. J Cell Biol 140: 1–15 Cullis PR, Fenske DB, Hope MJ (1996) Physical properties and functional roles of lipids in membranes. In: Vance DE, Vance JR (eds) Biochemistry of lipids, lipoproteins, and membranes. Elsevier, Amsterdam, pp. 1–33 Daleke DL (2007) Phospholipid flippases. J Biol Chem 282: 821–825 De Figueiredo P, Drecktrah D, Katzenellenbogen JA, Strang M, Brown WJ (1998) Evidence that phospholipase A2 activity is required for Golgi complex and trans Golgi network membrane tubulation. Proc Natl Acad Sci USA 95: 8642–8647 De Matteis M, Godi A, Corda D (2002) Phosphoinositides and the Golgi complex. Curr Opin Cell Biol 14: 434–447 Derganc J (2007) Curvature-driven lateral segregation of membrane constituents in Golgi cisternae. Phys Biol 4: 317–324 Derganc J, Mironov AA, Svetina S (2006) Physical factors that affect the number and size of Golgi cisternae. Traffic 7: 85–96 Dietrich C, Bagatolli LA, Volovyk ZN, Thompson NL, Levi M, Jacobson K, Gratton E (2001) Lipid rafts reconstituted in model membranes. Biophys J 80: 1417–1428  zaro-Die guez F, Vilella M (2006) Actin dynamics at the Golgi complex in Egea G, La mammalian cells. Curr Opin Cell Biol 18: 168–178 Engelman DM (2005) Membranes are more mosaic than fluid. Nature 438: 578–580 Fang M, Rivas MP, Bankaitis VA (1998) The contribution of lipids and lipid metabolism to cellular functions of the Golgi complex. Biochim Biophys Acta 1404: 85–100 Hamilton JA (2003) Fast flip-flop of cholesterol and fatty acids in membranes: implications for membrane transport proteins. Curr Opin Lipidol 14: 263–271 Hanada K, Pagano RE (1995) A Chinese hamster ovary cell mutant defective in the nonendocytic uptake of fluorescent analogs of phosphatidylserine: isolation using a cytosol acidification protocol. J Cell Biol 128: 793–798  Heinrich V, Bo zic B, Svetina S, Zekš B (1999) Vesicle deformation by an axial load: from elongated shapes to tethered vesicles. Biophys J 76: 2056–2071 Helfrich W (1973) Elastic properties of lipid bilayers – theory and possible experiments. Z Naturforsch (C) 28: 693–703

The geometry of organelles

*

329

Hwang WC, Waugh RE (1997) Energy of dissociation of lipid bilayer from the membrane skeleton of red blood cells. Biophys J 72: 2669–2678 Israelachvili JN, Marcelja S, Horn RG (1980) Physical principles of membrane organization. Q Rev Biophys 13: 121–200 Kearns BG, McGee TP, Mayinger P, Gedvilaite A, Phillips SE, Kagiwada S, Bankaitis VA (1997) Essential role for diacylglycerol in protein transport from the yeast Golgi complex. Nature 387: 21–22 Killian JA (1998) Hydrophobic mismatch between proteins and lipids in membranes. Biochim Biophys Acta 1376: 401–415 Koster G, VanDuijn M, Hofs B, Dogterom M (2003) Membrane tube formation from giant vesicles by dynamic association of motor proteins. Proc Natl Acad Sci USA 100: 15583–15588 Ladinsky MS, Mastronarde DN, McIntosh JR, Howell KE, Staehelin LA (1999) Golgi structure in three dimensions: functional insights from the normal rat kidney cell. J Cell Biol 144: 1135–1149 Lipowsky R (1992) Budding of membranes induced by intramembrane domains. J Phys II (France) 2: 1825–1840  Majhenc J, Bo zic B, Svetina S, Zekš B (2004) Phospholipid membrane bending as assessed by the shape sequence of giant oblate phospholipid vesicles. Biochim Biophys Acta 1664: 257–266 Malhotra V, Yaffe MP (2005) Membranes and organelles: regulating the size, shape, and plasticity of cellular compartments. Curr Opin Cell Biol 17: 343–344 Markin V (1981) Lateral organization of membranes and cell shapes. Biophys J 36: 1–19 € hlig A, Kubelt J, Herrmann € tter HG, Wu € stner D, Mu € ller P, Ho Marx U, Lassmann G, Holzhu A (2000) Rapid flip-flop of phospholipids in endoplasmic reticulum membranes studied by a stopped-flow approach. Biophys J 78: 2628–2640 McMahon HT, Gallop JL (2005) Membrane curvature and mechanisms of dynamic cell membrane remodelling. Nature 438: 590–596 € bereiner HG (1994) Budding transitions of fluid-bilayer Miao L, Seifert U, Wortis M, Do vesicles: the effect of area-difference elasticity. Phys Rev E 49: 5389–5407 Mitra K, Ubarretxena-Belandia I, Taguchi T, Warren G, Engelman DM (2004) Modulation of the bilayer thickness of exocytic pathway membranes by membrane proteins rather than cholesterol. Proc Natl Acad Sci USA 101: 4083–4088 Pomorski T, Menon AK (2006) Lipid flippases and their biological functions. Cell Mol Life Sci 63: 2908–2921 Rawicz W, Olbrich KC, McIntosh T, Needham D, Evans E (2000) Effect of chain length and unsaturation on elasticity of lipid bilayers. Biophys J 79: 328–339 Roux A, Cuvelier D, Nassoy P, Prost J, Bassereau P, Goud B (2005) Role of curvature and phase transition in lipid sorting and fission of membrane tubules. EMBO J 24: 1537–1545 Seemann J, Jokitalo E, Pypaert M, Warren G (2000) Matrix proteins can generate the higher order architecture of the Golgi apparatus. Nature 407: 1022-1026 Seifert U, Berndl K, Lipowsky R (1991) Shape transformations of vesicles: phase diagram for spontaneous- curvature and bilayer-coupling models. Phys Rev A 44: 1182–1202 Seifert U (1993) Curvature-induced lateral phase segregation in two-component vesicles. Phys Rev Lett 70: 1335–1338 Siegel DP, Kozlov MM (2004) The Gaussian curvature elastic modulus of N-monomethylated dioleoylphosphatidylethanolamine: relevance to membrane fusion and lipid phase behavior. Biophys J 87: 366–374 Simons K, Ikonen E (1997) Functional rafts in cell membranes. Nature 387: 569–572 Singer SJ, Nicolson GL (1972) The fluid mosaic model of the structure of cell membranes. Science 175: 720–731

330

*

J. Derganc et al.

 Svetina S, Zekš B (1989) Membrane bending energy and shape determination of phospholipid vesicles and red blood cells. Eur Biophys J 17: 101–111  Svetina S, Zekš B (2002) Shape behavior of lipid vesicles as the basis of some cellular processes. Anat Rec 268: 215–225  Svetina S, Zekš B, Waugh RE, Raphael RM (1998) Theoretical analysis of the effect of the transbilayer movement of phospholipid molecules on the dynamic behavior of a microtube pulled out of an aspirated vesicle. Eur Biophys J 27: 197–209 Trucco A, Polishchuk RS, Martella O, Di Pentima A, Fusella A, Di Giandomenico D, San Pietro E, Beznoussenko GV, Polishchuk EV, Baldassarre M, Buccione R, Geerts WJ, Koster AJ, Burger KNJ, Mironov AA (2004) Secretory traffic triggers the formation of tubular continuities across Golgi sub-compartments. Nat Cell Biol 6: 1071–1081 Upadhyaya A, Sheetz MP (2004) Tension in tubulovesicular networks of Golgi and endoplasmic reticulum membranes. Biophys J 86: 2923–2928 Van Helvoort A, Van Meer G (1995) Intracellular lipid heterogeneity caused by topology of synthesis and specificity in transport. Example: sphingolipids. FEBS Lett 369: 18–21 Voeltz GK, Prinz WA (2007) Sheets, ribbons and tubules—how organelles get their shape. Nat Rev Mol Cell Biol 8: 258–264  Waugh RE, Song J, Svetina S, Zekš B (1992) Local and nonlocal curvature elasticity in bilayer membranes by tether formation from lecithin vesicles. Biophys J 61: 974–982 Weigert R, Silletta MG, Spanò S, Turacchio G, Cericola C, Colanzi A, Senatore S, Mancini R, Polishchuk EV, Salmona M, Facchiano F, Burger KNJ, Mironov A, Luini A, Corda D (1999) CtBP/BARS induces fission of Golgi membranes by acylating lysophosphatidic acid. Nature 402: 429–433 Wirtz KW (2006) Phospholipid transfer proteins in perspective. FEBS Lett 580: 5436–5441 Ziherl P, Svetina S (2005) Nonaxisymmetric phospholipid vesicles: rackets, boomerangs, and starfish. Europhys Lett 70: 690–696 Zimmerberg J, Kozlov MM (2006) How proteins produce cellular membrane curvature. Nat Rev Mol Cell Biol 7: 9–19

Main transport steps

ER-to-Golgi transport

*

333

ER-to-Golgi transport Fatima Verissimo and Rainer Pepperkok

Biosynthetic cargo enters the secretory pathway at the endoplasmic reticulum (ER) where it is properly folded before it becomes sorted and concentrated into transport carriers that move towards the Golgi complex. The basic protein machinery involved in bidirectional transport between the ER and the Golgi complex has long been identified and is fairly well characterized (Lee et al. 2004). The vesicular coat complex COPII mediates ER exit, while retrograde transport from the Golgi complex to the ER involves the function of the vesicular coat complex COPI. In the following paragraphs we focus on our current knowledge on the spatial and temporal organization of ER-to-Golgi transport at the molecular level.

The spatial organization of ER-exit The first step in ER-to-Golgi transport is the exit of secretory cargo from the ER, which is mediated by the vesicular coat complex COPII and its regulators (Lee et al. 2004). COPII components are involved in the selection and concentration of secretory cargo at the ER and the membrane deformation that leads to the synthesis of 60–80 nm-sized COPII-coated transport vesicles (Lee et al. 2004). In Saccharomyces cerevisiae, COPII vesicle budding occurs across the entire ER (Rossanese et al. 1999), while in Pichia pastoris and metazoan ER exit is highly spatially organized occurring only at distinct domains of ribosome free ER known as transitional ER (tER) or ER-exit sites (ERES) (Palade 1975; Orci 1991; Rossanese et al. 1999). Intriguingly, the number and spatial organization of ERES in the different organisms appears to correlate with the organization of their Golgi apparatus. S. cerevisiae does not have ERES and the Golgi exists as a series of scattered individual net-like cisternae (Preuss et al. 1992, see also Fig. 1). The protozoan Trypanosoma brucei has only one ERES and one stacked Golgi complex adjacent to it (He 2007). P. pastoris has only few ERES all with a single stacked Golgi complex nearby (Bevis et al. 2002; Rossanese et al. 1999, see Fig. 1). In mammalian cells, up to hundreds of ERES are distributed throughout the cytoplasm with a concentration in the juxtanuclear Golgi region (Fig. 2). They are typically long-lived and although highly mobile, they do not translocate long distances (Stephens et al. 2000, Stephens 2003). Two different models exist for the organization of the ER–Golgi interface in plant cells. In the “secretory unit” model, of the numerous Golgi stacks existing in plant cells each pairs with one ERES and both move as a “unit” together along the ER (daSilva et al. 2004, see also Fig. 1). In contrast to this

334

*

F. Verissimo and R. Pepperkok

Figure 1. Existing ER-to-Golgi transport models in different species. (A) In S. cerevisiae COPIIcoated vesicles bud across the entire ER and move cargo to the Golgi complex and COPI vesicles return material to the ER (Lee et al. 2004). (B) In P. pastoris, T. brucei and plants cells the ER-toGolgi interface is organized as “secretory units”, where ERES are closely linked with a stacked Golgi (He 2007; Bevis et al. 2002; Rossanese et al. 1999; Robinson et al. 2007). In plants the “secretory units” are mobile (Robinson et al. 2007). (C) In mammalian cells, COPII-vesicles bud from the ERES and deliver cargo to the ER-to-Golgi intermediate compartment (ERGIC) or from vesicular tubular clusters (VTCs) de novo by homotypic fusion. COPI-coated VTCs deliver cargo to the Golgi and COPI dependent retrograde traffic occurs from the Golgi and ERGIC/VTCs (Stephens and Pepperkok 2001; Appenzeller-Herzog and Hauri 2006).

model it has also been proposed that ERES in some plant cells do not undergo long-range movement but transiently interact with highly mobile Golgis (Robinson et al. 2007). In many of the studies mentioned above ERES have been defined as the sites on the ER where the COPII complex becomes enriched and COPII budding profiles or vesicles accumulate. While this is clearly revealing their identity, the machinery involved in their biogenesis, maintenance and organization has so far remained largely elusive. An attractive hypothesis is that ERES are relatively stable structures and might be structurally organized by several proteins including the COPII complex. In this context it has been demonstrated that secretory cargo has a role in stabilizing COPII on ER membranes (Forster et al. 2006). Several studies in yeast (Connerly et al. 2005), and mammalian cells (Bhattacharyya and Glick 2007; Iinuma et al. 2007; Watson et al. 2006) have suggested that Sec16 plays a major role in the organization and maintenance of ERES. The Sec23p-interacting protein p125 has also been proposed to play a role in the organization of ERES (Shimoi et al. 2005). During mitosis COPII proteins dissociate from ER membranes and ERES seem to disappear (Hammond and Glick 2000; Stephens 2003, see also Fig. 2 and Quick time video 1). Coincident with this, ER-exit ceases during mitosis (Prescott et al. 2001) and it is thus tempting to speculate, that at least during mitosis disassembly of ERES is directly coupled to the block in ER-to-Golgi transport. It could also be shown that phosphorylation is necessary for the disassembly of ERES during mitosis (Kano et al. 2004). However, the mechanisms that regulate this process are not known.

ER-to-Golgi transport

*

335

Figure 2. ERES disassemble during mitosis. (A) and (D) life cell imaging of cells expressing the YFP-tagged COPII marker Sec31p showed that ERES are distributed throughout the cytoplasm with a concentration in the region of the juxta–nuclear Golgi region. (B) During mitosis COPII proteins dissociate from ER membranes and ERES seem to disappear. (C) After cell division ERES reappear.

ER-to-Golgi transport carriers In S. cerevisiae, secretion of cargo from the ER to the Golgi involves the formation of 60–80 nm COPII-coated transport vesicles that fuse with the ciscistaernae of the Golgi to deliver their cargo (Lee et al. 1994, see also Fig. 1). In these cells several highly mobile cis-cistaernae exist, which have the potential to come in close proximity of the ER to take up the content of the COPII vesicles. Therefore, cytoskeleton and associated motor proteins, which typically facilitate long-range transport, do not appear to play a role in this specie for ER-to-Golgi transport. In mammalian cells the situation is more complex, not least because the places of vesicle synthesis, the ERES, are numerous, but also several tens of microns away from the centrally positioned Golgi complex, raising the question how these carriers ever reach the Golgi complex. Advanced electron microscopy studies including tomography analyses have clearly demonstrated the existence of free COPII-coated vesicles at ERES in mammalian cells (Mironov et al. 2003; Zeuschner et al. 2006). Paradoxically, in one study most of these COPII vesicles seemed to lack secretory cargo

336

*

F. Verissimo and R. Pepperkok

(Mironov et al. 2003). Live cell imaging of ER-to-Golgi transport of different GFP-tagged secretory cargo combined with electron microscopy analyses revealed that large ER-to-Golgi transport carriers of different shape and size exist in mammalian cells (see, e.g., Stephens and Pepperkok 2001; Mironov et al. 2003; and Quick time video 2). This raises the questions, what are the contributions of these different transport carriers in ER-to-Golgi transport and why do they exist in mammalian cells, but not in yeast and how are they formed? Early time-lapse studies in living cells have demonstrated that the longrange ER-to-Golgi transport carriers are not the small COPII-coated vesicles but rather 300–500 nm-sized structures (VTCs for vesicular tubular clusters) that form in close proximity to ERES (see review by Stephens and Pepperkok 2001, see also Fig. 1). VTCs contain cargo amounts equivalent to that of several COPII-coated vesicles and are transported along microtubules powered by the microtubule motor protein dynein (Presley et al. 1997). While VTCs are transported to the Golgi complex, they are devoid of COPII but coated with COPI (Shima et al. 1999; Stephens et al. 2000). Microinjection of inactivating anti-COPI antibodies did not interfere with VTC formation but blocked their subsequent transport to the Golgi complex along microtubules (Shima et al. 1999). This has lead to the hypothesis that formation of recycling COPI vesicles at the VTC level is necessary for their long-range ER-to-Golgi transport demonstrating a tight molecular linkage between anterograde and retrograde transport (see Stephens and Pepperkok 2001). A number of distinct possibilities how the VTCs may be synthesized in mammalian cells have been proposed. First, small COPII vesicles, which rapidly uncoat (Barlowe et al. 1994), may undergo homotypic fusion to form a VTC de novo. In vitro experiments have shown that homotypic fusion of COPII vesicles takes place (Xu and Hay 2004) and experiments in permeabilized cells indicate that the same process might also occur in intact cells (Bentley et al. 2006). Alternatively, VTCs could be formed by the fusion of COPII vesicles with a pre-existing transport carrier or membrane structure at the ER–Golgi boundary. One such structure could be the ERGIC, which has been proposed to be an immobile organelle that functions as a major site of anterograde and retrograde sorting at the ER–Golgi interface (reviewed in Appenzeller-Herzog and Hauri 2006, see also Fig. 1). Ben-Tekaya et al. have provided evidence by live cell imaging of the well established ERGIC marker p53, that ERGIC might indeed be a static structure from which VTC-like transport carriers are released to be transported to the Golgi complex (Ben-Tekaya et al. 2005). ERGIC partly colocalises with COPI and COPII. However, it is not yet clear which vesicular coat complex is responsible for the formation of the carriers from the ERGIC to the Golgi apparatus. To elucidate these differences additional work using multi-labeling imaging in living cells and different cargo molecules will be necessary. Live cell microscopy of GFP-labeled secretory cargo has also revealed that besides vesicles and VTCs, tubular transport intermediates (TTis) might play a

ER-to-Golgi transport

*

337

role in ER-to-Golgi transport (Simpson et al. 2006). The precise contribution of the TTis to ER-to-Golgi transport and the molecular mechanisms of their synthesis are not clear. It has been proposed that there is a correlation between the amount of cargo to be transported and the number and size of TTis appearing in a cell (Simpson et al. 2006). This may suggest that the formation of TTis is a mechanism to compensate for transiently elevated cargo levels at donor membranes. Interestingly, correlative light and EM studies have proposed that COPII negative TTis are formed next to COPII labeled ERES (Mironov et al. 2003). In contrast to this, expression of GTPbound Sar1p has been shown to induce the formation of cargo-containing tubules (Lee et al. 2005) suggesting that COPII may be at least indirectly involved in TTi formation. Independently of how they are formed, EM tomography studies of thick (400 nm) cryo-sections clearly demonstrated the co-existence of vesicular and tubular carriers directly adjacent to ERES (Zeuschner et al. 2006). The same experimental approach has also revealed the existence of tubular connections between the ER and Golgi complex (Trucco et al. 2004; Moreau et al. 2007). Presently it is not clear to which extent each of these morphologically distinct transport carriers operating between the ER and Golgi complex contribute to cargo transport or why they exist. One explanation may be the existence of specialized types of cargo that need to be secreted. Extremely large cargo molecules such as procollagen, which forms upon its folding rod shaped structures greater than 300 nm in length (see Lamande and Bateman 1999), are physically unable to fit into 60–80 nm COPII vesicles. Therefore, the need to transport these cargo molecules necessitates the synthesis of transport carriers that fit their size. Electron microscopy analyses indicated that procollagen leaves the ER at sites devoid of COPII (Mironov et al. 2003). A more recent study has however demonstrated, that ER-exit of procollagen is dependent on the small GTPase Sar1 (Starkuviene and Pepperkok 2007). One explanation for this discrepancy might be that the antibodies used in the former localization study failed to detect the COPII component under view on the respective carriers. It has therefore been speculated that size adapted COPII carriers may form with cargo-specific COPII homologues (Fromme and Schekman 2005). In line with this hypothesis recent work suggested that COPII vesicles are flexible in size and that the number of Sec13/31p hetero-tetramers involved determine the vesicle size (Stagg et al. 2006). Furthermore, Zeuschner et al. (2006) showed that large vesicles and tubules coated with COPII exist close to ERES. The contribution of the different COPII isoforms that co-exist in cells to the production of large COPII-coated vesicles or TTis is presently not known.

Transport along microtubules and arrival at the Golgi complex Once VTCs or other ER-to-Golgi carriers are formed in the proximity of peripheral ERES they must be transported up to several tens of microns to the Golgi

338

*

F. Verissimo and R. Pepperkok

complex (see Quick Time video 2). Due to this size of eukaryotic cells and their “congested” cytosol, diffusion would be too slow to efficiently move the transport carriers from the ER to the Golgi complex. Instead, transport of the VTCs from the ER to the Golgi complex has been shown to occur by their movement along microtubules (Presley et al. 1997; Scales et al. 1997; Simpson et al. 2006). The motor protein powering these movements is the dynein/ dynactin complex (Presley et al. 1997). In higher plants Golgi and ER tubules movements are actin dependent and treatment with actin interfering drugs, such as cytochalasin or latrunculin B, leads to a cessation in ER and Golgi  et al. 2007). While these observations clearly demonstrate the motility (Boutte importance of the actin cytoskeleton in ER-to-Golgi transport in plant cells a direct involvement of the cytoskeleton in the movement of transport carriers  et al. 2007). has not been demonstrated (Boutte Interestingly, COPII and COPI, directly interact with regulatory components of the cytoskeleton-based motility machinery (Watson et al. 2005; Chen et al. 2005). It has been demonstrated that p150glued, a component of the dynactin complex, directly interacts with Sec23p, a component of the COPII coat (Watson et al. 2005). Therefore, p150glued might be a critical factor that coordinates the synthesis of a COPII transport carrier with its subsequent transport towards the Golgi complex along microtubules. The interaction of cdc42 with the COPI complex has been demonstrated to be critical for the recruitment of dynein to COPI vesicles (Chen et al. 2005) and thus their transport along microtubules. In any case, minus end directed translocation of the transport carriers along microtubules in mammalian cells guarantees that they will finally arrive at the spatially restricted Golgi area, which is in close proximity of the microtubule organizing centre. Once the ER-derived transport carriers have arrived at the Golgi complex the question arises how do they deliver their cargo. In vitro experiments have suggested that ER-derived COPII-coated vesicles contain all the necessary components for homotypic fusion and VTC formation (Xu and Hay 2004). Mechanisms that prevent fusion of COPIIcoated vesicles with the ER also exist (Kamena and Spang 2004). Therefore, one possibility could be that several VTCs arriving at the Golgi area undergo homotypic fusion. Coupled to subsequent recycling of the transport machinery by COPI coated carriers this could be a mechanism to form a cis-cistaernae de novo. Alternatively, heterotypic fusion between VTCs and a pre-Golgi structure such as ERGIC or the cis-Golgi network might occur. Although both possibilities are conceptually different mechanisms that would result in different ways of Golgi biogenesis or intra-Golgi transport, in both cases SNARES and tethers, key components of the membrane fusion machinery, play a central role in cargo delivery (Cai et al. 2007; Jahn and Scheller 2006). Remarkably, components necessary for Golgi organization are also components of the transport machinery (Lowe and Barr 2007) and vice versa. Docking of the VTCs with the cis-Golgi is dependent on p115 (Moyer et al. 2001) and a mechanism that couples membrane tethering and fusion at the

ER-to-Golgi transport

*

339

ERGIC and cis-Golgi, was suggested to be mediated through the interaction of the tether GM130 with the SNARE syntaxin 5 (Diao et al. 2008). Interestingly, p115 participates in reassembly of post-mitotic Golgi fragments (DiracSvejstrup et al. 2000) and GM130 has been described to have a role in membrane fusion during Golgi assembly allowing lateral cisternal fusion and enzyme equilibration (Puthenveedu et al. 2006). From the above it is clear that some components of COPII vesicles play a role in Golgi organization. However their precise role in Golgi biogenesis and how the linkage of transport and Golgi biogenesis is temporally and spatially regulated still needs further investigation.

Outlook ER-to-Golgi transport remains an open challenge with a number of questions waiting for an answer. Most importantly, it needs to be determined to which extent the distinct ER-to-Golgi transport carriers contribute to cargo transport. Are distinct carriers modulated or induced by the type and amount of cargo to be transported? Are there distinct protein machineries underlying the transport of distinct specialized cargo molecules such as procollagen? Current efforts attempt to identify the complete machinery of the secretory pathway (Gilrichst et al. 2006; Bard et al. 2006). Such large scale approaches together with live cell experiments that allow fine modulation of each of the early secretory pathway steps will help to address the questions above and further disclose the spatial and temporal organization of ER-to-Golgi transport. Acknowledgements. We would like to apologise to our colleagues whose work we were ~o para a unable to cite due to space limitations. F.V. has been supported by the Funda¸ca ^ ncia e Tecnologia fellowship SFRH/BPD/22014/2005. Cie

References Appenzeller-Herzog C, Hauri HP (2006) The ER–Golgi intermediate compartment (ERGIC): in search of its identity and function. J Cell Sci 119: 2173–2183 Bard F, Casano L, Mallabiabarrena A, Wallace E, Saito K, Kitayama H, Guizzunti G, Hu Y, et al. (2006) Functional genomics reveals genes involved in protein secretion and Golgi organization. Nature 439: 604–607 Barlowe C, Orci L, Yeung T, Hosobuchi M, Hamamoto S, Salama N, Rexach MF, Ravazzola M, Amherdt M, Schekman R (1994) COPII: a membrane coat formed by Sec proteins that drive vesicle budding from the endoplasmic reticulum. Cell 77: 895–907 Ben-Tekaya H, Miura K, Pepperkok R, Hauri HP (2005) Live imaging of bidirectional traffic from the ERGIC. J Cell Sci 118: 357–367 Bentley M, Liang Y, Mullen K, Xu D, Sztul E, Hay JC (2006) SNARE status regulates tether recruitment and function in homotypic COPII vesicle fusion. J Biol Chem 281: 38825–38833 Bevis BJ, Hammond AT, Reinke CA, Glick BS (2002) De novo formation of transitional ER sites and Golgi structures in Pichia pastoris. Nat Cell Biol 4: 750–756

340

*

F. Verissimo and R. Pepperkok

Bhattacharyya D, Glick BS (2007) Two mammalian Sec16 homologues have nonredundant functions in endoplasmic reticulum (ER) export and transitional ER organization. Mol Biol Cell 18: 839–849  Y, Vernhettes S, Satiat-Jeunemaitre B (2007) Involvement of the cytoskeleton in Boutte the secretory pathway and plasma membrane organisation of higher plant cells. Cell Biol Int 31: 649–654 Cai H, Yu S, Menon S, Cai Y, Lazarova D, Fu C, Reinisch K, Hay JC, Ferro-Novick S (2007) TRAPPI tethers COPII vesicles by binding the coat subunit Sec23. Nature 445: 941–994 Chen JL, Fucini RV, Lacomis L, Erdjument-Bromage H, Tempst P, Stamnes M (2005) Coatomer-bound Cdc42 regulates dynein recruitment to COPI vesicles. J Cell Biol 169: 383–389 Connerly PL, Esaki M, Montegna EA, Strongin DE, Levi S, Soderholm J, Glick BS (2005) Sec16 is a determinant of transitional ER organization. Curr Biol 15: 1439–1447 DaSilva LL, Snapp EL, Denecke J, Lippincott-Schwartz J, Hawes C, Brandizzi F (2004) Endoplasmic reticulum export sites and Golgi bodies behave as single mobilesecretory units in plant cells. Plant Cell 16: 1753–1771 Diao A, Frost L, Morohashi Y, Lowe M (2008) Coordination of golgin tethering and SNARE assembly: GM130 binds syntaxin 5 in a p115-regulated manner. J Biol Chem 283: 6957–6967 Dirac-Svejstrup AB, Shorter J, Waters MG, Warren G (2000) Phosphorylation of the vesicle-tethering protein p115 by a casein kinase II-like enzyme is required for Golgi reassembly from isolated mitotic fragments. J Cell Biol 150: 475–488 Forster R, Weiss M, Zimmermann T, Reynaud EG, Verissimo F, Stephens DJ, Pepperkok R (2006) Secretory cargo regulates the turnover of COPII subunits at single ER exit sites. Curr Biol 16: 173–179 Fromme JC, Schekman R (2005) COPII-coated vesicles: flexible enough for large cargo? Curr Opin Cell Biol 17: 345–352 Gilchrist A, Au CE, Hiding J, Bell AW, Fernandez-Rodriguez J, Lesimple S, Nagaya H, Roy L, Gosline SJ, Hallett M, et al (2006) Quantitative proteomics analysis of the secretory pathway. Cell 127: 1265–1281 Hammond AT, Glick BS (2000) Dynamics of transitional endoplasmic reticulum sites in vertebrate cells. Mol Biol Cell 11: 3013–3030 He CY (2007) Golgi biogenesis in simple eukaryotes. Cell Microbiol 9: 566–572 Iinuma T, Shiga A, Nakamoto K, O’Brien MB, Aridor M, Arimitsu N, Tagaya M, Tani K (2007) Mammalian Sec16/p250 plays a role in membrane traffic from the endoplasmic reticulum. J Biol Chem 282: 17632–17639 Jahn R, Scheller RH (2006) SNAREs engines for membrane fusion. Nat Rev Mol Cell Biol 7: 631–643 Kamena F, Spang A (2004) Tip20p prohibits back-fusion of COPII vesicles with the endoplasmic reticulum. Science 304: 286–289 Kano F, Tanaka AR, Yamauchi S, Kondo H, Murata M (2004) Cdc2 kinase-dependent disassembly of endoplasmic reticulum (ER) exit sites inhibits ER-to-Golgi vesicular transport during mitosis. Mol Biol Cell 15: 4289–4298  SR, Bateman JF (1999) Procollagen folding and assembly: the role of endoLamande plasmic reticulum enzymes and molecular chaperones. Semin Cell Dev Biol 10: 455–464 Lee MC, Miller EA, Goldberg J, Orci L, Schekman R (2004) Bi-directional protein transport between the ER and Golgi. Annu Rev Cell Dev Biol 20: 87–123 Lee MC, Orci L, Hamamoto S, Futai E, Ravazzola M, Schekman R (2005) Sar1p N-terminal helix initiates membrane curvature and completes the fission of a COPII vesicle. Cell 122: 605–617 Lowe M, Barr FA (2007) Inheritance and biogenesis of organelles in the secretory pathway. Nat Rev Mol Cell Biol 8: 429–439

ER-to-Golgi transport

*

341

Mironov AA, Mironov Jr AA, Beznoussenko GV, Trucco A, Lupetti P, Smith JD, et al (2003) ER-to-Golgi carriers arise through direct en bloc protrusion and multistage maturation of specialized ER exit domains. Dev Cell 5: 583–594 Moyer BD, Allan BB, Balch WE (2001) Rab1 interaction with a GM130 effector complex regulates COPII vesicle cis-Golgi tethering. Traffic 2: 268–276 Orci L, Ravazzola M, Meda P, Holcomb C, Moore HP, Hicke L, Schekman R (1991) Mammalian Sec23p homologue is restricted to the endoplasmic reticulum transitional cytoplasm. Proc Natl Acad Sci USA 88: 8611–8615 Palade G (1975) Intracellular aspects of the process of protein synthesis. Science 189: 347–358 Prescott AR, Farmaki T, Thomson C, James J, Paccaud JP, Tang BL, Hong W, Quinn M, Ponnambalam S, Lucocq J (2001) Evidence for prebudding arrest of ER export in animal cell mitosis and its role in generating Golgi partitioning intermediates. Traffic 2: 321–335 Preuss D, Mulholland J, Franzusoff A, Segev N, Botstein D (1992) Characterization of the Saccharomyces Golgi complex through the cell cycle by immunoelectron microscopy. Mol Biol Cell 3: 789–803 Presley JF, Cole NB, Schroer TA, Hirschberg K, Zaal KJ, Lippincott-Schwartz J (1997) ER-toGolgi transport visualized in living cells. Nature 389: 81–85 Puthenveedu MA, Bachert C, Puri S, Lanni F, Linstedt AD (2006) GM130 and GRASP65dependent lateral cisternal fusion allows uniform Golgi-enzyme distribution. Nat Cell Biol 8: 238–248 Robinson DG, Herranz MC, Bubeck J, Pepperkok R and Ritzenthaler Ch (2007) Membrane Dynamics in the Early Secretory Pathway. Crit Rev Plant Sci 26: 199–225 Rossanese OW, Soderholm J, Bevis BJ, Sears IB, O’Connor J, Williamson EK, Glick BS (1999) Golgi structure correlates with transitional endoplasmic reticulum organization in Pichia pastoris and Saccharomyces cerevisiae. J Cell Biol 145: 69–81 Scales SJ, Pepperkok R, Kreis TE (1997) Visualization of ER-to-Golgi transport in living cells reveals a sequential mode of action for COPII and COPI. Cell 90: 1137–1148 Shima DT, Scales SJ, Kreis TE, Pepperkok R (1999) Segregation of COPI-rich and anterograde-cargo-rich domains in endoplasmic-reticulum-to-Golgi transport complexes. Curr Biol 9: 821–824 Shimoi W, Ezawa I, Nakamoto K, Uesaki S, Gabreski G, Aridor M, Yamamoto A, Nagahama M, Tagaya M, Tani K (2005) p125 is localized in endoplasmic reticulum exit sites and involved in their organization. J Biol Chem 280: 10141–10148 Simpson JC, Nilsson T, Pepperkok R (2006) Biogenesis of tubular ER-to-Golgi transport intermediates. Mol Biol Cell 17: 723–737 € rkan C, Fowler DM, LaPointe P, Foss TR, Potter CS, Carragher B, Balch WE Stagg SM, Gu (2006) Structure of the Sec13/31 COPII coat cage. Nature 439: 234–238 Erratum in: Nature 442: 840 Starkuviene V, Pepperkok R (2007) Differential requirements for ts-O45-G and procollagen biosynthetic transport. Traffic 8: 1035–1051 Stephens DY, Lin-Marq N, Pagano A, Pepperkok R, Paccaud J-P (2000) COPIcoated ER-toGolgi transport complexes segregate from COPII in close proximity to ER exit sites. J Cell Sci 113: 2177–2185 Stephens DJ, Pepperkok R (2001) Illuminating the secretory pathway: when do we need vesicles? J Cell Sci 114: 1053–1059 Stephens DJ (2003) De novo formation, fusion and fission of mammalian COPII-coated endoplasmic reticulum exit sites. EMBO Rep 4: 210–217 Xu D, Hay JC (2004) Reconstitution of COPII vesicle fusion to generate a pre-Golgi intermediate compartment. J Cell Biol 167: 997–1003 Zeuschner D, Geerts WJC, Van Donselaar E, Humbel BM, Slot JW, Koster AJ, Klumperman J (2006) Immuno-electron tomography of ER exit sites reveals the existence of free COPII-coated transport carriers. Nat Cell Biol 8: 377–383

342

*

A. A. Mironov and G. V. Beznoussenko

Intra-Golgi transport Alexander A. Mironov and Galina V. Beznoussenko

In this chapter, we discuss mechanisms of intra-Golgi transport and models describing this process, analyze the pros and cons for each of these models in details.

The vesicular model (VM) To explain intra-Golgi transport, 10 different models have been proposed (Table 1). The most known is VM. At the level of the Golgi apparatus (GA), this model consists of four major steps. The first step is the formation of COPIcoated buds. During the second step, the buds that undergo fission were transformed into coated vesicles that then undergo uncoating. The third step is vesicle movement towards the distal Golgi cisterna and interacting with tethering molecules (Haas and Barr 2007). Finally, the vesicle docking and fusion with the distal cisterna occur. This step is mediated by SNARE proteins (Hong 2005; Fig. 1A). The strongest observation in favour of the vesicular model derives from the analysis of yeast temperature-sensitive mutants, where blockage of function of Sec18, Sec7 and Sec22 at the non-permissive temperature, induces accumulation of 50 nm vesicles. Analysis of double-mutant strains showed that vesicle accumulation can be avoided by the mutations Sec12, Sec13, Sec61 and Sec23 that stop ER-to-Golgi transport (Kaiser and Schekman 1990). VM gives three main predictions. 1. The size of vesicles and cargo should correspond to each other. 2. Cargoes undergo concentration in the COPIcoated buds and COPI-dependent vesicles. 3. There should be a system for the correct targeting of vesicles otherwise the vesicles will be dispersed within the cytosol. However, these predictions are not fulfilled. 1. The strongest objection against VM is intra-Golgi transport of particles that are larger in size than COPI-dependent vesicles (reviewed in Mironov et al. 1997). Indeed, in a number of diverse cell types, supramolecular aggregates are found in the GA/TGN: alga scales (Becker et al. 1995; Dupree and Sherrier 1998), apolipoprotein E-containing particles in liver cells (Claude 1970; Dahan et al. 1994), lipid droplets in enterocytes after oil feeding (Sabesin and Frase 1977), aggregates of procollagen in fibroblasts (Leblond 1989; Bonfanti et al. 1998), eccentric electron-dense spherical bodies in epithelial cells of the seminal vesicle (Clermont et al. 1992), asymmetric membrane thickenings of the apical plasmalemma of uroepithelial cells (Severs and Hicks 1979) and casein submicelles in lactating mammary gland cells (Clermont et al. 1993). These aggregates

Intra-Golgi transport

*

343

Table 1. Models of intra-Golgi transport 1. Models based on a single transport mechanism 1.1. Progression model. 1.2. Dissociation (vesicular) model. 1.3. Diffusion model. 1.3.1. Constant connections. 1.3.2. Bolus. 1.3.3. Kiss-and-run. 2. Combinatory models 2.1. Combination of progression and dissociation 2.1.1. Megavesicles for large aggregates and vesicle for small cargoes. 2.2. Combination of progression and dissociation recylcing 2.2.1. Cisterna maturation–progression þ retrograde dissociation carriers. 2.2.2. Cisterna maturation–progression þ retrograde diffusion of GE along connections (Snake-like model). 2.3. Combination of progression and kiss-and-run 2.3.1. Carrier maturation as the asymmetric kiss-and-run mechanism. 2.4. Combination of three mechanisms 2.4.1. Cisterna maturation–progression þ recycling of Golgi proteins by vesicles þ diffusion of cargo along continuity.

do not undergo dismantling during their transport through the GA (Bonfanti et al. 1998) and are too big to be carried by the 52-nm COPI vesicles. 2. There should be at least no significant depletion of cargoes from COPIdependent vesicles. A. Orci et al. (1997) claimed that they found enrichment of proinsulin in COPI vesicles. However, in fact, they observed a depletion of proinsulin not only in peri-Golgi round profiles, but also in COPI-coated buds. For instance, Table 3 there in shows the LD of proinsulin in COPIcoated buds at cis pole is 166  37, at trans pole 145  58. In cisternae of the cis half the density is 261  37, at trans half – 235  32. In order to fit VM, Orci et al. proposed the existence of two populations of vesicles and only using this assumption estimated that there is 1.4–1.8 enrichment of proinsulin in COPI-dependent vesicles. B. Griffiths et al. (1995) had found “much of the Golgi-associated G protein (from VSV). . . in COP/Rab1 positive buds.” However, in this study, cells were permeabilized with streptolysin O before fixation and then incubated in a permeabilization buffer in the presence of GTPgS. Only after this procedure was the colocalization of VSVG and COPI observed. GTPgS would be able to enter the cytoplasm and induce (in the presence of residual amount of cytosol still remaining in the cytoplasm) the artificial vesiculation of tubules containing VSVG (as it has been described in vitro, see Weidman et al. 1993). C. In GTP COPI-dependent light fraction isolated after incubation of isolated Golgi membranes with cytosol and ATP, the density of pIgR

344

*

A. A. Mironov and G. V. Beznoussenko

Figure 1. Scheme of the vesicular model (A) and the cisterna maturation model (B). (1a) Steadystate. (2A, 3A, 4A) Formation of COPI-coated buds. (5A, 6A) Fission of COPI-coated buds and their uncoating. (7A) Tethering of COPI vesicles to the next cisternae. (8A, 9A) Fusion of COPI vesicles and delivery of cargo (dots) to the next cisternae. (1B) Formation of ER-to-Golgi carriers (EGC, structures with sticks). (2B, 3B) Fusion of EGCs with the formation of the new cis-cisterna. (4B, 5B) Formation and COPI-coated buds and concentration of Golgi enzymes (dots) there. (6B) Uncoating. (7B, 8B) Tethering of COPI-dependent vesicles. (9B) Fusion of vesicles with the proximal cisternae and formation of Golgi-to-PM carriers from the last cisterna.

was found to be 1.8-fold higher than in original isolated Golgi membranes (Lanoix et al. 1999). This could be a result of the presence of cisternal fragments in the light fraction where this protein was found depleted in Golgi vesicles (Kweon et al. 2004). 3. There should be a system for the correct targeting of vesicles, otherwise the vesicles will be dispersed within the cytosol. At least, the ability of

Intra-Golgi transport

*

345

Table 2. List of cargo proteins that are depleted in COPI-dependent vesicles Cargo

Cell type

Reference

1. Apo-lipoproteins E and B 2. Lipoprotein granules Procollagen-I Procollagen-I Alga scales Lipid droplets Eccentric electron-dense spherical bodies Asymmetric membrane thickenings of the apical PM Casein submicelles

Hepatocytes Hepatocytes Chicken fibroblasts Human fibroblasts Algae Enterocytes Epithelial cells of the seminal vesicle Uroepithelial cells

Dahan et al. (1994) Claude (1970) Bonfanti et al. (1998) Mironov et al. (2001) Becker et al. (1995) Sabesin and Frase (1977) Clermont et al. (1992)

Lactating mammary gland cells Uukuniemi virus

Rambourg et al. (1993)

Herpes simplex virus

Di Lazzaro et al. (1995)

Exocrine pancreatic cells Exocrine pancreatic cells Hepatocytes Pancreatic endocrine cells Human fibroblast

Oprins et al. (2001) Oprins et al. (2001) Dahan et al. (1994) Orci et al. (1997) Mironov et al. (2001)

Isolated Golgi membranes; liver; 1.5-fold depletion NRK cells

Orci et al. (1993, 1989)

Virus particles with 100–120-nm diameter Virus particles with 200-nm diameters Amylase Chymotrypsinogen Albumin Proinsulin Soluble secretory peroxidase VSVG VSVG VSVG

VSVG VSVG Polymeric immunoglobulin receptor

Human fibroblasts; microinjection of aSNAP mutant CHO cells; isolation of vesicles from infected cells CHO cells; experiments in vitro Hepatocytes

Severs and Hicks (1979)

Jantii et al. (1997)

Martinez-Menarguez et al. (2001) Mironov et al. (2001); Kweon et al. (2004) Love et al. (1998) Pullikuth and Weidman (2002) Dahan et al. (1994)

vesicle filled with a reporter cargo to diffuse throughout the cytosol was the main assumption for the famous heterokaryon experiments demonstrating dissociative passage of the cargo from the donor to acceptor Golgi (Rothman et al. 1984). If the diffusion were unlimited and the targeting system were lacking, vesicles would fill the entire cytosol. In contrast, it has been shown that the ability of >50-nm particles (i.e. vesicle) to diffuse throughout the cytosol is very limited (Luby-Phelps 1994). Thus, there are two possibilities. If the diffusion of COPI vesicles were unlimited the explanation of the experiments with heterokaryon would be correct. If the diffusion were limited the experiments would be incorrect.

346

*

A. A. Mironov and G. V. Beznoussenko

Next, Orci et al. (1998) have found that COPI vesicles are on strings. In order to fit these observations, the authors of the heterokaryon experiment dramatically modified VM and instead proposed the percolatingvesicles model (Pelham and Rothman 2000; Orci et al. 2000b). According to this, the cargo-containing COPI vesicles may “percolate up and down” the stack, limited by tethering mainly to transfers between adjacent cisternae in the stack, and largely unable to distinguish up from down because the same SNAREs are present throughout. The authors explain that these vesicles would migrate in a stochastic process resembling a random walk, giving rise to the fast track of anterograde transport. Net flow in the cisto-trans direction would result because entry and exit are restricted to opposite faces, but would of course be less efficient than for unidirectional moving vesicles. However, within the framework of this scheme, the role of COPI vesicles for intra-Golgi transport in yeast could be questioned because different Golgi compartments are situated far away from each other. Some evidence against the VM has been obtained in cell free assays. For instance, intra-Golgi apparatus transfer has not been observed with Golgi apparatus fully immobilized on nitrocellulose (Hartel-Schenk et al. 1991). Intra-Golgi transport can occur without participation of COPI vesicles (Orci et al. 1991; Elazar et al. 1994; Taylor et al. 1994; Happe et al. 2000). Finally, the use of 50–60 nm vesicles as transport carriers is very unprofitable energetically (especially if other models can explain intra-Golgi transport, see below). Indeed, for the formation of one 52-nm vesicle a cell needs to use about 20 coatomer complexes and, thus, to hydrolyze at least 20 2.6 ¼ 50 molecules of GTP (Stamnes et al. 1998). Next, the existence of so-called on-parallel transport also represent definite difficulties for plausible explanations within the frames of the vesicular paradigm. In the case of vesicular transport the overall turnover of the Golgi membrane would be enormously high. Rothman et al. (1991) have calculated that in regulated secretory cells the speed of turnover of the plasma membrane (PM, counted on the basis of the assumption that COPI-vesicles function as transport carrier) is 100-fold higher than it has been determined experimentally.

The progression model The initial version of the progression model proposed that the membrane flow is unidirectional and coupled to the flow of secretory products (reviewed in Mironov et al. 1997, 2005) and that large domains compatible in size with the distal compartments (i.e. the Golgi cisternae) and filled with cargoes are formed on the proximal compartment and transported as carriers to the distal compartment along the secretory pathway. The obvious prediction from this dissociation model is that all Golgi cisternae are equal in their membrane composition and that PM should contain Golgi enzymes. However, Golgi enzymes were not found at least at

Intra-Golgi transport

*

347

the basolateral PM, although they were found at apical plasma membrane of some specialized cells such as enterocytes (Velasco et al. 1993). Moreover, in all animal cells that have been studied, different cisternae exhibit a distinct and stable enzymatic composition. There is a gradient of Golgi enzymes through Golgi stacks indicating that Golgi cisternae are not equivalent to each other. Strikingly, the gradient of enzymes corresponded to the order of processing of glycoproteins by Golgi glycosidases and transferases (Rabouille et al. 1995; see also Chapter 2.9). These contradictions forced researchers to modify the progression model and propose the cisterna maturation–progression model (CMPM, Bannykh and Balch 1997; Mironov et al. 1997). According to it, both soluble and membrane cargoes should be excluded from retrograde transport carriers being always within the lumen of cisternae, whereas resident proteins should be enriched there. The carriers exiting from the ER converge towards the Golgi, fuse with each other, and form a new cis-cisterna. The non-secretory material [e.g. the endoplasmic reticulum (ER)- and the intermediate compartment (IC)-resident proteins, SNAREs] is retrieved from the cis-Golgi into the ER, while at the same time the defining components of the medial compartment (e.g. medial Golgi enzymes) flow backward into the cis-elements. For instance, this new cis-cisterna generates COPI-coated vesicles carrying the ER and IC residents back to the ER/IC and fuses with COPI-coated vesicles derived from the old cis-Golgi cisterna and carrying cis-Golgi residents. The cis-compartment thus acquires medial Golgi features and, in effect, becomes a medial compartment. After several rounds of such events, the new cis-Golgi cisterna acquires cis-Golgi residents, whereas the old cis-Golgi cisterna undergoes transformation into a medial-Golgi cisterna. The process repeats itself (the medial compartment becoming trans) until the cargo reaches the trans-Golgi network (TGN). Finally, when resident proteins recycle back to the trans-Golgi cisternae, the secretory material in the TGN domain (secretory granules, or Golgi-to-PM carriers; GPC) moves towards PM, and fuses with it (Fig. 1B). Thus, there would be no stable compartments along the secretory pathway, and the cargo-containing proximal compartment gradually matures into the distal compartment. The second variant of CMPM implies that the recycling of the resident proteins occurs by the synchronous diffusion of Golgi resident proteins along the intercisternal connections that could be restricted for diffusion of some membrane cargoes such as VSVG (Trucco et al. 2004).

Evidence against the cisterna maturation model Although the cisterna maturation models can explain a significant fraction of the experimental data, it is not sufficient to explain all the experimental data. 1. According to CMPM, the speed of intra-Golgi transport of all cargoes and the speed of Golgi enzymes recycling should be equal. However, albumin

348

*

A. A. Mironov and G. V. Beznoussenko

moves through the Golgi faster than VSVG (Trucco et al. submitted) and the speed of Golgi enzyme recycling is 7–10-fold slower than membrane transport (Hoe et al. 1995). 2. Obviously, the loss of the function of COPI should force Golgi enzymes to move distally and intra-Golgi transport should be blocked. However, in yeast, inhibition or deletion of COPI function does not appear to affect the localization of Golgi enzymes (Gaynor et al. 1998). Also in mammals, in LDL F cells with COPI mutants the blockage of coatomer function does not lead to the redistribution of Golgi enzymes to the PM (Guo et al. 1994). Similarly, microinjection of anti-bCOP antibody does not induce the shift of Golgi enzymes to the PM (Pepperkok et al. 1993). Moreover, maturation of Golgi compartments exists also when the function of COPI is blocked in yeast (Matsuura-Tokita et al. 2006). 3. A significant concentration of Golgi enzymes in vesicles in vivo and in vitro is a prerequisite for the good performance of CMPM (Glick et al. 1997). However, there is no convincing evidence that COPI-dependent vesicles are concentrated but even not depleted of Golgi enzymes and sugar transporters. In a vast majority of the immuno EM papers based on cryosectioning, post-embedding or immunoperoxidase preembedding techniques the density of labelling upon cisternae is much higher than over nearby round profiles (Lucocq et al. 1995; Orci et al. 2000b; Cosson et al. 2002; Kweon et al. 2004; also see for example Fig. 7 in Velasco et al. 1993). Using several different approaches we demonstrated that Golgi enzymes are depleted in peri-Golgi 50–60 nm vesicles (Kweon et al. 2004). Finally, one of the main problem for this variant of CMPM is the different concentration of different SNAREs in COPI vesicles (Orci et al. 2000a). If one SNARE from the same SNARE complex will be relatively more concentrated in COPI vesicles very soon the stoichiometric balance of SNAREs within the same SNARE complex would be impaired. Actually, some evidence in support of this possibility has been presented. Martinez-Menarguez et al. (2001) demonstrate the presence of an almost equal (without a statistically significant difference) concentration of mannosidase II (Man II) in all Golgi cisternae and in COPI-coated peri-Golgiolar round profiles (PGRP) located within 200-nm distance at the level of cisternae (0.76 and 1.13, respectively). The authors wrote that indeed, Man II concentration in peri-Golgi vesicles is at least the same as in the Golgi stack (page 1218). One of the explanations of this contradiction between Cosson et al. (2002) and Martinez-Menarguez et al. (2001) data could be our observations (Kweon et al. 2004) that perforated zones of cisternae could be relatively enriched in Golgi enzymes. Samples studied by Klumperman group cells could contain more perforated zones and tubules enriched in Golgi enzymes appearing on sections as round profiles. In the case of Martinez-Menarguez et al. (2001), the round profiles would contain higher proportion of cross-sections of the perforated zones enriched in Golgi enzymes.

Intra-Golgi transport

*

349

Isolation of COPI vesicles in vitro Isolation of COPI-dependent vesicles enriched in Golgi enzymes could provide the proof in favour of the CMPM (Lanoix et al. 1999). However, this enrichment is caused by the presence in the light vesicular fraction fragments of Golgi cisternae (Kweon et al. 2004). Recent report (Malsam et al. 2005) shows that a small (10% input) fraction of 60-nm vesicles enriched in CASP, giantin, ManI and ManII observed after incubation of isolated Golgi membranes with purified coatomer, recombinant myristoylated ARF, GTP and a regeneration system. However, in the absence of ARFGAP, this assay system mimics similar in vitro systems based on the addition of GTPgS, and, thus, an artificial vesiculation of perforated zones of Golgi cisternae could have been induced (Weidman et al. 1993). Recently, the confirmation of these findings were published (Gilchrist et al. 2006). However, from the paper it is not clear how the authors made normalization of protein concentration and whether the experimental procedures used for the biochemical analysis of Golgi vesicle and for their EM study are fully identical. Finally, in this paper, ultra-thin sections (60 nm) were cut at various depths through the entire pellet. This means that the systematic examination of the entire pellet where as we have shown (Kweon et al. 2004) most of cisternal fragments were concentrated at the bottom was not performed.

The CMPM with connection-dependent recycling of Golgi enzymes The recycling of Golgi enzymes could occur through the intercisternal connections (Trucco et al. 2004). This gives the second variant of CMPM. This model is Golgi-matrix-dependent, because it should be a mechanism for the recycling of the matrix receptors or Golgi enzymes per se should be involved in the attachment of cisternae to each other. Also it should be a mechanism providing higher affinity of matrix receptor at the cis side of a Golgi stack, for instance, recycling of matrix proteins from the trans to the cis side of the stack. Alternatively, cytosolic tails of cis-Golgi enzymes have to have higher affinity to each other than cytosolic tails of trans enzymes. Thus, if it is possible to acutely inhibit the function of the Golgi matrix, the transport of cargo through the Golgi should be blocked. However, this mechanistic explanation faces some difficulties. GM-130 and GRASP-65 or -55 are not found inside Golgi stacks between all medical cisternae (Marra et al. 2007). If the mechanism forcing the cis membrane to go towards the cis side existed, the model should be sensitive to the absence of stacks. However, in insect S2 cells, a double depletion of dGRASP and dGM130 led to the quantitative conversion of Golgi stacks into clusters of vesicles and tubules, often featuring single cisternae. However, this disruption of Golgi architecture was not accompanied by the disorganization of tER sites or the inhibition

350

*

A. A. Mironov and G. V. Beznoussenko

of anterograde transport (Kondylis et al. 2005). Thus, normal stacks are not necessary for intra-Golgi transport. Moreover, this model cannot explain fast intra-Golgi transport in S. cerevisiae, where Golgi stacks do not exist. Next, this model requires the departure of post-Golgi carriers from the trans-most cisternae. However, the departure occurs from the last three trans cisternae (see Chapter 3.4). Another problem of this model is the fact that intercisternal connections are very narrow, curved (Marsh et al. 2004; Trucco et al. 2004), and not always regularly present within the Golgi stacks (our unpublished observations). The diffusion of membrane proteins through these connections and luminal cargo and ions could be quite different due to problems of geometry for membrane proteins and especially their aggregates which might not be able to diffuse along the highly curved tubes. Finally, this model predicts that if different amount of cargo would move synchronously through the Golgi, the number of Golgi cisternae should be different. Our data demonstrated that under the conditions of the miniwave protocol the number of cisternae in the stack is increased for only one cisterna whereas when the maxi-wave protocol is used this augmentation is two (Trucco et al. 2004).

The carrier maturation–progression model The carrier maturation–progression model (Mironov et al. 2005) represents the modification of the progression model. This model implies that cargo domain remains distinct and does not intermix with the Golgi domains containing Golgi resident proteins such as glycosylation enzymes and sugar transporters. Thus, the cargo domain preserves its identity moving through the Golgi without intermixing with the genuine Golgi cisternae like megavesicles (Volchuk et al. 2000) but the fusion with the distal compartment occurs before its fission from the proximal one. In this case, there is no necessity for the retrograde dissociation carriers (i.e. vesicles or intercisternal connections). However, the model works only if one assumes intralipid segregation of membrane proteins and/or intraluminal self-aggregation of soluble proteins due to a change of, for instance, pH, or concentration of other ions.

The kiss-and-run model (KARM) The idea of transport by lateral diffusion along the membrane continuity derives from the diagram published by Robertson (1962). The diffusion mechanism could be based not only on the constantly existing wide connections but also on narrow connections in combination with the bolus-like mechanism (Ayala 1994; Fig. 2). The main problems of the lateral diffusion model are the presence of SNARE complexes within all steps of the secretory and endocytic pathway and the existence of gradients across the Golgi. The bolus-like mechanism was adapted for intra-Golgi transport in the form of

Intra-Golgi transport

*

351

Figure 2. Scheme of the bolus-model. Upper plate. Formation of the varicosity filled with membrane (double ended arrows) and soluble (hexagons) cargoes and attachment of microtubule (MT) motor (structure with circles) to MT. Lower plate. Movement of the varicosity along the MT.

peristaltic movement of membranes (Griffiths 2000). However, until now the molecular mechanisms responsible for such movements are missing. However, if the connections are transient, one will deal with the kiss-andrun model that does not face problem due to existence of SNAREs. KARM assumes that compartments fuse with each other, and then become separated from each other. KARM has been proposed for synaptic vesicles, for the fusion of secretory granules (SGs) with the PM in neuroendocrine cells, fusion between endosomes and lysosomes (see Chapter 3.11). It seems that KARM could be useful for explanation of intra-Golgi transport. For its normal function, KARM has several requirements. 1. To ensure that correct compartments will fuse with each other it is necessary to have a mechanism for SNARE sorting along the secretory and endocytic pathways. 2. There should be working mechanism for concentration of SNARE in sites through which two compartments fuse with each other. 3. The cells have to have a mechanism to break connections. 4. The connections between organelles should be thin. If connections become thick it will be necessary to have their fission initially to make them thin. On the other hand, if connections become thick, the ionic composition of two compartments involved in kiss-and-run mechanism will be easily equilibrated. 5. Cells should have a mechanism to stimulate fusion at the defined time. 6. It should be a gradient of ionic pumps of other protein machineries regulating the concentration of ions along the secretory and endocytic pathway able to create linear gradients. 6. Kiss-and-run mechanism actually means that there is no specific retrograde transport and this transport occurs simultaneously with the anterograde transport. In the framework of the kiss-and-run model, three main coats have different functions. The important role of coat-dependent concentration through cytosolic domain of membrane proteins is the concentration of SNAREs at correct places of endomembrane compartments.

352

*

A. A. Mironov and G. V. Beznoussenko

Figure 3. Scheme of the kiss-and-run model. Temporal fusion of two compartments and then their separation. During the fusion state, the cargo (dots) moves from the compartment 2 to the compartment 1.

Kiss-and-run mechanism could exist in two forms. One when two organelles temporally fuse with each other and then, become separated (Fig. 3). The second mechanism of kiss-and-run mechanism could be the following. The temporal fusion between two organelles in one site with the consecutive fission in another places, thus, making some membrane displacement from one organelle to another. The carrier maturation model is the individual case (the asymmetric KARM) within the frame of kiss-and-run mechanism. In the framework of kiss-and-run model, the fission of membrane tubules connecting different Golgi compartments becomes one of the most important mechanisms. There are two possibilities. The first is that such machineries do not exist and only local temperature fluctuation could regulate fission. However, recently it has been shown that several proteins pretend to be in such a fission mechanism. For BARS, the most convincing data are about its role in fission of COPI-coated bud from Golgi membranes (Yang et al. 2005). BARS-dependent mechanism is interchangeable with endophilin A (Yang et al. 2006). The same mechanism could be responsible for the fission of intercisternal connections. Next mechanism might be dynamin, considered as being involved into the fission of clathrin-dependent vesicles from the PM (see Chapter 3.9).

Combined models There are two combined models of intra-Golgi transport, namely, percolating vesicles model in combination with megavesicles (Pelham and Rothman 2000) and the synthetic model (Mironov et al. 1998) representing the combination of the cisternal maturation mechanism based on COPI vesicle recycling and the diffusion mechanism along the spiral Golgi ribbon. The first scheme implies that cargo-containing COPI vesicles may “percolate up and down” the stack, limited by tethering mainly to transfer between adjacent cisternae in the stack, and largely unable to distinguish up from down because the same soluble N-ethylmaleimide-sensitive factor

Intra-Golgi transport

*

353

attachment protein receptors are present throughout. The authors explain that these vesicles would migrate in a stochastic process resembling a random walk, giving rise to the fast track of anterograde transport. Net flow in the cisto-trans direction would result because entry and exit are restricted to opposite faces, but would of course be less efficient than for unidirectional moving vesicles.

Role of COPI vesicles The main challenge for CPM and KID models is the elucidation of function for coat complexes. The most enigmatic issue is the role of COPI vesicles. First of all, COPI vesicles are not the obligatory feature of all cells. In some eukaryotes intracellular transport can occur without generation of COPI-dependent vesicles. For instance, in minimal secretory (Microsporidia) system, 50–60 nm vesicles (both COPI- and COPII-dependent) do not exist at all because both COPI and COPII machineries in this parasite are reduced (Beznoussenko et al. 2007). Several roles have been proposed for COPI-dependent vesicles. COPI vesicles could serve for the regulation of the cisterna shape or the generation of COPI-coated buds in order to facilitate subsequent uncoating of Golgi membranes. Transformation of COPI-coated buds into vesicles could accelerate uncoating of Golgi membranes from COPI. Finally, COPI vesicles could regulate formation of intercisternal connections by extraction of Qb SNAREs from cisternal membranes. COPI coat could destroy connections by formation of varicosities along membrane tubules. One of the most difficult problems for the models of intra-Golgi transport is the mode of SNARE recycling. Few SNARE complexes participate in intra-Golgi transport. One is composed of syntaxin 5, membrin, Ykt6 and Bet1. Another one consists of syntaxin 5, GOS28, Ykt6 and GS15. Additional SNARE complex operating at the cis pole of the Golgi consists of syntaxin 5, membrin, Sec22 and Bet1. If COPI vesicles function as anterograde carriers it should be a mechanism for SNARE recycling. Problems with CMPM were discussed above. In contrast, KARM could explain the situation with SNARE recycling on the basis of different characteristics of their trans-membrane domains and their interactions with different coats (Mironov et al. 2005).

Conclusions Today we have two main models competing for the correct explanation of the observations related to intra-Golgi transport: the cisterna maturation– progression models equipped with the recycling of Golgi enzymes along the intercisternal continuities and kiss-and-run mechanism that could be asymmetric for membrane cargoes. Each model has its own problems. The main problem is the function of COPI vesicles. An important feature of the KARM model is that it is able to explain the concentration of not only

354

*

A. A. Mironov and G. V. Beznoussenko

albumin or RSPs, but also of Golgi enzymes in defined Golgi compartments. This model explains why there are no Golgi enzymes in COPI-dependent vesicles. KARM can deal with the problem of SNARE recycling and the rare presence of intercisternal connections within the Golgi stacks. Thus, the kiss-and-run model represents the most promising model of intra-Golgi transport.

Abbreviations CMPM CSP ER GA GPC IC KARM PM RSP SNARE TGN VM

cisterna maturation-progression model constitutive secretory proteins endoplasmic reticulum Golgi apparatus Golgi-to-PM carrier intermediate compartment kiss-and-run model plasma membrane regulated secretory protein soluble N-ethylmaleimide-sensitive factor attachment protein receptor trans-Golgi network vesicular model

References Ayala J (1994) Transport and internal organization of membranes: vesicles, membrane networks and GTP-binding proteins. J Cell Sci 107 (Pt 4): 753–763 Bannykh SI, Balch WE (1997) Membrane dynamics at the endoplasmic reticulum–Golgi interface. J Cell Biol 138: 1–4 Becker B, Bolinger B, Melkonian M (1995) Anterograde transport of algal scales through the Golgi complex is not mediated by vesicles. Trends Cell Biol 5: 305–307 Beznoussenko GV, Dolgikh VV, Seliverstova EV, Semenov PB, Tokarev YS, Trucco A, Micaroni M, Di Giandomenico D, Auinger P, Senderskiy IV, Skarlato SO, Snigirevskaya ES, Komissarchik YY, Pavelka M, De Matteis MA, Luini A, Sokolova YY, Mironov AA (2007) Analogs of the Golgi complex in microsporidia: structure avesicular mechanisms of function. J Cell Sci 120: 1288–1298  rguez JA, Martella O, Fusella A, Baldassarre Bonfanti L, Mironov AA Jr, Martínez-Mena M, Buccione R, Geuze HJ, Mironov AA, Luini A (1998) Procollagen traverses the Golgi stack without leaving the lumen of cisternae: evidence for cisternal maturation. Cell 95(7): 993–1003 € gger B, Sandhoff R, Wegehingel S, Gorgas K, Malsam J, Helms JB, Lehmann WD, Bru Nickel W, Wieland FT (2000) Evidence for segregation of sphingomyelin and cholesterol during formation of COPI-coated vesicles. J Cell Biol 151(3): 507–518 Claude A (1970) Growth and differentiation of cytoplasmic membranes in the course of lipoprotein granule synthesis in the hepatic cell. I. Elaboration of elements of the Golgi complex. J Cell Biol 47: 745–766 Clermont Y, Rambourg A, Hermo L (1992) Segregation of secretory material in all elements of the Golgi apparatus in principal epithelial cells of the rat seminal vesicle. Anat Rec 232: 349–358

Intra-Golgi transport

*

355

Clermont Y, Xia L, Rambourg A, Turner JD, Hermo L (1993) Transport of casein submicelles and formation of secretion granules in the Golgi apparatus of epithelial cells of the lactating mammary gland of rat. Anat Rec 235: 363–373 Cosson P, Amherdt M, Rothman JE, Orci L (2002) A resident Golgi protein is excluded from peri-Golgi vesicles in NRK cells. Proc Natl Acad Sci USA. 99: 12831–12834 Dahan S, Ahluwalia JP, Wong L, Posner BI, Bergeron JJ (1994) Concentration of intracellular hepatic apolipoprotein E in Golgi apparatus saccular distensions and endosomes. J Cell Biol 127(6 Pt 2): 1859–1869 Di Lazzaro C, Campadelli-Fiume G, Torrisi MR (1995) Intermediate forms of glycoconjugates are present in the nevelope of herpes simplex virions during their transport along the exocytic pathway. Virology 214: 619–623 Dupree P, Sherrier DJ (1998) The plant Golgi apparatus. Biochim Biophys Acta 1404(1–2): 259–270 Elazar Z, Orci L, Osterman J, Amherdt M, Tanigawa G, Rothman J (1994) ADP-ribosylation factor and coatomer couple fusion to vesicle budding. J Cell Biol 124: 415–424 Gaynor EC, Graham TR, Emr SD (1998) COPI in ER/Golgi and intra-golgi transport: do yeast COPI mutants point the way? Biochim Biophys Acta 1404(1–2): 33–51 Gilchrist A, Au CE, Hiding J, Bell AW, Fernandez-Rodriguez J, Lesimple S, Nagaya H, Roy L, Gosline SJ, Hallett M, Paiement J, Kearney RE, Nilsson T, Bergeron JJ (2006) Quantitative proteomics analysis of the secretory pathway. Cell 127(6): 1265–1281 Glick BS, lston ET, Oster G (1997) A cisternal maturation mechanism can explain the asymmetry of the Golgi stack. FEBS Lett 414: 177–181 Griffiths G, Pepperkok R, Locker JK, Kreis TE (1995) Immunocytochemical localization of beta-COP to the ER–Golgi boundary and the TGN. J Cell Sci 108(Pt 8): 2839–2856 Griffiths G (2000) Gut thoughts on the Golgi complex, Traffic 1: 738–745 Guo Q, Vasile E, Krieger M (1994) Disruptions in Golgi structure and membrane traffic in a conditional lethal mammalian cell mutant are corrected by epsilon-COP. J Cell Biol 125: 1213–1224 Haas AK, Barr FA (2007) COP sets TRAPP for vesicles. Dev Cell 12(3): 326–327 Happe S, Cairns M, Roth R, Heuser J, Weidman P (2000) Coatomer vesicles are not required for inhibition of Golgi transport by G-protein activators. Traffic 1(4): 342–353 Hartel-Schenk S, Minnifield N, Reutter W, Hanski C, Bauer C, Morre DJ (1991) Distribution of glycosyltransferases among Golgi apparatus subfractions from liver and hepatomas of the rat. Biochim Biophys Acta, 1115: 108–122 Hoe MH, Slusarewicz P, Misteli T, Watson R, Warren G (1995) Evidence for recycling of the resident medial/trans Golgi enzyme, N-acetylglucosaminyltransferase I, in ldlD cells. J Biol Chem 270(42): 25057–25063 Hong W (2005) SNAREs and traffic. Biochim Biophys Acta 1744(3): 493–517 Jantii J, Hilden P, Ronka H, Makiranta V, Keranen S, Kuismanen E (1997) Immunocytochemical analysis of Uukunieli virus budding compartments: role of the intermediate compartment and the Golgi stack in virus maturation. J Virol 71(2): 1162–1172 Kaiser CA, Schekman R (1990) Distinct sets of SEC genes govern transport vesicle formation and fusion early in the secretory pathway. Cell 61(4): 723–733 Kondylis V, Spoorendonk KM, Rabouille C (2005) dGRASP localization and function in the early exocytic pathway in Drosophila S2 cells. Mol Biol Cell 16(9): 4061–4072 Kweon HS, Beznoussenko GV, Micaroni M, Polishchuk RS, Trucco A, Martella O, Di Giandomenico D, Marra P, Fusella A, Di Pentima A, Berger EG, Geerts WJ, Koster AJ, Burger KN, Luini A, Mironov AA (2004) Golgi enzymes are enriched in perforated zones of golgi cisternae but are depleted in COPI vesicles. Mol Biol Cell 15: 4710–4724 Lanoix J, Ouwendijk J, Lin CC, Stark A, Love HD, Ostermann J, Nilsson T (1999) GTP hydrolysis by arf-1 mediates sorting and concentration of Golgi resident enzymes into functional COPI vesicles. EMBO J 18: 4935–4948

356

*

A. A. Mironov and G. V. Beznoussenko

Leblond CP (1989) Synthesis and secretion of collagen by cells of connective tissue, bone, and dentin. Anat Rec 224(2): 123–138 Love HD, Lin CC, Short CS, Ostermann J (1998) Isolation of functional Golgi-derived vesicles with a possible role in retrograde transport. J Cell Biol 140: 541–551 Luby-Phelps K (1994) Physical properties of cytoplasm. Curr Opin Cell Biol 6(1): 3–9 Lucocq J, Berger E, Hug C (1995) The pathway of Golgi cluster formation in okadaic acidtreated cells. J Struct Biol 115(3): 318–330 Malsam J, Satoh A, Pelletier L, Warren G (2005) Golgin tethers define subpopulations of COPI vesicles. Science 307: 1095–1098 Marra P, Salvatore L, Mironov A Jr, Di Campli A, Di Tullio G, Trucco A, Beznoussenko G, Mironov A, De Matteis MA (2007) The biogenesis of the Golgi ribbon: the roles of membrane input from the ER and of GM130. Mol Biol Cell 18: 1595–1608 Marsh BJ, Volkmann N, McIntosh JR, Howell KE (2004) Direct continuities between cisternae at different levels of the Golgi complex in glucose-stimulated mouse islet beta cells. Proc Natl Acad Sci USA 101(15): 5565–5570 Martinez-Menarguez JA, Prekeris R, Oorschot VM, Scheller R, Slot JW, Geuze HJ, Klumperman J (2001) Peri-Golgi vesicles contain retrograde but not anterograde proteins consistent with the cisternal progression model of intra-Golgi transport. J Cell Biol 155: 1213–1224 Matsuura-Tokita K, Takeuchi M, Ichihara A, Mikuriya K, Nakano A (2006) Live imaging of yeast Golgi cisternal maturation. Nature 441(7096): 1007–1010 Mironov AA, Weidman P, Luini A (1997) Variations on the intracellular transport theme: maturing cisternae and trafficking tubules. J Cell Biol 138: 481–484 Mironov A Jr, Luini A, Mironov A (1998) A synthetic model of intra-Golgi traffic. FASEB J 12: 249–252 Mironov AA, Beznoussenko GV, Nicoziani P, Martella O, Trucco A, Kweon HS, Di Giandomenico D, Polishchuk RS, Fusella A, Lupetti P, Berger EG, Geerts WJ, Koster AJ, Burger KN, Luini A (2001) Small cargo proteins and large aggregates can traverse the Golgi by a common mechanism without leaving the lumen of cisternae. J Cell Biol 155: 1225–1238 Mironov AA, Beznoussenko GV, Polishchuk RS, Trucco A (2005) Intra-Golgi transport. A way to a new paradigm? BBA Mol Cell Res 1744: 340–350 Oprins A, Rabouille C, Posthuma G, Klumperman J, Geuze HJ, Slot JW (2001) The ER to Golgi interface is the major concentration site of secretory proteins in the exocrine pancreatic cell. Traffic 2: 831–838 Orci L, Malhotra V, Amherdt M, Serafini T, Rothman JE (1989) Dissection of a single round of vesicular transport: sequential intermediates for intercisternal movement in the Golgi stack. Cell 56(3): 357–368 Orci L, Tagaya M, Amherdt M, Perrelet A, Donaldson JG, Lippincott-Schwartz J, Klausner RD, Rothman JE (1991) Brefeldin A, a drug that blocks secretion, prevents the assembly of non-clathrin-coated buds on Golgi cisternae. Cell 64: 1183–1195 Orci L, Palmer DJ, Ravazzola M, Perrelet A, Amherdt M, Rothman JE (1993) Budding from Golgi membranes requires the coatomer complex of non-clathrin coat proteins. Nature 362(6421): 648–652 Orci L, Stamnes M, Ravazzola M, Amherdt M, Perrelet A, Sollner TH, Rothman JE (1997) Bidirectional transport by distinct populations of COPI-coated vesicles. Cell 90(2): 335–349 Orci L, Perrelet A, Rothman JE (1998) Vesicles on strings: morphological evidence for processive transport within the Golgi stack. Proc Natl Acad Sci USA 95(5): 2279–2283 Orci L, Amherdt M, Ravazzola M, Perrelet A, Rothman JE (2000a) Exclusion of Golgi residents from transport vesicles budding from Golgi cisternae in intact cells. J Cell Biol 150: 1263–1270 Orci L, Ravazzola M, Volchuk A, Engel T, Gmachl M, Amherdt M, Perrelet A, Sollner TH, Rothman JE (2000b) Anterograde flow of cargo across the golgi stack potentially

Intra-Golgi transport

*

357

mediated via bidirectional “percolating” COPI vesicles. Proc Natl Acad Sci USA 97 (19): 10400–10405 Pelham HR, Rothman JE (2000) The debate about transport in the Golgi – two sides of the same coin? Cell 102: 713–719 Pepperkok R, Scheel J, Horstmann H, Hauri HP, Griffiths G, Kreis TE. (1993) Beta-COP is essential for biosynthetic membrane transport from the endoplasmic reticulum to the Golgi complex in vivo. Cell 74(1): 71–82 Pullikuth AK, Weidman PJ (2002) In vitro transport on cis and trans sides of the Golgi involves two distinct types of coatomer and ADP-ribosylation factor-independent transport intermediates. J Biol Chem 277: 50355–50364  tien M, Olivier L (1993) Modulation of the Golgi Rambourg A, Clermont Y, Chre apparatus in stimulated and nonstimulated prolactin cells of female rats. Anat Rec 235(3): 353–362 Robertson JD (1962) The membrane of the living cell. Sci Am 206: 65–72 Rothman JE, Miller RL, Urbani LJ (1984) Intercompartmental transport in the Golgi complex is a dissociative process: facile transfer of membrane protein between two Golgi populations. J Cell Biol 99: 260–271 Rothman JE, (1994) Mechanisms of intracellular protein transport. Nature 372: 55–63 Rothman S, Liebow C, Grendell J (1991) Nonparallel transport and mechanisms of secretion. Biochim Biophys Acta 1071(2): 159–173 Sabesin SM, Frase S (1977) Electron microscopic studies of the assembly, intracellular transport, and secretion of chylomicrons by rat intestine. J Lipid Res 18: 496–511 Severs NJ, Hicks RM (1979) Analysis of membrane structure in the transitional epithelium of rat urinary bladder. 2. The discoidal vesicles and Golgi apparatus: their role in luminal membrane biogenesis. J Ultrastruct Res 69: 279–296 € llner TH, Rothman JE (1998) ADP-ribosylation Stamnes M, Schiavo G, Stenbeck G, So factor and phosphatidic acid levels in Golgi membranes during budding of coatomer-coated vesicles. Proc Natl Acad Sci USA 95(23): 13676–13680 Taylor TC, Kanstein M, Weidman P, Melan¸con P (1994) Cytosolic ARFs are required for vesicle formation but not for cell-free intra-Golgi transport/evidence for coated vesicle-independent transport. Mol Biol Cell 5: 237–252 Trucco A, Polishchuk RS, Martella O, Di Pentima A, Fusella A, Di Giandomenico D, San Pietro E, Beznoussenko GV, Polishchuk EV, Baldassarre M, Buccione R, Geerts WJ, Koster AJ, Burger KN, Mironov AA, Luini A (2004) Secretory traffic triggers the formation of tubular continuities across Golgi sub-compartments. Nat Cell Biol 6(11): 1071–1081 Velasco A, Hendricks L, Moreman KW, Tulsiani DRP, Touster O, Farquhar MG (1993) Celltype dependent variations in the subcellular distribution of a-mannosidase I and II. J Cell Biol 122: 39–51 € gger B, Rivera VM, Clackson T, Perrelet A, Volchuk A, Amherdt M, Ravazzola M, Bru € llner TH, Rothman JE, Orci L (2000) Megavesicles implicated in the rapid transport So of intracisternal aggregates across the Golgi stack. Cell 102: 335–348 Weidman P, Roth R, Heuser J (1993) Golgi membrane dynamics imaged by freeze-etch electron microscopy: views of different membrane coatings involved in tubulation versus vesiculation. Cell 75(1): 123–133 Yang JS, Lee SY, Spanò S, Gad H, Zhang L, Nie Z, Bonazzi M, Corda D, Luini A, Hsu VW (2005) A role for BARS at the fission step of COPI vesicle formation from Golgi membrane. EMBO J 24(23): 4133–4143 Yang JS, Zhang L, Lee SY, Gad H, Luini A, Hsu VW (2006) Key components of the fission machinery are interchangeable. Nat Cell Biol 8(12): 1376–1382

358

*

Z. Z. Lieu et al.

Structure and domain organization of the trans-Golgi network Zi Zhao Lieu, Merran C. Derby and Paul A. Gleeson

Introduction The trans-Golgi network (TGN) is a unique compartment located at the exit face of the Golgi stack (Griffiths and Simons 1986) typically associated with a large amount of vesicular and tubular membranes (Ladinsky et al. 1994; Mollenhauer and Morre 1998; Roth and Taatjes 1998). The TGN varies morphologically between different cell types, considered to be a consequence of the differences in secretory activity (Clermont et al. 1995; Gu et al. 2001). Like other Golgi cisternae, the TGN contains a number of resident enzymes involved in the processing of cargo molecules, such as glycosyltransferases involved in the addition of terminal sugars (Rabouille et al. 1995), several proprotein convertases including furin (Thomas 2002) and tyrosine sulphation enzymes. However, unique functions and characteristics clearly distinguish the TGN from the other Golgi cisterna. Firstly, the TGN sorts various mature cargo proteins and lipids into membrane carriers destined for the plasma membrane, endosomes, secretory granules, or earlier Golgi cisternae or the ER (Keller and Simons 1997; Traub and Kornfeld 1997; Opat et al. 2001; Sannerud et al. 2003; Young et al. 2005). Secondly, the TGN receives cargo from various endosomal locations (Bos et al. 1993; Pavelka et al. 1998; Medigeshi and Schu 2003; Shewan et al. 2003). Thus the TGN presents a trafficking hub where the secretory and endocytic pathways merge. Thirdly, the TGN responds differently to treatments which block membrane transport. For example, treatment with brefeldin-A results in the membranes of the Golgi stack rapidly fusing into the ER, while the TGN tubulates and merges with the recycling endosomal system (Chege and Pfeffer 1990; Lippincott-Schwartz et al. 1991). Fourthly, an intimate relationship between the recycling endosome and the TGN has been highlighted by the recent identification of the Rab linker protein, Rab6Ip1, which interacts with Rab6 from the TGN and Rab11 from the recycling endosome (Miserey-Lenkei et al. 2007). Indeed, whether the TGN should be considered as a Golgi subcompartment or as an endocytic compartment is debatable. How is the identity of the TGN established? The identity of the TGN is likely to be defined by membrane components delivered to the TGN from both anterograde and retrograde transport, together with the recruitment from the cytosol of coat proteins, regulatory GTPases, fusion and tethering factors and motor proteins. Over the past 5–10 years, membrane cell biologists have

Structure and domain organization of the trans-Golgi network

*

359

realised the importance of the recruitment of cytosolic components in defining the identity of intracellular compartments (Behnia and Munro 2005) and the generation of individual functional membrane domains within one compartment.

Trafficking at the TGN At least five different pathways are known to transport proteins from the TGN to the plasma membrane and endosomes, and at least two distinct pathways are known to transport proteins from the TGN to earlier Golgi cisternae (Roth 2004). Since a variety of membrane carriers, containing the correct cargo molecules, can be generated from the one continuous membrane bilayer of TGN, it follows that the membrane composition of this compartment will not be homogeneous but is likely to consist of a variety of membrane domains where particular membrane-bound carriers originate. There is evidence emerging that the TGN consists of functionally distinct subdomains to manage the complexity of traffic within this compartment (Gleeson et al. 2004). Early electron microscopic studies revealed that the TGN was highly tubular in nature (Griffiths and Simons 1986), providing the potential for the separation of sorting and transport activities. More recently, electron microscopy tomography has revealed that the TGN contains vesicle budding domains or ‘exit sites’, which give rise to distinctly coated vesicle carriers and uncoated vesicles and tubules (White et al. 2001; Ladinsky et al. 2002). These cargo domains are enriched in cargo and decorated with vesicle budding machinery and are also devoid of resident Golgi proteins. Visualization of the trafficking of cargo molecules from the TGN has dramatically illustrated the dynamics of transport into and from the Golgi apparatus. A range of different membrane carriers or post-Golgi carriers (PGCs) emerge from the TGN, from conventional transport vesicles to highly dynamic tubular membrane structures which can be greater than 10 mm in length (Hirschberg et al. 1998; Toomre et al. 1999; Polishchuk et al. 2000). Live cell imaging studies have revealed that a variety of membrane cargo are transported in these tubular PGCs from the TGN to the cell surface, including VSV-G, Nras, Connexins and E-cadherin (Hirschberg et al. 1998; Choy et al. 1999; Lock et al. 2005; Thomas et al. 2005). Live cell imaging has also revealed the segregation of proteins within the TGN. The separation of apical and basolateral membrane proteins within the TGN prior to the emergence of transport carriers has been visualised within the TGN by live imaging. For example, dual imaging of cyan and yellow fluorescently-tagged basolateral (VSV-G) and apical (GPI) cargo revealed that they are segregated into separate domains on Golgi membranes, and thus VSV-G and GPI are found on different PGCs (Keller et al. 2001). An emerging view proposes that membrane proteins are segregated into distinct membrane domains of the TGN, domains which then emerge as tubular extensions and which can subsequently break away to form distinct PGCs.

360

*

Z. Z. Lieu et al.

Several retrograde transport pathways from the endosomal compartments to the TGN have been identified (Sannerud et al. 2003; Bonifacino and Rojas 2006). Retrograde transport in the endocytic route to the TGN is important for the recycling of endogenous proteins such as mannose-6-phosphate receptor (M6PR), transmembrane peptidases, SNAREs, and the glucose transporter GLUT4 as well as the internalisation of bacterial and plant toxins (Ghosh et al. 1998, 2003; Lewis et al. 2000; Shewan et al. 2003; Sandvig and Van Deurs 2005; Bonifacino and Rojas 2006). There are a number of t-SNAREs located on the TGN which are required for the fusion of transport intermediates involved in these retrograde transport pathways and it is likely that this docking and fusion machinery is also restricted to distinct domains of the TGN membranes. Attention is now turning to exploring the trafficking events in a variety of specialised cells. One recent fascinating study has identified at least two distinct specialised cytokine secretion pathways in activated T helper cells (Huse et al. 2006) and the regulation of these pathways is likely to reflect the capacity of the TGN to segregate these cytokines into distinct domains of the TGN.

Maintenance of TGN subdomains Functional subdivision of the TGN implies biochemical heterogeneity of membrane domains. As with other organelles, small G proteins together with lipids are likely to play central roles in defining membrane identity and the organisation of membrane subdomains. Small G proteins act as switches to regulate recruitment of proteins to the cytosolic face of membranes. In the inactive, or GDP-bound state, many G proteins are located in the cytosol. On binding GTP, activated G proteins become membrane associated, a process that may be dependent on the localisation of specific GTP exchange factors (GEFs). In turn, membrane-bound activated G proteins recruit a range of effector molecules, including coat adaptor proteins, tethering molecules, and motor proteins to the membrane. A given G protein can recruit a distinct array of protein complexes and thereby regulate the molecular architecture of a subdomain to specify distinct trafficking events. The involvement of G proteins and peripheral membrane proteins, rather than integral membrane proteins that would have to transported from the ER, has the advantage that protein association with membranes is highly regulated and can be transient. Therefore functional domains can be readily formed and dissociated in a highly dynamic manner. The sorting and trafficking functions of the TGN have similarities to the early endosome. The generation of subdomains of the early endosome has provided a paradigm for understanding the establishment of the membrane subdomains. Like the TGN, the early endosome has a highly dynamic structure comprising of thin tubules and larger vesicles, as well as membrane invaginations and tubular and multivesicular elements (Gruenberg 2001). Different SNARE proteins are concentrated in different regions of the early

Structure and domain organization of the trans-Golgi network

*

361

endosomal membrane. For example, both syntaxin 7 and syntaxin 13 are associated with early endosomal membranes, with syntaxin 7 concentrated on the cisternal-like region and syntaxin 13 on the tubular region (Prekeris et al. 1999). This distribution of SNARE proteins would restrict syntaxin 7- and syntaxin 13-mediated fusion events to distinct regions of the early endosomal membrane, contributing to the subdomain maintenance. The small G protein, Rab5, is critical to the establishment of subdomains of the early endosome. Rab5 regulates the production of a local pool of PtdIns(3)P at the early endosomal membrane through its interaction with various phosphate kinases and phosphatases, resulting in the amplification of the recruitment of Rab5 and its effectors recruitment such as EEA1. Significantly, effectors for Rab5 include GEFs for downstream Rabs (Rink et al. 2005). This finding has lead to the proposal of a ‘Rab cascade’; an ordered recruitment of Rab proteins, mediated by Rab effectors, that accounts for different sets of Rab proteins on different endosomal organelles (Rink et al. 2005). The cross-talk between different small G proteins and their effectors are likely to play fundamental roles in the establishment and regulation of distinct subdomains of a compartment. The recycling endosome also regulates distinct trafficking pathway which is reflected by the division of this compartment into domains akin to the early endosome (Manderson et al. 2007; Thompson et al. 2007). Small G proteins, in particular members of the Rab11 family, play a critical role in defining different functional domains of the recycling endosome (Lapierre et al. 2003). Given the intimate relationship between the recycling endosome and the TGN it is likely that similar mechanisms are used define distinct functional regions of this distal Golgi compartment.

Small G proteins and lipids of the TGN A number of small G proteins have been identified at the TGN (Table 1). These include a number of Rabs, including the well-characterised Rab6 and Rab8 and other more poorly characterised Rabs including Rab10, Rab14, Rab30, Rab31. In addition, there are also a number of members of the ADP-ribosylation factor (Arf) family which are specifically located on the TGN. Arf1 is the most well characterised member and GTP-bound Arf1 recruits effectors to the Golgi including COPI, clathrin adaptor protein I (API) and Golgi-localised, g-ear containing, Arf-binding proteins (GGAs) (Crottet et al. 2002; Heldwein et al. 2004; Traub 2005). AP-1 and Arf appear to interact directly, however it is likely that AP-1 requires both Arf-GTP and PtdIns(4)P for efficient membrane binding, hence AP-1 recognises Arf-GTP in the context of local PtdIns(4)P. GGAs are recruited to the TGN by the interaction of membrane-bound-Arf-GTP with their GAT domain. PtdIns(4)P is also a key regulator of GGA binding (Wang et al. 2007). The combinatorial binding to both Arf1 and PtdIns (4)P ensures fidelity in the recruitment of defined effectors and is likely to restrict the location of complex assembly to particular sites on the membranes where both PtdIns(4)P and Arf1 are spatially co-localised.

362

*

Z. Z. Lieu et al.

Table 1. Small G proteins located at the TGN G proteins

Localisation

Effector molecules and function

Refs.

Arl1

TGN

TGN golgins, golgin 97 and p230/golgin 245. Involved in both anterograde and retrograde transport

(Lowe et al. 1996; Lu et al. 2004)

Arl5a

Golgi/TGN

Unknown

(Lin et al. 2002; Gillingham and Munro 2007)

Arl5b

Golgi/TGN

Unknown

(Gillingham and Munro 2007)

ARFRP1

TGN

Binds to hsys1 and recruits Arl1 to TGN membrane

(Panic et al. 2003b; Setty et al. 2003; Behnia et al. 2004; Shin et al. 2005)

ARF1

TGN/endosome

Membrane recruitment of COP I, GGA and AP-1

(D’Souza-Schorey and Chavrier 2006)

Rab6

Golgi/TGN

Retrograde transport of shiga toxin

(Mallard et al. 2002; Fridmann-Sirkis et al. 2004; Del Nery et al. 2006)

Rab8

TGN

Binds to AP-1B in MDCK cells and is involved in basolateral transport

(Ang et al. 2003)

Rab10?

TGN (early stage of polarization)/ endosome

Unknown effector, disrupts anterograde transport of VSVG in MDCK cells

(Schuck et al. 2007)

Rab14?

TGN

Unknown effectors

(Junutula et al. 2004; Proikas-Cezanne et al. 2006)

Rab30/Rab31

TGN

Possible effector includes TGN46. Involved in anterograde transport of VSVG

(Ng et al. 2007)

Cdc42

TGN/Golgi

Regulatory role in constitutive protein exit from the TGN

(Musch et al. 2001)

Structure and domain organization of the trans-Golgi network

*

363

The complete Arf family includes G proteins that do not have ADP-ribosylation factor activity but share with the Arf members a number of structural features (Tamkun et al. 1991; Pasqualato et al. 2002; Kahn et al. 2005, 2006). The more distantly related members are called the Arf-like (Arl) proteins (Clark et al. 1993; Pasqualato et al. 2002). Some of the members of this poorly characterised Arl family have been shown to be involved in membrane transport (Gillingham and Munro 2007). Arl1 has been shown to be enriched on the TGN (Lowe et al. 1996; Lu et al. 2001) and is involved in membrane transport. Overexpression of Arl1 in mammalian cells results in dramatic changes in the structure of the Golgi and a block in transport along the secretory pathway (Lu et al. 2001; Van Valkenburgh et al. 2001). In the yeast, Saccharomyces cerevisiae ARL1-null cells also showed several defects in membrane traffic, including TGN to endosome transport (Rosenwald et al. 2002). The membrane recruitment of Arl1 is in turn dependent on another member of the Arl small G protein family, namely Arl3p/ARFRP1 (Gangi Setty et al. 2003; Panic et al. 2003b; Behnia et al. 2004). Arl3p is important in TGN trafficking in yeast and deletion of the Arl3 counterpart in mice is embryonically lethal (Mueller et al. 2002). TGN Arls may also include Arl5a and Arl5b (Table 1). The Arl family was mainly identified by genomic analysis and many of the Arl members have yet to be characterised, thus it is possible there are additional Arl members localised at the TGN. The prospect then arises of an “Arl-casade” on TGN membranes, analogous to the Rab-casade of endosomes, which promotes the generation of functional subdomains. Membrane attachment of the Arf and Arls is commonly mediated by N-terminal myristoylation and the nucleotide state of these G proteins is tightly coupled to membrane binding (Lee et al. 1997; Lu et al. 2001; Pasqualato et al. 2002). Hence these small G proteins recycle between the cytosolic GDP-bound form and a GTP membrane-bound form. A search for putative effectors of Arl1 by yeast two-hybrid screening, using GTP-restricted Arl1, identified a number of interactive components including p230/golgin-245 and golgin-97 (Lu et al. 2001; Van Valkenburgh et al. 2001), two coiled-coil proteins previously known to be targeted specifically to the of the TGN (Gleeson et al. 1996; Luke et al. 2003b). This key finding initiated studies in a number of laboratories, from which emerged a relationship between Arls and TGN-tethering molecules that is critical for Golgi structure and membrane transport.

Golgins of the TGN Long coiled-coil proteins of the Golgi apparatus have been grouped together as a family called golgins. Members of the golgin family have been shown to be important in the biogenesis of membranes of the Golgi stack and in the regulation of membrane transport (Gillingham and Munro 2003; Lupashin and Sztul 2005; Short et al. 2005). Many golgins are peripheral membrane proteins that are recruited to the donor membrane in a highly regulated

364

*

Z. Z. Lieu et al.

manner (Gillingham and Munro 2003). A number of golgin molecules have been found specifically associated with the TGN; these include Bicaudal-D1 and D2 (Matanis et al. 2002) and four mammalian GRIP domain proteins (Kjer-Nielsen et al. 1999a; Munro and Nichols 1999). The functions of these golgins are being actively investigated. Bicaudal D1 and D2 have been shown to be involved in Rab6-mediated COPI-independent retrograde transport from the TGN (Matanis et al. 2002; Young et al. 2005). On the other hand the GRIP domain golgins represent a unique subfamily of golgins and work over the past few years have shown they have central functions in maintaining TGN structure and in membrane trafficking.

GRIP domain proteins The first of the mammalian TGN golgins was identified using sera from a patient with an systemic autoimmune disease, known as p230/golgin-245 (Kooy et al. 1992). The subsequent identification of a novel Golgi localisation domain near the C-terminus of p230 lead to the identification of a number of coiled-coil proteins in a range of organisms showing a modestly conserved C-terminal targeting domain, called the GRIP domain (Kjer-Nielsen et al. 1999a,b; Munro and Nichols 1999). GRIP domain sequences have been identified in mammals, flies, worms, plants, yeast and parasites. The identification of a member of this family of TGN golgins in the primitive eukaryotic cell, Leishmania mexicana (McConville et al. 2002), indicates that these GRIP domain proteins play a central role in the function of the TGN. Aside from the abundance of predicted coiled-coil structures, the only sequence similarity shared by members of this family is within this GRIP domain, therefore it is quite possible that the functions of the golgin family members will be diverse. Whereas there is a single GRIP domain protein in yeast there are four human golgins with GRIP domains, namely p230/golgin-245, golgin-97, GCC185 and GCC88 (GCC for Golgi localised Coiled-Coil protein), and all four mammalian TGN golgins are localised specifically to TGN membranes (Kooy et al. 1992; Fritzler et al. 1995; Erlich et al. 1996; Gleeson et al. 1996; Griffith et al. 1997; Luke et al. 2003b). In each case the isolated GRIP domain is sufficient and necessary for TGN localisation of these proteins.

Recruitment of GRIP domain proteins to TGN membranes Recruitment of GRIP domain proteins to the TGN is G protein dependent. Arl1, which is localised to the TGN, binds to the GRIP domains of the mammalian GRIP domain proteins, golgin-97 and p230/golgin-245 and also Arl1p binds the sole yeast GRIP domain protein, Imh1p (Gangi Setty et al. 2003; Jackson 2003; Lu and Hong 2003; Panic et al. 2003b). Structural studies of the Arl1-GTP complex with the GRIP domain of p230/golgin-245 have revealed that the isolated GRIP domain forms a homodimer that interacts with two Arl1-GTP molecules (Panic et al. 2003a; Wu et al. 2004). The interaction with Arl1 is directly responsible for the Golgi recruitment of GRIP domain proteins in both

Structure and domain organization of the trans-Golgi network

*

365

mammalian cells and yeast (Gangi Setty et al. 2003; Jackson 2003; Lu and Hong 2003; Panic et al. 2003b). There has been some controversy as to whether the GRIP domains of GCC88 and GCC185 can also bind Arl1. Using a number of in vitro and in vivo approaches our laboratory and others have demonstrated that the GRIP domains of GCC88 and GCC185 have different membrane binding properties from the GRIP domains of golgin-97 and p230/golgin-245 (Lu and Hong 2003; Derby et al. 2004; Reddy et al. 2006). Arl1-GTP fused to the early endosomal PI (3)P binding protein, SNX3, efficiently redirects endogenous golgin-97 and p230/golgin-245 to the endosomes. On the other hand, the Arl1-SNX3 fusion protein is unable to redirect GCC88 and GCC185 to endosomes, which strongly suggests that these two GRIP domain proteins are unable to bind Arl1 in vivo (Derby et al. 2004). The finding that the GRIP domains of GCC185 and GCC88 do not bind to Arl1 in vivo is consistent with yeast two-hybrid analysis, in which GTP-restricted Arl1 interacted poorly with the GRIP domains of either GCC88 or GCC185 (Lu and Hong 2003). Pull down assays, on the other hand had indicated that, at high concentrations, the GRIP domains of GCC88 and GCC185 can bind to glutathione beads loaded with GST-Arl1(Q71L) (Panic et al. 2003a), although using a similar strategy another study failed to detect an interaction between Arl1-GTP and GCC185 (Reddy et al. 2006). In view of the concentrations of the two species used in the pull down assays, low affinity interactions may be detected that are not physiologically relevant. As the GRIP domains of GCC185 and GCC88 have been shown to bind to TGN membranes in a G protein dependent manner (Luke et al. 2003a), G proteins other than Arl1 are necessary for the membrane recruitment of these two golgin family members. The possibility that GRIP sequences of GCC88 and GCC185 may interact with other members of the Arl family is currently being investigated by our laboratory.

TGN domain specificity of golgins The mammalian TGN golgins have a high percentage of a-helical heptad repeats and exist as dimers. Even through all four TGN golgins are found in the same cell, the four mammalian GRIP-domain proteins self-interact to form homodimers exclusively (Luke et al. 2005). Therefore, each TGN golgin has the potential to function independently from the other members of the GRIP domain family. An independent function for each golgin is supported by the finding that mammalian GRIP domain proteins are localised to distinct subdomains of the TGN. Firstly, expression of pairs of GRIP domains resulted in distinct, non-overlapping golgin-positive structures (Lock et al. 2005; Luke et al. 2005). Secondly, live imaging studies have shown the GRIP domains of p230 and golgin-97 to be recruited to distinct tubular subdomains of the TGN (Lock et al. 2005). Thirdly, overexpression of different full length TGN golgins has been shown to generate golgin-specific phenotypes. Overexpression of full length GCC88 results in extensive membrane structures that arise from the TGN (Luke et al. 2003b), whereas overexpression of full length GCC185

366

*

Z. Z. Lieu et al.

resulted in the dispersal of TGN-derived tubular structures (Derby et al. 2004). The TGN-derived membranes in the golgin phenotypes showed enrichment in each case for a particular combination of TGN components (Derby et al. 2004). Thus, the four different TGN golgins may contribute to the maintenance of distinct subdomains of the mammalian TGN.

Functions of mammalian golgins in membrane trafficking and maintenance of TGN structure The recruitment of TGN golgins to distinct domains points to specific roles for each golgin in the trafficking functions of the TGN. The recent application of RNAi in particular has demonstrated that the TGN golgins play key roles in the regulation of both anterograde and retrograde membrane transport pathways as well as maintaining the structural integrity of the TGN (see Fig. 1).

Anterograde transport There were early suggestions from EM and live imaging that p230 may be associated with TGN-derived transport carriers (Gleeson et al. 1996; Brown et al. 2001). Additional indirect evidence for a role of p230 in anterograde

Figure 1. TGN golgins and transport pathways. Shown is a summary of the anterograde and retrograde transport pathways which are regulated by the different TGN golgins, as defined by individual cargoes. Also indicated is the proposal that each TGN golgin defines a specific subdomain of the TGN. Golgins are indicated by the long coil-coiled molecules, Arl1 by the orange squares and unidentified small G proteins by a red square and blue triangle.

Structure and domain organization of the trans-Golgi network

*

367

transport was indicated by studies where overexpression of p230 domains inhibited the TGN-to-cell-surface transport of a GPI-anchored protein (Kakinuma et al. 2004). More compelling evidence comes from recent work that demonstrates that p230 is associated with tubulovesicular carriers loaded with TNFa (Lock et al. 2005). Furthermore, silencing p230 expression inhibits the transport of membrane TNFa, but not E-cadherin, to the plasma membrane (Lieu et al. 2008), strongly suggesting that p230 is an essential component of a population of TGN-derived tubules which defines a membrane transport pathway to the cell surface. Although both p230 and golgin-97 are effectors of Arl1, co-expression of full length GFP-p230 and myc-golgin-97 revealed a distinct and non-overlapping localisation of each golgin (Lock et al. 2005). Co-expression of the two GRIP domains showed that even the isolated GRIP domains are able to specify binding to distinct membrane domains (Lock et al. 2005). These findings indicate that the GRIP domains contain localisation determinants in addition to those specifying Ar1l binding. Distinct spatial segregation of p230 and golgin-97 is also reflected in their function. Golgin-97, but not p230, is associated with distinct membrane extensions of the TGN loaded with E-cadherin from the Golgi (Lock et al. 2005). siRNA knock down of golgin-97 selectively blocked exit of E-cadherin cargo from the TGN (Lock et al. 2005). Given that post-Golgi exocytic trafficking of E-cadherin involves transport via the recycling endosome to the cell surface, golgin-97 is likely to be a regulator of transport step from the TGN to the recycling endosome. Thus it appears that golgin-97 and p230 are essential components of distinct sets of post-Golgi or tubulovesiclar carriers which emerge from the TGN to transport E-cadherin and TFNa, respectively. How these TGN golgins contribute to the biogenesis of cargo loaded tubulovesicular carriers is not yet clear. The identification of the golgin interactive partners, in addition to Arl1, will be required to better define the molecular mechanism for the regulation of this post-Golgi trafficking pathway.

Retrograde transport In contrast to the Arl1-dependent TGN golgins, GCC88 and GCC185 function in the regulation of endocytic transport rather than exocytic transport. By overexpressing GCC88 it was noted that the recycling membrane protein, TGN38 was associated with GCC88-labeled TGN membranes (Luke et al. 2003b), implying a role for the GCC88 subdomain of the transport of TGN38. Recently, our laboratory has shown that silencing GCC88 resulted in a block in retrograde transport of not only TGN38, but also a fusion protein of mannose-6-phosphate receptor M6P-R (Lieu et al. 2007). In GCC88-depleted cells the cargo accumulated in the early endosomes. In contrast the transport of Shiga toxin to the TGN was unaffected in GCC88-depleted cells, indicating the presence of more than one retrograde transport pathway from the early endosomes. Furthermore, depletion of GCC88 resulted in a perturbation of the intracellular distribution of the TGN t-SNARE, syntaxin 6 (Lieu et al. 2007). The trafficking defect of GCC88 depletion could be rescued by overexpression

368

*

Z. Z. Lieu et al.

of syntaxin 6, illustrating a direct link between the mislocalisation of this t-SNARE and the block in retrograde transport. Syntaxin 6 is a t-SNARE and along with syntaxin 16 and Vti1a, pair with the v-SNARE VAMP4 (or VAMP3) (Mallard et al. 2002). These findings suggest that GCC88 is an effector molecular that mediates the recruitment of SNARE molecules required for the docking and fusion of transport intermediates derived from the early endosomes. Notably, GCC88 is the first example of a TGN-tethering molecule that can influence the localisation of a SNARE at the TGN. The basis of the interaction between GCC88 and syntaxin 6 is likely to involve regions of GCC88 other than the GRIP domain. The Arl1-independent GCC185 golgin has been shown to play a role in the regulation of endosomal to TGN membrane transport (Derby et al. 2007). In contrast to GCC88, GCC185 depletion resulted in a block in shiga toxin trafficking but had no affect on the retrograde transport of TGN38 (Derby et al. 2007). GCC185 depletion resulted in an accumulation of Shiga toxin in Rab11 positive recycling endosomes, which implies that GCC185 is required for transport of shiga toxin between the recycling endosome and the TGN. Analysis of the retrograde transport pathways in GCC88- and GCC185-depleted cells has thereby revealed two independent retrograde transport pathways which are regulated by the Arl1 independent TGN golgins. Depletion of either GCC88 or GCC185 also resulted in perturbation in the distribution of the M6P-R (Derby et al. 2007; Lieu et al. 2007), suggesting that the trafficking of the M6P-R utilises multiple pathways. From the analysis of M6P-R recycling, Reddy et al. (2006) suggested that GCC185 may be involved in Rab9-mediated transport of M6P-R from the late endosomes to the TGN. Given these findings it is possible that GCC185 may be involved in multiple endosome to Golgi pathways, and more than one pathway may contribute to M6P-R recycling. Early preliminary studies also implicated golgin-97 and p230/golgin-245 in the regulation of retrograde membrane trafficking between the endosomal system and TGN (Lu et al. 2004; Yoshino et al. 2005), although other studies have not supported a role for, at least p230, in retrograde transport (Lieu et al. 2007). It remains to be established whether there is a direct contribution by either of the Arl-dependent TGN golgins, golgin-97 or p230, to retrograde transport, or whether their functions are restricted to anterograde transport.

Maintenance of Golgi structure There is evidence that TGN golgins, in particular p230/golgin-245 (Yoshino et al. 2003) and GCC185 (Derby et al. 2007), contribute to the maintenance of TGN structure. The most compelling evidence has been reported for GCC185. Depletion of GCC185 in HeLa cells using either siRNA and microRNA also resulted in fragmentation of the Golgi (Derby et al. 2007). The individual Golgi fragments contained both cis- and trans-markers and moreover electron microscopy showed the presence of Golgi stacks in the cytoplasm of RNAi transfected cells. The key finding that the microtubule binding protein,

Structure and domain organization of the trans-Golgi network

*

369

CLASP, is recruited to the TGN by the golgin GCC185 and is required for Golgi associated micotubule arrays (Efimov et al. 2007) provides a molecular basis for a direct relationship between this TGN golgin, cytoskeletal interactions and the maintanence of the Golgi ribbon. The dual functions of GCC185 in maintenance of the Golgi ribbon and regulation of retrograde transport is likely to reflect the ability of this golgin to recruit multiple components to the Golgi apparatus. Clearly it will be important to identity the full set of effector molecules of GCC185.

Conclusion The molecular landscape underlying the architecture of the TGN is now beginning to be resolved. The recruitment of peripheral membrane proteins plays a critical role in defining the identity and organization of the TGN. Multiple resident small G proteins are pivotal in these processes by orchestrating the recruitment of arrays of effector molecules which provide the opportunity to establish distinct functional subdomains. The GRIP domain TGN golgins have emerged as key molecules for regulating TGN function. RNAi analyses in cultured cells has demonstrated that each TGN golgin has a unique role in the regulation of membrane transport. More molecular information is now required on the nature of the golgin complexes that regulate these pathways. In addition, it is likely that the full appreciation of the four mammalian GRIP domain proteins will only be revealed by studying their function in a range of differentiated and specialised cell types.

References Ang AL, Folsch H, Koivisto UM, Pypaert M, Mellman I (2003) The Rab8 GTPase selectively regulates AP-1B-dependent basolateral transport in polarized Madin-Darby canine kidney cells. J Cell Biol 163: 339–350 Behnia R, Munro S (2005) Organelle identity and the signposts for membrane traffic. Nature 438: 597–604 Behnia R, Panic B, Whyte JR, Munro S (2004) Targeting of the Arf-like GTPase Arl3p to the Golgi requires N-terminal acetylation and the membrane protein Sys1p. Nat Cell Biol 6: 405–413 Bonifacino JS, Rojas R (2006) Retorgrade transport from endosomes to the trans-Golgi network. Nat Rev Mol Cell Biol 7: 568–579 Bos K, Wraight C, Stanley KK (1993) TGN38 is maintained in the trans-Golgi network by a tyrosine-containing motif in the cytoplasmic domain. EMBO J 12: 2219–2228 Brown DL, Heimann K, Lock J, Kjer-Nielsen L, Van Vliet C, Stow JL, Gleeson PA (2001) The GRIP domain is a specific targeting sequence for a population of trans-Golgi network derived tubulo-vesicular carriers. Traffic 2: 336–344 Chege NW, Pfeffer SR (1990) Compartmentation of the Golgi complex: brefeldin-A distinguishes trans-Golgi cisternae from the trans-Golgi network. J Cell Biol 111: 893–899 Choy E, Chiu VK, Silletti J, Feoktistov M, Morimoto T, Michaelson D, Ivanov IE, Phillips MR (1999) Endomembrane trafficking of Ras: the CAAX motif targets proteins to the ER and Golgi. Cell 98: 69–80

370

*

Z. Z. Lieu et al.

Clark J, Moore L, Krasinskas A, Way J, Battey J, Tamkun J, Kahn RA (1993) Selective amplification of additional members of the ADP-ribosylation factor (ARF) family: cloning of additional human and Drosophila ARF-like genes. Proc Natl Acad Sci USA 90: 8952–8956 Clermont Y, Rambourg A, Hermo L (1995) Trans-Golgi network (TGN) of different cell types: three-dimensional structural characteristics and variability. Anat Rec 242: 289–301 Crottet P, Meyer DM, Rohrer J, Spiess M (2002) ARF1.GTP, tyrosine-based signals, and phosphatidylinositol 4,5-bisphosphate constitute a minimal machinery to recruit the AP-1 clathrin adaptor to membranes. Mol Biol Cell 13: 3672–3682 D’Souza-Schorey C, Chavrier P (2006) ARF proteins: roles in membrane traffic and beyond. Nat Rev Mol Cell Biol 7: 347–358 Del Nery E, Miserey-Lenkei S, Falguieres T, Nizak C, Johannes L, Perez F, Goud B (2006) Rab6A and Rab6A0 GTPases play non-overlapping roles in membrane trafficking. Traffic 7: 394–407 Derby MC, Lieu ZZ, Brown D, Stow JL, Goud B, Gleeson PA (2007) The trans-Golgi network golgin, GCC185, is required for endosome-to-Golgi transport and maintenance of Golgi structure. Traffic 8: 758–773 Derby MC, Van Vliet C, Brown D, Luke MR, Lu L, Hong W, Stow JL, Gleeson PA (2004) Mammalian GRIP domain proteins differ in their membrane binding properties and are recruited to distinct domains of the TGN. J Cell Sci 117: 5865–5874 Efimov A, Kharitonov A, Efimova N, Loncarek J, Miller PM, Andreyeva N, Gleeson P, Galjart N, Maia AR, McLeod IX, Yates JR III, Maiato H, Khodjakov A, Akhmanova A, Kaverina I (2007) Asymmetric CLASP-dependent nucleation of noncentrosomal microtubules at the trans-Golgi network. Developmental Cell 12: 917–930 Erlich R, Gleeson PA, Campbell P, Dietzsch E, Toh BH (1996) Molecular characterization of trans-Golgi p230 – a human peripheral membrane protein encoded by a gene on chromosome 6p12-22 contains extensive coiled-coil alpha-helical domains and a granin motif. J Biol Chem 271: 8328–8337 Fridmann-Sirkis Y, Siniossoglou S, Pelham HR (2004) TMF is a golgin that binds Rab6 and influences Golgi morphology. BMC Cell Biol 5: 18 Fritzler MJ, Lung CC, Hamel JC, Griffith KJ, Chan EKL (1995) Molecular characterization of golgin-245, a novel Golgi complex protein containing a granin signature. J Biol Chem 270: 31262–31268 Gangi Setty SR, Shin ME, Yoshino A, Marks MS, Burd CG (2003) Golgi recruitment of GRIP domain proteins by Arf-like GTPase 1 is regulated by Arf-like GTPase 3. Curr Biol 13: 401–404 Ghosh P, Dahms NM, Kornfeld S (2003) Mannose 6-phosphate receptors: new twists in the tale. Nat Rev Mol Cell Biol 4: 202–212 Ghosh RN, Mallet WG, Soe TT, McGraw TE, Maxfield FR (1998) An endocytosed TGN38 chimeric protein is delivered to the TGN after trafficking through the endocytic recycling compartment in CHO cells. J Cell Biol 142: 923–936 Gillingham AK, Munro S (2003) Long coiled-coil proteins and membrane traffic. Biochim Biophys Acta 1641: 71–85 Gillingham AK, Munro S (2007) The small G proteins of the Arf family and their regulators. Annu Rev Cell Dev Biol 23: 579–611 Gleeson P, Lock JG, Luke MR, Stow JL (2004) Domains of the TGN: coats, tethers and G proteins. Traffic 5: 315–326 Gleeson PA, Anderson TJ, Stow JL, Griffiths G, Toh BH, Matheson F (1996) p230 is associated with vesicles budding from the trans-Golgi network. J Cell Sci 109: 2811–2821 Griffith KJ, Chan EKL, Lung CC, Hamel JC, Guo XY, Miyachi K, Fritzler MJ (1997) Molecular cloning of a novel 97-kDa Golgi complex autoantigen associated with Sjogrens-syndrome. Arthritis Rheum 40: 1693–1702

Structure and domain organization of the trans-Golgi network

*

371

Griffiths G, Simons K (1986) The trans Golgi network: sorting at the exit site of the Golgi complex. Science 234: 438–443 Gruenberg J (2001) The endocytic pathway: a mosaic of domains. Nat Rev Mol Cell Biol 2: 721–730 Gu F, Crump CM, Thomas G (2001) Trans-Golgi network sorting. Cell Mol Life Sci 58: 1067–1084 Heldwein EE, Macia E, Wang J, Yin HL, Kirchhausen T, Harrison SC (2004) Crystal structure of the clathrin adaptor protein 1 core. Proc Natl Acad Sci USA 101: 14108–14113 Hirschberg K, Miller CM, Ellenberg J, Presley JF, Siggia ED, Phair RD, Lippincott-Schwartz J (1998) Kinetic analysis of secretory protein traffic and characterization of golgi to plasma membrane transport intermediates in living cells. J Cell Biol 143: 1485–1503 Huse M, Lillemeier BF, Kuhns MS, Chen DS, Davis MM (2006) T cells use two directionally distinct pathways for cytokine secretion. Nat Immunol 7: 247–255 Jackson CL (2003) Membrane traffic: Arl GTPases get a GRIP on the Golgi. Curr Biol 13: R174–R176 Junutula JR, De Maziere AM, Peden AA, Ervin KE, Advani RJ, Van Dijk SM, Klumperman J, Scheller RH (2004) Rab14 is involved in membrane trafficking between the Golgi complex and endosomes. Mol Biol Cell 15: 2218–2229 Kahn RA, Cherfils J, Elias M, Lovering RC, Munro S, Schurmann A (2006) Nomenclature for the human Arf family of GTP-binding proteins: ARF, ARL, and SAR proteins. J Cell Biol 172: 645–650 Kahn RA, Volpicelli-Daley L, Bowzard B, Shrivastava-Ranjan P, Li Y, Zhou C, Cunningham L (2005) Arf family GTPases: roles in membrane traffic and microtubule dynamics. Biochem Soc Trans 33: 1269–1272 Kakinuma T, Ichikawa H, Tsukada Y, Nakamura T, Toh BH (2004) Interaction between p230 and MACF1 is associated with transport of a glycosyl phosphatidyl inositolanchored protein from the Golgi to the cell periphery. Exp Cell Res 298: 388–398 Keller P, Simons K (1997) Post-Golgi biosynthetic trafficking. J Cell Sci 110(Pt 24): 3001–3009 Keller P, Toomre D, Diaz E, White J, Simons K (2001) Multicolour imaging of post-Golgi sorting and trafficking in live cells. Nat Cell Biol 3: 140–149 Kjer-Nielsen L, Teasdale RD, Van Vliet C, Gleeson PA (1999a) A novel Golgi-localisation domain shared by a class of coiled-coil peripheral membrane proteins. Curr Biol 9: 385–388 Kjer-Nielsen L, Van Vliet C, Erlich R, Toh BH, Gleeson PA (1999b) The Golgi-targeting sequence of the peripheral membrane protein p230. J Cell Sci 112: 1645–1654 Kooy J, Toh BH, Pettitt JM, Erlich R, Gleeson PA (1992) Human autoantibodies as reagents to conserved Golgi components – characterization of a peripheral, 230-kDa compartment-specific Golgi protein. J Biol Chem 267: 20255–20263 Ladinsky MS, Kremer JR, Furcinitti PS, McIntosh JR, Howell KE (1994) HVEM tomography of the trans-Golgi network: structural insights and identification of a lace-like vesicle coat. J Cell Biol 127: 29–38 Ladinsky MS, Wu CC, McIntosh S, McIntosh JR, Howell KE (2002) Structure of the Golgi and distribution of reporter molecules at 20 C reveals the complexity of the exit compartments. Mol Biol Cell 13: 2810–2825 Lapierre LA, Dorn MC, Zimmerman CF, Navarre J, Burnette JO, Goldenring JR (2003) Rab11b resides in a vesicular compartment distinct from Rab11a in parietal cells and other epithelial cells. Exp Cell Res 290: 322–331 Lee FJ, Huang CF, Yu WL, Buu LM, Lin CY, Huang MC, Moss J, Vaughan M (1997) Characterization of an ADP-ribosylation factor-like 1 protein in Saccharomyces cerevisiae. J Biol Chem 272: 30998–31005 Lewis MJ, Nichols BJ, Prescianotto-Baschong C, Riezman H, Pelham HR (2000) Specific retrieval of the exocytic SNARE Snc1p from early yeast endosomes. Mol Biol Cell 11: 23–38

372

*

Z. Z. Lieu et al.

Lieu ZZ, Derby MC, Teasdale RD, Hart C, Gunn P, Gleeson PA (2007) The Golgin, GCC88, is required for efficient retrograde transport of cargo from the early endosomes to the trans-Golgi network. Mol Biol Cell 18(12): 4979–4991 Lieu ZZ, Lock JG, Hammond LA, La Gruta NL, Stow JL, Gleeson PA (2008) A trans-Golgi network golgin is required for the regulated secretion of TNF in activated macrophages. Proc Natl Acad Sci 105: 3351–3356 Lin CY, Li CC, Huang PH, Lee FJ (2002) A developmentally regulated ARF-like 5 protein (ARL5), localized to nuclei and nucleoli, interacts with heterochromatin protein 1. J Cell Sci 115: 4433–4445 Lippincott-Schwartz J, Yuan L, Tipper C, Amherdt M, Orci L, Klausner RD (1991) Brefeldin A’s effects on endosomes, lysosomes, and the TGN suggest a general mechanism for regulating organelle structure and membrane traffic. Cell 67: 601–616 Lock J, Houghton F, Gleeson PA, Stow JL (2005) E-cadherin transport from the TGN in tubulovesicular carriers is selectively regulated by golgin-97. Traffic 6: 1142–1156 Lowe SL, Wong SH, Hong W (1996) The mammalian ARF-like protein 1 (Arl1) is associated with the Golgi complex. J Cell Sci 109(Pt 1), 209–220 Lu L, Hong W (2003) Interaction of Arl1-GTP with GRIP domains recruits autoantigens Golgin-97 and Golgin-245/p230 onto the Golgi. Mol Biol Cell 14: 3767–3781 Lu L, Horstmann H, Ng C, Hong W (2001) Regulation of Golgi structure and function by ARF-like protein 1 (Arl1). J Cell Sci 114: 4543–4555 Lu L, Tai G, Hong W (2004) Autoantigen Golgin-97, an effector of Arl1 GTPase, participates in traffic from the endosome to the TGN. Mol Biol Cell 15(10): 4426–4443 Luke MR, Brown DL, Stow JL, Gleeson PA (2003a) Two mammalian GRIP-domain proteins, GCC88 and golgin97, exist as homo-dimers and are recruited to distinct sub-compartments of the TGN. Mol Biol Cell 14 (supplement): 106a Luke MR, Houghton F, Perugini MA, Gleeson PA (2005) The trans-Golgi network GRIP-domain proteins form alpha-helical homodimers. Biochem J 388: 835–841 Luke MR, Kjer-Nielsen L, Brown DL, Stow JL, Gleeson PA (2003b) GRIP domain-mediated targeting of two new coiled-coil proteins, GCC88 and GCC185, to subcompartments of the trans-Golgi network. J Biol Chem 278: 4216–4226 Lupashin V, Sztul E (2005) Golgi tethering factors. Biochim Biophys Acta 1744: 325–339 Mallard F, Tang BL, Galli T, Tenza D, Saint-Pol A, Yue X, Antony C, Hong W, Goud B, Johannes L (2002) Early/recycling endosomes-to-TGN transport involves two SNARE complexes and a Rab6 isoform. J Cell Biol 156: 653–664 Manderson AP, Kay JG, Hammond LA, Brown DL, Stow JL (2007) Subcompartments of the macrophage recycling endosome direct the differential secretion of IL-6 and TNFalpha. J Cell Biol 178: 57–69 Matanis T, Akhmanova A, Wulf P, Del Nery E, Weide T, Stepanova T, Galjart N, Grosveld F, Goud B, De Zeeuw CI, Barnekow A, Hoogenraad CC (2002) Bicaudal-D regulates COPI-independent Golgi–ER transport by recruiting the dynein–dynactin motor complex. Nature Cell Biol. 4: 986–992 McConville MJ, Ilgoutz SC, Teasdale RD, Foth BJ, Matthews A, Mullin KA, Gleeson PA (2002) Targeting of the GRIP domain to the trans-Golgi network is conserved from protists to animals. Eur J Cell Biol 81: 485–495 Medigeshi GR, Schu P (2003) Characterization of the in vitro retrograde transport of MPR46. Traffic 4: 802–811 Miserey-Lenkei S, Waharte F, Boulet A, Cuif MH, Tenza D, El Marjou A, Raposo G, Salamero J, Heliot L, Goud B, Monier S (2007) Rab6-interacting protein 1 Links Rab6 and Rab11 function. Traffic 8: 1385–1403 Mollenhauer HH, Morre DJ (1998) The tubular network of the Golgi apparatus. Histochem Cell Biol 109: 533–543 Mueller AG, Moser M, Kluge R, Leder S, Blum M, Buttner R, Joost HG, Schurmann A (2002) Embryonic lethality caused by apoptosis during gastrulation in mice lacking the gene of the ADP-ribosylation factor-related protein 1. Mol Cell Biol 22: 1488–1494

Structure and domain organization of the trans-Golgi network

*

373

Munro S, Nichols BJ (1999) The GRIP domain- a novel Golgi-targeting domain found in several coiled-coil proteins. Curr Biol 9: 377–380 Musch A, Cohen D, Kreitzer G, Rodriguez-Boulan E (2001) cdc42 regulates the exit of apical and basolateral proteins from the trans-Golgi network. EMBO J 20: 2171–2179 Ng EL, Wang Y, Tang BL (2007) Rab22B’s role in trans-Golgi network membrane dynamics. Biochem Biophys Res Commun 361: 751–757 Opat AS, Houghton F, Gleeson PA (2001) Steady-state localization of a medial-Golgi glycosyltransferase involves transit through the trans-Golgi network. Biochem J 358: 33–40 Panic B, Perisic O, Veprintsev DB, Williams RL, Munro S (2003a) Structural basis for Arl1-dependent targeting of homodimeric GRIP domains to the Golgi apparatus. Mol Cell 12: 863–874 Panic B, Whyte JR, Munro S (2003b) The ARF-like GTPases Arl1p and Arl3p act in a pathway that interacts with vesicle-tethering factors at the Golgi apparatus. Curr Biol 13: 405–410 Pasqualato S, Renault L, Cherfils J (2002) Arf, Arl, Arp and Sar proteins: a family of GTP-binding proteins with a structural device for ‘front-back’ communication. EMBO Rep 3: 1035–1041 Pavelka M, Ellinger A, Debbage P, Loewe C, Vetterlein M, Roth J (1998) Endocytic routes to the Golgi apparatus. Histochem Cell Biol 109: 555–570 Polishchuk RS, Polishchuk EV, Marra P, Alberti S, Buccione R, Luini A, Mironov AA (2000) Corrective light-electron microscopy reveals the tubular–saccular ultrastructure of carriers operating between the Golgi apparatus and plasma membrane. J Cell Biol 148: 45–58 Prekeris R, Yang B, Oorschot V, Klumperman J, Scheller RH (1999) Differential roles of syntaxin 7 and syntaxin 8 in endosomal trafficking. Mol Biol Cell 10: 3891–3908 Proikas-Cezanne T, Gaugel A, Frickey T, Nordheim A (2006) Rab14 is part of the early endosomal clathrin-coated TGN microdomain. FEBS Lett 580: 5241–5246 Rabouille C, Hui N, Hunte F, Kieckbusch R, Berger EG, Warren G, Nilsson T (1995) Mapping the distribution of Golgi enzymes involved in the construction of complex oligosaccharides. J Cell Sci 108(Pt 4): 1617–1627 Reddy JV, Burguete AS, Sridevi K, Ganley IG, Nottingham RM, Pfeffer SR (2006) A functional role for the GCC185 golgin in mannose 6-phosphate receptor recycling. Mol Biol Cell 17: 4353–4363 Rink J, Ghigo E, Kalaidzidis Y, Zerial M (2005) Rab conversion as a mechanism of progression from early to late endosomes. Cell 122: 735–749 Rosenwald AG, Rhodes MA, Van Valkenburgh H, Palanivel V, Chapman G, Boman A, Zhang CJ, Kahn RA (2002) ARL1 and membrane traffic in Saccharomyces cerevisiae. Yeast 19: 1039–1056 Roth J, Taatjes DJ (1998) Tubules of the trans Golgi apparatus visualized by immunoelectron microscopy. Histochem Cell Biol 109: 545–553 Roth M (2004) New candidates for vesicle coat proteins. Nat Cell Biol 6: 384–385 Sandvig K, Van Deurs B (2005) Delivery into cells: lessons learned from plant and bacterial toxins. Gene Ther 12: 865–872 Sannerud R, Saraste J, Goud B (2003) Retrograde traffic in the biosynthetic-secretory route: pathways and machinery. Curr Opin Cell Biol 15: 438–445 Schuck S, Gerl MJ, Ang A, Manninen A, Keller P, Mellman I, Simons K (2007) Rab10 is involved in basolateral transport in polarized Madin-Darby canine kidney cells. Traffic 8: 47–60 Setty SR, Shin ME, Yoshino A, Marks MS, Burd CG (2003) Golgi recruitment of GRIP domain proteins by Arf-like GTPase 1 is regulated by Arf-like GTPase 3. Curr Biol 13: 401–404 Shewan AM, Van Dam EM, Martin S, Luen TB, Hong W, Bryant NJ, James DE (2003) GLUT4 recycles via a trans-Golgi network (TGN) subdomain enriched in Syntaxins 6

374

*

Z. Z. Lieu et al.

and 16 but not TGN38: involvement of an acidic targeting motif. Mol Biol Cell 14: 973–986 Shin HW, Kobayashi H, Kitamura M, Waguri S, Suganuma T, Uchiyama Y, Nakayama K (2005) Roles of ARFRP1 (ADP-ribosylation factor-related protein 1) in post-Golgi membrane trafficking. J Cell Sci 118: 4039–4048 Short B, Haas A, Barr FA (2005) Golgins and GTPases, giving identity and structure to the Golgi apparatus. Biochim Biophys Acta 1744: 383–395 Tamkun JW, Kahn RA, Kissinger M, Brizuela BJ, Rulka C, Scott MP, Kennison JA (1991) The arflike gene encodes an essential GTP-binding protein in Drosophila. Proc Natl Acad Sci USA 88: 3120–3124 Thomas G (2002) Furin at the cutting edge: from protein traffic to embryogenesis and disease. Nat Rev Mol Cell Biol 3: 753–766 Thomas T, Jordan K, Simek J, Shao Q, Jedeszko C, Walton P, Laird DW (2005) Mechanisms of Cx43 and Cx26 transport to the plasma membrane and gap junction regeneration. J Cell Sci 118: 4451–4462 Thompson A, Nessler R, Wisco D, Anderson E, Winckler B, Sheff D (2007) Recycling endosomes of polarized epithelial cells actively sort apical and basolateral cargos into separate subdomains. Mol Biol Cell 18: 2687–2697 Toomre D, Keller P, White J, Olivo J-C, Simons K (1999) Dual-color visualisation of trans-Golgi network to plasma membrane traffic along microtubules in living cells. J Cell Sci 112: 21–33 Traub LM (2005) Common principals in clathrin-mediated sorting at the Golgi and the plasma membrane. Biochim Biophys Acta 1744: 415–437 Traub LM, Kornfeld S (1997) The trans-Golgi network: a late secretory sorting station. Curr Opin Cell Biol 9: 527–533 Van Valkenburgh H, Shern JF, Sharer JD, Zhu X, Kahn RA (2001) ADP-ribosylation factors (ARFs) and ARF-like 1 (ARL1) have both specific and shared effectors: characterizing ARL1-binding proteins. J Biol Chem 276: 22826–22837 Wang J, Sun HQ, Macia E, Kirchhausen T, Watson H, Bonifacino JS, Yin HL (2007) PI4P promotes the recruitment of the GGA adaptor proteins to the trans-Golgi network and regulates their recognition of the ubiquitin sorting signal. Mol Biol Cell 18: 2646–2655 White J, Keller P, Stelzer EH (2001) Spatial partitioning of secretory cargo from Golgi resident proteins in live cells. BMC Cell Biol 2: 19 Wu M, Lu L, Hong W, Song H (2004) Structural basis for recruitment of GRIP domain golgin-245 by small GTPase Arl1. Nat Struct Mol Biol 11: in press Yoshino A, Bieler BM, Harper DC, Cowan DA, Sutterwala S, Gay DM, Cole NB, McCaffery JM, Marks MS (2003) A role for GRIP domain proteins and/or their ligands in structure and function of the trans Golgi network. J Cell Sci 116: 4441–4454 Yoshino A, Setty SR, Poynton C, Whiteman EL, Saint-Pol A, Burd CG, Johannes L, Holzbaur EL, Koval M, McCaffery JM, Marks MS (2005) tGolgin-1 (p230, golgin-245) modulates Shiga-toxin transport to the Golgi and Golgi motility towards the microtubule-organizing centre. J Cell Sci 118: 2279–2293 Young J, Stauber T, del Nery E, Vernos I, Pepperkok R, Nilsson T (2005) Regulation of microtubule-dependent recycling at the trans-Golgi network by Rab6A and Rab6A0 . Mol Biol Cell 16: 162–177

Golgi-to-PM transport

*

375

Golgi-to-PM transport Roman S. Polishchuk, Alberto Luini and Alexander A. Mironov

A detailed description of the structure and function of Golgi exit sites (GES) is given in Chapters 1.2 and 3.3. In this Chapter, the mechanisms of the cargo exit from the Golgi apparatus (GA) towards the plasma membrane (PM) will be described, while also considering the structure of the Golgi-to-PM carriers (GPCs) that are formed. Transport from the GA to endosomes and to lysosomes, and cargo sorting to the apical and basolateral PM surfaces in polarized cells are also discussed in separated chapters.

Models of transport After its passage through the Golgi stack, cargo then exits from the GA toward its final destination. The process of exit can be described using three main models: (1) the vesicular model, posing that there is the formation of small, coat-dependent, vesicles that serve as individual carriers, or form vesicular aggregates/clusters of vesicles, or fuse with each other to produce larger carriers (Rothman and Wieland 1996; Polishchuk et al. 2003); (2) the “classical” cisterna maturation model, which assumes that the last cisternae of the Golgi stack gradually lose their resident Golgi proteins, and then break down into GPCs; (3) the carrier maturation model, which requires that cargo moving through the GA preserves its identity within its membrane domain, which itself undergoes separation from the bulk of the Golgi membranes at the trans side of the Golgi stack (see Chapters 1 and 3.3). Allan and Vale (1994) first demonstrated the existence of these cargo domains in isolated Golgi membranes from rat liver that were placed on substrate covered by polymerized microtubules (MTs). When these membranes attached to MTs were incubated with cell cytosol, the formation of blobs was induced; these moved away from the GA along polymerized MTs, with the help of kinesin. The blobs contained chylomicrons or lipid micelles, typical cargoes that exit from the GA in hepatocytes. During their departure, the cargo domains preserved their connections with the GA for some time. The straight tubules that were seen could have been a result of the stretching (drawing out) of the membranes behind the GPCs they moved along the MTs. Thus these particles moving from the GA remained connected with it by a thin tubule. Finally, there is also the lateral diffusion model, which implies that a constant physical continuity exists between the GA (at least for the trans-

376

*

R. S. Polishchuk et al.

Golgi) and the PM, which appears not to be the case. The main problem for this diffusion model is the presence of SNAREs at each step of secretory pathway. This would indicate that the connections should be transient, a case that corresponds with the kiss-and-run model (see Chapter 3.2). Reconstitution of Golgi-to-PM transport in vitro has not given direct evidence in favour of vesicular carriers (reviewed in Pimplikar et al. 1994). Several attempts to isolate TGN-derived transport intermediates have provided only “the possibility that isolated vesicles represent carriers en route to the cell surface” (De Curtis and Simons 1989). Furthermore, there is the strongest objection against the vesicular model that stems from observations showing the intra-Golgi, and thus post-Golgi, transport of large supramolecular aggregates (see Chapter 3.2). Within the framework of the cisterna maturation model, the removal of the resident proteins from the last Golgi cisterna could occur by coat-dependent vesicular carriers. If this is the case, the concentrations of cargoes in GPCs should be higher than in the Golgi membranes themselves, due to the elimination of volume by vesicles carrying resident proteins back to the proximal cisternae of the stack. However, this appears not to be the case (Polishchuk et al. 2003). To resolve this problem in the framework of the “classical” cisterna maturation scheme, one can envisage the presence of a more distal, post-Golgi compartment from which another population of retrograde vesicles arrives to dilute the content of newly forming GPCs. However, the delivery of vesicles to GPCs budding from the GA has never been detected. Therefore, this would argue against this model. Thus, additional work is necessary to resolve this question.

The structure of GPCs In general, the GPC life cycle can be said to consist of five stages: (i) formation; (ii) maturation; (iii) separation from the Golgi/fission; (iv) transition through the cytosol towards the cell periphery; and (v) docking and fusion with the PM (Polishchuk et al. 2000, 2003). GPCs are formed from membrane domains of the GA that lack resident Golgi enzymes, and they are initially known as “GPC precursors” (Wacker et al. 1997; Hirschberg et al. 1998; Keller et al. 2001; Liljedahl et al. 2001; Polishchuk et al. 2003; Puertollano et al. 2003). Precursors of GPCs-containing VSVG–GFP are composed of bundles of anastomotic tubular–saccular membrane in continuity with the parent Golgi membranes (Polishchuk et al. 2003), similar to the bulging domains described by Ladinsky et al. (2002). GPC precursors do not concentrate constitutive cargoes such as VSVG. Using the 40–32 C temperature shift for the synchronization of tsVSVG, Hirschberg et al. (1998) and Polishchuk et al. (2003) have demonstrated that the concentration of secretory proteins in GPCs does not differ from that in the Golgi cisternae. In contrast, other data suggest that concentrat-

Golgi-to-PM transport

*

377

ing of VSVG does occur, and that it depends on the cytosolic tail of VSVG (Pimplikar et al. 1994) and is AP1B or AP3-dependent (Ang et al. 2003). This discrepancy might be related to the cargo load within the GA using different synchronization protocols (Mironov et al. 2001), as well as to the levels of expression of the adaptor proteins, which vary from one cell type to another. The GPC precursors dock onto and are extruded along MTs (Keller et al. 2001; Liljedahl et al. 2001; Polishchuk et al. 2003). This has been shown to be mechanistically possible for both artificial and natural membranes (Roux et al. 2002). The protruding tubular precursors of the GPCs have a particular molecular composition that is different from that of the COPI-positive Golgi cisternae, trans-Golgi/TGN or the bulk TGN. Precursors of GPCs destined to the basolateral PM generally contain Golgi- and the TGN-resident proteins such as Man6PR, syntaxin 6 and furin, and they are enriched in cargo proteins (Polishchuk et al. 2003). Thus these GPC precursors remain in physical continuity with the GA for some time after their departure from the Golgi zone. For instance, precursors of the GPCs carrying PCI are seen as cisternal distensions that are surrounded by highly perforated cisternae, which appear as a tubular network that most probably is a part of the TGN. In serial sections, these GPCs are often connected with one of the last three Golgi cisternae. Correlative light electron microscopy has revealed that GPCs that serve for constitutive transport of diffusible cargoes are neither small isolated vesicles nor clusters of vesicles and even the GPC precursors located outside of the Golgi stacks are, in many cases, connected to the medial cisternae by tubules. Finally, it is established that VSVG-containing tubulo-vesicular structures do not arise from the fusion of small vesicles (Polishchuk et al. 2003). Our data (Trucco et al. 2004) also suggest that VSVG can depart from the last 2–3 cisternae of the Golgi stack. A straight thin tubule connecting a budding GPC and the GA could be a result of stretching (drawing out) of the membrane behind the post-Golgi carrier as it moves along a MT (Allan and Vale 1994). Similar conclusions regarding the exit of cargo from multiple cisternae at the trans pole of the stack were formulated after an examination of the 20 C temperature block by EM tomography (Ladinsky et al. 2002). Thus, the structure of these GPC precursors and the exclusion of the role of membrane fusion in the generation of GPCs argue against a role for coat-dependent vesicles.

Elimination of Golgi enzymes from GPC precursors A precursor destined for conversion into a GPC undergoes a maturation that allows it to leave the stack. This process of GPC maturation is accompanied by changes in their composition and morphology. The very first step in GPC formation coincides with the segregation of cargo

378

*

R. S. Polishchuk et al.

proteins from Golgi-resident proteins, and especially Golgi glycosylation enzymes and sugar transporters. Mature GPCs are completely devoid of Golgi enzymes (Polishchuk et al. 2003) and sugar transporters (Fusella et al. submitted). The mechanisms responsible for the elimination of Golgi-resident proteins from maturing GPCs remain among those most enigmatic. One consideration is that this elimination of Golgi-resident proteins from a GPC precursor might occur due to the input of thicker membranes from the endosomal system into the cargo-containing Golgi domain. This mechanism is based on the concept that when transmembrane domains that are restricted by flanking hydrophilic segments are shorter than the thickness of the membrane, the presence of this membrane protein in such a bilayer is energetically unfavourable. In non-polarized cells, where the apical domain of the PM does not exist, the endosomes represent a single compartment that has greater membrane thickness than that of the GA (Mitra et al. 2004, see Chapter 2.16). The GA receives input of cholesterol/ sphingolipid-rich membranes mostly from endocytic system. Indeed, using perfringolysin O as a cholesterol-sensing probe, strong labelling for cholesterol was shown to be associated with internal vesicles of multivesicular bodies (Mobius et al. 2002). Therefore, the fusion of cargo-containing Golgi domains with membranes of endosomal origin in the TGN area (see Chapter 1.2) would allow cholesterol and other glycolipids with saturated acyl chains to diffuse into the GPC precursors. This would lead to an increase in the membrane thickness of the GPC precursors, which would force the Golgi enzymes and cis SNAREs (but not trans SNAREs) to move from this thicker membrane into the Golgi cisternae along tubular connections, thus, removing the Golgi-resident proteins. Another possible mechanism of cargo segregation from the Golgi enzymes relates to the ability of cargo and Golgi enzymes to undergo selfaggregation under specific ionic conditions of the luminal fluid. Narrow connections between a domain with a low pH and a low Ca2 þ concentration (as for the endosomes; see Chapter 2.8) and a domain where the pH is higher and the concentration of Ca2 þ is elevated in comparison with the endosomes (see Chapter 2.8) could attract the Golgi enzymes rather than cargoes. Thus, these possible mechanisms for the recycling of the resident Golgi proteins need at least temporary connections between GPC precursors and the Golgi cisternae.

The fissioning of GPCs In order to go outside of the GA, the precursors of the GA should loose their connections with the GA. This could be realized with the help of membrane fission mechanisms. However, the fission mechanisms working at the GES remain unclear.

Golgi-to-PM transport

*

379

One possibility could be that there are no specific mechanisms for membrane fission at the level of GES. Indeed, a MT-mediated pulling force on the tubules could be the primary mechanism for the non-specific fission of tubules connecting GPCs and the GA. This fissioning of GPCs has been seen in vitro although rarely (Allan and Vale 1994). Based on in vitro data, membranes under tension have recently been proposed to have an important role in fission (Roux et al. 2006). The boundaries between the different lipid domains at the TGN represent fission-prone areas, as lipid-phase segregation strongly stimulates the fission of tubular membranes budding from liposomes (Schuk and Simons 2004; Roux et al. 2005). Another mechanism that might be involved in the fission of GPCs relates to the rows of fenestrae that divide the precursors of GPCs and the rest of the exit cisternae (Ladinsky et al. 2002). In spermatides (Ho et al. 1999), the fenestrated zones undergo rupture from their edges, thereby forming strings of connected, but partially separated, small spheres, which usually remained connected to the edges of these saccules. How this rupture process occurs and what proteins are involved into it remain unclear. Live-cell imaging has demonstrated that fission of GPCs from a tubule emanating from the Golgi mass does not take place randomly. CLEM has provided evidence to suggest that the fission occurs at the thinnest parts of a GPC precursor (Polishchuk et al. 2003). The movement of kinesins along microtubules can thus create tension within a GPC precursor that will facilitate this unspecific fission process. However, if MT-mediated pulling is the main factor responsible for fission, there is then the problem of explaining transport in cells deprived of MTs (Trucco et al. 2004), which appear almost normal. Thus, there should be additional protein-dependent mechanisms. Several potential protein candidates for an involvement in GPC fission from the GA are being investigated, including protein kinase D (PKD; Liljedahl et al. 2001), dynamin (Cao et al. 2000; Kreitzer et al. 2000; De Matteis and Luini 2008; see also Chapter 2.15), phosphatidylinositol-transfer proteins (Simon et al. 1998), phospholipase D (PLD; Ktistakis et al. 1996), CtBP1-S/ BARS (Weigert et al. 1999; De Matteis and Luini 2008) and phosphoinositide kinases (Godi et al. 1999). Gbg, diacylglycerol (DAG), protein kinase C (PKC)h and PKD act in series to regulate events that can ultimately lead to membrane fission during transport carrier formation from the TGN (Bard and Malhotra 2006). The possible fissioning roles of CtBP1-S/BARS depend on the cargo (Corda et al. 2006; De Matteis and Luini 2008). Interestingly, when the transport of GPCs from the TGN was blocked by transfection of a mutant PKD (Liljedahl et al. 2001), these GPC precursors appeared as complex tubular–reticular structures at the EM level, like those seen in control cells (Polishchuk et al. 2003). However, till now the function of these proteins as the specific fission machines at the level of GES has to be confirmed by independent studies. Thus, what is the specific fission machine involved in the formation of GPCs remains unclear.

380

*

R. S. Polishchuk et al.

Delivery of GPCs to the plasma membrane After its exit from the GA, an anterograde cargo is transported by the GPCs, to moving towards its destination sites. In general, most GPCs are transported away from the GA towards the periphery of the cell until they have fused (Palazzo et al. 2001). According to Hirschberg et al. (1998), GPCs can move as separated entities, or in few cases as a varicosity, along the tubular continuity of the putative post-Golgi compartment (the bolus mechanism). Mature GPCs-containing VSVG represent vacuolar or tubular structures of variable sizes, with some as small as 200 nm (Keller et al. 2001), and they exhibit bonafide fenestrae (Polishchuk et al. 2003). The GPCs that ferry GFP–GPI are large tubular–saccular containers (Polishchuk et al. 2004). In neurons, synaptic, apical PM and TGN proteins are all transported from the TGN to the axonal PM via tubulo-vesicular structures of various sizes (Nakata et al. 1998). Clathrin, AP1, AP3, and GGAs are excluded from GPCs ferrying VSVG (Polishchuk et al. 2003). GPCs label for kinesin (Polishchuk et al. 2003). However, kinesin immunostaining did not accompany TGN-derived tubular extensions after treatment with brefeldin A, but instead remained at the site occupied by the TGN (Johnson et al. 1996). Finally, cytoplasmic dynein and myosin I, but not kinesin, have been found on the small vesicular membranes containing TGN38, a wellknown TGN marker, and isolated from polarized intestinal epithelial cells (Fath et al. 1994).

Visualization of GPC movement in live cells Observations in live cells have revealed that tubulo-vesicular projections and particles extend rather quickly away from existing elements toward the cell periphery (Cooper et al. 1990). At the immunofluorescence level, GPCs are frequently seen as “pleiomorphic” structures as they have a tubulo- (vacuolar) saccular appearance. Even globular structures are frequently stretched into tubular shapes (Hirschberg et al. 1998; Polishchuk et al. 2000; Liljedahl et al. 2001). Both small vacuoles (<250 nm) and larger vesicular–tubular structures (>1.5 mm long) move rapidly out of the Golgi complex along curvilinear pathways, with average speeds of approximately 0.7 mM/s (Wacker et al. 1997; Hirschberg et al. 1998; Nakata et al. 1998; Toomre et al. 1999; Polishchuk et al. 2000). A chimera of VSVG tagged with GFP has been shown to exit the TGN in tubules (Hirschberg et al. 1998). These tubules range from a few hundred nanometers to 1.7 mm in length and are therefore significantly larger than typical coated vesicles (50–100 nm) (Polishchuk et al. 2000). The exit of cargo from the GA depends on the characteristics of the cargo. The drainage of the GA from aggregates of procollagen I (non-diffusible cargo) take less than 40 min, whereas drainage of the GA from VSVG, a

Golgi-to-PM transport

*

381

diffusible cargo, is slower. Indeed, VSVG can still be found in the GA for as long as 1 h after the release of the 15 C block (Lin et al. 1999). However, when a cell synthesizes both small (VSVG) and supramolecular (procollagen-I) cargo proteins, these cargoes can leave the GA within the same GPCs, suggesting that both of these cargoes use the same mechanisms to exit from the GA (Polishchuk et al. 2003).

The role of microtubules The en bloc protrusion of tubular GPC precursors takes place along MTs by virtue of a mechanical force that is generated by motors, probably of the kinesin family (Kreitzer et al. 2000, 2003). GPCs can use a MT/kinesin-dependent motor system for their delivery to the PM. MT-dependent movement of cargo domains from the GA towards the plus end of MTs has been demonstrated in vitro (Allan and Vale 1994). Here, isolated Golgi membranes from rat liver were placed on a substrate covered by polymerized microtubules. Incubation of these membranes attached to microtubules with cytosol induced the formation of blobs, which moved away from the GA. However, the independence of TGN-to-PM trafficking from MTs has been clearly demonstrated in many cell types (for review, see Johnson et al. 1996). When the GA is completely fragmented, the speed of Golgi-to-PM transport is almost the same (Hirschberg et al. 1998). GPCs can also form when microtubules have been destroyed by nocodazole treatment, and their movement from the GA is not detectable (Trucco et al. 2004). In contrast, when the GA forms a ribbon, the delivery of GPCs to the cell periphery needs the function of microtubules and kinesin. Indeed, blockade of kinesin function by microinjection of an inhibitory antibody (Kreitzer et al. 2000) or expression of headless mutant (Nakata and Hirokawa 2003) prevent GPC formation from the GA. Thus, it is likely that MT- and kinesin-mediated movements of GPCs are needed for the delivery of the cargo to the correct surface domain of the cell, because in the absence of the MTs, the correct targeting of cell surface proteins is strongly impaired (Kreitzer et al. 2000). In cell-free systems, kinesin added to Golgi membranes or liposomes together with MTs induced the formation of tubule-like membranes that are similar to GPC precursors (Roux et al. 2002), while a block in kinesin function by microinjection of an inhibitory antibody (Kreitzer et al. 2000) or expression of the headless kinesin mutant (Nakata and Hirokawa 2003) prevented GPC formation from the GA. Kinesin has been seen to be associated with the tip of GPC precursors, although it can also attach to other points along the GPC precursor membrane (Polishchuk et al. 2003). Different members of the kinesin superfamily (Kamal et al. 2000; Nakata and Hirokawa 2003; Teng et al. 2005), and also the other microtubule motor dynein (Tai et al. 1999), have been shown to drive post-Golgi transport of specific cargo to various destinations. In the same cells, different GPCs filled with different cargoes can be delivered by different kinesins. KIF5 has been

382

*

R. S. Polishchuk et al.

shown to carry VSVG-containing GPCs to the axon, while KIF17A provided Kv2.1 ion channel delivery to the dendrites (Nakata and Hirokawa 2003). A number of neuronal proteins, such as bAPP, GAP43 and Vamp-2, require KIF5 for their correct targeting (Nakata and Hirokawa 2003). Finally, cytoplasmic dynein and myosin I, but not kinesin, have been found on the small vesicular membranes containing TGN38, a well-known TGN marker, and isolated from intestinal epithelial cells (Fath et al. 1994). Selection of specific cargoes by motors may be driven by various mechanisms, such as the recruitment of kinesin to GPCs. The simplest way would through direct interactions of motor proteins with a cargo (Tai et al. 1999; Kamal et al. 2000; Teng et al. 2005) or with components of the sorting machinery at the TGN (Nakagawa et al. 2000). Alternatively, adaptor proteins may serve as bridges between motors and cargo. For example, KIF13A transports the mannose-6-phosphate receptor through an interaction with the AP-1 complex (Nakagawa et al. 2000). Finally both motor and cargo can be associated with the same specific lipid microdomain, as are, for instance, KIFC3 and apically targeted annexin XIIIb. Another mechanism could be the use of a piece of the TGN with kinesin already attached. Precursors of the GPCs can be enveloped by the piece of the TGN. This gives the cargo domain the ability to use kinesin attached to the TGN as a motor for the delivery of GPCs towards the PM. Thus, post-Golgi transport can occur without MT-based delivery. However, this mechanism helps when MTs function normally.

Fusion with the PM The delivery of GPCs to the PM represents another separate and very important step. In contrast to ER-to-Golgi transit, transport from the TGN to the PM is rather standard and takes some 10 min (for review, see Musch et al. 1996). In general, most GPCs are transported away from the GA towards the periphery of the cell until they are to fuse with (Palazzo et al. 2001). GPCs that arrive at the PM represent pieces of tubular network or exhibit a rather simple vacuole-like morphology (Polishchuk et al. 2000). It appears (see detailed analysis in Chapter 3.10) that GPCs being delivered to the basolateral PM need to stop, in order to interact with endosomes according to the kiss-and-run mechanism described in Chapter 3.2. Many GPCs near the PM have been seen to be immobile for several minutes. While some eventually fused to the PM, the majority detached and moved back to the cell interior, suggesting that docking may be reversible. When GPCs are approaching the PM, they usually pause for 45 s before fusion. On average, the pre-fusion tethering phase (movement <2 mm) was 39  33 s. Tethering or docking for extensive periods of time (>5 min) is common to many GPCs, and it often precedes fusion, and is reversible (Toomre et al. 2000).

Golgi-to-PM transport

*

383

After delivery of GPCs to the PM, the direct fusion of GPC and PM occurs (Polishchuk et al. 2000; White et al. 2001). Although most GPCs fused completely, GPCs have also been seen to be capable of partial fusion, and in particular at the tips of larger tubulo-vesicular GPCs (Toomre et al. 2000). The fusion events are not homogenously distributed at the PM. Only a few fusion events occur in the very outer cell periphery, a region that often contains few microtubules in PtK2 cells. A few GPCs appear to move or tentatively fuse in the peripheral microtubule-poor, actin-rich lamella of PtK2 cells (Toomre et al. 1999). Delivery of the GPCs is not uniform and frequently it is directed to rapidly growing membrane surfaces. In motile cells, this process is restricted to the leading edge (Schmoranzer et al. 2003; Polishchuk et al. 2004). Such hot spots, where GPCs “prefer” to fuse, have been detected in PTK cells grown even under steady-state conditions (Keller et al. 2001). Delivery of membranes to the leading edge is important for cell locomotion. For instance, treatment with brefeldin A inhibits fibroblast motility (Bershadsky and Futerman 1994). Inhibition of PKD on the TGN has been seen to inhibit directed cell motility and the retrograde flow of surface markers and filamentous actin; inhibition of PKD elsewhere (in cells expressing the non-TGN-targeted PKD-kd-P155G) in the cell neither blocks anterograde membrane transport nor cell motile functions. Exogenous activation of Rac1 in PKD-kd-expressing cells restored lamellipodial dynamics independent of membrane trafficking. However, lamellipodial activity was delocalized from a single leading edge, and directed cell motility was not fully recovered (Prigozhina and Waterman-Storer 2006). Thus, delivery of GPCs occurs mostly to hot spots on the PM.

Conclusion Golgi exit sites are specialized for the exit of different cargoes. Membrane cargoes destined to the basolateral PM exit from the last two COPI-positive cisternae of the stack, whereas secretory cargo exit from the GA from both the last two COPI-positive cisternae (procollagen I) and from extended domains of the free TGN (regulatory secretory proteins [RSPs] and albuminlike cargo). In contrast, cargo destined to the endosomes, lysosomes and the APM use other domains of the free TGN for the formation of precursors of post-Golgi carriers. A scheme of Golgi-to-PM transport is presented in Fig. 1. Analysis of post-Golgi trafficking towards the PM has provide evidence in favour of the carrier maturation scheme of exit from the GA. Additionally, the Golgi-to-PM transport of both constitutive secretory proteins and regulatory secretory proteins occurs through endosomes, according to the kiss-and-run

384

*

R. S. Polishchuk et al.

Figure 1. Structure of the Golgi exit sites (GES) and mechanisms of post-Golgi transport Scheme. At the GES, the formation of two types of Golgi (G)-to-basolateral PM (BLPM) carriers takes place. One (dash arrow 1) is for membrane cargoes such as VSVG (arrowheads in the lumen). The other one (solid arrow 6) is for aggregates of luminal cargoes, such as procollagen I (straight lines in the lumen). These carriers are formed from the last two (trans) COPI-positive cisterna of the stack and they can move towards the basolateral PM alone or by using a piece (shown as oval structures attached to the carrier) of the attached TGN (AT) as a vehicle. Additionally, Golgi-to-basolateral PM carriers are formed from the TGN for soluble cargoes, such as albumin and a1-antitrypsin (solid arrow 2) and secretory granules (SGs, dash arrow 3). The Golgi-to-apical PM (solid arrow 4) and Golgi-to-endosome carriers (dash arrow 5) containing clathrin (arrowheads)-coated buds are derived from the TGN as well. The trans ER is attached to AT. Movement from the GA towards the basolateral PM involves interactions with endosomes (E) in order to exchange their sets of SNAREs into another one suitable for fusion with the PM and fission (double arrows of tubules (dash lines) connecting GPCs with cisternae. The apical PM (solid line) is separated form the basolateral PM (dash line) by tight junctions (TJ). The ER (the structure with grey content) marked by ribosomes (pictured as black dots) close to the cis-side of the Golgi complex contains the ER exit site (ERES).

mechanism (see Chapter 3.10). Now, as the structure and function, and many of the molecular mechanisms involved into post-Golgi transport became less enigmatic, a lot of work is still necessary to provide the full picture of the function of Golgi-to-PM transport.

Abbreviations GA GES GPC MT PM

Golgi apparatus Golgi exit site Golgi-to-PM carrier microtubule plasma membrane

Golgi-to-PM transport

*

385

References Allan V, Vale R (1994) Movement of membrane tubules along microtubules in vitro: evidence for specialized sites of motor attachment. J Cell Sci 107(Pt 7): 1885–1897 Ang AL, Folsch H, Koivisto UM, Pypaert M, Mellman I (2003) The Rab8 GTPase selectively regulates AP-1B-dependent basolateral transport in polarized Madin-Darby canine kidney cells. J Cell Biol 163: 339–350 €a €tta € J, Ka €a €ria €inen L, Kuismanen E (2001) Inhibition of the membrane Band AM, Ma fusion machinery prevents exit from the TGN and proteolytic processing by furin. FEBS Lett 505(1): 118–124 Bard F, Malhotra V (2006) The formation of TGN-to-plasma-membrane transport carriers. Annu Rev Cell Dev Biol 22: 439–455 Bershadsky AD, Futerman AH (1994) Disruption of the Golgi apparatus by brefeldin A blocks cell polarization and inhibits directed cell migration. Proc Natl Acad Sci USA 91: 5686–5689 Cao H, Thompson HM, Krueger EW, McNiven MA (2000) Disruption of Golgi structure and function in mammalian cells expressing a mutant dynamin. J Cell Sci 113(Pt 11): 1993–2002 Cooper MS, Cornell-Bell AH, Chernjavsky A, Dani JW, Smith SJ (1990) Tubulovesicular processes emerge from trans-Golgi cisternae, extend along microtubules, and interlink adjacent trans-Golgi elements into a reticulum. Cell 61(1): 135–145 Corda D, Colanzi A, Luini A (2006) The multiple activities of CtBP/BARS proteins: the Golgi view. Trends Cell Biol 16(3): 167–173 De Curtis I, Simons K (1989) Isolation of exocytis vesicles from BHK cells. Cell 58: 719–727 De Matteis MA, Luini A (2008) Exiting the Golgi complex. Nat Rev Mol Cell Biol 9(4): 273–284 Fath KR, Trimbur GM, Burgess DR (1994) Molecular motors are differentially distributed on Golgi membranes from polarized epithelial cells. J Cell Biol 126 (3): 661–675 Godi A, Pertile P, Meyers R, Marra P, Di Tullio G, Iurisci C, Luini A, Corda D, De Matteis MA (1999) ARF mediates recruitment of PtdIns-4-OH kinase-beta and stimulates synthesis of PtdIns(4,5)P2 on the Golgi complex. Nat Cell Biol 1(5): 280–287 Griffiths G, Fuller SD, Back R, Hollinshead M, Pfeiffer S, Simons K (1989) The dynamic nature of the Golgi complex. J Cell Biol 108: 277–297 Hirschberg K, Miller CM, Ellenberg J, Presley JF, Siggia ED, Phair RB, Lippincott-Schwartz J (1998) Kinetic analysis of secretory protein traffic and characterization of Golgi to plasma membrane transport in living cells. J Cell Biol 143: 1485–1503 Ho HC, Tang CY, Suarez SS (1999) Three-dimensional structure of the Golgi apparatus in mouse spermatids: a scanning electron microscopic study. Anat Rec 256(2): 189–194 Johnson KJ, Hall ES, Boekelheide K (1996) Kinesin localizes to the trans-Golgi network regardless of microtubule organization. Eur J Cell Biol 69(3): 276–287 Kamal A, Stokin GB, Yang Z, Xia CH, Goldstein LS (2000) Axonal transport of amyloid precursor protein is mediated by direct binding to the kinesin light chain subunit of kinesin-I. Neuron 28(2): 449–459 Keller P, Toomre D, Díaz E, White J, Simons K (2001) Multicolour imaging of post-Golgi sorting and trafficking in live cells. Nat Cell Biol 3(2): 140–149 Kreitzer G, Marmorstein A, Okamoto P, Vallee R, Rodriguez-Boulan E (2000) Kinesin and dynamin are required for post-Golgi transport of a plasma-membrane protein. Nat Cell Biol 2(2): 125–127 Kreitzer G, Schmoranzer J, Low SH, Li X, Gan Y, Weimbs T, Simon SM, Rodriguez-Boulan E (2003) Three-dimensional analysis of post-Golgi carrier exocytosis in epithelial cells. Nat Cell Biol 5(2): 126–136

386

*

R. S. Polishchuk et al.

Ktistakis NT, Brown HA, Waters MG, Sternweis PC, Roth MG (1996) Evidence that phospholipase D mediates ADP ribosylation factor-dependent formation of Golgi coated vesicles. J Cell Biol 134: 295–306 Ladinsky MS, Wu CC, McIntosh S, McIntosh JR, Howell K (2002) Structure of the Golgi and distribution of reporter molecules at 20 C reveals the complexity of the exit compartments. Mol Biol Cell 13: 2810–2825 Liljedahl M, Maeda Y, Colanzi A, Ayala I, Van Lint J, Malhotra V (2001) Protein kinase D regulates the fission of cell surface destined transport carriers from the trans-Golgi network. Cell 104(3): 409–420 Lin CC, Love HD, Gushue JN, Bergeron JJ, Ostermann J (1999) ER/Golgi intermediates acquire Golgi enzymes by brefeldin A-sensitive retrograde transport in vitro. J Cell Biol 147: 1457–1472 Matlin KS, Simons K (1983) Reduced temperature prevents transfer of a glycosylation. Cell 34: 233–243 Mironov AA, Beznoussenko GV, Nicoziani P, Martella O, Trucco A, Kweon HS, Di Giandomenico D, Polishchuk RS, Fusella A, Lupetti P, Berger EG, Geerts WJ, Koster AJ, Burger KN, Luini A (2001) Small cargo proteins and large aggregates can traverse the Golgi by a common mechanism without leaving the lumen of cisternae. J Cell Biol 155: 1225–1238 Mitra K, Ubarretxena-Belandia I, Taguchi T, Warren G, Engelman DM (2004) Modulation of the bilayer thickness of exocytic pathway membranes by membrane proteins rather than cholesterol. Proc Natl Acad Sci USA 101(12): 4083–4088 € bius W, Ohno-Iwashita Y, Van Donselaar EG, Oorschot VM, Shimada Y, Fujimoto T, Mo Heijnen HF, Geuze HJ, Slot JW (2002) Immunoelectron microscopic localization of cholesterol using biotinylated and non-cytolytic perfringolysin O. J Histochem Cytochem 50(1): 43–55 € sch A, Xu H, Shields D, Rodriguez-Boulan E (1996) Transport of vesicular stomatitis Mu virus G protein to the cell surface is signal mediated in polarized and nonpolarized cells. J Cell Biol 133(3): 543–558 Nakagawa T, Setou M, Seog D, Ogasawara K, Dohmae N, Takio K, Hirokawa N (2000) A novel motor, KIF13A, transports mannose-6-phosphate receptor to plasma membrane through direct interaction with AP-1 complex. Cell 103: 569–581 Nakata T, Terada S, Hirokawa N (1998) Visualization of the dynamics of synaptic vesicle and plasma membrane proteins in living axons. J Cell Biol 140: 659–674 Nakata T, Hirokawa N (2003) Microtubules provide directional cues for polarized axonal transport through interaction with kinesin motor head. J Cell Biol 162(6): 1045–1055 Palazzo AF, Joseph HL, Chen YJ, Dujardin DL, Alberts AS, Pfister KK, Vallee RB, Gundersen GG (2001) Cdc42, dynein, and dynactin regulate MTOC reorientation independent of Rho-regulated microtubule stabilization. Curr Biol 11(19): 1536–1541 Pimplikar SW, Ikonen E, Simons K (1994) Basolateral protein transport in streptolysin O-permeabilized MDCK cells. J Cell Biol 125(5): 1025–1035 Polishchuk EV, Di Pentima A, Luini A, Polishchuk RS (2003) Mechanism of constitutive export from the Golgi: bulk flow via the formation, protrusion, and en bloc cleavage of large trans-Golgi network tubular domains. Mol Biol Cell 14: 4470–4485 Polishchuk R, Di Pentima A, Lippincott-Schwartz J (2004) Delivery of raft-associated, GPI-anchored proteins to the apical surface of polarized MDCK cells by a transcytotic pathway. Nat Cell Biol 6(4): 297–307 Polishchuk RS, Polishchuk EV, Marra P, Alberti S, Buccione R, Luini A, Mironov AA (2000) Correlative light-electron microscopy reveals the tubular–saccular ultrastructure of carriers operating between Golgi apparatus and plasma membrane. J Cell Biol 148(1): 45–58 Prigozhina NL, Waterman-Storer CM (2006) Decreased polarity and increased random motility in PtK1 epithelial cells correlate with inhibition of endosomal recycling. J Cell Sci 119(Pt 17): 3571–3582

Golgi-to-PM transport

*

387

Puertollano R, Van der Wel NN, Greene LE, Eisenberg E, Peters PJ, Bonifacino JS (2003) Morphology and dynamics of clathrin/GGA1-coated carriers budding from the trans-Golgi network. Mol Biol Cell 14(4): 1545–1557 Rothman JE, Wieland FT (1996) Protein sorting by transport vesicles. Science 272(5259): 227–234 Roux A, Cappello G, Cartaud J, Prost J, Goud B, Bassereau P (2002) A minimal system allowing tubulation with molecular motors pulling on giant liposomes. Proc Natl Acad Sci USA 99(8): 5394–5399 Roux A, Cuvelier D, Nassoy P, Prost J, Bassereau P, Goud B (2005) Role of curvature and phase transition in lipid sorting and fission of membrane tubules. EMBO J 24(8): 1537–1545 Roux A, Uyhazi K, Frost A, De Camilli P (2006) GTP-dependent twisting of dynamin implicates constriction and tension in membrane fission. Nature 441(7092): 528–531 Schmoranzer J, Kreitzer G, Simon SM (2003) Migrating fibroblasts perform polarized, microtubule-dependent exocytosis towards the leading edge. J Cell Sci 116: 4513–4519 Schuck S, Simons K (2004) Polarized sorting in epithelial cells: raft clustering and the biogenesis of the apical membrane. J Cell Sci 117(Pt 25): 5955–5964 Simon JP, Morimoto T, Bankaitis VA, Gottlieb TA, Ivanov IE, Adesnik M, Sabatini DD (1998) An essential role for the phosphatidylinositol transfer protein in the scission of coatomer-coated vesicles from the trans-Golgi network. Proc Natl Acad Sci USA 95(19): 11181–11186 Tai AW, Chuang JZ, Bode C, Wolfrum U, Sung CH (1999) Rhodopsin’s carboxy-terminal cytoplasmic tail acts as a membrane receptor for cytoplasmic dynein by binding to the dynein light chain Tctex-1. Cell 97(7): 877–887 Teng J, Rai T, Tanaka Y, Takei Y, Nakata T, Hirasawa M, Kulkarni AB, Hirokawa N (2005) The KIF3 motor transports N-cadherin and organizes the developing neuroepithelium. Nat Cell Biol 7: 474–482 Toomre D, Keller P, White J, Olivo JC, Simons K (1999) Dual-color visualization of transGolgi network to plasma membrane traffic along microtubules in living cells. J Cell Sci 112: 21–33 Toomre D, Steyer JA, Keller P, Almers W, Simons K (2000) Fusion of constitutive membrane traffic with the cell surface observed by evanescent wave microscopy. J Cell Biol 149(1): 33–40 Trucco A, Polishchuk RS, Martella O, Di Pentima A, Fusella A, Di Giandomenico D, San Pietro E, Beznoussenko GV, Polishchuk EV, Baldassarre M, Buccione R, Geerts WJ, Koster AJ, Burger KN, Mironov AA, Luini A (2004) Secretory traffic triggers the formation of tubular continuities across Golgi sub-compartments. Nat Cell Biol 6(11): 1071–1081 € mer A, Migala A, Almers W, Gerdes HH (1997) MicrotubuleWacker I, Kaether C, Kro dependent transport of secretory vesicles visualized in real time with a GFP-tagged secretory protein. J Cell Sci 110(Pt 13): 1453–1463 Wasmeier C, Hutton JC (1996) Molecular cloning of phogrin, a protein-tyrosine phosphatase homologue localized to insulin secretory granule membranes. J Biol Chem 271: 18161–18170 Weigert R, Silletta MG, Spanò S, Turacchio G, Cericola C, Colanzi A, Mancini R, Polishchuk EV, Salmona M, Facchiano F, Burger KNJ, Mironov A, Luini A, Corda D (1999) CtBP/ BARS induces fission of Golgi membranes by acylating lysophosphatidic acid. Nature 402: 429–433 White J, Keller P, Stelzer EH (2001) Spatial partitioning of secretory cargo from Golgiresident proteins in live cells. BMC Cell Biol 2: 19

388

*

G. A. Mardones et al.

Protein transport from the trans-Golgi network to endosomes Gonzalo A. Mardones, Roman S. Polishchuk and Juan S. Bonifacino

After traversing the endoplasmic reticulum (ER), the ER–Golgi intermediate compartment (ERGIC) and the cisternae of the Golgi complex, newly synthesized proteins reach the trans-Golgi network (TGN), from where they can be sorted to several destinations: the extracellular space, different domains of the plasma membrane, regulated secretory granules and lysosome-related organelles. In addition, various types of protein have been shown to undergo direct transport from the TGN to endosomes. These include a subpopulation of lysosomal membrane proteins (Harter and Mellman 1992), endocytic receptors (Futter et al. 1995; Leitinger et al. 1995), processing peptidases (Molloy et al. 1999), and intracellular sorting receptors with their cargos (Le Borgne and Hoflack 1998; Ghosh et al. 2003). The latter are by far the best characterized and will be the main subject of this chapter.

Biosynthetic sorting of acid hydrolases by mannose 6-phosphate receptors Newly-synthesized acid hydrolase precursors are core-glycosylated on specific asparagine residues by en-bloc transfer of preformed high-mannose oligosaccharide chains in the ER (Kornfeld and Mellman 1989). Upon transport to the Golgi complex, specific features on the surface of the hydrolase precursors direct modification of one or more mannose residues to mannose 6-phosphate (M6P) by a two-step reaction (Lazzarino and Gabel 1988). M6P moieties serve as ligands for two transmembrane mannose 6-phosphate receptors (MPR), a cation-dependent MPR (CD-MPR) and a cation-independent MPR (CI-MPR) (Fig. 1). The hydrolase–receptor complexes become concentrated within areas of the TGN that are coated with clathrin and various clathrinassociated proteins. Among the latter are adaptor proteins, which (i) are recruited to membranes by interaction with the small GTP-binding protein, Arf1, and the phosphoinositide, PI4P, (ii) interact with signals present within the cytosolic tails of the MPRs, (iii) bind to clathrin and promote its polymerization to form the outer layer of a protein coat, and (iv) recruit accessory proteins that regulate various aspects of coat function. This eventually leads to the formation of clathrin-coated vesicles (CCVs) or pleiomorphic transport carriers (PTCs) (see below) that deliver the hydrolase–receptor complexes to endosomes. The acid pH of the endosomal lumen induces the dissociation of the hydrolase precursors from the MPRs. The released hydrolase precursors are then transported to lysosomes with the fluid phase, and undergo

Protein transport from the trans-Golgi network to endosomes

*

389

Figure 1. Schematic representation of the trafficking of MPRs. Mannose 6-phosphate residues on acid hydrolase precursors allow binding to two mannose 6-phosphate receptors (MPR), the cation-dependent MPR (CD-MPR) and the cation-independent MPR (CI-MPR). The hydrolase– receptor complexes concentrate within areas of the TGN coated with clathrin and its adaptors (GGAs and AP-1), leading to formation of clathrin-coated vesicles (CCV) or pleiomorphic transport carriers (PTC) that deliver their content to endosomes. Hydrolase precursors dissociate from the MPRs at the acid pH of the endosomal lumen (H þ ), and are transported with the fluid phase to lysosomes, where they are exposed to a more acidic pH and undergo proteolytic maturation. The unoccupied MPRs are recycled through the tubular endosomal network (TEN) to the TGN for further rounds of hydrolase sorting.

proteolytic processing to the mature, active forms. The MPRs, on the other hand, are retrieved to the TGN to mediate further rounds of acid hydrolase sorting (Fig. 1).

Adaptors and signals involved in sorting MPRs from the TGN to endosomes Two types of clathrin adaptor, the heterotetrameric adaptor protein complex 1 (AP-1) and the monomeric Golgi-localized, gamma-ear containing, ADP ribosylation factor binding proteins (GGAs), have been implicated in MPR sorting from the TGN to endosomes (Glickman et al. 1989; Puertollano et al. 2001; Doray et al. 2002b) (Fig. 2). AP-1 comprises four subunits, b1, g, m1, and s1,

390

*

G. A. Mardones et al.

Figure 2. Coat proteins, accessory proteins and binding motifs involved in protein sorting from the TGN to endosomes. A clathrin triskelion is formed by three clathrin light chains (CLC) and three clathrin heavy chains (CHC) having terminal domains (TD) that bind to the “hinge” segments of either the b1 subunit of AP-1 or the GGAs. AP-1 comprises four subunits, b1, g, m1, and s1. The N-terminal portions of b1 and g, and the entire m1 and s1 subunits form a “core” domain. The C-terminal portions of b1 and g end in globular “ear” domains. The m1 subunit binds to the YXXØ motif in the MPRs (Y is tyrosine, X is any amino acid, and Ø is an amino acid with a bulky hydrophobic side chain; key residues highlighted in blue). The GGAs comprise VHS, GAT and GAE domains. The VHS domain binds a DXXLL motif (D is aspartate, X is any amino acid, and L is leucine; key residues highlighted in blue) in the MPRs, and in the indicated proteins. Serine residues that are potential phosphorylation sites near the DXXLL motif, to increase the affinity for the interaction with the VHS domain, are shown in red. The VHS domain engages in auto-inhibitory binding to phosphorylated serine residues upstream of internal DXXLL motifs within the hinge domains of GGAs. The position of the transmembrane domain (Tm) and the number of residues before and after the signals is indicated. CD-MPR, cation-dependent mannose 6-phosphate receptor; CI-MPR, cation-independent mannose 6-phosphate receptor; LRP3, low-density-lipoprotein-receptor-related protein 3; SorLA, sorting-protein-related receptor containing low-density-lipoprotein-receptor class A repeats. In brackets are examples of GAE-binding sequences that have been inferred from interaction and crystallographic analyses. Residues that conform to the CG[PDE][CLM] tetrapeptide motif (where C is an aromatic residue, G is glycine, P is proline, D is aspartate, E is glutamate, L is leucine, and M is methionine) are indicated in green. Underlined residues do not match the canonical motif. The numbering corresponds to the amino acid sequence of the human proteins.

the latter three of which occur as two (A, B) or three (A–C) isoforms. The Nterminal portions of b1 and g, plus the entire m1 and s1 subunits, assemble into a folded “core” domain (Heldwein et al. 2004), whereas the C-terminal portions

Protein transport from the trans-Golgi network to endosomes

*

391

of b1 and g extend as largely disordered “hinge” segments that end in globular “ear” (or “appendage”) domains (Nogi et al. 2002). The GGAs (i.e., GGA1, GGA2, and GGA3 in humans) are single chain polypeptides comprising folded VHS, GAT, and GAE domains that are linked by largely disordered segments. The link connecting the GAT and GAE domains is quite long and is also referred to as “hinge” (Boman et al. 2000; Dell’Angelica et al. 2000; Hirst et al. 2000). The MPRs use several signals in their cytosolic tails for sorting from the TGN to endosomes (Bonifacino and Traub 2003; Ghosh et al. 2003) (Fig. 2). One signal conforms to the YXXØ consensus motif (Y is tyrosine, X is any amino acid, and Ø is an amino acid with a bulky hydrophobic side chain) and binds to the m1 subunit of the AP-1 complex (Johnson and Kornfeld 1992; Ohno et al. 1995; Honing et al. 1997; Ghosh and Kornfeld 2004). Another signal fits the consensus DXXLL (D is aspartate and L is leucine) and binds to the VHS domain of the GGAs (Puertollano et al. 2001; Takatsu et al. 2001; Zhu et al. 2001). The crystal structure of m1 in complex with a YXXØ signal has not yet been solved, although it is expected to resemble that of the homologous m2–YXXØ complex (Owen and Evans 1998). On the other hand, the crystal structures of the VHS domains of GGA1 and GGA3 in complexes with DXXLL signals from the CD- and CI-MPRs have been solved (Misra et al. 2002; Shiba et al. 2002a). The VHS domains of both proteins are right-handed superhelices of eight a-helices. Both the CDand CI-MPR peptides bind in an extended conformation to a groove formed by helices 6 and 8. The critical D and LL residues bind to an electropositive pocket and two shallow hydrophobic pockets, respectively, whereas the X residues and flanking residues point away from the VHS domain or are disordered. Other transmembrane proteins such as Sortilin, b-secretase, and SorLA also have DXXLL-type signals in their cytosolic tails and are believed to be sorted in a similar manner from the TGN to endosomes (Nielsen et al. 2001; Takatsu et al. 2001; Jacobsen et al. 2002; He et al. 2003) (Fig. 2).

Regulation of TGN sorting by adaptor phosphorylation The sorting efficiency of the MPRs seems to be regulated by phosphorylation of serine residues near the DXXLL signal. These serine residues are within a context that fits the consensus motif for casein kinase II (CKII) substrates, and are indeed phosphorylated by CKII both in vivo and in vitro (Meresse et al. 1990; Mauxion et al. 1996). X-ray crystallography has shown that phosphorylation of a serine residue upstream of the CI-MPR DXXLL signal allows electrostatic interactions between one of the negatively charged phosphate oxygen atoms and two positively charged residues on the VHS domain, increasing the binding affinity (Kato et al. 2002). A similar regulation is likely to occur for b-secretase (Shiba et al. 2004). Another form of regulation seems to result from auto-inhibitory binding of the VHS domain to CKII-phosphorylated serine residues upstream of internal DXXLL motifs within the hinge domains of GGA1 and GGA3 (Doray et al. 2002a) (Fig. 2). This indicates that GGA1 and GGA3 may become activated by a dephosphorylation event that

392

*

G. A. Mardones et al.

displaces its own DXXLL ligand and frees the VHS domain for interaction with signals from the cytosolic tails of the MPRs (Ghosh and Kornfeld 2003). GGA3 also undergoes phosphorylation on two other serine residues in the hinge region that regulate its conformation and association with membranes (Kametaka et al. 2005).

Functions of different TGN adaptor domains Despite their different quaternary structures, both the core domain of AP-1 and the VHS–GAT domains of the GGAs participate in membrane recruitment and cargo recognition. The hinge domains of both types of adaptor bind clathrin. Finally, the ear domain of AP-1-g and the GAE domains of the GGAs are homologous, eight-stranded immunoglobulin-like b-sandwich domains that bind sequences conforming to the CG[PDE][CLM] motif (C is an aromatic residue, G is glycine, P is proline, D is aspartate, E is glutamate, L is leucine, and M is methionine) or related motifs on accessory proteins (Fig. 2). Among the proteins that are known to bind GAE and/or ear domains through such motifs are Rabaptin-5, g-synergin, NECAP-1 and NECAP-2, Aftiphilin, gBAR, Clint/Enthoprotin/EpsinR, p56 and GAK (Page et al. 1999; Hirst et al. 2000; Takatsu et al. 2000; Doray and Kornfeld 2001; Kalthoff et al. 2002; Shiba et al. 2002b; Wasiak et al. 2002; Collins et al. 2003; Hirst et al. 2003; Lui et al. 2003; Mattera et al. 2003; Mills et al. 2003; Ritter et al. 2003; Mattera et al. 2004; Neubrand et al. 2005; Kametaka et al. 2007). These proteins are thought to regulate coat function, vesicle formation or fusion with endosomes, or to be adaptors for other cargos.

Cooperation of AP-1 and GGAs in MPR sorting at the TGN It is not clear why two different clathrin adaptors, AP-1 and the GGAs, are involved in MPR sorting at the TGN. Two possibilities are that they function (i) in parallel, packaging MPRs into different carriers, and (ii) cooperatively, leading to MPR sorting into the same carriers. A recent model has expanded on this latter model by proposing that AP-1 and the GGAs act sequentially (Ghosh and Kornfeld 2003). This model postulates that GGAs bind the MPRs and facilitate their entry into forming AP-1-containing CCVs. The bases for this model are EM data showing co-localization of GGAs and AP-1 in clathrincoated buds and vesicles at the TGN, and binding studies showing direct interaction between the hinge domain of the GGAs and the ear domain of AP-1-g. According to this model, the vectoriality of the process is ensured by dephosphorylation/phosphorylation events. As mentioned above, the cytosolic pool of GGAs exists in a “closed”, auto-inhibitory state induced by phosphorylation. After recruitment of the GGAs to the TGN by Arf1-GTP, a protein phosphatase activity associated with the cytosolic tail of MPRs dephosphorylates the GGAs, relieving auto-inhibition and allowing the “open” state of the GGAs to interact with the DXXLL signals on the cytosolic tail of MPRs.

Protein transport from the trans-Golgi network to endosomes

*

393

Within forming CCVs, the hinge domain of the GGAs interacts with the ear domain of AP-1-g. An AP-1-associated CKII then phosphorylates the cytosolic tail of MPRs and GGAs, restoring the GGAs “closed”, auto-inhibitory state. Additional phosphorylation of the m1 subunit of AP-1 generates high avidity binding for phosphorylated MPRs, leading to the transfer of MPRs from GGAs to AP-1. Finally, the MPRs are incorporated into CCVs by AP-1, and the autoinhibited, phosphorylated GGAs dissociate from the ear domain of AP-1-g and are released to the cytosol. This model, however, does not explain how the GGAs remain largely associated to CCVs and PTCs en route to endosomes, as seen in live-cell imaging studies (see below).

Nature of TGN-to-endosome transport carriers Early electron microscopy (EM) studies showed the presence of MPRs and their cargo enzymes within clathrin-coated areas of the TGN, as well as in a population of small (60–100 nm diameter), spherical CCVs found in the vicinity of the TGN (Griffiths et al. 1988; Klumperman et al. 1993; Le Borgne et al. 1996). In addition, MPRs were detected biochemically in CCVs purified by subcellular fractionation (Campbell et al. 1983; Sahagian and Steer 1985). This indicates that typical CCVs, morphologically similar to those that mediate receptor-mediated endocytosis from the plasma membrane, participate in MPR transport from the TGN to endosomes (Fig. 1). However, recent live-cell imaging experiments using proteins tagged with variants of the green fluorescent protein (GFP) have revealed diversity in the types of carrier that deliver MPRs to endosomes. Confocal laser scanning microscopy (CLSM) of cells expressing cyan fluorescent protein (CFP)-CI-MPR and CD-MPR-CFP showed that these proteins bud from the TGN on PTCs of varying sizes and shapes (Puertollano et al. 2003; Waguri et al. 2003; Polishchuk et al. 2006) (Fig. 3). Budding is regulated by Arf1 and results in segregation of exiting MPRs from secretory proteins (CFP–VSVG protein), plasma membrane receptors (Rhodamine-transferrin bound to its receptor), and resident Golgi proteins (CFP–galactosyl transferase) (Puertollano et al. 2003; Waguri et al. 2003). These PTCs also contain associated AP-1-YFP, YFP-GGA1 and YFP-clathrin (Huang et al. 2001; Puertollano et al. 2003; Waguri et al. 2003). Before losing their coats, some PTCs undergo long-range translocation toward the peripheral cytoplasm for distances of 10–20 mm, with average speeds of 1 mm/s, and in a microtubule-dependent fashion. Eventually the PTCs merge with tubular–vesicular endosomal elements containing endocytosed transferrin (CFP-CI-MPR) and albumin (CD-MPR-CFP) (Puertollano et al. 2003; Waguri et al. 2003). Until recently the ultrastructure of these PTCs remained a mystery, mostly due to the dynamic nature of these organelles. Indeed, they move rapidly out of the Golgi complex and fuse with endosomes. To overcome this limitation, correlative light-EM (CLEM) was employed to reveal the ultrastructure of the long-range PTCs moving from the TGN to peripheral endosomes (Polishchuk

394

*

G. A. Mardones et al.

Figure 3. Ultrastructure of TGN-to-endosome transport carriers. (A) HeLa cells were transfected with a plasmid encoding GFP–GGA1, incubated for 30 min with TRITC-dextran, and imaged by confocal fluorescence microscopy. Sequence of time-lapse frames shows three PTCs (arrowheads) that move from the juxtanuclear area in a living cell that was quickly fixed during the course of observation. (B) After fixation all three PTCs (indicated by arrows and arrowhead in inset) were found again within the same area of the cell outlined by box. Note the apparent pleiomorphism of the GFP-GGA1-containing PTCs (green) and the absence of TRITC-dextran (red) within them. (C) Immuno-EM image of a thin section reveals the same three PTCs with grape-like (arrow), saccular (open arrow) and vesicular (arrowhead) shape within the area corresponding to the box and inset in panel B. (D) Three-dimensional reconstruction of the GFPGGA1-containing PTCs indicated by the arrow and arrowhead in B and C. (E) Different classes of PTC shape are shown in schematic drawings (with membrane in black and coats in red) and EM images. Images are adapted from Polishchuk et al. (2006).

Protein transport from the trans-Golgi network to endosomes

*

395

et al. 2006). CLEM is a powerful technique that allows ultrastructural analysis of moving, fluorescently tagged organelles in vivo. The organelles are first identified in living cells using light microscopy and, after fixation, examined by EM in serial sections with further three-dimensional reconstruction (Mironov et al. 2000; Polishchuk et al. 2000). CLEM analysis of PTCs captured shortly after their exit from the TGN revealed great variability of their shapes, which could be assigned to four main classes: (i) vesicular, (ii) tubular, (iii) saccular, and (iv) grape-like (Fig. 3). Consistent with light microscopy observations in living cells (Puertollano et al. 2003), the sizes of PTCs identified in thin sections ranged from 100 nm (vesicles) to 1200 nm (all other structures) with an average diameter of 400 nm (Polishchuk et al. 2006). Notably, the use of CLEM demonstrated that some of these carriers are indeed typical CCVs, although these represent a minority (5%). In contrast the majority of PTCs (58%) consist of a few (typically 2–3) clathrin-coated buds (comprising 50% of PTC surface area) interconnected by tubular elements and, therefore, belong to the grape-like class of structures. These PTCs are similar to TGN domains that contain MPR–GFP and GFP–GGA1 and might therefore correspond to small pieces of TGN that are pulled along microtubules towards the cell periphery. The probability of forming a CCV versus a more complex PTC likely depends on how soon scission factors act on the necks of the forming PTCs after assembly of the coat. Investigation of PTCs upon their arrival to endosomes revealed that their basic ultrastructure remains mostly unchanged as they translocate from the TGN to the target membrane (Polishchuk et al. 2006). Thus, live-cell imaging combined with electron microscopy revealed a new type of clathrin-coated carriers that operate in delivery of cargo from the Golgi complex to the endosomal system.

Dynamics of TGN clathrin coats Quantitative analysis of fluorescence recovery after photobleaching (FRAP) allowed the study of the dynamics of coat recruitment from the cytosol to the TGN in live cells. In this methodology, fluorescence recovery indicates that the irreversibly photobleached fluorescent molecules exchange with unbleached pools of the protein in another part of the cell. On the other hand, lack of recovery indicates that the fluorescently tagged proteins are stably associated with the bleached region. FRAP experiments showed that AP-1, GGA1 and clathrin exchange between the cytosol and TGN membranes very rapidly, with half-times from 10 s to 15 s (Wu et al. 2003). Moreover, the cycling of adaptors was found to be faster than that of clathrin. When budding of CCVs was blocked without affecting the structure of clathrin-coated pits, as in incubation of cells at 20 C (Ladinsky et al. 2002), the exchange of AP-1, GGA1 and clathrin with their cytosolic pools continued unchanged. In contrast, under conditions that affect the structure of clathrin-coated pits, as in K þ depletion or hypertonic sucrose treatment, clathrin exchange was blocked, though not AP-1 or GGA1 exchange (Wu et al. 2003). These observations imply that clathrin-

396

*

G. A. Mardones et al.

coated domains of the TGN are dynamic and that, after its recruitment by the adaptors, exchange of clathrin is uncoupled to exchange of adaptors.

Mechanism of formation and functional advantages of PTCs There are several possible explanations for the diverse morphology of PTCs. It is known that the isolated membrane lipid constituents of the Golgi complex have the ability to tubulate (Roux et al. 2002). Different kinetics of fission while membranes elongate during the budding process may produce PTCs of different sizes, from CCVs to larger carriers. The formation of larger carriers may result from the ability of clathrin to polymerize into tubular cages, in particular when GGA1 is present (Zhang et al. 2007). Another possibility is that PTCs form by protrusion and en-bloc cleavage of TGN tubular domains (Polishchuk et al. 2003). This may result from membrane maturation/progression processes in the exocytic pathway, with stochastic fragmentation caused by the pulling force of molecular motors. The pleiomorphism of the PTCs could facilitate the packaging of different types of cargo, from small to large proteins, protein assemblies, aggregates, or lipoprotein particles. The ability of the PTCs to translocate for long distances would allow for the distribution of cargo to peripheral areas of the cells.

Possible existence of a population of retrograde PTCs Live-cell imaging of proteins involved in TGN-to-endosome transport by conventional CLSM showed mainly long-range movement of PTCs towards the periphery of the cells. Fluorescently tagged CD-MPR, CI-MPR, GGAs, the g subunit of AP-1, clathrin, and epsinR, were all observed in association with anterograde PTCs (Huang et al. 2001; Hirst et al. 2003; Puertollano et al. 2003; Waguri et al. 2003). The advent of new light microscopy technologies with better temporal and spatial resolution has allowed the observation of very fast transport processes. Spinning disk confocal microscopy (SDCM), in particular, revealed another population of GGA1-containing PTCs that move from the periphery to the center of the cell (Mardones et al. 2007). Likewise, using SDCM, retrograde PTCs containing fluorescently tagged CD-MPR, AP-1g and clathrin were also observed (Mardones GA, unpublished observations). The number and kinetic properties of anterograde and retrograde transport carriers, containing CD-MPR and coat proteins, are virtually the same. A possible explanation for the existence of these retrograde PTCs is that TGN coat proteins have an additional role in retrograde transport. This would be consistent with accumulation of MPRs in peripheral endosomes of m1Adeficient fibroblasts (Meyer et al. 2000). This later observation has in fact challenged the notion that AP-1 functions in sorting at the TGN in favor of AP-1 functioning in the retrieval of MPRs from endosomes (Meyer et al. 2000). An alternative is that GGAs and/or the other coat proteins play no role in

Protein transport from the trans-Golgi network to endosomes

*

397

retrograde transport of MPRs but remain passively associated with retrograde PTCs after their forward sorting function is fulfilled. A last possibility would be an additional function of the same set of coat proteins in the retrieval of a different type of cargo to the TGN.

Unresolved issues and perspectives Despite the progress in the elucidation of the mechanisms of TGN-to-endosome transport described above, many important issues remain to be addressed. Among these are: (i) the exact sequence of steps by which the GGAs and AP-1 effect sorting of MPRs and their cargo acid hydrolases at the TGN, (ii) the roles of most accessory proteins that bind to the GGAs and AP-1, (iii) the mechanism of scission of TGN-derived CCVs and PTCs, (iv) the exact contributions of CCVs and PTCs to TGN-to-endosome transport, (v) the role of cytoskeletal elements in budding and translocation of CCVs and PTCs, (vi) the mechanism of fusion of CCVs and PTCs with endosomes, (vii) the function of retrograde PTCs, and (viii) the possible functions of GGAs, AP-1 and their accessory proteins on endosomes and retrograde PTCs. Continued biochemical, genetic and morphological analyses of the molecular machinery and organellar carriers involved in TGN-to-endosome transport promises to resolve these issues as well as shed more light on the physiological and pathological processes mediated by this critical trafficking route. Acknowledgements. Work in the authors’ laboratories was supported by the Intramural Program of NICHD, NIH, and Telethon Italy.

References Boman AL, Zhang C, Zhu X, Kahn RA (2000) A family of ADP-ribosylation factor effectors that can alter membrane transport through the trans-Golgi. Mol Biol Cell 11: 1241–1255 Bonifacino JS, Traub LM (2003) Signals for sorting of transmembrane proteins to endosomes and lysosomes. Annu Rev Biochem 72: 395–447 Campbell CH, Fine RE, Squicciarini J, Rome LH (1983) Coated vesicles from rat liver and calf brain contain cryptic mannose 6-phosphate receptors. J Biol Chem 258: 2628–2633 Collins BM, Praefcke GJ, Robinson MS, Owen DJ (2003) Structural basis for binding of accessory proteins by the appendage domain of GGAs. Nat Struct Biol 10: 607–613 Dell’Angelica EC, Puertollano R, Mullins C, Aguilar RC, Vargas JD, Hartnell LM, Bonifacino JS (2000) GGAs: a family of ADP ribosylation factor-binding proteins related to adaptors and associated with the Golgi complex. J Cell Biol 149: 81–94 Doray B, Bruns K, Ghosh P, Kornfeld SA (2002a) Autoinhibition of the ligand-binding site of GGA1/3 VHS domains by an internal acidic cluster-dileucine motif. Proc Natl Acad Sci USA 99: 8072–8077 Doray B, Ghosh P, Griffith J, Geuze HJ, Kornfeld S (2002b) Cooperation of GGAs and AP-1 in packaging MPRs at the trans-Golgi network. Science 297: 1700–1703

398

*

G. A. Mardones et al.

Doray B, Kornfeld S (2001) Gamma subunit of the AP-1 adaptor complex binds clathrin: implications for cooperative binding in coated vesicle assembly. Mol Biol Cell 12: 1925–1935 Futter CE, Connolly CN, Cutler DF, Hopkins CR (1995) Newly synthesized transferrin receptors can be detected in the endosome before they appear on the cell surface. J Biol Chem 270: 10999–11003 Ghosh P, Dahms NM, Kornfeld S (2003) Mannose 6-phosphate receptors: new twists in the tale. Nat Rev Mol Cell Biol 4: 202–212 Ghosh P, Kornfeld S (2003) Phosphorylation-induced conformational changes regulate GGAs 1 and 3 function at the trans-Golgi network. J Biol Chem 278: 14543–14549 Ghosh P, Kornfeld S (2004) The cytoplasmic tail of the cation-independent mannose 6phosphate receptor contains four binding sites for AP-1. Arch Biochem Biophys 426: 225–230 Glickman JN, Conibear E, Pearse BM (1989) Specificity of binding of clathrin adaptors to signals on the mannose-6-phosphate/insulin-like growth factor II receptor. EMBO J 8: 1041–1047 Griffiths G, Hoflack B, Simons K, Mellman I, Kornfeld S (1988) The mannose 6-phosphate receptor and the biogenesis of lysosomes. Cell 52: 329–341 Harter C, Mellman I (1992) Transport of the lysosomal membrane glycoprotein lgp120 (lgp-A) to lysosomes does not require appearance on the plasma membrane. J Cell Biol 117: 311–325 He X, Zhu G, Koelsch G, Rodgers KK, Zhang XC, Tang J (2003) Biochemical and structural characterization of the interaction of memapsin 2 (beta-secretase) cytosolic domain with the VHS domain of GGA proteins. Biochemistry 42: 12174–12180 Heldwein EE, Macia E, Wang J, Yin HL, Kirchhausen T, Harrison SC (2004) Crystal structure of the clathrin adaptor protein 1 core. Proc Natl Acad Sci USA 101: 14108–14113 Hirst J, Lui WW, Bright NA, Totty N, Seaman MN, Robinson MS (2000) A family of proteins with gamma-adaptin and VHS domains that facilitate trafficking between the trans-Golgi network and the vacuole/lysosome. J Cell Biol 149: 67–80 Hirst J, Motley A, Harasaki K, Peak Chew SY, Robinson MS (2003) EpsinR: an ENTH domain-containing protein that interacts with AP-1. Mol Biol Cell 14: 625–641 Honing S, Sosa M, Hille-Rehfeld A, Von Figura K (1997) The 46-kDa mannose 6phosphate receptor contains multiple binding sites for clathrin adaptors. J Biol Chem 272: 19884–19890 Huang F, Nesterov A, Carter RE, Sorkin A (2001) Trafficking of yellow-fluorescentprotein-tagged mu1 subunit of clathrin adaptor AP-1 complex in living cells. Traffic 2: 345–357 Jacobsen L, Madsen P, Nielsen MS, Geraerts WP, Gliemann J, Smit AB, Petersen CM (2002) The sorLA cytoplasmic domain interacts with GGA1 and -2 and defines minimum requirements for GGA binding. FEBS Lett 511: 155–158 Johnson KF, Kornfeld S (1992) The cytoplasmic tail of the mannose 6-phosphate/insulinlike growth factor-II receptor has two signals for lysosomal enzyme sorting in the Golgi. J Cell Biol 119: 249–257 Kalthoff C, Groos S, Kohl R, Mahrhold S, Ungewickell EJ (2002) Clint: a novel clathrinbinding ENTH-domain protein at the Golgi. Mol Biol Cell 13: 4060–4073 Kametaka S, Mattera R, Bonifacino JS (2005) Epidermal growth factor-dependent phosphorylation of the GGA3 adaptor protein regulates its recruitment to membranes. Mol Cell Biol 25: 7988–8000 Kametaka S, Moriyama K, Burgos PV, Eisenberg E, Greene LE, Mattera R, Bonifacino JS (2007) Canonical interaction of cyclin G-associated kinase with adaptor protein 1 regulates lysosomal enzyme sorting. Mol Biol Cell 18: 2991–3001 Kato Y, Misra S, Puertollano R, Hurley JH, Bonifacino JS (2002) Phosphoregulation of sorting signal-VHS domain interactions by a direct electrostatic mechanism. Nat Struct Biol 9: 532–536

Protein transport from the trans-Golgi network to endosomes

*

399

Klumperman J, Hille A, Veenendaal T, Oorschot V, Stoorvogel W, Von Figura K, Geuze HJ (1993) Differences in the endosomal distributions of the two mannose 6-phosphate receptors. J Cell Biol 121: 997–1010 Kornfeld S, Mellman I (1989) The biogenesis of lysosomes. Annu Rev Cell Biol 5: 483–525 Ladinsky MS, Wu CC, McIntosh S, McIntosh JR, Howell KE (2002) Structure of the Golgi and distribution of reporter molecules at 20 C reveals the complexity of the exit compartments. Mol Biol Cell 13: 2810–2825 Lazzarino DA, Gabel CA (1988) Biosynthesis of the mannose 6-phosphate recognition marker in transport-impaired mouse lymphoma cells. Demonstration of a two-step phosphorylation. J Biol Chem 263: 10118–10126 Le Borgne R, Griffiths G, Hoflack B (1996) Mannose 6-phosphate receptors and ADPribosylation factors cooperate for high affinity interaction of the AP-1 Golgi assembly proteins with membranes. J Biol Chem 271: 2162–2170 Le Borgne R, Hoflack B (1998) Protein transport from the secretory to the endocytic pathway in mammalian cells. Biochim Biophys Acta 1404: 195–209 Leitinger B, Hille-Rehfeld A, Spiess M (1995) Biosynthetic transport of the asialoglycoprotein receptor H1 to the cell surface occurs via endosomes. Proc Natl Acad Sci USA 92: 10109–10113 Lui WW, Collins BM, Hirst J, Motley A, Millar C, Schu P, Owen DJ, Robinson MS (2003) Binding partners for the COOH-terminal appendage domains of the GGAs and gamma-adaptin. Mol Biol Cell 14: 2385–2398 Mardones GA, Burgos PV, Brooks DA, Parkinson-Lawrence E, Mattera R, Bonifacino JS (2007) The trans-Golgi network accessory protein p56 promotes long-range movement of GGA/clathrin-containing transport carriers and lysosomal enzyme sorting. Mol Biol Cell 18: 3486–3501 Mattera R, Arighi CN, Lodge R, Zerial M, Bonifacino JS (2003) Divalent interaction of the GGAs with the Rabaptin-5-Rabex-5 complex. EMBO J 22: 78–88 Mattera R, Ritter B, Sidhu SS, McPherson PS, Bonifacino JS (2004) Definition of the consensus motif recognized by gamma-adaptin ear domains. J Biol Chem 279: 8018–8028 Mauxion F, Le Borgne R, Munier-Lehmann H, Hoflack B (1996) A casein kinase II phosphorylation site in the cytoplasmic domain of the cation-dependent mannose 6-phosphate receptor determines the high affinity interaction of the AP-1 Golgi assembly proteins with membranes. J Biol Chem 271: 2171–2178 Meresse S, Ludwig T, Frank R, Hoflack B (1990) Phosphorylation of the cytoplasmic domain of the bovine cation-independent mannose 6-phosphate receptor. Serines 2421 and 2492 are the targets of a casein kinase II associated to the Golgi-derived HAI adaptor complex. J Biol Chem 265: 18833–18842 Meyer C, Zizioli D, Lausmann S, Eskelinen EL, Hamann J, Saftig P, Von Figura K, Schu P (2000) mu1A-adaptin-deficient mice: lethality, loss of AP-1 binding and rerouting of mannose 6-phosphate receptors. EMBO J 19: 2193–2203 Mills IG, Praefcke GJ, Vallis Y, Peter BJ, Olesen LE, Gallop JL, Butler PJ, Evans PR, McMahon HT (2003) EpsinR: an AP1/clathrin interacting protein involved in vesicle trafficking. J Cell Biol 160: 213–222 Mironov AA, Polishchuk RS, Luini A (2000) Visualizing membrane traffic in vivo by combined video fluorescence and 3D electron microscopy. Trends Cell Biol 10: 349–353 Misra S, Puertollano R, Kato Y, Bonifacino JS, Hurley JH (2002) Structural basis for acidic-cluster-dileucine sorting-signal recognition by VHS domains. Nature 415: 933–937 Molloy SS, Anderson ED, Jean F, Thomas G (1999) Bi-cycling the furin pathway: from TGN localization to pathogen activation and embryogenesis. Trends Cell Biol 9: 28–35

400

*

G. A. Mardones et al.

Neubrand VE, Will RD, Mobius W, Poustka A, Wiemann S, Schu P, Dotti CG, Pepperkok R, Simpson JC (2005) Gamma-BAR, a novel AP-1-interacting protein involved in postGolgi trafficking. EMBO J 24: 1122–1133 Nielsen MS, Madsen P, Christensen EI, Nykjaer A, Gliemann J, Kasper D, Pohlmann R, Petersen CM (2001) The sortilin cytoplasmic tail conveys Golgi–endosome transport and binds the VHS domain of the GGA2 sorting protein. EMBO J 20: 2180–2190 Nogi T, Shiba Y, Kawasaki M, Shiba T, Matsugaki N, Igarashi N, Suzuki M, Kato R, Takatsu H, Nakayama K, Wakatsuki S (2002) Structural basis for the accessory protein recruitment by the gamma-adaptin ear domain. Nat Struct Biol 9: 527–531 Ohno H, Stewart J, Fournier MC, Bosshart H, Rhee I, Miyatake S, Saito T, Gallusser A, Kirchhausen T, Bonifacino JS (1995) Interaction of tyrosine-based sorting signals with clathrin-associated proteins. Science 269: 1872–1875 Owen DJ, Evans PR (1998) A structural explanation for the recognition of tyrosine-based endocytotic signals. Science 282: 1327–1332 Page LJ, Sowerby PJ, Lui WW, Robinson MS (1999) Gamma-synergin: an EH domaincontaining protein that interacts with gamma-adaptin. J Cell Biol 146: 993–1004 Polishchuk EV, Di Pentima A, Luini A, Polishchuk RS (2003) Mechanism of constitutive export from the golgi: bulk flow via the formation, protrusion, and en bloc cleavage of large trans-golgi network tubular domains. Mol Biol Cell 14: 4470–4485 Polishchuk RS, Polishchuk EV, Marra P, Alberti S, Buccione R, Luini A, Mironov AA (2000) Correlative light-electron microscopy reveals the tubular–saccular ultrastructure of carriers operating between Golgi apparatus and plasma membrane. J Cell Biol 148: 45–58 Polishchuk RS, San Pietro E, Di Pentima A, Tete S, Bonifacino JS (2006) Ultrastructure of long-range transport carriers moving from the trans Golgi network to peripheral endosomes. Traffic 7: 1092–1103 Puertollano R, Aguilar RC, Gorshkova I, Crouch RJ, Bonifacino JS (2001) Sorting of mannose 6-phosphate receptors mediated by the GGAs. Science 292: 1712–1716 Puertollano R, Van der Wel NN, Greene LE, Eisenberg E, Peters PJ, Bonifacino JS (2003) Morphology and dynamics of clathrin/GGA1-coated carriers budding from the trans-Golgi network. Mol Biol Cell 14: 1545–1557 Ritter B, Philie J, Girard M, Tung EC, Blondeau F, McPherson PS (2003) Identification of a family of endocytic proteins that define a new alpha-adaptin ear-binding motif. EMBO Rep 4: 1089–1095 Roux A, Cappello G, Cartaud J, Prost J, Goud B, Bassereau P (2002) A minimal system allowing tubulation with molecular motors pulling on giant liposomes. Proc Natl Acad Sci USA 99: 5394–5399 Sahagian GG, Steer CJ (1985) Transmembrane orientation of the mannose 6-phosphate receptor in isolated clathrin-coated vesicles. J Biol Chem 260: 9838–9842 Shiba T, Kametaka S, Kawasaki M, Shibata M, Waguri S, Uchiyama Y, Wakatsuki S (2004) Insights into the phosphoregulation of beta-secretase sorting signal by the VHS domain of GGA1. Traffic 5: 437–448 Shiba T, Takatsu H, Nogi T, Matsugaki N, Kawasaki M, Igarashi N, Suzuki M, Kato R, Earnest T, Nakayama K, Wakatsuki S (2002a) Structural basis for recognition of acidic-cluster dileucine sequence by GGA1. Nature 415: 937–941 Shiba Y, Takatsu H, Shin HW, Nakayama K (2002b) Gamma-adaptin interacts directly with Rabaptin-5 through its ear domain. J Biochem 131: 327–336 Takatsu H, Katoh Y, Shiba Y, Nakayama K (2001) Golgi-localizing, gamma-adaptin ear homology domain, ADP-ribosylation factor-binding (GGA) proteins interact with acidic dileucine sequences within the cytoplasmic domains of sorting receptors through their Vps27p/Hrs/STAM (VHS) domains. J Biol Chem 276: 28541–28545 Takatsu H, Yoshino K, Nakayama K (2000) Adaptor gamma ear homology domain conserved in gamma-adaptin and GGA proteins that interact with gamma-synergin. Biochem Biophys Res Commun 271: 719–725

Protein transport from the trans-Golgi network to endosomes

*

401

Waguri S, Dewitte F, Le Borgne R, Rouille Y, Uchiyama Y, Dubremetz JF, Hoflack B (2003) Visualization of TGN to endosome trafficking through fluorescently labeled MPR and AP-1 in living cells. Mol Biol Cell 14: 142–155 Wasiak S, Legendre-Guillemin V, Puertollano R, Blondeau F, Girard M, De Heuvel E, Boismenu D, Bell AW, Bonifacino JS, McPherson PS (2002) Enthoprotin: a novel clathrin-associated protein identified through subcellular proteomics. J Cell Biol 158: 855–862 Wu X, Zhao X, Puertollano R, Bonifacino JS, Eisenberg E, Greene LE (2003) Adaptor and clathrin exchange at the plasma membrane and trans-Golgi network. Mol Biol Cell 14: 516–528 Zhang F, Yim YI, Scarselletta S, Norton M, Eisenberg E, Greene LE (2007) Clathrin adaptor GGA1 polymerizes clathrin into tubules. J Biol Chem 282: 13282–13289 Zhu Y, Doray B, Poussu A, Lehto VP, Kornfeld S (2001) Binding of GGA2 to the lysosomal enzyme sorting motif of the mannose 6-phosphate receptor. Science 292: 1716–1718

402

*

R. Castino and C. Isidoro

The transport of soluble lysosomal hydrolases from the Golgi complex to lysosomes Roberta Castino and Ciro Isidoro

Based on proteomic analyses the lysosome contains over 50 soluble lysosomal hydrolases plus several accessory integral membrane proteins (Journet et al. 2002; Sleat et al. 2005; Kolman et al. 2005; Czupalla et al. 2006). These proteins are synthesized at the rough endoplasmic reticulum and travel along the secretory pathway till they reach the trans-Golgi network (TGN) where they are sorted and delivered to pre-lysosomal compartments. Mannose-6-Phosphate (M6P) is the sorting signal for the majority, though not all, of the proteins destined to lysosomes. M6P groups have, in fact, been found also in proteins with no obvious lysosomal localization (Sleat et al. 2006). On the other hand, proteins not bearing the M6P group can also be targeted to the lysosome and proteins bearing the M6P group can be delivered to the lysosome independently of this group (MPR-independent transport). Particularly intriguing is the finding of the M6P group on secreted polypeptides such as the hydrolase plasma a-fucosidase or the hormone precursor Vasopressin-Neurophysisn 2 Copeptin and on endoplasmic reticulum (ER)-resident proteins such as the chaperone Microsomal Stress 70 Protein ATPase or the Sulfatase-modifying Factor 2 Precursor (Sleat et al. 2006). The functional significance of M6P in non-lysosomal resident protein remains at present a conundrum. Whatever the significance, this means that the presence of M6P does not impose the ineluctable targeting of a protein toward the endosomal compartment and, at one time, it reveals that generation of M6P occurs at a site relatively distant from where the MPRs bind their ligands and that other receptors specific for peptide signal may rescue these proteins before they encounter the MPRs. In a specular way, the presence of M6P does not ineluctably impose the MPRdependent targeting to the lysosome, as lysosomal proteins bearing the M6P can reach the lysosome through mechanisms that are independent of MPR shuttles. The functional relevance of MPR-independent segregation mechanisms is more obvious for proteins not bearing the M6P. In yeast, for instance, the transport of proteins to the vacuoles (equivalent to the mammalian lysosomes) is not dependent on the presence of M6P. In the following paragraphs we will focus on the biogenesis of M6P and on the pathways for lysosomal segregation of soluble lysosomal hydrolases that rely or not on the presence of M6P.

How and where is the M6P formed? In mammalian cells, the precursor of lysosomal enzymes is co-translationally glycosylated by en bloc transfer of a preformed oligosaccharide chain

The transport of soluble lysosomal hydrolases

*

403

containing nine mannose residues onto the NH2-group of the Asn-X-Ser/Thr tripeptide. Lysosomal proteins in general contain several N-glycosylation sites, though not all of them are necessarily glycosylated. For instance, the Niemann-Pick type C protein possesses three potential N-glycosylation sites of which the first (from the N-terminus) is never glycosylated, the second always contains a high-mannose type oligosaccharide and the third is either not glycosylated or bears a high-mannose or a complex type oligosaccharide (Chikh et al. 2005). After removal of the three terminal glucoses, the glycosylated lysosomal proenzyme moves to the Golgi complex where the N-linked oligosaccharides are subsequently modified in order to acquire the mannose-6-phosphate (M6P) recognition marker. The biosynthesis of M6P occurs in two steps: in the first step N-acetylglucosaminyl 1-phosphate residues are transferred to up to two mannose residues thus forming phosphodiester groups at C-6 hydroxyls in high-mannose oligosaccharides; in the second step, the N-acetylglucosamine residues covering the phosphate-mannosyl groups are removed to expose the M6P monoester (Traub and Kornfeld 1997; Faulhaber et al. 1998). Ectopic human cathepsin D synthesized by transfected Baby Hamster Kidney cells was shown to acquire the M6P with equal efficiency as the endogenous parental enzyme even when expressed at very high level, indicating that the two cathepsins were equally recognized as substrates for the phosphorylation and the uncovering reactions and that both these reactions were not saturable (Isidoro et al. 1998). High-mannose oligosaccharides on lysosomal enzymes are heterogeneous with respect to the number and the position of the mannose residues that are phosphorylated. Of the nine mannoses present in each oligosaccharide, only five are potential sites of phosphorylation (Varki and Kornfeld 1980). However, a maximum of two of these sites are effectively phosphorylated. In general, more than one oligosaccharide in the lysosomal hydrolase polypeptide are phosphorylated, though not necessarily all of them. In arylsulphatase A, for instance, of the three N-linked oligosaccharides phosphorylation is restricted to only the ones attached at Asn-158 and Asn-350 (Sommerlade et al. 1994). Proteomic studies have mapped the peptide sites at which the high-mannose oligosaccharides are phosphorylated (Sleat et al. 2006). The phosphorylation is catalyzed by the UDP-N-acetyl-D-glucosamine: lysosomal-enzyme N-acetylglucosaminephosphotransferase (hereafter simply referred to as the ‘phosphotransferase’). This enzyme is composed of homodimers of three subunits, of which the g-subunit recognizes the substrate (i.e., the lysosomal pro-enzyme) and the a–b subunits explicate the enzymatic activity (Bao et al. 1996; Tiede et al. 2005). Recognition of the substrate is based on protein domains flanking the glycosylation site and, in particular, on certain critical lysine residues in the lysosomal proenzyme (Baranski et al. 1990, 1991; Cuozzo et al. 1995; Yaghootfam et al. 2003; Steet et al. 2005). Studies in which the transport of newly synthesized lysosomal proenzymes was inhibited by lowering the temper-

404

*

R. Castino and C. Isidoro

ature or by the ionophore carbonyl cyanide m-chlorophenylhydrazone revealed that mannosyl phosphorylation takes place at two sites in an ordered compartmentalized sequence: at pre-Golgi level a first Nacetylglucosamine-1-phosphate is transferred onto a mannose residue of the a-1,6 branch, and at Golgi level, upon removal of a mannose residue from the a-1,6 branch, a second N-acetylglucosamine-1-phosphate is further added onto a mannose residue of the a-1,3 branch (Lazzarino and Gabel 1988, 1989). Subsequent studies on the formation of M6P in KDELtagged cathepsin D (Pelham 1988; Isidoro et al. 1996) and arylsulphatase A (Dittmer and Von Figura 1999) chimeras confirmed that phosphorylation of high-mannose oligosaccharides proceeds in a stepwise fashion and terminates prior to reaching the retrieval site by the KDEL receptor, i.e., the trans-Golgi. When a thiol-mediated ER retention signal was attached to cathepsin D, the oligosaccharides did not acquire the phosphate indicating that the first phosphorylation event occurs at a post-ER site (Isidoro et al. 1996). Based on these observation it is assumed that mannosyl phosphorylation takes place at both an ER–Golgi intermediate compartment and early (cis)-Golgi compartment (Fig. 1).

Figure 1. Schematic representation of the compartmentalized reactions leading to the formation of M6P. Upon exit from the ER, N-glycosylated lysosomal pro-hydrolases (LysH) acquire the mannose-6-phosphate group trough a series of reactions that involve a two-site phosphorylation and the removal of GlcNAc. These reactions occur in distinct compartments along the secretory pathway (see text for details). At TGN level the M6P-tagged LysH encounter the MPR. Some ER resident proteins are allowed to reach the trans-Golgi (where they acquire complex-type sugars; this is the case of KDEL-tagged proteins), other ER proteins are allowed to reach the TGN, where the uncovering enzyme resides; these proteins acquire the M6P, yet are retrieved back into the ER before they encounter the MPR (see the text for further explanations).

The transport of soluble lysosomal hydrolases

*

405

Lysosomal pro-enzymes bearing one or two mannosyl-phosphodiester(s) are substrates of the N-acetylglucosamine-1-phosphodiester a-N-acetylglucosaminidase (so called ‘uncovering enzyme’), which removes the covering N-acetyl-D-glucosamine (Glc-NAc) residue and exposes the mannosyl-phosphomonoester (Varki and Kornfeld 1980). This reaction has been shown to occur in a compartment beyond the site of action of brefeldin A (Radons et al. 1990; Sampath et al. 1992), a fungal macrolide that disassembles the Golgi stacks and causes the retrograde transport of certain Golgi enzymes to the ER (Lippincott-Scwartz et al. 1989, 1990). In the presence of brefeldin A, cathepsin D underwent the first step (i.e., the mannose phosphorylation with formation of mono- and di-phosphodiesters), but not the second step (i.e., the removal of the covering GlcNAc), of the M6P biogenetic pathway (Radons et al. 1990). This is similar to the behaviour of KDEL-tagged cathepsin D (Pelham 1988; Isidoro et al. 1996). An additional proof of the complete spatial separation of the compartments harbouring the phosphotransferase and the uncovering enzyme arose from a study showing that ammonium chloride could impair the uncovering reaction in phosphorylated procathepsin D (Isidoro et al. 1990). Of note, in cultured cells ammonium chloride greatly increased the secretion of procathepsin D bearing ‘covered’ M6P residues, suggesting that in the presence of this drug the formation of secretory vesicles could bypass the compartment harbouring the uncovering enzyme (Isidoro et al. 1990). More recently, it has been definitely shown that the uncovering enzyme resides in the trans-Golgi network, i.e., the same compartment where the M6P-receptors encounter their ligands (Rohrer and Kornfeld 2001; Nair et al. 2005). This study demonstrated that the uncovering enzyme may travel to the plasma-membrane and be retrieved to the TGN via coated vesicles, and that differently from galactosyl-transferase (a trans-Golgi enzyme) it does not relocate into the ER after brefeldin A treatment (Rohrer and Kornfeld 2001). Proteomic analyses have revealed the presence of the M6P groups on proteins that do not exhibit a lysosomal localization or lysosomal-associated function and that rather reside in the ER (Sleat et al. 2006), an occurrence compatible with their transport along the secretory pathway up to the compartment harbouring the uncovering enzyme and their retrieval back prior to encounter with the MPR. This fact would suggest the existence of a further spatial separation within the TGN between the sites of the uncovering reaction and the MPR binding. It is to note that the ER-resident M6Pcontaining proteins so far characterized do not contain the KDEL retrieval signal (Sleat et al. 2006), which would have imposed the retrieval back to the ER from a site (the trans-Golgi) upstream to the site of uncovering (the TGN) (Pelham 1988). To complicate this picture, two of the three oligosaccharides in the precursor of TGF-b were shown to contain M6P (Purchio et al. 1988) and a single M6P group was also found on the secretory hydrolase plasma a-fucosidase (Sleat et al. 2006).

406

*

R. Castino and C. Isidoro

Mechanisms of TGN to lysosome transport depending on M6P The soluble lysosomal proenzymes tagged with M6P are high-affinity ligands for the 300 kDa cation-independent and the 46 kDa cation-dependent M6P receptors known as CI-MPR and CD-MPR, respectively (Kornfeld and Melman 1989; Von Figura 1991). Both MPRs are integral type 1 transmembrane proteins that shuttle between the TGN and acid pre-lysosomal organelles, yet they differ in their structure, membrane trafficking, M6Pbinding affinity and for their preferences for substrates. From a structural point of view, the two MPRs essentially differ for the length and binding capability of the extracellular portion, which is of approx. 2270 amino acids organised in 15 homologous domains and of 154 amino acids forming one single domain in the CI-MPR and the CD-MPR, respectively. Two high-affinity binding sites (domains 1–3 and 9) and one low-affinity binding site (domain 5) for M6P have been located in the CI-MPR, whereas only one M6P-binding site (homologous to domain 9 of CI-MPR) exists in the CD-MPR (Lobel et al. 1988; Tong et al. 1989; Tong and Kornfeld 1989; Olson et al. 2004; Reddy et al. 2004). Unlike the CD-MPR or domain 9 of the CI-MPR, domain 5 exhibits a much higher affinity for M6P-GlcNAc than M6P monoester, which suggests the possibility for the CI-MPR to target phosphodiester-containing lysosomal enzymes to the lysosome (Chavez et al. 2007). The CI-MPR, unlike the CD-MPR, binds also the precursor of the transforming growth factor-b (Purchio et al. 1988) and interacts, at binding sites distinct from those of lysosomal enzymes, with a number of non-M6P-containing polypeptides that include insulin-like growth factor II, plasminogen, and retinoic acid (Devi et al. 1988; Leksa et al. 2002; Kang et al. 1997). While both MPRs traffic along the secretory pathway and can reach the plasma-membrane, only the CI-MPR can perform the endocytosis of extracellular lysosomal enzymes. The two MPRs also show different environmental requirements (pH, presence of divalent cations for CD-MPR) for optimal binding of the substrate (Tong et al. 1989; Olson et al. 2008). The role of MPRs in lysosomal segregation of soluble lysosomal pro-enzymes was confirmed in ‘in vitro’ studies with cells devoid of both MPRs (Ludwig et al. 1994; Pohlman et al. 1995). Mouse embryonic fibroblasts that are double-deficient of both MPRs missort the majority (85%) of soluble lysosomal enzymes into the medium. Overexpression (up to five times more than physiological level) of human CD-MPR only partially corrected the missorting, more than one-third of the newly synthesized lysosomal proteins being still secreted. However, a two-fold over-expression (with respect to physiological level) of the CI-MPR completely restored an efficient transport to lysosomes; in this circumstance almost one-fourth of the lysosomal pro-enzymes was segregated via a secretion-recapture pathway sensitive to extracellular M6P (Kasper et al. 1996). Altogether, these data support the view that both MPRs are needed

The transport of soluble lysosomal hydrolases

*

407

to guarantee the correct sorting of the total subset of soluble lysosomal hydrolases (Pohlman et al. 1995; Munier-Lehman et al. 1996), with the CI-MPR as the major player (Sohar et al. 1998). The number and the position of M6P groups within the oligosaccharides and the aminoacid context of the phosphorylated oligosaccharides are critical determinants of the binding affinity of the lysosomal enzyme toward either MPR (Dittmer et al. 1997; Olson et al. 1999, 2008). Therefore, distinct lysosomal enzymes can exhibit different affinities for the two receptors (Sleat and Lobel 1997; Qian et al. 2008; Pohlman et al. 1995). CI-MPR is known to bind preferentially lysosomal enzymes bearing oligosaccharides with two uncovered phosphate residues. For instance, procathepsin D is sorted efficiently via CI-MPR and poorly via CD-MPR (Pohlman et al. 1995). Proteomic analysis of M6P-tagged protein in the serum of mice deficient for either MPR revealed that CREG (cellular repressor of E1A-stimulated genes 1) and heparanase were abnormally missorted in the absence of the CD-MPR, whereas cathepsin D and prosaposin were abnormally missorted in the absence of the CI-MPR (Qian et al. 2008) (Fig. 2). Lysosomal pro-enzyme recognition by MPRs may also involve molecular portions other than the oligosaccharides. When human cathepsin D was over-expressed in Baby Hamster Kidney cells the endogenous cathepsin D was preferentially segregated in the lysosomal pathway while a large proportion of the human procathepsin D was diverted into the medium and not rescued

Figure 2. Schematic representation of the MPR-dependent and MPR-independent pathways for the lysosomal delivery of the following lysosomal proteins: cathepsin D (CD, ) ; Acid Phosphatase ( ); B-GlucoCerebroside (B-GlcCer ); prosaposin (pSap, ). The alternative receptors Sortilin and LIMP2 are also shown. Further details are given in the text.

408

*

R. Castino and C. Isidoro

via endocytosis (Isidoro et al. 1990). This occurred despite both the endogenous and the ectopic cathepsins were tagged with M6P with comparable efficiency (Isidoro et al. 1998). In the transfected cells cultivated in the presence of ammonium chloride, which interferes with the MPR-dependent targeting of lysosomal enzymes (Hasilik and Neufeld 1980), only the endogenous cathepsin D was largely missorted (Isidoro et al. 1990, 1998). These results are compatible with the involvement of, as yet unknown, peptide portions that confer a species-specific preferential interaction in the binding to MPRs. MPRs bind their ligands at TGN and (in the case of CI-MPR) plasmamembrane levels and deliver them to sorting pre-lysosomal organelles that identify with multivesicular bodies (MVB) and early endosomes (Hirst et al. 1998; Van Meel and Klumperman 2008). MVB are characterized by the intraluminal presence of small vesicles (50–80 nm in diameter) in which ligand-occupied receptors are sequestered, and are regarded to as late endosomes (Van Meel and Klumperman 2008). The mild acidity of these sorting compartments is sufficient to allow the MPR-ligand dissociation so that along with the maturation of late endosome into lysosome the empty receptors are retrieved back to TGN and are made available for new cycles of TGN to endosomes transport. This assumption was based on the observation that lysosomotropic amines (e.g., ammonium chloride and chloroquine) and monensin, which cause alkalinisation of the vacuolar compartments, stimulates the missorting of M6P-bearing lysosomal enzymes (Braulke et al. 1987). Yet, it has been recently shown that in cells lacking the CI-MPR, the CD-MPR retains the capacity to perform the lysosomal segregation of M6P-bearing enzymes even when endosomal acidification was abrogated by ammonium chloride or monensin (Probst et al. 2006).

Mechanisms of TGN to lysosome transport not depending on M6P The existence in mammalian cells of mechanisms for the transport of lysosomal proteins from TGN to endosomes independent of MPRs was suggested in the early 1980s based essentially on two observations: (1) impairment of endosomal acidification does not completely prevent the lysosomal delivery of lysosomal proenzymes (Hasilik and Neufeld 1980; Braulke et al. 1987); and (2) in certain non-mesenchymal tissues of I-cell disease (mucolipidosis II) patients (who genetically lack the phosphotransferase and therefore cannot tag lysosomal proenzymes with M6P) most lysosomal enzymes are still targeted correctly to lysosomes (Owada and Neufeld 1982; Waheed et al. 1982; Glickman and Kornfeld 1993). The latter phenotype has recently been reproduced in a mouse model of I-cell disease (Gelfman et al. 2007). The finding that lysosomes also contain hydrolases

The transport of soluble lysosomal hydrolases

*

409

whose precursor does not acquire the M6P (e.g., acid phosphatase and b-glucocerebrosidase) further demonstrates the existence of MPR-independent mechanisms of TGN to lysosomes transport (Aerts et al. 1988; Waheed et al. 1988). The fact that certain cell types (hepatocytes and thymocytes, but not fibroblasts) from knockout mice lacking both MPRs contain certain lysosomal enzymes at levels comparable to those of normal mice clearly indicate that at least a subset of lysosomal enzymes are trafficked (also) via a pathway not mediated by MPRs that is cell type-specific (Dittmer et al. 1999). These ‘alternative’ pathways are now being elucidated. Lysosomal acid phosphatase is synthesized as a type I trans-membrane precursor that follows a plasma-membrane-endocytosis route to reach the lysosomes wherein it is processed into a soluble active form (Waheed et al. 1988). b-Glucocerebrosidase was shown to associate with vesicular membranes soon after its synthesis (Rijnboutt et al. 1991) and to reach the lysosomes in an MPR-independent manner in I-cell fibroblasts (Van Dongen et al. 1985). Very recently the molecular basis for such membrane association has been clarified: the lysosomal integral membrane protein type 2 (LIMP-2) acts as the physiological carrier by interacting through its luminal domain with b-glucocerebrosidase at the ER level; the complex is then trafficked all the way along the secretory pathway to endosomes and reaches the lysosome where the lysosomal enzyme is discharged because of the acid pH (Reczek et al. 2007). Procathepsins L and D, though bearing the M6P group, can also be transported to lysosomes in a M6P-independent manner. Both these proteins have been shown associated with membranes in fibroblasts, macrophages and hepatocytes (Diment et al. 1988; McIntyre and Erickson 1991; Rijnboutt et al. 1991). The pro-peptide of procathepsin L and procathepsin D plays a critical role in the association to membranes (McIntyre and Erickson 1991; Rijnboutt et al. 1991). Procathepsin D was shown to form a transient non-covalent complex with prosaposin, the precursor of the sphingolipid activator proteins (sap) A, B and C (Zhu and Conner 1994). This complex formed at ER level in a M6P-independent manner, continued in the Golgi complex and dissociated at the level of dense lysosomes (Zhu and Conner 1994). Of note, prosaposin has been shown to interact in a M6P-independent manner with and to be transported to lysosomes by the multiligand receptor Sortilin (Lefrancois et al. 2003). Very recently, Sortilin has been shown to have similar subcellular distribution and identical trafficking pathways from TGN to endosomes and back as the CI- and CD-MPRs (Mari et al. 2008). This observation poses the question as to whether M6P-dependent and M6P-independent transport are completely distinct processes, or whether they actually intersect during the normal transport of lysosomal enzymes.  del Acknowledgements. Work in the authors’ laboratory are supported by Universita Piemonte Orientale, Regione Piemonte and Lega Italiana per la Lotta contro i Tumori (Novara). Thanks are due to Drs Eeva-Liisa Eskelinen and Alexander Mironov for critical reading of the manuscript.

410

*

R. Castino and C. Isidoro

References Aerts JM, Schram AW, Strijland A, Van Weely S, Jonsson LM, Tager JM, Sorrell SH, Ginns EI, Barranger JA, Murray GJ (1988) Glucorebrosidase, a lysosomal enzyme that does not undergo oligosaccharide phosphorylation. Biochim Biophys Acta 964: 303–308 Bao M, Booth JL, Elmendorf BJ, Canfield WM (1996) Bovine UDP-N-acetylglucosamine: lysosomal-enzyme N-acetylglucosamine-1-phosphotransferase. I. Purification and subunit structure. J Biol Chem 271: 31437–31445 Baranski TJ, Faust PL, Kornfeld S (1990) Generation of a lysosomal enzyme targeting signal in the secretory protein pepsinogen. Cell 63: 281–291 Baranski TJ, Koelsch G, Hartsuck JA, Kornfeld S (1991) Mapping and molecular modeling of a recognition domain for lysosomal enzyme targeting. J Biol Chem 266: 23365–23372 Braulke T, Gartung C, Hasilik A, Von Figura K (1987) Is movement of mannose 6phosphate-specific receptor triggered by binding of lysosomal enzymes? J Cell Biol 104: 1735–1742 Chavez CA, Bohnsack RN, Kudo M, Gotschall RR, Canfield WM, Dahms NM (2007) Domain 5 of the cation-independent mannose 6-phosphate receptor preferentially binds phosphodiesters (mannose 6-phosphate N-acetylglucosamine ester) Biochemistry. 46: 12604–12617 Chikh K, Rodriguez C, Vey S, Vanier MT, Millat G (2005) Niemann-Pick type C disease: subcellular location and functional characterization of NPC2 proteins with naturally occurring missense mutations. Hum Mutat 26: 20–28 Cuozzo JW, Tao K, Wu QL, Young W, Sahagian GG (1995) Lysine-based structure in the proregion of procathepsin L is the recognition site for mannose phosphorylation. J Biol Chem 270: 15611–15619 Czupalla C, Manusukoski H, Riedl T, Krause E, Hoflack B (2006) Proteomic analysis of lisosoma acid hydrolases secreted by osteoclasts: implications for lytic enzyme transport and bone metabolism. Mol Cell Proteomics 5: 134–143 Devi GR, Byrd JC, Slentz DH, MacDonald RG (1998) An insulin-like growth factor II (IGF-II) affinity-enhancing domain localized within extracytoplasmic repeat 13 of the IGF-II/ mannose 6-phosphate receptor. Mol Endocrinol 12: 1661–1672 Diment S, Leech MS, Stahl PD (1988) Cathepsin D is membrane-associated in macrophage endosomes. J Biol Chem 263: 6901–6907 Dittmer F, Von Figura K (1999) Phosphorylation of arylsulphatase A occurs through multiple interactions with the UDP-N-acetylglucosamine-1-phosphotransferase proximal and distal to its retrieval site by the KDEL receptor. Biochem J 340: 729–736 Dittmer F, Pohlmann R, Von Figura K (1997) The phosphorylation pattern of oligosaccharides in secreted procathepsin D is glycosylation site-specific and independent of the expression of mannose 6-phosphate receptors. J Biol Chem 272: 852–858 Faulhaber J, Fensom A, Hasilik A (1998) Abnormal lysosomal sorting with an enhanced secretion of cathepsin D precursor molecules bearing monoester phosphate groups. Eur J Cell Biol 77: 134–140 Gelfman CM, Vogel P, Issa TM, Turner CA, Lee WS, Kornfeld S, Rice DS (2007) Mice lacking alpha/beta subunits of GlcNAc-1-phosphotransferase exhibit growth retardation, retinal degeneration, and secretory cell lesions. Invest Ophthalmol Vis Sci 48: 5221–5228 Glickman JN, Kornfeld S (1993) Mannose 6-phosphate-independent targeting of lysosomal enzymes in I-cell disease B lymphoblasts. J Cell Biol 123: 99–108 Hasilik A, Neufeld EF (1980) Biosynthesis of lysosomal enzymes in fibroblasts. Phosphorylation of mannose residues. J Biol Chem 255: 4946–4950

The transport of soluble lysosomal hydrolases

*

411

Hirst J, Futter CE, Hopkins CR (1998) The kinetics of mannose 6-phosphate receptor trafficking in the endocytic pathway in HEp-2 cells: the receptor enters and rapidly leaves multivesicular endosomes without accumulating in a prelysosomal compartment. Mol Biol Cell 9: 809–816 Isidoro C, Radons J, Baccino FM, Hasilik A (1990) Suppression of the “uncovering” of mannose-6-phosphate residues in lysosomal enzymes in the presence of NH4Cl. Eur J Biochem 191: 591–597 Isidoro C, Maggioni C, Demoz M, Pizzagalli A, Fra AM, Sitia R (1996) Exposed thiols confer localization in the endoplasmic reticulum by retention rather than retrieval. J Biol Chem 271: 26138–26142 Isidoro C, Baccino FM, Hasilik A (1998) Human and hamster procathepsin D, although equally tagged with mannose-6-phosphate, are differentially targeted to lysosomes in transfected BHK cells. Cell Tissue Res 292: 303–310 Journet A, Chapel A, Kieffer S, Roux F, Garin J (2002) Proteomic analysis of human lysosomes: application to monocytic and breast cancer cells. Proteomics 2: 1026–1040 Kang JX, Li Y, Leaf A (1997) Mannose 6-phosphate/insulin-like growth factor-II is a receptor for retinoic acid. Proc Natl Acad Sci USA 95: 13671–13676 Kasper D, Dittmer F, Von Figura K, Pohlmann R (1996) Neither type of mannose 6phosphate receptor is sufficient for targeting of lysosomal enzymes along intracellular routes. J Cell Biol 134: 615–623 Kollman K, Mutenda KE, Balleininger M, Eckermann E, Von Figura K, Schmidt B, Lubke T (2005) Identification of novel lysosomal matrix proteins by proteome analysis. Proteomics 5: 3966–3978 Kornfeld S, Melman I (1989) The biogenesis of lysosomes. Annu Rev Cell Biol 5: 483–525 Lazzarino DA, Gabel CA (1988) Biosynthesis of the mannose 6-phosphate recognition marker in transport-impaired mouse lymphoma cells. J Biol Chem 263: 10118–10126 Lazzarino DA, Gabel CA (1989) Mannose processing is an important determinant in the assembly of prosphorylated high mannose-type oligosaccharides. J Biol Chem 264: 5051–5023 Lefrancois S, Zeng J, Hassan AJ, Canuel M, Morales CR (2003) The lysosomal trafficking of sphingolipid activator proteins (SAPs) is mediated by sortilin. EMBO J 22: 6430–6437 r S, Cebecauer M, Hilgert I, Breuss J, Weidle UH, Horejsí V, Binder BR, Leksa V, Goda Stockinger H (2002) The N terminus of mannose 6-phosphate/insulin-like growth factor 2 receptor in regulation of fibrinolysis and cell migration. J Biol Chem 277: 40575–40582 Lippincott-Schwartz J, Yuan LC, Bonifacino JS, Klausner RD (1989) Rapid redistribution of Golgi proteins into the ER in cells treated with brefeldin A: evidence for membrane cycling from Golgi to ER. Cell 56: 801–813 Lippincott-Schwartz J, Donaldson JG, Schweizer A, Berger EG, Hauri HP, Yuan LC, Klausner RD (1990) Microtubule-dependent retrograde transport of proteins into the ER in the presence of brefeldin A suggests an ER recycling pathway. Cell 60: 821–836 Lobel P, Dahms NM, Kornfeld S (1988) Cloning and sequence analysis of the cationindependent mannose 6-phosphate receptor. J Biol Chem 263: 2563–2570 Ludwig T, Munier-Lehmann H, Bauer U, Hollinshead M, Ovitt C, Lobel P, Hoflack B (1994) Differential sorting of lysosomal enzymes in mannose 6-phosphate receptor-deficient fibroblasts. EMBO J 13: 3430–3437 Mari M, Bujny MV, Zeuscner D, Geerts WJC, Griffith J, Petersen CM, Cullen PJ, Klumperman J, Geuze HJ (2008) SNX1 defines and early endosomal recycling exit for sortilin and mannose 6-phosphate receptors. Traffic 9: 380–393 McIntyre GF, Erickson AH (1991) Procathepsins L and D are membrane-bound in acidic microsomal vesicles. J Biol Chem 266: 15438–15445

412

*

R. Castino and C. Isidoro

Munier-Lehmann H, Mauxion F, Bauer U, Lobel P, Hoflack B (1996) Re-expression of the mannose 6-phosphate receptors in receptor-deficient fibroblasts. Complementary function of the two mannose 6-phosphate receptors in lysosomal enzyme targeting. J Biol Chem 271: 15166–15174 Nair P, Schaub BE, Huang K, Chen X, Murphy RF, Griffith JM, Geuze HJ, Rohrer J (2005) Characterization of the TGN exit signal of the human mannose 6-phosphate uncovering enzyme. J Cell Sci 271: 15166–15174 Olson LJ, Hancock MK, Dix D, Kim JJ, Dahms NM (1999) Mutational analysis of the binding site residues of the bovine cation-dependent mannose 6-phosphate receptor. J Biol Chem 274: 36905–36911 Olson LJ, Dahms NM, Kim JJ (2004) The N-terminal carbohydrate recognition site of the cation-independent mannose 6-phosphate receptor. J Biol Chem 279: 34000–34009 Olson LJ, Hindgaul O, Dahms NM, Kim JJP (2008) Structural insight into the mechanism of pH-dependent ligand binding and release by the cation-dependent mannose 6-phosphate receptor. JBC http://www.jbc.org/cgi/doi/10.1074/jbc.M708994200 Owada M, Neufeld EF (1982) Is there a mechanism for introducing acid hydrolases into liver lysosomes that is independent of mannose 6-phosphate recognition? Evidence from I-cell disease. Biochem Biophys Res Commun 105: 814–820 Pelham HR (1988) Evidence that luminal ER proteins are sorted from secreted proteins in a post-ER compartment. EMBO J 7: 913–918 Pohlman R, Wendland M, Boeker C, Von Figura H (1995) The two mannose 6-phosphate receptors transport distinct complements of lysosomal proteins. J Biol Chem 270: 27311–27318 Probst OC, Ton P, Svoboda B, Gannon A, Schuhmann W, Wieser J, Pohlmann R, Mach L (2006) The 46-kDa mannose 6-phosphate receptor does not depend on endosomal acidification for delivery of hydrolases to lysosomes. J Cell Sci 119: 4935–4943 Purchio AF, Cooper JA, Brunner AM, Lioubin MN, Gentry LE, Kovacina KS, Roth RA Marquardt H (1988) Identification of mannose 6-phosphate in two asparaginelinked sugar chains of recombinant transforming growth factor-beta 1 precursor. J Biol Chem 1988 263: 14211–14215 Qian M, Sleat DE, Zheng H, Moore D, Lobel P (2008) Proteomics analysis of serum mutant mice reveals lysosomal proteins selectively transported by each of the two mannose 6-phosphate receptors. Mol Cell Proteomics 7: 58–70 Radons J, Isidoro C, Hasilik A (1990) Brefeldin A prevents uncovering but not phosphorylation of the recognition marker in cathepsin D. Biol Chem Hoppe-Seyler 371: 567–573 Reczek D, Schwake M, Schroder J, Hughes H, Blanz J, Jin Xiaoying Brondyk W, Van Patten S, Edmunds T, Saftig P (2007) LIMP-2 is a receptor for lysosomal mannose-6-phosphate-independent targeting of b-glucocerebrosidase. Cell 131: 770–783 Reddy ST, Chai W, Childs RA, Page JD, Feizi T, Dahms NM (2004) Identification of a low affinity mannose 6-phosphate-binding site in domain 5 of the cation-independent mannose 6-phosphate receptor. J Biol Chem 279: 38658–38667 Rijnboutt S, Kal AJ, Geuze HJ, Aerts H, Strous GJ (1991) Mannose 6-phosphate-independent targeting of cathepsin D to lysosomes in HepG2 cells. J Biol Chem 266: 23586–23592 Rohrer J, Kornfeld R (2001) Lysosomal hydrolase mannose 6-phosphate uncovering enzyme resides in the trans-Golgi network. Mol Biol Cell 12: 1623–1631 Sampath D, Varki A, Freeze HH (1992) The spectrum of incomplete N-linked oligosaccharides synthesized by endothelial cells in the presence of brefeldin. J Biol Chem 267: 4440–4455 Sleat DE, Lobel P (1997) Ligand binding specificities of the two mannose 6-phosphate receptors. J Biol Chem 272: 731–738

The transport of soluble lysosomal hydrolases

*

413

Sleat DE, Lackland H, Wang Y, Sohar I, Xiao G, Li H, Lobel P (2005) The human brain mannose 6-phosphate glycoproteome: a complex mixture composed of multiple isoforms of many soluble lysosomal proteins. Proteomics 5: 1520–1532 Sleat DE, Zheng H, Qian M, Lobel. P (2006) Identification of sites of mannose 6-phosphorylation on lysosomal proteins. Mol Cell Proteomics 5: 686–701 Sohar I, Sleat D, Gong Liu C, Ludwig T, Lobel P (1998) Mouse mutants lacking the cationindependent mannose 6-phosphate/insulin-like growth factor II receptor are impaired in lysosomal enzyme transport: comparison of cation-independent and cation-dependent mannose 6-phosphate receptor-deficient mice. Biochem J 330: 903–908 Sommerlade HJ, Selmer T, Ingendoh A, Gieselmann V, Von Figura K, Neifer K, Schmidt B (1994) Glycosylation and phosphorylation of arylsulfatase A. J Biol Chem 269: 20977–20981 Steet R, Lee WS, Kornfeld S (2005) Identification of the minimal lysosomal enzyme recognition domain in cathepsin D. J Biol Chem 280: 33318–33323 Tiede S, Storch S, Lubke T, Henrissat B, Bargal R, Raas-Rothschild A, Braulke T (2005) Mucopolidosis II is caused by mutations in GNPTA encoding tha alpha/beta GlcNac1-phosphotransferase. Nat Med 11: 1109–1112 Tong PY, Gregory W, Kornfeld S (1989) Ligand interactions of the cation-independent mannose 6-phosphate receptor. J Biol Chem 264: 7962–7969 Tong PY, Kornfeld S (1989) Ligand interactions of the cation-independent mannose 6phosphate receptor. J Biol Chem 264: 7970–7975 Traub LM, Kornfeld S (1997) The trans-Golgi network: a late secretory sorting station. Curr Opin Cell Biol. 9: 527–533 Van Dongen JM, Willemsen R, Ginns EI, Sips HJ, Tager JM, Barranger JA, Reuser AJ (1985) The subcellular localization of soluble and membrane-bound lysosomal enzymes in I-cell fibroblasts: a comparative immunocytochemical study. Eur J Cell Biol 39: 179–189 Van Meel E, Klumperman J (2008) Imaging and imagination: understanding the endolysosomal system. Histochem Cell Biol 129: 253–266 Varki A, Kornfeld S (1980) Identification of a rat liver alpha-N-acetylglucosaminyl phosphodiesterase capable of removing “blocking” alpha-N-acetylglucosamine residues from phosphorylated high mannose oligosaccharides of lysosomal enzymes. J Biol Chem 255: 8398–8401 Von Figura K (1991) Molecular recognition and targeting of lysosomal proteins.Curr Opin Cell Biol 3: 642–646 Waheed A, Pohlmann R, Hasilik A, Von Figura K, Van Elsen A, Leroy JG (1982) Deficiency of UDP-N-acetylglucosamine:lysosomal enzyme N-acetylglucosamine-1-phosphotransferase in organs of I-cell patients. Biochem Biophys Res Commun 105: 1052–1058 Waheed A, Gottschalk S, Hille A, Kreutler C, Pohlmann R, Braulke T, Hauser H, Geuze H, Von Figura K (1988) Human lysosomal acid phosphatase is transported as a transmembrane protein to lysosomes in transfected baby hamster kidney cells. EMBO J 7: 2351–2358 Yaghootfam A, Schestag F, Dierks T, Gieselmann V (2003) Recognition of arylsulfatase A and B by the UDP-N-acetylglucosamine:lysosomal enzyme N-acetylglucosaminephosphotransferase. J Biol Chem 278: 32653–32661 Zhu Y, Conner GE (1994) Intermolecular association of lysosomal protein precursors during biosynthesis. J Biol Chem 269: 3846–3851

414

*

E.-L. Eskelinen and A. A. Mironov

Transport of lysosomal membrane proteins from the Golgi complex to lysosomes Eeva-Liisa Eskelinen and Alexander A. Mironov

Definition and morphology of lysosomes Lysosomes are intracellular organelles with an acid interior (pH <5) bounded by a single membrane and containing many lysosomal hydrolases that are optimally active at an acid pH (Kornfeld and Mellman 1989). Lysosomes are mannose 6-phosphate receptor (MPR) negative and devoid of labeling for phosphorylated lysosomal enzymes, but positive for fluid phase endocytic markers if they are internalized for a long (>4 h) time (Ludwig et al. 1991). The lysosomal limiting membrane contains a high concentration of integral membrane proteins including the lysosome-associated membrane protein 1 (LAMP-1), LAMP-2, and lysosomal integral membrane proteins 1 and 2 (LIMP-1/CD63 and LIMP-2) (Eskelinen et al. 2003). Mature lysosomes may occasionally contain a clathrin coat (Traub et al. 1996). Usually, lysosomes exhibit spherical or ovoid shape (Fig. 1). However, sometimes, i.e., in cultured smooth muscle cells (Robinson et al. 1986) or in phorbol ester-stimulated macrophages (Robinson and Karnovsky 1991), lysosome can form interconnected networks, which suggests that they can constantly fuse with each other. Late endosomes (also called pre-lysosomes) are sometimes difficult to distinguish from lysosomes. Late endosomes are highly acidic, contain a similar concentration of the membrane proteins LAMP-1 and LAMP-2 as lysosomes, and have detectable concentration of lysosomal hydrolases (Griffiths et al. 1988, 1990). Late endosomes are capable of exhibiting significant acid hydrolase activity, and they can contain a lot of indigestible material. Individual late endosomes can have different pH values from 4, or even below, up to neutrality (Butor et al. 1995). There are several models describing the biogenesis of lysosomes. According to the maturation hypothesis, lysosomes are formed from late endosomes (or pre-lysosomes) by maturation due to changes in the composition of their contents and limiting membranes (Murphy 1991). The preexisting-organelle hypothesis predicts that lysosomes, early endosomes and late endosomes are distinct organelles, which are not transformed into each other by maturation. Vesicles shuttle between these organelles and transport proteins and lipids (Griffiths and Gruenberg 1991). This model predicts that new lysosomes form via membrane transport from the Golgi complex as so-called primary

Transport of lysosomal membrane proteins from the Golgi complex to lysosomes

*

415

Figure 1. Transmission electron microscopic image of lysosomes in a mouse embryonic fibroblast. Four lysosomes and one late endosome-like vesicle (on the right) in a row that is probably arranged on a microtubule (indicated by the arrowheads). The insert shows the third lysosome from left at a higher magnification, to demonstrate the internal concentric membranes typical for lysosomes.

lysosomes, i.e., lysosomes that have not yet seen any degradation substrates (De Duve 1963). Distribution of lysosomes in cells depends on intact microtubules and the activity of microtubule motors. The small GTPase Rab7 and its effectors RILP and ORP1L, and the microtubule motor complex dynein–dynactin mediate the movements and subcellular positioning of lysosomes (Cantalupo et al. 2001; Johansson et al. 2007; Jordens et al. 2001). It also seems that LAMPs are needed for the microtubule-dependent movement of lysosomes (Huynh et al. 2007), although it is unclear at present whether LAMPs are physically linked to the microtubular motor complexes.

Lysosomal membrane proteins The limiting membranes of lysosomes and late endosomes are (Harter and Mellman 1992) enriched in highly glycosylated transmembrane proteins named lysosome-associated membrane proteins (LAMPs; also known as LGPs), and lysosomal integral membrane proteins (LIMPs). Among the most abundant proteins are two type I transmembrane proteins, LAMP-1 and LAMP-2, as well as LIMP-2 that spans the membrane twice with both its amino and carboxy terminal ends in the cytoplasm, and the tetraspanin LIMP1/CD63 (Fig. 2) (Eskelinen et al. 2003; Fukuda 1991; Hunziker and Geuze 1996). LAMP-1 and LAMP-2 are major protein components of the lysosomal membrane, which are estimated to contribute to about 50% of all proteins of the lysosomal membrane. LAMP-1 and LAMP-2 have similar domain structure and biochemical properties and possess 37% amino acid sequence homology but

416

*

E.-L. Eskelinen and A. A. Mironov

Figure 2. Schematic drawing demonstrating the structures of LAMPs, LIMP-1 and LIMP-2. The grey bar represents the lysosomal limiting membrane. N, amino terminus; C, carboxy terminus. The forks indicate N-linked carbohydrates. The number and position of the carbohydrates are hypothetical.

they are encoded by two separate genes that are localized on different chromosomes. LAMPs are transmembrane proteins with a large, heavily glycosylated luminal domain, one transmembrane domain, and a short Cterminal cytoplasmic tail.

Glycosylation of lysosomal membrane proteins Lysosomal membrane proteins are heavily glycosylated and contain predominantly N-linked oligosaccharides of the complex type. O-linked oligosaccharides are present as well, but to a smaller extent. N-glycosylation seems to be important for the stability of the proteins in the lysosomal membrane (Barriocanal et al. 1986; Kundra and Kornfeld 1999). LAMPs represent the major glycoproteins carrying poly-N-acetyllactosamine (Wang et al. 1991). The molecular mass of the polypeptide backbone of human LAMP-1 and LAMP-2 is 40–45 kDa, however after glycosylation the mass of the glycoproteins is approximately 120 kDa (Carlsson et al. 1988; Mane et al. 1989). Human LAMP-1 was estimated to have 18, and human LAMP-2 16, N-linked carbohydrate chains (Carlsson et al. 1988). The numbers of N-linked carbohydrate chains predicted from the cDNA sequences of LAMP-1 and LAMP-2 in different species range from 16 to 23. In addition, both LAMPs have O-linked carbohydrates in the “hinge” region of the luminal domain (Carlsson et al. 1993). LAMPs and LIMPs form a continuous carbohydrate lining on the inner leaflet of the lysosomal limiting membrane, generating a glycocalyx. Therefore, LAMPs were believed to function in the maintenance of the structural integrity of the lysosomal membrane by protecting it from the hostile luminal

Transport of lysosomal membrane proteins from the Golgi complex to lysosomes

*

417

environment. However, recent observations are inconsistent with this notion. Depletion of N-glycans with endoglycosidase H caused rapid degradation of LAMP-1 and LAMP-2, yet no changes in lysosomal integrity were noted (Kundra and Kornfeld 1999). Further, cells completely deficient in both LAMP-1 and LAMP-2 still have intact lysosomes (Eskelinen et al. 2004).

Transport of lysosomal membrane proteins from the Golgi to lysosomes Newly synthesized LAMPs and LIMPs are transported from the TGN to endosomes and lysosomes via two alternative routes: the intracellular route directly from the TGN to endosomes and from there to lysosomes, or the indirect route from TGN to the cell surface and from there via endocytosis to endosomes and finally to lysosomes (Fig. 3). There is consensus that both pathways contribute to the delivery of the LAMPs to lysosomes and that the fraction of the LAMPs expressed at the cell surface at steady state is low, e.g., 0–3% of the total LAMP-1 (Harter and Mellman 1992). However, estimates of the fraction of newly synthesized LAMPs that traffic via the plasma mem€ ning brane en route to lysosomes vary widely from 4 to 70% of the total (Ho

Figure 3. Schematic drawing of intracellular targeting routes of lysosomal membrane proteins. The direct transport route from the TGN to lysosomes is indicated by red arrows, and the indirect route from the TGN to the plasma membrane and then via endosomes to lysosomes is indicated by blue arrows. The main locations of the four adaptor protein complexes are indicated. EE, early endosome; LE, late endosome; Lys, lysosome.

418

*

E.-L. Eskelinen and A. A. Mironov

and Hunziker 1995). Recent results show that, unlike previously thought, about half of newly synthesized LAMPs pass through the plasma membrane on their way to lysosomes (Janvier and Bonifacino 2005). Lysosomal targeting is saturable (Harter and Mellman 1992), which suggests that it is receptor or coat dependent. Lysosomal membrane proteins contain one or more lysosomal targeting motifs in their cytosolic tails. The lysosomal targeting depends on either a tyrosine-based (LAMP-1, LAMP-2, LIMP1) or a di-leucine-based (LIMP-2) sorting signal (Hunziker et al. 1996; Le Borgne et al. 1998; Peters and Von Figura 1994). These sorting signals often conform to the consensus sequences YXXØ (tyrosine motif, where Ø is a _ hydrophobic amino acid) or EXXXLL (di-leucine motif; E may be replaced by D, and L by I or V), which both interact with tetrameric adaptor protein (AP) complexes and thereby mediate incorporation into transport vesicles (reviewed in Bonifacino and Glick (2004)). Like other tyrosine-based, YXXØ-type sorting signals, the lysosomal targeting motifs interact with the m subunits (i.e., m1, m2, m3, and m4) of the four heterotetrameric adaptor protein complexes (i.e., AP-1, AP-2, AP-3, and AP-4, respectively). AP-1, AP-2, and to some extent AP-3 are parts of coats that contain the scaffolding protein clathrin, whereas AP-4 is part of a non-clathrin coat (reviewed in Janvier and Bonifacino (2005)). AP-2 mediates clathrin-dependent endocytosis from the plasma membrane, whereas AP-1, AP-3, and AP-4 participate in protein sorting from the TGN and/ or endosomes to lysosomes. Lysosomal YXXØ-motifs are typically preceded by a glycine and located at the end of short carboxy terminal cytosolic tails. LAMPs have short (10–11 amino acid residues) C-terminal cytosolic tails that end with a GYXXØ motif (G is glycine, Y is tyrosine, and Ø is a bulky hydrophobic amino acid). Mutational studies have shown that the glycine is necessary for efficient direct transport of LAMP-1 from TGN to lysosomes, thus avoiding transport via the cell surface (Harter and Mellman 1992; Honing and Hunziker 1995). In addition to the amino acid sequence, also the exact placement of the motif relative to the transmembrane domain is critical for efficient biosynthetic targeting of the LAMPs to lysosomes (Rohrer et al. 1996). Changes in the amino acid composition of the targeting motif, and in the position of the motif relative to the transmembrane domain, result in increased expression of the LAMPs at the cell surface. A recent study showed that the AP-2 adaptors, associated with the plasma membrane coated pits, are required for efficient delivery of LAMPs to lysosomes, which supports the importance of the indirect transport route from TGN to lysosomes (Janvier and Bonifacino 2005). In this study the other adaptor proteins had a less important role in the lysosomal targeting. This recent result is supported by an older study showing that LAMP-1 and LAMP-2 molecules in the TGN are packed into vesicles that are distinct from clathrincoated vesicles containing MPR and AP-1 adaptor proteins (Karlsson and Carlsson 1998). Further, cells deficient in functional AP-1 exhibit normal localization of LAMPs to lysosomes (Meyer et al. 2000). In addition to the

Transport of lysosomal membrane proteins from the Golgi complex to lysosomes

*

419

indirect, AP-2 dependent route, LAMP-1, LIMP-2 and LIMP-1/CD63 can also be targeted from the TGN to lysosomes via an intracellular route that is dependent on the AP-3 adaptor complex (Honing and Hunziker 1995; Le Borgne et al. 1998; Rous et al. 2002). A di-leucine-based motif in the cytoplasmic tail of LIMP-2 and tyrosine mediate selective binding of AP-3 (Honing et al. 1998). Transport of newly synthesized LAMPs to lysosome takes a long time. Approximately 50–60% of LAMP-1, LAMP-2 and LIMP-1/CD63 was found in lysosomal fractions after 6 h of chase (Janvier and Bonifacino 2005). After translocation into the endoplasmic reticulum, lysosomal glycoproteins traverse the cisternae of the Golgi, where the majority of the oligosaccharide chains are converted to the complex type. About 30 min after synthesis half of the newly synthesized lysosomal glycoproteins have reached the trans-Golgi network (reviewed in Peters and Von Figura (1994)). This speed is substantially slower than that of other type I membrane glycoproteins, such as VSVG (Mironov et al. 2001).

Functions of LAMPs and LIMPs LAMP-2 undergoes alternative splicing that produces isoforms 2A, 2B, and 2C (Eskelinen et al. 2005; Hatem et al. 1995). LAMP-2A is predominant in the liver, while LAMP-2B is most abundant in skeletal muscle (Konecki et al. 1995). The cytoplasmic tail of LAMP-2A can interact with cytosolic proteins, such as RNAse A and GAPDH (Cuervo and Dice 1996). LAMP-2A is thought to act as a receptor in chaperone-mediated autophagy, a sequence motif-mediated, direct transport of cytosolic proteins through the lysosomal membrane for degradation (Cuervo and Dice 1998). RNAse A and GAPDH are both substrates of chaperone-mediated autophagy. LAMP proteins, LAMP-2 in particular, are required for fusion of lysosomes with phagosomes and autophagosomes. LAMP-1 and LAMP-2 are delivered to phagosomes and autophagosomes during their maturation and are needed for the recruitment of Rab7. In LAMP-deficient cells, phagosomes acquire early endosomal marker proteins but no Rab7, RILP, or other late endosomal €ger et al. 2004). and lysosomal markers (Binker et al. 2007; Huynh et al. 2007; Ja LAMP-2 mutations in humans cause a hereditary lysosomal disease called Danon disease that is characterized by accumulation of large autophagic vacuoles in heart and skeletal muscles, fatal cardiomyopathy, myopathy and mental retardation (Nishino et al. 2000). LAMP-2 deficient mice show the same phenotype (Tanaka et al. 2000). In addition to a role in lysosome motility, LAMP-2 also has a role in intracellular cholesterol traffic. Unesterified cholesterol accumulates in late endosomes and lysosomes of cells deficient in LAMP-2 or both LAMP-1 and LAMP-2 (Eskelinen et al. 2004). In addition to LAMP-2, also other lysosomal membrane proteins have been assigned with specific functions. LIMP-2 was recently shown to act as a receptor in the transport of newly synthesized b-glucocerebrosidase from the Golgi to lysosomes (Reczek et al. 2007). The recently discovered, specific functions of the lysosomal membrane proteins LAMP-2 and LIMP-2 show that,

420

*

E.-L. Eskelinen and A. A. Mironov

unlike previously assumed, lysosomal membrane proteins are not only structural components of the lysosomal membrane.

Contradictions on the transport of lysosomal membrane proteins It has been thought that most lysosomal membrane proteins are transported from the TGN to lysosomes via an intracellular route (Fukuda 1991; Harter and Mellman 1992; Honing and Hunziker 1995). This model predicts that lysosomal membrane proteins are sorted by adaptor proteins and packed into clathrin-coated vesicles in the TGN, and that the generated transport carriers fuse with endosomes or lysosomes. However, published data challenges this prediction. LAMP-1 and LAMP-2 are not enriched in clathrin-coated pits and vesicles derived from the TGN (Karlsson and Carlsson 1998). Analysis on Nycodenz gradients revealed that LAMP-carriers generated from the TGN in vitro were distinct from vesicles containing g-adaptin and MPR. Moreover, both types of carriers migrated differently from carriers containing proteins destined for the plasma membrane. Whereas wortmannin both in vitro and in vivo inhibited the production of g-adaptin/MPR-containing carriers, this drug had no effect on the generation of LAMP-carriers and on the sorting of LAMPs (Karlsson and Carlsson 1998). Thus it seems that during the incubation of the isolated Golgi with cytosol, three types of carriers are generated: LAMPcontaining carriers, g-adaptin/MPR-containing carriers, and carriers with proteins destined for the plasma membrane (Karlsson and Carlsson 1998). Furthermore, in the absence of AP-1, AP-3, and AP-4, LAMPs can be transported to lysosomes (Janvier and Bonifacino 2005; Reusch et al. 2002; Simmen et al. 2002), suggesting that transport of LAMPs is independent of AP-1, AP-3, and AP-4. As described above, a recent study showed that the AP-2 adaptors, associated with the plasma membrane coated pits, are required for efficient delivery of LAMPs to lysosomes (Janvier and Bonifacino 2005). Together these studies indicate that a major proportion of LAMPs are transported from the TGN to lysosomes via the plasma membrane in an AP-2 -dependent manner. Another contradiction concerns the basic idea of cytosolic targeting signals. Targeting information is thought to locate in the cytosolic tails of membrane proteins. These targeting signals then bind to different adaptor proteins which in turn mediate incorporation of the protein to a transport vesicle. This model is challenged by findings showing that targeting information in the cytosolic domain of TGN38, a trans Golgi membrane protein, could be overridden by the presence of the lumenal and transmembrane domains of LAMP-1 (lgp120) (Reaves et al. 1998). In contrast, the presence of the transmembrane and cytosolic domains of TGN38 was sufficient to deliver the lumenal domain of LAMP-1 to the trans-Golgi network. These results indicate that, in addition to the sorting signals in the cytosolic tails, also the transmembrane and luminal domains of membrane proteins contain sorting information. The authors propose that aggregation of the LAMP-1 luminal

Transport of lysosomal membrane proteins from the Golgi complex to lysosomes

*

421

domains in the acidic environment of endosomes might play a role in the subsequent targeting of the chimeric protein to lysosomes. However, the interpretation of the results is complicated by the fact the LAMP-1/TGN38 chimeras were expressed in cells that also express endogenous wild type LAMP-1 and TGN38. If the luminal domains of LAMP-1 form aggregates, these aggregates are likely to contain both wild type LAMP-1, containing the correct cytosolic targeting signals, and chimeric LAMP-1, containing the targeting signals of TGN38. However, the results demonstrate that targeting of membrane proteins to their correct locations is not solely dictated by the targeting motifs in the cytosolic tail. Also the transmembrane and luminal domains may play a role in correct intracellular targeting.

Abbreviations AP LAMP LIMP MPR TGN

adaptor protein lysosome-associated membrane protein lysosomal integral membrane protein mannose 6-phosphate receptor trans Golgi network

Acknowledgments. Work in the Eskelinen laboratory is funded by Biocentrum Helsinki, Helsinki University Foundations, The Academy of Finland, and The Ehrnrooth Foundation.

References Barriocanal JG, Bonifacino JS, Yuan L, Sandoval IV (1986) Biosynthesis, glycosylation, movement through the Golgi system, and transport to lysosomes by an N-linked carbohydrate-independent mechanism of three lysosomal integral membrane proteins. J Biol Chem 261: 16755–16763 Binker MG, Cosen-Binker LI, Terebiznik MR, Mallo GV, McCaw SE, Eskelinen EL, Willenborg M, Brumell JH, Saftig P, Grinstein S, Gray-Owen SD (2007) Arrested maturation of Neisseria-containing phagosomes in the absence of the lysosomeassociated membrane proteins, LAMP-1 and LAMP-2. Cell Microbiol 9: 2153–2166 Bonifacino JS, Glick BS (2004) The mechanisms of vesicle budding and fusion. Cell 116: 153–166 Butor C, Griffiths G, Aronson NN Jr, Varki A (1995) Co-localization of hydrolytic enzymes with widely disparate pH optima: implications for the regulation of lysosomal pH. J Cell Sci 108(Pt 6): 2213–2219 Cantalupo G, Alifano P, Roberti V, Bruni CB, Bucci C (2001) Rab-interacting lysosomal protein (RILP): the Rab7 effector required for transport to lysosomes. EMBO J 20: 683–689 Carlsson SR, Lycksell PO, Fukuda M (1993) Assignment of O-glycan attachment sites to the hinge-like regions of human lysosomal membrane glycoproteins lamp-1 and lamp-2. Arch Biochem Biophys 304: 65–73 Carlsson SR, Roth J, Piller F, Fukuda M (1988) Isolation and characterization of human lysosomal membrane glycoproteins, h-lamp-1 and h-lamp-2. Major sialoglycoproteins carrying polylactosaminoglycan. J Biol Chem 263: 18911–18919 Cuervo AM, Dice JF (1996) A receptor for the selective uptake and degradation of proteins by lysosomes. Science 273: 501–503

422

*

E.-L. Eskelinen and A. A. Mironov

Cuervo AM, Dice JF (1998) Lysosomes, a meeting point of proteins, chaperones, and proteases. J Mol Med 76: 6–12 De Duve C (1963) The lysosome concept. In: De Reuck AV, Cameron MP (eds.) Ciba foundation symposium on lysosomes. Churchill, London. pp. 1–31 Eskelinen EL, Cuervo AM, Taylor MRG, Nishino I, Blum JS, Dice JF, Sandoval IV, LippincottSchwartz J, August JT, Saftig P (2005) Unifying nomenclature for the isoforms of the lysosomal membrane protein LAMP-2. Traffic 6: 1058–1061 Eskelinen EL, Schmidt CK, Neu S, Willenborg M, Fuertes G, Salvador N, Tanaka Y, € llmann-Rauch R, Hartmann D, Heeren J, et al (2004) Disturbed cholesterol traffic Lu but normal proteolytic function in LAMP-1/LAMP-2 double deficient fibroblasts. Mol Biol Cell 15: 3132–3145 Eskelinen EL, Tanaka Y, Saftig P (2003) At the acidic edge: emerging functions for lysosomal membrane proteins. Trends Cell Biol 13: 137–145 Fukuda M (1991) Lysosomal membrane glycoproteins. Structure, biosynthesis, and intracellular trafficking. J Biol Chem 266: 21327–21330 Griffiths G, Gruenberg J (1991) The arguments for pre-existing early and late endosomes. Trends Cell Biol 1: 5–9 Griffiths G, Hoflack B, Simons K, Mellman I, Kornfeld S (1988) The mannose 6-phosphate receptor and the biogenesis of lysosomes. Cell 52: 329–341 Griffiths G, Matteoni R, Back R, Hoflack B (1990) Characterization of the cationindependent mannose 6-phosphate receptor-enriched prelysosomal compartment in NRK cells. J Cell Sci 95(Pt 3): 441–461 Harter C, Mellman I (1992) Transport of the lysosomal membrane glycoprotein lgp120 (lgp-A) to lysosomes does not require appearance on the plasma membrane. J Cell Biol 117: 311–325 Hatem CL, Gough NR, Fambrough DM (1995) Multiple mRNAs encode the avian lysosomal membrane protein LAMP-2, resulting in alternative transmembrane and cytoplasmic domains. J Cell Sci 108: 2093–2100 € ning S, Hunziker W (1995) Cytoplasmic determinants involved in direct lysosomal Ho sorting, endocytosis, and basolateral targeting of rat lgp120 (lamp-1) in MDCk cells. J Cell Biol 128: 464–473 Honing S, Hunziker W (1995) Cytoplasmic determinants involved in direct lysosomal sorting, endocytosis, and basolateral targeting of rat lgp120 (lamp-I) in MDCK cells. J Cell Biol 128: 321–332 Honing S, Sandoval IV, Von Figura K (1998) A di-leucine-based motif in the cytoplasmic tail of LIMP-II and tyrosinase mediates selective binding of AP-3. EMBO J 17: 1304–1314 Hunziker W, Geuze HJ (1996) Intracellular trafficking of lysosomal membrane proteins. Bioessays 18: 379–389 Hunziker W, Simmen T, Honing S (1996) Trafficking of lysosomal membrane proteins in polarized kidney cells. Nephrologie 17: 347–350 Huynh KK, Eskelinen EL, Scott CC, Malevanets A, Saftig P, Grinstein S (2007) LAMP proteins are required for fusion of lysosomes with phagosomes. EMBO J 26: 313–324 €ger S, Bucci C, Tanida I, Ueno T, Kominami E, Saftig P, Eskelinen EL (2004) Role for Rab7 Ja in maturation of late autophagic vacuoles. J Cell Sci 117: 4837–4848 Janvier K, Bonifacino JS (2005) Role of the endocytic machinery in the sorting of lysosome-associated membrane proteins. Mol Biol Cell 16: 4231–4242 Johansson M, Rocha N, Zwart W, Jordens I, Janssen L, Kuijl C, Olkkonen VM, Neefjes J (2007) Activation of endosomal dynein motors by stepwise assembly of Rab7-RILP-p150Glued, ORP1L, and the receptor betalll spectrin. J Cell Biol 176: 459–471 Jordens I, Fernandez-Borja M, Marsman M, Dusseljee S, Janssen L, Calafat J, Janssen H, Wubbolts R, Neefjes J (2001) The Rab7 effector protein RILP controls lysosomal

Transport of lysosomal membrane proteins from the Golgi complex to lysosomes

*

423

transport by inducing the recruitment of dynein–dynactin motors. Curr Biol 11: 1680–1685 Karlsson K, Carlsson SR (1998) Sorting of lysosomal membrane glycoproteins lamp-1 and lamp-2 into vesicles distinct from mannose 6-phosphate receptor/gamma-adaptin vesicles at the trans-Golgi network. J Biol Chem 273: 18966–18973 Konecki DS, Foetisch K, Zimmer KP, Schlotter M, Lichter-Konecki U (1995) An alternatively spliced form of the human lysosome-associated membrane protein-2 gene is expressed in a tissue-specific manner. Biochem Biophys Res Commun 215: 757–767 Kornfeld S, Mellman I (1989) The biogenesis of lysosomes. Annu Rev Cell Biol 5: 483–525 Kundra R, Kornfeld S (1999) Asparagine-linked oligosaccharides protect Lamp-1 and Lamp-2 from intracellular proteolysis. J Biol Chem 274: 31039–31046 Le Borgne R, Alconada A, Bauer U, Hoflack B (1998) The mammalian AP-3 adaptor-like complex mediates the intracellular transport of lysosomal membrane glycoproteins. J Biol Chem 273: 29451–29461 Ludwig T, Griffiths G, Hoflack B (1991) Distribution of newly synthesized lysosomal enzymes in the endocytic pathway of normal rat kidney cells. J Cell Biol 115: 1561–1572 Mane SM, Marzella L, Bainton DF, Holt VK, Cha Y, Hildreth JE, August JT (1989) Purification and characterization of human lysosomal membrane glycoproteins. Arch Biochem Biophys 268: 360–378 Meyer C, Zizioli D, Lausmann S, Eskelinen EL, Hamann J, Saftig P, Von Figura K, Schu P (2000) Mue1A-adaptin-deficient mice: lethality, loss of AP-1 binding and rerouting of mannose 6-phosphate receptors. EMBO J 19: 2193–2203 Mironov AA, Beznoussenko GV, Nicoziani P, Martella O, Trucco A, Kweon HS, Di Giandomenico D, Polishchuk RS, Fusella A, Lupetti P, et al (2001) Small cargo proteins and large aggregates can traverse the Golgi by a common mechanism without leaving the lumen of cisternae. J Cell Biol 155: 1225–1238 Murphy RF (1991) Maturation models for endosome and lysosome biogenesis. Trends Cell Biol 1: 77–82 Nishino I, Fu J, Tanji K, Yamada T, Shimojo S, Koori T, Mora M, Riggs JE, Oh SJ, Koga Y, et al (2000) Primary LAMP-2 deficiency causes X-linked vacuolar cardiomyopathy and myopathy (Danon disease). Nature 406: 906–910 Peters C, Von Figura K (1994) Biogenesis of lysosomal membranes. FEBS Lett 346: 146–150 Reaves BJ, Banting G, Luzio JP (1998) Lumenal and transmembrane domains play a role in sorting type I membrane proteins on endocytic pathways. Mol Biol Cell 9: 1107–1122 Reczek D, Schwake M, Schroder J, Hughes H, Blanz J, Jin X, Brondyk W, Van Patten S, Edmunds T, Saftig P (2007) LIMP-2 is a receptor for lysosomal mannose-6-phosphate-independent targeting of beta-glucocerebrosidase. Cell 131: 770–783 Reusch U, Bernhard O, Koszinowski U, Schu P (2002) AP-1A and AP-3A lysosomal sorting functions. Traffic 3: 752–761 Robinson JM, Karnovsky MJ (1991) Rapid-freezing cytochemistry: preservation of tubular lysosomes and enzyme activity. J Histochem Cytochem 39: 787–792 Robinson JM, Okada T, Castellot JJ Jr, Karnovsky MJ (1986) Unusual lysosomes in aortic smooth muscle cells: presence in living and rapidly frozen cells. J Cell Biol 102: 1615–1622 Rohrer J, Schweizer A, Russell D, Kornfeld S (1996) The targeting of Lamp1 to lysosomes is dependent on the spacing of its cytoplasmic tail tyrosine sorting motif relative to the membrane. J Cell Biol 132: 565–576 Rous BA, Reaves BJ, Ihrke G, Briggs JA, Gray SR, Stephens DJ, Banting G, Luzio JP (2002) Role of adaptor complex AP-3 in targeting wild-type and mutated CD63 to lysosomes. Mol Biol Cell 13: 1071–1082

424

*

E.-L. Eskelinen and A. A. Mironov

Simmen T, Honing S, Icking A, Tikkanen R, Hunziker W (2002) AP-4 binds basolateral signals and participates in basolateral sorting in epithelial MDCK cells. Nat Cell Biol 4: 154–159 € llmann-Rauch R, Janssen PML, Tanaka Y, Guhde G, Suter A, Eskelinen EL, Hartmann D, Lu Blanz J, Von Figura K, Saftig P (2000) Accumulation of autophagic vacuoles and cardiomyopathy in LAMP-2 -deficient mice. Nature 406: 902–906 Traub LM, Bannykh SI, Rodel JE, Aridor M, Balch WE, Kornfeld S (1996) AP-2-containing clathrin coats assemble on mature lysosomes. J Cell Biol 135: 1801–1814 Wang WC, Lee N, Aoki D, Fukuda MN, Fukuda M (1991) The poly-N-acetyllactosamines attached to lysosomal membrane glycoproteins are increased by the prolonged association with the Golgi complex. J Biol Chem 266: 23185–23190

Retrograde endosome-to-TGN Transport

*

425

Retrograde endosome-to-TGN transport Lei Lu and Wanjin Hong

Introduction Eukaryotic cells are characterized by the existence of many membranebound organelles with each having a unique structural composition and carrying out specific cellular functions. Many of these organelles are parts of the secretory and endocytic pathways and are integrated by dynamic interorganellar membrane trafficking events. Endocytosis is the process of uptaking of molecules from the cell surface into intracellular compartments, through which cargos at the plasma membrane (PM), such as solutes, lipids, receptors (and their bound ligands), toxins and viruses, are internalized by vesicles, which either fuse with existing endosomes or undergo homotypic fusion to form endosomes. Recent advances in this area have established the existence of multiple routes/modes of endocytosis and the resulting diverse types of endosomes. Thus, aside from the classic early endosome (EE), which is mostly derived from clathrin-dependent endocytosis and is characterized by the presence of Rab5 GTPase and PtdIns(3)P, there are endosomes generated from non-clathrin mediated endocytic routes such as “ARF6 endosome” and “caveolin endosome” (Pelkmans et al. 2001) from ARF6- and caveolae-dependent endocytosis, respectively. Endosomes of non-clathrin endocytic origin are less understood currently. Different types of endosomes can undergo heterotypic fusion to deliver cargos derived from non-clathrin-dependent endocytosis to the classic endocytic pathway (Naslavsky et al. 2003). The clathrin-dependent endocytic pathway is extensively researched. Typically, plasma membrane proteins with cytosolic sorting motifs, such as tyrosine-based motif of TGN38 and di-leucine-based motif of transferrin receptor (TfR) are internalized through the recognition of sorting motifs by adaptor protein complex 2 (AP2) to assemble the clathrin coat. Once endocytosed, the clathrin coated vesicles lose its coat and homotypically fuse to form the EE or fuse with existing EE (Fig. 1). From the EE, cargos are sorted into the following major trafficking routes: (I) membrane proteins such as TfR (or together with bound apo-transferrin) can be recycled back to the PM either directly (route 3) or through the recycling endosome (RE) (route 4); (II) receptor-ligand complexes, such as LDLR (low density lipoprotein receptor)– LDL complex and EGF receptor (EGFR)–EGF complex are sorted into the late endosome/multivesicular body (LE/MVB) and finally to the lysosome for degradation (routes 5–6); (III) In polarized epithelial cells, receptors

426

*

L. Lu and W. Hong

Figure 1. The general organization of the endocytic pathway and its link with the Golgi apparatus by two retrograde endosome-to-TGN trafficking events. After clathrin-mediated (route 1) or clathrin-independent (route 2) endocytosis from the plasma membrane, cargos are delivered to the early endosome. From the EE, cargos, such as transferrin receptor, can be directly recycled back to the PM (route 3) or indirectly through the recycling endosome (route 4). Other cargoes, such as EGF receptor and LDL receptor complexed with the respective ligand, are sorted to the late endosome (route 5) and subsequently to the lysosome (route 6). The LE is characterized by the existence of internal vesicles and is also referred to as multivesicular body. The endocytic pathway is linked to the TGN by two retrograde trafficking events: one, used by TGN38, is direct traffic from the RE (and, to lesser extents, possibly also EE) to the TGN (colored green); while the other, employed by furin, is from the LE to the TGN (colored pink). At the TGN, cargos are delivered to various post-Golgi destinations such as the PM (route 7), to the LE (route 8) or to the EE/RE (route 9). Additional retrograde transport from the TGN back to the endoplasmic reticulum is also indicated (route 10).

such as poly-IgA receptor complexed with poly-IgA, traverse the cell through transcytosis (reviewed by Tuma and Hubbard 2003); (IV) There are increasing number of proteins as exampled by TGN38 (or TGN46 in human) and furin that be delivered to the trans-Golgi network (TGN) of the Golgi apparatus (Gosh et al. 1998; Mallet et al. 1999) (routes indicated by heavy green and pink arrowed-lines, respectively). The endosome-to-TGN transport is emerging as one of the major pathways for both endogenous and exogenous proteins. Bacteria toxins, such as Cholera toxin and Shiga toxin, plant toxins such as ricin, and viruses such as SV40, have all evolved to be able to hijack this trafficking route to evade the disastrous fate of lysosomal degradation. The Golgi complex is a polarized stack of membrane cisternae. The cisternal maturation model proposes that anterograde secrectory cargos passively move forward along the Golgi stack as the cisternae matures

Retrograde endosome-to-TGN Transport

*

427

progressively from cis, medial to trans cisternae, and finally to the TGN, which is a major sorting station to route cargos to various post-Golgi compartments, such as endosomes, the PM and secretory granules; at the same time, the residential Golgi enzymes are recycled through the retrograde transport to maintain the polarized distribution of Golgi enzymes. Recently, classical secretory cargo VSV-G was found to traverse the RE before reaching the PM in polarized MDCK cells, suggesting the existence of constant membrane flow from the TGN to the RE (Ang et al. 2004). In yeast, all known late Golgi membrane proteins cycle through endocytic compartments (Wilsbach and Payne 1993; Bryant and Stevens 1997; Lewis et al. 2000). The endosome-to-TGN transport can be viewed as a recycling process of maintaining TGN in the face of constant anterograde trafficking. This is exampled by the transport of mannose-6-phosphate receptor (M6PR). Newly synthesized lysosomal hydrolases are post-translationally modified to contain the mannose-6-phosphate sorting signal. At the TGN, the hydrolases are selectively recognized by the M6PRs and sorted to the endosome, where acidic environment causes dissociation of bound hydrolases from M6PRs. The hydrolases are delivered to the lysosome, whereas the M6PRs are recycled from the endosome backed to the TGN via the endosome-to-TGN retrograde trafficking events (Dahms et al. 1989; Kornfeld 1989). The retrieval of M6PRs from the endosome is crucial for continuous delivery of lysosomal hydrolases, as a failure in recycling causes M6PRs to be delivered to the lysosome and degraded, resulting in a defect in lysosomal delivery of hydrolases. Thus, the endosome-toTGN trafficking is an essential pathway for eukaryotic cells to retrieve lipids and proteins from endosomes to maintain the identity and functionality of the TGN. Elucidating the molecular mechanisms governing endosome-to-TGN traffic will greatly contribute to our understanding of the intracellular fate of lipids, receptors, toxins, viral pathogens and therapeutic reagents. One example is the proteolytic generation of Ab peptide, the formation and aggregation of which are involved in neuropathogenesis of Alzheimer’s disease (AD). Ab peptide is produced by sequential cleavage of amyloid precursor protein (APP) by b- and g-secretases, which is abrogated by a-secretase (reviewed by Annaert and De Strooper 2002). These secretases and their substrate, APP, are potentially shuttling among the PM, endosomes and the TGN (Lammich et al. 1999; Huse et al. 2000; Rechards et al. 2006; Stephens and Austen 1996). The endosome-to-TGN trafficking is probably instrumental in regulating Ab peptide production. Knowing the mechanism of endosome-to-TGN transport will likely help us design new therapeutic strategies to combat AD. The molecular mechanisms governing endosome-to-TGN transport is now beginning to be revealed, with the increasing recognition of the importance of this trafficking process and recent technical advances, especially in vitro transport assays and live cell imaging techniques.

428

*

L. Lu and W. Hong

Two parallel routes of endosome-to-TGN transport There are at least two trafficking itineraries from the EE to the TGN, represented by two TGN proteins – TGN38 and furin (Ghosh et al. 1998; Mallet et al. 1999): one involving the RE, while the other employing the LE (Fig. 1).

RE–TGN TGN38 (or TGN46 in human) is a TGN-enriched type I transmembrane protein cycling between the TGN and the PM. TGN38 is internalized by clathrindependent process from the PM through the recognition of its tyrosine-based sorting motif by AP2 and is delivered to the EE and then the RE before being recycled back to the TGN (Ghosh et al. 1998). The RE, which can be labeled by the TfR, is enriched in the pericentriolar region and in close proximity to the Golgi apparatus in many cell types (Hopkins and Trowbridge 1983; Yamashiro and Maxfield 1984; Lippincott-Schwartz et al. 1991). TGN38 slowly leaves the TGN for the PM to complete the cycling journey. This slow kinetics maintains the predominant localization of TGN38 at the TGN at the steady state. Although it is possible that a fraction of TGN38 may travel directly from the EE to the TGN, it is generally believed that RE–TGN route is the principle one. CI (cation-independent)-M6PR/insulin-like growth factor receptor/ MPR300 and CD (cation-dependent)-M6PR/MPR47 are among the most widely used makers for post-Golgi trafficking studies. In addition to their presence in the LE and TGN, they are also found on the PM, EE, and RE (Willingham et al. 1981; Geuze et al. 1984; Griffiths et al. 1988; Press et al. 1998). Although CIM6PR is an example of protein traveling via the other LE–TGN retrograde transport (Dahms et al. 1989), recent studies suggest that it also uses the RE– TGN retrograde transport (Lin et al. 2004) as the trafficking of CD-M6PR depends on the same protein machinery as in EE/RE–TGN pathway (Medigeshi and Schu 2003). There is a growing list of cargos employing this RE–TGN retrograde pathway and they are summarized in Table 1. One example is a fourtransmembrane protein GLUT4 (glucose transporter 4). In fat and muscle cells, GLUT4 actively mediates glucose uptake at the PM upon stimulation of insulin. In 3T3-L1 adipocytes, GLUT4 is internalized through both di-leucine and tyrosine-based signals by clathrin-dependent endocytosis, and subsequently transported to the TGN via the EE/RE (Shewan et al. 2003). Specific lipids can also follow this retrograde pathway (Table 1). Some sphingolipids, such as globotriaosyl ceramide (Gb3), glucosylceramide and sphingomyelin, use this pathway to reach the Golgi apparatus as shown by their fluorescent analogs and sphingolipid binding toxin—STxB and CTxB (Mallard et al. 1998;  et al. 2001), although there are other sphingolipids, such as lactosylcerBabia amide, use the other LE–TGN pathway (Choudhury et al. 2002). Exogenous proteins, including protein toxins, such as Shiga toxin, Cholera toxin, ricin, have evolved ways to usurp this pathway for entering cells without encountering the degradative LE/lysosome. Shiga toxin and Cholera

Retrograde endosome-to-TGN Transport

*

429

Table 1. A summary of currently known endogenous proteins, lipids and exogenous cargos that utilize endosome-to-TGN trafficking pathways, which include two routes: EE/RE–TGN and LE-TGN. “Endo-TGN” indicates the detailed pathway is not clear Endogenous proteins TGN38 (TGN46) M6PRs (CI- or CD-)

EE/RE–TGN EE/RE–TGN

GLUT4 SCAMP1–4 GPP130 GPP73 GS15 Vamp4 b-secretase Furin Sortilin APP SorLA/LR11 LAT

EE/RE–TGN EE/RE–TGN EE/RE–TGN EE/RE–TGN EE/RE–TGN EE/RE–TGN EE/RE–TGN LE–TGN Endo-TGN Endo-TGN Endo-TGN Endo-TGN

(Gosh et al. 1998) (Lin et al. 2004; Medigeshi and Schu 2003) (Willingham et al. 1981; Geuze et al. 1984; Griffiths et al. 1988; Press et al. 1998) (Shewan et al. 2003) (Castle and Castle 2005) (Puri et al. 2002) (Puri et al. 2002) (Tai et al. 2004) (Tran et al. 2007) (Huse et al. 2000) (Mallet et al. 1999) (Nielsen et al. 2001) (Stephens and Austen 1996) (Anderson et al. 2005) (Brignatz et al. 2005)

Lipids Globotriaosyl ceramide Gb3 Glucosylceramide Sphingomyelin Lactosylceramine

EE/RE–TGN EE/RE–TGN EE/RE–TGN LE–TGN

(Mallard et al. 1998)  et al. 2001) (Babia  et al. 2001) (Babia (Choudhury et al. 2002)

Exogenous cargoes Shiga toxin Cholera toxin Ricin

EE/RE–TGN EE/RE–TGN EE/RE–TGN

Nef-MHC-I complex Pseudomonas Exotoxin A gp41 (HIV-1)

EE/RE–TGN Endo-TGN LE–TGN

(Mallard et al. 1998) (our unpublished result) (Iversen et al. 2001; Moisenovich et al. 2004) (Blagoveshchenskaya et al. 2002) (Smith et al. 2006) (Blot et al. 2003; Murray et al. 2005)

LE–TGN

CI Cation-independent; CD cation-dependent; M6PR mannose-6-phosphate receptor; GLUT4 glucose transporter 4; SCAMP1–4 secretory carrier membrane proteins 1–4; APP amyloid precursor protein; LAT Linker of activated T cells; EE early endosome; RE recycling endosome; LE later endosome; endo endosome; TGN trans-Golgi network.

toxin are AB5 structured protein toxins secreted from bacteria Shigella dysenteriae and Vibrio cholerae, respectively (reviewed by Sandvig and Van Deurs 2002). Both toxins are composed of two non-covalently linked subunits: A-subunit and homopentameric B subunits. Ricin, a plant toxin from castor bean, consists of a single A subunit and a single B subunit. In these three toxins, the A subunit has the toxic enzymatic activity upon translocation into the cytosol while the B subunit has the ability to govern the intracellular

430

*

L. Lu and W. Hong

trafficking of the toxins. Shiga toxin and Cholera toxin B-subunits (STxB and CTxB) are responsible for binding to the cell surface receptor—Gb3 and ganglioside GM1, respectively, which are sphingolipids presumably residing in lipid microdomain (Falguieres et al. 2001; Nichols et al. 2001) and directing the holotoxins to follow the retrograde traffic from the PM to the endosomes and then the TGN (Sandvig et al. 1992). A significant amount of STxB and CTxB enters cells through clathrin-dependent endocytosis, although multiple clathrin-independent routes were also reported (Sandvig et al. 1989; Torgersen et al. 2001; Nichols et al. 2001; Saint-Pol et al. 2004). In HeLa cells, a high percentage of STxB is found in the TfR positive EE and RE shortly after internalization (Mallard et al. 1998, 2002). From EE/RE, STxB is rapidly delivered to the TGN (Mallard et al. 1998). CTxB probably utilizes the similar pathway. The B subunit of ricin binds to the cell surface glycoproteins or glycolipids containing galactose and is also delivered to the Golgi through endosome-to-TGN trafficking (reviewed by Sandvig and Van Deurs 2002). Ricin may use both the RE and LE for transport to the TGN (Iversen et al. 2001; Moisenovich et al. 2004). From the Golgi apparatus, STxB, CTxB and ricin are slowly transported to the endoplasmic reticulum (ER) for exporting their A subunits into the cytosol (reviewed by Sandvig and Van Deurs 2002). Because of their trafficking itineraries, STxB, CTxB and ricin become invaluable tools for the studying of retrograde endosome-to-TGN transport. Viral pathogens, such as HIV-1, can hijack this pathway to sabotage cellular defense system (Blagoveshchenskaya et al. 2002). MHCI is a type I transmembrane protein serving as an antenna to present cytosolic pathogen’s peptide on the cell surface. In normal cells, MHCI is internalized by non-clathrin, dynamin-independent, ARF6-dependent endocytic pathway to “ARF6 endosome”, and then recycled back to PM. To evade the surveillance of host immune system, HIV-1 is able to down-regulate the surface MHCI. By recruiting PACS-1 and phosphatidylinositol 3-kinase, HIV-1 Nef protein uses RE–TGN pathway to divert MHCI to the TGN, thus reducing the surface level of MHCI. The RE–TGN route is also conserved in lower eukaryotes such as the budding yeast Saccharomyces cerevisiae. In yeast, all late Golgi (equivalent to TGN of mammalian cells) membrane proteins cycle through endocytic compartments (Wilsbach and Payne 1993; Bryant and Stevens 1997; Lewis et al. 2000). Many of them, such as Snc1p and Chs3p, follow the itinerary from the EE/RE to the late Golgi and often serve as markers for this pathway (Holthuis et al. 1998; Lewis et al. 2000; Valdivia et al. 2002).

LE–TGN Furin is a type I transmembrane protein with endopeptidase activity responsible for proteolytic processing of secretory and membrane proteins at the TGN. From the EE, furin employs the LE en route to the TGN (Mallet et al. 1999) (Fig. 1). Like TGN38, furin slowly flows out of the TGN to the PM, resulting in its steady state accumulation in the TGN. As mentioned above, M6PRs use the LE–TGN pathway (in addition to the RE–TGN route) for retrieval after

Retrograde endosome-to-TGN Transport

*

431

releasing hydrolases at the LE. The sorting of M6PRs probably takes place at the lysobisphosphatidic acid (LBPA)-enriched internal membrane of LE/MVB (Kobayashi et al. 1998). Sortilin may travel with the same itineraries as CIM6PR using both RE–TGN and LE–TGN routes. In fact, the transmembrane and cytosolic domains of sortilin can substitute for those of CI-M6PR to generate a chimeric protein, which fully restores the lysosomal trafficking of soluble hydrolases in CI-M6PR-deficient cells (Nielsen et al. 2001). Endogenous glycosphingolipids, such as lactosylceramide, can also travel from the LE to the TGN (Choudhury et al. 2002). Some retroviruses have evolved elaborate mechanisms to take advantage of the LE–TGN pathway in order to assemble virion particles. A good example is HIV-1. HIV-1 envelope glycoprotein (Env) consists of two subunits: receptor binding subunit gp120 and membrane anchoring subunit gp41. In transfected or infected HeLa cells, HIV-1 transmembrane protein gp41 was found to cycle continuously between the cell surface and the TGN with steady state enrichment in the TGN (Blot et al. 2003). A tyrosine-based motif of gp41 is responsible for clathrin-dependent internalization. Subsequently, gp41 is delivered to the TGN via the LE (Blot et al. 2003; Murray et al. 2005). Other enveloped viruses, such as Ebola, Marburg, and measles virus may also employ the LE-TGN transport for their assembly (Murray et al. 2005). Yeast has an equivalent route from the pre-vacuolar compartment (PVC, the late endosome in yeast) to the late Golgi. Late Golgi membrane proteins, such as Kex2p (functional ortholog of furin), dipeptidyl aminopeptidase-A (DPAP-A) and Vps10p, cycle between the PVC and the late Golgi and often serve as markers for LE–TGN transport (Piper et al. 1995). Vps10p is the receptor for vacuolar hydrolase CPY (carboxy peptidase Y) and its role is functionally equivalent to M6PRs in the mammal. Vps10p cycles between the late Golgi and PVC to deliver CPY to the vacuole. Many yeast mutants which cause missorting of CPY, in fact have disrupted retrograde trafficking of Vps10p from the PVC to the late Golgi, resulting in the delivery and degradation of Vps10p in the vacuole. In addition to M6PRs and sortlin, other proteins may also use both RE–TGN and LE–TGN routes. For instance, the AB type bacteria toxin—Pseudomonas Exotoxin A has recently been shown to reach the TGN via both pathways (Smith et al. 2006). The pathway that a protein adopts may be dependent on cell type. One example is provided by STxB, which is delivered to the lysosome for degradation in human monocytes, while it escapes the LE and is delivered to ER via RE–TGN in HeLa cells (Falguieres et al. 2001).

Models for endosome-to-TGN transport Similar to transport in the secretory pathway, endosome-to-TGN transport is believed to use shuttling intermediates in the form of vesicles or tubules, and/or to employ the process of compartment maturation. The vesicle-mediated endosome-to-TGN transport is supported by the observation

432

*

L. Lu and W. Hong

Figure 2. The general model of vesicle-mediated transport envisions four major steps. During the budding (I), the transmembrane receptor captures the soluble cargo in the lumen of a donor compartment. The cytoplasmic domain of the cargo receptor or membrane cargoes somehow promotes the GDP-to-GTP exchange of small GTPases of the ARF/Arl/Sar family via guanine nucleotide exchange factor. GTP-loaded GTPase coordinates the recruitment of coat proteins to drive the formation of coated vesicles. Upon/during the completion of the budding, the GTPase is inactivated by the hydrolysis of its bound GTP, resulting in the shedding of coat proteins. Protein machinery important for subsequent events, such as v-SNARE and motor proteins, is also incorporated into the vesicle. During the translocation/targeting (II) process, cytoskeletal network facilitates the movement of vesicles to the vicinity of the acceptor compartment. The tethering (III) process, which is mediated by GTPases of the Rab family and tethering factors, ensures that the vesicle is positioned precisely in the acceptor compartment. During the docking/ fusion (IV), the v-SNARE on the transport vesicle pairs with the t-SNARE complex on the target compartment to drive the fusion of the vesicle with the target compartment, releasing the cargo proteins to the acceptor compartment. The dissociation of the v–t SNARE complex is powered by ATP hydrolysis of NSF/a-SNAP, so that the v-SNARE and the cargo receptor can be recycled back to the donor compartment.

of Rab9-containing vesicles (Barbeto et al. 2002). However, tubule-mediated transport and/or membrane maturation may act in parallel. Vesicle-mediated transport generally involves four major steps (Fig. 2). The 1st step (I), budding, is the biogenesis of a vesicle from a donor compartment. It involves cargo selection and vesicle budding, executed by the coat proteins and regulatory proteins. In this step, the GTP-bound form of ARF family small GTPase such as Sar1 and ARF1 recruits coat proteins, such as COPI, COPII or AP-1 onto the membrane (reviewed by Scales et al. 2000; Antonny and Schekman 2001; Boehm and Bonifacino 2001). Both coats and adaptors select the transmem-

Retrograde endosome-to-TGN Transport

*

433

brane cargos by interacting with their cytosolic motifs. After the completion of the budding, the GTPase is inactivated by the hydrolysis of its bound GTP, resulting in the shedding of coats. In the 2nd step (II), targeting, the vesicles move along the actin or microtubule cytoskeleton to the vicinity of the acceptor compartment. The next step is tethering (III), in which tethering molecules (or complexes) on the acceptor compartment anchor vesicles in the precise vicinity of an acceptor compartment (Pfeffer 1999). Rab small GTPases and their effectors regulate both step II and III (Zerial and McBride 2001). The tethering molecules loosely attach to the vesicle at relative long distances, ensuring the last step (IV), docking and fusion, in which the content of vesicle is released into the lumen of acceptor compartment due to fusion of the vesicle with the acceptor compartment. SNARE (soluble NSF attachment protein receptor) family and Sec1/Munc18 (SM) proteins are key players of the docking and fusion reactions (Chen and Scheller 2001; Hong 2005). The endosome-to-TGN transport uses this general mode of vesicle-mediate traffic for transporting cargoes and/or for maturing a compartment.

The methodology to study endosome-to-TGN transport As in other membrane trafficking steps, in vitro transport assays reconstituting endosome-to-TGN traffic have proven to be instrumental for the identification of essential players. The commonly used endosome-to-TGN transport assays utilize enzymatic reactions specific to the TGN. In the earliest version, a unique trans-Golgi/TGN enzymatic reaction, which is the addition of sialic acid to M6PRs, was utilized (Goda and Pfeffer 1988). The endosomes containing M6PRs lacking sialic acid were isolated from sialyltransferase-negative mutant CHO cells and incubated with wild type Golgi membranes. The acquisition of sialic acid by M6PRs indicates the transport from the endosome to the TGN. The recently developed assays all took advantage of tyrosine sulfation, an enzymatic reaction which is carried out by tyrosyl protein sulfotransferase (TPST) in the lumen of TGN (reviewed by Moore 2003). The sulfation reaction has been shown to be much faster than the overall transport process, which makes this reaction suitable for monitoring the TGN arrival of test proteins bearing tyrosine sulfation sites (Itin et al. 1997). A reporter protein can be engineered to contain artificial tyrosine sulfation sites at the luminal domain of protein traveling from endosomes to the TGN, such as CD-M6PR, STxB or ricin (Itin et al. 1997; Johannes et al. 1997; Tai et al. 2004; Iversen et al. 2001). The STxB-based RE–TGN transport assay is illustrated in Fig. 3. After synchronization of the reporter protein at the RE, the plasma membrane of sulphate-starved cells is selectively permeabilized. Endosome-to-TGN transport is reconstituted by supplying S35 sulfate, ATP regenerating system, exogenous cytosol and testing antibody or recombinant protein. The amount of S35 sulfate incorporated into STxB reflects the extent of transport of STxB from the RE to the TGN. Chimeric proteins consisting of the lumenal (extracellular) domain of Tac or CD8, transmembrane and cytoplasmic domains of recycling proteins,

434

*

L. Lu and W. Hong

Figure 3. Schematic diagram illustrating STxB-based in vitro EE/RE–TGN transport assay. When recombinant STxB (drawn as homo-pentamer) is added to the medium of cultured cells, it binds to the cell surface receptor (a certain type of sphingolipid) and is then internalized and delivered to the EE. At 18 C, the EE/RE–TGN traffic is arrested, resulting in accumulation of internalized STxB in the EE/RE. After selective perforation of the PM, the semi-intact cells are supplied with exogenous cytosol, ATP regenerating system, and 35S-sulfate in the presence of reagents such as antibody or fusion protein. Upon incubation at 37 C, a synchronized transport of STxB from the EE/RE to the TGN is reconstituted. Upon delivery to the lumen of the TGN, tyrosyl protein sulfotransferase (TPSP), a TGN-specific enzyme, incorporates 35S-sulfate into the engineered sulfation sites of STxB. The extent of sulfation of STxB serves as a measure for EE/RE–TGN transport.

such as CI-M6PR, TGN38, furin and sortilin, are also widely used in this area (Ghosh et al. 1998; Mallet et al. 1999; Seaman 2004; Hirst et al. 2004). The vast selection of antibodies against Tac and CD8 makes the labeling and tracking of these chimeras convenient in cultured mammalian cells. Experiments have shown that the itineraries of these chimeric proteins are faithfully directly by the cytosolic targeting motifs of the recycling proteins.

The molecular machinery of endosome-to-TGN transport The molecular machinery of endosome-to-TGN transport is evolutionally conserved. The powerful yeast genetics has revealed many components regulating endosome-to-TGN transport, which greatly aided the discovery and characterization of mammalian counterparts. Since the two trafficking directions (anterograde and retrograde) between endosomes and the TGN are tightly coupled, inhibition of transport in either direction could indirectly lead to the disruption of the other. For example, deletion of Vps35p disrupted the delivery of CPY from the late Golgi to the PVC, due to actual inhibition of recycling of CPY receptor, Vps10p (Seaman et al. 1997). Thus multiple research approaches are necessary to confirm a protein’s role in endosome-to-TGN

Retrograde endosome-to-TGN Transport

*

435

transport. As in other transport events, machinery of endosome-to-TGN transport involves SNAREs, coat proteins, small GTPases (Rabs and Arfs) and tethering proteins.

SNAREs Most SNAREs are tail-anchored membrane proteins (reviewed by Hong 2005). According to the SNARE hypothesis, v-SNARE on the vesicle specifically pairs with its cognate t-SNARE on the target compartment. The pairing of v–t SNAREs brings the two sheets of membranes in such a close proximity that the fusion event occurs subsequently. Testing the role of Golgi localized SNAREs in endosome-to-TGN transport using the assays described above has been fruitful (Mallard et al. 2002; Tai et al. 2004; Wang et al. 2005). In mammlian cells, there are at least 12 SNAREs residing on different compartments of the Golgi complex. These include syntaxin 5, syntaxin 6, syntaxin 16, GS27/membrin, Sec22b/ERS24, Bet1, GS28/ GOS28, Ykt6, Vamp4, GS15, Vti1a, GS32/SNAP-29, and syntaxin 10 (reviewed by Hong 2005). A screening of the TGN localized SNAREs identified that syntaxin 6, syntaxin 16, Vti1a, Vamp4 (and Vamp3) participate in RE–TGN transport of STxB (Mallard et al. 2002). Syntaxin 6, syntaxin 16 and Vti1a are believed to form a t-SNARE complex and uses either Vamp3 or Vamp4 as its cognate v-SNARE on the vesicle. Syntaxin 16 is also important for TGN delivery of TGN38 and CI-M6PR (Saint-Pol et al. 2004). Syntaxin 5, GS28, Ykt6, GS15, GS27, Sec22 and Bet1 SNAREs have been implicated in the ER–Golgi or early stage of intra-Golgi transport (Hay et al. 1998; Zhang et al. 2001; Xu et al. 2002). A systematic screening of the remaining Golgi SNAREs revealed that syntaxin 5/GS28/Ykt6/GS15 SNARE complex is also involved in STxB-based EE/RE–TGN transport (Tai et al. 2004). Syntaxin 5 and GS28, originally thought to be enriched in the early Golgi cisternae, are now shown to be present in every Golgi cisternae (Hay et al. 1998; Orci et al. 2000; Volchuk et al. 2004). GS15 is recently shown to have an increasing concentration across the cisternae toward the trans-Golgi (Volchuk et al. 2004) and may cycle between endosomes and TGN (Tai et al. 2004). The finding that anti-syntaxin 5 and 16 antibodies had additive inhibitory effect in vitro suggests that there exist at least three populations of vesicles from the RE having Vamp3, Vamp4 or GS15 as the v-SNARE for fusion with the TGN (Mallard et al. 2004; Tai et al. 2004). In yeast, Tlg1p, Tlg2p and Gos1p are functional homologs to mammalian syntaxin 6, 16 and GS28, respectively. Yeast lacking Tlg1p, Tlg2p or Gos1p exhibits defects in delivery of Snc1p from the EE to the late Golgi (Lewis et al. 2000; Siniossoglou and Pelham 2001) and in retrieval of TGN resident proteins such as Kex2p (Holthuis et al. 1998). Yeast components of the Sed5p (ortholog of syntaxin 5) complex have also been found to participate in protein recycling from endosomes to the Golgi. Gos1p, Ykt6p (ortholog of Ykt6), and a dominant activator of Sed5p, Sly1p, function as multi-copy suppressors of ric1D and ypt6D cells that exhibit defects in protein retrieval from the endosomes to the TGN (Bensen et al. 2001).

436

*

L. Lu and W. Hong

NSF and a-SNAP are essential regulators of SNAREs by dissociating SNARE complex after fusion. They are required for in vitro endosome–TGN transport (Itin et al. 1997; Mallard et al. 2002). The SM proteins bind SNAREs directly to regulate the functionality of SNAREs. In yeast, studies have shown that the SM protein, Vps45p, participates in the assembly of the TGN SNARE complex by binding to Tlg2p (Syntaxin 16) (Bryant and James 2001). Sly1p and Vps45p may regulate syntaxin 5 and 16 SNARE complexes, respectively, in endosome–TGN transport.

Coat proteins COPI COPI coat consists of seven subunits coatomers: a, b, b0 , g, d, e and z-COP (Rothman and Wieland 1996). Although COPI functions primarily at the early secretory pathway, some subunits such as b and e COPs, are reported to be on endosomes and appear to participate in sorting cargos out of the EE and the biogenesis of MVB (Aniento et al. 1996; Gu et al. 1997; Daro et al. 1997; Piguet et al. 1999). The role of COPI in endosome-to-TGN pathway is implied in yeast. Yeast with mutant a- or g-COP accumulates Snc1p in the EE (Lewis et al. 2000). Yeast COPI has been shown to have a weak interaction with TRAPP II complex, a potential tethering complex in RE–TGN transport (Cai et al. 2005).

Clathrin coat Clathrin coat has been observed on the PM, TGN and endosomes (Stoorvogel et al. 1996). The evidence supporting clathrin’s involvement in retrograde transport to the TGN comes from the finding that interference of clathrin either by siRNA or antibody prevents transport of STxB from the EE/RE to the Golgi (Saint-Pol et al. 2004). Clathrin itself does not have intrinsic affinity for membranes. Adaptor proteins are responsible for connecting clathrin coat with cargos. Adaptor proteins 1–4 (APs) are heterotetromeric protein complex. AP1 is localized to endosomes and the TGN. Recruitment of AP1 to the membrane is dependent of ARF1-GTP, as is the case of COPI coats and GGAs (gamma-adaptin homology-Golgi-associated Arf-binding) (Dittie et al. 1996). It is most plausible that AP1 and clathrin coat participate in both anterograde and retrograde directions of transport between endosomes and the TGN. AP1 clathrin coat (together with GGA) may act at the TGN to sort membrane proteins such as M6PRs to endosomes in anterograde manner. AP1 and CI-M6PR are found on clathrin-coated TGN membrane (Ahle et al. 1988; Klumperman et al. 1993). The transport of Lamp1 from the TGN to the lysosome is abolished when its AP1 interacting motif is mutated (Honing et al. 1996). In live cell imaging, AP1 was detected on the CI-M6PR positive transport intermediate en route from the TGN to the endosomes (Waguri et al. 2003). Recent experiments also indicate a role of AP1 and clathrin in retrograde trafficking. Mammalian cells with inactivated μ1A gene have compromised retrieval of CD-M6PR back to the Golgi, implying that AP1 functions in the retrograde trafficking from the RE and/or LE to the TGN

Retrograde endosome-to-TGN Transport

*

437

(Meyer et al. 2000). Furthermore, upon siRNA-mediated depletion of AP1, CD8-furin chimera relocates to peripheral endosomes and the PM, instead of the TGN (Hirst et al. 2004). Supporting that view, yeast Chs3p, which cycles between the EE and the late Golgi, relocates to the PM when AP1 is disrupted (Valdivia et al. 2002). Interestingly, AP1 depletion study suggests that it is not essential for the retrograde trafficking of STxB to the TGN (Saint-Pol et al. 2004), which implies that AP1 may selectively transport a subset of cargos from the endosome to the TGN. AP1 may work through the “connector” protein, phosphofurin acidic cluster sorting protein 1 (PACS1), which simultaneously interacts with AP1 and an acidic residue-cluster targeting motif on cargo proteins (Wan et al. 1998; Crump et al. 2001). Acidic clusters recognized by PACS1 have been found in the cytoplasmic domain of an increasing number of TGN membrane proteins, including furin, PC6B, M6PRs, polycistin-2 (TRP family cation channel), nephrocystin, VAMP4, herpes virus envelope glycoprotein and HIV-1 Nef (Wan et al. 1998; Piguet et al. 2000; Xiang et al. 2000; Crump et al. 2003; Hinners et al. 2003; Kottgen et al. 2005; Schermer et al. 2005; Tran et al. 2007). The acidic clusters often contain consensus sites for phosphorylation by casein kinase 2 (CK2), and dephosphorylation by protein phosphatase 2A (PP2A) (Wan et al. 1998; Molloy et al. 1998). In many cases, the binding of PACS1 requires phosphorylation of the acidic cluster. The phosphorylatable acidic clusters of furin and M6PRs are crucial for their retrieval from endosomes, as mutations of these regions abolish their TGN localization (Wan et al. 1998). Monomeric adaptor proteins such as GGAs, epsins and AP180 are able to connect cargos with clathrin coats. By interacting with clathrin, the ear domain of AP2 complex, phosphoinositides and sorting signals of cargos, epsins function as cargo adaptors in mediating clathrin-dependent endocytosis on the PM (Bonifacino and Traub 2003). EpsinR (epsin related) is an epsin family protein present on the Golgi and endosomes (Mills et al. 2003; Hirst et al. 2003; Saint-Pol et al. 2004). EpsinR is a multivalent adaptor connecting clathrin and cargo. EpsinR can bind both g ear of AP1 and PtdIns4P. The binding of PtdIns4P, which is enriched at TGN, is mediated by its ENTH domain (epsin N-terminal homology). EpsinR is proposed to be an adaptor facilitating clathrin-dependent transport from the endosome to the TGN, as depletion of epsinR inhibited the retrograde trafficking of STxB, TGN38 and CI-M6PR to the TGN (Saint-Pol et al. 2004). Vti1b, a SNARE involved in late endosome fusion, but not its isoform Vti1a, was found to specifically bind epsinR (Hirst et al. 2004). Depleting epsinR caused the redistribution of Vti1b from the TGN to peripheral endosomes, suggesting that epsinR is an adaptor protein for cargos, such as Vti1b, to recycle from endosomes to the TGN.

GTPases Many trafficking events are generally regulated by GTPases. The involvement of GTPases in endosome-to-TGN transport is demonstrated by in vitro transport assay, in which GTPgs is a potent inhibitor for the STxB transport (Tai et al.

438

*

L. Lu and W. Hong

2004). GTPases can be classified as heterotrimeric GTPases, such as Gabg, and monomeric GTPases, which are further subclassified into large, such as dynamin, and Ras-like small GTPases. The Ras-like small GTPases consist of five families: Ras, Rab, Rho, ARF, and Ran. GTPases switch between GTP-bound active and GDP-bound inactive states. The activation of a small GTPase, which is the exchange of GDP to GTP, is dependent on the guanine nucleotide exchange factors (GEFs); while the inactivation process – the hydrolysis of bound GTP to GDP – is facilitated by GTPase activating proteins (GAPs). When in GTP-bound form, they interact with downstream effectors and thus regulate the trafficking process. Rab and ARF family small GTPases are the major regulators for protein trafficking events, including endosome-to-TGN transport. Rabs form the largest Ras-like small GTPase family with more than 60 members identified in human and 11 (Ypts) in yeast (Pereira-Leal et al. 2000). Ypt6p is a Golgi localized Rab GTPase in yeast. It has an important role in endosome-to-late Golgi transport. In ypt6 mutant yeast, the recycling of Kex2p and Vps10p is impaired, resulting in their degradation in the vacuole (Tsukada et al. 1999; Siniossoglou et al. 2000; Bensen et al. 2001). Ypt6p is also essential for endosome-to-late Golgi delivery of Snc1p. The genetic screening of the suppressor of loss of Ypt6 yielded many proteins which are putative components of the machinery regulating end some-to-late Golgi trafficking, including Imh1p, Ric1p/Rgp1p and GARP/VFT complex. The guanine nucleotide exchange of Ypt6p from GDP to GTP depends on Ric1p/Rgp1p protein complex (Siniossoglou et al. 2000; Siniossoglou and Pelham 2001). The GTP-bound Ypt6p is able to recruit its effector GARP/VFT complex (detailed below) to the late Golgi. By interacting directly with the N-terminus of t-SNARE Tlg1p, the Golgi localized GARP/VFT complex is likely to activate Tlg1p SNARE. Ypt6pGTP also recruits Sgm1p, a Golgin family protein having long coiled-coil structure (Fig. 4). The homolog of Ypt6p in mammals is Rab6 and its isoform Rab6A0 . Both Rab6A and Rab6A0 are enriched at the TGN. Rab6A0 is an alternative splicing variant of Rab6A, differing in only three amino acids at the effector binding switch II region (Echard et al. 2000). This sequence variation results in different functions and effectors for Rab6A0 . Rab6A regulates TGN to ER retrograde trafficking of STxB and some endogenous proteins (Martinez et al. 1997; Girod et al. 1999; White et al. 1999). Rab6A0 , on the other hand, regulates the RE–TGN transport (Mallard et al. 2002). Rab6A0 has two effectors—RabIP2A and Rab6IP2B (Rab6 interacting protein 2). Overexpression of either RabIP2A or Rab6IP2B prevents the transport of STxB to the TGN (Monier et al. 2002). Similar to Ypt6p, Rab6A0 and Rab6A interact with the mammalian Sgm1p homolog, Golgin TMF/ARA160 (Fridmann-Sirkis et al. 2004). Enriched in the RE and TGN, Rab11 functions in recycling proteins from the RE back to the PM (Urbe et al. 1993; Ullrich et al. 1996). It also regulates the RE–TGN transport of STxB and TGN38 (Wilcke et al. 2000; Mallard et al. 2002). The Rab11 ortholog in yeast is an essential pair of GTPases – Ypt31/32p.

Retrograde endosome-to-TGN Transport

*

439

Figure 4. A working model to depict the tethering and fusion machinery on the TGN responsible for endosome-to-TGN transport. The tethering event is likely regulated by the small GTPases, such as Arl1, ArfRP1, Rab6A0 /Ypt6p and their regulatory proteins. The identity of the guanine nucleotide exchange factor for Arl1 is yet to be resolved, although Ysl2p is a candidate. The signaling cascade from the transmembrane protein Sys1p to the activation of ArfRP1GTPase is suggested to activate Arl1 so that Arl1-GTP is recruited to the TGN, which in turn recruits cytosolic Golgin-97 and Golgin-245 onto the Golgi membrane via interaction of Arl1-GTP with the C-terminal conserved GRIP domain of the Golgins. The activation of Arl1GTP is balanced by its inactivation, which is likely facilitated by candidate Arl1’s GTPase activating proteins (GAPs) (yeast Gcs1p is a candidate). Ric1p/Rgp1p protein complex in yeast has been discovered to have GEF activity toward Ypt6p /Rab6A0 . The active Ypt6p/Rab6A0 may recruit yeast Sgm1p and mammalian TMF onto the membrane of the TGN. The long fiber-like Golgin molecules are speculated to selectively tether the incoming transport vesicles by interacting with unknown receptors on the vesicles. Multi-protein tethering complexes such as COG, GARP/VFT and TRAPPII are likely to coordinate with the Golgins during the tethering process. The SNARE complexes Vamp4 (or Vamp3)/Syn6/Syn16/Vti1a and GS15/Syn5/GS28/ Ykt6 are likely to mediate the fusion of properly tethered vesicles with the TGN membrane in mammalian cells.

Ypt31p is localized at the late Golgi (Sciorra et al. 2005). Aside from their function in anterograde trafficking from the late Golgi to the endosome (Jedd et al. 1997), Ypt31/32p, together with its effector Rcy1p, participates in the retrograde recycling of Snc1p and Kex2p from the endosomes to the late Golgi (Chen et al. 2005). Rcy1p is an F-box protein and, together with Skp1p, form a non-SCF (Skp1p-cullin-F-box) complex, which is required for the recycling of Snc1p (Galan et al. 2001; Chen et al. 2005). Rcy1p directly interacts with Snc1p (Chen et al. 2005). F-box proteins are known for its role in ubiquitin-dependent degradation by binding to substrates (reviewed by Deshaies et al. 1999), raising the possibility that Ypt31/32 and Rcy1p–Skp1p

440

*

L. Lu and W. Hong

signaling cascade regulates an ubiquitin-dependent event to facilitate endosome-to-TGN transport. Rab7 and Rab9 are both enriched in the LE. However, Rab7 and Rab9 have distinct distributions: they are either on different population of endosomes or they reside on distinct membrane subdomains of the same endosome (Barbero et al. 2002). Rab7 mainly functions in the transport from the EE to the LE and then to the lysosome. CI-M6PR-based in vitro transport assays established that Rab9 functions in LE-to-TGN trafficking (Lombardi et al. 1993; Riederer et al. 1994). CI-M6PR was found in the Rab9 positive domain of the LE. Live cell imaging also supports a role of Rab9 in LE-to-TGN transport (Barbero et al. 2002). The role of Rab9 in LE-to-TGN traffic is probably mediated by its effectors. Two effectors have been discovered for Rab9: TIP47 and p40. TIP47 is detected on both the LE and lipid droplets (Diaz and Pfeffer 1998; Wolins et al. 2001) and originally identified from yeast-two-hybrid screening using cytosolic domain of M6PR as the bait (Diaz and Pfeffer 1998). By binding to the cytosolic domain of transmembrane cargos, such as M6PRs and HIV-1 Env gp41, TIP47 is proposed to be a cargo adaptor to regulate transport from the LE to the TGN (Diaz and Pfeffer 1998; Carroll et al. 2001; Barbero et al. 2002; Blot et al. 2003). This cargo adaptor activity of TIP47 is regulated by Rab9, as Rab9-GTP increases the binding of TIP47 to CI-M6PR (Carroll et al. 2001). P40 is a kelch repeat containing 40 kDa protein identified in yeast-two-hybrid screening using active Rab9 as bait (Diaz et al. 1997) and it selectively binds Rab9-GTP. Recombinant p40 stimulates in vitro transport of M6PR from the LE to the TGN (Diaz et al. 1997). The membrane association of p40 is probably regulated through the phosphorylation by its newly discovered kinase binding partner PIKfyve (Ikonomov et al. 2003). ARF family of small GTPases is divided into two groups: ARF and Arl (ARF-like). ARFs reside at the Golgi complex (ARF1-5), the endosomes (ARF1 and ARF6) and the PM (ARF6) and participate in a variety of pre-, intra-, postGolgi trafficking events (reviewed by Donaldson and Honda 2005). ARF1, the most studied ARF family member, has multiple functions. ARF1 recruits COPI, AP1 and GGA1/2/3 onto the membrane (Bednarek et al. 1995; Stamnes and Rothman 1993; Traub et al. 1993; Boman et al. 2000) and it may also regulate AP2, AP3 and AP4 clathrin coats (Dell’Angelica et al. 2000; Hirst et al. 2000; Takatsu et al. 2000; West et al. 1997; Ooi et al. 1998; Hirst et al. 1999). ARF1 also regulates the metabolism of phospholipids on Golgi membranes, by activating phospholipase D or recruiting PtdIns 4-kinase b and PtdIns 4-phosphate 5-kinase on the Golgi membrane (Brown et al. 1993; Godi et al. 1999). Activation of ARF1 leads to the elevated level of PA, PtdIns4P and PtdIns (4,5)P2 on the Golgi membrane, leading to the remodeling of membrane traffic and cytoskeleton. Thus ARF1 participates in multiple steps in both secretory and endocytic pathways, although its role in endosome–TGN trafficking has not been clearly demonstrated. Arls, including Arl1-8, ArfRP1 and ARD1, form another group (reviewed by Burd et al. 2004). Among them, Arl1 and ArfRP1 are recently reported to be

Retrograde endosome-to-TGN Transport

*

441

trans-Golgi/TGN localized (Lu et al. 2001; Shin et al. 2005). Arl1 regulates the Golgi structure and secretory function (Lu et al. 2001). GRIP domain-containing Golgins identified as Arl1 interacting proteins using yeast-two-hybrid screening are downstream effectors of Arl1 (Lu et al. 2001; Van Valkenburg et al. 2001). Arl1-GTP specifically recruits GRIP Golgins, Golgin-97 and Golgin-245/trans-Golgi p230, to the Golgi membrane by interacting with their C-terminal GRIP domains (Lu and Hong 2003). Transport of STxB or CTxB to the TGN is dependent of Arl1 (Lu et al. 2004), suggesting that Arl1 regulates RE–TGN transport. Arl1 effector, Golgin-97, is likely to function as a tethering factor for RE–TGN transport. Yeast has two Arls – Arl1p and Arl3p. Arl1p, the ortholog of mammalian Arl1, is responsible for Golgi targeting of Imh1p, the sole GRIP Golgin in yeast (Panic et al. 2003a, b; Setty et al. 2003). The involvement of Arl1 in retrograde transport from the EE/RE to the TGN is well conserved in yeast. Deletion of Arl1p in yeast resulted in missorting of CPY, due to impaired retrograde recycling of Vps10p to the late Golgi (Bonangelino et al. 2002). Arl1p is genetically or biochemically implicated in interaction with Ric1p/Rgp1p complex, Ypt6p, and the GARP/VFT complex (Bensen et al. 2001; Panic et al. 2003a, b). Guanine nucleotide binding state of Arl1p could be regulated by recently identified GEF candidate – Mon2p/Ysl2p (Jochum et al. 2002) and GAP – Gcs1p (Liu et al. 2005). This is supported by the recent finding that mon2/ysl2 deletion in yeast caused retrieval defect of Snc1p and Tlg2p (Efe et al. 2005). The Golgi targeting of Arl1p is regulated by Arl3p, the yeast ortholog of ArfRP1 (Panic et al. 2003a, b; Setty et al. 2003). Arl3 deletion causes missorting of CPY (Bonangelino et al. 2002). Similarly, mammalian ArfRP1 may act in endosome-to-TGN transport as internalized TGN38 antibody and Shiga toxin are arrested at endosomes in HeLa cells expressing dominant-negative mutant ArfRP1-T31N (Shin et al. 2005). Although the GTPase signaling cascade of ARFRP1/Arl3p ! Arl1 ! GRIP Golgins seems to regulate endosome-to-TGN traffic (Fig. 4), we do not know how ArfRP1/Arl3p regulates Golgi targeting of Arl1. One possibility is that ArfRP1/Arl3p activates the GEF for Arl1 (Panic et al 2003a, b; Setty et al. 2003). Arl3p has a unique post-translational modification which is acetylation at its N-terminus. This is different from myristoylation for other ARF family GTPases. The interaction between acetylated N-terminus of Arl3p with Sys1p, a multi-transmembrane protein, is essential for Golgi targeting of Arl3p (Behnia et al. 2004; Setty et al. 2004), indicating another layer of regulation for this pathway.

Tethering proteins Tethering is vaguely defined as a process of loosely restraining or capturing transport vesicles at the vicinity of the acceptor compartment before SNAREmediated short-ranged docking and subsequent fusion of vesicles with the acceptor compartment (Pfeffer 1999). There are two classes of tethering molecules: Golgin-like coiled-coil proteins predicted to have extended long rod-like structure and multi-subunits tethering complexes. Golgins are a group of heterogenous peripheral or membrane Golgi proteins with extensive coiled-

442

*

L. Lu and W. Hong

coil regions exposed in the cytoplasm (reviewed by Gillingham and Munro 2003). There are more than a dozen of Golgins in human and about five in yeast. The majority of human Golgins were reported as autoantigens in autoimmune diseases. Golgins are known to form homodimers and adopt an extended rod-like or fibrous structures with hinges. The lengths of some Golgins are comparable to the diameters of transport vesicles. For example, EM imaging shows that Uso1p, the yeast ortholog of p115, has a length of 150 nm (Yamakawa et al. 1996). The long-rod structure of Golgins makes them good candidates as the tethering molecules for incoming vesicles. GRIP Golgins, including Golgin-97, Golgin-245, GCC88, GCC185 in mammal and Imh1p (Sys3p) in yeast are characterized by long coiled-coil regions and a C-terminal GRIP domain, which is responsible for their TGN targeting (Barr 1999; Kjer-Nielsen et al. 1999; Munro and Nichols 1999; Luke et al. 2003). Consisting of about 50–60 residues, the GRIP domain of Golgin-245 assembles into an array of three helices, which is capable of homodimerization and subsequently interacting with two molecules of Arl1-GTP (Lu and Hong 2003; Setty et al. 2003; Panic et al. 2003a, b; Wu et al. 2004a). Arl1-GTP is necessary and sufficient for the recruitment of its effectors Golgin-97 and Golgi-245 onto the Golgi membrane (Fig. 4). Unlike Golgin-97 and Golgin-245, the GRIP domains of GCC88 and GCC185 do not interact efficiently with Arl1-GTP (Derby et al. 2004). Overexpression of GRIP domain of Golgin-245 resulted in the mis-localization of TGN46 in the endosomes, suggesting GRIP Golgins may regulate the endosomal retrieval of TGN46 to the TGN (Yoshino et al. 2005). Golgin-97 was shown to participate in RE–TGN transport of shiga toxin (Lu et al. 2004). When endogenous Golgin-97 was neutralized by polyclonal antibody, competed by recombinant GRIP protein or knock-downed by siRNA, transport of STxB to the TGN was inhibited. Importantly, the temporal requirement of Golgin-97 is significantly earlier than syntaxin16, implying Golgin-97 acts as a tethering molecule and functions prior to SNAREs in RE–TGN transport of STxB (Lu et al. 2004). This is in line with previous study of Imh1p. Deletion of Imh1 in the ypt6D background yeast caused defect in the retrieval of Kex2p from the endosome to the late Golgi and a massive accumulation of vesicles (Tsukada et al. 1999). Recent studies suggest that other two GRIP Golgins, GCC88 and GCC185, are involved in endosome-toTGN transport as well (Reddy et al. 2006; Derby et al. 2007; Lieu et al. 2007). Interestingly, it has been shown that GCC88 and GCC185 are selective toward their cargo proteins. GCC88 and GCC185 were found to be required for TGN38 and Shiga toxin endosome–TGN trafficking, respectively (Derby et al. 2007; Lieu et al. 2007). The findings support the existence of multiple populations of vesicles in RE–TGN trafficking, which utilize different Golgins or SNAREs (see also SNAREs) to target to different domains of the TGN. In addition to the four GRIP Golgins, Rab6 effector, TMF/ARA160, (Sgm1p in yeast) may serve as the tethering molecule in this process (Fridmann-Sirkis et al. 2004; Matanis et al. 2002; Short et al. 2002). Multi-subunit complexes, including COG (conserved oligomeric Golgi complex), TRAPP (transport protein particle) and GARP/VFT (Golgi-associated

Retrograde endosome-to-TGN Transport

*

443

retrograde protein/VPS fifty three) complexes, also regulate tethering process of various transport events (reviewed by Whyte et al. 2002). COG complex is a cytosolic hetero-octameric complex consisting of COG1-8, which are conserved from yeast to human (Whyte and Munro 2001). Combinatory studies using EM, biochemistry and yeast genetics revealed that these eight subunits form a two-lobe structure (Ungar et al. 2002, 2005; Oka et al. 2005; Loh and Hong 2004). COG complex is localized at cis/medial Golgi (Suvorova et al. 2001; Ungar et al. 2002). In yeast and mammals, COG complex has been implicated in multiple trafficking pathways, such as ER-to-Golgi, intra-Golgi and post-Golgi trafficking events in both anterograde and retrograde directions. COG may physically interact with Golgi SNAREs, Ypt1p and COPI coat (Suvorova et al. 2002). Mutations in subunits of COG complex cause pleiotropic defects in Golgi-associated glycosylation of proteins and lipids in mammalian cells and yeast (Kingsley et al. 1986; Reddy and Krieger 1989; Whyte and Munro 2001; Suvorova et al. 2002; Bruinsma et al. 2004). In fact, mutation in COG7 has been reported to result in a human disease called congenital disorder of glycosylation (Wu et al. 2004a). The current working model favors for the COG complex to act in tethering the retrograde trafficking vesicles. The previously observed anterograde transport defects may be a result from the defect in recycling of Golgi enzymes and SNAREs. The retrograde trafficking mechanism of COG complex may also extend to the endosome. Firstly, the yeast COG3 ortholog, Grd20p is localized to late Golgi, implying COG complex resides also in the late Golgi (Spelbrink and Nothwehr 1999). Secondly, COG complex may physically interact with COPI and GS28 (Zolov and Lupashin 2005), both participating in endosomal retrograde transport. Thirdly, compromising the COG complex results in defect in endosome– TGN trafficking. For example, mutation in the yeast COG complex caused the mis-localiazation of Snc1p and Kex2p (Whyte and Munro 2001; Spelbrink and Nothwehr 1999). In mammals, COG subunit depletion by siRNA led to the accumulation of COG complex-dependent (CCD) vesicles carrying type II Golgi transmembrane proteins (GEARS), including v-SNAREs, GS15 and GS28, and cisGolgi glycoprotein GPP130 (Oka et al. 2004; Zolov and Lupashin 2005). Depletion of COG also inhibited endsome–TGN trafficking of STxB. It is interesting to note that, the GEARS, such as GS15, GS28 and GPP130, are regulators of STxB trafficking from the endosome to the TGN (Tai et al. 2004; Natarajan and Linstedt 2004). As GS15, GS28 and GPP130 cycle between endosomes and the Golgi (Tai et al. 2004; Linstedt et al. 1997), it is possible that some of CCD vesicles are derived from the RE and accumulate due to the defect in tethering function of COG complex. Together, there studies suggest that COG complex could facilitate the tethering process during endosome–TGN transport. There are two forms of TRAPP complexes, TRAPP I and II. TRAPP I complex of yeast consists of seven subunits, Bet3p, Bet5p, Trs20p, Trs23p, Trs31p, Trs33p and Trs85p, which are also shared by TRAPP II complex (Sacher et al. 1998, 2000). TRAPPII complex has three extra subunits: Trs65p, Trs120p and Trs130p (Sacher et al. 1998, 2000). From yeast mutational analysis and mammalian in

444

*

L. Lu and W. Hong

vitro transport assay, TRAPP complex has been proposed to function in the cis-Golgi for tethering ER-derived COPII vesicles (Sacher et al. 2001; Loh et al. 2005). Disruption of TRAPP causes general Golgi secretion defects (Sacher et al. 2001; Cai et al. 2005). TRAPP complex can biochemically function as a GEF for Ypt1p, an early Golgi Rab GTPase, and Ypt31/32p at the late Golgi (Jones et al. 2000; Wang et al. 2000). It seems that TRAPPI and II complexes could act as tethers at two different locations – cis- and trans-Golgi/TGN, respectively. Density gradient and cellular localization study positioned TRAPPII complex at the late Golgi and early endosome (Cai et al. 2005). Loss of function of Trs120p or Trs130p, two of the three subunits unique to TRAPPII, displayed defects of retrieval of Snc1p and Chs3p, resulting in both being accumulated in the EE (Cai et al. 2005). These findings suggest that TRAPPII, different from TRAPPI complex, probably function in endosome–TGN trafficking. TRAPPII may initiate the Ypt31/32p Rab GTPase signaling cascade and/or serve as a tethering factor for retrograde incoming vesicles (Cai et al. 2005). GARP/VFT complex is a late Golgi-associated tetramer consisting of Vps51p, 52p, 53p and 54p in yeast (Conibear et al. 2003; Reggiori et al. 2003). Yeast cells lacking any one of the four genes display retrieval defect for Kex2p and Vps10p from the PVC to the late Golgi (Conibear and Stevens 2000; Conibear et al. 2003), indicating GARP/VFT complex is essential for endosome–TGN trafficking. The Golgi localization of GARP/VFT complex is regulated by Ypt6p and this complex could in turn regulate the SNARE assembly at the late Golgi by binding directly to t-SNARE, Tlg1p (see SNAREs). Arl1p, a trans-Golgi/TGN small GTPase essential for RE–TGN transport, was found to physically interact with GARP/VFT complex (Panic et al. 2003a), but it is not clear how this interaction contributes to endosome–TGN trafficking. The human GARP/VFT ortholog complex has been identified and is likely to have a conserved function (Liewen et al. 2005).

Sorting nexins Phox (PX) domain is known to bind phosphoinositides, especially PtdIns(3)P (Cheever et al. 2001; Kanai et al. 2001; Xu et al. 2001). 15 and 47 PX domain proteins in yeast and human, respectively, have been discovered (reviewed by Seet and Hong 2006). PX domain proteins are engaged in diverse functions including immune defense, cellular signaling and membrane trafficking. Sorting nexins are among the PX-domain containing peripheral membrane proteins involved in membrane trafficking. Studies have implicated sorting nexins in modulating endosomal transport of various receptors for recycling and degradation. For instance, SNX1 was identified by its ability to bind specifically the cytosolic domain of receptors at the PM, such as EGFR and PAR-1, and promote their trafficking to the lysosome for degradation (Kurten et al. 1996; Haft et al. 1998; Cozier et al. 2002; Zhong et al. 2002; Wang et al. 2002). The implication of sorting nexins in endosome–TGN trafficking was originally from studies of yeast sorting nexins – Snx4p, Snx41p, Snx42p, Mvp1p, Grd19p/SNX3p, Vps5p and Vps17p. All yeast 15 sorting nexins have been

Retrograde endosome-to-TGN Transport

*

445

demonstrated to bind PtdIns(3)P (Yu and Lemmon 2001). As PtdIns(3)P is enriched in endosomes, most of these sorting nexins reside and act in endosomes. Snx4p, Snx41p and Snx42p are required for the cycling of Snc1p, but not Tlg1p, Tlg2p and Chs3p, from the EE to the late Golgi. Snx4p may interact with Snc1p as it can be cross-linked to Snc1p (Hettema et al. 2003). Mvp1p is required for retrieval of Kex2p and Vps10p from the PVC to the late Golgi (Luo and Chang 1997). Grd19p/SNX3p is required for the retrieval of Kex2p, DPAP-A and Pep12p from the late endosome to the late Golgi, although it was found that Grd19p only interacted with the cytosolic domain of DPAP-A (Voos and Stevens 1998). The finding that the retrieval of Vps10p is dependent on Mvp1p, but not Grd19p (Ekena et al. 1995; Voos and Stevens 1998), raises the hypothesis that different sorting nexins may be required for sorting distinctive sets of proteins. Retromer is one of the key protein complexes in endosome-to-TGN transport (Seaman et al. 1997, 1998). The yeast retromer is a conserved pentameric complex consisting of two sorting nexins, Vps5p and Vps17p (forming a heterodimer), with another subcomplex consisting of Vps26p, Vps29p and Vps35p (Horazdovsky et al. 1997; Nothwehr and Hindes 1997; Seaman et al. 1998). Vps5p and Vps17p are essential for maintaining the proper localization of Vps10p (Horazdovsky et al. 1997; Nothwehr and Hindes 1997). The mammalian orthologs of Vps5p and Vps17p are SNX1 and probably SNX2, respectively (Griffin et al. 2005). SNX1 and SNX2 play both distinct and overlapping functions and, together, they are essential for normal development (Schwarz et al. 2002). Mammalian Vps26, Vps29 and Vps35 form a similar subcomplex to interact with SNX1 and/or SNX2 (Haft et al. 2000; Gullapalli et al. 2004; Carlton et al. 2005; Griffin et al. 2005). In mammalian cells, retromer is localized to tubular–vesicular profiles of the EE/RE (Arighi et al. 2004; Seaman et al. 2004). This targeting is achieved through the PtdIns(3)P and membrane curvature of the EE/RE sensed by the PX and BAR domains of SNX1/2, respectively (Carlton et al. 2004). The BAR domain of SNX1 also provides the mechanical force to drive the membrane tabulation of the EE/RE (Carlton et al. 2004). Hence, the membrane association of retromer depends on PtdIns(3)P, which is generated by PtdIns-3 kinases. Indeed, it has been demonstrated that endosomal targeting of retromer is regulated by Vps34p, the only PtdIns-3 kinase in yeast (Burda et al. 2002). The subcomplex consisting of Vps26p, Vps29p and Vps35p is likely involved in cargo selection, as Vps35p binds the cytosolic domain of membrane cargos, such as DPAP-A and Vps10p (Notwehr et al. 2000). In DPAP-A, this signal consists of an aromatic motif – FXFXD with both phenylanaline residues being required. The mammalian Vps35 can bind the cytosolic domain of CI-M6PR in yeast-two-hybrid assay (Arighi et al. 2004), which functionally mirrors the interaction of Vps35p with Vps10p and DPAP-A (Notwehr et al. 2000). Using CD8 chimera expressing HeLa cells and antibody internalization assay, Vps26 was shown to be essential for the endosome–TGN transport of CI-M6PR and sortilin, but not furin. Depletion of VPS26 and SNX1 by siRNA resulted in accumulation of CI-M6PR in endosomes and enhanced

446

*

L. Lu and W. Hong

degradation in the lysosome (Arighi et al. 2004; Carlton et al. 2004; Seaman 2004). The missorting of CI-M6PR significantly affected the lysosomal targeting of Cathepsin D, the lumenal cargo of CI-M6PR (Seaman 2003). Given the recent evidence that CI-M6PR and sortilin follow both the RE–TGN and LE-TGN itineraries en route to the TGN, while furin employs preferentially the LE–TGN pathway (Lin et al. 2004; Mallet and Maxfield 1999), the selective effect of retromer depletion on CI-M6PR and sortilin, but not furin, suggests that retromer may function in RE-TGN but not the LE–TGN pathway. However, in yeast, retromer is crucial for the recycling of Vps10p from the PVC to the late Golgi (Notwehr et al. 2000); but the EE–TGN recycling of Snc1p is normal in vps5, vps17 and vps35 null mutants (Lewis et al. 2000). Therefore, retromer could function in both RE–TGN and LE–TGN pathways, depending on the cargo proteins and cell type. Retromer may also regulate endosome – TGN transport of other proteins such as sortilin (Seaman 2004) and b-secretase/BACE/memapsin-2 (He et al. 2005). The pathogenesis of Alzheimer’s disease involves the accumulation of Ab peptide, which is generated by sequential cleavage of APP by b and g-secretases. b-secretase is a type I integral membrane protease cycling between the PM and the TGN via the EE/RE (Huse et al. 2000; Walter et al. 2001). Depletion of Vps26 caused the redistribution of b-secretase from the Golgi apparatus to endosomes, implying retromer regulates the retrograde trafficking of b-secretase and may play a role in the progression of AD (He et al. 2005). Supporting retromer’s relevance to the disease, downregulation of retromer has been reported in AD patients (Small et al. 2005).

Conclusions The endosome–TGN transport is emerging as one of the major post-Golgi trafficking events. It integrates the secretory and endocytic pathways and regulates the distribution of many proteins functioning on the PM, endosomes and the Golgi. Despite enormous progresses made in the past decades in our understanding of endosome–TGN transport, the mechanistic basis for the different use of the RE–TGN and LE–TGN routes for different cargoes remains elusive. More defined understanding about the molecular mechanisms and the physiological and pathological relevance of the endosome– TGN transport is poised to come.

References Ahle S, Mann A, Eichelsbacher U, Ungewickell E (1988) Structural relationships between clathrin assembly proteins from the Golgi and the plasma membrane. EMBO J 7: 919–929 Andersen OM, Reiche J, Schmidt V, Gotthardt M, Spoelgen R, Behlke J, Von Arnim CA, Breiderhoff T, Jansen P, Wu X, Bales KR, Cappai R, Masters CL, Gliemann J, Mufson EJ, Hyman BT, Paul SM, Nykjaer A, Willnow TE (2005) Neuronal sorting protein-

Retrograde endosome-to-TGN Transport

*

447

related receptor sorLA/LR11 regulates processing of the amyloid precursor protein. Proc Natl Acad Sci 102: 13461–13466 Aniento F, Gu F, Parton RG, Gruenberg J (1996) An endosomal beta COP is involved in the pH-dependent formation of transport vesicles destined for late endosomes. J Cell Biol 133: 29–41 Ang AL, Taguchi T, Francis S, Folsch H, Murrells LJ, Pypaert M, Warren G, Mellman I (2004) Endosomes can serve as intermediates during transport from the Golgi to the plasma membrane of MDCK cells. J Cell Biol 167: 531–543 Annaert W, De Strooper B (2002) A cell biological perspective on Alzheimer’s disease. Annu Rev Cell Dev Biol 18: 25–51 Antonny B, Schekman R (2001) ER export: public transportation by the COPII coach. Curr Opin Cell Biol 13: 438–443 Arighi CN, Hartnell LM, Aguilar RC, Haft CR, Bonifacino JS (2004) Role of the mammalian retromer in sorting of the cation-independent mannose 6-phosphate receptor. J Cell Biol 165: 123–133  T, Ledesma MD, Saffrich R, Kok JW, Dotti CG, Egea G (2001) Endocytosis of NBDBabia sphingolipids in neurons: exclusion from degradative compartments and transport to the Golgi Complex. Traffic 2: 395–405 Barbero P, Bittova L, Pfeffer SR (2002) Visualization of Rab9-mediated vesicle transport from endosomes to the trans-Golgi in living cells. J Cell Biol 156: 511–518 Barr FA (1999) A novel Rab6-interacting domain defines a family of Golgi-targeted coiled-coil proteins. Curr Biol 9: 381–384 Bednarek SY, Ravazzola M, Hosobuchi M, Amherdt M, Perrelet A, Schekman R, Orci L (1995) COPI- and COPII-coated vesicles bud directly from the endoplasmic reticulum in yeast. Cell 83: 1183–1196 Behnia R, Panic B, Whyte JR, Munro S (2004) Targeting of the Arf-like GTPase Arl3p to the Golgi requires N-terminal acetylation and the membrane protein Sys1p. Nat Cell Biol 6: 405–413 Bensen ES, Yeung BG, Payne GS (2001) Ric1p and the Ypt6p GTPase function in a common pathway required for localization of trans-Golgi network membrane proteins. Mol Biol Cell 12: 13–26 Blagoveshchenskaya AD, Thomas L, Feliciangeli SL, Hung C-H, Thomas G (2002) HIV-1 Nef downregulates MHC-I by a PACS-1- and PI3Kregulated ARF6 endocytic pathway. Cell 111: 853–866 Blot G, Janvier K, Le Panse S, Benarous R, Berlioz-Torrent C (2003) Targeting of the human immunodeficiency virus type 1 envelope to the trans-Golgi network through binding to TIP47 is required for env incorporation into virions and infectivity. J Virol 77: 6931–6945 Boehm M, Bonifacino JS (2001) Adaptins: the final recount. Mol Biol Cell 12: 2907–2290 Bonangelino CJ, Chavez EM, Bonifacino JS (2002) Genomic screen for vacuolar protein sorting genes in Saccharomyces cerevisiae. Mol Biol Cell 13: 2486–2501 Bonifacino JS, Traub LM (2003) Signals for sorting of transmembrane proteins to endosomes and lysosomes. Annu Rev Biochem 72: 395–447 Boman AL, Zhang C, Zhu X, Kahn RA (2000) A family of ADP-ribosylation factor effectors that can alter membrane transport through the trans-Golgi. Mol Biol Cell 11: 1241–1255 Brignatz C, Restouin A, Bonello G, Olive D, Collette Y (2005) Evidences for ubiquitination and intracellular trafficking of LAT, the linker of activated T cells. Biochim Biophys Acta 1746: 108–115 Brown HA, Gutowski S, Moomaw CR, Slaughter C, Sternweis PC (1993) ADP-ribosylation factor, a small GTP-dependent regulatory protein, stimulates phospholipase D activity. Cell 75: 1137–1144

448

*

L. Lu and W. Hong

Bruinsma P, Spelbrink RG, Nothwehr SF (2004) Retrograde transport of the mannosyltransferase Och1p to the early Golgi requires a component of the COG transport complex. J Biol Chem 279: 39814–39823 Bryant NJ, Stevens TH (1997) Two separate signals act independently to localize a yeast late Golgi membrane protein through a combination of retrieval and retention. J Cell Biol 136: 287–297 Bryant NJ, James DE (2001) Vps45p stabilizes the syntaxin homologue Tlg2p and positively regulates SNARE complex formation. EMBO J 20: 3380–3388 Burd CG, Strochlic TI, Gangi Setty SR (2004) Arf-like GTPases: not so Arf-like after all. Trends Cell Biol 14: 687–694 Burda P, Padilla SM, Sarkar S, Emr SD (2002) Retromer function in endosome-to-Golgi retrograde transport is regulated by the yeast Vps34 PtdIns 3-kinase. J Cell Sci 115: 3889–3900 Cai H, Zhang Y, Pypaert M, Walker L, Ferro-Novick S (2005) Mutants in trs120 disrupt traffic from the early endosome to the late Golgi. J Cell Biol 171: 823–833 Carlton JG, Bujny MV, Peter BJ, Oorschot VM, Rutherford A, Arkell RS, Klumperman J, McMahon HT, Cullen PJ (2005) Sorting nexin-2 is associated with tubular elements of the early endosome, but is not essential for retromer-mediated endosome-toTGN transport. J Cell Sci 118: 4527–4539 Carroll KS, Hanna J, Simon I, Krise J, Barbero P, Pfeffer SR (2001) Role of the Rab9 GTPase in facilitating receptor recruitment by TIP47. Science 292: 1373–1377 Castle A, Castle D (2005) Ubiquitously expressed secretory carrier membrane proteins (SCAMPs) 1–4 mark different pathways and exhibit limited constitutive trafficking to and from the cell surface. J Cell Sci 118: 3769–3780 Cheever ML, Sato TK, De Beer T, Kutateladze TG, Emr SD, Overduin M (2001) Phox domain interaction with PtdIns(3)P targets the Vam7 t-SNARE to vacuole membranes. Nat Cell Biol 3: 613–618 Chen YA, Scheller RH (2001) SNARE-mediated membrane fusion. Nat Rev Mol Cell Biol 2: 98–106 Chen SH, Chen S, Tokarev AA, Liu F, Jedd G, Segev N (2005) Ypt31/32 GTPases and their novel F-box effector protein Rcy1 regulate protein recycling. Mol Biol Cell 16: 178–192 Choudhury A, Dominguez M, Puri V, Sharma DK, Narita K, Wheatley CL, Marks DL, Pagano RE (2002) Rab proteins mediate Golgi transport of caveola-internalized glycosphingolipids and correct lipid trafficking in Niemann-Pick C cells. J Clin Invest 109: 1541–1550 Conibear E, Stevens TH (2000) Vps52p, Vps53p, and Vps54p form a novel multisubunit complex required for protein sorting at the yeast late Golgi. Mol Biol Cell 11: 305–323 Conibear E, Cleck JN, Stevens TH (2003) Vps51p mediates the association of the GARP (Vps52/53/54) complex with the late Golgi t-SNARE Tlg1p. Mol Biol Cell 14: 1610–1623 Cozier GE, Carlton J, McGregor AH, Gleeson PA, Teasdale RD, Mellor H, Cullen PJ (2002) The phox homology (PX) domain-dependent, 3-phosphoinositide-mediated association of sorting nexin-1 with an early sorting endosomal compartment is required for its ability to regulate epidermal growth factor receptor degradation. J Biol Chem 277: 48730–48736 Crump CM, Xiang Y, Thomas L, Gu F, Austin C, Tooze SA, Thomas G (2001) PACS-1 binding to adaptors is required for acidic cluster motif-mediated protein traffic. EMBO J 20: 2191–2201 Crump CM, Hung CH, Thomas L, Wan L, Thomas G (2003) Role of PACS-1 in trafficking of human cytomegalovirus glycoprotein B and virus production. J Virol 77: 11105–11113 Dahms NM, Lobel P, Kornfeld S (1989) Mannose 6-phosphate receptors and lysosomal enzyme targeting. J Biol Chem 264: 12115–12118

Retrograde endosome-to-TGN Transport

*

449

Daro E, Sheff D, Gomez M, Kreis T, Mellman I (1997) Inhibition of endosome function in CHO cells bearing a temperature-sensitive defect in the coatomer (COPI) component epsilon-COP. J Cell Biol 139: 1747–1759 Dell’Angelica EC, Puertollano R, Mullins C, Aguilar RC, Vargas JD, Hartnell LM, Bonifacino JS (2000) GGAs: a family of ADP ribosylation factor-binding proteins related to adaptors and associated with the Golgi complex. J Cell Biol 149: 81–94 Derby MC, Van Vliet C, Brown D, Luke MR, Lu L, Hong W, Stow JL, Gleeson PA (2004) Mammalian GRIP domain proteins differ in their membrane binding properties and are recruited to distinct domains of the TGN. J Cell Sci 117: 5865–5874 Derby MC, Lieu ZZ, Brown D, Stow JL, Goud B, Gleeson PA (2007) The trans-Golgi network golgin, GCC185, is required for endosome-to-Golgi transport and maintenance of Golgi structure. Traffic 8: 758–773 Deshaies RJ (1999) SCF and Cullin/Ring H2-based ubiquitin ligases. Annu Rev Cell Dev Biol 15:435–467 Diaz E, Schimmoller F, Pfeffer SR (1997) A novel Rab9 effector required for endosometo-TGN transport. J Cell Biol 138: 283–290 Diaz E, Pfeffer SR (1998) TIP47: a cargo selection device for mannose 6-phosphate receptor trafficking. Cell 93: 433–443 Dittie AS, Hajibagheri N, Tooze SA (1996). The AP-1 adaptor complex binds to immature secretory granules from PC12 cells, and is regulated by ADP-ribosylation factor. J Cell Biol 132: 523–536 Donaldson JG, Honda A (2005) Localization and function of Arf family GTPases. Biochem Soc Trans 33: 639–642 Echard A, De Opdam FJM, Leeuw HJPC, Jollivet F, Savelkoul P, Hendriks W, Voorberg J, Goud B, Fransen JAM (2000) Alternative splicing of the human Rab6A gene generates two close but functionally different isoforms. Mol Biol Cell 3: 3819–3833 Efe JA, Plattner F, Hulo N, Kressler D, Emr SD, Deloche O (2005) Yeast Mon2p is a highly conserved protein that functions in the cytoplasm-to-vacuole transport pathway and is required for Golgi homeostasis. J Cell Sci 118: 4751–4764 Ekena K, Stevens TH (1995) The Saccharomyces cerevisiae MVP1 gene interacts with VPS1 and is required for vacuolar protein sorting. Mol Cell Biol 15: 1671–1678 Falguieres T, Mallard F, Baron C, Hanau D, Lingwood C, Goud B, Salamero J, Johannes L (2001) Targeting of Shiga toxin B-subunit to retrograde transport route in association with detergent-resistant membranes. Mol Biol Cell 12: 2453–2468 Fridmann-Sirkis Y, Siniossoglou S, Pelham HR (2004) TMF is a golgin that binds Rab6 and influences Golgi morphology. BMC Cell Biol 5: 5–18 Galan JM, Wiederkehr A, Seol JH, Haguenauer-Tsapis R, Deshaies RJ, Riezman H, Peter M (2001) Skp1p and the F-box protein Rcy1p form a non-SCF complex involved in recycling of the SNARE Snc1p in yeast. Mol Cell Biol 21: 3105–3117 Geuze HJ, Slot JW, Strous GJ, Hasilik A, Von Figura K (1984) Ultrastructural localization of the mannose 6-phosphate receptor in rat liver. J Cell Biol 98: 2047–2054 Ghosh RN, Mallet WG, Soe TT, McGraw TE, Maxfield FR (1998) An endocytosed TGN38 chimeric protein is delivered to the TGN after trafficking through the endocytic recycling compartment in CHO cells. J Cell Biol 142: 923–936 Gillingham AK, Munro S (2003). Long coiled-coil proteins and membrane traffic. Biochim Biophys Acta 1641: 71–85 Girod A, Storrie B, Simpson JC, Johannes L, Goud B, Roberts LM, Lord JM, Nilsson T, Pepperkok R (1999) Evidence for a COPI-independent transport route from the Golgi complex to the endoplasmic reticulum. Nat Cell Biol 1: 423–430 Goda Y, Pfeffer SR (1988) Selective recycling of the mannose 6-phosphate/IGF-II receptor to the trans-Golgi network in vitro. Cell 55: 309–320 Godi A, Pertile P, Meyers R, Marra P, Di Tullio G, Iurisci C, Luini A, Corda D, De Matteis MA (1999) ARF mediates recruitment of PtdIns-4-OH kinase-beta and stimulates synthesis of PtdIns(4,5)P2 on the Golgi complex. Nat Cell Biol 1: 280–287

450

*

L. Lu and W. Hong

Griffin CT, Trejo J, Magnuson T (2005) Genetic evidence for a mammalian retromer complex containing sorting nexins 1 and 2. Proc Natl Acad Sci 102: 15173–15177 Griffiths G, Hoflack B, Simons K, Mellman I, Kornfeld S (1988) The mannose 6-phosphate receptor and the biogenesis of lysosomes. Cell 52: 329–341 Grimmer S, Iversen TG, Van Deurs B, Sandvig K (2000) Endosome to Golgi transport of ricin is regulated by cholesterol. Mol Biol Cell 11: 4205–4216 Gu F, Aniento F, Parton RG, Gruenberg J (1997) Functional dissection of COP-I subunits in the biogenesis of multivesicular endosomes. J Cell Biol 139: 1183–1195 Gullapalli A, Garrett TA, Paing MM, Griffin CT, Yang Y, Trejo J (2004) A role for sorting nexin 2 in epidermal growth factor receptor down-regulation: evidence for distinct functions of sorting nexin 1 and 2 in protein trafficking. Mol Biol Cell 15: 2143–2155 Haft CR, De la Luz Sierra M, Barr VA, Haft DH, Taylor SI (1998) Identification of a family of sorting nexin molecules and characterization of their association with receptors. Mol Cell Biol 18: 7278–7287 Haft CR, Sierra ML, Bafford R, Lesniak MA, Barr VA, Taylor SI (2000) Human orthologs of yeast vacuolar protein sorting proteins Vps26, Vps29, and Vps35: assembly into multimeric complexes. Mol Biol Cell 11: 4105–4116 Hay JC, Klumperman J, Oorschot V, Steegmaier M, Kuo CS, Scheller RH (1998) Localization, Dynamics, and protein interactions reveal distinct roles for ER and Golgi SNAREs. J Cell Biol 141: 1489–1502 He X, Li F, Chang WP, Tang J (2005) GGA proteins mediate the recycling pathway of memapsin 2 (BACE). J Biol Chem 280: 11696–11703 Hettema EH, Lewis MJ, Black MW, Pelham HR (2003) Retromer and the sorting nexins Snx4/41/42 mediate distinct retrieval pathways from yeast endosomes. EMBO J 22: 548–557 Hinners I, Wendler F, Fei H, Thomas L, Thomas G, Tooze SA (2003) AP-1 recruitment to VAMP4 is modulated by phosphorylation-dependent binding of PACS-1. EMBO Rep 4: 1182–1189 Hirst J, Lui WW, Bright NA, Totty N, Seaman MN, Robinson MS (2000) A family of proteins with g-adaptin and VHS domains that facilitate trafficking between the trans-Golgi network and the vacuole/lysosome. J Cell Biol 149: 67–80 Hirst J, Bright NA, Rous B, Robinson MS (1999) Characterization of a fourth adaptorrelated protein complex. Mol Biol Cell 10: 2787–2802 Hirst J, Motley A, Harasaki K, Peak Chew SY, Robinson MS (2003) EpsinR: an ENTH domain-containing protein that interacts with AP-1. Mol Biol Cell 14: 625–641 Hirst J, Miller SE, Taylor MJ, Von Mollard GF, Robinson MS (2004) EpsinR is an adaptor for the SNARE protein Vti1b. Mol Biol Cell 15: 5593–5602 Holthuis JC, Nichols BJ, Pelham HR (1998) The syntaxin Tlg1p mediates trafficking of chitin synthase III to polarized growth sites in yeast. Mol Biol Cell 9: 3383–3397 Hong W (2005) SNAREs and traffic. Biochim Biophys Acta 1744: 493–517 Honing S, Griffith J, Geuze HJ, Hunziker W (1996) The tyrosine-based lysosomal targeting signal in lamp-1 mediates sorting into Golgi-derived clathrin-coated vesicles. EMBO J 15: 5230–5239 Hopkins CR, Trowbridge IS (1983) Internalization and processing of transferrin and the transferrin receptor in human carcinoma A431 cells. J Cell Biol 97: 508–521 Horazdovsky BF, Davies BA, Seaman MN, McLaughlin SA, Yoon S, Emr SD (1997) A sorting nexin-1 homologue, Vps5p, forms a complex with Vps17p and is required for recycling the vacuolar protein-sorting receptor. Mol Biol Cell 8: 1529–1541 Huse JT, Pijak DS, Leslie GJ, Lee VM, Doms RW (2000) Maturation and endosomal targeting of beta-site amyloid precursor protein-cleaving enzyme. The Alzheimer’s disease beta-secretase. J Biol Chem 275: 33729–33737 Ikonomov OC, Sbrissa D, Mlak K, Deeb R, Fligger J, Soans A, Finley RL Jr, Shisheva A (2003) Active PIKfyve associates with and promotes the membrane attachment of

Retrograde endosome-to-TGN Transport

*

451

the late endosome-to-trans-Golgi network transport factor Rab9 effector p40. J Biol Chem 19: 50863–50871 Itin C, Rancano C, Nakajima Y, Pfeffer SR (1997) A novel assay reveals a role for soluble N-ethylmaleimide-sensitive fusion attachment protein in mannose 6-phosphate receptor transport from endosomes to the trans Golgi network. J Biol Chem 272: 27737–27744 Iversen TG, Skretting G, Llorente A, Nicoziani P, Van Deurs B, Sandvig K (2001) Endosome to Golgi transport of ricin is independent of clathrin and of the Rab9- and Rab11GTpases. Mol Bio Cell 12: 2099–2107 Jedd G, Mulholland J, Segev N (1997) Two new Ypt GTPases are required for exit from the yeast trans-Golgi compartment. J Cell Biol 137:563–580 Jochum A, Jackson D, Schwarz H, Pipkorn R, Singer-Kruger B (2002) Yeast Ysl2p, homologous to Sec7 domain guanine nucleotide exchange factors, functions in endocytosis and maintenance of vacuole integrity and interacts with the Arf-Like small GTPase Arl1p. Mol Cell Biol 22: 4914–4928 Johannes L, Tenza D, Antony C, Goud B (1997) Retrograde transport of KDEL-bearing Bfragment of Shiga toxin. J Biol Chem 272: 19554–19561 Jones S, Newman C, Liu F, Segev N (2000) The TRAPP complex is a nucleotide exchanger for Ypt1 and Ypt31/32. Mol Biol Cell 11: 4403–4411 Kanai F, Liu H, Field SJ, Akbary H, Matsuo T, Brown GE, Cantley LC, Yaffe MB (2001) The PX domains of p47phox and p40phox bind to lipid products of PI(3)K. Nat Cell Biol 3: 675–678 Kingsley DM, Kozarsky KF, Segal M, Krieger M (1986) Three types of low density lipoprotein receptor-deficient mutant have pleiotropic defects in the synthesis of N-linked, O-linked, and lipid-linked carbohydrate chains. J Cell Biol 102: 1576–1585 Kjer-Nielsen L, Teasdale RD, Van Vliet C, Gleeson PA (1999) A novel Golgi-localisation domain shared by a class of coiled-coil peripheral membrane proteins. Curr Biol 9: 385–388 Klumperman J, Hille A, Veenendaal T, Oorschot V, Stoorvogel W, Von Figura K, Geuze HJ (1993) Differences in the endosomal distributions of the two mannose 6-phosphate receptors. J Cell Biol 121: 997–1010 Kobayashi T, Beuchat MH, Lindsay M, Frias S, Palmiter RD, Sakuraba H, Parton RG, Gruenberg J (1999) Late endosomal membranes rich in lysobisphosphatidic acid regulate cholesterol transport. Nat Cell Biol 1: 113–118 Kornfeld S (1989) The biogenesis of lysosomes. Annu Rev Cell Biol 5: 483–525 Kottgen M, Benzing T, Simmen T, Tauber R, Buchholz B, Feliciangeli S, Huber TB, Schermer B, Kramer-Zucker A, Hopker K, Simmen KC, Tschucke CC, Sandford R, Kim E, Thomas G, Walz G (2005) Trafficking of TRPP2 by PACS proteins represents a novel mechanism of ion channel regulation. EMBO J 24: 705–716 Kurten RC, Cadena DL, Gill GN (1996) Enhanced degradation of EGF receptors by a sorting nexin, SNX1. Science 272: 1008–1010 Lammich S, Kojro E, Postina R, Gilbert S, Pfeiffer R, Jasionowski M, Haass C, Fahrenholz F (1999) Constitutive and regulated alpha-secretase cleavage of Alzheimer’s amyloid precursor protein by a disintegrin metalloprotease. Proc Natl Acad Sci 96: 3922–3927 Lewis MJ, Nichols BJ, Prescianotto-Baschong C, Riezman H, Pelham HR (2000) Specific retrieval of the exocytic SNARE Snc1p from early yeast endosomes. Mol Biol Cell 11: 23–38 Liewen H, Meinhold-Heerlein I, Oliveira V, Schwarzenbacher R, Luo G, Wadle A, Jung M, Pfreundschuh M, Stenner-Liewen F (2005) Characterization of the human GARP (Golgi associated retrograde protein) complex. Exp Cell Res 306: 24–34 Lieu ZZ, Derby MC, Teasdale RD, Hart C, Gunn P, Gleeson PA (2007) The Golgin GCC88 is required for efficient retrograde transport of Cargo from the early endosomes to the trans-Golgi network. Mol Biol Cell 18: 4979–4991

452

*

L. Lu and W. Hong

Lin SX, Mallet WG, Huang AY, Maxfield FR (2004) Endocytosed cation-independent mannose 6-phosphate receptor traffics via the endocytic recycling compartment en route to the trans-Golgi network and a subpopulation of late endosomes. Mol Biol Cell 15: 721–733 Linstedt AD, Mehta A, Suhan J, Reggio H, Hauri HP (1997) Sequence and overexpression of GPP130/GIMPc: evidence for saturable pH-sensitive targeting of a type II early Golgi membrane protein. Mol Biol Cell 8: 1073–1087 Lippincott-Schwartz J, Yuan L, Tipper C, Amherdt M, Orci L, Klausner RD (1991) Brefeldin A’s effects on endosomes, lysosomes, and the TGN suggest a general mechanism for regulating organelle structure and membrane traffic. Cell 67: 601–616 Liu YW, Huang CF, Huang KB, Lee FJ (2005) Role for Gcs1p in regulation of Arl1p at transGolgi compartments. Mol Biol Cell 16: 4024–4033 Loh E, Hong W (2004) The binary interacting network of the conserved oligomeric Golgi tethering complex. J Biol Chem 279: 24640–24648 Loh E, Peter F, Subramaniam VN, Hong W (2005) Mammalian Bet3 functions as a cytosolic factor participating in transport from the ER to the Golgi apparatus. J Cell Sci 118: 1209–1222 Lombardi D, Soldati T, Riederer MA, Goda Y, Zerial M, Pfeffer SR (1993) Rab9 functions in transport between late endosomes and the trans Golgi network in vitro. EMBO J 12: 677–682 Lu L, Horstmann H, Ng C, Hong W (2001) Regulation of Golgi structure and function by ARF-like protein 1 (Arl1). J Cell Sci 114: 4543–4555 Lu L, Hong W (2003) Interaction of Arl1-GTP with GRIP domains recruits autoantigens Golgin-97 and Golgin-245/p230 onto the Golgi. Mol Biol Cell 14: 3767–3781 Lu L, Tai G, Hong W (2004) Autoantigen Golgin-97, an effector of Arl1 GTPase, participates in traffic from the endosome to the trans-Golgi network. Mol Biol Cell 15: 4426–4443 Luke MR, Kjer-Nielsen L, Brown DL, Stow JL, Gleeson PA (2003) GRIP domain-mediated targeting of two new coiled-coil proteins, GCC88 and GCC185, to subcompartments of the trans-Golgi network. J Biol Chem 278: 4216–4226 Luo W, Chang A (1997) A Novel genes involved in endosomal traffic in yeast revealed by suppression of a targeting-defective plasma membrane ATPase mutant. J Cell Biol 138: 731–746 Mallard F, Antony C, Tenza D, Salamero J, Goud B, Johannes L (1998) Direct pathway from early/recycling endosomes to the Golgi apparatus revealed through the study of shiga toxin B-fragment transport. J Cell Biol 143: 973–990 Mallard F, Tang BL, Galli T, Tenza D, Saint-Pol A, Xu Y, Antony C, Hong W, Goud B, Johannes L (2002) Early/recycling endosomes-to-TGN transport involves two SNARE complexes and a Rab6 isoform. J Cell Biol 156: 653–664 Matanis T, Akhmanova A, Wulf P, Del Nery E, Weide T, Stepanova T, Galjart N, Grosveld F, Goud B, De Zeeuw CI, Barnekow A, Hoogenraad CC (2002) Bicaudal-D regulates COPI-independent Golgi-ER transport by recruiting the dynein-dynactin motor complex. Nat Cell Biol 4: 986–992 Mallet WG, Maxfield FR (1999) Chimeric forms of furin and TGN38 are transported with the plasma membrane in the trans-Golgi network via distinct endosomal pathways. J Cell Biol 146: 345–359 Martinez O, Antony C, Pehau-Arnaudet G, Berger EG, Salamero J, Goud B (1997) GTPbound forms of rab6 induce the redistribution of Golgi proteins into the endoplasmic reticulum. Proc Natl Acad Sci 94: 1828–1833 Medigeshi GR, Schu P (2003) Characterization of the in vitro retrograde transport of MPR46. Traffic 4: 802–811 Mills IG, Praefcke GJ, Vallis Y, Peter BJ, Olesen LE, Gallop JL, Butler PJ, Evans PR, McMahon HT (2003) EpsinR: an AP1/clathrin interacting protein involved in vesicle trafficking. J Cell Biol 160: 213–222

Retrograde endosome-to-TGN Transport

*

453

Moisenovich M, Tonevitsky A, Maljuchenko N, Kozlovskaya N, Agapov I, Volknandt W, Bereiter-Hahn J (2004) Endosomal ricin transport: involvement of Rab4- and Rab5positive compartments. Histochem Cell Biol 121: 429–439 Monier S, Jollivet F, Janoueix-Lerosey I, Johannes L, Goud B (2002) Characterization of novel Rab6-interacting proteins involved in endosome-to-TGN transport. Traffic 3: 289–397 Molloy SS, Thomas L, Kamibayashi C, Mumby MC, Thomas G (1998) Regulation of endosome sorting by a specific PP2A isoform. J Cell Biol 142: 1399–1411 Moore KL (2003) The Biology and Enzymology of Protein Tyrosin O-sulfation. J Biol Chem 278: 24243–24246 Munro S, Nichols BJ (1999) The GRIP domain – a novel Golgi-targeting domain found in several coiled-coil proteins. Curr Biol 9: 377–380 Murray JL, Mavrakis M, McDonald NJ, Yilla M, Sheng J, Bellini WJ, Zhao L, Le Doux JM, Shaw MW, Luo CC, Lippincott-Schwartz J, Sanchez A, Rubin DH, Hodge TW (2005) Rab9 GTPase is required for replication of human immunodeficiency virus type 1, filoviruses, and measles virus. J Virol 79: 11742–11751 Naslavsky N, Weigert R, Donaldson JG (2003) Convergence of Non-clathrin- and Clathrin-derived Endosomes Involves Arf6 Inactivation and Changes in Phosphoinositides. Mol Biol Cell 14: 417–431 Natarajan R, Linstedt AD (2004) A cycling cis-Golgi protein mediates endosome-to-Golgi traffic. Mol Biol Cell 15: 4798–4806 Nichols BJ, Kenworthy AK, Polishchuk RS, Lodge R, Roberts TH, Hirschberg K, Phair RD, Lippincott-Schwartz J (2001) Rapid cycling of lipid raft markers between the cell surface and Golgi complex. J Cell Biol 153: 529–541 Nichols BJ (2002) A distinct class of endosome mediates clathrin-independent endocytosis to the Golgi complex. Nat Cell Biol 4: 374–378 Nielsen MS, Madsen P, Christensen EI, Nykjaer A, Gliemann J, Kasper D, Pohlmann R, Petersen CM (2001) The sortilin cytoplasmic tail conveys Golgi–endosome transport and binds the VHS domain of the GGA2 sorting protein. EMBO J 20: 2180–2190 Nothwehr SF, Hindes AH (1997) The yeast VPS5/GRD2 gene encodes a sorting nexin-1like protein required for localizing membrane proteins to the late Golgi. J Cell Sci 110: 1063–1072 Notwehr SF, Ha S-A, Bruinsma P (2000) Sorting of yeast membrane proteins into an endosome-to-Golgi pathway involves direct interaction of their cytoplasmic domains with Vps35p. J Cell Biol 151: 297–309 Oka T, Ungar D, Hughson FM, Krieger M (2004) The COG and COPI complexes interact to control the abundance of GEARs, a subset of Golgi integral membrane proteins. Mol Biol Cell 15: 2423–2435 Oka T, Vasile E, Penman M, Novina CD, Dykxhoorn DM, Ungar D, Hughson FM, Krieger M (2005) Genetic analysis of the subunit organization and function of the COG complex: studies of Cog5 and Cog7 deficient mammalian cells. J Biol Chem 280: 32736–32745 Ooi CE, Dell’Angelica EC, Bonifacino JS (1998) ADP-Ribosylation factor 1 (ARF1) regulates recruitment of the AP-3 adaptor complex to membranes. J Cell Biol 142: 391–402 Panic B, Whyte JR, Munro S (2003a) The ARF-like GTPases Arl1p and Arl3p act in a pathway that interacts with vesicle-tethering factors at the Golgi apparatus. Curr Biol 13: 405–410 Panic B, Perisic O, Veprintsev DB, Williams RL, Munro S (2003b) Structural basis for Arl1dependent targeting of homodimeric GRIP domains to the Golgi apparatus. Mol Cell 12: 863–874 Pelkmans L, Kartenbeck J, Helenius A (2001) Nat Cell Biol 3: 473–483 Pereira-Leal JB, Seabra MC (2000) The mammalian Rab family of small GTPases: definition of family and subfamily sequence motifs suggests a mechanism for functional specificity in the Ras superfamily. J Mol Biol 3: 1077–1087

454

*

L. Lu and W. Hong

Pfeffer SR (1999) Transport-vesicle targeting: tethers before SNAREs. Nat Cell Biol 1: E17–E22 Piguet V, Gu F, Foti M, Demaurex N, Gruenberg J, Carpentier JL, Trono D (1999) Nefinduced CD4 degradation: a diacidic-based motif in Nef functions as a lysosomal targeting signal through the binding of beta-COP in endosomes. Cell 97: 63–73 Piguet V, Wan L, Borel C, Mangasarian A, Demaurex N, Thomas G, Trono D (2000) HIV-1 Nef protein binds to the cellular protein PACS-1 to downregulate class I major histocompatibility complexes. Nat Cell Biol 2: 163–167 Piper RC, Cooper AA, Yang H, Stevens TH (1995) VPS27 controls vacuolar and endocytic traffic through a prevacuolar compartment in Saccharomyces cerevisiae. J Cell Biol 131: 603–617 Press B, Feng Y, Hoflack B, Wandinger-Ness (1998) A Mutant Rab7 causes the accumulation of cathepsin D and cation-independent mannose 6-phosphate receptor in an early endocytic compartment. J Cell Biol 140: 1075–1089 Puri S, Bachert C, Fimmel CJ, Linstedt AD (2002) Cycling of early Golgi proteins via the cell surface and endosomes upon lumenal pH disruption. Traffic 3: 641–653 Rechards M, Xia W, Oorschot V, Van Dijk S, Annaert W, Selkoe DJ, Klumperman J (2006) Presenilin-1-mediated retention of APP derivatives in early biosynthetic compartments. Traffic 7: 354–364 Reddy P, Krieger M (1989) Isolation and characterization of an extragenic suppressor of the low-density lipoprotein receptor-deficient phenotype of a Chinese hamster ovary cell mutant. Mol Cell Biol 9: 4799–4806 Reddy JV, Burguete AS, Sridevi K, Ganley IG, Nottingham RM, Pfeffer SR (2006) A functional role for the GCC185 golgin in mannose 6-phosphate receptor recycling. Mol Biol Cell 17: 4353–4363 Reggiori F, Wang CW, Stromhaug PE, Shintani T, Klionsky DJ (2003) Vps51 is part of the yeast Vps fifty-three tethering complex essential for retrograde traffic from the early endosome and Cvt vesicle completion. J Biol Chem 278: 5009–5020 Riederer MA, Soldati T, Shapiro AD, Lin J, Pfeffer SR (1994) Lysosome biogenesis requires mannose 6-phosphate receptor recycling from endosomes to the trans-Golgi network. J Cell Biol 125: 573–582 Rothman JE, Wieland FT (1996) Protein sorting by transport vesicles. Science 272: 227–234 Sacher M, Jiang Y, Barrowman J, Scarpa A, Burston J, Zhang L, Schieltz D, Yates JR 3rd, Abeliovich H, Ferro-Novick S (1998) TRAPP, a highly conserved novel complex on the cis-Golgi that mediates vesicle docking and fusion. EMBO J 17: 2494–2503 Sacher M, Barrowman J, Schieltz D, Yates JR 3rd, Ferro-Novick S (2000) Identification and characterization of five new subunits of TRAPP. Eur J Cell Biol 79: 71–80 Sacher M, Barrowman J, Wang W, Horecka J, Zhang Y, Pypaert M, Ferro-Novick S (2001) TRAPP I implicated in the specificity of tethering in ER-to-Golgi transport. Mol Cell 7: 433–442 Saint-Pol A, Yelamos B, Amessou M, Mills IG, Dugast M, Tenza D, Schu P, Antony C, McMahon HT, Lamaze C, Johannes L (2004) Clathrin adaptor epsinR is required for retrograde sorting on early endosomal membranes. Dev Cell 6: 525–538 Sandvig K, Olsnes S, Brown JE, Petersen OW, Van Deurs B (1989) Endocytosis from coated pits of Shiga toxin: a glycolipid-binding protein from Shigella dysenteriae 1. J Cell Biol 108: 1331–1343 Sandvig K, Garred O, Prydz K, Kozlov JV, Hansen SH, Van Deurs B (1992) Retrograde transport of endocytosed Shiga toxin to the endoplasmic reticulum. Nature 358: 510–512 Sandvig K, Van Deurs B (2002) Transport of protein toxins into cells: pathways used by ricin, cholera toxin and Shiga toxin. FEBS Lett 529: 49–53 Scales SJ, Gomez M, Kreis TE (2000) Coat proteins regulating membrane traffic. Int Rev Cytol 195: 67–144

Retrograde endosome-to-TGN Transport

*

455

Schermer B, Hopker K, Omran H, Ghenoiu C, Fliegauf M, Fekete A, Horvath J, Kottgen M, Hackl M, Zschiedrich S, Huber TB, Kramer-Zucker A, Zentgraf H, Blaukat A, Walz G, Benzing T (2005) Phosphorylation by casein kinase 2 induces PACS-1 binding of nephrocystin and targeting to cilia. EMBO J 24: 4415–4424 Schwarz DG, Griffin CT, Schneider EA, Yee D, Magnuson T (2002) Genetic analysis of sorting nexins 1 and 2 reveals a redundant and essential function in mice. Mol Biol Cell 13: 3588–3600 Sciorra VA, Audhya A, Parsons AB, Segev N, Boone C, Emr SD (2005) Synthetic genetic array analysis of the PtdIns 4-kinase Pik1p identifies components in a Golgi-specific Ypt31/rab-GTPase signaling pathway. Mol Biol Cell 16: 776–793 Seaman MNJ, Marcusson EG, Cereghino JL, Emr SD (1997) Endosome to Golgi retrieval of the vacuolar protein sorting receptor, Vps10p, requires the function of the VPS29, VPS30, and VPS35 gene products. J Cell Biol 137: 79–92 Seaman MNJ, McCaffery JM, Emr SD (1998) A membrane coat complex essential for endosome-to-Golgi retrograde transport in yeast. J Cell Biol 142: 665–681 Seaman MNJ (2004) Cargo-selective endosomal sorting for retrieval to the Golgi requires retromer. J Cell Biol 165: 111–122 Seet LF, Hong W (2006) The Phox (PX) domain proteins and membrane traffic. Biochim Biophys Acta 1761: 878–896 Setty SR, Shin ME, Yoshino A, Marks MS, Burd CG (2003) Golgi recruitment of GRIP domain proteins by Arf-like GTPase 1 is regulated by Arf-like GTPase 3. Curr Biol 13: 401–404 Setty SR, Strochlic TI, Tong AH, Boone C, Burd CG (2004) Golgi targeting of ARF-like GTPase Arl3p requires its Nalpha-acetylation and the integral membrane protein Sys1p. Nat Cell Biol 6: 414–419 Shewan AM, Van Dam EM, Martin S, Luen TB, Hong W, Bryant NJ, James DE (2003) GLUT4 Recycles via a trans-Golgi Network (TGN) Subdomain enriched in syntaxins 6 and 16 but not TGN 38, involvement of an acidic targeting motif. Mol Biol Cell 14: 973–986 Shin HW, Kobayashi H, Kitamura M, Waguri S, Suganuma T, Uchiyama Y, Nakayama K (2005) Roles of ARFRP1 (ADP-ribosylation factor-related protein 1) in post-Golgi membrane trafficking. J Cell Sci 118: 4039–4048 Short B, Preisinger C, Schaletzky J, Kopajtich R, Barr FA (2002) The Rab6GTPase regulates recruitment of the dynactin complex to Golgi membranes. Curr Biol 12: 1792–1795 Siniossoglou S, Peak-Chew SY, Pelham HR (2000) Ric1p and Rgp1p form a complex that catalyses nucleotide exchange on Ypt6p. EMBO J 19: 4885–4894 Siniossoglou S, Pelham HR (2001) An effector of Ypt6p binds the SNARE Tlg1p and mediates selective fusion of vesicles with late Golgi membranes. EMBO J 20: 5991– 5998 Smith DC, Spooner RA, Watson PD, Murray JL, Hodge TW, Amessou M, Johannes L, Lord JM, Roberts LM (2006) Internalized Pseudomonas exotoxin A can exploit multiple pathways to reach the endoplasmic reticulum. Traffic 7: 379–393 Small SA, Kent K, Pierce A, Leung C, Kang MS, Okada H, Honig L, Vonsattel JP, Kim TW (2005) Model-guided microarray implicates the retromer complex in Alzheimer’s disease. Ann Neurol 58: 909–919 Spelbrink RG, Nothwehr SF (1999) The yeast GRD20 gene is required for protein sorting in the trans-Golgi network/endosomal system and for polarization of the actin cytoskeleton. Mol Biol Cell 10: 4263–4281 Stamnes MA, Rothman JE (1993) The binding of AP-1 clathrin adaptor particles to Golgi membranes requires ADP-ribosylation factor, a small GTP-binding protein. Cell 73: 999–1005 Stephens DJ, Austen BM (1996) Metabolites of the beta-amyloid precursor protein generated by beta-secretase localise to the trans-Golgi network and late endosome in 293 cells. J Neurosci Res 46: 211–225

456

*

L. Lu and W. Hong

Stoorvogel W, Oorschot V, Geuze HJ (1996) A novel class of clathrin-coated vesicles budding from endosomes. J Cell Biol 132: 21–33 Suvorova ES, Kurten RC, Lupashin VV (2001) Identification of a human orthologue of Sec34p as a component of the cis-Golgi vesicle tethering machinery. J Biol Chem 276: 22810–22818 Suvorova ES, Duden R, Lupashin VV (2002) The Sec34/Sec35p complex, a Ypt1p effector required for retrograde intra-Golgi trafficking, interacts with Golgi SNAREs and COPI vesicle coat proteins. J Cell Biol 157: 631–643 Tai G, Lu L, Wang TL, Tang BL, Goud B, Johannes L, Hong W (2004) Participation of the syntaxin 5/Ykt6/GS28/GS15 SNARE complex in transport from the early/recycling endosome to the trans-Golgi network. Mol Biol Cell 15: 4011–4022 Takatsu H, Yoshino K, Nakayama K (2000) Adaptor gamma ear homology domain conserved in gamma-adaptin and GGA proteins that interact with gamma-synergin. Biochem Biophys Res Comm 271:719–725 Torgersen ML, Skretting G, Van Deurs B, Sandvig K (2001) Internalization of cholera toxin by different endocytic mechanisms J Cell Sci 114: 3737–3747 Tran TH, Zeng Q, Hong W (2004) VAMP4 cycles from the cell surface to the trans-Golgi network via sorting and recycling endosomes. J Cell Sci 120: 1028–1041 Traub LM, Ostrom JA, Kornfeld S (1993) Biochemical dissection of AP-1 recruitment onto Golgi membranes. J Cell Biol 123: 561–573 Tsukada M, Will E, Gallwitz D (1999) Structural and functional analysis of a novel coiledcoil protein involved in Ypt6 GTPase-regulated protein transport in yeast. Mol Biol Cell 10: 63–75 Tuma PL, Hubbard AL (2003) Transcytosis: crossing cellular barriers. Physiol Rev 83: 871–932 Ullrich O, Reinsch S, Urbe S, Zerial M, Parton R (1996) Rab11 regulates recycling through the pericentriolar recycling endosome. J Cell Biol 135: 913–924 Ungar D, Oka T, Brittle EE, Vasile E, Lupashin VV, Chatterton JE, Heuser JE, Krieger M, Waters MG (2002) Characterization of a mammalian Golgi-localized protein complex, COG, that is required for normal Golgi morphology and function. J Cell Biol 157: 405–415 Ungar D, Oka T, Vasile E, Krieger M, Hughson FM (2005) Subunit map of the Conserved Oligomeric Golgi complex. J Biol Chem 280: 32729–32735 Urbe S, Huber LA, Zerial M, Tooze SA, Parton RG (1993) Rab11, a small GTPase associated with both constitutive and regulated secretory pathways in PC12 cells. FEBS Lett 334: 175–182 Valdivia RH, Baggott D, Chuang JS, Schekman RW (2002) The yeast clathrin adaptor protein complex 1 is required for the efficient retention of a subset of late Golgi membrane proteins. Dev Cell 2: 283–294 Van Valkenburgh H, Shern JF, Sharer JD, Zhu X, Kahn RA (2001) ADP-ribosylation factors (ARFs) and ARF-like 1 (ARL1) have both specific and shared effectors: characterizing ARL1-binding proteins. J Biol Chem 276: 22826–22837 Volchuk A, Ravazzola M, Perrelet A, Eng WS, Di Liberto M, Varlamov O, Fukasawa M, € llner TH, Rothman JE, Orci L (2004) Countercurrent distribution of two Engel T, So distinct SNARE complexes mediating transport within the Golgi stack. Mol Biol Cell 15: 1506–1518 Voos W, Stevens TH (1998) Retrieval of resident late-Golgi membrane proteins from the prevacuolar compartment of Saccharomyces cerevisiae is dependent on the function of Grd19p. J Cell Biol 140: 577–590  Y, Uchiyama Y, Dubremetz JF, Hoflack B (2003) Waguri S, Dewitte F, Le Borgne R, Rouille Visualization of TGN to endosome trafficking through fluorescently labeled MPR and AP-1 in living cells. Mol Biol Cell 14: 142–155 Walter J, Fluhrer R, Hartung B, Willem M, Kaether C, Capell A, Lammich S, Multhaup G, Haass C (2001) Phosphorylation regulates intracellular trafficking of beta-secretase. J Biol Chem 276: 14634–14641

Retrograde endosome-to-TGN Transport

*

457

Wan L, Molloy SS, Thomas L, Liu G, Xiang Y, Rybak SL, Thomas G (1998) PACS-1 defines a novel gene family of cytosolic sorting proteins required for trans-Golgi network localization. Cell 94: 205–216 Wang W, Sacher M, Ferro-Novick S (2000) TRAPP stimulates guanine nucleotide exchange on Ypt1p. J Cell Biol 151: 289–296 Wang Y, Tai G, Lu L, Johannes L, Hong W, Luen Tang B (2005) Trans-Golgi network syntaxin 10 functions distinctly from syntaxins 6 and 16. Mol Membr Biol 22: 313–325 Wang Y, Zhou Y, Szabo K, Haft CR, Trejo J (2002) Down-regulation of protease-activated receptor-1 is regulated by sorting nexin 1. Mol Biol Cell 13: 1965–1976 West MA, Bright NA, Robinson MS (1997) The role of ADP-ribosylation factor and phospholipase D in adaptor recruitment. J Cell Biol 138: 1239–1254 White J, Johannes L, Mallard F, Girod A, Grill S, Reinsch S, Keller P, Tzschaschel B, Echard A, Goud B, Stelzer EH (1999) Rab6 coordinates a novel Golgi to ER retrograde transport pathway in live cells. J Cell Biol 147: 743–760 Whyte JR, Munro S (2001) The Sec34/35 Golgi transport complex is related to the exocyst, defining a family of complexes involved in multiple steps of membrane traffic. Dev Cell 1: 527–537 Whyte JR, Munro S (2002) Vesicle tethering complexes in membrane traffic. J Cell Sci 115: 2627–2637 Willingham MC, Pastan IH, Sahagian GG, Jourdian GW, Neufeld EF (1981) Morphologic study of the internalization of a lysosomal enzyme by the mannose 6-phosphate receptor in cultured Chinese hamster ovary cells. Proc Natl Acad Sci 78: 6967–6971 Wilsbach K, Payne GS (1993) Dynamic retention of TGN membrane proteins in Saccharomyces cerevisiae. Trends Cell Biol 3: 426–432 Wilcke M, Johannes L, Galli T, Mayau V, Goud B, Salamero J (2000) Rab11 regulates the compartmentalization of early endosomes required for efficient transport from early endosomes to the trans-Golgi network. J Cell Biol 151: 1207–1220 Wolins NE, Rubin B, Brasaemle DL (2001) TIP47 associates with lipid droplets. J Biol Chem 276: 5101–5108 Wu M, Lu L, Hong W, Song H (2004a) Structural basis for recruitment of GRIP domain golgin-245 by small GTPase Arl1. Nat Struct Mol Biol 11: 86–94 Wu X, Steet RA, Bohorov O, Bakker J, Newell J, Krieger M, Spaapen L, Kornfeld S, Freeze HH (2004b) Mutation of the COG complex subunit gene COG7 causes a lethal congenital disorder. Nat Med 10: 518–523 Xiang Y, Molloy SS, Thomas L, Thomas G (2000) The PC6B cytoplasmic domain contains two acidic clusters that direct sorting to distinct trans-Golgi network/endosomal compartments. Mol Biol Cell 11: 1257–1273 Xu Y, Hortsman H, Seet L, Wong SH, Hong W (2001) SNX3 regulates endosomal function through its PX-domain-mediated interaction with PtdIns(3)P. Nat Cell Biol 3: 658–666 Xu Y, Martin S, James DE, Hong W (2002) GS15 forms a SNARE complex with syntaxin 5, GS28 and Ykt6 and is implicated in traffic in the early cisternae of the Golgi apparatus. Mol Biol Cell 13: 3493–3507 Yahara N, Ueda T, Sato K, Nakano A (2001) Multiple roles of Arf1 GTPase in the yeast exocytic and endocytic pathways. Mol Biol Cell 12: 221–238 Yamakawa H, Seog DH, Yoda K, Yamasaki M, Wakabayashi T (1996) Uso1 protein is a dimer with two globular heads and a long coiled-coil tail. J Struct Biol 116: 3563–3565 Yamashiro DJ, Maxfield FR (1984) Acidification of endocytic compartments and the intracellular pathways of ligands and receptors. J Cell Biochem 26: 231–246 Yoshino A, Setty SR, Poynton C, Whiteman EL, Saint-Pol A, Burd CG, Johannes L, Holzbaur EL, Koval M, McCaffery JM, Marks MS (2005) tGolgin-1 (p230, golgin245) modulates Shiga-toxin transport to the Golgi and Golgi motility towards the microtubule-organizing centre. J Cell Sci May 118: 2279–2293

458

*

L. Lu and W. Hong

Yu JW, Lemmon MA (2001) All phox homology (PX) domains from Saccharomyces cerevisiae specifically recognize phosphatidylinositol 3-phosphate. J Biol Chem 276: 44179–44184 Zerial M, McBride H (2001) Rab proteins as membrane organizers. Nat Rev Mol Cell Biol 2: 107–117 Zhang T, Hong W (2001) Ykt6 forms a SNARE complex with syntaxin 5, GS28, and Bet1 and participates in a late stage in endoplasmic reticulum-Golgi transport. J Biol Chem 276: 27480–27487 Zhong Q, Lasar CS, Tronchere H, Sato T, Meerloo T, Yeo M, Songyang Z, Emr SD, Gill GN (2002) Endosomal localization and function of sorting nexin 1. Proc Natl Acad Sci 99: 6767–6772 Zolov SN, Lupashin VV (2005) Cog3p depletion blocks vesicle-mediated Golgi retrograde trafficking in HeLa cells. J Cell Biol 168: 747–759

Retrograde plasma membrane-to-Golgi apparatus transport

*

459

Retrograde plasma membrane-to-Golgi apparatus transport Margit Pavelka and Adolf Ellinger

Introduction Traffic between the plasma membrane and the Golgi apparatus includes multiple routes, which are parts of the biosynthetic-secretory system, of the endocytic system, and of circuits allowing the recycling of molecules from the plasma membrane to the Golgi apparatus, and vice versa from the Golgi apparatus to cell surfaces, both from and to apical, and basolateral cell surface areas (for review e.g. Bonifacino and Rojas 2006; de Matteis and Luini 2008; Farquhar and Palade 1998; Maxfield and McGraw 2004; Pelkmans et al. 2005; Sannerud et al. 2003; Starr et al. 2007). None of these transport routes can be seen separately, since multiple connections and crossing areas exist. This chapter focuses on the retrograde routes from the plasma membrane to the Golgi apparatus. With reference to the main topic of the book, it deals particularly with the influences of retrograde transport on the Golgi apparatus organization and architecture. The Golgi apparatus not only is a central station in the traffic of the biosynthetic system, and takes up newly synthesized membranes and contents coming from the endoplasmic reticulum (ER) at the cis side of its stacks but receives input from the endocytic system as well (for review e.g. Berger 1985; Farquhar and Hauri 1997; Glick 2000; Malhotra and Mayor 2006; Marsh and Howell 2002; Mellman and Warren 2000; Pavelka et al. 2008; Pelham and Rothman 2000; Puthenveedu and € sch 2005; Roth 1997; Storrie 2005). Linstedt 2005; Rodriguez-Boulan and Mu Endocytic import into the Golgi apparatus mainly takes place at the trans side of the stacks, where an endocytic trans-Golgi network (endocytic TGN) is formed, parts of which subsequently become integrated into the stacks of Golgi cisternae. The organization and architecture of the Golgi apparatus is influenced in response to endocytic flow; it is one of the main questions, how these changes mutually influence the other Golgi functions and dynamics. Harmful substances, such as plant and bacterial toxins, are known to travel routes that involve the Golgi apparatus (for review e.g. Sandvig and Van Deurs 2005). Furthermore, the Golgi apparatus is particularly interesting for the development of strategies for targeted drug delivery to the interior of cells (for review e.g. Tarrago-Trani and Storrie 2007).

460

*

M. Pavelka and A. Ellinger

Retrograde transport routes from the plasma membrane to the Golgi apparatus Retrograde recycling routes Anterograde traffic of newly synthesized molecules to the final destinations at the plasma membrane, in the extracellular space, in secretory granules, endosomes, and lysosomes, as well as anterograde traffic in recycling circuits, is counterbalanced by retrograde transport, and a backward flow of membranes and membrane constituents to the Golgi apparatus and trans-Golgi network (TGN). Examples are the transport routes of resident Golgi, TGN, and plasma membrane glycoproteins, and receptors, which are involved in the sorting of newly synthesized molecules out of the trans-Golgi apparatus and TGN to their specific sites of action. In this bi-directional transport, receptors recycle to the TGN and Golgi apparatus, where they may be modified, and from where they again start to be reused in further rounds (Bonifacino and Rojas 2006; Rohn et al. 2000; Snyder and Rogers 1985; Volz et al. 1995). The cationindependent and the cation-dependent mannose-6-phosphate receptors (CI- and CD-M6PR), the multi-ligand receptor sortilin, TGN38, GPP130, and GP73, the transmembrane endoproteases furin and carboxypeptidase D, and SNAREs (soluble N-maleimide-sensitive fusion protein/NSF/attachment protein receptor) are prominent examples (Banting and Ponnambalam 1997; Ghosh et al. 1998; 2003; Hettema et al. 2003; Hong 2005; Mari et al. 2008; Molloy et al. 1999; Puri et al. 2002; Stanley and Howell 1993; Varlamov and Fricker 1998). At least two independent retrograde routes from early endosomes to the TGN exist. One way to the TGN uses routes via late endosomes (Barbero et al. 2002; Carroll et al. 2001; Lombardi et al. 1993; Mallet and Maxfield 1999), the other one bypasses late endosomes, and leads directly from early and/or recycling endosomes to the TGN (Mallard et al. 1998, 2002). Among the multiple constituents of the molecular transport machineries, which have been characterized during the past years, are PACS1 (phosphofurin acidic cluster sorting protein 1), and AP-1 (Crump et al. 2001; Meyer et al. 2000), Rab9 and TIP47 (tail-interacting protein 47 kD; Diaz et al. 1997), EpsinR (Saint-Pol et al. 2004), the t-SNAREs syntaxin 16 and 5 (Amessou et al. 2007), and the two retromer subcomplexes, a membranebound coat that consists of the phosphoinositide-binding proteins sorting nexin 1 and possibly sorting nexin 2, and the cargo-binding proteins Vps26, Vps29, and Vps35 (Arighi et al. 2004; Gokool et al. 2007; Popoff et al. 2007; Restrepo et al. 2007; Rojas et al 2007; Seaman 2004, 2005; for review Bonifacino and Rojas 2006). In the retrograde plasma membrane to TGN traffic, complex pleomorphic endosomal compartments are involved consisting of vacuolar, tubular, and vesicular parts; at the same vacuolar early endosome, various complex carrier compartments can form: tubular endosomal networks (Bonifacino and Rojas 2006), tubular sorting endosomes (Peden et al. 2004), and the recently

Retrograde plasma membrane-to-Golgi apparatus transport

*

461

described endosome-to-TGN carriers (ETCs; Mari et al. 2008), which are clearly different from the others, and characterized by non-branched tubules and vesicles. Vacuolar parts of early endosomes exhibit bilayered coats composed of clathrin and Hrs (hepatocyte-growth-factor-regulated tyrosine kinase substrate), which are thought to contain the ESCRT machinery that targets proteins to the intraluminal vesicles to be sorted to the multivesicular-late endosomal–lysosomal route (Hurley and Emr 2006; Sachse et al. 2002). It is proposed that the vacuolar endosomal part receives at the one hand endocytic cargo from the plasma membrane, and at the other hand, biosynthetic cargo, e.g. lysosomal enzymes, from the TGN. Acidification of the interior leads to a release of the cargo from the receptors. The vacuolar endosomal compartments are suggested to mature into late endosomes, whereas from tubular endosomal networks (TEN; Bonifacino and Rojas 2006) being connected to the vacuolar regions, cargos are sorted to different destinations, to the TGN and Golgi apparatus along the retrograde routes, to the plasma membrane for recycling and transcytosis, and to specialized storage compartments, such as melanosomes. Different transport machineries, including the retromer complex, are present at the tubular networks. The pericentriolar endocytic recycling compartment (Ghosh et al. 1998; Maxfield and McGraw 2004; Ullrich et al. 1996) is considered as possibly being a specialized subdomain of the endocytic tubular networks. Concerning its central sorting role, TEN has been suggested to be a “mirror image” of the TGN (Bonifacino and Rojas 2006). Furthermore, there exists a similarity with the endocytic trans-Golgi network, which develops after internalization of wheat germ agglutinin (Pavelka et al. 1998, 2008; Vetterlein et al. 2002; Figs. 1–4).

Retrograde routes of internalized molecules Retrograde pathways from the cell surface to the TGN and Golgi apparatus are used not only by recycling molecules and plasma membrane glycoproteins for reentry into the biosynthetic system to be modified and repaired prior to reuse (Snyder and Rogers 1985; Volz et al. 1995; reviewed in Bonifacino and Rojas 2006) but are also travelled by extracellular ligands, including harmful substances, such as bacterial and plant toxins (for review Sandvig and Van Deurs 2005); furthermore, retrograde plasma membrane to Golgi apparatus routes provide tracks for drug delivery to the interior of cells (for review Tarragó-Trani and Storrie 2007). It is known since more than two decades (e.g. Gonatas et al. 1977, 1983; Pelham et al. 1992; Sandvig and Brown 1987; Van Deurs et al. 1986, 1987) that lectins, and bacterial and plant toxins, such as the Shiga and Cholera toxins, the Pseudomonas exotoxin A, and ricin, misuse physiologic retrograde pathways within cells, and along such routes are transported to the TGN and Golgi apparatus, to the ER, and eventually, to the cytosol. Different retrograde pathways are utilized by different toxins, and there appear to be fundamental differences in the sorting into the retrograde pathways of different toxins, e.g. Shiga and cholera toxins (Bujny et al. 2007; Chinnapen et al. 2007; Feng

462

*

M. Pavelka and A. Ellinger

Figure 1. (a,b) HepG2 hepatoma cell – WGA-internalization 30 min. (a) Endosomes containing reaction products for internalized WGA (black arrows) are accumulated close to small stacks of Golgi apparatus cisternae (G); endosomal networks (white arrows) are visible nearby. The arrowhead indicates a tiny bridge interconnecting two endosomes. Internalized WGA also is apparent at limited regions of stacked Golgi cisternae (curved arrow). ER trans-Golgi endoplasmic reticulum; asterisks ER export sites. 22,125. (b) A stack of Golgi cisternae is accompanied by an early endocytic TGN, which consists of prominent large, frequently clathrin coated elements (arrows), interconnected by thinner and branched network structures. The curved arrow indicates WGA-uptake into a cisterna of the stack. 33,600.

Retrograde plasma membrane-to-Golgi apparatus transport

*

463

Figure 2. HepG2 hepatoma cell – WGA internalization 30 min. Several Golgi apparatus stacks are associated with an extended endocytic TGN, which consists of “free” parts (white arrows), and other parts that are attached to the transmost cisterna of a Golgi stack (black arrows). The endocytic TGN interconnects several individual stacks of cisternae, thus forming a Golgi stack ribbon. ER trans-Golgi endoplasmic reticulum; MVB multivesiculated bodies; N nucleus.

21,300.

et al. 2004; Massol et al. 2004; Torgersen et al. 2001); even one class of toxin not only uses one route, as has been shown with Pseudomonas exotoxin A, which is transported to the endoplasmic reticulum along multiple pathways (Smith et al. 2006). Sandvig et al. (1992) were the first, who demonstrated that Shiga toxin is transported from the cell surface en route to the ER. Shiga toxin contains an enzymatically active A-subunit non-covalently linked to a pentamer of B-chains, which binds to the glycosphingolipid Gb3, and mediates its transport. The intracellular route of the toxin after internalization involves early endosomes, the TGN, the Golgi apparatus, and the ER, from

464

*

M. Pavelka and A. Ellinger

Figure 3. (a,b) HepG2 hepatoma cells – WGA internalization 30 min (a) and 45 min (b). WGA is apparent within the cisternae of these Golgi stacks. The reactions are confined to limited regions of two cisternae in a, and are contained in all of the stacked cisternae in b. TGN Endocytic transGolgi network; MVB multivesiculated bodies; N nucleus. 24,400 (a), 32,500 (b).

where after cleavage the A-subunit is retrotranslocated to the cytosol, the site of its toxic effect by inactivation of ribosomes, and inhibition of protein synthesis (Johannes and Goud 1998, 2000; Sandvig and Van Deurs 1994, 1996,

"

Figure 4. (a–f) Different views of a 3D-model of an endocytic TGN of a HepG2 hepatoma cell after 30 min WGA-internalization, as revealed from electron tomography analyses. The stacked cisternae are coloured in yellow, the endocytic TGN is orange-coloured, and trans-Golgi ER is coloured in blue. In a, b, c, and d, the model is shown from different sides; in c and d, the ER has been removed; e and f show details of c and d. The endocytic TGN shown in this model consists of “free” parts (arrows), and another part, which is attached to the transmost Golgi cisterna (white arrows). In both parts, large cylindrical elements are interconnected by bridges, and broad tubular, and crest-like architectures.

Retrograde plasma membrane-to-Golgi apparatus transport

*

465

466

*

M. Pavelka and A. Ellinger

2002, 2005). During the past years, multiple details concerning uptake mechanisms, and the retrograde transport of Shiga toxin and ricin became elucidated (Amessou et al. 2007; Garred et al. 1995; Grimmer et al. 2006; Johannes et al. 1997; Lauvrak et al. 2004, 2006; Mallard et al. 1998; Rapak € mer et al. 2007; Skanland et al. 2007; Slominska-Wojewodzka et al. 1997; Ro et al. 2006; Tai et al. 2004; Torgersen et al. 2007; Utskarpen et al. 2006, 2007; Yoshino et al. 2005). From early endosomes, the toxins are transported to the TGN along a direct pathway, which is travelled by endogenous proteins as well (for review Bonifacino and Rojas 2006). It has been shown that the retromer complex is required for efficient transport of Shiga toxin from early endosomes to the TGN and Golgi apparatus (Buijy et al. 2007; Popoff et al. 2007; Utskarpen et al. 2007). The endosome-to-TGN and Golgi transport of ricin is facilitated by depletion of sphingolipid (Grimmer et al. 2006). Furthermore, there is a dependence on Rab6A and Rab6A0 (Utskarpen et al. 2006). The syntaxins 5 and 16, which are involved in the retrograde traffic of mannose-6-phosphate receptors, are also required for efficient retrograde transport of Shiga toxin, as well as for transport of ricin and cholera toxin (Amessou et al. 2007). Furthermore, regulatory functions of phosphoinositides (for review De Matteis and Godi 2004), and signalling events (Pelkmans et al. 2005; Von Zastrow and Sorkin 2007) are particularly interesting in connection with retrograde plasma membrane to Golgi trafficking. It has been shown recently that retrograde transports of Shiga toxin and ricin are phosphoinositide-regulated (Skanland et al. 2007; Utskarpen et al. 2007), and that the phosphoinositide-binding proteins sorting nexins 1 and 2, which are parts of the retromer complex (Seaman 2004, 2005), are required for efficient transport of Shiga toxin to the Golgi apparatus (Buijy et al. 2007; Popoff et al. 2007; Utskarpen et al. 2007). Shiga toxin has been shown to be an active player in its own transport, mediating both its internalization, and intracellular transport. Signalling cascades are triggered upon binding of the toxin to the plasma membrane or its entry into cells (Ikeda et al. 2000; Lauvrak et al. €lchli et al. 2008). The tyrosine kinase Syk is activated, by which 2006; Wa clathrin phosphorylation and uptake of Shiga toxin is induced (Lauvrak et al. 2006); activation of protein kinase Cd by the Shiga toxin is important for endosome-to-Golgi transport (Torgersen et al. 2007). By modifying Ca2þ homeostasis, Shiga toxin is proposed to recruit the MAP-kinase p38 to €lchli et al. endosomes, thus regulating transport to the Golgi apparatus (Wa 2008).

Changes of Golgi apparatus and TGN induced by retrograde traffic, and entry of internalized molecules into Golgi cisternae Internalized molecules not only are transported to the TGN but may enter the cisternae of the Golgi apparatus stacks as well (for review Pavelka et al.

Retrograde plasma membrane-to-Golgi apparatus transport

*

467

1998, 2008). Although this is well established, and for several toxins it has been demonstrated ultrastructurally very early, that not only the TGN is involved in retrograde transport but the toxins also are taken up into the stacked cisternae of the Golgi apparatus (e.g. Sandvig et al. 1992; Van Deurs et al. 1987), the mechanisms involved in these transport processes are still poorly understood. In a HepG2 hepatoma cell model, detailed ultrastructural endocytosis studies have been performed with wheat germ agglutinin labelled with horseradish peroxidase (WGA-HRP; Pavelka et al. 1998; Vetterlein et al. 2002). The findings revealed that endocytic transport into the Golgi apparatus is a complex multistep process, during which the Golgi architecture changes, the trans-Golgi side is reorganized, an endocytic TGN is formed, parts of which become integrated into the Golgi stacks. Wheat germ agglutinin is an N-acetyl-glucosamine- and sialic acid-binding lectin (Goldstein and Hayes 1978), and is a particularly interesting substance € ck for developments of drug delivery systems (Lochner et al. 2003; Weissenbo et al. 2004). For many years, it is known that WGA is transported to the Golgi apparatus in retrograde direction (e.g. Gonatas et al. 1977; Stieber et al. 1984). Wheat germ agglutinin reacts with multiple binding sites at the cell surface, and is internalized in large amounts. For endocytosis studies, in which particular attention is drawn to the Golgi apparatus, HepG2 hepatoma cells are especially well suited, since prior to, and in the initial phases of WGA endocytosis, the Golgi apparatus is small and inconspicuous, and its reorganizations during WGA-endocytosis can clearly be examined and visualized. Wheat germ agglutinin is internalized via clathrin-coated vesicles and possibly other mechanisms as well, and is rapidly transported to the Golgi apparatus. Early endosomes, in thin sections appearing as organelles with round profiles, accumulate at the trans-Golgi side (Fig. 1a). Some of the endosomes appear to contact each other; fine bridges are visible (Fig. 1a). Subsequently, an endocytic trans-Golgi network (endocytic TGN) is formed, which contains elements, the dimensions and shapes of which resemble the earlier endosomes. These prominent large elements are interconnected by thinner branched network-forming structures (Figs.1a,b, 2). Parts of the endocytic TGN attach to trans-Golgi cisternae, thus becoming integrated parts of the Golgi stacks (Fig. 2). Fully formed, the endocytic TGN consists of two different but continuous portions, being either attached to the transmost Golgi cisterna (Fig. 2, black arrows), or residing “free” apart from the stack without contacting the cisternae (Fig. 2, white arrows). This organization of the endocytic TGN leads to interconnections of the former small Golgi stacks, present prior to endocytosis, and causes the formation of Golgi stack ribbons. The endocytic TGN is often closely associated with trans-Golgi-ER (Figs. 1a, 2). Electron tomography analyses and 3D-reconstructions of the endocytic TGN clarified that the prominent large elements mostly are cylindrical compartments, which are interconnected by bridges, and broad tubular, and crest-like architectures (Fig. 4).

468

*

M. Pavelka and A. Ellinger

Concomitantly with, and in part, prior to the formation of an endocytic TGN, internalized WGA appears within cisternae of Golgi apparatus stacks (Figs. 1a,b, 3a). At early endocytosis stages, the WGA-reactions are confined to limited regions of the cisternae, whereas stacks that contain high amounts of WGA within the lumina of all cisternae are mainly found at later periods of WGA endocytosis, e.g. after 45–60 min (Fig. 3b). Very rarely, internalized WGA can be found within cisternae of the endoplasmic reticulum. At each of the internalization times, WGA appears within multivesiculated bodies (Figs. 2, 3a,b), indicating that internalized WGA not only is transported to the Golgi apparatus but is sorted to the late endosomal–lysosomal pathway as well. The new insights into the subsequent steps of uptake of internalized WGA into the Golgi apparatus, and the endocytosis-induced Golgi reorganizations provided the base for developments of precise time schedules for a regulated retrograde transport of internalized WGA into the endoplasmic reticulum by treatment with Brefeldin A (Vetterlein et al. 2003). Multiple questions are open, and are in the center of current research: the mechanisms underlying the formation of the endocytic TGN, and its attachment to the stacked Golgi cisternae, the relationships between endocytic TGN and TGN involved in secretion, and the relationships between Golgi exit sites in the secretory system (de Matteis and Luini 2008) and Golgi entrance sites in the endocytic system, the delivery of cargo, and its retrograde transport within the Golgi stacks, signalling and the importance of contact points, lipid transfer (de Matteis et al. 2007; Hanada et al. 2007), and mechanisms of direct endocytic TGN-to-ER traffic via trans-Golgi associated ER. Does anterograde secretory traffic influence retrograde transport, and vice versa? Does retrograde flow, and uptake of internalized molecules into the Golgi apparatus influence its size and localization? Main questions also concern the formation of Golgi ribbons, and as to whether the organization of the ribbons induced by retrograde flow are comparable with those built by membrane input from the ER (Marra et al. 2007).

Concluding remarks Retrograde plasma membrane to Golgi apparatus transport plays a key role in the physiologic life of cells, the maintenance of cellular homeostasis, and in cell pathologic mechanism as well. Detailed insights into retrograde pathways to the Golgi apparatus are essential for the understanding of intoxication processes (e.g. Sandvig and Van Deurs 2005), for assessment of effects and side effects of drugs (e.g. Sandoval and Molitoris 2004), and for the development of strategies for targeted drug delivery to the interior of cells (e.g. El Alaoui et al. 2007; Johannes and Decaudin 2005; Kreitman 2006; € ck et al. 2004). Smith et al. 2002; Tarragó-Trani and Storrie 2007; Weissenbo These important roles put retrograde routes to the Golgi apparatus into the center of molecular cell biologic and medical sciences, and will be

Retrograde plasma membrane-to-Golgi apparatus transport

*

469

driving forces for future research projects. One of the great challenges will be to correlate ultrastructural results, which have elucidated multiple details of the functional Golgi apparatus architecture (e.g. Marsh 2005; Marsh et al. 2004; Mironov et al. 2005; Mogelsvang and Howell 2006; Mogelsvang et al. 2004; Pavelka et al. 1998, 2008; Rambourg and Clermont 1997; Trucco et al. 2004; Vetterlein et al. 2002), with molecular cell biologic research on the machineries and regulatory mechanisms of retrograde transport to, and into the Golgi apparatus (e.g. Amessou et al. 2007; Bonifacino and Rojas 2006; de Matteis et al. 2007; Hanada et al. 2007; Levine and Loewen 2006; Marra et al. 2007; Missiaen et al. 2007; Von Zastrow and Sorkin. 2007; €lchli et al. 2008). Wa Acknowledgements. The authors cordially thank all colleagues involved in the work presented in this review, Anita Aichinger, Peter Auinger, Ulrich Kaindl, Jedrzej Kosiuk, € ller, Majid Niapir, Julia Beatrix Mallinger, Claudia Meisslitzer-Ruppitsch, Josef Neumu Riess, Elfriede Scherzer, Monika Vetterlein, and Christoph Weiss. Parts of the studies were performed within the Associate Membership to the Network of Excellence 3D-EM € rgen M. Plitzko, and colleagues, Max in cooperation with Wolfgang Baumeister, Ju Planck Institute of Biochemistry, Martinsried, Germany. Parts of the studies were performed in cooperation with Alexander Mironov, Alberto Luini, and colleagues from the Consorzio Mario Negri Sud, S. Maria Imbaro, Italy.

References Amessou M, Fradagrada A, Falguieres T, Lord JM, Smith DC, Roberts LM, Lamaze C, Johannes L (2007) Syntaxin 16 and syntaxin 5 are required for efficient retrograde transport of several exogenous and endogenous cargo proteins. J Cell Sci 120: 1457–1468 Arighi CN, Hartnell LM, Aguilar RC, Haft CR, Bonifacino JS (2004) Role of the mammalian retromer in sorting of the cation-independent mannose 6-phosphate receptor. J Cell Biol 165: 123–133 Banting G, Ponnambalam S (1997) TGN38 and its orthologues: roles in post-TGN vesicle formation and maintenance of TGN morphology. Biochim Biophys Acta 1355: 209–217 Barbero P, Bittova L, Pfeffer SR (2002) Role of Rab9 GTP-mediated vesicle transport from endosomes to the trans-Golgi in living cells. J Cell Biol 156: 511–518 Berger EG (1985) Mini-review: how Golgi-associated glycosylation works. Cell Biol Int Rep 9: 407–417 Bonifacino JS, Rojas R (2006) Retrograde transport from endosomes to the trans-Golgi network. Nat Rev Mol Cell Biol 7: 568–579 Bujny MV, Popoff V, Johannes L, Cullen PJ (2007) The retromer component sorting nexin-1 is required for efficient retrograde transport of Shiga toxin from early endosome to the trans Golgi network. J Cell Sci 120: 2010–2021 Carrol KS, Hanna J, Simon I, Krise J, Barbero P, Pfeffer SR (2001) Role of Rab9-GTPase in facilitating receptor recruitment by TIP47. Science 292: 1373–1376 Chinnapen DJF, Chinnapen H, Saslowsky D, Lencer WI (2007) Rafting with cholera toxin: endocytosis and trafficking from the plasma membrane to ER. FEMS Microbiol Lett 266: 129–137 Crump CM, Xiang X, Thomas L, Gu F, Austin C, Tooze SA, Thomas G (2001) PACS-1 binding to adaptors is required for acidic cluster motif-mediated protein traffic. EMBO J 20: 2191–2201 De Matteis MA, Godi A (2004) PI-loting membrane traffic. Nat Cell Biol 6: 487–492

470

*

M. Pavelka and A. Ellinger

De Matteis MA, Luini A (2008) Exiting the Golgi complex. Nat Rev Mol Cell Biol 9: 273–284 De Matteis MA, Di Campli A, D’Angelo G (2007) Lipid transfer proteins in membrane trafficking at the Golgi complex. Biochim Biophys Acta 1771: 761–768 Diaz E, Schimmoller F, Pfeffer SR (1997) A novel Rab9 effector required for endosometo-TGN transport. J Cell Biol 138: 283–290 El Alaoui A, Schmidt F, Amessuo M, Sarr M, Decaudin D, Florent JC, Johannes L (2007) Shiga toxin-mediated retrograde delivery of a topoisomerase I inhibitor prodrug. Angew Chem Int Ed 46: 6469–6472 Farquhar MG, Hauri H-P (1997) Protein sorting and vesicular traffic in the Golgi appa€ user, Basel, pp 63–129 ratus. In: Berger EG, Roth J (eds) The Golgi apparatus. Birkha Farquhar MG, Palade GE (1998) The Golgi apparatus: 100 years of progress and controversy. Trends Cell Biol 8: 2–10 Feng Y, Jadhav AP, Rodighiero CH, Fujinaga Y, Kirchhausen T, Lencer WI (2004) Retrograde transport of cholera toxin from the plasma membrane to the endoplasmic reticulum requires the trans-Golgi network but not the Golgi apparatus in Exo2-treated cells. EMBO J 5: 596–601 Garred O, Van Deurs B, Sandvig K (1995) Furin-induced cleavage and activation of Shiga toxin. J Biol Chem 270: 10817–10821 Ghosh P, Dahms NM, Kornfeld S (2003) Mannose-6 phosphate receptors: new twists in the tale. Nat Rev Mol Cell Biol 4: 202–212 Ghosh RN, Mallet WG, Soe TT, McGraw TE, Maxfield FR (1998) An endocytosed TGN38 chimeric protein is delivered to the TGN after trafficking through the endocytic recycling compartment in CHO cells. J Cell Biol 142: 923–936 Glick BS (2000) Organization of the Golgi apparatus. Curr Opin Cell Biol 12: 50–56 Gokool S, Tattersall D, Seaman MNJ (2007) EHD 1 Interacts with retromer to stabilize SNX1 tubules and facilitate endosome-to-Golgi retrieval. Traffic 8: 1873–1886 Goldstein IJ, Hayes CE (1978) The lectins: carbohydrate-binding proteins of plants and animals. Adv Carbohydr Chem Biochem 35: 127–340 Gonatas NK, Kim SU, Stieber A, Avrameas S (1977) Internalization of lectins in neuronal GERL. J Cell Biol 73: 1–13 Gonatas NK, Stieber A, Gonatas JO, Mommoi T, Fishman PH (1983) Endocytosis of exogenous GM1 ganglioside and cholera toxin by neuroblastoma cells. Mol Cell Biol 3: 91–101 Grimmer S, Spilsberg B, Hanada K, Sandvig K (2006) Depletion of sphingolipids facilitates endosome to Golgi transport of ricin. Traffic 7: 1243–1253 Hanada K, Kumagai K, Tomishige N, Kawano M (2007) CERTand intracellular trafficking of ceramide. Biochim Biophys Acta 1771: 644–653 Hettema EH, Lewis MJ, Black MW, Pelham HR (2003) Retromer and sorting nexins Snx4/ 41/42 mediate distinct retrieval pathways from yeast endosomes. EMBO J 22: 548–557 Hong W (2005) SNAREs and traffic. Biochim Biophys Acta 1744: 493–517 Hurley JA, Emr SD (2006) The ESCRT complexes: structure and mechanism of a membrane trafficking network. Annu Rev Biophys Biomol Struct 35: 277–298 Ikeda M, Gunji Y, Yamasaki S, Takeda Y (2000) Shiga toxin activates p38 MAP kinase through cellular Ca(2þ ) increase in Vero cells. FEBS Lett 485: 94–98 Johannes L, Goud B (1998) Surfing on a retrograde wave: how does Shiga toxin reach the endoplasmic reticulum? Trends Cell Biol 8: 158–162 Johannes L, Goud B (2000) Facing inward from compartment shores: how many pathways were we looking for? Traffic 1: 119–123 Johannes L, Decaudin D (2005) Protein toxins: intracellular trafficking for targeted therapy. Gene Ther 12: 1360–1368 Johannes L, Tenza D, Antony C, Goud B (1997) Retrograde transport of KDEL-bearing B-fragment of Shiga toxin. J Biol Chem 272: 19554–19561

Retrograde plasma membrane-to-Golgi apparatus transport

*

471

Kreitman RJ (2006) Immunotoxin for targeted cancer therapy. AAPS J 8: E532–E551 Lauvrak SU, Torgersen ML, Sandvig K (2004) Efficient endosome-to-Golgi transport of Shiga toxin is dependent on dynamin and clathrin. J Cell Sci 117: 2321–2331 € lchli S, Iversen TG, Slagsvold HH, Torgersen ML, Spilsberg B, Sandvig K Lauvrak SU, Wa (2006) Shiga toxin regulates its entry in a Syk-dependent manner. Mol Biol Cell 17: 1096–1109 Levine T, Loewen C (2006) Inter-organelle contact sites: through a glass, darkly. Curr Opin Cell Biol 18: 371–378 Lochner N, Pittner F, Wirth M, Gabor F (2003) Wheat germ agglutinin binds to the epidermal growth factor receptor of artificial CaCo-2 membranes as detected by silver nanoparticle enhanced fluorescence. Pharma Res 20: 833–839 Lombardi D, Soldati T, Riederer MA, Goda Y, Zerial M, Pfeffer SR (1993) Rab9 functions in transport between late endosomes and trans Golgi network. EMBO J 12: 677–682 Malhotra V, Mayor S (2006) Cell biology: the Golgi grows up. Nature 441: 939–940 Mallard F, Antony C, Tenza D, Salamero J, Goud G, Johannes L (1998) Direct pathway from early/recycling endosomes to the Golgi apparatus revealed trough the study of Shiga toxin-B fragment transport. J Cell Biol 143: 973–990 Mallard F, Tang BL, Galli T, Tenza D, Saint-Pol A, Yue X, Antony C, Hong W, Goud B, Johannes L (2002) Early/recycling endosomes-to-TGN transport involves two SNARE complexes and a Rab6 isoform. J Cell Biol 156: 653–664 Mallet WG, Maxfiled FR (1999) Chimeric forms of furin and TGN38 are transported from the plasma membrane to the trans-Golgi network via distinct endosomal pathways. J Cell Biol 146: 345–359 Mari M, Bujny MV, Zeuschner D, Geerts WJC, Griffith J, Petersen CM, Cullen PJ, Klumperman J, Geuze HJ (2008) SNX1 defines an early endosomal recycling exit for sortilin and mannose 6-phosphate receptors. Traffic 9: 380–393 Marra P, Salvatore L, Mironov A Jr, Di Campli A, Di Tullio G, Trucco A, Beznoussenko G, Mironov A, De Matteis MA (2007) The biogenesis of the Golgi ribbon: the roles of membrane input from the ER and of GM130. Mol Biol Cell 18: 1595–1608 Marsh BJ (2005) Lessons from tomographic studies of the mammalian Golgi. Biochim Biophys Acta 1744: 273–292 Marsh BJ, Howell KE (2002) The mammalian Golgi – complex debates. Nat Rev Mol Cell Biol 3: 789–795 Marsh BJ, Volkman N, McIntosh JR, Howell KE (2004) Direct continuities between cisternae at different levels of the Golgi complex in glucose-stimulated mouse islet beta cells. Proc Natl Acad Sci USA 101: 5565–5570 Massol RH, Larsen JE, Fujinaga Y, Lencer WI, Kirchhausen T (2004) Cholera toxin toxicity does not require functional Arf6- and dynamin-dependent endocytic pathways. Mol Biol Cell 15: 3631–3641 Maxfield FR, McGraw TE (2004) Endocytic recycling. Nat Rev Mol Cell Biol 5: 121–132 Mellman I, Warren G (2000) The road taken: past and future foundation of membrane traffic. Cell 100: 99–112 Meyer C, Zizioli D, Lausman S, Eskelinen EL, Hamann J, Saftig P, Von Figura K, Schu P (2000) m1A-adaptin-deficient mice: lethality, loss of AP-1 binding and rerouting of mannose-6-phosphate receptors. EMBO J 19: 2193–2203 Mironov AA, Beznoussenko GV, Polishchuk RS, Trucco A (2005) Intra-Golgi transport: a way to a new paradigm? Biochim Biophys Acta 1744: 340–350 Missiaen L, Dode L, Vanoevelen J, Raeymaekers L, Wuytack F (2007) Calcium in the Golgi apparatus. Cell Calcium 41: 405–416 Mogelsvang S, Howell KE (2006) Global approaches to study Golgi function. Curr Opin Cell Biol 18: 438–443 Mogelsvang S, Marsh BJ, Ladinsky MS, Howell KE (2004) Predicting function from structure: 3D structure studies of the mammalian Golgi complex. Traffic 5: 338–345

472

*

M. Pavelka and A. Ellinger

Molloy SS, Anderson ED, Jean F, Thomas G (1999) Bi-cycling the furin pathway: from TGN localization to pathogen activation and embryogenesis. Trends Cell Biol 9: 28–35 Pavelka M, Ellinger A, Debbage P, Loewe C, Vetterlein M, Roth J (1998) Endocytic routes to the Golgi apparatus. Histochem Cell Biol 109: 555–570 € ller J, Ellinger A (2008) Retrograde traffic in the biosyntheticPavelka M, Neumu secretory route. Histochem Cell Biol 129: 277–288 Peden AA, Oorschot V, Hesser BA, Austin CD, Scheller RH, Klumperman J (2004) Localization of the AP-3 adapter complex defines a novel endosomal exit site for lysosomal membrane proteins. J Cell Biol 164: 1065–1076 Pelham HRB, Roberts LM, Lord JM (1992) Toxin entry: how reversible is the secretory pathway? Trends Cell Biol 2: 183–185 Pelham HRB, Rothman JE (2000) The debate about transport in the Golgi – two sides of the same coin? Cell 102: 713–719 Pelkmans L, Fava E, Grabner H, Hannus M, Haberman B, Krausz E, Zerial M (2005) Genome-wide analysis of human kinases in clathrin- and caveolae/raft-mediated endocytosis. Nature 436: 78–86 Popoff V, Mardones GA, Tenza D, Rojas R, Lamaze C, Bonifacino JS, Raposo G, Johannes L (2007) The retromer complex and clathrin define an early endosomal retrograde exit site. J Cell Sci 120: 2022–2031 Puri S, Bachert C, Fimmel CJ, Linstedt AD (2002) Cycling of early Golgi proteins via the cell surface and endosomes upon luminal pH disruption. Traffic 3: 641–653 Puthenveedu MA, Linstedt AD (2005) Subcompartmentalizing the Golgi apparatus. Curr Opin Cell Biol 17: 369–375 Rambourg A, Clermont Y (1997) Three-dimensional structure of the Golgi apparatus in €user, Basel, mammalian cells. In: Berger EG, Roth J (eds) The Golgi apparatus. Birkha pp 37–61 Rapak A, Falsnes PO, Olsnes S (1997) Retrograde transport of mutant ricin to the endoplasmic reticulum with subsequent translocation to the cytosol. Proc Natl Acad Sci USA 94: 3783–3788 Restrepo R, Zhao X, Peter H, Zhang B, Arvan P, Nothwehr SF (2007) Structural features of Vps35p involved in interaction with other subunits of the retromer complex. Traffic 8: 1841–1853 € sch A (2005) Protein sorting in the Golgi complex: shifting Rodriguez-Boulan E, Mu paradigms. Biochim Biophys Acta 1744: 455–464 Rohn WM, Rouille Y, Waguri S, Hoflack B (2000) Bi-directional trafficking between the trans-Golgi network and the endosomal/lysosomal system. J Cell Sci 113: 2093–2101 Rojas R, Kametaka S, Hafr CR, Bonifacino JS (2007) Interchangeable but essential functions of SNX1 in the association of retromer with endosomes and the trafficking of mannose 6-phosphate receptors. Mol Cell Biol 27: 1112–1124 € mer W, Berland L, Chambon V, Gaus K, Windschiegl B, Tenza D, Aly MRE, Fraisier V, Ro Florent JC, Perrais D, Lamaze C, Raposo G, Steinem C, Sens P, Bassereau P, Johannes L (2007) Shiga Toxin induces tubular membrane invagination for its uptake into cells. Nature 450: 670–675 Roth J (1997) Topology of glycosylation in the Golgi apparatus. In: Berger EG, Roth J €user, Basel, pp 131–161 (eds) The Golgi apparatus. Birkha  S, Oorschot V, Strous GJ, Klumperman J (2002) Bilayered clathrin coats Sachse M, Urbe on endosomal vacuoles are involved in protein sorting towards lysosomes. Mol Biol Cell 13: 1313–1328 Saint-Pol A, Yelamos B, Amessou M, Mills IG, Dugast M, Tenza D, Schu P, Antony C, McMahon HT, Lamaze C, Johannes L (2004) Clathrin adaptor EpsinR is required for retrograde sorting on early endosomal membranes. Dev Cell 6: 525–538 Sandoval RM, Molitoris BA (2004) Gentamycin traffics retrograde through the secretory pathway and is released in the cytosol via the endoplasmic reticulum. J Physiol Renal Physiol 286: F617–F624

Retrograde plasma membrane-to-Golgi apparatus transport

*

473

Sandvig K, Brown JE (1987) Ionic requirements for entry of Shiga toxin from Shigella dysenteriae 1 into cells. Infect Immun 55: 298–303 Sandvig K, Van Deurs B (1994) Endocytosis and intracellular sorting of ricin and Shiga toxin. FEBS Lett 346: 99–102 Sandvig K, Van Deurs B (1996) Endocytosis, intracellular transport, and cytotoxic action of Shiga toxin and ricin. Physiol Rev 76: 949–966 Sandvig K, Van Deurs B (2002) Membrane traffic exploited by protein toxins. Annu Rev Cell Dev Biol 18: 1–24 Sandvig K, Van Deurs B (2005) Delivery into cells: lessons learned from plant and bacterial toxins. Gene Ther 12: 865–872 Sandvig K, Garred O, Prydz K, Kozlov J, Hansen SH, Van Deurs B (1992) Retrograde transport of endocytosed Shiga toxin to the endoplasmic reticulum. Nature 358: 510–512 Sannerud R, Saraste J, Goud B (2003) Retrograde traffic in the biosynthetic-secretory route: pathways and machinery. Curr Opin Cell Biol 15: 438–445 Seaman MN (2004) Cargo-selective endosomal sorting for retrival to the Golgi requires retromer. J Cell Biol 165: 111–122 Seaman MN (2005) Recycle your receptors with retromer. Tends Cell Biol 15: 68–75 €lchli S, Utskarpen A, Wandinger-Ness A, Sandvig K (2007) PhosphoSkanland SS, Wa inositide-regulated retrograde transport of ricin: crosstalk between hVps34 and sorting nexins. Traffic 8: 297–309 € lchli S, Sandvig K (2006) EDEM is involved in Slominska-Wojewodzka M, Gregers TF, Wa retrotranslocation of ricin from the endoplasmic reticulum to the cytosol. Mol Biol Cell 17: 1664–1675 Smith DC, Lord JM, Roberts LM, Tartour E, Johannes L (2002) 1st class ticket to class I: protein toxins as pathfinders for antigen presentation. Traffic 3: 697–704 Smith DC, Spooner RA, Watson PD, Murray JL, Hodge TW, Amessou M, Johannes L, Lord JM, Roberts LM (2006) Internalized pseudomonas exotoxin A can exploit multiple pathways to reach the endoplasmic reticulum. Traffic 7: 379–393 Snyder MD, Rogers OC (1985) Intracellular movement of cell surface receptors after endocytosis: resialylation of asialo-transferrin receptor in human erythroleukemia cells. J Cell Biol 100: 826–834 Stanley KK, Howell KE (1993) TGN38/41: a molecule on the move. Trends Cell Biol 3: 252–255 Stieber A, Gonatas JO, Gonatas NK (1984) Differences between the endocytosis of horseradish peroxidase and its conjugate with wheat germ agglutinin by cultured fibroblasts. J Cell Physiol 119: 71–76 Storrie B (2005) Maintenance of Golgi apparatus structure in the face of continuous protein recycling to the endoplasmic reticulum: making ends meet. Int Rev Cytol 244: 69–94 Tai GH, Lu L, Wang TL, Tang BL, Goud B, Johannes L, Hong WJ (2004) Participation of the syntaxin 5/Ykt6/GS28/GS15 SNARE complex in transport from the early/recycling endosome to the trans-Golgi network. Mol Biol Cell 15: 4011–4022 Tarrago-Trani MT, Storrie B (2007) Alternate routes for drug delivery to the cell interior: pathways to the Golgi apparatus and endoplasmic reticulum. Adv Drug Deliv Rev 59: 782–797 Torgersen ML, Skretting G, Van Deurs B, Sandvig K (2001) Internalization of cholera toxin by different endocytic mechanisms. J Cell Sci 114: 3737–3747 €lchli S, Grimmer S, Skanland SS, Sandvig K (2007) Protein kinase Torgersen ML, Wa Cd is activated by Shiga toxin and regulates its transport. J Biol Chem 282: 16317–16328 Trucco A, Polishchuk RS, Martella O, Di Pentima A, Fusella A, Di Giandomenico D, San Pietro E, Beznoussenko GV, Polishchuk EV, Baldassare M, Buccione R, Geerts WJ, Koster AJ, Burger KN, Mironov AA, Luini A (2004) Secretory traffic triggers the

474

*

M. Pavelka and A. Ellinger

formation of tubular continuities across Golgi subcompartments. Nat Cell Biol 6: 1071–1081 Ullrich OS, Reinsch S, Urbe S, Zerial M, Parton RG (1996) Rab11 regulates recycling through the pericentriolar recycling endosome. J Cell Biol 135: 913–924 €lchli S, Sandvig K (2006) Transport of ricin Utskarpen A, Slagsvold HH, Iversen TG, Wa from endosomes to the Golgi apparatus is regulated by Rab6A and Rab6A´. Traffic 7: 663–672 Utskarpen A, Slagsvold HH, Dyve AB, Skanland SS, Sandvig K (2007) SNX1 and SNX2 mediate retrograde transport of Shiga toxin. Biochim Biophys Res Commun 358: 566–570 Van Deurs B, Tonnensen TI, Petersen OW, Sandvig K, Olsnes S (1986) Routing of internalized ricin and ricin conjugates in the Golgi complex. J Cell Biol 102: 37–47 Van Deurs B, Petersen OW, Olsnes S, Sandvig K (1987) Delivery of internalized ricin from endosomes to cisternal Golgi elements is a discontinuous, temperature-sensitive process. Exp Cell Res 171: 137–152 Varlamov O, Fricker LD (1998) Intracellular trafficking of metallocarboxypeptidase D in AtT-20 cells: localization to the trans-Golgi network and recycling from the cell surface. J Cell Sci 111: 877–885 € ller J, Pavelka M (2002) Golgi apparatus and TGN during Vetterlein M, Ellinger A, Neumu endocytosis. Histochem Cell Biol 117: 143–150 € ller J, Pavelka M (2003) Brefeldin A-regulated Vetterlein M, Niapir M, Ellinger A, Neumu retrograde transport into the endoplasmic reticulum of internalised wheat germ agglutinin. Histochem Cell Biol 120: 121–128 Volz B, Orberger G, Porwoll S, Hauri H-P, Tauber R (1995) Selective reentry of recycling cell surface glycoproteins to the biosynthetic pathway in human hepatocarcinoma HepG2 cells. J Cell Biol 130: 537–551 Von Zastrow M, Sorkin A (2007) Signaling in the endocytic pathway. Curr Opin Cell Biol 19: 436–445 €lchli S, Skanland SS, Gregers TF, Lauvrak SU, Torgersen ML, Ying M, Kuroda S, Wa Maturana A, Sandvig K (2008) The MAP kinase p38 links Shiga toxin dependent signaling and trafficking. Mol Biol Cell 19: 95–104 € ck A, Bogner E, Wirth M, Gabor F (2004) Binding and uptake of wheat germ Weissenbo agglutinin-grafted PLGA-nanospheres by Caco-2 monolayers. Pharma Res 21: 1917–1923 Yoshino A, Setty SR, Poynton C, Whiteman EL, Saint-Pol A, Burd CG, Johannes L, Holzbaur EL, Koval M, McCaffery JM, Marks MS (2005) tGolgin-1(p230, golgin245) modulates Shiga-toxin transport to the Golgi and Golgi motility towards the microtubule-organizing centre. J Cell Sci 18: 2279–2293

Interactions between endocytosis and secretory transport

*

475

Interactions between endocytosis and secretory transport Galina V. Beznoussenko, Margit Pavelka and Alexander A. Mironov

Here, we describe data showing that exo- and endocytosis functionally are closely related to each other and the inhibition/activation of one process affects another.

Interactions between two pathways Secretory and endocytic pathways are closely integrated with each other (Ang et al. 2004). Usually, alteration of one pathway affects the function of another. For instance, stimulation of the endocytosis augments the intracellular traffic and exocytosis. In particular, the activation of endocytosis by ligand–receptor interaction or by PKC activation leads to the acceleration of Golgi-to-plasma membrane (PM) traffic (De Matteis et al. 1993; Buccione et al. 1996). And, vice versa, stimulation of exocytosis augments the endocytosis. Indeed, the enhanced Golgi-to-PM traffic after activation of protein synthesis accelerates endocytosis and PM-to-trans-Golgi network (TGN) transport (Sandvig and Van Deurs 1996). Massive fusion of synaptic vesicles with presynaptic membrane activates endocytosis at the level of synapses (Liu and Robinson 1995; Roos and Kelly 1999). Compensatory endocytosis can occur through a clathrin-mediated mechanism (Gundelfinger et al. 2003). In adrenal chromaffin cells, rapid endocytosis occurring immediately after an exocytosis burst is complete within 20 s and is dynamin-dependent but clathrin-independent (Artalejo et al. 1995). In RBL cells, the physiological stimulation of the IgE receptor or direct activation of PKC leads to the missorting of the aspartic protease cathepsin D and secretion of its immature form (Baldassarre et al. 2000). This additionally suggests the interference between different transport routes. Similarly, blockage of endocytosis inhibits exocytosis. In Saccharomyces cerevisiae, when End3p-mediated endocytosis is impaired the intracellular transport of proteins involved into spore wall formation is blocked (Morishita and Engebrecht 2005). Reduction of endocytic activity by withdrawal of serum (containing growth factors) from culture medium decreases the rate of intercellular traffic including TGN–PM step (Liou et al. 1997) and results in the decline of membrane delivery to the PM and decrease of the size of the Golgi complex (GC, Mironov and Mironov 1998). In the temperature-sensitive End4 mutant of Chinese hamster ovary cells with impaired fluid-phase endocytosis, the secretory traffic is also impaired

476

*

G. V. Beznoussenko et al.

and the hemagglutinin of influenza virus (Wang et al. 1990) and Sindbis virus envelope glycoproteins (Presley et al. 1991) failed to reach the PM. Secretion of total soluble protein into the medium is also strongly reduced at the temperature restricted for End4 (Wang et al. 1990). And again, vice versa, inhibition of biosynthetic pathway causes the inhibition of endocytosis (Hicke and Riezman 1996). The inhibition of intracellular traffic in temperature sensitive mutants greatly reduces endocytosis (Hicke et al. 1997). Additionally, the evidence (although indirect) that connections between endosomal and exocytic systems exist has been obtained. Indeed, labelled lipids incorporated at the PM can diffuse to the GC and endoplasmic reticulum (ER) without participation of the aqueous phase at low temperature and even in aldehyde-fixed cells (but not after OsO4 fixation) and when either vesicular or protein transport were blocked. This implies the existence of membrane continuities (at the least transient) between the PM and the GC (Pagano et al. 1989). Mutual influence of secretory and endocytic pathways could be executed by the following ways. 1. Through signalling initiated after endocytosis or exocytosis (see Chapter 2.13). 2. Through Ca2þ release during endocytosis or exocytosis (see Chapter 2.8). 3. Through enhanced membrane flow to the GC from the PM or to the PM from the GC. 4. Through alteration of protein synthesis, although this mechanism is slower.

Transport of secretory proteins through endosomes One of the important aspects of the problem of interaction between the endocytic and exocytic pathways is the observation that several secretory proteins in their way towards the PM pass through the endosomal system (Fishman and Fine 1985; Stoorvogel et al. 1988; Rudick et al. 1993; Sariola et al. 1995; Futter et al. 1995; Leitinger et al. 1995; Brachet et al. 1999; Orzech et al. 2000; Ang et al. 2003; Lock and Stow 2005). Here, we will describe only the recent observations. In TfR-transfected MDCK cells, the newly synthesized VSV-G after exiting from TGN enters transferrin-positive endosomes in significant amount only when TfR is over expressed. It occurs within a few minutes after exit from the trans-Golgi network. Precipitation of DAB inside the endosomal lumen resulting in elimination of endosomal function blocks transport of VSVG to the PM (Ang et al. 2004). Experiments with TfR tagged with photo activated green fluorescent protein (PA-GFP) confirm these observations. Time-lapse imaging revealed

Interactions between endocytosis and secretory transport

*

477

that the photo activated fluorescence signals of PA-GFP–TfR moved into the Tfn-labelled early Tfn- and EEA1-positive endosomal compartments within 60–90 s (Luo et al. 2006). The passage of some basolateral cargoes through endosomes has confirmation at the level of molecular mechanisms. Indeed, by 3–5 min after microinjection of an antibody against the medium subunit of m1B into polarized FRT cells, two newly synthesized basolateral proteins (VSVG and TfR) tagged with green fluorescent protein (GFP) exited the TGN normally but became blocked at the level of recycling endosomes. In contrast, LDLR is transported normally (Cancino et al. 2007). These observations suggest that VSVG and TfR (but not LDLR) pass through recycling endosomes and this normal passage need a function of AP-1. Also glycosylphosphatidyl inositol-linked proteins seem to bypass endosomes (Futter et al. 1995). On the other hand, in living cells, VSVG-positive Golgi-to-PM carriers (GPCs) were not found to be mixed with endosomes before reaching the PM (Keller et al. 2001) and VSVG present in GPCs does not colocalize with GGA/AP1 (Puertollano et al. 2003). Only in 5% of images both Tfn-ferritin and VSV G could be detected in irregularly shaped elongated tubular elements in the cytoplasm of cells, which were carefully examined to ensure that no G protein was present on the PM (Hedman et al. 1987). The rarity of these coincidences suggests the short life of the common structures (fused exo- and endocytic carriers) as it has been suggested by the kiss-and-run model of intracellular transport (see Chapters 3.2 and 3.4). In cells possessing regulated secretion, the constitutively secreted proteins (CSPs) pass through endosomes (Arvan et al. 1991; Kuliawat and Arvan 1992; Turner and Arvan 2000; Huang et al. 2001; Feng and Arvan 2003; see also Chapter 3.11). For instance, in INS-1 cells, a fraction of 1-protease inhibitor is processed to a stable, protease-resistant fragment, which accumulates intracellularly in a post-Golgi compartment that can be identified as the endosome/lysosome system (Feng and Arvan 2003). In insulin-secreting beta cells, constitutive secretion of the lysosomal enzyme procathepsin B represents constitutive-like secretion via an endosomal intermediate (Turner and Arvan 2000), although the role of endosomes as intermediates in the release of bona fide secretory proteins is less well studied (Millar et al. 2000). In yeast, where the transport is polarized and the PM of yeast is the analogue of the APM (Riezman 1993), it seems that one branch of the secretory pathway transits through endosomes before reaching the cell surface (Harsay and Schekman 2002). In yeast, the resident Golgi protein KEX recycles through the endocytic system (Riezman 1993). The internal organelle from which Golgi proteins recycle is an endosomal compartment. The fact that budding and retrieval from this endosomal compartment requires clathrin, represents an additional evidence in favour of close links between these two systems. Thus, there is a consensus that most of constitutive secretory proteins pass through endosomes on their way towards the basolateral PM. However, it is

478

*

G. V. Beznoussenko et al.

not completely clear whether apically directed constitutive secretory proteins also go through endosomes (see Chapter 3.14).

The necessity of two transport steps during Golgi-to-PM transport One could envisage that GPCs cannot be delivered to the BLPM directly and they need a stop in order to interact with endosomes. Indeed, careful analysis of our own movies and published movies showing movement of GPCs in live cells revealed that no one GPC can fuse with the PM during one step without any stop and consecutive hovering, no transport event when GPC departs from the Golgi mass and without stops moves towards the peripheral area and immediately fuses with the PM can be seen. In the majority of cases, GPCs stop their fast movement along MT at least once and start to hover either during their centrifugal movement or after arrival to the peripheral areas. Then, GPCs restore their travelling after some time of hovering around the site, where they stopped. On the other hand, it seems that the structure of GPCs immediately after their formation differs from that after their delivery to the PM. Indeed, careful examination of published images revealed that GPCs located near the GC are characterized by rather complicated morphology. In contrast, GPCs observed near the BLPM exhibit rather simple morphology being similar to elongated vacuoles or tubules (compare Figs. 6 and 7 in Polishchuk et al. (2000) and Figs. 5 and 8 in Polishchuk et al. (2003)). In cells microinjected with the inhibitory antibody against NSF, a SNARE regulator, GPCs are formed at a usual rate and with normal dynamics in spite of the inhibition of membrane fusion. However, under these conditions, even GPCs observed near the BLPM exhibited complex structure but not just as GPC located near the PM in control cells and exhibiting a simple vacuole-like appearance in cells where membrane fusion is not blocked (compare Fig. 5 in Polishchuk et al. (2003) and Fig. 7 in Polishchuk et al. (2000)). Thus, it seems that membrane fusion is necessary for the maturation of GPCs. Thus, the movements of GPCs filled with VSVG to the cell periphery are accompanied by the alteration of their structure and this process is fusion-dependent. Is the interaction between GPCs with endosomes really necessary and important? There are several possibilities. 1. Exchange of SNAREs obtained during formation of GPCs into SNAREs necessary for the fusion of GPCs with the BLPM (see Chapter 2.1). There is no enough SNAP25 and syntaxin 1, 3 or 4 at the level of the GC and thus, their acquisition into GPCs could be dependent on fusion–fission with endosomes. 2. Additional acidification after fusion with late endosomes could promote protein multimerization and condensation. 3. The thick membranes of endosomes are used for the complete exclusion of Golgi enzymes from the GPCs (see Chapter 3.4).

Interactions between endocytosis and secretory transport

*

479

COPI and endosomes An additional similarity between endocytic and exocytic pathways is that both of them depend of the function of COPI. The facts, that COPI is functionally important for both exocytic and endocytic systems, evidence ones more in favour of the functional relation of these compartments (Mellman 1996). Endosomes recruit ARF, COPI and clathrin (Whitney et al. 1995). GTPgS stimulates COP association to endosomes but inhibits formation of multivesicular bodies (MVB, Aniento et al. 1996). ARF1 is also involved in recruiting AP3 and COPI subunits to endosomal membrane buds in PC12 cells (Faundez et al. 1998; Gu and Gruenberg 2000). Recently, it has been demonstrated that activated ARF1 is present at the plasma membrane and recruits ARHGAP10, a GAP for Cdc42 (Kumari and Mayor 2008). However, COPI-coated buds were not found on the endosomal membranes (Mellman 1996). COP binding onto endosomal membranes still occurs even when e-COP was missing. Moreover, although pH-dependent binding of COP to endosomal membranes occurs, bound coatomer complex is non-functional. However, a small COP subset only, consisting of b0 , b, and z-COP was then recruited onto early endosomes. In the absence of e-COP, COP failed to interact with endosomes (Gu et al. 1997). It seems that COP association to endosomal membranes is itself pHdependent both in vitro and in vivo (Aniento et al. 1996) as is the recruitment of the b, b0 and z-COP sub complex in the absence of e-COP (Gu et al. 1997). Neutralization of the endosomal pH causes very similar effect: early endosomes form clusters of the tubules (50–60 nm diameter) that typically lack multivesicular domains and both MVB formation and transport to late endosomes are inhibited (Aniento et al. 1996). Neutralization of the pH, however, does not cause the release of pre-bound COP. Neutralization of pH in endosomes inhibits the stimulatory effects of GTPgS on COP recruitment. COP recruitment by endosomes was partially inhibited when the endosomal pH was neutralized with nigericin before addition of GTPgS when compared with GTPgS alone (Gu et al. 1997). There are also several evidences in favour of the role of coatomer for endosomes. Reduction in activity or levels of ARF1 specifically inhibits GPI-AP and fluid-phase endocytosis without affecting other clathrin-dependent or -independent endocytic pathways (Kumari and Mayor 2008). There are several reports about the dependency of endosomal function on COP I machinery. For instance, in ldlF cells, EES to LES transport is impaired at restrictive temperature. Transferrin recycling and accumulation of HRP were markedly inhibited after e-COP degradation in ldlF F cells (Daro et al. 1997). In vitro MVBs formation from donor (animal) early endosomal membranes is dependent on b-COP (Aniento et al. 1996). In yeast, Arf mutants in which function of coatomer is impaired exhibited an abnormal morphology of endosomes, which appear as hollow spheres of

480

*

G. V. Beznoussenko et al.

interconnected tubules. Transport from endosome to vacuole, an analogue of lysosome in yeast, is inhibited. Similar alterations of endosome structure and function were observed under the action of BFA. ARF1 is required for maintenance not only of yeast GC but also of endosome structure and function. Brefeldin A similarly perturbed endosome morphology and also inhibited transport of some cargo from endosomal structures to the vacuole (Gaynor et al. 1998). In animal cells, the absence of a functional COP coat inhibits MVB biogenesis and therefore also inhibits the accumulation of internal membranes and transport from early to late endosomes. Without functional coatomer, early endosomal membranes are changed into typical clusters of thin tubules (Gu et al. 1997). The delivery of ricin to lysosomes is strongly inhibited in ldlF cells grown at the no permissive temperature. Interestingly, whereas in control cells (ldlF cells preincubated at the permissive temperature) 45% of the total amount of sulphated ricin was transported to the ER, only minimal amounts of ricin, which was both sulphated and glycosylated, were observed in ldlF cells preincubated at the no permissive temperature, thus suggesting that the transport of ricin from the GC to the ER is severely inhibited in ldlF cells lacking e-COP and supporting the idea that ricin is able to reach the ER and intoxicate these cells bypassing the GC. Thus, ricin was

Figure 1. Scheme showing interactions between the exocytosis and endocytosis. Augmentation of Ca2þ concentration and stimulation of cell signalling both at the Golgi level (transportdependent) and at the level of the PM (endocytosis-dependent) induces similar effect within another pathway (arrows 1 and 2). CCB clathrin-coated buds; ER endoplasmic reticulum; ERES ER exit sites; ES endosomes; G the Golgi; GPC Golgi-to-PM carriers; LY lysosomes; MT microtubules; Nu nucleus; PM plasma membrane; RE recycling endosomes.

Interactions between endocytosis and secretory transport

*

481

able to intoxicate ldlF cells depleted of e-COP in the presence of brefeldin A. In contrast, control cells were completely protected against ricin by brefeldin A (Llorente et al. 2003). The role of COPI for endosome function seems mysterious. However, these observations could be easily explained within the kiss-and-run model of intracellular transport (see Chapters 3.2 and 3.4). Indeed, if exocytic carriers interact with endosomal structures through fusion–fission, some COPI coat could leak to endosomal structures. On the other hand, if the regulation of kiss-and-run mechanism by COPI is impaired, fusion–fission events between exocytic carriers would be altered too and as a result the function of endosomes becomes impaired as well. Finally, function of COPI could be important for the segregation of lipid and proteins into these two domains at the level of sorting endosomes. Thus, the secretory and endocytic pathways are closely related to each other (the possible scheme of these interactions is shown in Fig. 1).

Abbreviations BL CSP EM ER GC GFP GPC MVB PM RS RSP TfR TGN VSVG

basolateral constitutively secreting protein electron microscopy endoplasmic reticulum Golgi complex green fluorescent protein Golgi-to-PM carriers multivesicular body plasma membrane regulatory secretion regulatory secretory protein transferrin receptor trans-Golgi network G proteins of vesicular stomatitis virus.

References Ang AL, Folsch H, Koivisto UM, Pypaert M, Mellman I (2003) The Rab8 GTPase selectively regulates AP-1B-dependent basolateral transport in polarized Madin-Darby canine kidney cells. J Cell Biol 163: 339–350 Ang AL, Taguchi T, Francis S, Folsch H, Murrells LJ, Pypaert M, Warren G, Mellman I (2004) Recycling endosomes can serve as intermediates during transport from the Golgi to the plasma membrane of MDCK cells. J Cell Biol 167: 531–543 Aniento F, Gu F, Parton RG, Gruenberg J (1996) An endosomal bCOP is involved in the pH-dependent formation of transport vesicles destined for alte endosomes. J Cell Biol 133: 29–41 Artalejo CR, Henley JR, McNiven MA, Palfrey HC (1995) Rapid endocytosis coupled to exocytosis in adrenal chromaffin cells involves Ca2 þ , GTP, and dynamin but not clathrin. Proc Natl Acad Sci USA 92(18): 8328–8332

482

*

G. V. Beznoussenko et al.

Arvan P, Kuliawat R, Prabakharan D, Zavacki A-M, Elahi D, Wang S, Pilkey D (1991) Protein discharge from immature secretory granules displays both regulated and constitutive characteristics. J Biol Chem 266: 14171–14174 Baldassarre M, Dragonetti A, Marra P, Luini A, Isidoro C, Buccione R (2000) Regulation of protein sorting at the TGN by plasma membrane receptor activation. J Cell Sci 113 (Pt 4): 741–748 hau-Arnaudet G, Desaymard C, Raposo G, Amigorena S (1999) Early Brachet V, Pe endosomes are required for major histocompatiblity complex class II transport to peptide-loading compartments. Mol Biol Cell 10(9): 2891–2904 Buccione R, Bannykh S, Santone I, Baldassare M, Facchiano F, Bozzi Y, Di Tullio D, Mironov A, Luini A, De Matteis MA (1996) Regulation of constitutive exocytic transport by membrane receptors. A biochemical and morphometric study. J Biol Chem 271: 3523–3533 Cancino J, Torrealba C, Soza A, Yuseff MI, Gravotta D, Henklein P, Rodriguez-Boulan E, lez A (2007) Antibody to AP1B adaptor blocks biosynthetic and recycling Gonza routes of basolateral proteins at recycling endosomes. Mol Biol Cell 18(12): 4872–4884 Daro E, Sheff D, Gomez M, Kreis T, Mellman I (1997) Inhibition of endosome function in CHO cells bearing a temperature-sensitive defect in the coatomer (COPI) component epsilon-COP. J Cell Biol 139: 1747–1759 De Matteis MA, Santini G, Kahn RA, Di Tullio G, Luini A (1993) Receptor and protein kinase C-mediated regulation of ARF binding to the Golgi complex. Nature 364: 818–821 Faundez V, Horng JT, Kelly RB (1998) A function for the AP3 coat complex in synaptic vesicle formation from endosomes. Cell 93: 423–432 Feng L, Arvan P (2003) The trafficking of alpha 1-antitrypsin, a post-Golgi secretory pathway marker, in INS-1 pancreatic beta cells. J Biol Chem 278: 31486–31494 Fishman JB, Fine RE (1985) A Golgi-derived coated vesicles can contain both newly synthesized acetylcholintransferase and internalized transferrin. J Cell Biol 98: 423a Futter CE, Connolly CN, Cutler DF, Hopkins CR (1995) Newly synthesized transferrin receptors can be detected in the endosome before they appear on the cell surface. J Biol Chem 270(18): 10999–11003 Gaynor EC, Chen CY, Emr SD, Graham TR (1998) ARF is required for maintenance of yeast Golgi and endosome structure and function. Mol Biol Cell 9(3): 653–670 Gu F, Aniento F, Parton RG, Gruenberg J (1997) Functional dissection of COP-I subunits in the biogenesis of multivesicular endosomes. J Cell Biol 139(5): 1183–1195 Gu F, Gruenberg J (2000) ARF1 regulates pH-dependent COP functions in the early endocytic pathway. J Biol Chem 275: 8154–8160 Gundelfinger ED, Kessels MM, Qualmann B (2003) Temporal and spatial coordination of exocytosis and endocytosis. Nat Rev Mol Cell Biol 4(2): 127–139 Harsay E, Schekman R (2002) A subset of yeast vacuolar protein sorting mutants is blocked in one branch of the exocytic pathway. J Cell Biol 156(2): 271–285 Hedman K, Goldenthal KL, Rutherford AV, Pastan I, Willingham MC (1987) Comparison of the intracellular pathways of transferrin recycling and vesicular stomatitis virus membrane glycoprotein exocytosis by ultrastructural double-label cytochemistry. J Histochem Cytochem 35(2): 233–243 Hicke L, Riezman H (1996) Ubiquitination of a yeast plasma membrane receptor signals its ligand-stimulated endocytosis. Cell 84(2): 277–287 Hicke L, Zanolari B, Pypaert M, Rohrer J, Riezman H (1997) Transport through the yeast endocytic pathway occurs through morphologically distinct compartments and requires an active secretory pathway and Sec18p/N-ethylmaleimide-sensitive fusion protein. Mol Biol Cell 8(1): 13–31

Interactions between endocytosis and secretory transport

*

483

Huang AY, Castle AM, Hinton BT, Castle JD (2001) Resting (basal) secretion of proteins is provided by the minor regulated and constitutive-like pathways and not granule exocytosis in parotid acinar cells. J Biol Chem 276: 22296–22306 Keller P, Toomre D, Díaz E, White J, Simons K (2001) Multicolour imaging of post-Golgi sorting and trafficking in live cells. Nat Cell Biol 3(2): 140–149 Kuliawat R, Arvan P (1992) Protein targeting via the ‘constitutive-like’ secretory pathway in isolated pancreatic islets: passive sorting in the immature granule compartment. J Cell Biol 118: 521–529 Kumari S, Mayor S (2008) ARF1 is directly involved in dynamin-independent endocytosis. Nat Cell Biol 10(1): 30–41 Leitinger B, Hille-Rehfeld A, Spiess M (1995) Biosynthetic transport of the asialoglycoprotein receptor H1 to the cell surface occurs via endosomes. Proc Natl Acad Sci USA 92(22): 10109–10113 Liou JC, Yang RS, Fu WM (1997) Regulation of quantal secretion by neurotrophic factors at developing motoneurons in Xenopus cell cultures. J Physiol 503(Pt 1): 129–139 Liu JP, Robinson PJ (1995) Dynamin and endocytosis. Endocr Rev 16(5): 590–607 Llorente A, Lauvrak SU, Van Deurs B, Sandvig K (2003) Induction of direct endosome to endoplasmic reticulum transport in Chinese hamster ovary (CHO) cells (LdlF) with a temperature-sensitive defect in epsilon-coatomer protein (epsilon-COP). J Biol Chem 278(37): 35850–35855 Lock JG, Stow JL (2005) Rab 11 in recycling endosomes regulates the sorting and basolateral transport of E-cadherin. Mol Biol Cell 16: 1744–1755 Luo H, Nakatsu F, Furuno A, Kato H, Yamamoto A, Ohno H (2006) Visualization of the post-Golgi trafficking of multiphoton photoactivated transferrin receptors. Cell Struct Funct 31(2): 63–75 Mellman I (1996) Endocytosis and molecular sorting. Annu Rev Cell Dev Biol 12: 575–625 Miaczynska M, Zerial M (2002) Mosaic organization of the endocytic pathway. Exp Cell Res 272: 8–14 Millar CA, Meerloo T, Martin S, Hickson GRX, Shimwell NJ, Wakelam MJO, James DE, Gould GW (2000) Adipsin and the glucose transporter GLUT4 traffic to the cell surface via independent pathways in adipocytes. Traffic 1(2): 141–151 Mironov AA Jr, Mironov AA (1998) Estimation of subcellular organelle volume from ultrathin sections through centrioles with a discretized version of vertical rotator. J Microsc 192: 29–36 Morishita M, Engebrecht J (2005) End3p-mediated endocytosis is required for spore wall formation in Saccharomyces cerevisiae. Genetics 170(4): 1561–1574 Orzech E, Cohen S, Weiss A, Aroeti B (2000) Interactions between the exocytic and endocytic pathways in polarized Madin-Darby canine kidney cells. J Biol Chem 275 (20): 15207–15219 Pagano RE, Sepanski MA, Martin OC (1989) Molecular trapping of a fluorescent ceramide analogue at the Golgi apparatus of fixed cells: interaction with endogenous lipids provides a trans-Golgi marker for both light and electron microscopy. J Cell Biol 109(5): 2067–2079 Polishchuk RS, Polishchuk EV, Marra P, Alberti S, Buccione R, Luini A, Mironov AA (2000) Correlative light-electron microscopy reveals the tubular–saccular ultrastructure of carriers operating between Golgi apparatus and plasma membrane. J Cell Biol 148(1): 45–58 Polishchuk EV, Di Pentima A, Luini A, Polishchuk RS (2003) Mechanism of constitutive export from the Golgi: bulk flow via the formation, protrusion, and en bloc cleavage of large trans-Golgi network tubular domains. Mol Biol Cell 14: 4470–4485 Presley JF, Draper RK, Brown DT (1991) Defective transport of Sindbis virus glycoproteins in End4 mutant Chinese hamster ovary cells. J Virol 65(3): 1332–1339

484

*

G. V. Beznoussenko et al.

Puertollano R, Van der Wel NN, Greene LE, Eisenberg E, Peters PJ, Bonifacino JS (2003) Morphology and dynamics of clathrin/GGA1-coated carriers budding from the trans-Golgi network. Mol Biol Cell 14(4): 1545–1557 Riezman H (1993) Three clathrin-dependent budding steps and cell polarity. Trends Cell Biol 3: 330–332 Roos J, Kelly RB (1999) The endocytic machinery in nerve terminals surrounds sites of exocytosis. Curr Biol 9(23): 1411–1414 Rudick VL, Rudick MJ, Munoz-Medellin DA, Brun-Zinkernagel AM, Chang IF (1993) Routing of a secretory protein to the endocytic compartment in transfected Madin Darby canine kidney cells. Cell Mol Biol Res 39(8): 773–788 Sandvig K, Van Deurs B (1996) Endocytosis, intracellular transport, and cytotoxic action of Shiga toxin and ricin. Physiol Rev 76(4): 949–966 Sariola M, Saraste J, Kuismanen E (1995) Communication of post-Golgi elements with early endocytic pathway: regulationof endoproteolytic cleavage of Semliki Forest virus p62 precursor. J Cell Sci 108: 2465–2475 Stoorvogel W, Geuze HJ, Griffith JM, Strous GJ (1988) The pathways of endocytosed transferrin and secretory protein are connected in the trans-Golgi reticulum. J Cell Biol 106(6): 1821–1829 Turner MD, Arvan P (2000) Protein traffic from the secretory pathway to the endosomal system in pancreatic b-cells. J Biol Chem 275: 14025–14030 Wang RH, Colbaugh PA, Kao CY, Rutledge EA, Draper RK (1990) Impaired secretion and fluid-phase endocytosis in the End4 mutant of Chinese hamster ovary cells. J Biol Chem 265(33): 20179–20187 Whitney JA, Gomez M, Sheff D, Kreis TE, Mellman I (1995) Cytoplasmic coat proteins involved in endosome function. Cell 83(5): 703–713

Origins of the regulated secretory pathway

*

485

Origins of the regulated secretory pathway Alexander A. Mironov and Peter Arvan

Introduction Modes of transport of soluble (or luminal) secretory proteins synthesized in the endoplasmic reticulum (ER) could be divided into two groups. The socalled constitutive secretory pathway (CSP) is common to all eukaryotic cells, constantly delivering constitutive soluble secretory proteins (CSSPs) linked to the rate of protein synthesis but largely independent of external stimuli. In regulated secretion, protein is sorted from the Golgi into storage/secretory granules (SGs) whose contents are released when stimuli trigger their final fusion with the plasma membrane (Hannah et al. 1999). In highly specialized exocrine, endocrine, and neural cells, known as professional secretors, the SGs represent the dominant route of secretory protein trafficking, superimposed upon other post-Golgi protein trafficking pathways, even though regulated secretory proteins (RSPs) and some polysaccharides are released at low levels until exocytosis is amplified in response to a stimulus. In SGs, RSPs are stored at high concentration (Castle and Castle 1996) in order to economize cytoplasmic space and prevent premature activation of zymogen forms. The regulated secretory pathway delays release of the secretory product under unstimulated conditions so that, upon demand, protein release rate can quickly exceed protein synthesis rate, potentially resulting in massive release of protein product (in most cases, regulated secretory proteins have to be converted from condensed to soluble form after fusion of SGs with the plasmalemma). Usually the stimulatory signal occurs via binding of a stimulating ligand to its cell surface receptor, which induces entry of Ca2 þ into the cytosol from the extracellular space and from internal storage sites (see Chapter 2.8). Regulated secretion is present in some protists but not found in plants and yeast. However, the isolation of mutants lacking SGs in multicellular animal organisms has never been reported, suggesting an important role of the regulated secretory pathway (Borgonovo et al. 2006). In Tetrahymena thermophila, a master gene for SG biogenesis has not yet been found, but several genes for SG cargoes have been identified, including a family of five proteins all of which are essential for the normal assembly of the granule core (Bowman et al. 2005; Cowan et al. 2005; Borgonovo et al. 2006). In endocrine, exocrine and mast cells, it is typical to have 50% of all translation devoted to the synthesis of SG proteins, and the SGs may occupy an overwhelming

486

*

A. A. Mironov and P. Arvan

fraction of cytoplasm volume. Thus, there is a veritable river of protein flow through the secretory pathway to newly made SGs. SGs are typically distinguished by their morphological appearance and high immunoreactivity for cell-type-specific content proteins, such as peptide hormones, as well as other more widespread cargoes, such as chromogranin A (CgA), B (CgB) and secretogranins II–VI (Taupenot et al. 2003) as well as prohormone convertases 1 (PC1) and 2 (PC2) and carboxypeptidase E (CPE) (Steiner 1998). Yet deletion of various SG genes in mice (reviewed below) neither precludes expression of other SG genes nor the biogenesis of SGs (although processing of cargoes and/or their secretion can be affected, Borgonovo et al. 2006), indicating that multiple gene products contribute to granule structure and function. One feature that tends to be shared by such proteins is their ability to undergo multimerization/condensation under intraluminal conditions prevailing in young SGs, which includes a slightly acidic pH and an increased divalent cation concentration (Laine and Lebel 1999). Although condensation of RSPs has mostly been described in the trans-Golgi elements (“condensing vacuoles”) and immature SGs (ISGs) (Tooze et al. 1987), it has occasionally also been found in cis- and medial-Golgi cisternae (Rambourg et al. 1984; Slot et al. 1997), and rarely in the ER (Geuze and Slot 1980). High Ca2þ concentration and a pH  6.5 in specialized parts of the trans-Golgi network (TGN) especially favour initiation of multimerization of selected SG cargo such as granins. These in turn may form a nidus upon which condensation of other RSPs may follow, driven by multiple interactions among RSPs and helper/assembly factors within SGs. The origins of the RSP seem to derive from the constitutive secretory pathway (CSP). Although no visible granules are apparent, Chavez et al. (1996) suggested that constitutive secretory cells such as CHO or L cells have a cryptic regulated pathway. A significant fraction of [35S] sulfate-labelled free glycosaminoglycan chains were stored intracellularly and showed stimulated secretion after treatments with phorbol ester or those that increase cytoplasmic calcium. As for the morphological appearance of granules, the basic message is that formation of SG structures is to a large extent the result of selforganizing properties of RSP cargo. Moreover, although surely not an essential feature of the regulated secretory pathway (Arvan and Halban 2004), many RSPs are endoproteolytically processed from proprotein precursors by specialized granule processing enzymes not expressed in cells outside of the neuroendocrine system. Such processing can be critical for the biological activity of the stored products (Beuret et al. 2004) and may also change the biophysical properties of RSPs (Zhang et al. 2001), facilitating multimerization in the intragranular ionic environment. Once formed, the majority of SGs accumulate in the cytosol, creating an undocked vesicle pool. ISGs then migrate to the cell periphery with cytoskeleton assistance (Rudolf et al. 2001). ISGs, disconnected from the TGN, acquire at least a partial stimulated exocytosis response to secretagogues (Rindler et al. 2001). Indeed, SGs are subsequently recruited to the plasma membrane

Origins of the regulated secretory pathway

*

487

where they become tethered and primed for fusion with the plasma membrane upon external stimulation (Martin 2003), to deliver their cargo to the extracellular space. This is the process of regulated exocytosis, controlled by Ca2 þ (using synaptotagmin) and mediated by soluble N-ethylmaleimide-sensitive factor attachment protein receptors (SNAREs), which is similar to that occurring at neuronal synapses (DeCamilli and Jahn 1990). Under unstimulated conditions, SGs are long-lived, resulting in cytoplasmic accumulation (Alberts et al. 2002).

Morphology of the regulated secretory pathway Inoverview, theformationofSGsrequiresdenovosynthesisandtranslocationof the RSPs and mediators into the ER and transport to the Golgi and subsequently the TGN, where secretory vesicles are formed. Nevertheless, the morphology of the RSP is different in different cells. The sizes and location of granule formation appears to be related to dilations of condensed material within the interconnected network of Golgi membranes (Rambourg et al. 1993). In the Golgi of the epithelial cells of the lactating rat mammary gland, casein submicelles could be seenalreadywithinthecis-Golgi,becomingmoredistendedintranscisternae.In the Golgi of principal epithelial cells of rat seminal vesicles, the first detectable aggregation of RSPs begins in the medial cisternae. In neuroendocrine cells, peptide hormone precursors and helper/assembly proteins such as granins gradually develop into an electron-dense core in the SG lumen(Arvan andHalban2004).Thesedense-corevesicles inneuroendocrine and endocrine cells show considerable size variation, have a diameter >100 nm and can contain more than one cargo molecule (Sollner 2003). In bovine adrenal chromaffin cells, atrial natriuretic factor is packaged with catecholamines in large dense-core vesicles (Nguyen et al. 1988) with a mean diameter of 380 nm (Duncan et al. 2003). The electron density of the granule core depends upon the internally stored contents and may be less evident in exocrine cells. Additionally, SGs accumulated in differentiated cell lines are typically smaller than those found in their cognate cells in vivo (Feng and Arvan 2003). Condensing vacuoles, still attached to the Golgi stack, contain a denser aggregation of contents than found in the Golgi stacks, while mature SGs (MSGs) contain compact dense cores (Clermont et al. 1993). In one group of cells where SGs do not appear associated with the Golgi, Clermont et al. (1995) observed the TGN as the tubular network connected with last two Golgi cisternae. In a second group of cells such as prolactin-producing cells of lactating rats (Rambourg et al. 1992) ISGs form within the trans-most cisterna and the TGN is smaller. In a third group of cells, ISGs are formed as distensions of most Golgi cisternae and the non-attached, free TGN is even smaller. TGN tubules are almost invisible in cells in which the ISGs start forming from the cismost cisternae (Clermont et al. 1995). ISGs bud from the trans-Golgi network by a process that is said to exhibit superficial similarity to virus budding (Alberts et al. 2002). EM studies have

488

*

A. A. Mironov and P. Arvan

found patches of both clathrin (Andresen and Moore 2001) and coatomer (Fig. 5c in Martinez-Menarguez et al. 1999; and see Andresen and Moore 2001) on the surface of ISGs. Whether ISGs are connected to other elements of the TGN is not always evident. According to Rambourg et al. (2001), ‘in the trans-most Golgi elements, nodular swellings reach their final size and remain interconnected by flattened anastomosed membranous-tubules’. In endothelial cells, connections between forming Weibel–Palade bodies and the TGN are more evident after high pressure freezing than after chemical fixation (Zenner et al. 2007). These images imply that liberation of ISGs may occur by rupture of the tubular areas rather than by ISG budding from saccular elements (Rambourg et al. 2001). ISGs detached from the TGN may still have some characteristics of that area, including lysosomal enzymes (Kuliawat et al. 1997) and clathrin coats attached in patches on the surface (Orci et al. 1987a,b; Tooze and Tooze 1986). In some cell types, the size of SGs is reported to increase during granule maturation, implying that immature SGs (ISGs) may fuse and increase in size while granule contents becomes denser (Clermont et al. 1992).

Birth of the secretory granule at the Golgi complex Formation of MSGs has been divided into two main processes. The first is ISG formation from the Golgi complex and its operation as an advanced sorting station beyond the TGN. The second is the ability of ISGs to use molecular motors to move to the cell periphery that accompanies maturation to MSGs (Rudolf et al. 2001). Useful indicators of the MSG compartment includes prolonged storage in unstimulated cells, the inclusion of specific v-SNAREs (Meldolesi et al. 2004), and acquisition of special markers (e.g., cysteine string protein (Chamberlain et al. 1996)). As ISGs mature, their intraluminal pH becomes more acidic (Duncan et al. 2003), and Ca2 þ more elevated (Winkler and Westhead 1980). This change in ionic environment facilitates conversion of prohormones to mature peptide hormones (Orci 1985; Orci et al. 1986) and facilitates their intragranular storage (Cowley et al. 2000). Along with this, as described above (Dittie et al. 1997), certain proteins present both at the TGN and in ISGs (Orci et al. 1987a,b; Beuret et al. 2004) disappear during granule maturation (Dittie et al. 1997), linked to the loss of clathrin-coated membrane (Kuliawat et al. 1997) with a subset of removed proteins delivered to the endosome–lysosome system (Feng and Arvan 2003). ISGs represent a minority of granules, yet are indistinguishable from MSGs by light microscopy (Rudolf et al. 2001). However, electron microscopic studies have suggested that in some cells the dense core of MSGs is larger in diameter than that of ISGs (Tooze et al. 1991) consistent with homotypic ISG fusion (Wendler et al. 2001), while in other cells MSGs are smaller than ISG (Arvan and Halban 2004). One concept invokes the formation of small free ISGs that undergo homotypic fusion to produce larger granules as a key step in MSG biogenesis (reviewed in Arvan et al. 2002). This reaction can be reconstituted

Origins of the regulated secretory pathway

*

489

in vitro (Wendler et al. 2001), although it has been difficult to confirm in live cells (Clermont et al. 1992; Rindler et al. 2001). A second school of thought views the initial creation of condensing vacuoles at the TGN as larger dilations, which, during and after disconnection from the Golgi complex, proceed to lose membrane and selected contents (i.e., maturation, Sesso et al. 1980). Although there is a heterogeneous picture on the formation and deployment of SGs, a detailed staging of SG life cycle involves: 1. synthesis of RSPs at the ER and their intra-Golgi transport to form ISGs. 2. Maturation of ISGs. 3. Delivery of SGs to the site of the fusion. 4. Priming of SGs. 5. Fusion of SGs with the PM. 6. Elimination of SG membranes from the PM. We now review these areas in the sections that follow (Fig. 1).

Assays of intra-Golgi transport to secretory granules One way to follow protein transport to and through the Golgi complex is to measure endo-H resistance of RSPs. For glycoproteins stored within the lumen of secretory granules, an endo-H-resistant pool of secretory protein generally

Figure 1. Formation of secretory granules (SGs) at the Golgi exit site (Two types of SGs are formed. The first (ASGs) serves for the delivery of RSPs towards the apical plasma membrane (APM). The second type of SGs (BSGs) delivers RSPs to the bosolateral plasma membrane (BLPM). In some cells, the formation of SGs starts in the distensions of COPI-positive Golgi cisternae (1). In other cells, SGs begin to form in the distensions (2) of the attached and free TGN. SGs containing clathrin-coated buds are immature (AISGs and BISGs). Movement of SGs towards the BLPM and their maturation involves interactions with endosomes (E) possibly using kiss-and-run mechanism. APM (solid line) is separated form the BLPM (dash line) by tight junctions (TJ). The ER (the structure with grey content) is marked by ribosomes (pictured as black dots) close to the cis-side of the Golgi contains ER exit site (ERES).

490

*

A. A. Mironov and P. Arvan

becomes increasingly evident, with the size of this pool at steady state approximating the fraction of molecules stored in secretory granules (Stahl et al. 1996). Thus, despite the slowness of ER exit for some secretory proteins (especially those that are difficult to fold), granule half-life in unstimulated cells is sufficiently long that the majority of intracellular regulated secretory proteins are endo-H-resistant (Lara-Lemus et al. 2006). Another way to follow transport is to follow synchronous granulation or re-granulation (Handley et al. 2007). For instance in rat parotid acinar cells, existing secretory granules are almost entirely released immediately after isoproterenol injection, and 60 min later, newly formed granules are observed at the concave trans-face of the Golgi apparatus. Alternatively, formation of new SGs bearing fluorescent cargo has been followed after release from the 20 C block of protein exit from the TGN. This temperature block causes the trans-Golgi region to become markedly extended (Tamaki and Yamashina 2002). In another approach, after 60 min of restored temperature, fluorescent RSP [hCgB-GFP (S65T)] has completely exited from the TGN into ISGs (Rudolf et al. 2001). Both approaches suggest that intra-Golgi protein transport takes less than 60 min. However, precise measurements of the rate of intra-Golgi transport of RSPs have yet to be reported. Several groups have attempted to determine requirements for SG formation fromtheGolgi complexusingcell-freesystems andinpermeabilizedcells (Austin and Shields 1996). For example, pro-opiomelanocortin (POMC) is packaged into granules before its endoproteolytic cleavage (Tooze et al. 1987; Tanaka et al. 1997; Andresen and Moore 2001). The requirements identified to obtain POMC processing (a marker ofISG maturation) are inclusion ofcytosol and acidification of the granule lumen (at least to pH 6.2). Probably this reflects more than just a requirement for acidification, because processing of secretogranins II by PC2 cannot be activated merely by acidification alone. ARF1 enhances the POMC processing(maturation)assay,duringwhichV-ATPasecontributescompartmental acidification for activation of prohormone convertases, and further incubation at 37 C allows free granule release. Granule release is dependent on cytosol and ATP and is effectively inhibited by BFA or 100 mM GTPgS, implicating the activity of a high affinity GTP-binding protein (Dittie et al. 1997; Urbe et al. 1997a,b).

Membrane mechanisms for packaging regulated secretory proteins It is well known that while transiting along the entire secretory pathway, RSPs become progressively concentrated. Concentration is evident already upon exit from the ER (Bendayan et al. 1980; Oprins et al. 2001), continues in the Golgi complex, and proceeds further in cells in which the proteins are efficiently packaged into SGs (Geuze et al. 1979; Geuze and Slot 1980; Slot and Geuze 1983). While prolactin may be 200 times more concentrated in dense core SGs than in the lumen of the ER (Farquhar et al. 1978), amylase

Origins of the regulated secretory pathway

*

491

concentration in exocrine SGs is more moderate, engaging small concentration steps at the cis-side (Postuma et al. 1988) or trans-side (Bendayan 1982) of the Golgi stack. In any event, formation of SGs is directly linked to ongoing transport of RSPs from earlier stages of the secretory pathway. While brefeldin A (BFA) does not affect already existing SGs, BFA causes marked redistribution of RSPs from the Golgi cluster to dilated cisternae of the RER, and blocks granule biogenesis. After BFA washout, Golgi stacks regenerate and shortly thereafter granule formation is re-initiated (Haang and Arvan 1994; Tamaki and Yamashina 2002). Linked to these observations, one hypothesis has implied that concentration of RSPs might occur by COPI vesicles; however, evidence indicates that COPI-coated membranes are mostly depleted of RSPs such as proinsulin in pancreatic beta cells (Orci et al. 1997) or amylase in exocrine cells (MartinezMenarguez et al. 1999). COPI is especially enriched in the membranous network of vesicular tubular clusters (VTCs), but this appears insufficient to explain concentration during RSP transport along the secretory pathway except perhaps to the extent that the concentration event may occur by exclusion (Martinez-Menarguez et al. 1999), resulting in increased protein concentration in the remaining cis-Golgi compartment. For amylase and chymotrypsinogen, an increase in concentration between ER and cis-Golgi compartment was reported as 3.7 and 57.6-fold, respectively (Oprins et al. 2001). If this were to occur solely from ‘concentration by exclusion’ (i.e., without other volume-removal mechanisms), then the initial VTC volume might need to be 57.6-fold larger than that of the final ER exit carriers, which would also need to exhibit unique surface-to-volume characteristics that render such a possibility unlikely. Indeed, after synchronization of transport, the surface area and volume of each Golgi cisterna is equal at any moment in time, and the volume of the VTC compartment is equal to the volume of the whole stack (Trucco et al. 2004). Thus, despite a definite role, Golgi cisternal maturation–progression may not be the sole concentrating mechanism for RSPs. A detailed analysis of various models of intra-Golgi transport is described in Chapter 3.2. For the present discussion, we briefly note that increasing consideration is being given to a “kiss-and-run” model of intra-Golgi transport, which invokes a series of coordinated fusions and fission. This leads to residual connectivity along the secretory pathway, which may account for the fact that labelled lipids may diffuse through the Golgi complex even in aldehydefixed cells – suggesting structural continuities. In addition, this model may help to explain why SG biogenesis is blocked at 20 C (Kuliawat and Arvan 1992; Rudolf et al. 2001; Xu and Shields 1994). The “kiss-and-run” model may be extended to include ISG biogenesis and maturation, implying that Golgi dilations (Ladinsky et al. 2002) fuse with the TGN, inducing concentration and carrier maturation by removal of nongranule resident proteins. The “kiss-and-run” hypothesis is based on the

492

*

A. A. Mironov and P. Arvan

assumption that several separate and distinct compartments co-exist in the secretory pathway and that membrane fusion events (homotypic fusions may be included, but the model especially requires heterotypic fusions) along with consecutive fissions, can be used to generate compartmental maturation. For such a system to work: (1) compartments of distinct ionic environment must be narrowly connected to limit admixing, and (2) condensation of different RSPs along these interconnected but distinct compartmental domains must create a diffusion gradient along the tubular connections. Certainly, long processes/tubules emanating from the TGN have indeed been described (Polishchuk and Mironov 2004; Fig. 2). If intercisternal connections mediated by continuous fusion–fission events are used for granule formation

Origins of the regulated secretory pathway

*

493

and RSP concentration, then predictions of this model may include that: (1) in properly designed studies to stage advancement of RSPs in the secretory pathway, microinjection of mutant aSNAP should be expected to block MSG formation, and (2) upon treatment with ONO or pyrophenone [inhibitors of PLA2a (see also Chapter 3.2.)], MSG generation should similarly be inhibited.

Sorting of RSPs to secretory granules Secretory proteins ultimately follow regulated, constitutive, or constitutivelike secretory pathways. The delivery of different proteins to specific postGolgi carriers and SGs involves processes collectively termed protein sorting and targeting (Arvan and Castle 1998). This occurs mostly at the TGN level. There are three principles driving this sorting: (1) selective entry of RSPs into nascent SGs due to their binding to membrane or cargo receptors, (2) selective retention of these proteins in maturing secretory granules due to multimerization/condensation of RSPs in the ionic environment of SGs, and (3) refinement of membrane and content components by membrane removal mechanisms (Shennan 1996; Martinez-Menarguez et al. 1999). It seems most likely that a mosaic of these sorting mechanisms may contribute to RSP sorting in various regulated secretory cell types (Gorr et al. 2001; Day and Gorr 2003; Kim et al. 2003; Huh et al. 2003). U Figure 2. A schematic representing kiss-and-run mechanism of intra-Golgi transport of newlysynthesized RSPs. The ER-to-post-Golgi transport pathway is pictured as a row of compartments beginning from the ER (left, the most elongated oval filled with RSPs, black dots). Golgi cisternae are pictured as less elongated ovals in the middle, whereas the post-Golgi compartment (oval on the right) is placed above the Golgi cisternae. Each of these compartments could be successively connected or disconnected from each other. Thus in this model, there is alternating fusion and fission between Golgi cisternae and post-Golgi compartments. Initially (upper left panel), the Golgi cisternae and the post-Golgi compartments are shown as empty such as would be seen during a pulse-chase experiment, although in the steady state each of these compartments contains abundant protein undergoing transit. During the second stage of transport (right upper panel), fusion delivers cargo (black dots) to the most proximal Golgi cisterna. For ease of understanding the model, presentation of the ER-to-Golgi transit step is schematized and highly simplified. During the next stage (middle left panel), the first Golgi cisternae becomes disconnected from more proximal membranes whereas the first and second Golgi cisterna become transiently connected with cargo flowing between them. After several further reiterations (the middle right panel), as consecutive compartments become connected and disconnected, cargo reaches the post-Golgi compartment (black dots in the upper oval on the right). In this post-Golgi compartment where ionic pumps would be localized, cargo undergoes precipitation; the aggregated state limits cargo backflow to proximal compartments. The transient nature of the connections between the post-Golgi compartment and other compartments helps to prevent the specialized ionic environment from being dissipated. At the conclusion of intracellular transport, the near-final distributions of newly-synthesized RSPs are shown in the lower panels, with nearly all newly-synthesized cargo molecules becoming precipitated in the post-Golgi compartment (the compartment now called ISGs, or “1” in the Figure). In summary, in this model, the combination of alternating fusion and fission between Golgi cisternae, and aggregation of RSPs, are driving forces that facilitate transport of RSPs for storage in secretory granules.

494

*

A. A. Mironov and P. Arvan

The cargo receptor-based model called “sorting-for-entry” is analogous to receptor-mediated endocytosis and M6PR-dependent lysosomal proenzyme transport (Schmid 1997, see also Chapter 3.6), where cargo binds to receptors that in turn recruit a cytosolic coat. In indirect support of this model, the propeptide of prosomatostatin (Stoller and Shields 1989) a charged/hydrophobic patch (Dikeakos et al. 2007), and a disulfide-bonded loop segment of chromogranin B (CgB) (Kromer et al. 1998; Glombik et al. 1999) and dibasic processing sites (Felliciangeli and Kitabgi 2002) have been proposed as being necessary and sufficient to mediate sorting, acting as putative ligands. An amphipathic loop of POMC was also reported to be required for efficient sorting (Cool et al. 1995). Some major caveats would exist if this were to operate as a sole conserved sorting mechanism (Irminger et al. 1997). First, the failure to discover a sorting signal universal to all RSPs (Halban and Irminger 1994) excludes this as a single general mechanism. Second, luminal protein cargo abundance (often exceeding 100 mg/ml) in conjunction with restricted availability of membrane, limits accessible surface area to only a small fraction of luminal protein. Third, in ‘sorting for entry’ models, CSSPs are portrayed as excluded from granules (reviewed in Beuret et al. 2004; Lara-Lemus et al. 2006), although they may actually be selectively captured for entry into their own export carriers (Lara-Lemus et al. 2006; Rustom et al. 2002) as is seen in apical sorting in epithelial cells (Hansen et al. 2000; Ikonen 2001). Finally, receptor-mediated sorting for entry would lead one to anticipate a significant step-up in luminal cargo concentration during ISG formation, which has not been found, at least not in the exocrine pancreas (Oprins et al. 2001) (although this remains to be determined in AtT20 cells (Dhavantari and Loh 2000)), PC12 cells (Wendler et al. 2001; Feliciangeli and Kitabgi 2002), or Neuro2A cells (Cool et al. 1997)). While it does seem likely that at least some RSPs bind to specific sites on the luminal side of the ISG membrane, the puzzle to be solved in this model is to understand the stoichiometry of cargo receptors to the vast majority of RSPs stored in secretory granules. To the extent that true cargo receptors exist, they must be membrane proteins or luminal proteins engaged tightly with granule lipids. The lipid composition of secretory granules is more cholesterol-rich than that of more proximal secretory pathway compartments, and this lipid composition is likely to be critical to proper formation of outbound membrane carriers (Wang et al. 2000; Tooze et al. 2001). Some RSPs interact with the luminal leaflet of the granule membrane (Dhanvantari and Loh 2000; Glombik et al. 1999; Tooze et al. 2001), including association with “lipid rafts” (Wasmeier et al. 2002; Dhavantari and Loh 2000; Blazquez et al. 2000, 2001). In one example, SgIII association with cholesterol-rich microdomains can facilitate CgA association that can in turn associate with other RSPs (Hosaka et al. 2002; Taupenot et al. 2002a,b). The prohormone-convertases PC1, PC2, and PC5/6A also associate with lipid rafts (Blazquez et al. 2000, 2001; Dikeakos et al. 2007a,b; Lou et al. 2007) where they can contribute to sorting prohormones with dibasic processing sites (Feliciangeli et al. 2001; Brakch et al. 2002;

Origins of the regulated secretory pathway

*

495

Lacombe et al. 2005; Garcia et al. 2005; Mulcahy et al. 2005). The most studied example is CPE, a protein that exhibits association with lipid rafts in the TGN, interacting with membranes via its C-terminal domain (Dhanvantari and Loh 2000), and had at one time been proposed as necessary to efficiently deliver proinsulin into newly forming ISGs (Dhavantari and Loh 2000; Cawley et al. 2003; Dhavantari et al. 2003). Indeed, CPE can co-aggregate with POMC, prolactin, and insulin at an acidic pH in vitro (Cool et al. 1997; Rindler 1998), which has served as a basis for proposing it as a sorting receptor for POMC (Cool et al. 1997) and other peptide hormones. However, not only is sorting of CgA not disturbed in Neuro-2a cells depleted of CPE (Normant and Loh 1998) but entry of proinsulin into ISGs is also not impaired in CPE-deficient Cpefat/Cpefat mice (Irminger et al. 1997; Varlamov et al. 1997). Moreover, in the 4 years since the homozygous CPE null mice have become available (Niamh et al. 2004), there are no indications of an inability of secretory granules to form or for newly synthesized prohormones to enter ISGs in these mice. However, CPE has more recently been proposed as a retention factor for prohormones during maturation of ISGs (Lou et al. 2005). Thus, CPE like many RSPs can serve as helper/assembly factors, facilitating protein condensation during SG biogenesis. Granin-family members including chromogranin A (CgA) (Hosaka et al. 2005), CgB, secretogranin II (SgII), SgIII, and 7B2, are also targeted to SGs (Winkler and Fischer-Colbrie 1992; Song and Fricker 1995; Dannies 1999; Tooze et al. 2001; Day and Gorr 2003) and also have been implicated in RSP condensation (Natori and Huttner 1996). Both CgA and CgB associate with inositol 1,4,5-trisphosphate receptors in a pH-dependent manner (Thrower et al. 2003), although the significance of this is unclear. In CgA-deficient, POMC-expressing PC12 cells, storage of POMC in SGs appears impaired (Kim et al. 2001), but in tissues, granin family members probably exhibit functional redundancy. Mice lacking SgIII reveal no major obvious effects on viability, fertility, or locomotor behaviour (Kingsley et al. 1990). Perhaps this is because CgA is still efficiently targeted to SGs in cells, such as primary adrenal chromaffin cells, in which SgIII is lacking (Hosaka et al. 2002; Taupenot et al. 2002a,b). However, absence of CgA also generates a largely unremarkable mouse phenotype (Hendy et al. 2006), again speaking to redundancy between granins as well as non-granin RSPs. By contrast, overexpression of CgA (Kim et al. 2001), CgB (Huh et al. 2003); or various RSPs can induce formation of SG-like structures with an electron-dense core even in cells lacking the traditional regulated secretory pathway (Michaux et al. 2003; Beuret et al. 2004). Indeed, transient expression in the COS-1 fibroblast cell line of several RSPs as well as SgII and CgB, but not alpha 1-antitrypsin, accumulate with varying efficiency in granule-like cytoplasmic structures devoid of markers of the endoplasmic reticulum, endosomes, and lysosomes, but partially co-localizing with TGN46 (Beuret et al. 2004). Expression of CgA (Kim et al. 2001) or CgB (Huh et al. 2003) induces related morphological structures, although 80%

496

*

A. A. Mironov and P. Arvan

of CgB–GFP is released from Vero cells within 2 h after synthesis (Wacker et al. 1997) with a storage of pro-Vasopressin in heterologous cells that is even lower (Beuret et al. 2004). While expression of granins or other RSPs in heterologous cells does not induce gene expression of other SG components (Kim et al. 2001; Huh et al. 2003; Beuret et al. 2004), overexpression of CgB may increase the sorting and processing of selected RSPs in regulated secretory cells (Natori and Huttner 1996). CgA forms dimers at pH 7.5 and tetramers at pH 5.5 (Yoo and Lewis 1992). The granin multimers formed in a weakly acidic, high Ca2þ -intragranular milieu can in turn facilitate RSP condensation (Yoo 2000; Yoo et al. 2001; Taupenot et al. 2002a,b; Dannies 1999; Thrower et al. 2003). Thus, like CPE, granins are also SG helper/assembly factors (Kim et al. 2001, 2003; Day and Gorr 2003), but endocrine cells clearly engage more than one gene product regulating the process of granule biogenesis and optimizing RSP storage efficiency (Beur et al. 2004).

RSP polymerization RSPs appear to have evolved to lose solubility via multimerization in the intraluminal space (Kuliawat et al. 2000; Lee et al. 2001; Arvan et al. 2002; Song and Fricker 1995; Yoo 1996), indicating that differences in solubility may play an important role in protein trafficking. The main site of RSP condensation is the TGN with continuation in ISGs (although multimers can form within earlier compartments – see above sections). Condensed proteins often accumulate in dilated regions of the TGN possessing tubular interconnections (Clermont et al. 1995), which can result in formation of a concentration gradient (Bauerfeind and Huttner 1993) that, according to the “kiss-and-run” model (see above) can lead to formation of ISGs. In such a view, one might expect no important structural requirements for luminal protein entry into ISGs (El Meskini et al. 2001; Molinete et al. 2000) other than these are proteins that had not otherwise been removed via TGN tubule sequestration or fragmentation. Multimeric protein assembly is one of the factors regulating efficiency of RSP storage by mechanisms that might be considered as the culmination of the intra-Golgi transport process. In the extreme case, it would be plausible for >99% of RSPs to be sorted by multimerization with <1% (i.e., an undetectable fraction) of molecules sorted by receptor interactions, underscoring the potential quantitative importance of passive associations in the appropriate intraluminal environment (reviewed in Arvan and Halban 2004). Thus, the bioactivity of substoichiometric amounts of SG assembly/helper factors in the Golgi complex is augmented by the predisposition of RSPs to multimerize under intraluminal TGN conditions (Chanat and Huttner 1991; Colomer et al. 1996). However, the consequences for protein sorting in the secretory pathway as a result of higher-order protein assembly may be less predictable than for receptor-mediated trafficking events. This is because there may be equilibria between soluble and multimeric states, competitions at the luminal

Origins of the regulated secretory pathway

*

497

leaflet and in the soluble phase for both heterotypic and homotypic protein interactions, mixed complexes containing similar proteins in different stoichiometric ratios, and different post-translationally modified forms of the same gene product exhibiting different properties (Zhang et al. 2001). Remarkably, most condensed RSPs rapidly dissolve when the ionic environment and concentrations of the intragranular space are reversed (Arvan et al. 1984). Multimerization interactions may be homotypic or heterotypic. In endocrine cells, an increase in SG density is accompanied by membrane remodelling (Eaton et al. 2000) and even crystallization of internal cargo (Greider et al. 1969; Kuliawat and Arvan 1992), although actual crystal formation is not required for storage (Arvan and Halban 2004). In exocrine cells, certain sulphated proteoglycans may facilitate efficient storage of amylase in parotid SGs (Venkatesh and Gorr 2002) and may contribute to heterotypic RSP storage in the exocrine pancreas (De Lisle 2002; Venkatesh et al. 2004), even as they play no clear role in sorting and storage of endocrine RSPs (Gorr 1996). Differences in aggregation of individual proteins (Colomer et al. 1996; Kleene et al. 1999) can result in different RSPs being sorted to distinct intracellular locations (Klumperman et al. 1996). In exocrine cells two distinct proteins initially appearing in ISGs at one stoichiometric ratio are finally stored in MSGs at a different stoichiometric ratio (Von Zastrow and Castle 1987). In somatomammotrophs of the bovine pituitary, distinct prolactin and growth hormone granules appear in the same cells, and occasionally appear in separate aggregates within single granules (Fumagalli and Zanini 1985; Hashimoto et al. 1987). In bag cell neurons of Aplysia californica, the precursor of egglaying hormone is cleaved into distinct C-terminal and N-terminal products that are packaged into separate granules, with sorting that appears to occur by homotypic multimerization (Sossin et al. 1990). The partitioning of luminal volume between various outbound pathways from the TGN is likely to be a factor in these various regulated secretory cell types (Arvan and Castle 1998; Thiele et al. 1997) and between cell lines (Rindler et al. 2001) and their cognate cell types in vivo. Protein structure optimized for the distinct intracompartmental ionic environment of SGs is an important factor in RSP multimer formation and storage. Simple engineering of hexahistidine epitope tags onto secretory proteins can enhance calcium-induced RSP condensation within the SG environment (Gorr 1996; Oliver et al. 1997; Jain et al. 2000). The “granin loop” (Song and Fricker 1995) is another feature likely designed for condensation in the high calcium, lower pH environment. This ability to undergo calcium- and low-pH-induced condensation has been demonstrated for both endocrine (Bell-Parikh et al. 2001; Jain et al. 2000), and exocrine (Dartsch et al. 1998; De Lisle 2002) RSPs. A CgA–GFP fusion protein is trafficked to dense-core granules in PC12 cells (Taupenot et al. 2002a,b) and exposure to bafilomycin A1 causes a substantial decrease in detectable granules (Taupenot et al. 2005), while such treatment of pituitary cells causes POMC and prolactin to accumulate in larger vacuolar structures (Henomatsu et al. 1993; Schoonderwoert

498

*

A. A. Mironov and P. Arvan

et al. 2000). Thus, in order to induce condensation of RSPs, SGs must generate an acidic pH (reviewed in Wu et al. 2001). Mechanisms that SGs may use to achieve a more acidic pH than the TGN may also include a low transmembrane proton leak and a favourable buffering capacity at low pH. Additionally, calcium buffering (e.g., CgB binds >90 Ca2þ /mol with a Kd of 1.5 mM) promotes RSP condensation as well as providing calcium for stimulated release to the cytosol (Yoo et al. 2001).

Insulin storage as an example The “sorting by retention” model was developed from studies of trafficking of proinsulin-derived peptides (the major peptides manufactured in pancreatic beta cells) (Arvan and Halban 2004). Proinsulin forms hexamers but they have very poor self-association properties to form higher-order complexes (Orci 1985). Nevertheless the vast majority of newly synthesized proinsulin in beta cells is directed into ISGs (Rhodes and Halban 1987). Despite hypotheses to the contrary (Orci et al. 1984a,b; Fredman et al. 2000; Osterbye et al. 2001), the existence of a specific TGN-based sorting mechanism for entry into ISGs has not been found for proinsulin (Glombik and Gerdes 2000). In pancreatic beta cells, proinsulin rather than conversion intermediates or final processing products) is initially packaged in ISGs. Thus the earliest detectable granules are proinsulin-rich, replete with electron-pale material, carry a (discontinuous) clathrin coat, and have a weakly acidic milieu (Orci et al. 1984b, 1986). In the absence of stimulation, proinsulin is converted to insulin, which is retained for storage in granules, whereas in the presence of secretagogues there is stimulated release of both proinsulin and newly made insulin. The majority of endogenous proinsulin is recovered in the soluble phase prior to its conversion to insulin within ISGs. By contrast, intragranular insulin polymerizes as a function of both prevailing proton and zinc concentrations (Kuliawat and Arvan 1994). Even in yeast, endoproteolytic processing in the late Golgi to form single chain insulin from a fusion protein precursor promotes intracellular retention, preventing rapid secretion of the precursor (Zhang et al. 2001). In mammalian beta cells, after proteolytic processing, insulin becomes insoluble within granules whereas C-peptide remains completely soluble. During the processing period, lysosomal procathepsin B, long recognized as an insulin granule component (Docherty et al. 1984), is actively removed from beta cell ISGs (Kuliawat and Arvan 1994). Nevertheless, the majority of granule C-peptide is not removed from islet ISGs but remains behind in mature granules (Kuliawat and Arvan 1992) because constitutive-like protein traffic conveys only a very minor fraction of ISG volume (reviewed in Arvan and Halban 2004). Indeed, it has been questioned in primary beta cells whether insulin is significantly better retained in granules than proinsulin (Molinete et al. 2001), although such a result does seem clear in cell culture models (Kuliawat et al. 2000). Further, a point mutant of proinsulin, His-

Origins of the regulated secretory pathway

*

499

B10Asp, that is defective for hexamerization, tends to be associated with enhanced constitutive or constitutive-like secretion (Carroll et al. 1988; Gross et al. 1989), and is found at excessive levels in the circulation of human patients bearing the mutation (Chan et al. 1987). A proinsulin in which dibasic endoprotease cleavage sites were mutated to residues that cannot be cleaved was reported to be sorted efficiently to SGs (Halban and Irminger 2003), consistent with another report that a mutant bearing no C-peptide whatsoever is efficiently sorted to SGs (Powell et al. 1988). These findings have been used to argue in favour of selective sorting for proinsulin entry into ISGs rather than efficient storage by condensation of insulin. However upon further study, it was found that the latter molecules (lacking C-peptide) were quantitatively misfolded, bearing mispaired disulfide bonds (Liu et al. 2003). Surprisingly, this unsuspected misfolding does not prevent transit of the mutants to the TGN and beyond (Liu et al. 2003), but it would imply that a putative proinsulin sorting receptor (Dhanvantari et al. 2003) would need also to be able to recognize a completely non-native ligand. We think this is possible, but unlikely. All in all, the insulin example tends to support the sorting by retention model; favouring the view that polymeric assembly of newly synthesized insulin (Huang and Arvan 1994, 1995) helps to enhance its storage in maturing secretory granules (Kuliawat and Arvan 1992). This model is discussed further in the next section.

Entrance, exit, and avoidance of non-granule proteins in ISGs The ‘sorting by retention’ model invokes both multimerization/condensation of specific RSPs to facilitate retention in ISGs along with membranous removal of proteins that are not destined to be stored in SGs (Arvan and Halban 2004). Thus, the sorting-by-retention hypothesis does not require that specific sorting information is used at the entry step into forming SGs, thereby rendering the SG itself as an important sorting station. Indeed, lysosomal proenzymes, and endosomal and lysosomal membrane proteins en route, and luminal and membrane proteins thought of as markers of the constitutive secretory pathway, can also enter ISGs (for review see (Arvan and Castle 1998)). Newly synthesized lysosomal proenzymes en route to the endosomal system enter ISGs (Kuliawat and Arvan 1994; Klumperman et al. 1998; Turner and Arvan 2000). To refine the composition of MSGs, these proteins must be removed as ISGs mature (Kuliawat and Arvan 1994); indeed, procathepsin B is virtually quantitatively removed during ISG maturation. When lysosomal proenzymes are defective for specific M6PR-mediated recognition, their entry into granules is even more abundant – although removal from maturing granules is blocked (Kuliawat and Arvan 1992). Although challenged by some (Rindler et al. 2001), the bulk of the evidence favours that there also can be substantial entrance of soluble secretory proteins from the TGN into ISGs (Arvan and Castle 1998). This entry seems

500

*

A. A. Mironov and P. Arvan

to require neither N-glycans (Kelly et al. 1983) nor lipid raft association (Feng and Arvan 2003). For instance, SEAP, a truncated form of alkaline phosphatase lacking the glycosyl phosphatidylinositol membrane anchor, has been described as a CSSP (Gorr 1996; Harrison et al. 1996; Molinete et al. 2000) although at least in pancreatic beta cells, SEAP abundantly enters ISGs (LaraLemus et al. 2006). Indeed, there are no data indicating that initial entry of SEAP into b-cell secretory granules is any less efficient than that of endogenous proinsulin, although ultimate storage of the two proteins in SGs differs (Arvan and Halban 2004). While a CgB–eGFP fusion protein is targeted to ISGs (Kaether et al. 1997), soluble secretory GFP (a protein that is not normally targeted to the secretory pathway but that can be introduced therein merely by virtue of the presence of a cleavable signal sequence) also abundantly enters SGs (El Meskini et al. 2001; Molinete et al. 2006). This form of GFP not only enters insulin secretory granules, but it also appears to be stored and secreted with efficiency that is comparable to that of authentic insulin (Molinete et al. 2006). However, the ability of eGFP to form disulfide-linked oligomers may affect the outcome of these experiments (Jain et al. 2001; Feng and Arvan 2003). Nevertheless, entrance from the TGN into ISGs of “marker proteins of the constitutive pathway” is not limited to fusion proteins or genetically engineered GFP-containing constructs (Feng and Arvan 2003). How would CSSPs be prevented from entry into ISGs? Several experiments have been performed to date (Tooze and Huttner 1990; Chanat and Huttner 1991), and thus far, only very few identified proteins have been found to meet the criteria of failing to enter ISGs. Recently, a genetically truncated version of Cab45 (a protein that binds Ca2 þ and normally resides in the Golgi lumen, Scherer et al. 1996) was described as the first bona fide constitutive secretory marker to be identified without demonstratably entering ISGs. Similar to other members of the CREC family, loss of information encoded in the Cterminal region of Cab45 causes the protein to lose its intracellular retention within the Golgi complex (Honore and Vorum 2000). Indeed, Cab308myc is no longer colocalizes with the GM130 Golgi marker, but rather is distributed to vesicular organelles at the cell periphery that are non-overlapping with the immunoreactive distribution of insulin (SGs) or proinsulin (ISGs) (Lara-Lemus et al. 2006). Importantly, unlike SEAP or even alpha 1-antitrypsin (Feng and Arvan 2003), at no time does Cab308myc ever enter a stimulus-releasable compartment – at least not in INS1 insulinoma cells, and apparently also not in AtT20 pituitary cells (Lara-Lemus et al. 2006). How does Cab308myc avoid entry into ISGs? Interestingly, upon permeabilization of organelle membranes with saponin, SEAP behaves as a typical soluble secretory protein, being fully extracted by the detergent treatment, whereas Cab308myc remains associated with the luminal aspect of secretory pathway membranes destined to become transport intermediates for the constitutive secretory pathway (LaraLemus et al. 2006). At the time of exocytosis, the membrane associations of such secretory proteins must be reversed to account for their free release to the extracellular environment upon exocytosis (Schlegel et al. 2001).

Origins of the regulated secretory pathway

*

501

In addition to this constitutive secretory pathway conveyed directly from the TGN to the cell surface in regulated secretory cells (Orci et al. 1987a,b), the constitutive-like secretory pathway involves two limbs: one from ISGs (and the TGN) to the endosomal system followed by a second from endosomes to the cell surface (Turner and Arvan 2000). Constitutive-like secretion of lysosomal procathepsin B appears to proceed via an endosomal intermediate (Turner and Arvan 2000), although the role of endosomes as intermediates in the release of bona fide secretory proteins has been less well studied (Millar et al. 2000). In our view, the important aspect of the constitutive-like secretory pathway is not the quantitative contribution of its contents to overall secretion, but rather the notion that a membrane trafficking pathway out of ISGs is likely targeted to endosomes rather than direct retrograde transport to the Golgi/TGN. Membrane recycling between exocytic and endocytic pathways is described in greater detail in Chapter 3.10, but here, we wish to emphasize that these relationships are especially prominent in regulated secretory cells. Once reaching endosomes, a fraction of secretory protein is likely to be channelled to lysosomes for degradation (Neerman-Arbez and Halban 1993) with another fraction conveyed to the extracellular space, creating constitutive-like secretion (Kuliawat and Arvan 1992). Support continues to grow for the idea that unstimulated secretion of proteins that had already traversed, but were no longer contained within, the secretory granule compartment (Arvan et al. 1991; Huang et al. 2001) is likely to reflect membrane trafficking events occurring via an endosomal intermediate (Feng and Arvan 2003).

Role of clathrin In neuroendocrine cells, there is clear concentration of clathrin at regions of the TGN membrane engaging in granulogenesis (Orci et al. 1984a,b) and on parts of the membrane of ISGs. In both TGN dilations and detached ISGs, most clathrin is accumulated on clathrin-coated buds (for review see Halban and Irminger 1994). Clathrin coat, mannose phosphate receptors (M6PRs), furin, and AP-1 adaptors present at the TGN and ISGs (Orci et al. 1987a,b; Dittie et al. 1996; Klumperman et al. 1998) disappear in MSGs. Excess membranes and proteins such as the ubiquitously expressed TGN/ endosomal membrane protease furin, are thought to be removed from maturing granules by budding of clathrin-coated ISG-derived vesicles (Dittie et al. 1996, 1997; Klumperman et al. 1998). Indeed, a list of proteins removed from ISGs may include some CSSPs as well as lysosomal enzymes and endo/ lysosomal membrane proteins such as M6PR. Thus, sorting of lysosomal enzymes from the ISGs uses the same system involving the M6PR as elsewhere in the TGN (Kuliawat et al. 1997). GGA (Golgi-associated g-ear-containing ADP-ribosylation-factor-binding protein) is an adaptor protein recruiting clathrin to ISGs (Kakhlon et al. 2006). Only after reaching a relatively advanced stage of morphological maturation do SGs lose the clathrin coat

502

*

A. A. Mironov and P. Arvan

(Tooze and Tooze 1986). Neither the dominant negative clathrin hub fragment (Molinete et al. 2001) nor arginine/lysine analogs (Kuliawat and Arvan 1992) block creation of new granules but both demonstrate that clathrin function is needed for ISG maturation. What might greatly advance this model of ISG maturation would be if clathrin-dependent vesicles after budding from ISGs could be isolated in vitro after incubation with cytosol and appropriate budding factors. Clathrin might also be invoked in the sorting-for-entry model, in which RSPs bind specific membrane receptors. However, it has been shown that TGN and trans-Golgi-derived clathrin-coated buds contain Golgi resident proteins (Velasco et al. 1993) but excluding secretory protein cargoes (Salamero et al. 1990); and experiments with yeast mutants reveal that clathrin can play a direct role in the retention of Golgi resident proteins (Seeger and Payne 1992). Lui-Roberts et al. (2005) have recently proposed a novel scaffolding role for an AP-1/clathrin coat in initial endothelial cell granule formation. However, unless we invoke significant cell type-specific differences, such a model cannot account for the fact that expression of the dominant negative clathrin Hub peptide does not affect the efficiency of delivery of peptide prohormones to SGs (Molinete et al. 2001); and clathrin depletion does not affect prohormone processing (Andresen and Moore 2001).

Role of cytoskeleton in transport of SG ISGs and MSGs have somewhat different localizations. The cell cortex is the site of priming and docking of SGs (Robinson and Martin 1998). The fraction of granules docked at the plasmalemma varies with cell type (e.g., in PC12 cells, the majority of MSGs is localized in close proximity to the plasma membrane (Baneriee et al. 1996)); such granules can be recruited to undergo stimulated exocytosis (Martin and Kowalchyk 1997). While certainly not the only factors, both microtubules and F-actin synergistically and consecutively affect SG cortical localization. Microtubules are implicated particularly in fast transport of ISGs from the TGN to the PM (Rudolf et al. 2001). Newly made ISGs may leave the TGN in straight trajectories towards the plasmalemma with maximal velocities of up to 2 mm/s (Rudolf et al. 2001). This straight, unidirectional transport contrasts with the anterograde transport of constitutive carriers which is also microtubule-dependent but may occur in a random bidirectional manner (Wacker et al. 1997) – although in NGF-differentiated PC12 cells, SGs may also move both anterogradely and retrogradely (Lochner et al. 1998; Rudolf et al. 2001). In the presence of nocodazole to disrupt microtubules, ISGs do not accumulate in the TGN area but reach the PM more indirectly (Rudolf et al. 2001). Using a phogrin-dsRed fusion protein to follow SG movement (Wasmeier and Hutton 1996), anterograde transport along microtubules was shown to require the ATP-dependent activity of the conventional kinesin motor (Varadi

Origins of the regulated secretory pathway

*

503

et al. 2002). Using another RSP tagged with the dsRed-E5 ‘timer’ protein (a fluorophore that progressively shifts its fluorescence emission over 16 h, Terskikh et al. 2000), it was found that newly generated SGs are quickly mobilized to the plasma membrane (Duncan et al. 2003) and, as a function of age, their speed and direction changes with some granules moving rapidly in a saltatory manner along straight lines, indicating transport along tracks. Actin filaments engage SGs at slightly later stages of maturation. In neuroendocrine chromaffin and PC12 cells, completion of ISG-to-MSG maturation is kinetically linked to F-actin binding (Rudolf et al. 2001). Specifically, upon shift from 20 C (a temperature at which detachment of ISGs from the TGN is blocked, Kuliawat and Arvan 1992) to 37 C, GFP-labelled ISGs move within seconds in a microtubule-dependent fashion to the F-actin-rich cell cortex where they mature as measured by progressive loss of furin signal (Rudolf et al. 2001). SGs undergo both F-actin-dependent and independent immobilization (Rudolf et al. 2001), and interaction of SGs with cortical F-actin (Trifaro and Vitale 1993) remains preserved in vitro in preparations of plasma membrane sheets (Martin and Kowalchyk 1997; Avery et al. 2000).

SNAREs and RABs in the regulated secretory pathway A complete review of these proteins is beyond the scope of this chapter. However, it is necessary here to at least recognize that SNARE proteins move dynamically during compartmental genesis and maturation, while selected Rab proteins are acquired on the cytosolic face of the organelle. Syntaxin 6 (Syn6) is predominantly distributed in endosomes and the TGN (Kuliawat et al. 2004) although Syn6 is also present in ISGs but not MSGs (Bock et al. 1997; Klumperman et al. 1998; Steegmaier et al. 1999; Eaton et al. 2000; Hinners et al. 2003). Homotypic ISG–ISG fusion, proposed as a step in MSG biogenesis (Tooze et al. 1991; Urbe et al. 1998) engages Syn6 (Wendler et al. 2001) and synaptotagmin IV (Ahras et al. 2006). Interestingly, the process of secretory granule maturation is SNAP-dependent (Chamberlain et al. 1995) suggesting involvement of (forward or retrograde) membrane fusion machinery. The final stages of SG lifespan (Nemoto et al. 2001; Thorigrave;n et al. 2004; Sørensen 2004) require SNAREs to trigger exocytosis in the presence of Ca2þ . Synaptotagmin (a Ca2þ – and phospholipid-binding protein) is involved in setting the Ca2þ dependence of the fusion process (Sorensen 2004). In sea urchin eggs, SNARE complexes primarily increase the Ca2þ sensitivity of fusion between plasma membrane and cortical exocytic vesicles (Coorssen et al. 1998; Johns et al. 2001). In turn, SNARE proteins undergo assembly into functionally active complexes following a rise in Ca2þ , with formation of the tSNARE complex preceding formation of the trans vSNARE–tSNARE assembly. The details of tSNARE complex formation have not been characterized in vivo, thus it is not certain which represents the core complex precursor in vivo (Stojikovic 2005).

504

*

A. A. Mironov and P. Arvan

Assembly of the SG–plasma membrane SNARE complex consists of two main steps. The first involves a high affinity syntaxin interaction with the N-terminal SNARE motif of SNAP25 (An and Almers 2004). This complex assembles reversibly during transient elevations in Ca2þ induced by depolarization of cells and tolerates a mutation that blocks formation of other kinds of syntaxin–SNAP25 complexes. This complex might function as the plasma membrane “docking receptor” for SGs containing VAMP, to be followed by trans-SNARE formation during a Ca2þ-dependent priming step of exocytosis (Stojikovic 2005). However, at steady state, certain molecules such as VAMP4 are distributed in the TGN and ISGs but not in MSGs (Bock et al. 1997; Klumperman et al. 1998; Steegmaier et al. 1999; Eaton et al. 2000; Hinners et al. 2003) eliminating them from consideration as partners in the final stages of SG docking. Thus, some SNARE proteins in ISGs function only in a prelude to actual stimulus-dependent exocytosis (Regazziet al. 1996; Wheeler et al. 1996). Another important class of regulators are the Rabs, especially members of the Rab3 and Rab27 families. In PC12 cells, endogenous Rab3a colocalizes with SGII-marked SGs, and transiently expressed EGFP–Rab3a (and ECFP–Rab27a) preferentially localizes to newly made ISGs dissociated from the TGN within 20 min after release of the 20 C temperature block (Handley et al. 2007). Experiments using fluorescence recovery after photo bleaching have suggested that Rab3a might continue to rapidly exchange between granulebound and cytosolic forms, yet exocytosis per se does not drive immediate release or dispersal of Rab3a from granule membranes (Handley et al. 2007). A potentially important effector of Rab3a is Noc2, whose deficiency in genetically altered mice results in diminished insulin secretion from pancreatic beta cells as well as significant defects in exocrine granule exocytosis (Matsumoto et al. 2004). The Rab27 subfamily consists of two homologs with closely related but non-identical function in the regulated secretory pathway. Rab27a is known to be important for the biogenesis and function of lysosome-related organelles such as melanosomes (Barral et al. 2002) although it has also been implicated in endocrine SG docking at the plasma membrane (Kasai et al. 2005). Rab27b may be expressed in a somewhat more narrow range of tissues than Rab27a. Loss of Rab27b function seems to be associated with a prominent decrease in the numbers of platelet granules and a bleeding disorder (Tolmachova et al. 2007; Tsuboi and Fukuda 2006), as well as abnormal movement and secretion of mast cell granules (Mizuno et al. 2007). Nevertheless, Rab27b may also play a role in granule exocytosis in the brain and in certain endocrine and exocrine tissues (Zhao et al. 2002; Imai et al. 2004; Gomi et al. 2007).

Exocytosis of SGs As with the previous section, such a topic extends well beyond the scope of this chapter. Here, we merely wish to note that design of the secretory granule

Origins of the regulated secretory pathway

*

505

must build in how the SG will be deployed to complete the anterograde transport pathway with depletion of granule contents and selective handling of SG membranes (Borgonovo et al. 2006). Exocytosis of SG contents to the extracellular space is mediated by complete or partial fusion (“kiss-and-run”, with a variable duration of <1 s to >10 s (Rutter and Tsuboi 2004) of SG continuities with the plasma membrane (Tsuboi and Rutter 2003; Tsuboi et al. 2004). In its most minimal form, a narrow pore allows only selective release of low-molecular-mass species (ATP, GABA, etc.) from SGs. Electrophysiological recordings appear unable to detect fusion events with such a minimal pore size. In a greater form, the fusion opening proceeds to limited admixing of vesicle membrane proteins with those of the plasma membrane (Tsuboi et al. 2004) while constituents of the dense core are extruded. Live cell imaging of fluorescent probes such as the lipophilic dye FM-1-43 appears able to detect a multiplicity of such interactions between SGs and the PM (Leung et al. 2002). To be available for either form of fusion with the plasma membrane, SG docking and priming events must take place first. Priming occurs in two stages, resulting in generation of slowly releasable and rapidly releasable pools of granules. Recent findings indicate that all transitional stages leading up to and including the fusion event itself have potential reversibility (Martin 2003). While priming requires only a moderate increase in Ca2 þ as well as the presence of MgATP, SG exocytosis demands a more substantial elevation in Ca2 þ and exhibits a considerably longer latency than that known for synaptic vesicles. SG exocytosis can occur from either the slowly or rapidly releasable primed pool of SGs, but with different rates and latency. Thus, the extent of content release is variable with respect to both the fraction of cellular granules involved, and the fraction of cargo of individual SGs undergoing release – both depend on the nature and strength of the secretagogue as well as key second messengers (Stojikovic 2005) and their effectors (Tsuboi et al. 2006). One of the molecules that may limit full fusion and diffusional intermixing of SG membranes with the plasmalemma is dynamin. When simultaneously imaged with a fluorescent vesicle cargo (NPY-mRFP), dynamin-1-EGFP is observed to arrive at sites of exocytosis synchronously with the onset of release events, and to linger for 2–3 s after the peak of release (Artalejo et al. 2002), where it may be engaged in closing the fusion pore (Tsuboi et al. 2004). Dynamin recruitment is also likely to be a key feature to the nearly intact recapture of SG membranes to the cytoplasm following exocytosis (Taraska et al. 2003). However, the signals that recruit dynamin to SG exocytotic sites remain unclear. Local increases in PtdIns(4,5)P2 and increased local Ca2þ (Wiser et al. 1999; Emmanouilidou et al. 1999) might each play roles in recruiting dynamin to the sites of potential binding partners, including members of the SNX family (sorting nexins, especially SNX9; Soulet et al. 2005) as well as phospholipase D (Lee et al. 2006), syndapin-1/PACSIN (Anggono et al. 2006), and syntaxin-1 (Galas et al. 2000).

506

*

A. A. Mironov and P. Arvan

Conclusions This chapter on origins of the SG highlights the creation and utilization of a novel organelle that has evolved with the increasing complexity of eukaryotic cells and been assigned a highly specialized task for regulated cargo delivery to the cell surface. Much of the task begins in the Golgi complex, and is critically linked to the specific biophysical behaviour of content proteins that helps to drive protein condensation for storage (see Fig. 1). Regardless of model invoked for the origins of the regulated secretory pathway, the specialization and refinement of the SG membrane appears to require ongoing membrane fusion and fission events. Herein, we present the idea that “kiss-and-run” – a model that has grown in popularity to account for many exocytotic events – may be equally applicable in earlier intracellular compartments along the anterograde transport pathway, and could be intimately associated with Golgi structure and sorting mechanisms for granule biogenesis (Fig. 2).

Abbreviations BFA Cg CPE CS CSSP ER ISG MSG PC1 PC2 POMC RSP SG SgII TGN VTC

brefeldin A chromogranin carboxypeptidase E constitutive secretory constitutive soluble secretory protein endoplasmic reticulum immature SGs mature SG prohormone convertases 1 prohormone convertases 2 pro-opiomelanocortin regulatory secretory protein storage/secretory granule secretogranin II trans-Golgi network vesicular tubular cluster

References Ahras M, Otto GP, Tooze SA (2006) Synaptotagmin IV is necessary for the maturation of secretory granules in PC12 cells. J Cell Biol 173(2): 241–251 Alberts B, Johnson A, Lewis J, Raff M, Roberts K, Walter P (2002) Molecular biology of the cell, 4th edn. Garland Science, New York An SJ, Almers W (2004) Tracking SNARE complex formation in live endocrine cells. Science 306: 1042–1046 Andresen JM, Moore HP (2001) Biogenesis of processing-competent secretory organelles in vitro. Biochemistry 40(43): 13020–13030

Origins of the regulated secretory pathway

*

507

Anggono V, Smillie KJ, Graham ME, Valova VA, Cousin MA, Robinson PJ (2006) Syndapin I is the phosphorylation-regulated dynamin I partner in synaptic vesicle endocytosis. Nat Neurosci 9(6): 752–760 Artalejo CR, Elhamdani A, Palfrey HC (2002) Sustained stimulation shifts the mechanism of endocytosis from dynamin-1-dependent rapid endocytosis to clathrin- and dynamin-2-mediated slow endocytosis in chromaffin cells. Proc Natl Acad Sci USA 99(9): 6358–6363 Arvan P, Rudnick G, Castle JD (1984) Osmotic properties and internal pH of isolated rat parotid secretory granules. J Biol Chem 259: 13567–13572 Arvan P, Castle D (1998) Sorting and storage during secretory granule biogenesis: looking backward and looking forward. Biochem J 332: 593–610 Arvan P, Kuliawat R, Prabakharan D, Zavacki A-M, Elahi D, Wang S, Pilkey D (1991) Protein discharge from immature secretory granules displays both regulated and constitutive characteristics. J Biol Chem 266: 14171–14174 Arvan P, Zhang BY, Feng L, Liu M, Kuliawat R (2002) Lumenal protein multimerization in the distal secretory pathway/secretory granules. Curr Opin Cell Biol 14(4): 448–453 Arvan P, Halban PA (2004) Sorting ourselves out: seeking consensus on trafficking in the beta-cell. Traffic 5: 53–61 Austin CD, Shields D (1996) Prosomatostatin processing in permeabilized cells. Calcium is required for prohormone cleavage but not formation of nascent secretory vesicles. J Biol Chem 271: 1194–1199 Avery J, Ellis D, Lang T, Holroyd P, Riedel D, Henderson R, Edwardson M, Jahn R (2000) A cell-free system for regulated exocytosis in PC12 cells. J Cell Biol 148: 317–324 Banerjee A, Barry VA, Bibhuti RD, Martin TFJ (1996) N-ethylmaleimide-sensitive factor acts at a prefusion ATP-dependent step in Ca2þ -activated exocytosis. J Biol Chem 271: 20223–20226 Barral DC, Ramalho JS, Anders R, Hume AN, Knapton HJ, Tolmachova T, Collinson LM, Goulding D, Authi KS, Seabra MC (2002) Functional redundancy of Rab27 proteins and the pathogenesis of Griscelli syndrome. J Clin Invest 110: 247–257 Bauerfeind R, Huttner WB (1993) Biogenesis of constitutive secretory vesicles, secretory granules and synaptic vesicles. Curr Opin Cell Biol 5(4): 628–635 Bell-Parikh LC, Eipper BA, Mains RE (2001) Response of an integral granule membrane protein to changes in pH. J Biol Chem 276: 29854–29863 Bendayan M, Roth J, Perrelet A, Orci L (1980) Quantitative immunocytochemical localization of pancreatic secretory proteins in subcellular compartments of the rat acinar cell. J Histochem Cytochem 28: 149–160 Bendayan M (1982) Double immunocytochemical labeling applying the protein A-gold technique. J Histochem Cytochem 30: 81–85 Beuret N, Stettler H, Renold A, Rutishauser J, Spiess M (2004) Expression of regulated secretory proteins is sufficient to generate granule-like structures in constitutively secreting cells. J Biol Chem 279(19): 20242–20249 Blazquez M, Thiele C, Huttner WB, Docherty K, Shennan KIJ (2000) Involvement of the membrane lipid bilayer in sorting prohormone convertase 2 into the regulated secretory pathway. Biochem J 349: 843–852 Blazquez M, Docherty K, Shennan KI (2001) Association of prohormone convertase 3 with membrane lipid rafts. J Mol Endocrinol 27(1): 107–116 Bock JB, Klumperman J, Davanger S, Scheller RH (1997) Syntaxin 6 functions in transGolgi network vesicle trafficking. Mol Biol Cell 8: 1261–1271 Borgonovo B, Ouwendijk J, Solimena M (2006) Biogenesis of secretory granules. Curr Opin Cell Biol 18: 365–370 Bowman GR, Smith DG, Siu KWM, Pearlman RE, Turkewitz AP (2005) Genomic and proteomic evidence for a second family of dense core granule cargo proteins in Tetrahymena thermophila. J Eukaryot Microbiol 52: 291–297

508

*

A. A. Mironov and P. Arvan

Brakch N, Allemandou F, Cavadas C, Grouzmann E, Brunner HR (2002) Dibasic cleavage site is required for sorting to the regulated secretory pathway for both pro- and neuropeptide Y. J Neurochem 81: 1166–1175 Carroll RJ, Hammer RE, Chan SJ, Swift HH, Rubenstein AH, Steiner DF (1988) A mutant human proinsulin is secreted from islets of Langerhans in increased amounts via an unregulated pathway. Proc Natl Acad Sci USA 85: 8943–8947 Castle JD, Castle AM (1996) Two regulated secretory pathways for newly synthesized parotid salivary proteins are distinguished by doses of secretagogues. J Cell Sci 109: 2591–2599 Cawley NX, Rodriguez YM, Maldonado A, Loh YP (2003) Trafficking of mutant carboxypeptidase E to secretory granules in a {beta}-cell line derived from Cpefat/ Cpefat Mice. Endocrinology 144: 292–298 Cawley NX, Zhou J, Hill JM, Abebe D, Romboz S, Yanik T, Rodriguiz RM, Wetsel WC, Loh YP (2004) The carboxypeptidase E knockout mouse exhibits endocrinological and behavioral deficits. Endocrinology 145: 5807–5819 Chabin-Brion K, Marceiller J, Perez F, Settegrana C, Drechou A, Durand G, Pous C (2001) The Golgi complex is a microtubule-organizing organelle. Mol Biol Cell 12: 2047–2060 Chamberlain LH, Roth D, Morgan A, Burgoyne RD (1995) Distinct effects of alpha-SNAP, 14-3-3 proteins, and calmodulin on priming and triggering of regulated exocytosis. J Cell Biol 130(5): 1063–1070 Chamberlain LH, Henry J, Burgoyne RD (1996) Cysteine string proteins are associated with chromaffin granules. J Biol Chem 271(32): 19514–19517 Chan SJ, Seino S, Gruppuso PA, Schwartz R, Steiner DF (1987) A mutation in the B chain coding region is associated with impaired proinsulin conversion in a family with hyperproinsulinemia. Proc Natl Acad Sci USA 84: 2194–2197 Chanat E, Huttner WB (1991) Milieu-induced, selective aggregation of regulated secretory proteins in the trans-Golgi network. J Cell Biol 115(6): 1505–1519 Chavez RA, Miller SG, Moore HP (1996) A biosynthetic regulated secretory pathway in constitutive secretory cells. J Cell Biol 133(6): 1177–1191 Clermont Y, Rambourg A, Hermo L (1992) Segregation of secretory material in all elements of the Golgi apparatus in principal epithelial cells of the rat seminal vesicle. Anat Rec 232: 349–358 Clermont Y, Xia L, Rambourg A, Turner JD, Hermo L (1993) Transport of casein submicelles and formation of secretion granules in the Golgi apparatus of epithelial cells of the lactating mammary gland of rat. Anat Rec 235: 363–373 Clermont Y, Rambourg A, Hermo L (1995) Trans-Golgi network (TGN) of different cell types: three-dimensional structural characteristics and variability. Anat Rec 242(3): 289–301 Colomer V, Kicska GA, Rindler MJ (1996) Secretory granule content proteins and the luminal domains of granule membrane proteins aggregate in vitro at mildly acidic pH. J Biol Chem 271(1): 48–55 Cool DR, Fenger M, Snell CR, Loh YP (1995) Identification of the sorting signal motif within pro-opiomelanocortin for the regulated secretory pathway. J Biol Chem 270 (15): 8723–8729 Cool DR, Normant E, Shen FS, Chen HC, Pannel L, Zhang Y, Loh YP (1997) Carboxypeptidase E is a regulated secretory pathway sorting receptor: genetic obliteration leads to endocrine disorders in Cpefat mice. Cell 88: 73–83 Coorssen JR, Blank PS, Tahara M, Zimmerberg J (1998) Biochemical and functional studies of cortical vesicle fusion: the SNARE complex and Ca2þ sensitivity. J Cell Biol 143(7): 1845–1857 Cowan AT, Bowman GR, Edwards KF, Emerson JJ, Turkewitz AP (2005) Genetic, genomic, and functional analysis of the granule lattice proteins in Tetrahymena secretory granules. Mol Biol Cell 16: 4046–4060

Origins of the regulated secretory pathway

*

509

Cowley DJ, Moore YR, Darling DS, Joyce PB, Gorr SU (2000) N- and C-terminal domains direct cell type-specific sorting of chromogranin A to secretory granules. J Biol Chem 275: 7743–7748 Dannies PS (1999) Protein hormone storage in secretory granules: mechanisms for concentration and sorting. Endocr Rev 20(1): 3–21 Dartsch H, Kleene R, Kern HF (1998) In vitro condensation-sorting of enzyme proteins isolated from rat pancreatic acinar cells. Eur J Cell Biol 75: 211–222 Day R, Gorr SU (2003) Secretory granule biogenesis and chromogranin A: master gene, on/off switch or assembly factor? Trends Endocrinol Metab 14(2): 10–13 De Lisle RC (2002) Role of sulfated O-linked glycoproteins in zymogen granule formation. J Cell Sci 115: 2941–2952 De Matteis MA, Luini A (2008) Exiting the Golgi complex. Nat Rev Mol Cell Biol 9(4): 273–284 DeCamili P, Jahn R (1990) Pathways to regulated exocytosis in neurons. Annu Rev Physiol 52: 625–645 Dhanvantari S, Loh YP (2000) Lipid raft association of carboxypeptidase E is necessary for its function as a regulated secretory pathway sorting receptor. J Biol Chem 275: 29887–29893 Dhanvantari S, Shen FS, Adams T, Snell CR, Zhang C, Mackin RB, Morris SJ, Loh YP (2003) Disruption of a receptor-mediated mechanism for intracellular sorting of proinsulin in familial hyperproinsulinemia. Mol Endocrinol 17: 1856–1867 Dikeakos JD, Lacombe MJ, Mercure C, Mireuta M, Reudelhuber TL (2007a) A hydrophobic patch in a charged alpha-helix is sufficient to target proteins to dense core secretory granules. J Biol Chem 282: 1136–1143 Dikeakos JD, Mercure C, Lacombe MJ, Seidah NG, Reudelhuber TL (2007b) PC1/3, PC2 and PC5/6A are targeted to dense core secretory granules by a common mechanism. FEBS J 274: 4094–4102 Dittie AS, Hajibagheri N, Tooze SA (1996) The AP-1 adaptor complex binds to immature secretory granules from PC12 cells, and is regulated by ADP-ribosylation factor. J Cell Biol 132(4): 523–536 Dittie AS, Thomas L, Thomas G, Tooze SA (1997) Furin is sorted into the regulated secretory pathway in neuroendocrine cells, interacts with the AP-1 complex, and is removed during granule maturation by a casein kinase II dependent mechanism. EMBO J 16: 4859–4870 Docherty K, Hutton JC, Steiner DF (1984) Cathepsin B-related proteases in the insulin secretory granule. J Biol Chem 259: 6041–6044 Duncan RR, Greaves J, Wiegand UK, Matskevich I, Bodammer G, Apps DK, Shipston MJ, Chow RH (2003) Functional and spatial segregation of secretory vesicle pools according to vesicle age. Nature 422(6928): 176–180 Eaton BA, Haugwitz M, Lau D, Moore HP (2000) Biogenesis of regulated exocytotic carriers in neuroendocrine cells. J Neurosci 20: 7334–7344 El Meskini R, Jin L, Marx R, Bruzzaniti A, Lee J, Emeson R, Mains R (2001) A signal sequence is sufficient for green fluorescent protein to be routed to regulated secretory granules. Endocrinology 142: 864–873 Emmanouilidou E, Teschemacher A, Pouli AE, Nicholls LI, Seward EP, Rutter GA (1999) Imaging Ca2þ concentration changes at the secretory vesicle surface with a recombinant targeted cameleon. Curr Biol 9(16): 915–918 Farquhar MG, Reid JJ, Daniell LW (1978) Intracellular transport and packaging of prolactin: a quantitative electron microscope autoradiographic study of mammotrophs dissociated from rat pituitaries. Endocrinology 102: 296–311 Feliciangeli S, Kitabgi P, Bidard JN (2001) The role of dibasic residues in prohormone sorting to the regulated secretory pathway. A study with proneurotensin. J Biol Chem 276: 6140–6150

510

*

A. A. Mironov and P. Arvan

Feliciangeli S, Kitabgi P (2002) Insertion of dibasic residues directs a constitutive protein to the regulated secretory pathway. Biochem Biophys Res Commun 290: 191–196 Feng L, Arvan P (2003) The trafficking of alpha 1-antitrypsin, a post-Golgi secretory pathway marker, in INS-1 pancreatic beta cells. J Biol Chem 278: 31486–31494 Fontijn RD, Goud B, Echard A, Jollivet F, Van Marle J, Pannekoek H, Horrevoets AJ (2001) The human kinesin-like protein RB6K is under tight cell cycle control and is essential for cytokinesis. Mol Cell Biol 21: 2944–2955 Fredman P, Mansson JE, Rynmark BM, Josefsen K, Ekblond A, Halldner L, Osterbye T, Horn T, Buschard K (2000) The glycosphingolipid sulfatide in the islets of Langerhans in rat pancreas is processed through recycling: possible involvement in insulin trafficking. Glycobiology 10: 39–50 Fumagalli G, Zanini A (1985) In cow anterior pituitary, growth hormone and prolactin can be packed in separate granules of the same cell. J Cell Biol 100: 2019–2024 Galas MC, Chasserot-Golaz S, Dirrig-Grosch S, Bader MF (2000) Presence of dynamin– syntaxin complexes associated with secretory granules in adrenal chromaffin cells. J Neurochem 75: 1511–1519 Garcia AL, Han SK, Janssen WG, Khaing ZZ, Ito T, Glucksman MJ, Benson DL, Salton SR (2005) A prohormone convertase cleavage site within a predicted alpha-helix mediates sorting of the neuronal and endocrine polypeptide VGF into the regulated secretory pathway. J Biol Chem 280: 41595–41608 Geuze HJ, Slot JW (1980) Disproportional immunostaining patterns of two secretory proteins in guinea pig and rat exocrine pancreatic cells. An immunoferritin and fluorescence study. Eur J Cell Biol 21: 93–100 Geuze HJ, Slot JW, Tokuyasu KT (1979) Immunocytochemical localization of amylase and chymotrypsinogen in the exocrine pancreatic cell with special attention to the Golgi complex. J Cell Biol 82: 697–707 € mer, A, Salm T, Huttner WB, Gerdes HH (1999) The disulfide-bonded Glombik MM, Kro loop of chromogranin B mediates membrane binding and directs sorting from the trans-Golgi network to secretory granules. EMBO J 18(4): 1059–1070 Glombik MM, Gerdes HH (2000) Signal-mediated sorting of neuropeptides and prohormones: secretory granule biogenesis revisited. Biochimie 82: 315–326 Gomi H, Mori K, Itohara S, Izumi T (2007) Rab27b is expressed in a wide range of exocytic cells and involved in the delivery of secretory granules near the plasma membrane. Mol Biol Cell 18: 4377–4386 Gorr SU (1996) Differential storage of prolactin, granins (chromogranin B and secretogranin II), and constitutive secretory markers in rat pituitary GH4C1 cells. J Biol Chem 271: 3575–3580 Gorr SU, Jain RK, Kuehn U, Joyce PB, Cowley DJ (2001) Comparative sorting of neuroendocrine secretory proteins: a search for common ground in a mosaic of sorting models and mechanisms. Mol Cell Endocrinol 172: 1–6 Greider MH, Howell SL, Lacy PE (1969) Isolation and properties of secretory granules from rat islets of Langerhans. II. Ultrastructure of the beta granule. J Cell Biol 41: 162–166 Gross DJ, Halban PA, Kahn CR, Weir GC, Villa-Komaroff L (1989) Partial diversion of a mutant proinsulin (B10 aspartic acid) from the regulated to the constitutive secretory pathway in transfected AtT-20 cells. Proc Natl Acad Sci USA 86(11): 4107–4111 Halban PA, Irminger JC (1994) Sorting and processing of secretory proteins. Biochem J 299(Pt 1): 1–18 Halban PA, Irminger JC (2003) Mutant proinsulin that cannot be converted is secreted efficiently from primary rat beta-cells via the regulated pathway. Mol Biol Cell 14: 1195–1203

Origins of the regulated secretory pathway

*

511

Handley MT, Haynes LP, Burgoyne RD (2007) Differential dynamics of Rab3A and Rab27A on secretory granules. J Cell Sci 120: 973–984 Hannah MJ, Schmidt AA, Huttner WB (1999) Synaptic vesicle biogenesis. Annu Rev Cell Dev Biol 15: 733–798 Hansen GH, Nies-Christiansen L-L, Thorsen E, Immerdal L, Danielsen EM (2000) Cholesterol depletion of enterocytes. Effect on the Golgi complex and apical membrane trafficking. J Biol Chem 275(7): 5136–5142 Harrison TM, Chidgey MA, Uff S (1996) Novel markers for constitutive secretion used to show that tissue plasminogen activator is sorted to the regulated pathway in transfected PC12 cells. Cell Biol Int 20: 293–299 Hashimoto S, Fumagalli G, Zanini A, Meldolesi J (1987) Sorting of three secretory proteins to distinct secretory granules in acidophilic cells of cow anterior pituitary. J Cell Biol 105(4): 1579–1586 Hendy GN, Li T, Girard M, Feldstein RC, Mulay S, Desjardins R, Day R, Karaplis AC, Tremblay ML, Canaff L (2006) Targeted ablation of the chromogranin a (Chga) gene: normal neuroendocrine dense-core secretory granules and increased expression of other granins. Mol Endocrinol 20(8): 1935–1947 Henomatsu N, Yoshimori T, Yamamoto A, Moriyama Y, Tashiro Y (1993) Inhibition of intracellular transport of newly synthesized prolactin by bafilomycin A1 in a pituitary tumor cell line, GH3 cells. Eur J Cell Biol 62(1): 127–139 Hinners I, Wendler F, Fei H, Thomas L, Thomas G, Tooze SA (2003) AP-1 recruitment to VAMP4 is modulated by phosphorylation-dependent binding of PACS-1. EMBO Rep 4: 1–8 Honore B, Vorum H (2000) The CREC family, a novel family of multiple EF-hand, lowaffinity Ca2þ -binding proteins localised to the secretory pathway of mammalian cells. FEBS Lett 466(1): 11–18 Hosaka M, Watanabe T, Sakai Y, Uchiyama Y, Takeuchi T (2002) dentification of a chromogranin A domain that mediates binding to secretogranin III and targeting to secretory granules in pituitary cells and pancreatic beta-cells. Mol Biol Cell 13(10): 3388–3399 Hosaka M, Suda M, Sakai Y, Izumi T, Watanabe T, Takeuchi T (2004) Secretogranin III binds to cholesterol in the secretory granule membrane as an adapter for chromogranin A. J Biol Chem 279: 3627–3634 Hosaka M, Watanabe T, Sakai Y, Kato T, Takeuchi T (2005) Interaction between secretogranin III and carboxypeptidase E facilitates prohormone sorting within secretory granules. J Cell Sci 118: 4785–4795 Huang IF, Arvan P (1994) Formation of the insulin-containing secretory granule core occurs within immature beta-granules. J Biol Chem 269: 20838–20844 Huang XF, Arvan P (1995) Intracellular transport of proinsulin in pancreatic beta-cells. Structural maturation probed by disulfide accessibility. J Biol Chem 270(35): 20417–20423 Huang AY, Castle AM, Hinton BT, Castle JD (2001) Resting (basal) secretion of proteins is provided by the minor regulated and constitutive-like pathways and not granule exocytosis in parotid acinar cells. J Biol Chem 276: 22296–22306 Huh YH, Jeon SH, Yoo SH (2003) Chromogranin B-induced secretory granule biogenesis: comparison with the similar role of chromogranin A. J Biol Chem 278(42): 40581–40589 € nnichsen B, Shima DT, Nilsson T, Warren G (1997) An isoform of Hui N, Nakamura N, So the Golgi t-SNARE, syntaxin 5, with an endoplasmic reticulum retrieval signal. Mol Biol Cell 8(9): 1777–1787 Ikonen E (2001) Roles of lipid rafts in membrane transport. Curr Opin Cell Biol 13(4): 470–477 Imai A, Yoshie S, Nashida T, Shimomura H, Fukuda M (2004) The small GTPase Rab27B regulates amylase release from rat parotid acinar cells. J Cell Sci 117: 1945–1953

512

*

A. A. Mironov and P. Arvan

Irminger JC, Verchere CB, Meyer K, Halban PA (1997) Proinsulin targeting to the regulated pathway is not impaired in carboxypeptidase E-deficient Cpefat/Cpefat mice. J Biol Chem 272(44): 27532–27534 Jain RK, Joyce PB, Gorr SU (2000) Aggregation chaperones enhance aggregation and storage of secretory proteins in endocrine cells. J Biol Chem 275(35): 27032–27036 Jain RK, Joyce PB, Molinete M, Halban PA, Gorr SU (2001) Oligomerization of green fluorescent protein in the secretory pathway of endocrine cells. Biochem J 360: 645–649 Johns LM, Levitan ES, Shelden EA, Holz RW, Axelrod D (2001) Restriction of secretory granule motion near the plasma membrane of chromaffin cells. J Cell Biol 153: 177–190 Kaether C, Salm T, Glombik M, Almers W, Gerdes H-H (1997) Targeting of green fluorescent protein to neuroendocrine secretory granules: a new tool for real time studies of regulated protein secretion. Eur J Cell Biol 74: 133–142 Kakhlon O, Sakya P, Larijani B, Watson R, Tooze SA (2006) GGA function is required for maturation of neuroendocrine secretory granules. EMBO J 25: 1590–1602 Kasai K, Ohara-Imaizumi M, Takahashi N, Mizutani S, Zhao S, Kikuta T, Kasai H, Nagamatsu S, Gomi H, Izumi T (2005) Rab27a mediates the tight docking of insulin granules onto the plasma membrane during glucose stimulation. J Clin Invest 115: 388–396 Kelly RB, Buckley KM, Burgess TL, Carlson SS, Caroni P, Hooper JE, Katzen A, Moore HP, Pfeffer SR, Schroer TA (1983) Membrane traffic in neurons and peptide-secreting cells. Cold Spring Harbor Symp Quant Biol 48(Pt2): 697–705 Kim T, Tao-Cheng JH, Eiden LE, Loh YP (2001) Chromogranin A, an “on/off” switch controlling dense-core secretory granule biogenesis. Cell 106(4): 499–509 Kim T, Tao-Cheng JH, Eiden LE, Loh YP (2003) The role of chromogranin A and the control of secretory granule genesis and maturation. Trends Endocrinol Metab 14 (2): 56–57 Kingsley DM, Rinchik EM, Russell LB, Ottiger H-P, Sutcliffe JG, Copeland NG, Jenkins NA (1990) Genetic ablation of a mouse gene expressed specifically in brain. EMBO J 9(2): 395–399 Kleene R, Kastner B, Rosser R, Kern H (1999) Complex formation among rat pancreatic secretory proteins under mild alkaline pH conditions. Digestion 60(4): 305–313 Klumperman J, Spijker S, Van Minnen J, Sharp-Baker H, Smit AB, Geraerts WPM (1996) Cell type-specific sorting of neuropeptides: a mechanism to modulate peptide composition of large dense-core vesicles. J Neurosci 16(24): 7930–7940 Klumperman J, Kuliawat R, Griffith JM, Geuze HJ, Arvan P (1998) Mannose 6-phosphate receptors are sorted from immature secretory granules via AP-1, clathrin, and syntaxin 6-positive vesicles. J Cell Biol 141: 359–371 € mer A, Glombik MM, Huttner WB, Gerdes HH (1998) Essential role of the disulfideKro bonded loop of chromogranin B for sorting to secretory granules is revealed by expression of a deletion mutant in the absence of endogenous granin synthesis. J Cell Biol 140(6): 1331–1346 Kuliawat R, Arvan P (1992) Protein targeting via the ‘constitutive-like’ secretory pathway in isolated pancreatic islets: passive sorting in the immature granule compartment. J Cell Biol 118: 521–529 Kuliawat R, Arvan P (1994) Distinct molecular mechanisms for protein sorting within immature secretory granules of pancreatic b-cells. J Cell Biol 126: 77–86 Kuliawat R, Klumperman J, Ludwig T, Arvan P (1997) Differential sorting of lysosomal enzymes out of the regulated secretory pathway in pancreatic beta-cells. J Cell Biol 137: 595–608 Kuliawat R, Prabakaran D, Arvan P (2000) Proinsulin endoproteolysis confers enhanced targeting of processed insulin to the regulated secretory pathway. Mol Biol Cell 11: 1959–1972

Origins of the regulated secretory pathway

*

513

Kuliawat R, Kalinina E, Bock J, Fricker L, McGraw TE, Kim SR, Zhong J, Scheller R, Arvan P (2004) Syntaxin-6 SNARE involvement in secretory and endocytic pathways of cultured pancreatic b–cells. Mol Biol Cell 15(4): 1690–1701 Lacombe MJ, Mercure C, Dikeakos JD, Reudelhuber TL (2005) Modulation of secretory granule-targeting efficiency by cis and trans compounding of sorting signals. J Biol Chem 280: 4803–4807 Ladinsky MS, Wu CC, McIntosh S, McIntosh JR, Howell K (2002) Structure of the Golgi and distribution of reporter molecules at 20 C reveals the complexity of the exit compartments. Mol Biol Cell 13: 2810–2825 Lane J, Allan V (1998) Microtubule-based membrane movement. Biochim Biophys Acta 1376(1): 27–55 Laine J, Lebel D (1999) Efficient binding of regulated secretory protein aggregates to membrane phospholipids at acidic pH. Biochem J 338: 289–294 Lara-Lemus R, Liu M, Turner MD, Scherer P, Stenbeck G, Iyengar P, Arvan P (2006) Lumenal protein sorting to the constitutive secretory pathway of a regulated secretory cell. J Cell Sci 119: 1833–1842 Lee MS, Zhu YL, Chang JE, Dannies PS (2001) Acquisition of Lubrol insolubility, a common step for growth hormone and prolactin in the secretory pathway of neuroendocrine cells. J Biol Chem 276: 715–721 Lee CS, Kim IS, Park JB, Lee MN, Lee HY, Suh PG, Ryu SH (2006) The phox homology domain of phospholipase D activates dynamin GTPase activity and accelerates EGFR endocytosis. Nat Cell Biol 8(5): 477–484 Leung YM, Sheu L, Kwan E, Wang G, Tsushima R, Gaisano H (2002) Visualization of sequential exocytosis in rat pancreatic islet beta cells. Biochem Biophys Res Commun 292(9): 980–986 Liu M, Ramos-Castaneda J, Arvan P (2003) Role of the connecting peptide in insulin biosynthesis. J Biol Chem 278: 14798–14805 Lochner JE, Kingma M, Kuhn S, Meliza CD, Cutler B, Scalettar BA (1998) Real-time imaging of the axonal transport of granules containing a tissue plasminogen activator/green fluorescent protein hybrid. Mol Biol Cell 9: 2463–2476 Lou H, Kim SK, Zaitsev E, Snell CR, Lu B, Loh YP (2005) Sorting and activity-dependent secretion of BDNF require interaction of a specific motif with the sorting receptor carboxypeptidase E. Neuron 45: 245–255 Lou H, Smith AM, Coates LC, Cawley NX, Loh YP, Birch NP (2007) The transmembrane domain of the prohormone convertase PC3: a key motif for targeting to the regulated secretory pathway. Mol Cell Endocrinol 267: 17–25 Lui-Roberts WW, Collinson LM, Hewlett LJ, Michaux G, Cutler DF (2005) An AP-1/clathrin coat plays a novel and essential role in forming the Weibel–Palade bodies of endothelial cells. J Cell Biol 170: 627–636 Martin TFJ, Kowalchyk JA (1997) Docked secretory vesicles undergo Ca2 þ -activated exocytosis in a cell-free system. J Biol Chem 272: 14447–14453 Martin TF (2003) Tuning exocytosis for speed: fast and slow modes. Biochim Biophys Acta 1641: 157–165  rguez JA, Geuze HJ, Slot JW, Klumperman J (1999) Vesicular tubular Martínez-Mena clusters (VTCs) between ER and Golgi mediate concentration of soluble secretory proteins by exclusion from COPI-coated vesicles. Cell 98: 81–90 Matsumoto M, Miki T, Shibasaki T, Kawaguchi M, Shinozaki H, Nio J, Saraya A, Koseki H, Miyazaki M, Iwanaga T, Seino S (2004) Noc2 is essential in normal regulation of exocytosis in endocrine and exocrine cells. Proc Natl Acad Sci USA 101: 8313–8318 Meldolesi J, Chieregatti E, Luisa Malosio ML (2004) Requirements for the identification of dense-core granules. Trends Cell Biol 14: 13–19 Michaux G, Hewlett LJ, Messenger SL, Goodeve AC, Peake IR, Daly ME, Cutler DF (2003) Analysis of intracellular storage and regulated secretion of 3 von Willebrand disease-causing variants of von Willebrand factor. Blood 102: 2452–2458

514

*

A. A. Mironov and P. Arvan

Millar CA, Meerloo T, Martin S, Hickson GRX, Shimwell NJ, Wakelam MJO, James DE, Gould GW (2000) Adipsin and the glucose transporter GLUT4 traffic to the cell surface via independent pathways in adipocytes. Traffic 1(2): 141–151 Mizuno K, Tolmachova T, Ushakov DS, Romao M, Abrink M, Ferenczi MA, Raposo G, Seabra MC (2007) Rab27b regulates mast cell granule dynamics and secretion. Traffic 8: 883–892 Molinete M, Lilla V, Jain R, Joyce PBM, Gorr S-U, Ravazzola M, Halban PQ (2000) Trafficking of non-regulated secretory proteins in insulin secreting (INS-1) cells. Diabetologia 43: 1157–1164 Molinete M, Dupuis S, Brodsky FM, Halban PA (2001) Role of clathrin in the regulated secretory pathway of pancreatic beta-cells. J Cell Sci 114: 3059–3066 Mulcahy LR, Vaslet CA, Nillni EA (2005) Prohormone-convertase 1 processing enhances post-Golgi sorting of prothyrotropin-releasing hormone-derived peptides. J Biol Chem 280: 39818–39826 Natori S, Huttner WB (1996) Chromogranin B (secretogranin I) promotes sorting to the regulated secretory pathway of processing intermediates derived from a peptide hormone precursor. Proc Natl Acad Sci USA 93(9): 4431–4436 Neerman-Arbez M, Halban PA (1993) Novel, non-crinophagic, degradation of connecting peptide in transformed pancreatic beta cells. J Biol Chem 268(22): 16248–16252 Nemoto Y, Wenk MR, Watanabe M, Daniell L, Murakami T, Ringstad N, Yamada H, Takei K, De Camilli P (2001) Identification and characterization of a synaptojanin 2 splice isoform predominantly expressed in nerve terminals. J Biol Chem 276(44): 41133–41142 Nguyen TT, Ong H, De Lean A (1988) Secretion and biosynthesis of atrial natriuretic factor by cultured adrenal chromaffin cells. FEBS Lett 231: 393–396 Normant E, Loh YP (1998) Depletion of carboxypeptidase E, a regulated secretory pathway sorting receptor, causes misrouting and constitutive secretion of proinsulin and proenkephalin, but not chromogranin A. Endocrinology 139(4): 2137–2145 Oliver PM, Fox JE, Kim R, Rockman HA, Kim HS, Reddick RL, Pandey KN, Milgram SL, Smithies O, Maeda N (1997) Hypertension, cardiac hypertrophy, and sudden death in mice lacking natriuretic peptide receptor A. Proc Natl Acad Sci USA 94(26): 14730–14735 Oprins A, Rabouille C, Posthuma G, Klumperman J, Geuze HJ, Slot JW (2001) The ER to Golgi interface is the major concentration site of secretory proteins in the exocrine pancreatic cell. Traffic 2: 831–838 Orci L, Ravazzola M, Perrelet A (1984a) Proinsulin associates with Golgi membranes of pancreatic B cells. Proc Natl Acad Sci USA 81: 6743–6746 Orci L, Halban P, Amherdt M, Ravazzola M, Vassalli J-D, Perrelet A (1984b) A clathrincoated, Golgi-related compartment of the insulin secreting cell accumulates proinsulin in the presence of monensin. Cell 39: 39–47 Orci L (1985) The insulin factory. a tour of the plant surroundings and a visit to the assembly line. Diabetologia 28: 528–546 Orci L, Ravazzola M, Amherdt M, Madsen O, Perrelet A, Vassalli JD, Anderson RGW (1986) Conversion of proinsulin to insulin occurs coordinately with acidification of maturing secretory vesicles. J Cell Biol 103: 2273–2281 Orci L, Ravazzola M, Storch MJ, Anderson RG, Vassalli JD, Perrelet A (1987a) Proteolytic maturation of insulin is a post-Golgi event which occurs in acidifying clathrin-coated secretory vesicles. Cell 49(6): 865–868 Orci L, Ravazzola M, Amherdt M, Perrelet A, Powell SK, Quinn DL, Moore H-PH (1987b) The trans-most cisternae of the Golgi complex: a compartment for sorting of secretory and plasma membrane proteins. Cell 51(6): 1039–1051 Orci L, Stamnes M, Ravazzola M, Amherdt M, Perrelet A, Sollner TH, Rothman JE (1997) Bidirectional transport by distinct populations of COPI-coated vesicles. Cell 90(2): 335–349

Origins of the regulated secretory pathway

*

515

Osterbye T, Jorgensen KH, Fredman P, Tranum-Jensen J, Kaas A, Brange J, Whittingham JL, Buschard K (2001) Sulfatide promotes the folding of proinsulin, preserves insulin crystals, and mediates its monomerization. Glycobiology 11: 473–479 Polishchuk RS, Mironov AA (2004) Structural aspects of Golgi function. Cell Mol Life Sci 61(2): 146–158 Posthuma G, Slot JW, Geuze HJ (1988) Immunogold determination of amylase concentrations in pancreatic subcellular compartments. Eur J Cell Biol 46: 327–335 Powell SK, Orci L, Craik CS, Moore HPH (1988) Efficient targeting to storage granules of human proinsulins with altered propeptide domain. J Cell Biol 106: 1843–1851 Rambourg A, Segretain D, Clermont Y (1984) Tridimensional architecture of the Golgi apparatus in the atrial muscle cell of the rat. Am J Anat 170: 163–179 Rambourg A, Clermont Y, Chretien M, Olivier L (1992) Formation of secretory granules in the Golgi apparatus of prolactin cells in the rat pituitary gland: a stereoscopic study. Anat Rec 232: 169–179  tien M, Olivier L (1993) Modulation of the Golgi Rambourg A, Clermont Y, Chre apparatus in stimulated and nonstimulated prolactin cells of female rats. Anat Rec 235(3): 353–362 Rambourg A, Jackson CL, Clermont Y (2001) Three dimensional configuration of the secretory pathway and segregation of secretion granules in the yeast Saccharomyces cerevisiae. J Cell Sci 114(Pt 12): 2231–2239 Regazzi R, Sadoul K, Meda P, Kelly RB, Halban PA, Wollheim CB (1996) Mutational analysis of VAMP domains implicated in Ca2 þ -induced insulin exocytosis. EMBO J 15: 6951–6959 Rhodes CJ, Halban PA (1987) Newly synthesized proinsulin/insulin and stored insulin are released from pancreatic B cells predominantly via a regulated, rather than constitutive, pathway. J Cell Biol 105: 145–153 Rindler MJ (1998) Carboxypeptidase E, a peripheral membrane protein implicated in the targeting of hormones to secretory granules, co-aggregates with granule content proteins at acidic pH. J Biol Chem 273(47): 31180–31185 Rindler MJ, Colomer V, Jin Y (2001) Immature granules are not major sites for segregation of constitutively secreted granule content proteins in NIT-1 insulinoma cells. Biochem Biophys Res Commun 288: 1071–1077 Robinson LJ, Martin TFJ (1998) Docking and fusion in neurosecretion. Curr Opin Cell Biol 10: 483–492 Rudolf R, Salm T, Rustom A, Gerdes HH (2001) Dynamics of immature secretory granules: role of cytoskeletal elements during transport, cortical restriction, and F-actindependent tethering. Mol Biol Cell 12(5): 1353–1365 Rustom A, Bajohrs M, Kaether C, Keller P, Toomre D, Corbeil D, Gerdes HH (2002) Selective delivery of secretory cargo in Golgi-derived carriers of nonepithelial cells. Traffic 3: 279–288 Rutter GA, Tsuboi T (2004) Kiss and run exocytosis of dense core secretory vesicles. Neuroreport 15(1): 79–81 Salamero J, Sztul E, Howell K (1990) Exocytic transport vesicles generated in vitro from the trans-Golgi network carry secretory and plasma membrane proteins. Proc Natl Acad Sci USA 87(19): 7717–7721 Schlegel A, Arvan P, Lisanti MP (2001) Caveolin-1 binding to endoplasmic reticulum (ER) membranes and entry into the regulated secretory pathway are regulated by serine phosphorylation. Protein sorting at the level of the ER. J Biol Chem 276: 4398–4408 Schmid SL (1997) Clathrin-coated vesicle formation and protein sorting: an integrated process. Annu Rev Biochem 66: 511–548 Scherer PE, Lederkremer GZ, Williams S, Fogliano M, Baldini M, Lodish HF (1996) Cab45, a novel (Ca2þ )-binding protein localized to the Golgi lumen. J Cell Biol 133(2): 257–268

516

*

A. A. Mironov and P. Arvan

Schoonderwoert VT, Holthuis JC, Tanaka S, Tooze SA, Martens GJ (2000) Inhibition of the vacuolar H þ -ATPase perturbs the transport, sorting, processing and release of regulated secretory proteins. Eur J Biochem 267(17): 5646–5654 Seeger M, Payne G (1992) Selective and immediate effects of clathrin heavy chain mutations on Golgi membrane protein retention in Saccharomyces cerevisiae. J Cell Biol 118(3): 531–540 Sesso A, Assis JE, Kuwajima VY, Kachar B (1980) Freeze-fracture and thin-section study of condensing vacuoles in rat pancreatic acinar cells. Acta Anat 108: 521–539 Shennan KI (1996) Intracellular targeting of secretory proteins in neuroendocrine cells. Biochem Soc Trans 24: 535–539 Slot JW, Geuze HJ (1983) Immunoelectron microscopic exploration of the Golgi complex. J Histochem Cytochem 31: 1049–1056 Slot JW, Garutti G, Martin S, Oorschot V, Posthuma G, Kraegen EW, Laybutt R, Thibault G, James DE (1997) Glucose transporter (GLUT-4) is targeted to secretory granules in rat atrial cardiomyocytes. J Cell Biol 137: 1243–1254 Sollner TH (2003) Regulated exocytosis and SNARE function. Mol Membr Biol 20: 209–220 Song L, Fricker LD (1995) Calcium- and pH-dependent aggregation of carboxypeptidase E. J Biol Chem 270(14): 7963–7967 Sørensen JB (2004) Formation, stabilisation and fusion of the readily releasable pool of secretory vesicles. Pflugers Arch 448(4): 347–362 Sossin WS, Fisher JM, Scheller RH (1990) Sorting within the regulated secretory pathway occurs in the trans-Golgi network. J Cell Biol 110: 1–12 Soulet F, Yarar D, Leonard M, Schmid SL (2005) SNX9 regulates dynamin assembly and is required for efficient clathrin-mediated endocytosis. Mol Biol Cell 16(4): 2058–2067 Stahl LE, Wright RL, Castle JD, Castle AM (1996) The unique proline-rich domain of parotid proline-rich proteins functions in secretory sorting. J Cell Sci 109: 1637–1645 Steegmaier M, Klumperman J, Foletti DL, Yoo JS, Scheller RH (1999) Vesicle-associated membrane protein 4 is implicated in trans-Golgi network vesicle trafficking. Mol Biol Cell 10: 1957–1972 Steiner DF (1998) The proprotein convertases. Curr Opin Chem Biol 2(2): 31–39 Stojilkovic SS (2005) Ca2 þ -regulated exocytosis and SNARE function. Trends Endocrinol Metabol 16(3): 81–83 Stoller TJ, Shields D (1989) The propeptide of preprosomatostatin mediates intracellular transport and secretion of alpha-globin from mammalian cells. J Cell Biol 108(5): 1647–1655 Tamaki H, Yamashina S (2002) Structural integrity of the Golgi stack is essential for normal secretory functions of rat parotid acinar cells: effects of brefeldin A and okadaic acid. J Histochem Cytochem 50: 1611–1623 Tanaka S, Yora T, Nakayama K, Inoue K, Kurosumu K (1997) Proteolytic processing of pro-opiomelanocortin occurs in acidifying secretory granules of AtT-20 cells. J Histochem Cytochem 45: 425–436 Taraska JW, Perrais D, Ohara-Imaizumi M, Nagamatsu S, Almers W (2003) Secretory granules are recaptured largely intact after stimulated exocytosis in cultured endocrine cells. Proc Natl Acad Sci USA 100: 2070–2075 Taupenot L, Harper KL, Mahapatra NR, Parmer RJ, O’Connor DT (2002a) Intracellular protein trafficking into catecholamine storage vesicles: novel chimeric photoproteins visualized by deconvolution fluorescence microscopy. Ann NY Acad Sci 971: 262–265 Taupenot L, Harper KL, Mahapatra NR, Parmer RJ, Mahata SK, O’Connor DT (2002b) Identification of a novel sorting determinant for the regulated pathway in the secretory protein chromogranin A. J Cell Sci 115: 4827–4841

Origins of the regulated secretory pathway

*

517

Taupenot L, Harper KL, O’Connor DT (2003) The chromogranin–secretogranin family. N Engl J Med 348: 1134–1149 Taupenot L, Harper KL, O’Connor DT (2005) Role of H þ -ATPase-mediated acidification in sorting and release of the regulated secretory protein chromogranin A: evidence for a vesiculogenic function. J Biol Chem 280(5): 3885–3897 Terskikh A, Fradkov A, Ermakova G, Zaraisky A, Tan P, Kajava AV, Zhao X, Lukyanov S, Matz M, Kim S, Weissman I, Siebert P (2000) Fluorescent timer: protein that changes color with time. Science 290(5496): 1585–1588 Thiele C, Gerdes HH, Huttner WB (1997) Protein secretion: puzzling receptors. Curr Biol 7(8): R496–R500 Thorn P, Fogarty KE, Parker I (2004) Zymogen granule exocytosis is characterized by long fusion pore openings and preservation of vesicle lipid identity. Proc Natl Acad Sci USA 101(17): 6774–6779 Thrower EC, Choe CU, So SH, Jeon SH, Ehrlich BE, Yoo SH (2003) A functional interaction between chromogranin B and the inositol 1,4,5-trisphosphate receptor/Ca2þ channel. J Biol Chem 278(50): 49699–49706 Tolmachova T, Abrink M, Futter CE, Authi KS, Seabra MC (2007) Rab27b regulates number and secretion of platelet dense granules. Proc Natl Acad Sci USA 104(4): 5872–5877 Tooze J, Tooze SA (1986) Clathrin-coated vesicular transport of secretory proteins during the formation of ACTH-containing secretory granules in AtT20 cells. J Cell Biol 103: 839–850 Tooze J, Hollinshead M, Frank R, Burke B (1987) An antibody specific for an endoproteolytic cleavage site provides evidence that pro-opiomelanocortin is packaged into secretory granules in AtT20 cells before its cleavage. J Cell Biol 105: 155–162 Tooze J, Tooze SA, Fuller SD (1987) Sorting of progeny coronavirus from condensed secretory proteins at the exit from the trans-Golgi network of AtT2O cells. J Cell Biol 105: 1215–1226 Tooze SA, Huttner WB (1990) Cell-free protein sorting to the regulated and constitutive secretory pathways. Cell 60: 837–847 Tooze S, Flatmark T, Tooze J, Huttner WB (1991) Characterization of the immature secretory granule, an intermediate in granule biogenesis. J Cell Biol 115(6): 1491–1503 Tooze S, Martens GJM, Huttner WB (2001) Secretory granule biogenesis: rafting to the SNARE. Trends Cell Biol 11(3): 116–122 Trifaró J-M, Vitale ML (1993) Cytoskeleton dynamics during neurotransmitter release. Trends Neurobiol Sci 16: 466–472 Trucco A, Polishchuk RS, Martella O, Di Pentima A, Fusella A, Di Giandomenico D, San Pietro E, Beznoussenko GV, Polishchuk EV, Baldassarre M, Buccione R, Geerts WJ, Koster AJ, Burger KN, Mironov AA, Luini A (2004) Secretory traffic triggers the formation of tubular continuities across Golgi sub-compartments. Nat Cell Biol 6 (11): 1071–1081 Tsuboi T, Rutter GA (2003) Multiple forms of “kiss-and-run” exocytosis revealed by evanescent wave microscopy. Curr Biol 13(7): 563–567 Tsuboi T, McMahon HT, Rutter GA (2004) Mechanisms of dense core vesicle recapture following “kiss-and-run” (“cavicapture”) exocytosis in insulin-secreting cells. J Biol Chem 279(45): 47115–47124 Tsuboi T, Ravier MA, Parton LE, Rutter GA (2006) Sustained exposure to high glucose concentrations modifies glucose signaling and the mechanics of secretory vesicle fusion in primary rat pancreatic b-cells. Diabetes 55(4): 1057–1065 Tsuboi T, Fukuda M (2006) Rab3A and Rab27A cooperatively regulate the docking step of dense-core vesicle exocytosis in PC12 cells. J Cell Sci 119: 2196–2203 Turner MD, Arvan P (2000) Protein traffic from the secretory pathway to the endosomal system in pancreatic b-cells. J Biol Chem 275: 14025–14030

518

*

A. A. Mironov and P. Arvan

Urbe S, Tooze SA, Barr FA (1997a) Formation of secretory vesicles in the biosynthetic pathway. Biochim Biophys Acta 1358(1): 6–22 Urbe S, Dittie AS, Tooze SA (1997b) pH-dependent processing of secretogranin II by the endopeptidase PC2 in isolated immature secretory granules. Biochem J 321(Pt 1): 65–74 Urbe S, Page LJ, Tooze SA (1998) Homotypic fusion of immature secretory granules during maturation in a cell-free assay. J Cell Biol 143: 1831–1844 Varadi A, Ainscow EK, Allan VJ, Rutter GA (2002) Involvement of conventional kinesin in glucose-stimulated secretory granule movements and exocytosis in clonal pancreatic b-cells. J Cell Sci 115(Pt 21): 4177–4189 Varlamov O, Fricker LD, Furukawa H, Steiner DF, Langley SH, Leiter EH (1997) Beta-cell lines derived from transgenic Cpe(fat)/Cpe(fat) mice are defective in carboxypeptidase E and proinsulin processing. Endocrinology 138: 4883–4892 Velasco A, Hendricks L, Moreman KW, Tulsiani DRP, Touster O, Farquhar MG (1993) Celltype dependent variations in the subcellular distribution of a-mannosidase I and II. J Cell Biol 122: 39–51 Venkatesh SG, Gorr SU (2002) A sulfated proteoglycan is necessary for storage of exocrine secretory proteins in the rat parotid gland. Am J Physiol Cell Physiol 283: C438–C445 Venkatesh SG, Cowley DJ, Gorr SU (2004) Differential aggregation properties of secretory proteins that are stored in exocrine secretory granules of the pancreas and parotid glands. Am J Physiol Cell Physiol 286(2): C365–C371 Von Zastrow M, Castle JD (1987) Protein sorting among two distinct export pathways occurs from the content of maturing exocrine storage granules. J Cell Biol 105: 2675–2684 € mer A, Migala A, Almers W, Gerdes HH (1997) MicrotubuleWacker I, Kaether C, Kro dependent transport of secretory vesicles visualized in real time with a GFP-tagged secretory protein. J Cell Sci 110(Pt 13): 1453–1463 Wang Y, Thiele C, Huttner WB (2000) Cholesterol is required for the formation of regulated and constitutive secretory vesicles from the trans-Golgi network. Traffic 1: 952–962 Wasmeier C, Hutton JC (1996) Molecular cloning of phogrin, a protein-tyrosine phosphatase homologue localized to insulin secretory granule membranes. J Biol Chem 271: 18161–18170 Wasmeier C, Bright NA, Hutton JC (2002) The lumenal domain of the integral membrane protein phogrin mediates targeting to secretory granules. Traffic 3(9): 654–665 Wendler F, Page L, Urbe S, Tooze SA (2001) Homotypic fusion of immature secretory granules during maturation requires syntaxin 6. Mol Biol Cell 12: 1699–1709 Wheeler MB, Sheu L, Ghai M, Bouquillon A, Grondin G, Weller U, Beaudoin AR, Bennett MK, Trimble WS, Gaisano H (1996) Characterization of SNARE protein expression in beta cell lines and pancreatic islets. Endocrinology 137: 1340–1348 Winkler H, Westhead E (1980) The molecular organization of adrenal chromaffin granules. Neuroscience 5(11): 1803–1823 Winkler H, Fischer-Colbrie R (1992) The chromogranins A and B: the first 25 years and future perspectives. Neuroscience 49(3): 497–528 Wiser O, Trus M, Hernandez A, Renstrom E, Barg S, Rorsman P, Atlas D (1999) The voltage sensitive Lc-type Ca2 þ channel is functionally coupled to the exocytotic machinery. Proc Natl Acad Sci USA 96(1): 248–253 Wu MM, Grabe M, Adams S, Tsien RY, Moore HP, Machen TE (2001) Mechanisms of pH regulation in the regulated secretory pathway. J Biol Chem 276: 33027–33035 Xu H, Shields D (1994) Prosomatostatin processing in permeabilized cells. J Biol Chem 269: 22875–22881 Yoo SH, Lewis MS (1992) Effects of pH and Ca2 þ on monomer–dimer and monomer–tetramer equilibria of chromogranin A. J Biol Chem 267(16): 11236–11241

Origins of the regulated secretory pathway

*

519

Yoo SH (1996) pH- and Ca(2 þ )-dependent aggregation property of secretory vesicle matrix proteins and the potential role of chromogranins A and B in secretory vesicle biogenesis. J Biol Chem 271(3): 1558–1565 Yoo SH (2000) Coupling of the IP3 receptor/Ca2 þ channel with Ca2 þ storage proteins chromogranins A and B in secretory granules. Trends Neurosci 23(9): 424–428 Yoo S, Oh Y, Kang M, Huh Y, So S, Park H, Park H (2001) Localization of three types of the inositol 1,4,5-trisphosphate receptor/Ca(2 þ ) channel in the secretory granules and coupling with the Ca(2 þ ) storage proteins chromogranins A and B. J Biol Chem 276 (49): 45806–45812 Zenner HL, Collinson LM, Michaux G, Cutler DF (2007) High-pressure freezing provides insights into Weibel-Palade body biogenesis. J Cell Sci 120(Pt12): 2117–2125 Zhang B-Z, Chang A, Kjeldsen TB, Arvan P (2001) Intracellular retention of newlysynthesized insulin in yeast is caused by endoproteolytic processing in the Golgi complex. J Cell Biol 153: 1187–1197 Zhao S, Torii S, Yokota-Hashimoto H, Takeuchi T, Izumi T (2002) Involvement of Rab27b in the regulated secretion of pituitary hormones. Endocrinology 143(5): 1817–1824

520

*

€ller and A. Ellinger J. Neumu

Secretion and endocytosis in endothelial cells € ller and Adolf Ellinger Josef Neumu

Although the cell machinery for uptake and transport of nutrients and lipids in EC follows the general routes and main rules observed in other cell types, there are several unique features that can be demonstrated in endothelial cells. This fact is due to the impact of the endothelium in the transport of proteins, glucose and lipids from blood circulation to the surrounding tissue. Furthermore, the exchange of the blood gases oxygen and carbon dioxide between capillaries and the surrounding tissues is a main task of endothelium. For this, a high specialization of microvascular architecture and function can be found in different tissues and organs (for reviews see Aird 2008; Andersson et al. 2004; Craig et al. 1998; Ekataksin and Kaneda 1999; Firth 2002; Hendrickx et al. 2004; Hudetz 1997; Lehmann et al. 2000; Levick 1995; Lim et al. 2003; Michel 1995; Ohta et al. 1992; Pittman 1995; Yano et al. 2007). A third functional aspect of endothelium is the maintenance of blood hemostasis. The constitutive as well as the regulated secretion of von Willebrand factor in relation to the posttranslational modification in the ER and in the Golgi apparatus will be discussed later in this article. The sorting machinery of the Golgi apparatus is bypassed in EC by transcytosis where the caveolar system is involved. The central function of transcytosis is the transport of albumin and lipids, hormones and peptides that bind avidly to albumin (Minshall et al. 2002). Caveolae form a unique endocytic as well as exocytic compartment at the cell surface and are capable of importing molecules which are delivered at the abluminal side of the endothelium. Feng et al. (2002) showed impressively the ultrastructural features of transendothelial transport in venules, lymphatic vessels and tumor-associated microvessels in man and animals including 3D models, obtained from serial ultra-thin sections. Most of the in vitro investigations have been done using two dimensional cell cultures of human veins such as HUVEC (Bachetti and Morbidelli 2000; Ito et al. 2005; Li et al. 2006; Menzel et al. 1997; Ochiai et al. 1999), microvascular EC from skin and other organs as well as from the arterial vascular tree such as from aorta or coronary vessels (Bagavandoss and Wilks 1991; Binion et al. 2000; Carley et al. 1992; Carson and Haudenschild 1986; Cheng and Kramer 1989; Clough 1991; DeBault et al. 1979; Diglio et al. 1982; Dorovini-Zis et al. 1991; Dunky et al. 1997; Johnson et al. 2002; Kern et al. 1983; Lamszus et al. 1999; Lehmann et al. 2000; Michel and Curry € ller et al. 2001; Stein and St Clair 1988; Stolz and Jacobson 1999; Neumu 1991; Tscheudschilsuren et al. 2002; Watkins et al. 2004; Weston et al.

Secretion and endocytosis in endothelial cells

*

521

2003). For particular questions on the endothelial functions three-dimensional systems using gels containing extracellular matrix proteins (Aird 2006; Kim 2005; Ment et al. 1997; Mueller-Klieser 1997; Nicosia and Ottinetti 1990; Risbud et al. 2002; Scherberich and Beretz 2000) or spheroids (KunzSchughart et al. 1998; Mueller-Klieser 1997; Oudar 2000) have been developed. In this article, particular aspects of protein and lipid synthesis, as well as the regulated exocytosis of VWF and WPB-related proteins will be discussed. All of these processes are more or less connected to the import, export and sorting machinery of the Golgi apparatus, the central topic of this book.

The Golgi apparatus in EC With exception of endothelium in high endothelial venules, EC are highly polarized flat cells with a maximal height of approximately 10 mm at the perinuclear region. The Golgi apparatus can be impressively demonstrated by FM and LSM using short chain ceramide analogous conjugated with fluorescent dyes such as BODIPY or NBD (Martin and Pagano 1994; Puri et al. 2001). The visualization of ceramide in the Golgi apparatus and the TGN can be done by in situ immunofluorescence, or by interaction of labeled ceramide attached to defatted albumin. The association of the ceramide analogues with intracellular membranes of endosomes and the Golgi apparatus remains unclear. Internalization studies with labeled lipid analogues have shown a redistribution of lipids at the plasma membrane or in the forming endosomes within seconds after initiation of endocytosis (Chen et al. 1997). Physiologically, ceramide is synthesized de novo in the ER while the synthesis of the sphingolipid species sphingomyelin, sphingosine, glucosylceramide and galactosylceramide takes place in the Golgi apparatus (Ohanian and Ohanian 2001). Own experiments using HDMEC revealed the same extended Golgi apparatus in perinuclear localization as it has been described by Pagano (1994) in fibroblasts (Fig. 1). After binding to EC at 4 C for 15–30 min and subsequent rewarming at 37 C, the ceramide analogues are internalized by the cells within a few minutes and enriched at the trans-Golgi side as well as in the TGN. Using LSM and TEM, it could be demonstrated that the uptake of HRP-conjugated WGA is delayed after incubation with short chain ceramide (Dillinger 2003). Figure 2 shows the incorporation of WGA after 30 min binding to EC at 4 C conditions and incubation at 37 C for 5 and 30 min, respectively. These results are in agreement with those of Chen et al. (1995) who showed that ceramide is able to inhibit endocytosis. In addition, the results of Abousalham et al. (2002), demonstrate an effect of C2-ceramide in CHO cells by decreasing ARF-1 and PKC-a binding to Golgi-enriched membranes and inhibition of COP1 vesicle formation. Sphingolipids also seem to stabilize the Golgi apparatus: short chain ceramides blocked the brefeldin-A

522

*

€ller and A. Ellinger J. Neumu

Figure 1. Overview micrograph of a culture of HDMECs, incubated for 30 min at 4 C with BODIPY-FL-ceramide followed by 30 min incubation at 37 C. In the fluorescence microscope uptake of ceramide by the cells into the endocytic compartment (arrow) and rapid transfer to the Golgi apparatus and the TGN is seen (arrowhead).

induced disassembly of the Golgi apparatus in NRK-cells (Fukunaga et al. 2000). Although only the ultrastructure provides a detailed morphological investigation of cultured EC, the rather flat appearance of the Golgi apparatus allows studying only some aspects of this organelle in a single section. Figure 3 shows an overview of the Golgi apparatus in a pericentriolar region (arrow). Several Golgi stacks (arrowheads) in connection with the vesiculo-tubular TGN-compartment are visible. Golgi staining using BODIPY ceramide is also a suitable tool for correlative studies both at the light microscopical and ultrastructural level applying the photoconversion technique (Ladinsky et al. 1994; Meisslitzer-Ruppitsch et al. 2008). The Golgi apparatus is multiply engaged in the biosynthesis of adhesion molecules, essential for cell–cell interactions and signaling during recruitment of leukocytes in the microvasculature of inflamed tissues (mainly in postcapillary venules). These events lead to rolling, tethering and peripedesis or extravasation of inflammatory cells. In tumorgenesis, cells can migrate inside of vessels in order to disseminate to other organs. Of particular interest is the release of nitric oxide which plays a pivotal role in vascular hemostasis. It has multiple functions in the vessels such as regula-

Secretion and endocytosis in endothelial cells

*

523

Figure 2. Uptake of FITC-labeled WGA (a and b) and horseradish peroxidase-conjugated WGA (c and d) by cultures of HDMEC. The WGA uptake was visualized using FM or TEM respectively. For these experiments, the cells were incubated for 5 (a and c) as well as for 30 min (b and d) at 4 C with the lectin followed by a warming up period of 30 min at 37 C. Note the continuous staining of the plasma membrane still after 5 min of incubation (arrowheads; a and c). b An additional vesicular staining pattern after 30 min of postincubation (arrows) but also a faint accumulation in the perinuclear Golgi/TGN area. The TEM image shows that this staining is present in endosomal (arrows) and lysosomal (arrowheads) compartments in close vicinity to the Golgi apparatus (arrows; d). In contrast to other cell systems such as HepG2 cells, in HDMECs, the staining could never be observed in the transGolgi cisterns.

tion of the vascular tone, it is anti-inflammatory, antiproliferative and antithrombotic (for reviews see: Braam and Verhaar (2007); Schulz et al. (2008)). Therefore, it has a variety of implications in vascular diseases. NO is formed from L-arginine by the enzymatic action of the endothelial nitric oxide synthase (eNOS) with the help of some cofactors. The expression and function of eNOS is important for the release of NO by the endothelium. There are two localizations of eNOS, the trans-Golgi cisterns and the caveolae in the plasma membrane. A defective function of the Golgi apparatus and the impairment of the traffic from TGN to the caveolar system could be a crucial point for altered NO release in hypercholesterolemia (Jin 2006; Zhang et al. 2006).

524

*

€ller and A. Ellinger J. Neumu

Figure 3. TEM image showing the perinuclear localization of the Golgi apparatus in cultured HDMEC using chemical fixation and conventional embedding. Note the centriole in the center of the Golgi system (arrows) and the Golgi stacks (arrowheads); flat sectioned region (F).

The secretory pathway of endothelium-specific organelles – Weibel–Palade bodies (WPB) Among the various proteins, synthesized by EC, the von Willebrand factor is of particular interest. VWF is synthesized as a 350 kDa monomer that includes a signal sequence (22 residues), a pro-peptide (741 residues) and the mature protein (2,050 residues), summarized by Metcalf et al. (2008). The biosynthetic pathway, mainly involving the Golgi apparatus and the TGN, includes several steps, summarized in Table 1. The monomers are translocated into the ER, glycosylated at 12 N-linked sites and dimerized by forming disulphide bonds at the C-terminus (Titani et al. 1986). This process occurs as oxidoreductase reaction at low pH while the pro-peptide functions as a chaperone (Sadler 2005). The processed VWF is subsequently transported to the Golgi apparatus, where O-linked glycosylation and modifications of N-linked oligosaccharides occur (Michaux and Cutler 2004); further multimerization takes place in the TGN. Subsequently, a transient linkage between the pro-peptide and the D0 -D3 domain of the mature protein is applied, which is later on cleaved by the endopeptidase furin, an enzyme

Secretion and endocytosis in endothelial cells

*

525

Table 1. Synthetic pathway of VWF and formation of WPB ER

1. Formation of VWF monomers containing the pro-peptide including a signal peptide and the domains D1 and D2 and the mature molecule including the domains D0 , D3, A1-A3, D4, B1-B3, C1, C2 and CK; 2. dimerization at the C-terminus; 3. initial N-linked glycosylation

ER and cis-Golgi

1. Formation of disulphide bonds by transient linkage between the pro-peptide and the D0 -D3 domains of the mature molecule; 2. O-linked glycosylation and modification of Nlinked oligosaccharides

trans-Golgi and TGN

Formation of N-terminal interchain bonds by cleavage of the transient linkage between the pro-peptide and the D0 -D3 domains of the mature molecule by the endopeptidase furin

TGN

1. Initiation of tubulation at low pH and high concentration of Ca2þ providing a high affinity between the VWF domains D1D2 and D0 D3 and allowing the formation of disulphide bonds by juxtaposing two D3 domains. Linkage between the pro-peptide and the mature VWF molecule by non-covalent bonds in a 1: 1 stoichiometric ratio; 2. formation of AP-1/ clathrin coats; 3. targeting of P-selectin to the membrane and enrichment of luminal proteins

Formation of early electron-lucent WPB

1. Budding of early WPB from TGN after assembly with an AP-1/clathrin coat; 2. formation of the rod like shape by progressive tubulation of the VWF

Transition into mature electron-dense WPB

1. Tight twisted tubulation of VWF; 2. Membrane targeting of CD63 by the AP3/clathrin complex; 3. recruitment of Rab27a and Rab3d; 4. new formation of bridges between WPB and TGN Formation of early and late WPB can occur simultaneously

localized in the TGN (Hosaka et al. 1991). The pro-peptide and the mature VWF molecule remain linked by non-covalent bonds in a 1:1 stoichiometric ratio (Voorberg et al. 1990). The multimerization is leading to the formation of large molecules with a size of up to 20 MDa. The VWF multimers are constitutively secreted by EC or transferred via the TGN in rod like organelles, where they are organized into a longitudinally striation pattern with a microtubular-like structure. These organelles have been originally described in 1964 by Weibel and Palade (Weibel 1964). The shape of the WPB comprises a length up to 5 mm and a diameter ranging from 100–300 nm. These striation pattern of VWF multimers is discussed to occur due to selective coaggregation of stored components in the TGN or due to the function of sorting receptor(s) (Michaux and Cutler 2004). Some membrane-associated proteins have been demonstrated in WPB: integral membrane proteins such as the adhesion molecule P-selectin, the tetraspanin CD63 and the fucosyl transferase VI as well as the small GTPases Rab27a and Rab3d. The lumen of WPB comprises several proteins like the

526

*

€ller and A. Ellinger J. Neumu

chemokine interleukin-8, the tissue type plasminogen activator, the endothelin-1 and the endothelin-converting enzyme, eotaxin-3, angiopoietin-2, and osteoprotegerin (Metcalf et al. 2008). In own experiments using endothelial progenitor derived cells, differentiated from hematopoetic stem cells, a high number of newly formed multi€ ller shaped WPB with bendings and bifurcations could be observed (Neumu et al. 2006 2007). The formation of WPB proceeds inside of enlarged pockets of the TGN in vicinity of the Golgi apparatus (Fig. 4a–d). ET investigations of high pressure fixed and cryosubstituted WPB have been recently presented by Zenner et al. (2007) and Valentijn et al. (2008). The authors report also about multishaped WPB, thought to be a sign of homeotypic fusion of these organelles. ET could also reveal that the VWF tubules follow a twisted

Figure 4. Demonstration of characteristic features of WPB using TEM and ET. a Ultrathin section of an EPDC in an early stage of WPB formation. Note the tubulation inside of enlarged pockets of the TGN; (b) shows irregularly shaped WPB; (c) a bifurcated WPB (arrow). In d a slice out of ET volume is shown demonstrating the twisting course of the VWF tubules (arrowheads) which is interrupted at the kink point (arrow).

Secretion and endocytosis in endothelial cells

*

527

longitudinal orientation (Valentijn et al. 2008). Our own findings about connecting stalks between the mature WPB and the TGN are supported by the work of (Zenner et al. 2007). In the review of Metcalf et al. (2008), different steps of WPB formation and assembly of VWF have been described: (a) dimerization at the C-terminus taking place in the ER; (b) formation of disulphide bonds at the N-terminus; (c) polymerization of dimers occurring due to the formation of disulphide bonds replacing the transient disulphide bonds between the cleaved pro-peptide and the D3 domain of the mature VWF molecule; (d) tubulation of VWF by ionic interaction between the mature VWF molecule and the pro-peptide. It is important to point out that although the ER would provide a favourable environment for the formation of disulphide bonds between the D3 domains of the VWF subunits, the low pH and high Ca2þ concentration of the Golgi apparatus and the TGN facilitate the affinity between the VWF domains D1D2 and D0 D3 allowing the formation of disulphide bonds by juxtaposing two D3 domains (Huang et al. 2008). The initial formation of WPB requires a cytoplasmic machinery that involves the heterotetrameric adaptor protein 1 (AP1) and clathrin (Lui-Roberts et al. 2005). These authors showed that the surface of immature WPB is partially or completely covered by a clathrin coat, a finding controversially discussed in the literature (Valentijn et al. 2008). Nevertheless, this coat seems to be important for tubulation of VWF and formation of the rod-like shape of the organelle. Reaching this shape, the WPB appear increasingly electron-dense in TEM and loose their clathrin coat. In contrast to the secretory proteins angiopoietin-2, eotaxin, osteoprotegerin and P-selectin, proteins such as CD63, Rab27A and Rab3D are recruited after budding of WPB from TGN. The luminal domain of P-selectin is able to bind to the D0 -D3 domain of the VWF and to target them into the newly formed WPB. Hence, this second wave involves the adaptor protein-3 (AP-3) complex and CD63 allowing the recruitment of proteins into the mature WPB. Nevertheless, the full function of CD63, a constituent of late endosomes and lysosomes and present at the limiting membrane of WPB remains unclear (Vischer and Wagner 1993). Its interaction with several adhesion molecules suggests a possible role in inflammatory processes where activated EC bind leukocytes via counter receptors. These leukocytes transmigrate through the vessel wall into the extravasal tissue. The significance of the AP-3 independent pathway of P-selectin in the formation of Weibel–Palade bodies by budding from TGN in contrast to the AP-3 dependent mechanism involving CD63 has been clearly demonstrated by (Harrison-Lavoie et al. 2006). This second pathway might be involved in processes, analyzed in our own experiments using electron tomography where bridging connections between WPB and the TGN could be demonstrated. We assume that such connections could manage the transport of WPB constituents (Fig. 5a–e). In addition, a partial colocalization between fluorolabeled VWF and the trans-Golgi and TGN marker TGN-46, could be demonstrated in the LSM (Fig. 5c).

528

*

€ller and A. Ellinger J. Neumu

Figure 5. Relationships between TGN and WPB in LSM and ET. a and b demonstrate different slices from an ET volume showing a newly formed electron-lucent WPB with irregular courses of the VWF tubules (arrow). In b two coats of the connecting TGN are visible (arrowheads). c shows a partial colocalization (in yellow; arrowheads) of TGN-46, FITC-conjugated (in green; small arrow) and VWF visualized using an indirect immunofluorescence with Alexa Fluor 568 (in red; large arrow) in LSM. WPB can be recognized as long rod-like structures. d demonstrates a slice out of ET volume showing the interaction of TGN membranes with a WPB without any coat (arrowhead). In e, a slice out of ET volume shows a mature WPB (star) and a bridge (arrow) consisting of TGN and an immature WPB (x). The respective model is shown in f.

Secretion and endocytosis in endothelial cells

*

529

Another constituent of WPB is the chemokine IL-8. Bierings et al. (2007) could demonstrate by using cell fractionation and density gradient centrifugation of HUVEC that VWF can direct the storage of IL-8 to WPB. The low pH at the trans-Golgi/TGN compartment facilitates the association of VWF and IL-8 and allows its targeting to the WPB. Exocytosis of WPB results in the release of long strings of VWF polymers providing an adhesion site for activated platelets. This attachment occurs spontaneously via the platelet GP Ib-a receptor, a component of the GP Ib-IX-V integrin complex. After secretion of the ultra-large VWF multimers, they are cleaved by a limited proteolysis by the plasma metalloprotease ADAMTS-13 (Dong et al. 2002). Exocytosis of WPB can be induced by several scretagogues such as thrombin or histamine (Rondaij et al. 2006, 2008). Upon stimulation of EC, the phosphatidyl inositol pathway is activated, resulting in a rise of cytoplasmic Ca2þ . Thrombin signaling is mediated by the G protein-coupled protease-activated receptor (PAR). Both, vasopressin, acting via the V2R receptor as well as epinephrine binding to the b2adrenergic receptor follow the c-AMP signaling pathway (van Mourik et al. 2002). Until now, the exocytotic machinery concerning the WPB could not sufficiently be elucidated. A regulatory role exerts the GTPases Rab3D and Ra1a. In addition, it has been shown, that the SNARE proteins syntaxin-4, VAMP3 and SNAP23 play a crucial role in the exocytosis of WPB (Matsushita et al. 2003; Lowenstein et al. 2005; Fu et al. 2005; Sehgal and Mukhopadhyay 2007).

Abbreviations BODIPY EC ET ER FM HDL HDMEC HRP HUVEC LDL LSM NBD SNARE NSF TGN TEM VWF WGA WPB

4,4-difluoro-4-bora-3a,4a-diaza-s-indacene endothelial cells electron tomography endoplasmic reticulum fluorescence microscopy high density lipoproteins human dermal microvascular endothelial cells horseradish peroxidase human umbilical vein endothelial cells low density lipoproteins confocal laser scanning microscope N-{[6-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino]hexanoyl} SNAP receptor protein (SNAP: soluble NSF attachment protein N-ethylmaleimide sensitive factor) trans-Golgi network transmission electron microscopy von Willebrand factor wheat germ agglutinin Weibel-Palade bodies

530

*

€ller and A. Ellinger J. Neumu

Acknowledgements. The authors gratefully acknowledge the excellent technical assistance of Mrs. Beatrix Mallinger, Mrs. Elfriede Scherzer as well as Mr. Ulrich Kaindl for image processing and Mr. Jedrzej Kosiuk for performing the ET model.

References Abousalham A, Hobman TC, Dewald J, Garbutt M, Brindley DN (2002) Cell-permeable ceramides preferentially inhibit coated vesicle formation and exocytosis in Chinese hamster ovary compared with Madin-Darby canine kidney cells by preventing the membrane association of ADP-ribosylation factor. Biochem J 361: 653–661 Aird WC (2006) Mechanisms of endothelial cell heterogeneity in health and disease. Circ Res 98: 159–162 Aird WC (2008) Endothelium in health and disease. Pharmacol Rep 60: 139–143 Andersson U, Malmer B, Bergenheim AT, Brannstrom T, Henriksson R (2004) Heterogeneity in the expression of markers for drug resistance in brain tumors. Clin Neuropathol 23: 21–27 Bachetti T, Morbidelli L (2000) Endothelial cells in culture: a model for studying vascular functions. Pharmacol Res 42: 9–19 Bagavandoss P, Wilks JW (1991) Isolation and characterization of microvascular endothelial cells from developing corpus luteum. Biol Reprod 44: 1132–1139 Bierings R, van den Biggelaar M, Kragt A, Mertens K, Voorberg J, van Mourik JA (2007) Efficiency of von Willebrand factor-mediated targeting of interleukin-8 into Weibel–Palade bodies. J Thromb Haemost 5: 2512–2519 Binion DG, Rafiee P, Ramanujam KS, Fu S, Fisher PJ, Rivera MT, Johnson CP, Otterson MF, Telford GL, Wilson KT (2000) Deficient iNOS in inflammatory bowel disease intestinal microvascular endothelial cells results in increased leukocyte adhesion. Free Radic Biol Med 29: 881–888 Braam B, Verhaar MC (2007) Understanding eNOS for pharmacological modulation of endothelial function: a translational view. Curr Pharm Des 13: 1727–1740 Carley WW, Niedbala MJ, Gerritsen ME (1992) Isolation, cultivation, and partial characterization of microvascular endothelium derived from human lung. Am J Respir Cell Mol Biol 7: 620–630 Carson MP, Haudenschild CC (1986) Microvascular endothelium and pericytes: high yield, low passage cultures. In Vitro Cell Dev Biol 22: 344–354 Chen CS, Martin OC, Pagano RE (1997) Changes in the spectral properties of a plasma membrane lipid analog during the first seconds of endocytosis in living cells. Biophys J 72: 37–50 Chen CS, Rosenwald AG, Pagano RE (1995) Ceramide as a modulator of endocytosis. J Biol Chem 270: 13291–13297 Cheng YF, Kramer RH (1989) Human microvascular endothelial cells express integrinrelated complexes that mediate adhesion to the extracellular matrix. J Cell Physiol 139: 275–286 Clough G (1991) Relationship between microvascular permeability and ultrastructure. Prog Biophys Mol Biol 55: 47–69 Craig LE, Spelman JP, Strandberg JD, Zink MC (1998) Endothelial cells from diverse tissues exhibit differences in growth and morphology. Microvasc Res 55: 65–76 DeBault LE, Kahn LE, Frommes SP, Cancilla PA (1979) Cerebral microvessels and derived cells in tissue culture: isolation and preliminary characterization. In Vitro 15: 473–487 Diglio CA, Grammas P, Giacomelli F, Wiener J (1982) Primary culture of rat cerebral microvascular endothelial cells. Isolation, growth, and characterization. Lab Invest 46: 554–563

Secretion and endocytosis in endothelial cells

*

531

Dillinger A (2003) Die Interaktion von Lipidenund Proteinen bei der Endozytose €ren Zellkulturen und Zelllinien). Inst f Histologie und (Untersuchung an prima €t und Inst f Anthropologie der Fakulta €t fu €r Embryologie der Medizinische Fakulta €t Wien, Wien, p 75 Naturwissenschaften und Mathematik. Universita Dong JF, Moake JL, Nolasco L, Bernardo A, Arceneaux W, Shrimpton CN, Schade AJ, McIntire LV, Fujikawa K, Lopez JA (2002) ADAMTS-13 rapidly cleaves newly secreted ultralarge von Willebrand factor multimers on the endothelial surface under flowing conditions. Blood 100: 4033–4039 Dorovini-Zis K, Prameya R, Bowman PD (1991) Culture and characterization of microvascular endothelial cells derived from human brain. Lab Invest 64: 425–436 € ller J, Menzel J (1997) Interactions of lymphocytes from patients with Dunky A, Neumu psoriatic arthritis or healthy controls and cultured endothelial cells. Clin Immunol Immunopathol 85: 297–314 Ekataksin W, Kaneda K (1999) Liver microvascular architecture: an insight into the pathophysiology of portal hypertension. Semin Liver Dis 19: 359–382 Feng D, Nagy JA, Dvorak HF, Dvorak AM (2002) Ultrastructural studies define soluble macromolecular, particulate, and cellular transendothelial cell pathways in venules, lymphatic vessels, and tumor-associated microvessels in man and animals. Microsc Res Tech 57: 289–326 Firth JA (2002) Endothelial barriers: from hypothetical pores to membrane proteins. J Anat 200: 541–548 Fu J, Naren AP, Gao X, Ahmmed GU, Malik AB (2005) Protease-activated receptor-1 activation of endothelial cells induces protein kinase Calpha-dependent phosphorylation of syntaxin 4 and Munc18c: role in signaling p-selectin expression. J Biol Chem 280: 3178–3184 Fukunaga T, Nagahama M, Hatsuzawa K, Tani K, Yamamoto A, Tagaya M (2000) Implication of sphingolipid metabolism in the stability of the Golgi apparatus. J Cell Sci 113 ( Pt 18): 3299–3307 Harrison-Lavoie KJ, Michaux G, Hewlett L, Kaur J, Hannah MJ, Lui-Roberts WW, Norman KE, Cutler DF (2006) P-selectin and CD63 use different mechanisms for delivery to Weibel–Palade bodies. Traffic 7: 647–662 Hendrickx J, Doggen K, Weinberg EO, Van Tongelen P, Fransen P, De Keulenaer GW (2004) Molecular diversity of cardiac endothelial cells in vitro and in vivo. Physiol Genomics 19: 198–206 Hosaka M, Nagahama M, Kim WS, Watanabe T, Hatsuzawa K, Ikemizu J, Murakami K, Nakayama K (1991) Arg-X-Lys/Arg-Arg motif as a signal for precursor cleavage catalyzed by furin within the constitutive secretory pathway. J Biol Chem 266: 12127–12130 Huang RH, Wang Y, Roth R, Yu X, Purvis AR, Heuser JE, Egelman EH, Sadler JE (2008) Assembly of Weibel–Palade body-like tubules from N-terminal domains of von Willebrand factor. Proc Natl Acad Sci USA 105: 482–487 Hudetz AG (1997) Blood flow in the cerebral capillary network: a review emphasizing observations with intravital microscopy. Microcirculation 4: 233–252 Ito Y, Hasuda H, Terai H, Kitajima T (2005) Culture of human umbilical vein endothelial cells on immobilized vascular endothelial growth factor. J Biomed Mater Res A 74: 659–665 Jin ZG (2006) Where is endothelial nitric oxide synthase more critical: plasma membrane or Golgi? Arterioscler Thromb Vasc Biol 26: 959–961 Johnson EK, Schelling ME, Quitadamo IJ, Andrew S, Johnson EC (2002) Cultivation and characterization of coronary microvascular endothelial cells: a novel porcine model using micropigs. Microvasc Res 64: 278–288 Kern PA, Knedler A, Eckel RH (1983) Isolation and culture of microvascular endothelium from human adipose tissue. J Clin Invest 71: 1822–1829

532

*

€ller and A. Ellinger J. Neumu

Kim JB (2005) Three-dimensional tissue culture models in cancer biology. Semin Cancer Biol 15: 365–377 Kunz-Schughart LA, Kreutz M, Knuechel R (1998) Multicellular spheroids: a threedimensional in vitro culture system to study tumour biology. Int J Exp Pathol 79: 1–23 Ladinsky MS, Kremer JR, Furcinitti PS, McIntosh JR, Howell KE (1994) HVEM tomography of the trans-Golgi network: structural insights and identification of a lace-like vesicle coat. J Cell Biol 127: 29–38 Lamszus K, Schmidt NO, Ergun S, Westphal M (1999) Isolation and culture of human neuromicrovascular endothelial cells for the study of angiogenesis in vitro. J Neurosci Res 55: 370–381 Lehmann I, Brylla E, Sittig D, Spanel-Borowski K, Aust G (2000) Microvascular endothelial cells differ in their basal and tumour necrosis factor-alpha-regulated expression of adhesion molecules and cytokines. J Vasc Res 37: 408–416 Levick JR (1995) Microvascular architecture and exchange in synovial joints. Microcirculation 2: 217–233 Li N, Eljaafari A, Bensoussan D, Wang Y, Latger-Cannard V, Serrurier B, Boura C, Kennel A, Stoltz J, Feugier P (2006) Human umbilical vein endothelial cells increase ex vivo expansion of human CD34( þ ) PBPC through IL-6 secretion. Cytotherapy 8: 335–342 Lim YC, Garcia-Cardena G, Allport JR, Zervoglos M, Connolly AJ, Gimbrone MA Jr, Luscinskas FW (2003) Heterogeneity of endothelial cells from different organ sites in T-cell subset recruitment. Am J Pathol 162: 1591–1601 Lowenstein CJ, Morrell CN, Yamakuchi M (2005) Regulation of Weibel–Palade body exocytosis. Trends Cardiovasc Med 15: 302–308 Lui-Roberts WW, Collinson LM, Hewlett LJ, Michaux G, Cutler DF (2005) An AP-1/clathrin coat plays a novel and essential role in forming the Weibel–Palade bodies of endothelial cells. J Cell Biol 170: 627–636 Martin OC, Pagano RE (1994) Internalization and sorting of a fluorescent analogue of glucosylceramide to the Golgi apparatus of human skin fibroblasts: utilization of endocytic and nonendocytic transport mechanisms. J Cell Biol 125: 769–781 Matsushita K, Morrell CN, Cambien B, Yang SX, Yamakuchi M, Bao C, Hara MR, Quick RA, Cao W, O’Rourke B, Lowenstein JM, Pevsner J, Wagner DD, Lowenstein CJ (2003) Nitric oxide regulates exocytosis by S-nitrosylation of N-ethylmaleimide-sensitive factor. Cell 115: 139–150 € ller J, Freissmuth M, Meisslitzer-Ruppitsch C, Vetterlein M, Stangl H, Maier S, Neumu Pavelka M, Ellinger A (2008) Electron microscopic visualization of fluorescent signals in cellular compartments and organelles by means of DAB-photoconversion. Histochem Cell Biol (in press) Ment LR, Stewart WB, Scaramuzzino D, Madri JA (1997) An in vitro three-dimensional coculture model of cerebral microvascular angiogenesis and differentiation. In Vitro Cell Dev Biol Anim 33: 684–691 € ller J, Sengoelge G, Reihsner R (1997) Effects of aminoguanidine on Menzel EJ, Neumu adhesion molecule expression of human endothelial cells. Pharmacology 55: 126–135 Metcalf DJ, Nightingale TD, Zenner HL, Lui-Roberts WW, Cutler DF (2008) Formation and function of Weibel–Palade bodies. J Cell Sci 121: 19–27 Michaux G, Cutler DF (2004) How to roll an endothelial cigar: the biogenesis of Weibel–Palade bodies. Traffic 5: 69–78 Michel CC (1995) Renal medullary microcirculation: architecture and exchange. Microcirculation 2: 125–139 Michel CC, Curry FE (1999) Microvascular permeability. Physiol Rev 79: 703–761 Minshall RD, Tiruppathi C, Vogel SM, Malik AB (2002) Vesicle formation and trafficking in endothelial cells and regulation of endothelial barrier function. Histochem Cell Biol 117: 105–112

Secretion and endocytosis in endothelial cells

*

533

Mueller-Klieser W (1997) Three-dimensional cell cultures: from molecular mechanisms to clinical applications. Am J Physiol 273: C1109–C1123 € ller J, Dunky A, Burtscher H, Jilch R, Menzel JE (2001) Interaction of monocytes Neumu from patients with psoriatic arthritis with cultured microvascular endothelial cells. Clin Immunol 98: 143–152 € ller J, Kosiuk J, Ellinger A, Vetterlein M, Pavelka M (2007) Demonstration of Neumu multishaped Weibel–Palade bodies in early matured endothelial cells derived from cord blood stem cells using correlative microscopy and 3D-EM tomography. In: Engel A (ed) Network of Excellence – 3DEM, 3rd Annual Meeting, 23–25, January 2007, Palma de Mallorca € ller J, Neumu € ller-Guber SE, Lipovac M, Mosgoeller W, Vetterlein M, Pavelka M, Neumu Huber J (2006) Immunological and ultrastructural characterization of endothelial cell cultures differentiated from human cord blood derived endothelial progenitor cells. Histochem Cell Biol 126: 649-664 Nicosia RF, Ottinetti A (1990) Modulation of microvascular growth and morphogenesis by reconstituted basement membrane gel in three-dimensional cultures of rat aorta: a comparative study of angiogenesis in matrigel, collagen, fibrin, and plasma clot. In Vitro Cell Dev Biol 26: 119–128 Ochiai K, Omura M, Mochizuki A, Ito M, Tomioka H (1999) Human umbilical vein endothelial cells support interleukin-3- and interleukin-5-induced eosinophil differentiation from cord blood CD34 þ cells. Int Arch Allergy Immunol 120 Suppl 1: 2–6 Ohanian J, Ohanian V (2001) Sphingolipids in mammalian cell signalling. Cell Mol Life Sci 58: 2053–2068 Ohta Y, Okada S, Toda I, Ike H (1992) Scanning electron microscopic studies of the oral mucosa and its microvasculature: a review of the palatine mucosa and its microvascular architecture in mammals. Scanning Microsc 6: 463–474 Oudar O (2000) Spheroids: relation between tumour and endothelial cells. Crit Rev Oncol Hematol 36: 99–106 Pittman RN (1995) Influence of microvascular architecture on oxygen exchange in skeletal muscle. Microcirculation 2: 1–18 Puri V, Watanabe R, Singh RD, Dominguez M, Brown JC, Wheatley CL, Marks DL, Pagano E (2001) Clathrin-dependent and -independent internalization of plasma membrane sphingolipids initiates two Golgi targeting pathways. J Cell Biol 154: 535–547 Risbud MV, Karamuk E, Moser R, Mayer J (2002) Hydrogel-coated textile scaffolds as three-dimensional growth support for human umbilical vein endothelial cells (HUVECs): possibilities as coculture system in liver tissue engineering. Cell Transplant 11: 369–377 Rondaij MG, Bierings R, Kragt A, van Mourik JA, Voorberg J (2006) Dynamics and plasticity of Weibel–Palade bodies in endothelial cells. Arterioscler Thromb Vasc Biol 26: 1002–1007 Rondaij MG, Bierings R, van Agtmaal EL, Gijzen KA, Sellink E, Kragt A, Ferguson SS, Mertens K, Hannah MJ, van Mourik JA, Fernandez-Borja M, Voorberg J (2008) Guanine exchange factor RalGDS mediates exocytosis of Weibel–Palade bodies from endothelial cells. Blood 112(1): 56–63 Sadler JE (2005) New concepts in von Willebrand disease. Annu Rev Med 56: 173–191 Scherberich A, Beretz A (2000) Culture of vascular cells in tridimensional (3D) collagen: a methodological review. Therapie 55: 35–41 Schulz E, Jansen T, Wenzel P, Daiber A, Munzel T (2008) Nitric oxide, tetrahydrobiopterin, oxidative stress, and endothelial dysfunction in hypertension. Antioxid Redox Signal 10: 1115–1126 Sehgal PB, Mukhopadhyay S (2007) Dysfunctional intracellular trafficking in the pathobiology of pulmonary arterial hypertension. Am J Respir Cell Mol Biol 37: 31–37

534

*

€ller and A. Ellinger J. Neumu

Stein GH, St Clair JA (1988) Human microvascular endothelial cells: coordinate induction of morphologic differentiation and twofold extension of life span. In Vitro Cell Dev Biol 24: 381–387 Stolz DB, Jacobson BS (1991) Macro- and microvascular endothelial cells in vitro: maintenance of biochemical heterogeneity despite loss of ultrastructural characteristics. In Vitro Cell Dev Biol 27A: 169–182 Titani K, Kumar S, Takio K, Ericsson LH, Wade RD, Ashida K, Walsh KA, Chopek MW, Sadler JE, Fujikawa K (1986) Amino acid sequence of human von Willebrand factor. Biochemistry 25: 3171–3184 Tscheudschilsuren G, Aust G, Nieber K, Schilling N, Spanel-Borowski K (2002) Microvascular endothelial cells differ in basal and hypoxia-regulated expression of angiogenic factors and their receptors. Microvasc Res 63: 243–251 Valentijn KM, Valentijn JA, Jansen KA, Koster AJ (2008) A new look at Weibel–Palade body structure in endothelial cells using electron tomography. J Struct Biol 161: 447–458 Van Mourik JA, Romani de Wit T, Voorberg J (2002) Biogenesis and exocytosis of Weibel–Palade bodies. Histochem Cell Biol 117: 113–122 Vischer UM, Wagner DD (1993) CD63 is a component of Weibel–Palade bodies of human endothelial cells. Blood 82: 1184–1191 Voorberg J, Fontijn R, van Mourik JA, Pannekoek H (1990) Domains involved in multimer assembly of von willebrand factor (vWF): multimerization is independent of dimerization. Embo J 9: 797–803 Watkins MT, Al-Badawi H, Russo AL, Soler H, Peterson B, Patton GM (2004) Human microvascular endothelial cell prostaglandin E1 synthesis during in vitro ischemiareperfusion. J Cell Biochem 92: 472–480 Weibel ER, Palade GE (1964) New cytoplasmic components in arterial endothelia. J Cell Biol 23: 101–112 Weston GC, Cann L, Rogers PA (2003) Myometrial microvascular endothelial cells express oxytocin receptor. Bjog 110: 149–156 Yano K, Gale D, Massberg S, Cheruvu PK, Monahan-Earley R, Morgan ES, Haig D, von Andrian UH, Dvorak AM, Aird WC (2007) Phenotypic heterogeneity is an evolutionarily conserved feature of the endothelium. Blood 109: 613–615 Zenner HL, Collinson LM, Michaux G, Cutler DF (2007) High-pressure freezing provides insights into Weibel–Palade body biogenesis. J Cell Sci 120: 2117–2125 Zhang Q, Church JE, Jagnandan D, Catravas JD, Sessa WC, Fulton D (2006) Functional relevance of Golgi- and plasma membrane-localized endothelial NO synthase in reconstituted endothelial cells. Arterioscler Thromb Vasc Biol 26: 1015–1021

Formation of mucin granules

*

535

Formation of mucin granules Juan Perez-Vilar

In this chapter, the formation of the mucin granule, the characteristic secretory granule found in mucus/goblet cells, is discussed in view of our knowledge on the structure and biosynthesis of gel-forming mucins, granule intraluminal organization and the theoretical framework provided by the main sorting/trafficking models thought to operate in the secretory pathway. Contrary to most reviews on intracellular trafficking, a soluble protein cargo, rather than membrane components, occupies the central stage of our discussion. Considering the large sizes and properties of mucin precursors, it is very difficult not to envision an active role for these macromolecules during its intracellular trafficking and storage.

Introduction The mucosa of the gastrointestinal, urogenital, respiratory, auditory and visual systems, the gill in fishes and the epidermis in amphibians are covered by a gel-like, viscous secretion known as mucus. Mucus not only protects against noxious, chemical and biological, agents but also lubricates and hydrates the underlying epithelia. Although mucus is a complex, aqueous mixture of proteins, ions and other components, a family of structurally related glycoproteins, known as gel-forming mucins, is largely responsible for mucus viscoelastic and adhesive properties (Dekker et al. 2002; Perez-Vilar and Hill 1999, 2004). Proteins structurally similar to animal and human gelforming mucins are also found in invertebrates (Lang et al. 2007), which highlights the functional importance of these glycoproteins during metazoan evolution. Indeed, mucin gene expression and secretion, mucus composition and mucus/goblet cell differentiation are subjected to regulation by a broad range of luminal/epithelial conditions/factors, including bacterial and viral infection, presence of toxic chemicals and inflammatory mediators, mechanical stimuli, etc. It is not surprising, therefore, that altered mucin expression and secretion impede normal epithelia function and are well known contributors to important human pathologies. For example, increased mucus production and obstruction of small airways are observed in patients with chronic obstructive pulmonary disease, bronchitis, asthma or cystic fibrosis (e. g., Rose and Voynow 2006). Considering the large sizes and biochemical/biophysical complexity of gelforming mucins, it is unavoidable to wonder on the mechanisms that make possible mucin intracellular transport and storage. The formation of the

536

*

J. Perez-Vilar

0.5–2.0 mm mucin granule is virtually an unexplored field and subject of continuous speculation mostly based on data obtained in other cell types. Indeed, the same views and presumptions of 13 years ago (e.g., Forstner 1995) can be found essentially unaltered in current reviews. This is not surprising though, since working with mucins, and mucin-producing cells, is technically very challenging and time consuming. Nevertheless, our understanding on the biosynthesis and structure of mucin macromolecules is deep enough to make reasonable and experimentally testable speculations regarding to the fate of intracellular mucins. Moreover, a novel experimental approach has allowed a direct exploration of the mucin granule lumen in living goblet cells and to gain some information on its organization (PerezVilar 2007). One of the author’s aims in writing this chapter is to bring to the reader’s attention the notion that the properties of predominant cargo proteins, at least those as large as mucins, must be considered for a full understanding of the morpho-functional organization and dynamic of the organelles involved in the secretory pathway.

Goblet/mucus cells Mucins are synthesized in and secreted from goblet and mucus cells located in the surface epithelia and submucosal glands, respectively. In the mucus/goblet cell, numerous, closely apposed 0.5–2 mm wide mucin granules occupy most of the apical cytoplasm, leaving the nucleus constrained in the basal pole of the cell (for reviews see Neutra et al. 1984; Specian and Olivier 1991). Other organelles and cytoskeletal elements fill the intergranular space, although the Golgi complex and endosomes/ lysosomes largely reside in the periphery of the granular mass (e.g., Shimomura et al. 1996; Perez-Vilar et al. 2005a, b). A layer of microtubule/filament-rich cytoplasm encircles the granule mass, forming a cuplike structure known as theca, while actin filaments separate the granule mass from the apical membrane (Neutra et al. 1984; Specian and Olivier 1991). Allegedly, this compact organization helps to extrude large numbers of granules during massive compound exocytosis, which involves fusion of granules between themselves and with the plasma membrane, leading to the formation of the typical “omega” structures viewed with the electron microscope (Neutra et al. 1984). However, more recent studies support the existence of apocrine- and merocrine-type mechanisms of secretion (e.g., Puchelle et al. 1991; Lethem et al. 1993; PerezVilar et al., 2005a). Albeit recent promising developments (e.g., Li et al. 2001), our current knowledge on the mechanisms regulating mucin (constitutive and regulated) secretion still is fragmented and largely based on indirect evidence or notions demonstrated in other cell types with regulated secretion but not yet corroborated, or convincingly demonstrated, in goblet cells.

Formation of mucin granules

*

537

Gel-forming mucins Biochemical properties There are six main biochemical properties of gel-forming mucins that must be considered when trying to understand the intracellular transport and storage of mucins (Perez-Vilar and Hill 1999). First, mucins are amongst the largest proteins known with thousands of amino acid residues per monomer and molecular weights in the millions. Second, mucins are organized into different protein domains (Fig. 1A). Some of these domains can be found in other protein families, not functionally related to mucins. Third, mucins have prominent, centrally located and highly glycosylated domains, known as the mucin domains, O-glycosylated domains or TRD, which consist of mostly tandemly repeated sequences rich in serine and threonine residues. The latter are covalently bound to mucin-type O-linked oligosaccharides, which are very diverse and provide mucins with the capability to bind many microorganisms and chemical compounds (e.g., Roussel and Lamblin 2003; Spiro 2002). Moreover, the high number of O-glycan chains in the mucin domains constrains the free rotation of peptide bonds, which explains the rather extended structure and large hydrodynamic sizes of mucins (e.g., Gerken 1993; Yakubov et al. 2007). The lack of stable secondary structure (e.g., Eckhardt et al. 1987) plus the high density of O-glycans in the mucin domains are the two critical factors that make mucins highly flexible, random coils macromolecules suited for forming mucus gels. Fourth, O-linked oligosaccharides in mucins are typically sialylated and sulfated (Spiro 2002, Brockhausen 2003), making mucins highly polyanionic proteins. The high density of negative charges also contributes to the stiffness of mucin polypeptides, although mucin extended conformation is largely determined by the GalNAc residues that initiate the mucin-type O-glycan chains (Gerken 1993). Most important, mucin negatively charged rests are important for the organization of the mucin granule (Perez-Vilar 2007) and likely its formation (see Sections “Mucin granule intralumenal organization” and “Mucins and the formation of mucin granules”). Fifth, gel-forming mucins are assembled into linear (see Note 1) disulfide-linked oligomers/multimers (Fig. 1B) (Perez-Vilar and Hill 1999; Perez-Vilar and Mabolo 2007). Formation of interchain disulfide bonds involves several highly conserved, cystine-rich protein domains that usually are N- but not O-glycosylated. Six, mucin solutions are molecularly polydisperse due to differences in the glycosylation pattern, the degree of oligomerization, proteolytic processing (e.g., Lidell et al. 2003) and the existence of different mucin gene alleles (e.g., Vinall et al. 2000).

Biophysical properties Two biophysical properties of mucins are of especial interest for our discussion (Verdugo 1990, 1991; Gerken 1993; Bansil et al. 1995; Hong et al. 2005; Bansil and Turner 2007; Perez-Vilar 2007). First, mucins oligomers/multimers in

538

*

J. Perez-Vilar

Figure 1. Different levels of structural organization in gel-forming mucins. From the biochemical point of view (panel A), mucin polypeptides have very complex multi-domain structures and glycosylation patterns, and thousands of amino acids per monomer. Shown are the schematic structures (not at scale) of two human mucins, MUC2 and MUC5AC. The monomers are assembled into disulfide-linked oligomers/multimers (panel B) with contour lengths of several microns. MUC5AC oligomers/multimers are basically very long, linear (see Note 1) and flexible random coils. Hence, in the physical sense (panel C), mucin oligomers/multimers can be defined as polymers of N freely-jointed Kuhn segments (see text for details). At this level, the biochemical features of mucins are secondary and, indeed, a number of properties can be described in terms of N. Mucin chains occupy large spherical volumes (V) and continually change from one conformation to another. As mucin concentration increases, the chains start to overlap and eventually form entangled networks or gel-like structures (panel D). The volume of polymeric gels can be modulated by phase transitions, which shrink or swell the gel upon changes in environmental parameters (pH, ions, temperature, etc). Phase transitions are likely exploited during mucin granule biogenesis and exocytosis (see text).

Formation of mucin granules

*

539

solution behave like extended, flexible random coil polymers (Fig. 1C). Accordingly, mucin chains in aqueous solutions have many possible conformations, which to a great extent determine mucin properties while in solution. The mucin concentration, the solvent, ion content, temperature, etc will restrict the number of conformations available to the mucin chain. Biophysical studies have shown that the minimal polypeptide size in a mucin chain that can be considered rigid, i.e., the Kuhn segment (l), is 30–60 nm. Moreover, the average diameter (d) of the mucin chain, which is directly proportional to the length of the O-linked oligosaccharide chains, is 7–15 nm. Therefore, a mucin oligomer/multimer chain in solution can be mathematically modeled by a chain comprising N freely jointed, cylindrical segments of length l and width d, where N ¼ l1 L, and L is the contour distance, i.e., the end-to-end length of the mucin oligomeric/multimeric chain. For example, a gel-forming mucin such as human MUC5AC has 5000 amino acid residues per monomer, up to 16 monomers per oligomer, l  0.03 mm, L  8 mm and N ¼ 8/0.03  266. Hence, such a MUC5AC disulfide-linked oligomer is a physical polymer with 266 topological subunits. The parameter N is very important since it determines the different conformations a chain can have, which itself determines the equilibrium and dynamic properties of the polymer. Moreover, it allows us to anticipate reasonably the topological issues that a mucin macromolecule creates during its intracellular transport and storage (see Perez-Vilar 2007) (see Sections “Biosynthesis and intracellular trafficking” to “Mucins and the formation of mucin granules”). Interestingly, while mucins and genomic DNA l values are within an order of magnitude, which suggest that the flexibility of these two macromolecules is similar, DNA N values are of course several orders of magnitude higher. Hence the need for a complex molecular machinery for packing/unpacking DNA molecules within the nucleus seems obvious. Second, due to mucin large sizes and flexibility, mucin chains begin to overlap and form entanglements at very low concentrations (Figs. 1D, 2). For instance, 16-mer MUC5AC oligomers would start to form entanglements when the total volume of the mucin chains is just over 0.8% of the total solution volume (the rationale behind this and other quantitative estimations given below can be found in Perez-Vilar 2007). Accordingly, mucin oligomers form entangled networks at the concentrations found in the native mucus (1–2%) and likely inside the cell. The mucin networks can be further stabilized by non-covalent interactions in environments such as the gastric lumen (e.g., Cao et al. 2005) or the mucin granule lumen (Perez-Vilar 2007) (see Section “Mucin granule intraluminal organization”). Mucin–mucin interactions via non-O-glycosylated domains, Ca2þ , other proteins, etc. are among the putative cross-linkers proposed. Despite the cross-linker involved, even a highly entangled polymeric network in water is by definition a (hydro) gel, i.e., has viscoelastic properties, which in the case of native or purified mucin solutions can range from a very viscous, gel-like but readily dispersible solution (e.g., saliva) to a semi-solid, elastic gel only dispersible by strong chaotropic

540

*

J. Perez-Vilar

Figure 2. Polymer entanglements. Above certain threshold concentration, large/flexible polymers overlap and form entanglements (see text for details). Shown are different types of simple entanglements between two polymeric chains (panel A). The entanglements constrain the movement of the polymeric chain (panel B). Cross-linkers such as Ca2þ , link proteins or direct mucin interactions can stabilize the entanglements and create new-interchain links (panel C) and eventually increase the degree of cross-linking and, accordingly, the elasticity of the network.

reagents (e.g., sputum). However, the formation of a dense tangled network does not necessarily result in the disappearance of the network interstitial space (“pores”). Indeed, large proteins can diffuse through the pores of the mucin networks found in mucus secretion (e.g., Olmsted et al. 2001, Shen et al. 2006). For instance, a highly concentrated mucin polymer solution in which 36% of the volume is occupied by the polymer chains will form a very dense entangled network with 10 nm-wide pores.

Gel volume (phase) transitions A critical property of a polymeric gel is its capability to undertake a reversible volume transition between a swollen and a collapsed phase (Fig. 1D) (Li and

Formation of mucin granules

*

541

Tanaka 1992). Volume gel phase transitions can be continuous or discontinuous pending on the solvent and environmental factors in which the polymers are embedded. In the latter case, the changes in gel volume can be massive (up to 1000-fold), although not necessarily rapid or sudden, and most important they result from small changes in the external conditions that affect the gel, including solvent, pH, temperature, ionic composition, etc. All these parameters influence the forces acting in the gel, including chain–chain affinity, osmotic pressure and rubber elasticity. In this respect the gel is able to “sense” changes in its surroundings and shrinks or swells accordingly. This property makes gels and phase transitions solid candidates to play a critical role during intracellular trafficking of flexible macromolecules such as mucins (Verdugo 1990, 1991; Perez-Vilar 2007), especially considering the luminal changes taking place along the cis–trans axis of the GC and later during mucin granule formation. Moreover, intracellular compartments gain or loose water by regulating the luminal osmotic pressure, i.e., pumping ions in/out of the compartments. Note that a gel volume phase transition is not a rare physical phenomenon but a universal property of gels, irrespective of its chemical nature (Li and Tanaka 1992). The first indication that a volume phase transition could be involved in mucin secretion was obtained by studies focused on mucin granule discharge. Thus, the swelling of the “nascent” mucus (i.e., granule lumen) upon mucin granule discharge could be visualized by video enhanced microscopy (Verdugo 1990, 1991). The kinetics of the process followed the same curve of the swelling of spherical synthetic gels when the solvent was changed (Tanaka and Filmore 1979), which in both cases is characterized by the relation: t

R2 D

ð1Þ

where t is the characteristic time of swelling of the gel, i.e., the time taken by the gel to reach a radius half the value of its final radius (R), i.e., once the gel is fully swollen, and D is the collective diffusion of the gel into the aqueous medium (see Note 2) (Tanaka and Filmore 1979). Since the swelling of these synthetic gels can be explaining by a volume phase transition, it was proposed that the same mechanism could explain mucin gel expansion during granule discharge (Verdugo 1990, 1991). The displacement of the intragranular Ca2þ , which cross-links and shields the polyanionic mucin chains in the granule lumen, by extra-granular Kþ or Naþ , both unable to cross-link the mucins, likely allows the repulsion among mucin chains, i.e., alters the chain–chain interactions, eventually resulting in gel swelling. Indeed, the mucus swelling kinetic showed a high Hill coefficient, which suggests it is a highly cooperative process. Because phase transitions are reversible, a volume phase transition could also explain the condensation of the mucin matrix during granule formation (see Section “Mucin and the formation of mucin granules”) (Perez-Vilar 2007).

542

*

J. Perez-Vilar

Biosynthesis and intracellular trafficking Endoplasmic reticulum Mucin monomers are synthesized in the endoplasmic reticulum (ER), where they are N-glycosylated and likely C-mannosylated, and form disulfide-linked dimers (Fig. 3) (Perez-Vilar and Hill 1999, Perez-Vilar and Mabolo 2007). Studies with deglycosylated mucins suggest mucin precursors in the ER have a smaller Kuhn segment (15 nm) (e.g., Gerken 1993, Hong 2005). Accordingly, these precursors would be more flexible and globular than its O-glycosylated counterparts in the GC, mucin granules or the mucus. Nevertheless, these rather globular precursors still are very large macromolecules with contour lengths in the mm range and end-to-end radius over a hundred nm (Fig. 3).

Figure 3. Biosynthesis of gel-forming mucins. Mucin biosynthesis is a sequential process that starts in the ER and ends in the TGC. In parallel to biochemical changes, the mucin precursors biophysical properties also change (see right column). See text for further details. #aa, number of amino acid residues; L, contour length; l, Kuhn segment, N, number of Kuhn segments; and R, end-to-end distance (a measure of the length of the chain in solution).

Formation of mucin granules

*

543

Two basic mechanisms have been proposed for explaining secretory protein exit from the ER (see Chapter 3.1 in this volume and references therein): (a) intraluminal concentration inside 60 nm-wide, COPII-coated vesicles, which bud from ER exit sites and; (b) accumulation into large (>0.5 mm long) ER domains, which are later separated from the ER. Whether 60 nm vesicular carriers are suitable for transporting mucin dimeric precursors is questionable in view of the large hydrodynamic sizes of these macromolecules. In a diluted aqueous medium, i.e., inter chain interactions are negligible, and without non-covalent intra chain interactions between the Kuhn segments (see Section “Biophysical properties”), the spherical volume (V) occupied by a polymer, a parameter known as the pervade volume, can be approximated by the following expression (Doi and See 1995):  4=5  p l d 1=5 N0:588 3 V 4

ð2Þ 2:45 3 where l, d and N are the Kuhn segment, chain width and number of Kuhn segments, respectively. For a 10,000 amino acid residue MUC5AC dimer (l  15 nm; L  1 mm; N  66), the corresponding R and V values would be 180 nm and 0.05 mm3, respectively. Note that the latter value is larger than the volume of a 60-nm ER vesicular carrier (0.0001 mm3). Hence, if mucins dimers are transported out of the ER by vesicular carriers, the mucin chains must be necessarily forming a highly condensed network. However, even in the most hypothetical favorable case scenario, i.e., assuming that inter and intra chain attractive forces are so strong that the aqueous medium is virtually excluded from the network interstitial space, it can be demonstrated, assuming a mucin chain is a cylinder of length L and width d, that less than a few dozens of unglycosylated mucin dimers could be packed within a single 60 nm COPII vesicle. Hence, in the absence of experimental proof against it, it seems reasonable to presume that mucin dimeric chains are transported out of the ER by large ER domains instead of vesicular carriers.

Golgi complex In the different compartments of the Golgi complex (GC), mucin dimers are Oglycosylated and sulfated, and once they reach the acidic trans-Golgi compartments (TGC), sialylated and assembled into disulfide-linked oligomers/multimers (Fig. 3) (Perez-Vilar and Hill 1999, Perez-Vilar and Mabolo 2007). As discussed in detail elsewhere (Perez-Vilar and Mabolo 2007), formation of disulfide inter-dimer bonds is a rare post-translational modification for two main reasons: (a) it requires an acidic pH, which does not favor the reaction mechanism leading to disulfide bond formation; and (b) it occurs in a compartment in which resident thiol/disulfide oxido/reductases are not present. Indeed, the available data suggest that formation of mucin inter-dimeric disulfide bonds in the trans-GC is a self-catalyzed process (Perez-Vilar and Hill 1998, 1999; Perez-Vilar and Mabolo 2007; Mabolo and Perez-Vilar, unpub-

544

*

J. Perez-Vilar

lished results). Proteolytic processing of the mucin chains may take place in the late GC compartments, although some studies suggest it occurs in the ER as well (e.g., Lidell et al. 2003, Lidell and Hansson 2006). Nevertheless, this proteolytic processing is cryptic, meaning that it is detected only when disulfide bonds are reduced, and hence it will not alter the size of the mucin precursors. In parallel to the biochemical changes, the polymeric properties of the mucin chains change as they traverse the different compartments of the GC (Fig. 3) (Perez-Vilar 2007). First, addition of GalNAc residues to the mucin Ser/ Thr-rich central regions limits the rotation of the peptide bonds, increasing the size of the Kuhn segments, which diminishes the flexibility of the mucin chains. Thus, a transition from a globular to a rather extended conformation takes place upon initiation of O-glycosylation. Second, the synthesis of larger, disulfide-linked macromolecules means a larger V per chain and, in consequence, an increase in the extent of entanglement. Third, as larger mucin chains are assembled, a significant increase in intraluminal viscosity would be expected as this property is directly proportional to N3. Fourth, extensive sulphation and sialylation of N- and O-linked oligosaccharide chains in mucins and, eventually, the conversion of mucins to highly hydrophilic, polyanionic macromolecules. The same topological considerations discussed above for the transport of mucin precursors out of the ER are valid for intra-Golgi transport (see Chapter 3.2 in this volume and references therein). Moreover, since extensive O-glycosylation results in mucin precursors with larger Kuhn segments, O-glycosylated mucin dimers would be less flexible and with larger pervade volumes that their non-O-glycosylated counterparts in the ER (Fig. 3). It appears unlikely, therefore, that intra-GC transport of mucin precursors is mediated by 50–60 nm vesicular carriers, as proposed by the shuttle vesicular model and derivatives (see Chapter 3.2). Similarly difficult to envision are intra-GC transport models based in the directional transport of cargo proteins through 50–60 nm continuities connecting adjacent GC cisternae. Hence, the intra-GC transport of mucin macromolecules can be better explained following the postulates of cisternal-maturation/progression-based models, although it is not currently possible to favor one of the derivatives of these models over the others.

Intraluminal diffusion As expected for a highly specialized secretory cell type, the ER in mucus/goblet cells is very extensive whereas the GC is well developed in different stacks with prominent cisternae each. Therefore, most likely both organelles are rich in mucin precursors as also suggested by the high intracellular signal intensities obtained with different antibodies against mucins and irrespective of the cytochemical method used (e.g., Deschuyteneer et al. 1988, Roth et al. 1994). Since a random coil of the size of mucin dimers would overlap when the chains occupy just 3% and 1%, in the case of unglycosylated and O-glycosylated macromolecules, respectively, it can be assumed that these chains are forming

Formation of mucin granules

*

545

entangled networks with gel-like, i.e., viscoelastic, properties. In the absence of strong non-covalent interactions, flexible polymers in an entangled network are moving continually, i.e., the network is highly dynamic. Indeed, the diffusion of individual mucin chains in mucus can be reasonably explained assuming the chains move by a reptation-like mechanism (Fig. 4) (De Gennes 1979). Mucin intraluminal diffusion through the entangled network must be considered critical to facilitate the interactions between mucin precursors and modifying enzymes. The reptation model predicts that intra-gel diffusion is inversely proportional to N2 and, hence, mucin precursors will diffuse faster in the ER, cis-GC and medial-GC compartments than the multimeric mucin chains in the transGC. In principle, mucin intraluminal diffusion would be important for three main reasons. First, it would make possible interactions between mucin precursors and modifying enzymes (glycosyltransferases, glycosidases, etc.), which usually are integral membrane proteins. Second, the intralumenal mucin networks would likely fill the aqueous lumen of the compartment in which they are confined, irrespective of its morphology. The basis for this assertion is the large hydrodynamic sizes of mucins, especially once they are fully O-glycosylated and sulfated, which make them highly hydrophilic, and the continuous diffusion of mucin chains. In this respect, mucin networks would be key components, at least quantitatively, of the intraluminal matrices of the ER and GC in mucus/Goblet cells. Third, mucin intraluminal networks, as any gel, would respond to chances in the luminal environment (e.g., pH, specific ions, etc.) by changing its volume.

Mucin chains and the lumen of the trans-Golgi complex compartments As secretory proteins in general and mucin precursors in particular are terminally modified in the TGC, there are gradual changes across the cis–trans axis affecting membrane and luminal components (see Chaps. 1.2 and 2.10–2.13 and references therein). Among the later, increased intraluminal [Ca2þ ] and [Hþ ] can be considered critical for three main interrelated reasons. First, these ions can reduce the strong repulsive forces generated by the “nascent” polyanionic macromolecules, preventing their hydration, the swelling of the entangled mucin network and eventually the swelling of the lumen. Not surprisingly, the swelling of TGC can be induced by simply incubating cells with compounds that increase the intraluminal pH such as monensin, ammonium chloride, etc. (e.g., Tartakoff 1983). Second, some gelforming mucins, e.g., gastric mucins, are able to interact non-covalently by pH-dependent protein–protein interactions (e.g., Burgoyne et al. 2003). Third, Ca2þ not only can shield negative charges but also cross-links mucin chains, stabilizing the entanglements and creating new cross-links. In this respect, it is important to mention that while Ca2þ -induced contraction of

546

*

J. Perez-Vilar

Figure 4. Mucin chain reptation. Inside an entangled network, chains of large, flexible polymers such as mucins are free to reptate, as shown schematically in this cartoon. Reptation is likely critical for the interaction between mucin precursors and modifying enzymes during mucin biosynthesis along the different compartments of the secretory pathway.

purified mucin macromolecules have been reported (e.g., Varma et al. 1990), proof of Ca2þ -driven mucin chain collapse still is controversial (Verdugo 1991). It has to be considered, however, that in most cases mucins had to be solubilized with chaotropic agents and usually relative low concentrations of mucins were used. Hence, conditions similar to the granule lumen (i.e., very high concentrations of mucins and Ca2þ ) have not been tested. In any case, purified native mucins interact in the presence of increasing [Ca2þ ] and gelforming mucins such as gastric mucins can aggregate at acidic pH (e.g., Cao et al. 1999, Hong et al. 2005). Since 2/3 of the amino acid residues in mucins are O-glycosylated, and accordingly negatively charged, the possibility of Ca2þ -mediated interactions along most of the polypeptide chains in different mucin oligomers, i.e., via a zipper-like mechanism, seems possible. As mentioned above, the intraluminal environment in the trans-Golgi would be expected to be quite viscous considering the length and charge density of the

Formation of mucin granules

*

547

mucin chains, which together would reduce the diffusion rate of the water molecules. Under this scenario, the lumen of the trans-GC in the mucus/Goblet cell would be more viscous and less diffusible than the lumen in the medialand cis-GC, and ER.

Mucin granule intralumenal organization The current model explaining the mucin granule matrix was first derived from studies focused on mucin granule discharge using video enhanced microscopy (reviewed in Verdugo 1990, 1991), and recently revised and extended by studies measuring the intragranular diffusion of a soluble protein marker under different experimental conditions (reviewed in Perez-Vilar 2007). Fundamental to our current understanding of mucin granule organization was the generation of goblet cells expressing a fluorescent protein that is sorted to and accumulated in the mucin granule (Fig. 6) (Perez-Vilar et al. 2005a, b). The fluorescent protein polypeptide comprises a NH2-terminal signal peptide, a central EGFP (Enhanced Green Fluorescence Protein) and the CK domain of MUC5AC in the COOH-terminal side. The accumulation of this protein tracer in the granule made possible to study the granule lumen by FRAP (Fluorescence Recovery After Photobleaching) (Perez-Vilar et al. 2005a, b, 2006). FRAP provides a conceptually simple and straightforward method to characterize organelles in live cells (Snapp et al. 2003). In a typical FRAP experiment, an area of interest is bleached at high laser power and the recovery of the fluorescence, i.e., the diffusion of the non-bleached molecules inside the bleached area, followed over time (Fig. 5) (Movie 1 in Supplemental file). Important information can be derived from the fluorescence recovery curves that are pertinent to the environment where the fluorescent tracer is confined. Since covalent binding between the fusion protein and the endogenous MUC5AC has not yet been proved, it is presumed that the recombinant protein is a soluble tracer for the lumen of the granule. The results obtained from the FRAP analyses in combination with results obtained by early video enhanced microscopy studies are consistent with three major conclusions (Fig. 7) (reviewed in Perez-Vilar 2007). First, the granule lumen comprises two components: (a) a pH-dependent immobile, condensed (mucin) matrix meshwork, which an average pore size of at least 5–10 nm; and (b) a very viscous, but diffusible, fluid phase in which the meshwork is embedded. Proteins and ions in the fluid phase can diffuse through the meshwork pores, and interact with the mucin chains. For example, a significant fraction of the intragranular pool of the tracer fusion protein was immobile while the other very slowly diffused. This differential distribution may reflect the stationary state of a dynamic interaction between the fusion protein and the mucin matrix. Thus, if the dissociation constant of this interaction is rather small, an immobile fraction of the fusion

548

*

J. Perez-Vilar

Figure 5. FRAP analysis of the mucin granule lumen using a confocal laser microscope. Schematic representation of a FRAP analysis. In brief, large (>1.1 mm) mucin granules are imaged in the equatorial plane under unsaturated emission conditions using a 63 /1.4 NA oil immersion objective. A circular area corresponding to the laser beam (0.5 mm) is irreversible bleached with a brief pulse (0.02–0.07 s) at 70% laser power (40 mW laser) and 100% transmission. The recovery of the fluorescence over time is visualized and recorded periodically (0.1–0.35 s) at 70% laser power and 0.5% transmission. Two of the three possible outcomes are shown: (a) no recovery of fluorescence, which would mean the fluorescent tracer intragranular pool is entirely immobilized; and (b) partial recovery of fluorescence, which is what is usually observed. From the FRAP curves, i.e., fluorescence emission (F) versus time (t) curves, it can be obtained two parameters: (a) the mobile fraction (Mf), which informs on the percent of the tracer that is able to diffuse; and (b) the characteristic diffusion time (t1/2), i.e., the time needed to reach 50% of the asymptotic F values, a parameter which is inversely proportional to the diffusion of the tracer across the bleached area. With the latter parameter, geometric boundaries and other empirical parameters it is possible to solve the pertinent diffusion equation and eventually obtain the apparent or effective coefficient of diffusion of the tracer (Deff) (see Perez-Vilar et al. 2005). Performing FRAP analyses under different experimental conditions can reveal changes in these parameters, which together can inform on the environment in which the tracer is embedded.

protein would be always present. Alternatively, if one or more saturable binding sites are involved in this interaction, an intragranular free pool of the fusion protein would permanently exist above certain concentration. In

Formation of mucin granules

*

549

Figure 6. HT29-18N2 cells expressing a fusion protein between EGFP and the MUC5AC-CKdomain. Shown is a confocal image with several goblet cells with the fluorescent protein accumulated in the mucin granules (green). The phase contrast image is also shown. n, nucleus.

any case, the mobile fusion protein appears to diffuse homogenously, suggesting that the mucin framework organization is similar along the granule lumen. Moreover, the fusion protein diffuses 200–300-fold faster in the ER lumen (Deff  2.4 mm2 s1 ) or the extracellular mucus (Deff  3 mm2 s1 ) than in the granule lumen (Deff  0.014 mm2 s1 ) (Perez-Vilar et al. 2005a; Mabolo and Perez-Vilar, unpublished observations). Assuming protein diffusion within a single plane, these Deff values mean that the time for one fusion protein molecule to move across the entire surface of a 2-mm granule would be in the same range that the time it would need for moving across the entire ER system. Considering that the ER lumen has a high concentration of proteins, the granule lumen appears be an extremely crowded and viscous environment with most of the available space occupied by the mucinmeshwork. In such an environment, it is difficult to envision the occurrence of enzymatic reactions consistent with the classical notion that secretory (mucin) granules are just storage organelles. It must be noted, however, that protein diffusion within cellular compartments is a complex process that not only is inversely proportional to the luminal viscosity and protein Stokes radius, and directly proportional to temperature as suggested by the Stokes–Einstein expression, but many other factors, including interactions, protein geometry, mass and charge, compartment geometry, etc. (e.g., Periasamy and Verkman 1998). Second, charge density (i.e., the degree of mucin/glycoprotein sialylation and sulfation) of the mucin matrix and other intragranular glycoproteins, and also the length of their O-glycan chains, determine the mobility

550

*

J. Perez-Vilar

Figure 7. Mucin granule intralumenal organization. Shown is a cartoon illustrating the organization of the mucin granule matrix as suggested by the FRAP studies. The granule matrix comprises a mucin meshwork in which the entangled mucin chains are mainly stabilized by interchain cross-links involving Ca2þ . The possibility of protein–protein interactions via nonglycosylated domains cannot be discarded at this moment though. The meshwork interstitial space (pores) are filled with Ca2þ and Hþ . In this soluble compartment, proteins (very slowly) diffuse and interact with protein-rich regions of the matrix he matrix. See text for further details.

Formation of mucin granules

*

551

of secretory proteins through the matrix pores (Perez-Vilar et al. 2006, Perez-Vilar 2007). Moreover, mucin O-glycans affect the accessibility of luminal proteins to the matrix protein-rich regions. To some extent these conclusions could have been anticipated on the basis that the granule lumen is dominated by a condensed mucin network and the O-glycosylated domains are predominant in mucin polypeptides (see Section “Biochemical properties”). Third, the structural integrity of the mucin disulfide-rich domains (i.e., D-, CS, C- and CK-domains) appears not to be essential to maintain the intragranular mucin meshwork (Perez-Vilar et al. 2006). Thus, disruption of mucin intra- and interchain disulfide bonds did not disorganize the intragranular matrix, which suggests this structure comprises rather an immobile mucin frame mainly maintained by interchain links involving the O-glycosylated regions (see Note 3 at the end of the text). Because the mucin O-glycosylated domains are highly polyanionic, any divalent or multivalent cationic compound would be an effective cross-linker. Ca2þ , which is found at a high concentration in the mucin granule lumen (e.g., Kuver et al. 2002, Chin et al. 2002), is presumed to be the cross-linker keeping together the mucin meshwork by binding to sialyl and sulfate rests in adjacent mucin chains (Verdugo 1991, Perez-Vilar 2007). In this respect, the intragranular matrix would be simply an entangled mucin network stabilized by acidic pH and high concentration of Ca2þ , which presumably cross-links mucin oligomeric/multimeric chains. The dynamic nature of an entangled mucin gel-like solution, shown by its capability to shrink/swell and to flow and anneal, provides a single physical mechanism to explain condensation of the matrix during mucin granule biogenesis (Perez-Vilar 2007), swelling during granule discharge (Verdugo 1990) and flowing/annealing during mucus formation (Perez-Vilar 2008). The meshwork model assumes that the intragranular gel-forming mucins are both the main secretory protein cargo and the critical structural component of the mucin granule lumen, i.e., the granule matrix. This assumption is supported by studies showing that mice lacking mMuc2, the major intestinal gel-forming mucin, have intestinal goblet cells with very small secretory granules, which contain another mucin granule-specific cargo protein (Velcich et al. 2002). Nevertheless, it must be noted that mucin expression and secretion are not sufficient to generate mucin granules since there are mucin-producing cells that do not have a goblet cell phenotype. Whether or not the rate of mucin gene expression is the critical factor for mucin granule formation is presently unknown. Cargo/matrix proteins can be found in certain cells (e.g., von Willebrand factor in Weibel-Palade bodies of endothelial cells, proteoglycans in chondrocytes, etc.) but not in others (e.g., hormones-secreting cells). In the latter, cargo proteins are usually intermingled with polyanionic matrix proteins (e.g., chromogranins) (Burgoyne and Morgan 2003). Since these matrix proteins are able to aggregate at acidic pH and high [Ca2þ ], it is possible that the same principles discussed for mucin granule lumen are also valid for the lumen in these secretory granules.

552

*

J. Perez-Vilar

Mucins and the mechanisms of secretory granule formation Doubtless, the TGC, and specifically the trans-Golgi network (TGN), are the most sophisticated cisternae of the GC in terms of protein/membrane trafficking. In polarized cells, the TGN is the place where specific proteins destined to the apical and basolateral domains of the plasmatic membrane, regulated and constitutive secretory granules/vesicles, and lysosomes/endosomes are properly sorted (see Chapters 3.3–3.5, 3.11 and 3.13 and references therein). Sorting of cargo proteins is the first step during the formation of regulated secretory granules, which has been proposed to proceed through one or more of three basic mechanisms: (a) Receptor-mediated aggregation of regulated proteins; (b) Ca2þ /pH-dependent selective aggregation of regulated proteins; and (c) Sorting out of non-regulated proteins in trans-Golgi/post-Golgi compartments (e.g., immature granules) back to their final destinations. The latter mechanism does not directly affect the luminal secretory protein since its molecular specificity is given by the luminal and membrane proteins that are recognized and transported to their final destinations. Although there are no current studies addressing this issue in goblet/mucus cells, note that this mechanism would be compatible with any of the other two sorting mechanisms that are discussed below. The first mechanism assumes the existence of specific receptors, perhaps, but not necessarily, exclusive for each cargo protein. This exquisite molecular refinement is also this mechanism main problem since it demands receptor concentrations in the range of the cargo protein concentrations besides adding a new level of complexity (i.e., a variety of receptors). A recent variant of this basic model comes from studies of prohormone secretion (e.g., Kim et al. 2006). This model suggests that different prohormones aggregate in the TGC with granulocytic proteins (e.g., granins), and that the resulting macromolecular aggregates will further concentrate in budding vesicles via specific receptors for the granulocytic proteins. Thus, expression of a variety of receptors in the pertinent cells is not required since only a few could manage the aggregation and sorting of different cargo proteins. Nevertheless, in the case of an entangled polymer network such as a mucin gel, it is difficult to envision how an interaction with membrane receptors can concentrate first and then pull out discrete parts of a heavily entangled gel. Indeed, in order to disperse and disentangle a polymeric entangled network, i.e., to overcome the energy that keeps the gel together, more solvent must be added, a significant mechanical force must be applied, or both. These requirements make receptor-based sorting mechanisms unlikely to operate alone during mucin granule formation. The second mechanism is by definition charge-dependent, which means that two different anionic proteins will aggregate and eventually divert to or become (immature) granules. The molecular specificity of such a mechanism is

Formation of mucin granules

*

553

necessarily low and presumably proportional to the density of negative charges in the corresponding secretory proteins, i.e., negatively charged amino acid residues, sialylated/sulfated rests in N- and O-glycans, or both. As discussed in the preceding sections, mucins in the TGC are transformed into polyanionic polymers and, accordingly, an interaction with increasing Ca2þ would be physically inevitable and not a mere suggestion. However, the question arises whether interchain cross-linking through Ca2þ is sufficient for mucin granule formation or, alternatively, sorting into the granule requires somehow further molecular recognition among the mucin chains. Several lines of evidence suggest the latter possibility could be correct. As discussed above (see Section “Mucin granule intralumenal organization”), protein–protein interactions are not absolutely required to maintain the granule organization (Perez-Vilar et al. 2006), which is consistent with the notion that inter-strand links via multivalent ions (Ca2þ ) are sufficient to hold the matrix together at acidic pH (Verdugo 1991, Perez-Vilar 2007). However, when the intragranular pH is raised, reduction of intragranular disulfide bonds resulted in granule matrix disorganization and partial discharge (Perez-Vilar et al. 2006), i.e., the links holding the meshwork at acidic pH were insufficient as the pH was raised. These results are consistent with a role for protein–protein interactions during the early stages of matrix assembly in the TGC lumen, which likely is less acidic than the granule lumen. As the granule lumen is increasingly acidified, the Ca2þ -dependent inter-strand links are quantitatively more important, implying that Ca2þ -dependent mucin–mucin interactions are stronger at acidic pH. Further studies aimed to characterize the underlying reasons for the accumulation of the EGFP-mucin CK-domain in the lumen of mucin granules showed that this fusion protein is only sorted to mucin granules and not to other regulated secretory granules, nor it is able to generate secretory granules in non-mucin producing cell lines (Mabolo and Perez-Vilar, unpublished observations). These observations suggest the CK-domain does not contain a self-sufficient sorting-signal. Hence, considering that mucin CKdomains are not O-glycosylated but just have a single N-glycan chain, the accumulation of this protein in mucin granules can only be explained if it interacts with the endogenous mucins during its intracellular trafficking. In contrast, our ongoing studies suggest that other non-O-glycosylated domains in the mucin polypeptides comprise sorting signals able to direct EGFP to mucin granules and at least another regulated secretory granule unrelated to the latter (Mabolo and Perez-Vilar, unpublished observations). Taken together the above results support the notion that somehow protein–protein interactions are needed during early stages of mucin granule formation.

Mucins and the formation of mucin granules We would like to propose a simplified phenomenological hypothesis for mucin granule formation in the TGC that takes into account the information we have reviewed in the preceding sections.

554

*

J. Perez-Vilar

Assumptions The hypothesis is built around two main assumptions. First, mucin precursors are post-translationally modified in the GC as individual cisternae mature following the cis-to-trans axis (see Section “Biosynthesis and intracellular trafficking”). This assumption does not favor a particular cisternal-maturation/progression mechanism (see Chapter 3.2 in this volume), but simply emphasizes the notions that mucin precursors are: (a) too large to be transported via 60 nm vesicular carriers or 60–100 nm inter-cisternal connections; (b) transported through the GC via entire cisterna or, alternatively, large cisternal domains; and (c) modified as the polarity of the GC is also generated, which means that in each cisterna a mixture of different mucin precursor species are present. Although the hypothesis can be framed assuming mucin flow via entire cisternae or, alternatively, via large cisternal carriers or domains (e.g., tubules), only the latter will be discussed here. A critical review on this and related cisternal maturation/progression models can be found in the literature (Mironov et al. 2005) or elsewhere in this volume (see Chaps. 3.1 and 3.2). By assuming that mucin precursors are transported via large cisternal domain carriers, it is implied the existence of fusion/fission events of these carriers with/from the donor/acceptor cisternae (Fig. 8). However, with respect to the mucin granule formation, either of the above two scenarios converges into the same hypothesis. The second assumption considers that mucin polymers are forming entangled networks within each cisterna and, accordingly, are subject to the physical principles governing the dynamics and solution properties of chemical polymers and polymeric networks (see Sections “Biophysical properties” and “Biosynthesis and intracellular trafficking”).

Hypothesis In its simplest form, the hypothesis defines a cisterna as a dynamic mixture of incoming, biochemically unmodified (in terms of the specific modifications pertinent to the compartment) mucin precursors and departing, biochemically modified mucin precursors. Similarly, within the framework provided by cisternal domain carrier-based maturation/progression models, it can be defined incoming and departing cisternal carrier domains that transport the corresponding mucin precursors (Fig 7). In each GC cisterna, or functional equivalent set of cisternae (as defined by a similar content of active enzymes, ion channels, etc.), at least two different but energetically stable intraluminal mucin phases would be present: one enriched in incoming mucin precursors and another in departing (modified) mucin precursors (Fig. 8). The underlying mechanism for such a luminal phase separation could be found in the polymeric nature of mucin chains and the fact that when two polymer solutions are mixed they will separate overtime into two different phases containing one polymer type each. This energy-driven separation process, known as phase-partitioning (e.g., Zaslavsky 1994), occurs even between polymers with minimal structural differences. Intra-

Formation of mucin granules

*

555

Figure 8. Flow of mucin precursors and maturation/progression of GC cisternae. Cartoon depicting the flow of mucin precursors in parallel with the progression of GC cisternae along the cis-to-trans axis. Each cisterna has an incoming domain (IC) and a departing domain (DP) where mucin precursors entry and exit the cisterna, respectively. The departing cisternal domains are detached from the exit sites and move to the next cisterna where the precursors are discharged following the fusion between the cisternal domain in transit and the acceptor cisterna. In each cistern, distinct phases enriched in incoming and outgoing mucin precursors can be found.

lumenal mucin precursors are highly polydisperse in terms of glycosylation, charge and/or length (see Section “Biosynthesis and intracellular trafficking”) (Fig. 3) and, hence, partition in different phases within a single cisterna is a reasonable possibility. Note that the unidirectional incoming flow of mucin precursors into the acceptor cisterna lumen would be important for two main reasons. First, it would prevent the modified precursors in the same compartments to flow back since the source of incoming protein is uninterrupted and the separated phases are energetically favored. Second, it will provide a gradient for the directional diffusion of the incoming mucin chains. It is amply accepted that protein transport within secretory compartments is largely, if not exclusively, attained by diffusion. However, diffusion is by definition a non-directional process, meaning that a protein will have the same probability to move to any direction, unless there is a concentration gradient. Protein diffusion is obviously limited in living cell compartments by geometrical (imposed by the

556

*

J. Perez-Vilar

compartment boundaries), charge and steric constrains, protein crowding, interactions, etc. But irrespective of these inherent constrains, under our hypothesis, there is always a concentration gradient of incoming mucin chains in the regions where the cisternal domain carriers fuse with acceptor cisternal membrane. Some kind of hydraulic force pushing the incoming macromolecules (perhaps driven indirectly by components of the cytoskeleton) could also be in place. In any case, the incoming mucin precursors would diffuse away following the concentration gradient since incoming and departing (modified) mucin precursors are different polymeric species. Moreover, as the incoming mucin chains are modified, they spend more time diffusing within the more energetically favorable, departing mucin phase, which also contribute to the amplitude of the gradient. In this way, a directional mucin flow within each cisterna is generated that goes from the site where the incoming cisternal domains are fused to the cisternal regions where the modified mucin precursors depart in cisternal domain carriers. Molecular differences between intraluminal mucin precursors would be especially notorious in the trans-Golgi cisternae. The incoming phase would be enriched in rather uncharged, entangled polyanionic network, comprising disulfide-linked dimers with N- and O-glycan chains not terminally modified. As the mucin chains are progressively sialylated and sulfate, their motilities would be reduced as the negatively charges are shielded by increasing [Hþ ] and [Ca2þ ], and the cross-linking degree among mucin chains increases by Ca2þ - and protein–protein-mediated interactions (see Sections “Mucin chains and the lumen of the trans-Golgi complex compartments” to “Mucins and the mechanisms of secretory granule formation”). Thus, an increasingly charged polyanionic mucin gel is assembled (see Section “Mucin chains and the lumen of the trans-Golgi complex compartments”) and eventually a welldefined gel phase would segregate form the incoming gel phase (Fig. 9). The changes in intraluminal Ca2þ and pH would make possible (discontinuous or partial) volume phase transitions (see Section “Biophysical properties”), which would result in mucin network condensation. Whether mucin gel condensation takes place gradually in different trans-Golgi compartments, individual compartments (e.g., the TGN), or post-Golgi compartments (e.g., immature granules) is unknown. Our early studies in living goblet cells co-expressing the mucin CK-domain-EGFP fusion protein and fusion protein between the sorting signals of galactosyltransferase and RFP did not reveal intraluminal concentration of the fusion protein in the GC cisternae where this enzyme is located (Perez-Vilar et al. 2005a, b). Moreover, similar studies with another trans-Golgi protein marker (chlatrin) also supported the notion that mucin condensation likely takes place in a post-Golgi compartment (Mabolo and Perez Vilar, unpublished observations) (pathway I in Fig. 9). In parallel to the phase partition of the luminal mucins and the condensation of the charged mucin gel, a partition of the membrane components

Formation of mucin granules

*

557

Figure 9. Intraluminal reorganization of mucin precursors during mucin granule formation. Schematic representation of the putative changes taking place in the TGC lumen during the early stages of mucin granule formation. As for the other cisternae, the incoming mucin precursors diffuse into the trans-cisternae where they are extensively modified. The structural differences (charge, length, etc.) between incoming and modified (departing) precursors result in their progressive separation into different, more energetically stable, phases. In the TGC, however, the luminal changes are relatively drastic, including the formation of stabilized polyanionic mucin networks, acidic environment, and increased amounts of Ca2þ . These luminal changes may create favorable conditions for the mucin network to condense, although not necessarily to collapse. Condensation may take place after the intraluminal phases are separated into different compartments (pathway I) or prior this separation occurs (pathway II). See text for further details. By cisternal maturation we understand all those processes that could modify the size (e.g., membrane retrieval or lipid synthesis) or the organization (e.g., formation of membrane subdomains) of GC cisternae.

558

*

J. Perez-Vilar

around the luminal phases would be expected, so that the fragmentation of the departing cisternal domain carrier could take place. From a luminal point of view, and irrespective of the recruitment of cytosolic factors needed for the formation of the carrier, the key question is whether the luminal mucin gradient somehow affects the flow of membrane proteins and ultimately the creation of a cisternal domain prone to be separated from the remaining cisterna. There are two tentative mechanisms that could contribute to create such transient membrane domains. First, some membrane mucins expressed in goblet cells (e.g., Stanley and Phillips 1999) have very long O-glycosylated domains facing the GC lumen that could bind via Ca2þ with the corresponding domains in luminal gel-forming mucins. Similarly, membrane proteoglycans with polyanionic glycosaminoglycans could also form such complexes with luminal mucins. Indirect proof suggests proteoglycans might be present in the mucin granule lumen (Perez-Vilar et al. 2006). Second, specific glycosyltransferases could be diverted to the departing cisternal domains through interactions with mucin precursor substrates. To envision this last model it is necessary to consider that: (a) O-glycosylated domains in mucins comprises thousands of Ser/Thr that are acceptor sites for mucin-type O-glycans (see Section “Biochemical properties”); and (b) glycosyltransferases (except for some core polypeptide glycosyltransferases) are (weak) lectins. Hence, if the intraluminal conditions are not optimal for the glycosyltransferase activity but the appropriate mucin substrate is present, relatively stable mucin–glycosyltransferases complexes could be assembled, especially considering the extended structure and the large number of potential binding sites per mucin chain. These lectin-type interactions could lead to the directional diffusion of specific glycosyltransferases within the membrane plane toward the departing cisternal domain. Once discharged into the (biochemically) optimal cisterna, the transfer of the specific monosaccharide would prevent the corresponding glycosyltransferase to further bind to the mucin chains. Therefore, the flow of mucin–glycosyltransferase complexes might contribute as well to the sorting of glycosyltransferases along the GC. Of course, such a mechanism would be a specific, rather than a general, sorting mechanism since all Golgi resident enzymes come from one source (the ER) and accordingly in certain, but not all, cisternae non-active enzymes encounter their specific substrates.

Conclusions Some of the intracellular mucin precursors have molecular weight in the millions. From a physical sense, mucin precursors are flexible polymers that are likely organized into entangled networks or gels. Mucin sizes and macromolecular organization make vesicular carrier-based trafficking models unlikely to play a role during the transport of mucin precursors along the GC and beyond. As mucin precursors traverses the different compartments of the secretory pathway, from the ER to the TGN, mucin biophysical and

Formation of mucin granules

*

559

biochemical properties change. Formation of polyanionic mucin gels takes place in the acidic compartments of the GC, which can be shielded and further cross-linked by intraluminal Ca2þ and protein–protein interactions. Volume phase transitions might be critical at this stage of mucin granule formation by driving mucin condensation. In parallel, intralumenal areas of the TGN rich in polyanionic, condensed mucin gel might be originated by phase partitioning. Current efforts from our laboratory are focused in obtained proof for intralumenal phase partitioning in living cells. If proved, this could be the first step to develop a more physically-oriented notion regarding protein and membrane flow along the secretory pathway. However, many fundamental questions regarding protein flow in general and mucin in particular still are unanswered. Of interest is the exquisite coupling of mucin flow and post-translational modifications, especially considering that most enzymes are type II membrane proteins and mucin chains are assembled into entangled gel-like solutions. Is mucin reptative diffusion the only mechanism to assure complete post-translational modification? What factors are recruited to the incoming and departing domains, including the one that will result in immature granules? Where volume transitions of the intralumenal gels do occur?, etc. Acknowledgements. The work in the author’s laboratory was supported by grants from the University of North Carolina Research Council, National Institute of Health (NIDDK) and the North American Cystic Fibrosis Foundation.

Notes. Note 1. Biophysical and electron microscopy studies are consistent with mucin forming linear oligomers/multimers such in the case of MUC5B, MUC5AC and MUC6. However, biochemical studies with recombinant mucins have provided strong evidence that pMUC19 (Perez-Vilar et al. 1998), the porcine counterpart of MUC19, and MUC2 (Gold et al. 2002) form branched oligomers/multimers. In both cases, however, the monomeric subunits are bound by linkages linking their respective NH- and COOH-terminal regions. Note 2. Contrary to the common belief that the expansion of a gel is due to the diffusion of the solute into the interstitial space, Tanaka and collaborators demonstrated that it is the gel that diffuses into the solute while swelling. Hence the need for a collective diffusion of the gel. The swelling starts in the outer regions and continues to the interior of the gel though. Note 3. The fact that a phosphine reducing compound was employed for these studies and that disulfide reduction was done at acidic pH minimizes the possibility of interchain disulfide shuffling, which further supports the notion that (Ca2þ -mediated) interchain interactions involving the mucin domains are sufficient to hold the granule mucin matrix. However, these results do not discard that protein–protein interactions among the non-O-glycosylated domains are present in unaltered granules. If this is the case, it might mean that intragranular mucin chains are organized following a certain geometrical, rather than a random, pattern. Since mucins are flexible polymers, this macromolecular organization would be within the realm of fractal, rather than Euclidean, geometry.

560

*

J. Perez-Vilar

References Bansil R, Stanley E, LaMont JT (1995) Mucin biophysics. Annu Rev Physiol 57: 635–657 Bansil R, Turner BS (2006) Mucin structure, aggregation, physiological functions and biomemical applications. Curr Opin Colloid Interf Sci 11: 164–170 Brockhausen I (2003) Sulphotransferases acting on mucin-type oligosaccharides. Biochem Soc Trans 31: 318–325 Burgoyne RD, Morgan A (2003) Secretory granule exocytosis. Physiol Rev 83: 581–632 Cao X, Bansil R, Bhaskar KR, Turner BS, LaMont JT, Niu N, Afdhal NH (1999) pHdependent conformational change of gastric mucin leads to sol–gel transition. Biophys J 76: 1250–1258 Chin WC, Quesada I, Nguyen T, Verdugo P (2002) Oscillations of pH inside the secretory granule control the gain of Ca2þ release for signal transduction in goblet cell exocytosis. Novartis Found Symp 248: 132–141 De Gennes PG (1979) Scalling concepts in polymer physics. Cornell University Press, USA Dekker J, Rossen JW, Buller HA, Einerhand AW (2002) The MUC family: an obituary. Trends Biochem Sci 27: 126–131 Deschuyteneer M, Eckhardt AE, Roth J, Hill RL (1988) The subcellular location of apomucin and nonreducing terminal N-acetylgalactosamine in porcine submaxillary glands. J Biol Chem 263: 2452–2459 Doi M, See H (1995) Introduction to polymer physics. Oxford University Press, USA Eckhardt AE, Timpte CS, Abernethy JL, Toumadje A, Johnson WC, Hill RL (1987) Structural properties of porcine submaxillary mucin. J Biol Chem 262: 11339–11344 Forstner G (1995) Signal transduction, packaging and secretion mucins. Annu Rev Physiol 57: 585–605 Gerken TA (1993) Biophysical approaches to salivary mucin structure, conformation and dynamics. Crit Rev Oral Biol Med 4, 261–270 Gold K, Johansson ME, Lidell ME, Morgelin M, Karlsson H, Olson FJ, Gum JR, Kim YS, Hansson GC (2002) The N terminus of the MUC2 mucin forms trimers that are held together within a trypsin-resistant core fragment. J Biol Chem 277: 47248–47256 Hong Z, Chasan B, Bansil R, Turner BS, Bhaskar KR, Afdhal NH (2005) Atomic force microscopy reveals aggregation of gastric mucin at low pH. Biomacromolecules 6, 3458–3466 Kim T, Gondre-Lewis MC, Arnaoutova I, Loh YP (2006) Dense-core secretory granule biogenesis. Physiology (Bethesda) 21: 124–133 Kuver R, Klinkspoor JH, Osborne WRA, Lee SP (2000) Mucus granule exocytosis and CFTR expression in gallbladder epithelium. Glycobiology 10: 149–157 Lang T, Hansson GC, Samuelsson T (2007) Gel-forming mucins appeared early in metazoan evolution. Proc Natl Acad Sci USA 104: 16209–16214 Lethem MI, Dowell ML, Van Scott M, Yankaskas JR, Egan T, Boucher RC, Davis CW (1993) Nucleotide regulation of goblet cells in human airway epithelial explants: normal exocytosis in cystic fibrosis. Am J Respir Cell Mol Biol 9, 315–322 Li Y, Tanaka T (1992) Volume phase transitions of gels. Annu Rev Mater Sci 22: 243–277 Li Y, Martin LD, Spizz G, Adler KB (2001) MARCKS protein is a key molecule regulating mucin secretion by human airway epithelial cells in vitro. J Biol Chem 276: 40982–40990 Lidell ME, Johansson ME, Hansson GC (2003) An autocatalytic cleavage in the C terminus of the human MUC2 mucin occurs at the low pH of the late secretory pathway. J Biol Chem 278: 13944–13951 Lidell ME, Hansson GC (2006) Cleavage in the GDPH sequence of the C-terminal cysteinerich part of the human MUC5AC mucin. Biochem J 399: 121–129 McCullagh CM, Jamieson AM, Blackwell J, Gupta R (1995) Viscoelastic properties of human tracheobronchial mucin in aqueous solution. Biopolymers 35: 149–159

Formation of mucin granules

*

561

Michaux G, Cutler DF (2004) How to roll and endothelial cigar: the biogenesis of WeibelPalade bodies. Traffic 5: 69–78 Mironov AA, eznoussenko GV, Polishchuk RS, Trucco A (2005) Intra-Golgi transport: a way to a new paradigm? Biochem Biophys Acta 1744: 340–350 Neutra MR, Phillips TL, Phillips TE (1984) Regulation of intenstinal goblet cells in situ, in mucosal explants and in the isolated epithelium. Ciba Found Symp 109: 20–39 Olmsted SS, Padgett JL, Yudin AI, Whaley KJ, Moench TR, Gone RA (2001) Diffusion of macromolecules and virus-like particles in human cervical mucus. Biophys J 81: 1930–1937 Perez-Vilar J, Hill RL (1998) Identification of the half-cystine residues in porcine submaxillary mucin critical for multimerization through the D-domains. Roles of the CGLCG motif in the D1 and D3-domains. J Biol Chem 273: 34527–34534 Perez-Vilar J, Hill RL (1999) The structure and assembly of secreted mucins. J Biol Chem 274: 31751–31754 Perez-Vilar J, Hill RL (2004) Mucin family of glycorpoetins. In: Lennarz WJ, Lane MD. Encyclopedia of Biological Chemistry. Academic Press/Elseview, New York, 2: 758–764 Perez-Vilar J, Eckhardt AE, DeLuca A, Hill RL (1998a) Porcine submaxillary mucin forms disulfide-linked multimers through its amino-terminal D domains. J Biol Chem 273: 14442–14449 Perez-Vilar J, Olsen JC, Chua M, Boucher RC (2005a) pH-dependent intraluminal organization of mucin granules in live human mucus/goblet cells. J Biol Chem 280: 16868–16881 Perez-Vilar J (2007) Mucin granule intraluminal organization. Am J Respir Cell Mol Biol 36: 183–190 Perez-Vilar J, Mabolo R (2007) Gel-forming Mucins. Notions from in vitro Studies. Histol Histopathol 22: 455–464 Perez-Vilar J, Mabolo R, McVaugh CT, Bertozzi CR, Boucher RC (2006) Mucin granule intraluminal organization in living mucus/goblet cells. Roles of protein post-translational modifications and secretion. J Biol Chem 281: 4844–4855 Perez-Vilar J, Ribeiro CM, Salmon WC, Mabolo R, Boucher RC (2005b) Mucin granules are in close contact with tubular elements of the endoplasmic reticulum. J Histochem Cytochem 53: 1305–1309 Periasamy N, Verkman AS (1998) Analysis of fluorophore diffusion by continuous distributions of diffusion coefficients application to photobleaching measurements of multicomponent and anomalous diffusion. Biophys J 75: 557–567 Plaisancie P, Ducroc R, El Homsi, Tosacas A, Guilmeau S, Zoghbi S, Thibaudeau O, Bado A (2006) Am J Physiol Gastrointest Liver Physiol 290: G805–G812 Puchelle E, Beorchia A, Menager M, Zahm JM, Ploton I (1991) Three-dimensional imaging of the mucus secretory process in the cryofixed frog respiratory epithelium. Biol Cell 72: 159–166 Rose MC, Voynow JA (2006) Respiratory tract mucin genes and mucin glycoproteins in health and disease. Physiol Rev 86: 245–278 Roth J, Wang Y, Eckhardt AE, Hill RL (1994) Subcellular localization of the UDP-Nacetyl-D-galactosamine: polypeptide N-acetylgalactosaminyltransferase-mediated O-glycosylation reaction in the submaxillary gland. Proc Natl Acad Sci USA 91: 8935–8939 Roussel P, Lamblin G (2003) The glycosylation of airway mucins in cystic fibrosis and its relationship with lung infection by Pseudomonas aeruginosa. Adv Exp Med Biol 535: 17–32 Sellers LA, Allen A, Morris ER, Ross-Murphy SB (1988) Submaxillary mucins. Intermolecular interactions and gel-forming potential of concentrated solutions. Biochem J 256: 599–607

562

*

J. Perez-Vilar

Shen H, Hu Y, Saltzman WM (2006) DNA diffusion in mucus: effect of size, topology of DNAs, and transfection reagents. Biophys J 91: 639–644 Shimomura S, Hisanmatsu K, Fujii Y, Ohno S (1996) An ultrastructural study of goblet cells in rat nasal mucosa as revealed by the quick-freezing method. J Anat 188: 651–659 Snapp E, Altan N, Lippincott-Schwartz J (2003) Fluorescence recovery after photobleaching. In: Bonifacino J, Dasso M, Harford J, Lippincott-Schwartz J, Yamada K, Morgan KS (eds) Current protocols in cell biology. Unit 21.1 John Wiley & Sons, Inc, New York Specian RD, Oliver MG (1991) Functional biology of intestinal goblet cells. Am J Physiol 260: 183–193 Spiro RG (2002) Protein glycosylation: nature, distribution, enzymatic formation, and disease implications of glycopeptide bonds. Glycobiology 12: 43R–56R Stanley CM, Phillips TE (1999) Selective secretion and replenishment of discrete mucin glycoforms from intestinal goblet cells. Am J Physiol 277: G191–G200 Strous GJ, Dekker J (1992) Mucin-type glycoproteins. Crit Rev Biochem Mol Biol 27: 57–92 Tanaka T, Filmore DJ (1979) Kinetics of swelling of gels. J Chem Phys 70: 1214–1218 Tartakoff AM (1983) Perturbation of vesicular traffic with the carboxylic ionophore monensin. Cell 32: 1026–1028 Varma BK, Demers A, Jamieson AM, Blackwell J, Jentoft N (1990) Light scattering studies of the effect of Ca2þ on the structure of porcine submaxillary mucin. Biopolymers 29: 441–448 Velcich A, Yang W, Heyer J, Fragale A, Nicholas C, Viani S, Kucherlapati R, Lipkin M, Yang K, Augenlicht L (2002) Colorectal cancer in mice genetically deficient in the mucin Muc2. Science 295: 1726–1729 Verdugo P (1990) Goblet cells secretion and mucogenesis. Annu Rev Physiol 52: 157–176 Verdugo P (1991) Mucin exocytosis. Am Rev Respir Dis 144: S33–S37 Vinall LE, Pratt WS, Swallow DM (2000) Detection of mucin gene polymorphism. Methods Mol Biol 125: 337–350 Yakubov GE, Papagiannopoulos A, Rat E, Easton RL, Waigh TA (2007) Molecular structure and rheological properties of short-side-chain heavily glycosylated porcine stomach mucin. Biomacromolecules 8: 3467–3477 Zaslavsky BY (1994) Aqueous two-phase partitioning. Physical chemistry and analytical applications. Mercel Dekker Inc, New York, NY

Golgi apparatus and epithelial cell polarity

*

563

Golgi apparatus and epithelial cell polarity Lessons from 3 decades of MDCK cells Sylvie Deborde, Diego Gravotta, Aparna Lakkaraju and Enrique Rodriguez-Boulan

The MDCK model, introduced three decades ago (Cereijido et al. 1978; Rodriguez-Boulan et al. 1978; Rodriguez-Boulan and Pendergast 1980), allowed the elucidation of biosynthetic and recycling routes of epithelial cells and the compartments that sort apical and basolateral proteins as the trans Golgi network (TGN) and recycling endosomes (RE). These compartments were originally believed to operate in separate biosynthetic and recycling routes but discoveries over the past decade have revealed that TGN and RE cooperate in biosynthetic protein sorting. TGN and RE display clathrindependent and clathrin-independent exocytic routes to the plasma membrane (PM), equivalent to the clathrin-dependent and clathrin-independent endocytic routes at the PM. In epithelial cells, clathrin-mediated exocytosis (CME) from TGN and RE is utilized only by basolateral PM proteins. With the exception of AP1B, which sorts a subgroup of basolateral proteins in RE, the clathrin adaptors and regulatory proteins involved in CME remain largely unknown. The clathrin-independent routes from TGN and RE to the apical membrane remain enigmatic; their generation involves a clustering event mediated by proteinaceous motifs or by N- or O-linked carbohydrates, which may or may not promote lipid raft association, as well as the key participation of the MT and actin cytoskeletons.

Biosynthetic and recycling routes in polarized epithelial cells Figure 1 represents a diagram of the biosynthetic and recycling routes in MDCK cells (Rodriguez-Boulan et al. 2005). The observation that viral envelope glycoproteins behave as apical or basolateral markers in MDCK cells (Cereijido et al. 1978; Rodriguez-Boulan et al. 1978; Rodriguez-Boulan and Pendergast 1980) led to the discovery of the key role played by the Golgi apparatus in the sorting of apical and basolateral plasma membrane (PM) proteins and the identification of the TGN as a major sorting station in the secretory route (Fig. 1, routes 1 and 2) (Griffiths and Simons 1986; RodriguezBoulan and Powell 1992; Simons and Wandinger-Ness 1990). Several laboratories mapped the endocytic routes from apical and basolateral membranes in the 1980s. Their collective work demonstrated that apical and basolateral proteins are endocytosed into separate early endosomes, apical sorting endosomes (ASE) and basal sorting endosomes (BSE) (Fig. 1, routes 6 and 7) (Breitfeld et al. 1989; Mostov et al. 2003; Rodriguez-Boulan et al. 2005). At ASE and BSE, soluble proteins are sorted from membrane proteins and

564

*

S. Deborde et al.

Figure 1. Trafficking routes and sorting signals of apical and basolateral plasma membrane proteins in MDCK cells. Apical and basolateral routes are depicted in blue and red, respectively. Degradative route is shown in orange. Routes 1 and 2 are from TGN to plasma membrane; route 3 is from TGN to endosomes, lysosomes; routes 4 and 5 are from endosome to plasma membrane; routes 6 and 7 are endocytotic routes, from plasma membrane to endosome; route 8 is from TGN to CRE. ARE Apical recycling endosome, ASE apical sorting endosome, BSE basolateral sorting endosomes, CRE common recycling endosomes, LE late endosome, Lys lysosome, TGN trans-Golgi network.

transferred to common late endosomes (LE) for degradation in Lysosomes (Lys) (Bomsel et al. 1989) (Fig. 1, route 3). Apical and basolateral membrane proteins were shown to mix in common RE (Odorizzi et al. 1996; Brown et al. 2000) and to sort there into separate recycling routes to apical (route 4) and basolateral (route 5) membranes. An important outcome was the identification by Mostov, Apodaca, Dunn and coworkers of an intermediate, more alkaline endosomal compartment, the apical recycling endosome (ARE) (Apodaca et al. 1994; Brown et al. 2000) in the apical route. As research extended into other epithelial cells, variations in the polarized intracellular routes became evident. MDCK cells display characteristic “direct routes” from TGN to apical and basolateral membranes, but liver cells utilize a transcytotic route for most apical membrane proteins (Hubbard et al. 1989; Mostov et al. 2003), with the exception of certain polytopic transporters (Kipp

Golgi apparatus and epithelial cell polarity

*

565

and Arias 2000). Other epithelial cell types (e.g., intestine and retinal pigment epithelium) target apical proteins to variable extents via direct and transcytotic routes (Rodriguez-Boulan et al. 2005). The molecular mechanisms underlying these targeting variations are unclear; it has been proposed that the proteolipid MAL (VIP17) may be responsible for some of them in the liver (Ramnarayanan et al. 2007). By contrast, basolateral proteins utilize a direct route in all epithelial cell types studied to date.

Recycling endosomes and post-Golgi trafficking The early paradigm that dominated the epithelial polarity field for 20 years posited that sorting of apical and basolateral proteins occurs in different compartments in the biosynthetic and recycling routes, respectively the TGN and RE. The tubulo-vesicular conglomerates that form the TGN and RE are located very close to each other in the peri-centrosome area of epithelial cells (Fig. 2), as initially shown in non-polarized cells (Maxfield and McGraw 2004). This early paradigm has been replaced, over the past decade, by a new paradigm, according to which TGN and RE cooperate in the sorting of apical and basolateral proteins in the biosynthetic route. Reports from several laboratories have described a trans-endosomal route for several newly synthesized plasma membrane proteins, e.g., TfR (Cancino et al. 2007; Futter et al. 1995; Gravotta et al. 2007) asialoglycoprotein receptor (Leitinger et al. 1995), pIgR (Orzech et al. 2000), E-cadherin (Lock and Stow 2005) and more recently VSVG protein (Ang et al. 2004) (reviewed in Ellis et al. 2006). Certain apical proteins, such as the raft-associated apical marker influenza hemagglutinin have also been found to move through endosomes on the way to the PM (Cresawn et al. 2007). Recent work in collaboration with Alfonso Gonzalez’s laboratory has shown that certain newly synthesized basolateral plasma membrane proteins move quantitatively and with very fast kinetics from the TGN to RE (Fig. 1, route 8), where AP1B, a key basolateral sorting adaptor, sorts them to the basolateral membrane (Cancino et al. 2007; Gan et al. 2002; Gravotta et al. 2007). These experiments demonstrate a key role for RE in apical-basolateral protein sorting in the biosynthetic route. MDCK cells appear to shift from a trans-endosomal mode of delivery of basolateral proteins to a direct delivery route from TGN to PM as they establish full polarity (Gravotta et al. 2007), a process that may be partially controlled by Rab 8 (Henry and Sheff 2008).

Basolateral signals resemble endocytic motifs The discovery of apical and basolateral sorting signals (Fig. 1) in the late 1980s and 1990s constituted a major paradigm shift in the secretion field. Up to 1990, the dominant view was that transport from ER to the plasma membrane was by bulk flow, with specialized signal-mediated routes to divert hydrolases to the endolysosomal compartment (Pfeffer and Rothman 1987). Translated

566

*

S. Deborde et al.

Figure 2. Recycling endosomes and TGN in MDCK cells. Confocal image of MDCK cells microinjected with cDNA encoding NCAM–GFP and incubated at 20 C for 2 h for Golgi block. Ten minutes after release of Golgi block by transfer to 32 C, cells were stained for TGN38 and transferrin receptor (TfR), markers for the TGN and recycling endosomes, respectively. Note that the TGN38 label encircles a bright green spot (NCAM–GFP), colocalizing with TfR.

to epithelial cells, this view postulated that transport to the basolateral membrane was a default route, with only the apical route requiring specialized sorting information (Simons and Wandinger-Ness 1990). The discovery of apical and basolateral sorting signals dramatically altered this landscape, leading to a new paradigm (Rodriguez-Boulan et al. 2004). The new model, based on currently available evidence, posits that sorting of apical and basolateral proteins is mediated by hierarchically organized apical and basolateral signals, with the latter usually dominant (Keller and Simons 1997; Rodriguez-Boulan et al. 2005). The discovery of basolateral sorting signals in 1991 by four laboratories (Brewer and Roth 1991; Casanova et al. 1991; Hunziker et al. 1991; Le Bivic et al. 1991; Matter et al. 1992) demonstrated a striking similarity to endocytic motifs in their structure and topological localization to the cytoplasmic domain of basolateral proteins (Matter and Mellman 1994; Rodriguez-Boulan

Golgi apparatus and epithelial cell polarity

*

567

et al. 2005) (Fig. 1). Basolateral sorting signals currently known fall into five different groups: 1. Tyrosine motifs of the form Yxxf found in VSV G protein, LDL receptor (LDLR) and mutant forms of p75NTR and HA and NPXY, the endocytic signal of LDLR discovered by Brown and Goldstein, which doubles as a “weak” basolateral signal (Brewer and Roth 1991; Le Bivic et al. 1991; Matter et al. 1992). 2. Dileucine motifs (LL) with nearby acidic patches, found in Fc receptor (Hunziker and Fumey 1994) that also behave as endocytic signals. 3. Monoleucine motifs with nearby acidic patches, a novel basolateral sorting motif recently identified by our laboratory in CD147, a chaperone for lactate transporters (Deora et al. 2004). 4. PxxP motifs, such as found in EGF receptor (He et al. 2002). 5. Pleomorphic motifs that do not yet fall into established classes, e.g., N-CAM (Le Gall et al. 1997), transferrin receptor (TfR) (Odorizzi and Trowbridge 1997) and PIg-R (Aroeti and Mostov 1994). For transport proteins formed by two or more subunits, with the transporting subunit usually a multispan transmembrane protein, the sorting signals have only started to be elucidated (reviewed by Muth and Caplan 2003).

Clathrin and clathrin adaptors are key regulators of basolateral polarity The biological rationale for the similarity of basolateral and endocytic motifs was recently explained by the discovery that clathrin plays a key role in basolateral sorting (Deborde et al. 2008). Clathrin depletion in MDCK cells, mediated by RNA interference (RNAi), resulted in selective loss of polarity of basolateral proteins with practically all known types of basolateral signals (Fig. 3). Biochemical assays demonstrated that clathrin-depleted MDCK cells missorted basolateral proteins in the biosynthetic and recycling routes. Live imaging assays demonstrated that chronic clathrin depletion by RNAi (Fig. 3) or acute cross-linking of clathrin by a bi-functional reagent slowed down the exit of basolateral proteins, but not apical proteins, from the TGN and promoted missorting of basolateral proteins into apical carrier vesicles. Thus, an important task for the future is to identify the clathrin adaptors that participate in the sorting of the various classes of basolateral proteins. Clathrin operates in tandem with approximately twenty clathrin adaptors (Bonifacino and Traub 2003; Kirchhausen 1999, 2000; Owen et al. 2004), with characteristic subcellular distributions, clathrin and cargo binding sites, as well as binding sites for membrane phosphoinositides. The AP subfamily is hetero-tetrameric and includes the plasma membrane adaptor AP2 (a, b2, m2, s2), and three TGN/endosomal adaptors: AP1 (g, b1, m1, s1), AP3 (d, b3, m3, s3) and AP4 (e, b4, m4, s4). Yeast 2 hybrid (Ohno et al. 1995) and yeast

568

*

S. Deborde et al.

Figure 3. Clathrin knock-down depolarizes basolateral proteins and delays their exit from Golgi complex. A Confocal microscopy of control and clathrin-depleted cells. B Domain-selective biotinylation of control and clathrin-depleted cells. MDCK cells were confluent on filters for at least 3 days after treatment with luciferase (LF) or clathrin heavy chain (CHC) siRNAs. Some markers were endogenous (TfR (b), E-cadherin, Naþ , Kþ -ATPase b subunit (Na K ATPase) and gp135). NCAM and CD147 were permanently transfected. VSVG and human TfR (a) were expressed by adenoviral infection. Bl, basolateral; Ap, apical. C Clathrin depletion by siRNA inhibits the exit of VSVG–GFP from the TGN. Experiments were performed in subconfluent MDCK cells expressing sialyl transferase–RFP (ST–RFP) after a shift from 20 C to 32 C. D Quantification of the exit of NCAM–GFP, VSVG–GFP and p75–GFP in control and clathrindepleted cells. Perinuclear GFP was quantified as a fraction of the total cell GFP. Data are shown as Mean  S.E.M. n > 10; at least two experiments per condition. (From Deborde et al. Nature 2008.)

Golgi apparatus and epithelial cell polarity

*

569

3 hybrid assays (Janvier et al. 2003) demonstrated that tyrosine motifs of the Yxxf type (endocytic and basolateral motifs) interact with the m subunit of AP adaptors whereas dileucine motifs of the type [DE]XXXL[LI], found in lysosomal integral membrane proteins (LIMPs), bind to g and s1 subunits of AP1 and to d and s3 subunits of AP3. The GGA subfamily of clathrin adaptors (for Golgi-localized, Gamma-ear containing, Arf-binding) is composed of monomeric adaptors with homology to the g subunit of AP1 (Dell’Angelica et al. 2000; Hirst et al. 2000; Puertollano et al. 2001; Zhu et al. 2001). GGAs localize to TGN and endosomes and mediate transport of mannose 6-phosphate (Man6P) receptor out of TGN and trafficking of ubiquitinated cargo (Puertollano and Bonifacino 2004). Crystallographic studies have elucidated the details of the interaction of the m subunits of APs with tyrosine motifs and of GGAs with dileucine motifs in Man6P receptor (Bonifacino 2004; Owen et al. 2004). To date only two AP adaptors have been implicated in basolateral sorting, AP1B (Ohno et al. 1999) and AP4 (Simmen et al. 2002). The epithelial-specific adaptor AP1B has been quite extensively characterized, but the sorting role of AP4 remains unclear. AP1B differs from AP1A (formerly AP1) just in the possession of an epithelial-specific m1B subunit, instead of the ubiquitous m1A (Fig. 4). Caplan and coworkers (1998) made the puzzling discovery that the pig epithelial cell line LLC-PK1, believed to originate in the kidney proximal convoluted tubule (PCT) missorts several basolateral proteins to the apical surface. Mellman and coworkers (1999, 2001) showed that this was due to the absence of AP1B and could be corrected by transfection of m1B. Based on its EM localization, they proposed that AP1B participates in biosynthetic sorting of basolateral proteins at the Golgi complex (Folsch et al. 2001). However, our laboratory demonstrated that m1B localizes to endosomes and not to the TGN (Gan et al. 2002). Furthermore, using quantitative surface delivery assays in LLC-PK1 cells, we showed that AP1B does not control the biosynthetic route of LDLR; rather, AP1B regulates the post-endocytic recycling of LDLR and TfR to the basolateral membrane (Gan et al. 2002). More recently, however, we showed that knock-down of m1B disrupts the biosynthetic routes of TfR in recently confluent (1–2 days) MDCK cells but not in 4–5 days confluent monolayers (Gravotta et al. 2007). Experiments with a function-blocking antibody against m1B blocked the biosynthetic routes of TfR but not of LDLR at recycling endosomes in recently confluent FRT cells (Cancino et al. 2007). These results are consistent with a model in which AP1B sorts basolateral proteins in RE located at the cross-roads of the biosynthetic and recycling routes. They also show that some basolateral proteins follow an AP1B-independent route to the basolateral membrane which may reflect direct transport between TGN and the PM. The identities of the adaptors involved in this route are still unknown. The dual role of some basolateral signals as endocytic signals is extremely useful for the cell, as many basolateral proteins are fast recycling receptors involved in the internalization of essential metabolites from the blood (e.g.,

570

*

S. Deborde et al.

Figure 4. Knock-down of the clathrin adaptor AP1B in MDCK cells. A Diagrammatic representation of AP1A and AP1B, they differ only in the m subunit. B MDCK cells permanently knockeddown for m1B missort TfR apically. Confluent MDCK cells were infected with adenovirus encoding human TfR. Cells were fixed and immunostained for TfR, ZO-1 and b-catenin and examined by confocal microscopy. (From Gravotta et al., PNAS 2007.)

TfR and LDLR). As basolateral signals are usually dominant over apical signals, it is not surprising that most apical proteins have no efficient endocytic motifs and thus endocytose very slowly. A notable exception to this rule is megalin (Takeda et al. 2003), a highly endocytic apical protein in the PCT that performs a crucial role in the recovery by endocytosis of many proteins in the kidney ultrafiltrate.

Golgi apparatus and epithelial cell polarity

*

571

Apical trafficking: sorting signals, lipid rafts and motors Apical sorting signals are more heterogeneous than basolateral signals and may be found in the ectoplasmic, transmembrane or cytoplasmic domains of the protein (Rodriguez-Boulan et al. 2005) (Fig. 1). Our laboratory discovered the first apical sorting signal, the glycosylphosphatidyl inositol (GPI) anchor of GPI-anchored proteins, simultaneously with Brown and Rose (Brown et al. 1989; Lisanti et al. 1988, 1989; Powell et al. 1991). A second group of apical sorting signals includes the transmembrane domains of integral membrane proteins such as influenza HA and Neuraminidase (Kundu et al. 1996; Lin et al. 1998; Scheiffele 1997). A third group of apical signals includes luminal N-glycans (Fiedler and Simons 1995) and O-glycans (Yeaman et al. 1997). A fourth group includes cytoplasmic determinants such as, e.g., Rhodopsin, which displays binding sites for cytoplasmic dynein (Tai et al. 1999) and Megalin, which unusually appears to use tyrosine motifs as apical sorting signals (Marzolo et al. 2003; Takeda et al. 2003). The demonstration that GPI-anchors behave as apical sorting signals provided the first experimental support for the “raft hypothesis”, as GPIanchors associate with lipid microdomains or “rafts”, believed to a major mechanism for apical routing (Simons and Ikonen 1997; Simons and Van Meer 1988). How lipid rafts contribute to apical sorting remains controversial. Early observations from our laboratory (Hannan et al. 1993) (Fig. 5) and recent work (Paladino et al. 2004) suggest that the ability of GPI anchors to act as apical sorting signals depends on being clustered, although the natural clustering mechanism remains unknown (Rodriguez-Boulan et al. 2005; Helms and Zurzolo 2004; Simons and Vaz 2004). Clustering might promote interaction

Figure 5. Apical sorting of GPI-anchored proteins require a clustering event. A MDCK cells selected for resistance to the lectin concanavalin A (ConAR-MDCK) missort the GPI-anchored protein gD1-DAF (right panels), which is apical in control MDCK cells (left panels), as shown by a biotin polarity assay. B Fluorescence recovery after photobleaching (FRAP) shows that the population of gD1-DAF recently arrived at the PM from the Golgi complex is immobile in control MDCK cells but not in ConAR-MDCK. (Adapted from Hannan et al. (1992).)

572

*

S. Deborde et al.

with MT motors, directly or via specific adaptors. It has been shown that polarized epithelial cells have a non-centrosomal arrangement of vertical MT (Bacallao et al. 1989; Gilbert and Rodriguez-Boulan 1991; Musch 2004), whose disruption promotes selective mistargeting of apical membrane proteins (Gilbert et al. 1991; Van Zeijl and Matlin 1990). Some apical proteins are not associated with rafts, an example is p75 neurotrophin receptor (p75) which utilizes its O-glycans as an apical targeting signal (Yeaman et al. 1997). Like GPI-anchored proteins, p75 may require clustering for apical targeting, which is likely mediated by the lectin galectin 3 (Delacour et al. 2006, 2007). It is well established that p75 leaves the TGN in long tubules and vesicles that move along MT tracks (Kreitzer et al. 2000). Both microtubule and actin cytoskeletons appear to be involved in the exit of p75 vesicles from the TGN. Recently, the plus end kinesin family member KIF5B was identified as the MT motor involved in TGN exit of p75 in fully polarized MDCK cells; interestingly this kinesin is not involved in the exit of p75 from the TGN of non-polarized cells (Jaulin et al. 2007). KIF5B moves p75 carriers likely along a population of dynamic MT that emerge from the supranuclear MT organizing compartment (MTOC) towards the apical surface. Other data indicate that the Rho-family GTPase cdc42 selectively stimulates the exit of p75 from the TGN, whereas it inhibits the exit of basolateral proteins (Kroschewski et al. 1999; Cohen et al. 2001; Musch et al. 2001). Recent results indicate that a population of TGN actin filaments regulated by the downstream effectors of cdc42, LIMK1 and cofilin, is required for the formation and fission of p75 tubular transporters (Salvarezza et al. submitted). These actin filaments may be required for the fission of p75 carriers, a process mediated by dynamin 2 (Bonazzi et al. 2005; Kreitzer et al. 2000). Interestingly, the exit of basolateral proteins requires different fission proteins, e.g., BARS (Bonazzi et al. 2005; Kreitzer et al. 2000) and protein kinase D (Yeaman et al. 2004). Future work must establish the connection between p75 and the actin cytoskeleton: for instance, could the interaction between p75 and galectin 3 promote apical sorting via interaction with the actin cytoskeleton? In particular, it will be important to study the role of membrane curvaturesensing proteins containing BAR domains or N-terminal amphipathic helices, alone or in combination (N-BAR proteins), such as epsins, syndapins, sorting nexins, amphiphysin, endophilins and Arfaptin. Some members of this protein group localize to TGN (e.g., syndapins, Epsin R) but their role in protein exit pathways from TGN/endosomes and polarized trafficking has not been studied in detail.

Changes in post-Golgi routes of MDCK cells during polarization: hot spot for targeting of basolateral proteins Using high resolution live imaging protocols developed by our laboratory (Kreitzer et al. 2000) we showed that MDCK cells dramatically reorganize

Golgi apparatus and epithelial cell polarity

*

573

their targeting routes to the plasma membrane when they polarize. These studies visualized the exit of apical and basolateral proteins linked to GFP from the TGN in vesicular and tubular transport intermediates and their delivery and fusion with the apical and basolateral plasma membrane. Total internal reflection (evanescent) fluorescence microscopy (TIRFM), a technique that permits high resolution imaging of events within 0.1 mm of the basal membrane in contact with the coverslip demonstrated that subconfluent MDCK cells fuse apical and basolateral post-Golgi vesicles with the basal membrane. These experiments strongly suggested that apical and basolateral exocytic routes are targeted randomly to the entire plasma membrane of non-polarized MDCK cells. In striking contrast, in polarized MDCK cells, neither apical nor basolateral post-Golgi carriers fused with the basal membrane. Spinning disc confocal microscopy (SDCM, a fast, real-time type of confocal microscopy) demonstrated that post-Golgi basolateral carrier vesicles fused exclusively with a “hot spot” at the level of the junctional area (Fig. 6). In nice agreement with these data, we found that the basolateral t-SNARE syntaxin 4 was randomly localized to the entire cell surface in subconfluent MDCK cells, but was restricted to the lateral membrane after MDCK cells polarize (Kreitzer et al. 2003). By contrast, post-Golgi vesicles carrying the apical marker p75-GFP were never seen fusing with the lateral membrane or basal membranes in polarized MDCK cells, unless MT were

Figure 6. Basolateral proteins are delivered to a hot spot in the lateral membrane. Fusion sites of post-Golgi carriers for basolateral membrane proteins change with MDCK polarization. Nonpolarized MDCK cells expressing LDLR–GFP were studied by TIR-FM and SDCM to visualize the sites of fusion of post-Golgi vesicles with the plasma membrane. A TIRFM shows post-Golgi vesicles fusing with the basal surface in non-polarized MDCK cells. B In polarized MDCK cells, vesicles carrying LDLR–GFP fuse with the junctional region of the lateral membrane, as shown by the SDCM. (From Kreitzer et al. Nature Cell Biology, 2003.)

574

*

S. Deborde et al.

depolymerized with nocodazole, which promoted rapid randomization of the apical T-SNARE syntaxin 3 (Kreitzer et al. 2003). Our findings were generally confirmed by a recent study that used regular scanning confocal microscopy to follow the kinetics of arrival of apical (GFP–GPI) and basolateral (VSVG–GFP) markers to their respective membranes (Hua et al. 2006). While these authors confirmed our observation that basolateral proteins are delivered to the lateral membrane in polarized MDCK cells, they failed to observe preferential delivery to the junctional hot spot. This likely reflects the lower time and space resolution of the confocal live imaging techniques used in this study. The discovery of a “hot spot” for basolateral carrier vesicle fusion was very exciting, as it had been predicted by previous biochemical work in collaboration with James Nelson and Richard Scheller (Grindstaff et al. 1998). This work had shown that the peri-junctional area is a localization site for the exocyst, a multiprotein tethering particle that plays a key role in polarized secretion to the daughter cell in the budding yeast S. cerevisiae. Like the yeast secretory route, the basolateral route of epithelial cells also relies on the actin regulator cdc42, rabs, SNAREs and myosin (Kroschewski et al. 1999; Cohen et al. 2001; Musch et al. 2001). An additional ‘hot spot’ has been recently reported in the lower part of the lateral membrane for the basolateral protein TGF-alpha receptor and its associated adaptor Naked2 (Li et al. 2007). Interestingly, the targeting of this complex seems to be independent from AP1B and other previously described adaptors.

Future directions The MDCK model has been very useful for identifying polarized trafficking routes of plasma membrane proteins in epithelial cells. The discovery that clathrin is broadly involved in basolateral sorting opens the way to identifying clathrin adaptors involved in the sorting of specific subsets of basolateral proteins. The cooperation between TGN and RE needs to be further characterized in terms of both sorting signals and decoding mechanisms that mediate incorporation into apical or basolateral carrier vesicles. The clustering event that normally contributes to apical sorting needs to be further investigated. The crucial role of microtubules and microtubule motors in apical targeting requires the screening of the large population of kinesins expressed by epithelial cells for their participation in specific apical routes. Also important is to identify the mechanisms that dictate whether the same apical protein follows either a transcytotic or a direct route in different types of epithelial cells. Undoubtedly, the study of the molecular bases of epithelial polarity will remain very exciting for many years to come. Acknowledgements. This work was supported by NIH grants GM34107 and EY08538 to ERB, by funds from the Dyson Foundation and Research to Prevent Blindness Foundation to ERB, and by a National Research Service Award from NIH to AL.

Golgi apparatus and epithelial cell polarity

*

575

References Ang AL, Taguchi T, Francis S, Folsch H, Murrells LJ, Pypaert M, Warren G, Mellman I (2004) Recycling endosomes can serve as intermediates during transport from the Golgi to the plasma membrane of MDCK cells. J Cell Biol 167: 531–543 Apodaca G, Katz LA, Mostov KE (1994) Receptor-mediated transcytosis of IgA in MDCK cells is via apical recycling endosomes. J Cell Biol 125: 67–86 Aroeti B, Mostov KE (1994) Polarized sorting of the polymeric immunoglobulin receptor in the exocytotic and endocytotic pathways is controlled by the same amino acids. EMBO J 13: 2297–2304 Bacallao R, Antony C, Dotti C, Karsenti E, Stelzer EH, Simons K (1989) The subcellular organization of Madin-Darby canine kidney cells during the formation of a polarized epithelium. J Cell Biol 109: 2817–2832 Bomsel M, Prydz K, Parton RG, Gruenberg J, Simons K (1989) Endocytosis in filter grown Madin-Darby canine kidney cells. J Cell Biol 109: 3243–3258 Bonazzi M, Spano S, Turacchio G, Cericola C, Valente C, Colanzi A, Kweon HS, Hsu VW, Polishchuck EV, Polishchuck RS, Sallese M, Pulvirenti T, Corda D, Luini A (2005) CtBP3/ BARS drives membrane fission in dynamin-independent transport pathways. Nat Cell Biol 7: 570–580 Bonifacino JS (2004) The GGA proteins: adaptors on the move. Nat Rev Mol Cell Biol 5: 23–32 Bonifacino JS, Traub LM (2003) Signals for sorting of transmembrane proteins to endosomes and lysosomes. Annu Rev Biochem 72: 395–447 Breitfeld PP, Casanova JE, Simister NE, Ross SA, Mckinnon WC, Mostov KE (1989) Sorting signals. Curr Opin Cell Biol 1: 617–623 Brewer CB, Roth MG (1991) A single amino acid change in the cytoplasmic domain alters the polarized delivery of influenza viral hemagglutin. J Cell Biol 114: 413–421 Brown DA, Crise B, Rose JK (1989) Mechanism of membrane anchoring affects polarized expression of two proteins in MDCK cells. Science 245: 1499–1501 Brown PS, Wang E, Aroeti B, Chapin SJ, Mostov KE, Dunn KW (2000) Definition of distinct compartments in polarized Madin-Darby canine kidney (MDCK) cells for membranevolume sorting, polarized sorting and apical recycling. Traffic 1: 124–140 Cancino J, Torrealba C, Soza A, Yuseff MI, Gravotta D, Henklein P, Rodriguez-Boulan E, Gonzalez A (2007) Antibody to AP1B adaptor blocks biosynthetic and recycling routes of basolateral proteins at recycling endosomes. Mol Biol Cell 18: 4872–4884 Casanova JE, Apodaca G, Mostov KE (1991) An autonomous signal for basolateral sorting in the cytoplasmic domain of the polymeric immunoglobulin receptor. Cell 66: 65–75 Cereijido M, Robbins ES, Dolan WJ, Rotunno CA, Sabatini DD (1978) Polarized monolayers formed by epithelial cells on a permeable and translucent support. J Cell Biol 77: 853–880 Cohen D, Musch A, Rodriguez-Boulan E (2001) Selective control of basolateral membrane protein polarity by cdc42. Traffic 2: 556–564 Cresawn KO, Potter BA, Oztan A, Guerriero CJ, Ihrke G, Goldenring JR, Apodaca G, Weisz OA (2007) Differential involvement of endocytic compartments in the biosynthetic traffic of apical proteins. EMBO J 26: 3737–3748 Deborde S, Perret E, Gravotta D, Deora A, Salvarezza S, Schreiner R, Rodriguez-Boulan E (2008) Clathrin is a key regulator of basolateral polarity. Nature 452: 719–723 Delacour D, Cramm-Behrens CI, Drobecq H, Le Bivic A, Naim HY, Jacob R (2006) Requirement for galectin-3 in apical protein sorting. Curr Biol 16: 408–414 Delacour D, Greb C, Koch A, Salomonsson E, Leffler H, Le Bivic A, Jacob R (2007) Apical sorting by galectin-3-dependent glycoprotein clustering. Traffic 8: 379–388

576

*

S. Deborde et al.

Dell’Angelica EC, Puertollano R, Mullins C, Aguilar RC, Vargas JD, Hartnell LM, Bonifacino JS (2000) GGAs: a family of ADP ribosylation factor-binding proteins related to adaptors and associated with the Golgi complex. J Cell Biol 149: 81–94 Deora AA, Gravotta D, Kreitzer G, Hu J, Bok D, Rodriguez-Boulan E (2004) The basolateral targeting signal of CD147 (EMMPRIN) consists of a single leucine and is not recognized by retinal pigment epithelium. Mol Biol Cell 15: 4148–4165 Ellis MA, Potter BA, Cresawn KO, Weisz OA (2006) Polarized biosynthetic traffic in renal epithelial cells: sorting, sorting, everywhere. Am J Physiol Renal Physiol 291: F707–F713 Fiedler K, Simons K (1995) The role of N-glycans in the secretory pathway. Cell 81: 309–312 Folsch H, Ohno H, Bonifacino JS, Mellman I (1999) A novel clathrin adaptor complex mediates basolateral targeting in polarized epithelial cells. Cell 99: 189–198 Folsch H, Pypaert M, Schu P, Mellman I (2001) Distribution and function of AP-1 clathrin adaptor complexes in polarized epithelial cells. J Cell Biol 152: 595–606 Futter CE, Connolly CN, Cutler DF, Hopkins CR (1995) Newly synthesized transferrin receptors can be detected in the endosome before they appear on the cell surface. J Biol Chem 270: 10999–11003 Gan Y, McGraw TE, Rodriguez-Boulan E (2002) The epithelial-specific adaptor AP1B mediates post-endocytic recycling to the basolateral membrane. Nat Cell Biol 4: 605–609 Gilbert T, Le Bivic A, Quaroni A, Rodriguez-Boulan E (1991) Microtubular organization and its involvement in the biogenetic pathways of plasma membrane proteins in Caco-2 intestinal epithelial cells. J Cell Biol 113: 275–288 Gilbert T, Rodriguez-Boulan E (1991) Induction of vacuolar apical compartments in the Caco-2 intestinal epithelial cell line. J Cell Sci 100: 451–458 Gravotta D, Deora A, Perret E, Oyanadel C, Soza A, Schreiner R, Gonzalez A, RodriguezBoulan E (2007) AP1B sorts basolateral proteins in recycling and biosynthetic routes of MDCK cells. Proc Natl Acad Sci USA 104: 1564–1569 Griffiths G, Simons K (1986) The trans Golgi network: sorting at the exit site of the Golgi complex. Science 234: 438–443 Grindstaff KK, Yeaman C, Anandasabapathy N, Hsu SC, Rodriguez-Boulan E, Scheller RH, Nelson WJ (1998) Sec6/8 complex is recruited to cell–cell contacts and specifies transport vesicle delivery to the basal–lateral membrane in epithelial cells. Cell 93: 731–740 Hannan LA, Lisanti MP, Rodriguez-Boulan E, Edidin M (1993) Correctly sorted molecules of a GPI-anchored protein are clustered and immobile when they arrive at the apical surface of MDCK cells. J Cell Biol 120: 353–358 He C, Hobert M, Friend L, Carlin C (2002) The epidermal growth factor receptor juxtamembrane domain has multiple basolateral plasma membrane localization determinants, including a dominant signal with a polyproline core. J Biol Chem 277: 38284–38293 Helms JB, Zurzolo C (2004) Lipids as targeting signals: lipid rafts and intracellular trafficking. Traffic 5: 247–254 Henry L, Sheff DR (2008) Rab8 regulates basolateral secretory, but not recycling, traffic at the recycling endosome. Mol Biol Cell 19: 2059–2068 Hirst J, Lui WW, Bright NA, Totty N, Seaman MN, Robinson MS (2000) A family of proteins with gamma-adaptin and VHS domains that facilitate trafficking between the trans-Golgi network and the vacuole/lysosome. J Cell Biol 149: 67–80 Hua W, Sheff D, Toomre D, Mellman I (2006) Vectorial insertion of apical and basolateral membrane proteins in polarized epithelial cells revealed by quantitative 3D live cell imaging. J Cell Biol 172: 1035–1044 Hubbard AL, Stieger B, Bartles JR (1989) Biogenesis of endogenous plasma membrane proteins in epithelial cells. Annu Rev Physiol 51: 755–770

Golgi apparatus and epithelial cell polarity

*

577

Hunziker W, Fumey C (1994) A di-leucine motif mediates endocytosis and basolateral sorting of macrophage IgG Fc receptors in MDCK cells. EMBO J 13: 2963–2969 Hunziker W, Harter C, Matter K, Mellman I (1991) Basolateral sorting in MDCK cells requires a distinct cytoplasmic domain determinant. Cell 66: 907–920 Janvier K, Kato Y, Boehm M, Rose JR, Martina JA, Kim BY, Venkatesan S, Bonifacino JS (2003) Recognition of dileucine-based sorting signals from HIV-1 Nef and LIMP-II by the AP-1 gamma-sigma1 and AP-3 delta-sigma3 hemicomplexes. J Cell Biol 163: 1281–1290 Jaulin F, Xue X, Rodriguez-Boulan E, Kreitzer G (2007) Polarization-dependent selective transport to the apical membrane by KIF5B in MDCK cells. Dev Cell 13: 511–522 Keller P, Simons K (1997) Post-Golgi biosynthetic trafficking. J Cell Sci 110: 3001–3009 Kipp H, Arias IM (2000) Newly synthesized canalicular ABC transporters are directly targeted from the Golgi to the hepatocyte apical domain in rat liver. J Biol Chem 275: 15917–15925 Kirchhausen T (1999) Adaptors for clathrin-mediated traffic. Annu Rev Cell Dev Biol 15: 705–732 Kirchhausen T (2000) Three ways to make a vesicle. Nat Rev Mol Cell Biol 1: 187–198 Kreitzer G, Marmorstein A, Okamoto P, Vallee R, Rodriguez-Boulan E (2000) Kinesin and dynamin are required for post-Golgi transport of a plasma-membrane protein. Nat Cell Biol 2: 125–127 Kreitzer G, Schmoranzer J, Low S-H, Li X, Gan Y, Weimbs T, Simon S, Rodriguez-Boulan E (2003) Three-dimensional analysis of post-Golgi carrier exocytosis in epithelial cells. Nat Cell Biol 5: 126–136 Kroschewski R, Hall A, Mellman I (1999) Cdc42 controls secretory and endocytic transport to the basolateral plasma membrane of MDCK cells. Nat Cell Biol 1: 8–13 Kundu A, Avalos RT, Sanderson CM, Nayak DP (1996) Transmembrane domain of influenza virus neuramidase, a type II protein, posesses an apical sortin signal in polarized MDCK cells. J Virol 70: 6508–6515 Le Bivic A, Sambuy Y, Patzak A, Patil N, Chao M, Rodriguez-Boulan E (1991) An internal deletion in the cytoplasmic tail reverses the apical localization of human NGF receptor in transfected MDCK cells. J Cell Biol 115: 607–618 Le Gall AH, Powell SK, Yeaman CA, Rodriguez-Boulan E (1997) The neural cell adhesion molecule expresses a tyrosine-independent basolateral sorting signal. J Biol Chem 272: 4559–4567 Leitinger B, Hille-Rehfeld A, Spiess M (1995) Biosynthetic transport of the asialoglycoprotein receptor H1 to the cell surface occurs via endosomes. Proc Natl Acad Sci USA 92: 10109–10113 Li C, Hao M, Cao Z, Ding W, Graves-Deal R, Hu J, Piston DW, Coffey RJ (2007) Naked2 acts as a cargo recognition and targeting protein to ensure proper delivery and fusion of TGF-alpha containing exocytic vesicles at the lower lateral membrane of polarized MDCK cells. Mol Biol Cell 18: 3081–3093 Lin S, Naim HY, Rodriguez AC, Roth MG (1998) Mutations in the middle of the transmembrane domain reverse the polarity of transport of the influenza virus hemagglutinin in MDCK epithelial cells. J Cell Biol 142: 51–57 Lisanti M, Caras IP, Davitz MA, Rodriguez-Boulan E (1989) A glycophospholipid membrane anchor acts as an apical targeting signal in polarized epithelial cells. J Cell Biol 109: 2145–2156 Lisanti M, Sargiacomo M, Graeve L, Saltiel A, Rodriguez-Boulan E (1988) Polarized apical distribution of glycosyl phosphatidylinositol anchored proteins in a renal epithelial line. Proc Nat Acad Sci USA 85: 9557–9561 Lock JG, Stow JL (2005) Rab11 in recycling endosomes regulates the sorting and basolateral transport of E-cadherin. Mol Biol Cell 16: 1744–1755 Marzolo MP, Yuseff MI, Retamal C, Donoso M, Ezquer F, Farfan P, Li Y, Bu G (2003) Differential distribution of low-density lipoprotein-receptor-related protein (LRP)

578

*

S. Deborde et al.

and megalin in polarized epithelial cells is determined by their cytoplasmic domains. Traffic 4: 273–288 Matter K, Hunziker W, Mellman I (1992) Basolateral sorting of LDL receptor in MDCK cells: the cytoplasmic domain contains two tyrosine-dependent targeting determinants. Cell 71: 741–753 Matter K, Mellman I (1994) Mechanisms of cell polarity: sorting and transport in epithelial cells. Curr Opin Cell Biol 6: 545–554 Maxfield FR, McGraw TE (2004) Endocytic recycling. Nat Rev Mol Cell Biol 5: 121–132 Mostov K, Su T, ter Beest M (2003) Polarized epithelial membrane traffic: conservation and plasticity. Nat Cell Biol 5: 287–293 Musch A (2004) Microtubule organization and function in epithelial cells. Traffic 5: 1–9 Musch A, Cohen D, Kreitzer G, Rodriguez-Boulan E (2001) cdc42 regulates the exit of apical and basolateral proteins from the trans-Golgi network. EMBO J 20: 2171–2179 Muth TR, Caplan MJ (2003) Transport protein trafficking in polarized cells. Annu Rev Cell Dev Biol 19: 333–366 Odorizzi G, Pearse A, Domingo D, Trowbridge IS, Hopkins CR (1996) Apical and basolateral endosomes of MDCK cells are interconnected and contain a polarized sorting mechanism. J Cell Biol 135: 139–152 Odorizzi G, Trowbridge IS (1997) Structural requirements for basolateral sorting of the human transferrin receptor in the biosynthetic and endocytic pathways of MadinDarby canine kidney cells. J Cell Biol 137: 1255–1264 Ohno H, Stewart J, Fournier MC, Bosshart H, Rhee I, Miyatake S, Saito T, Gallusser A, Kirchhausen T, Bonifacino JS (1995) Interaction of tyrosine-based sorting signals with clathrin-associated proteins. Science 269: 1872–1875 Ohno H, Tomemori T, Nakatsu F, Okazaki Y, Aguilar RC, Foelsch H, Mellman I, Saito T, Shirasawa T, Bonifacino JS (1999) Mu1B, a novel adaptor medium chain expressed in polarized epithelial cells. FEBS Lett 449: 215–220 Orzech E, Cohen S, Weiss A, Aroeti B (2000) Interactions between the exocytic and endocytic pathways in polarized Madin-Darby canine kidney cells. J Biol Chem 275: 15207–15219 Owen DJ, Collins BM, Evans PR (2004) Adaptors for clathrin coats: structure and function. Annu Rev Cell Dev Biol 20: 153–191 Paladino S, Sarnataro D, Pillich R, Tivodar S, Nitsch L, Zurzolo C (2004) Protein oligomerization modulates raft partitioning and apical sorting of GPI-anchored proteins. J Cell Biol 167: 699–709 Pfeffer SR, Rothman JE (1987) Biosynthetic protein transport and sorting by the endoplasmic reticulum and Golgi. Annu Rev Biochem 56: 829–852 Powell SK, Cunningham BA, Edelman GM, Rodriguez-Boulan E (1991) Targeting of transmembrane and GPI-anchored forms of N-Cam to opposite domains of a polarized epithelial cell. Nature 353: 76–77 Puertollano R, Bonifacino JS (2004) Interactions of GGA3 with the ubiquitin sorting machinery. Nat Cell Biol 6: 244–251 Puertollano R, Randazzo PA, Presley JF, Hartnell LM, Bonifacino JS (2001) The GGAs promote ARF-dependent recruitment of clathrin to the TGN. Cell 105: 93–102 Ramnarayanan SP, Cheng CA, Bastaki M, Tuma PL (2007) Exogenous MAL reroutes selected hepatic apical proteins into the direct pathway in WIF-B cells. Mol Biol Cell 18: 2707–2715 Rodriguez-Boulan E, Kreibich G, Sabatini DD (1978) Spatial orientation of glycoproteins in membranes of rat liver rough microsomes. I. Localization of lectin-binding sites in microsomal membranes. J Cell Biol 78: 874–893 Rodriguez-Boulan E, Kreitzer G, Musch A (2005) Organization of vesicular traffic in epithelia. Nat Rev Mol Cell Biol 6: 233–247

Golgi apparatus and epithelial cell polarity

*

579

Rodriguez-Boulan E, Musch A, Le Bivic A (2004) Epithelial trafficking: new routes to familiar places. Curr Opin Cell Biol 16: 436–442 Rodriguez-Boulan E, Pendergast M (1980) Polarized distribution of viral envelope proteins in the plasma membrane of infected epithelial cells. Cell 20: 45–54 Rodriguez-Boulan E, Powell SK (1992) Polarity of epithelial and neuronal cells. Annu Rev Cell Biol 8: 395–427 Roush DL, Gottardi CJ, Naim HY, Roth MG, Caplan MJ (1998) Tyrosine-based membrane protein sorting signals are differentially interpreted by polarized Madin-Darby canine kidney and LLC-PK1 epithelial cells. J Biol Chem 273: 26862–26869 Scheiffele P, Roth MG, Simons K (1997) Interaction of influenza virus hemagglutinin with sphingolipid-cholesterol membrane rafts via its transmembrane domain. EMBO J 16: 5501–5508 Simmen T, Honing S, Icking A, Tikkanen R, Hunziker W (2002) AP-4 binds basolateral signals and participates in basolateral sorting in epithelial MDCK cells. Nat Cell Biol 4: 154–159 Simons K, Ikonen E (1997) Functional rafts in cell membranes. Nature 387: 569–572 Simons K, Van Meer G (1988) Lipid sorting in epithelial cells. Biochemistry 27: 6197–6202 Simons K, Vaz WL (2004) Model systems, lipid rafts, and cell membranes. Annu Rev Biophys Biomol Struct 33: 269–295 Simons K, Wandinger-Ness A (1990) Polarized sorting in epithelia. Cell 62: 207–210 Tai AW, Chuang JZ, Bode C, Wolfrum U, Sung CH (1999) Rhodopsin’s carboxy-terminal cytoplasmic tail acts as a membrane receptor for cytoplasmic dynein by binding to the dynein light chain Tctex-1. Cell 97: 877–887 Takeda T, Yamazaki H, Farquhar MG (2003) Identification of an apical sorting determinant in the cytoplasmic tail of megalin. Am J Physiol Cell Physiol 284: C1105–C1113 Van Zeijl MJAH, Matlin KS (1990) Microtubule perturbation inhibits intracellular transport of an apical membrane glycoprotein in a substrate-dependent manner in polarized Madin-Darby canine kidney epithelial cells. Cell Regul 1: 921–936 Yeaman C, Grindstaff KK, Nelson WJ (2004) Mechanism of recruiting Sec6/8 (exocyst) complex to the apical junctional complex during polarization of epithelial cells. J Cell Sci 117: 559–570 Yeaman C, Le Gall AH, Baldwin AN, Monlauzeur L, Le Bivic A, Rodriguez-Boulan E (1997) The O-glycosylated stalk domain is required for apical sorting of neurotrophin receptors in polarized MDCK cells. J Cell Biol 139: 929–940 Zhu Y, Doray B, Poussu A, Lehto VP, Kornfeld S (2001) Binding of GGA2 to the lysosomal enzyme sorting motif of the mannose6-phosphate receptor. Science 292: 1716–1718

580

*

Y. Wang

Golgi apparatus inheritance Yanzhuang Wang

Introduction: Golgi structure, function and inheritance General structure The Golgi apparatus is a membrane-bound organelle found in essentially all eukaryotic cells, including plants, animals, and fungi. The Golgi stack, with its arrangement of flattened cisternae with fenestrated rims, is the most unique feature of this organelle (Ladinsky et al. 1999). In most single cell low eukaryotes, each cell has one single Golgi stack (Shorter and Warren 2002); in plants and Drosophila, multiple copies of Golgi stacks are scattered throughout the entire cytosol (Nebenfuhr and Staehelin 2001; Segui-Simarro and Staehelin 2006; Stanley et al. 1997). This is in contrast to mammalian cells, where stacks are often interconnected to form a ribbon-like structure, which is localized adjacent to the nucleus (Ladinsky et al. 1999; Rambourg et al. 1987). The fact that Golgi cisternae are not stacked inbudding yeast(Saccharomyces cerevisiae) suggests that stacking is not absolutely essential for Golgi function. However, Golgi stacking is a pronounced feature of cellular organization in all metazoan eukaryotes as well as in many unicellular eukaryotes. This makes a strong argument that stacking has important functional consequences. Several hypotheses can explain the biological significance of Golgi stacking in higher eukaryotic cells. The close arrangement of cisternae ensures the movement of the vesicles from one cisterna to another in the most efficient and rapid manner; and local tethering proteins on the Golgi membrane hold the vesicles so that they cannot diffuse away (Lupashin and Sztul 2005). In addition, an ordered stack of cisternae is required to carry out the sequential modifications as cargo proteins pass from cisterna to cisterna. It has been observed that the enzymes that modify the N-linked oligosaccharides bound to many cargo molecules are distributed across the stack (in the cis to trans direction) in the order in which they operate on the cargo passing through the stack (Kornfeld and Kornfeld 1985). In multicellular organisms, modification of membrane and secretory proteins can be critical. So far the effect of Golgi cisternal stacking on the rate of protein transport through the Golgi has not been tested because of the lack of molecular tools to modify the stacking state of the Golgi.

The function of the Golgi The primary function of the Golgi apparatus is to process membrane and secretory proteins synthesized in the endoplasmic reticulum (ER), and sort them to their proper destination, such as the endosomal/lysosomal system,

Golgi apparatus inheritance

*

581

the plasma membrane or outside of the cell, by regulated or constitutive secretion (Farquhar and Palade 1998; Marsh and Howell 2002; Mellman and Simons 1992). Vesicles from the ER fuse with the cis Golgi network at the cis face of the Golgi apparatus and transfer their soluble protein content into the Golgi lumen. The proteins are then transported through the stack towards the trans Golgi network at the trans face and are modified on the way. Modifications include glycosylation, phosphorylation, sulfation and proteolysis (Goldfischer 1982). Three models describe the process of trafficking through the Golgi stack: cisternal maturation, in which proteins are transported by progression of the cisternae, while Golgi enzymes are recycled via retrograde transport of COPI vesicles (Bonfanti et al. 1998; Losev et al. 2006; Malhotra and Mayor 2006; Matsuura-Tokita et al. 2006; Orci et al. 1997), vesicular transport, in which Golgi cisternae remain stable while cargo is transported by COPI vesicles (Elsner et al. 2003), or transient tubular connections between the cisternae (Marsh et al. 2004; Trucco et al. 2004). Despite its complicated morphology and function, the Golgi apparatus is dynamic, capable of rapid disassembly and reassembly during mitosis (Shorter and Warren 2002) or upon drug treatment (Ho et al. 1989; Lippincott-Schwartz et al. 1989; Misumi et al. 1986; Takizawa et al. 1993).

Biological techniques used to study Golgi inheritance Light and electron microscopy The inheritance of the Golgi apparatus was first investigated in the early 1900s. One particularly good example is the study performed by Ludford in 1924, in which the Golgi was described as forming ‘rodlets’ that were scattered throughout the dividing cell and that, at a late stage in cell division, re-associated to form two separate Golgi complexes, one in each daughter cell (Ludford 1924). Studies using electron microscopy (EM) showed fragmentation of the characteristic stack-like organization of the Golgi at the onset of mitosis, and Golgi fragments were subsequently distributed to each daughter cell (Lucocq et al. 1989; Maul and Brinkley 1970). In the late 1990s, when green fluorescent proteins and live cell imaging techniques became available, it was shown that the Golgi undergoes extensive fragmentation at the beginning of mitosis (Altan-Bonnet et al. 2006; Axelsson and Warren 2004; Shima et al. 1997; Zaal et al. 1999). At the end of mitosis, the dispersed Golgi fragments fuse into larger Golgi fragments that finally re-form the pericentriolar ribbon structure. These studies provided the initial insights into Golgi behavior during cell division. In addition, microscopic and biochemical studies were often combined with pharmacological treatments in order to change Golgi behavior and protein trafficking through it. One such chemical is Brefeldin A (BFA), a fungal metabolite that redistributes the Golgi membranes into the ER. BFA works by inactivating the ADP-ribosylation factor-1 (ARF1) and thereby its recruitment to Golgi membranes (Doms et al. 1989). Inactivation of ARF1 in turn prevents recruitment of coatomer (and hence the formation of

582

*

Y. Wang

COPI vesicles) and leads to the redistribution of Golgi membranes to the ER (Klausner et al. 1992; Lippincott-Schwartz et al. 1989).

In vitro reconstitution assays Biochemical reconstitution experiments have provided powerful tools to dissect biological processes. Two basic experimental approaches have been taken to reconstitute mitotic Golgi disassembly and reassembly and thus to deconstruct the molecular processes. The semi-intact cell assay involves cells that are permeabilized with mild detergent (e.g., digitonin), washed with 1 M KCl to remove endogenous cytosol and peripheral membrane proteins, and then incubated with cytosol prepared from mitotic or interphase cells (Acharya et al. 1998; Colanzi et al. 2000; Kano et al. 2000). Cells can then be analyzed for changes in Golgi morphology by immunofluorescence and electron microscopy, or biochemistry. So far, this approach has reconstituted the disassembly phase, especially the initial breakdown of the Golgi ribbon, and thus has been used to identify the mitotic kinases that regulate the disassembly of the Golgi (Colanzi et al. 2000; Kano et al. 2000). The second method, the in vitro Golgi disassembly/reassembly assay, is a powerful technique for studying the factors involved in the morphological changes of the Golgi during mitosis. In the assay, structural changes of the purified stacks (Wang et al. 2006) in response to addition of mitotic or interphase cytosol or purified proteins are monitored by electron microscopy (Misteli and Warren 1994; Rabouille et al. 1998, 1995a, b). Briefly, highly purified rat liver Golgi stacks (RLG) are incubated in mitotic cytosol to generate mitotic Golgi fragments (MGF). After separation from cytosol by centrifugation through a sucrose cushion, these fragments will reassemble into Golgi stacks upon incubation with interphase cytosol or purified cytosolic components. This provides a readily manipulable biochemical system within which the sequence of morphological events can be precisely followed by quantitative EM or biochemical analysis. The fact that the Golgi apparatus can be disassembled and reassembled in vitro demonstrates the innate self-organizing potential of this organelle (Misteli 2001). This approach has contributed to the discovery and examination of much of the currently identified proteins that mediate Golgi membrane tethering (Satoh et al. 2003; Shorter and Warren 1999), fusion (Kondo et al. 1997; Rabouille et al. 1998, 1995a; Wang et al. 2004) and Golgi cisternal stacking (Barr et al. 1997; Shorter et al. 1999; Wang et al. 2003). Further improvement of this assay using purified proteins will allow the identification of the minimal machinery and the key components that control mitotic Golgi disassembly and post-mitotic reassembly.

Molecular mechanism of Golgi structure formation Golgi stack formation How the Golgi cisternae form stacks has been a long standing question. Identifying the proteins involved in Golgi stacking is vital to understand the

Golgi apparatus inheritance

*

583

structure and biogenesis of this organelle. Electron microscopic studies on cells and isolated Golgi membranes have described filamentous structures that bridge the gaps between adjacent cisternae (Franke et al. 1972). Stacking requires protein–protein interactions since it can be disrupted by mild proteolysis (Cluett and Brown 1992; Nilsson et al. 1994, 1996). The original studies that investigated such stacking factors focused on Golgi peripheral proteins such as p115. The GM130–p115–giantin complex, which is known to tether COPI transport vesicles to the Golgi, may initiate stack formation, as p115 can link cisternae via bridging giantin and GM130 in adjacent cisternae (Shorter and Warren 1999). However, this is not likely the major stacking mechanism, as p115 can be removed from Golgi membranes using high salt without unstacking Golgi cisternae (Cluett and Brown 1992; Waters et al. 1992). The development of the in vitro Golgi disassembly/reassembly assay led to the discovery of GRASP65, the first identified Golgi stacking protein. Specifically, GRASP65 was identified as a Golgi protein that is only accessible to the alkylating reagent N-ethylmaleimide (NEM) when the Golgi stacks are disassembled (Barr et al. 1997). Furthermore, adding GRASP65 antibodies to the in vitro Golgi reassembly reaction inhibited restacking of Golgi cisternae (Barr et al. 1997). Consistent with this, microinjection of GRASP65 antibodies into mitotic cells interfered with Golgi stack formation in the daughter cells (Wang et al. 2003). Subsequent studies using RNAi in Drosophila cells showed that knockdown of dGRASP (the sole GRASP protein in Drosophila) caused, at least partially, disassembly of the Golgi stacks into single cisternae and vesicles (Kondylis et al. 2005). Similar studies in mammalian cells indicated that GRASP65 knockdown reduced the average number of cisternae per stack from 6–3 (Sutterlin et al. 2005), although a different report using the same RNA oligos showed that Golgi ribbon formation rather than stacking was affected (Puthenveedu et al. 2006), indicating that stacking may not be the sole function of GRASP65. Finally, expression of a caspase-resistant form of GRASP65 partially prevented Golgi fragmentation in apoptotic cells (Lane et al. 2002). Perhaps related to this, knockdown of GRASP65 by RNAi also caused apoptosis (Sutterlin et al. 2005). Nevertheless, these studies provide convincing evidence that GRASP65 is essential for Golgi stacking (Fig. 1). GRASP55, a homolog of GRASP65, was identified by database searching (Shorter et al. 1999). Its GRASP domain at the N-terminus is 80% similar and 66% identical to the GRASP domain of GRASP65. GRASP55 has a serine/proline-rich (SPR) domain at the C-terminus that, although less conserved than the GRASP domain, contains a number of potential phosphorylation sites similar to those in GRASP65. GRASP proteins are attached to the membrane through lipid modifications, as both GRASP65 and GRASP55 are myristoylated (Barr et al. 1997; Shorter et al. 1999), and GRASP55 is also palmitoylated (Kuo et al. 2000). GRASP65 is mainly localized to the cis side of the Golgi stack, while GRASP55 is primarily

584

*

Y. Wang

Figure 1. Golgi cisternal stacking controlled by GRASP65 phosphorylation. (A) Schematic structure of Golgi stack assembly. GRASP65 dimers from adjacent cisternal membranes form oligomers that hold the cisternae together. Phosphorylation by CDK1 (with cyclin B1) and pololike kinase (Plk) breaks GRASP65 oligomers and thus unstacks the cisternae. Conversely, dephosphorylation of GRASP65 by PP2A restacks the cisternae. (B–D) Representative EM images of rat liver Golgi stacks were either untreated (B), or treated with CDK1 and Plk kinases (C) at 37 C for 20 min. Membranes re-isolated from (C) were further treated with interphase cytosol (IC) for 60 min at 37 C (D) Bar, 0.5 mm. (E) Quantitation of B–D to estimate the percentage of single or stacked cisternal membranes. The results represent the mean of three independent experiments  SEM.

localized to medial cisternae, suggesting that these two proteins may contribute to stacking of different regions of the Golgi (Pfeffer 2001). However, the mechanisms by which the two GRASP proteins are targeted to and maintained in different regions of the Golgi stack are not understood. GRASP65 binds GM130, while GRASP55 interacts with Golgin-45, both via PDZ-like interactions (Barr et al. 1998; Short et al. 2001). GM130 binds Rab1, and Golgin-45 interacts specifically with Rab2; both interactions are essential for Golgi structure and function (Moyer et al. 2001; Short et al. 2001; Weide et al. 2001). In Saccharomyces cerevisiae, in which the Golgi cisternae are not stacked, there is only one GRASP protein, GRH1/Ydr517w, which has a higher homology to mammalian GRASP55 (23% overall identity) than to GRASP65 (Behnia et al. 2007). It is involved in protein trafficking rather than stacking, suggesting that GRASP65 may have evolved from GRASP55 by gene duplication. So far the relative contributions of the two GRASP proteins to Golgi structure formation in mammalian cells, the establishment of the polarity of the Golgi stack, and the regulation of Golgi biogenesis during the cell cycle are not clear.

Golgi apparatus inheritance

*

585

Golgi ribbon formation The formation and positioning of the Golgi ribbon is thought to be dependent on microtubule organization and the centrosome (Barr and Egerer 2005; Infante et al. 1999; Linstedt 2004; Rios et al. 2004). The centrosome contains g-tubulin ring complexes, which nucleate polarized microtubule assembly. This creates a radial array of microtubules with minus ends embedded in the centrosome and plus ends extended in all directions towards the cell periphery. Golgi membrane-associated minus end-directed microtubule motors, mainly dynein, move the Golgi stacks inward and cause concentration of the Golgi stacks around the centrosome. It is indicated that GMAP-210 binds both the Golgi and microtubules and thus is involved in pericentriolar positioning of the Golgi (Barr and Egerer 2005; Infante et al. 1999; Linstedt 2004; Liu et al. 2007; Rios et al. 2004). Vesicles or other carriers originate at the cell periphery and the ER, track along microtubules to allow collection of the cargo proteins in the Golgi apparatus positioned at the cell center, and then move back towards the cell periphery along microtubules to various destinations including the plasma membrane. Thus the ribbon formation and localization of the Golgi may facilitate protein transport in mammalian cells. While positioned and compacted into the pericentriolar region by the microtubule network, the stacks are interconnected by tubular structures. So far there is no information available on how these tubular connections are formed at the molecular level. A few Golgi proteins have been implicated in Golgi ribbon formation. These include the Golgi stacking proteins GRASP65 and GRASP55 (Feinstein and Linstedt 2007; Puthenveedu et al. 2006), and the Golgi matrix proteins GM130 and golgin-84 (Diao et al. 2003; Puthenveedu et al. 2006). The evidence for GRASP65 in lateral membrane fusion that generates a Golgi ribbon came from a knockdown experiment of GRASP65 by RNAi which led to the fragmentation of the Golgi ribbon (Puthenveedu et al. 2006). However, this might be a secondary effect of Golgi unstacking caused by the interruption of GRASP65 function in stack formation (Kondylis et al. 2005; Sutterlin et al. 2005; Wang et al. 2003). In addition, knockdown of GRASP65 affects spindle dynamics and may cause cell death (Sutterlin et al. 2005). Since Golgi ribbon formation and localization depends on the microtubule network (Rios et al. 2004), and the Golgi is fragmented in apoptotic cells (Chiu et al. 2002; Lane et al. 2002; Lowe et al. 2004; Machamer 2003; Mancini et al. 2000), it is thus necessary to determine whether the Golgi ribbon breakdown observed in this study (Puthenveedu et al. 2006) is caused by interruption of the microtubule cytoskeleton or by apoptotic response. GRASP55 is phosphorylated in vitro by the MAP kinase (mitogen-activated protein kinase) ERK2 (the extracellular-signal-regulated kinase) (Jesch et al. 2001), which is indicated in breakdown of the Golgi ribbon at the onset of mitosis (Feinstein and Linstedt 2007; Kano et al. 2000), suggesting that GRASP55 might be involved in Golgi ribbon disassembly. In addition, expression of a non-phosphorylatable GRASP55 mutant interrupted mitotic

586

*

Y. Wang

Golgi fragmentation, e.g. Golgi stack unlinking, as well as G2/M transition (Feinstein and Linstedt 2007), further indicating a role for GRASP55 (or mitotic Golgi fragmentation itself) in cell cycle control. The underlying mechanisms are not clear. Two golgins, GM130 and golgin-84, are also involved in Golgi ribbon formation, as knockdown of either protein caused fragmentation of the Golgi (Diao et al. 2003; Puthenveedu et al. 2006). Both GM130 and golgin-84 are involved in tethering vesicles to the Golgi (Malsam et al. 2005). The GM130–p115–giantin complex tethers vesicles at the cis side of the Golgi (Nelson et al. 1998), while the golgin-84 and CASP complex functions in a similar way at the trans side of the Golgi (Malsam et al. 2005). Both GM130 and golgin-84 bind to Rab1 (Diao et al. 2003; Moyer et al. 2001; Satoh et al. 2003). How these vesicle tethering proteins link the stacks together to form a ribbon is so far unknown. The processes of Golgi ribbon and stack formation might be linked, as GRASP65 binds GM130, and disruption of this tethering complex may change the size of the Golgi stacks by altering protein trafficking (Seemann et al. 2000b). Consistently, it has been shown that golgin-84 depletion leads to a reduction of the overall size of the Golgi to approximately 25% of the volume of a normal Golgi apparatus (Diao et al. 2003), which may indirectly affect Golgi ribbon formation. Clearly, the direct involvement of these golgins in Golgi ribbon formation needs to be validated.

Diverse mechanisms of Golgi biogenesis Strategies of Golgi inheritance The mitotic division of cells requires the duplication and partitioning of all cellular components; without a Golgi apparatus the cell will not function. Golgi inheritance includes membrane growth during interphase and partitioning into the daughter cells during cell division. Membrane growth possibly occurs by insertion of membrane lipid and proteins into the preexisting Golgi membranes, which thus expand in size. In mammalian cells, the mechanism for the generation of new Golgi stacks within the Golgi ribbon is unclear. During cell division, the Golgi needs to be inherited accurately by the two daughter cells. There are two major strategies used to facilitate partitioning of subcellular components into the two daughter cells during cell division. One is a stochastic strategy, based on probability (Warren and Wickner 1996). This is particularly suited to organelles present in multiple dispersed copies such as mitochondria in mammalian cells. The accuracy of partitioning depends on the number of copies (the more, the better) and their distribution (the more evenly distributed, the better). Another strategy is an ordered portioning strategy (Shima et al. 1998). This strategy is used for low copy organelles such as the chromosomes and centrosomes and often involves the mitotic spindle and associated astral microtubules. Several organisms have been used as model systems to study the mechanism of Golgi

Golgi apparatus inheritance

*

587

inheritance, mostly focused on one of the basic questions: whether the new Golgi forms de novo. This includes two aspects: can a new functional Golgi structure be generated from the ER exit sites independent of the existing Golgi apparatus, or does it also need materials or a template from the existing Golgi? So far the results indicate that the mechanisms of Golgi inheritance used by different organisms are quite diverse.

Golgi inheritance in single cell eukaryotes In most single cell eukaryotes that generally contain only one single Golgi stack per cell (He 2007), inheritance of the Golgi normally occurs through the duplication of this stack; each stack then moves into each daughter cell. The generation of the new Golgi can occur through two different mechanisms: de novo formation or templated growth (Munro 2002). The de novo formation model (Fig. 2) proposes that the new Golgi can form using materials provided by the ER; the existing Golgi plays no role in the generation of the new Golgi. The templated growth model proposes the opposite—the new Golgi is generated from the old Golgi, which is done either by lateral growth followed by medial fission, or by providing a template from the existing Golgi for the new Golgi to grow. As discussed below, there are several cases in which the

Figure 2. Schematic models of Golgi division in single cell eukaryotes. (A) The templated model. Found in Toxoplasma gondii and Trichomonas foetus, the existing Golgi grows laterally and is divided in the middle to generate two new Golgi stacks. Although the ER exit site (ERES) may provide materials for lateral growth, it plays no role in Golgi division. (B) The Golgi de novo biogenesis model shown in Pichia pastoris. The ERES duplicates and provides all the materials required for the generation of the new Golgi, while the existing Golgi plays no role in this process. (C) The hybrid model that occurs in Trypanosoma brucei. The generation of the new Golgi depends on both the ERES and the old Golgi. The old Golgi provides materials (such as structural proteins and enzymes), directly or indirectly by ER–Golgi recycling, that may function as a template or seed for the new Golgi to grow from, while the duplicated ERES provides materials (such as cargo proteins and lipids) for the new Golgi to expand in size.

588

*

Y. Wang

generation of the new Golgi depends on both the old Golgi and the ER, possibly as a combination of both mechanisms (Fig. 2). Strictly speaking, Golgi biogenesis de novo means that a structurally and functionally defined Golgi forms independent of the existing Golgi; however, this has not been vigorously investigated in most reports described below. Several single cell parasites have been shown to generate a new Golgi stack by templated growth; one of the better characterized models is Toxoplasma gondii. Each parasite has a single Golgi stack. This single copy Golgi stack grows laterally, and undergoes medial fission during cell division, so that each daughter cell obtains one functional Golgi (Pelletier et al. 2002). This result indicates that the existing Golgi provides the template and materials for generation of the new Golgi (Fig. 2A). A similar mechanism has also been used in another protozoan, Trichomonas foetus (Benchimol et al. 2001). It is apparent that lateral growth of the Golgi is mediated by accumulation of lipids and proteins, presumably by receiving these components from the ER, although some lipids can be synthesized locally on the Golgi. So far it is unknown what controls the medial fission process. In the fission yeast Pichia pastoris, Golgi can form de novo from ER exit sites (Fig. 2B). Unlike Saccharomyces cerevisiae, which does not have a stacked Golgi, Pichia has a stacked Golgi with 2–5 stacks per cell, which are associated with or localized adjacent to the ER exit sites (also called transitional ER, e.g. tER). In Pichia, the biogenesis of tER sites is tightly linked to the biogenesis of the Golgi, and both compartments can form de novo (Losev et al. 2006; Matsuura-Tokita et al. 2006). Using time-lapse video microscopy to follow GFP-tagged proteins of a number of residents of the tER and the Golgi through mitosis, it was shown that the Golgi often fuse with one another, but they maintain a consistent average size through shrinkage after fusion and growth following de novo formation; late Golgi elements often move away from tER sites towards regions of polarized growth. By the time the daughter is ready to separate from the mother, both cells have 2–3 stacks. This model suggests that Golgi inheritance depends on the duplication of the tER, although the molecular mechanism of tER duplication is currently unknown. It is also not known what components are provided by the tER for the generation of the new Golgi; it could include some proteins from the old Golgi that are recycled through the ER. In addition, vesicles could also arrive directly from existing Golgi stacks; peripheral membrane proteins are probably in equilibrium between cytoplasmic pools and so could also be delivered directly (Munro 2002). In Trypanosoma brucei, each cell has a single Golgi stack that lies adjacent to the basal body (Hartmann et al. 2006). The cell builds a new Golgi adjacent to the basal body by a hybrid mechanism of de novo biogenesis and templated growth (Hartmann et al. 2006; He et al. 2004, 2005; Ho et al. 2006). The de novo synthesis requires materials transported from the ER exit site. During cell division, the single ER exit site is also duplicated. Duplication of the ER exit site and the Golgi are coordinated by centrin2, a protein that is specifically

Golgi apparatus inheritance

*

589

associated with the Golgi. Another centrin, centrin1, is associated with the basal body. Depletion of centrin2 inhibits duplication of the Golgi (He et al. 2005). This centrin2-containing structure appears to mark the site for new Golgi assembly; it forms a bi-lobed structure, and one lobe associates with the old Golgi, and the other one indicates the place where the new ER exit site and the new Golgi form. Currently the molecular nature of this centrin2-containing structure is not clear. In Trypanosoma, generation of the new Golgi also requires transfer of old material from the existing Golgi (He et al. 2004), and thus can be considered as a hybrid model of the templated growth and de novo biogenesis (Fig. 2C). It is not known what the old Golgi provides. Thus it is difficult to discriminate whether it is a true de novo assembly or whether the materials provided by the old Golgi function as a template. It can be generally outlined that Golgi biogenesis depends on both the old Golgi, which may provide materials that function as a template or a seed for the new Golgi, such as matrix proteins and Golgi enzymes, and the ER exit site, which provides lipids and proteins for the growth of the new Golgi, such as cargo proteins. It is necessary to investigate the nature of these materials provided by the ER and the old Golgi and their roles in Golgi inheritance in all the aforementioned model systems.

Golgi inheritance in plants In plants, fungi and other lower organisms, the Golgi stacks are not linked with each other to form a ribbon due to the lack of a centrosome and a radial arrangement of the microtubule network. Instead, plant Golgi stacks distribute in the cytosol and move along actin filaments using myosin motors (Brandizzi et al. 2002; Nebenfuhr and Staehelin 2001). In Arabidopsis thaliana, each interphase cell contains about 35 Golgi stacks, and this number doubles during G2 phase immediately prior to mitosis (Segui-Simarro and Staehelin 2006); exactly how this is controlled is not known. In addition, the Golgi maintains its stacked structure and function in secretion during cell division, as this is needed for formation of the new cell wall between the two daughter cells. It has been suggested that Golgi stacks accumulate in the vicinity of the phragmoplast, the cytoskeletal structure responsible for cell plate assembly (Nebenfuhr et al. 2000). As a consequence, the Golgi is partitioned between nascent daughter cells as pre-existing discrete units (Nebenfuhr et al. 2000).

Golgi inheritance in interphase mammalian cells The mechanism that governs the increase in the number of Golgi stacks is still a mystery. Many EM and fluorescence images from disparate organisms give the impression that this proceeds by lateral growth followed by medial fission (Garcia-Herdugo et al. 1988; Hager et al. 1999; Troyer and Cameron 1980). A few studies have been performed to determine Golgi biogenesis in interphase cells. For example, microsurgery of cells can generate cytoplasts that contain

590

*

Y. Wang

no Golgi; these pieces cannot generate new Golgi and cannot deliver proteins to the correct location (Pelletier et al. 2000). In another study, horseradish peroxidase (HRP) was tagged to a Golgi enzyme and thus was targeted into the Golgi lumen, which allowed Golgi inactivation by chemical cross-linking of diaminobenzidine (DAB) in the Golgi lumen (Jollivet et al. 2007). Under this condition, new membranes of vesicular–tubular structures accumulated adjacent to the old Golgi. Some Golgi matrix proteins were recruited onto these membranes, but others such as giantin and GS-15 were absent, and this new compartment did not display a normal Golgi morphology nor transport proteins. Both studies suggest that the formation of a new functionally and structurally defined Golgi apparatus requires a pre-existing functional Golgi apparatus.

Mitotic Golgi disassembly and post-mitotic reassembly during the mammalian cell cycle In mammalian cells, over a hundred Golgi stacks are restricted to a tight juxtanuclear location, where they are interconnected to form a ribbon. Division of the Golgi during each cycle of cell division occurs through unique disassembly and reassembly processes (Lucocq et al. 1987, 1989; Lucocq and Warren 1987). The Golgi is disassembled at the onset of mitosis, by breakdown of the Golgi ribbon and dispersal of the stacks, which then undergo further vesiculation, yielding vesicles that are distributed throughout the cytoplasm (Check 2002). These vesicles are observed under the fluorescence microscope as a Golgi haze in mitotic cells, although an alternative view suggests that the Golgi haze represents Golgi proteins merged with the ER (see below). The dispersed distribution of these vesicles is one mechanism that ensures equal distribution of the Golgi membranes into the two daughter cells. Due to the limitation of light microscopy to distinguish the disassembled structures, disagreement exists about the mechanism of mitotic Golgi disassembly (Barr 2004; Check 2002; Rossanese and Glick 2001). One hypothesis proposes that mitotic Golgi disassembly phenocopies the effect of BFA that causes fusion of the Golgi membrane with the ER (Klausner et al. 1992), and thus the Golgi is a derivative of the ER and Golgi division depends on ER division (Altan-Bonnet et al. 2003, 2004, 2006; Reinke et al. 2004; Storrie 2005; Zaal et al. 1999). A second idea proposes that mitotic Golgi disassembly is mediated by generation of vesicles, so the Golgi is an autonomous organelle whose division is independent of the ER (Axelsson and Warren 2004; Jesch 2002; Jokitalo et al. 2001; Lucocq et al. 1989; Misteli and Warren 1995a; Pecot and Malhotra 2004; Seemann et al. 2000a, 2002; Shorter and Warren 2002; Xiang et al. 2007). Despite the controversy, it is often observed that mitotic Golgi disassembly is not complete: not all the Golgi membranes are fragmented into vesicles. Instead, some remain as mitotic membrane clusters, detected as large irregular dots by immunocytochemistry methods (Axelsson and Warren 2004; Jesch and Linstedt 1998; Misteli and Warren 1995a). It has been shown that at least

Golgi apparatus inheritance

*

591

some of these mitotic Golgi clusters are associated with and concentrated near the spindle pole bodies suggesting that they are inherited by the two daughter cells in a more ordered way during anaphase and telophase (Shima et al. 1998). The separation of the two spindle poles may allow the ordered partitioning of the mitotic Golgi clusters into the two daughter cells. The ordered inheritance strategy used by mammalian cells is a second mechanism to ensure equal distribution of Golgi membranes into the two daughter cells. Some Golgi matrix proteins are more concentrated in the mitotic clusters. These proteins may function as templates in this process (Seemann et al. 2000a, 2002). During telophase, Golgi vesicles are distributed equally between daughter cells, where they are assembled into stacks and ribbons. Due to Golgi fragmentation and inhibition of membrane fusion, secretion is inactivated between prometaphase and telophase and is resumed during telophase. It is generally believed that almost all the disassembled Golgi membranes are reassembled into new Golgi stacks after mitosis, although some of the vesicles may directly fuse with the plasma membrane. Fusion of secretory vesicles with the plasma membrane may facilitate closure of the intercellular bridge during abscission (Gromley et al. 2005; Robinson and Spudich 2000). This naturally occurring phenomenon of Golgi disassembly and reassembly provides a unique opportunity to deconstruct the molecular mechanisms behind the biogenesis and maintenance of Golgi architecture.

The biological significance of Golgi disassembly and reassembly during the mammalian cell cycle Several reasons can explain the biological importance of mitotic Golgi fragmentation. The first concerns the mechanism of Golgi partitioning between the two daughter cells during the cell cycle. The distribution of Golgi membranes over a larger space in the cell is expected to aid in the even distribution of this organelle into the daughter cells (Shorter and Warren 2002). The second hypothesis concerns the release of “mitotic” components stored on the Golgi during interphase, which are important in the promotion of cytokinesis (Altan-Bonnet et al. 2003; Colanzi et al. 2003). For example, Nir2, a Golgi-associated protein in interphase cells, has been shown to move to the cleavage furrow and is essential for cytokinesis (Litvak et al. 2004). Thus, in addition to directly providing membrane for cleavage furrow invagination (Gromley et al. 2005), it is likely that the Golgi mediates the coordinated release of proteins required for cytokinesis and other cytoplasmic events as the cell progresses through mitosis. In addition, recent studies indicate a role for mitotic Golgi fragmentation in cell cycle progression; inhibition of breakdown of the Golgi ribbon in mammalian cells via a functional block of proteins involved in this process (GRASP65, GRASP55 and BARS) results in the arrest of the cell cycle at the G2 stage (Colanzi et al. 2007; Hidalgo Carcedo et al. 2004; Sutterlin et al. 2002; Yoshimura et al. 2005), although different views exist (Uchiyama et al. 2003).

592

*

Y. Wang

Molecular mechanisms of mitotic Golgi disassembly in mammalian cells Initial Golgi ribbon breakdown It is generally agreed that the initial step of mitotic Golgi disassembly is the breakdown of the Golgi ribbon, which is regulated by kinases. Much of the knowledge on kinase regulation of initial Golgi ribbon fragmentation was contributed by the semi-intact cell assay. Semi-intact normal rat kidney (NRK) cells were treated with purified kinases, or with mitotic cytosol in the presence of kinase inhibitors. This has indicated a role for Polo-like kinase 1 (Plk1) and MEK1 in the initial fragmentation of the Golgi ribbon into its constituent stacks during early prophase (Acharya et al. 1998; Colanzi et al. 2000; Sutterlin et al. 2001). Plk1 phosphorylates GRASP65 at multiple residues during mitosis (Lin et al. 2000), leading to GRASP65 de-oligomerization and thus Golgi unstacking (Wang et al. 2003), which may enhance Golgi ribbon fragmentation. Inhibition of Plk1 inhibited mitotic Golgi fragmentation in semi-intact NRK cells (Sutterlin et al. 2001). In a similar assay using Madin-Darby canine kidney (MDCK) cells, sequential roles for MEK1 and cyclin-dependent kinase 1 (CDK1, also called cdc2) were perceived during mitotic fragmentation (Kano et al. 2000). MEK1 was required for the initial breakdown of the Golgi ribbon into large punctuated fragments, which appear to be Golgi stacks by EM. After this phase, CDK1 was required for the subsequent transformation of these stacks into dispersed Golgi vesicles (Kano et al. 2000). Several hypotheses have been proposed for MEK1 regulation of Golgi ribbon breakdown. One hypothesis suggests that the MEK1 activity required for the Golgi fragmentation occurring uniquely in mitosis does not require activation of its classical substrates ERK1 or ERK2; instead, it is through activation of a novel Golgi-associated ERK activity (Acharya et al. 1998; Colanzi et al. 2000). Consistently, tyrosine-phosphorylated ERK2 specifically accumulates in the nucleus and on the Golgi in late G2/prophase, and elevation of tyrosine-phosphorylated ERK2 by increased MEK1 expression appears to disrupt the Golgi apparatus at the light microscope level (Cha and Shapiro 2001). These effects were independent of ERK2 kinase activity, but did require the tyrosine phosphorylation of ERK2. Further understanding of the role of MEK1 in mitotic Golgi fragmentation awaits the identification of the substrate(s) on the Golgi. A second hypothesis on the role of MEK1 in Golgi fragmentation proposes that this is mediated by phosphorylation of GRASP55, a substrate of ERK2 at least in vitro (Jesch et al. 2001). Expression of a non-phosphorylatable GRASP55 mutant inhibited Golgi ribbon fragmentation in early mitosis (Feinstein and Linstedt 2007). How the Golgi stacking factor GRASP55 is involved in Golgi ribbon formation and the underlying mechanism have yet to be determined. In addition, a recent study showed that a novel isoform of ERK, ERK1c, regulates mitotic Golgi fragmentation (Aebersold et al. 2004; Shaul and Seger 2006). Whether ERK1c is a general

Golgi apparatus inheritance

*

593

kinase that regulates Golgi structure through phosphorylation of GRASP55 is so far unknown. Recent studies showed that Golgi ribbon fragmentation, which converts it into isolated stacks by fission of inter-stack connecting tubules, occurs at the onset of G2/M transition. This process involves the Brefeldin A-dependent ADP-ribosylation substrate (BARS), also known as the fission protein Cterminal-binding protein 1, short form (CtBP1). BARS is a key component of mitotic Golgi fragmentation in permeabilized NRK cells (Hidalgo Carcedo et al. 2004). Inhibition of BARS activity inhibited fragmentation of the Golgi ribbon as well as G2/M transition and mitotic entrance (Colanzi et al. 2007). So far it is unknown whether BARS and MEK1 act in the same pathway.

Golgi membrane unstacking Eventually Golgi disassembly is mediated by unstacking of the cisternae and vesiculation and/or tubulation of the unstacked cisternae. It was observed that the number of cisternae per stack decreases as Golgi disassembly proceeds, both in vivo and in vitro (Misteli and Warren 1994, 1995b). As described above, Golgi stacking is mediated by two stacking factors, GRASP65 localized at the cis side and GRASP55 localized towards the medial cisternae. The disassembly of the Golgi apparatus during mitosis indicates that the Golgi structural proteins must be modified or changed at this time to allow membrane fragmentation to occur. GRASP65 is phosphorylated by two mitotic kinases, CDK1 and Plk (Lin et al. 2000; Wang et al. 2003), and its dephosphorylation is mediated by protein phosphatase PP2A (Tang et al. 2008). Phosphorylation of GRASP65 by purified recombinant kinases leads to Golgi membrane unstacking, both in vivo and in vitro (Wang et al. 2003). Correspondingly, dephosphorylation of GRASP65 restacks Golgi cisternae. Furthermore, GRASP65 forms dimers, but the two GRASP65 proteins in the homodimer could not be separated by kinase treatment, suggesting that the homodimer might be the basic functional unit. In fact, the GRASP65 homodimer is capable of forming oligomers between molecules residing on adjacent Golgi membranes and these trans-oligomers are capable of holding the cisternal membrane together (Wang et al. 2005, 2003). Oligomerization of GRASP65 is regulated by phosphorylation; mitotic phosphorylation of GRASP65 leads to de-oligmerization and thus Golgi cisternal unstacking. Post-mitotic GRASP65 dephosphorylation leads to GRASP65 re-oligomerization and thus Golgi stack formation (Figs. 1, 3). Consistently, expression of a mutant GRASP65 that is capable of oligomerization but is not mitotically regulated inhibits mitotic Golgi fragmentation (Wang et al. 2005). The second stacking factor, GRASP55, is phosphorylated by ERK2 during mitosis (Jesch et al. 2001), but whether this induces Golgi unstacking is not known.

COPI-dependent Golgi membrane vesiculation There are two pathways involved in fragmentation of the Golgi cisternae: a COPI-dependent pathway and a COPI-independent pathway (Fig. 3). The

594

*

Y. Wang

Figure 3. Illustration of the processes and proteins involved in Golgi stack disassembly and reassembly during the mammalian cell cycle. At the onset of mitosis, after fragmentation of the Golgi ribbon (not shown) into individual stacks, there are two major processes that disassemble the Golgi stacks into mitotic Golgi fragments: unstacking and vesiculation. Unstacking is mediated by protein phosphorylation of Golgi stacking proteins by mitotic kinases (e.g. CDK1 and polo-like kinase), which disrupt the GRASP65 oligomer and thus unstack the Golgi. Phosphorylation of golgins (i.e. GM130 and golgin-84) also interrupts vesicle fusion. Vesiculation of the cisternal membranes is through a COPI-dependent mechanism, mediated by ARF1 and the coatomer complex, and possibly a COPI-independent pathway whose mechanism is so far unknown. Post-mitotic Golgi reassembly is mediated by formation of single cisternae by membrane fusion and the subsequent restacking. Cisternal membrane fusion requires two AAA ATPases, p97 and NSF (N-ethylmaleimide-sensitive fusion protein), each of which functions together with specific adaptor proteins. Membrane fusion mediated by the p97 pathway requires mono-ubiquitination of Golgi membrane proteins during mitosis. Restacking of the newly formed Golgi cisternae requires dephosphorylation of Golgi stacking proteins by the protein phosphatase PP2A, which allows re-oligomerization of GRASP65 and thus membrane stacking.

COPI-dependent pathway proceeds as COPI vesicles continue to bud from all levels of the Golgi stack (Sonnichsen et al. 1996), but are unable to tether and so fuse with their target membrane (Misteli and Warren 1994; Nakamura et al. 1997). In general, this pathway likely consumes the peripheral rims of cisternae (Misteli and Warren 1994, 1995b; Sonnichsen et al. 1996). However, since Golgi cisternae are unstacked during mitosis, more surface area is available for budding. The COPI-dependent pathway can consume about 80% of unstacked Golgi membranes, leaving the trans-Golgi network intact (Tang et al. 2008). Besides the increased vesicle budding enhanced by cisternal unstacking, vesicle fusion might be inhibited during mitosis. As stated previously, the GM130–p115–giantin tethering complex functions to hold the vesicles to the Golgi membranes so that these vesicles do not diffuse away (Lowe 2000; Lowe et al. 1998; Nakamura et al. 1997). At the onset of mitosis, GM130 is

Golgi apparatus inheritance

*

595

phosphorylated by CDK1 on serine 25, which is located in the p115 binding site, leading to disruption of the GM130–p115–giantin tethering complex (Lowe 2000; Lowe et al. 1998; Nakamura et al. 1997). As a result, COPI vesicles diffuse away from the Golgi region and accumulate in the cytosol (Collins and Warren 1992; Fernandez and Warren 1998; Stuart et al. 1993). Furthermore, microinjection of recombinant cyclin B-CDK1 into interphase cells induces GM130 phosphorylation and a mitotic-like disassembly of the Golgi apparatus (Wang et al. 2003). Besides GRASP65 and GM130, a few more Golgi matrix proteins are also regulated by phosphorylation during mitosis. p115 phosphorylation by a casine kinase II (CKII)-like kinase enhances p115–GM130 interaction. Inhibition of p115 phosphorylation during mitosis may contribute to the interruption of the tethering complex (Sohda et al. 1998), while phosphorylation of p115 may be essential for post-mitotic Golgi reassembly, as it has been shown that p115 can catalyze SNARE complex formation for membrane fusion (Dirac-Svejstrup et al. 2000; Shorter et al. 2002). Another protein, p47, a cofactor of the AAA ATPase p97 that is involved in post-mitotic Golgi membrane fusion, is also phosphorylated by CDK1 during mitosis (Kondo et al. 1997; Mayr et al. 1999). How this affects p97/p47 activity is unclear, but may preclude p47 binding to the Golgi t-SNARE syntaxin 5, which is essential for p97/p47 function in Golgi homotypic fusion (Rabouille et al. 1998). Rab1 is also phosphorylated by CDK1 during mitosis, and this induces increased Rab1 membrane association (Bailly et al. 1991). Rab1 phosphorylation may alter interactions with its effectors GM130 and p115, and thus affect COPI vesicletethering reactions (Allan et al. 2000; Moyer et al. 2001; Weide et al. 2001). Golgin-84 is phosphorylated during mitosis, which might cause fragmentation of the Golgi ribbon (Diao et al. 2003).

COPI-independent membrane fragmentation The COPI-independent pathway converts the cisternal cores into a heterogeneous array of tubulovesicular profiles (Misteli and Warren 1994). The existence of this pathway was indicated by treatment of purified Golgi membranes with mitotic cytosol depleted of coatomer (Misteli and Warren 1994), which is consistent with the observation that some of the Golgi membranes remain as mitotic clusters during mitosis (Misteli and Warren 1995a). Although the mechanism mediating this conversion is unknown, it might be related to BARS (Weigert et al. 1999). BARS is involved in Golgi membrane fission and breakdown of the Golgi ribbon at the on set of mitosis (Colanzi et al. 2007; Hidalgo Carcedo et al. 2004). To what extent the COPI-independent pathway contributes to Golgi fragmentation is unclear. There is speculation that these mitotic clusters of tubulovesicular structures may function as templates for formation of the new Golgi in the daughter cells (Seemann et al. 2000a, 2002). In the presence of BFA, Golgi enzymes and contents return to the ER via a COPI-independent retrograde transport pathway that may utilize tubules as the transport vector (Klausner et al. 1992; Lippincott-Schwartz et al. 1989).

596

*

Y. Wang

Coincidentally, BARS is an ADP ribosylation substrate induced by BFA (Spano et al. 1999; Weigert et al. 1997, 1999). Whether the COPI-independent pathway shares a similar mechanism as BFA treatment is an interesting question to investigate.

Molecular mechanisms of post-mitotic Golgi reassembly in mammalian cells Compared to the mechanism of mitotic Golgi disassembly, fewer studies have been performed concerning the molecular mechanism of post-mitotic Golgi reassembly. The morphological changes of the Golgi during the post-mitotic reassembly process in later mitosis through cytokinesis have been documented by light and electron microscopy. Much of our understanding on the molecular mechanism of Golgi reassembly was contributed by the in vitro reconstitution assays. In general, there are essentially two processes for formation of new stacks (Fig. 3): membrane fusion that is mediated by two AAA ATPases, NSF and p97, each of which works together with its adaptor proteins (Acharya et al. 1995; Kondo et al. 1997; Meyer et al. 1998, 2000; Rabouille et al. 1995a, 1998), and Golgi membrane restacking, mediated by dephosphorylation and re-oligomerization of the Golgi stacking proteins (Wang et al. 2003, 2005).

NSF-dependent cisternal membrane regrowth So far the best characterized membrane fusion machinery is the four-helical bundle SNAREpin complex. This assembly of the v-/t-SNARE complex induces membrane fusion (Rothman and Warren 1994; Weber et al. 1998). The disassembly of the SNARE complex is catalyzed by N-ethylmaleimide-sensitive factor (NSF), with its cofactors a- and g-SNAP (soluble NSF attachment protein). The ATPase activity of NSF is required for SNARE disassembly. However, this NSF ATPase activity is not required during Golgi reassembly, suggesting a different role for NSF in membrane fusion (Muller et al. 1999). The fact that purified NSF, a- and g-SNAPs and p115 can be used to replace interphase cytosol for reassembly of mitotic Golgi fragments into Golgi cisternae strongly suggests that NSF and its cofactors are capable of fusing Golgi membranes (Rabouille et al. 1995a). In the NSF pathway, the tethering factor p115 plays an important role in regulating the membrane fusion process. Besides its role in vesicle tethering as part of the GM130–p115–giantin complex, p115 interacts with syntaxin 5 and thus may catalyze SNAREpin formation and membrane fusion (Shorter et al. 2002). Another essential component of NSF-driven Golgi reassembly is GATE-16, a small protein that contains a ubiquitin fold decorated by two additional N-terminal helices. It binds NSF and GOS-28 (a Golgi v-SNARE) and is involved in intra-Golgi transport (Nagahama et al. 1996; Sagiv et al. 2000) and post-mitotic cisternal regrowth (Muller et al. 2002).

Golgi apparatus inheritance

*

597

p97-dependent cisternal membrane regrowth Compared to NSF-dependent membrane fusion, the mechanism of p97 (also referred to as valosin-containing protein, VCP) -mediated cisternal regrowth is less clear. The evidence that p97 can mediate Golgi membrane fusion came from the experiment in which interphase cytosol in the in vitro reassembly assay was replaced by purified p97 and p47 (Kondo et al. 1997; Rabouille et al. 1995a, 1998). Incubation of mitotic Golgi fragments with p97/p47 generated long, single cisternae. This result suggests that p97 may induce homotypic fusion for extension of cisternal length. Thus, NSF may mediate fusion of existing cisternal membranes with COPI vesicles, while p97 may mediate fusion between Golgi cisternae or tubular structures. So far the equilibrium between the p97/NSF pathways is unknown, but the NSF and p97 pathways of Golgi reassembly contribute non-additively to cisternal regrowth (Rabouille et al. 1998). p97 is an abundant and highly conserved protein (Peters et al. 1990). It was discovered that p97 could bind to syntaxin 5 in vitro (Rabouille et al. 1998), which led to the proposal that p97 might carry out a reaction similar to the SNARE complex disassembly mediated by NSF, but on different SNAREs (Patel et al. 1998; Rabouille et al. 1998). Although p97 has been shown to release the t-SNARE syntaxin 5 from a complex with p47 and VCIP135 (Uchiyama et al. 2002), p97-mediated membrane fusion, in contrast to NSF, does not require the Golgi v-SNARE Gos-28 (Rabouille et al. 1998). One hypothesis is that p97mediated membrane fusion functions through t-/t-SNARE pairing instead of v-/t-SNARE complex formation; the t-/t-SNARE pairing was suggested previously for Ufe1, a yeast SNARE involved in ER membrane fusion (Patel et al. 1998), and syntaxin 6 for homotypic membrane fusion of immature secretory granules (Wendler et al. 2001). So far there is no evidence that syntaxin 5 forms such a complex, nor that p97/p47 complex can dissociate it. p97 has broader functions than NSF as it has several cofactors, including p47 (Kondo et al. 1997; Meyer et al. 1998), VCIP135 (Uchiyama et al. 2002; Wang et al. 2004), p37 (Uchiyama et al. 2006), the Ufd1p/Npl4p complex (Meyer et al. 2000, 2002) and SVIP (Nagahama et al. 2003). It is believed that different adaptor proteins direct p97 to perform these varied cellular activities (Meyer et al. 2002; Woodman 2003). Similar to p47, p37 forms a complex with p97 in the cytosol and targets it to the Golgi and ER. p37 is required for Golgi and ER biogenesis, as knockdown of this protein causes fragmentation of the Golgi. In an in vitro Golgi reassembly assay, the p97/p37 complex has membrane fusion activity. In contrast to the p97/p47 pathway, this pathway requires p115–GM130 tethering and SNARE GS15, but not syntaxin 5 (Uchiyama et al. 2006). Although both p47 and p37 can direct p97 activity toward Golgi membrane fusion, the underlying mechanism awaits elucidation. One surprising finding in the field is the role of ubiquitination in regulation of Golgi membrane dynamics during the cell cycle (Meyer et al. 2002; Wang et al. 2004). Although poly-ubiquitination is normally linked with targeting proteins for degradation by the proteasome, it is now appreciated that

598

*

Y. Wang

mono-ubiqitination also has regulatory functions in membrane sorting and for Golgi inheritance. Ubiquitination occurs during mitotic Golgi disassembly, which is required for subsequent Golgi reassembly (Meyer et al. 2002). It has been shown that the p97/p47 complex binds mono-ubiquitin and this interaction is required for p97-mediated membrane fusion (Meyer et al. 2002); inhibition of this interaction also inhibited p97-mediated Golgi membrane fusion (Wang et al. 2004). Another strong piece of evidence for the involvement of ubiquitination in Golgi inheritance is the discovery of the p97/p47 cofactor VCIP135 as a deubiquitinating enzyme that is required for reformation of the Golgi complex in an in vitro assay (Wang et al. 2004). This suggests that ubiquitination – through cycles of adding and removing ubiquitin to substrates – is necessary for Golgi reassembly, rather than ubiquitin-dependent proteolysis (Wang et al. 2004). The exact role played by ubiquitin in this process is unknown, but perhaps mon-oubiquitin linked to a Golgi substrate could be recognized as a label on the fragmented Golgi membranes that are dispersed in the cytosol during mitosis. Thus the interaction between the mono-ubiquitin on the Golgi fragments and the p97/p47 membrane fusion machinery may bring the membrane together, and after removal of the ubiquitin by the deubiquiting enzyme VCIP135, these membranes would ultimately fuse to form the cisternae (Fig. 3). Similarly, it has also been shown that NSF can bind to GATE-16, a ubiquitin-like protein (Nagahama et al. 1996; Sagiv et al. 2000), suggesting a general mechanism for ubiquitination in regulation of Golgi membrane dynamics during the cell cycle. A number of questions remain: one concerns the identity of the ubiquitin ligase, and the second concerns the identity of the substrate(s) on the Golgi membranes. Answers to these questions will help to determine the role of ubiquitination in Golgi inheritance.

Stacking Golgi cisternae The cisternae formed by NSF- and p97-mediated membrane fusion begin to align with each other and form stacks. At the initial steps of stack formation, the formation of giantin–p115–GM130 tethers may play a critical role in linking membrane together (Shorter and Warren 1999), which is regulated by a Rab-GTPase (Shorter et al. 2002). Golgi stack formation is also inhibited by microcystin (Rabouille et al. 1995b), suggesting a role for protein dephosphorylation in this process. GM130 is dephosphorylated in telophase (Lowe et al. 2000); this allows its interaction with p115 and formation of the tethering complex, which is important for membrane fusion. Membrane tethering mediated by p115 dimers through the interactions with GM130 on the Golgi membranes and giantin on the vesicles is thought to be important for the initial step in stack formation (Shorter and Warren 1999). Although p115 and GM130 are involved in an early stage of stack formation, the GRASP proteins play a much more critical role in stacking. Both GRASP65 and GRASP55 are phosphorylated during mitosis, and GRASP65 is dephosphorylated by PP2A after mitosis; so far the identity of

Golgi apparatus inheritance

*

599

the phosphatase that dephosphorylates GRASP55 after mitosis is unknown. GRASP65 is capable of oligomerization, which is sufficient to hold the cisternae to form stacks (Wang et al. 2003, 2005). Phosphorylation of GRASP65 by CDK1 and Plk disrupts the oligomers and leads to Golgi membrane unstacking; post-mitotic dephosphorylation of GRASP65 leads to re-oligomerization of this protein and thus stack formation (Figs. 1, 3). As GRASP65 is mainly localized at the cis side, it is not known whether the stack formation is sequential, such as from cis to trans, or simultaneous to form the entire stack. At the end of mitosis as the spindle disassembles and new microtubule network forms, the Golgi stacks are transported to the cell center, where they are interconnected to form a ribbon.

Summary and future perspectives Study on Golgi inheritance in single cell organisms holds much promise for unraveling important aspects of these processes, since the Golgi stack is a single-copy organelle in these organisms and can be readily monitored using GFP technology. The complexity comes from the lack of conservation of the mechanism even in single cell organisms. One challenging task is to determine the role of the existing Golgi in the generation of the new Golgi, as this will help to conclude whether the mechanism of Golgi inheritance is evolutionarily conserved. In mammalian cells, the formation of the Golgi ribbon and the disassembly and reassembly processes during cell division increase the complexity. Fluorescence microscopy and live cell imaging in combination with pharmacological treatments have documented the morphological changes of the Golgi during the cell cycle under different conditions. The in vitro assays that reconstitute the Golgi disassembly and reassembly processes have led to several breakthroughs. However, the conclusions of the two sets of techniques are not always consistent, and thus it is a major challenge to determine the true mechanism. Furthermore, it is also necessary to determine whether the mechanism of Golgi structure formation is conserved between different species. For example, the roles of the Golgi matrix proteins, which are important for formation of the mammalian Golgi, have not been documented in lower eukaryotes and plants, and so far no homologues of GM130 or the GRASP proteins have been found in plants. Clearly, finding the mechanism of structure formation is a prerequisite for understanding the inheritance of this organelle. For example, how is Golgi polarity maintained? What keeps the Golgi a relatively stable structure, while at the same time allowing it to fulfill its function in highly dynamic membrane transport, or during cisternal maturation? What controls Golgi duplication in lower organisms and when does this happen? How is the Golgi structure linked to cell cycle control? There is no doubt that all these questions must be addressed by experimental approaches. In addition, studies on Golgi biogenesis also provide basic knowledge on many more specific and important issues, such as Golgi biogenesis in abnormal cell growth such as cancer, or under certain

600

*

Y. Wang

stress conditions in diseases. These newly immerging questions will stimulate the development of this field as they are directly related to other important questions that are relevant to human health and development. Acknowledgements. The author wishes to thank Graham Warren, Joachim Seemann and Yi Xiang for critical reading of the manuscript, and Kari Mar for help with text editing.

References Acharya U, Jacobs R, Peters JM, Watson N, Farquhar MG, Malhotra V (1995) The formation of Golgi stacks from vesiculated Golgi membranes requires two distinct fusion events. Cell 82: 895–904 Acharya U, Mallabiabarrena A, Acharya JK, Malhotra V (1998) Signaling via mitogenactivated protein kinase kinase (MEK1) is required for Golgi fragmentation during mitosis. Cell 92: 183–192 Aebersold DM, Shaul YD, Yung Y, Yarom N, Yao Z, Hanoch T, Seger R (2004) Extracellular signal-regulated kinase 1c (ERK1c), a novel 42-kilodalton ERK, demonstrates unique modes of regulation, localization, and function. Mol Cell Biol 24: 10000–10015 Allan BB, Moyer BD, Balch WE (2000) Rab1 recruitment of p115 into a cis-SNARE complex: programming budding COPII vesicles for fusion. Science 289: 444–448 Altan-Bonnet N, Phair RD, Polishchuk RS, Weigert R, Lippincott-Schwartz J (2003) A role for Arf1 in mitotic Golgi disassembly, chromosome segregation, and cytokinesis. Proc Natl Acad Sci USA 100: 13314–13319 Altan-Bonnet N, Sougrat R, Lippincott-Schwartz J (2004) Molecular basis for Golgi maintenance and biogenesis. Curr Opin Cell Biol 16: 364–372 Altan-Bonnet N, Sougrat R, Liu W, Snapp EL, Ward T, Lippincott-Schwartz J (2006) Golgi inheritance in mammalian cells is mediated through endoplasmic reticulum export activities. Mol Biol Cell 17: 990–1005 Axelsson MA, Warren G (2004) Rapid, ER-independent diffusion of the mitotic Golgi haze. Mol Biol Cell 15: 1843–1852 Bailly E, McCaffrey M, Touchot N, Zahraoui A, Goud B, Bornens M (1991) Phosphorylation of two small GTP-binding proteins of the Rab family by p34cdc2 Nature 350: 715–718 Barr FA (2004) Golgi inheritance: shaken but not stirred. J Cell Biol 164: 955–958 Barr FA, Egerer J (2005) Golgi positioning: are we looking at the right MAP? J Cell Biol 168: 993–998 Barr FA, Nakamura N, Warren G (1998) Mapping the interaction between GRASP65 and GM130, components of a protein complex involved in the stacking of Golgi cisternae. EMBO J 17: 3258–3268 Barr FA, Puype M,Vandekerckhove J, Warren G (1997) GRASP65, a protein involved in the stacking of Golgi cisternae. Cell 91: 253–262 Behnia R, Barr FA, Flanagan JJ, Barlowe C, Munro S (2007) The yeast orthologue of GRASP65 forms a complex with a coiled-coil protein that contributes to ER to Golgi traffic. J Cell Biol 176: 255–261 Benchimol M, Ribeiro KC, Mariante RM, Alderete JF (2001) Structure and division of the Golgi complex in Trichomonas vaginalis and Tritrichomonas foetus. Eur J Cell Biol 80: 593–607 Bonfanti L, Mironov AA Jr, Martinez-Menarguez JA, Martella O, Fusella A, Baldassarre M, Buccione R, Geuze HJ, Mironov AA, Luini A (1998) Procollagen traverses the Golgi stack without leaving the lumen of cisternae: evidence for cisternal maturation. Cell 95: 993–1003

Golgi apparatus inheritance

*

601

Brandizzi F, Snapp EL, Roberts AG, Lippincott-Schwartz J, Hawes C (2002) Membrane protein transport between the endoplasmic reticulum and the Golgi in tobacco leaves is energy dependent but cytoskeleton independent: evidence from selective photobleaching. Plant Cell 14: 1293–1309 Cha H, Shapiro P (2001) Tyrosine-phosphorylated extracellular signal – regulated kinase associates with the Golgi complex during G2/M phase of the cell cycle: evidence for regulation of Golgi structure. J Cell Biol 153: 1355–1367 Check E (2002) Cell biology: will the real Golgi please stand up. Nature 416: 780–781 Chiu R, Novikov L, Mukherjee S, Shields D (2002) A caspase cleavage fragment of p115 induces fragmentation of the Golgi apparatus and apoptosis. J Cell Biol 159: 637–648 Cluett EB, Brown WJ (1992) Adhesion of Golgi cisternae by proteinaceous interactions: intercisternal bridges as putative adhesive structures. J Cell Sci 103: 773–784 Colanzi A, Carcedo CH, Persico A, Cericola C, Turacchio G, Bonazzi M, Luini A, Corda D (2007) The Golgi mitotic checkpoint is controlled by BARS-dependent fission of the Golgi ribbon into separate stacks in G2 EMBO J 26: 2465–2476 Colanzi A, Deerinck TJ, Ellisman MH, Malhotra V (2000) A specific activation of the mitogen-activated protein kinase kinase 1 (MEK1) is required for Golgi fragmentation during mitosis. J Cell Biol 149: 331–339 Colanzi A, Suetterlin C, Malhotra V (2003) Cell-cycle-specific Golgi fragmentation: how and why? Curr Opin Cell Biol 15: 462–467 Collins RN, Warren G (1992) Sphingolipid transport in mitotic HeLa cells. J Biol Chem 267: 24906–24911 Diao A, Rahman D, Pappin DJ, Lucocq J, Lowe M (2003) The coiled-coil membrane protein golgin-84 is a novel rab effector required for Golgi ribbon formation. J Cell Biol 160: 201–212 Dirac-Svejstrup AB, Shorter J, Waters MG, Warren G (2000) Phosphorylation of the vesicle-tethering protein p115 by a casein kinase II-like enzyme is required for Golgi reassembly from isolated mitotic fragments. J Cell Biol 150: 475–488 Doms RW, Russ G, Yewdell JW (1989) Brefeldin A redistributes resident and itinerant Golgi proteins to the endoplasmic reticulum. J Cell Biol 109: 61–72 Elsner M, Hashimoto H, Nilsson T (2003) Cisternal maturation and vesicle transport: join the band wagon! Mol Membr Biol 20: 221–229 Farquhar MG, Palade GE (1998) The Golgi apparatus: 100 years of progress and controversy. Trends Cell Biol 8: 2–10 Feinstein TN, Linstedt AD (2007) Mitogen-activated protein kinase kinase 1-dependent Golgi unlinking occurs in G2 phase and promotes the G2/M cell cycle transition. Mol Biol Cell 18: 594–604 Fernandez CJ, Warren G (1998) In vitro synthesis of sulfated glycosaminoglycans coupled to inter-compartmental Golgi transport. J Biol Chem 273: 19030–19039 Franke WW, Kartenbeck J, Krien S, VanderWoude WJ, Scheer U, Morre DJ (1972) Interand intracisternal elements of the Golgi apparatus. A system of membrane-tomembrane cross-links. Z Zellforsch Mikrosk Anat 132: 365–380 lez-Reyes JA, Garcua-Navarro F, Navas P (1988) Growth Garcia-Herdugo G, Gonza kinetics of the Golgi apparatus during the cell cycle in onion root meristems. Planta 175: 305–312 Goldfischer S (1982) The internal reticular apparatus of Camillo Golgi: a complex heterogeneous organelle, enriched in acid, neutral, and alkaline phosphatases, and involved in glycosylation, secretion, membrane flow, lysosome formation, and intracellular digestion. J Histochem Cytochem 30: 717–733 Gromley A, Yeaman C, Rosa J, Redick S, Chen CT, Mirabelle S, Guha M, Sillibourne J, Doxsey SJ (2005) Centriolin anchoring of exocyst and SNARE complexes at the midbody is required for secretory-vesicle-mediated abscission. Cell 123: 75–87

602

*

Y. Wang

Hager KM, Striepen B, Tilney LG, Roos DS (1999) The nuclear envelope serves as an intermediary between the ER and Golgi complex in the intracellular parasite Toxoplasma gondii. J Cell Sci 112: 2631–2638 Hartmann J, Hu K, He CY, Pelletier L, Roos DS, Warren G (2006) Golgi and centrosome cycles in Toxoplasma gondii. Mol Biochem Parasitol 145: 125–127 He CY (2007) Golgi biogenesis in simple eukaryotes. Cell Microbiol 9: 566–572 He CY, Ho HH, Malsam J, Chalouni C, West CM, Ullu E, Toomre D, Warren G (2004) Golgi duplication in Trypanosoma brucei. J Cell Biol 165: 313–321 He CY, Pypaert M, Warren G (2005) Golgi duplication in Trypanosoma brucei requires Centrin2. Science 310: 1196–1198 Hidalgo Carcedo C, Bonazzi M, Spano S, Turacchio G, Colanzi A, Luini A, Corda D (2004) Mitotic Golgi partitioning is driven by the membrane-fissioning protein CtBP3/BARS. Science 305: 93–96 Ho HH, He CY, De Graffenried CL, Murrells LJ, Warren G (2006) Ordered assembly of the duplicating Golgi in Trypanosoma brucei. Proc Natl Acad Sci USA 103: 7676–7681 Ho WC, Allan VJ, Van Meer G, Berger EG, Kreis TE (1989) Reclustering of scattered Golgi elements occurs along microtubules. Eur J Cell Biol 48: 250–263 Infante C, Ramos-Morales F, Fedriani C, Bornens M, Rios RM (1999) GMAP-210, A cisGolgi network-associated protein is a minus end microtubule-binding protein. J Cell Biol 145: 83–98 Jesch SA (2002) Inheriting a structural scaffold for Golgi biosynthesis. Bioessays 24: 584–587 Jesch SA, Lewis TS, Ahn NG, Linstedt AD (2001) Mitotic phosphorylation of Golgi reassembly stacking protein 55 by mitogen-activated protein kinase ERK2. Mol Biol Cell 12: 1811–1817 Jesch SA, Linstedt AD (1998) The Golgi and endoplasmic reticulum remain independent during mitosis in HeLa cells. Mol Biol Cell 9: 623–635 Jokitalo E, Cabrera-Poch N, Warren G, Shima DT (2001) Golgi clusters and vesicles mediate mitotic inheritance independently of the endoplasmic reticulum. J Cell Biol 154: 317–330 Jollivet F, Raposo G, Dimitrov A, Sougrat R, Goud B, Perez F (2007) Analysis of De Novo Golgi complex formation after enzyme-based inactivation. Mol Biol Cell 18: 4637–4647 Kano F, Takenaka K, Yamamoto A, Nagayama K, Nishida E, Murata M (2000) MEK and Cdc2 kinase are sequentially required for Golgi disassembly in MDCK cells by the mitotic Xenopus extracts. J Cell Biol 149: 357–68 Klausner RD, Donaldson JG, Lippincott-Schwartz J (1992) Brefeldin A: insights into the control of membrane traffic and organelle structure. J Cell Biol 116: 1071–1080 Kondo H, Rabouille C, Newman R, Levine TP, Pappin D, Freemont P, Warren G (1997) p47 is a cofactor for p97-mediated membrane fusion. Nature 388: 75–78 Kondylis V, Spoorendonk KM, Rabouille C (2005) dGRASP localization and function in the early exocytic pathway in Drosophila S2 cells. Mol Biol Cell 16: 4061–4072 Kornfeld R, Kornfeld S (1985) Assembly of asparagine-linked oligosaccharides. Ann Rev Biochem 54: 631–664 Kuo A, Zhong C, Lane WS, Derynck R (2000) Transmembrane transforming growth factor-alpha tethers to the PDZ domain-containing, Golgi membrane-associated protein p59/GRASP55. EMBO J 19: 6427–6439 Ladinsky MS, Mastronarde DN, McIntosh JR, Howell KE, Staehelin LA (1999) Golgi structure in three dimensions: functional insights from the normal rat kidney cell. J Cell Biol 144: 1135–1149 Lane JD, Lucocq J, Pryde J, Barr FA, Woodman PG, Allan VJ, Lowe M (2002) Caspasemediated cleavage of the stacking protein GRASP65 is required for Golgi fragmentation during apoptosis. J Cell Biol 156: 495–509

Golgi apparatus inheritance

*

603

Lin CY, Madsen ML, Yarm FR, Jang YJ, Liu X, Erikson RL (2000) Peripheral Golgi protein GRASP65 is a target of mitotic polo-like kinase (Plk) and Cdc2. Proc Natl Acad Sci USA 97: 12589–94 Linstedt AD (2004) Positioning the Golgi apparatus. Cell 118: 271–272 Lippincott-Schwartz J, Yuan LC, Bonifacino JS, Klausner RD (1989) Rapid redistribution of Golgi proteins into the ER in cells treated with brefeldin A: evidence for membrane cycling from Golgi to ER. Cell 56: 801–813 Litvak V, Argov R, Dahan N, Ramachandran S, Amarilio R, Shainskaya A, Lev S (2004) Mitotic phosphorylation of the peripheral Golgi protein Nir2 by Cdk1 provides a docking mechanism for Plk1 and affects cytokinesis completion. Mol Cell 14: 319–330 Liu Z, Vong QP, Zheng Y (2007) CLASPing microtubules at the trans-Golgi Network. Dev Cell 12: 839–840 Losev E, Reinke CA, Jellen J, Strongin DE, Bevis BJ, Glick BS (2006) Golgi maturation visualized in living yeast. Nature 441: 1002–1006 Lowe M (2000) Membrane transport: tethers and TRAPPs. Curr Biol 10: R407–R409 Lowe M, Gonatas NK, Warren G (2000) The mitotic phosphorylation cycle of the cisGolgi matrix protein GM130. J Cell Biol 149: 341–356 Lowe M, Lane JD, Woodman PG, Allan VJ (2004) Caspase-mediated cleavage of syntaxin 5 and giantin accompanies inhibition of secretory traffic during apoptosis. J Cell Sci 117: 1139–1150 Lowe M, Rabouille C, Nakamura N, Watson R, Jackman M, Jamsa E, Rahman D, Pappin DJ, Warren G (1998) Cdc2 kinase directly phosphorylates the cis-Golgi matrix protein GM130 and is required for Golgi fragmentation in mitosis. Cell 94: 783–793 Lucocq JM, Berger EG, Warren G (1989) Mitotic Golgi fragments in HeLa cells and their role in the reassembly pathway. J Cell Biol 109: 463–474 Lucocq JM, Pryde JG, Berger EG, Warren G (1987) A mitotic form of the Golgi apparatus in HeLa cells. J Cell Biol 104: 865–874 Lucocq JM, Warren G (1987) Fragmentation and partitioning of the Golgi apparatus during mitosis in HeLa cells. EMBO J 6: 3239–3246 Ludford RJ (1924) The distribution of the cytoplasmic organs in transplantable tumour cells, with special reference to dictyokinesis. Proc R Soc Lond 97: 50–60 Lupashin V, Sztul E (2005) Golgi tethering factors. Biochim Biophys Acta 1744: 325–339 Machamer CE (2003) Golgi disassembly in apoptosis: cause or effect? Trends Cell Biol 13: 279–281 Malhotra V, Mayor S (2006) Cell biology: the Golgi grows up. Nature 441: 939–940 Malsam J, Satoh A, Pelletier L, Warren G (2005) Golgin tethers define subpopulations of COPI vesicles. Science 307: 1095–1098 Mancini M, Machamer CE, Roy S, Nicholson DW, Thornberry NA, Casciola-Rosen LA, Rosen A (2000) Caspase-2 is localized at the Golgi complex and cleaves golgin-160 during apoptosis. J Cell Biol 149: 603–612 Marsh BJ, Howell KE (2002) The mammalian Golgi – complex debates. Nat Rev Mol Cell Biol 3: 789–795 Marsh BJ, Volkmann N, McIntosh JR, Howell KE (2004) Direct continuities between cisternae at different levels of the Golgi complex in glucose-stimulated mouse islet beta cells. Proc Natl Acad Sci USA 101: 5565–5570 Matsuura-Tokita K, Takeuchi M, Ichihara A, Mikuriya K, Nakano A (2006) Live imaging of yeast Golgi cisternal maturation. Nature 441: 1007–1010 Maul GG, Brinkley BR (1970) The Golgi apparatus during mitosis in human melanoma cells in vitro. Cancer Res 30: 2326–2335 Mayr P SM, Allan VJ, Woodman PG (1999) Phosphorylation of p97 (VCP) and p47 in vitro by p34cdc2 kinase. Eur J Cell Biol 78: 224–232 Mellman I, Simons K (1992) The Golgi complex: in vitro veritas? Cell 68: 829–840 Meyer HH, Kondo H, Warren G (1998) The p47 co-factor regulates the ATPase activity of the membrane fusion protein, p97. FEBS Lett 437: 255–257

604

*

Y. Wang

Meyer HH, Shorter JG, Seemann J, Pappin D, Warren G (2000) A complex of mammalian ufd1 and npl4 links the AAA-ATPase, p97, to ubiquitin and nuclear transport pathways. EMBO J 19: 2181–2192 Meyer HH, Wang Y, Warren G (2002) Direct binding of ubiquitin conjugates by the mammalian p97 adaptor complexes, p47 and Ufd1-Npl4. EMBO J 21: 5645–5652 Misteli T (2001) The concept of self-organization in cellular architecture. J Cell Biol 155: 181–185 Misteli T, Warren G (1994) COP-coated vesicles are involved in the mitotic fragmentation of Golgi stacks in a cell-free system. J Cell Biol 125: 269–282 Misteli T, Warren G (1995a) Mitotic disassembly of the Golgi apparatus in vivo. J Cell Sci 108: 2715–2727 Misteli T, Warren G (1995b) A role for tubular networks and a COP I-independent pathway in the mitotic fragmentation of Golgi stacks in a cell-free system. J Cell Biol 130: 1027–1039 Misumi Y, Misumi Y, Miki K, Takatsuki A, Tamura G, Ikehara Y (1986) Novel blockade by brefeldin A of intracellular transport of secretory proteins in cultured rat hepatocytes. J Biol Chem 261: 11398–11403 Moyer BD, Allan BB, Balch WE (2001) Rab1 interaction with a GM130 effector complex regulates COPII vesicle cis-Golgi tethering. Traffic 2: 268–276 Muller JM, Rabouille C, Newman R, Shorter J, Freemont P, Schiavo G, Warren G, Shima DT (1999) An NSF function distinct from ATPase-dependent SNARE disassembly is essential for Golgi membrane fusion. Nat Cell Biol 1: 335–340 Muller JM, Shorter J, Newman R, Deinhardt K, Sagiv Y, Elazar Z, Warren G, Shima DT (2002) Sequential SNARE disassembly and GATE-16-GOS-28 complex assembly mediated by distinct NSF activities drives Golgi membrane fusion. J Cell Biol 157: 1161–1173 Munro S (2002) More than one way to replicate the Golgi apparatus. Nat Cell Biol 4: E223–E224 € llner Nagahama M, Orci L, Ravazzola M, Amherdt M, Lacomis L, Tempst P, Rothman JE, So TH (1996) A v-SNARE implicated in intra-Golgi transport. J Cell Biol 133: 507–516 Nagahama M, Suzuki M, Hamada Y, Hatsuzawa K, Tani K, Yamamoto A, Tagaya M (2003) SVIP is a novel VCP/p97-interacting protein whose expression causes cell vacuolation. Mol Biol Cell 14: 262–273 Nakamura N, Lowe M, Levine TP, Rabouille C, Warren G (1997) The vesicle docking protein p115 binds GM130, a cis-Golgi matrix protein in a mitotically regulated manner. Cell 89: 445–455 Nebenfuhr A, Frohlick JA, Staehelin LA (2000) Redistribution of Golgi stacks and other organelles during mitosis and cytokinesis in plant cells. Plant Physiol 124: 135–152 Nebenfuhr A, Staehelin LA (2001) Mobile factories: Golgi dynamics in plant cells. Trends Plant Sci 6: 160–167 Nelson DS, Alvarez C, Gao YS, Garcia-Mata R, Fialkowski E, Sztul E (1998) The membrane transport factor TAP/p115 cycles between the Golgi and earlier secretory compartments and contains distinct domains required for its localization and function. J Cell Biol 143: 319–31 Nilsson T, Hoe MH, Slusarewicz P, Rabouille C, Watson R, Hunte F, Watzele G, Berger EG, Warren G (1994) Kin recognition between medial Golgi enzymes in HeLa cells. EMBO J 13: 562–574 Nilsson T, Rabouille C, Hui N, Watson R, Warren G (1996) The role of the membranespanning domain and stalk region of N-acetylglucosaminyltransferase I in retention kin recognition and structural maintenance of the Golgi apparatus in HeLa cells. J Cell Sci 109: 1975–1989 Orci L, Stamnes M, Ravazzola M, Amherdt M, Perrelet A, Sollner TH, Rothman JE (1997) Bidirectional transport by distinct populations of COPI-coated vesicles. Cell 90: 335–349

Golgi apparatus inheritance

*

605

Patel SK, Indig FE, Olivieri N, Levine N, Latterich M (1998) Organelle membrane fusion: a novel function for the syntaxin homolog Ufe1p in ER membrane fusion. Cell 92: 611–620 Pecot MY, Malhotra V (2004) Golgi membranes remain segregated from the endoplasmic reticulum during mitosis in mammalian cells. Cell 116: 99–107 Pelletier L, Jokitalo E, Warren G (2000) The effect of Golgi depletion on exocytic transport. Nat Cell Biol 2: 840–846 Pelletier L, Stern CA, Pypaert M, Sheff D, Ngo HM, Roper N, He CY, Hu K, Toomre D, Coppens I, Roos DS, Joiner KA, Warren G (2002) Golgi biogenesis in Toxoplasma gondii. Nature 418: 548–552 Peters JM, Walsh MJ, Franke WW (1990) An abundant and ubiquitous homo-oligomeric ring-shaped ATPase particle related to the putative vesicle fusion proteins Sec18p and NSF. EMBO J 9: 1757–1767 Pfeffer SR (2001) Constructing a Golgi complex. J Cell Biol 155: 873–875 Puthenveedu MA, Bachert C, Puri S, Lanni F, Linstedt AD (2006) GM130 and GRASP65dependent lateral cisternal fusion allows uniform Golgi-enzyme distribution. Nat Cell Biol 8: 238–248 Rabouille C, Kondo H, Newman R, Hui N, Freemont P, Warren G (1998) Syntaxin 5 is a common component of the NSF- and p97-mediated reassembly pathways of Golgi cisternae from mitotic Golgi fragments in vitro. Cell 92: 603–610 Rabouille C, Levine TP, Peters JM, Warren G (1995a) An NSF-like ATPase, p97, and NSF mediate cisternal regrowth from mitotic Golgi fragments. Cell 82: 905–914 Rabouille C, Misteli T, Watson R, Warren G (1995b) Reassembly of Golgi stacks from mitotic Golgi fragments in a cell-free system. J Cell Biol 129: 605–618 Rambourg A, Clermont Y, Hermo L, Segretain D (1987) Tridimensional structure of the Golgi apparatus of nonciliated epithelial cells of the ductuli efferentes in rat: an electron microscope stereoscopic study. Biol Cell 60: 103–115 Reinke CA, Kozik P, Glick BS (2004) Golgi inheritance in small buds of Saccharomyces cerevisiae is linked to endoplasmic reticulum inheritance. Proc Natl Acad Sci USA 101: 18018–18023 Rios RM, Sanchis A, Tassin AM, Fedriani C, Bornens M (2004) GMAP-210 recruits gammatubulin complexes to cis-Golgi membranes and is required for Golgi ribbon formation. Cell 118: 323–335 Robinson DN, Spudich JA (2000) Towards a molecular understanding of cytokinesis. Trends Cell Biol 10: 228–237 Rossanese OW, Glick BS (2001) Deconstructing Golgi inheritance. Traffic 2: 589–596 Rothman JE, Warren G (1994) Implications of the SNARE hypothesis for intracellular membrane topology and dynamics. Curr Biol 4: 220–233 Sagiv Y, Legesse-Miller A, Porat A, Elazar Z (2000) GATE-16, a membrane transport modulator interacts with NSF and the Golgi v-SNARE GOS-28. EMBO J 19: 1494–1504 Satoh A, Wang Y, Malsam J, Beard MB, Warren G (2003) Golgin-84 is a rab1 binding partner involved in Golgi structure. Traffic 4: 153–161 Seemann J, Jokitalo E, Pypaert M, Warren G (2000a) Matrix proteins can generate the higher order architecture of the Golgi apparatus. Nature 407: 1022–1026 Seemann J, Jokitalo EJ, Warren G (2000b) The role of the tethering proteins p115 and GM130 in transport through the Golgi apparatus in vivo. Mol Biol Cell 11: 635–645 Seemann J, Pypaert M, Taguchi T, Malsam J, Warren G (2002) Partitioning of the matrix fraction of the Golgi apparatus during mitosis in animal cells. Science 295: 848–851 Segui-Simarro JM, Staehelin LA (2006) Cell cycle-dependent changes in Golgi stacks vacuoles clathrin-coated vesicles and multivesicular bodies in meristematic cells of Arabidopsis thaliana: a quantitative and spatial analysis. Planta 223: 223–236 Shaul YD, Seger R (2006) ERK1c regulates Golgi fragmentation during mitosis. J Cell Biol 172: 885–897

606

*

Y. Wang

Shima DT, Cabrera-Poch N, Pepperkok R, Warren G (1998) An ordered inheritance strategy for the Golgi apparatus: visualization of mitotic disassembly reveals a role for the mitotic spindle. J Cell Biol 141: 955–966 Shima DT, Haldar K, Pepperkok R, Watson R, Warren G (1997) Partitioning of the Golgi apparatus during mitosis in living HeLa cells. J Cell Biol 137: 1211–1228 Short B, Preisinger C, Korner R, Kopajtich R, Byron O, Barr FA (2001) A GRASP55-rab2 effector complex linking Golgi structure to membrane traffic. J Cell Biol 155: 877–883 Shorter J, Beard MB, Seemann J, Dirac-Svejstrup AB, Warren G (2002) Sequential tethering of Golgins and catalysis of SNAREpin assembly by the vesicle-tethering protein p115. J Cell Biol 157: 45–62 Shorter J, Warren G (1999) A role for the vesicle tethering protein p115, in the post-mitotic stacking of reassembling Golgi cisternae in a cell-free system. J Cell Biol 146: 57–70 Shorter J, Warren G (2002) Golgi architecture and inheritance. Annu Rev Cell Dev Biol 18: 379–420 Shorter J, Watson R, Giannakou ME, Clarke M, Warren G, Barr FA (1999) GRASP55, a second mammalian GRASP protein involved in the stacking of Golgi cisternae in a cell-free system. EMBO J 18: 4949–4960 Sohda M, Misumi Y, Yano A, Takami N, Ikehara Y (1998) Phosphorylation of the vesicle docking protein p115 regulates its association with the Golgi membrane. J Biol Chem 273: 5385–5388 Sonnichsen B, Watson R, Clausen H, Misteli T, Warren G (1996) Sorting by COP I-coated vesicles under interphase and mitotic conditions. J Cell Biol 134: 1411–1425 Spano S, Silletta MG, Colanzi A, Alberti S, Fiucci G, Valente C, Fusella A, Salmona M, Mironov A, Luini A, Corda D, Spanfo S (1999) Molecular cloning and functional characterization of brefeldin A-ADP-ribosylated substrate. A novel protein involved in the maintenance of the Golgi structure. J Biol Chem 274: 17705–17710 Stanley H, Botas J, Malhotra V (1997) The mechanism of Golgi segregation during mitosis is cell-type-specific. Proc Natl Acad Sci USA 94: 14467–14470 Storrie B (2005) Maintenance of Golgi apparatus structure in the face of continuous protein recycling to the endoplasmic reticulum: making ends meet. Int Rev Cytol 244: 69–94 Stuart RA, Mackay D, Adamczewski J, Warren G (1993) Inhibition of intra-Golgi transport in vitro by mitotic kinase. J Biol Chem 268: 4050–4054 Sutterlin C, Hsu P, Mallabiabarrena A, Malhotra V (2002) Fragmentation and dispersal of the pericentriolar Golgi complex is required for entry into mitosis in mammalian cells. Cell 109: 359–369 Sutterlin C, Lin CY, Feng Y, Ferris DK, Erikson RL, Malhotra V (2001) Polo-like kinase is required for the fragmentation of pericentriolar Golgi stacks during mitosis. Proc Natl Acad Sci USA 98: 9128–9132 Sutterlin C, Polishchuk R, Pecot M, Malhotra V (2005) The Golgi-associated protein GRASP65 regulates spindle dynamics and is essential for cell division. Mol Biol Cell 16: 3211–3222 Takizawa PA, Yucel JK, Veit B, Faulkner DJ, Deerinck T, Soto G, Ellisman M, Malhotra V (1993) Complete vesiculation of Golgi membranes and inhibition of protein transport by a novel sea sponge metabolite, ilimaquinone. Cell 73: 1079–1090 Tang D, Mar K, Warren G, Wang Y (2008) Molecular mechanism of mitotic Golgi disassembly and reassembly revealed by a defined reconstitution assay. J Biol Chem 283: 6085–6094 Troyer D, Cameron ML (1980) Spermiogenesis in Lumbricid earthworms revisited. I – Function and fate of centrioles, fusion of organelles and organelle movements. Biol Cell 37: 279–286 Trucco A, Polishchuk RS, Martella O, Di Pentima A, Fusella A, Di Giandomenico D, San Pietro E, Beznoussenko GV, Polishchuk EV, Baldassarre M, Buccione R, Geerts WJ, Koster AJ, Burger KN, Mironov AA, Luini A (2004) Secretory traffic triggers the

Golgi apparatus inheritance

*

607

formation of tubular continuities across Golgi sub-compartments. Nat Cell Biol 6: 1071–1081 Uchiyama K, Jokitalo E, Kano F, Murata M, Zhang X, Canas B, Newman R, Rabouille C, Pappin D, Freemont P, Kondo H (2002) VCIP135, a novel essential factor for p97/p47mediated membrane fusion, is required for Golgi and ER assembly in vivo. J Cell Biol 159: 855–866 Uchiyama K, Jokitalo E, Lindman M, Jackman M, Kano F, Murata M, Zhang X, Kondo H (2003) The localization and phosphorylation of p47 are important for Golgi disassembly-assembly during the cell cycle. J Cell Biol 161: 1067–1079 Uchiyama K, Totsukawa G, Puhka M, Kaneko Y, Jokitalo E, Dreveny I, Beuron F, Zhang X, Freemont P, Kondo H (2006) p37 is a p97 adaptor required for Golgi and ER biogenesis in interphase and at the end of mitosis. Dev Cell 11: 803–816 Wang Y, Satoh A, Warren G (2005) Mapping the functional domains of the Golgi stacking factor GRASP65. J Biol Chem 280: 4921–4928 Wang Y, Satoh A, Warren G, Meyer HH (2004) VCIP135 acts as a deubiquitinating enzyme during p97-p47-mediated reassembly of mitotic Golgi fragments. J Cell Biol 164: 973–978 Wang Y, Seemann J, Pypaert M, Shorter J, Warren G (2003) A direct role for GRASP65 as a mitotically regulated Golgi stacking factor. EMBO J 22: 3279–3290 Wang Y, Taguchi T, Warren G (2006) Purification of rat liver Golgi stacks. In: Celis J (ed) Cell biology: a laboratory handbook, 3rd edn. Vol. 2. Elsevier Science (USA), San Diego. pp. 33–39 Warren G, Wickner W (1996) Organelle inheritance. Cell 84: 395–400 Waters MG, Clary DO, Rothman JE (1992) A novel 115-kD a peripheral membrane protein is required for intercisternal transport in the Golgi stack. J Cell Biol 118: 1015–1026 Weber T, Zemelman BV, McNew JA, Westermann B, Gmachl M, Parlati F, Sollner TH, Rothman JE (1998) SNAREpins: minimal machinery for membrane fusion. Cell 92: 759–772 Weide T, Bayer M, Koster M, Siebrasse JP, Peters R, Barnekow A (2001) The Golgi matrix protein GM130: a specific interacting partner of the small GTPase rab1b. EMBO Rep 2: 336–341 Weigert R, Colanzi A, Mironov A, Buccione R, Cericola C, Sciulli MG, Santini G, Flati S, Fusella A, Donaldson JG, Di Girolamo M, Corda D, De Matteis MA, Luini A (1997) Characterization of chemical inhibitors of brefeldin A-activated mono-ADP-ribosylation. J Biol Chem 272: 14200–14207 Weigert R, Silletta MG, Spano S, Turacchio G, Cericola C, Colanzi A, Senatore S, Mancini R, Polishchuk EV, Salmona M, Facchiano F, Burger KN, Mironov A, Luini A, Corda D (1999) CtBP/BARS induces fission of Golgi membranes by acylating lysophosphatidic acid. Nature 402: 429–433 Wendler F, Page L, Urbe S, Tooze SA (2001) Homotypic fusion of immature secretory granules during maturation requires syntaxin 6. Mol Biol Cell 12: 1699–1709 Woodman PG (2003) p97, a protein coping with multiple identities. J Cell Sci 116: 4283–4290 Xiang Y, Seemann J, Bisel B, Punthambaker S, Wang Y (2007) Active ADP-ribosylation factor-1 (ARF1) is required for mitotic Golgi fragmentation. J Biol Chem 282: 21829–21837 Yoshimura S, Yoshioka K, Barr FA, Lowe M, Nakayama K, Ohkuma S, Nakamura N (2005) Convergence of cell cycle regulation and growth factor signals on GRASP65. J Biol Chem 280: 23048–23056 Zaal KJ, Smith CL, Polishchuk RS, Altan N, Cole NB, Ellenberg J, Hirschberg K, Presley JF, Roberts TH, Siggia E, Phair RD, Lippincott-Schwartz J (1999) Golgi membranes are absorbed into and reemerge from the ER during mitosis. Cell 99: 589–601

Peculiarities of intracellular transport in different organisms

Features of the plant Golgi apparatus

*

611

Features of the plant Golgi apparatus Chris Hawes, Anne Osterrieder and Imogen Sparkes

Introduction The plant Golgi apparatus (GA) as its counterparts in mammalian, insect and fungal cells is a multifunctional organelle not only receiving and modifying cargo delivered from the endoplasmic reticulum (ER) for export, but also synthesising lipids and many of the complex polysaccharides of the cell wall (Neumann et al. 2003). It is also likely that the organelle acts as one of several destinations for endocytosed material (Fowke et al. 1991). The GA in a plant cell is composed of numerous, sometimes many hundreds, individual stacks of cisternae being approximately 1 mm in diameter (Fig. 1). These superficially resemble the stacks reported within insect cells such as Drosophila (Kondylis et al. 2005), with the major difference in that in elongate or mature cells individual stacks demonstrate an actin-based motility, showing a range of movements associated with the surface of ER tubules (Boevink et al. 1998). In this chapter we will concentrate on some of the features that make the plant Golgi unique, such as the motility, the interface with the ER and mechanisms of transport within and from the stack.

Figure 1. Pathways from the Golgi in plant cells: a map showing the major pathways from the Golgi apparatus (G) in plants and excluding the endocytic pathway. (1) Uncoated secretory vesicles carry material to the plasma membrane. (2) Dense vesicles carry storage proteins to the protein storage vacuole (PSV). (3) Unknown carriers may transport stromal proteins from the Golgi to chloroplasts (C). (4) Clathrin coated vesicles (CCVs) carry cargo destined for the lytic vacuole (LV) via a prevacuolar compartment (PVC). The ER and nucleus (N) are also depicted in the schematic.

612

*

C. Hawes et al.

Golgi body motility – the missing ER link? The development of fluorescent protein technology and plant transformation techniques has resulted in an explosion of data pertaining to plant Golgi body dynamics over the past decade, most of which have been derived from studies on tobacco epidermal and BY-2 cells. In 1998 Boevink et al. provided the first evidence of Golgi body dynamics in vivo in tobacco leaf epidermal cells, where Golgi bodies containing a fluorescent fusion to the signal anchor sequence of a rat sialyl transferase appeared to display a range of motilities; stationary, slow, fast, unidirectional and bidirectional. The protein and membrane trafficking equilibrium between the ER and Golgi bodies, through anterograde and retrograde trafficking, functionally links the two compartments within the cell. Studies correlating plant Golgi body movement with the cortical ER network indicate that they are associated with the three way junctions, and a vast majority appear to move along the ER tubular network (see movie A online*). Occasionally, Golgi stacks appear to ‘break free’ of the underlying cortical ER network resulting in the ER tubules following the same track to make new tubular connections (Brandizzi et al. 2002). These early studies unexpectedly indicated that there was an intimate association between the highly motile Golgi bodies and the ER, and that they could potentially affect ER remodelling. The observations raised many questions as to the nature of the Golgi and ER motors, the relationship between cargo exit sites on the ER to the Golgi bodies, and the mechanisms for tethering Golgi to the ER. Drug inhibition studies perturbing actin or tubulin polymerisation dynamics have shown that, plant Golgi body movement is dependent upon actin, not € hr et al. 1999). This alongside microtubules (Brandizzi et al. 2002; Nebenfu imaging Golgi stacks on fluorescent actin networks (Brandizzi et al. 2002) has led to speculation that the active movement of Golgi bodies is driven by a myosin motor, rather than a microtubule-based motor such as kinesin. Even though arabidopsis is predicted to contain 17 genes encoding for myosins (Reddy et al. 2001), there are few reports of Golgi associated myosins and only one report of a myosin tail fragment partially labelling plant Golgi stacks (Li € hr 2007). However, over expression and knockdown of plant and Nebenfu myosin XI-K in tobacco and XI-2 (MYA2) and XI-K in arabidopsis, has been shown to perturb Golgi, peroxisome and mitochondrial movement, suggesting a role of these specific motors in small organelle motility (Avisar et al. 2008; Peremyslov et al. 2008; Sparkes et al. 2008). Other cytoskeletonassociated proteins have been reported associated with the plant Golgi such as the actin-binding protein KATAMARI 1/MURUS3 (Tamura et al. 2005) and a kinesin-13A (Lu et al. 2005), although neither appeared to have a role in Golgi motility. Studies in mammals have shown that several myosin and kinesin motors are associated with the Golgi stack, and are required for stack maintenance rather than motility (Allan et al. 2002). Perhaps there is a similar interplay between kinesins and myosins in plants?

Features of the plant Golgi apparatus

*

613

Depolymerisation of the actin cytoskeleton does not immediately affect the overall structure of the ER network, but does appear to stop movement and growth of ER tubules as well as Golgi movement (Boevink et al. 1998). An interesting hypothesis suggested that Golgi bodies physically linked to the ER membrane are motile due to the flow of the ER membrane itself and are effectively dragged around the surface of the ER (Hawes and SatiatJeunemaitre 2005). This hypothesis was tested using a fluorescent protein tool to monitor the movement of specific activated pools of protein within the ER membrane. Photoactivatable GFP (Patterson and Lippincott-Schwartz 2002), which can be ‘switched on’ to order, was spliced to the transmembrane domain of calnexin and demonstrated active vectorial actin-dependent movement of the chimeric protein in the ER membrane. Upon actin depolymerisation movement was no longer active but restricted to diffusion. By monitoring the movement of this ER membrane protein with fluorescent Golgi bodies in the same region, it was shown that the Golgi displayed similar rates and directional movement to the underlying ER membrane protein marker (Runions et al. 2006). These results implicated the ER in the active movement of Golgi bodies, and beg the question, as to whether the movement of the ER and Golgi bodies could be controlled by the same motor, with the ER membrane dragging the Golgi attached to exit sites over its surface, or vice-versa. Or, is the movement controlled by a combination of Golgi-associated and ER-associated motors? Another interesting question is the nature of the structures holding the Golgi stacks together during these highly dynamic movements in the cell, which one might expect to literally tear the stacks apart. Candidates that might play a role in holding the stacks together are Golgi matrix proteins with coiled-coil domains (Latijnhouwers et al. 2007, see later section).

The ER–Golgi interface In many cell types it is clear from the expression of ER and Golgi targeted fluorescent protein constructs that there is a very close relationship between the two organelles. Work on tobacco leaf epidermal cells has led to the development of the concept of the Golgi and the components of the ER exit sites forming a mobile secretory unit which collects secretory cargo from the ER (daSilva et al. 2004). This hypothesis was based on the observation that the COPII coat initiating GTPase Sar1 almost exclusively co-located with the Golgi on the ER of tobacco leaf cells. However, this model is not universally recognised, as it has also been suggested from work on tobacco suspension culture cells that not all Golgi stacks are associated with exit sites (Yang et al. 2005) and that the whole of the ER surface is export competent, and the Golgi travel across it from exit site to exit site (Robinson et al. 2007). This Golgi/ER relationship appears most developed in highly vacuolated cells such as elongate root cortical cells, those of the hypodermis and the leaf epidermis (movies A, C on online*). In more isodiametric meristematic cells it is clear that

614

*

C. Hawes et al.

in many cases many numerous Golgi stacks are not associated with the ER and show little movement (Hawes unpublished data, See movie B online*). How such Golgi receive cargo from the ER is unknown. Do they dock onto ER exit sites or are there long range carriers transporting cargo from the ER to Golgi? It could even be possible that these Golgi are formed on the cortical ER, break loose and act as independent organelles synthesising complex wall carbohydrates and maintaining their membrane equilibrium via an endocytic pathway (Hawes and Satiat-Jeunemaitre 2005). However, to date there is no live cell imaging data to substantiate any of these theories. Transport of membrane cargo from the Golgi to the ER has been demonstrated by the use of fluorescence recovery after photobleaching (FRAP, Brandizzi et al. 2002; daSilva et al. 2004). It is clear that material can transfer between the two organelles when the Golgi is static or motile and this transfer does require energy but is independent of the cytoskeleton. Database mining has shown that most of the proteins of mammalian ER exit sites are also present in plants (Robinson et al. 2007) and biochemically that a COPII-mediated transport of cargo may occur between the ER and Golgi (Phillipson et al. 2001). Sec12, the exchange factor for the small GTPase Sar1p which initiates COPII coat formation, appears to be distributed over the entire ER membrane, whereas it has been shown that Sar1p itself, depending on the isoform of the protein, when expressed as a fluorescent protein construct can co-locate with the Golgi (Brandizzi et al. 2002) or be distributed over the ER (Robinson et al. 2007). The COPII coat protein complexes certainly exist in plants as demonstrated from the expression of fluorescent Sar1 constructs (daSilva et al. 2004; Yang et al. 2005) and Sec24 constructs (Hanton et al. 2007a) in tobacco leaves. However, it is still a matter of debate as to whether COPII vesicles functioning as the ER-to-Golgi vectors are actually formed. There are few reports in the literature of such vesicles in higher plants and then only from meristematic cells in high pressure frozen freeze-substituted material (Donohoe et al. 2007). COPII components were identified by immunolabelling with a Sar1 antibody, although the specificity of the antibody was not demonstrated. In systems which display an extremely close relationship between the ER and Golgi, it is debatable as to whether a vesicle-based vector as opposed to some form of direct continuity, initiated by COPII-induced curvature of the ER membrane, would be necessary to affect cargo transfer (Hawes and SatiatJeunemaitre 2005). Indeed the exclusivity of COPII vesicles as the vector for ER exit is also being challenged in mammalian cells (Watson and Stephen 2005; Tang et al. 2005). Interestingly, in an ultrastructural study of Golgi reformation in BY-2 cells after BFA treatments there was no evidence of COPII vesicles in the regions of vesicular tubular clusters, which represented the first stages of stack construction (Langhans et al. 2007). These clusters did not label with antisera to the COPII protein Sec13, but did contain ARF1 and g-COP. Whatever the nature of the transport vectors, tethering and fusion proteins are likely to be required to mediate fusion between the ER derived

Features of the plant Golgi apparatus

*

615

transport vector and cis-Golgi membranes. A number of plant SNARE homologues have been located to the Golgi and to the ER including those representing the well known Q- and R-SNARES of the ER Golgi interface, Sec22, Sed5 (AtSYP31), Memb 11(Bos 1) and Bet 11 (Chatre et al. 2005; Moreau et al. 2006). What is remarkable about the individual cisternal stacks in plant cells is, that not only can they traffic over the ER surface at speeds of several microns per second, but they can also travel within major cytoplasmic streams, such as found in trans-vacuolar strands, root hairs and pollen tubes at many microns a second and still maintain their three-dimensional organisation. Thus, the concept of a matrix that organises Golgi stack structure is even more compelling in plants than in mammals. To date, a number of potential matrix proteins have been identified from the arabidopsis database (Latijnhouwers et al. 2005a, Fig. 2) and a number of these co-locate with Golgi markers when expressed as fluorescent protein fusions (Renna et al. 2005; Latijnhouwers et al. 2005b, 2007; Stefano et al. 2006, Fig. 2). Some of these appear to be located at or towards the cis-face of the Golgi stacks including the plant homologues of p115, golgin-84 and CASP. One model of the plant ER–Golgi interface would hypothesise that cisGolgi matrix proteins may form a structural bridge between the ER and Golgi, thus preventing the Golgi stacks from drifting away from the ER export sites. Preliminary evidence (Osterrieder and Hawes unpublished) has shown that after deconstruction of GFP-marked Golgi in tobacco leaf epidermal cells by the controlled expression of a GTP-locked form of Sar1p (Sar1p-GTP) ER to Golgi transport is inhibited and the Golgi stacks disappear, giving a Brefeldin A phenotype of fluorescent ER. In these cells, if fluorescent fusions of Sar1GTP (daSilva et al. 2004), CASP or golgin-84 (from Latijnhouwers et al. 2007) are also expressed, they can be seen as punctate structures on the ER surface after dissolution of the Golgi. This implies that both exit site proteins and cisGolgi matrix proteins may mark exit sites on the ER even in the absence of Golgi stacks.

Biogenesis of the Golgi stack As discussed by Langhans et al. (2007) it is often assumed that Golgi bodies replicate by fission, but without much published evidence, although conventional EM studies have suggested that the plant-like stacks in the filamentous Oomycetes indeed divide by fission (Bracker et al. 1996). There is no evidence of breakdown of the Golgi and segregation of components into daughter cells during mitosis as occurs in animal cells (Altan-Bonnet et al. 2006), so there is no need to reassemble the system during the division process. However, it is known that after dissolution and reabsorption of Golgi membranes into the ER after Brefeldin A treatment, on removal of the drug Golgi stacks can reform de novo from the ER (Saint-Jore et al. 2002; Langhans et al. 2007). This phenomenon is independent of the cytoskeleton and does not require the

616

*

C. Hawes et al.

Figure 2. Plant Golgi matrix proteins: Confocal images showing transient co-expression of the Golgi marker ST-mRFP/YFP (shown in magenta) and GFP-tagged golgins (shown in green) in wild-type tobacco leaf epidermal cells. (A) GFP-AtCASP (cis/medial Golgi located) and ST-mRFP. (B) GFP-golgin-84 (cis/medial Golgi located) and ST-mRFP. (C) AtGRIP-GFP (trans-Golgi located) and ST-mRFP. (D) GFP-TMF (trans-Golgi located) and ST-mRFP. (E) p115-GFP (cis/medial-Golgi located) and ST-YFP. Scale bars¼2 mm.

Features of the plant Golgi apparatus

*

617

biosynthesis of new Golgi proteins. Thus, it is tempting to speculate that the ER, as may be the case in mammalian cells (Puri and Linstedt 2003), has the capacity to generate new Golgi stacks de novo. This may be the mechanisms by which Golgi number is maintained in a population of dividing cells or the ratio of Golgi to cytoplasm is kept constant in growing cells (Hawes and SatiatJeunemaitre 2005). In support of this hypothesis it has been demonstrated that ER export sites can be generated de novo by the over expression of a Golgi-targeted membrane cargo (Hanton et al. 2007a). Expression of the HDEL receptor homologue ERD2 and a chimeric type I transmembrane protein TMcCCASP with a diacidic ER export motif resulted in an increase in punctate sites on the ER labelled by the COPII protein Sec24, whereas expression of a secretory cargo sec-GFP had no effect on the number of exit sites (Hanton et al. 2007a). One then has to assume that de novo formation of export sites results in the rapid formation of Golgi stacks as evidenced by the fact that the Sec24 marker always co-localised with ERD2, which is a well established Golgi marker. Some more recent data shows that in higher plant cells the Golgi may not only be able to form de novo from the ER but also have the ability to divide by fission (Langhans et al. 2007). In BY-2 cells after BFA treatment Golgi membranes were redistributed into the ER as expected. After washout of the drug the Golgi reassembled from the ER, first forming a tubulo-vesicular cluster of membrane that then differentiated into discrete cisternae. Surprisingly, the Golgi stacks continued to grow beyond their normal diameter of 0.65 mm until they reached 1.25 mm and at that stage in development divided by fission. Whether during the normal growth process of these cells Golgi form in this manner is unknown. What is worth noting though is, that during the early stages of Golgi differentiation from the ER, there was no evidence of the formation of discrete COPII vesicles.

Transport within and exit from the Golgi The mechanisms by which membrane and cargo traverse the cisternae on Golgi stacks are by no means clear. Unfortunately, the plant community has adopted the tendency to apply whatever the current model in vogue for the mammalian Golgi to explain the mechanics of the plant Golgi (Staehelin and Moore 1995). The original proponents of the cisternal maturation model were plant scientists working on Golgi-derived algal cell surface scales (Becker et al. 1995), as this was the only model that could explain how large micron diameter scales could traverse the Golgi stack and also the exclusion of smaller scales from peripheral Golgi vesicles (Hawes and Satiat-Jeunemaitre 1996). This model was superseded by almost universal adoption in the 1980s of the vesicle trafficking model stating that ER-to-Golgi transport is mediated by COPII vesicles and anterograde transport within the stack is mediated by COPI coated vesicles (Balch et al. 1984; Bonifacino and Glick 2004). Although since the discovery of the retrograde action of COPI vesicles (Letourneur et al. 1994)

618

*

C. Hawes et al.

a number of different mechanisms explaining trans-Golgi traffic have been proposed including tubular connections between individual cisternae (Trucco et al. 2004). As in mammalian cells the cisternal margins of plant Golgi stacks are populated by a number of vesicle types depending on the biosynthetic activity of the tissue. It is widely accepted that COPI vesicles are active in plants and the majority of COPI components and regulating molecules such as ARF1 and its exchange factors have been localised to the Golgi (Richter et al. 2007; Robinson et al. 2007). As yet no evidence has been published demonstrating the direction of transport of these vesicles, although it has been suggested from morphological examination that, similar to mammalian cells, two morphologically distinct COPI vesicle populations may exist (Donohoe et al. 2007). The existence of Golgi tubules that may be responsible for cargo transport as seen by conventional EM fixation and selective staining (Cunningham et al. 1966; Harris and Oparka 1983), has yet to be confirmed by cryo-techniques. It has been shown that some secretory cargos such as seed storage proteins can accumulate into dense transport vesicles attached to cisternal rims as early as the cis-cisternae, but probably do not exit the stacks until they reach the trans-face (Hillmer et al. 2001; Robinson et al. 2005). Unless such large vesicles are tethered and swing between cisternae as they traverse the stack, such data provides reasonable evidence in support of the cisternal maturation model. Exit from the Golgi stack is towards the trans-face or at the trans-Golgi network (TGN). Whilst transport of proteins to the lytic and storage vacuole is likely to be receptor-mediated and facilitated by a number of different targeting mechanisms (Robinson et al. 2005; Hinz et al. 2007), to date no targeting mechanisms have been revealed for transport of proteins to the plasma membrane which has been suggested to be the default pathway out of the Golgi (Hadlington and Denecke 2000). Indeed one of the key features of the plant Golgi is heavy secretion to the cell surface due to the production of many of the complex polysaccharide components of the cell wall (Hawes et al. 1996; Staehelin and Moore 1995). However, transport does not appear to be restricted just to the vacuoles and plasma membrane, as convincing evidence has been published suggesting an ER-to-chloroplast route for the transport of a chloroplast stroma located carbonic anhydrase, which is via the Golgi (Villarejo et al. 2005). Therefore, yet another unknown post-Golgi targeting mechanism must exist in plants. The exact nature of the trans-Golgi network in plants is still a matter for debate (Hawes and Satiat-Jeunemaitre 2005). Certainly a clathrin coated tubular network can often be seen attached to the trans-Golgi and often a similar network is observed separate from the Golgi. Evidence shows that such a network, marked by specific SNAREs (Uemura et al. 2004) may also be part of the early endosomal system (Dettmer et al. 2006; Hanton et al. 2007b; Lam et al. 2007). It is possible that the TGN is formed at the Golgi and at the right

Features of the plant Golgi apparatus

*

619

developmental stage sloughs off to form an endosomal/prevacuolar organelle in its own right (Hawes and Satiat-Jeunemaitre 2005).

Conclusions From this brief review we can conclude that although at the molecular level many homologues of the genes and proteins found in and associated with mammalian and fungal (yeast) Golgi are present in plants, the phenotypic manifestation of these has resulted in the evolution of an organelle displaying many unique features. There are however many questions which remain unanswered about the mechanics of the plant Golgi stack including the nature of the proposed ER-cis-Golgi tether, the role of the matrix proteins in maintaining cisternal integrity, the sorting, packaging and targeting mechanisms for exocytosed cargo and of course the exact source of the motive force behind Golgi movement. Acknowledgments. We thank the BBSRC for grant funding which supports much of the work in our laboratory, Maita Latijnhouwers and Claudine Carvalho for cloning the matrix proteins and making fluorescent protein fusions, and John Runions for generating dual coloured arabidopsis plants. *Movies can be viewed online at: www.springer.com/springerwiennewyork/lifeþ sciences/book/978-3-211-76309-4

References Allan VJ, Thompson HM, McNiven MA (2002) Motoring around the Golgi. Nat Cell Biol 4: E236–E242 Altan-Bonnet N, Sougrat R, Liu W, Snapp EL, Ward T, Lippincott-Schwartz J (2006) Golgi inheritance in mammalian cells is mediated through endoplasmic reticulum export activities. Mol Biol Cell 17: 990–1005 Avisar D, Prokhnevsky AJ, Makarova KS, Koonin EV, Dolja VV (2008) Myosin XI-K is required for rapid trafficking of Golgi stacks, peroxisomes and mitochondria in leaf cells of Nicotiana benthamiana. Plant Physiol DOI: 10.1104/pp.107.113647 Balch WE, Glick BS, Rothman JE (1984) Sequential intermediates in the pathway of intercompartmental transport in a cell-free system. Cell 39: 525–536 € linger B, Melkonian M (1995) Anterograde transport of algal scales through Becker B, Bo the Golgi complex is not mediated by vesicles. Trends Cell Biol 5: 305–307 Boevink P, Oparka K, Santa-Cruz S, Martin B, Betteridge A, Hawes C (1998) Stacks on tracks: the plant Golgi apparatus traffics on and actin/ER network. Plant J 15: 441–447 Bonifacino JS, Glick BS (2004) The mechanisms of vesicle budding and fusion. Cell 116: 153–166  DJ, Grove SN (1996) Structure, differentiation, and multiplication of Bracker CE, Morre Golgi apparatus in fungal hyphae. Protoplasma 194: 250–274 Brandizzi F, Snapp EL, Roberts AG, Lippincott-Schwartz J, Hawes C (2002) Membrane protein transport between the endoplasmic reticulum and the Golgi in tobacco leaves is energy dependent but cytoskeleton independent: evidence from selective photobleaching. Plant Cell 14: 1293–1309 Chatre L, Brandizzi F, Hocquellet A, Hawes C, Moreau P (2005) Sec22 and Memb11 are v-SNAREs of the anterograde endoplasmic reticulum – Golgi pathway in tobacco leaf epidermal cells. Plant Physiol 139: 1244–1254

620

*

C. Hawes et al.

 DJ, Mollenhauer HH (1966) Structure of isolated plant Golgi Cunningham WP, Morre apparatus revealed by negative staining. J Cell Biol 28: 169–179 Dettmer J, Hong-Hermesdorf A, Stierhof YD, Schumacher K (2006) Vacuolar H+-ATPase activity is required for endocytic and secretory trafficking in Arabidopsis. Plant Cell 18: 715–730 Donohoe BS, Kang B-H, Staehelin LA (2007) Identification and characterization of COPIa- and COPIB-type vesicle classes associated with plant and algal Golgi. PNAS 104: 163–168 Fowke LC, Tanchak MA, Galway ME (1991) Ultrastructural cytology of the endocytotic pathway in plants. In: Hawes C, Coleman JOD, Evans DE (eds) Endocytosis, exocytosis and vesicle traffic in plants. OUP, Oxford, pp 15–40 Hadlington JL, Denecke J (2000) Sorting of soluble proteins in the secretory pathway of plants. Curr Opin Plant Biol 3: 461–468 Hanton SL, Chatre L, Renna L, Matheson LA, Brandizzi F (2007a) De novo formation of plant endoplasmic reticulum export sites is membrane cargo induced and signal mediated. Plant Physiol 143: 1640–1650 Hanton S, Matheson L, Chatre L, Rossi M, Brandizzi F (2007b) Post-Golgi protein traffic in the plant secretory pathway. Plant Cell Rep 26: 1431–1438 Harris N, Oparka K (1983) Connections between dictyosomes, ER and GERL in cotyledons of mung bean (Vigna radiata L). Protoplasma 114: 92–102 Hawes C, Faye L, Satiat-Jeunemaitre B (1996) The Golgi apparatus and pathways of vesicle trafficking. In Smallwood M, Know JP, Bowles DJ (eds) Membranes: specialized functions in plants. Bios Scientific Publishers, Oxford, pp 337–365 Hawes C, Satiatt-Jeunemaitre B (1996) Stacks of questions: how does the plant Golgi work? Trends Plant Sci 1: 395–401 Hawes C, Satiat-Jeunemaitre B (2005) The plant Golgi apparatus: going with the flow. Biochim Biophys Acta 1744: 93–107 Hillmer S, Movafeghi A, Robinson DG, Hinz G (2001) Vacuolar storage proteins are sorted in the cis-cisternae of the pea cotyledon Golgi apparatus. J Cell Biol 152: 41–50 Hinz G, Colanesi S, Hillmer S, Rogers JC, Robinson DG (2007) Localization of vacuolar transport receptors and cargo proteins in the Golgi apparatus of developing Arabidopsis embryos. Traffic 8: 1452–1464 Kondylis V, Spoorendonk KM, Rabouille C (2005) dGRASP localisation and function in the early exocytotic pathway in Drosophila S2 Cells. Mol Biol Cell 16: 4061–4072 Lam SK, Tse YC, Robinson DG, Jiang L (2007) Tracking down the elusive early endosome. Trends Plant Sci 11: 497–504 Langhans M, Hawes C, Hillmer S, Hummel E, Robinson D (2007) Golgi regeneration after Brefeldin A treatment in BY-2 cells entails stack enlargement and cisternal growth followed by division. Plant Physiol 145: 527–538 Latijnhouwers M, Hawes C, Carvalho C (2005a) Holding it all together? Candidate proteins for the plant Golgi matrix. Curr Opin Plant Biol 8: 632–639 Latijnhouwers M, Hawes C, Carvalho C, Oparka K, Gillingham AK, Boevink P (2005b) An Arabidopsis GRIP domain protein locates to the trans-Golgi and binds the small GTPase ARL1. Plant J 44: 459–470 Latijnhouwers M, Gillespie T, Boevink P, Kriechbaumer V, Hawes C, Carvalho CM (2007) Localization and domain characterization of Arabidopsis golgin candidates. J Exp Bot 58: 4373–4386 Letourneur F, Gaynor EC, Hennecke S, Demolliere C, Duden R, Emr SD, Riezman H, Cosson P (1994) Coatomer is essential for retrieval of dilysine-tagged proteins to the endoplasmic reticulum. Cell 79: 1199–1207 € hr A (2007) Organelle targeting of myosin XI is mediated by two globular Li J, Nebenfu tail domains with separate cargo binding sites. J Biol Chem 282: 20593–20602

Features of the plant Golgi apparatus

*

621

Lu L, Lee Y-R J, Pan R, Maloof JN, Liu B (2005) An internal motor kinesin is associated with the Golgi apparatus and plays a role in trichome morphogenesis in Arabidopsis. Mol Biol Cell 16: 811–823 Moreau P, Brandizzi F, Hanton S, Chatre L, Melser S, Hawes C, Satiat-Jeunemaitre B (2006) The plant ER-Golgi interface: a highly structured and dynamic membrane complex. J Exp Bot 58: 49–64 € hr A, Gallagher LA, Dunahay TG, Frohlick JA, Mazukiewicz AM, Meehl JB, Nebenfu Staehelin LA (1999) Stop and go movements of plant Golgi stacks are mediated by the acto-mysin system. Plant Physiol 121: 1127–1141 Neumann U, Brandizzi F, Hawes C (2003) Protein transport in plant cells: in and out of the Golgi. Ann Bot 92: 167–180 Patterson GH, Lippincott-Schwartz J (2002) A photoactivatable GFP for selective photolabeling of proteins and cells. Science 297: 1873–1877 Peremyslov VV, Prokhnevsky AI, Avisar D, Dolja VV (2008) Two class XI myosins function in organelle trafficking and root hair development in Arabidopsis thaliana Plant Physiol DOI:10.1104/pp.107.113654 Phillipson BA, Pimpl P, daSilva LL, Crofts AJ, Taylor JP, Movafeghi A, Robinson DG, Denecke J (2001) Secretory bulk flow of proteins is efficient and COPII dependent. Plant Cell 13: 2005–2020 Puri S, Linstedt AD (2003) Capacity of the Golgi apparatus for biogenesis from the endoplasmic reticulum. Mol Biol Cell 14: 5011–5018 Reddy ASN, Day IS (2001) Analysis of the myosins encoded in the recently completed Arabidopsis thaliana genome sequence. Genome Biology 2: 1–18 Renna L, Hanton SL, Stefano G, Bortolotti L, Misra V, Brandizzi F (2005) Identification and characterisation of AtCasp, a plant transmembrane Golgi matrix protein. Plant Mol Biol 58: 109–122 € rgens G Richter S, Geldner N, Schrader J, Wolters H, Stierhof Y -D, Rios G, Robinson D, Ju (2007) Functional diversification of closely related ARF-GEFs in protein secretion and recycling. Nature 448: 488–493 Robinson DG, Herranz M-C, Bubeck J, Peppercock R, Ritzenthaler C (2007) Membrane dynamics in the early secretory pathway. Crit Revs Plant Sci 26: 199–225 Robinson DG, Oliviusson P, Hinz G (2005) Protein sorting to the storage vacuoles of plants: a critical appraisal. Traffic 6: 615–625 € hner T, Hawes C (2006) Photoactivation of GFP reveals protein Runions J, Brach T, Ku dynamics within the endoplasmic reticulum membrane. J Exp Bot 57: 43–50 Saint-Jore CM, Evins J, Batoko H, Brandizzi F, Moore I, Hawes C (2002) Redistribution of membrane proteins between the Golgi apparatus and endoplasmic reticulum in plants is reversible and not dependent on cytoskeletal networks. Plant J 29: 661–678 DaSilva LLP, Snapp EL, Denecke J, Lippincott-Schwartz J, Hawes C, Brandizzi F (2004) Endoplasmic reticulum export sites and Golgi bodies behave as single mobile secretory units in plant cells. Plant Cell 16: 1753–1771 Sparkes IA, Teanby NA, Hawes CR (2008) Truncted myosin XI tail fusions inhibit peroxisome, Golgi, and mitochondrial movement in tobacco leaf epidermal cells: a genetic tool for the next generation. J Exp Bot 59: 2499–2512 Staehelin LA, Moore I (1995) The plant Golgi apparatus: structure, functional organization and trafficking mechanisms. Ann Rev Plant Physiol Mol Biol 46: 261–288 Stefano G, Renna L, Hanton L, Chatre L, Haas TA, Brandizzi F (2006) ARL1 plays a role in the binding of the GRIP domain of a peripheral matrix protein to the Golgi apparatus in plant cells. Plant Mol Biol 61: 431–449 Tamura K, Shimada T, Kondo M, Nishimura M, Hara-Nishimura I (2005) KATAMARI 1/ MURUS3 is a novel Golgi membrane protein that is required for endomembrane organisation in Arabidopsis. Plant Cell 17: 1764–1776 Tang BL, Wang Y, Ong YW, Hong W (2005) COPII and exit from the endoplasmic reticulum. Biochim Biophys Acta 1744: 293–303

622

*

C. Hawes et al.

Trucco A, Polishchuk RS, Martella O, Di Pentima A, Fusella A, Di Glandomenico D, San Pietro E, Beznoussenko GV, Polishchuk E, Baldassarre M, Buccione R, Geerts WJC, Koster AJ, Burger KNJ, Mironov AA, Luini A (2004) Secretory traffic triggers the formation of tubular continuities across Golgi sub-compartments. Nature Cell Biology 6: 1071–1081 Uemura T, Ueda T, Ohniwa RL, Nakano A, Takeyasu K, Sata MH (2004) Systematic analysis of SNARE molecules in Arabidopsis: dissection of the post-Golgi network in plant cells. Cell Struct Funct 29: 49–65 n S, Larsson S, De jardin A, Monne  M, Rudhe C, Karlsson J, Jansson S, Villarejo A, Bure Lerouge P, Rolland N, Von Heijne G, Grebe M, Bako L, Samuelsson G (2005) Evidence for a protein transported through the secretory pathway en route to the higher plant chloroplast. Nat Cell Biol 7: 1224–1231 Watson P, Stephens DJ (2005) ER-to-Golgi transport: form and formation of vesicular and tubular carriers. Biochim Biophys Acta 1744: 304–315 Yang YD, el Amawi R, Bubeck J, Pepperkok R, Ritzenthaler C, Robinson DG (2005) Visualization of COPII and Golgi dynamics in Nicotiana tabacum BY-2 cells provides evidence for transient association of Golgi stacks with ER exit sites. Plant Cell 17: 1513–1531

Yeast Golgi apparatus

*

623

Yeast Golgi apparatus Akihiko Nakano

Introduction There is no doubt that the budding yeast Saccharomyces cerevisiae has played leading roles in elucidating many problems of protein secretion and membrane traffic. The problem of the Golgi apparatus is not an exception; the yeast system has contributed much, but in a little unique way. The Golgi apparatus of yeast was first described as a subject of molecular cell biology in the analysis of S. cerevisiae sec mutants (Novick et al. 1980, 1981). Two temperature-sensitive mutants, sec7 and sec14, accumulated peculiar membrane structures at the restrictive temperature, which were named Berkeley bodies according to the place of discovery. Furthermore, the sec7 mutant cells built up what looked like stacks of the Golgi cisternae, when the cells were incubated under a certain condition (low glucose), leading to the conclusion that the Berkeley bodies were deformed Golgi structures (Novick et al. 1981). These mutants accumulate secretory and vacuolar proteins in highly glycosylated forms, which proved to be Golgimodified (Esmon et al. 1981; Stevens et al. 1982). The SEC7 and SEC14 genes turned out to encode a guanine nucleotide exchange factor (GEF) for Arf GTPases (Achstetter et al. 1988; Sata et al. 1998) and a phospholipid transfer protein (Bankaitis et al. 1989, 1990), respectively. Many other gene products are now known to function in the trafficking in and around the Golgi (Table 1). However, extensive description of these proteins is not the purpose of this short review.

Compartmentalization Like other organisms, the yeast Golgi apparatus can be functionally divided into early (cis), medial and late (trans) regions. Early glycosylation enzymes such as a1,6-mannosyltransferases (Och1, Mnn9, etc.) define the cis region and later enzymes such as a1,2-mannosyltransferase (Mnn2) and a1,3-mannosyltransferase (Mnn1) define later (medial and trans) compartments. Components of trafficking machinery also reside in different regions of the Golgi. Golgi-to-ER retrieval receptors Erd2 (HDEL receptor) (Semenza et al. 1990) and Rer1 (retrieval receptor for ER transmembrane proteins) (Sato et al. 1995, 1997, 2001) are in the cis Golgi and an Arf GEF Sec7 is mostly localized to the trans Golgi (Franzusoff et al. 1991). SNAREs also distribute differently; Sed5 (Hardwick and Pelham 1992) is mainly in the cis part and the distribution

624

*

A. Nakano

Table 1. Important yeast proteins involved in trafficking in and around the Golgi GTPase GEF GAP COPII coat COPI coat Clathrin and AP-1 Retromer SNARE Tether TRAPP I TRAPP II Cargo receptor Retrieval receptor p21 family Others

Sar1, Arf1, Arf2, Ypt1, Ypt31, Ypt32, Ypt6, Arl1, Arl3 Sec12, Gea1, Gea2, Sec7, TRAPP I, TRAPP II Sec23, Gcs1, Glo3, Age1, Age2, Gyp1 Sec23, Sec24, Sec13, Sec31, Sfb2/Iss1, Sfb3/Lst1 Ret1, Sec26, Sec27, Sec21, Ret2, Sec28, Ret3 Chc1, Clc1, Apl2, Apl4, Apm1/2, Aps1, Gga1, Gga2 Vps5, Vps17, Vps26, Vps29, Vps35 Bet1, Bos1, Sec22, Sed5, Ykt6, Sft1, Gos1, Vti1, Tlg1, Tlg2 Uso1, Imh1, Cog1–Cog8 Bet3, Bet5, Trs20, Trs23, Trs31, Trs33, Trs85 TRAPPI þ Trs65, Trs120, Trs130 Erv14, Erv26, Erv29, Emp46, Emp47 Erd2, Rer1 Emp24, Erv25, Erp1–Erp6 Sec14, Sec16, Sly1, Yip1, Yif1, Yos1, Grh1

of Gos1 (McNew et al. 1998) peaks in the medial region. This order of locations has been determined by mutual colocalization studies either by biochemical fractionation or by microscopic methods using antibodies or fusions to fluorescent proteins.

Unique features During these compartmentation studies, researchers began to realize that yeast Golgi cisternae behave differently according to their nature of cis/ medial/trans. For example, in subcellular fractionation by density gradient ultracentrifugation, late cisternae tend to behave denser than early cisternae (Whitters et al. 1994). Immunofluorescence microscopy often showed nonoverlapping patterns for different Golgi proteins (Franzusoff et al. 1991; Hardwick and Pelham 1992). In other words, individual Golgi cisternae of the yeast Saccharomyces appeared to be separate entities and behave independently from each other. Electron micrography also indicated that wild-type yeast cells seldom exhibit stacked Golgi cisternae (Preuss et al. 1992), even though sec7 mutant cells appear to pile up stacks of membranes. As will be described in the next section (see Chapter 4.3) in more detail, three-dimensional electron microscopic studies revealed amazing characteristics of the yeast Golgi. Saccharomyces Golgi cisternae, which do not form stacks, are not simple flattened bladders or saccules but are elaborate networks of tubules (Rambourg et al. 1995, 2001).

Stacked or unstacked In the meantime, studies on other yeast species such as Schizosaccharomyces pombe and Pichia pastoris described that these yeasts display

Yeast Golgi apparatus

*

625

stacked Golgi, more reminiscent to those of higher eukaryotes. Indeed in P. pastoris, Golgi stacks are easily observed by conventional electron microscopy unlike the case of S. cerevisiae (Glick 1996). Glick’s group took advantage of this fact and started studies on the behaviors of the Golgi in Pichia. It is a big mystery why some organisms such as S. cerevisiae do not require stacked Golgi like other organisms. More remarkable organization of the Golgi ribbons as seen in higher animal cells can be explained by the action of the dynein motor, which brings immature Golgi units toward the perinuclear microtubules organizing center (centrosomal region), and is considered to be important for highly polarized secretion and cell motility. However, the reason why each Golgi unit takes a stacked structure in most organisms is quite puzzling, when one learns that the Saccharomyces yeast can live and secrete proteins quite efficiently without stacks. Perhaps stacked cisternae are advantageous over scattered ones in the organization of cisternal maturation (see below). Regarding stacking, a S. cerevisiae homolog (Grh1) of GRASP65, which is implicated in vesicle tethering and stack formation in mammalian cells, has recently been characterized (Behnia et al. 2007), but its role in the organization and function of the Golgi remains unclear.

A model to prove cisternal maturation That the cisternae do not stack is quite a queer nature of the Saccharomyces Golgi, but at the same time it provides us with a great opportunity for live imaging. Because cis, medial and trans cisternae are scattered in the cytoplasm of S. cerevisiae, individual single cisternae are much easier to follow under optic microscopes than in other type of cells with stacked Golgi. The demonstration of cisternal maturation (Losev et al. 2006; Matsuura-Tokita et al. 2006), which settled a long-lasting big debate on how cargo molecules are transported within the Golgi, was an achievement taking this advantage. Two groups, Ben Glick’s and of myself, expressed two Golgi markers with different fluorescent colors, GFP and RFP, simultaneously in living yeast cells and observed the behaviors of individual Golgi cisternae. The results clearly showed that cis cisternae form de novo, change their nature from cis to trans over time, and then dissipate (Fig. 1). The time required for cis-totrans maturation is around 5 min, which is consistent with the rate of cargo transport within the yeast Golgi (Losev et al. 2006). To enable these observations, much effort was made to develop new microscopic systems. The system combining the high-speed spinning-disk confocal scanner and the ultrahigh-sensitivity HARP cameras was particularly powerful and could achieve extremely high resolution in time (5 ms in 2D) and space (50 nm in 3D) with the help of deconvolution analysis (Nakano 2002, 2004; MatsuuraTokita et al. 2006). As shown in Supplementary Materials (sMovie 1 and sFig. 1), such a high-resolution observation unveiled amazingly dynamic

626

*

A. Nakano

Figure 1. A new scheme of Golgi cisternal maturation. As proposed by classic cisternal maturation models, cis cisternae form de novo from ER-derived cargo-carrying vesicles (COPII vesicles) by transferring cis components from the pre-existing compartments. cis cisternae then mature to medial and then to trans cisternae by abandoning earlier components and acquiring later ones. Transfer of Golgi enzymes and other resident proteins (such as SNAREs) is probably mediated by both COPI vesicles and tubular connections. One important feature of maturation that has been unveiled by high-resolution live imaging is that the maturing cisternae undergo very dynamic mixing and re-segregation of membranes (sFig. 1).

mixing and re-segregation of membrane components within a maturing Golgi cisterna.

Transitional ER and Golgi biogenesis Pichia Golgi has several characteristics that differ from Saccharomyces Golgi. They are bigger in size and smaller in number. Furthermore, when the transitional ER (tER; also called ER exit site or ERES) is visualized by COPII markers, Pichia tER represents distinct sites along the ER and Pichia Golgi is almost always located in front of the tER (Rossanese et al. 1999; Bevis et al. 2002), indicating that the vesicle formation from the ER immediately leads to the de novo formation of the Golgi. This raises a very interesting question as to how and why these yeast cells organize the Golgi in different fashions. One key molecule is Sec16, which is involved in the COPII vesicle budding from the ER, because a Pichia sec16 temperature-sensitive mutant shows disorganized tER and disrupted Golgi structures at the restrictive temperature (Connerly et al. 2005). This finding suggests that Pichia Sec16 plays a critical role in the organization of tER and thus the biogenesis of the Golgi. Our recent observation indicates that S. cerevisiae also displays close association of cis Golgi with tER (Matsuura-Tokita and Nakano, unpublished). The difference between the two yeasts may be partly due to different extensiveness of the organization of tER.

Yeast Golgi apparatus

*

627

References Achstetter T, Franzusoff A, Field C, Schekman R (1988) SEC7 encodes an unusual, high molecular weight protein required for membrane traffic from the yeast Golgi apparatus. J Biol Chem 263: 11711–11717 Bankaitis VA, Malehorn DE, Emr SD, Greene R (1989) The Saccharomyces cerevisiae SEC14 gene encodes a cytosolic factor that is required for transport of secretory proteins from the yeast Golgi complex. J Cell Biol 108: 1271–1281 Bankaitis VA, Aitken JR, Cleves AE, Dowhan W (1990) An essential role for a phospholipid transfer protein in yeast Golgi function. Nature 347: 561–562 Behnia R, Barr FA, Flanagan JJ, Barlowe C, Munro S (2007) The yeast orthologue of GRASP65 forms a complex with a coiled-coil protein that contributes to ER to Golgi traffic. J Cell Biol 176: 255–261 Bevis BJ, Hammond AT, Reinke CA, Glick BS (2002) De novo formation of transitional ER sites and Golgi structures in Pichia pastoris. Nat Cell Biol 4: 750–756 Connerly PL, Esaki M, Montegna EA, Strongin DE, Levi S, Soderholm J, Glick BS (2005) Sec16 is a determinant of transitional ER organization. Curr Biol 15: 1439–1447 Esmon B, Novick P, Schekman R (1981) Compartmentalized assembly of oligosaccharides on exported glycoproteins in yeast. Cell 25: 451–460 Franzusoff A, Redding K, Crosby J, Fuller RS, Schekman R (1991) Localization of components involved in protein transport and processing through the yeast Golgi apparatus. J Cell Biol 112: 27–37 Glick BS (1996) Cell biology: alternatives to baker’s yeast. Curr Biol 6: 1570–1572 Hardwick KG, Pelham HR (1992) SED5 encodes a 39-kD integral membrane protein required for vesicular transport between the ER and the Golgi complex. J Cell Biol 119: 513–521 Losev E, Reinke CA, Jellen J, Strongin DE, Bevis BJ, Glick BS (2006) Golgi maturation visualized in living yeast. Nature 441: 1002–1006 Matsuura-Tokita K, Takeuchi M, Ichihara A, Mikuriya K, Nakano A (2006) Live imaging of yeast Golgi cisternal maturation. Nature 441: 1007–1010 € llner TH (1998) McNew JA, Coe JG, Søgaard M, Zemelman BV, Wimmer C, Hong W, So Gos1p, a Saccharomyces cerevisiae SNARE protein involved in Golgi transport. FEBS Lett 435: 89–95 Nakano A (2002) Spinning-disk confocal microscopy – a cutting-edge tool for imaging of membrane traffic. Cell Struct Funct 27: 349–355 Nakano A (2004) Yeast Golgi apparatus – dynamics and sorting. Cell Mol Life Sci 61: 186–191 Novick P, Field C, Schekman R (1980) Identification of 23 complementation groups required for post-translational events in the yeast secretory pathway. Cell 21: 205–215 Novick P, Ferro S, Schekman R (1981) Order of events in the yeast secretory pathway. Cell 25: 461–469 Preuss D, Mulholland J, Franzusoff A, Segev N, Botstein D (1992) Characterization of the Saccharomyces Golgi complex through the cell cycle by immunoelectron microscopy. Mol Biol Cell 3: 789–803  pe  s F (1995) Three-dimensional structure of Rambourg A, Clermont Y, Ovtracht L, Ke tubular networks, presumably Golgi in nature, in various yeast strains: a comparative study. Anat Rec 243: 283–293 Rambourg A, Jackson CL, Clermont Y (2001) Three dimensional configuration of the secretory pathway and segregation of secretion granules in the yeast Saccharomyces cerevisiae. J Cell Sci 114: 2231–2239 Rossanese OW, Soderholm J, Bevis BJ, Sears IB, O’Connor J, Williamson EK, Glick BS (1999) Golgi structure correlates with transitional endoplasmic reticulum

628

*

A. Nakano

organization in Pichia pastoris and Saccharomyces cerevisiae. J Cell Biol 145: 69–81 Sata M, Donaldson JG, Moss J, Vaughan M (1998) Brefeldin A-inhibited guanine nucleotide-exchange activity of Sec7 domain from yeast Sec7 with yeast and mammalian ADP ribosylation factors. Proc Natl Acad Sci USA 95: 4204–4208 Sato K, Nishikawa S, Nakano A (1995) Membrane protein retrieval from the Golgi apparatus to the endoplasmic reticulum (ER): characterization of the RER1 gene product as a component involved in ER localization of Sec12p. Mol Biol Cell 6: 1459–1477 Sato K, Sato M, Nakano A (1997) Rer1p as common machinery for the endoplasmic reticulum localization of membrane proteins. Proc Natl Acad Sci USA 94: 9693–9698 Sato K, Sato M, Nakano A (2001) Rer1p, a retrieval receptor for endoplasmic reticulum membrane proteins, is dynamically localized to the Golgi apparatus by coatomer. J Cell Biol 152: 935–944 Semenza JC, Hardwick KG, Dean N, Pelham HR (1990) ERD2, a yeast gene required for the receptor-mediated retrieval of luminal ER proteins from the secretory pathway. Cell 61: 1349–1357 Stevens T, Esmon B, Schekman R (1982) Early stages in the yeast secretory pathway are required for transport of carboxypeptidase Y to the vacuole. Cell 30: 439–448 Whitters EA, McGee TP, Bankaitis VA (1994) Purification and characterization of a late Golgi compartment from Saccharomyces cerevisiae. J Biol Chem 269: 28106–28117

Yeast Golgi apparatus

*

629

Supplement Figure 1. Four-dimensional imaging of a yeast Golgi cisterna. The maturation of a Golgi cisterna from medial to trans (indicated by arrowhead in sMovie 1) is shown for a series of time points (indicated in seconds). By high-speed and high-resolution confocal measurement and deconvolution, a spatial resolution as high as 50 nm has been achieved. Removal of red components in the form of vesicles and entry of green components in the forms of vesicles and sometimes tubular connection with an adjacent cisterna can be recognized. Taken from the data of supplementary movie 6 of Matsuura-Tokita et al. (2006).

630

*

A. Rambourg et al.

Morphodynamics of the yeast Golgi apparatus  pe  s and Jean-Marc Verbavatz Alain Rambourg, Jean Daraspe, Fran¸cois Ke

Introduction Using metallic impregnation, Camillo Golgi (1898) discovered in the cytoplasm of nerve ganglion cells a new cell organelle that formed an extensive ticulaire interne”. With the empirical perinuclear network : the “appareil re and non-specific methods used for its demonstration, the “Golgi apparatus” was not easily detected in living cells (reviews in Beams and Kessell (1968), Whaley and Dauwalder (1979), Farquhar and Palade (1981) Mollenhauer and  (1991), Berger (1997)). It was the merit of Dalton and Felix (1954) to Morre demonstrate with the electron microscope that the Golgi apparatus was not an artefact. It appeared instead as a system of stacks of closely apposed lamellae or saccules usually encountered in the juxtanuclear area of mammalian cells. With the improvement of the fixation and staining techniques, it became clear that an organelle made up of stacked flattened saccules (cisternae) and vesicles could be seen in most cells. Histochemical techniques at the light microscope level indicated that in some cells (Leblond 1950) the Golgi area reacted with the periodic acid-Schiff technique known to detect glycoproteins in tissues (Leblond et al. 1957). The development of histochemical techniques for the detection of carbohydrates at the ultrastructural level confirmed the presence of these molecules in this organelle and showed in addition that there was a staining gradient across the Golgi stacks (Rambourg et al. 1969) from a poorly reactive “forming” (Mollenhauer and Whaley 1963) or “cis” face (Ehrenreich et al. 1973) to an intensely stained “mature” (Mollenhauer and Whaley 1963) or “trans” face (Ehrenreich et al. 1973). Subsequent radioautographic and biochemical studies revealed that the Golgi apparatus was indeed involved in the elaboration of complex carbohydrates. Thus, biochemical studies on the synthesis of asparagine-linked complex oligosaccharides also indicated that carbohydrates were progressively added from the “cis” to the “trans” face of the Golgi apparatus (review in Kornfeld and Kornfeld 1985). Electron microscope (EM) studies using immunolabeling and cytochemistry usually confirmed these data at the ultrastructural level (reviewed in Roth (1991)) and showed in particular that enzymes involved in terminal glycosylation steps, i.e., galactosyl and sialyl transferases were found in trans-Golgi elements. The observation that Golgi subcompartments could be recognized by the localization of enzymes involved in specific steps of the glycolysation of

Morphodynamics of the yeast Golgi apparatus

*

631

proteins and lipids led to the development of cell-free assays allowing the reconstitution of protein transport through the Golgi apparatus (Fries and Rothman 1980). This technique proved particularly helpful to identify cytosol and membrane factors involved in transport of macromolecules to and through the Golgi stacks and the results were usually thought to support the so-called “vesicular model” according to which transport through the Golgi stacks occurs through the budding, pinching off, targeting and fusion of transport vesicles (reviews in Farquhar and Palade (1981), Farquhar (1985), Rothman (1994)).  In the early 1960s, Mollenhauer et al. (Review in Mollenhauer and Morre (1991)), studying Golgi stacks in thin sections of plant cells or Golgi fractions after negative staining, discovered that, in addition to flattened cisternae and vesicles, Golgi stacks contained tubular projections forming networks of tubules at the periphery of the saccules. They noticed that secretory granules increased in size from the cis to the trans face of the stack and were released at the trans face by rupture of the peripheral tubular network. They also postulated that the loss of saccules from the trans face was counterbalanced by the formation of new saccules at the opposite or cis face of the stack. Thus, in this model, of “saccular maturation”, saccular membrane and secretory products were said to migrate through the Golgi stack as the saccules moved and matured from the cis- to the trans- face of the Golgi apparatus. The formation of new cisternae on the cis face, in contrast, was believed to occur by fusion of small vesicles originating from the rough endoplasmic reticulum. In the above-mentioned EM studies, the use of ultrathin sections only permitted observations on small fragments of the Golgi apparatus which, in some cells, can be extensive and widespread throughout the cytoplasm. It was therefore difficult to provide a comprehensive description of the organelle. In order to overcome this difficulty, Rambourg (1969) examined thick sections, in which cell organelles were selectively stained. The observations confirmed the presence of tubular networks at the periphery of the trans saccules of the Golgi apparatus. Furthermore, when the cis element of the Golgi apparatus of mammalian cells was selectively impregnated with osmium and observed at low magnifications in conventional or high-voltage electron microscopes, it consisted, as previously reported by Golgi, of a single ribbon-like structure (Rambourg et al. 1974). In all cells, this ribbon displayed numerous branches and interconnections making up a single network that in some cells appeared as a spheroidal mass at one pole of the nucleus while, in other cells, it formed a perinuclear network that could extend at considerable distances from the nucleus to reach the periphery of the cell (Rambourg et al. 1974). When all elements of the Golgi apparatus were impregnated with techniques which enhanced the contrast of intracellular membranes, the Golgi ribbon displayed a structural heterogeneity when observed at higher magnifications. Some regions consisted of closely apposed and poorly fenestrated cisternae (saccules) strictly parallel to each other that were referred to as “compact” or “saccular regions” which were continuous with highly fenestrated cisternae

632

*

A. Rambourg et al.

and/or membranous tubules bridging the adjacent compact saccular zones and, thus, referred to as “non-compact” or “tubular” regions (Rambourg et al. 1979; Rambourg and Clermont 1997). Biochemical and genetic studies have extensively used the budding yeast Saccharomyces cerevisiae to investigate various biological processes which are thought to be highly conserved in eukaryotic cells. In yeast, like in plants and mammalian cells, the Golgi apparatus is involved in secretion. Furthermore, temperature-sensitive secretion mutants (sec) have been shown to block traffic at various stages and to accumulate proteins at various points along the secretory pathway in the endoplasmic reticulum (ER), in the Golgi apparatus or at the terminal stages of the secretory process, after release of secretion granules (vesicles) in the cytoplasm (Schekman 1992). Yet, until recently no adequate correlation between biochemical and structural data at the EM level was performed. In double-immunofluorescence labeling experiments, Franzusoff et al. (1991) observed in S. cerevisiae sec7 mutants a significant colocalization of Sec7 protein (Sec7p) implicated in protein transport through the Golgi apparatus and Kex2p, involved in processing within a late Golgi compartment. When examined at the light microscope the markers appeared as small dots dispersed throughout the cytoplasm. Preuss et al. (1992) characterized membrane compartments responsible for Golgi function in wild-type S. cerevisiae by immunofluorescence and immunogold staining at the electron microscope level. These compartments were labeled with antibodies specific for alpha 1–6 mannose linkages, Sec7p or Ypt1p, all markers of the yeast Golgi apparatus. When observed at the electron microscope, elements labeled with each of these antibodies appeared as disk-like structures apparently surrounded by small vesicles. In contrast to what is usually observed in plant cells or in compact zones of the Golgi ribbon of mammalian cells, yeast Golgi were seen as single, isolated cisternae, generally not arranged into parallel stacks. Due to differences in staining properties of yeast cytoplasm compared with mammalian and plant cells, the observation in the electron microscope of intracellular membranes stained with conventional contrasting techniques was relatively difficult in yeast. Techniques using oxidizing agents such as potassium permanganate (Kaiser and Schekman 1990; Preuss et al. 1992) or a mixture of osmium tetroxide and potassium ferrocyanide (reduced osmium) (Rambourg et al. 1993, 2001) after glutaraldehyde fixation were found to preferentially enhance the contrast of intracytoplasmic membranes against a mostly unstained background. These techniques not only delineated membranes but also stained the content of ER cisternae, Golgi-like elements and secretion granules. Furthermore, with reduced osmium, the staining increased from the nuclear envelope to the secretion granules and cell wall. It was thus postulated that reducing substances such as carbohydrates which are progressively added to proteins during their maturation, were responsible for the staining of the content of cell organelles and thus could facilitate their identification along the secretory pathway (Rambourg et al. 1993, 2001).

Morphodynamics of the yeast Golgi apparatus

*

633

Another problem inherent to the examination of tissue sections by transmission electron microscopy is that the image is only a projection, i.e., the shadow of three-dimensional structures present in the section. This problem becomes crucial when relatively thick sections are observed. To overcome this difficulty, the exact 3D organization of structures can be studied from stereo micrographs obtained by tilting the goniometric stage of the electron microscope at two angles chosen as a function of the section thickness and the final magnification of the pictures. A 3D view of the structure is then obtained by placing the two photographs side by side and examining them with a stereoscope. This method has been used to study the 3D structure of the Golgi apparatus in several types of mammalian cells (reviewed in Rambourg and Clermont (1997)). The objective of the present chapter is to describe the structure of the Golgi apparatus by stereoelectron microscopy in the yeast S. cerevisiae and its modifications in various mutants.

The secretory pathway in wild-type yeast cells In wild-type strains of the yeast S. cerevisiae post-fixed in reduced osmium  pe s et al. 2005), the nucleus is surrounded by a (Rambourg et al. 2001; Ke faintly stained cisterna perforated by nuclear pores. This nuclear envelope is continuous with stained sheets and ribbons (Fig. 1a, ER) which extend within the surrounding cytoplasm and may occasionally establish connections with a more or less fenestrated continuous sheet located just beneath the plasma membrane (Fig. 1a, arrows). Since these elements were in continuity with the nuclear envelope, they were thought to be part of the endoplasmic reticulum in accordance with previous observations by Preuss et al. (1991). In yeast, any region of ER elements may display continuities with membranous tubules forming networks of various sizes and staining density (some of these tubular networks display a wide polygonal mesh (Fig. 1b, TeR)). Their staining density approximates that of the ER elements from which they were derived. The transition between non-perforated ER elements and tubules of the wide-meshed network is not abrupt but usually occurs by a progressive fenestration of the ER element (Fig. 1c), thereby suggesting that such networks are distal parts of the ER. At their other end, these ER derived tubular networks are often connected to more intensely stained networks with narrower meshes and dilations (nodules) located at the intersections of membranous tubules (Fig. 1a, G; b, G; c, SG). Some of these distensions are filled with an intensely stained material and are similar in size and texture to nearby secretion granules (Fig. 1d). Strings of strongly stained granules are often present in the vicinity of the small and more intensely reactive tubular networks. Upon closer examination, the stained granules making up these strings are interconnected by thin, poorly reactive, bridges seemingly deprived of secretory material. The presence in yeast buds of granules still interconnected (Fig. 1d,

634

*

A. Rambourg et al.

Figure 1. Thin sections of wild-type S. cerevisiae cells grown at 24 C. stained with reduced osmium silver proteinate (from Rambourg et al. 2001) (a) shows a nucleus (N) delimited by a lightly stained nuclear envelope in continuity with ER sheets (ER) interconnecting it with a peripheral subplasmalemmal ER indicated by oblique arrows. Tubular Golgi elements (G) are connected either with the nuclear envelope or with ER sheets. Nodular dilations corresponding to segregating secretion granules (g) may be recognized at the intersections of the tubular Golgi elements. L: lipid droplets. (b) An ER sheet (ER) is continuous with a wide-meshed polygonal tubular network (TeR). Faintly stained granules(g) are segregated at the interconnecting points of this network. More strongly stained granules are seen at the interconnecting points of the narrow-meshed polygonal Golgi network (G) seen in side view. SG : secretion granules; V : vacuole. (c) Stained granules (arrows) are present at the intersections of tubules located at the periphery of a non-perforated ER cisterna (ER). (d) In the lower part of the picture, a Golgi network (G) with stained granules is seen in face view. In the upper half, another Golgi network in side view is continuous with an ER sheet indicated by an horizontal arrow.

G) and free granules with tubular residues still attached to them strongly suggest that in yeast, secretory granules are liberated in the cytoplasm by fragmentation of tubular networks (Rambourg et al. 1993, 2001). Furthermore, as the Golgi apparatus is known to be involved in the segregation and maturation of secretion granules, these tubular networks with nodular dilations should be identified as Golgi elements. Thus, in wild-type yeast, the secretory pathway appears to consist of a continuous membranous system along which membrane transformations such as fenestration and tubulization would lead to segregation of the secretory material into nodules. These nodules are of increasing sizes and staining densities, and progressively transform into secretion granules to be released in the cytoplasm by rupture of interconnecting tubules (Rambourg et al. 2001). The three-dimensional structure of Golgi elements was not exactly similar in wild-type strains of other yeast species. In Zygosaccharomyces rouxii or Schizosaccharomyces pombe fixed and stained using the same conditions as for S. cerevisiae, Golgi elements appear as polygonal networks of intensely stained membranous tubules. Like in S. cerevisiae, distensions filled with stained material at the intersections of membranous tubules suggested that

Morphodynamics of the yeast Golgi apparatus

*

635

secretion granules formed by fragmentation of tubular networks (Rambourg et al. 1995b). In Hansenula polymorpha (Rambourg et al. 1995), Pichia pastoris (Rambourg et al. 1995; Rossanese et al. 1999) or in Schizosaccharomyces pombe (Chappell and Warren 1989; Rambourg et al. 1995), networks of anastomosed tubules are closely superimposed to each other and form parallel arrays reminiscent of Golgi stacks seen in mammalian cells. In other strains such as Kluyveromyces lactis, Candida albicans, and Candida parapsilosis (Rambourg et al. 1995), the situation is intermediate and the cytoplasm contains only small arrays with two or at most three superimposed layers of membranous tubules. In yeast, membranous tubules forming parallel arrays are clearly continuous with each other. They are frequently found in close proximity to the ER to which they are occasionally connected. Furthermore, small vesicles in the 30–50 nm range as observed in mammalian cells are rarely encountered in most yeast strains.

The sec7 mutants: a continuous membrane flow As mentioned above, temperature-sensitive mutants have been isolated that block protein trafficking at various points of the secretory pathway. When grown at permissive temperature, they behave like wild-type strains. Upon shifting to restrictive temperature, they gradually accumulate various types of structures (e.g. small vesicles, secretory granules, etc.), depending on the location of the block along the secretory pathway. Upon return at a permissive temperature, the phenotype of the wild-type cells is usually progressively restored. The location at which these sec mutations restrict protein transport have been determined by examination of nascent glycoproteins accumulated at restrictive temperature (Novick et al. 1981). Two mutations, sec7 and sec14, block protein transport from the yeast Golgi apparatus and were shown to accumulate structures, the so-called Berkeley bodies, initially identified as Golgi structures (Esmon et al. 1981). In sec7-1 mutants grown at the permissive temperature (24 C), like in wildtype cells, Golgi elements consisted in independent networks of tubules with intensely stained dilations at the intersections of their interconnected membranous tubules (Fig. 2, 0 min). As early as 5 min after shifting the mutants to the restrictive temperature (37 C) in low glucose (0.1%) medium, there was a post-Golgi block in the secretory pathway (Rambourg et al. 1993). The secretion granules decreased in number while the networks of Golgi tubules increased in size and lost their distensions (Fig. 2, 5 min). At 15 min, the tubules which were interconnected in all directions progressively formed closely superimposed and parallel tubular arrays (Fig. 2, 15 min). These parallel arrays, however, remained interconnected. Between 15 and 30 min, the parallel elements making up these arrays were less and less perforated and finally transformed into stacks of 5–8 flattened and poorly perforated cisternae resembling those commonly observed in thin sections of the Golgi apparatus of animal or plant cells (Fig. 2, 30 min). The closely packed

636

*

A. Rambourg et al.

Figure 2. Modifications of the Golgi networks in sec7-1 mutant at various time intervals following a shift to the non-permissive temperature of 37 C. At 0 min, a stained granule (Sg) is present at the interconnection of tubules making up a Golgi network. After 5 min, the stained granules are no longer visible in an elaborate Golgi network made up of interconnected tubules. After 15 min, the Golgi network appears as a stack of four closely superimposed parallel polygonal tubular networks, the meshes of which are indicated by vertical black or white arrows. At 30 min the stack of tubular networks is transformed in a stack of flattened cisternae. At 60 min a tubular ER starts to accumulate at one side of the Golgi stack (G). Reduced osmium lead citrate staining. From Rambourg et al. (1993).

elements frequently interlocked and interconnections were often observed between stacks. At later time intervals, the size of stacks hardly increased but ER membranes started to accumulate on one side of the stacks and formed large reticulated spherical bodies (Fig. 2, 60 min). At all stages of these structural transformations, the Golgi networks remained connected to the ER. It was thus postulated that in the yeast S. cerevisiae, structurally continuous membranes are permanently flowing down the secretory pathway and accumulate upon downstream blocking (Rambourg et al. 1993). In sec7-4 mutant cells grown at the permissive temperature (20 C), the morphology of the secretory pathway (Fig. 3a) was identical to that observed in wild-type cells. Post-Golgi secretion granules were scattered throughout the cytoplasm and accumulated in the bud. They displayed a staining intensity similar to that of the cell wall. Polygonal tubular networks showing more intensely stained nodules at tubules intersections were identified as Golgi networks elaborating pro-secretory granules at various stages of maturation. The ER appeared either as sheet-like cisternae or networks of interconnected tubules with polygonal meshes significantly wider that those making up the

Morphodynamics of the yeast Golgi apparatus

*

637

Figure 3. Modifications of the Golgi networks in sec7-4 mutant at various time intervals following a shift to the non-permissive temperature of 37 C. (a) At 20 C, the Golgi apparatus (G) consists of a system of interconnected tubules. At the intersections of these tubules small nodules are intensely stained and correspond to maturing prosecretory granules (oblique arrow). A curved sheet-like ER cisterna (ER) is seen at right of the Golgi network. Mature secretion granules are interspersed in the cytoplasm (vertical arrow) and accumulate (SG) in the bud. (b) After 5 min at 37 C, a polygonal array of membranous tubules makes up the tubular ER (T). At the extremity of the tubular ER, irregularly dilated and intensely stained interconnected tubules (large vertical arrow) show a tendency to form parallel arrays. At this early stage, a few secretory granules (sg) are still present in the cytoplasm. (c) At 20 min after shifting the cells at the restrictive temperature, intensely stained units (SS) consist of parallel arrays of interconnected flattened saccules which may give rise at their extremities or be transformed into spherical sheetlike structures (L) which may be incorporated into the vacuole V or degenerate (arrows). (d) After 40 min, stacks of saccules may still be observed. They show however a tendency to desintegrate into irregularly stained loosely and irregularly anastomosed membranous tubules (T) and numerous irregularly stained spherical structures (L) interspersed throughout the cytoplasm. N : nucleus. Reduced osmium lead citrate staining. Adapted from Deitz et al. (2000).

Golgi networks. These ER networks were deprived of stained nodules at their intersections and were thus referred to as ER tubular networks. As soon as 5 min after shifting the cells at 37 C (Fig. 3b) the morphology of the mutant was distinctly altered. Irregularly dilated and intensely stained tubules started to form arrays in continuity with tubular ER. A few secretion granules were interspersed throughout the cytoplasm but nodular networks were no longer visible. At later time intervals (Fig. 3c), secretion granules disappeared and tubular ER were less frequently encountered. Structures consisting of parallel arrays of strongly stained flattened saccules were mainly observed in the

638

*

A. Rambourg et al.

cytoplasm. Ultimately, these flattened saccules appeared to give rise to hollow spherical sheet-like elements interspersed throughout the cytoplasm (Fig. 3d). Some of these elements, presumably autophagosomes (Reggiori et al. 2004) were incorporated in the vacuole or seemed to degenerate in situ within the cytoplasm. According to Deitz et al. (2000), the strongly stained and irregularly dilated elements that proliferate aberrantly and remain connected to the tubular ER at presumptive exit sites would correspond to Golgi-like elements which accumulate as the result of a block of a continuous membranous flow along the secretory pathway upstream of the block observed in sec7-1 mutant cells. In sec14 mutants, the cells grown at restrictive temperature were originally thought to accumulate Golgi stacks like in sec7 mutants. Yet in contrast to the latter, the various steps of glycosylation of secretory proteins remained seemingly unaltered in sec14 mutants, which, in addition, accumulated secretory granules. In S. cerevisiae, the SEC14sc gene encodes a protein (Sec14psc), a phosphoinositol/phosphatidylcholine transfer protein, one of the main functions of which would be to control the phosphatidyl content of yeast Golgi membranes (McGee et al. 1994). At the permissive temperature or after shifting the mutant at 37 C for 2 min, (Fig. 4) the cytoplasm was relatively poor in cell organelles. The endoplasmic reticulum consisted mainly of poorly fenestrated flattened cisternae interconnecting the nuclear envelope with the peripheral subplasmmal cisterna (Fig. 4, ER). Sacculotubular elements (Fig. 4a) and more or less extensive of thin anastomosed membranous tubules (Fig. 4b, arrow-

Figure 4. sec14 mutant grown at the permissive temperature (25 C) or shifted for 2 min to the restrictive temperature (37 C). At 25 C, the permissive temperature (a) poorly fenestrated cisternae of endoplasmic reticulum (ER) are well developed and form a system of interconnected sheets. A few cisternae appear as fenestrated sheets (asterisks). More intensely stained nodular tubules (open arrows), and a few intensely stained secretion granules (oblique arrow) are also seen. After 2 min at 37°C (b), faintly stained networks of tubular ER (arrow heads) are connected to ER cisternae (ER). V: vacuoles. Reduced osmium lead citrate staining. From Rambourg et al. (1996).

Morphodynamics of the yeast Golgi apparatus

*

639

heads) were located at close proximity of the ER cisternae. They were occasionally continuous with the latter and displayed the same staining intensity. Thus, they were thought to correspond to tubular ER. More intensely stained networks of membranous tubules showed nodular dilations typical of Golgi structures (Fig. 4a, empty arrow). Some secretion granules were occasionally observed within the cytoplasm but most of them were located within growing buds. At 5 and 10 min after transferring sec14 mutants to the restrictive temperature (Fig. 5a), the sheets of ER were reduced in size but the incidence of membranous tubules increased considerably. Such tubules showed a tendency to form large and irregular networks distributed throughout the cytoplasm. Well stained dilations were frequently present along the larger tubules and thus contributed to impart to the networks a nodular aspect (Fig. 5a, arrow). Occasionally, anastomosed tubules formed spherical structures that were referred to as fenestrated spheres (Fig. 5, S). At these time intervals, an

Figure 5. sec14 mutant shifted for 10 min (a) or for 60 min (b) and (c) to the restrictive temperature (37 C). At 10 min, membranous tubules with intensely stained dilations (oblique arrow) were often seen. Occasionally, anastomosed tubules formed spherical structures (S). There was also an increase in the number of secretory-like granules interspersed throughout the cytoplasm (sg). At later time the membranous tubules with heavy stained dilations were still observed and occasionally formed heavily stained parallel arrays (SS). Fenestrated spheres (S) were more frequently .encountered as well as the secretory-like granules (sg) dispersed within the cytoplasm. Numerous vacuoles (v) with a size and approximating that of the fenestrated spheres were also observed. N : nucleus. Reduced osmium lead citrate staining. See Rambourg et al. (1996).

640

*

A. Rambourg et al.

additional feature was the increase in number of secretion-like granules irregular in size and shape which were no longer located within growing buds but were distributed throughout the cytoplasm (Fig. 5a, sg). At later times, sheets of ER and faintly stained networks of tubular ER were rarely observed. At 45 and 60 min (Fig. 5b,c), they were almost absent. The more heavily stained networks of nodular tubules were less numerous than at the 5–10 min intervals. They occasionally formed stacks or parallel arrays (Fig. 5c). Fenestrated spheres were frequently encountered (S, Fig. 5b, c). Secretion-like granules (sg, Fig. 5) were numerous and dispersed throughout the cytoplasm. Numerous spherical bodies of a size range similar to that of the fenestrated spheres (v, Fig. 5b) contained a material resembling that observed in the vacuoles. The observations indicate that as soon as 5 min after shifting the mutant cells to the restrictive temperature, intensely stained granules start to accumulate in the cytoplasm. This result is compatible with biochemical results indicating that in sec14 mutants, invertase present in secretory granules remained intracellular, while acquiring all their sugar moieties (Novick et al. 1980; Esmon et al. 1981; Bankaitis et al. 1989). Between 10 and 30 min, stained nodular membranous tubules formed networks that also accumulated and the number of secretory-like granules increased in the cytoplasm, while the reticular sheets and poorly stained networks of tubular ER started to decrease. Later, the number of granules continued to increase whereas there was a decrease in the number of nodular membranous tubules. These data are thus consistent with the proposal that in yeast, secretory granules are formed by rupture of nodular membranous tubules which would be the structural equivalents of the trans-Golgi network observed in mammalian cells. In sec14 mutants secretion granules continued to be formed and accumulated in the cytoplasm. Thus, parallel arrays reminiscent of Golgi stacks as observed in sec7 mutants where the formation of secretion granules is blocked, were rarely encountered in sec14 mutants. This observation supports the proposal that the formation of parallel arrays is the result of a partial block or slowing down of a continuous membranous flow (Rambourg et al. 1993) upstream of secretion granule accumulation. In addition, the number of fenestrated spheres continued to increase at time intervals where nodular tubules start to disappear. It is therefore postulated that these fenestrated spheres, like the secretory granules, are derived from the nodular tubular networks. At later time intervals, all structural intermediates between fenestrated spheres and small vacuolelike spherical bodies, particularly numerous at the 45–120 min intervals can be seen (Figs. 15 and 21 in Rambourg et al. 1996). The latter structures are thus thought to arise by transformation of the fenestrated spheres. They would then correspond to autophagosomes such as those described in the small neurons of the rat spinal ganglion in which acid phosphatase reactive bodies have been reported to form by a progressive transformation of portions of trans-Golgi networks (Rambourg and Clermont 1991).

Morphodynamics of the yeast Golgi apparatus

*

641

Effect of brefeldin A in yeast erg6 mutants In plant and animal cells, the fungal metabolite brefeldin A (BFA) inhibits transport from the endoplasmic reticulum to the Golgi apparatus. BFA effects have been extensively studied in erg6 mutants of the yeast S. cerevisiae, where the deletion of an ergosterol-processing enzyme gene renders the cells permeant to the drug. In these mutants, Golgi elements were similar to the ones described in wild-type cells. Addition of BFA induced within minutes the disappearance of secretion granules and tubular networks with strongly stained dilations that resemble secretion granules, i.e. the Golgi units (Rambourg et al., 1995a, 2001). In contrast, networks consisting of thinner membranous tubules increased in size. They were connected with the nuclear envelope (Fig. 6) or other ER cisternae and extended into the surrounding cytoplasm (Fig. 6c) to form cage-like structures in which the bars consisted of anastomosed tubules. At later time intervals, such cage-like structures were no longer observed and ple€ıomorphic membranous bodies mainly composed of stacked ER cisternae (like in Fig. 6b) accumulated in the cytoplasm. Biochemical analysis of protein transport in erg6 mutant cells indicated that after a brief exposure of these cells to BFA, the vacuolar carboxypeptidase Y, the periplasmic invertase and the secreted pheromone a-factor accumulated as both the core glycosylated (ER) and alpha-1,6 mannosylated (early glycosylation step taking place specifically in the Golgi) forms. At later time intervals, in contrast, only the accumulation of unprocessed, unglycosylated cytoplasmic precursors of the vacuolar carboxypeptidase Y and of the ER luminal BiP protein could be observed, indicating a partial disruption of protein translocation into the ER, a non-specific transport defect possibly due to some kind of cell degeneration (Graham et al., 1993). It was thus proposed that at early intervals, BFA induced the disappearance

a

b

c

Figure 6. erg6 strain treated for 5 min with BFA. (a) A tubular network with wide polygonal meshes is seen next to the nucleus (N).This continuous network may be observed in face view (vertical arrow) or in cross section (horizontal arrow). Note the intense staining of the coat at left of the picture. (b) A tubular network with small dilations at the intersections of narrow irregular meshes (oblique arrow) is continuous at one side with a wide-meshed tubular network (vertical arrow) and is connected at right with a parallel array of anostomosed tubules seen in profile (S). Part of the nucleus is seen at left. (c) Next to the nucleus (N) a large tubular network forms an ovoid mass. Reduced osmium silver proteinate staining. From Rambourg et al. (2001).

642

*

A. Rambourg et al.

of structures equivalent to the maturing (trans) part of the Golgi apparatus and a block at the exit of its forming (cis) part (Rambourg et al., 1995, 2001). This result strongly supports that tubular networks that accumulate in BFAtreated cells correspond to the yeast cis-Golgi.

Structure and function of the yeast Golgi apparatus It is sometimes assumed that in yeasts like in mammalian cells, the Golgi apparatus might consist of stacks of saccules interspersed throughout the cytoplasm. Indeed, stacks consisting of several closely apposed parallel structures are observed in the yeasts Schizosaccharomyces pombe (Chappell and Warren 1989) and Pichia pastoris (Rambourg et al. 1995b; Rossanese et al. 1999; Bevis et al. 2002; Mogelsvang et al. 2003). Yet, stacks of saccular cisternae are rarely found in wild-type S. cerevisiae. How can we reconcile such observations with the current models of Golgi structure and function based on the concept of distinct saccular cisternae ? The only situation in which such structures have been clearly identified is found in the thermosensitive sec7 and sec14 secretory mutants at the restrictive temperature. Furthermore, this situation provided the opportunity to understand how strictly tubular networks can be transformed in stacks of parallel flattened saccules, due to membrane accumulation along a continuous secretion flow. Such stacks of saccules resembled those observed in mammalian cells. Like in spermatids (Clermont et al. 1994) and Sertoli cells (Rambourg et al. 1979), connections between the saccules were seen but in contrast to spermatid cells, Golgi associated vesicles were not present. Thus, in the particular case of S. cerevisiae, no saccular migration (maturation) or vesicular transport can be seen but rather the formation of saccules as the result of what appears as a continuous membrane flow. When a sec18 thermosensitive mutant was transferred for 10 min at the restrictive temperature of 37 C, the Golgi tubular networks and secretory granules were no longer observed and small tubulovesicular fragments accumulated in the cytoplasm (Morin-Ganet et al. 2000). When the block was released by returning the cells to the permissive temperature of 25 C, isolated tubulovesicular fragments disappeared, appeared to fuse as tubulovesicular aggregates and secretion granules started to reappear. Then, tubular Golgi networks progressively increased in number with a concomitant decrease of the tubulovesicular aggregates while secretory granules accumulated in the bud. It is thus concluded that tubular networks are not static structures, but that they are constantly renewed by fusion of vesiculotubular elements presumably originating from the endoplasmic reticulum. Tubular networks thus appear to represent the cis-element of the Golgi apparatus. It should be stressed however that if tubular networks arise by fusion of vesiculotubular elements, isolated vesicles and tubules are rarely encountered in wild-type cells. In contrast, tubular networks with dilations reminis-

Morphodynamics of the yeast Golgi apparatus

*

643

cent of pro-secretory granules are frequently in continuity with the nuclear envelope or the subplasmalemmal reticulum. In BFA-treated yeast cells, networks containing large strongly stained nodules resembling mature secretion granules are not observed. Moreover, in sec18 and sec23 mutants switched from restrictive to permissive temperature, nodular networks reappear later than the cis-Golgi tubular networks (Morin-Ganet et al. 2000). These nodular network would thus correspond to the trans-tubular parts of the Golgi apparatus of mammalian cells, known to liberate prosecretory granules into the surrounding cytoplasm of secretory cells (Rambourg and Clermont 1997). As reported by Rambourg et al. (2001), it should be mentioned, however, that, in wild-type cells of S. cerevisiae, the three-dimensional organization of these “trans” networks may vary from one cell to another or even in the same cell. They may be scattered throughout the cytoplasm as independent units but also they may form a continuous structure with subplasmalemmal ER or with the nuclear envelope at one end, and give rise to intensely stained secretion granules at the other end. In some cases, extensive tubular networks are lacking and secretion granules arise directly from the perforated periphery of poorly stained ER sheets. Furthermore, when the sec21-3 thermosensitive mutant was placed for 20 min at the restrictive temperature of 37°C, the secretory pathway was blocked at the exit of the ER, which started to accumulate as clusters of interconnected ribbon-like elements. When the block was released, no vesicular or vesiculotubular fragments were ever observed. Tubular networks of various sizes and staining densities appeared in the cytoplasm as soon as 5 min after release of the block while granules were seemingly released by rupture of such networks or even started to form at the edges of ER fenestrae (Rambourg et al. 2001). Thus the Golgi apparatus of S. cerevisiae appears as extremely flexible, exhibiting an ultrastructure that heavily depends on membrane flow rates, rather than on a complex static organization. Finally, despite yeast species Pichia pastoris appeared to exhibit stacks of saccules reminiscent of the mammalian cell Golgi apparatus on sections and by electron tomography (Rambourg et al. 1995b; Mogelsvang et al. 2003), preliminary observations by stereo-electron microscopy and by electron tomography of thick sections (0.25 mm) of P. pastoris cells stained with reduced osmium and silver proteinate strongly suggest that the Golgi apparatus also consist of a system of tightly interconnected membranous tubules which appear continuous with the nuclear envelope or endoplasmic reticulum sheets in these cells (not shown). Furthermore, like in S. cerevisiae, small vesicles or isolated tubules were rarely encountered in the cytoplasm. If these observations are confirmed, they would support a dynamic secretion model for yeast cells, in which vectorial membrane flow from the endoplasmic reticulum would lead, through a defined series of membrane transformations, to the formation of tubular networks, segregation and liberation of

644

*

A. Rambourg et al.

secretion granules and/or formation of autophagosomes and lysosome-like bodies, along a continuous corpus. In this model, the apparent ultrastructure of the Golgi apparatus would mostly depend on the difference between the membrane flow rate upstream and downstream of each compartment along the secretory pathway. This could account for the major differences observed between yeast strains and upon mutations of proteins involved in the regulation of secretion.

References Bankaitis VA, Malehorn DE, Emr SD, Greene R (1989) The Saccharomyces cerevisiae SEC14 gene encodes a cytosolic factor that is required for transport of secretory proteins from the yeast Golgi complex. J Cell Biol 108: 1271–1281 Beams HW, Kessell RG (1968) The Golgi apparatus: structure and function. Int Rev Cytology 23: 209–276 Berger EG (1997) The Golgi apparatus : from discovery to contemporary studies. In: € ser Verlag, Boston pp. 1–36 Berger EG, Roth J (ed) The Golgi apparatus. Birkhau Bevis BJ, Hammond AT, Reinke CA, Glick BS (2002) De novo formation of transitional ER sites and Golgi structures in Pichia pastoris. Nat Cell Biol 4(10): 750–756 Clermont Y, Rambourg A, Hermo L (1994) Connections between the various elements of the cis- and mid-compartments of the Golgi apparatus of early rat spermatids. Anat Record 240: 469–480 Chappell Th G, Warren G (1989) A galactosyltransferase from the fission Yeast Schizosaccharomyces pombe. J Cell Biol 109: 2693–2702 Dalton AJ, Felix MD (1954) Cytological and cytochemical characteristics of the Golgi substance of epithelial cells of the epididymis—in situ, in homogenates and after isolation. Am J Anatomy 94: 171–208  pe  s F, Franzusoff A (2000) Sec7p directs the transitions Deitz SB, Rambourg A, Ke required for yeast Golgi biogenesis. Traffic 1: 172–183 Ehrenreich JH, Bergeron JM, Siekevitz P, Palade GE (1973) Golgi fractions prepared from rat liver homogenates. I. Isolation procedure and morphological characterization. J Cell Biol 59: 65–72 Esmon B, Novick P, Schekman R (1981) Compartmentalized assembly of oligosaccharides on exported glycoproteins in yeast. Cell 25: 451–460 Farquhar MG (1985) Progress in unraveling pathways of Golgi traffic. Ann Rev Cell Biol 1: 447–1488 Farquhar MG, Palade GE (1981) The Golgi apparatus (complex) – 1954–1981 – from artefact to center stage. J Cell Biol 91: 77–103s Franzusoff A, Redding K, Crosby J, Fuller RS, Schekman R (1991) Localization of components involved in protein transport and processing through the yeast Golgi apparatus. J Cell Biol 112: 27–37 Fries E, Rothman JE (1980) Transport of vesicular stomatitis viral glycoprotein in a cell-free extract. Proc Natl Acad Sci USA 77: 3870–3874 Golgi C (1898) Sur la structure des cellules nerveuses des ganglions spinaux. Arch Ital Biologie 30: 278–286 Graham TR, Scott PA, Emr SD (1993) Brefeldin A reversibly blocks early but not late protein transport steps in the yeast secretory pathway. EMBO J 12: 869–877 Kaiser CA, Schekman R (1990) Distinct sets of SEC genes govern transport vesicle formation and fusion early in the secretory pathway. Cell 61: 723–733  pe  s F, Rambourg A, Satiat-Jeunemaître B (2005) Morphodynamics of the secretory Ke pathway. Int rev cytol 242: 55–120

Morphodynamics of the yeast Golgi apparatus

*

645

Kornfeld R, Kornfeld S (1985) Assembly of asparagine-linked oligosaccharides. Ann Rev Biochem 54: 631–664 Leblond CP (1950) Distribution of periodic acid-reactive carbohydrates in the adult rat. Am J Anat 86: 1–25 Leblond CP, Glegg RE, Eidinger D (1957) Presence of carbohydrates with free 12-glycol groups in sites stained by the periodic acid-Schiff technique. J Histochem Cytochem 5: 445–458 McGee TP, Skinner HB, Whitters EA, Henry SA, Bankaitis VA (1994) A phosphatidylcholine transfer protein controls the phosphatidyl content of yeast Golgi membranes. J Cell Biol 124: 273–287  DJ (1991) Perspectives on Golgi apparatus and function. J Electr Mollenhauer HH, Morre Microsc Technique 17: 2–14 Mollenhauer HH, Whaley WG (1963) An observation on the Golgi apparatus. J Cell Biol 17: 222–225  pe s F (2000) Morphogenesis Morin-Ganet MN, Rambourg A, Deitz SB, Franzusoff A, Ke and dynamics of the yeast Golgi apparatus. Traffic 1: 56–68 Novick P, Field C, Scheckman R (1980) Identification of 23 complementation groups required for post-translational events in the yeast secretory pathway. Cell 21: 205–215 Preuss D, Mulholland J, Franzusoff A, Segev N, Botstein D (1992) Characterization of the Saccharomyces Golgi complex through the cell cycle by immunoelectron microscopy. Mol Biol Cell 3: 789–803 Preuss D, Mulholland J, Kaiser Ch A, Orlean P, Albright Ch, Rose MD, Robbins Ph W, Botstein D (1991) Structure of the yeast endoplasmic reticulum : localization of ER proteins using immunofluorescence and immunoelectron microscopy. Yeast 7: 891  lectronique de coupes Rambourg A (1969) L’appareil de Golgi : examen en microscopie e paisses (0,5–1 m), colore es par le me  lange chlorhydrique-phosphotungstique. C R e rie D) 269: 2125–2127 Acad Sci Paris (se Rambourg A, Clermont Y (1991) Three-dimensional structure of cytidine monophosphatase reactive trans-Golgi elements in spinal ganglion cells of the rat. Anat Rec 232: 25–35 Rambourg A, Clermont Y (1997) Three-dimensional structure of the Golgi apparatus in € ser Verlag. mammalian cells. In: Berger EG, Roth J (ed). The Golgi Apparatus. Birkhau Boston pp. 37–62 Rambourg A, Clermont Y, Hermo L (1979) Three-dimensional architecture of the Golgi apparatus in the Sertoli cell of the rat. Am J Anat 154: 455–476 pe s F (1993) Modulation of the Golgi apparatus in Rambourg A, Clermont Y, Ke Saccharomyces cerevisiae sec7 mutants as seen by three-dimensional electron microscopy. Anat Rec 237: 441–452 Rambourg A, Clermont Y, Marraud A (1974) Three-dimensional structure of the osmium-impregnated Golgi apparatus as seen in the high voltage electron microscope. Am J Anat 140: 27–46  pe s F (1996) Transformations of Rambourg A, Clermont Y, Nicaud JM, Gaillardin C, Ke membrane-bound organelles in sec14 mutants of the yeasts Saccharomyces cerevisiae and Yarrowia lipolytica. Anat Rec 245: 447–458  pe  s F (1995a) Effects of Brefeldin A on the Rambourg A, Clermont Y, Jackson CL, Ke three-dimensional structure of the Golgi apparatus in a sensitive strain of Saccharomyces cerevisiae. Anat Rec 241: 1–9 pe s F (1995b) Three-dimensional structure of Rambourg A, Clermont Y, Ovtracht L, Ke tubular networks, presumably Golgi in nature, in various yeast strains : a comparative study. Anat Rec 243: 283–293 Rambourg A, Hernandez W, Leblond CP (1969) Detection of complex carbohydrates in the Golgi apparatus of rat cells. J Cell Biol 40: 395–414

646

*

A. Rambourg et al.

Rambourg A, Jackson C, Clermont Y (2001) Three dimensional configuration of the secretory pathway and segregation of secretion granules in the yeast Saccharomyces cerevisiae. J Cell Science 114: 2231–2239 Reggiori F, Wang Ch-W, Nair U, Shintani T, Abeliovich H, Klionsky DJ (2004) Early stages of the secretory pathway, but not endosomes, are required for Cvt vesicle and autophagosome assembly in Saccharomyces cerevisiae. Mol Biol Cell 15: 2189–2204 Rossanese OW, Soderholm J, Bevis BJ, Sears IB, O'Connor J, Williamson EK, Glick BS (1999) Golgi structure correlates with transitional endoplasmic reticulum organization in Pichia pastoris and Saccharomyces cerevisiae. J Cell Biol 145: 69–81 Roth J (1991) Localization of glycosylation sites in the Golgi apparatus using immunolabeling and cytochemistry. J Electr Microsc Technique 17: 121–131 Rothman JE (1994) Mechanisms of intracellular protein transport. Nature 372: 55–63 Schekman R (1992) Genetic and biochemical analysis of vesicular traffic in yeast. Curr Opi Cell Biol 4: 587–592 Wattenberg BW (1991) Analysis of protein transport through the Golgi in a reconstituted cell-free system. J Electr Microsc Tech 17: 150–131 Whaley WG, Dauwalder M (1979) The Golgi apparatus, the plasma membrane, and functional integration. Int Rev Cytology 58: 199–245

Structure and function of the Golgi organelle in parasitic protists

*

647

Structure and function of the Golgi organelle in parasitic protists Y. Y. Sokolova and A. A. Mironov

At least three aspects make study of endomembrane systems of parasitic protists scientifically important. First, unicellular parasites are promising objects to clarify Golgi functions, as well as mechanisms of intracellular transport because they possess essentially reduced secretory machinery, often coupled with hypertrophied secretion of certain proteins. A comparative molecular analysis of their compact genomes is helpful in determining the minimal set of genes for certain transport and secretory functions. In addition, small cell size of many protists facilitates morphological analysis, particularly, three-dimensional reconstruction of the secretory compartment (Beznoussenko et al. 2007; Cooke et al. 2004; Joiner and Roos 2002; Lujan and Touz 2003; Overath and Engstler 2004; Plattner 1993). Second, knowledge about specific mechanisms of function of endomembrane system of pathogenic protists could be helpful in development of pharmaceutical drugs. Third, the reconstruction of the ancestral nature of intracellular trafficking depends on the topology and rooting of the eukaryotic tree (Richards and Cavalier-Smith 2005). Thus, analysis of protists by reconstructing ancestral states could help in understanding of the origin and evolution not only of the secretory system, but also the eukaryotic cell per se. Basing on similarity of secretory machinery in yeast and mammalian cells it was concluded that most components of the transport systems are conservative, because they exist “from yeasts to men” (Horazdovsky et al. 1995; Roos et al. 1994; Rothblatt et al. 1994), and it is not surprising given the fact that mammals (a class of Metazoa) and Fungi, belong to the “crown” of the Eukaryotic Tree of Life and represent the sister groups (Fig. 1). Therefore, studies on Golgi organization of systematically diverse groups of parasitic protists belonging to “basic” lineages shed more light on the level of conservatism/divergence of this organelle in living organisms. The aim of this review was to summarize data on the structure and function of the Golgi apparatus (GA) in several unrelated groups of protists with the parasitic life style. The following systematic groups formerly regarded as Protozoa (Levine et al. 1980) are considered: the “true protozoans” – Parabasalia (Trichomonas)1, Diplomonada (Giardia), Entamoebidae (Entamoeba), Apicomplexa (Toxoplasma, Plasmodium), and Kinetoplastida (Trypanosoma, Leishmania), and Microsporidia (Nosema, Encephalitozoon).

1 The most well-known or well-studied representative of the group is given in brackets

648

*

Y. Y. Sokolova and A. A. Mironov

Figure 1. Tree of life (Dacks et al. 2003; Adl et al. 2005; Arisue et al. 2005). The scheme reflects the conventional concept of monophyletic origin of the eukaryotic cell from prokaryotic and divergence of major lineages of eukaryotes (Adl et al. 2005). Monophyletic origin of all supretaxa except Escavata is supported by independent methods of phylohgenetic analyses based on both comparison of ultrastructural characters and sequences analyses of genes encoding various macromolecules (>20). Mono- or paraphyletic origin of Escavata was not confirmed in all cases, and this taxon might be polyphyletic (Arisue et al. 2005). Consequence of divergence events cannot be resolved by the methods used, but the most probable locations for the tree rout are branches leading to Parabasalia/Diplomonada clade and to Opisthokonta. Putative positions for the rout are marked with cartoon fur-trees. The taxons lacking Golgi dictiosomes, are placed in boxes limited by solid lines. Biochemical, structural or molecular evidence for presence of a functionally active Golgi complex have been obtained for representatives of all taxa except Retortamonada and Oxymonada, marked by the dashed boxes. It is presumed that the cell of the last common ancestor contained Golgi dictiosomas, and in the course of evolution Golgi complex has shifted its morphology beyond recognition at least five times (strikethrough paper stacks).

Golgi apparatus and its function in Trichomonas (Parabasalia) Flagellated parabasalids are known as simbionts and parasites of insects and mammals. Experiments with Tritrichomonas foetus and Trichomonas vaginalis infecting urogenital tracts of humans and cattle, provided the majority of

Structure and function of the Golgi organelle in parasitic protists

*

649

data on biochemistry and cell biology of trichomonads. The representatives of this group were in the focus of attention of cell biologists mainly because they possess hydrogenosomes, the anaerobic precursors of mitochondria (Andersson and Kurland 1999), and also as the presumably most ancient group of protists (Cavalier-Smith and Chao 1996; Keeling et al. 2000). Golgi of Trichomonas spp. is large and consists of 8–12 prominent cisterns located in vicinity of the nucleus. Golgi is differentiated into four classical subcompartments (cis, intermediate, trans and trans-Golgi network). Cis-cistern is connected with the parabasal filament and located in close proximity to the ER (Benchimol et al. 2001). GA together with parabasal filament forms so called “parabasal body”, the conspicuous character of all parabasalids. The major GA function is glycosylation of adhesins, the surface proteins, which mediate the parasite adhesion to the surface of an epithelial cell, and therefore, are being responsible for pathogenesis (Benchimol et al. 2001). Three-dimensional reconstruction of serial sections and immunofluorescence demonstrate that trichomonad’s GA do not undergo disassembly during mitosis, like Golgi organelle in mammalian cells; it is not sensitive to nocodazole and other anti-tubulin agents, though tubulin components can be detected by immunofluorescence. Other than in mammalians tubulin isoforms and/or tubulin-associated proteins were assumed to interact with GA in Trichomonas (Benchimol et al. 2001). Interesting, just before mitotic division GA elongates and then each cistern divides in halves. The process starts with the surface (trans) cisterns and moves towards the deepest cis-cisterns. It was demonstrated that the Golgi complex, parabasal filament, flagella with their basal bodies, center of organization of microtubules, and axostyle replicate simultaneously with the nuclear genome in the interphase, and segregate in mitosis. Thus, trichomonad’s GA, follows the cytoskeleton elements in their duplication, segregation and migration to the daughter cells (Ribeiro et al. 2000). Similar behavior of GA in mitosis was described for Toxoplasma gondii zoites (Pelletier et al. 2002).

Organization of intracellular secretory traffic in Giardia (Diplomonada) Diplomonads, including Giardia spp, the parasites of humans and livestock, were also considered once as the ancient group of eukaryotes (Hashimoto et al. 1994, 1998; Sogin et al. 1989), a “missing link between pro- and eukaryotes” (Kabnick and Peattie 1991). Though many of their “ancient” characters were found later to be due to reduction in response to the parasitic lifestyle, diplomonads are still believed to belong to one of the earliest eukaryotic lineages (Roger 1999). Giardia lamblia is a unicellular intestinal parasite and a leading cause of diarrhea disease in humans worldwide (Adam 2001). The life cycle of Giardia consists of two distinct phases: a flagellate vegetative trophozoite and a cyst with a wall adapted to the survival in the environment. Synthesis and

650

*

Y. Y. Sokolova and A. A. Mironov

secretion of the protective cell wall are essential for transmission of Giardia infectious stages. Endomembrane system of the Giardia trophozoites is composed of the perinuclear envelope and rough ER, transitional elements, the ERGIC zone, putative tubular–vesicular elements, Golgi-like smooth perinuclear membrane stacks, small (50–80 nm) vesicles, many of which are coated, and lysosome-like peripheral granules containing acid phosphatases (Adam 2001; Elmendorf et al. 2003). Occasional flattened cisternae and clefts in EM sections of encysting cells likely represent organized smooth ER induced by weak homotypic interaction of ER membrane proteins (Snapp et al. 2003). Trophozoites lack mitochondria, peroxisomes, secretory granules, and conventional Golgi apparatus (Adam 2001; Marti et al. 2003a). During encystation Giardia trophozoites secrete a fibrillar extracellular matrix of glycans and cyst wall proteins on the cell surface. Secretory proteins contain signal sequences. The bulk of newly synthesized material is exported from the RER as a pulse during the first 5–8 h after induction, and accumulates in a set of approximately spherical encystation-specific vesicles (ESVs), the specialized Golgi-like compartments generated de novo before secretion (Hell and Marti 2004; Marti et al. 2003b). ESVs are coated with clathrin and are connected with the ER. These post-ER vesicles neither have morphological characteristics of Golgi cisternae nor sorting functions. Like conventional Golgi cisternae, ESVs are sensitive to brefeldin A and associate with two Golgi markers, ’COPI and GiYip1 (Marti et al. 2003a, 2003b). The generation of vesicular–tubular clusters, cis-Golgi compartments, and formation of ESVs seems to require the small GTPase Sar1p (Stefanic et al. 2006). There are indications for aggregation of cyst wall material in the enlarged ER cisternae, which could subsequently transform into large carrier compartments and nascent ESVs (Lanfredi-Rangel et al. 2003). This bulk transport of the cyst material is similar to the export of procollagen in mammalian cells (Mironov et al. 2003). The Golgi consists of 3–20 parallel cisterns, appeared only at the late encysting stage of trophozoites. At this period, expression of GA enzymes, such as galactosyl transferase and N-acetylgalactosamine transferase, dramatically increases (Lujan and Touz 2003). Recently several proteosome subunits and HSP70-BiP have been found In Giardia. BiP is exported to ESVs and retrieved via its C-terminal KDEL signal from ESVs (Stefanic et al. 2006). Transitional ER regions and early ESVs were co-localized with the coat protein COP II, and maturing ESVs—with COPI (Marti et al. 2003a). Two syntaxin homologues associated with intraGolgi membrane traffic, have been identified in trophozoites (Dacks and Doolittle 2002); ARF homologues have been localized in the vicinity of nuclei of the vegetative and encysting trophozoites; Rab 1 protein responsible for vesicle transport to the membrane target, was described in ER and peripheral granules. It was shown that ARF and coatomer proteins COPI and COPII were sensitive to brefeldin A (BFA). BFA treatment of both vegetative

Structure and function of the Golgi organelle in parasitic protists

*

651

and encysting trophozoites blocked protein transport and induced disassembly of NBD-ceramide labeled structures (Lujan et al. 1995a, b, c; Lujan and Touz 2003). Thus, Giardia trophozoites exemplify reduction of the Golgi structure and function. However, GA emerges at the certain stage of the parasite life cycle when exocytosis of proteins and polysaccharides is vital. Noteworthy, the genome of Giardia contains almost all key proteins involved in the intracellular transport.

Secretory traffic in Entamoeba histolytica trophozoites Entamoebas are another group of parasitic protists of presumably ancient ancestry; they lack mitochondria and Golgi dictiosomes. The very presence of ER in Entamoeba histolytica was questioned (Ghosh et al. 1999). E. histolytica is a human pathogen; contamination occurs through ingestion of cysts. Ingested cysts are differentiated into trophozoites and invade intestinal epithelium and other organs, in particular liver. Entamoeba cells secrete numerous surface glycoproteins and lectins, which regulate adhesion and invasive properties. In addition, during the encystation entamoebas synthesize proteinases and other molecules directly connected with the virulence, and a chitinase. Secretory pathways of these compounds were studied in connection with surveys for the factors of pathogenicity and drug targets. Though structural organization of GA in Entamoeba is still obscure, conventional mechanisms were shown to be involved in secretory transport of most molecules. Series of biochemical data proved existence of the ER and GA in Entamoeba. Signal sequences, which direct the synthesized proteins into distinct cellular compartments of trophozoites (KDEL, N-terminal signal, etc.) appeared to be structurally similar to the signal peptides of mammalian cells (Ghosh et al. 1999). EM studies, using better preservation conditions to improve visualization of endomembranes, revealed as well the presence of ER and Golgi elements in Entamoeba spp. (Chavez-Munguia et al. 2000). Large spheres located in vicinity of the nucleus and connected by the membrane network, were identified as a part of Golgi complex basing on immunolocalization of ARF and others molecules associated with vesicle transport. Finally, Bredeston et al. (2005) experimentally proved that fraction of “large vesicles” functions as GA; ER and Golgi compartments share glycosylation functions; and that chemical structure of enzyme transporters in E.histalitica is similar to one of the higher eukaryotes. There are evidence for the presence of sugar transporters in Entamoeba cells (Bredeston et al. 2005). In E. histalitica, like in mammalian cells, GA and the transport of most molecules, is sensitive to BFA. On the other hand, traffic of several de novo synthesized proteinases induced by external stimuli, occurred to be insensitive to BFA. These data suggest that two distinct transport systems occur in E. histolytica, one similar to classical membrane protein transport and

652

*

Y. Y. Sokolova and A. A. Mironov

another – independent of BFA and inducible by external stimuli (ManningCela et al. 2003). The Entamoeba histolytica Genome Annotation Database (http://www. tigr.org) includes all major genes responsible for intracellular transport, processing and synthesis of the secretory pathway proteins.

Golgi apparatus and secretory transport systems in Apicomplexa Phyllum Apicomplexa is composed of exclusively obligate intracellular parasites. Representatives of the genera Toxoplasma, Plasmodium, Eimeria, Cryptosporidium, Sarcocystis, all belonging to the class Coccidia, are ubiquitous parasites of humans and livestock. Life cycle of all Apicomplexa includes a motile invasive stage, a zoite, which penetrates the target host cell and rapidly multiplies there. Intracelluar organization of zoites is similar in all coccidian genera. Ultrastructurally a zoite may be regarded as a simplified model of the eukaryotic cell. It contains one nucleus, one mitochondrion, one rudiment plastid (apicoplast), a compact network of endoplasmic reticulum, a single dictiosome composed of one (Plasmodium) or 3–5 (Toxoplasma, Eimeria) cisterns, and a complex of apical secretory organelles (Hager et al. 1999).

Secretory traffic in Toxoplasma gondii zoites Due to specific replication mechanisms (Hager et al. 1999; Hu et al. 2001) T. gondii zoites are strictly polarized; they contain an extensive microtubular network and multiple microtubule organizing centers (Morrissette and Sibley 2002). A centrally located nucleus divides the cell in two parts. ER is reduced and located posterior, behind the nucleus, and the perinuclear space makes an essential part of the total ER volume. Small coated vesicles budding off the apical side of the nuclear envelope move to the Golgi complex located in close vicinity. HDEL motive responsible for recycling of ER resident proteins from AG to ER, was localized by immunocytochemical and genetic markers just above the nucleus; thus it was proved that nucleus envelope served as ERGIC (ER–Golgi intermediate compartment ) (Hager et al. 1999). The presence of a single Golgi apparatus in T. gondii was noted in early EM studies and was recently confirmed by three-dimensional reconstruction of serial EM thin sections (Pelletier et al. 2002). In T. gondii, the single Golgi is located apical to the nucleus, adjacent to the centrosomes, and closely associated with a single ER exit site, which appears to be a specialized region of the nuclear envelope (Hager et al. 1999; Hartmann et al. 2006; He 2007). Each subsequent zoite division splits GA in halves. By video fluorescence microscopy and three-dimensional reconstructions of serial thin sections it was demonstrated that the new GA grows by autonomous duplication of the old one, which raises the possibility that the Golgi is a paired structure analogous to centrioles (Pelletier et al. 2002; see Chapter X).

Structure and function of the Golgi organelle in parasitic protists

*

653

The GA has been shown to be essential for the formation of three specialized secretory organelles, found in zoites of all apicomlexans: rhoptries, micronemes and dense granules (Joiner and Roos 2002). Dense granules distributed uniformly all over the cytoplasm, rhoptries and micronemes located apically. Secretion from these organelles is critical for parasite invasion and establishment of intracellular infections. Morphologically and functionally dense granules resemble mature secretory granules in sectertory cells of mammals. Rhoptries display features of both endosomes and secretory granules (Bishop and Woodmane 2000; Foussard et al. 1991). T. gondii zoites were the first parasitic protists, in which tyrosine-dependent mechanisms of protein sorting was identified (Joiner et al. 1990; Mordue et al. 1999; Sibley et al. 1985). The Toxoplasma genome (http://toxoDB.org) possesses all seven of the predicted COPI subunit homologues, found in mammalian cells. BetaCOP has amino acid insertions specific to T. gondii and a C-terminal insertion that is unique to apicomplexan parasites (Smith et al. 2007). Apicoplast beta-COP changed at most, which lead to perhaps an overall alteration of function. Forty eight residues of the beta-COP are likely to be functionally important, because they exhibit subtle yet specific amino acid changes among apicomplexans, kinetoplastids, and fungi (Smith et al. 2007). Protein sorting and transport at early stages of the secretory pathway in Toxoplasma are regulated mainly by conservative mechanisms similar to those in yeast and mammalian cells: insert of hydrophobic motives in cytoplasmic domains of cargo proteins; COP2 coating; COP1-dependent recycling of resident proteins; formation of adaptin complexes, etc. (Ajioka et al. 1998; Liendo et al. 2001; Stedman et al. 2003). Signal sequences, such as the bipartite terminal NH2 domain, which mediates co-translational translocation of the plastid proteins into ER and subsequent post-translation translocation into the apicoplast, have been revealed by a series of molecular and genetic studies (Roos et al. 1999; Waller et al. 2000; Yung et al. 2001). In T. gondii protein transport via Golgi is inhibited by low temperatures, BFA and nocodazole, a ubiquitous inhibitor of microtubule assembly (Soldati et al. 1998; Stokkermans et al. 1996). Vesicles budding from distal parts of trans-cisterns are coated with clathrin (Liendo et al. 2001). Many surface antigens of T. gondii have a conservative transmembrane glycosylphosphatidyl inositol (GPI) motive which serves as a signal for building in into the plasma membrane (Karsten et al. 1998). It was demonstrated that in T. gondii Rab6 GTPase mediated the retrograde transport of proteins from dense granules to GA. T. gondii mutants with aberrant Rab6 expression demonstrated abnormal cytokinesis (multiple instead of binary fission) suggesting the role of Rab6 in coupling of mitosis and cytokinesis (Stedman et al. 2003). Figure 2 summarizes current interpretation of mechanisms underlying secretory traffic in zoites of Apicomplexa.

654

*

Y. Y. Sokolova and A. A. Mironov

Figure 2. Scheme of post-Golgi protein transport in zoites of Apicomplexa, summarizing the existing experimental data (according Joiner and Roos 2002). Protein traffic from ER to Golgi organelle and through Golgi cisternae occurs in vesicles coated with coatomer proteins COP1 and COP2. It is regulated by conservative signal sequences and by small GTPases of Rab family. Soluble proteins of dense granules (DG) are transported from Golgi to DG by default in a signal independent way, whereas transport of membrane proteins is regulated by length of the transmembrane domain. Proteins of rhoptries (ROP) and micronemes (MN) are transported from Golgi via the specialized rhoptries precursor compartment. Transmembrane proteins of ROP and MN contain signal motifs for tyrosin- and adaptin-dependent sorting. Targeting soluble MN proteins requires binding with transmembrane escort proteins. Proteins targeted to apicoplast, possess NH2-terminal domain, which directs protein transport first to AG, and then, after cleavage of the terminal peptide, into the apicoplast lumen, using transport peptide, homologous to one found in plants. It is not clear, whether all proteins synthesized in Golgi are transported through apicoplast (black dashed arrows). Direction of products processed in apicoplast is yet unknown. Solid gray lines indicate traffic routes proved by direct experiments, black dashed lines—hypothetical pathways.

Erythrocyte stage of Plasmodium and extracellular protein transport Structure and function of the Golgi in Plasmodium falciparum, the intraerythrocyte parasite that causes malaria in humans, depends on its life cycle. In human erythrocytes P. falciparum resides inside a parasitophorous vacuole (PV), and exports the synthesized proteins through the plasmalemma and the parasitophorous vacuole membrane (PVM) into erythrocyte cytoplasm. Previous results suggesting that this parasite exports its exoicytic system into the host cells (Banting et al. 1995; (Lauer et al. 1997) have been recently argued (Adisa et al. 2007; Struck et al. 2005). In the intraerythrocyte stage, the P. falciparum ER consists of the nuclear envelope with two small protrusions that develop into an extended reticular network as the parasite enlarges (Van Dooren et al. 2005). Neither structural,

Structure and function of the Golgi organelle in parasitic protists

*

655

nor biochemical approaches detected presence of typical Golgi in the asexual erythrocyte stages of the plasmodium life cycle – trophozoites and the ring stage. As the ring stage develops, the ER expands from the nuclear membrane to form a reticulum throughout the cell. When parasite divides, the ER forms around the individual dividing nuclei but remains connected with the ER of other developing merozoites until very late in schizogony (Adisa et al. 2007; Van Dooren et al. 2005). In 3D, the Golgi-like compartment of P. falciparum is consisted of 1–3 tubular or flattened cisterns surrounded be vesicles of different size budding off the perinuclear space (Bannister et al. 2004). The Golgi is initially present at one or two foci that multiply as the parasite matures (Adisa et al. 2007). Many proteins involved in the exocytosis, such as: P. falciparum ERC (PfERC), PfBip (Kumar et al. 1988; La Greca et al. 1997; Van Dooren et al. 2005), the COPI protein, Pfbeta-COP (Adisa et al. 2001), PfERD2 – the retrieval receptor homologues to the signal recognition particle (Van Wye et al. 1996), and others (Gardner et al. 2002), have been identified, although the machinery for protein glycosylation is minimal. Coatomer proteins COPI, COPII, and other elements of the vesicle fusion machinery and intracellular traffic, such as Sar1p, Sec 31p, Sec 23, Pf NSF, etc. have been immunolocalized. The PfSar1p defines a network of membranes wrapped around parasite nuclei (Cooke et al. 2004; Hayashi et al. 2001). PfBet3p is largely present as a membrane- or cytoskeleton-bound pool (Adisa et al. 2007). PfGRASP colocalized with the cis Golgi marker ERD2 (Struck et al. 2005), but apart from the trans-Golgi marker PfRab6, suggesting that the cisand trans-Golgi compartments are spatially separated in P. falciparum cells (Adisa et al. 2007). Transport of most proteins is inhibited by BFA via interaction with ARF. Resident proteins of the parasitophorous vacuole (PV), as well as transit proteins directed to erythrocytes, are released into the PV lumen from the transport carriers. Electron microscopy revealed characteristic vesicles encircled by double membrane (DMV), budding off ER in the parasite cell. External membrane of vesicles fuses with the parasite plasmalemma and releases “daughter” internal vesicles which, in turn, fuse with PVM and discharge the content into the erythrocyte cytoplasm (Cooke et al. 2004; Olliaro and Castelli 1997). Application of green fluorescent protein and luciferase in combination with transfection of P. falcipatum erythrocyte stages revealed signals targeting parasite proteins to different membrane compartments (Klemba et al. 2004; Tabe et al. 1984). It was shown that translocation of soluble parasitic proteins through PVM requires energy in a form of ATP, and most likely the ATP-dependent translocators are involved (Ansorge et al. 1996). Plasmodium creates its own membrane system in the infected erythrocyte, but some important for the parasite pathogenesis proteins were detected free in erythrocyte cytoplasm, not surrounded by membrane envelopes.

656

*

Y. Y. Sokolova and A. A. Mironov

These proteins pass through PV as polypeptides lacking secondary structure (unfolded proteins). In the erythrocyte cytoplasm these proteins undergo folding; they accumulate in the specialized cell regions and interact with the erythrocyte resident proteins (Taraschi et al. 2001, 2003). Characteristic membrane compartments arise in the erythrocyte cytoplasm whilst parasite matures inside PV. In the ring stage PVM produces short finger-shaped protrusions. These protrusions bud off the PVM and give rise to small membrane carriers (Bannister et al. 2003, 2004). Elongated membrane cisterns with dense walls and electron transparent content known as Maurer’s Clefts (MCs), appear inside the erythrocyte. Role of MCs in transport of parasite antigens has been studied in connection with Pf EMP1 (P. falciparum erythrocyte memebrane protein). PfEMP 1 protein is a polymorphic integral protein, which mediates adhesive properties of infected erythrocytes. Adhesion of erythrocytes to the vascular surface causes the lethal syndromes in malaria patients. It is assumed that MCs serve as a transit depot while transporting of PfEMP1. This protein is anchored to the MC membrane by the C-terminal domain. Also it has been demonstrated that some cytosolic proteins are transported to the erythrocyte membrane in complexes adhered to the MC membrane (Cooke et al. 2004). Another membrane structure, a tubular–vesicular network (TVN) connected with PVM, was revealed in cytoplasm of infected erythrocytes. The membranes of TVN and PV lack electron dense coating typical for MCs, and they have similar antigen composition. TVN is assumed to take part in transporting the components of the erythrocyte PM from the environment to the parasite cell (Lauer et al. 1997). Many elements of vesicular transport machinery of the parasite origin have been immunolocalized in the cytoplasm of the infected erythrocytes, including proteins of coatomer complexes Sar1p, Sec 31p, Sec 23, Pf NSF, etc. (Cooke et al. 2004; Hayashi et al. 2001).

Golgi apparatus in Kinetoplastida Trypanosoma brucei, T. cruzi, and Leishmania spp. are the causative agents of devastating diseases—sleeping sickness, Chagas’ disease, and human visceral leishmaniasis correspondingly (Overath and Engstler 2004), and in these species the secretory apparatus was especially thoroughly studied. Basically, secretory apparatus in parasitic kinetoplastids, as well as in their close relatives, free-living Euglenoidea, is organized in a similar way as in mammalian cells. Golgi apparatus of kinetoplastids is composed of one dictiosome with 6–12 flat cisterns, divided in four standard compartments (cis, medial, trans, and trans-Golgi network), easily visualized by the marker proteins (Duszenko et al. 1988; McConville et al. 2002a, b; Weise et al. 2000). Pharmacological and cell biological studies suggest an intimate relationships between the Golgi and the basal bodies (Field et al. 2000; He et al. 2004). Recently it was shown that in dividing flagellates GA was formed de novo

Structure and function of the Golgi organelle in parasitic protists

*

657

from ER in ERGIC compartment (Golgi export sites), which is located in trypanosomes close to the nucleus (He et al. 2004). Thus Golgi biogenesis in T. bruceri supports the de novo biogenesis model, versus the template model exemplified by Toxoplasma gondii (He 2007; Chapter X). Transport of endoproteases, protein-like phophatases, lipophosphoglucans and membrane glucose transporters in Leishmania spp.; L-mannosidase in T. cruzi, and variable surface glycoproteins (VSGs) in Trypanosoma spp. was in the focus of numerous studies in search for targets of potential vaccines (for review see Becker and Melkonian (1996)). Trypanosomatids are characterized by an extremely high velocity of endocytosis, exocytosis, sorting and concentration of the exported and recycled proteins. Every minute 107 surface glycoproteins undergo recycling from ER, the site of their synthesis, to PM and back. Ten percent of VSGs synthesized by T. brucei, contain the anchoring GPI motif, which provides building of these proteins into the membrane (versus only 0.5% of surface proteins in mammalian cells), and all of them are glycosylized (Overath and Engstler 2004). In trypanosomids N-glycosylation, insertion of GPI anchors, and other elements of protein processing usually (but not always) take place in GA and use mechanisms and enzyme systems described for higher eukaryotes (Parodi 1993; Rubotham et al. 2005). Immune co-localization of these proteins and GA markers demonstrated that T. brucei VSGs directed to the cell surface by the standard pathway, were revealed in the ER, GA, trans-Golgi network and, finally at the PM surface (Duszenko et al. 1988). Fifty-fold increase in VSGs concentration was recorded during their transportation to PM (Grunfelder et al. 2002). It remains unresolved why monezin, a ionophore of monovalent cations, which inhibits GA–PM transport in most cellular system (Mollenhauer et al. 1990), does not effect VSGs secretion, although it causes adequate alterations in GA morphology (swelling of trans-Golgi compartment), and blocks N-glycan synthesis (Bangs et al. 1986; Duszenko et al. 1988). It was suggested that additional pathway for VSGs synthesis and recycling may exist (Ferguson et al. 1986; Grunfelder et al. 2002). T. brucei Rab1, 2, 18 and X2 all localize to the Golgi (Ackers et al. 2005), but show different distribution patterns in bloodstream (BSF) versus insect (procyclic) forms of the parasite (Dhir et al. 2004; Field et al. 2000). For example, TbRab18 is expressed only in BSF (Jeffries et al. 2002). Depletion of ARL1, the Golgi-localized ARF-like protein, disrupts Golgi stacks and leads to cell death only in the BSF cells (Price et al. 2005). Developmentally regulated trafficking of the lysosomal membrane protein P67 (Alexander et al. 2002) and procyclin (Engstler and Boshart 2004), a major surface glycoprotein in procyclic cells, has also been reported. Furthermore, different effects of actin depletion on Golgi morphology have been noted in BSF, but not in procyclic cells (Garcıa-Salcedo et al. 2004). Sequencing of genes encoding some proteins of kinetoplastid secretory pathway revealed remarkable similarity to genes for analogous proteins in

658

*

Y. Y. Sokolova and A. A. Mironov

yeast and mammalian cells (Bangs et al. 1993; Descoteaux et al. 1995; Meissner et al. 2002; Ryan et al. 1993).

Structure and function of the Golgi-like organelle in Microsporidia Microsporidia are intracellular parasites that infect all taxa of animals with bilateral symmetry (Bilateria). Basing on ultrasrtuctural data which showed no mitochondria, hydrogenosomes or peroxisomes, no stacked Golgi dictyosomes, no 9 þ 2 structures, the 70S ribosomes of the prokaryote type and the ancient type of mitosis, and on early SSUrDNA-inferred phylogenies, Microsporidia were once believed to be an early diverged lineage of Eukaryotes evolved before acquisition of mitochondria (Vossbrinck et al. 1987; Vossbrinck and Woese 1986). The later research evidenced though that microsporidia are highly derived rather than primitively simple (Keeling and Slamovits 2004). Three groups of facts ruined the hypothesis of the ancient ancestry: (i) it was discovered that microsporidia possess genes for mitochondria-targeted proteins (Germot et al. 1996), and structurally recognizable mitochondria relicts, the mitosomes (Williams et al. 2002); (ii) molecular phylogenetic analyses inferred from several genes, showed that microsporidia do not evolve early in the eukaryotic evolution, but are either Fungi or their close relatives (Fischer and Palmer 2005; Keeling 2003; Keeling et al. 2000; Thomarat et al. 2004); (iii) information from completed (Encephalitozoan cuniculi (Katinka et al. 2001)), and oncoming Spraguea lophii and Nosema locustae genome projects revealed high homology with yeasts alongside with specific genome organization, which reflected reductive evolution of these organisms due to parasitic lifestyle (Fedorov and Hartman 2004; Katinka et al. 2001; Keeling 2001). Currently it is widely accepted that microsporidia (phylum Microsporidia Balbiani 1882) are highly specialized lineage of Fungi (Adl et al. 2005; Arisue et al. 2005; Keeling 2003; Richards and Cavalier-Smith 2005; Stechmann and Cavalier-Smith 2003). Mirosporidia life cycle consists of a proliferative stage (meronts and sporonts), sporogenic stage (sporoblasts), and spores, the infectious stage and the only one that can survive in the environment. Ultrastructurally meronts can be characterized by the absence of extracellular envelopes and a very few membrane structures in their cytoplasm. Sporonts are defined by appearance of the electron dense envelope outside the plasma membrane, the precursor of the exospore (the outer layer of the spore wall). GA is hardly visible in meronts and early sporonts. In the latter it appears as a conglomerate of membrane profiles of about 30 nm in diameter (Vavra and Larsson 1999) accumulated in the vicinity of ER cisterns deriving from the perinuclear space (Sokolova et al. 2001). At late proliferative stages the conglomerate grows in size, includes more elongated profiles, and

Structure and function of the Golgi organelle in parasitic protists

*

659

forms a “tubular cluster”, which looses connection with the nucleus, migrates to the distal end of the cell and eventually transforms into trans-Golgi tubular network associated with polar filament proteins (PTP) pathway in prespore stages. Application of osmium impregnation techniques, which results in deposition of insoluble osmium specifically inside the lumens of cis-Golgi, suggested the portions of perinuclear space with associated elongated ER cisterns, conglomerates of 30 nm- membrane profiles and “tubular clusters” belong to the cis-Golgi compartment (Sokolova et al. 2001). At the sporoblast stage morphogenesis of the internal structures of the spore takes place. Sporoblasts possess a well-developed voluminous Golgi, portion of which function as a container for the polar filament proteins (PFP). The spore (1–25 mm in length depending on the species) is defined by appearance of the electron transparent endospore (the inner layer of the spore wall). Spores are equipped with a unique set of organelles – the extrusion apparatus, which functions when the spore content is injected into a host cell (Vavra and Larsson 1999). The central role in the process of infection belongs to the polar filament (PF). When the spore is activated to firing, PFPs undergo self- assemblage causing transformation of the filament into a tube (Keohane and Weiss 1999; Weidner 1982; Weidner and Byrd 1982). This tube

Figure 3. Sections through secretory compartment in microsporidia proliferative stages and spores. (a). Golgi organelle in meronts can be visualized as conglomerates of 20–40 nm membrane profiles (arrows) connected with the perinuclear space and ER cisterns. (b). In sporonts, conglomerates grow in size and transforms into a “tubular cluster” (arrow), which eventually transforms into trans-Golgi tubular network associated with polar filament pathway in pre-spore stages. (c). Late Golgi compartment in a developing spore is composed of tubular networks (TN1 and TN2) and membrane containers with polar filament elements (PF and APF). Polar filament protein-containing profiles are budding off the TN2 (arrow). APF – apical portion of the PF; ER – endoplasmic reticulum; N – parasite nuclei arranged in diplokaria; PF – polar filament; PP – primordial polaroplast (a part of the infection machinery, presumably ER derivate); SW – spore wall; TN 1 and TN2–trans Golgi tubular networks. Figure 3c is provided by E. Seliverstova

660

*

Y. Y. Sokolova and A. A. Mironov

delivers the sporoplasm into the target cell, like a syringe. In sporoblasts and mature spores PFPs reside in membrane-bound tubular compartment seen on cross-sections as rows of coils underlying the spore surface. PF coils are often associated with tubular networks, well seen in immature spores and sporoblasts (Fig. 3). Staining with thiamin pyrophosphatase (Takvorian and Cali 1994; YS, unpublished observations) proved the PFP-containing profiles and tubular networks to be homologous to trans-Golgi compartment of other eukaryotes. Actually the whole GA, which in spore is represented by the transGolgi compartment, is being transformed into the polar filament. Interestingly, the membrane contours surrounding PF coils remain inside the spore shell after the sporoplasm discharge; Ca2 þ influx seemingly plays a pivotal role in PF discharge which is inhibited by calcium channel antagonists and calmodulin inhibitors (Keohane and Weiss 1999; Pleshinger and Weidner 1985; Weidner 1982; Weidner and Byrd 1982), thus resembling a specialized version of exocytosis analogous to trichocyst discharge in ciliates (Plattner 1993; Plattner et al. 1991). Quick-freezing cryosubstitution and chemical fixation, followed by 3-D tomography demonstrated that the Golgi analogs of at least two microsporidia Paranosema grylli and P. locustae appeared as 300-nm networks of thin (25–40-nm in diameter), branching or varicose tubules, that displayed histochemical features of a Golgi (Beznoussenko et al. 2007). Interestingly, that Golgi-like structures of microsporidia never displayed vesicles, even when the membrane fusion was inhibited. These tubular networks were connected to the perinuclear space (Snigirevskaya et al. 2006), endoplasmic reticulum, the plasma membrane and the forming polar tube (Beznoussenko et al. 2007). They were positive for microsporidian Sec13, subunits of COP and analogs of giantin and GM130. The spore-wall and polar tube proteins were transported from the ER to the target membranes through these tubular networks, within which they underwent concentration and glycosylation (Beznoussenko et al. 2007). Importantly, the intracellular transport of secreted proteins in microsporidia occurs by a progression mechanism that does not involve the participation of vesicles generated by coat proteins I and II (Beznoussenko et al. 2007). We believe that the model of “avesicular” transport is not unique to microsporidia. Analyses of the minute (2.9 Mbp) and completely sequenced genome of the mammalian microsporidium E. cuniculi (Katinka et al. 2001) and partly sequenced genome of P. (Antonospora) locustae parasitizing insects (genome size 5.4 Mbp) (http://jbpc.mbl.edu/ Nosema/index.html) revealed all the most important protein machineries that are involved in co-translational translocation of polypeptide chains into the ER lumen, in intracellular transport, and exocytosis, well-characterized in yeast and mammalian cells; although some of these machineries lack non-essential components. E. cuniculi has a limited number of proteins involved in intracellular transport (Katinka et al. 2002), including two subunits of Sec61, the proteins

Structure and function of the Golgi organelle in parasitic protists

*

661

responsible for the incorporation of polypeptide chains into the ER lumen, and seven enzymes involved in glycosylation. Among these, five are glycosyltransferases. Among six SNAREs, there are two R-SNAREs (SNC2 and synaptobrevin) and four Q-SNAREs (syntaxin 5, VAMP, Bos1 and Vti1). Importantly, Microsporidia and yeast SNAREs are very similar; and the latter, in turn, display high degree of sequence similarity to the mammalian SNAREs (Von Mollard et al. 1997). The machinery responsible for the dismantling of SNARE complexes is represented by only one protein, Sec18 (homologous to mammalian NSF), although there are no proteins that are homologous to SNAPs (Katinka et al. 2001). This means that the SNARE machinery in microsporidia might work slower than in mammalian cells. The minimal set of Rab proteins (Ypt1, Rab1b, Rab5, Ytp6 and Rab10) and the Rab-GDP dissociation inhibitor are present in E. cuniculi genome. Instead of the seven subunits of COPI typical for mammalian and plant cells, microsporidia have only six, with the epsilon-COPI missing (Katinka et al. 2001). In mammalian cells, epsilon-COPI mediates generation of COPI vesicles (Guo et al. 1994). The ARF machinery in microsporidia is also limited by only two ARFs and one exchange factor for ARF. Another important gene missing, besides epsilon-COPI, is ARF-GAP, instead E. cuniculi genome contains an ARFlike protein that can partially replace ARF-GAP (Lu et al. 2001). On the other hand, among four known subunits of the COPII machinery, only three, Sec13, Sec23, and Sec31, are found in microsporidia (Katinka et al. 2001). Although Sar1p is present, Sec12 that operates as a Sar1 exchange factor is absent. Finally, microsporidia lack clathrin, which can be explained by lack of lysosomes and absence of endocytosis. It can be concluded that microsporidia exploit the similar basic machinery for protein transport and secretion as mammalian and yeast cells marked by extreme reduction of non-functional components. Additionally, analysis of microsporidia cells provides evidence in favor of minimal role of the coated vesicles as transport carriers.

Golgi organelle on the phylogenetic tree of eukaryotes Once, dictiosomes were considered as an important argument in phylogenetic reconstructions of the Tree of Life, not less important than flagella or mitochondria. Basing on the presence or absence of the stacked Golgi cisterns the kingdom Protozoa was divided into two subkingdoms: Adictyiozoa and Dictyozoa (Cavalier-Smith 1993). It is just a curious episode of only historical value; such taxons do not exist anymore. Indeed, presence or absence of typical dictiosomes cannot serve as a reliable character to differentiate major taxons, because structure of Golgi organelle my vary not only between closely related taxons (stacks are absent in Oxymonad and present in Trymastix (Dacks and Doolittle 2001)) but within the life span of the same species (like in Giardia and Plasmodium).

Homo Saccharomyces Encephalitozoon Dictyostelium Entamoeba Mastigamoeba Arabidopsis Chlamydomonas Porphyra Phytophthora Giardia Trypanosoma Naegleria

Metazoa, Bilateria Fungi Microsporidia Eumycetozoa Entamoebidae Mastigamoebidae Embryophyta Chlorophyceae Rhodophyceae Stramenopiles Diplomonadida Kinetoplastida Heterolobosea

Snap 25b þ þ NR NR NR ? þ NR NR NR NR NR ?

Synt-axins þ þ þ þ þ ? þ þ þ þ þ þ ?

þ þ þ þ þ ? þ þ þ þ þ þ ?

Rab þ þ þ NR þ ? þ þ NR þ þ þ ?

Arf-gap þ þ þ NR þ NR þ þ þ þ þ þ þ

b-COP þ þ þ þ þ þ þ þ þ þ þ þ NR

Vps

Proteins involved in endomembrane traffica

þ þ NR þ þ þ þ þ þ þ þ þ NR

AP

þ þ þ þ þ ? þ þ NR þ þ þ ?

Sec1

a

Syntaxins – the proteins of SNARE family, encoded by genes homologous to sso, sed5, pep12 Saccharomyces); Snap25 – gene for SNAP (soluble NSFatachment protein, essential element of fusion machinery), homologous sec9 in Saccharomyces); rab–proteins of GTPase family, which participate in ER – Golgi transport; arf-gap – proteins of GTPase family, participate in intra-Golgi transport; beta-COP is a subunit of coatomer protein, participate in the retrograte ER – Golgi transport; Vps – complex of retromer proteins, encoded by Vps26 and ps35 genes, involved in protein recycling in trans-Golgi network; AP – family of adaptor proteins, involved in coating of clathrin vesicules; Sec1 – a protein, which mediates anchoring of SNARE (SNAP receptors) to the membrane. b Snap 25 was detected only in mammalian cells, higher plants and yeast and, therefore, can hardly regarded as a conservative protein. NR – not revealed; “?”– not examined.

Genus

*

Taxon

Table 1. Proteins associated with the Golgi apparatus in different groups of eukaryotes. (Based on data from genome projects (Dacks and Doolittle 2002; Dacks et al. 2003; Katinka et al. 2001; Fedorov and Hartmann 2004), and original papers on Naegleria and Mastigamoeba (Dacks et al. 2003))

662 Y. Y. Sokolova and A. A. Mironov

Structure and function of the Golgi organelle in parasitic protists

*

663

In three of six supertaxa defined by the new classification of protists (Adl et al. 2005) there are at minimum eight groups of species, in which GA does not display stack organization (Table 1, Fig. 1). However, molecular–genetic and/or biochemical data point to the presence of the organelle with Golgi functions in all lineages examined, with the exceptions of Oxymonada and Retortamonada. So far no proves has been obtained for these organisms, but in the closely related taxa GA was revealed either in hidden (Giardia) or apparent (Trimastix) form (Dacks et al. 2003). The fact that the organelle with Golgi functions presents in all contemporary groups of eukaryots suggests that a “common eukaryotic ancestor” possessed GA, as well as mitochondria, introns, splicing mechanisms, and all basic biological and molecular features of an eukaryotic cell (Dacks and Doolittle 2001). The simplicity of “early diverged lineages” most likely is a consequence of secondary loss of cell structures, for example, as result of switching to parasitism. The complexity of organization and conservatism of proteins and genes involved in secretory traffic in different eukaryotic groups (Table 1 and 2) strongly suggest that GA arose once in the evolution of the eukaryotic cell before acquisition of coat machinery.

Table 2. Some features of the Golgi in Protists Feature

Parabasalia Diplomonada Entamoebidae

Apicomplexa Kinetoplastida

Microsporidia

Segregation from the ER Periodical continuity with the ER Periodical continuity with the post-Golgi Network of smooth and varicose tubules Presence of more than two compartments Stacked disk-like cisternae COPI vesicles Clathrin vesicles Movement by actin (fragmented Golgi) Movement by microtubules H þ -ATP pump Golgi glycosidases Nucleotide transporters Matrix proteins Sar1/COPII ARF/COPI AP/clathrin SNAREs

þ þ

þ þ

þ þ

þ

þ

þ

þ

þ

þ

þ

þ

þ

þ þ þ 

þ þ þ 

   

þ þ þ þ þ þ ? þ þ

 þ þ þ  þ ? ? þ

 þ Reduced þ  Reduced Reduced  Reduced

664

*

Y. Y. Sokolova and A. A. Mironov

References Ackers JP, Dhir V, Field MC (2005) A bioinformatic analysis of the RAB genes of Trypanosoma brucei. Mol Biochem Parasitol 141: 89–97 Adam RD (2001) Biology of Giardia lamblia. Clin Microbiol Rev 14: 447–475 Adisa A, Albano FR, Reeder J, Foley M, Tilley L (2001) Evidence for a role for a Plasmodium falciparum homologue of Sec31p in the export of proteins to the surface of malaria parasite-infected erythrocytes. J Cell Sci 114: 3377–3386 Adisa A, Frankland S, Rug M, Jackson K, Maier AG, Walsh P, Lithgow T, Klonis N, Gilson PR, Cowman AF, Tilley L (2007) Re-assessing the locations of components of the classical vesicle-mediated trafficking machinery in transfected Plasmodium falciparum. J Parasitol 37: 1127–1141 Adl SM, Simpson AGB, Farmer MA, Andersen RA, Anderson OR, Barta JR, Browser SS, Brugerolle G, Fensome RA, Fredericq S, James TY, Karpov S, Kugrens P, Krug J, Lane CE, Lewis LA, Lodge J, Lynn DH, Mann DG, McCourt RM, Mendoza L, Moestrup O, Mozley-Standridge SE, Nerad TA, Shearer CA, Smirnov AV, Spiegel FW, Taylor M (2005) The new higher level classification of eukaryotes with emphasis on the taxonomy of protists. J Eukaryot Microbiol 52: 399–451 Ajioka JW, Boothroyd JC, Brunk BP, Hehl A, Hillier L, Manger ID, Marra M, Overton GC, Roos DS, Wan KL, Waterston R, Sibley L D(1998) Gene discovery by EST sequencing in Toxoplasma gondii reveals sequences restricted to the Apicomplexa. Genome Res 8: 18–28 Alexander DL, Schwartz KL, Balber AE, Bangs JD (2002) Developmentally regulated trafficking of the lysosomal membrane protein p67 in Trypanosoma brucei. J Cell Sci 115: 3243–3263 Allan D, Kallen KJ (1994) Is plasma membrane lipid composition defined in the exocytic or the endocytic pathway? Trends Cell Biol 4: 350–353 Allan RB, Balch WE (1999) Cell biology – Protein sorting by directed maturation of Golgi compartments. Science 285: 63–66 Allan V (1995) Membrane traffic motors. FEBS Letters 369: 101–106 Andersson SGE, Kurland CG (1999) Origins of mitochondria and hydrogenosomes. Curr Opin Microbiol 2: 535–541 Ansorge I, Benting J, Bhakdi S, Lingelbach K (1996) Protein sorting in Plasmodium falciparum-infected red blood cells permeabilized with the pore-forming protein streptolysin O. Biochem J 315: 307–314 Arisue N, Hasegawa M, Hashimoto T (2005) Root of the eukaryota tree as inferred from combined maximum likelihood analyses of multiple molecular sequence data. Mol Biol Evol 22: 409–420 Arisue N, Sachez LB, Weiss LM, Muller M, Hashimoto T (2002) Mitochondrial-type hsp70 genes of the amitochondriate protists, Giardia intestinalis, Entamoeba histolytica and two microsporidians. Parasitol Int 51: 9–16 Bahl A, Brunk B, Coppel RL, Crabtree J, Diskin SJ, Fraunholz MJ, Grant GR, Gupta D, Huestis RL, Kissinger JC, Labo P, Li L, McWeeney SK, Milgram AJ, Roos DS, Schug J, Stoeckert Jr CJ (2002) PlasmoDB: the Plasmodium genome resource. An integrated database providing tools for accessing, analyzing and mapping expression and sequence data (both finished and unfinished). Nucleic Acids Res 30: 87–90 Baldi DL, Andrews KT, Waller RF, Roos DS, Howard RF, Crabb BS, Cowman AF (2000) RAP1 controls rhoptry targeting of RAP2 in the malaria parasite Plasmodium falciparum. EMBO J 19: 2435–2443 Bangs JD, Andrews NW, Hart GW, Englund PT (1986) Posttranslational modification and intracellular-transport of a trypanosome variant surface glycoprotein. J Cell Biol 103: 255–263

Structure and function of the Golgi organelle in parasitic protists

*

665

Bangs JD, Hereld D, Krakow JL, Hart GW, Englund PT (1985) Rapid processing of the carboxyl terminus of a trypanosome variant surface glycoprotein. Proc Natl Acad Sci USA 82: 3207–3211 Bangs JD, Uyetake L, Brickman MJ, Balber AE, Boothroyd JC (1993) Molecular cloning and cellular localization of a Bip homolog in Trypanosoma brucei – divergent ER retention signals in a lower eukaryote. J Cell Sci 105: 1101–1113 Bannister LH, Hopkins JM, Dluzewski AR, Margos G, Williams IT, Blackman MJ, Kocken CH, Thomas AW, Mitchell GH (2003) Plasmodium falciparum apical membrane antigen 1 (PfAMA-1) is translocated within micronemes along subpellicular microtubules during merozoite development. J Cell Sci 116: 3825–3834 Bannister LH, Hopkins JM, Margos G, Dluzewski AR, Mitchell GH (2004) Three-dimensional ultrastructure of the ring stage of Plasmodium falciparum: evidence for export pathways. Microsc Microanal 10: 551–562 Banting G, Benting J, Lingelbach K (1995) A minimalist view of the secretoryp pathway in Plasmodium falciparum. Trends Cell Biol 5: 340–343 Becker B, Melkonian M (1996) The secretory pathway of protists: spatial and functional organization and evolution. Microbiol Rev 60: 697–721 Beh CT, Rose MD (1995) 2 Redundant systems maintain levels of resident proteins within the yeast endoplasmic reticulum. Proc Natl Acad Sci USA 92: 9820–9823 Benchimol M, Ribeiro KC, Mariante RM, Alderete JF (2001) Structure and division of the Golgi complex in Trichomonas vaginalis and Tritrichomonas foetus. Eur J Cell Biol 80: 593–607 Bevis BJ, Glick BS, Mogelsvang S, Staehelin LA (2002) De novo formation of transitional ER and Golgi compartments in Pichia pastoris. Mol Biol Cell 13: 283A Beznoussenko GV, Dolgikh VV, Seliverstova EV, Semenov PB, Tokarev YS, Trucco A, Micaroni M, Di Giandomenico D, Auinger P, Senderskiy IV, Skarlato SO, Snigirevskaya ES, Komissarchik YY, Pavelka M, De Matteis MA, Luini A, Sokolova YY, Mironov AA (2007) Analogs of the Golgi complex in microsporidia: structure and avesicular mechanisms of function. J Cell Sci 120: 1288–1298 Beznoussenko GV, Mironov AA (2002) Models of intracellular transport and evolution of the Golgi complex. Anat Res 268: 226–238 Biderre C, Metenier G, Vivares CP (1998) A small spliceosomal-type intron occurs in a ribosomal protein gene of the microsporidian Encephalitozoon cuniculi. Mol Biochem Parasitol 94: 283–286 Bishop N, Woodmane P (2000) ATPase-defective mammalian VPS4 localizes to aberrant endosomes and impairs cholesterol trafficking. Mol Biol Cell 11: 227–239 Boman AL, Kahn RA (1995) Arf proteins – the membrane traffic police. Trends Biochem Sci 20: 147–150 Bonifacino JS, Dell’Angelica EC (1999) Molecular bases for the recognition of tyrosinebased sorting signals. J Cell Biol 145: 923–926 Bradley PJ, Boothroyd JC (2001) The pro region of Toxoplasma ROP1 is a rhoptrytargeting signal. Int J Parasitol 31: 1177–1186 Brecht S, Carruthers VB, Ferguson DJP, Giddings OK, Wang G, Jakle U, Harper JM, Sibley LD, Soldati D (2001) The toxoplasma micronemal protein MIC4 is an adhesin composed of six conserved apple domains. J Biol Chem 276: 4119–4127 Bredeston LM, Caffaro CE, Samuelson J, Hirschberg CB (2005) Golgi and endoplasmic reticulum functions take place in different subcellular compartments of Entamoeba histolytica. J Biol Chem 280: 32168–32176 Bretscher MS, Munro S (1993) Cholesterol and the Golgi-apparatus. Science 261: 1280–1281 Cavalier-Smith T (1983) A 6-kindom classification and aunited phylogeny. In: Schwemmler W, Schenk H (eds) Endocytobiology II. de Gruyter, Berlin, pp. 265–279 Cavalier-Smith T (1991) Cell evolution. In: Osawa S, Honjo T (eds) Evolution of life. Springer-Verlag, Tokyo, pp. 271–304

666

*

Y. Y. Sokolova and A. A. Mironov

Cavalier-Smith T (1993) Kingdom Protozoa and its 18 phyla. Microbiol Rev 57: 953–994 Cavalier-Smith T (1999) Principles of protein and lipid targeting in secondary symbiogenesis: Euglenoid, dinoflagellate, and sporozoan plastid origins and the eukaryote family tree. J Eukaryot Microbiol 46: 347–366 Cavalier-Smith T (2002) The phagotrophic origin of eukaryotes and phylogenetic classification of protozoa. Int J Sys Evol Microbiol 52: 297–354 Cavalier-Smith T (2004) Only six kingdoms of life. Proc Royal Soc Biol Sci 271: 1251–1262 Cavalier-Smith T, Chao EE (1996) Molecular phylogeny of the free-living archezoan Trepomonas agilis and the nature of the first eukaryote. J Mol Evol 43: 551–562 Chavez-Munguia B, Espinosa-Cantellano M, Castanon G, Martinez-Palomo A (2000) Ultrastructural evidence of smooth endoplasmic reticulum and Golgi-like elements in Entamoeba histolytica and Entamoeba dispar. Arch Med Res 31:S165–S167 Clayton C, Hausler T, Blattner J (1995) Protein trafficking in kinetoplastid Protozoa. Microbiol Rev 59: 325–344 Cole NB, Lippincottschwartz J (1995) Organization of organelles and membrane traffic by microtubules. Cur Opin Cell Biol 7: 55–64 Conibear E, Stevens TH (1995) Vacuolar biogenesis in yeast – sorting out the sorting Proteins. Cell 83: 513–516 Cooke BM, Lingelbach K, Bannister LH, Tilley L (2004) Protein trafficking in Plasmodium falciparum-infected red blood cells. Trends Parasitol 20: 581–589 Dacks JB, Davis LAM, Sjogren AM, Andersson JO, Roger AJ, Doolittle WF (2003) Evidence for Golgi bodies in proposed ‘Golgi-lacking’ lineages. Proc Royal Soc Biol Sci 270: S168–S171 Dacks JB, Doolittle WF (2001) Reconstructing/deconstructing the earliest eukaryotes: how comparative genomics can help. Cell 107: 419–425 Dacks JB, Doolittle WF (2002) Novel syntaxin gene sequences from Giardia, Trypanosoma and algae: implications for the ancient evolution of the eukaryotic endomembrane system. J Cell Sci 115: 1635–1642 De Duve C (1990) The primitive phagocyte. In: Nardon P, Gianiazzi-Pearson V, Grenier AM, Margulis L, Smith DC (eds) Endocytobiology IV. Institut National de la Recherche Agronomique, Paris, pp. 511–514 Descoteaux A, Luo Y, Turco SJ, Beverley SM (1995) A specialized pathway affecting virulence glycoconjugates of Leishmania. Science 269: 1869–1872 Dhir V, Goulding D, Field MC (2004) TbRAB1 and TbRAB2 mediate trafficking through the early secretory pathway of Trypanosoma brucei. Mol Biochem Parasitol 137: 253–265 Dubremetz JF, Achbarou A, Bermudes D, Joiner KA (1993) Kinetics and pattern of organelle exocytosis during Toxoplasma gondii host–cell interaction. Parasitol Res 79: 402–408 Duden R, Griffiths G, Frank R, Argos P, Kreis TE (1991) Beta-Cop, a 110 kDa protein associated with non-clathrin-coated vesicles and the Golgi complex, shows homology to beta-adaptin. Cell 64: 649–665 Duszenko M, Ivanov IE, Ferguson MAJ, Plesken H, Cross GAM (1988) Intracellular transport of a variant surface glycoprotein in Trypanosoma brucei. J Cell Biol 106: 77–86 Elmendorf HG, Dawson SC, McCaffery M (2003) The cytoskeleton of Giardia lamblia. Int J Parasitol 33: 3–28 Engstler M, Boshart M (2004) Cold shock and regulation of surface protein trafficking convey sensitization to inducers of stage defferenciation in Trypanosoma brucei. Genes Dev 18: 2798–2811 € nfelder CG, Gu € nzel M, Heddergott N, Overath P (2005 ) Engstler M, Weise F, Bopp K, Gru The membrane-bound histidine acid phosphatase TbMBAP1 is essential for endocytosis and membrane recycling in Trypanosoma brucei. J Cell Sci 118: 2105–2118

Structure and function of the Golgi organelle in parasitic protists

*

667

Fast NM, Doolittle WF (1999) Trichomonas vaginalis possesses a gene encoding the essential spliceosomal component, PRP8. Mol Biochem Parasitol 99: 275–278 Fast NM, Roger AJ, Richardson CA, Doolittle WF (1998) U2 and U6 snRNA genes in the microsporidian Nosema locustae: evidence for a functional spliceosome. Nucleic Acids Res 26: 3202–3207 Fedorov A, Hartman H (2004) What does the microsporidian E. cuniculi tell us about the origin of the eukaryotic cell? J Mol Evol 59: 695–702 Felsenstein J (1978) Cases in which parsimony or compatibility methods will be positively misleading. Systematic Zool 27: 401–410 Ferguson MAJ, Duszenko M, Lamont GS, Overath P, Cross GAM (1986) Biosynthesis of Trypanosoma brucei variant surface glycoproteins N-glycosylation and addition of a phosphatidylinositol membrane anchor. J Biol Chem 261: 356–362 Ferguson MAJ, Haldar K, Cross GAM (1985) Trypanosoma brucei variant surface glycoprotein has a Sn-1,2-dimyristyl glycerol membrane anchor at Its COOH terminus. J Biol Chem 260: 4963–4968 Fichera ME, Roos DS (1997) A plastid organelle as a drug target in apicomplexan parasites. Nature 390: 407–409 Field H, Sherwin T, Smith AC, Gull K, Field MC (2000) Cell-cycle and developmental regulation of TbRAB31 localisation, a GTP-locked RAB protein from Trypanosoma brucei. Mol Biochem Parasitol 106: 21–35 Fischer WM, Palmer JD (2005) Evidence from small-subunit ribosomal RNA sequences for a fungal origin of Microsporidia. Mol Phylogenet Evol 36: 606–622 Foussard F, Leriche MA, Dubremetz JF (1991) Characterization of the lipid content of Toxoplasma gondii rhoptries. Parasitology 102: 367–370 rez-Morga D, Gijón P, Dilbeck V, Pays E, Nolan DP (2004 ) A Garcıa-Salcedo JA, Pe differential role for actin during the life cycle of Trypanosoma brucei EMBO J l 23: 780–789 Gardner MJ, Hall N, Fung E, White O, Berriman M, Hyman RW, Carlton JM, Pain A, Nelson KE, Bowman S, Paulsen IT, James K, Eisen JA, Rutherford K, Salzberg SL, Craig A, Kyes S, Chan MS, Nene V, Shallom SJ, Suh B, Peterson J, Angiuoli S, Pertea M, Allen J, Selengut J, Haft D, Mather MW, Vaidya AB, Martin DM, Fairlamb AH, Fraunholz MJ, Roos DS, Ralph SA, McFadden GI, Cummings LM, Subramanian GM, Mungall C, Venter JC, Carucci DJ, Hoffman SL, Newbold C, Davis RW, Fraser CM, Barrell B (2002) Genome sequence of the human malaria parasite Plasmodium falciparum. Nature 419: 498–511 Gaut JR, Hendershot LM (1993) The modification and assembly of proteins in the endoplasmic reticulum. Curr Opin Cell Biol 5: 589–595 Germot A, Philippe H, LeGuyader H (1996) Presence of a mitochondrial-type 70-kDa heat shock protein in Trichomonas vaginalis suggests a very early mitochondrial endosymbiosis in eukaryotes. Proc Natl Acad Sci USA 93: 14614–14617 Germot A, Philippe H, LeGuyader H (1997) Evidence for loss of mitochondria in Microsporidia from a mitochondrial-type HSP70 in Nosema locustae. Mol Biochem Parasitol 87: 159–168 Ghosh S, Field J, Rogers R, Hickman M, Samuelson J (2000) The Entamoeba histolytica mitochondrion-derived organelle (crypton) contains double-stranded DNA and appears to be bound by a double membrane. Infect Immun 68: 4319–4322 Ghosh SK, Field J, Frisardi M, Rosenthal B, Mai ZM, Rogers R, Samuelson J (1999) Chitinase secretion by encysting Entamoeba invadens and transfected Entamoeba histolytica trophozoites: Localization of secretory vesicles, endoplasmic reticulum, and Golgi apparatus. Infection and Immunity 67: 3073–3081 Grunfelder CG, Engstler M, Weise F, Schwarz H, Stierhof YD, Boshart M, Overath P (2002) Accumulation of a GPI-anchored protein at the cell surface requires sorting at multiple intracellular levels. Traffic 3: 547–559

668

*

Y. Y. Sokolova and A. A. Mironov

Guo Q, Vasile E, Krieger M (1994) Disruptions in Golgi structure and membrane traffic in a conditional lethal mammalian cell mutant are corrected by epsilon-COP. J Cell Biol 125: 1213–1224 Hager KM, Striepen B, Tilney LG, Roos DS (1999) The nuclear envelope serves as an intermediary between the ER and Golgi complex in the intracellular parasite Toxoplasma gondii. J Cell Sci 112: 2631–2638 Hartman H, Fedorov A (2002) The origin of the eukaryotic cell: a genomic investigation. Proc Natl Acad Sci USA 99: 1420–1425 Hartmann J, Hu K, He CY, Pelletier L, Roos DS, Warren G (2006) Golgi and centrosome cycles in Toxoplasma gondii. Mol Biochem Parasitol 145: 125–127 Hashimoto T, Nakamura Y, Nakamura F, Shirakura T, Adachi J, Goto N, Okamoto K, Hasegawa M (1994) Protein phylogeny gives a robust estimation for early divergences of eukaryotes: phylogenetic place of a mitochondria-lacking protozoan, Giardia lamblia. Mol Biol Evol 11: 65–71 Hashimoto T, Sanchez LB, Shirakura T, Muller M, Hasegawa M (1998) Secondary absence of mitochondria in Giardia lamblia and Trichomonas vaginalis revealed by valyltRNA synthetase phylogeny. Proc Natl Acad Sci USA 95: 6860–6865 Hayashi M, Taniguchi S, Ishizuka Y, Kim HS, Wataya Y, Yamamoto A, Moriyama Y (2001) A homologue of N-ethylmaleimide-sensitive factor in the malaria parasite Plasmodium falciparum is exported and localized in vesicular structures in the cytoplasm of infected erythrocytes in the brefeldin A-sensitive pathway. J Biol Chem 276: 15249–15255 He CY (2007) Golgi biogenesis in simple eukaryotes. Cell Microbiol 9: 566–572 He CY, Ho HH, Malsam H, Chalouni C, West CM, Ullu E, Toomre D, Warren G (2004) Golgi duplication in Trypanosoma brucei. J Cell Biol 165: 313–321 He CY, Shaw MK, Pletcher CH, Striepen B, Tilney LG, Roos DS (2001) A plastid segregation defect in the protozoan parasite Toxoplasma gondii. EMBO J 20: 330–339 Hehl AB, Marti M, Kohler P (2000) Stage-specific expression and targeting of cyst wall protein-green fluorescent protein chimeras in Giardia. Mol Biol Cell 11: 1789–1800 Hell AB, Marti M (2004) Mol Microbiol 53: 19–28 Hoppe HC, Joiner KA (2000) Cytoplasmic tail motifs mediate endoplasmic reticulum localization and export of transmembrane reporters in the protozoan parasite Toxoplasma gondii. Cell Microbiol 2: 569–578 Horazdovsky BF, Dewald DB, Emr SD (1995) Protein transport to the yeast vacuole. Curr Opin Cell Biol 7: 544–551 Hu K, Mann T, Striepen B, Beckers CJM, Roos DS, Murray JM (2001) Daughter cell assembly in the protozoan parasite Toxoplasma gondii. Mol Biol Cell 12: 143A Itin C, Kappeler F, Linstedt AD, Hauri HP (1995) A novel endocytosis signal related to the KKXX ER-retrieval signal. EMBO J 14: 2250–2256 Jeffries TR, Morgan GW, Field MC (2002) TbRAB18, a developmentally regulated Golgi GTPase from Trypanosoma bucei. Mol Biochem Parasitol 121: 63074 Joiner KA, Fuhrman SA, Miettinen HM, Kasper LH, Mellman I (1990) Toxoplasma gondii fusion competence of parasitophorous vacuoles in Fc-receptor transfected fibroblasts. Science 249: 641–646 Joiner KA, Roos DS (2002) Secretory traffic in the eukaryotic parasite Toxoplasma gondii: less is more. J Cell Biol 157: 557–563 Kabnick KS, Peattie DA (1991) Giardia – a missing link between Prokaryotes and Eukaryotes. Am Sci 79: 34–43 Kahn RA, Kern FG, Clark J, Gelmann EP, Rulka C (1991) Human ADP ribosylation factors – a functionally conserved family of Gtp-binding proteins. J Biol Chem 266: 2606–2614 Kallen KJ, Allan D (1995) Sphingomyelin synthesis in endosomal compartments. Trends Cell Biol 5: 149

Structure and function of the Golgi organelle in parasitic protists

*

669

Karsten V, Qi HL, Beckers CJM, Reddy A, Dubremetz JF, Webster P, Joiner KA (1998) The protozoan parasite Toxoplasma gondii targets proteins to dense granules and the vacuolar space using both conserved and unusual mechanisms. J Cell Biol 141: 1323–1333 Katinka MD, Duprat S, Cornillot E, Metenier G, Thomarat F, Prensier G, Barbe V, Peyretaillade E, Brottier P, Wincker P, Delbac F, El Alaoui H, Peyret P, Saurin W, Gouy M, Weissenbach J, Vivares CP (2001) Genome sequence and gene compaction of the eukaryote parasite Encephalitozoon cuniculi. Nature 414: 450–453 Keeling PJ (2001) Parasites go the full monty. Nature 414: 401–402 Keeling PJ (2003) Congruent evidence from alpha-tubulin and beta-tubulin gene phylogenies for a zygomycete origin of microsporidia. Fungal Genet Biol 38: 298–309 Keeling PJ, Fast NM (2002) Microsporidia: Biology and evolution of highly reduced intracellular parasites. Ann Rev Microbiol 56: 93–116 Keeling PJ, Luker MA, Palmer JD (2000) Evidence from beta-tubulin phylogeny that microsporidia evolved from within the fungi. Mol Biol Evol 17: 23–31 Keeling PJ, Slamovits CH (2004) Simplicity and complexity of microsporidian genomes. Eukaryot Cell 3: 1363–1369 Keohane EM, Weiss LM (1999) The structure, function, and composition of the microsporidian polar tube. In: Wittner M, Weiss LM (eds) The Microsporidia and Microsporidiosis. ASM press, Washington, D.C., pp. 196–224 Klemba M, Beatty W, Gluzman I, Goldberg DE (2004) Trafficking of plasmepsin II to the food vacuole of the malaria parasite Plasmodium falciparum. J Cell Biol 164: 47–56 Kohler S, Delwiche CF, Denny PW, Tilney LG, Webster P, Wilson RJM, Palmer JD, Roos DS (1997) A plastid of probable green algal origin in apicomplexan parasites. Science 275: 1485–1489 Ktistakis NT, Brown HA, Sternweis PC, Roth MG (1995) Phospholipase-D Is present on Golgi-enriched membranes and its activation by ADP-ribosylation factor is sensitive to Brefeldin-A. Proc Natl Acad Sci USA 92: 4952–4956 Kumar N, Syin CA, Carter R, Quakyi R, Miller LH (1988) Plasmodium falciparum gene encoding a protein similar to the 78-kDa rat glucose-regulated stress protein. Proc Natl Acad Sci USA 85: 6277–6281 La Greca N, Hibbs AR, CR, Foley M, Tilley L (1997) Identification of an endoplasmic reticulum resident protein with multiple EF-hand motifs in asexual stage of Plasmodium falciparum. Mol Biochem Parasitol 89: 283–293 Lanfredi-Rangel A, Attias M, Reiner DS, Gillin FD, De Souza W (2003) Fine structure of the biogenesis of Giardia lamblia encystation secretory vesicles. J Struct Biol 143: 153–163 Lauer SA, Rathod PK, Ghori N, Haldar K (1997) A membrane network for nutrient import in red cells infected with the malaria parasite. Science 276: 1122–1125 Levine ND, Corliss JO, Cox FEG, Deroux G, Grain J, Honigberg BM, Leedale GF, Loeblich AR, Lom J, Lynn D, Merinfeld EG, Page FC, Poljansky G, Sprague V, Vavra J, Wallace FG (1980) A newly revised classification of the Protozoa. J Protozool 27: 37–58 Liendo A, Stedman TT, Ngo HM, Chaturvedi S, Hoppe HC, Joiner KA (2001) Toxoplasma gondii ADP-ribosylation factor 1 mediates enhanced release of constitutively secreted dense granule proteins. J Biol Chem 276: 18272–18281 Lu L, Horstmann H, Ng C, Hong W (2001) Regulation of Golgi structure and function by ARF-like protein 1 (Arl1). J Cell Sci 114: 4543–4555 Lujan HD, Marotta A, Mowatt MR, Sciaky N, Lippincottschwartz J, Nash TE (1995a) Developmental induction of Golgi structure and function in the primitive eukaryote Giardia lamblia. J Biol Chem 270: 4612–4618 Lujan HD, Mowatt MR, Chen GZ, Nash TE (1995b) Isoprenylation of proteins in the protozoan Giardia lamblia. Mol Biochem Parasitol 72: 121–127

670

*

Y. Y. Sokolova and A. A. Mironov

Lujan HD, Mowatt MR, Wu JJ, Lu Y, Lees A, Chance MR, Nash TE (1995c) Purification of a variant-specific surface protein of Giardia lamblia and characterization of its metalbinding Properties. J Biol Chem 270: 13807–13813 Lujan HD, Touz MC (2003) Protein trafficking in Giardia lamblia. Cell Microbiol 5: 427–434 Mallabiabarrena A, Jimenez MA, Rico M, Alarcon B (1995) A tyrosine-containing motif mediates ER retention of Cd3-epsilon and adopts a helix-turn structure. EMBO J 14: 2257–2268 Manning-Cela R, Marquez C, Franco E, Talamas-Rohana P, Meza I (2003) BFA-sensitive and insensitive exocytic pathways in Entamoeba histolytica trophozoites: their relationship to pathogenesis. Cell Microbiol 5: 921–932 Marti M, Lit YJ, Schraner EM, Wild P, Kohler P, Hehl AB (2003a) The secretory apparatus of an ancient eukaryote: protein sorting to separate export pathways occurs before formation of transient Golgi-like compartments. Mol Biol Cell 14: 1433–1447 Marti M, Regos A, Li YJ, Schraner EM, Wild P, Muller N, Knopf LG, Hehl AB (2003b) An ancestral secretory apparatus in the protozoan parasite Giardia intestinalis. J Biol Chem 278: 24837–24848 McConville MJ, Ilgoutz SC, Teasdale RD, Foth BJ, Matthews A, Mullin KA, Gleeson PA (2002a) Targeting of the GRIP domain to the trans-Golgi network is conserved from protists to animals. Eur J Cell Biol 81: 485–495 McConville MJ, Mullin KA, Ilgoutz SC, Teasdale RD (2002b) Secretory pathway of trypanosomatid parasites. Microbiol Mol Biol Rev 66: 122–154 Meissner M, Reiss M, Viebig N, Carruthers VB, Toursel C, Tomavo S, Ajioka JW, Soldati D (2002b) A family of transmembrane microneme proteins of Toxoplasma gondii contain EGF-like domains and function as escorters. J Cell Sci 115: 563–574 Mironov AA, Beznoussenko GV, Polishchuk RS, Trucco A (2005) Intra-Golgi transport: a way to a new paradigm? Biochem Bioph Acta 1744: 340–350 Mironov AA, Beznoussenko GV, Trucco A, Lupetti P, Smith JD, Geerts WJC, Koster AJ, Burger KNJ, Martone ME, Deerinck TJ, Ellisman MH, Luini A (2003) ER-to-Golgi carriers arise through direct en bloc protrusion and multistage maturation of specialized ER exit domains. Dev Cell 5: 583–594 Mollenhauer HH, Morre DJ, Rowe LD (1990) Alteration of intracellular traffic by monensin – mechanism, specificity and relationship to toxicity. Biochem Biophys Acta 1031: 225–246 Mordue DG, Hakansson S, Niesman I, Sibley LD (1999) Toxoplasma gondii resides in a vacuole that avoids fusion with host cell endocytic and exocytic vesicular trafficking pathways. Exp Parasitol 92: 87–99 Morrissette NS, Sibley LD (2002 ) Cytoskeleton of apicomplexan parasites Microbiol Mol Biol Rev 66: 21–38 Ohashi M, Devries KJ, Frank R, Snoek G, Bankaitis V, Wirtz K, Huttner WB (1995) A role for phosphatidylinositol transfer protein in secretory vesicle formation. Nature 377: 544–547 Olliaro P, Castelli F (1997) Plasmodium falciparum: an electronmicroscopy study of caveolae and trafficking between the parasite and the extracellular medium. Int J Parasitol 27: 1007–1012 Overath P, Engstler M (2004) Endocytosis, membrane recycling and sorting of GPIanchored proteins: Trypanosoma brucei as a model system. Mol Microbiol 53: 735–744 Parodi AJ (1993) N-Glycosylation in trypanosomatid Protozoa. Glycobiology 3: 193–199 Patterson DJ (1994) Protozoa: evolution and systematics. In: Haussman K, Hulsman N (eds) Progress in Protozoology, Proceedings of the 9th International Congress on Protozoology. Fischer Verlag, Stuttgart, pp. 1–14 Pelham HRB (1995) Sorting and retrieval between the endoplasmic reticulum and Golgi apparatus. Curr Opin Cell Biol 7: 530–535

Structure and function of the Golgi organelle in parasitic protists

*

671

Pelletier L, Stern CA, Pypaert M, Sheff D, Ngo HM, Roper N, He CY, Hu K, Toomre D, Coppens I, Roos DS, Joiner KA, Warren G (2002) Golgi biogenesis in Toxoplasma gondii. Nature 418: 548–552 Philippe H, Lopez P, Brinkmann H, Budin K, Germot A, Laurent J, Moreira D, Muller M, Le Guyader H (2000) Early-branching or fast-evolving eukaryotes? An answer based on slowly evolving positions. Proc R Soc Biol Sci 267: 1213–1221 Plattner H (1993) Membrane traffic in Protozoa. JAI Press, Greenwich Plattner H, Lumpert CJ, Knoll G, Kissmehl R, Hohne B, Momayezi M, Glasalbrecht R (1991) Stimulus secretion coupling in Paramecium cells. Eur J Cell Biol 55: 3–16 Pleshinger J, Weidner E (1985) TThe microsporidian spore invasion tube. 4 . Discharge activation begins with PH-triggered Ca2 þ influx. J Cell Biol 100: 1834–1838 Preisig HR, Anderson OR, Corliss JO, Moestrup O, Powell MJ, Roberson RW, Wetherbee R (1994) Terminology and nomenclature of protist cell-surface structures. Protoplasma 181: 1–28 Price HP, Panethymitaki C, Goulding D, Smith DF (2005) Functional analysis of TbARL1, an N-myristoylated Golgi protein essential for viability in bloodstream trypanosomes. J Cell Sci 118: 831–841 Rambourg A, Clermont Y, Hermo L, Segretain D (1987) Tridimensional structure of the Golgi apparatus of nonciliated epithelial cells of the ductuli efferentes in rat—an electron microscope stereoscopic Study. Biol Cell 60: 103–116 Ribeiro KC, Monteiro-Leal LH, Benchimol M (2000) Contributions of the axostyle and flagella to closed mitosis in the protists Tritrichomonas foetus and Trichomonas vaginalis. J Eukaryot Microbiol 47: 481–492 Richards TA, Cavalier-Smith T (2005) Myosin domain evolution and the primary divergence of eukaryotes. Nature 436: 1113–1118 Roger AJ (1999) Reconstracting early events in eukaryotic evolution. Am Nat 154: S146–S163 Roger AJ, Morrison HG, Sogin ML (1999 ) Primary structure and phylogenetic relationships of a malate dehydrogenase gene from Giardia lamblia. J Mol Evol 48: 750–755 Roger AJ, Svard SG, Tovar J, Clark CG, Smith MW, Gillin FD, Sogin ML (1998) A mitochondrial-like chaperonin 60 gene in Giardia lamblia: evidence that diplomonads once harbored an endosymbiont related to the progenitor of mitochondria. Proc Natl Acad Sci USA 95: 229–234 Roos DS, Crawford MJ, Donald RGK, Kissinger JC, Klimczak LJ, Striepen B (1999) Origin, targeting, and function of the apicomplexan plastid. Curr Opin Microbiol 2: 426–432 Roos DS, Donald RGK, Morrissette NS, Moulton ALC (1994) Molecular tools for genetic dissection of the protozoan parasite Toxoplasma gondii. Methods Cell Biol 45: 27–63 Rossanese OW, Soderholm J, Bevis BJ, Sears IB, O’Connor J, Williamson EK, Glick BS (1999) Golgi structure correlates with transitional endoplasmic reticulum organization in Pichia pastoris and Saccharomyces cerevisiae. J Cell Biol 145: 69–81 Rothblatt J, Novick P, Stevens TH (1994) Guidebook to the secretory pathway. Oxford University Press, Oxford Rubotham J, Woods K, Garcia-Salcedo JA, Pays E, Nolan DP (2005) Characterization of two protein disulfide isomerases from the endocytic pathway of bloodstream forms of Trypanosoma brucei. J Biol Chem 280: 10410–10418 Ryan KA, Garraway LA, Descoteaux A, Turco SJ, Beverley SM (1993) Isolation of virulence genes directing surface glycosyl-phosphatidylinositol synthesis by functional complementation of Leishmania. Proc Natl Acad Sci USA 90: 8609–8613 Salama NR, Schekman RW (1995) The role of coat proteins in the biosynthesis of secretory proteins. Curr Opin Cell Biol 7: 536–543 Schekman R (1994) Translocation gets a push. Cell 78: 911–913

672

*

Y. Y. Sokolova and A. A. Mironov

Seemann J, Pypaert M, Taguchi T, Malsam J, Warren G (2002) Partitioning of the matrix fraction of the Golgi apparatus during mitosis in animal cells. Science 295: 848–851 Serafini T, Orci L, Amherdt M, Brunner M, Kahn RA, Rothman JE (1991) ADP-ribosylation factor is a subunit of the coat of Golgi-derived COP-coated vesicles—a novel role for a GTP-binding protein. Cell 67: 239–253 Serafini T, Stenbeck G, Brecht A, Lottspeich F, Orci L, Rothman JE, Wieland FT (1991) A coat subunit of Golgi-derived non-clathrin-coated vesicles with homology to the clathrin-coated vesicle coat protein beta-adaptin. Nature 349: 214–220 Shaw MK, Roos DS, Tilney LG (1998) Acidic compartments and rhoptry formation in Toxoplasma gondii. Parasitology 117: 435–443 Sibley LD, Weidner E, Krahenbuhl JL (1985) Phagosome acidification blocked by intracellular Toxoplasma gondii. Nature 315: 416–419 Simon S (1993) Translocation of proteins across the endoplasmic reticulum. Curr Opin Cell Biol 5: 581–588 Slamovits CH, Fast NM, Law JS, Keeling PJ (2004) Genome compaction and stability in microsporidian intracellular parasites. Curr Biol 14: 891–896 Smith SS, Pfluger SL, Hjort E, McArthur AG, Hager KM (2007) Molecular evolution of vesicle coat compartment betaCOP in Toxoplasma gondii. Mol Phylogenet Evol 44: 1284–1294 Snapp EL, Hegde RS, Francolini M, Lombardo F, Colombo S, Pedrazzini E, Borgese N, Lippincott-Schwartz J (2003) Formation of stacked ER cisternae by low affinity protein interactions. J Cell Biol 163: 257–269 Sogin ML (1991) Early evolution and the origin of eukaryotes. Curr Opin Gen Dev 1: 457–463 Sogin ML, Gunderson JH, Elwood HJ, Alonso RA, Peattie DA (1989) Phylogenetic meaning of the Kingdom concept—an unusual ribosomal-RNA from Giardia lamblia. Science 243: 75–77 Sogin ML, Silberman JD (1998) Evolution of the protists and protistan parasites from the perspective of molecular systematics. Int J Parasitol 28: 11–20 Sokolova Y, Snigirevskaya E, Morzhina E, Skarlato S, Mironov A, Komissarchik Y (2001) Visualization of early Golgi compartments at proliferate and sporogenic stages of a microsporidian Nosema grylli. J Eukaryot Microbiol: 86S-87S Soldati D, Lassen A, Dubremetz JF, Boothroyd JC (1998) Processing of Toxoplasma ROP1 protein in nascent rhoptries. Mol Biochem Parasitol 96: 37–48 Sollner T, Whitehart SW, Brunner M, Erdjumentbromage H, Geromanos S, Tempst P, Rothman JE (1993) SNAP receptors implicated in vesicle targeting and fusion. Nature 362: 318–324 Stamnes MA, Rothman JE (1993) The binding of Ap-1 clathrin adapter particles to Golgi membranes requires ADP-ribosylation factor, a small GTP-binding protein. Cell 73: 999–1005 Stechmann A, Cavalier-Smith T (2003) Phylogenetic analysis of eukaryotes using heatshock protein Hsp90. J Mol Evol 57: 408–419 Stedman TT, Sussmann AR, Joiner KA (2003) Toxoplasma gondii Rab6 mediates a retrograde pathway for sorting of constitutively secreted proteins to the Golgi complex. J Biol Chem 278: 5433–5443 Stefanic S, Palm D, Svard SG, Hehl AB (2006) Organelle proteomics reveals cargo maturation mechanisms associated with Golgi-like encystation vesicles in the early-diverged protozoan Giardia lamblia. J Biol Chem 281: 7595–7604 Stokkermans TJW, Schwartzman JD, Keenan K, Morrissette NS, Tilney LG, Roos DS (1996) Inhibition of Toxoplasma gondii replication by dinitroaniline herbicides. Exp Parasitol 84: 355–370 Striepen B, Soldati D, Garcia-Reguet N, Dubremetz JF, Roos DS (2001) Targeting of soluble proteins to the rhoptries and micronemes in Toxoplasma gondii. Mol Biochem Parasitol 113: 45–53

Structure and function of the Golgi organelle in parasitic protists

*

673

Striepen B, White MW, Li C, Guerini MN, Malik SB, Logsdon JM, Liu C, Abrahamsen MS (2002) Genetic complementation in apicomplexan parasites. Proc Natl Acad Sci USA 99: 6304–6309 Struck NS, S. d SD, Langer C, Marti M, Pearce AF, Cowman AF, Gilberger TW (2005) Redefining the Golgi complex in Plasmodium falciparum using the novel Golgi marker PfGRASP. J Cell Sci 118: 5603–5613 Tabe L, Krieg P, Strachan R, Jackson D, Wallis E, Colman A (1984) Segregation of mutant ovalbumins and ovalbumin-globin fusion proteins in Xenopus oocytes—identification of an ovalbumin signal sequence. J Mol Biol 180: 645–666 Takvorian PM, Cali A (1994) Enzyme histochemical identification of the Golgi apparatus in the microsporidian, Glugea stephani. J Eukaryot Microbiol. 41: 63S-64S Taraschi TF, O’Donnell M, Martinez S, Schneider T, Trelka D, Fowler VM, Tilley L, Moriyama Y (2003) Generation of an erythrocyte vesicle transport system by Plasmodium falciparum malaria parasites. Blood 102: 3420–3426 Taraschi TF, Trelka D, Martinez S, Schneider T, O’Donnell ME (2001) Vesicle-mediated trafficking of parasite proteins to the host cell cytosol and erythrocyte surface membrane in Plasmodium falciparum infected erythrocytes. Int J Parasitol 31: 1381–1391 Thomarat F, Vivares P, Gouy M (2004) Phylogenetic analysis of the complete genome sequence of Encephalitozoon cuniculi supports the fungal origin of microsporidia and reveals a high fequency of fast-eolving genes. J Mol Evol 59: 780–791 Tovar J, Leon-Avila G, Sanchez LB, Sutak R, Tachezy J, Van der Giezen M, Hernandez M, Muller M, Lucocq JM (2003) Mitochondrial remnant organelles of Giardia function in iron-sulphur protein maturation. Nature 426: 172–176 Traub LM, Ostrom JA, Kornfeld S (1993) Biochemical dissection of Ap-1 recruitment onto Golgi membranes. J Cell Biol 123: 561–573 Van Dooren GG, Marti M, Tonkin SJ, Stimmler LM, Cowman AF, McFadden GI (2005) Development of the endoplasmatic reticulum, mitochondrion and apicoplast during the asexual life stage of Plasmodium falciparum. Mol Microbiol 57: 405–419 Van Meer G (1989) Lipid traffic in animal cells. Ann Rev Cell Biol 5: 247– 275 Van Wye J, Ghori N, Webster P, Mitschler RR, Elmendorf HG, Haldar K (1996) Identification and localization of Rab6, separation of Rab6 from Erd2 and implications for an unstacked Golgi in Plasmodium falciparum. Mol Biochem Parasitol 83: 107–120 Vavra J, Larsson RJI (1999) Structure of the Microsporidia. In: Wittner M, Weiss LM (eds) The Microsporidia and Microsporidiosis. ASM press, Washington, D.C., pp. 7–84 Von Mollard GF, Nothwehr S F, Stevens TH (1997) The yeast v-SNARE Vti1p mediates two vesicle transport pathways through interactions with the t-SNAREs Sed5p and Pep12p. J Cell Biol 137: 1511–1524 Vossbrinck CR, Maddox JV, Friedman S, Debrunner-Vossbrinck BA, Woese CR (1987) Ribosomal-RNA sequence suggests Microsporidia are extremely ancient eukaryotes. Nature 326: 411–414 Vossbrinck CR, Woese CR (1986) Eukaryotic ribosomes that lack a 5.8s RNA. Nature 320: 287–288 Waller RF, Keeling PJ, Donald RGK, Striepen B, Handman E, Lang-Unnasch N, Cowman AF, Besra GS, Roos DS, McFadden GI (1998) Nuclear-encoded proteins target to the plastid in Toxoplasma gondii and Plasmodium falciparum. Proc Natl Acad Sci USA 95: 12352–12357 Waller RF, Reed MB, Cowman AF, McFadden GI (2000) Protein trafficking to the plastid of Plasmodium falciparum is via the secretory pathway. EMBO J 19: 1794–1802 Weidner E (1982) The microsporidian spore invasion tube. 3. Tube extrusion and assembly. J Cell Biol 93: 976–979 Weidner E, Byrd W (1982) The microsporidian spore invasion tube. 2. Role of calcium in the activation of invasion tube discharge. J Cell Biol 93: 970–975

674

*

Y. Y. Sokolova and A. A. Mironov

Weise F, Stierhof YD, Kuhn C, Wiese M, Overath P (2000) Distribution of GPI-anchored proteins in the protozoan parasite Leishmania, based on an improved ultrastructural description using high-pressure frozen cells. J Cell Sci 113: 4587–4603 Wetherbee R, Andersen RA, Pickett-Heaps J (1994) The protistan cell surface. SpringerVerlag, Vienna Williams BAP, Hirt RP, Lucocq JM, Embley TM (2002) A mitochondrial remnant in the microsporidian Trachipleistophora hominis. Nature 418: 865–869 Williams BAP, Slamovits CH, Patron NJ, Fast NM, Keeling PJ (2005) A high frequency of overlapping gene expression in compacted eukaryotic genomes. Proc Natl Acad Sci USA 102: 10936–10941 Woese CR, Fox GE (1977) Concept of cellular evolution. J Mol Evol 10: 1–6 Yung S, Unnasch TR, Lang-Unnasch N (2001) Analysis of apicoplast targeting and transit peptide processing in Toxoplasma gondii by deletional and insertional mutagenesis. Mol Biochem Parasitol 118: 11–21

Evolution of the Golgi complex

*

675

Evolution of the Golgi complex  spa  r Je kely Ga

Introduction The Golgi complex evolved very early during the origin of the eukaryotic cell. It is present in every eukaryote living today, including parasitic lineages formerly considered as Golgi-lacking (Bredeston et al. 2005; Dacks et al. 2003; Dacks and Doolittle 2002; Marti et al. 2003), and thus was present in the last common eukaryotic ancestor (LCEA). Its origin traces back to a protoeukaryotic stage and was a significant step in the prokaryote–eukaryote transition, one of the major transitions in the history of the biosphere ry 1995). The Golgi evolved into a compartment (Maynard Smith and Szathma with a central role in the modification and sorting of secreted and membrane proteins as well as proteins destined for other intracellular compartments (e.g. lysosomes). To understand the evolutionary history of the Golgi complex we have to analyse the history of its individual constituents. The availability of dozens of eukaryotic full genome sequences coupled to an understanding of eukaryote phylogeny allow us to reconstruct the ancestral state of the Golgi complex in the LCEA and to gain insights into the diversification of Golgi functions in different lineages.

Comparative genomics and Golgi evolution in crown eukaryotes The phylogenetic context of comparative genomic reconstructions The availability of full genome sequences from dozens of eukaryotic organisms allows us to reconstruct the gene inventory of the LCEA and the history of gene losses and duplications in descendants of the LCEA. These studies can give us an idea of the timing of events during the history of eukaryotic gene families in the eukaryotic crown group (i.e. the LCEA and all its descendants). The reconstruction is simple if we deal with genes that can be found in all eukaryotic genome sequences. In such cases we can be sure that this gene or gene family was also present in the LCEA. The inference of the presence of a gene in the LCEA can be more difficult if a gene is not present in all eukaryotic genomes. In such cases we have to know the phylogeny of eukaryotes and the position of the root of the eukaryotic tree. According to the current consensus on eukaryote phylogeny there are six major monophyletic eukaryotic superphyla (Adl et al. 2005; Keeling et al.

676

*

G. J ekely

Figure 1. The six major eukaryotic superphyla and the likely position of the root of the eukaryotic tree. Representative organisms for each superphylum are shown.

2005; Simpson and Roger 2004) that encompass most or all of the known eukaryotic diversity (Fig. 1). There is no widely held consensus regarding the rooting of the eukaryotic tree (Brinkmann and Philippe 2007), but the most prominent one, based on rare genomic rearrangements (Richards and Cavalier-Smith 2005; Stechmann and Cavalier-Smith 2002), places the root between unikonts (opisthokonts and amoebozoa, ancestrally with one cilium) and bikonts (plants, chromalveolates and excavates, ancestrally with two cilia). Assuming that this rooting is correct, we can infer the ancestral presence of a gene in the LCEA if we can identify its orthologues in both unikont and bikont taxa. However, if the phyletic distribution is not universal for a given gene, we always have to keep in mind that the reconstructions rest on the accepted rooting of the eukaryotic tree. There are other caveats that one has to keep in mind, such as the possibility of horizontal gene transfer (HGT) between distinct lineages. HGT is expected to have occurred for example during the evolution of several major algal groups (e.g. chromophyte algae) that derived from secondary endosymbiotic events between a photosynthetic eukaryotic alga and a non-photosynthetic eukaryotic host (Archibald et al. 2003; Deane et al. 2000; Gould et al. 2006).

Evolution of vesicle coating and tethering complexes The broad phyletic distribution of several proteins and protein complexes indispensable for Golgi function indicates their ancestral presence in the LCEA. These include all major vesicle coating complexes and their regulators as well as the Golgi tethering complexes and the Golgi fusion machinery.

Evolution of the Golgi complex

*

677

The major vesicle coating complexes, COPI, COPII, and clathrin, are all ancient, widely distributed, and were most likely present in the LCEA (Dacks and Doolittle 2001; Dacks and Field 2007; Field et al. 2007; Schledzewski et al. 1999). Besides, three subunits of the retromer complex (Vps26, Vps29, and Vps35), are also ancient (Dacks et al. 2003; Nakada-Tsukui et al. 2005; Oliviusson et al. 2006), as well as the four adaptor protein (AP) complexes (Boehm and Bonifacino 2001; Dacks and Field 2007; Field et al. 2007). In contrast, GGA proteins that contain a domain homologous to the ear domain of g-adaptin are evolutionarily younger, being only present in fungi and animals (Boehm and Bonifacino 2001; Field et al. 2007). The Golgi tethering complexes that contribute to the specificity of vesicle fusion events are also widespread and thus ancient. These include the TRAPPI complex, two subunits (Trs120p and Trs130p) of TRAPPII, the Dsl1 complex, the Vps52p/53p/54 subunits of the GARP complex, and the COG complex (Koumandou et al. 2007). Besides, the four SNARE families mediating membrane fusion events (including Golgi SNAREs), all major small GTPase families (Sar1, Arf, SRb, Rab, Ran, Ras, Rho/Rac) as well as regulators of their nucleotide cycle (GAPs and GEFs) were also present in the LCEA (Dacks and Doolittle 2001, kely 2003; Mouratou et al. 2005). 2002; Je Starting with a complex Golgi in the LCEA, an important factor that shaped the diversity of the organelle is gene loss and the simplification of trafficking pathways and compartment morphology. This trend can be observed in several independently evolving parasitic lineages. Microsporidia, a divergent parasitic fungal lineage, reduced or lost several gene families involved in membrane traffic (reduced Rab and SNARE families, no clathrin). As a consequence, they lack vesicular carriers altogether and only possess a tubular Golgi (Beznoussenko et al. 2007; Beznoussenko and Mironov 2002; Mironov et al. 2007). The intestinal parasite Giardia intestinalis also dramatically reduced its secretory apparatus. It lost several of the trafficking complexes (no COG, GARP, TRAPPII, reduced TRAPPI, and Dsl1 complexes) and lost Golgi stacking. The reduction of the trafficking machinery correlates with the extreme n et al. simplification of Golgi morphology also seen in this organism (Luja 1995; Marti et al. 2003). Quite contrary to the drastic reduction of Golgi trafficking and morphology in parasitic lineages Golgi complexity increased in some lineages, most prominently in parallel with the advent of multicellularity in animals and plants. The independent origin of multicellularity in both land plants and animals was paralleled with an expansion of the SNARE proteins of the secretory pathway (Dacks and Doolittle 2002; Sanderfoot 2007). The Rab small GTPase family also expanded from 5–20 Rab family members in protists to 25–60 members in multicellular eukaryotes independently in animals and plants (Lal et al. 2005). Interestingly, some amoeboid protists also largely extended their Rab repertoire. The excavate Trichomonas vaginalis, a sexually transmitted pathogen has 65 Rab proteins. A similar independent diversifi-

678

*

G. J ekely

cation of Rab families is observed in the enteric protozoan parasite Entamoeba histolytica (Saito-Nakano et al. 2005) and the soil amoeba Dictyostelium discoideum (Eichinger et al. 2005; Lal et al. 2005).

Evolution of ER and Golgi glycosylation The Golgi is the major site of protein glycosylation in eukaryotes. The known glycosyltransferases (GT) sequences from both eukaryotes and prokaryotes have been classified into 90 distinct GT families (GT1–GT90), each having up to several thousand members (Coutinho et al. 2003a) (http://afmb.cnrs-mrs.fr/ CAZY/index.html). Despite the close sequence similarity within these families, the enzymes belonging to one GT family can show up to 14 distinct experimentally demonstrated activities (Coutinho et al. 2003a). More sensitive sequence similarity searches using PSI-BLAST and structural comparisons uncovered a deep evolutionary relationship between several distinct GT families. About 75% of all GTs can be included into one of only three monophyletic GT superfamilies (Liu and Mushegian 2003) (GT-A, GT-B, GT-C), each including enzymes with dozens of distinct activities. Protein glycosylation is ubiquitous in eukaryotes (Becker and Melkonian 1996). Several GT families trace back to the LCEA and many have prokaryotic ancestors. The glycosylation machineries also diversified considerably in eukaryotes ranging from the extremes of complete loss of N-glycosylation (microsporidia) to gene family expansion into the hundreds (plants, animals). The distribution of individual Golgi GTs is very patchy and, assuming a unikont–bikont rooting, is replete with independent losses. Plants, animals, and choanoflagellates retained a largely overlapping set of GT families while fungi and several chromalveolate, excavate and amoebozoan protists lost them or modified them beyond recognition by non-iterated BLAST searches (Table 1). Mucin-type O-glycosylation that generates GalNAc-a-Ser/Thr O-glycans (Wilson 2002) is characteristic of animals (Spiro 2002). The polypeptide (pp) aGalNAc transferases (GT27 family) that initiate O-glycosylation are widespread in animals but can also be found in the choanoflagellate, Monosiga brevicollis, and in two apicomplexan animal parasites, Toxoplasma gondii (Stwora-Wojczyk et al. 2004) and Cryptosporidium sp. (Templeton et al. 2004), both belonging to the chromalveolates (Table 1). The isolated presence of aGalNAcT in these parasites suggests that they may have acquired it by HGT from their animal host (Templeton et al. 2004). Two other protists, Trypanosoma cruzi and Dictyostelium discoideum, have O-glycans that are attached as GlcNAc-a-Ser/Thr instead of GalNAc-a-Ser/Thr (Jung et al. 1998; Previato et al. 1998). These O-linked glycans are synthesized by pp aGlcNAc transferases that prefer UDP-GlcNAc over UDP-GalNAc as donor substrate (Ercan and West 2005), and that are distantly related to the animal pp aGalNAcT (West et al. 2004). It is likely that the common ancestor of these two types of enzymes traces back to the LCEA and it had GlcNAc transferase activity. GalNAc transferase activity presumably evolved later, sometime before the animal–choanoflagellate common ancestor.

Evolution of the Golgi complex

*

679

Table 1. Phyletic distribution of ER and Golgi glycosyltransferase families in five eukaryotic superphyla

The presence/absence of the various GT families was established either by reciprocal BLAST searches at the NCBI (e-value cut-off 0.02), species-specific databases (Paramecium: http://paramecium.cgm.cnrs-gif.fr/db/index; Tetrahymena: http://tigrblast.tigr. org/er-blast/index.cgi?project¼ttg; Ostreococcus, Monosiga: http://genome.jgi-psf. org) or is based on the CAZy database (http://www.cazy.org/). The upper part of the table shows the core N-glycan structure found in the respective species. The species name abbreviations used are: Gi – Giardia intestinalis; Lm – Leishmania major; Tv –Trichomonas vaginalis; Tb – Trypanosoma brucei; Pf – Plasmodium falciparum; Pt – Paramecium tetraurelia; Tg – Toxoplasma gondii; Tt – Tetrahymena thermophila; Cr – Chlamydomonas reinhardtii; At – Arabidopsis thaliana; Ol – Ostreococcus lucimarinus; Eh – Entamoeba histolytica; Dd – Dictyostelium discoideum; Ec – Encephalitozoon cuniculi; Sc – Saccharomyces cerevisiae; Hs – Homo sapiens; Nv – Nematostella vectensis; Mb – Monosiga brevicollis.

O-mannosylation forming a Man-a-Ser/Thr linkage is found in Opisthokonts (Table 1). It is initiated in the ER by protein O-mannosyltransferases (POMT1, POMT2 in human, PMT1-7 in yeast, GT39 family). In yeast O-mannosylated proteins are further processed in the Golgi by mannosyltransferases (fungal specific GT15 family, e.g. yeast MNT1) (Ernst and Prill 2001). O-mannosylation is a sorting determinant for cell surface delivery in yeast (Proszynski et al. 2004) and is also required for cell wall stability and viability (Goto 2007). Several POMT/PMT homologues can also be found in the anaerobic parasite, Trichomonas vaginalis a member of the Excavata. The isolated presence of these enzymes in this parasite may be due to HGT from its host. If it is the case, the pathway evolved in opisthokonts.

680

*

G. J ekely

O-fucosylation forming Fuc-a-Ser/Thr linkage on EGF repeats of various proteins has been described in animals (Harris and Spellman 1993). In Drosophila this modification occurs in the ER (Luo and Haltiwanger 2005). O-fucosyltransferases catalyzing the reaction can only be found in animals, including the cnidarian Nematostella vectensis, as well as choanoflagellates (Table 1). Proteoglycan biosynthesis is initiated by the formation of a Xyl-b-Ser bond by protein-O-xylosyltransferase (GT14). This enzyme is localised to the late ER and early Golgi compartments (Vertel et al. 1993). Xylosyltransferases can be found in animals, choanoflagellates, in Dictyostelium and plants, indicating their ancestral presence in the LCEA (Reiter 2002). One GT family (GT34) is only present in plants and fungi and contains plant Golgi xylosyltransferases (e.g. ATX1) that are involved in the biosynthesis of xyloglucan, a major hemicellulose in the cell wall of land plants (Faik et al. 2002; Reiter 2002), by forming a-1,6-linkages on a b-1,4-glycan backbone. The related fungal Golgi mannosyltransferases (yeast Mnn10p, Mnn11p) are involved in the biosynthesis of mannan, an N-glycan structure built of a backbone of about 50 mannoses and short side branches (Jungmann et al. 1999; Munro 2001). Closely related enzymes are lacking from animals. N-glycosylation with GlcNAc-b-Asn bonds is the most common form of protein modification. It is phyletically more widespread in eukaryotes and its core pathway is related to the N-glycosylation system creating Glc-b-Asn and GalNAc-b-Asn bonds in archaebacteria (Fig. 2). This core N-glycan synthesis pathway has already been present in the LCEA, as demonstrated by its wide phyletic distribution (Samuelson et al. 2005). Similarly to the vesicle trafficking complexes, gene loss and pathway simplification represent an important trend in glycan evolution. These secondary simplifications involve the loss of Golgi GT families and different degrees of shortening of the core N-glycan precursor due to the loss of the ER monosaccharyltransferases (Kelleher and Gilmore 2006; Samuelson et al. 2005). Fungi lost several ancestral GT families and rely on GT families that are either fungal specific (GT15), or are shared only by plants but not animals (GT34, GT71). This is reflected in the relative simplicity of the N-glycans of yeast that occur in two basic forms only, both generated by attaching mannose residues to the trimmed core N-glycan (Munro 2001). However, yeasts have retained the complete ER core N-glycan synthesis machinery, and the trimmed N-glycan core serves as the substrate for the Golgi mannosyltransferases. The microsporidian Encephalitozoon cuniculi is much more reduced, as it completely lacks the N-glycosylation machinery (including ALG7 and STT3) and the N-glycan processing apparatus (GTs and lectins) (Samuelson et al. 2005). Microsporidia only have O-glycosylation, and retained an ER dolichylphosphate-mannose-protein mannosyltransferase (GT39).

Evolution of the Golgi complex

*

681

Figure 2. Synthesis of the core glycan structure of N-linked glycoproteins in the ER. Synthesis of the dolichol-PP-GlcNAc2Man5 precursor starts on the cytoplasmic side of the ER. Dolichol-PPGlcNAc2Man5 is flipped into the ER lumen by Rft1. The subsequent action of 6-membrane monosaccharyltransferases (GT families 22, 57–59) makes dolichol-PP-GlcNAc2Man9Glc3. The oligosaccharide is transferred from the lipid carrier to the nascent polypeptides in an N-X-T/Smotif-dependent manner by the oligosaccharyltransferase (OST), a heterooligomeric transmembrane protein complex with one catalytic subunit, the 13-transmembrane STT3 protein (Helenius and Aebi 2004; Kelleher and Gilmore 2006).

G. intestinalis and the apicomplexan Plasmodium falciparum, the causative agent of malaria, also almost completely lost the N-glycosylation machinery. Both organisms have lost all ALG GTs except ALG7 and also lost the flippase that flips the half made precursor from the cytoplasmic to the lumenal side of the ER. They retained the STT3 subunit of the OST, indicating that the short dolichol-PP-GlcNAc2 precursor that is synthesized is transferred to proteins. Given the sequential nature of glycan synthesis, the severe truncation of the ER core glycan structure suggests that no further modification of N-glycans takes place in the Golgi of these organisms. Consistent with this, the enzymes for further N-glycan processing and the N-glycan binding lectins (ERGIC-53, calreticulin, calnexin) are missing from the genome of G. intestinalis and P. falciparum.

682

*

G. J ekely

At the other end of the spectrum lay vertebrates and land plants that greatly expanded some of their GT families. The core synthesis machinery in the ER is unaltered, but several glycosyltransferases evolved that decorate the core oligosaccharide in the Golgi complex. These expanded GT families in plants synthesize the complex polysaccharides of their cell walls. In Arabidopsis about 50 enzymes have been estimated to be involved in pectin synthesis alone (Coutinho et al. 2003b). Humans have 29 ppGaNTases forming the mucin-type GalNAc-a-Ser/Thr, many of which show tissue specific expression patterns (Ten Hagen et al. 2003). The elucidation of the role of these tissue specific activities in multicellular development and the generation of cell type diversity is an important challenge for the future.

Evolution of cargo sorting Some of the trafficking signals and receptors operating in the secretory pathway are ancient and their presence can be inferred in the LCEA. The KDEL receptor for retrieval of ER proteins from the Golgi is present in every eukaryote. Dileucin motifs that regulate the targeting of membrane proteins from the Golgi to the lysosomes in animals (Bonifacino and Traub 2003; Letourneur and Klausner 1992) are also conserved in Trypanosomes (Allen et al. 2007), indicating their potential ancient presence in eukaryotes. The C-terminal tyrosine-based sorting signals can be found in animals, plants and protists, indicating their ancestral nature. Some lectins recognising glycan structures in the ER and Golgi are also ancient. This includes the mannosebinding lectin ERGIC-53 found in the ER–Golgi intermediate compartment (Fiedler and Simons 1994). The retrieval of membrane proteins to the ER by KKxx-COOH motifs (Becker and Melkonian 1996; Teasdale and Jackson 1996) is also ancient. Such dileucine motif-based sorting signals are also present in the OSTcomplex in mammalian cells. It is not the catalytic OST subunit but three other subunits that are responsible for ER retrieval. One of these subunits (OST48) uses a dilysine signal formed by two lysines at positions  3 and  5 (Fu and Kreibich 2000). This motif is conserved in Plasmodium OST48 and a tandem dilysin motif can be found in the OST48 of fungi and the free-living ciliate, Tetrahymena thermophila, indicating the ancestral presence of this retrieval mechanism for TM proteins in the LCEA (Fig. 3). The sorting of acid hydrolases from the TGN to the vacuole/lysosome seems to be less conserved. The sorting receptors in animals and plants are nonhomologous, and employ different mechanisms of cargo recognition (Masclaux et al. 2005). In plants, the vacuolar targeting signals are short motives in the cargo peptide sequence. The most common is the NPIR motif usually located at the N terminus of cargo proteins that is recognised by members of the BP-80 family (Robinson et al. 2005) with no relatives in animals and fungi. In mammals, lysosomal sorting of hydrolases depends on mannose-6-phosphate residues in their oligosaccharides recognised by mannose-6-phosphate receptors (MPR). The evolution of this recognition system

Evolution of the Golgi complex

*

683

Figure 3. Conservation of the dilysine-based sorting motif in the very C-terminus of OST48. Alignments of the very C-termini are shown. GenBank IDs: Homo sapiens NP_005207, Tribolium castaneum XP_969003, Nematostella vectensis XP_001632938, Saccharomyces cerevisiae EDN62965, Schizosaccharomyces pombe NP_588153, Dictyostelium discoideum XP_629963, Entamoeba histolytica XP_648945, Trichomonas vaginalis XP_001316950, Plasmodium falciparum XP_001352067, Plasmodium chabaudi XP_740758, Paramecium tetraurelia XP_001425674, Tetrahymena thermophila XP_001013892, Cryptosporidium parvum XP_626385, Chlamydomonas reinhardtii XP_001699044, Oryza sativa NP_001059165, Arabidopsis thaliana AAK59450.

is relatively recent. It is absent from Drosophila (Dennes et al. 2005) and other insects, but its absence is due to secondary loss. MPR homologues showing sequence similarity across their entire length to both vertebrate cationdependent (CD-MPR, about 270 amino acids) and cation-independent (CI-MPR, about 2,300 amino acids) MPRs can be found in the cnidarian sea anemone, Nematostella vectensis (Putnam et al. 2007), and also in the choanoflagellate, Monosiga brevicollis. Choanoflagellates are the protist sister group to animals (King 2004; Lang et al. 2002), and the fact that they contain MPRs indicates that this sorting pathway evolved before the animal– choanoflagellate last common ancestor. Mannose-6-phosphate-dependent sorting may also function in yeast, where an MPR-like receptor has been identified with a role in lysosomal enzyme sorting. However, whether it is a bona fide cargo receptor is not known (Whyte and Munro 2001). In Dictyostelium discoideum, an amoebozoan protist, some lysosomal enzymes have N-linked oligosaccharides containing mannose-6-phosphate in a phosphomethyldiester linkage (Man-6-P-OCH3) that can also bind to mammalian MPR (Souza et al. 1997). There is, however, no clear MPR homolog in Dictyostelium. Mannose-6-phosphate-dependent lysosomal sorting therefore seems to have evolved somewhere in the Opistokhont lineage, before the animal–choanoflagellate split. The mechanisms of cytoplasmic sorting of MPR homologues is also likely conserved in Opisthokonts. MPRs in mammals are sorted by a C-terminal DXXLL-type dileucine motif that is flanked by an acidic cluster of residues and

684

*

G. J ekely

Figure 4. Conservation of sorting motifs in MPRs in Opisthokonta. The acidic cluster and DXXLL-type C-terminal sorting motif is partially conserved in the choanoflagellate (Monosiga brevicollis) and chidarian (Nematostella vectensis) cation-dependent (A) and cation-independent (B) MPRs. Alignments of the very C-termini are shown. GenBank IDs: CD-MPRs Xenopus laevis AAI57733, Homo sapiens BAD96512, Strongylocentrotus purpuratus XP_001201595, Takifugu rubripes AAZ08350, Nematostella vectensis EDO46637, Monosiga brevicollis EDQ87334; CI-MPRs Danio rerio NP_001034716, Homo sapiens AAF16870, Gallus gallus NP_990301, Monosiga brevicollis EDQ91968.

is recognised by GGA proteins, a class of Arf-dependent clathrin adaptors (see above). The acidic cluster and one of the leucines is also conserved in choanoflagellate, cnidarian, and sea urchin MPRs, indicating the probable conservation of the Arf, GGA, and clathrin-dependent sorting mechanism in these organisms (Fig. 4). The clathrin-dependent sorting of receptors for acid hydrolases is more conserved than the receptors themselves and their lumenal recognition mechanisms. Beside MPRs the plant BP-80-like receptors, as well as yeast Vps10p, an unrelated receptor involved in carboxypeptidase sorting, are also sorted in a clathrin-dependent manner. Clathrin coat assembly also generally involves an interaction between tyrosine or dileucine sorting motifs in the cytoplasmic tail of the receptors and adaptor complexes, as well as GGA proteins (the latter only in fungi and animals). In plants vacuolar sorting mechanisms further diversified since there are two distinct vacuolar compartments, a storage vacuole and a lytic vacuole. Both vacuole types employ a distinct targeting mechanism (Robinson et al. 2005).

Early evolution of the Golgi in the prokaryote–eukaryote transition Common origin of eukaryotic and archaebacterial N-glycosylation systems The membrane-bound machinery that synthesizes and attaches the core oligosaccharide of glycoproteins to nascent polypeptide chains in the endoplasmic reticulum (ER) of eukaryotes (see above) is related to the N-glycosyla-

Evolution of the Golgi complex

*

685

tion system found in archaebacteria (Eichler and Adams 2005). Archaebacteria have the catalytic STT3 subunit of the OST (COG1287) that catalyzes oligosaccharide transfer onto proteins in an N-X-T/S-motif-dependent manner (Igura et al. 2007). They also use a dolichol-phosphate or pyrophosphate carrier (Kelleher and Gilmore 2006; Lechner et al. 1985; Wieland et al. 1985) coupled to a large diversity of glycan structures not found in eukaryotes. Archaebacteria contain the closest prokaryotic relatives of the ALG7 glycosyltransferase (UDP-GlcNAc:dolichol-phosphate GlcNAc-1-phosphate transferase), the first enzyme in the synthesis of dolichol-PP-linked glycans (Samuelson et al. 2005). Besides, the Archaeoglobus fulgidus genome contains two gene clusters, putatively involved in protein glycosylation, containing STT3-related genes, several glycosyltransferases and transporters that are presumably involved in the flipping of the dolichol-P/PP-linked oligosaccharides (Eichler and Adams 2005). The incontestable homology between the archaebacterial and eukaryotic N-glycosylation systems and their lack from most eubacteria (except for Campylobacter sp. that acquired the pathway via HGT (Samuelson et al. 2005)) indicate that a similar system was present in the last common ancestor of archaebacteria and eukaryotes. In archaebacteria the first steps in the synthesis of the dolichol-P/PP-linked oligosaccharide occur in the cytoplasm. The half-made precursor is subsequently flipped to the extracellular face. The transfer of the oligosaccharide onto proteins occurs at the cell surface as the nascent polypeptide chain co-translationally translocates across the PM through the translocon (proteinconducting channel) (Eichler and Adams 2005). In eukaryotes oligosaccharide transfer occurs in the ER lumen that corresponds to the extracellular side of archaebacteria. This strongly suggests that the topology of eukaryotic endomembranes originated via a single inward budding step from a precursor state similar to the one in modern archaebacteria.

Origin of secretory endomembranes by invagination Secretory endomembranes probably evolved very early during the history of eukaryotes (Becker and Melkonian 1996; Cavalier-Smith 2002; Devos et al. kely 2003, 2007). The primitive endomembranes of the proto-eukary2004; Je ote must have initially been continuous with the PM and formed a network of invaginations or tubules. The sorting of membrane proteins could already have evolved at this stage as the translocon, the glycosylation machinery and other proteins were selectively enriched in these invaginations (Fig. 5). This system could have been the evolutionary precursor of the ancient and universal ER retention and retrieval systems for soluble (K/HDEL-COOH) and TM proteins (KKxx-COOH) (Becker and Melkonian 1996; Teasdale and Jackson 1996). Such mechanism could have retained and restricted the activity of the translocon and the OST complex in the primitive ER generating a membrane network with a distinct composition. ER retention of the originally monomeric STT3 could have evolved by the recruitment of additional subunits with the conserved KKxx motifs (see above).

686

*

G. J ekely

Figure 5. Hypothetical stages in the evolution of eukaryotic secretory endomembranes in the proto-eukaryotic (stem eukaryote) lineage.

The maintenance of this tubular reticulum was likely dependent on the developing eukaryotic cytoskeleton (Cavalier-Smith 2002). Cytoskeletal motors may also have had an active role in generating and maintaining membrane tubules as happens in modern eukaryotes. The engulfment and maintenance of the first secretory endomembranes was probably also facilitated by the common ancestor of the small GTPases Arf/Sar1/SRb, proteins with indispensable roles in ER and Golgi transport and co-translational kely 2003). A tubular network could targeting of ribosomes to the ER (Je have been able to sustain membrane transport functions even without vesicle budding (an analogous tubular endomembrane system evolved secondarily in microsporidia). Cargo sorting could have already evolved in such tubules and could have been under the control of coating complexes that bound and curved membranes without inducing vesicle budding (Mironov et al. 2007). At a later stage the evolving reticulum of secretory endomembranes topologically segregated from the PM. Before topological separation occurred the membrane re-fusion machinery had to evolve to maintain transport functions. The evolution of membrane fusion probably coincided with the origin of the SNARE family of membrane fusion proteins (Dacks and Doolittle 2002). The evolution of membrane fission evolved to separate membrane tubules, generate vesicles and to allow the equal distribution of the fragmented endomembranes during cell division. There are diverse mechanisms of membrane fission that operate in the Golgi. Most prominently it is the coat complexes that can pinch off vesicles from the growing membrane tubules. Another mechanism operating in the TGN relies on the localised synthesis of dyacylglycerol (DAG) that induces local curvature of

Evolution of the Golgi complex

*

687

the membrane. DAG can be synthesized by several pathways, most of which are ancient.

Evolution of Golgi identity In eukaryotes endomembrane compartments have distinct and well-defined contents (cargo) that have to be delivered from one membrane compartment to the other or to the extracellular surface. The membrane compartments have to be tagged to be identifiable for directional transport (via attachment of molecular motors) and for fusion with the correct target membrane. This necessitates a molecular barcoding system where a given molecular label on the cytoplasmic side corresponds to a given cargo composition inside the membrane carriers. This correlation between the outside and inside components gives every compartment a unique identity. The evolution of the eukaryotic cell is partly a history of the diversification of transport compartments with a unique identity. Most of the gene families and complexes that determine compartment identity evolved via paralogous gene duplications before the LCEA. This is evident from the homology between coat complexes (clathrin, adaptin, COPI, and most likely also COPII), syntaxins, syntaxin-binding proteins families (SM proteins), and small GTPases (but not between distinct tethering complexes) kely 2003; Je kely and (Dacks and Doolittle 2004; Devos et al. 2004, 2006; Je Arendt 2006; Koumandou et al. 2007; Schledzewski et al. 1999). The increase in the number of distinct compartments can thus be interpreted as a process of duplication events made possible by the duplication and divergence of the genes regulating membrane identity (Becker and Melkonian 1996; Cavalierkely 2003). The distinct Smith 2002; Dacks and Field 2007; Devos et al. 2004; Je identity of the ER and the Golgi complex probably evolved by the subdivision of the first secretory endomembrane domain by the evolution of distinct, but paralogous, molecular tags (Sar1–Arf, COPII–COPI). This subdivision could have occurred parallel with the increase in the complexity of glycosylation steps and the increase in cargo destinations (e.g. PM, phagosome, spore wall, cilium). Secretory membrane domains of distinct composition were established by the evolution of dynamic forward sorting of cargo into the trans compartment (towards the PM) and reverse sorting of the enzymes and fusion proteins into the cis compartment. The cargo and the corresponding molecular tags could have been segregated into cis and a trans compartments as the steady state of a dynamic process involving constant forward and reverse sorting of specific components. Besides the distribution of cargo other factors contribute to the generation of compartment identity. Distinct compartments are also labelled by their unique lipid composition. The Golgi is marked by the combined presence of PtdIns(4)P and Arf (Munro 2005). Such a system of cargo-independent identity-determination probably evolved after the cargo-dependent mechanism (but before the LCEA) and evolved sharper delineation of membrane domains and compartments.

688

*

G. J ekely

References Adl SM, Simpson AG, Farmer MA, Andersen RA, Anderson OR, Barta JR, Bowser SS, Brugerolle G, Fensome RA, Fredericq S et al (2005) The new higher level classification of eukaryotes with emphasis on the taxonomy of protists. J Eukaryot Microbiol 52: 399–451 Allen CL, Liao D, Chung WL, Field MC (2007) Dileucine signal-dependent and AP-1independent targeting of a lysosomal glycoprotein in Trypanosoma brucei. Mol Biochem Parasitol 156: 175–190 Archibald JM, Rogers MB, Toop M, Ishida K, Keeling PJ (2003) Lateral gene transfer and the evolution of plastid-targeted proteins in the secondary plastid-containing alga Bigelowiella natans. Proc Natl Acad Sci USA 100: 7678–7683 Becker B, Melkonian M (1996) The secretory pathway of protists: spatial and functional organization and evolution. Microbiol Rev 60: 697–721 Beznoussenko GV, Dolgikh VV, Seliverstova EV, Semenov PB, Tokarev YS, Trucco A, Micaroni M, Di Giandomenico D, Auinger P, Senderskiy IV et al (2007) Analogs of the Golgi complex in microsporidia: structure and avesicular mechanisms of function. J Cell Sci 120: 1288–1298 Beznoussenko GV, Mironov AA (2002) Models of intracellular transport and evolution of the Golgi complex. Anat Rec 268: 226–238 Boehm M, Bonifacino JS (2001) Adaptins: the final recount. Mol Biol Cell 12: 2907–2920 Bonifacino JS, Traub LM (2003) Signals for sorting of transmembrane proteins to endosomes and lysosomes. Annu Rev Biochem 72: 395–447 Bredeston LM, Caffaro CE, Samuelson J, Hirschberg CB (2005) Golgi and endoplasmic reticulum functions take place in different subcellular compartments of Entamoeba histolytica. J Biol Chem 280: 32168–32176 Brinkmann H, Philippe H (2007) The diversity of eukaryotes and the root of the eukaryotic tree. Adv Exp Med Biol 607: 20–37 Cavalier-Smith T (2002) The phagotrophic origin of eukaryotes and phylogenetic classification of Protozoa. Int J Syst Evol Microbiol 52: 297–354 Coutinho PM, Deleury E, Davies GJ, Henrissat B (2003a) An evolving hierarchical family classification for glycosyltransferases. J Mol Biol 328: 307–317 Coutinho PM, Stam M, Blanc E, Henrissat B (2003b) Why are there so many carbohydrate-active enzyme-related genes in plants? Trends Plant Sci 8: 563–565 € gren AM, Andersson JO, Roger AJ, Doolittle WF (2003) Evidence Dacks JB, Davis LA, Sjo for Golgi bodies in proposed ‘Golgi-lacking’ lineages. Proc Biol Sci 270 Suppl 2: S168–S171 Dacks JB, Doolittle WF (2001) Reconstructing/deconstructing the earliest eukaryotes: how comparative genomics can help. Cell 107: 419–425 Dacks JB, Doolittle WF (2002) Novel syntaxin gene sequences from Giardia Trypanosoma and algae: implications for the ancient evolution of the eukaryotic endomembrane system. J Cell Sci 115: 1635–1642 Dacks JB, Doolittle WF (2004) Molecular and phylogenetic characterization of syntaxin genes from parasitic protozoa. Mol Biochem Parasitol 136: 123–136 Dacks JB, Field MC (2007) Evolution of the eukaryotic membrane-trafficking system: origin, tempo and mode. J Cell Sci 120: 2977–2985 Deane JA, Fraunholz M, Su V, Maier UG, Martin W, Durnford DG, McFadden GI (2000) Evidence for nucleomorph to host nucleus gene transfer: light-harvesting complex proteins from cryptomonads and chlorarachniophytes. Protist 151: 239–252 Dennes A, Cromme C, Suresh K, Kumar NS, Eble JA, Hahnenkamp A and Pohlmann R (2005) The novel Drosophila lysosomal enzyme receptor protein mediates lysosom-

Evolution of the Golgi complex

*

689

al sorting in mammalian cells and binds mammalian and Drosophila GGA adaptors. J Biol Chem 280: 12849–12857 Devos D, Dokudovskaya S, Alber F, Williams R, Chait BT, Sali A, Rout MP (2004) Components of coated vesicles and nuclear pore complexes share a common molecular architecture. PLoS Biol 2: e380 Devos D, Dokudovskaya S, Williams R, Alber F, Eswar N, Chait BT, Rout MP, Sali A (2006) Simple fold composition and modular architecture of the nuclear pore complex. Proc Natl Acad Sci USA 103: 2172–2177 Eichinger L, Pachebat JA, Glockner G, Rajandream MA, Sucgang R, Berriman M, Song J, Olsen R, Szafranski K, Xu Q et al (2005) The genome of the social amoeba Dictyostelium discoideum. Nature 435: 43–57 Eichler J, Adams MW (2005) Posttranslational protein modification in Archaea. Microbiol Mol Biol Rev 69: 393–425 Ercan A, West CM (2005) Kinetic analysis of a Golgi UDP-GlcNAc:polypeptide-Thr/Ser N-acetyl-alpha-glucosaminyltransferase from Dictyostelium. Glycobiology 15: 489–500 Ernst JF, Prill SK (2001) O-glycosylation. Med Mycol 39 Suppl 1: 67–74 Faik A, Price NJ, Raikhel NV, Keegstra K (2002) An Arabidopsis gene encoding an alpha-xylosyltransferase involved in xyloglucan biosynthesis. Proc Natl Acad Sci USA 99: 7797–7802 Fiedler K, Simons K (1994) A putative novel class of animal lectins in the secretory pathway homologous to leguminous lectins. Cell 77: 625–626 Field MC, Gabernet-Castello C and Dacks JB (2007) Reconstructing the evolution of the endocytic system: insights from genomics and molecular cell biology. Adv Exp Med Biol 607: 84–96 Fu J, Kreibich G (2000) Retention of subunits of the oligosaccharyltransferase complex in the endoplasmic reticulum. J Biol Chem 275: 3984–3990 Goto M (2007) Protein O-glycosylation in fungi: diverse structures and multiple functions. Biosci Biotechnol Biochem 71: 1415–1427 Gould SB, Sommer MS, Kroth PG, Gile GH, Keeling PJ, Maier UG (2006) Nucleus-tonucleus gene transfer and protein retargeting into a remnant cytoplasm of cryptophytes and diatoms. Mol Biol Evol 23: 2413–2422 Harris RJ, Spellman MW (1993) O-linked fucose and other post-translational modifications unique to EGF modules. Glycobiology 3: 219–224 Helenius A, Aebi M (2004) Roles of N-linked glycans in the endoplasmic reticulum. Annu Rev Biochem 73: 1019–1049 Igura M, Maita N, Kamishikiryo J, Yamada M, Obita T, Maenaka K, Kohda D (2008) Structure-guided identification of a new catalytic motif of oligosaccharyltransferase. Embo J 27: 234–243 kely G (2003) Small GTPases and the evolution of the eukaryotic cell. Bioessays 25: Je 1129–1138 kely G (2007) Origin of eukaryotic endomembranes: a critical evaluation of different Je model scenarios. Adv Exp Med Biol 607: 38–51 kely G, Arendt D (2006) Evolution of intraflagellar transport from coated vesicles and Je autogenous origin of the eukaryotic cilium Bioessays 28(2): 191–198 Jung E, Gooley AA, Packer NH, Karuso P, Williams KL (1998) Rules for the addition of Olinked N-acetylglucosamine to secreted proteins in Dictyostelium discoideum – in vivo studies on glycosylation of mucin MUC1 and MUC2 repeats. Eur J Biochem 253: 517–524 Jungmann J, Rayner JC, Munro S (1999) The Saccharomyces cerevisiae protein Mnn10p/ Bed1p is a subunit of a Golgi mannosyltransferase complex. J Biol Chem 274: 6579–6585 Keeling PJ, Burger G, Durnford DG, Lang BF, Lee RW, Pearlman RE, Roger AJ, Gray MW (2005) The tree of eukaryotes. Trends Ecol Evol 20: 670–676

690

*

G. J ekely

Kelleher DJ, Gilmore R (2006) An evolving view of the eukaryotic oligosaccharyltransferase. Glycobiology 16: 47R–62R King N (2004) The unicellular ancestry of animal development. Dev Cell 7: 313–325 Koumandou VL, Dacks JB, Coulson RM, Field MC (2007) Control systems for membrane fusion in the ancestral eukaryote; evolution of tethering complexes and SM proteins. BMC Evol Biol 7: 29 Lal K, Field MC, Carlton JM, Warwicker J, Hirt RP (2005) Identification of a very large Rab GTPase family in the parasitic protozoan Trichomonas vaginalis. Mol Biochem Parasitol 143: 226–235 Lang BF, O’Kelly C, Nerad T, Gray MW, Burger G (2002) The closest unicellular relatives of animals. Curr Biol 12: 1773–1778 Lechner J, Wieland F, Sumper M (1985) Biosynthesis of sulfated saccharides N-glycosidically linked to the protein via glucose. Purification and identification of sulfated dolichyl monophosphoryl tetrasaccharides from halobacteria. J Biol Chem 260: 860–866 Letourneur F, Klausner RD (1992) A novel di-leucine motif and a tyrosine-based motif independently mediate lysosomal targeting and endocytosis of CD3 chains. Cell 69: 1143–1157 Liu J, Mushegian A (2003) Three monophyletic superfamilies account for the majority of the known glycosyltransferases. Protein Sci 12: 1418–1431  n HD, Marotta A, Mowatt MR, Sciaky N, Lippincott-Schwartz J, Nash TE (1995) Luja Developmental induction of Golgi structure and function in the primitive eukaryote Giardia lamblia. J Biol Chem 270: 4612–4618 Luo Y, Haltiwanger RS (2005) O-fucosylation of notch occurs in the endoplasmic reticulum. J Biol Chem 280: 11289–11294 € s A, Li Y, Schraner EM, Wild P, Mu € ller N, Knopf LG, Hehl AB (2003) An Marti M, Rego ancestral secretory apparatus in the protozoan parasite Giardia intestinalis. J Biol Chem 278: 24837–24848 Masclaux FG, Galaud JP, Pont-Lezica R (2005) The riddle of the plant vacuolar sorting receptors. Protoplasma 226: 103–108 ry E (1995) The major transitions in evolution. Oxford Maynard Smith J, Szathma University Press, New York Mironov AA, Banin VV, Sesorova IS, Dolgikh VV, Luini A, Beznoussenko GV (2007) Evolution of the endoplasmic reticulum and the Golgi complex. Adv Exp Med Biol 607: 61–72 Mouratou B, Biou V, Joubert A, Cohen J, Shields DJ, Geldner N, Jurgens G, Melancon P, Cherfils J (2005) The domain architecture of large guanine nucleotide exchange factors for the small GTP-binding protein Arf. BMC genomics 6: 20 Munro S (2001) What can yeast tell us about N-linked glycosylation in the Golgi apparatus? FEBS Lett 498: 223–227 Munro S (2005) The Golgi apparatus: defining the identity of Golgi membranes. Curr Opin Cell Biol 17: 395–401 Nakada-Tsukui K, Saito-Nakano Y, Ali V, Nozaki T (2005) A retromerlike complex is a novel Rab7 effector that is involved in the transport of the virulence factor cysteine protease in the enteric protozoan parasite Entamoeba histolytica. Mol Biol Cell 16: 5294–5303 Oliviusson P, Heinzerling O, Hillmer S, Hinz G, Tse YC, Jiang L, Robinson DG (2006) Plant retromer, localized to the prevacuolar compartment and microvesicles in Arabidopsis may interact with vacuolar sorting receptors. Plant Cell 18: 1239–1252 Previato JO, Sola-Penna M, Agrellos OA, Jones C, Oeltmann T, Travassos LR, MendoncaPreviato L (1998) Biosynthesis of O-N-acetylglucosamine-linked glycans in Trypanosoma cruzi. Characterization of the novel uridine diphospho-N-acetylglucosamine:polypeptide N-acetylglucosaminyltransferase-catalyzing formation of N-acetylglucosamine alpha1 ! O-threonine. J Biol Chem 273: 14982–14988

Evolution of the Golgi complex

*

691

Proszynski TJ, Simons K, Bagnat M (2004) O-glycosylation as a sorting determinant for cell surface delivery in yeast. Mol Biol Cell 15: 1533–1543 Putnam NH, Srivastava M, Hellsten U, Dirks B, Chapman J, Salamov A, Terry A, Shapiro H, Lindquist E, Kapitonov VV et al (2007) Sea anemone genome reveals ancestral eumetazoan gene repertoire and genomic organization. Science 317: 86–94 Reiter WD (2002) Biosynthesis and properties of the plant cell wall. Curr Opin Plant Biol 5: 536–542 Richards TA, Cavalier-Smith T (2005) Myosin domain evolution and the primary divergence of eukaryotes. Nature 436: 1113–1118 Robinson DG, Oliviusson P, Hinz G (2005) Protein sorting to the storage vacuoles of plants: a critical appraisal. Traffic 6: 615–625 Saito-Nakano Y, Loftus BJ, Hall N, Nozaki T (2005) The diversity of Rab GTPases in Entamoeba histolytica. Exp Parasitol 110: 244–252 Samuelson J, Banerjee S, Magnelli P, Cui J, Kelleher DJ, Gilmore R, Robbins PW (2005) The diversity of dolichol-linked precursors to Asn-linked glycans likely results from secondary loss of sets of glycosyltransferases. Proc Natl Acad Sci USA 102: 1548–1553 Sanderfoot A (2007) Increases in the number of SNARE genes parallels the rise of multicellularity among the green plants. Plant Physiol 144: 6–17 Schledzewski K, Brinkmann H, Mendel RR (1999) Phylogenetic analysis of components of the eukaryotic vesicle transport system reveals a common origin of adaptor protein complexes 1, 2, and 3 and the F subcomplex of the coatomer COPI. J Mol Evol 48: 770–778 Simpson AG, Roger AJ (2004) The real ‘kingdoms’ of eukaryotes. Curr Biol 14: R693–R696 Souza GM, Mehta DP, Lammertz M, Rodriguez-Paris J, Wu R, Cardelli JA, Freeze HH (1997) Dictyostelium lysosomal proteins with different sugar modifications sort to functionally distinct compartments. J Cell Sci 110 ( Pt 18): 2239–2248 Spiro RG (2002) Protein glycosylation: nature, distribution, enzymatic formation, and disease implications of glycopeptide bonds. Glycobiology 12: 43R–56R Stechmann A, Cavalier-Smith T (2002) Rooting the eukaryote tree by using a derived gene fusion. Science 297: 89–91 Stwora-Wojczyk MM, Kissinger JC, Spitalnik SL, Wojczyk BS (2004) O-glycosylation in Toxoplasma gondii: identification and analysis of a family of UDP-GalNAc:polypeptide N-acetylgalactosaminyltransferases. Int J Parasitol 34: 309–322 Teasdale RD, Jackson MR (1996) Signal-mediated sorting of membrane proteins between the endoplasmic reticulum and the Golgi apparatus. Annu Rev Cell Dev Biol 12: 27–54 Templeton TJ, Iyer LM, Anantharaman V, Enomoto S, Abrahante JE, Subramanian GM, Hoffman SL, Abrahamsen MS, Aravind L (2004) Comparative analysis of apicomplexa and genomic diversity in eukaryotes. Genome Res 14: 1686–1695 Ten Hagen KG, Fritz TA, Tabak LA (2003) All in the family: the UDP-GalNAc:polypeptide N-acetylgalactosaminyltransferases. Glycobiology 13: 1R–16R Vertel BM, Walters LM, Flay N, Kearns AE and Schwartz NB (1993) Xylosylation is an endoplasmic reticulum to Golgi event. J Biol Chem 268: 11105–11112 West CM, Van Der Wel H, Sassi S, Gaucher EA (2004) Cytoplasmic glycosylation of protein-hydroxyproline and its relationship to other glycosylation pathways. Biochim Biophys Acta 1673: 29–44 Whyte JR, Munro S (2001) A yeast homolog of the mammalian mannose 6-phosphate receptors contributes to the sorting of vacuolar hydrolases. Curr Biol 11: 1074–1078 Wieland F, Paul G, Sumper M (1985) Halobacterial flagellins are sulfated glycoproteins. J Biol Chem 260: 15180–15185 Wilson IB (2002) Glycosylation of proteins in plants and invertebrates. Curr Opin Struct Biol 12: 569–577

General conclusions

Concluding remarks, questions, and perspectives

*

695

Concluding remarks, questions, and perspectives Alexander A. Mironov and Margit Pavelka

During these years, beginning from the discovery of the intracellular transport by Jamieson and Palade in 1967, a huge work has been accomplished in deciphering of the molecular mechanisms involved in the intracellular transport. Compared to what was known even 20 years ago, scientists have gained an extraordinary amount of information about the mechanisms of intracellular transport. A major part of the molecular mechanisms has already been unraveled. These insights have unified the several, previously independent subfields in trafficking area into a single coherent discipline – intracellular transport. We tried to summarize most of the known facts and overlap these facts with the main models of transport, and map all these functions, and possibly develop a universal idea about the mechanism of intracellular transport. The mechanisms proposed here can be generally applied to all the steps of the intracelluar transport, and are not restricted to one. This wholistic approach helps us to better comprehend the functioning and evolution of the intracellular transport systems. Now, it could be important to outline the main wholistic questions that remain within the area of intracellular transport. Within the ER-to-Golgi transport step, the main questions are the mechanism of COPII-dependent concentration of cargo, the precise role of COPII coat and COPII-dependent vesicles, mechanisms of the carrier formation at the ER exit sites. It is not clear whether the exit of cargo from the ER, maturation of the ER-to-Golgi carriers and their centralization need membrane fusion and how the COPII/COPI-system functions during the cargo exit from the ER. Important questions are related to the role of ER-to-Golgi connections and better characterization of the intermediate compartment. At the Golgi apparatus level, the main wholistic questions are the mechanisms of the function of the cis Golgi network and the Golgi exit sites; what could be the mechanism of intra-Golgi transport at steady state and whether membranes of the cargo domain undergo maturation during their progression from the cis to medial Golgi and then to the trans Golgi network. There remain many questions related to the problem of stacking. What is the role for Golgi matrix? Are there matrix receptors and if so, how do they recycle? Why are so-called Golgi matrix proteins not found between cisternae? If these matrix proteins have other functions, what glues Golgi cisternae? Why are cisternae flat and narrow? Why the number of medial cisternae is so constant in the same cell type? Why does Golgi cisternae have perforations, and why are the perforations not present in the center of cisternae?

696

*

A. A. Mironov and M. Pavelka

It is not clear what is the exact role of membrane fusion during transport, and whether membrane fusion is necessary to enter the medial Golgi, and required for the exit out of the Golgi apparatus, and whether the cargo domain changes its SNARE composition during progression through the Golgi stacks. How many fusion events occur during intra-Golgi transport? It remains to study how the SNARE machinery is organized within the secretory pathway. What is the role for Rabs? What is the role of Ca2þ flux for the transport and what is its source? The molecular mechanisms involved into Golgidependent signalling should be deciphered better. Many questions are linked with the intercisternal connections. What is the role of intercisternal connections, and how are connections formed? Can lipids diffuse along intercisternal connections, and are connections suitable for diffusion of soluble cargoes? Is there any pH gradient across the stack and if yes, how is this gradient preserved in the presence of intercisternal connections? What is the mechanism responsible for the break down of connections? The most enigmatic issue is the role of COPI-vesicles. One should prove or disprove, whether COPI-vesicles are carriers for anterograde or retrograde cargoes, or find other functions for these structures. The precise organization of the Golgi exit sites remains to be specified. What is the role for fission here? Has fission an important role for the departure of post-Golgi carriers? If yes, what are the molecular machines responsible for fission? The precursors of all carriers departing from the Golgi exit site should be identified together with the molecular mechanisms involved in these processes. The maturation of these carriers during their travelling towards the sites of their destinations needs additional examination. We hope that this book will be useful for scientists working in other fields of the cell biology, for students studying cell biology and for general readers giving them a possibility to have all necessary information in one place. Moreover, we think that some chapters of the book could be useful also for the researchers working in the field of intracellular transport. Sometimes, the overview of the field could help in the determination of the future research directions. The future stage of the development of the field could be the creation of the web site devoted to the intracellular trafficking, where scientists could have all important information already in the form of extended reviews on all proteins involved and all steps of transport. It could be done in the framework of Wikipedia or other platforms.

Contributors

Contributors in alphabetical order Arvan Peter Division of Metabolism, Endocrinology & Diabetes University of Michigan Medical Center 1150 W. Medical Center Drive Ann Arbor MI 48109 USA Banfield David Department of Biology The Hong Kong University of Science and Technology Clear Water Bay Knowloon Hong Kong, SAR of China Berger Eric G. Institute of Physiology University of Zürich Winterthurerstrasse 190 8057 Zürich Switzerland Beznoussenko Galina V. Department of Cell Biology and Oncology Consorzio Mario Negri Sud Via Nazionale 8 66030 Santa Maria Imbaro (Chieti) Italy Bonifacino Juan S. Cell Biology and Metabolism Program National Institute of Child Health and Human Development National Institutes of Health Bethesda MD 20892 USA Cao Hong Center for Basic Research in Digestive Diseases and Department of Biochemistry and Molecular Biology Mayo Clinic College of Medicine 200 First Street Southwest Rochester MN 55905 USA Castino Roberta Department of Medical Sciences University “A. Avogadro” Via Solaroli 17 28100 Novara Italy

*

697

698

*

Contributors

Claas Christoph BIOQUANT University of Heidelberg Im Neuenheimer Feld 267 69120 Heidelberg Germany Colley Karen J. Department of Biochemistry and Molecular Genetics University of Illinois at Chicago College of Medicine 900 S. Ashland Avenue M/C 669 Chicago IL 60607 USA Daraspe Jean IBITEC-S CEA and CNRS URA2096 91191 Gif-sur-Yvette and LRA17V University Paris-Sud 11 91405 Orsay France De Matteis Maria A. Department of Cell Biology and Oncology Consorzio Mario Negri Sud Via Nazionale 8 66030 Santa Maria Imbaro (Chieti) Italy Deborde Sylvie Margaret Dyson Vision Research Institute and Department of Cell and Developmental Biology Weill Medical College of Cornell University 1300 York Ave New York 10065 USA Derby Merran C. Department of Biochemistry and Molecular Biology and Bio21 Molecular Science and Biotechnology Institute The University of Melbourne Melbourne Victoria 3010 Australia Derganc Jure Institute of Biophysics, Faculty of Medicine University of Ljubljana Lipiceva 2 1000 Ljubljana Slovenia Egea Gustavo Department Biologia Cellular i Anatomia Patològica Facultat de Medicina Universitat de Barcelona C /Casanova, 143 08036 Barcelona Spain

Contributors Ellinger Adolf Department of Cell Biology and Ultrastructure Research Center for Anatomy and Cell Biology Medical University of Vienna Schwarzspanierstraße 17 1090 Vienna Austria Eskelinen Eeva-Liisa Department of Biological and Environmental Sciences Division of Biochemistry University of Helsinki Viikinkaari 5D, 00014 Helsinki Finland Gleeson Paul A. Department of Biochemistry and Molecular Biology and Bio21 Molecular Science and Biotechnology Institute The University of Melbourne Melbourne Victoria 3010 Australia Gravotta Diego Margaret Dyson Vision Research Institute and Department of Cell and Developmental Biology Weill Medical College of Cornell University 1300 York Ave New York 10065 USA Hauri Hans-Peter Biozentrum University of Basel Klingelbergstrasse 70 4056 Basel Switzerland Hawes Chris School of Life Sciences Oxford Brookes University Oxford, OX3 OBP UK Hong Wanjin Cancer and Developmental Cell Biology Division (CDCBD) Institute of Molecular and Cell Biology (IMCB), A*STAR 61 Biopolis Drive Singapore 138673 Singapore Hsu Victor W. Harvard Medical School Brigham and Women’s Hospital One Jimmy Fund Way, Smith 538 Boston MA 02115 USA

*

699

700

*

Contributors

Isidoro Ciro Department of Medical Sciences University “A. Avogadro” Via Solaroli 17 28100 Novara Italy

Jékely Gáspár Max Planck Institute for Developmental Biology Spemannstrase 35 72076 Tübingen Germany

Kahn Richard A. Department of Biochemistry Emory University School of Medicine 1510 Clifton Rd. Atlanta GA 30322 USA

Képès François Epigenomics Project, Genopole CNRS & University of Evry Tour Évry2 91034 ÉVRY cedex France

Koegler Eva Biozentrum University of Basel Klingelbergstrasse 70 4056 Basel Switzerland Lakkaraju Aparna Margaret Dyson Vision Research Institute and Department of Cell and Developmental Biology Weill Medical College of Cornell University 1300 York Ave New York 10065 USA Lee Stella Y. Division of Rheumatology, Immunology and Allergy Brigham and Women’s Hospital Harvard Medical School One Jimmy Fund Way, Smith 528 Boston MA 02115 USA

Contributors Lieu Zi Zhao Department of Biochemistry and Molecular Biology Bio21 Molecular Science and Biotechnology Institute The University of Melbourne Melbourne Victoria 3010 Australia Lu Lei Cancer and Developmental Cell Biology Division (CDCBD) Institute of Molecular and Cell Biology (IMCB), A*STAR 61 Biopolis Drive Singapore 138673 Singapore Luini Alberto Department of Cell Biology and Oncology Consorzio Mario Negri Sud Via Nazionale 8 S. Maria Imbaro (Chieti) 66030 Italy Lupashin Vladimir Department of Physiology and Biophysics College of Medicine University of Arkansas for Medical Sciences Little Rock AR 72205 USA Mardones Gonzalo A. Cell Biology and Metabolism Program National Institute of Child Health and Human Development National Institutes of Health Bethesda MD 20892 USA McNiven Mark A. Center for Basic Research in Digestive Diseases and Department of Biochemistry and Molecular Biology Mayo Clinic College of Medicine 200 First Street Southwest Rochester MN 55905 USA Micaroni Massimo Laboratory of Intracellular Traffic Department of Cell Biology and Oncology Consorzio Mario Negri Sud Via Nazionale 8 66030 Santa Maria Imbaro (Chieti) Italy

*

701

702

*

Contributors

Mironov Alexander A. Laboratory of Intracellular Traffic Department of Cell Biology and Oncology Consorzio Mario Negri Sud Via Nazionale 8 66030 S. Maria Imbaro (Chieti) Italy Nakano Akihiko Department of Biological Sciences, Graduate School of Science University of Tokyo, 7-3-1 Hongo, Bunkyo-ku Tokyo 113-0033 Japan and Molecular Membrane Biology Laboratory RIKEN Discovery Research Institute, 2-1 Hirosawa, Wako Saitama 351-0198 Japan Neumüller Josef Department of Cell Biology and Ultrastructure Research Center for Anatomy and Cell Biology Medical University of Vienna Schwarzspanierstraße 17 1090 Vienna Austria Nyfeler Beat Biozentrum University of Basel Klingelbergstrasse 70 4056 Basel Switzerland Osterrieder Anne School of Life Sciences Oxford Brookes University Oxford, OX3 OBP UK Pavelka Margit Department of Cell Biology and Ultrastructure Research Center for Anatomy and Cell Biology Medical University of Vienna Schwarzspanierstrasse 17 1090 Vienna Austria Pepperkok Rainer Cell Biology Cell Biophysics Unit, EMBL Meyerhofstraße 1 69117 Heidelberg Germany

Contributors Perez-Vilar Juan Cystic Fibrosis/Pulmonary Research and Treatment Center University of North Carolina at Chapel Hill Chapel Hill NC 27599 USA Polishchuk Roman S. Department of Cell Biology and Oncology Consorzio Mario Negri Sud Via Nazionale 8 66030 Santa Maria Imbaro (Chieti) Italy Rambourg Alain IBITEC-S CEA and CNRS URA2096 91191 Gif-sur-Yvette and LRA17V University Paris-Sud 11 91405 Orsay France Reiterer Veronika Biozentrum University of Basel Klingelbergstrasse 70 4056 Basel Switzerland Ríos Rosa M. Departamento de Señalización Celular CSIC-Centro Andaluz de Biomedicina y Medicina Regenerativa 41092 Sevilla Spain Rizzuto Rosario Department of Experimental and Diagnostic Medicine Section of General Pathology University of Ferrara Via Borsari 46 44100 Ferrara Italy Rodriguez-Boulan Enrique Margaret Dyson Vision Research Institute and Department of Cell and Developmental Biology Weill Medical College of Cornell University 1300 York Ave New York 10065 USA Rohrer Jack Institute of Physiology University of Zürich Winterthurerstrasse 190 8057 Zürich Switzerland

*

703

704

*

Contributors

Sallese Michele Unit of Genomic Approaches to Membrane Traffic Laboratory of Membrane Traffic Department of Cell Biology and Oncology Consorzio Mario Negri Sud Via Nazionale 8 66030 Santa Maria Imbaro (Chieti) Italy Sato Ken Department of Life Sciences, Graduate School of Arts and Sciences University of Tokyo Komaba, Meguro-ku Tokyo 153-8902 Japan Scanu Tiziana Department of Cell Biology and Oncology Consorzio Mario Negri Sud Via Nazionale 8 66030 Santa Maria Imbaro (Chieti) Italy Sokolova Yuliya Y. Tulane National Primate Research Center Division of Microbiology 18703 Three Rivers Road Covington LA 704333 USA and Laboratory of Microbiological Control All-Russian Institute for Plant Protection Sh. Podbelskogo, 3 189620 St. Petersburg – Pushkin Russia Sparkes Imogen School of Life Sciences Oxford Brookes University Oxford, OX3 OBP UK Starkuviene Vytante BIOQUANT University of Heidelberg Im Neuenheimer Feld 267 69120 Heidelberg Germany Svetina Saša Jozef Stefan Institute Jamova 39 and Institute of Biophysics, Faculty of Medicine University of Ljubljana Lipiceva 2 1000 Ljubljana Slovenia

Contributors Ungar Daniel Department of Biology University of York York, YO10 5YW UK Verbavatz Jean-Marc IBITEC-S CEA and CNRS URA2096 91191 Gif-sur-Yvette and LRA17V University Paris-Sud 11 91405 Orsay France Verissimo Fatima Cell Biology Cell Biophysics Unit, EMBL Meyerhofstraße 1 69117 Heidelberg Germany Wang Yanzhuang Department of Molecular, Cellular and Developmental Biology University of Michigan, 830 North University Avenue Ann Arbor MI 48109-1048 USA Weller Shaun Center for Basic Research in Digestive Diseases and Department of Biochemistry and Molecular Biology Mayo Clinic College of Medicine 200 First Street Southwest Rochester MN 55905 USA Wilson Cathal Department of Cell Biology and Oncology Consorzio Mario Negri Sud Via Nazionale 8 66030 Santa Maria Imbaro (Chieti) Italy Yang Jia-Shu Division of Rheumatology, Immunology and Allergy, Brigham and Women’s Hospital, and Department of Medicine Harvard Medical School One Jimmy Fund Way, Smith 528 Boston MA 02115 USA Zhao Weihan Department of Biochemistry and Molecular Genetics University of Illinois at Chicago College of Medicine 900 S. Ashland Avenue M/C 669 Chicago IL 60607 USA

*

705

706

*

Index

Index A N-Acetylglucosaminyltransferases 169 Acid hydrolases 214, 388, 389, 397, 414, 682, 684 Acid phosphatase 250, 407, 409, 640, 650 Actin 72, 90, 91, 107, 224, 226, 270, 271, 277, 278–286, 288, 289, 307, 309–311, 325, 338, 383, 433, 502, 503, 536, 563, 572, 574, 589, 611–613, 657, 663 Adaptins 109, 111, 420, 436, 653, 654, 677, 687 Adaptors 23, 27, 72, 82, 87, 90, 93, 109–112, 254, 257, 283, 284, 302, 311, 360, 361, 377, 382, 388, 389, 391, 392, 395, 396, 417–421, 425, 432, 436, 437, 440, 501, 527, 563, 565, 567, 569, 570, 572, 574, 594, 596, 662, 677, 684 Adaptor protein complex1, AP-1 109, 111, 112, 215, 254, 257, 284, 311, 361, 362, 382, 389–393, 395–397, 417, 418, 420, 432, 460, 477, 501, 502, 525, 624 Adhesion 171–173, 191, 254, 255, 257, 258, 320, 326, 522, 525, 527, 529, 649, 651, 656 Antiporter 163, 193, 195 Apical 9, 18, 28, 146, 150, 212, 223, 226, 261, 276, 277, 281, 283, 285, 288, 305, 306, 322, 342, 345, 347, 359, 375, 378, 380, 382, 384, 459, 478, 489, 494, 536, 552, 563–574, 652, 653, 659 Apicomplexa 647, 652–654, 663, 678, 681

Arf GAPs 54, 108, 111, 112, 661, 662 Arf, ADP-ribosylation factor 54, 90, 106, 107–114, 252, 257, 280, 309, 361, 363, 436, 479, 501, 569, 581, 623, 662, 677, 684, 686, 687 Arf-dependent adaptors 110, 112 ARF1 54, 87–92, 95, 97–99, 106– 108, 253, 257, 259, 260, 280, 282, 285, 309, 310, 361, 362, 388, 392, 393, 432, 436, 440, 479, 480, 490, 581, 594, 614, 618, 624 ARFGAP1 89, 92, 93, 96, 98, 111 Arfrp1 107, 113, 362, 363, 439–441 Arls 3, 106, 107, 114, 228, 363, 365, 368, 440 Arl1s 107, 112–114, 228, 231, 362– 368, 439–442, 624, 657 Attached CGN 18, 21, 230, 238, 239 Avoidance of non-granule proteins 499

B BAPTA 144, 150–152, 256 BARS 89, 92–94, 233, 250, 305–307, 323, 352, 379, 572, 591, 593, 595, 596 Basolateral 18, 146, 150, 212, 261, 277, 281, 283, 285, 305, 306, 322, 347, 359, 362, 375, 377, 382–384, 459, 477, 552, 563–574 Bet3 131–137, 139, 624 Beta-glucocerebrosidase 409, 419 Bifunctional enzymes 176, 177 Blood groups 171, 172, 176, 177 Brefeldin A (BFA) 88, 110, 491, 581, 615, 641, 650

Index

C Ca2+ binding protein 145, 147–148, 256, 257 Ca2+ channel 149, 152, 254 Ca2+ pump 147, 148, 154 Calcium 3, 143, 144, 147, 149, 153, 208, 212–214, 216, 255, 257, 486, 497, 498, 660 Calmodulin 145, 148, 150–152, 660 CALNUC 147–149, 256 Camillo Golgi 7, 630 Cargo receptor 82, 211, 216, 280, 432, 493, 494, 624, 683 Carrier maturation–progression model 350 Cathepsin 212, 403–405, 407, 408, 446, 475 Caveolae 425, 520, 523 Cazy database 163, 174, 679 Cisterna maturation–progression model 10, 11, 353, 354, 347 Cis-Golgi network 21, 32, 228, 239, 321, 327, 338 Cis-, medial- and trans-Golgi 19 Cis-perforated cisternae of the intermediate compartment 21 Cisternal maturation 11, 94, 95, , 178, 352, 426, 491, 544, 554, 557, 581, 599, 617, 618, 625, 626 Classical cisterna maturation model 375 Clathrin 8, 10, 23, 26, 27, 32, 87, 90, 93, 110, 146, 154, 215, 280, 283, 301, 302, 309, 311, 325, 352, 361, 380, 384, 388–390, 392, 393, 395, 396, 414, 418, 420, 425, 426, 428, 430, 431, 436, 437, 440, 461, 462, 466, 467, 475, 477, 479, 488, 498, 501, 502, 525, 527, 563, 567–570, 574, 611, 618, 650, 653, 661, 662, 663, 677, 684, 687 Clathrin-coated vesicles (CCV) 8, 10, 23, 27, 215, 307, 388, 389, 392, 393, 395–397, 420, 467, 611

*

707

Coat protein complex II 78 Coatomer 44, 54, 87–92, 94–98, 152, 253, 257, 275, 280, 346, 348, 349, 436, 479, 480, 488, 581, 594, 595, 650, 654–656, 662 COG 3, 56, 120–126, 178, 439, 442, 443, 677 COG complex 56, 57, 120–127, 178, 443, 677 Combined models 352 Compartments along the secretory pathway 16, 30, 32, 347, 404 Compensatory endocytosis 475 Conference in Pavia 9 Constitutively secreted proteins 154, 477 Constitutive secretory (CS) pathway 485, 486, 499–501 COPI 3, 8, 10, 11, 16, 18, 21–24, 26, 29, 32, 54, 55, 58, 59, 87–99, 109, 110, 122, 123, 135, 144–146, 149, 151, 152, 200, 212, 235, 236, 253, 256, 257, 260, 261, 275, 280, 325, 326, 333, 334, 336, 338, 342–345, 347–349, 352, 353, 361, 364, 377, 383, 384, 432, 436, 440, 443, 479, 481, 489, 491, 583, 593–596, 617, 618, 624, 650, 653, 655, 661, 677, 687 COPI and endosomes 479 COPI vesicles 10, 18, 23, 24, 88, 89, 92, 94–98, 126, 180, 236, 275, 316, 317, 326, 334, 336, 338, 343–346, 348, 349, 352, 353, 491, 581, 582, 594, 595, 597, 617, 618, 626, 661, 663, 696 COPII 3, 7, 9, 16, 17, 18, 21, 23, 32, 46, 47, 59, 72, 79–83, 88, 97, 107, 112, 133, 152, 199–212, 216, 235, 325, 333–339, 353, 432, 444, 543, 613, 614, 617, 624, 626, 650, 655, 661, 663, 677, 687, 695 COPII-coated buds 16 Condensing vacuoles 486, 487, 489

708

*

Index

Correlative light-electron microscopy (CLEM) 21, 379, 393, 395 Cortactin 3, 280–282, 301–303, 306–307, 309–311 Cytoplasmic domain 53, 81, 90, 94, 97, 138, 165, 178, 179, 212, 229, 230, 432, 433, 437, 566, 571, 653 Cytoplasmic dynein 225, 273, 274, 310, 380, 382, 571 Cytoskeleton 3, 19, 30, 107, 248, 249, 270, 271, 274, 277, 279, 280, 283, 284–286, 288, 289, 326, 335, 338, 433, 440, 486, 502, 556, 572, 585, 613–615, 649, 655, 686

D Delivery of GPCs 380–383 Diffusible cargoes 11, 377 Dileucine motif 215, 567, 569, 682, 683 Diplomonads 647, 649, 648, 662, 663 Disassembly 55, 74, 132, 152, 248, 249, 257, 261, 271, 276, 334, 522, 581–583, 585, 590–599, 649, 651 Discoveries in the field of intracellular transport 7 Dynactin 225, 226, 231, 274, 275, 338, 415 Dynamin 3, 280, 284, 301–307, 309–311, 325, 352, 379, 430, 438, 475, 505, 572

E EGTA 143, 144, 150–152 Electron tomography 464, 467, 527, 529, 643 Elimination of Golgi enzymes from GPC 377 Endocytic system 7, 378, 459, 468, 477, 479 Endocytic TGN 9, 26, 459, 462, 463, 467, 468

Endocytosis 3, 8, 25, 66, 70, 107, 154, 247, 256, 280, 307, 310, 311, 393, 406, 408, 409, 417, 418, 425, 426, 428, 430, 437, 467, 468, 475, 476, 479, 480, 494, 520, 521, 570, 657, 661 Endophilin B 89, 92–94 Endoplasmic reticulum 611 Endosome 3, 9, 19, 23, 25, 26, 30, 51, 70, 71, 108, 109, 112, 114, 135, 144–146, 154, 209, 214–216, 238, 251, 252, 259, 285, 309, 351, 358– 363, 365, 367, 368, 375, 378, 382– 384, 388–397, 408, 409, 414, 415, 417–421, 425–431, 433–446, 460– 463, 466, 467, 476–481, 488, 489, 495, 501, 503, 521, 527, 556, 552, 563–566, 569, 572, 653 Endosomal compartments 144, 253, 360, 402, 460, 461, 477, 564 Endothelial cells 488, 520, 529, 551 Entamoebida 647, 662, 663 Epithelial cells 19, 21, 26, 223, 226, 272, 276, 277, 283, 287, 305, 342, 345, 380, 382, 425, 487, 494, 563–566, 569, 572, 574, 649 ER 1, 3, 7–9, 16–19, 21–23, 27, 29, 30, 32, 46, 49, 51, 54, 56–59, 70, 72, 78–83, 88, 94, 96–98, 106–108, 125, 133–139, 143–151, 153–155, 163, 165, 167, 174, 176, 178, 180, 195, 198, 200, 201, 207–214, 216, 217, 224–227, 229, 230, 232, 234, 235, 237–240, 248–252, 254–256, 258– 262, 270, 271, 273–275, 277, 279– 282, 284–286, 288, 306, 314, 321, 322, 324, 327, 333–339, 342, 344, 347, 430, 431, 435, 438, 443, 444, 459, 461–464, 467, 468, 476, 480, 481, 485–487, 489–491, 493, 506, 520, 521, 524, 525, 527, 529, 542– 545, 547, 549, 558, 565, 574, 580– 582, 585, 587–590, 595, 597, 611– 615, 617–619, 623, 626, 636–641, 643, 650–655, 657–663, 678, 679, 682, 684–687, 695

Index

ER-Golgi intermediate compartment (ERGIC) 51, 83, 108, 177, 209, 211–214, 216, 217, 388, 404, 652, 682 ER-Golgi transport 18, 46, 56, 136 ER–Golgi connections 17, 18 ER–Golgi interface 16, 18, 333, 336, 613, 615 ER-to-Golgi carriers 16, 18, 21, 23, 32, 145, 146, 151, 227, 275, 337, 344, 695 ER exit sites 9, 23, 29, 81, 106, 133, 177, 224, 239, 240, 324, 333, 480, 543, 587, 588, 613, 614, 695 ER export 21, 82, 198–200, 212, 214, 462, 615, 617 ER quality control 210, 213, 216 ERAD 210, 211, 214, 217 ERES 8, 21, 32, 81, 146, 231, 240, 270, 271, 273, 277, 333–337, 384, 480, 489, 587, 626 ERGIC 3, 51, 56, 209, 211, 212, 214, 216, 334, 336, 338, 339, 388, 650, 652, 657 Evidence against the cisterna maturation model 347 Evolution 3, 9, 73, 107, 138, 177, 535, 619, 647, 648, 658, 663, 675–678, 680, 682, 684, 686, 687, 695 Exocytosis 7, 25, 32, 144, 152, 154, 178, 256, 277, 475, 476, 480, 485– 487, 500, 502–505, 521, 529, 536, 538, 563, 651, 655, 657, 660 Exocytosis of SGs 504

F Fissioning of GPCs 378, 379 Flippases 324, 681 FRAP 225, 395, 547, 548, 550, 571, 614 Function of the Golgi apparatus 25, 270, 523, 580, 647

*

709

Function of the Golgi exit site 27 Fusion 1, 10, 24, 26, 30–32, 43, 49– 53, 55, 57, 58, 66, 72, 73, 83, 98, 130, 132–136, 138, 139, 143–146, 150, 152–154, 177, 199, 200, 233– 236, 238–240, 251, 253, 256, 259, 271, 275, 277, 318, 325, 326, 334, 336, 338, 339, 342, 344, 350–352, 358, 360, 361, 365, 367, 368, 376, 377, 378, 382–384, 392, 397, 419, 425, 432–437, 439, 441, 460, 475, 478, 481, 485, 487–489, 492, 493, 497, 498, 500, 502, 503, 505, 506, 526, 536, 547–549, 553–556, 573, 574, 582, 585, 588, 590, 591, 594–598, 612, 614, 631, 642, 655, 660, 662, 676, 677, 686, 687, 695, 696 Fusion with the PM 152, 382, 384

G Gagosome 177 Galactosylation 125, 171, 196 GDI displacement factor 67, 74 GDP–GTP exchange factors 67 Gel phase transitions 541 General principles of intracellular transport 29 GGA(s) (1-3) 27, 215, 361, 389–397, 436, 437, 440, 569, 624 GlcNAc2Man9Glc3 681 Glycosidases 1, 162, 207, 208, 347, 545, 663 Glycosylation 1, 8, 20, 22, 25, 87, 95, 124–127, 150, 153, 161, 162, 164–168, 170, 174, 176–181, 190, 191, 195–197, 200–202, 207, 234, 235, 238, 271, 350, 378, 403, 416, 443, 524, 525, 537, 538, 555, 581, 623, 630, 638, 641, 649, 651, 655, 657, 660, 661, 678, 685, 687 N-Glycosylation 8, 28, 125, 161, 167, 168, 179, 207, 403, 416, 657, 678, 680, 681, 684, 685

710

*

Index

Glycosyltransferases 27, 125, 127, 161–165, 167, 170, 171, 177, 181, 190, 196, 200–201, 207, 234, 255, 279, 358, 545, 558, 661, 678, 679, 682, 685 Goblet cells 19, 21, 26, 535, 536, 544, 545, 547, 549, 551, 556, 558 Golgi biogenesis 338–339, 584, 586, 588, 589, 599, 626, 657 Golgi cisternae 9, 11, 17, 18, 21–24, 26, 27, 29, 87, 94, 95, 120, 138, 143–146, 148, 149, 168, 170, 171, 173, 190, 200, 227, 235, 236, 238, 239, 259, 279, 305, 307, 316, 326, 327, 346–350, 358, 359, 376–378, 435, 459, 462, 466–468, 486, 487, 489, 493, 556, 580–584, 593, 594, 596–598, 623–625, 650, 654, 695 Golgi complex 3, 88, 91, 97, 132, 133, 136, 143, 155, 247–249, 251–262, 276, 278, 282, 307, 314, 315, 325, 333–338, 384, 388, 393, 395, 396, 402, 403, 409, 414, 426, 435, 440, 442, 475, 481, 488–491, 496, 500, 506, 536, 543, 545, 556, 568–569, 571, 581, 598, 648, 649, 651, 654, 675, 677, 682, 687 Golgi enzymes 3, 18–20, 53, 95, 126, 127, 148, 162, 178, 200, 223, 225, 235, 236, 344, 346–349, 353, 354, 376–378, 405, 427, 443, 478, 581, 589, 595, 626 Golgi lumen 8, 25, 145, 148, 149, 193, 194, 256, 500, 581, 590 Golgi ministacks 224, 225 Golgi ribbon 9, 24, 28, 126, 223–227, 229, 232–235, 237, 239, 240, 251, 270, 271, 276, 277, 288, 352, 369, 468, 582, 583, 585, 586, 590–595, 599, 625, 631, 632 Golgi stack 9, 11, 18, 19, 21, 22, 25, 26, 28, 95, 143, 147, 152, 175–177, 223, 224, 226, 233, 234, 236–240, 251, 270, 272, 273, 277–279, 281, 284, 285, 288, 305, 326, 333,

347–350, 354, 358, 363, 368, 375, 377, 405, 426, 463, 464, 467, 468, 487, 591, 522, 524, 581–596, 598, 599, 612–615, 617–619, 625, 630, 631, 635, 636, 638, 640, 657, 677, 696 Golgi vesicles 23, 193, 197, 302, 344, 348, 573, 591, 592, 617 Golgi-localized, gamma-ear containing, ADP ribosylation factor binding proteins, GGAs 389 Golgins 3, 22, 26, 56, 112, 113, 122, 223, 227–229, 231, 232, 235–240, 362–369, 439, 441, 442, 586, 594, 616 GPC precursors 376–379, 381 Gradients within a Golgi stack 19 Granule lumen 490, 536, 539, 541, 546–549, 551, 553, 558 Granule organization 547, 553 GRIP domain 11, 113, 228, 364, 365, 367–369, 439, 441, 442 GTPase activating proteins 67, 108, 438, 439

H Habc domain 44, 45, 59 Heterokaryon experiments 10, 345 Heterotrimeric G proteins 109, 248, 254, 256, 259–261 History 3, 9, 93, 675, 685, 687 History of models of intracellular transport 9 Hot spots on the PM 383

I Inheritance 3, 284, 288, 580, 581, 586, 587–589, 591, 598, 599 Intercisternal connections 9, 24, 29, 347, 349, 350, 352–354, 492, 696 Intermediate filaments 271, 286, 287

Index

Intracellular traffic 1, 2, 16, 143, 283, 475, 476, 535, 539, 541, 542, 553–555, 647, 655, 696, 701, 702 Intracellular transport 1–3, 7–9, 29–31, 66, 94, 130, 143, 145, 150, 154, 235, 236, 238, 240, 253, 466, 475, 477, 481, 493, 535, 537, 539, 509, 647, 651, 652, 660, 695, 696 Intra-Golgi transport 3, 8–11, 21, 28, 29, 70, 72, 88, 89, 94–96, 109, 144, 145, 150, 209, 234, 236, 238, 254, 256, 338, 342, 343, 346–348, 350–354, 435, 489, 490, 491, 493, 496, 544, 596, 662, 696 In vitro reconstitution experiments 1 IP3 145, 147, 149, 150, 250 Isolation of COPI vesicles 349

K KDEL receptor 96, 251, 259, 404, 682 Kinetoplastida 647, 656, 662, 663 Kiss-and-run model 25, 350–352, 354, 376, 477, 481, 491, 496, 506

L Labeled lipid analogues 521 Lateral diffusion model 31, 350, 375 Lateral segregation 326 LCEA (last common eukaryotic ancestor) 675–678, 680, 682, 687 Lectins 3, 7, 124, 196, 285, 461, 558, 651, 680–682 Live-cell imaging 379, 393, 395, 396, 614 Longin 44–46, 58 Lysosome 8, 32, 97, 146, 153, 167, 214–216, 226, 238, 250, 28, 351, 375, 383, 388, 389, 402, 406, 408, 409, 414, 415, 417–421, 425, 426– 428, 431, 436, 440, 444, 446, 460, 477, 480, 488, 495, 501, 504, 527,

*

711

536, 552, 564, 644, 650, 661, 675, 682 Lysosomal membrane protein 298, 414–420, 499, 501, 657 Lysosomal sorting 209, 682, 683

M Macromolecules 1, 535–537, 539, 541–545, 546, 556, 631, 648 Macular corneal dystrophy 173 Mannose 6-P receptor 402–404 Mannose 6-phosphate receptors, MPRs 211, 214, 216, 309, 360, 367, 382, 388–393, 395–397, 402, 406–409, 414, 421, 427, 429, 460, 466, 682 Mannosidases 19, 21, 126, 162, 167–170, 181, 207, 209–211, 216, 217, 287, 348, 657 Matrix proteins 3, 22, 72, 230, 236, 349, 521, 551, 585, 589, 590, 591, 595, 599, 613, 615, 616, 619, 663, 695 Membrane 425–427, 429–433, 435–437, 439–441, 444–446 Membrane attachment of Rabs 66 Membrane curvature 92, 108, 109, 310, 315–318, 321, 322, 324, 326, 445, 572 Membrane flow 163, 346, 427, 476, 559, 635, 642–644 Membrane input from the endoplasmic reticulum 227 Membrane shape 315, 318–321, 323, 327 Membrane spontaneous curvature 318–320, 322, 323, 327 Membrane traffic 66, 71, 73, 93, 106–110, 112, 130, 134, 138, 139, 144, 226, 234, 247–249, 251–262, 271, 275, 276, 283, 286, 288, 305, 363, 364, 366–368, 383, 406, 425, 433, 440, 444, 501, 562, 612, 623, 650, 662, 677, 704

712

*

Index

Microsporidia 253, 647, 658–663, 677, 678, 680, 686 Microtubules 18, 22, 224, 270, 271, 276, 283, 325, 336, 337, 338, 375, 379, 381, 383, 395, 415, 480, 302, 574, 585, 586, 612, 625, 649, 663 Microtubule-organizing centre 223, 241, 338 Mint(s) (1-3) 109 Mitosis 25, 112, 230, 231, 236, 237, 239, 248, 249, 334, 335, 581, 582, 584, 585, 588–596, 598, 599, 615, 649, 653, 658 Models of intra-Golgi transport 11, 29, 236, 238, 342, 343, 352, 353, 491 Molecular switches 67, 121, 274 Molecular tools 2, 580 Morphodynamics 630 Morphology of the Golgi apparatus 19 Motor proteins 26, 72, 73, 235, 270, 273, 274, 358, 360, 382, 432 Movement 10, 22, 25, 224, 226, 273–275, 282, 283, 285, 286, 334, 338, 342, 351, 379, 380–382, 384, 396, 415, 432, 478, 489, 502, 504, 540, 580, 611–614, 619, 663 Mucins 535–539, 541–547, 551–556, 558–560 Mucin gels 541, 551, 552, 556, 558, 559 Mucolipidosis 168, 408 Multimerization/condensation 486, 493, 499 Multi-spanning membrane protein 200 Myosin 72, 226, 227, 280, 282–284, 286, 288, 289, 310, 380, 382, 574, 589, 612

N Non-diffusible cargoes 11 Nucleotide sugar transporter 192, 193, 195

190,

O OST (oligosaccharyltransferase) 167, 207, 216, 681–683, 685

P Palmitoylation 55, 58 Parasitic proteists 647, 651, 653, 655 Phospholipase A2 250, 323 Phospholipase C 250, 253, 323 Phospholipase D 250, 251 Phosphorylation 22, 55, 57, 90, 171, 180, 181, 190, 214, 215, 229, 239, 253, 255, 311, 334, 390, 391–393, 403, 405, 437, 440, 466, 581, 583, 584, 592–596, 599 Phosphotransferase 168, 214, 403, 405, 408 Phyletic distribution 676, 679, 680 Physiology of traffic 1 Pichia pastoris 80, 223, 333, 587, 588, 624, 635, 642, 643 Plant 2, 3, 7, 18, 19, 120, 124, 177, 208, 211, 224, 237, 270, 271, 285, 286, 288, 333, 334, 338, 360, 364, 426, 429, 459, 461, 485, 580, 589, 599, 611, 612, 614–619, 631, 632, 635, 641, 654, 661, 662, 676, 677, 678, 680, 682, 684 Plant and bacterial toxins 459 Pleiomorphic transport carriers (PTCs) 388, 389, 393–397 Plus-end motor kinesin 226 Polarity 3, 192, 223, 277, 285, 286, 554, 563, 565, 567, 571, 574, 584, 599 Polysialic acid 173 Processing glycosidases 162 Progression model 9–11, 29, 31, 343, 346, 347, 350, 353, 354, 554 Prosaposin 407, 409 Protein folding 1, 96, 207–210, 216, 219

Index

Protein kinase A 57, 259 Protein kinase C 252, 260, 379 Protein traffic 211, 498, 654

R Rabs 3, 66–74, 121, 122, 228, 361, 435, 438, 503, 504, 574, 696 Rab cycle 73 Rab effectors 69, 71, 361 Rab escort protein 66, 74 Rabs family of proteins 66 Ras 25, 66–69, 79, 106, 254, 255, 257, 258, 438, 677 Reassembly 236, 288, 309, 339, 581–583, 590, 591, 594–599 Regulatory secretory proteins (RSPs) 9, 383, 485–487, 489–499, 502 Retrograde recycling routes 460 Retrograde routes 459, 460, 461, 468 Retrograde transport 26, 32, 94– 98, 251, 258, 259, 333, 336, 347, 351, 358, 360, 362, 364, 366–369, 396, 397, 405, 426, 427, 428, 436, 441, 443, 459, 460, 466–469, 501, 581, 595, 653 Role of cargoes 28 Role of COPI for endosome function 481 Role of COPI vesicles 24, 88, 94, 346, 353, 696 RSP polymerization 496

S Saccharomyces cerevisiae 22, 66, 79, 191, 223, 270, 233, 363, 430, 475, 580, 584, 588, 623, 632, 679, 683 Sar1 79–83, 106, 107, 112, 138, 337, 432, 613–615, 624, 650, 655, 656, 661, 663, 677, 686, 687 Sec12 79–82, 342, 614, 624, 661

*

713

Sec13/31 79–83, 337 SEC7 22, 91, 342, 623, 624, 632, 635–638, 640, 642 SEC14 623, 624, 638–640, 642 Sec16 80, 81, 334, 624, 626 Sec23/24 79–83 SERCA 147–149, 255 Secretory pathway 3, 16, 21, 23, 25, 30, 32, 59, 66, 70, 72, 127, 133, 143, 147, 154, 161, 163, 173, 177, 199, 207, 211–213, 254, 270, 283, 284, 306, 314, 315, 325, 333, 339, 346, 347, 363, 376, 402, 404, 405, 406, 409, 431, 436, 477, 485, 486, 499, 500, 501, 503, 504, 506, 524, 535, 536, 546, 558, 559, 632–636, 638, 643, 644, 651–653, 657, 677, 682, 696 Shiga toxin 68, 144, 155, 256, 362, 367, 368, 426, 428–430, 441, 442, 463, 466 Sialylation 20, 125, 171, 191, 200, 544, 549 Signalling platform 247–249 Small G proteins 360–363, 366, 369 SNARE 3, 7–9, 23, 24, 31, 43–59, 66, 72, 73, 98, 121, 122, 126, 130, 134– 139, 143, 151–153, 234, 235, 237, 239, 240, 259, 275, 326, 338, 339, 342, 346–348, 350, 351, 353, 354, 360, 361, 367, 368, 376, 378, 384, 432, 433, 435–439, 441–444, 460, 478, 487, 488, 503, 504, 529, 573, 574, 595–597, 615, 618, 623, 624, 626, 661–663, 677, 686, 696 SNARE-motif 44, 47, 50, 53, 56, 58, 59 i-SNAREs 53 Qa-SNAREs 43, 44, 47, 49, 52, 56 Qb-SNAREs 24, 44, 47 Qc-SNAREs 44, 47, 49 t-SNAREs 47, 50, 360, 460 v-SNAREs 43, 47, 49, 443, 488 Sorting 1, 2, 8, 25, 28, 47, 53, 55, 59, 67, 87, 90, 92, 95, 96, 99, 110–112,

714

*

Index

126, 153, 167, 177, 199, 207, 209, 211, 212, 214, 216, 217, 234, 238, 247, 260, 281, 283, 284, 288, 311, 321, 326, 336, 351, 359, 360, 375, 382, 388, 389, 390–392, 396, 397, 402, 406–408, 418, 420, 425, 427, 428, 431, 436, 437, 444, 445, 460, 461, 466, 481, 488, 493–499, 501, 502, 505, 506, 520, 521, 525, 535, 552, 553, 556, 558, 563–567, 569, 571, 572, 574, 598, 619, 650, 653, 654, 657, 675, 679, 682–687 Sorting of RSPs 493 SPCA 147–149, 151, 254, 255, 256 Spectrin 225, 253, 279, 286, 310, 325 Spondyloepiphyseal dysplasia 137, 173 Stacking 22, 229–231, 236, 237, 580, 582–585, 592–594, 596, 598, 625, 677, 695 Stacks in G2 cells 225 Stem domain 165 Stimulation of exocytosis 475 Stimulation of endocytosis 475 Storage/secretory granules (SGs) 22, 26, 351, 384, 485–491, 493, 495, 497–500, 502–505 Structure of GPCs 376, 478 Structure of Rabs 68 Structure of the Golgi exit site 21, 25, 384 Sulfotransferase 173, 174, 190

T Tethering 31, 44, 50, 55–57, 72, 83, 96, 120–122, 126, 130, 134–139, 227, 228, 235, 236, 338, 342, 344, 346, 352, 358, 360, 363, 368, 382, 432, 433, 435, 436, 439, 441–444, 522, 574, 580, 582, 586, 594–598, 612, 614, 625, 676, 677, 687 The “sorting by retention” model 499

Tn-syndrome 175, 176 Topography of glycosylation enzymes 166 Traffic 163, 164, 170, 171, 177–180, 215, 359, 459, 564, 565, 569, 571, 572, 574 Transbilayer area asymmetry 317, 327 TRANSCET 26, 27, 29, 32 Trans-Golgi 19, 26–28, 52, 53, 112, 136, 147, 150, 168, 170, 171, 173, 177–180, 209, 213, 215, 217, 228, 232, 238, 240, 241, 254, 280, 325, 347, 377, 404, 405, 433, 435, 441, 444, 460, 462–464, 467, 468, 486, 490, 502, 521, 523, 525, 527, 529, 543, 545, 546, 556, 616, 618, 630, 655, 657, 659, 660, 662 Trans-Golgi Network (TGN) 3, 18, 25, 26, 32, 71, 108, 144, 146, 155, 215, 228, 241, 251, 275, 280, 301, 302, 347, 354, 358, 388, 402, 405, 419, 420, 426, 429, 459, 460, 461, 464, 467, 475, 476, 481, 486, 487, 506, 529, 552, 564, 594, 618, 640, 649, 656, 657, 662 Transitional ER 81, 333, 588, 626, 650 Transport 1–3, 7–11, 17, 18, 21, 25, 26, 28–32, 43, 46, 47, 49–52, 55–58, 66, 68, 70–72, 74, 78, 81, 87–99, 109, 120, 122, 126, 130, 132–139, 143–150, 152–154, 164, 167, 168, 176–178, 181, 190–201, 207, 209, 211–217, 224, 225, 229–231, 234– 238, 240, 251, 252, 254, 255–261, 271, 273–275, 279–288, 302, 305– 307, 309–311, 324, 333–339, 342, 343, 346–354, 358–360, 362–364, 366–369, 375–377, 379, 381–384, 388, 389, 393, 394, 396, 397, 402, 403, 405, 406, 408, 409, 414, 417– 420, 425–428, 430–446, 459–461, 463, 466–469, 475–481, 485, 487, 489–491, 493, 494, 496, 500–503,

Index

505, 506, 520, 527, 535, 537, 539, 544, 554, 555, 558, 565, 565–567, 569, 573, 580, 581, 583, 585, 590, 595, 596, 599, 611, 614, 615, 617, 618, 625, 631, 632, 635, 641, 642, 647, 650–657, 660–662, 686, 687, 695, 696 Transport models 11, 32, 236, 334, 544 Transport of secretory proteins through endosomes 476 Transmembrane domain 1, 43, 50, 53, 56–58, 161, 165, 178, 228, 233, 234, 256, 260, 324, 327, 338, 390, 416, 418, 420, 571, 613, 654 TRAPP 3, 44, 130–139, 235, 436, 439, 442–444, 624, 677 Tubules 16–18, 21, 24–27, 30, 82, 93, 95, 154, 224–226, 230, 234, 251, 273, 282, 286, 301, 302, 305, 307, 309, 323, 324, 337, 338, 343, 348, 349, 352, 353, 359, 360, 367, 375, 377, 379, 380, 384, 431, 461, 478– 480, 487, 492, 526, 528, 554, 572, 593, 595, 611–613, 618, 624, 631– 643, 660, 663, 685, 686 Two transport steps during Golgi-toPM transport 478

U Ubiquitination 58, 594, 597, 598 Uncovering enzyme 168, 214, 404, 405 UDPGlcNAc: lysosomal enzyme-1phosphotransferase 168 UDPGalNAc: polypeptide GalNAc-Ts 174

*

715

144–146, 151–154, 177, 180, 191, 193, 195, 197, 199, 210, 211, 212, 215, 216, 233–236, 252, 273, 275, 281, 283–286, 301, 302, 304, 305, 307, 311, 314, 316, 317, 319, 325, 326, 333, 334–339, 342–354, 359, 360, 375–378, 380, 388, 389, 392, 395, 405, 408, 414, 148, 420, 425, 426, 431–433, 435, 439, 441–444, 461, 467, 475, 487, 491, 501–503, 505, 543, 552, 567, 572–574, 580– 583, 585, 586, 588, 590–592, 594, 595, 597, 598, 611, 614, 617, 618, 626, 629–632, 635, 642, 643, 650– 655, 660, 661, 663, 686, 695, 696 Vesicle fusion 55, 72, 132, 133, 139, 574, 594, 655, 677 Vesicle tethering 44, 55, 56, 121, 126, 134, 135, 138, 586, 596, 625 Vesicle scission 82, 307 Vesicle transport 74, 274, 650, 651 Vesicular model 8, 10, 30–32, 235, 236, 238, 342, 343, 344, 354, 375, 376, 544, 631 Vesicular–tubular clusters 16, 143, 650 Visualization of GPC 380 Von Willebrand factor 520, 524, 529, 551 VTCs 16, 143, 334, 336–338, 491

W Weibel-Palade bodies 488, 524, 527, 529, 551 Wheat germ agglutinin 461, 467, 529

Y V Vesicles 2, 8–11, 16, 18, 19, 21, 23, 24, 27, 29–32, 43, 46, 49, 52, 53, 56– 59, 67, 71, 72, 78–83, 87, 88, 89, 92– 98, 120, 122, 126, 127, 133–139,

Yeast 2, 3, 8, 10, 18, 19, 22, 28, 43, 45–47, 49, 51–58, 67, 72, 73, 78, 79, 81, 83, 91, 94, 97, 98, 107, 113, 120– 124, 126, 127, 130–139, 145, 148, 152, 177, 191–193, 195, 197, 199,

716

*

Index

209–211, 224, 228, 237, 253, 256, 270, 283, 284, 288, 302, 334, 336, 342, 346, 348, 363–365, 402, 427, 430, 431, 434, 435–446, 477, 479, 480, 485, 498, 502, 567, 574, 580, 588, 597, 619, 623–626, 629, 630,

632–637, 640–644, 647, 653, 658, 660–662, 679, 680, 683, 684

Z Zone of exclusion

19