Rice biology in the genomics era (Biotechnology in Agriculture and Forestry, Vol. 62)

Biotechnology in Agriculture and Forestry Volume 62 Series Editors Prof. Dr. Toshiyuki Nagata (Managing Editor) IT Res...

Author: Hiro-Yuki Hirano | Atsushi Hirai

164 downloads 1158 Views 4MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Biotechnology in Agriculture and Forestry Volume 62

Series Editors Prof. Dr. Toshiyuki Nagata (Managing Editor) IT Research Center, Hosei University, 2-17-1 Fujimicho, Chiyoda-ku, Tokyo 102-8160, Japan; Graduate School of Science, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan Prof. Dr. Horst Lörz Biozentrum Klein Flottbek, Molekulare Phytopathologie und Genetik, Universität Hamburg, Ohnhorststr. 18, 22609 Hamburg, Germany Prof. Dr. Jack M. Widholm 285A E.R. Madigan Laboratory, Department of Crop Sciences, University of Illinois, 1201 W. Gregory, Urbana, IL 61801, USA

Biotechnology in Agriculture and Forestry Volume 29: Plant Protoplasts and Genetic Engineering V (1994) Volume 30: Somatic Embryogenesis and Synthetic Seed I (1995) Volume 31: Somatic Embryogenesis and Synthetic Seed II (1995) Volume 32: Cryopreservation of Plant Germplasm I (1995) Volume 33: Medicinal and Aromatic Plants VIII (1995) Volume 34: Plant Protoplasts and Genetic Engineering VI (1995) Volume 35: Trees IV (1996) Volume 36: Somaclonal Variation in Crop Improvement II (1996) Volume 37: Medicinal and Aromatic Plants IX (1996) Volume 38: Plant Protoplasts and Genetic Engineering VII (1996) Volume 39: High-Tech and Microprogation V (1997) Volume 40: High-Tech and Microprogation VI (1997) Volume 41: Medicinal and Aromatic Plants X (1998) Volume 42: Cotton (1998) Volume 43: Medicinal and Aromatic Plants XI (1999) Volume 44: Transgenic Trees (1999) Volume 45: Transgenic Medicinal Plants (1999) Volume 46: Transgenic Crops 1I (1999) Volume 47: Transgenic Crops II (2001) Volume 48: Transgenic Crops III (2001) Volumes 1-48 were edited by Y.P.S. Bajaj† Volume 49: Somatic Hybridization in Crop Improvement II (2001) T. Nagata and Y.P.S. Bajaj (Eds.) Volume 50: Cryopreservation of Plant Germplasm II (2002) L.E. Towill and Y.P.S. Bajaj (Eds.) Volume 51: Medicinal and Aromatic Plants XII (2002) T. Nagata and Y. Ebizuka (Eds.) Volume 52: Brassicas and Legumes: From Genome Structure to Breeding (2003) T. Nagata and S. Tabata (Eds.) Volume 53: Tobacco BY-2 Cells (2004) T. Nagata, S. Hasezawa, and D. Inzé (Eds.) Volume 54: Brassica (2004) E.C. Pua and C.J. Douglas (Eds.) Volume 55: Molecular Marker Systems in Plant Breeding and Crop Improvement (2005) H. Lörz and G.Wenzel (Eds.) Volume 56: Haploids in Crop Improvement II (2005) C.E. Palmer,W.A. Keller, and K.J. Kasha (Eds.) Volume 57: Plant Metabolomics (2006) K. Saito, R.A. Dixon, and L.Willmitzer (Eds.) Volume 58: Tobacco BY-2 Cells: From Cellular Dynamics to Omics (2006) T. Nagata, K. Matsuoka, and D. Inzé (Eds.) Volume 59: Transgenic Crops IV (2007) E.C. Pua and M.R. Davey (Eds.) Volume 60: Transgenic Crops V (2007) E.C. Pua and M.R. Davey (Eds.) Volume 61: Transgenic Crops VI (2007) E.C. Pua and M.R. Davey (Eds.) Volume 62: Rice Biology in the Genomics Era (2008) H.-Y. Hirano, A. Hirai, Y. Sano and T. Sasaki (Eds.)

Hiro-Yuki Hirano • Atsushi Hirai • Yoshio Sano Takuji Sasaki Editors

Rice Biology in the Genomics Era

Prof. Dr. Hiro-Yuki Hirano Graduate School of Science University of Tokyo Hongo, Bunkyo-ku Tokyo 113-8654 Japan [email protected]

Prof. Dr. Atsushi Hirai Faculty of Agriculture Meijo University Shiogamaguchi, Tenpaku-ku Nagoya 468-8502 Japan [email protected]

Prof. Dr. Yoshio Sano Graduate School of Agriculture Hokkaido University Kita 9, Nishi 9 Sapporo 060-8589 Japan [email protected]

Dr. Takuji Sasaki National Institute of Agrobiological Sciences 1-2, Kannondai 2-chome, Tsukuba Ibaraki 305-8602 Japan [email protected]

ISBN 978-3-540-74248-7

e-ISBN 978-3-540-74250-0

Biotechnology in Agriculture and Forestry ISSN 0934-943X Library of Congress Control Number: 2007938410 © Springer-Verlag Berlin Heidelberg 2008 This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, roadcasting, reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer. Violations are liable to prosecution under the German Copyright Law. The use of general descriptive names, 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. Cover Design: WMXDesign GmbH, Heidelberg, Germany Printed on acid-free paper 9 8 7 6 5 4 3 2 1 springer.com

Preface

Rice (Oryza sativa) is one of the most important staple food crops in the world. Breeding efforts to improve the agronomical quality of rice have been conducted, and studies on rice from the viewpoint of basic biological interest have also been carried out. In 1991, a book entitled Rice (edited by Dr. Bajaj) was published as the 14th volume in the series Biotechnology in Agriculture and Forestry (BAF), detailing rice research activities at that time, and focusing mainly on cell and tissue culture and genetic variability. Studies on rice have fundamentally advanced since then, whose outcomes are mentioned below. This is a good reason to compile a new volume on rice. The situation regarding rice research has markedly changed in the last 16 years. First, the genomic sequences of rice were completely determined by the International Rice Genome Sequencing Project in 2004. Since the genome sequence of Arabidopsis thaliana had been determined in 2000, rice became the second species in the seed plants to have its genome well understood. Second, the technology to transform rice by the Agrobacterium-mediated method was developed and is now established. In classical phytopathology, Poaceae (including rice) has not been considered as a host for Agrobacterium. This transformation method is relatively easy and reproducible as compared to conventional transformation methods using protoplasts, and is now widely used in rice research. As a result, the Agrobacterium-mediated transformation method allows rice researchers to analyze the functions of the gene of interest by producing transgenic rice. Third, molecular techniques and tools to isolate genes and to analyze their functions have become available. Many important genes have been isolated by map-based cloning methods as well as tagging methods using endogenous transposons and exogenous elements. In addition, reverse genetic studies have become possible by screening of knockout lines, by making knockdown lines and by using the TILLING (targeting induced local lesions in genomes) method. Genome-wide expression analyses such as microarray and MPSS (massively parallel signature sequencing) for rice research have also developed and will provide a wealth of information in rice in the form of functional genomic studies, as in other model organisms. Such technical advances promote rice from a local experimental material to a model organism not only for applied research to improve crops, but also for basic research to understand molecular and cellular activities in monocots. As a result, many excellent studies that uncover critical biological function have been published v

vi

Preface

recently and some of them have had a huge impact on plant sciences. For example, the receptor for gibberellin (GA), a key phytohormone, was discovered in rice research, and studies on GA signaling in rice have advanced much more than in Arabidopsis. A large number of other basic studies to reveal gene function and evolution of rice and applied studies to improve the yield and quality of rice are now being developed. Considering these situations, we have compiled a volume on the recent advances in rice research, many of which have been greatly facilitated by the information derived from the complete sequence of the rice genome. Therefore, this book is a completely new edition on rice in the genomics era. It is comprised of four sections and 26 chapters. Each chapter is written by expert scientists in their respective fields of rice research. The grass family includes many important crops other than rice, such as wheat, maize and barley, providing most of the staple food for human-beings in the world. The alignment of the cereal genome showed extensive conservation of gene order or synteny among grasses. This syntenic relationship has already been exploited in map-based cloning of agronomic genes using homologues in rice. Therefore, results from rice research will have a great impact on other cereal crops as well. According to information from the series editors of BAF, a new volume on maize is being prepared, which would complement the contents of this volume. We hope that this book will become an valuable resource for fundamental and applied research in rice as well as a reference for broad-spectrum research in other plants. Tokyo, October 2007

Hiro-Yuki Hirano Atsushi Hirai Yoshio Sano Takuji Sasaki

Contents

Section I I.1

I.2

I.3

Genome-wide and Genome-based Research

The Rice Genome Sequence as an Indispensable Tool for Crop Improvement ................................................................. Takuji Sasaki, Jianzhong Wu, Hiroshi Mizuno, Baltazar A. Antonio, and Takashi Matsumoto

3

1 Introduction ....................................................................................... 2 Major Features of the Rice Genome ................................................. 3 Comparative Analysis of Genome Structure Within Oryza sativa ... 4 Comparative Analysis of Genome Structure in the Genus Oryza ..... 5 Comparative Analysis of Genome Structure in the Grass Family .... 6 Conclusion......................................................................................... References ...............................................................................................

3 4 6 7 9 10 11

Bioinformatics and Database of the Rice Genome ............................. Hisataka Numa, Tsuyoshi Tanaka, and Takeshi Itoh

13

1 Introduction ....................................................................................... 2 Gene-ﬁnding in the Rice Genome ..................................................... 3 Comparison of Gene Prediction Programs........................................ 4 Prediction of Gene Functions ............................................................ 5 The Rice Annotation Project Database ............................................. References ...............................................................................................

13 13 14 16 17 20

Sequencing-based Measurements of mRNA and Small RNA ........... Kan Nobuta and Blake C. Meyers

23

1 Introduction ....................................................................................... 2 mRNA MPSS Data............................................................................ 3 Small RNA MPSS ............................................................................. 4 Conclusion......................................................................................... References ...............................................................................................

23 25 32 34 35

vii

viii

I.4

Contents

Microarray-based Approaches to Rice Transcriptome Analysis ...... Lei Li and Xing Wang Deng

37

1 2 3

37 37

Introduction ....................................................................................... Microarray Platforms and Methods .................................................. Applications of Microarrays in Rice Transcriptome Analysis..................................................................... 4 Technical Challenges and Resources for Using Rice Microarray Systems .......................................................................... 5 Future Perspectives ........................................................................... References ............................................................................................... I.5

High-throughput Transcriptome Analysis in Rice from a Genomic Perspective................................................................. Shoshi Kikuchi 1 Introduction ....................................................................................... 2 The Rice Transcriptome: From EST Collection to cDNA-based Microarray Systems................................................. 3 Rice Full-length cDNA Clone Collection Project ............................. 4 Oligoarray Systems ........................................................................... 5 Gene Family Analysis ....................................................................... 6 Future Perspectives ........................................................................... References ...............................................................................................

I.6

I.7

39 45 48 49

53 53 53 55 61 63 64 65

Active Transposons in Rice ................................................................... Tetsuya Nakazaki, Ken Naito, Yutaka Okumoto, and Takatoshi Tanisaka

69

1 Introduction ....................................................................................... 2 Detection of Transposon Sequences in Rice ..................................... 3 Active Transposons of Rice .............................................................. 4 Conclusion......................................................................................... References ...............................................................................................

69 70 72 77 77

Homologous Recombination-dependent Gene Targeting and an Active DNA Transposon nDart-promoted Gene Tagging for Rice Functional Genomics ............................................... Yasuyo Johzuka-Hisatomi, Masahiko Maekawa, Kyoko Takagi, Chang-Ho Eun, Takaki Yamauchi, Zenpei Shimatani, Nisar Ahmed, Hiroko Urawa, Kazuo Tsugane, Rie Terada, and Shigeru Iida 1 Introduction ....................................................................................... 2 Homologous Recombination-dependent Gene Targeting ................. 3 Active DNA Transposon nDart-promoted Gene Tagging................. 4 Current Status and Future Prospects ................................................. References ...............................................................................................

81

81 84 89 91 92

Contents

I.8

I.9

ix

T-DNA Tagging Lines ............................................................................ Gynheung An

95

1 Introduction ....................................................................................... 2 Generation of T-DNA Insertional Mutants ....................................... 3 Generation of Activation-tagging Mutants........................................ 4 Entrapment Tagging .......................................................................... 5 DNA Pool Screening for Reverse Genetics....................................... 6 Establishment of Flanking Sequence Tag (FST) Databases.............. 7 Obstacles in T-DNA Insertional Mutagenesis ................................... 8 Future Prospects ................................................................................ References ...............................................................................................

95 95 96 98 99 100 103 103 104

Frequent DNA Transfer Among Mitochondrial, Plastid and Nuclear Genomes of Rice During Evolution .................. Mikio Nakazono and Atsushi Hirai

107

1 2 3

Introduction ....................................................................................... Plastid-to-Mitochondrion DNA Transfer Events .............................. Mitochondrion-to-Nucleus or Plastid-to-Nucleus DNA Transfer Events ................................................................................................ 4 DNA Segments Found Commonly in all Three Genomes ................ 5 Nucleus-to-Mitochondrion DNA Transfer Events ............................ 6 Conclusions and Future Perspectives ................................................ References ............................................................................................... Section II II.1

II.2

107 109 110 113 114 114 115

Signal Transduction and Development

Hormonal Signal Transduction in Rice ............................................. Ayako Nakamura and Makoto Matsuoka

121

1 Introduction...................................................................................... 2 GA Signaling in Rice ....................................................................... 3 Brassinosteroid Signaling in Rice.................................................... 4 Auxin Signaling in Rice................................................................... 5 Future Perspectives .......................................................................... References..............................................................................................

121 122 126 129 131 132

Rice Heterotrimeric G Protein Signaling .......................................... Yukimoto Iwasaki, Hisaharu Kato, Yukiko Fujisawa, and Katsuyuki Oki

135

1 Introduction...................................................................................... 2 Genes and Translation Products for Rice Heterotrimeric G Proteins......................................................................................... 3 Mutants with Defects in the Subunit genes Encoding Plant Heterotrimeric G Proteins .......................................................

135 135 138

x

Contents

4

Analysis of Signaling Mediated by Rice Heterotrimeric G Proteins Using a Constitutively Active Form of the a Subunit ............................................................................... 5 Molecules Interacting with Arabidopsis Gα ................................... 6 Signaling Pathways Mediated by Plant Heterotrimeric G Proteins......................................................................................... 7 Further Perspectives ......................................................................... References.............................................................................................. II.3

II.4

141 145 146

Genetic Control of Embryogenesis in Rice ........................................ Yutaka Sato

149

1 Introduction...................................................................................... 2 Molecular Markers Used in the Analysis of Rice Embryogenesis .. 3 Analysis of Organogenesis During Embryogenesis ........................ 4 Future Perspectives .......................................................................... References..............................................................................................

149 152 153 158 158

Photoperiodic Flowering in Rice ........................................................ Takeshi Izawa

163

1 2

163

Introduction ..................................................................................... The Importance of Photoreception for Photoperiodic Flowering in Rice ............................................................................ 3 Quantitative Trait Loci Analyses Reveal the Conserved Floral Pathway ................................................................................. 4 A Unique Floral Pathway Exists in Rice ......................................... 5 The Importance of Floral Inhibition Control Under Long-day Conditions........................................................................................ 6 Night-break Experiments in Rice .................................................... 7 The Roles of Circadian Clocks in Floral Induction in Rice Are Still Unknown .............................................................. 8 Differences Between Long-day and Short-day Plants ..................... 9 Future Perspectives .......................................................................... References ............................................................................................. II.5

138 139

164 165 166 167 169 170 171 173 174

Genetic Regulation of Meristem Maintenance and Organ Speciﬁcation in Rice Flower Development ..................... Hiro-Yuki Hirano

177

1 Introduction...................................................................................... 2 Structure of the Inﬂorescence and Flower in Rice........................... 3 Floral Meristem ............................................................................... 4 Floral Organ Development .............................................................. 5 Future Perspectives .......................................................................... References..............................................................................................

177 178 179 183 186 187

Contents

II.6

II.7

xi

Genetic Dissection of Sexual Reproduction in Rice (Oryza sativa L.).................................................................................... Ken-Ichi Nonomura and Shinichiro Yamaki

191

1 Introduction...................................................................................... 2 Reproductive Organ Development................................................... 3 Sporogenesis .................................................................................... 4 Meiosis............................................................................................. 5 Gametogenesis ................................................................................. 6 Future Perspectives on Studies in Rice Sexual Reproduction ......... References..............................................................................................

191 192 195 196 200 201 202

Molecular Studies on Cytoplasmic Male Sterility-associated Genes and Restorer Genes in Rice ..................................................... Sota Fujii, Tomohiko Kazama, and Kinya Toriyama

205

1 Introduction...................................................................................... 2 Types of CMS in Rice...................................................................... 3 Mitochondrial Chimeric Genes........................................................ 4 Molecular Cloning of Rf1 and Molecular Mapping of Other Rf Genes............................................................................. 5 Approaches Toward Understanding the Mechanism of CMS Induction ............................................................................ 6 Conclusion ....................................................................................... References.............................................................................................. Section III III.1

III.2

205 205 208 210 213 214 214

Evolution and Ecology

Phylogeny and Biogeography of the Genus Oryza........................... Duncan A. Vaughan, Song Ge, Akito Kaga, and Norihiko Tomooka

219

1 Introduction .................................................................................... 2 Phylogeny ....................................................................................... 3 Oryza Biogeography ...................................................................... References ............................................................................................

219 220 225 231

Chromosome and Genome Evolution in Rice .................................. Nori Kurata

235

1 2 3 4 5

235 235 237 239

6

Introduction .................................................................................... Chromosomes in the Different Genomes ....................................... Classiﬁcation of Genomes Based on Chromosome Pairing ........... Genome Duplication in the Genus Oryza ...................................... Genome-wide Evolution Among Oryza Species: Genome Size, Retrotransposon and Centromere ........................... Regional Genome Evolution Among Oryza Species .....................

240 241

xii

III.3

III.4

III.5

Contents

7 Genome Evolution in the Poaceae Family ..................................... References ............................................................................................

242 243

Genetics of Speciation in Rice ........................................................... Yohei Koide, Kazumitsu Onishi, Akira Kanazawa, and Yoshio Sano

247

1 Introduction .................................................................................... 2 Genetic Mechanisms of Post-mating Barriers................................ 3 Evolution of Reproductive Isolation .............................................. 4 Conclusions and Perspectives ........................................................ References ............................................................................................

247 248 254 256 256

Genetic Diversity in Wild Relatives of Rice and Domestication Events ................................................................. Hong-Wei Cai, Masahiro Akimoto, and Hiroko Morishima

261

1 Introduction .................................................................................... 2 Genetic Diversity of AA-genome Wild Taxa ................................. 3 Differentiation Within O. ruﬁpogon ............................................... 4 Genetics of Domestication ............................................................. References ............................................................................................

261 262 265 268 273

Rice Retroposon, p-SINE, and Its Use for Classiﬁcation and Identiﬁcation of Oryza Species ................................................... Hisako Ohtsubo, Suguru Tsuchimoto, Jian-Hong Xu, Chaoyang Cheng, Marcia Y. Koudo, Nori Kurata, and Eiichi Ohtsubo 1 Introduction .................................................................................... 2 p-SINE Families in the Oryza Genus ............................................. 3 Phylogenetic Analysis of Oryza Species Based on SINE Insertion Analysis ............................................................ 4 “SINE Code” Is an Excellent Tool for Identiﬁcation of Oryza Species............................................................................. 5 Final Remarks ................................................................................ 6 Future Perspectives ......................................................................... References ............................................................................................

Section IV IV.1

277 278 280 285 288 289 289

Improvement of Rice

Detection and Molecular Cloning of Genes Underlying Quantitative Phenotypic Variations in Rice..................................... Toshio Yamamoto and Masahiro Yano 1 2

277

Introduction .................................................................................... Molecular Cloning of Major QTLs with Agronomic Values ..........................................................................

295 295 295

Contents

3 Identiﬁcation of Major QTLs with Agronomic Value.................... 4 Systematic Genetic and Molecular Analysis and Utilization of Natural Variations ...................................................................... 5 Future Prospects ............................................................................. References ............................................................................................ IV.2

xiii

299 302 305 306

Rice Yielding and Plant Hormones .................................................. Motoyuki Ashikari and Tomoaki Sakamoto

309

1 Introduction .................................................................................... 2 Breeding of Semi-dwarf Rice Varieties Led to Dramatically Increased Grain Production ............................................................ 3 Gibberellins Regulate Plant Height ................................................ 4 Brassinosteroids Control Leaf Angle and Biomass Production ...................................................................................... 5 Cytokinins Regulate Grain Number ............................................... 6 Future Perspectives: Accumulation of Yielding-related Genes for Breeding......................................................................... References ............................................................................................

309

IV.3 Regulation of Iron and Zinc Uptake and Translocation in Rice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Takanori Kobayashi and Naoko K. Nishizawa Iron (Fe) and Zinc (Zn) Uptake from the Rhizosphere Mediated by Mugineic Acid Family Phytosiderophores (MAs) .............................................................. 2 Uptake of Fe2+ and Zn2+ Ions from the Rhizosphere ...................... 3 Fe and Zn Translocation Inside the Plant Body ............................. 4 Regulation of Expression of Fe Uptake-related Genes .................. 5 Future Perspectives ........................................................................ References ............................................................................................

310 310 314 315 316 318

321

1

IV.4

321 326 327 330 332 332

Abiotic Stress ...................................................................................... Takayuki Ohnishi, Mikio Nakazono, and Nobuhiro Tsutsumi

337

1 Introduction .................................................................................... 2 Signal Transduction........................................................................ 3 Particular Stresses .......................................................................... 4 Future Perspectives ........................................................................ References ............................................................................................

337 337 345 351 352

xiv

IV.5

Contents

Health-promoting Transgenic Rice: Application of Rice Seeds as a Direct Delivery System for Bioactive Peptides in Human Health................................................................. Fumio Takaiwa, Lijun Yang, and Hiroshi Yasuda 1 2

Introduction .................................................................................... Plant Production System as a Bioreactor for Foreign Recombinant Proteins .................................................................... 3 Characteristics of the Rice Seed Production System as a Bioreactor ................................................................................ 4 Production Tools Required for Expression of Recombinant Proteins or Peptides in Rice Seed ................................................... 5 Creation of Functional Foods that Contribute to Human Health ............................................................................ References ............................................................................................ Index ................................................................................................................

357 357 358 358 359 363 371 375

Contributors

Nisar Ahmed Research Institute for Bioresources, Okayama University, Kurashiki 710-0046, Japan, e-mail: [email protected] Masahiro Akimoto School of Agriculture, Obihiro University of Agriculture and Veterinary Medicine, Inada-cho, Obihiro 080-8555, Japan, e-mail: [email protected] Gynheung An National Research Laboratory of Plant Functional Genomics, Division of Molecular and Life Sciences, Pohang University of Science and Technology (POSTECH), Pohang 790-784, Korea, e-mail: [email protected] Baltazar A. Antonio National Institute of Agrobiological Sciences, 1-2, Kannondai 2-chome, Tsukuba, Ibaraki 305-8602, Japan Motoyuki Ashikari Bioscience and Biotechnology Center, Nagoya University, Furoucho, Chikusaku, Nagoya, Aichi 464-8601 Japan, e-mail: [email protected] Hong-Wei Cai College of Agronomy and Biotechnology, China Agricultural University, Yuanmingyuan West Road, Beijing 100094 China, e-mail: [email protected] Chaoyang Cheng Institute of Molecular and Cellular Biosciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-0032, Japan Xing Wang Deng Department of Molecular, Cellular and Developmental Biology, Yale University, New Haven, Connecticut 06520, USA, e-mail: [email protected] Chang-Ho Eun National Institute for Basic Biology, Okazaki 444-8585, Japan, e-mail: [email protected]

xv

xvi

Contributors

Sota Fujii Graduate School of Agricultural Science, Tohoku University, Sendai 981-8555, Japan, e-mail: [email protected] Yukiko Fujisawa Department of Bioscience, Fukui Prefectural University, 4-1-1 Kenjyojima, Matsuoka, Eiheiji-cho, Yoshida-gun, Fukui 910-1195, Japan Song Ge Key State Laboratory of Systematic and Evolutionary Botany, Institute of Botany, Chinese Academy of Sciences, Beijing 100093, China, e-mail: [email protected] Atsushi Hirai School of Agriculture, Meijo University, 1-501 Shiogamaguchi, Tenpaku, Nagoya, Aichi 468-8502, Japan, e-mail: [email protected] Hiro-Yuki Hirano Graduate School of Science, The University of Tokyo, Hongo, Tokyo 113-8654, Japan, e-mail: [email protected] Shigeru Iida National Institute for Basic Biology, Okazaki 444-8585, Japan, e-mail: [email protected] and Department of Basic Biology, School of Life Science, The Graduate University for Advanced Studies, Okazaki 444-8585, Japan Takeshi Itoh Division of Genome and Biodiversity Research, National Institute of Agrobiological Sciences, 2-1-2 Kannondai, Tsukuba, Ibaraki 305-8602, Japan, e-mail: [email protected] Yukimoto Iwasaki Department of Bioscience, Fukui Prefectural University, 4-1-1 Kenjyojima, Matsuoka, Eiheiji-cho, Yoshida-gun, Fukui 910-1195, Japan, e-mail: [email protected] Takeshi Izawa National Institute of Agrobiological Sciences, Tsukuba, Japan, e-mail: [email protected] Yasuyo Johzuka-Hisatomi National Institute for Basic Biology, Okazaki 444-8585, Japan, e-mail: [email protected] Akito Kaga National Institute of Agrobiological Sciences, 2-1-2 Kannondai, Tsukuba, Ibaraki 305-8602, Japan, e-mail: [email protected]

Contributors

xvii

Akira Kanazawa Laboratory of Cell Biology and Manipulation, Graduate School of Agriculture, Hokkaido University, Kita-9, Nishi-9, Kita-ku, Sapporo 060-8589, Japan, e-mail: [email protected] Hisaharu Kato Department of Bioscience, Fukui Prefectural University, 4-1-1 Kenjyojima, Matsuoka, Eiheiji-cho, Yoshida-gun, Fukui 910-1195, Japan Tomohiko Kazama Graduate School of Life Science, Tohoku University, Sendai 980-8577, Japan, e-mail: [email protected] Shoshi Kikuchi Plant Genome Research Unit, Division of Genome and Biodiversity Research, National Institute of Agrobiological Sciences (NIAS), Kan’non dai 2-1-2, Tsukuba Ibaraki 305-8602, Japan, e-mail: [email protected] Takanori Kobayashi Graduate School of Agricultural and Life Sciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan, e-mail: [email protected] and Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Corporation, Japan Yohei Koide Laboratory of Plant Breeding, Graduate School of Agriculture, Hokkaido University, Kita-9, Nishi-9, Kita-ku, Sapporo 060-8589, Japan, e-mail: [email protected] Marcia Y. Koudo Institute of Molecular and Cellular Biosciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-0032, Japan Nori Kurata National Institute of Genetics, Yata 1111, Mishima, Japan 411-8540, e-mail: [email protected] Lei Li Department of Molecular, Cellular and Developmental Biology, Yale University, New Haven, Connecticut 06520, USA, e-mail: [email protected] Masahiko Maekawa Research Institute for Bioresources, Okayama University, Kurashiki 710-0046, Japan, e-mail: [email protected] Takashi Matsumoto National Institute of Agrobiological Sciences, 1-2, Kannondai 2-chome, Tsukuba, Ibaraki 305-8602, Japan

xviii

Contributors

Makoto Matsuoka Bioscience and Biotechnology Center, Nagoya University, Nagoya, Aichi 464-8601, Japan, e-mail: [email protected] Blake C. Meyers Delaware Biotechnology Institute, University of Delaware, 15 Innovation Way, Room 230, Newark, DE 19711, USA, e-mail: [email protected] Hiroshi Mizuno National Institute of Agrobiological Sciences, 1-2, Kannondai 2-chome, Tsukuba, Ibaraki 305-8602, Japan Hiroko Morishima Saiwai-cho 15-2, Hiratsuka 254-0804, Japan, e:mail: [email protected] Ken Naito Plant Biology Department, University of Georgia, 4505 Miller Plant Sciences Building, Athens, GA, 30602, USA e-mail: [email protected] Ayako Nakamura Bioscience and Biotechnology Center, Nagoya University, Nagoya, Aichi 464-8601, Japan, e-mail: [email protected] Tetsuya Nakazaki Graduate School of Agriculture, Kyoto University, Kitashirakawa, Sakyoku, Kyoto 606-8502, Japan, e-mail: [email protected] Mikio Nakazono Laboratory of Plant Molecular Genetics, Graduate School of Agricultural and Life Sciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan, e-mail: [email protected] Naoko K. Nishizawa Graduate School of Agricultural and Life Sciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan, e-mail: [email protected] and Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Corporation, Japan Kan Nobuta Delaware Biotechnology Institute, University of Delaware, 15 Innovation Way, Room 230, Newark, DE 19711, USA, e-mail: [email protected] Ken-Ichi Nonomura Experimental Farm/Plant Genetics Lab., National Institute of Genetics, Yata 1111, Mishima, Shizuoka 411-8540, Japan, e-mail: [email protected]

Contributors

Hisataka Numa Division of Genome and Biodiversity Research, National Institute of Agrobiological Sciences, 2-1-2 Kannondai, Tsukuba, Ibaraki 305-8602, Japan, e-mail: [email protected] Takayuki Ohnishi Graduate School of Agricultural and Life Sciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan, e-mail: [email protected] Eiichi Ohtsubo Institute of Molecular and Cellular Biosciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-0032, Japan, e-mail: [email protected] Hisako Ohtsubo Institute of Molecular and Cellular Biosciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-0032, Japan, e-mail: [email protected] Katsuyuki Oki Department of Bioscience, Fukui Prefectural University, 4-1-1 Kenjyojima, Matsuoka, Eiheiji-cho, Yoshida-gun, Fukui 910-1195, Japan Yutaka Okumoto Graduate School of Agriculture, Kyoto University, Kitashirakawa, Sakyoku, Kyoto 606-8502, Japan, e-mail: [email protected] Kazumitsu Onishi Laboratory of Plant Breeding, Graduate School of Agriculture, Hokkaido University, Kita-9, Nishi-9, Kita-ku, Sapporo 060-8589, Japan, e-mail: [email protected] Tomoaki Sakamoto Institute for Advanced Research, Nagoya University, Furoucho, Chikusaku, Nagoya, Aichi 464-8601, Japan, e-mail: [email protected] Yoshio Sano Laboratory of Plant Breeding, Graduate School of Agriculture, Hokkaido University, Kita-9, Nishi-9, Kita-ku, Sapporo 060-8589, Japan, e-mail: [email protected] Takuji Sasaki National Institute of Agrobiological Sciences, 1-2, Kannondai 2-chome, Tsukuba, Ibaraki 305-8602, Japan, e-mail: [email protected] Yutaka Sato Graduate School of Bioagricultural Sciences, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8601, Japan, e-mail: [email protected]

xix

xx

Contributors

Zenpei Shimatani Department of Basic Biology, School of Life Science, The Graduate University for Advanced Studies, Okazaki 444-8585, Japan, e-mail: [email protected] Kyoko Takagi National Institute for Basic Biology, Okazaki 444-8585, Japan, e-mail: [email protected] and Plant Breeding Laboratory, Graduate School of Agriculture, Hokkaido University, Sapporo 060-8589, Japan Fumio Takaiwa Transgenic Crop Research and Development Center, National Institute of Agrobiological Sciences, Kannondai 2-1-2, Tsukuba Ibaraki 305-8602, Japan, e-mail: [email protected] Tsuyoshi Tanaka Division of Genome and Biodiversity Research, National Institute of Agrobiological Sciences, 2-1-2 Kannondai, Tsukuba, Ibaraki 305-8602, Japan, e-mail: [email protected] Takatoshi Tanisaka Graduate School of Agriculture, Kyoto University, Kitashirakawa, Sakyoku, Kyoto 606-8502, Japan, e-mail: [email protected] Rie Terada National Institute for Basic Biology, Okazaki 444-8585, Japan, e-mail: [email protected] and Department of Basic Biology, School of Life Science, The Graduate University for Advanced Studies, Okazaki 444-8585, Japan Norihiko Tomooka National Institute of Agrobiological Sciences, 2-1-2 Kannondai, Tsukuba, Ibaraki 305-8602, Japan, e-mail: [email protected] Kinya Toriyama Graduate School of Agricultural Science, Tohoku University, Sendai 981-8555, Japan, e-mail: [email protected] Suguru Tsuchimoto Institute of Molecular and Cellular Biosciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-0032, Japan, e-mail: [email protected] Kazuo Tsugane National Institute for Basic Biology, Okazaki 444-8585, Japan, e-mail: [email protected] and Department of Basic Biology, School of Life Science, The Graduate University for Advanced Studies, Okazaki 444-8585, Japan

Contributors

Nobuhiro Tsutsumi Graduate School of Agricultural and Life Sciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan, e-mail: [email protected] Hiroko Urawa National Institute for Basic Biology, Okazaki 444-8585, Japan, e-mail: [email protected] Duncan A. Vaughan National Institute of Agrobiological Sciences, 2-1-2 Kannondai, Tsukuba, Ibaraki 305-8602, Japan, e-mail: [email protected] Jianzhong Wu National Institute of Agrobiological Sciences, 1-2, Kannondai 2-chome, Tsukuba, Ibaraki 305-8602, Japan Jian-Hong Xu Institute of Molecular and Cellular Biosciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-0032, Japan Shinichiro Yamaki Department of Life Science, Graduate University for Advanced Studies/Sokendai, Yata 1111, Mishima, Shizuoka 411-8540, Japan Toshio Yamamoto National Institute of Agrobiological Sciences, Tsukuba, Ibaraki 305-8602, Japan, e-mail: [email protected] Takaki Yamauchi National Institute for Basic Biology, Okazaki 444-8585, Japan, e-mail: [email protected] and Graduate School of Science and Technology, Chiba University, Matsudo 271-8510, Japan Lijun Yang Transgenic Crop Research and Development Center, National Institute of Agrobiological Sciences, Kannondai 2-1-2, Tsukuba Ibaraki 305-8602, Japan, e-mail: [email protected] Masahiro Yano National Institute of Agrobiological Sciences, Tsukuba Ibaraki 305-8602, Japan, e-mail: [email protected] Hiroshi Yasuda Transgenic Crop Research and Development Center, National Institute of Agrobiological Sciences, Kannondai 2-1-2, Tsukuba Ibaraki 305-8602, Japan, e-mail: [email protected] and National Agriculture Research Center for Hokkaido Region, Hitsujigaoka 1, Toyohira-ku, Sapporo Hokkaido 062-8555, Japan

xxi

Section I

Genome-wide and Genome-based Research

I.1

The Rice Genome Sequence as an Indispensable Tool for Crop Improvement Takuji Sasaki1(* ü ), Jianzhong Wu1, Hiroshi Mizuno1, Baltazar A. Antonio1, and Takashi Matsumoto1

1

Introduction

Rice is one of the three major staple food crops in the world. In 2004, world rice production representing total rice yield was about 450 million tons. This indicates that the introduction of high-yielding rice varieties coupled with improvements in agricultural practices over the last three decades have made a major impact in the form of increased rice production. However, although the global production and consumption of rice is currently at an equilibrium (http://worldfood.apionet.or.jp/index-e.html), the majority of the population in rice-producing areas, particularly in many Asian and African countries, is still suffering from hunger, malnutrition and extreme poverty. Although a stable supply of rice is closely associated with the agricultural policies of each country and the natural environment that supports agricultural activities, the scientific community can play a major role in maintaining a sustainable agriculture system for the benefit of mankind. Therefore, continuous efforts towards innovative research should be encouraged in order to address the many aspects related to increasing rice productivity in the midst of all the obstacles associated with rice cultivation, such as land scarcity, depleted water resources and climatic changes. Fortunately, recent advances in the field of genomics may offer new opportunities to tackle problems associated with crop improvement and agricultural productivity. Characterization and sequencing of the genome has been carried out in many organisms including plants as well as many scientifically and practically important eukaryotes. This approach has led to a thorough analysis of the relationship between gene activity and the genomic structure of each target species that characterizes it. It has also facilitated the identification of genes corresponding to phenotypes, which differentiates one species from another within the same genus or other closely related genera. Genomics has already revolutionized many aspects of basic research which has resulted in many surprising biological discoveries not only in

1

National Institute of Agrobiological Sciences, 1–2, Kannondai 2-chome, Tsukuba, Ibaraki 305-8602, Japan e-mail: [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

3

4

T. Sasaki et al.

agriculture, but also in medicine and other biologically based industries. In the case of major cereal crops such as rice and maize, large-scale genome analyses have been carried out since the early 1990s with the aim of identifying genes practically important to agriculture. These include genes related to yield potential, biotic and abiotic stress tolerance, heading date, and other agronomically important traits which are either orthologous among cereal crops or unique to each crop. Rice is the first cereal crop to be completely decoded (International Rice Genome Sequencing Project 2005). The high-quality map-based sequence of the entire rice genome is now available in the public domain. With the completion of sequencing, the next challenge to the scientific community would be to determine the function of about 37,000 predicted genes in rice and to utilize this information in identifying agronomically important genes not only in cultivated rice, but also in wild relatives of rice and in other cereal crops.

2

Major Features of the Rice Genome

The japonica rice variety “Nipponbare” was chosen as a common template for complete and accurate decoding of the Oryza sativa genome. This is because Nipponbare was one of the parent cultivars of the F2 population used for linkage analysis to construct a high-density molecular genetic map of rice (Harushima et al. 1998). The same cultivar was also used as a resource for the construction of cDNA libraries in order to generate a catalog of rice ESTs. About 6,000 of these ESTs were used as genetic and/or PCR-based markers to reconstruct the rice genome using large-sized rice genomic DNA fragments ligated in yeast artificial chromosome (YAC), bacterial artificial chromosome (BAC), and P1-derived artificial chromosome (PAC) vectors. Additionally, Nipponbare could be easily regenerated from callus, an important advantage for genetic transformation. The International Rice Genome Sequencing Project (IRGSP), a consortium of publicly funded laboratories from ten countries, was formed in 1998 to pursue the sequencing of the genome. In 2004, the IRGSP succeeded in completely decoding the rice genome sequence with 99.99% accuracy (International Rice Genome Sequencing Project 2005). The major features of the rice genome based on the high-quality genome sequence are as follows: 1. The genome size of Nipponbare is 389 Mb including the unsequenced regions or gaps in the physical map, which were measured by fiber-FISH and conventional FISH. 2. The genome consists of 37,544 non-transposable element-related genes, which were predicted ab initio by FGENESH after masking the repeat sequences. 3. Twenty nine percent of the genes (10,837) are amplified at least once in tandem within an adjacent 5 Mb region of the genome. 4. Chloroplast and mitochondrial insertions contribute 0.20–0.24% and 0.18–0.19% of the nuclear genome, respectively.

I.1 The Rice Genome Sequence as an Indispensable Tool for Crop Improvement

5

5. Class I and class II transposons occupy 19.4% and 13% of the rice genome, respectively. 6. A total of 18,828 class 1 simple sequence repeats (SSR) were identified as di, tri, and tetra-nucleotide SSRs. A detailed annotation of the completed rice genome sequence was carried out using rice full-length cDNAs as basic tools. This effort resulted in the accurate prediction of 29,550 expressed loci including the unmapped-mRNA clusters (The Rice Annotation Project, 2007; http://rapdb.dna.affrc.go.jp/). The details of this annotation and curation system are described by Ito et al. in Chapter I.2 of this book.

2.1

Characterization of Rice Centromeres

Rice centromeres have been characterized by the presence of a highly repetitive 155–165 bp satellite DNA, CentO, along with centromere-specific retrotransposons. Using CentO as a probe, the size of the centromeres of chromosome 4 and 8 were estimated cytogenetically as 50–100 kb (Cheng et al. 2002). Therefore, physical mapping of these genomic regions by assignment of fingerprint contigs using DNA markers as anchors and chromosome walking strategies has been a major problem. The structures of the centromeres of chromosomes 4 and 8 were found to be strikingly different from each other. In the case of the 124 kb physical map of chromosome 4 centromere, 18 tracts of 379 tandemly arrayed CentO units were identified (Zhang et al. 2004). On the other hand, in the case of chromosome 8 centromere, the core region of 30 kb consisted of only three tracts of 452 CentO units. The genomic regions between each CentO consisted of retroelements such as RIRE7, solo LTR, or Ty3-gypsy (Wu et al. 2004; Ma and Bennetzen 2005). Although there is no reasonable explanation based on biological function or the divergence and/or evolution of the centromeres, the dynamic amplification and/or recombination of centromeric repeats and accumulation of retrotransposons in the centromere could account for the difference between chromosome 4 and 8 centromeres (Ma and Jackson 2007; Ma et al. 2007). A comparative analysis of the centromere structure among the twelve chromosomes in several rice ecotypes should be carried out to address this phenomenon (Mizuno et al. 2007). A detailed genomic characterization has also been performed on the centromere of chromosome 3. The size of the CentO-rich core region, however, was too large (about 480 kb), which makes it unrealistic to fully sequence. Therefore, cytogenetic analysis by fiber-FISH using CentO as a probe was performed to reveal several long CentO clusters interrupted by non-CentO blocks within the core. The resolution of fiber-FISH allowed for further interpretation of the arrangement of CentO of chromosome 3 (The Rice Chromosome 3 Sequencing Consortium, 2005). The centromere structure of chromosome 5 has also been recently elucidated (unpublished). The physical map of chromosome 5 includes two PAC clones covering the centromere with 4 kb and 65 kb CentO clusters in the sequence and shows similar structural characteristics to that of chromosome 8.

6

3

T. Sasaki et al.

Comparative Analysis of Genome Structure Within Oryza sativa

The genus Oryza includes more than 120,000 varieties cultivated throughout the world, as well as many wild rice species. The cultivar Nipponbare belongs to Oryza sativa ssp. japonica, which is mainly adapted to temperate regions and cultivated in limited areas such as Japan, Korea, Taiwan, and some parts of Europe. The subspecies indica, on the other hand, is widely cultivated in many countries, including China, India, Thailand, Indonesia, Senegal, and the USA. Therefore, indica rice has a wider range of genetic diversity compared with japonica rice. Furthermore, indica rice varieties account for about 90% of total rice production worldwide. This suggests that the genome sequence information of an indica rice variety is just as desirable and important in agriculture as the genome sequence of a japonica rice variety. Phylogenetic analysis has shown the existence of subgroups among indica rice varieties (Garris et al. 2005). So far, the genome sequence of 93-11, an indica rice variety cultivated in China, has been obtained by whole-genome shotgun sequencing (Yu et al. 2002). The assembled draft sequence totals 466 Mb, which is much larger than the estimated genome size for Nipponbare. Since it is only a draft sequence, sequence fragments of transposable elements may have been redundantly incorporated into nucleotide sequences of euchromatic regions. To confirm the actual genome size of 93-11 would require a similar map-based sequencing strategy as performed in Nipponbare. The difference in genome size between the subspecies of cultivated rice was clarified by in silico physical mapping of indica rice variety Kasalath, using the Nipponbare physical map as a standard (Katagiri et al. 2004). A BAC library of Kasalath was constructed and the end-sequence of each BAC clone was analyzed. After removing the BACs carrying repetitive or unrealistically distant pairs of end sequences, each pair of end sequences was blasted to the Nipponbare genome sequence. As a result, a total of 12,170 paired BACs were mapped to the 12 chromosomes, representing 450 contigs and a total length of 308.5 Mb. To further confirm the alignment of Kasalath BACs along the Nipponbare chromosome 1, PCR primers were designed based on the Nipponbare EST sequences and used for mapping the Kasalath BACs. Of the 306 Kasalath BACs corresponding to the minimum tiling path of chromosome 1, a total of 290 BACs were confirmed as containing the marker sequences derived from expressed genes. The mapping results indicate that the difference in the genome size of the two rice subspecies is rather small. However, the frequency of Kasalath BACs mapped in chromosomes 4, 9, 11, and 12 was relatively low. In the case of chromosomes 4 and 9, the low recovery could be attributed to a large amount of heterochromatin in these chromosomes. Detailed sequence comparison of a 2.3 Mb region of chromosome 4 between Nipponbare and a Chinese indica variety, Guangluai 4, revealed the colinearity of gene order and more insertions in japonica than in indica (Feng et al. 2002). On the other hand, in the case of chromosomes 11 and

I.1 The Rice Genome Sequence as an Indispensable Tool for Crop Improvement

7

12, the low frequency of mapped Kasalath BAC clones recovery may be due to the difference in the genome structure of Nipponbare and Kasalath. A 1.35 Mb region in chromosome 11 and a 0.69 Mb region in chromosome 12 of Nipponbare did not generate corresponding Kasalath BAC contigs. In chromosome 11, this region was found to be rich in disease resistance genes and varied from one cultivar to another (Leister et al. 1998). To facilitate the identification of allelic divergence among O. sativa, the genetic diversity among 30,000 rice germplasm accessions at the Gene Bank of the National Institute of Agrobiological Sciences (NIAS) in Japan was evaluated. After the first screening based on the available data such as regional origin and morphological appearance of each accession, 332 accessions were selected and evaluated for polymorphism using 179 loci of RFLPs (Kojima et al. 2005). The RFLP data were subjected to cluster analysis and 67 groups were recognized. A single accession from each of the 67 groups was selected. These 67 accessions retained 91% of the alleles in the original 332 accessions. This core collection is now available at the NIAS Gene Bank (http://www.gene.affrc.go.jp/index_j.php) and can be used as an important resource for detailed genetic studies and for further improvement of O. sativa cultivars.

4

Comparative Analysis of Genome Structure in the Genus Oryza

The completion of an accurate and precise map-based genome sequence of Oryza sativa cultivar Nipponbare could open the door in understanding genome-wide diversity based on the nucleotide sequence. The genus Oryza is composed of 23 species categorized into 10 genome types (Khush 1997). Among them, the wild Oryza species represent various closely related species, which show a wide range of diversity. These could provide important resources that could elucidate the evolution of cultivated rice and provide a detailed analysis of divergence in the genus Oryza. The present O. sativa cultivars have been bred to produce more panicles as well as many of the agronomic characteristics inherent to wild rice, such as tolerance to biotic and abiotic stresses, some of which must have been lost in the process of domestication and breeding. Gene transfer has been successful between species with the AA genome. The rice bacterial blight disease-resistant gene Xa21 from O. longistaminata has been transferred to O. sativa by ordinary crossing (Khush et al. 1991) and the resulting progenies have been widely used for breeding new rice varieties resistant to Xanthomonas infection (Singh et al. 2001). Other desirable genes conferring traits such as blast resistance or insect resistance have been successfully transferred from a wild species to cultivated rice (Amante-Bordeos et al. 2004). Gene transfer to O. sativa from species other than the AA genome has also been possible by embryo rescue technique (Multani et al. 2004).

8

T. Sasaki et al.

The Oryza Map Alignment Project (OMAP, http://www.omap.org/index.html) was established to provide a platform for understanding genome diversity in the genus Oryza. In this project, 12 deep-coverage large-insert BAC libraries that represent the 10 genome types of the genus Oryza, namely AA, BB, BBCC, CC, CCDD, EE, FF, GG, HHJJ, and HHKK, were constructed (Ammiraju et al. 2006). Then BAC end sequences were collected and fingerprints were assembled to make contigs. Each contig was mapped to the reference Nipponbare genome using end sequences as markers for landing. The physical map of chromosome 3 of O. nivara (AA genome) showed that 76 BAC contigs were aligned on the pseudomolecules of O. sativa and the existence of tremendous synteny between them (Wing et al. 2005). Based on the distances of BAC end sequences, chromosome 3 of O. sativa was found to be approximately 21% larger that that of O. nivara. Genome-wide mapping of O. punctata (BB genome) BAC clones was also carried out (Kim et al. 2007). As a result, 176 fingerprint contigs were aligned onto the 12 O. sativa pseudomolecules covering 408 Mb which corresponds to 96% of the O. punctata genome (genome size = 418 Mb). Comparison of the physical maps of O. sativa and O. punctata showed one inversion of 1.1 Mb which was cytogenetically confirmed on chromosome 8, and one transposition of 400 kb on chromosome 6 of O. punctata which was confirmed by checking the assembly of fingerprint using the overgo hybridization method. Other genomic regions are syntenic between these two species, although five chromosomes showed an increase in size, whereas six chromosomes showed a decrease in size at seven other regions. Such comparative physical maps between O. sativa and other Oryza species should be useful for the analysis of corresponding genes among the different species in the genus Oryza and may provide valuable information on the divergence and evolution of major genes in rice. The genome structure of the centromere of wild Oryza species was compared with that of O. sativa in order to understand the evolutionary relationships among Oryza species. Although almost all of the Oryza species carry the centromerespecific CentO unit of 155 bp repeat sequence, the amount varies from species to species. Oryza species of AA, BBCC, and CCDD genomes are rich in CentO, the EE genome species O. australiensis has a low copy number of CentO, whereas six different species of CC, FF, or GG genomes have no CentO in their genomes (Lee et al. 2005). By immunoprecipitation with centromere-specific histon H3 variant CenH3, novel repeat sequences CentO-C1 (126 bp) and CentO-C2 (366 bp), and Cent-F (154 bp) were identified in CC and FF genomes, respectively. CentO-C2 and Cent-F have no sequence matches to that of CentO, but Cent-C1 has significant sequence similarity with CentO. Phylogenetic analysis of CentO and CentO-C1 has indicated the divergence of these two components at around the time of divergence of CentO and CentC, a maize-centromere core-specific repeat unit. These findings suggest several patterns of divergence and evolution of Oryza and other related grass species. At present, it remains unclear why the centromere functions equally among Oryza species in spite of the rapid divergence of core nucleotide sequence units. Furthermore, it is unclear why the centromeres diverged in closely related species even if they function equally.

I.1 The Rice Genome Sequence as an Indispensable Tool for Crop Improvement

5

9

Comparative Analysis of Genome Structure in the Grass Family

The colinearity of gene arrangement among the members of the family Poaceae has been clarified by cross mapping of expressed genes (Devos and Gale 1997). Although the grass species are highly diverse in morphology, growth habit, and genome size, they share a common genome structure inherited from an ancestral grass genome about 70 million years ago. Since the genetically identified synteny indicates conservation of gene content and gene order, it is necessary to validate the microsynteny at the genomic sequence level to identify commonly existing genes among grass species. The map-based rice genome sequence information is therefore an indispensable tool to understanding the genome structure of major cereal crops. Studies on micro-colinearity have been carried out in rice, maize, and sorghum with the aim of elucidating the nature, timing, and lineages of genic rearrangements that have differentiated the selected regions after the divergence of each species from the common ancestral genome. Among the genes selected were alcohol dehydrogenase 1 (adh1) (Tarchini et al. 2000), r1/b1 genes which regulate the expression level of several enzymes of the anthocyanin biosynthetic pathway (Swigonova et al. 2005), and orp1/orp2 genes which encode the beta subunit of tryptophan syntase (Ma et al. 2005). The sequences of rice orthologous genomic regions were analyzed and compared with those of maize and sorghum. The results indicated significant conservation of gene content and order in large segments of these grass genomes, although several gene rearrangements were also observed. This genetic micro-colinearity, however, can also be violated even within the same species. The 100 kb bz genomic region of two different maize lines, McC and B73, was sequenced, and unexpected differences of genes were observed between them (Fu and Dooner 2002). This difference was confirmed in more than 2.8 Mb allelic chromosomal regions of Mo17 and B73 (Brunner et al. 2005) and was due to the insertion of large numbers of long terminal repeatretrotransposons called helitron (Morgante et al. 2005). On average, more than 50% of the compared sequence is noncolinear. The shared sequences in these two maize lines clearly showed colinearity with the gene order of rice. Several duplicated regions in the maize genome have also been sequenced and aligned with rice chromosome 3 (Bruggman et al. 2006). The aligned sequences indicate that the short arm of this rice chromosome is highly colinear with the short arm of maize chromosome 1 and the inverted long arm of maize chromosome 9, whereas the long arm of rice chromosome 3 is colinear with the long arm of maize chromosome 1 and the inverted short arm of maize chromosome 5. These data clearly demonstrate the conservation of gene order between maize and rice, and provide evidence of the ancient tetraploid origin of the maize genome. Other grass species are now being targeted for genome analysis by wholegenome shotgun sequencing, clone-by-clone sequencing, or selective gene sequencing. A whole-genome shotgun sequence of sorghum has been obtained and is currently being assembled (http://www.phytozome.net/sorghum). The sequencing of

10

T. Sasaki et al.

Brachypodium distachyon, a temperate wild grass species which could be used as a model forage and turf crop, was started in 2007 (http://www.jgi.doe.gov/sequencing/ why/CSP2007/brachypodium.html). Moreover, clone-by-clone sequencing strategy is now being used for maize genome sequencing and the genome sequence is expected to be completed by 2008 (http://www.nsf.gov/news/news_summ. jsp?cntn_id=104608&org=olpa&from=news). Barley and wheat genomes share a similar structure; however, it has not yet been finalized which species will be targeted that will include both the diploid and hexaploid wheat. Detailed comparison among Poaceae genome sequences will be a basic resource for identification of agronomically important genes and may help in understanding the divergence and evolution of the grass species.

6

Conclusion

Rice, wheat, and maize are the major three staples supporting our daily life. It has taken more than 10,000 years to establish modern elite cultivars, which produce large numbers of seeds with high nutritional content and good eating quality. Undoubtedly, these cereal crops must be further improved to supply enough food to our ever-increasing world population. Furthermore, these crops are important sources of feed for livestock. With increasing consumption of meat products such as beef and pork, there is also a need to increase the supply of livestock feed. In addition, crops such as maize, sorghum, and sugarcane are now in demand as potential sources of ethanol, which could be used as a substitute for petroleum for fuel. Currently, world cereal crop output even on the most productive irrigated lands has reached nearly the maximum limit with current cultivars. Insect pests and diseases continue to counteract any increases in harvest even with high-yielding cultivars. Environmental degradation including pollution, increase in temperature due to global warming, and reductions in suitable arable land provide additional constraints. These factorsmake steps to maximize crop productivity especially important. One of the promising approaches to solving these problems is to understand the genomic systems of plants and modify their genetic composition so that their potential to meet our food supply demands can be maximized. Increasing yield potential and yield stability will come from a combination of biotechnology and improved conventional breeding. Both will be dependent on a high-quality genome sequence. Rice is the first cereal crop species to be completely decoded and this information is certain to accelerate the discovery of many agronomically important genes. Biotechnological tools for manipulating rice genetics have already made significant impact in addressing several constraints in plant breeding. Improved techniques and the increase in genomic information have resulted in great advances in rice biotechnology. Furthermore, although rice genome information could be used in understanding other cereal crops, it is still necessary to study each crop based on the specific

I.1 The Rice Genome Sequence as an Indispensable Tool for Crop Improvement

11

characteristics governed by each genomic system. The next challenge in genomics would involve elucidating the genome sequence information from other major cereal crops and developing a comprehensive program on crop improvement based on the genome information in order to secure a stable food supply for mankind.

References Amante-Bordeos A, Sitch LA, Nelson R, et al. (2004) Transfer of bacterial blight and blast resistance from the tetraploid wild rice Oryza minuta to cultivated rice, Oryza sativa. Theor Appl Genet 84:345–354 Ammiraju JSS, Luo M, Goicoechea JL, et al. (2006) The Oryza bacterial artificial chromosome library resource: construction and analysis of 12 deep-coverage large-insert BAC libraries that represent the 10 genome types of the genus Oryza. Genome Res 16:140–147 Bruggman R, Bharti AK, Gundlach H, et al. (2006) Uneven chromosome contraction and expansion in the maize genome. Genome Res 16:1241–1251 Brunner S, Fengler K, Morgante M, et al. (2005) Evolution of DNA sequence nonhomologies among maize inbreds. Plant Cell 17:343–360 Cheng Z, Dong F, Langdon T, et al. (2002) Functional rice centromeres are marked by a satellite repeat and a centromere-specific retrotransposon. Plant Cell 14:1691–1704 Devos KM, Gale MD (1997) Comparative genetics in the grasses. Plant Mol Biol 35:3–15 Feng Q, Zhang Y, Hao P, et al. (2002) Sequence and analysis of rice chromosome 4. Nature 420:316–320 Fu H, Dooner HK (2002) Intraspecific violation of genetic colinearity and its implications in maize. Proc Natl Acad Sci USA 99:9573–9578 Garris AJ, Tai TH, Coburn J, et al. (2005) Genetic structure and diversity in Oryza sativa L. Genetics 169:1631–1638 Harushima Y, Yano M, Shomura A, et al. (1998) A high-density rice genetic linkage map with 2275 markers using a single F2 population. Genetics 148:479–484 International Rice Genome Sequencing Project (2005) The map-based sequence of the rice genome. Nature 436:793–800 Katagiri S, Wu J, Ito Y, et al. (2004) End sequencing and chromosomal in silico mapping of BAC clones derived from an indica rice cultivar, Kasalath. Breed Sci 54:273–279 Khush GS (1997) Origin, dispersal, cultivation and variation of rice. Plant Mol Biol 35:25–34 Khush GS, Bacalangco E, Ogawa T (1991) A new gene for resistance to bacterial blight from O. longistaminata. Rice Genet Newsl 7:121–122 Kim H, SanMiguel P, Nelson W, et al. (2007) Comparative physical mapping between O. sativa (AA genome type) and O. punctata (BB genome type). Genetics 176:379–390 Kojima K, Ebana K, Fukuoka S, et al. (2005) Development of an RFLP-based rice diversity research set of germplasm. Breed Sci 55:431–440 Lee H-R, Zhang W, Langdon T, et al. (2005) Chromatin immunoprecipitation cloning reveals rapid evolutionary patterns of centromeric DNA in Oryza species. Proc Natl Acad Sci USA 102:11793–11798 Leister D, Kurth J, Laurie DA, et al. (1998) Rapid organization of resistance gene homologs in cereals. Proc Natl Acad Sci USA 95:370–375 Ma J, Bennetzen JL (2005) Recombination, rearrangement, reshuffling and divergence in a centromeric region of rice. Proc Natl Acad Sci USA 103:383–388 Ma J, Jackson SA (2007) Retrotransposon accumulation and satellite amplification mediated by segmental duplication facilitate centromere expansion in rice. Genome Res 16:251–259 Ma J, SanMiguel P, Lai J, et al. (2005) DNA rearragement in orthologous Orp regions of the maize, rice and sorghum genomes. Genetics 170:1209–1220

12

T. Sasaki et al.

Ma J, Wing RA, Bennetzen JL,et al. (2007) Plant centromere organization: a dynamic structure with conserved functions. Trends Genet 23:134–139 Mizuno H, Ito K, Wu J, et al. (2007) Identification and mapping of expressed genes, simple sequence repeats and transposable elements in centromeric regions of rice chromosomes. DNA Res 13:267–274 Morgante M, Brunner S, Pea G, et al. (2005) Gene duplication and exon shuffling by helitron-like transposons generate intraspecies diversity in maize. Nature Genet 37:997–1002 Multani DS, Jena KK, Brar DS, et al. (2004) Development of monosomic alien addition lines and introgression of genes from Oryza australiensis Domin. to cultivated rice O. sativa L. Theor Appl Genet 88:102–109 Singh S, Sidhu JS, Huang N, et al. (2001) Pyramiding three bacterial blight resistance genes (xa5, xa13 and Xa21) using marker-assisted selection into indica rice cultivar PR106. Theor Appl Genet 102:1011-1015 Swigonova Z, Bennetzen JL, Messing J (2005) Structure and evolution of the r/b chromosomal regions in rice, maize and sorghum. Genetics 169:891–906 Tarchini R, Biddle P, Wineland R, et al. (2000) The complete sequence of 340 kb of DNA around the rice Adh1-Adh2 region reveals interrupted colinearity with maize chromosome 4. Plant Cell 12:381–391 The Rice Annotation Project (2007) Curated genome annotation of Oryza sativa ssp. japonica and comparative genome analysis with Arabidopsis thaliana. Genome Res 17:175–185 The Rice Chromosome 3 Sequencing Consortium (2005) Sequence, annotation, and analysis of synteny between rice chromosome 3 and diverged grass species. Genome Res 15:1284–1291 Wing RA, Ammiraju JSS, Luo M, et al. (2005) The Oryza Map Alignment Project: the golden path to unlocking the genetic potential of wild rice species. Plant Mol Biol 59:53–62 Wu J, Yamagata H, Hayashi-Tsugane M, et al. (2004) Composition and structure of the centromeric region of rice chromosome 8. Plant Cell 16:967–976 Yu J, Hu S, Wang J, et al. (2002) A draft sequence of the rice genome (Oryza sativa L. ssp.indica). Science 296:79–92 Zhang Y, Huang Y, Zhang L, et al. (2004) Structural features of the rice chromosome 4 centromere. Nucl Acids Res 32:2023–2030

I.2

Bioinformatics and Database of the Rice Genome Hisataka Numa1, Tsuyoshi Tanaka1, and Takeshi Itoh1(* ü)

1

Introduction

After the completion of the genome sequencing of Asian cultivated rice (Oryza sativa ssp. japonica cv. Nipponbare) by the International Rice Genome Sequencing Project (IRGSP) (International Rice Genome Sequencing Project 2005), it was envisaged to decipher the underlying biological meanings of the genomic sequences. For this purpose, an indispensable step is to detect the locations of the genes and predict their precise exon–intron structures (Itoh 2007). This procedure is called gene-finding or gene prediction. Predicted genes will be the basis of further computational and experimental studies, such as molecular evolutionary analysis, transposon tagging and microarray experiments. In addition, databases are needed from which information about genome-wide analyses can be retrieved. In this chapter, we overview the bioinformatics methods and databases that are useful for data analysis of the rice genome.

2

Gene-finding in the Rice Genome

There are three methods to predict genes in a genomic sequence. First, the comparative method finds conserved regions between evolutionarily related species. In general, functional DNA such as a protein-coding gene is more conserved between species than the other nonfunctional regions. Therefore, a pairwise alignment of two related genome sequences identifies possible genic sequences that are highly conserved between the species examined. In addition, since nonsynonymous substitutions (dN) are much fewer than synonymous substitutions (dS), the dN/dS ratio is known to be a good indicator of protein-coding regions (Nekrutenko et al.

1

Division of Genome and Biodiversity Research, National Institute of Agrobiological Sciences, 2-1-2 Kannondai, Tsukuba, Ibaraki 305-8602, Japan e-mail: [email protected]; [email protected]; [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

13

14

H. Numa et al.

2002). The comparative method is quite powerful if genome sequences are available in several related species, as shown between human and mouse genomes (Batzoglou et al. 2000). This method, however, requires sufficient genomic sequence information from moderately divergent species, so that for the application of the comparative method to rice, genome-wide sequencing of other cereal species is required. Second, the homology-based method searches known complementary DNAs (cDNAs) or protein sequences for homologous regions in newly determined genomic DNA. In particular, regions identical to cDNAs are regarded as convincing candidates of bona fide genes, because their expressions are supported by experimental evidence. Since a large-scale sequencing effort has produced more than 30,000 full-length cDNA sequences in rice (Kikuchi et al. 2003), the homologybased method is markedly effective in predicting gene structures in the rice genome. For example, approximately 20,000 rice genes were identified by aligning the full-length cDNAs to the genome (Ohyanagi et al. 2006; The Rice Annotation Project 2007). The two aforementioned methods depend on known sequences, whilst the ab initio method uses merely statistical information obtained from known genes. Thus, an advantage of the ab initio method is that completely novel genes with no homologues in the current databases can be found. Moreover, even though a large number of known sequences are not available for a species of interest, this method is applicable when protein-coding sequences are obtained from closely related species. For these reasons, ab initio gene prediction programs have played an important role for gene finding in genome sequencing projects (Itoh 2007). However, ab initio predictions should be interpreted with caution, because they tend to give rise to a large number of false positives, although the sensitivity of programs is sufficiently high (Yao et al. 2005). In addition, pseudogenes may not be distinguishable from functional genes. It is thus important for users to know the accuracy of a program before conducting gene predictions. We will discuss this issue in the next section.

3

Comparison of Gene Prediction Programs

A number of programs are available online for rice gene predictions (Table 1). All of these programs are based on a Hidden Markov model except GlimmerM, which employs an Interpolated Markov model and a decision tree (Salzberg et al. 1999). Although the algorithms of the programs are quite similar, their results are not necessarily the same. Therefore, the prediction accuracy of ab initio prediction programs should be evaluated. To compare results of the programs in Table 1, we selected 100 protein-coding genes at random from the Rice Annotation Project Database (RAP-DB) (Ohyanagi et al. 2006). Each program was run with default parameters for rice, or monocots if a rice model was unavailable. Program performance was evaluated by two measures: sensitivity (fraction of true positives

I.2 Bioinformatics and Database of the Rice Genome

15

Table 1 Ab initio gene prediction programs for O. sativa Program URL BGF FGENESH

http://tlife.fudan.edu.cn/bgf/ http://sun1.softberry.com/berry.phtml?topic=fgene sh&group=programs&subgroup=gfind http://opal.biology.gatech.edu/GeneMark/eukhmm. cgi http://glocate.dna.affrc.go.jp/ http://www.tigr.org/tdb/glimmerm/glmr_form.html http://rgp.dna.affrc.go.jp/E/RiceHMM/index.html http://homepage.mac.com/iankorf/

GeneMark.hmm GLocate GlimmerM RiceHMM SNAP

in the genuine protein-coding genes) and specificity (fraction of true positives in all predicted genes). Both sensitivity and selectivity were examined at three levels; the nucleotide, exon and gene. Our comparison of the seven programs showed that BGF, FGENESH and GeneMark.hmm were generally more accurate at any level than the others (Table 2). It is noteworthy, however, that a number of genes were not reported by these three programs but could be detected by the others. These observations suggest that a combination of multiple ab initio prediction programs should be considered so that genuine genes are not missing from a predicted data set. For instance, to select the most probable gene structures from multiple predictions, a modified version of Combiner (Allen et al. 2004) was used for the Rice Annotation Project (The Rice Annotation Project 2007). If the homology-based method is combined with the ab initio method, it should further improve the predictions. Programs widely used for this purpose are listed in Table 3. Since more than one million expressed sequence tags (ESTs) of rice are available in DDBJ/EMBL/GenBank, these are of great help in modifying predicted exon–intron structures. For example, refer to RiceBLAST to align rice ESTs to a DNA sequence. FGENESH+ was designed to utilize both homology-based Table 2 Prediction accuracy of ab initio gene prediction programs for O. sativa Nucleotide level Exon level Gene level Program

Sna

Spa

Sn

Sp

Sn

Sp

BGF 0.93 0.89 0.77 0.75 0.34 0.34 FGENESH 0.90 0.88 0.78 0.74 0.37 0.36 GeneMark.hmm 0.93 0.86 0.74 0.67 0.35 0.30 GLocate 0.82 0.85 0.71 0.68 0.17 0.15 GlimmerM 0.65 0.81 0.36 0.44 0.12 0.08 RiceHMM 0.64 0.72 0.18 0.22 0.10 0.06 SNAP 0.72 0.85 0.44 0.53 0.22 0.19 a Sn and Sp represent sensitivity and specificity, respectively. Given the numbers of true positives (TP), actual positives (AP) and predicted positives (PP), Sn and Sp are defined as follows: Sn = TP/ AP and Sp = PP/AP. At the nucleotide level, PP is defined as the number of nucleotides that are correctly predicted as protein coding regions. At the exon level, PP means the number of predicted exons that are identical to actual exons. At the gene level, all exons included in one gene must be identical between predicted and actual genes

16

H. Numa et al.

Table 3 Web-based resources for gene prediction and data analysis Category Name URL Gene prediction

Databases

FGENESH+ JIGSAW RiceGAAS Agene Augustus GeneZilla GlimmerHMM DDBJ EMBL NCBI RAP-DB KOME Gramene Osa1 RISe OryzaBase

Similarity search

Annotation format

Tos17 insertion mutants RiceGE RetrOryza Gene Ontology RiceBLAST BLAST InterProScan GFF format DDBJ-XML BSML

http://www.softberry.com/ http://www.cbcb.umd.edu/software/jigsaw/ http://ricegaas.dna.affrc.go.jp/usr/ http://servers.binf.ku.dk/agene/ http://augustus.gobics.de/ http://www.genezilla.org/ http://cbcb.umd.edu/software/glimmerhmm/ http://www.ddbj.nig.ac.jp/Welcome-e.html http://www.ebi.ac.uk/embl/ http://www.ncbi.nlm.nih.gov/ http://rapdb.dna.affrc.go.jp/ http://cdna01.dna.affrc.go.jp/cDNA/ http://www.gramene.org/Oryza_sativa/ index.html http://www.tigr.org/tdb/e2k1/osa1/ http://rise.genomics.org.cn/rice/index2.jsp http://www.shigen.nig.ac.jp/rice/oryzabase/ top/top.jsp http://tos.nias.affrc.go.jp/ http://signal.salk.edu/cgi-bin/RiceGE http://www.retroryza.org/ http://www.geneontology.org/ http://riceblast.dna.affrc.go.jp/ http://www.ncbi.nlm.nih.gov/BLAST/ http://www.ebi.ac.uk/InterProScan/ http://www.sanger.ac.uk/Software/ formats/GFF/ http://www.xml.nig.ac.jp/index.html http://www.bsml.org/

and ab initio methods. If one would like to select the most probable genes from multiple sources, JIGSAW assists in finding appropriate gene structures (Allen and Salzberg 2005). For anyone not familiar with bioinformatics, an automated annotation system of rice, RiceGAAS, which has a web-based interface, is useful (Sakata et al. 2002). This system analyzes a DNA sequence submitted by a user and displays the results of automated annotation of protein-coding regions suggested by various programs.

4

Prediction of Gene Functions

The prediction of gene functions is one of the crucial steps of annotation, along with gene-finding. Assuming that similarity of DNA or amino acid sequences implies similarity of their functions, sequence similarity searches against existing databases should give valuable information about possible biological functions (Table 3). BLAST is recognized as one of the best programs for similarity searches

I.2 Bioinformatics and Database of the Rice Genome

17

(Altschul et al. 1997). The International Nucleotide Sequence Databases (DDBJ, EMBL and GenBank) maintain the public DNA databases and provide the BLAST search against the databases. Protein databases such as the UniProt Knowledgebase (UniProtKB) (The UniProt Consortium 2007) are also useful for the similarity search. Even though newly determined sequences do not show any significant similarity to known sequences in the databases by BLAST, small functional domains that are possibly helpful to infer functions may exist. InterPro is an integrated database of protein families and domains (Zdobnov and Apweiler 2001; Quevillon et al. 2005). It consists of a number of databases, such as Pfam (Finn et al. 2006), that can be searched simultaneously by InterProScan. The InterPro entries are linked to the identifiers of gene ontology (GO) (Ashburner et al. 2000), which is widely used to define and classify gene functions. Although similarity searches against protein databases are of great importance, their results should be examined carefully, because functions assigned by automated methods tend to contain irrelevant descriptions that were derived from electronic annotations. These descriptions are largely due to genome-wide sequencing projects, which have been generating enormous numbers of “new genes” that have not yet been experimentally validated. Thus, automated annotations might hamper further analyses unless the definitions of the homologues detected are not based on experimental evidence. With this in mind, curation is currently recognized as an indispensable process to construct a biological database (Misra et al. 2002; Camon et al. 2003). For example, in the Rice Annotation Project (RAP) all of the automated annotations of gene functions were manually curated so that biologically significant descriptions are given to the rice genes (The Rice Annotation Project 2007).

5

The Rice Annotation Project Database

To facilitate genome-wide analysis of rice, the Rice Annotation Project Database (RAP-DB) was developed (Fig. 1; Ohyanagi et al. 2006). The RAP-DB provides users with the IRGSP genome assembly and the RAP annotation as well as a number of functions to retrieve information about the rice genome. The annotation can be viewed through two different browsers, GBrowse and G-integra (Stein et al. 2002; Imanishi et al. 2004). The contents of the RAP-DB are listed in Table 4. GBrowse displays the rice genes and their functional descriptions. The rice genes are depicted in three tracks on the basis of their supporting evidence; transcripts with full-length cDNAs, predicted genes with ESTs, and genes predicted by pure ab initio methods. We assigned a systematic identifier to each loci (for details see http://rapdb.dna.affrc.go.jp/note.html#nomenclature). Whilst the genes supported by cDNAs are expected to be reliable, computationally predicted genes may contain errors. Detailed descriptions of the annotation are obtained by clicking on each transcript. In addition to GBrowse’s genomic view, G-integra displays comparative maps of rice and other cereal genes. cDNAs of the other species were mapped on the rice genome, and the alignments between the cDNA and the genome are available

18

H. Numa et al.

Fig. 1 Schematic views of the RAP-DB. a Genome viewer; b annotation detail viewer

in G-integra. GBrowse and G-integra are interlinked, and hence users can simultaneously grasp data displayed in the two browsers. The RAP-DB also contains annotation data created by Gnomon developed by the National Center for Biotechnology Information (NCBI) (Wheeler et al. 2007); this NCBI annotation can be compared with the RAP annotation by browsing GBrowse (the build 4 assembly). The Institute for Genomic Research (TIGR) has released rice genome

I.2 Bioinformatics and Database of the Rice Genome Table 4 Contents of the RAP-DB Viewer Category GBrowse

RAP annotations Link to other databases Other annotation data Mapping information

G-integra

Mapping information

19

Contents Gene function, Ortholog InterPro, GO, KOME, DDBJ, UniProt IRGSP, NCBI TIGR annotation, Flanking sequence tags (FSTs) of Tos17/T-DNA/Ds, BAC end sequences (BESs) by Gramene, Retrotransposons by RetrOryza ESTs of rice and other plants, Results of gene prediction programs, BESs, FSTs of Tos17/T-DNA/Ds

sequences assembled independently of IRGSP as well as their annotations (Yuan et al. 2005). The TIGR genes were mapped to the IRGSP genome and can be accessed by keywords in GBrowse as described below. To obtain annotation information in the RAP-DB, keyword search systems are prepared in both GBrowse and G-integra. Keywords should be any terms in the annotation data including the locus identifier, DDBJ/EMBL/GenBank accession number, descriptions of gene functions, Gene Ontology and InterPro. To search the genome or transcript sequences, BLAST and BLAT searches are useful; for example, a newly determined cDNA can be mapped to the rice genome by BLAT (Kent 2002). Results of BLAST and BLAT are hyperlinked to GBrowse so that users can browse a corresponding coordinate on the genome and look into the annotation information of a gene that is similar to a query sequence submitted. If downloading the entire data set is preferred, the following data are provided: the IRGSP genome assembly, the RAP annotation in the GFF format (Table 3), and the DNA or amino acid sequences of loci, transcripts and predicted genes in the FASTA format. The transcript sequences in the RAP-DB are, however, not necessarily identical to corresponding cDNA sequences because of polymorphisms, sequencing errors, etc. Here we show a database search using the RAP-DB. Protein kinases are known for their role in signal transductions. The rice genome contains a variety of protein kinase genes, but in many cases their functions remain to be clarified. A keyword search can detect 200 genes that contain a description of “protein kinase” in the RAP-DB, and 186 of them are similar to well-documented protein kinases registered in UniProtKB, whereas the other 14 possess only functional domain(s) of kinase in InterPro. Functions of these genes may be investigated by the transposon tagging method. For example, a mutant line (ND2051) created by Tos17 (Miyao et al. 2003) is found to disrupt a rice gene, Os07g0540800 (Fig. 1). ND2051 showed several phenotypes and they may be related to the function of Os07g0540800. In this way, a combination of large-scale computational and experimental analyses will lead to defining the function of individual genes. Development of the RAP-DB is still underway. Recently we added to the build 4 assembly new tracks for flanking sequence tags of 10 mutant lines, bacterial artificial

20

H. Numa et al.

chromosome (BAC)-end sequences derived from 13 Oryza species (Wing et al. 2005), and positions of LTR retrotransposons (Chaparro et al. 2007). In the future more tracks for information about genetics and functional genomics will be created. We anticipate that the RAP-DB will serve as a hub for rice genomics. Acknowledgements We thank Drs. Tatiana Tatusova and Alexandre Souvorov of NCBI for the NCBI gene models; Dr. Gynheung An of POSTECH for the information on T-DNA tagging lines; Dr. Olivier Panaud of the University of Perpignan for the information on RetrOryza on the IRGSP genome. This work was supported in part by a grant from the Special Coordination Funds for Promoting Science and Technology of the Ministry of Education, Culture, Sports, Science and Technology of Japan.

References Allen JE, Pertea M, Salzberg SL (2004) Computational gene prediction using multiple sources of evidence. Genome Res 14:142–148 Allen JE, Salzberg SL (2005) JIGSAW: integration of multiple sources of evidence for gene prediction. Bioinformatics 21:3596–3603 Altschul SF, Madden TL, Schaffer AA, et al. (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25:3389–3402 Ashburner M, Ball CA, Blake JA, et al. (2000) Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet 25:25–29 Batzoglou S, Pachter L, Mesirov JP, Berger B, Lander ES (2000) Human and mouse gene structure: comparative analysis and application to exon prediction. Genome Res 10:950–958 Camon E, Magrane M, Barrell D, et al. (2003) The Gene Ontology Annotation (GOA) project: implementation of GO in SWISS-PROT, TrEMBL, and InterPro. Genome Res 13:662–672 Chaparro C, Guyot R, Zuccolo A, Piegu B, Panaud O (2007) RetrOryza: a database of the rice LTR-retrotransposons. Nucleic Acids Res 35:D66–D70 Finn RD, Mistry J, Schuster-Bockler B, et al. (2006) Pfam: clans, web tools and services. Nucleic Acids Res 34:D247–D251 Imanishi T, Itoh T, Suzuki Y, et al. (2004) Integrative annotation of 21,037 human genes validated by full-length cDNA clones. PLoS Biol 2:856–875 International Rice Genome Sequencing Project (2005) The map-based sequence of the rice genome. Nature 436:793–800 Itoh T (2007) Rice genome annotation: beginnings of functional genomics. In: Upadhyaya NM (ed) Rice functional genomics: challenges, progress and prospects. Springer, Berlin Heidelberg New York, pp 21–31 Kent WJ (2002) BLAT – the BLAST-like alignment tool. Genome Res 12:656–664 Kikuchi S, Satoh K, Nagata T, et al. (2003) Collection, mapping, and annotation of over 28,000 cDNA clones from japonica rice. Science 301:376–379 Misra S, Crosby MA, Mungall CJ, et al. (2002) Annotation of the Drosophila melanogaster euchromatic genome: a systematic review. Genome Biol 3:83.1–83.22 Miyao A, Tanaka K, Murata K, et al. (2003) Target site specificity of the Tos17 retrotransposon shows a preference for insertion within genes and against insertion in retrotransposon-rich regions of the genome. Plant Cell 15:1771–1780 Nekrutenko A, Makova KD, Li WH (2002) The KA/KS ratio test for assessing the protein-coding potential of genomic regions: an empirical and simulation study. Genome Res 12:198–202 Ohyanagi H, Tanaka T, Sakai H, et al. (2006) The Rice Annotation Project Database (RAP-DB): hub for Oryza sativa ssp. japonica genome information. Nucleic Acids Res 34:D741–D744 Quevillon E, Silventoinen V, Pillai S, et al. (2005) InterProScan: protein domains identifier. Nucleic Acids Res 33:W116–W120

I.2 Bioinformatics and Database of the Rice Genome

21

Sakata K, Nagamura Y, Numa H, et al. (2002) RiceGAAS: an automated annotation system and database for rice genome sequence. Nucleic Acids Res 30:98–102 Salzberg SL, Pertea M, Delcher AL, Gardner MJ, Tettelin H (1999) Interpolated Markov models for eukaryotic gene finding. Genomics 59:24–31 Stein LD, Mungall C, Shu S, et al. (2002) The generic genome browser: a building block for a model organism system database. Genome Res 12:1599–1610 The Rice Annotation Project (2007) Curated genome annotation of Oryza sativa ssp. japonica and comparative genome analysis with Arabidopsis thaliana. Genome Res 17:175–183 The UniProt Consortium (2007) The Universal Protein Resource (UniProt). Nucleic Acids Res 35: D193–D197 Wheeler DL, Barrett T, Benson DA, et al. (2007) Database resources of the National Center for Biotechnology Information. Nucleic Acids Res 35:D5–D12 Wing RA, Ammiraju JS, Luo M, et al. (2005) The Oryza Map Alignment Project: the golden path to unlocking the genetic potential of wild rice species. Plant Mol Biol 59:53–62 Yao H, Guo L, Fu Y, et al. (2005) Evaluation of five ab initio gene prediction programs for the discovery of maize genes. Plant Mol Biol 57:445–460 Yuan Q, Ouyang S, Wang A, et al. (2005) The Institute for Genomic Research Osa1 rice genome annotation database. Plant Physiol 138:18–26 Zdobnov EM, Apweiler R (2001) InterProScan – an integration platform for the signature-recognition methods in InterPro. Bioinformatics 17:847–848

This page intentionally blank

I.3

Sequencing-based Measurements of mRNA and Small RNA Kan Nobuta1 and Blake C. Meyers1(* ü)

1 1.1

Introduction Rice Genomic Sequence Data

The whole-genome sequence of rice (Oryza sativa subsp. japonica cv Nipponbare) has been sequenced and the annotation of this genome has been performed and continues to be updated and improved by several organizations (Yuan et al. 2003; IRGSP 2005). In this chapter, we used the fourth annotation of the rice genome released from TIGR (The Institute of Genomic Research) to analyze our massively parallel signature sequencing (MPSS) data. In this version (TIGR v4.0), 62,827 gene models or transcription units (TUs) were identified, including 4,734 genes with multiple splice isoforms. Among the 55,890 genes, 13,237 were annotated as transposable element (TE)-related genes, while the rest (42,653) were annotated as protein-coding genes or non-TE genes. In addition to the japonica subspecies, a draft version of the genomic sequence of an indica subspecies (cv 93-11) has been completed and is available from the Beijing Genomic Institute (BGI) (Zhao et al. 2004). Because the genomic sequence of japonica rice is of such high quality and has been ordered and oriented into assembled “pseudochromosomes” with relatively few gaps, the japonica sequence has been used as a framework or scaffold for the alignment of the contigs of indica rice. This has enabled the identification of insertion/deletion sites (“indels”) and single nucleotide polymorphisms (SNPs), as well as large regions of sequence gaps missing primarily from the indica genome due to the draft status of that sequence (Feltus et al. 2004). We obtained the Nipponbare/93-11 alignment from TIGR (Drs. Wei Zhu and Robin Buell, personal communication). Based on this alignment, we mapped the gene annotations from the japonica-based TIGR rice genome (v4.0) onto the indica genome. This provided us with the starting material for comparisons and analyses of indica, japonica, and F1 1

Delaware Biotechnology Institute, University of Delaware, 15 Innovation Way, Room 230, Newark, DE 19711, USA e-mail: [email protected]; [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

23

24

K. Nobuta and B.C. Meyers

hybrid (of japonica and indica) MPSS. The resulting data are being used to examine gene expression based on our F1 hybrid MPSS data described below.

1.2

mRNA MPSS Technology and Plant Databases

The MPSS technology involves the cloning of distinct cDNA fragments onto microbeads and the parallel sequencing of millions of beads (Brenner et al. 2000). This sequencing reaction results in sequencing “reads” of 17 or 20 nucleotides derived from or “anchored by” the 3¢-most DpnII restriction site (GATC) of the cDNA fragments. In other words, every tag that is generated by MPSS from mRNA starts with a restriction site like DpnII. The number of occurrences of each of these nucleotide sequences (generally called “signatures” or “tags”) in a particular library represents the expression level of the transcripts; highly abundant transcripts may be represented thousands or even tens of thousands of times per million tags sequenced. Since the MPSS technology does not pre-select the transcripts or specific genes to be analyzed, unlike most microarray technologies, it is able to capture novel transcripts such as antisense transcripts, alternatively terminated isoforms, and intergenic transcripts (Meyers et al. 2004b). In order to create our rice mRNA MPSS database, the signatures were mapped onto the TIGR v4.0 genome and classified based on their position relative to annotated genes (Meyers et al. 2004a). The seven classes that we use can be grouped together to refer in more general terms to sense, antisense, and intergenic transcripts. In order to facilitate comparisons among libraries, each of which may have a substantially different total number of tags obtained by MPSS, the abundance value of each signature is normalized to one million TPM (transcripts per million). The details of this classification methodology as well as the full database schema showing how all of our various data are stored relative to one another can be found elsewhere (Meyers et al. 2004a; Nobuta et al. 2007a). We have generated 17 Arabidopsis mRNA MPSS databases, and from those data we have shown the capability of this technology to capture and measure the expression level of strongly and weakly expressed transcripts in Arabidopsis (Meyers et al. 2004b). In order to understand the transcriptional complexity in rice, 72 rice mRNA MPSS libraries were generated. The libraries were designed not only to answer basic science questions but also to address agronomically important questions. The libraries are grouped into five categories and our findings are described below (see Sect. 2).

1.3

Small RNA MPSS Technology and Databases

Small RNA molecules in plants and animals play a critical role in developmental stages and stress responses. Small interfering RNAs (siRNAs) and microRNAs (miRNAs) are the two major categories of these molecules, among several different

I.3 Sequencing-based Measurements of mRNA and Small RNA

25

types of small RNA molecules (Bartel 2004; Verdel et al. 2004). In order to characterize the complexity of these molecules, we modified the cloning steps of the regular MPSS technology, described above, to capture and sequence the small RNAs (Lu et al. 2005). First, size fractionation of RNA samples was performed using a polyacrylamide gel, and fragments 20 to 25 nucleotides in length were extracted. Second, RNA adaptors were ligated to the 5¢ and 3¢ ends of the small RNA molecules and the MPSS templates (cDNA) were generated by reverse transcriptase. These templates were treated in the same way as mRNA MPSS templates (cloned onto beads) and the parallel sequencing reactions were performed. While all the mRNA MPSS signatures start with GATC and the potential signatures can be pre-extracted from rice genome, small RNA MPSS signatures can start with any nucleotide sequence. Therefore, each signature was compared against the entire rice genome after the MPSS experiments and the coordinates of each perfectly matching location were recorded (Nobuta et al. 2007a). Since siRNAs are known to be associated with repetitive sequences, we ran RepeatMasker (http:// www.repeatmasker.org), Etandem, and Einverted (Rice et al. 2000), and stored the positional information in our databases. This allows us to analyze and calculate the correlation of small RNA and repetitive regions. The abundance of each signature was normalized to transcripts per quarter million (TPQ) instead of TPM, because the total number of small RNA MPSS signatures sequenced was roughly one quarter of that of mRNA signatures (Lu et al. 2005). The small RNA sequencing technology has been applied to wild-type Arabidopsis and an rdr2 mutant (Lu et al. 2005, 2006). From a collection of more than 1.6 million small RNAs that were sequenced from these two libraries, a large number of siRNAs were identified from wild-type tissue, but the rdr2 mutant tissue was enriched in miRNAs and provided the expression level of known and novel miRNAs (Lu et al. 2005, 2006). We have recently generated three rice small RNA libraries derived from various tissues. Here, we summarize the major preliminary findings from these libraries and discuss the potential application of these data to rice biotechnology and agriculture.

2

mRNA MPSS Data

Our rice mRNA MPSS libraries can be categorized into five different groups: untreated, abiotic-stressed, hybrid, biotic-stressed, and public libraries. The details of the libraries, such as number of sequencing runs, total signatures per library, or distinct signatures per library, can be found on our website (http://mpss.udel.edu). The public libraries are not discussed further in this chapter, but are available on our website and represent eight libraries from one experiment examining different varieties of varying seed quality, as well as a second experiment looking at insect damage. Together with our collaborators at the Ohio State University, Drs. Guo-liang Wang and R.C. Venu, we have obtained a total of 123,998,944 signatures that represent 419,156 distinct sequences (Fig. 1). After filtration, approximately one quarter of

26

K. Nobuta and B.C. Meyers

Reliable signatures 199,044 Total number of distinct signatures 419,156 Unreliable signatures 220,112

Hit to TIGR genome 121,484

Hits = 1 101,169

Hits > 1 20,315

NO hit to TIGR genome 77,560

Hit to TIGR genome 44,666

Hits = 1 37,314

Hits > 1 7,352 NO hit to TIGR genome 175,446

Fig. 1 Filter results for 72 mRNA MPSS libraries. A total of 419,156 distinct 17-base signatures from 72 rice mRNA MPSS libraries were processed according to three filters: “reliable,” “genomic match,” and “hits.” Numbers indicate distinct signatures that are separated by each filter or set of filters. The reliable filter, which separates signatures found in only one sequencing run, removes most sequencing errors, as does genome matching. The hit rate is an assessment of the number of perfect matches found in the genome for each distinct signature sequence

the distinct signatures (101,169) were defined as “reliable” and had a unique match to the rice genome (Fig. 1). Reliable signatures are those that have been identified from multiple sequencing runs; because we have four or more runs from each library, this filter is an effective way to remove sequencing errors and other poor quality data. Signatures with only a single match to the genome (“hits = 1”) are easiest to work with because the source of the expression data is not confounded by duplicated matches. Although the remaining signatures may contain variable information (e.g., alternative splicing, unannotated transcripts, expression of gene family members, etc.), here we have focused on the analyses of reliable signature with hits = 1.

2.1

Untreated mRNA Libraries

We have generated 18 untreated libraries, representing 12 diverse tissues and with four and two replicates from leaf and root, respectively. The majority of the signatures match to the sense-strand of the annotated genes (Table 1), as would be expected for normal protein-coding mRNA transcripts. These sense signatures correspond to 21,155 genes, which is less than half of the non-TE genes annotated by TIGR. Most of the genes with corresponding MPSS signatures have additional support, such as FL-cDNAs and/or ESTs, and have high similarity (BLASTP e-value < 10−7) to

I.3 Sequencing-based Measurements of mRNA and Small RNA Table 1 mRNA-derived MPSS signatures and corresponding genes from rice Description Total distinct Classa

27

Genes

1,2,5,7b 3,6c 4d

Gene-matched, sense strand 41,946 21,155 Gene-matched, antisense strand 8,911 6,871 Unannotated, intergenic region 12,211 — Total 63,068 22,101 a mRNA signatures were classified based on their genomic locations for the untreated MPSS libraries; only reliable signatures with one match to genome are indicated b 1, Exon sense; 2, 3′ UTR; 5, intron sense; 7, splice junction c 3, Exon antisense; 6, intron antisense d 4, Intergenic e Indicates totals calculated using signatures matching more than one genomic location. This represents an upper boundary for numbers in the “Total” row

Arabidopsis genes. In contrast, most of the genes without MPSS support have low similarity to Arabidopsis genes and do not have additional transcriptional support from other data sources. We have investigated the nature of the genes that have no MPSS support, and we found that many of them are associated with mutator-like elements (MULEs) (Nobuta et al. 2007b). MULEs represent one of several classes of gene-shuffling TEs that are known to transpose DNA fragments, including protein-coding regions, and form pseudogenes (Jiang et al. 2004; Juretic et al. 2005; Morgante et al. 2005). It is possible that the majority of the genes without any support (MPSS or other cDNA data) are pseudogenes and are not expressed in rice, or these may be expressed at very low levels or under specialized conditions. While this information is useful for genome annotation, the genes with MPSS support also represent a fundamentally important and informative data set for systems biology. Although similar information can be obtained from microarray technology, the deep sampling feature of MPSS allows the researchers to examine, in parallel, a wide range of gene expression levels. In addition, since MPSS is not subjected to cross hybridization, MPSS is capable of measuring the expression level of highly similar genes (if the corresponding MPSS signatures are polymorphic). For example, we have identified 34 TEs which are expressed only in pollen libraries, and for which the MPSS signatures are specific to a single copy. The expression of these TEs could have been missed or dismissed with microarray data analysis because most of the TEs have repetitive sequences and could cross hybridize to the same class of TEs. The analysis of such a set of genes will give us great insight to understand the mechanisms of tissue-specific gene expression in rice. Another distinct feature of MPSS technology, also important for systems biology, is that it can identify novel transcripts that are missed by annotation programs. We identified many genes with alternative termination and antisense transcripts (Table 1). In addition, more than 10,000 MPSS signatures were identified at intergenic regions (Table 1). These signatures could have derived from either (1) the mis-annotated 3′ untranslated region (UTR) of known genes, (2) novel protein-coding genes, or (3) non-coding RNAs (ncRNAs). Interestingly, there were 77 intergenic signatures

28

K. Nobuta and B.C. Meyers

which are expressed in all untreated libraries, suggesting the importance of the corresponding transcripts for “housekeeping”, as the activities of many constitutively expressed genes are called. We also investigated the correspondence of the intergenic signatures to known miRNAs and their non-coding RNA precursors from the Sanger miRNA registry (Griffiths-Jones 2004). Sixty-three intergenic mRNA signatures matched to these pol II-derived small RNA precursors and suggested the expression of these molecules in certain tissues and at defined developmental stages (Nobuta et al. 2007b). These results clearly suggest that substantial numbers of transcripts (both coding and non-coding) have yet to be identified in rice.

2.2

Abiotically Stressed Rice Libraries

Salt, cold, and drought stress can cause serious agricultural damage. We treated 2-week-old rice seedlings with these stresses and extracted RNA samples for MPSS. In order to understand the response to these stresses, the expression profiles of these libraries were compared to those of untreated young roots and leaves. Preliminary analyses of these libraries identified many transcripts, both annotated and novel, that are differentially expressed in the stressed tissues compared to the untreated tissues. The set of transcripts identified with this comparison may identify transcripts strongly induced or repressed under stress conditions, and engineering of these genes could ultimately lead to increased protection for rice from abiotic stress. Here we focus on the novel transcripts, especially antisense and alternative terminated transcripts, and discuss their potential functions. As mentioned in the Introduction, large numbers of alternative isoforms were identified and annotated in TIGR v4.0, based on cDNA data. Alternative termination, as well as alternative splicing, is an important mechanism for mRNA stability and to increase the diversity of proteins (Stamm et al. 2005; van Hoof and Green 2006). Since MPSS signatures represent the 3′-most DpnII site, MPSS data are an excellent source by which to identify transcripts with different lengths (3′ ends or termination sites) corresponding to the same gene. We took the advantage of this unique feature of MPSS and examined the effect of salt stress on the 3′ termination site. Many genes show substantial differences in both 3′ ends (based on the site of the MPSS signature) and expression levels between untreated and salt-treated leaves (Nobuta et al. 2007b). As shown in Fig. 2A, some genes lack potentially important domains in the shorter transcripts. This type of information is difficult to obtain with hybridization-based technologies and could be critical to understanding abiotic-stress responses in rice. Like alternative isoforms, natural antisense transcripts (NATs) are known to be important for mRNA stability (Jen et al. 2005; Wang et al. 2005). In order to investigate the function of antisense transcripts, we investigated the correlation of the expression level between the sense transcripts expressed in young leaves against the antisense transcripts expressed in the cold- treated young leaves. As shown in Fig. 2B, we identified sense/antisense pairs that show reciprocal expression. It is

I.3 Sequencing-based Measurements of mRNA and Small RNA

29

A

Os05g45180: MSP domain containing protein 1

2

3

4

5 5

NYL

6

7 4

3

8

2

1

0

0

0

0

36 TPM

NSL 26

0

0

0

0 TPM NYL: 101 TPM NL4: 57 TPM

B

Os02g05480: Mitogen-activated kinase homolog NTF3 1

2

3 NYL: 0 TPM NL4: 6 TPM

Fig. 2 Example of potential alternative termination and antisense transcript captured with MPSS. A Different transcripts and expression in young leaf (Nipponbare young leaf library, NYL) and salttreated young leaf (Nipponbare salt-treated leaf library, NSL) are indicated by triangles and corresponding numbers in boxes. Black and gray boxes represent exons and UTRs, respectively. In this particular case, the longer transcript, which corresponds to the annotated gene, is expressed in NYL, and the shorter one is expressed in NSL. The shorter transcript apparently lacks a portion of the 3¢ exons that contain a predicted transmembrane domain. B Black and gray triangles represent the sense and antisense transcript of this gene. In young leaf (NYL), the sense transcript is expressed at 101 TPM and the antisense transcript is not expressed. In contrast, the antisense transcript is expressed (6 TPM) in mature leaf (Nipponbare mature leaf 4 libraries combined, NL4) and the expression level of the sense transcript is reduced to 57 TPM from 101 TPM

possible that these transcription pairs form double-stranded RNA, inducing cleavage and degradation of both RNAs. In contrast to these transcripts, we identified many examples that did not show reciprocal expression patterns. These findings agreed with other reports and suggest the difficulty in making general statements about the function of NATs (Jen et al. 2005). Deep profiling of NATs will be essential to elucidating the function of this unusual regulatory system.

2.3

Hybrid Rice Libraries

The phenomenon known as “heterosis” is well documented not only in plants but also in other fields of biology. Heterosis, also known as hybrid vigor or outbreeding, is an increased strength of phenotype (such as vigor, strength, or yield) in the hybrid compared to either of its parents. One of the best examples is hybrid corn, which produces higher yields than the two inbred parental lines. Such corn sustains a multi-billion dollar seed business in the US. Like corn, inbred strains of rice also show heterosis (Luo et al. 2001). For example, the F1 hybrid of Nipponbare and 93-11 shows a significant increase in plant size and in number of tillers, as well as

30

K. Nobuta and B.C. Meyers

an increase in yield. Although there are several theories that explain this phenomenon, the molecular mechanism of heterosis is not well understood. In this post-genomic era, we are one step closer to understanding this agronomically important phenomenon. The genome sequence of the parent species and the deep expression profiling of F1 and its parents are important resources to assist in the elucidation of the mechanism of heterosis, although a more comprehensive analysis may require proteomics approaches to understand the interactions among proteins and protein complexes in the F1 progeny. As a first step, we took the advantage of the rice japonica and indica genome sequences, and generated four leaf, two root, and one meristem libraries from F1 hybrid rice and the same number of libraries from japonica and indica rice. MPSS offers some advantages over microarrays for this analysis because MPSS, as a sequence-based measurement, may be able to assess allele-specific expression in the F1 plants by taking advantage of SNPs and indels that distinguish the two parental alleles. The genome sequences of japonica and indica were aligned as described above in the Introduction. After the alignment, the MPSS signatures were categorized into several cases based on their presence within high-quality alignments, partial alignments, or the types of sequence polymorphisms that distinguish the two parental alleles. While the majority of the signatures have the same sequence between the two parent subspecies, we identified many signatures that have SNPs. Figure 3 is a graphical representation of the expression differences of japonica and indica signature pairs on chromosome 8 of F1 hybrid rice. The expression level of japonica and indica signatures in the F1 hybrid was compared with that of the same signatures in the parents and plotted along chromosome 8. The height difference between the round and triangle symbols indicates the differential expression between japonica and indica signatures (Fig. 3). Although some japonica and indica signature pairs show similar expression patterns, many

Log base 2 of the fold difference in F1 hybrid

8 F1/JPN 6

F1/IND

4 2 0 -2 -4 -6 -8 0 Mb

5 Mb

10 Mb

15 Mb

20 Mb

25 Mb

30 Mb

Coordinate on chromosome 8

Fig. 3 Allele-specific expression level in F1 hybrid. The expression level of signatures in F1 hybrid was compared with corresponding levels in the Nipponbare ( japonica, JPN) and 93-11 (indica, IND) parents. Only signatures that have SNPs and are located on chromosome 8 are shown. Coordinates of signatures are displayed on the X-axis and relative expression differences between japonica and indica signatures in F1 hybrid are indicated on the Y-axis in log2 scale. Gray bar around 13 Mb corresponds to the centromere

I.3 Sequencing-based Measurements of mRNA and Small RNA

31

are expressed quite differently in F1 hybrid (Fig. 3). The set of genes that show allele-specific expression (e.g., high expression of the japonica allele and low expression of the indica allele or vice versa, within a single F1 plant) may provide useful data to help elucidate the molecular basis of heterosis.

2.4

Biotically Stressed Rice Libraries

The most comprehensive set of MPSS libraries that we have generated is being used to understand the transcriptional component of the rice defense system which is employed against the two most devastating rice diseases: rice blast and bacterial blight, caused by Magnaporthe grisea and Xanthomonas oryzae pv oryzae, respectively. The experiments were designed using isogenic strains of rice that represent both susceptible and resistant plants to these diseases. In order to generate isogenic lines, Nipponbare was transformed with either the Xa21 resistance gene (R-gene) or the Pi-9 R-gene, and homozygous lines containing single-copy inserts were developed for each transgene (Wang et al. 1996; Qu et al. 2006). For the disease treatments, both susceptible and resistant plants were treated with the respective pathogens, and materials were collected at different time points after inoculation. Magnaporthe-treated libraries are particularly interesting because the pathogen also produces polyadenylated mRNA, providing the possibility of measuring expression profiles of both host and pathogen in a single sample. We believe that these data may help us to understand the transcriptional basis of the tactics used in both rice and Magnaporthe. A draft version of the genome sequence of Magnaporthe grisea has been completed using whole-genome shotgun sequencing; this has been made freely available from the Broad Institute at the Massachusetts Institute of Technology (Dean et al. 2005). The sequences in the MPSS libraries from the infected rice tissues were compared with both the rice and Magnaporthe genomes, and the MPSS signatures were classified as (1) rice-specific, (2) Magnaporthe-specific, and (3) common signatures. These three classes refer to the genome to which each sequence was perfectly matched. We identified 75,460 and 4,064 rice- and Magnaporthe-specific signatures, respectively, and 861 common signatures in our database. The relatively small number of Magnaporthe-specific signatures could be due to a dilution effect since the bulk of the material in the sample represented the rice leaves; however, it is also possible that this could be attributed to a lower transcriptional complexity of the fungus compared to the plant host, or to the large number of unsequenced gaps in the Magnaporthe genome. In each disease-treated rice library, we compared the expression levels of the rice-specific signatures to the mock-treated library and categorized them as up- or down-regulated transcripts. Table 2A displays the number of signatures that show 10-fold or more differences in abundance when the experiment and the control are compared. We performed the same analysis with Magnaporthe-specific signatures, taking advantage of other MPSS libraries that represent two different developmental stages in Magnaporthe, grown in the absence of rice (appressorium and mycelium).

32

K. Nobuta and B.C. Meyers

Table 2 Number of differentially expressed signatures in Magnaporthe-treated rice at various time points A Number of 10-fold up- or down-regulated signatures in resistant or susceptible rice in comparison to mock-treated rice 3h

6h

12 h

24 h

48 h

Resistant

Up 3,925 3,380 3,973 3,714 4,545 Down 4,183 5,153 4,827 5,205 3,768 Susceptible Up 4,325 3,446 3,668 4,049 4,094 Down 3,939 5,693 4,897 4,659 3,856 B Number of 10-fold up- or down-regulated Magnaporthe signatures in resistant or susceptible rice in comparison to Magnaporthe mycelium Resistant Susceptible

Up Down Up Down

3h

6h

12 h

24 h

48 h

11 1,543 7 1,535

17 1,546 34 1,492

31 1,517 20 1,513

37 1,500 23 1,538

18 1,538 10 1,541

The gene expression profile of the mycelium library was considered as the basal gene expression level during the infection and was compared with that of Magnaporthe-specific signatures derived from infected rice libraries (Table 2B). This analysis is quite preliminary, and the transcripts apparently “down-regulated” are probably an artifact of the dilution of Magnaporthe transcripts in the rice libraries mentioned above. However, there were numerous fungal-derived, up-regulated transcripts in the Magnaporthe-infected rice libraries compared to mycelium. Because these have changed above and beyond the dilution effect, these will be of significant interest in our follow-up experiments. The signatures and the transcripts identified here can be compared through our web interface in various ways, using our library analysis tools. Although it is necessary to perform additional biological experiments to understand the cascade of signal transduction events that results in the differential expression levels of these genes, the expression profiles of Magnaporthe and rice will help researchers identify critical players in these organisms under the conditions we have utilized, and these data will clearly add a new dimension to our understanding of host–pathogen interactions.

3

Small RNA MPSS

Since the discovery of miRNAs and siRNAs in the late 1990s, an explosion of research has taken place, although we are still in the early stages of understanding the complete biological significance of these molecules in eukaryotes. Small RNA molecules are known to be associated with a wide range of functions that include development, stress responses, and genome stability (Palatnik et al. 2003; Sunkar and Zhu 2004; Mallory et al. 2005). Both miRNAs and siRNAs regulate gene expression by forming double-stranded RNAs with their target, resulting in either

I.3 Sequencing-based Measurements of mRNA and Small RNA

33

cleavage and degradation of the target molecule or repression of translation. In addition, siRNAs play a role in directing DNA methylation in the nuclear genome and the production of heterochromatin. In both cases, the end result is a negative regulation of gene expression. Because of the high specificity of these RNAdirected events, the use of both siRNAs, and more recently the development of artificial miRNAs, has been and continues to be an important tool for applications in plant biotechnology (Baulcombe 2004; Schwab et al. 2006). Because of its high throughput, MPSS offers the advantage of deep sequencing compared to the low throughput of conventional sequencing approaches to discovering small RNA molecules (Lu et al. 2005, 2006). We have applied this technology

Rice inflorescence

Rice seedling

9 (5)

34 (7)

66 (29)

122 (102) 116 (70)

62,089 (934)

30,185 (62)

2,399 (4)

116,252 (145)

175 (60) 159 (113)

4,999 (835)

Arabidopsis inflorescence

34,087 (1,025)

Arabidopsis seedling

Fig. 4 Conservation of small RNAs in rice and Arabidopsis. A comparative genomics approach is being used to identify potential miRNAs. The small RNA signatures from rice inflorescence, rice seedling, Arabidopsis inflorescence, and Arabidopsis seedling libraries were compared. Numbers in each portion of the Venn diagram represent distinct signatures found in a set or subset of libraries. Signatures identified in Arabidopsis rdr2 mutant were compared for each section and are indicated in those boxes by numbers in parentheses

34

K. Nobuta and B.C. Meyers

to rice and generated three small RNA libraries: an inflorescence, seedling, and stem library. The total number of signatures sequenced and the number of distinct signatures identified from inflorescence was 1,731,548 and 201,764, respectively, and these were substantially greater than those of seedling and stem libraries (Nobuta et al. 2007b). The majority of these small RNAs correspond to repetitive regions, including annotated TEs; this suggests that these signatures represent siRNAs. Comparing the different rice tissues, the same trend was observed as for Arabidopsis, in which the inflorescence is much more complex than the seedling library, although the complexity of each of the libraries was greater in rice than in Arabidopsis (Lu et al. 2005). This is consistent with the larger genome size in rice that is the result of an expansion of transposons and retrotransposons, and these sequences are the primary sources (and targets) of endogenous siRNAs in plants. Since miRNAs are more consistently identified from library to library compared to siRNAs, and because miRNAs are much more highly conserved across plant species than siRNAs, we are using a comparative genomics approach to identify potential miRNAs. We compared the small RNA signatures that are expressed in inflorescence, seedling, and rdr2 mutant MPSS libraries from Arabidopsis with the rice libraries (Fig. 4). The proportion of the signatures from these libraries that were also identified in the rdr2 mutant increases as the number of compared libraries increases (Fig. 4). The signatures expressed in all five libraries (102 signatures) may be enriched for highly expressed and conserved miRNAs (Fig. 4). We believe that the application of MPSS data and other deep sequencing technologies, combined with genomic sequence data, provides a powerful tool to identify small RNAs such as miRNAs and siRNA clusters (possibly representing some epigenetically regulated loci) that play a major role in plant growth, development, and stress responses.

4

Conclusion

MPSS and other deep, short-read sequencing technologies that will soon be available have unique applications, including several advantages over other popular technologies such as microarrays and EST sequencing for transcriptional analyses. MPSS is capable of measuring the absolute expression level of almost all transcripts, including both mRNA and small RNA, and those that have not been previously annotated or identified. Here we have summarized our rice mRNA and small RNA data, and discussed the application of these data to rice biotechnology. At this point in time, our rice MPSS mRNA and small RNA data represent the largest amount of sequence- or tag-based profiling data and provide critical information for rice research. In the last few years, much more powerful tag-based sequencing technologies [e.g., 454 pyrosequencing technology (454), sequence-by-synthesis (SBS)] have been developed (Margulies et al. 2005). Some of these technologies promise much greater depth than MPSS, and at lower prices. As these technology platforms become more widely available, the types of applications we have described will be more common and used to address a wide variety of biological questions.

I.3 Sequencing-based Measurements of mRNA and Small RNA

35

References Bartel DP (2004) MicroRNAs: genomics, biogenesis, mechanism, and function. Cell 116: 281–297 Baulcombe D (2004) RNA silencing in plants. Nature 431:356–363 Brenner S, Johnson M, Bridgham J, et al. (2000) Gene expression analysis by massively parallel signature sequencing (MPSS) on microbead arrays. Nat Biotechnol 18:630–634 Dean RA, Talbot NJ, Ebbole DJ, et al. (2005) The genome sequence of the rice blast fungus Magnaporthe grisea. Nature 434:980–986 Feltus FA, Wan J, Schulze SR, Estill JC, Jiang N, Paterson AH (2004) An SNP resource for rice genetics and breeding based on subspecies indica and japonica genome alignments. Genome Res 14:1812–1819 Griffiths-Jones S (2004) The microRNA registry. Nucleic Acids Res 32:D109–D111 Jen CH, Michalopoulos I, Westhead D, Meyer P (2005) Natural antisense transcripts with coding capacity in Arabidopsis may have a regulatory role that is not linked to double-stranded RNA degradation. Genome Biol 6:R51 Jiang N, Bao Z, Zhang X, Eddy SR, Wessler SR (2004) Pack-MULE transposable elements mediate gene evolution in plants. Nature 431:569–573 Juretic N, Hoen DR, Huynh ML, Harrison PM, Bureau TE (2005) The evolutionary fate of MULE-mediated duplications of host gene fragments in rice. Genome Res 15:1292–1297 Lu C, Tej SS, Luo S, Haudenschild CD, Meyers BC, Green PJ (2005) Elucidation of the small RNA component of the transcriptome. Science 309:1567–1569 Lu C, Kulkarni K, Souret FF, et al. (2006) MicroRNAs and other small RNAs enriched in the Arabidopsis RNA-dependent RNA polymerase-2 mutant. Genome Res 16:1276–1288 Luo LJ, Li ZK, Mei HW, et al. (2001) Overdominant epistatic loci are the primary genetic basis of inbreeding depression and heterosis in rice. II. Grain yield components. Genetics 158:1755–1771 Mallory AC, Bartel DP, Bartel B (2005) MicroRNA-directed regulation of Arabidopsis AUXIN RESPONSE FACTOR17 is essential for proper development and modulates expression of early auxin response genes. Plant Cell 17:1360–1375 Margulies M, Egholm M, Altman WE, et al. (2005) Genome sequencing in microfabricated highdensity picolitre reactors. Nature 437:376 Meyers BC, Tej SS, Vu TH, et al. (2004a) The use of MPSS for whole-genome transcriptional analysis in Arabidopsis. Genome Res 14:1641–1653 Meyers BC, Vu TH, Tej SS, et al. (2004b) Analysis of the transcriptional complexity of Arabidopsis thaliana by massively parallel signature sequencing. Nat Biotechnol 22:1006–1011 Morgante M, Brunner S, Pea G, Fengler K, Zuccolo A, Rafalski A (2005) Gene duplication and exon shuffling by helitron-like transposons generate intraspecies diversity in maize. Nat Genet 37:997–1002 Nobuta K, Vemaraju K, Meyers BC (2007a) Methods for analysis of gene expression in plants using MPSS. In: Plant bioinformatics: methods and protocols (ed. D. Edwards). Humana Press, Totowa, pp 387–408 Nobuta K, Venu RC, Lu C, et al. (2007b) An expression atlas of rice mRNAs and small RNAs. Nat Biotech 25:473 Palatnik JF, Allen E, Wu X, et al. (2003) Control of leaf morphogenesis by microRNAs. Nature 425:257–263 IRGS (2005) The map-based sequence of the rice genome. Nature 436:793 Qu S, Liu G, Zhou B, et al. (2006) The broad-spectrum blast resistance gene Pi9 encodes a nucleotide-binding site-leucine-rich repeat protein and is a member of a multigene family in rice. Genetics 172:1901–1914 Rice P, Longden I, Bleasby A (2000) EMBOSS: the European Molecular Biology Open Software Suite. Trends Genet 16:276–277

36

K. Nobuta and B.C. Meyers

Schwab R, Ossowski S, Riester M, Warthmann N, Weigel D (2006) Highly specific gene silencing by artificial microRNAs in Arabidopsis. Plant Cell 18:1121–1133 Stamm S, Ben-Ari S, Rafalska I, et al. (2005) Function of alternative splicing. Gene 344:1 Sunkar R, Zhu J-K (2004) Novel and stress-regulated microRNAs and other small RNAs from Arabidopsis. Plant Cell 16:2001–2019 van Hoof A, Green PJ (2006) NMD in plants. In: Nonsense-mediated mRNA decay (ed. L.E. Maquat), pp. 1–6. Eurekah, Austin Verdel A, Jia S, Gerber S, et al. (2004) RNAi-mediated targeting of heterochromatin by the RITS complex. Science 303:672–676 Wang GL, Song WY, Ruan DL, Sideris S, Ronald PC (1996) The cloned gene, Xa21, confers resistance to multiple Xanthomonas oryzae pv. oryzae isolates in transgenic plants. Mol Plant Microbe Interact 9:850–855 Wang XJ, Gaasterland T, Chua NH (2005) Genome-wide prediction and identification of cis-natural antisense transcripts in Arabidopsis thaliana. Genome Biol 6:R30 Yuan Q, Ouyang S, Liu J, et al. (2003) The TIGR rice genome annotation resource: annotating the rice genome and creating resources for plant biologists. Nucleic Acids Res 31:229–233 Zhao W, Wang J, He X, et al. (2004) BGI-RIS: an integrated information resource and comparative analysis workbench for rice genomics. Nucleic Acids Res 32:D377–D382

I.4

Microarray-based Approaches to Rice Transcriptome Analysis Lei Li1 and Xing Wang Deng1(* ü)

1

Introduction

The advent of rice genome sequences and subsequent annotation is having two marked, complementary effects on the relatively new discipline of plant transcriptomics. First, the rice genome contains approximately 60,000 annotated protein-coding loci that top other full sequenced genomes. Although masking protein related transposable elements considerably reduces this number, abundant rice- or cereal-specific genes exist. Second, the application of various transcriptome-profiling techniques in rice has generated a large number of putative transcripts that have yet to be incorporated into the genome annotation. The unprecedented complexity and impending completeness of the transcriptome data are simultaneously demanding and creating opportunities for new avenues of discovery using novel, high-throughput approaches. In this chapter, we review the recent advances in rice microarray systems and the application of these systems in deep and comprehensive transcriptome analysis.

2

Microarray Platforms and Methods

For the scope of this chapter, microarrays are defined as transcript analysis systems involving hybridization of nucleic acid-derived targets to multiple DNA probes immobilized on solid surface. First described by Schena et al. (1995), the earlier microarrays have cDNA elements (such as PCR amplicons or EST clones) representing known or predicted genes. The elements are robotically spotted onto glass microscope slides coated with derivatized silane. Nowadays the DNA elements are increasingly produced by synthesis as single-stranded oligonucleotides 30–70 bases in length, depending on the array platform. Compared with the cDNA arrays,

1

Department of Molecular, Cellular and Developmental Biology, Yale University, New Haven, Connecticut 06520, USA e-mail: [email protected]; [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

37

38

L. Li and X.W. Deng

oligonucleotide-based arrays have improved consistency of data acquisition and increased robustness for cross-study comparison. Currently there are three commercial sources for microarrays in which elements are synthesized in situ. Most widespread are the GeneChips produced by Affymetrix, where the array elements are sets of 25-mer oligonucleotides. Synthesis of elements on the Affymetrix platform involves progressive photodeprotection of sites at oligonucleotide extension on the array surface through a predefined series of physical masks, followed by conventional DNA synthesis chemistries. In contrast, for NimbleGen arrays, photodeprotection is achieved by steering the UV light onto the array surface using a matrix of micromirrors integrated with a microchip, such that each micromirror is individually addressable and controllable (Singh-Gasson et al. 1999). Because no masks are required, NimbleGen arrays can include elements with variable length and thus offer greater flexibility than GeneChips. The third source is Agilent arrays produced through ink-jet-based deposition of nucleotide precursors at the locations of element synthesis (Hughes et al. 2001). As technologies advance, the number of elements accommodated on the arrays is increasing. For instance, genome tiling arrays (see below) from Affymetrix can hold as many as 6.4 million independent elements on a single array (Mockler and Ecker 2005). Microarrays have become the method of choice for high-throughput analysis of global gene expression and for genome-wide mapping of functional elements. A microarray experiment is typically done by starting with hybridization of the appropriate arrays to fluorescently labeled targets, which are specific preparations of nucleic acids purified from cells. The targets can be either cDNA or cRNA, with or without intervening amplification steps, for analyzing mRNA or non-coding transcript accumulation within the tissues of interest. For non-expression arrays, the targets are in the form of DNA fragments associated with certain proteins that are immunoprecipitated by antibodies recognizing the proteins. The hybridized microarrays are then scanned and the images captured. Data preprocessing, which includes image analysis, data normalization and transformation, is then carried out. Different methods of data preprocessing are used, depending on the array platform and application. For oligonucleotide genome tiling arrays and many non-expression arrays, single-color measurements are performed using special software supplied by the array manufacturer (e.g. Stolc et al. 2005). The other microarray platforms generally accommodate two-color fluorescence hybridization and scanning, with the Cy3 and Cy5 fluorescence measured separately and merged to produce a composite image (Fig. 1). Analysis of the array image to appropriately quantify the intensity values of the pixels corresponding to the site of hybridization is still a topic of vigorous inquiry, although most image-processing methods give satisfying results (Allison et al. 2006). Another important preprocessing step is data normalization and transformation, which allows cross-array comparison and the control of extraneous variation among experiments. Total signal output from an individual array is generally considered proportional to total nucleic acid input, and is routinely employed for data normalization for tiling and non-expression arrays, although this may not always be appropriate for expression arrays.

I.4 Microarray-based Approaches to Rice Transcriptome Analysis

39

Fig. 1 Data analysis for expression arrays. A Scanned image of a Yale–BGI rice oligonucleotide array hybridized with Cy3- and Cy5-labeled mRNA. B Volcano plot is used to simultaneously look at fold change and statistical significance of the preprocessed hybridization data. The name stems from the volcano shape of the plot. In the example illustrated, log2 (fold change) and minus log10 (p value) are shown on the abscissa and ordinate, respectively. C Hierarchical clustering display of differentially expressed genes in three rice RNA samples. Microarray expression data are often classified from hierarchical cluster analyses of both genes and samples

3

Applications of Microarrays in Rice Transcriptome Analysis

In rice transcriptome studies conducted so far, virtually all platforms have been employed to construct arrays. Below, we briefly review some of the available rice arrays and the results from using these arrays in transcriptome profiling. For the

40

L. Li and X.W. Deng

sake of simplicity, we divided all rice microarray systems into three types: surrogation expression arrays, genome tiling arrays, and SuperSAGE arrays, based on the strategy for designing the array elements and preparing the targets.

3.1

Rice Surrogation Expression Arrays

The term surrogation expression arrays refers to arrays in which the probes are selected from a predefined set of known or predicted genes. Typically the arrays are used for inference of differential gene expression in a test sample against the reference sample (Fig. 1). Following hybridization and data preprocessing, genes having fold change above a cut-off that is statistically significant are considered differentially expressed between two samples. The differentially expressed genes are then classified by either placing genes into pre-existing categories (supervised classification) or developing a set of categories into which the genes are placed (unsupervised classification). The earliest rice surrogation expression arrays contain rice ESTs as probes to detect and rank genes that present at different steady-state levels between biological samples. For example, the microarray developed by the Rice Microarray Project of Japan was a PCR amplicon array containing 1265 elements derived from rice EST clones (Yazaki et al. 2000). This array has been since updated to contain 8987 elements (Yazaki et al. 2003). Several other spotted cDNA arrays have been used to profile the rice transcriptome under various conditions. These studies, to name a few, include the use of a 1728-element array to study gene expression during the initial phase of salt stress in rice (Kawasaki et al. 2001); and 10K-element arrays to profile the rice transcriptome associated with pollination and fertilization (Lan et al. 2004), during anther development (Wang et al. 2005), at the early stage of nitrogen starvation (Lian et al. 2006), and in hybrid rice (Huang et al. 2006). By September 2003, the Rice Full-Length cDNA Consortium had obtained the complete sequences of 32,127 full-length cDNA clones (Kikuchi et al. 2003). These full-length cDNA sequences were used by Agilent to design 60-mer probes representing 22,000 rice transcripts. The first published work resulting from this array was the gene expression profiles of abscisic acid- and gibberellin-responsive genes in rice (Yazaki et al. 2004). Using the same array, Furutani et al. (2006) conducted global transcriptome profiling of the early stages of rice panicle development from phase transition to floral organ differentiation. Three hundred and fifty seven out of the 22,000 genes represented on the array were identified as differentially expressed in the early stages of panicle development. Based on temporal expression patterns, a fairly small number of genes, which were significantly enriched with transcription factors, were found to be upregulated in the shoot apical meristems immediately after phase transition (Furutani et al. 2006). Rice oligonucleotide arrays were also made on the Affymetrix GeneChip platform. According to the description in the Gene Expression Omnibus (GEO)

I.4 Microarray-based Approaches to Rice Transcriptome Analysis

41

database at the National Center for Biotechnology Information (Barrett et al. 2005), the current version of this array contains probes representing 51,279 transcripts. Early works using the rice GeneChip array include studies to profile rice gene expression patterns associated with grain filling (Zhu et al. 2003) and drought tolerance (Hazen et al. 2005). Using a half-genome rice microarray consisting of 23,000 elements, gene expression in rice aleurone cells was characterized. Approximately half of the array elements generated signal when hybridized with RNA from aleurone subjected to different treatments. Comparison with gene expression patterns in aleurone of the dwarf1 mutant indicates that changes in transcript abundance were smaller in the mutant (Bethke et al. 2006). In the public sector, the US National Science Foundation (NSF) supported a project in which a rice 45K oligonucleotide array was designed using gene models from the Institute for Genomic Research (TIGR) Rice Annotation Database that have EST and/or full-length cDNA support. This array contains 43,312 oligonucleotide probes corresponding to 44,974 japonica rice transcripts. A 70-mer microarray covering 41,754 annotated indica genes was developed through a Yale–Beijing Genome Institute (BGI) collaboration (Ma et al. 2005). Using the Yale–BGI array, expression of genes in six representative rice organs (seedling shoots, tillering-stage shoots and roots, heading and filling-stage panicles, and suspension culture cells) was analyzed. Expression of 86% of the 41,754 genes was detected. A large percentage of the rice gene models that lack significant Arabidopsis homologues were found expressed (Ma et al. 2005). Using the same microarray, expression of 20% rice genes was found regulated by white light (Jiao et al. 2005a). Global comparison of expression profiles between rice and Arabidopsis revealed a higher correlation of genome expression patterns in constant light than in darkness, suggesting that the genome expression profile of photomorphogenesis is more conserved (Jiao et al. 2005a). Most recently, this microarray was used to examine global transcriptional reorganization during the development of somatic embryos, shoots, and roots from cultured cells in rice (Su et al. 2006). MicroRNAs (miRNAs) are a class of evolutionarily conserved small non-coding RNAs that play important regulatory roles in genome expression by targeting mRNAs for cleavage or translational repression (Bartel 2004; He and Hannon 2004). Based on a previous method of small RNA sequencing (Wu et al. 1996), Liang et al. (2005) developed a quantum dot labeling procedure to directly label miRNA for hybridization with corresponding complementary oligonucleotide probes immobilized on glass slides, which can be detected and analyzed by measuring fluorescence of the quantum dots. Analysis of a prototype microarray containing probes interrogating 11 miRNA genes of rice showed that the detection limit for miRNA was ~0.4 fmol and the detection dynamic ranged by about two orders of magnitude (Liang et al. 2005). These results indicate that microarrays interrogating miRNA genes have the potential to profile transcription of miRNA genes, the success of which should help in deciphering miRNA functions and regulation, and contribute much to the studies of the overall transcriptome in rice.

42

3.2

L. Li and X.W. Deng

Rice Genome Tiling Arrays

The microarray systems described above employ a surrogation strategy; that is, the selection of a sequence optimal for the design of array elements relies on the availability of an annotation from which oligonucleotides are chosen to represent the known or predicted genes. However, there is abundant evidence that there are transcription units outside of the annotated protein-coding genes. For example, mapping of the initial collection of 28,469 rice full-length cDNAs indicates that approximately 24% of the cDNAs cannot be linked to a tentative protein function (Kikuchi et al. 2003). One approach to systematically identify all the transcripts from a genome is to use a set of tiling arrays in which all the genome sequences can be represented. Genomic tiling arrays involve the generation of a ‘tile path’ made up of progressive oligonucleotide tiles that represent a target genome region or the entire genome sequence (Fig. 2). These probes may overlap, lay end to end, or be spaced at regular intervals. The average nucleotide distance between the centers of neighboring probes are called the ‘step,’ which defines the resolution of the tiling arrays. Recent advances in microarray technologies, particularly the high-density oligonucleotide microarrays that contain short oligonucleotide probes synthesized in situ by photolithography, allow oligonucleotide arrays to be made with several hundred thousand to several million discrete features per array (Mockler and Ecker 2005). This makes it feasible to synthesize probes tiling complex genomes within a manageable number of arrays. These probes are immobilized on glass slides and are used to hybridize with fluorescence-labeled RNA samples. Hybridization intensity of each probe can be retrieved and integrated, leading to the identification of transcribed regions of the genome (Fig. 2). Genome tiling arrays have been used in model systems with full genome sequence available. For example, the first genome-wide transcription study using tiling microarrays was performed in Escherichia coli using 25-mer oligonucleotides with 6- and 30-nucleotide (nt) steps for intergenic and coding regions, respectively (Selinger et al. 2000). The first reported human whole genome tiling experiment involved 36-mer oligonucleotides with 46-nt steps to interrogate mRNA derived from human liver tissue (Bertone et al. 2004). In plant, the Arabidopsis thaliana genome was probed by 8-nt-step, 25-mer tiling microarrays (Yamada et al. 2003). Results from these studies and others show that for targets derived from known transcripts, such as mRNAs from annotated genes, the tiling array hybridization pattern identifies the transcriptional start and stop sites, the locations of introns, and the occurrence of alternative splicing (Yamada et al. 2003). Tiling array analysis therefore provides a valuable means of confirming gene annotations. On the other hand, it also reveals a large number of candidate transcripts, such as extensive transcripts of antisense to annotated genes, to which no conventional functions have yet been assigned (Yamada et al. 2003; Bertone et al. 2004). In rice, tiling array analysis started with chromosomes 4 and 10 (Jiao et al. 2005b; Li et al. 2005), which were among the first to be completely sequenced.

I.4 Microarray-based Approaches to Rice Transcriptome Analysis

43

Fig. 2 Application of genome tiling microarrays. Genome tiling arrays involve the generation of a virtual tile path representing a target genome that typically consists of oligonucleotide probes. These probes may overlap, lay end to end, or be spaced at regular intervals, and are immobilized on the surface of glass slides at a high feature density. Hybridization with fluorescence-labeled mRNA, non-coding RNA, or immunoprecipitated DNA samples generates signals that can be retrieved and analyzed to infer the identity and abundance of the input nucleic acids

In one study, chromosome 4 was tiled with PCR-generated probes representing on average 3 Kb genomic DNA fragments. Six representative rice organ types were examined using this microarray to catalog the transcribed regions and to reveal organ- and developmental stage-specific transcription patterns. This analysis provided expression support for 82% of the annotated gene loci of the chromosome and detected an additional 1643 loci with potential transcriptional activity (Jiao et al. 2005b). Comparison of the tiling array data with cytologically defined chromatin features indicated that in juvenile-stage rice the euchromatin is more

44

L. Li and X.W. Deng

actively transcribed than the transposon-rich heterochromatin. Interestingly, increased transcription of transposon-related genes in certain heterochromatic regions was observed in suspension-cultured cells as well as in tissues from adult plants (Jiao et al. 2005b). These results thus indicate a close correlation between transcriptional activity and chromosome organization that implies a layer of transcriptional regulation at the chromosome level. Recently, oligonucleotide tiling microarrays were developed based on the NimbleGen platform, representing both the japonica and indica genomes (Li et al. 2005, 2006, 2007; Stolc et al. 2005). The arrays contain 36-mer oligonucleotide probes with a 10-nt space on average between neighboring probes (Stolc et al. 2005). The rice tiling arrays were hybridized with a pooled mRNA target derived from seedling root, seedling shoot, panicle and suspension-cultured cells, and hybridization signals were correlated with the transcriptionally active regions (TARs) of the genome by alignment of the probes to the chromosomal coordinates. The tiling array data were used to detect transcription of the majority of the annotated gene models. For example, of the 43,914 non-transposable element protein-coding gene models from the improved indica whole genome shotgun sequence (Yu et al. 2005), transcription of 35,970 (81.9%) gene models was detected (Li et al. 2006). Consistent with the results from tiling microarray analysis in other model organisms, significant transcriptional activities were detected in the annotated intergenic regions of the rice genome (Li et al. 2006, 2007). Systematic scoring of indica tiling array data identified 5464 unique novel TARs in the intergenic regions using a set of stringent criteria (Li et al. 2006). These novel TARs were validated by several independent experimental means, including RT-PCR experiments, alignment against the rice ESTs, analysis of their coding content, and their association with simple sequence repeats (Li et al. 2006). Results from these analyses indicate that the novel TARs compositionally resemble the exonic regions and thus provide a reliable estimation of additional transcribed genomic loci beyond the predicted exons. In a similar vein, when coupled with procedures to specifically purify non-coding RNA species, tiling arrays can be used to identify the non-coding portion of the rice transcriptome (Fig. 2). Besides directly detecting components of the transcriptome, genome tiling arrays can be used to map other functional elements of the genome that are important for understanding gene regulation. For example, genome tiling arrays can be used in the ChIP-chip (chromatin immunoprecipitation coupled with microarray analysis) approach to identify the binding sites of rice transcription factors (TFs) that are important for regulation of gene expression. As yet another example, DNA and the core histone proteins are organized into nucleosomes that form the higher-ordered structure of chromatin in eukaryotes. Covalent modifications to DNA and the N-terminal tails of histones are conserved epigenetic markings involved in many important biological processes. Again, the genome-tiling arrays can be used in the ChIP-chip approach to map the sequences associated with DNA/histone modifications in rice (Fig. 2). Knowledge from these studies and others should help fill in the link between transcriptional landscape and epigenetic markings in the rice genome.

I.4 Microarray-based Approaches to Rice Transcriptome Analysis

3.3

45

The Rice SuperSAGE Array

Serial analysis of gene expression (SAGE) is a tag-based method that allows both qualitative and quantitative evaluation of thousands of genes in parallel without any prior sequence or annotation information (Velculescu et al. 1995). In rice, SAGE was first applied to profiling expressed genes in seedlings (Matsumura et al. 1999). Inherent to tag-based methods, one limitation of the SAGE procedure is that the assignment of 14 bp (21 bp in the LongSAGE procedure; Saha et al. 2002) tags to duplicated genes or repeated sequences is problematic, especially for complex genomes. To circumvent this limitation, a new procedure called SuperSAGE was developed in which the type III restriction endonuclease EcoP15I is used to generate tags of up to 26 bp from the 3¢ end of cDNAs (Matsumura et al. 2003). SuperSAGE has been successfully used in several expression profiling studies, including one in which the gene expression profiles of blast-infected rice leaves was investigated (Matsumura et al. 2003). The increased tag size also made possible the combination of SAGE and microarray to make the so-called SuperSAGE array through a direct tag-to-probe conversion approach (Matsumura et al. 2006). Because the SAGE tags are not limited to annotated or known transcripts, the SuperSAGE array is essentially an open-ended design that can be used to verify and profile all putative novel transcripts identified by SAGE studies. Recently, a prototype SuperSAGE array was developed on the NimbleGen platform that contains oligonucleotide probes based on SuperSAGE tags representing 1000 rice transcripts (Matsumura et al. 2006). When hybridized to labeled cRNA from rice leaves and suspension-cultured cell, the SuperSAGE array produced highly reproducible hybridization signals. The observed overall gene expression trend for the 1000 assayed genes in leaf and suspension-cultured cells was similar for both the SuperSAGE analysis and the SuperSAGE array. Specifically, comparison of hybridization signal values between the two tissue samples revealed that almost 90% of the statistically differentially presented tags were also detected as differentially expressed by the array (Matsumura et al. 2006). These results provide a proof-ofconcept for the SuperSAGE array, which should soon see broader applications in rice transcriptome profiling and related efforts.

4

Technical Challenges and Resources for Using Rice Microarray Systems

As for all large-scale, high-throughput methods, considerable importance should be attached to the characterization of the statistical significance of the results generated from microarray analysis. A number of statistical designs and tools have been developed for dealing with microarray experiments, as well as corresponding methods and software for analyzing the resultant data. Consensus on the statistical methods is emerging in many key areas of microarray analysis, such as experiment

46

L. Li and X.W. Deng

design, data preprocessing, inference, and other analyses. The reader is referred to reviews elsewhere that specifically cover statistical considerations of microarray experiments (e.g. Allison et al. 2006). As part of scientific discoveries, it is required that the microarray data be reliable and reproducible not only between platforms but also between laboratories. Thus, selection of the most effective microarray platform can be critical, depending on the purpose and scope of the studies pursued. Most microarray platforms provide reasonably robust and reproducible results for gene expression measurements. A number of cross-platform comparisons (e.g. Larkin et al. 2005) have been carried out for expression arrays, which highlight multiple issues that should be adequately considered. For genome tiling arrays and non-expression arrays, no cross-platform studies have yet been performed. Furthermore, the complex nature of microarray data, such as experimental conditions and target preparation, can have profound effects on the data obtained. Thus another important consideration in microarray studies is to include sufficient biological replicates, which allow the effects of measurement variability and biological differences between cases to be estimated and reduced. Many tools for microarray data preprocessing and analysis have been developed and can be used for rice microarray systems. For example, ExpressYourself is a fully integrated platform for processing microarray data in an automated fashion (Luscombe et al. 2003). It can be used to correct the background array signal, normalize the foreground signals from both Cy3 and Cy5 channels, filter problematic features, assess the quality of individual and replicate experiments, score levels of differential hybridization, and combine the results of replicate experiments (Luscombe et al. 2003). The modular architectural design of ExpressYourself has the advantage of allowing various types of microarray analysis algorithms to be readily incorporated as they are developed. The processed data from ExpressYourself are presented using a web-based graphical interface to facilitate comparison with the original array images (Luscombe et al. 2003). To compensate for the relative dearth of analytical tools tailored for NimbleGen arrays, a user-customized software bundle, call NMPP, was developed that integrates several established modules and algorithms for analyzing data from custom-designed NimbleGen arrays (Wang et al. 2006). The NMPP bundle uses a command-line-based integrative processing procedure that comprises five major functional components, namely the raw microarray data parsing and integrating module, the array spatial effect smoothing and visualization module, the probe level multi-array normalization module, the gene expression intensity summarization module, and the gene expression status inference module (Wang et al. 2006). Because all individual modules in NMPP can be independently customized, it is applicable to regular gene expression arrays and genome tiling arrays, as well as other types of non-expression arrays, for the purpose of raw data preprocessing. In comparison to the number of rice microarrays available, relatively few reports on its use have been published. This may reflect the combinative effect of the difficulty of analyzing large-scale data sets and the stringent standard for publishing

I.4 Microarray-based Approaches to Rice Transcriptome Analysis

47

microarray studies. Nevertheless, gene expression data in rice are accumulating in databases designed to hold raw and normalized data from expression profiles generated from various array platforms. For example, expression data obtained from the two cDNA-based microarrays developed by the Rice Microarray Project are in the Rice Expression Database (http://red.dna.affrc.go.jp/RED/). The Rice Oligonucleotide Microarray (ROMA) expression database at TIGR is designed to support expression data from all publicly available rice array platforms, including the NSF 45K oligonucleotide array, Agilent rice arrays, Affymetrix rice arrays, and the Yale–BGI oligonucleotide array (http://www.ricearray.org). It is noteworthy that the raw expression data deposited into ROMA are preprocessed using in-house methods and thus the resultant data may differ from that in other expression databases. Currently, there are two main repositories for general microarray data deposition and dissemination: the GEO database (Barrett et al. 2005) and ArrayExpress at the European Bioinformatics Institute (Brazma et al. 2003). These expression data repositories contain several expression platforms for plants but are not plantspecific and contain expression data for a wide range of species. Importantly, nonexpression microarray data in rice are probably only available in these two public repositories. Both public repositories provide basic search capabilities. For instance, GEO allows for the retrieval of gene profiles for several experiments (Barrett et al. 2005), while ArrayExpress offers the Expression Profiler online analysis tool (Brazma et al. 2003). Data mining tools are required to navigate the large quantity of microarray data now present in public databases. Most tools were developed for expression data to identify experiments of interest as well as expression profiles for genes of interest. Available on-line expression analysis tools provide some capability for expression data analysis. For example, Expression Profiler is a web-based platform for microarray gene expression and other functional genomics-related data analysis (Brazma et al. 2003). Data analysis components of Expression Profiler include expression data preprocessing, missing value imputation, filtering, clustering methods, visualization, significant gene finding, between-group analysis, and other statistical components that are integrated with the ArrayExpress repository of expression data (Brazma et al. 2003; Kapushesky et al. 2004). Several comprehensive analytic tools were developed specifically for the Arabidopsis expression data. For example, Genevestigator is a database and webbased data mining interface for GeneChip expression data to direct gene functional discovery. It supports bi-directional query of the database to retrieve expression patterns of individual genes throughout chosen experimental conditions or to identify genes specifically expressed in selected conditions (Zimmermann et al. 2004). Mapman was developed as a tool to display genomics data onto diagrams of biological processes and metabolic pathways (Thimm et al. 2004). These tools can be downloaded and installed locally to analyze large-scale Arabidopsis expression data. In the context of rice expression data analysis, these tools offer some capacity, allowing rice genes that have orthologs in Arabidopsis to be directly analyzed.

48

5

L. Li and X.W. Deng

Future Perspectives

Rice genome annotation and transcriptome profiling have benefited from studies utilizing microarray systems, as discussed above. Experimental data resulting from these microarray systems have been used to verify annotated gene models, to facilitate functional assignment of predicted genes, to generate candidate transcripts not included in current annotations, and to provide insight into the evolutionary relation of the rice transcriptome with those of other plant species. However, additional research efforts would be desirable in several areas of microarray studies to more effectively complement genome annotation efforts and to accelerate functional discovery. The first area would be the development of a true whole genome microarray to facilitate the generation of a reference expression data set of cell, tissue, or developmental-specific gene expression. Such a rice gene expression compendia would serve as a starting point for cataloging gene expression throughout the life cycle of rice, in mutants disrupting specific pathways, or under treatments soliciting specific transcription programs that will aid in the identification of expression and regulatory networks. In the vast majority of microarray experiments, RNA samples are typically prepared from tissues or organs that contain mixtures of many different cell types. Consequently, results obtained from these experiments provide very limited spatial resolution, leaving a blind spot in our understanding of the transcriptome (Galbraith and Birnbaum 2006). Thus, the second area for improving microarraybased transcriptome analysis in rice is to increase the spatial resolution, which requires a means to separate organs or tissues into their constituent cell types prior to RNA isolation. Two major techniques, fluorescence-activated cell sorting (Birnbaum et al. 2005) and laser capture microdissection (Asano et al. 2002), have been employed in conjunction with RNA amplification to deconvolute the transcriptome of specific cell types from the organs or tissues in plants. It is expected that cell-specific transcriptome profiling in rice will yield a rich resource for studies of gene function and regulation. Another area would be increased comparative analyses of microarray data from different plant species. Just as comparative genomics has enhanced genome annotations, comparative analyses of microarray data should prove valuable in elucidating plant transcriptome components, regulation, and evolution. Currently, plant gene expression data are spread and scattered in multiple databases, often separated according to species, which hinders cross-species comparisons. Identification of orthologs among different plant species and improvement of largescale databases and clustering tools should accelerate the comparative efforts. Coupled with comparative transcriptome analysis will be the improved structural annotation of plant genomes and an increased number of genes with a known function. These efforts constitute an important step towards identifying the transcriptome component common and unique to different plants that underlies their great physiological and chemical diversity. Finally, to fully exploit rice microarray data and to embrace a systems approach, alternative data types such as those generated from the non-expression application of

I.4 Microarray-based Approaches to Rice Transcriptome Analysis

49

genome tiling arrays will need to be integrated with the expression data. The integrated data should provide a hierarchy of information that is supposed to be processed in living cells through regulatory networks to program genome transcription. Identification of cis-regulatory sequences shared by co-regulated genes or bound by specific TFs and sequences associated with DNA or histone modifications not only fills in the link between transcriptional landscape and epigenetic markings, but also, to a first approximation, reveals the hardwired ‘control logic’ in the genome that encodes the transcription programs to be implemented by specific TFs and other functional elements. Integration of these data, in conjunction with computational modeling and model testing, should help to identify the motifs and design principles of the transcriptional regulatory networks in the rice genome.

References Allison DB, Cui X, Page GP, Sabripour M (2006) Microarray data analysis: from disarray to consolidation and consensus. Nat Rev Genet 7:55–65 Asano T, Masumura T, Kusano H, et al. (2002) Construction of a specialized cDNA library from plant cells isolated by laser capture microdissection: toward comprehensive analysis of the genes expressed in the rice phloem. Plant J 32:401–408 Barrett T, Suzek TO, Troup DB, et al. (2005) NCBI GEO: mining millions of expression profiles – database and tools. Nucleic Acids Res 33:D562–D566 Bartel DP (2004) MicroRNAs: genomics, biogenesis, mechanism, and function. Cell 116:281–297 Bertone P, Stolc V, Royce TE, et al. (2004) Global identification of human transcribed sequences with genome tiling arrays. Science 306:2242–2246 Bethke PC, Hwang YS, Zhu T, Jones RL (2006) Global patterns of gene expression in the aleurone of wild-type and dwarf1 mutant rice. Plant Physiol 140:484–498 Birnbaum K, Jung JW, Wang JY, et al. (2005) Cell type-specific expression profiling in plants via cell sorting of protoplasts from fluorescent reporter lines. Nat Methods 2:615–619 Brazma A, Parkinson H, Sarkans U, et al. (2003) ArrayExpress – a public repository for microarray gene expression data at the EBI. Nucleic Acids Res 31:68–71 Furutani I, Sukegawa S, Kyozuka J (2006) Genome-wide analysis of spatial and temporal gene expression in rice panicle development. Plant J 46:503–511 Galbraith DW, Birnbaum K (2006) Global studies of cell type-specific gene expression in plants. Annu Rev Plant Biol 57:451–475 Hazen SP, Pathan MS, Sanchez A, et al. (2005) Expression profiling of rice segregating for drought tolerance QTLs using a rice genome array. Funct Integr Genomics 5:104–116 He L, Hannon GJ (2004) MicroRNAs: small RNAs with a big role in gene regulation. Nature Rev Genet 5:522–531 Huang Y, Zhang L, Zhang J, et al. (2006) Heterosis and polymorphisms of gene expression in an elite rice hybrid as revealed by a microarray analysis of 9198 unique ESTs. Plant Mol Biol 62:579–591 Hughes TR, Mao M, Jones AR, et al. (2001) Expression profiling using microarrays fabricated by an ink-jet oligonucleotide synthesizer. Nat Biotechnol 19:342–347 Jiao Y, Ma L, Strickland E, Deng XW (2005a) Conservation and divergence of light-regulated genome expression patterns during seedling development in rice and Arabidopsis. Plant Cell 17:3239–3256 Jiao Y, Jia P, Wang X, et al. (2005b) A tiling microarray expression analysis of rice chromosome 4 suggests a chromosome-level regulation of transcription. Plant Cell 17:1641–1657 Kapushesky M, Kemmeren P, Culhane AC, et al. (2004) Expression Profiler: next generation – an online platform for analysis of microarray data. Nucleic Acids Res 32:W465–W470

50

L. Li and X.W. Deng

Kawasaki S, Borchert C, Deyholos M, et al. (2001) Gene expression profiles during the initial phase of salt stress in rice. Plant Cell 13:889–905 Kikuchi S, Satoh K, Nagata T, et al. (2003) Collection, mapping, and annotation of over 28,000 cDNA clones from japonica rice. Science 300:1566–1569 Lan L, Chen W, Lai Y, et al. (2004) Monitoring of gene expression profiles and isolation of candidate genes involved in pollination and fertilization in rice (Oryza sativa L.) with a 10K cDNA microarray. Plant Mol Biol 54:471–487 Larkin JE, Frank BC, Gavras H, Sultana R, Quackenbush J (2005) Independence and reproducibility across microarray platforms. Nat Methods 2:337–343 Li L, Wang X, Xia M, et al. (2005) Tiling microarray analysis of rice chromosome 10 to identify the transcriptome and relate its expression to chromosomal architecture. Genome Biol 6:R52 Li L, Wang X, Stolc V, et al. (2006) Genome-wide transcription analyses in rice using tiling microarrays. Nature Genet 38:124–129 Li L, Wang X, Sasidharan R, et al. (2007) Global identification and characterization of transcriptionally active regions in the rice genome. PLoS ONE 2:e294 Lian X, Wang S, Zhang J, et al. (2006) Expression profiles of 10,422 genes at early stage of low nitrogen stress in rice assayed using a cDNA microarray. Plant Mol Biol 60:617–631 Liang RQ, Li W, Li Y, et al. (2005) An oligonucleotide microarray for microRNA expression analysis based on labeling RNA with quantum dot and nanogold probe. Nucleic Acids Res 33:e17 Luscombe NM, Royce TE, Bertone P, et al. (2003) ExpressYourself: a modular platform for processing and visualizing microarray data. Nucleic Acids Res 31:3477–3482 Ma L, Chen C, Liu X, et al. (2005) A microarray analysis of the rice transcriptome and its comparison to Arabidopsis. Genome Res 15:1274–1283 Matsumura H, Nirasawa S, Terauchi R (1999) Transcript profiling in rice (Oryza sativa L.) seedlings using serial analysis of gene expression (SAGE). Plant J 20:719–726 Matsumura H, Reich S, Ito A, et al. (2003) Gene expression analysis of plant hostpathogen interactions by SuperSAGE. Proc Natl Acad Sci USA 100:15718–15723 Matsumura H, Bin Nasir KH, Yoshida K, et al. (2006) SuperSAGE array: the direct use of 26-base-pair transcript tags in oligonucleotide arrays. Nat Methods 3:469–474 Mockler TC, Ecker JR (2005) Applications of DNA tiling arrays for whole-genome analysis. Genomics 85:1–15 Saha S, Sparks AB, Rago C, et al. (2002) Using the transcriptome to annotate the genome. Nat Biotechnol 20:508–512 Schena M, Shalon D, Davis RW, Brown PO (1995) Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 270:467–470 Selinger DW, Cheung KJ, Mei R, et al. (2000) RNA expression analysis using a 30 base pair resolution Escherichia coli genome array. Nat Biotechnol 18:1262–1268 Singh-Gasson S, Green RD, Yue YJ, et al. (1999) Maskless fabrication of light-directed oligonucleotide microarrays using a digital micromirror array. Nat Biotechnol 17:974–978 Stolc V, Li L, Wang X, et al. (2005) A pilot study of transcription unit analysis in rice using oligonucleotide tiling-path microarray. Plant Mol Biol 59:137–149 Su N, He K, Jiao Y, et al. (2007) Distinct reorganization of the genome transcription associates with organogenesis of somatic embryo, shoots, and roots in rice. Plant Mol Biol 63(3):337–349 Thimm O, Blasing O, Gibon Y, et al. (2004) MAPMAN: a user-driven tool to display genomics data sets onto diagrams of metabolic pathways and other biological processes. Plant J 37:914–939 Velculescu VE, Zhang L, Vogelstein B, Kinzler KW (1995) Serial analysis of gene expression. Science 270:484–487 Wang X, He H, Li L, Chen R, Deng XW, Li S (2006) NMPP: a user-customized NimbleGen microarray data processing pipeline. Bioinformatics 22:2955–2957 Wang Z, Liang Y, Li C, et al. (2005) Microarray analysis of gene expression involved in anther development in rice (Oryza sativa L.). Plant Mol Biol 58:721–737 Wu TP, Ruan KC, Liu WY (1996) A fluorescence-labeling method for sequencing small RNA on polyacrylamide gel. Nucleic Acids Res 24:3472–3473

I.4 Microarray-based Approaches to Rice Transcriptome Analysis

51

Yamada K, Lim J, Dale JM, et al. (2003) Empirical analysis of transcriptional activity in the Arabidopsis genome. Science 302:842–846 Yazaki J, Kishimoto N, Nakamura K, et al. (2000) Embarking on rice functional genomics via cDNA microarray: use of 3¢ UTR probes for specific gene expression analysis. DNA Res 7:367–370 Yazaki J, Kishimoto N, Nagata Y, et al. (2003) Genomics approach to abscisic acid- and gibberellinresponsive genes in rice. DNA Res 10:249–261 Yazaki J, Shimatani Z, Hashimoto A, et al. (2004) Transcriptional profiling of genes responsive to abscisic acid and gibberellin in rice: phenotyping and comparative analysis between rice and Arabidopsis. Physiol Genomics 17:87–100 Yu J, Wang J, Lin W, et al. (2005) The genomes of Oryza sativa: a history of duplications. PLOS Biol 3:e38 Zhu T, Budworth P, Chen W, et al. (2003) Transcriptional control of nutrient partitioning during rice grain filling. Plant Biotechnol J 1:59–70 Zimmermann P, Hirsch-Hoffmann M, Hennig L, Gruissem W (2004) GENEVESTIGATOR. Arabidopsis microarray database and analysis toolbox. Plant Physiol 136:2621–2632

This page intentionally blank

I.5

High-throughput Transcriptome Analysis in Rice from a Genomic Perspective Shoshi Kikuchi1

1

Introduction

The major goals of transcriptome analysis in rice (Oryza sativa L.) are to identify the transcribed regions in the rice genome and to understand the mechanisms of gene expression at the transcription level. Starting with the vast collection of complementary DNA (cDNA) clones (expressed sequence tags, ESTs) in the early 1990s, more than 580,000 full-length cDNA (FL-cDNA) clones were collected from japonica subspecies and more than 35,000 of their complete sequences were published; in contrast, no large-scale full-length cDNA collection is currently available for indica subspecies. Sequence information for ESTs and FL-cDNA clones has contributed to the global annotation of rice genes. Using cDNA clones and sequence information for the FLcDNA clones, several kinds of microarray systems were established, along with databases storing the gene expression data from the microarray systems. Bioinformatics tools for the analysis of gene expression data and for data mining have also been developed. In this chapter, I will review and discuss the significant progress that has been achieved in rice transcriptome analysis.

2

The Rice Transcriptome: From EST Collection to cDNA-based Microarray Systems

In many organisms, the first approach to transcriptome research usually involves the collection of a large number of ESTs from many cDNA libraries, such as organ- or tissue-specific libraries and libraries that contain data on the responses to stress treatments. These ESTs are useful for discovering new genes, especially when searching for genes specific to certain developmental stages and tissues, and

1

Plant Genome Research Unit, Division of Genome and Biodiversity Research, National Institute of Agrobiological Sciences (NIAS), Kan’non dai 2-1-2, Tsukuba Ibaraki 305-8602, Japan e-mail: [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

53

54

S. Kikuchi

for gene complexity analyses using bodymap technology (Okubo et al. 1992; Kawamoto et al. 1996). Subsequently, EST information was also used to design probes for use in microarrays, and for sequence analysis of coding regions in the genome. However, because ESTs are usually derived from single sequencing, they are often short fragments (300 to 500 bp) and do not contain the whole open reading frame (ORF) of an expressed gene. In mammalian systems such as mice and humans, the technology for construction of full-length cDNA libraries has been well established and the isolation of full-length cDNAs has made a significant contribution to the annotation of gene structures in these organisms. The same technology has been used in the construction of FL-cDNA libraries for japonica rice, with approximately 380,000 FL-cDNA clones isolated (Kikuchi et al. 2003; Satoh et al., 2007). The Japanese Rice Genome Research Program (RGP) contributed extensively to the earliest stage of EST collection. The first major contribution was made by the large-scale, pre-genome-sequencing phase of the RGP (1991–1997), which contributed about 60,000 EST sequences from the ‘Nipponbare’ cultivar. Sequence data on each clone can be obtained via the Ministry of Agriculture, Forestry and Fisheries of Japan (MAFF) Rice cDNA Clone Overview page at http://bank.dna. affrc.go.jp/%7Eqxrice/hiho/ (Sasaki et al. 1994). Clustering analysis revealed that this collection originated from 10,000 independent cDNA groups. Protein coding analysis revealed that 25% of the clones had significant similarities to known proteins (Yamamoto and Sasaki 1997). More than 1.2 million ESTs from rice have been registered in the National Center for Biotechnology Information (NCBI) GenBank. The main purpose of this large-scale EST collection is the construction of a restriction fragment length polymorphism (RFLP) linkage map that will allow the construction of a physical map of the chromosomes and an understanding of the mechanisms of gene expression for various isozymes. Later, 6713 unique EST sequences from this collection were mapped to 4387 yeast artificial chromosome (YAC) clones from rice genomic DNA, generating 6591 mapped sites within the rice genome (Wu et al. 2002). The mapping results showed that chromosomes 1, 2, and 3 have relatively high EST densities (approximately twice those of chromosomes 11 and 12), and contain 41% of the total EST sites in the map. Most of the EST-dense regions are distributed in the distal regions of each chromosome arm. A further 86,136 ESTs were sequenced from nine rice cDNA libraries from the superhybrid cultivar ‘LYP9’ and its parents. This assembly of EST sequences yielded 13,232 contigs and 8976 singletons (Zhou et al. 2003). Updated information on indica ESTs and the mapping information for rice FL-cDNAs in the indica genome sequence can be viewed through the Beijing Genomics Institute’s Rice Information System (BGI–RIS; Zhao et al. 2004). Using the results of large-scale cDNA analysis to generate probes, microarray technology can be used to monitor gene expression profiles and to perform functional analysis of the rice genome. For this purpose, the Rice Microarray Project was started in April 1999 by the National Institute of Agrobiological Resources (NIAR) and the Society for Techno-Innovation of MAFF, in collaboration with 64 research institutes across Japan (Kikuchi 2007). The members of this project and their research themes are shown at the Rice Microarray Opening Site (RMOS;

I.5 High-throughput Transcriptome

55

http://cdna01.dna.affrc.go.jp/RMOS/index.html), which is administered by the National Institute of Agrobiological Sciences (NIAS). In this project, using semiunique ESTs produced by RGP as probes, 1265 (Yazaki et al. 2000) and 8987 (Yazaki et al. 2003) cDNA-based microarray systems have been established, and more than 1300 hybridization records have been deposited in the database (Yazaki et al. 2002). Gene expression data generated from these two sets of cDNA-based microarrays are shown at the Rice Expression Database (RED) website (http://red. dna.affrc.go.jp/RED/). The RMOS explains the experimental procedures used for microarray analysis and gives information on the available probes. The 1265- and 8987-cDNA arrays are pioneer microarray systems for rice, but the limited number of probes (corresponding to only one-quarter of the number of genes estimated to exist in rice) means that too few genes have been interrogated. A 22K oligoarray system described in Section 4 of this chapter has helped to overcome several problems, including the reproducibility of the microarray quality caused by the printing process (i.e., the construction of microarrays), cross-hybridization caused by unknown nucleotide sequences, and insufficient capacity to accept requests from users. Currently, twelve reports have been published for studies that used a 9K cDNA microarray system for various physiologies (Demura et al. 2002; Negishi et al. 2002; Akimoto-Tomiyama et al. 2003; Otsuki et al. 2003; Yazaki et al. 2003; Sugimoto et al. 2004; Yamaguchi et al. 2004; Yang et al. 2004; Kitanaga et al. 2006; Kondou et al. 2006; Suzuki et al. 2006; Wasaki et al. 2006)

3

Rice Full-length cDNA Clone Collection Project

The International Rice Genome Sequencing Project (IRGSP) was launched in 1997 following efforts to establish a catalog of rice genes (Sasaki et al. 1994), a highdensity linkage map (Harushima et al. 1998), and a YAC-based physical map (Sasaki et al. 1996). At that time, the rice EST collection was estimated to cover about onequarter to one-third of the genes in the rice genome. For complete information on transcripts, an enormous collection of FL-cDNA clones was required. These FLcDNA clones are necessary to identify exon–intron boundaries and gene-coding regions within genomic sequences and for comprehensive gene function analyses at the transcriptional and translational levels. At the beginning of 2000, with the joint collaboration of the Foundation for the Advancement of International Science (FAIS), the Institute of Physical and Chemical Research (RIKEN, and the NIAS under the supervision of the Bio-oriented Technology Research Advancement Institution, the Rice FL-cDNA project was launched. This project was the first joint collaboration focusing on rice biology using the technology for FL-cDNA collection from human and mouse genomes. From more than 60 different tissues, with and without several stress treatments, two methods were used to construct the FL-cDNA libraries: the oligo-capping method (Maruyama and Sugano 1994) and the biotinylated cap-trapper method (Carninci et al. 2000). The completed project collected more than 580,000 clones and randomly sequenced them from their 5′ and 3′ ends (Fig. 1).

56

S. Kikuchi

Starting materials for RNA preparation Seed, Leaf of seedlings, Root of seedlings, Mature leaf, Mature root, Panicle, Immature embryo, Callus Seedlings and calluses under several kinds of stress treatments polyA RNA preparation

Construction of FL-cDNA libraries by two different methods

Randomly picked up clones and single-pass sequence analysis FL-EST Select representative clones by clustering FL-cDNA 580K Full-length cDNAclones were picked up and 780K FL-EST, 35K FL-cDNAsequences were analyzed Fig. 1 Schematic illustration of the rice full-length cDNA (FL-cDNA) collection and sequencing project. As described in the text, the rice FL-cDNA project sampled more than 60 different tissues of japonica Nipponbare rice and constructed FL-cDNA libraries using two different methods (the oligo-capping method and biotinylated cap-trapper method). In total, about 580,000 clones were randomly collected and single-pass sequences for FL-ESTs were obtained from their 5′ and 3′ ends. Currently, 350,000 representative clones have been completely sequenced and the resulting data are available on the KOME website (http://cdna01.dna.affrc.go.jp/cDNA/)

By September 2003, 32,127 of 170,000 FL-cDNA clones were completely sequenced (Kikuchi et al. 2003), and 780,000 full-length ESTs (FL-ESTs) from the 580,000 FLcDNA clones had been sequenced (Satoh et al., 2007). All related FL-EST sequences were published in the DNA Data Bank of Japan in February 2006 (DDBJ accession CI000001–CI778739). Mapping of the FL-cDNAs to five rice genome assemblies (The Institute for Genomic Research, TIGR release 3 and 4, Yuan et al. 2005; IRGSP build 3 and 4, International Rice Genome Sequencing Project 2005; Beijing Genomics Institute’s 93-11 genome, Yu et al. 2005) revealed about 20,600 transcription units (TUs), about 6000 alternative splicing events, and many types of antisense transcripts (Osato et al. 2003). In contrast, mapping of the 780,000 FL-ESTs generated about 28,500 TUs in the pseudomolecules that originated from the japonica genome sequence and 27,900 TUs in the pseudomolecules that originated from the indica genome sequence. A new sequencing project for the FL-cDNA clones that have been collected but not completely sequenced was launched in 2005 as a 3-year project. In 2005, about 3000 clones were completely sequenced and the sequence and related information, such as the results of a homology search and protein coding, were

I.5 High-throughput Transcriptome

57

presented on the Knowledge-based Oryza Molecular Biological Encyclopedia (KOME) website by the end of December 2006. By the end of the 2006 fiscal year, about 5000 clones were newly completely sequenced. Using 35,188 sequences for the completely sequenced FL-cDNA clones and 778,739 single-pass terminal sequences from 578,473 FL-cDNA clones as query sequences, these FL-cDNAs have been mapped to the five published rice genome sequences (TIGR pseudomolecules release 4, IRGSP pseudomolecules builds 3 and 4, BGI’s 93-11 pseudomolecules, and Syngenta’s ‘Nipponbare’ pseudomolecules; Goff et al. 2002). The number of identified TUs on each chromosome for each pseudomolecule and differences in the results compared with those of TIGR release 4 are shown in Table 1. The table also summarizes the numbers of mapped and unmapped clones: 2186 FL-cDNA clones were not mapped in any of the five assemblies. In addition, 27,800 to 28,600 full-length transcription units (FL-TUs) were identified by mapping of FL-cDNAs and FL-ESTs. The number of FL-TUs generated from the four assemblies originating from ‘Nipponbare’ (TIGR4, IRGSP3, IRGSP4, and Syngenta) totaled more than 28,000, but the number of FL-TUs in 93-11 (indica) was less than 28,000. The mapping results for TIGR release 4 revealed 32,775 FL-cDNA sequences and 527,297 FL-EST sequences (225,229 sequences from 5′ ESTs and 496,998 sequences from 3′ ESTs) that were mapped non-redundantly to the assembled sequence. FL-cDNA and FL-EST generated a total of 28,564 FL-TUs in the TIGR release 4 assembly. Alignment with the predicted genes in the TIGR-predicted coding sequence (TIGR-CDS) in TIGR release 4 (number of loci: 55,890), revealed 23,193 that overlapped with the FL-TUs (AE: annotated and expressed) and 32,697 that were not mapped (ANE: annotated not expressed) (Table 2). In addition, 23,117 TU identified by FL-cDNA mapping overlapped with TIGR-CDS (AE), whereas 5447 were generated between TIGR-CDS, and were designated as non-annotated expressed (NAE). According to the information in TIGR’s rice genome annotation database (Osa1 DB; Yuan et al. 2005), 20,923 of the expressed genes annotated in the TIGR Osa1 database were covered by FL-TUs, but 3101 were not. Similarly, 13,237 TIGR-CDS were labeled as transposable element (TE)-related in Osa1 DB. Only 662 of these CDS were covered by FL-TUs, but 12,575 were designated as ANE genes. Of the 775 TIGR-CDS labeled as expressed TE-related, 428 were covered by FL-TUs but 347 were not. From these results, 85% of expressed genes in Osa1 DB were covered by the FL-cDNA/FL-EST collection, and almost all TE-related or non-annotated genes (95% of TE-related CDS and 94% of non-annotated CDS) were not covered by the FL-cDNA/FL-EST collection. Structural analyses of the NAE genes revealed that many have a short locus length, a small number of exons and introns, and a relatively long exon length. Many of the NAE genes are non-homologous to Arabidopsis genes: 22,943 FL-TUs determined by the FL-cDNA mapping results could be divided into highly homologous (HH), low homology (LH), and nonhomologous (NH) classes. HH genes are highly homologous with Arabidopsis genes (E value < e−50). LH genes have low homology with Arabidopsis genes (e−10 < E < e−50). NH genes have no homology with Arabidopsis genes. In total, 11,897 of 11,973 HH genes are AE types, versus 76 NAE types; 4759

FL-cDNA 5′ end FLEST 3′ end FLEST

a

28,564

Total 32730 32646 80 4 45 15 2397

28,541

32,745 212,539 484,358 4,021 3,198 3,567 2,530 2,305 2,293 2,185 1,934 1,605 1,528 1,683 1,692

IRGSP4

Sequences assembled contigs that were not assembled in 12 chromosomes

Both mapped Same Chr – same strand Same Chr – reverse strand Differential Chr. Mapped on only TIGR Unmapped on only TIGR Both unmapped

32,775 212,598 483,657 4,026 3,196 3,569 2,531 2,313 2,292 2,183 1,933 1,605 1,538 1,685 1,693

35,187 241,854 536,885 Chr1 Chr2 Chr3 Chr4 Chr5 Chr6 Chr7 Chr8 Chr9 Chr10 Chr11 Chr12 Chr0a

TIGR

Map-base cloning

Sequencing

All

japonica genome

Origin

Table 1 The result of FL-cDNA clone mapping to five rice genome assemblies

FL-cDNA locus

Comparison of FLcDNA mapping with TIGR4

Unmapped on all assemblies

32623 32611 10 2 152 17 2395 Mapped on all assemblies

28,576

32,640 211,564 482,909 4,039 3,215 3,566 2,534 2,310 2,290 2,193 1,939 1,574 1,536 1,675 1,705

IRGSP3

31741 30422 317 1002 1034 187 2225

28,477

31,928 208,606 482,665 4,050 3,186 3,597 2,477 2,338 2,262 2,165 1,912 1,545 1,502 1,486 1,523 434

Syngenta

93-11

2186

30162 28760 335 1067 2613 192 2220 29925

27,832

30,354 199,001 465,775 3,940 3,153 3,607 2,493 2,329 2,266 2,021 1,827 1,515 1,416 1,333 1,435 497

Whole shotgun

indica genome

58 S. Kikuchi

28564

FL-cDNA 5endFLEST 3endFLEST Mapped FL-cloned

5447

2967 11255 17841 21850

29808 201343 465816 511817

TE-related Exp Non-Exp 55890

Non-TE-related Exp Ann

32697

TE-related Exp Non-Exp

Non-TE-related Exp Ann

23193

TIGR4 CDSa

Exp-Ann Exp_non_Ann Non-Exp_Ann Non-Exp_non_Ann 12575

20122

Exp-Ann Exp_non_Ann Non-Exp_Ann Non-Exp_non_Ann 662

22531

347 12228

3101 5282

428 234

20923 16123

1733 1368 3549 13472

15380 5543 743 865

Each TIGR4 locus was classified into several categories: (non-) TE-related, for which CDS in the TIGR locus was (was not) a gene related to a transposable element; Exp, for which expression of the CDS was confirmed using rice transcriptome resources; (non-)Ann, for which the CDS was not a gene related to a transposable element and was (was not) a “hypothetical protein”

a

Total

ANE

NAE

23117

AE

FL-cDNA 5endFLEST 3endFLEST Mapped FL-cloned

FL-locus

TIGR4 annotation

Class

TIGR4 annotation

Table 2 Results of the comparison between FL-cDNA loci and TIGR4 loci I.5 High-throughput Transcriptome 59

Fig. 2 Front page: an example of the report page for a full-length cDNA (FL-cDNA) clone on the KOME website

60 S. Kikuchi

I.5 High-throughput Transcriptome

61

of 4903 LH genes are AE types, and 3650 of 6067 NH genes are AE types. NAE genes in the LH and NH categories thus totaled 144 and 2417 genes, respectively. These results suggest that about 59% and 82% of AE genes have high and low homology, respectively, to Arabidopsis genes, and that 18% of AE genes are rice-specific genes. In contrast, 91% of NAE genes are NH genes, and are thus likely to be ricespecific genes. Compared with the AE genes, the median length of TUs of NAE genes (by alignment with FL-cDNA sequences) is shorter. Median values of TU length for AE and NAE genes are 3352 and 1737 bp, respectively. Many NAE genes are intronless, single-exon genes. The average exon lengths in AE and NAE genes are 280 and 450 bp, respectively. This may have made it difficult to predict NAE genes in the DNA sequence. Gene expression analysis based on the analysis of copy number for each TU revealed that NAE genes have fewer transcripts than AE genes. Custom microarray analysis also supports this hypothesis (Satoh et al., 2007). These results suggest that NAE genes are structurally specific genes that are nonhomologous to Arabidopsis genes and have low expression. In addition to the mapping results for the FL-cDNA/FL-EST analysis, the KOME website provides BLASTN and BLASTX homology searches for sequences registered in GenBank, computer analyses of cellular locations, trans-membrane analyses, and gene ontology classifications of the putative proteins encoded by FLcDNAs (http://cdna01.dna.affrc.go.jp/cDNA/) (Fig. 2). Information from the 32,127 FL-cDNA clones was also used in Rice Annotation Program 1 (RAP1). In this activity, FL-cDNAs and other public ESTs were mapped and aligned to the rice genome sequence from IRGSP built 3 and 4, then annotations were added by hand. The first Annotation Jamboree meeting was held in December 2004 in Tsukuba. Details of the annotated genes are shown in the RAP database, RAP-DB (http:// rapdb.lab.nig.ac.jp/; Ohyanagi et al. 2006; Itoh et al. 2007).

4

Oligoarray Systems

The collection and complete sequencing of 32,127 rice full-length cDNA clones allowed NIAS researchers to increase the 8987-cDNA microarray to produce a new global rice array based on oligomicroarray technology. This was carried out in collaboration with Agilent Technologies, a private company with strong capabilities to synthesize 60-mer or 70-mer oligonucleotides as probes for microarray systems. Because about 22,000 probes can be printed on a single glass plate, only one probe per TU that mapped to the rice genome sequence was selected. Agilent Technologies designed 60-mer probe sequences from 29,100 FL-cDNA sequences by considering the melting temperature (Tm) and guanine–cytosine (GC) contents and removing the possibility of cross-hybridization. After several validation experiments using custom-prepared arrays and RNAs obtained from seeds, callus, seedlings, and other tissues, a final set of probe sequences was fixed. In November 2003, the 22K rice oligomicroarray version 1 (G4138A) was commercialized by Agilent Technologies, and is now being used by rice molecular biologists worldwide. In 2006, the 22K

62

S. Kikuchi

array was updated in collaboration with the Rice Genome Resource Center (RGRC) using the information in RAP-DB, to produce a 44K array. The new array will cover almost all expressed genes in rice, and should become the new standard oligoarray system for rice functional genomics. Many journals request registration of the data produced by microarray experiments in public databases, such as the NCBI Gene Expression Omnibus (GEO; http://www.ncbi.nlm.nih.gov/projects/geo/) and the Array Express repository (http://www.ebi.ac.uk/arrayexpress/). NIAS rice oligoarray version 1 was registered in GEO under the accession number GPL892. The first published gene expression analysis result using the oligoarray produced gene-expression profiles for rice genes that respond to abscisic acid and gibberellins (Yazaki et al. 2004). These data sets are registered in NCBI-GEO as gene series 661 (GSE661), samples 9853–9860 (GSM9853–9860), and platform GPL477 (22K custom oligoarray). Information from known and predicted gene models was used to construct the global rice gene expression microarray system. The Affymetrix GeneChip Array is one of the standard microarray systems based on the 25-mer probe system. According to the description in NCBI-GEO’s registration, this array contains probes to query 51,279 transcripts representing two rice cultivars, with approximately 48,564 japonica transcripts and 1260 transcripts representing the indica cultivar. This unique design was created within the Affymetrix GeneChip Consortia Program, and provides scientists with a single array that can be used for the study of rice. High-quality sequence data were derived from GenBank mRNAs, TIGR release 4 gene predictions, and the International Rice Genome sequencing project. The arrays were designed using NCBI UniGene Build #52 (May 7, 2004) incorporating predicted genes from GenBank and the TIGR Osa1 v2 data set (ftp://tigr.org; FASTA software, 89.3 Mb). A 70-mer microarray covering 41,754 annotated genes and a nontransposable-element rice gene model with and without experimental support were constructed (Ma et al. 2005), and the expression of genes in representative rice organs (seedling shoots, tillering-stage shoots and roots, heading- and filling-stage panicles, and suspension culture cells) was analyzed. Expression of 86% of the 41,754 genes was detected. A similar proportion of the genomes was expressed in the corresponding organs of rice and Arabidopsis. A large percentage of the rice gene models that lack significant homology with Arabidopsis genes were found to be expressed. The expression patterns of rice and Arabidopsis that best matched homologous genes in distinct functional groups revealed dramatic differences in their degree of conservation of expression between the species. These data show some basic similarities and differences between the Arabidopsis and rice transcriptomes. Since the commercialization of the 22K rice oligomicroarray system, few reports of its use have been published. The reason for this might be the large amount of gene expression data that exists, which makes data analysis difficult. In rice, many types of genomic information are available, such as map locations of probed genes, protein coding information, and promoter sequence information. To obtain such information, researchers must use data mining techniques, and to support this process it is important to have the ability to overlay these and other layers of genomic information, including the ability to relate these layers to classical plant

I.5 High-throughput Transcriptome

63

biochemical information. The development of these relationships will assist in the interpretation of gene functions. Comparisons of gene expression under various biotic and abiotic stresses are also important. To meet these needs, a RED II database is being established to cover the 22K microarray data and provide appropriate data mining tools. Furthermore, for the analysis of cis-elements in the promoter region of the genes that have similar expression profile, we have established a pipeline system for the cis-element search (http://hpc.irri.cgiar.org/tool/nias/ces).

5

Gene Family Analysis

After global collection of FL-cDNA clones, many genes are structurally annotated by RAP (Itoh et al. 2007). Biological and biochemical functional annotation must then follow. Currently, about 3000 rice genes have been assigned to metabolic pathways by the activity of Gramene’s Rice Cyc initiative (http://www.gramene.org/pathway/). Comparative gene family analyses have been carried out in many plant species for a long time, but it is important to also perform comparative analyses of gene families among many different kinds of organisms, from bacteria to animals and plants, and to understand how gene families have diverged among organisms. Initial efforts have focused on membrane transport genes and the protein related to the calcium-related signal transduction system (Nagata et al. 2004). Proteins encoded by the membrane transport genes have common characteristics such as membrane-spanning hydrophobic amino acid residues, and have been divided into three categories: ion channels, pumps, and secondary transporters. In total, 1300 genes have been annotated as membrane transport proteins. Ion channels are the gates in cell membranes that open in response to signals such as mechanical or electronic stimulation and ligand binding. Compared with other systems, the channel system is the fastest structure for transporting molecules without consuming energy. Therefore, animals have adapted these systems for signal transmission in nerves and muscles. The genes classified into this category have diverged considerably in animals, whereas the number of orthologs in plants is very small. About 200 rice genes have been annotated in this category, and they fall into the following main families and superfamilies: the major intrinsic protein (MIP) family, the voltage-gated ion channel (VIC) superfamily, the annexin family, the chloride channel family, and several small channel-protein families. ATP-dependent transport systems are classified as pump systems. The major roles of pump systems are to transport molecules in specific directions, independent of environmental conditions, and to transport ions to create a concentration gradient between the areas outside and inside the membrane (i.e., active transport). About 260 rice genes have been annotated with this category, into the following families: ATP-binding cassette (ABC) proteins, proton-pump F0F1 ATPases (F-ATPases), ATPases that form a phosphorylated intermediate form (P-ATPases), and proton pyrophosphatase (H+-PPase).

64

S. Kikuchi

Secondary transporter systems work via concentration gradients in co-transport molecules. In this system, it is efficient to use a few abundant molecules as common co-transporters for various transport substrates. Therefore, in accordance with the species of ion for which there is a gradient between the inside and outside of the cell, major co-transport molecules were selected and secondary transport systems were developed for them. There are more than 100 gene families in the secondary transport systems of all organisms, and various materials are transported by these families. In rice, more than 830 genes have been annotated as secondary transporters, including the following families and superfamilies: the amino acid– polyamine–organocation (APC) superfamily, the monovalent cation–proton antiporter 1 and 2 (CPA1 and CPA2) family, the major facilitator superfamily (MFS), and the amino acid–auxin permease (AAAP) family. In summary, about 1300 rice genes have been classified into the above-mentioned membrane transport gene families. The clone-by-clone information will be published on the KOME website. The same kind of analysis will also be done for genes encoding proteins related to the calcium signal-transduction system (Nagata et al. 2004). Thus far, 110 rice genes have been annotated as calciumbinding proteins, and they have been classified into 18 gene families, including genes for EF-hand (a calcium-binding motif, with the Ca ion located between helix E and helix F) proteins, Ca2+/phospholipid binding protein, gamma-carboxyglutamic acid, lipid-containing proteins, and Ca2+ storage proteins. Similar analyses focusing on transcription factors, protein kinases, and protein phosphorylases will follow.

6

Future Perspectives

Currently, more than 900,000 rice ESTs and FL-cDNAs have been sequenced and deposited in the NCBI databases. The public availability of these sequences has not only advanced the functional analysis of rice genes but also played an important role in rice genome annotation. The mapping and alignment of combined EST sequences and FL-cDNAs to the genome sequence have provided direct experimental evidence for many of the gene models predicted by computer programs. However, a considerable number of the gene models have not been confirmed by any experimental data. The problem of mis-prediction or mis-annotation of exon–intron structures by current computer programs for gene structural annotation is still a major challenge for rice genome biologists. Further collection and complete sequencing of FL-cDNA clones and the detailed comparison of gene models and FL-cDNA sequences will improve current rice genome annotation, as was shown in this chapter with the results of NAE genes. Recently, in the process of updating the 22K oligomicroarray system to a 44K array, probes were constructed based on the sequences of the predicted genes. These predicted-gene probes were subjected to hybridization analyses with RNA from four diverse tissue samples: seeds, the upper part of seedlings, roots of seedlings,

I.5 High-throughput Transcriptome

65

and callus. Based on the signals detected by each probe, the estimated number of expressed genes in rice appears to be around 42,000 (Satoh et al., 2007). During the process of updating the array, it was also found that many of the TE-related genes were ANE genes. An in-depth comparison of the structures of the TE-related genes and the truly expressed genes is also very important. This knowledge could be used to improve the implementation of gene-prediction programs. Micro-RNAs (miRNAs) and short interfering RNAs (siRNAs) are relatively new and important areas of research in transcriptomics. The rice FL-cDNA collection includes many of these small RNAs. However, to obtain comprehensive coverage of these small RNAs, a new collection might be required. Several types of microarray systems have been established for rice geneexpression analysis in recent years. However, because of the high expense associated with microarray systems, the technique has not become routine in ordinary molecular biology laboratories in the same way that northern-blot hybridization and the reverse-transcriptase polymerase chain reaction (RT-PCR) have. Therefore, microarray service centers should be established to perform the hybridizations for individual laboratories. The appropriate use of statistical and microarray analysis procedures and software packages for mining the large data sets is another obstacle for many molecular biologists. The development of simple, user-friendly, yet rigorously structured microarray analysis programs will promote more extensive use of the microarray system. As many rice gene expression data sets are accumulated, public databases should be established so that all these data sets regarding the output from different microarray platforms can be easily compared. In these databases, the experimental conditions should be described according to international standards, such as those for the Minimum Information About a Microarray Experiment (MIAME, http://www.mged.org/index.html). These databases should also promote the coordinated analysis of transcriptomic, proteomic, and metabolomic data, which will further advance the prediction and validation of gene functions. Acknowledgements The FL-cDNA project of Dr. Kikuchi’s laboratory was funded by a Rice Genome Full Length cDNA Library Construction Project grant from BRAIN (the Bio-oriented Technology Research Advancement Institution). The rice microarray project and the mapping and alignment of cDNA sequences to the rice genome sequence were supported by the Rice Genome Project in Japan.

References Akimoto-Tomiyama C, Sakata K, Yazaki J, et al. (2003) Rice gene expression in response to N-acetylchitooligosaccharide elicitor: comprehensive analysis by DNA microarray with randomly selected ESTs. Plant Mol Biol 52:537–551 Carninci P, Shibata Y, Hayatsu N, et al. (2000) Normalization and subtraction of cap-trapper-selected cDNAs to prepare full-length cDNA libraries for rapid discovery of new genes. Genome Res 10:1617–1630

66

S. Kikuchi

Demura T, Tashiro G, Horiguchi G, et al. (2002) Visualization by comprehensive microarray analysis of gene expression programs during transdifferentiation of mesophyll cells into xylem cells. Proc Natl Acad Sci USA 99:15794–15799 Goff SA, Ricke D, Lan T-H, et al. (2002) A draft sequence of the rice genome (Oryza sativa L. ssp. japonica). Science 296:92–100 Harushima Y, Yano M, Shomura A, et al. (1998) A high-density rice genetic linkage map with 2275 markers using a single F2 population. Genetics 148:479–494 International Rice Genome Sequencing Project (2005) The map-based sequence of the rice genome. Nature 436:793–800 Itoh T, Tanaka T, Barrero RA, et al. (2007) Curated genome annotation of Oryza sativa ssp. japonica and comparative genome analysis with Arabidopsis thaliana. Genome Res 17:175 Kawamoto S, Matsumoto Y, Mizuno K, Okubo K, Matsubara K (1996) Expression profiles of active genes in human and mouse livers. Gene 174:151–158 Kikuchi S (2007) Comprehensive analysis of rice gene expression by using the microarray system: what we have learned from the microarray project. In: Datta S (ed) Rice improvement in the genomics era. Haworth Press, Binghamton, New York (in press) Kikuchi S, Satoh K, Nagata T, et al. (2003) Collection, mapping, and annotation of over 28,000 cDNA clones from japonica rice. Science 301:376–379 Kitanaga Y, Jian C, Hasegawa M, et al. (2006) Sequential regulation of gibberellin, brassinosteroid, and jasmonic acid biosynthesis occurs in rice coleoptiles to control the transcript levels of anti-microbial thionin genes. Biosci Biotechnol Biochem 70:2410–2419 Kondou H, Ooka H, Yamada H, et al. (2006) Microarray analysis of gene expression at initial stages of rice seed development. Breeding Sci 56:235–242 Maruyama K, Sugano S (1994) Oligo-capping: a simple method to replace the cap structure of eukaryotic mRNAs with oligoribonucleotides. Gene 138:171–174 Nagata T, Iizumi S, Satoh K, et al. (2004) Comparative analysis of plant and animal calcium signal transduction element using plant full-length cDNA data. Mol Biol Evol 21:1855–1870 Negishi T, Nakanishi H, Yazaki J, et al. (2002) cDNA microarray analysis of gene expression during Fe-deficiency stress in barley suggests that polar transport of vesicles is implicated in phytosiderophore secretion in Fe-deficient barley roots. Plant J 30:83–94 Ohyanagi H, Tanaka T, Sakai H, et al. (2006) The rice annotation project database (RAP-DB): hub for Oryza sativa ssp. japonica genome information. Nucl Acids Res 34:741–744 Okubo K, Hori N, Matoba R, et al. (1992) Large scale cDNA sequencing for analysis of quantitative and qualitative aspects of gene expression. Nat Genet 3:167–168 Osato N, Yamada H, Satoh K, et al. (2003) Antisense transcripts with rice full-length cDNAs. Genome Biol 5:R5 Otsuki S, Ikeda A, Sunako T, et al. (2003) Novel gene encoding a Ca2+-binding protein and under hexokinase-dependent sugar regulation. Biosci Biotechnol Biochem 67:347–353 Sasaki T, Song J, Koga-Ban Y, et al. (1994) Toward cataloguing all rice genes: large-scale sequencing of randomly chosen rice cDNAs from a callus cDNA library. Plant J 6:615–624 Sasaki T, Yano M, Kurata, N, Yamamoto K (1996) The Japanese Rice Genome Research Program. Genome Res 6:661–666 Satoh K, Doi K, Nagata T, et al. (2007 or 8) Gene organization in rice revealed by full-length cDNA mapping and gene expression analysis through microarray PLoS One (in press) Sugimoto H, Kusumi K, Tozawa Y, et al. (2004) The virescent-2 mutation inhibits translation of plastid transcripts for the plastid genetic system at an early stage of chloroplast differentiation. Plant Cell Physiol 45:985–996 Suzuki M, Takahashi M, Tsukamoto T, et al. (2006) Biosynthesis and secretion of mugineic acid family phytosiderophores in zinc-deficient barley. Plant J 48:85–97 Wasaki J, Shinano T, Onishi K, et al. (2006) Transcriptomic analysis indicates putative metabolic changes caused by manipulation of phosphorus availability in rice leaves. J Exp Bot 57:2049–2059 Wu J, Maehara T, Shimokawa T, et al. (2002) A comprehensive rice transcript map containing 6,591 expressed sequence tag sites. Plant Cell 14:525–535

I.5 High-throughput Transcriptome

67

Yamaguchi T, Nakayama K, Hayashi T, et al. (2004) cDNA microarray analysis of rice anther genes under chilling stress at the microsporogenesis stage revealed two genes with DNA transposon Castaway in the 5′-flanking region. Biosci Biotechnol Biochem 68:1315–1323 Yamamoto K, Sasaki T (1997) Large-scale EST sequencing in rice. Plant Mol Biol 35:135–144 Yang GX, Jan A, Shen SH, et al. (2004) Microarray analysis of brassinosteroid- and gibberellinregulated gene expression in rice seedlings. Mol Genet Genomics 271:468–478 Yazaki J, Kishimoto N, Nakamura K, et al. (2000) Embarking on rice functional genomics via cDNA microarray: use of 3′ UTR probes for specific gene expression analysis. DNA Res 7:367–370 Yazaki J, Kishimoto N, Ishikawa M, Kikuchi S (2002) Rice expression database: the gateway to rice functional genomics. Trends Plant Sci 7:563–564 Yazaki J, Kishimoto N, Nagata Y, et al. (2003) Genomics approach to abscisic acid- and gibberellinresponsive genes in rice. DNA Res 10:249–261 Yazaki J, Shimatani Z, Hashimoto A, et al. (2004) Transcriptional profiling of genes responsive to abscisic acid and gibberellin in rice: phenotyping and comparative analysis between rice and Arabidopsis. Physiol Genomics 17:87–100 Yu J, Wang J, Lin W, et al. (2005) The genomes of Oryza sativa: a history of duplications. PLoS Biol 3:E38 Yuan Q, Ouyang S, Wang A, et al. (2005) The Institute for Genomic Research Osa1 rice genome annotation database. Plant Physiol 138:18–26 Zhao W, Wang J, He X, et al. (2004) BGI–RIS: an integrated information resource and comparative analysis workbench for rice genomics. Nucl Acids Res 32:377–382 Zhou Y, Tang J, Walker MG, et al. (2003) Gene identification and expression analysis of 86,136 expressed sequence tags (EST) from the rice genome. Genomics Proteomics Bioinformatics 1:26–42

This page intentionally blank

I.6

Active Transposons in Rice Tetsuya Nakazaki1, Ken Naito1, Yutaka Okumoto1, and Takatoshi Tanisaka1(* ü)

1

Introduction

The first ‘transposon’ – that of the ‘controlling element’ of maize – was discovered by Mclintock more than half a century ago. This element was detected as a factor controlling the mutable character of kernel pigmentation. In higher plants, many mutable traits have bee found, especially for genes involved in pigmentation and endosperm quality. Recent molecular analyses have revealed that many of these mutable traits are controlled by transposons (Fedoroff et al. 1983, 1984; Bonas et al. 1984; Brown et al. 1989; Inagaki et al. 1994). Transposons or transposable elements were originally defined as mobile genetic elements; that is, as DNA fragments with the ability to move to new chromosomal locations (this is known as transposition). With advances in the knowledge of the sequence structures of transposons, it has become clear that they include a number of DNA elements which had probably lost mobility because of mutations on the sequences or other factors. Transposable elements are divided into two groups according to the mode of propagation: retrotransposons (class I elements) and DNA transposable elements (transposons in a narrow sense, class II). While the former moves through RNA intermediate by the action of reverse transcriptase, the latter moves in a DNA form through a cut-and-paste mechanism. Therefore, only class II elements, transposons, cause mutable traits by precise excision from the silent alleles where they are inserted. In rice (Oryza sativa L.), retrotransposons were detected as active transposable elements (Hirochika et al. 1996), and their application to rice genomics and genetics has advanced considerably (Hirochika 2001). An active transposon was not detected for a long time in the rice genome. However, now we know of two such transposons. A number of analyses of maize mutable phenotypes caused by transposons have led to the grouping of transposons into several families (Peterson 1988), such

1 Graduate School of Agriculture, Kyoto University, Kitashirakawa, Sakyoku, Kyoto 606-8502, Japan e-mail: [email protected]; [email protected]; [email protected]; [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

69

70

T. Nakazaki et al.

as Ac-Ds, Spm/En-dSpm/I and MuDR-Mu1. Subsequently, from the DNA sequence analysis of active transposons, their independent classifications were categorized into larger groups termed superfamilies. At present, in plants the five main superfamilies of transposons that are recognized are hAT, CACTA, Mutator, PIF/ Harbinger and Tc1/mariner (Feschotte et al. 2002). These superfamilies are classified based on three criteria: the sequence of terminal inverted repeats (TIRs), the homology of their putative transposases and the size of target site duplication (TSD) (Xu and Dooner 2005). Therefore, we can detect transposons involved in the superfamilies, both of autonomous and non-autonomous elements, by their sequence nature without evidence of transposition ability.

2 2.1

Detection of Transposon Sequences in Rice Transposons Detected from Sequence Analyses of Some Loci

In rice, until the transposition ability of mPing (miniature Ping) was reported (Jiang et al. 2003; Kikuchi et al. 2003; Nakazaki et al. 2003), no active transposon had been detected. Then, several transposons, sequences with TIRs and TSDs, were reported in the 1990s. For example, the 233 bp sequence with imperfect 75 bp TIRs and 4 bp TSDs was found in the waxy locus of rice, designated Tnr1 (transposable element in rice #1) (Umeda et al. 1991). Later, this non-autonomous element was classified to the Tc1/mariner superfamily (Tenzen et al. 1994; Han et al. 2000). The 1536 bp sequence, Tnr3, was detected in a rice genomic clone as the non-autonomous element of the CACTA superfamiliy (Motohashi et al. 1996). In the Xa21 locus (Xanthomonas campestris pv. Oryzae resistance-21), three Ds-like sequences were found (Song et al. 1998) and considered as non-autonomous elements belonging to the hAT superfamily. Several non-autonomous elements of the Mutator superfamily were also screened out from genomic clones including Tnr2 (Mochizuki et al. 1993) and OsMus (Asakura et al. 2002). Tnr5 in the region of a Tnr1 element was reported as the sequence related to the Tourist family (Bureau and Wessler 1992; Han et al. 2000). This means that Tnr5 belongs to the PIF/Harbinger superfamily. Thus, from the sequence analysis of certain genomic regions, all kinds of non-autonomous elements belonging to the five main superfamilies were detected in the rice genome during the 20th century. Thereafter, autonomous elements or transposase-like sequences were also discovered in the rice genome (Song et al. 1998; Asakura et al. 2002).

2.2

Transposons in the Developed Sequence Database

Advances in sequence information databases have opened up another approach to salvage transposons. Such analyses showed that a great number of small elements with TIRs were inserted in gene regions (Bureau and Wessler 1994a, 1994b;

I.6 Active Transposons in Rice

71

Bureau et al. 1996). These elements were designated as miniature inverted-repeat transposable elements (MITEs), which were initially given to the two groups of small TIR elements, the Tourist and Stowaway families (Wessler et al. 1995). The Touristtype MITEs were detected in the untranslated regions of genes in diverse grass species as homologous elements with a 128 bp maize insertion sequence (Bureau and Wessler 1994b). Subsequently, the Stowaway type was detected in the genes of diverse flowering plants as homologous elements with a 257 bp sorghum sequence (Bureau and Wessler 1994a). The elements related to both groups were also detected in rice (Bureau et al. 1996; Turcotte et al. 2001). After the discovery of plant MITEs, MITE families were also found in several animals. Their distinct features are high copy number (more than several thousand), small size (less than ca. 600 bp) and a lack of related autonomous elements, serving to set them apart from previously described non-autonomous elements (Feschotte et al. 2002). Because MITEs are usually located in gene-rich regions and their transposition events occur in great numbers, they are believed to contribute much to the evolution of species, although little is known about the mechanism of their dramatic and quick amplification in the genome. At present, the Tourist and Stowaway groups are considered to be classified into the PIF/Harbinger and Tc1/mariner superfamilies, respectively, based on the sequence similarity of TIRs and TSDs (Feschotte et al. 2002), and also Mutator-derived new MITE families in rice have been reported (Yang and Hall 2003). Therefore, MITEs could be described as a group of small-sized non-autonomous elements with a large copy number or with potential for quick amplification in future. Decoding the genome sequences of the rice variety Nipponbare by the International Rice Genome Sequencing Project (IRGSP) also clarified the position and the number of transposons in the genome (IRGSP 2005). The IRGSP (2005) showed that transposable elements occupy at least 31.0% of the japonica rice genome, and about 164,000 copies of transposons are present. This number is more than 2.5 times that of class I elements (62,000 copies). However, the nucleotide contribution of transposons is less than that of class I elements. This is because of the small sizes of transposons, especially MITEs. The PIF/Harbinger superfamily, including the Tourist MITEs, shows the largest copy number among diverse transposons. This may imply the importance of this superfamily in constructing the rice genome, but further analysis is needed in this regard.

2.3

Use of Transposons as Markers

The transposition events of transposable elements cause a sequence variation in the genome. Sequence comparison at the Wx (Waxy) locus between O. sativa and O. glaberrima disclosed many substitutions in the sequence of the Tnr1 element inserted in the locus (Mochizuki et al. 1992). The Tnr1 element in the Waxy locus was referred to as Stowaway-Os3 (Bureau and Wessler 1994a). In addition, sequence comparison between Nipponbare and the indica variety IR36 showed that the Cht4 (Chitinase-4)

72

T. Nakazaki et al.

gene of the latter harbored an insertion of Wanderer in its 3¢ flanking region (Nakazaki and Ikehashi 1998). Thus polymorphisms of sequence or insertion sites of MITE are far higher in frequency than those in the unique regions, as discussed by Tenzen et al. (1994), suggesting that such polymorphisms for MITEs are used as DNA markers. Markers using insertion/deletion polymorphisms of MITEs are easily detected by the conventional PCR method. On the other hand, the transposon display (TD) method, applying the AFLP technique and detecting transposons in the genome, has been developed. In rice, by TD method, detection of insertion site polymorphisms of CACTA element, Rim 2/Hipa, has been reported (Kwon et al. 2005). Komori and Nitta (2003) reported the very high frequency of SNPs within the MITE regions. These SNPs also are useful for making new markers for closely related varieties.

3 3.1

Active Transposons of Rice mPing

mPing is the first active transposon isolated in rice, which was simultaneously reported by three independent groups (Jiang et al. 2003; Kikuchi et al. 2003; Nakazaki et al. 2003). From the structural nature of mPing, 430 bp long with 15 bp TIRs and 3 bp TSDs (TAA), it was considered a novel class of the Tourist-like MITE. The discovery of mPing represents the first isolation of active MITE from any organism, as well as the first active transposon from rice. In the two studies, mPing was initially picked up from the rice genome database as a candidate for an active MITE, followed by anther culture or tissue culture to investigate its activity. While Kikuchi et al. (2003) showed the activity of mPing in anther culture derived from the japonica cultivar Nipponbare, Jiang et al. (2003) showed the activity in tissue culture of the indica cultivar C5924. In the third case, mPing was detected in the course of research for a mutable character (gene) (Nakazaki et al. 2003). This study started from analysis of the mutable phenotype of ‘slender glume’ of a mutant strain IM294, which was induced by gamma-ray irradiation to seeds of the japonica rice variety Gimbozu (Fig. 1). This mutable phenotype was considered a single recessive mutation of the slg (slender glume) locus (Teraishi et al. 1999). Following map-based cloning, the Rurm1 (Rice ubiquitin modifier-1) locus with an insertion sequence (433 bp, mPing plus TSD), which was not observed in the original variety Gimbozu, was found to be a candidate of the slg allele. From studies using revertant and chimeric plants, the excision event of mPing was proved to cause reversion for the muntable phenotype. As a result, this study was the first finding of a rice transposon that is active in intact plants. Moreover, this study also provided the first direct evidence that the MITEs are capable of both insertion and excision. As mPing is a non-autonomous element lacking coding capacity, an autonomous element is considered to be necessary for its transposition. In the rice genome, two candidates for the autonomous element, called Ping and Pong, have been detected (Fig. 2). Ping is 5341 bp and shares 253 bp from the 5′ end and 177 bp from the 3′

I.6 Active Transposons in Rice

73

Fig. 1 Chimeric plants for glume shape that appeared in a slender glume mutant line IM294. a An example of a between-panicle chimera (left slender glume panicle, right normal glume panicle). b An example of a within-chimera (S:slender glume branch; N normal glume branch) (Nakazaki et al. 2003)

Pong Homology

Ping

mPing

ORF1 70%

80%

ORF2 87%

ORF1

70%

ORF2

TIR

Fig. 2 tructures of Pong, Ping and mPing. All the three elements share 15 bp terminal inverted repeats (TIRs). mPing has arisen from Ping by direct deletion event of internal sequences. Pong has relatively low homology to Ping besides TIRs

end, respectively, of its terminal sequences with mPing, indicating that mPing is a recent deletion derivative of Ping. Pong is the 5166 bp sequence with identical TIRs except one base with mPing. Both share similarity in two blocks of internal sequence corresponding to the two open reading frames (ORFs) of each element. Of these two ORFs, the second ORF encodes a putative DDE motif containing three acidic amino acids found at the catalytic core of the transposases of some organisms (Jiang et al. 2003). Hu et al. (2006) surveyed the distribution of Ping and Pong in 102 Asian cultivated rice varieties. While they detected Pong from all the

74

T. Nakazaki et al.

investigated varieties, Ping was detected only from the japonica varieties. We have no answer as to which element is an autonomous element of mPing. Although mPing is clearly a deletion derivative of Ping, it has been mobilized in the C5924 cells that lack Ping (Jiang et al. 2003). Recently, in the course of analyzing the copy number of mPing elements with the transposon display (TD) method (Casa et al. 2000, 2004; Jiang et al. 2003), a dramatic, burst amplification was discovered in Gimbozu (Fig. 3; Naito et al. 2006). The number of mPing in the Nipponbare genome was 51, and the same number of mPing was detected from the sequence database, while that in Gimbozu was 1163 (Naito et al. 2006). This is the first report showing the presence of more than 1000 mPing copies in a genome, and proves that mPing is a true MITE. PCR

Fig. 3 Copy-number estimation of mPing elements in Nipponbare (N) and Gimbozu (G) in the autoradiograph of a transposon display gel of mPing amplicons (Naito et al. 2006)

I.6 Active Transposons in Rice

75

analysis of 50 randomly chosen from the 1163 sites revealed that five loci were heterozygous for insertion. Such a situation could arise when an insertion allele has not yet been fixed in the population, and/or when mPing is still actively transposing. To examine whether mPing is active in Gimbozu, DNA was extracted from Gimbozu plants propagated by single seed descent, and was subjected to TD analysis (Fig. 4; Naito et al. 2006). As a result, new mPing insertions were recognized as bands, each of which appeared in an individual but not in its sib-lines or parent. Such bands are herein referred to as de novo insertions, and indicate that Gimbozu harbors the ability of activation of mPing in intact plants. The study also showed that sites of de novo mPing insertions were not preferential to any sites. The finding of active mPing in the anther culture line of Nipponbare (Kikuchi et al. 2003), where the mPing element is not capable of transposition in intact plants

Fig. 4 Transposon display analysis of mPing insertions of generational materials from Gimbozu. Arrows indicate de novo insertions detected in the progenies from a single plant of Gimbozu (P). BfaI + GCT, BfaI + TTC and BfaI + GAC each indicate the combination of restriction enzyme and selective nucleotides (cf. Naito et al. 2006)

76

T. Nakazaki et al.

(Naito et al. 2006), indicates that the Nipponbare genome possesses enough instruments for transposition of mPing. Recently, activation of mPing was observed by hydrostatic pressurization on seeds (Lin et al. 2006) and by hybridization with wild rice (Zizania latifolia Griseb.) (Shan et al. 2005). On the other hand, mPing in the genome of IM294 and Gimbozu is active in intact plants. Therefore, these two varieties/lines should posses some factors for activation of mPing other than the instruments for transposition such as transposase. This may be essential for understanding the burst amplification of MITE.

3.2

nDart1

Recently, the second active transposon nDart1 (non-autonomous DNA-based active rice transposon 1) was found in a variegated yellow leaf segregant, which exceptionally occurred in the F2 population from the cross between the indica native variety C-5052 and the japonica marker line H-126 (Maekawa et al. 1999; Tsugane et al. 2006). nDart1 was related to the mutability of the pyl-v (pale-yellow-leaf variegated) allele at the pyl locus. The wild-type allele of this locus was identical to the OsClP5 gene, which shared homology with Arabidopsis gene ClP5 controlling the P5 subunit of ATP-dependent caseinolytic protease, AlClP5. nDart1 is a 606 bp element with 19 bp perfect TIRs with 8 bp TSDs, and belongs to the hAT superfamily. From the segregation analysis for pyl-v, H-126 was found to carry a single factor related to the pyl-v phenotype, which might be an autonomous Dart (aDart) element of nDart1. Tsugane et al. (2006) also reported that there were many nDart1 and related elements in the Nipponbare genome: 18 nDart1 elements, 53 inactive Dart (iDart) elements that might be structurally similar to aDart but functionally deficient, and three defective Dart (dDart) elements that were derived from iDart by internal deletions. There were no identical sequences among the 53 iDart elements, and 29 of them were found to putatively encode transposase. From the study of the occurrence of excision events of nDart1 elements, it was suggested that Nipponbare harbors no active aDart element and that nDart1 elements differ in response to aDart. However, 5-azacytidine (demethylating agent) treatment to Nipponbare seeds caused excition of nDart1 elements. This indicates that Nipponbare also carries a dormant aDart element or iDart element(s), both of which are probably activated by the reduction of DNA methylation level. This finding is marvelous because it suggests the possibility to artificially control the transposition of nDart1 in the Nipponbare genome. It was reported that 5-azacytidine treatment to rice seedlings of Gimbozu induced heritable dwarfism (Sano et al. 1990). The cause of this mutational event is still unclear, but there is a possibility that this mutation is related to the nDart1 transposition event. Tsugane et al. (2006) also stated that another mutable gene, thl-m (thumbelina-m), which occurred in the pyl-v-line, was successfully identified by the transposon-tagging method. These findings lead us to expect construction of a new transposon-tagging system using the nDart1 family.

I.6 Active Transposons in Rice

4

77

Conclusion

As described above, the burst amplification of mPing from approximately 50 copies to more than 1000 in Gimbozu was established only in the last 100 years. This fact coincides with the prior evolutionary analyses of plant genomic sequences describing that MITE families (with several thousands of copies) arise from the burst of only a few elements over a very short period (Bureau and Wessler 1994; Feschotte et al. 2002; Jiang et al. 2003). On the other hand, in Gimbozu, although Ping is highly active, it has only seven copies. This indicates that the excision/insertion events of short transposons (MITEs) are far more likely to occur than those of long ones. Thus short transposons have a great advantage as a tool for transposon tagging. nDart is also expected to hold great potential for generating mutations. In addition, these ‘recently-activated elements’ have a great advantage as markers. Their insertion/deletion polymorphisms can be readily detected even in closely related strains. Even though Gimbozu and Nipponbare are japonica varieties, mPing insertion/deletion polymorphisms between them number more than 1000. If we integrate these markers with annotated rice-genome sequences, we would have a powerful weapon to undermine the subtle differences of proximate strains or varieties. A large fraction of genomes are comprised of transposons and are considered as a great source of diversification of genome sizes, structures and functions (Feshotte et al. 2002; Kazazian 2004; Naito et al. 2006). Further analyses of transposons will open a way to solving genomic evolution.

References Asakura N, Nakamura C, Ishii T, Kasai Y, Yoshida S (2002) A transcriptionally active maize MuDR-like transposable element in rice and its relatives. Mol Genet Genomics 268:321–330 Bonas U, Sommer H, Harrison BJ, Saedler H (1984) The transposable element Tam1 of Antirrhinum majus is 17 kb long. Mol Gen Genet 194:138–143 Brown JJ, Mattes MG, O’Reilly C, Shepherd NS (1989) Molecular characterization of rDt, a maize transposon of the “Dotted” controlling element system. Mol Gen Genet 215:239–244 Bureau TE, Wessler SR (1992) Tourist: a large family of small inverted repeat elements frequently associated with maize genes. Plant Cell 4:1283–1294 Bureau TE, Wessler SR (1994a) Stowaway: a new family of inverted repeat elements associated with the genes of both monocotyledonous and dicotyledonous plants. Plant Cell 6:907–916 Bureau TE, Wessler SR (1994b) Mobile inverted-repeat elements of the Tourist family are associated with the genes of many cereal grasses. Proc Natl Acad Sci USA 91:1411–1415 Bureau TE, Ronald PC, Wessler SR (1996) A computer-based systematic survey reveals the predominance of small inverted-repeat elements in wild-type rice genes. Proc Natl Acad Sci USA 93:8524–8529 Casa AM, Brouwer C, Nagel A, Wang L, Zhang Q, Kresovich S, Wessler SR (2000) The MITE family Heartbreaker (Hbr): molecular markers in maize. Proc Natl Acad Sci USA 97:10083–10089 Casa AM, Nagel A, Wessler SR (2004) MITE display. Methods Mol Biol 260:175–188 Fedoroff N, Wessler SR, Shure M (1983) Isolation of the transposable maize controlling elements Ac and Ds. Cell 35:235–242

78

T. Nakazaki et al.

Fedoroff N, Furtek DB, Nelson OE (1984) Cloning of the bronze locus in maize by a simple and generalizable procedure using the transposable controlling element Activator (Ac). Proc Natl Acad Sci USA 81:3825–3829 Feschotte C, Jiang N, Wessler SR (2002) Plant transposable elements: where genetics meets genomics. Nature Rev Genet 3:329–341 Han C-G, Frank MJ, Ohtsubo H, Ohtsubo E (2000) New transposable elements identified as insertions in rice transposon Tnr1. Genes Genet Syst 75:69–77 Hirochika H (2001) Contribution of the Tos17 retrotransposon to rice functional genomics. Curr Opin Plant Biol 4:118–122 Hirochika H, Sugimoto K, Otsuki Y, Tsugawa H, Kanda M (1996) Retrotransposons of rice involved in mutations induced by tissue culture. Proc Natl Acad Sci USA 93:7783–7788 Hu H, Mu J, Zhang H-J, Tao Y-Z, Han B (2006) Differentiation of a miniature inverted transposable element (MITE) system in Asian rice cultivars and its inference for a diphyletic origin of two subspecies of Asian cultivated rice. J Integrative Plant Biol 48:260–267 Inagaki Y, Hisatomi Y, Suzuki T, Kasahara K, Iida S (1994) Isolation of a Suppressor-Mutator/ Enhancer-like transposable element, Tpn1, from Japanese morning glory bearing variegated flowers. Plant Cell 6:375–383 IRGSP (International Rice Genome Sequencing Project) (2005) The map-based sequence of the rice genome. Nature 436:793–800 Jiang N, Bao Z, Zhang X, et al. (2003) An active DNA transposon family in rice. Nature 421:163–167 Kazazian Jr HH (2004) Mobile elements: drivers of genome evolution. Science 303:1626–1632 Kikuchi K, Terauchi K, Wada M, Hirano HY (2003) The plant MITE mPing is mobilized in anther culture. Nature 421:163–167 Komori T, Nitta N (2003) High frequency of sequence polymorphism in rice MITEs and application to efficient development of PCR-based markers. Breed Sci 53:85–92 Kwon SJ, Park KC, Kim JH, Lee JK, Kim NS (2005) Rim 2/Hipa CACTA transposon display; a new genetic marker technique in Oryza species. BMC Genet 6:15 Lin X, Long L, Shan X, Zhang S, Shen S, Liu B (2006) In planta mobilization of mPing and its putative autonomous element Pong in rice by hydrostatic pressurization. J Exp Bot 57:2313–2323 Maekawa M, Rikiishi K, Matsuura T, Noda K (1999) A marker line H-126, carries a genetic factor making chlorophyl mutation variagated. Rice Genet Newslett 16:61–62 Mochizuki K, Umeda M, Ohtsubo H, Ohtsubo E (1992) Characterization of a plant SINE, p-SINE1, in rice genomes. Jpn J Genet 67:155–166 Mochizuki K, Ohtsubo H, Hirano H, Sano Y, Ohtsubo E (1993) Classification and relationships of rice strains with AA genome by identification of transposable elements at nine loci. Jpn J Genet 68:205–217 Motohashi R, Ohtsubo E, Ohtsubo H (1996) Identification of Tnr3, a Suppressor-Mutator/ Enhancer-like transposable element from rice. Mol Gen Genet 250:148–152 Naito K, Cho E, Yang G, et al. (2006) Dramatic amplification of a rice transposable element during recent domestication. Proc Natl Acad Sci USA 103:17620–17625 Nakazaki T, Ikehashi H (1998) Genomic sequence and polymorphisms of a rice chitinase gene, Cht4. Breed Sci 48:371–376 Nakazaki T, Okumoto Y, Horibata A, et al. (2003) Mobilization of a transposon in the rice genome. Nature 421:170–172 Peterson PA (1988) The mobile element system in maize. In: Nelson OE (ed) Plant transposable element. Plenum Press, New York, pp 43–68 Sano H, Kamada I, Youssefian S, Katsumi M, Wabiko H (1990) A single treatment of rice seedling with 5-azacytidine induced heritable dwarfism and undermethylation of genomic DNA. Mol Gen Genet 220:441–447 Shan X, Liu Z, Dong Z, et al. (2005) Mobilization of the active MITE transposons mPing and Pong in rice by introgression from wild rice (Zizania latifolia Griseb.). Mol Biol Evol 22:976–990 Song WY, Pi LY, Bureau TE, Ronald PC (1998) Identification and characterization of 14 transposonlike elements in the noncoding regions of members of the Xa21 family of disease resistance genes in rice. Mol Gen Genet 258:449–456

I.6 Active Transposons in Rice

79

Tenzen T, Matsuda Y, Ohtsubo H, Ohtsubo E (1994) Transposition of Tnr1 in rice genomes to 5′PuTAPy-3′ sites, duplicating the TA sequence. Mol Gen Genet 245:441–448 Teraishi M, Okumoto Y, Hirochika H, Horibita A, Yamagata H, Tanisaka T (1995) Identification of a mutable slender glume gene in rice (Oryza sativa L.). Mol Gen Genet 261:487–494 Teraishi M, Okumoto Y, Hirochika H, Horibata A, Yamagata H, Tanisaka T (1999) Identification of mutable slender glume gene in rice (Oryza sativa L.). Mol Gen Genet 261:487–494 Teraishi M, Hirochika H, Okumoto Y, Horibata A, Yamagata H, Tanisaka T (2001) Identification of YAC clones containing the mutable slender glume locus slg in rice (Oryza sativa L.). Genome 44:1–6 Tsugane K, Maekawa M, Takagi K, et al. (2006) An active DNA transposon nDart causing leaf variegation and mutable dwarfism and its related elements in rice. Plant J 45:46–57 Turcotte K, Srinivasan S, Bureau T (2001) Survey of transposable elements from rice genomic sequences. Plant J 25:169–179 Umeda M, Ohtsubo H, Ohtsubo E (1991) Diversification of rice waxy gene by insertion of mobile DNA elements into introns. Jpn J Genet 66:569–586 Wessler SR, Bureau TE, White SE (1995) LTR-retrotransposons and MITEs: important players in the evolution of plant genomes. Curr Opin Genet Dev. 5:814–821 Xu Z, Dooner HK (2005) Mx-rMx, a new family of interacting transposons in the growing hAT superfamily of maize. Plant Cell 17:375–388 Yang G, Hall TC (2003) MDM-1 and MDM-2: two Mutator-derived MITE families in rice. J Mol Evol 56:255–264

This page intentionally blank

I.7

Homologous Recombination-dependent Gene Targeting and an Active DNA Transposon nDart-promoted Gene Tagging for Rice Functional Genomics Yasuyo Johzuka-Hisatomi1, Masahiko Maekawa2, Kyoko Takagi1,3, Chang-Ho Eun1, Takaki Yamauchi1,4, Zenpei Shimatani5, Nisar Ahmed2, Hiroko Urawa1, Kazuo Tsugane1,5, Rie Terada1,5, and Shigeru Iida1,5(* ü)

1

Introduction

Rice is an important staple food for more than half the world’s population, and it has become the first crop plant to have its 389-Mb genome sequenced (IRGSP 2005). Even though various functional genomic tools for elucidating the function of rice genes are available (Hirochika et al. 2004; Leung and An 2004; Sasaki et al. 2005; Upadhyaya 2007), developing new methods for characterizing genes of interest by both forward and reverse genetic approaches has become particularly important. This chapter contains a description of the current state of gene targeting mediated by homologous recombination (HR) and gene tagging promoted by a nonautonomous DNA-based active rice transposon, nDart, for rice functional genomics, both of which are being developed (Terada et al. 2002, 2007; Tsugane et al. 2006; Takagi et al. 2007). Gene targeting refers to the alteration of a specific DNA sequence in an endogenous gene at its original locus in the genome and, often, to

1 National Institute for Basic Biology, Okazaki 444-8585, Japan e-mail: [email protected] (Y.J.-H.), [email protected] (K. Takagi); [email protected] (C.-H.E.); [email protected] (T.Y.); [email protected] (H.U.); [email protected] (K. Tsugane); [email protected] (R.T.); [email protected] (S.I.) 2 Research Institute for Bioresources, Okayama University, Kurashiki 710-0046, Japan e-mail: [email protected] (M.M.), [email protected] (N.A.) 3

Plant Breeding Laboratory, Graduate School of Agriculture, Hokkaido University, Sapporo 060-8589, Japan (K. Takagi) 4

Graduate School of Science and Technology, Chiba University, Matsudo 271-8510, Japan (T.Y.)

5

Department of Basic Biology, School of Life Science, The Graduate University for Advanced Studies, Okazaki 444-8585, Japan e-mail: [email protected] (Z.S.), (K. Tsugane), (R.T.), (S.I.)

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

81

82

Y. Johzuka-Hisatomi et al.

the conversion of the endogenous gene into a designed sequence. Gene targeting has been shown to lead to both gene replacements and base changes (Reiss 2003; Iida and Terada 2004, 2005); by the end of 2006, there was only one report describing reproducible true gene targeting (TGT) in rice, and it dealt with the inactivation or knockout of the endogenous Waxy gene by gene disruption (Figs. 1a and 2; Terada et al. 2002). Because independent gene targeting events by HR should generate an identical genomic structure with a designed sequence alteration(s), the experimental demonstration of the capability for the reproducible isolation of recombinants with the anticipated gene structure would be very important. For gene tagging in rice, foreign elements, such as T-DNA or the maize DNA transposon, Ac/Ds and En/Spm, and the endogenous retrotransposon, Tos17, have

Fig. 1 Strategy for gene targeting in rice. a The Waxy gene. b The Adh2 gene. c The MET1-1 gene. d The DFR gene. Shaded boxes and filled arrowheads indicate homologous regions carried by the introduced T-DNA segments and their border sequences, respectively. The hpt and DT-A genes for hygromycin B resistance and diphtheria toxin A fragment are used as positive and negative markers, respectively. Small lollipops above the DFR pentagonals represent nonsense mutations. The CaMV 35S promoter is indicated by P35S

I.7 Gene Targeting and Tagging in Rice

83

Fig. 2 Gene targeting and ectopic recombination events. The anticipated TGT is generally regarded to occur via double crossovers at flanking homologous regions on the vector, and brackets under the maps indicate junction fragments generated by crossovers and used for screening homologous recombinant calli by PCR (Iida et al. 2007; Terada et al. 2007). The most efficient integration of a transgene in Agrobacterium-mediated transformation is BARI, in which integrated single T-DNA molecules contain the entire T-DNA segment with a well-conserved right border (dark-gray parts of arrowhead) and either conserved or slightly truncated left border (open part of arrowhead) sequences. It is noteworthy that such transformants are to be killed by the DT-A proteins expressed when the vectors in Fig. 1 are employed. Other symbols are as in Fig. 1

been systematically employed; moreover, tissue cultures are necessary to either introduce these foreign elements into rice calli or activate dormant Tos17 in the genome (Hirochika et al. 2004; Leung and An 2004; Upadhyaya 2007). While endogenous DNA transposon have been extensively used for gene tagging in maize, snapdragon, petunia, and three morning glory species (May and Martienssen 2003; Chopra et al. 2006), two active endogenous DNA transposon in rice, mPing and nDart, have been identified only recently. The 0.43-kb element mPing of the miniature inverted-repeat transposable element (MITE) family was shown to actively transpose in rice cell cultures (Jiang et al. 2003), in plants regenerated from anther-derived calli (Kikuchi et al. 2003), and in γ-ray-irradiated rice plants (Nakazaki at al. 2003). It was also reported very recently that mPing is active in a certain rice cultivar under natural growth conditions (Naito et al. 2006). Another active endogenous element, nDart (nonautonomous DNA-based active rice transposon), of the Ac/Ds or hAT family, is 0.6 kb in length and a causative element of a spontaneous mutable virescent allele, pyl-v (pale-yellow leaf-variegated), conferring pale-yellow leaves with dark-green sectors in the seedlings (Maekawa et al. 1999; Tsugane et al. 2006). The pyl-v allele is caused by the disruption of the nuclear-coded chloroplast protease gene, OsClpP5, due to insertion of a 607-bp element, named nDart1-0 (Fig. 3a). Subsequently, we showed that the transposition of nDart1-0 can be induced by crossing with a line containing an active autonomous aDart element and stabilized by segregating aDart under natural growth conditions. At present, the nDart/aDart system appears to be the only endogenous rice DNA transposon system whose transposition activity can be controlled under natural

84

Y. Johzuka-Hisatomi et al.

Fig. 3 Structures and distribution of nDart-related elements in the rice genome. a Structure of the mutable virescent allele pyl-v. Open and shaded boxes represent the OsClpP5 untranslated exon and coding regions, respectively. b Structures of nDart-related elements. Horizontal filled arrowheads and shaded or hatched boxes at both ends indicate the TIRs and subterminal regions, respectively, and numerals in parentheses indicate the copy numbers of a group of elements with closely related sequences. Pentagonal arrows within the iDart boxes represent open reading frames for potential transposase genes. Brackets with vertical arrows under the iDart/dDart boxes indicate that large deletions occurred during the generation of nDart or dDart elements, and black vertical lines within the Dart boxes attached to open and filled circles or open arrowheads indicate the microhomologies for the deletion formations. c Localization of nDart-related elements in the Nipponbare genome. Short and long bars represent nDart1 and iDart/dDart elements, respectively. Asterisk and pentagram indicate pyl-v and thl-m alleles, respectively

growth conditions without any artificial treatments, including tissue cultures. This feature of the nDart/aDart system may hold a considerable advantage over other systems for gene tagging in rice because no somaclonal variation due to tissue cultures (Kaeppler et al. 2000) is expected to occur.

2 2.1

Homologous Recombination-dependent Gene Targeting Current Status of Gene Targeting by Homologous Recombination in Higher Plants

In Agrobacterium-mediated transformation, which has been used in all of the successful gene targeting studies by HR in higher plants (Iida and Terada 2005; Iida et al. 2007), T-DNA appears to integrate randomly throughout the plant genome as

I.7 Gene Targeting and Tagging in Rice

85

a single molecule or multiple sequences ligated with each other in various orientations (Tzfira et al. 2004). The majority of the randomly integrated singlecopy T-DNA molecules mediated by nonhomologous end-joining (NHEJ) are known to contain the entire T-DNA segment with a well-conserved right border, and a left-border sequence that is either conserved or truncated by a few to around 100 bp (Tinland and Hohn 1995; Brunaud et al. 2002; Zhu et al. 2006); we termed this type of random integrations collectively as border-associated random integration (BARI) (Fig. 2; Iida et al. 2007; Terada et al. 2007). There appears to be another type of random integrations with relatively large deletions at both ends of the T-DNA segment without the border proximal regions, which is called borderindependent random integration (BIRI) (Fig. 2; Matsumoto et al. 1990; Terada et al. 2007). BIRI appears to occur much less frequently than BARI. Since a significant portion of single-stranded T-DNA imported into the plant nucleus can become double-stranded in Agrobacterium-mediated transformation (Tzfira et al. 2004), it has been speculated that such BIRI processes must share common recombination mechanisms with random-integration processes by direct DNA delivery methods using double-stranded DNA (Tinland and Hohn 1995; Sommers and Makarevitch 2004; Terada et al. 2007). While HR-mediated sequence-specific integration of a transgene generating TGT in mouse embryogenic stem (ES) cells has been reported to be 1% or higher of all integration events, most of which are the random integration of the transgene by NHEJ (Jasin et al. 1996), TGT in higher plants has been thought to be in the order of 0.01–0.1% compared with random integration; furthermore, not only random integration but also undesirable ectopic recombinations, such as one-sided invasion (OSI) and ectopic gene targeting (EGT), have been occasionally detected (Fig. 2; Reiss 2003; Iida and Terada 2005). OSI results from one homologous crossover and another NHEJ at the target locus, while EGT is thought to be generated by random integration of a recombinant molecule produced by homologous crossovers between the introduced transgene and a copy of the target sequence without altering the gene to be targeted (Fig. 2). To isolate targeted events efficiently among the overwhelming random integration events, one approach is to apply gene-specific direct selection or screening for the target genes; the modified Arabidopsis PPO gene for protoporphyrinogen oxidase was reproducibly isolated by gene-specific direct selection, through which targeted plants acquired herbicide resistance (Hanin et al. 2001), and the disrupted Arabidopsis Cruciferin gene (encoding a seed storage protein), by fusing in-frame with the gfp gene, was reproducibly obtained by gene-specific visual screening, through which targeted plants produced fluorescent seeds (Shaked et al. 2005). The targeting frequencies of the PPO and Cruciferin genes in the wild-type Arabidopsis were estimated to be approximately 7.2 × 10−4 and 5.6 × 10−3 TGT events per transformant, respectively. Moreover, the targeting frequencies of the Cruciferin gene in the transgenic Arabidopsis plants overexpressing the yeast RAD54 gene for an SWI2/SNF2 chromatin-remodeling protein were shown to be approximately 9.9 × 10−2 per transformant (Shaked et al. 2005).

86

Y. Johzuka-Hisatomi et al.

Another approach is to use a strong positive–negative selection for enriching homologous recombinants indirectly by reducing transformants with randomly integrated transgenes and to isolate TGT recombinants by polymerase chain reaction (PCR) screening (Fig. 2); the mutants having the single-copy Waxy modified in rice were reproducibly obtained in this way (Fig. 1a; Terada et al. 2002). In their vector used for the transformation of embryogenic rice calli by Agrobacterium, the strongly expressed hpt (hph) gene was flanked by targeting Waxy homologous sequences of 6.3–6.8 kb, and two inversely oriented DT-A genes fused with strong and constitutive promoters were placed at both ends of the T-DNA segment adjacent to its border sequences (Fig. 1a) in order to eliminate the overwhelming BARI events (Fig. 2). Six independent truly targeted Waxy knockout lines were isolated at a frequency of 6.5 × 10−4 per transformant or 9.4 × 10−3 per surviving callus, and all of their primary (T0) transgenic lines were heterozygous (one wild-type allele and another target allele). The targeted allele was shown to transmit into the selfed progeny (T1) in a Mendelian fashion. No ectopic events, including OSI, EGT, or an additional random integration of the positive hpt selection marker, were detected. Subsequently, we examined whether the presence of homologous and/or highly repetitive sequences carried by a gene in the multigene family causes a preclusive effect on gene targeting. We chose the Adh2 gene for gene targeting in a small multigene family because (1) no adh2 mutation in rice has been reported, (2) a highly repetitive Copia-like element is present 1.0 kb downstream of Adh2, (3) Adh2 resides in the middle of Adh1 and Adh3 in the genome, and (4) all three Adh genes are expressed significantly in rice calli, where somatic HR for gene targeting occurs (Terada et al. 2007). The hpt gene in the vector used is flanked by the 6.2-kb Adh2 promoter sequence and the 4.0-kb Adh2 and its adjacent region together with the 2.0-kb 3′-part of a Copia-like retroelement (Fig. 1b). Nine independent transformed calli carrying the anticipated Adh2 modification in the heterozygous condition were isolated with a frequency of 1.9 × 10−2 per surviving callus, and, subsequently, eight fertile transgenic plants without ectopic events were obtained (the remaining callus was accidentally lost), suggesting that no apparent preclusion by homologous and/or repetitive sequences was observed. Moreover, the surviving calli with positive–negative selection were found to contain truncated T-DNA segments including the active hpt gene that had been integrated into the genome by NHEJ processes through BIRI (Fig. 2). Thus, the targeting frequency estimated as targeted calli per surviving callus with positive– negative selection can be interpreted to be HR-promoted targeting per NHEJmediated BIRI event of the transgene(s) including the positive hpt selection marker (Iida et al. 2007; Terada et al. 2007). The targeting frequencies of the Waxy and Adh2 genes calculated as targeted calli per surviving callus with the positive–negative selection were approximately 1% and 2%, respectively (Terada et al. 2002, 2007), comparable with those in mouse ES cells (Jasin et al. 1996). Therefore, the previous concept, according to which the overwhelming occurrence of random integration of transgenes by NHEJ relative to targeted HR is the main obstacle to the development of an efficient gene targeting

I.7 Gene Targeting and Tagging in Rice

87

system in higher plants (Reiss 2003; Iida and Terada 2004; Tzfira and White 2005) can be reconciled because Agrobacterium-mediated transformation via BARI of T-DNA is a highly efficient NHEJ process (Fig. 2; Tzfira et al. 2004; Tzfira and Citovsky 2006); moreover, the targeting frequency estimated by targeted recombinants per BARI-mediated transformant may be too underestimated to calculate the proper targeting frequency in higher plants. Possible models for the generation of successful gene targeting events with positive–negative selection have been discussed (Iida and Terada 2004, 2005). In addition, because the moss Physcomitrella patens exhibiting highly frequent gene targeting capability has sometimes been regarded as a model organism concerning gene targeting in plant (Reiss 2003), considerably different characteristic features of gene targeting between the moss system and our rice system have been discussed (Terada et al. 2007).

2.2

An Attempt to Generate Knockin Mutants by Homologous Recombination

Both Waxy and Adh2 genes were disrupted by insertion of the positive hpt marker because all of the vectors introduced were “knockout” constructs (Fig. 1). By modifying the basic design of the vector used, it would be feasible to obtain “knockin” transformants; e.g., a reporter gene gus (or gfp) can be fused with a natural endogenous promoter at its original locus. Because almost all of the primary transgenic (T0) plants having Waxy or Adh2 modified carry only one copy of the hpt gene with the anticipated structure in the heterozygous condition (Terada et al. 2002, 2007), the introduction of a reporter gene into an endogenous promoter through knockin targeting would provide the most effective means to assay the promoter activity by minimizing the inter-individual variation of transgene expression that is often observed within populations of transgenic plants obtained by random integrations of the same transgene construct. The inter-individual variations are attributed mainly to the transgene copy number, epigenetic gene silencing caused by the generation of directly or inversely repeated transgenes, and insertion sites of the transgene (Matzke and Matzke 1998; Butaye et al. 2005). As an initial attempt, we chose one of two rice MET1 genes for DNA methylase involved in maintenance DNA methylation, MET1-1 on chromosome 3 (Teerawanichpan et al. 2004), and tried to obtain homologous recombinant calli having the gus reporter gene fused with the endogenous MET1-1 promoter (Fig. 1c). Such targeted calli were obtained at a frequency of 5.6 × 10−2 per surviving callus with positive–negative selection, and more than ten targeted calli gave invariable Gus staining in their intensity (T. Yamauchi, Y. Johzuka-Hisatomi, and R. Terada, unpublished). When a T-DNA-based vector containing the MET1-1 promoter fused with gus and the positive hpt selection marker without the negative DT-A selection marker was used, by way of contrast, the intensity of Gus staining in more than 15 transformed calli containing randomly integrated gus varied considerably (T. Yamauchi, unpublished). The results clearly indicate that the inter-individual variations can be minimized with a targeted knockin procedure.

88

Y. Johzuka-Hisatomi et al.

We are in the process of regenerating transgenic rice plants from these targeted knockin calli and of examining their Gus staining patterns.

2.3

An Attempt to Restore a Defect for Gene Therapy by Homologous Recombination

Actually, the strategy of the targeted knockin mentioned above not only leads to knockin targeting but also results in knockout targeting simultaneously, because the fusion of a reporter gene to an endogenous promoter of a target gene results in insertional disruption of the target gene. HR-dependent gene targeting can also be used for the restoration of a defect in a target gene, which is one of the approaches for gene therapy. Gene therapy is a technique for correcting defective genes and includes the following four approaches (http://www.ornl.gov/sci/techresources/ Human_Genome/medicine/genetherapy.shtml): (1) a normal gene may be inserted into a nonspecific location within the genome to replace a nonfunctional gene, which is the most common approach; (2) an abnormal gene could be replaced by a normal gene through HR; (3) the abnormal gene could be repaired through selective reverse mutation; and (4) the regulation of a particular gene could be altered. Because the genome of a cultivated rice plant bears various alleles (Olsen et al. 2006), it would be interesting to correct only a certain specific allele by gene targeting. As an initial step for such attempts, we chose the DFR gene encoding dihydroflavonol 4-reductase in anthocyanin biosynthesis as a model gene to be modified. The rice DFR gene for anthocyanin pigmentation is a unique gene consisting of three exons on chromosome 1 and corresponds to the A (Activator) locus for anthocyanin production; both cultivars, Nipponbare and T-65, carry an identical single-base alteration leading to a nonsense mutation at its exon 2 (Fig. 1d; Nakai et al. 1998; Maekawa et al. 2001; Furukawa et al. 2007). We employed calli produced from a T-65 derivative carrying the dominant Plw allele (M. Maekawa, unpublished) and introduced the wild-type DFR cDNA from Murasaki-ine (Nakai et al. 1998) fused with the CaMV 35S promoter by Agrobacterium-mediated transformation. The rice Pl (Purple leaf) locus on chromosome 4 encodes a basic helix–loop–helix (bHLH) transcriptional regulator and controls tissue-specific anthocyanin accumulation, and the dominant Plw allele comprises duplicated bHLH genes conferring anthocyanin pigmentation in almost all aerial tissues (Sakamoto et al. 2001). After obtaining the results that the introduced wild-type DFR cDNA complements the defect in the DFR gene and anthocyanin pigmentation occurs in transformed calli as well as various tissues in transgenic plants (R. Terada, H. Urawa, and M. Maekawa, unpublished), we constructed a vector carrying the wild-type DFR genomic sequence from Murasaki-ine and introduced an artificial nonsense mutation within the DFR exon 3 (Fig. 1d). Subsequently, calli produced from the same T-65 derivative containing the dominant Plw allele were transformed with the resulting vector by Agrobacterium. Thus, double crossovers (one crossover at the promoter regions and another at the DFR coding regions between the two

I.7 Gene Targeting and Tagging in Rice

89

nonsense mutations) would result in the wild-type DFR gene being driven from the CaMV 35S promoter. So far, we have been able to obtain a callus producing an anticipated junction fragment, and we are in the process of characterizing its genomic structure (Y. Johzuka-Hisatomi, M. Maekawa, R. Terada, and H. Urawa, unpublished). Although targeted restoration of defects in an endogenous gene by HR remains in its infancy, the attempt described here may help broaden the applications of gene targeting for functional genomic studies in rice.

3

Active DNA Transposon nDart-promoted Gene Tagging

All of the known transposon superfamilies contribute to at least 35% of the Nipponbare genome, and DNA transposon belonging to the hAT superfamily are reported to constitute approximately 0.4% of the genome (IRGSP 2005). The recently identified nDart-related elements belong to the hAT superfamily, and the 3.0 pseudomolecules of the Nipponbare genome in the IRGSP sequence (released in February 2005) were found to contain 18 copies of the 0.6-kb nDart-related sequences and approximately 60 copies of other nDart-related elements longer than 2 kb, some of which encode a putative transposase (Tsugane et al. 2006). The nDart-related elements can be classified into three subgroups of about 0.6-kb nonautonomous elements (nDart1-3, nDart1-101, and nDart1-201) and four subgroups of elements longer than 2 kb, which comprise epigenetically silenced inactive iDart and genetically defective dDart elements (iDart1/ dDart1, iDart2, iDart3, and iDart4) on the basis of their lengths and sequence characteristics (Fig. 3b). Once again, we searched for the nDart-related elements in the Nipponbare genome with the latest 4.0 pseudomolecules released in August 2005 (Takagi et al. 2007). The three previously reported iDart1 elements, iDart1-44, iDart1-46, and iDart1-48, on chromosome 8 were found to be the same element, designated as iDart1-44, and the two elements, iDart1-31 and iDart1-62, on chromosome 6 were also the same sequence, represented by iDart1-31. Thus, the copy numbers of the small nDart1 elements of 0.6 kb and longer iDart/dDart elements are 18 and 63, respectively, and both elements are distributed throughout the Nipponbare chromosomes (Fig. 3c). The mutable pyl-v line displaying variegated leaves also contains at least 18 nDart1 elements, 17 of which are in common with those carried by Nipponbare (Tsugane et al. 2006). Of these, 13 elements belong to the nDart1-3 subgroup, and one of them, nDart1-0, residing at the OsClpP5 gene in the pyl-v line, was found to transpose most actively. Two and three other nDart1 elements in the pyl-v line exhibited modest and slight transposition activities, respectively, whereas the remaining elements showed no activity at all, even though the active aDart element was present in the genome. The subterminal regions of the active and inactive nDart1 elements were found to be hypomethylated and hypermethylated, respectively (C.-H. Eun and K. Tsugane, unpublished), indicating that the DNA methylation state of the elements must be an important factor for their transposition activities. Since transposon display is a powerful technique to identify the integration site of transposons in gene tagging (Maes et al. 1999; Fukada-Tanaka et al. 2001), we

90

Y. Johzuka-Hisatomi et al.

have developed an efficient procedure by comparing the observed bands in transposon display with the anticipated virtual bands of the nDart-related elements based upon the available rice genome sequence. Even though it was rather difficult to visualize all of the nDart-related elements, probably due to the GC richness of the elements and their flanking regions, we were able to improve our transposon display protocol by optimizing the PCR amplification conditions and thus visualize all of the 0.6-kb nDart elements and over 80% of longer nDart-related elements (Takagi et al. 2007). The GC content of the 0.5-kb 5′-flanking regions of the longer nDart-related elements that failed to give PCR-amplified fragments was found to be 62.5% on average, which is much higher than the percentage of the successfully amplified elements (45.6%). The results suggest that visualization of the expected bands in the transposon display becomes more difficult when the flanking sequences of the elements bear higher GC content. Using the optimized transposon display procedure, we could identify several mutable alleles caused by the insertion of nDart 1-0 (M. Maekawa, K. Takagi, and K. Tsugane, unpublished), and one of them, thumbelina-mutable, which confers the mutable gibberellin-insensitive dwarf phenotype, was found to be a new allele of the GIBBELLERIN INSENSITIVE DWARF1 (GID1) gene encoding a soluble gibberellin receptor (Ueguchi-Tanaka et al. 2005; Tsugane et al. 2006). The results indicate that our nDart-promoted gene tagging system is effective. We also noticed that the 0.6-kb nDart1-0 and some other nDart1-3 subgroup elements were often found to be integrated into either the exon or 5′-UTR regions in aDart-containing lines (Tsugane et al. 2006; K. Takagi and M. Maekawa, unpublished). Because the transposition of the nDart1-3 subgroup elements is shown to be much more active than that of the remaining nDart-related elements and because all of the nDart1-3 subgroup elements in Nipponbare can be successfully visualized using our transposon display procedure (Tsugane et al. 2006; Takagi et al. 2007), we can efficiently identify newly tagged genes by the nDart1-3 subgroup elements. We are currently generating a large number of mutant lines tagged by nDart1 and its relatives. To generate such mutant lines, we have developed a population consisting of panicle-row lines of Pyl-r revertants derived from the selfed progeny of a pyl-v line carrying both aDart and nDart1-0. The starting pyl-v line used exhibits early heading and black purple-pigmented apiculus, which allow us to shorten the generation time and ensure the homogeneity of the mutant lines, respectively (M. Maekawa, unpublished). Although the frequency of revertants from the pyl-v line varied, we were able to obtain an average of 25 revertants per line (M. Maekawa and N. Ahmed, unpublished). As a preliminary investigation, we examined 866 panicle-row lines from 106 revertants derived from four plants and found 293 phenotypic mutants, which include those conferring chlorophyll defects, dwarfisms, or sterile phenotypes, with a frequency of 33.8% mutants per population studied. Because the frequency of phenotypic mutants detected among regenerated plants derived from Tos17-activated tissue cultures was estimated to be 43.7% mutants per population studied (Miyao et al. 2007), the apparent frequency of phenotypic mutants detected in the nDart-active lines appears to be slightly lower than that in the Tos17-activated lines. However, the tagging frequency of

I.7 Gene Targeting and Tagging in Rice

91

Tos17 was reported to be only 5–10% (Hirochika 2001) because the high occurrence of somaclonal variation, which refers to genetic and epigenetic changes induced by the tissue culture (Kaeppler et al. 2000), is inevitable. nDart-promoted gene tagging is likely to be more effective than Tos17-mediated gene tagging because no somaclonal variation should occur in the former tagging system without involving tissue culture.

4

Current Status and Future Prospects

We have briefly described here our efforts to develop two new approaches for the functional genomic characterization of rice genes: HR-dependent gene targeting and active DNA transposon nDart-promoted gene tagging. Because gene targeting by HR is thought to be an infrequent event in higher plants, various approaches, including the engineering of the host recombination and/or repair systems, have been attempted to enhance HR processes or to suppress NHEJ (Reiss 2003; Iida and Terada 2005; Shaked et al. 2005; Tzfira and White 2005; Iida et al. 2007). Although plant genes affecting HR have been reported (Britt and May 2003; Reiss 2003; Schuermann et al. 2005; Emmanuel et al. 2006; Endo et al. 2006; Kirik et al. 2006), we continue to pursue the use of wild-type rice to avoid the potential side effects caused by altering the recombination and/ or repair systems; our ultimate goal is to develop a general reverse genetic method to characterize an endogenous gene by modifying the gene of interest (Terada et al. 2002, 2007; Iida and Terada 2004, 2005; Iida et al. 2007). It should be emphasized here that our large-scale transformation procedure with a strong positive–negative selection, combined with subsequent PCR screening, is sufficiently effective to obtain transgenic rice plants carrying anticipated modifications in their target genes (Terada et al. 2002, 2007; Iida et al. 2007). Indeed, we have generated fertile transgenic rice plants having the Waxy and Adh2 genes disrupted by the introduction of the hpt gene (Terada et al. 2002, 2007). Using basically the same strategy, we were able to obtain more than three independent transgenic rice plants with either an altered Adh1 gene for alcohol dehydrogenase or a modified DDM1a (Os09g0442700) gene, one of two rice DDM1 genes for an SWI2/SNF2 chromatin-remodeling protein (Jeddeloh et al. 1999; IRGSP 2005; Y. Johzuka-Hisatomi and R. Terada, unpublished). Because tissue culture is necessary in almost all of the currently available reverse genetic procedures in rice (Hirochika et al. 2004; Leung and An 2004; Sasaki et al. 2005; Upadhyaya 2007), the occurrence of somaclonal variation appears to be inevitable and may hamper the efficient characterization of gene function. Indeed, the tagging efficiency with the endogenous retrotransposon Tos17 is reported to be very low due to the high occurrence of somaclonal variation because the tissue culture is a prerequisite for activating the dormant Tos17 element (Hirochika et al. 1996; Hirochika 2001). In this respect, our nDart-promoted gene tagging system appears to have considerable potential for elucidating the function

92

Y. Johzuka-Hisatomi et al.

of rice genes because no somaclonal variation is expected to occur and because the frequency of phenotypic mutants that appeared in the descendants of the Pyl-r revertants generated by excision of nDart1-0 from the OsClpP5 gene was only slightly lower than that in the derivatives of the Tos17-activated plants, as reported above. Therefore, both HR-dependent gene targeting and active DNA transposon nDart-promoted gene tagging should facilitate rice functional genomic studies, particularly those elucidating the function of a large number of putative genes annotated (The Rice Annotation Project 2007). Acknowledgments The work in our laboratories was supported by grants from the Ministry of Agriculture, Forestry, and Fisheries of Japan (Green Technology Project IP1007), the Program for Promotion of Basic Research Activities for Innovative Biosciences (PROBRAIN) from the Bio-oriented Technology Research Advancement Institution (BRAIN) in Japan, and the Ministry of Education, Culture, Sports, Science, and Technology of Japan. We also thank Atsushi Hoshino, Yoshishige Inagaki, Makoto Takano, Katsushi Yamaguchi, Satoru Moritoh, Gonghao Jiang, and Akemi Ono for discussions, Ikuo Nakamura for his encouragements, and Akio Miyao and Hirohiko Hirochika for sharing unpublished results. Y. J.-H., K.T., and Z.S. are recipients of a fellowship awarded by the Japan Society for the Promotion of Science for Young Scientists, and C.-H.E. and N.A. are recipients of a fellowship awarded by the Japan Society for the Promotion of Science for Foreign Researchers.

References Britt AB, May GD (2003) Re-engineering plant gene targeting. Trends Plant Sci 8:90–95 Brunaud V, Balzergue S, Dubreucq B, et al. (2002) T-DNA integration into the Arabidopsis genome depends on sequences of pre-insertion sites. EMBO Rep 3:1152–1157 Butaye KMJ, Cammue BPA, Delauré SL, De Bolle MFC (2005) Approaches to minimize variation of transgene expression in plants. Mol Breed 16:79–91 Chopra S, Hoshino A, Boddu J, Iida S (2006) Flavonoid pigments as tools in molecular genetics. In: Glotewold E (ed) The science of flavonoids. Springer, Berlin Heidelberg New York, pp. 147–173 Emmanuel E, Yehuda E, Melamed-Bessudo C, Avivi-Ragolsky N, Levy AA (2006) The role of AtMSH2 in homologous recombination in Arabidopsis thaliana. EMBO Rep 7:100–105 Endo M, Ishikawa Y, Osakabe K, et al. (2006) Increased frequency of homologous recombination and T-DNA integration in Arabidopsis CAF-1 mutants. EMBO J 25:5579–5590 Fukada-Tanaka S, Inagaki Y, Yamaguchi T, Iida S (2001) Simplified transposon display (STD): a new procedure for isolation of a gene tagged by a transposable element belonging to the Tpn1 family in the Japanese morning glory. Plant Biotechnol 18:143–149 Furukawa T, Maekawa M, Oki T, et al. (2007) The Rc and Rd genes are involved in proanthocyanidin synthesis in rice pericarp. Plant J 49:91–102 Hanin M, Volrath S, Bogucki A, Briker M, Ward E, Paszkowski J (2001) Gene targeting in Arabidopsis. Plant J 28:671–677 Hirochika H (2001) Contribution of the Tos17 retrotransposon to rice functional genomics. Curr Opin Plant Biol 4:118–122 Hirochika H, Sugimoto K, Otsuki Y, Tsugawa H, Kanda M (1996) Retrotransposons of rice involved in mutations induced by tissue culture. Proc Natl Acad Sci USA 93:7783–7788 Hirochika H, Guiderdoni E, An G, et al. (2004) Rice mutant resources for gene discovery. Plant Mol Biol 54:325–334 Iida S, Terada R (2004) A tale of two integrations, transgene and T-DNA: gene targeting by homologous recombination in rice. Curr Opin Biotechnol 15:132–138

I.7 Gene Targeting and Tagging in Rice

93

Iida S, Terada R (2005) Modification of endogenous natural genes by gene targeting in rice and other higher plants. Plant Mol Biol 59:205–219 Iida S, Johzuka-Hisatomi Y, Terada R (2007) Gene targeting by homologous recombination for rice functional genomics. In: Upadhyaya N (ed) Rice functional genomics – challenges, progress and prospects. Springer, Berlin Heidelberg New York, pp. 273–289 IRGSP (International Rice Genome Sequencing Project) (2005) The map-based sequence of the rice genome. Nature 436:793–800 Jasin M, Moynahan ME, Richardson C (1996) Targeted transgenesis. Proc Natl Acad Sci USA 93:8804–8808 Jeddeloh JA, Stokes TL, Richards EJ (1999) Maintenance of genomic methylation requires a SWI2/SNF2-like protein. Nat Genet 22:94–97 Jiang N, Bao Z, Zhang X, et al. (2003) An active DNA transposon family in rice. Nature 421:163–167 Kaeppler SM, Kaeppler HF, Rhee Y (2000) Epigenetic aspects of somaclonal variation in plants. Plant Mol Biol 43:179–188 Kikuchi K, Terauchi K, Wada M, Hirano HY (2003) The plant MITE mPing is mobilized in anther culture. Nature 421:167–170 Kirik A, Pecinka A, Wendeler E, Reiss B (2006) The chromatin assembly factor subunit FASCIATA1 is involved in homologous recombination in plants. Plant Cell 18:2431–2442 Leung H, An G (2004) Rice functional genomics: large-scale gene discovery and applications to crop improvement. Adv Agron 82:55–111 Maekawa M, Rikiishi K, Matsuura T, Noda K (1999) A marker line H-126, carries a genetic factor making chlorophyll mutation variegated. Rice Genet Newslett 16:61–62 Maekawa M, Himi E, Inagaki Y, Iida S, Noda K, Sakamoto W (2001) Anthocyanin activator A encodes DFR in rice anthocyanin biosynthetic pathway. Breed Res 3 (Suppl 1):229 (in Japanese) Maes T, De Keukeleire P, Gerats T (1999) Plant tagnology. Trends Plant Sci 4:90–96 Matsumoto S, Ito Y, Hosoi T, Takahashi Y, Machida Y (1990) Integration of Agrobacterium T-DNA into a tobacco chromosome: possible involvement of DNA homology between T-DNA and plant DNA. Mol Gen Genet 224:309–316 Matzke AJ, Matzke MA (1998) Position effects and epigenetic silencing of plant transgenes. Curr Opin Plant Biol 1:142–148 May BP, Martienssen RA (2003) Transposon mutagenesis in the study of plant development. Crit Rev Plant Sci 22:1–35 Miyao A, Iwasaki Y, Kitano H, et al. (2007) A large-scale collection of phenotypic data describing an insertional mutant population to facilitate functional analysis of rice genes. Plant Mol Biol 63:625–635 Naito K, Cho E, Yang G, et al. (2006) Dramatic amplification of a rice transposable element during recent domestication. Proc Natl Acad Sci USA 103:17620–17625 Nakai K, Inagaki Y, Nagata H, Miyazaki C, Iida S (1998) Molecular characterization of the gene for dihydroflavonol 4-reductase of Japonica rice varieties. Plant Biotechnol 15:221–225 Nakazaki T, Okumoto Y, Horibata A, et al. (2003) Mobilization of a transposon in the rice genome. Nature 421:170–172 Olsen KM, Caicedo AL, Polato N, McClung A, McCouch S, Purugganan MD (2006) Selection under domestication: evidence for a sweep in the rice Waxy genomic region. Genetics 173:975–983 Reiss B (2003) Homologous recombination and gene targeting in plant cells. Int Rev Cytol 228:85–139 Sakamoto W, Ohmori T, Kageyama K, et al. (2001) The Purple leaf (Pl) locus of rice: the Plw allele has a complex organization and includes two genes encoding basic Helix-Loop-Helix proteins involved in anthocyanin biosynthesis. Plant Cell Physiol 42:982–991 Sasaki T, Matsumoto T, Antonio BA, Nagamura Y (2005) From mapping to sequencing, postsequencing and beyond. Plant Cell Physiol 46:3–13 Schuermann D, Molinier J, Fritsch O, Hohn B (2005) The dual nature of homologous recombination in plants. Trends Genet 21:172–181

94

Y. Johzuka-Hisatomi et al.

Shaked H, Melamed-Bessudo C, Levy AA (2005) High-frequency gene targeting in Arabidopsis plants expressing the yeast RAD54 gene. Proc Natl Acad Sci USA 102:12265–12269 Somers D, Makarevitch I (2004) Transgene integration in plants: poking or patching holes in promiscuous genomes? Curr Opin Biotechnol 15:126–131 Takagi K, Ishikawa N, Maekawa M, Tsugane K, Iida S (2007) Transposon display for active DNA transposons in rice. Genes Genet Syst 82:109–122 Teerawanichpan P, Chandrasekharan MB, Jiang Y, Narangajavana J, Hall TC (2004) Characterization of two rice DNA methyltransferase genes and RNAi-mediated reactivation of a silenced transgene on rice callus. Planta 218:337–349 Terada R, Urawa H, Inagaki Y, Tsugane K, Iida S (2002) Efficient gene targeting by homologous recombination in rice. Nat Biotechnol 20:1030–1034 Terada R, Johzuka-Hisatomi Y, Saitoh M, Asao H, Iida S (2007) Gene targeting by homologous recombination as a biotechnological tool for rice functional genomics. Plant Physiol 144:846–856 The Rice Annotation Project (2007) Curated genome annotation of Oryza sativa ssp. japonica and comparative genome analysis with Arabidopsis thaliana. Genome Res 17:175–183 Tinland B, Hohn B (1995) Recombination between prokaryotic and eukaryotic DNA: integration of Agrobacterium tumefaciens T-DNA into the plant genome. Genet Eng 17:209–229 Tsugane K, Maekawa M, Takagi K, et al. (2006) An active DNA transposon nDart causing leaf variegation and mutable dwarfism and its related elements in rice. Plant J 45:46–57 Tzfira T, Citovsky V (2006) Agrobacterium-mediated genetic transformation of plants: biology and biotechnology. Curr Opin Biotechnol 17:147–154 Tzfira T, White C (2005) Towards targeted mutagenesis and gene replacement in plants. Trends Biotechnol 23:567–569 Tzfira T, Li J, Lacroix B, Citovsky V (2004) Agrobacterium T-DNA integration: molecules and models. Trends Genet 20:375–383 Ueguchi-Tanaka M, Ashikari M, Nakajima M, et al. (2005) GIBBERELLIN INSENSITIVE DWARF1 encodes a soluble receptor for gibberellin. Nature 437:693–698 Upadhyaya N (2007) Rice functional genomics – challenges, progress and prospects. Springer, Berlin Heidelberg New York Zhu Q-H, Ramm K, Eamens AL, Dennis ES, Upadhyaya NM (2006) Transgene structures suggest that multiple mechanisms are involved in T-DNA integration in plants. Plant Sci 171:308–322

I.8

T-DNA Tagging Lines Gynheung An1(* ü)

1

Introduction

Recent annotation of rice genome has identified nearly 40,000 genes (Feng et al. 2002; Goff et al. 2002; Sasaki et al. 2002; Yu et al. 2002). To facilitate discovery of their functional roles, several approaches have been developed. Current methods include antisense and RNAi technologies to decrease the expression of target genes, as well as insertional mutagenesis to knock out gene expression (Jeon et al. 2000). Random insertional mutagenesis has been achieved by transferred DNA (T-DNA) or transposons for large-scale analyses. This technique not only is efficient for identifying knockout mutants, but also can be employed for both promoter- and activation-tagging. The establishment of a large number of insertional mutants will accelerate these reverse-genetics approaches for studying rice gene function in this model monocot species. In addition, homologous recombination and viral-induced gene silencing are apparently effective techniques, although their popular usages are still awaited. The Targeting Induced Local Lesions in Genomes (TILLING) method has been adapted in rice in order to find point mutations in genes of interest (Henikoff et al. 2004). In this chapter we will review the progress in generation and characterization of T-DNA insertional lines and identification of gene function using the insertional element in rice.

2

Generation of T-DNA Insertional Mutants

Usefulness of T-DNA insertional mutant populations has been demonstrated in Arabidopsis (Azpiroz-Leehan and Feldmann 1997). The first large-scale collection of T-DNA tagged lines was established by the Arabidopsis knockout facility at the

1

National Research Laboratory of Plant Functional Genomics, Division of Molecular and Life Sciences, Pohang University of Science and Technology (POSTECH), Pohang 790-784, Korea e-mail: [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

95

96

Gynheung An

University of Wisconsin (Krysan et al. 1999). Consequently, several such mutant libraries have been generated and their insertion sites have been analyzed (Sessions et al. 2002; Szabados et al. 2002). More recently, over 225,000 independent T-DNA insertional lines of Arabidopsis were created that represent almost the entire gene space (Alonso et al. 2003). Compared with other insertion elements, T-DNA appears to insert relatively randomly into genome (Kolesnik et al. 2004; Sallaud et al. 2004). Therefore, this mutagen is suitable for achieving near-saturation mutagenesis. However, it is prerequisite that the genome size not be too large and that transformation frequency be high enough to ensure the generation of a large number of transgenic plants. Currently, Arabidopsis and rice are the only plant species that meet such demands. In Arabidopsis, due to the efficient in planta transformation method, obtaining a large number of transformants has been relatively easy. However, that technique has not been routinely applied to other species. In rice, the Agrobacterium-mediated cocultivation method is efficient enough to produce a large quantity of T-DNA insertional mutant plants (Hiei et al. 1994; Lee et al. 1999; Sallaud et al. 2003). For example, we have generated approximately 100,000 fertile lines tagged by T-DNA (Jeon et al. 2000; Jeong et al. 2002, 2006; Ryu et al. 2004). In addition, several groups in China, Taiwan, and Europe have independently produced T-DNA insertional mutant lines (Yin and Wang 2000; Sallaud et al. 2003, 2004; Wu et al. 2003; Sha et al. 2004; Chern et al. 2006; Zhang et al. 2006). Altogether, more than 400,000 T-DNA tags have now been generated in this species. Representative T-DNA tagging populations are presented in Table 1. These populations are large enough to find a knockout in a given gene at more than 95% probability, assuming that T-DNA is randomly inserted into a chromosome. However, the number of tagging lines necessary for saturating all the rice genes may actually be smaller than the estimated value because T-DNA insertions occur preferentially in gene-rich regions (Barakat et al. 2000; An et al. 2003; Wu et al. 2003; Sallaud et al. 2004; Jeong et al. 2006; Zhang et al. 2006).

3

Generation of Activation-tagging Mutants

A large number of rice genes are duplicated. At a minimum of 80% identity over 100 bp, 851 markers among 2,000 rice cDNA markers (41%) are single loci, 509 (24%) have two copies, and the remainder (35%) have three or more (Goff et al. 2002). In addition, there are multiple pathways to transmit the biotic and abiotic signals, and to generate cellular metabolites. Therefore, unless multiple mutants are generated, functional analysis of most of the genes is difficult. It is estimated that fewer than 10% of the genes tagged in Arabidopsis and rice are likely to generate a visible phenotypic change (Feldmann 1991; Jeon and An 2001). The activation-tagging approach is one method that complements the technologies needed for studying genes whose function cannot be resolved by insertional mutagenesis because of gene redundancy, a lethal phenotype in loss-of-function

I.8 T-DNA Tagging Lines

97

Table 1 Inventory of T-DNA insertional mutants in rice Number Character of lines Website Gene trap Activation/gene trap Enhancer trap Enhancer trap

50,000 50,000 40,000 230,000

Activation/gene trap

55,000

http://www.postech.ac.kr/life/pfg http://www.postech.ac.kr/life/pfg http://genoplante.com http://www.ricefgchina.org http://rmd.ncpr.cn http://trim.sinica.edu.tw

Reference Jeon et al. 2000 Jeong et al. 2002 Sallaud et al. 2004 Wu et al. 2003 Zhang et al. 2006 Chern et al. 2006

mutants, or phenotype expression only under specific conditions. Activation-tagging generates gain-of-function mutants and phenotypes different from that of loss-offunction mutants. An alternative to activation-tagging is the use of tissue-specific promoters that mis-express genes only during certain life-cycle phases or in targeted tissues (Weigel et al. 2000). For example, a chemical-inducible activation-tagging system has recently been used to identify genes controlling the vegetative-toembryonic transition (Zuo et al. 2002). A heat shock-tagging system also has been developed to induce transcription of flanking genomic sequences in response to heat treatment (Matsuhara et al. 2000). However, the genomic scale approach toward this system has not been reported. Activation-tagging has been effectively applied in Arabidopsis (Kardailsky et al. 1999; Lee et al. 2000; Weigel et al. 2000; Tani et al. 2004). In the vector, T-DNA carries strong activator elements that enhance the expression of genes adjacent to the insertion site. The cauliflower mosaic virus (CaMV) 35S enhancer or promoter sequences have been most frequently implemented as an activator. It is unclear whether the enhancer acts by quantitatively increasing the original expression patterns or by promoting ectopic or constitutive over-expression of the nearby gene. From the T-DNA activation-tagging pools of Arabidopsis, Weigel et al. (2000) have characterized over 30 dominant mutants with various phenotypes. Analysis of a subset has shown that the tagging vector causes over-expression of the gene immediately adjacent to the inserted enhancer. Activation-tagging has been successfully employed not only to study plant growth and development, but also to discover disease resistance genes (Tani et al. 2004). We have developed a binary vector that carries the tetramerized 35S enhancers within T-DNA. This vector has been used to generate activation-tagging pools for more than 50,000 individual transformants (Jeong et al. 2002, 2006). Examination of randomly selected tags in the intergenic regions has shown that at least half the test lines (59/112) displayed greater expression of the nearby genes (Jeong et al. 2006). In most of the increased lines, patterns after activation were similar to those in the wild type, maintaining their endogenous expression patterns. In some of the lines, the pattern changed, with ectopic expression being most frequently observed in the mature leaves. No good relationship was found between frequency of activation

98

Gynheung An

and distance from the 35S enhancers to the gene. Enhancement was equally observed both upstream and downstream of the tagged genes. Enhanced expression does not always result in phenotypic alteration (Weigel et al. 2000; Jeong et al. 2002). This differs from ectopic expression by strong promoters, e.g., actin or ubiquitin, which cause constitutive over-expression in all tissues. Nonetheless, the dominant mutation frequency of an activation-tagging population is much higher than that of a tag population generated by a simple insertion vector, thereby indicating that the activation-tagging strategy works in rice (Jeong et al. 2002). In fact, characterization of these dominant mutants has led to the identification of a number of genes tagged by the enhancer in rice (An G. unpublished data).

4

Entrapment Tagging

One of the major breakthroughs in using T-DNA as a DNA delivery vector was the development of binary vectors (Bevan 1984; An et al. 1985). Since the proteins needed for T-DNA transfer are encoded by the virulence genes located outside of T-DNA, they can be provided by a separate vector located within the same cell. Most of the sequences present in the native T-DNA can be replaced with foreign genes without altering the transfer ability. In some of the insertional vectors, T-DNA carries a promoterless reporter gene next to the T-DNA border (Pang et al. 1996; Jeon et al. 2000; Ryu et al. 2004). When these gene trap vectors are inserted into an exon, they can produce a fusion transcript between the endogenous gene and that reporter (Springer 2000). The presence of multiple splice sites aligned in all reading frames preceding the reporter gene allows expression of the reporter if insertion occurs in an intron (Sundaresan et al. 1995). This allows splicing from the donor sites in the disrupted gene to the acceptor sites in the reporter gene, resulting in fusion of the upstream exon sequences to the reporter regardless of insert position. Expression patterns of the tagged genes can be monitored by assaying reporter-gene activity. Since this feature allows us to study gene expressions at tissue and cellular levels, characterization of the tagged gene can be facilitated. However, the multiple splice sequences are not always efficiently functional. The b-glucuronidase (GUS) gene has frequently served as a reporter because of the easy detection of its gene product and the tolerance for N-terminal translational fusions in its enzyme activity (Jefferson et al. 1987). Various genes exhibiting tissue- or organ-preferential expression patterns have been identified using the reporter (Jeon et al. 2000). Although this reporter is valuable in generating T-DNA tagging populations, there are some drawbacks to this method. Samples are destroyed during the assay and the substrate is not equally penetrated by all cell types. Therefore, non-destructive reporter genes, such as green fluorescent protein (GFP) or luciferase, have been used for entrapment in plants. Tagging populations with the GFP gene have now been established for Arabidopsis and rice (Haseloff et al. 1997; Ryu et al. 2004; Johnson et al. 2005). Fluorescence microscopy and confocal microscopy are used for monitoring expression of the GFP reporter protein

I.8 T-DNA Tagging Lines

99

at cellular level. One of the main problems in using the GFP reporter is the detection of autofluorescence emitted from mesophyll. With its rapid turnover and half-life of 3 h, luciferase is well suited as a real-time reporter for in planta geneexpression studies (Thompson et al. 1991). However, these non-destructive reporter systems are not as sensitive as GUS because their activity is not cumulative. In addition, the luciferase assay requires a substrate, making it difficult to conduct in situ assays. As many as 30% of the tagged lines show activation of the reporter gene in Arabidopsis (Sundaresan 1996). In rice, at least 10% of the tagged lines are GUSpositive in the various organs (Jeon et al. 2000). Some tags are tissue- or organ-specific, while others are ubiquitous in all examined organs. If one includes reporter-gene activation caused by certain environmental conditions or chemicals, such as growth substances, total tagging efficiency is higher. Whereas the gene-trap vectors are useful in generating a fusion between the tagged gene and a reporter, enhancer-trap vectors are valuable tools in identifying regulatory elements. In the enhancer-trap system, the reporter gene is fused to a minimal 35S promoter, which is activated when T-DNA is inserted near an enhancer element. Enhancer trapping yields a higher frequency of reporter-gene activation. Enhancer-trap systems have been developed for rice (Greco et al. 2003; Wu et al. 2003). A modified enhancer system using the yeast transcription factor GAL4 has been applied during the construction of enhancer-trap lines in rice (Wu et al. 2003; Yang et al. 2004; Johnson et al. 2005). Enhancer-trapping frequency using the GAL4 system in rice ranged from 29% with GFP (Johnson et al. 2005) to 60–70% with GUS (Wu et al. 2003; Yang et al. 2004).

5

DNA Pool Screening for Reverse Genetics

If an insertional mutant is present in the flanking sequence tag (FST) database (see below), one could order the mutant from a contributor and utilize for functional analyses of the gene. However, if a mutant is not present in the database, screening mutant population by PCR could result in identification of a knockout mutant. The screening is also valuable when an additional allele is needed. A PCR-based screening method has been derived for identifying insertional mutants in Arabidopsis (Krysan et al. 1999). Pools of DNA were prepared from insertional mutant lines and they were used as templates to screen knockout mutants using a set of primers: a gene-specific primer and a primer located near the end of the insert. PCR products are run on an agarose gel, blotted onto a membrane, and hybridized with a gene-specific probe. Once a positive band is identified, DNAs from corresponding subpools and individuals are sequentially screened for identification of a line with the T-DNA insert. The DNA fragment flanking the insert element is then amplified and its sequence determined for verification. Because this approach is labor-intensive and time-consuming, DNA pools for a large number of lines are commonly used with this reverse-genetics tool.

100

Gynheung An

In Arabidopsis, pools of 1,000–5,000 lines have been efficiently utilized for identifying insertional mutations (Krysan et al. 1999). Because the rice genome is about four times larger than that of Arabidopsis, a pool of 1,000 lines is considered sufficient for PCR-based screening of its knockout mutations (Lee et al. 2003). Rice also has a higher amount of GC than does Arabidopsis, especially at the 5′ ends of its genes (Yu et al. 2002). These high-GC regions are difficult to amplify under normal PCR conditions. Therefore, use of a GC buffer and adjustment of the annealing temperature can improve the screening efficiency of such GC-rich genes (Lee et al. 2003). Nevertheless, several primers must be examined when analyzing large genes because PCR efficiency is quite low when the product is > 1 kb. Tos17 is an endogenous retroposon that becomes unstable during tissue culture. Therefore, an average of four new copies were generated in the rice genome of the T-DNA insertional mutant population. Since Tos17 also causes insertional mutation, one could look for Tos17 insertion in the gene-of-interest if DNA pools have already been established. This DNA pool screening strategy has been successfully utilized in Arabidopsis, petunia, and maize (Koes et al. 1995; Mena et al. 1996; Krysan et al. 1999; Parinov and Sundaresan 2000). For example, 17 insertions in 63 genes involved in signal transduction and ion transport, 47 insertions in 36 members of the R2R3 MYB gene family, and 22 mutations in 70 P450 genes have been isolated from Arabidopsis (Krysan et al. 1999). In petunia, a DNA pool has been generated from 1,000 individual plants with highly active dTph1 elements; PCR-screening has resulted in the identification of dTph1 insertions for 10 genes (Koes et al. 1995). Reverse geneticsscreening has also led to the isolation of a transposon-induced mutation in ZAG1, the maize homologue of AGAMOUS (Mena et al. 1996). In rice, PCR-screening for 12 MADS box genes of DNA pools prepared from 21,049 T-DNA tagged lines has enabled the identification of five insertions in four genes (Lee et al. 2003). The DNA pool size has been increased to 60,000 lines and the success rate is approximately 50%.

6

Establishment of Flanking Sequence Tag (FST) Databases

Establishment of databases for the insertion sites should facilitate the use of T-DNA tag lines. FSTs that flank insert elements have been obtained by various methods, such as thermal asymmetric interlaced PCR (TAIL PCR), inverse PCR (iPCR), or adaptor-ligation PCR. Although this strategy requires considerable effort and time, the FST databases can be easily shared with other scientists, facilitating the distribution of mutant materials and the analysis of gene functioning (Parinov et al. 1999; Tissier et al. 1999; Parinov and Sundaresan 2000; An et al. 2003; Jeong et al. 2006). With sequencing of their entire genomes nearly complete, FST databases for rice and Arabidopsis become the most powerful tool for systematically analyzing the functions of a large number

I.8 T-DNA Tagging Lines

101

of their genes (Parinov and Sundaresan 2000; Walbot 2000; Kumar and Hirochika 2001; Pan et al. 2003). Databases for T-DNA and Ds transposon insertion-site sequences have already been developed for Arabidopsis (Parinov et al. 1999; Tissier et al. 1999; Ortega et al. 2002; Sessions et al. 2002). In maize, the DNAs adjacent to transposed Ac elements have also been sequenced (Cowperthwaite et al. 2002). In rice, FST databases for T-DNA insertional lines have also been created. While we have reported the establishment of a database of more than 30,000 FST (An et al. 2003; Ryu et al. 2004; Jeong et al. 2006), others have described one comprising 1,009 tag end sequences (TES) (www.genomics.zju.edu.cn/ricetdna). Sallaud and colleagues have analyzed 7,480 TES from T-DNA tagging lines (Sallaud et al. 2004) and Zhang et al. (2006) reported establishment of 13,738 FST. The database is rapidly growing, and its public site currently contains more than 100,000 independent FST data generated from the T-DNA insertional lines (http:// signal.salk.edu/cgi-bin/RiceGE) (Table 2). Together with the data from Tos17 and transposable elements, current publicly available FST databases hit 25,103 genes, more than half the annotated genes. FST analyses also provide information on the distribution of insert elements in plant chromosomes. This information is useful in estimating how many additional insertional mutants are needed to saturate the mutagenesis and what method is more efficient in achieving the goal. Compared to transposons, T-DNA is less prone to hot spots (Miyao et al. 2003; Kolesnik et al. 2004; Sallaud et al. 2004). T-DNA insertion density is higher in four chromosomes (1, 2, 3, and 6) but lower in three chromosomes (9, 10, and 12) when compared with that generated by transposons (Sallaud et al. 2004). Overall GC content at the T-DNA insertion sites is close to that measured for the entire rice genome (An et al. 2003; Jeong et al. 2006), whereas Tos17 target regions have a narrow pattern of GC-content distribution (Miyao et al. 2003). T-DNAs are inserted into genic regions more frequently than into intergenic regions in rice genome. Especially, T-DNA prefers the promoter region near the start codon (An et al. 2003; Sallaud et al. 2004). This observation is different from Arabidopsis, in which T-DNA inserts more frequently in the intergenic regions (Sessions et al. 2002; Alonso et al. 2003). On the other hand, T-DNA insertions are almost equally distributed between exons and introns (An et al. 2003; Sallaud et al. 2004), whereas Ds appears to prefer introns. When functional classification of genes tagged by T-DNA is performed, T-DNA insertion has not been found to be biased toward a particular type of gene, whereas Tos17 insertions do occur preferentially in certain families such as “kinase-” and “disease-resistance” classes (An et al. 2003; Miyao et al. 2003). Although T-DNA and transposons differ in several ways, they have common advantages: (1) genome-wide distribution, (2) preferred insertion into gene-rich regions, and (3) low frequencies of integration within repetitive or pericentromeric regions. Along with possessing the most random distribution patterns, these advantages make T-DNA an ideal insertional mutagen for rice functional genomics.

PFG T-DNA

80,861 7,546 7,272 4,662 6,886 19,354

The Salk FST database

Mapped Exon Intron 5′ UTR Promoter Unique gene hits

17,934 2,097 1,949 417 684 3,697

RTIM Tos17 15,668 1,420 1,612 747 1,397 4,714

RMD T-DNA 6,959 978 879 330 561 2,640

TRIM T-DNA 7,173 517 640 343 567 2,001

707 58 68 15 38 179

Genoplante T- ZJ DNA T-DNA 588 156 76 40 53 312

CSIRO Ac/Ds

6,766 726 508 232 372 1,676

UC Davis Ac/Ds

1,045 231 147 70 92 516

137,701 11,535 10,504 6,463 9,578 25,103

GSNU Ds Total

Table 2 FST database in Salk. The Salk FST database (http://signal.salk.edu/cgi-bin/RiceGE) was updated on September 22, 2006. The 5′ UTR is the region between the start ATG codon and upstream 300 bp. The promoter is the region between −300 and −1000 bp from the start ATG. PFG, Postech T-DNA, Korea; RTIM, NIAS Tos17 insertion mutants, Japan; RMD, Rice Mutant Data, Huazhong Agricultural University, China; TRIM, Taiwan Rice Insertion Mutant, Academia Sinica, Taiwan; Genoplante oryza tag lines, France; ZJ, Zhejiang University, China; CSIRO Ac/Ds, Australia; UC Davis Ac/Ds, USA; GSNU Ds, Gyeongsang National University, Korea

102 Gynheung An

I.8 T-DNA Tagging Lines

103

7 Obstacles in T-DNA Insertional Mutagenesis One of the major difficulties in deducing gene function using the T-DNA insertional lines is the mutant phenotypes that are not necessarily due to T-DNA insertion. This is because other elements, such as endogenous transposons, are mobilized during the generation of T-DNA insertional mutants. Tissue-culturing also causes point mutations as well as small deletions and insertions. Therefore, it is necessary to confirm those observations by analyzing additional lines that contain a mutation in the target gene. Large databases for knockout mutants, not only from T-DNA but also from transposons, are extremely important for identifying multiple alleles in a given gene. DNA pools are also valuable in enriching alleles. If an allele is not present, antisense or RNAi approaches are useful. However, these alternative methods may affect the expression of genes that are structurally similar to the target gene. In addition, a small amount of expression persists in antisense or RNAi plants, a level that is often sufficient to produce a normal phenotype. In these cases, complementation of the mutant with wild-type cDNA or a genomic clone can prove whether the mutant phenotype is indeed due to the T-DNA insertion. Another obstacle in the reverse-genetics approach is gene duplication, in which redundancy may result in no phenotypic alteration. Researchers with the rice genome project have predicted that a large number of its genes are duplicated or polyploidized (Goff et al. 2002; Yu et al. 2002). Therefore, double or multiple mutations in a group of related genes are necessary for observing mutant phenotypes. To this end, numerous tagging populations and FST databases have been generated, and websites for those databases are being constructed (Hirochika et al. 2004; An et al. 2005a, 2005b). These databases will facilitate the identification of individual mutant lines of related genes. The maintenance and distribution of mutant seeds are also major problems. Primary mutant plants normally generate a small number of seeds. Therefore, seeds from the primary transgenic plants should be propagated before distribution. Because T-DNA tagging lines contain transgenic elements, only limited space is available for seed propagation. The resultant seed must then be stored in a controlled environment of low temperature and humidity because viability is rapidly reduced under warm, humid conditions. To ensure the safety of the seed during emergency situations, collections should be kept in multiple locations. Ideally, in addition to holding the original mutant stock in developed laboratories, a second copy should be maintained by a well-established international organization, such as the International Rice Research Institute.

8

Future Prospects

Since the number of insertional mutant lines is large enough to cover most of the rice genes, current effort is focused on determination of their FST. It is speculated that about 5 years are needed to determine the franking sequences of the currently

104

Gynheung An

established lines. If this process can be accelerated, elucidation of rice gene function will be facilitated. It is also valuable to develop a database for phenotypic alterations in mutant lines. The evaluation of tagging lines should include the assessment of traits for any visible phenotypes, such as abnormalities in morphology, growth rate, plant color, flowering time, and fertility. Genotyping each tagging line is also needed. Since most of the laboratories do not have a sufficient facility to grow rice to fertile plants, providing homozygotic mutants will save the time and effort in characterizing gene function. The resource can also be used for the generation of multiple mutants of closely related genes when a single insertion does not reveal the gene function. Acknowledgements The gene tagging project for functional genomics of rice in the An laboratory is supported, in part, by grants from the Crop Functional Genomic Center, the 21st Century Frontier Program (CG-1111), the Biogreen 21 Program, Rural Development Administration, and the National Research Laboratory Program.

References Alonso JM, Stepanova AN, Leisse TJ, et al. (2003) Genome-wide insertional mutagenesis of Arabidopsis thaliana. Science 301:653–657 An G, Watson BD, Stachel S, Gordon MP, Nester EW (1985) New cloning vehicles for transformation of higher plants. EMBO J 4:277–288 An G, Jeong D-H, Jung K-H, Lee S (2005a) Reverse genetic approaches for functional genomics of rice. Plant Mol Biol 59:111–123 An G, Lee S, Kim S-H, Kim S-R (2005b) Molecular genetics using T-DNA in rice. Plant Cell Physiol 46:14–22 An S, Park S, Jeong D-H, et al. (2003) Generation and analysis of end sequence database for T-DNA tagging lines in rice. Plant Physiol 133:2040–2047 Azpiroz-Leehan R, Feldmann KA (1997) T-DNA insertion mutagenesis in Arabidopsis: going back and forth. Trends Genet 13:152–156 Barakat A, Gallois P, Raynal M, et al. (2000) The distribution of T-DNA in the genomes of transgenic Arabidopsis and rice. FEBS Lett 471:161–164 Bevan MW (1984) Binary Agrobacterium vectors for plant transformation. Nucleic Acid Res 12:8711–8721 Chern C-G, Fan M-J, Yu S-M, et al. (2006) In: Proc 4th Int Rice Functional Genomics Symp, 25 October 2003, ManilaCowperthwaite M, Park W, Xu Z, Yan X, Maurais SC, Dooner HK (2002) Use of the transposon Ac as a gene-searching engine in the maize genome. Plant Cell 14:713–726 Feng Q, Zhang Y, Hao P, et al. (2002) Sequence and analysis of rice chromosome 4. Nature 420:316–320 Goff SA, Ricke D, Lan TH, et al. (2002) A draft sequence of the rice genome (Oryza sativa L. ssp. japonica). Science 296:92–100 Greco R, Ouwerkerk PB, de Kam RJ, et al. (2003) Transpositional behaviour of an Ac/Ds system for reverse genetics in rice. Theor Appl Genet 108:10–24 Haseloff J, Siemering KR, Prasher DC, Hodge S (1997) Removal of a cryptic intron and subcellular localization of green fluorescent protein are required to mark transgenic Arabidopsis plants brightly. Proc Natl Acad Sci USA 94:2122–2127 Henikoff S, Till BJ, Comai L (2004) TILLING. Traditional mutagenesis meets functional genomics. Plant Physiol 135:630–636

I.8 T-DNA Tagging Lines

105

Hiei Y, Ohta S, Komari T, Kumashiro T (1994) Efficient transformation of rice (Oryza sativa L.) mediated by Agrobacterium and sequence analysis of the boundaries of the T-DNA. Plant J 6:271–282 Hirochika H, Guiderdoni E, An G, et al. (2004) Rice mutant resources for gene discovery. Plant Mol Biol 54:325–334 Jefferson RA, Kavanagh TA, Bevan MW (1987) GUS fusions: Beta-glucuronidase as a sensitive and versatile gene fusion marker in higher plants. EMBO J 6:3901–3907 Jeon JS, An G (2001) Gene tagging in rice: a high throughput system for functional genomics. Plant Sci 161:211–219 Jeon JS, Lee S, Jung KH, et al. (2000) T-DNA insertional mutagenesis for functional genomics in rice. Plant J 22:561–570 Jeong D-H, An S, Kang H-G, et al. (2002) T-DNA insertional mutagenesis for activation tagging in rice. Plant Physiol 130:1636–1644 Jeong D-H, An S, Park S, et al. (2006) Generation of a flanking sequence-tag database for activationtagging lines in japonica rice. Plant J 45:123–132 Johnson AA, Hibberd JM, Gay G, et al. (2005) Spatial control of transgene expression in rice using the GAL4 enhancer trapping system. Plant J 41:779–789 Kardailsky I, Shukla VK, Ahn JH, et al. (1999) Activation tagging of the floral inducer FT. Science 286:1962–1965 Koes R, Souer E, van Houwelingen A, et al. (1995) Targeted gene inactivation in petunia by PCRbased selection of transposon insertion mutants. Proc Natl Acad Sci USA 92:8149–8153 Kolesnik T, Szeverenyi I, Bachmann D, et al. (2004) Establishing an efficient Ac/Ds tagging system in rice: large-scale analysis of Ds flanking sequences. Plant J 37:301–314 Krysan PJ, Young JC, Sussman MR (1999) T-DNA as an insertional mutagen in Arabidopsis. Plant Cell 11:2283–2290 Kumar A, Hirochika H (2001) Applications of retrotransposons as genetic tools in plant biology. Trends Plant Sci 6:127–134 Lee S, Jeon JS, Jung KH, An G (1999) Binary vectors for efficient transformation of rice. J Plant Biol 42:310–316 Lee S, Kim J, Son JS, et al. (2003) Systematic reverse genetic screening of T-DNA tagged genes in rice for functional genomic analyses: MADS-box genes as a test case. Plant Cell Physiol 244:1403–1411 Matsuhara S, Jingu F, Takahashi T, Komeda Y (2000) Heat-shock tagging: a simple method for expression and isolation of plant genome DNA flanked by T-DNA insertions. Plant J 22:79–86 Mena M, Ambrose BA, Meeley RB, Briggs SP, Yanofsky MF, Schmidt RJ (1996) Diversification of C-function activity in maize flower development. Science 274:1537–1540 Miyao A, Tanaka K, Murata K, et al. (2003) Target site specificity of the Tos17 retrotransposon shows a preference for insertion within genes and against insertion in retrotransposon-rich regions of the genome. Plant Cell 15:1771–1780 Ortega D, Raynal M, Laudie M, et al. (2002) Flanking sequence tags in Arabidopsis thaliana T-DNA insertion lines: a pilot study. C R Biol 325:773–780 Pan X, Liu H, Clarke J, Jones J, Bevan M, Stein L (2003) ATIDB: Arabidopsis thaliana insertion database. Nucleic Acids Res 31:1245–1251 Pang SZ, DeBoer DL, Wan Y, et al. (1996) An improved green fluorescent protein gene as a vital marker in plants. Plant Physiol 112:893–900 Parinov S, Sundaresan V (2000) Functional genomics in Arabidopsis: large-scale insertional mutagenesis complements the genome sequencing project. Curr Opin Biotechnol 11:157–161 Parinov S, Sevugan M, De Y, Yang WC, Kumaran M, Sundaresan V (1999) Analysis of flanking sequences from dissociation insertion lines: a database for reverse genetics in Arabidopsis. Plant Cell 11:2263–2270 Ryu C-H, You J-H, Kang H-G, et al. (2004) Generation of T-DNA tagging lines using a bidirectional gene trap vector and the establishment of an insertion-site database. Plant Mol Biol 54:489–502

106

Gynheung An

Sallaud C, Meynard D, van Boxtel J, et al. (2003) Highly efficient production and characterization of T-DNA plants for rice (Oryza sativa L.) functional genomics. Theor Appl Genet 106:1396–1408 Sallaud C, Gay C, Larmande P, et al. (2004) High throughput T-DNA insertion mutagenesis in rice: a first step towards in silico reverse genetics. Plant J 39:450–464 Sasaki T, Matsumoto T, Yamamoto K, et al. (2002) The genome sequence and structure of rice chromosome 1. Nature 420:312–316 Sessions A, Burke E, Presting G, et al. (2002) A high-throughput Arabidopsis reverse genetics system. Plant Cell 14:2985–2994 Sha Y, Li S, Pei Z, Luo L, Tian Y, He C (2004) Generation and flanking sequence analysis of a rice T-DNA tagged population. Theor Appl Genet 108:306–314 Springer PS (2000) Gene traps: tools for plant development and genomics. Plant Cell 12:1007–1020 Sundaresan V (1996) Horizontal spread of transposon mutagenesis: new uses of old elements. Trends Plant Sci 1:184–191 Sundaresan V, Springer P, Volpe T, et al. (1995) Patterns of gene action in plant development revealed by enhancer trap transposable element. Genes Dev 9:1797–1810 Szabados L, Kovacs I, Oberschall A, et al. (2002) Distribution of 1000 sequenced T-DNA tags in the Arabidopsis genome. Plant J 32:233–242 Tani H, Chen X, Nurmberg P, et al. (2004) Activation tagging in plants: a tool for gene discovery. Funct Integr Genomics 4:258–266 Thompson JF, Hayes LS, Lloyd DB (1991) Modulation of firefly luciferase stability and impact on studies of gene regulation. Gene 103:171–177 Tissier AF, Marillonnet S, Klimyuk V, et al. (1999) Multiple independent defective suppressormutator transposon insertions in Arabidopsis: a tool for functional genomics. Plant Cell 11:1841–1852 Walbot V (2000) Saturation mutagenesis using maize transposons. Curr Opin Plant Biol 3:103–107 Weigel D, Ahn JH, Blazquez MA, et al. (2000) Activation tagging in Arabidopsis. Plant Physiol 122:1003–1013 Wu C, Li X, Yuan W, et al. (2003) Development of enhancer trap lines for functional analysis of the rice genome. Plant J 35:418–427 Yang Y, Peng H, Huang H, et al. (2004) Large-scale production of enhancer trapping lines for rice functional genomics. Plant Sci 167:281–288 Yin Z, Wang G-L (2000) Evidence of multiple complex patterns of T-DNA integration into rice genome. Theor Appl Genet 100:461–470 Yu J, Hu S, Wang J, et al. (2002) A draft sequence of the rice genome (Oryza sativa L. ssp. indica). Science 296:79–92 Zhang J, Li C, Wu C (2006) RMD: a rice mutant database for functional analysis of the rice genome. Nucl Acids Res 34:D741–D744 Zuo J, Niu QW, Frugis G, Chua NH (2002) The WUSCHEL gene promotes vegetative-to-embryonic transition in Arabidopsis. Plant J 30:349–359

I.9

Frequent DNA Transfer Among Mitochondrial, Plastid and Nuclear Genomes of Rice During Evolution Mikio Nakazono1 and Atsushi Hirai2(* ü)

1

Introduction

Mitochondria and plastids are semiautonomous organelles that contain their own DNA. During evolution, it is thought that most of the genes of ancestral mitochondria and plastids, which appear to have originated from endosymbionts such as a proteobacterium and a cyanobacterium, respectively, have been transferred to the nuclear genome (as reviewed by Timmis et al. 2004). In addition to the transfer of functional genes from mitochondria and plastids to the nucleus, non-functional organelle-derived DNA segments are also found in the nuclear genome. Although most of the genes that function in mitochondria and plastids in higher plants have been transferred to the nuclear genome, the mitochondrial genome still contains 50–60 genes (Sugiyama et al. 2005) and the plastid genome still contains 100–150 genes (Sugiura and Takeda 2000). The mitochondrial genomes of animals, plants, fungi and unicellular eukaryotes show considerable variations in size, structure and organization. In vertebrates the mitochondrial genomes are single circles of about 16 kilo-basepairs (kbp), whereas in fungi the mitochondrial genomes vary in size from 18.9 to 176.3 kbp. In higher plants, the mitochondrial genome is much larger and more complicated than in animals and fungi. The genome varies from 200 kbp in Brassica to 2,500 kbp in muskmelon. The mitochondrial genomes of higher plants contain several recombination repeats (reviewed by Lonsdale et al. 1988). Recombination within the master chromosome that contains the entire complement of genetic information gives rise to a number of subgenomic molecules (Lonsdale et al. 1984). So far, the complete mitochondrial genomes have been sequenced in liverwort (Marchantia polymorpha; Oda et al. 1992),

1 Laboratory of Plant Molecular Genetics, Graduate School of Agricultural and Life Sciences, University of Tokyo, 1-1-1 Yayoi, Bunkyo, Tokyo 113–8657, Japan e-mail: [email protected] 2

School of Agriculture, Meijo University, 1–501 Shiogamaguchi, Tenpaku, Nagoya, Aichi 468–8502, Japan e-mail: [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

107

108

M. Nakazono and A. Hirai

Arabidopsis (Arabidopsis thaliana; Unseld et al. 1997), sugar beet (Beta vulgaris; Kubo et al. 2000), rice (Oryza sativa; Notsu et al. 2002; Tian et al. 2006), rapeseed (Brassica napus; Handa 2003), maize (Zea mays; Clifton et al. 2004), tobacco (Nicotiana tabaccum; Sugiyama et al. 2005) and wheat (Triticum aestivum; Ogihara et al. 2005). The gene contents and gene order vary widely among these mitochondrial genomes. The structure of the plastid genome is conserved among higher plants (Sugiura 1992). In general, the plastid genome consists of four segments: a large region of single copy genes (LSC), a small region of single copy genes (SSC), and two copies of an inverted repeat (IRA and IRB). Moreover, the sizes (110–160 kbp) of the plastid genome of higher plants are similar to each other (Sugiura 1992). In 1986, complete nucleotide sequences of the plastid genome of liverwort (Ohyama et al. 1986) and tobacco (Shinozaki et al. 1986) were determined, and in 1989 complete nucleotide sequence of the rice plastid genome was determined (Hiratsuka et al. 1989). So far, the complete plastid genome sequences of more than 40 species including Arabidopsis have been determined. Sequencing of the whole nuclear genome of rice was completed in 2005 (International Rice Genome Sequencing Project 2005), 5 years after completion of sequencing of the whole nuclear genome of Arabidopsis (The Arabidopsis Genome Initiative 2000). To date, the complete genome sequencing of three organelles (nucleus, mitochondrion and plastid) in two plant species (rice and Arabidopsis) has made it possible to investigate the transfer of DNA between organelles during evolution. The following sections summarize the evidence for inter-organellar DNA transfer events in mainly rice (Fig. 1). Detailed reviews of genome structures

nucleus

plastid

mitochondrion

Fig. 1 Schematic representation of the inter-organellar DNA transfer events in higher plants during evolution. Arrows indicate possible directions of DNA transfer

I.9 Frequent DNA Transfer Among Mitochondrial

109

and regulation of gene expressions in mitochondria and plastids in higher plants are given in other articles (Sugiura 1992; Sugiura and Takeda 2000; Binder and Brennicke 2003; Knoop 2004).

2

Plastid-to-Mitochondrion DNA Transfer Events

A specific feature of the mitochondrial genomes of higher plants is the presence of many regions homologous to sequences in plastid DNA (ptDNA). Similar sequences present in more than one subcellular genome have been termed examples of “promiscuous DNA”. The existence of promiscuous DNA in plant mitochondrial DNA (mtDNA) was first reported in maize. A 12-kbp portion of the inverted repeat of the ptDNA, encompassing genes for 16S rRNA, tRNAIle and tRNAVal, was identified in maize mtDNA (Stern and Lonsdale 1982). To date, a number of plastidderived sequences have been identified in several plant mtDNAs (reviewed by Schuster and Brennicke 1988). However, different ptDNA sequences are present at various locations in the mitochondrial genomes of different species. The sequences common to plastids and mitochondria are more than 90% homologous to one another. In contrast to the cases of higher plants, no plastid-derived sequences have been detected in the mitochondrial genomes of lower plants [e.g. liverwort (Oda et al. 1992) and green alga (Chlamydomonas reinhardtii; Gray and Boer 1988)], suggesting that DNA fragments were transferred only recently from the plastid to the mitochondrion during the evolution of plants. The transferred fragments contain many genes, but, in most cases, the genes are truncated and have been mutated by frame shifts or base substitutions. Therefore, most of the plastid genes are non-functional in the mitochondrion. However, it seems likely that several plastid genes for tRNAs are transcribed and used for the biosynthesis of proteins in plant mitochondria (Joyce and Gray 1989; Miyata et al. 1998). The first comprehensive survey of plastid-derived sequences in plant mtDNAs was conducted using the rice mtDNA clone bank (Nakazono and Hirai 1993). The mitochondrial fragments that were homologous to plastid DNA were detected by Southern hybridization and sequenced. Sixteen plastid-derived sequences were found to be dispersed throughout the rice mitochondrial genome (Nakazono and Hirai 1993). The subsequent complete sequencing of rice mtDNA revealed the presence of 17 fragments [i.e. the previously identified 16 fragments (Nakazono and Hirai 1993) plus one more fragment] of plastid-like sequences, ranging from 32 to 6,653 bp (Notsu et al. 2002). The total length of these sequences is equal to 6.3% (22,593 bp) of the rice mitochondrial genome and to 16.8% of the rice plastid genome (Fig. 2). The Arabidopsis mitochondrial genome was found to have plastidderived sequences ranging in size from 30 to 930 bp (Unseld et al. 1997). The sequences that were transferred from the plastid genome to the mitochondrial genome in rice were not the same as those in Arabidopsis. Thus, the transfers of segments of ptDNA seem to have occurred frequently from time to time. The mitochondrial genome appears to have been rearranged after the transfer of plastid

110

M. Nakazono and A. Hirai

rrn26 atpA rrn18

rice mtDNA 490,520 bp

rrn26 rice ptDNA 134,525 bp atpA

IR

IR

coxI nad3 rps12 trnfM

cob coxII

rrn18 atp9

atp6

Fig. 2 Plastid-derived DNA sequences in the rice mitochondrial genome. The inner circle is a physical map of PstI fragments of the rice plastid genome (Hirai et al. 1985). Thick lines indicate the inverted repeat sequence (IR). The outer circle is a predicted master circle of the rice mitochondrial genome (Iwahashi et al. 1992). Repeated sequences are indicated by boxes outside the master circle. Sequences homologous to ptDNA are indicated by colored triangles. Homologous sequences in the mitochondrial and plastid genomes are indicated by the same color

sequences as a result of recombination at those sequences (Nakazono and Hirai 1993; Watanabe et al. 1994). All the plastid-derived protein coding genes have many nucleotide alterations and none of them showed evidence of RNA editing (Zeltz et al. 1996; Notsu et al. 2002). Thus, these genes seem to have no functions in rice mitochondria. In some cases, however, a chloroplast-derived sequence can be used as a source of promoter sequences for a mitochondrial gene (e.g. rice nad9) (Nakazono et al. 1996). Moreover, several plastid-derived tRNA genes are transcribed and are most likely functional in rice mitochondria (Miyata et al. 1998).

3

Mitochondrion-to-Nucleus or Plastid-to-Nucleus DNA Transfer Events

The DNA transfer from the mitochondrial genome or the plastid genome to the nuclear genome seems to be an on-going event in higher plants. The transferred genes can be classified as functional genes and non-functional genes (i.e. pseudogenes).

I.9 Frequent DNA Transfer Among Mitochondrial

111

In order to become functional in the nucleus, the transferred genes must acquire regulatory elements such as a promoter sequence, terminator sequence and mitochondrion- or plastid-targeting signal. One way to acquire regulatory elements is to use pre-existing genes for mitochondrial or plastid proteins in the nuclear genome. In rice, the gene encoding mitochondrial ribosomal protein S11 (rps11) was transferred from the mitochondrion to the nucleus and subsequently became two copies (rps11-1 and rps11-2) by duplication in the nuclear genome during evolution (Kadowaki et al. 1996). One copy (rps11-1) acquired the mitochondrialtargeting signal from atpB, which encodes F1-ATP synthase β-subunit, and another copy (rps11-2) acquired the mitochondrial-targeting signal from coxVb, which encodes cytochrome c oxidase subunit Vb (Kadowaki et al. 1996). In rice and maize, mitochondrial rps14 was inserted into an intron of sdhB (also designated sdh2), which encodes mitochondrial succinate dehydrogenase subunit B in the nuclear genome (Figueroa et al. 1999; Kubo et al. 1999). The gene is co-transcribed with the sdhB gene and then rps14 mRNA and sdhB mRNA, both of which possess the same mitochondrial-targeting signal, are produced by alternative splicing (Fig. 3A; Figueroa et al. 1999; Kubo et al. 1999). It is likely that the transfer of rps14 to the nuclear genome occurred before divergence of rice and maize. In contrast,

rps14

A

B prototype rpl27

sdhB

promoter ancient & present chr. 4

TS Osspt16 Integration of rps14 into the intron of sdhB

Interchromosome duplication ancient chr. 8

TS Osspt16 alternative splicing

TS

Intrachromosome duplication

TS mRNA

mRNA

rpl27 present chr. 8

Fig. 3 A Example of an acquisition of a mitochondrial-targeting signal for the gene transferred from a mitochondrion to the nucleus. In rice, rps14 was integrated into the intron of sdhB in the nuclear genome during evolution. The rps14 gene is co-transcribed with the sdhB gene, and then rps14 mRNA and sdhB mRNA, which possess the same mitochondrial-targeting signal (TS), are produced by alternative splicing. B Example of an acquisition of a promoter sequence for the gene transferred from a mitochondrion to the nucleus. The rpl27 gene, which was transferred from mitochondrion to nucleus during evolution, was inserted upstream of Osspt16 on chromosome 4 of rice. The fragment containing the prototype of rpl27 and a part of Osspt16 was duplicated to chromosome 8 by interchromosomal duplication. Subsequently, the fragment containing the prototype of rpl27 and a promoter sequence of Osspt16 was tandem duplicated by intrachromosomal duplication. In this way, it was speculated that rice rpl27 acquired a promoter sequence in the nuclear genome during evolution

112

M. Nakazono and A. Hirai

Arabidopsis mitochondrial rps14, which is also located in the nuclear genome, does not have the features observed in rice and maize (Figueroa et al. 1999), suggesting that the transfer of rps14 occurred independently in Arabidopsis, rice and maize during evolution. Similarly, carrot rps10 and cotton sdh3 (also designated sdhC) acquired mitochondrial-targeting signals from nuclear-encoded mitochondrial hsp22 (Adams et al. 2000, 2001), and Arabidopsis sdh3 acquired a mitochondrial-targeting signal from nuclear-encoded mitochondrial hsp70 (Adams et al. 2001). Plastid-targeting signals may have been acquired in the same way that mitochondrial-targeting signals were acquired. The rice nuclear-encoded plastid genes rps9 and rpl12 are examples of genes that acquired plastid-targeting signals (Arimura et al. 1999). Another example is the rice orf160 gene, which encodes a protein of 160 amino acid residues (Ueda et al. 2006b). The part of the gene encoding the 60 C-terminal amino acid residues is derived from the mitochondrial rpl13 gene. The part encoding the 100 N-terminal amino acid residues is derived from another duplicated fragment of a pentatricopeptide repeat 564 ( ppr564) gene, although it has a different reading frame. As a result of the frameshift, the N-terminal sequence became a plastid-targeting signal (i.e. a transit peptide). These results indicate that the plastid-targeting signal was generated through duplication and subsequent frameshifting of a reading frame of a preexisting protein gene (Ueda et al. 2006b). On the other hand, rice rpl27, which is absent from the mitochondrial genome but is encoded in the nuclear genome, acquired a promoter sequence via inter- and intrachromosomal duplications from the rice spt16-related (Osspt16) gene (Fig. 3B; Ueda et al. 2006a). Hence, the promoter shuffling and the exon shuffling through inter- and intrachromosomal recombination or duplication events in the nuclear genome may be important for an acquisition of regulatory elements such as a promoter sequence and a mitochondrion- or plastidtargeting signal for the genes transferred from mitochondria or plastids. It is likely that the DNA migration from mitochondria or plastids to the nucleus uses either an RNA-mediated transfer process or a direct genomic DNA transfer process. With regard to the RNA-mediated DNA transfer, a well-studied example is the gene encoding cytochrome c oxidase subunit 2 (cox2), which migrated to the nucleus via an edited RNA intermediate in legumes (Nugent and Palmer 1991; Covello and Gray 1992). Adams et al. (1999) reported that transfer and activation of cox2 appeared to have occurred during recent legume evolution. An investigation using 392 legume genera revealed that many kinds of intermediate stages existed among lineages: e.g. (1) mitochondrial cox2 has been lost or has been nonfunctional in some lineages, (2) cox2 exists in both mitochondria and the nucleus and both copies are expressed in some lineages, and (3) a functional cox2 is retained in mitochondria and the nuclear copy has become nonfunctional in some lineages (Adams et al. 1999). In the case of genomic DNA transfer, the recent complete sequencing of nuclear genomes of rice and Arabidopsis revealed that many mitochondrial- or plastidderived genomic DNA fragments migrated to the nuclear genome. Most of the mitochondrial- or plastid-derived DNA segments in the nuclear genome are less than 1 kbp in length, whereas some of the nuclear-integrated DNA segments are more

I.9 Frequent DNA Transfer Among Mitochondrial

113

than 100 kbp. For example, a 131-kbp plastid-derived DNA segment and a 190-kbp mitochondrial-derived DNA segment are located on chromosome 10 and chromosome 12, respectively, of rice (The Rice Chromosome 10 Sequencing Consortium 2003; Ueda et al. 2005), and a 620-kbp mitochondrial-derived DNA segment is located on chromosome 2 of Arabidopsis (Lin et al. 1999; Stupar et al. 2001). The 131-kbp plastid-derived DNA is observed in the Oryza sativa subsp. japonica nuclear genome, but the corresponding sequence is absent in the O. sativa subsp. indica and O. rufipogon nuclear genome (Huang et al. 2005). This suggests that the 131-kbp DNA fragment was transferred from a plastid to the nucleus after divergence of the subspecies japonica and indica. The date of integration of the plastid-derived sequence has been estimated between 74,000 and 296,000 years ago (Huang et al. 2005). The frequency of the plastid-to-nucleus DNA transfer using transgenic tobacco was directly monitored (Huang et al. 2003; Stegemann et al. 2003). Huang et al. (2003) integrated the kanamycin-resistant neomycin phosphotransferase gene (neo) under the control of the nuclear-specific cauliflower mosaic virus (CaMV) 35S promoter into the plastid genome, and then screened kanamycin-resistant seedlings using about 250,000 progeny produced by fertilization of a wild-type female with pollen of the transgenic plants. As a result, 16 independent transfers of the neo gene from a plastid to the nucleus were detected, and the frequency of the plastid-to-nucleus DNA transfer was estimated as one transposition event in about 16,000 pollen grains (Huang et al. 2003). On the other hand, Stegemann et al. (2003) transformed a different region of the plastid genome of tobacco by integration of the neo gene under the control of the CaMV 35S promoter. They screened kanamycin-resistant cells by placing small leaf sections of the neo homoplastomic lines on kanamycin-containing medium, and consequently obtained 12 independent kanamycin-resistant regenerated plants, in which the kanamycin-resistance was conferred by transfer of the neo gene from a plastid to the nucleus. As a result, they estimated one transposition event in about 5,000,000 somatic cells for the frequency of the plastid-to-nucleus DNA transfer (Stegemann et al. 2003), which is approximately two orders of magnitude less than the frequency estimated by Huang et al. (2003). Both studies indicated that the bulk plastid DNA (not RNA) had been transferred from a plastid and had been recombined into the nuclear genome. Taken together, these results reveal that organelle-tonucleus DNA transfers occur frequently and thus are an on-going evolutional event.

4

DNA Segments Found Commonly in all Three Genomes

Six DNA segments of 1.5, 1.2, 0.9 and 0.1 kbp and two of 0.2 kbp in length are found in all three genomes of rice (Notsu et al. 2002). All of them appear to be of plastid origin. Three of the six sequences were probably transferred independently from a plastid to a mitochondrion or the nucleus. On the other hand, it seems likely that the other three sequences first migrated from the plastid genome to the mitochondrial genome, and then part of each fragment was transferred from the mitochondrial genome to the nuclear genome (Notsu et al. 2002).

114

5

M. Nakazono and A. Hirai

Nucleus-to-Mitochondrion DNA Transfer Events

Sixteen retrotransposon-related sequences of various sizes (96–1,555 bp) and three transposon-related sequences that apparently originated from related sequences in the nuclear genome are dispersed in the rice mitochondrial genome (Notsu et al. 2002). About 13.4% of the rice mitochondrial genome is nuclear-derived. Similarly, sequences homologous to nuclear retrotransposon and transposon sequences were identified in the Arabidopsis and sugar beet mitochondrial genomes (Knoop et al. 1996; Kubo et al. 2000). Interestingly, most of the nuclear-derived sequences in the mitochondrial genome originate from mobile elements such as retrotransposons and transposons. An exception is a 168-bp segment of a proteincoding gene in the sugar beet mitochondrial genome (Kubo et al. 2000). In contrast to the cases of higher plants, no nuclear-derived DNA sequence was detected in the liverwort mitochondrial genome (Oda et al. 1992). No similarity was found among the nuclear-derived sequences from rice, Arabidopsis and sugar beet, suggesting that the 19 DNA segments independently migrated from the nucleus to the mitochondrial genome during the evolution of rice. The transfer of DNA fragments to the nucleus might occur relatively often since foreign DNA could easily be imported into the nucleus through the nuclear pores. However, mitochondrial DNA is localized in the matrix space, which is surrounded by inner and outer membranes. Using isolated potato mitochondria, Koulintchenko et al. (2003) showed that DNA was actively imported into mitochondria via a permeability transition pore complex. This finding suggests that the plastid- or nuclear-derived DNA fragments were imported into mitochondria using this import system, which provided them a chance to be integrated into the mitochondrial genome.

6

Conclusions and Future Perspectives

Now that the mitochondrial, plastid and nuclear genomes of rice and Arabidopsis have been completely sequenced, there is strong evidence that DNA has frequently and dynamically been transferred among the three genomes during evolution. The DNA transfer among the three genomes is likely to be an on-going process. However, there is no evidence that foreign sequences (i.e. mitochondrial- or nuclear-derived sequences) were inserted into the plastid genome, possibly because plastids do not have a DNA import system on their membrane, unlike the case of mitochondria. This feature may partly contribute to the conserved structures of the plastid genomes. In contrast, the frequent and dynamic DNA transfer from or to the mitochondrial genome may contribute to the genetic fluidity and plasticity of the mitochondrial genome of higher plants during evolution. Although methods for transforming the plastid and nuclear genomes are available, so far no method has been developed for transforming the plant mitochondrial genome. Such a

I.9 Frequent DNA Transfer Among Mitochondrial

115

method is necessary to demonstrate DNA transfer between the three genomes directly and to understand the functional significance of DNA transfer events during the evolution of plant genomes. Acknowledgements The authors express their appreciation to Dr. M. Ueda (University of Tokyo) for his critical reading of this chapter and to Drs. N. Tsutsumi and S. Arimura (University of Tokyo) for their stimulating discussions.

References Adams KL, Song K, Roessler PG, et al. (1999) Intracellular gene transfer in action: dual transcription and multiple silencings of nuclear and mitochondrial cox2 genes in legumes. Proc Natl Acad Sci USA 96:13863–13868 Adams KL, Daley DO, Qiu Y-L, Whelan J, Palmer JD (2000) Repeated, recent and diverse transfers of a mitochondrial gene to the nucleus in flowering plants. Nature 408:354–357 Adams KL, Rosenblueth M, Qiu Y-L, Palmer JD (2001) Multiple losses and transfers to the nucleus of two mitochondrial succinate dehydrogenase genes during angiosperm evolution. Genetics 158:1289–1300 Arimura S, Takusagawa S, Hatano S, Nakazono M, Hirai A, Tsutsumi N (1999) A novel plant nuclear gene encoding chloroplast ribosomal protein S9 has a transit peptide related to that of rice chloroplast ribosomal protein L12. FEBS Lett 450:231–234 Binder S, Brennicke A (2003) Gene expression in plant mitochondria: transcriptional and posttranscriptional control. Phil Trans R Soc Lond B 358:181–189 Clifton SW, Minx P, Fauron CM-R, et al. (2004) Sequence and comparative analysis of the maize NB mitochondrial genome. Plant Physiol 136:3486–3503 Covello PS, Gray MW (1992) Silent mitochondrial and active nuclear genes for subunit 2 of cytochrome c oxidase (cox2) in soybean: evidence for RNA-mediated gene transfer. EMBO J 11:3815–3820 Figueroa P, Gómez I, Holuigue L, Araya A, Jordana X (1999) Transfer of rps14 from the mitochondrion to the nucleus in maize implied integration within a gene encoding the iron-sulphur subunit of succinate dehydrogenase and expression by alternative splicing. Plant J 18:601–609 Gray MW, Boer PH (1988) Organization and expression of algal (Chlamydomonas reinhardtii) mitochondrial DNA. Phil Trans R Soc Lond B 319:135–147 Handa H (2003) The complete nucleotide sequence and RNA editing content of the mitochondrial genome of rapeseed (Brassica napus L.): comparative analysis of the mitochondrial genomes of rapeseed and Arabidopsis thaliana. Nucleic Acids Res 31:5907—5916 Hirai A, Ishibashi T, Morikami A, Iwatsuki N, Shinozaki K, Sugiura M (1985) Rice chloroplast DNA: a physical map and the location of the genes for the large subunit of ribulose 1,5-bisphosphate carboxylase and the 32KD photosystem II reaction center protein. Theor Appl Genet 70:117–122 Hiratsuka J, Shimada H, Whittier R, et al. (1989) The complete sequence of the rice (Oryza sativa) chloroplast genome: intermolecular recombination between distinct tRNA genes accounts for a major plastid DNA inversion during the evolution of the cereals. Mol Gen Genet 217:185–194 Huang CY, Ayliffe MA, Timmis JN (2003) Direct measurement of the transfer rate of chloroplast DNA into the nucleus. Nature 422:72–76 Huang CY, Grünheit N, Ahmadinejad N, Timmis JN, Martin W (2005) Mutational decay and age of chloroplast and mitochondrial genomes transferred recently to angiosperm nuclear chromosomes. Plant Physiol 138:1723–1733 International Rice Genome Sequencing Project (2005) The map-based sequence of the rice genome. Nature 436:793–800

116

M. Nakazono and A. Hirai

Iwahashi M, Nakazono M, Kanno A, Sugino K, Ishibashi T, Hirai A (1992) Genetic and physical maps and a clone bank of mitochondrial DNA from rice. Theor Appl Genet 84:275–279 Joyce PBM, Gray MW (1989) Chloroplast-like transfer RNA genes expressed in wheat mitochondria. Nucleic Acids Res 17:5461–5476 Kadowaki K, Kubo N, Ozawa K, Hirai A (1996) Targeting presequence acquisition after mitochondrial gene transfer to the nucleus occurs by duplication of existing targeting signals. EMBO J 15:6652–6661 Knoop V (2004) The mitochondrial DNA of land plants: peculiarities in phylogenetic perspective. Curr Genet 46:123–139 Knoop V, Unseld M, Marienfeld J, et al. (1996) copia-, gypsy- and LINE-like retrotransposon fragments in the mitochondrial genome of Arabidopsis thaliana. Genetics 142:579–585 Koulintchenko M, Konstantinov Y, Dietrich A (2003) Plant mitochondria actively import DNA via the permeability transition pore complex. EMBO J 22:1245–1254 Kubo N, Harada K, Hirai A, Kadowaki K (1999) A single nuclear transcript encoding mitochondrial RPS14 and SDHB of rice is processed by alternative splicing: common use of the same mitochondrial targeting signal for different proteins. Proc Natl Acad Sci USA 96:9207–9211 Kubo T, Nishizawa S, Sugawara A, Itchoda N, Estiati A, Mikami T (2000) The complete nucleotide sequence of the mitochondrial genome of sugar beet (Beta vulgaris L.) reveals a novel gene for tRNACys(GCA). Nucleic Acids Res 28:2571–2576 Lin X, Kaul S, Rounsley S, et al. (1999) Sequence and analysis of chromosome 2 of the plant Arabidopsis thaliana. Nature 402:761–768 Lonsdale DM, Hodge TP, Fauron CM-R (1984) The physical map and organisation of the mitochondrial genome from the fertile cytoplasm of maize. Nucleic Acids Res 12:9249–9261 Lonsdale DM, Brears T, Hodge TP, Melville SE, Rottmann WH (1988) The plant mitochondrial genome: homologous recombination as a mechanism for generating heterogeneity. Phil Trans R Soc Lond B 319:149–163 Miyata S, Nakazono M, Hirai A (1998) Transcription of plastid-derived tRNA genes in rice mitochondria. Curr Genet 34:216–220 Nakazono M, Hirai A (1993) Identification of the entire set of transferred chloroplast DNA sequences in the mitochondrial genome of rice. Mol Gen Genet 236:341–346 Nakazono M, Nishiwaki S, Tsutsumi N, Hirai A (1996) A chloroplast-derived sequence is utilized as a source of promoter sequences for the gene for subunit 9 of NADH dehydrogenase (nad9) in rice mitochondria. Mol Gen Genet 252:371–378 Notsu Y, Masood S, Nishikawa T, et al. (2002) The complete sequence of the rice (Oryza sativa L.) mitochondrial genome: frequent DNA sequence acquisition and loss during the evolution of flowering plants. Mol Genet Genomics 268:434–445 Nugent JM, Palmer JD (1991) RNA-mediated transfer of the gene coxII from the mitochondrion to the nucleus during flowering plant evolution. Cell 66:473–481 Oda K, Yamato K, Ohta E, et al. (1992) Gene organization deduced from the complete sequence of liverwort Marchantia polymorpha mitochondrial DNA. J Mol Biol 223:1–7 Ogihara Y, Yamazaki Y, Murai K, et al. (2005) Structural dynamics of cereal mitochondrial genomes as revealed by complete nucleotide sequencing of the wheat mitochondrial genome. Nucleic Acids Res 33:6235–6250 Ohyama K, Fukuzawa H, Kohchi T, et al. (1986) Chloroplast gene organization deduced from complete sequence of liverwort Marchantia polymorpha chloroplast DNA. Nature 322:572–574 Schuster W, Brennicke A (1988) Interorganellar sequence transfer: plant mitochondrial DNA is nuclear, is plastid, is mitochondrial. Plant Sci 54:1–10 Shinozaki K, Ohme M, Tanaka M, et al. (1986) The complete nucleotide sequence of the tobacco chloroplast genome: its gene organization and expression. EMBO J 5:2043–2049 Stegemann S, Hartmann S, Ruf S, Bock R (2003) High-frequency gene transfer from the chloroplast genome to the nucleus. Proc Natl Acad Sci USA 100:8828–8833 Stern DB, Lonsdale DM (1982) Mitochondrial and chloroplast genomes of maize have a 12-kilobase DNA sequence in common. Nature 299:698–702

I.9 Frequent DNA Transfer Among Mitochondrial

117

Stupar RM, Lilly JW, Town CD, et al. (2001) Complex mtDNA constitutes an approximate 620-kb insertion on Arabidopsis thaliana chromosome 2: implication of potential sequencing errors caused by large-unit repeats. Proc Natl Acad Sci USA 98:5099–5103 Sugiura M (1992) The chloroplast genome. Plant Mol Biol 19:149–168 Sugiura M, Takeda Y (2000) Nucleic acids. In: Biochemistry and molecular biology of plants (eds B.B. Buchanan, W. Gruissem, and R.L. Jones). American Society of Plant Physiologists, Rockville, pp. 260–310 Sugiyama Y, Watase Y, Nagase M, et al. (2005) The complete nucleotide sequence and multipartite organization of the tobacco mitochondrial genome: comparative analysis of mitochondrial genomes in higher plants. Mol Genet Genomics 272:603–615 The Arabidopsis Genome Initiative (2000) Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature 408:796–815 The Rice Chromosome 10 Sequencing Consortium (2003) In-depth view of structure, activity, and evolution of rice chromosome 10. Science 300:1566–1569 Tian X, Zheng J, Hu S, Yu J (2006) The rice mitochondrial genomes and their variations. Plant Physiol 140:401–410 Timmis JN, Ayliffe MA, Huang CY, Martin W (2004) Endosymbiotic gene transfer: organelle genomes forge eukaryotic chromosomes. Nat Rev Genet 5:123–135 Ueda M, Tsutsumi N, Kadowaki K (2005) Translocation of a 190-kb mitochondrial fragment into rice chromosome 12 followed by the integration of four retrotransposons. Int J Biol Sci 1:110–113 Ueda M, Arimura S, Yamamoto MP, Takaiwa F, Tsutsumi N, Kadowaki K (2006a) Promoter shuffling at a nuclear gene for mitochondrial RPL27. Involvement of interchromosome and subsequent intrachromosome recombinations. Plant Physiol 141:702–710 Ueda M, Fujimoto M, Arimura S, Tsutsumi N, Kadowaki K (2006b) Evidence for transit peptide acquisition through duplication and subsequent frameshift mutation of a preexisting protein gene in rice. Mol Biol Evol 23:2405–2412 Unseld M, Marienfeld JR, Brandt P, Brennicke A (1997) The mitochondrial genome of Arabidopsis thaliana contains 57 genes in 366,924 nucleotides. Nat Genet 15:57–61 Watanabe N, Nakazono M, Kanno A, Tsutsumi N, Hirai A (1994) Evolutionary variations in DNA sequences transferred from chloroplast genomes to mitochondrial genomes in the Gramineae. Curr Genet 26:512–518 Zeltz P, Kadowaki K, Kubo N, Maier RM, Hirai A, Kössel H (1996) A promiscuous chloroplast DNA fragment is transcribed in plant mitochondria but the encoded RNA is not edited. Plant Mol Biol 31:647–656

This page intentionally blank

Section II

Signal Transduction and Development

II.1

Hormonal Signal Transduction in Rice Ayako Nakamura1 and Makoto Matsuoka1(* ü)

1

Introduction

The compact nature of the rice genome provides a distinct advantage in gene isolation and genomic sequencing in contrast to other cereal crops. Further, the rice genome shows apparent syntenies with many other cereal crops such as wheat, barley, and maize (reviewed by Devos 2005). These syntenies suggest that rice genomics has implications not only for rice genetic studies and breeding, but also for other crops. Thus, rice has been selected as a target species for genome research by a number of research groups, and the International Rice Genome Sequencing Project (IRGSP) was launched in 1998, with the participation of ten countries (Sasaki et al. 2002). In 2004, the IRGSP declared the completion of the whole rice genome sequence, which provides very useful information for studying rice biology. Using this advantage and others in rice biology such as tagging libraries, transformation techniques, and full-length cDNA clones (An et al. 2005; Sasaki et al. 2005), we and other groups have been studying the signal transduction mechanisms of some phytohormones such as gibberellin (GA), brassinosteroid (BR), and auxin. Because dwarf characteristics are favored in plant breeding, many rice dwarf mutants have been identified and some have been used in the analysis of GA and BR (Ashikari and Matsuoka 2002). The sd1 mutant is a good example of how rice GA-related mutants have contributed to progress in basic science and breeding programs. This mutant enabled a dramatic increase in the yield of rice, which contributed significantly to global food security in the 1960s, during the time referred to as the “green revolution” (Khush 1999). We identified the SD1 gene and determined that the gene encodes GA20 oxidase, which catalyzes the final steps in GA biosynthesis (Ashikari et al. 2002; Sasaki et al. 2002). Erect leaves caused by BR deficiency or insensitivity also increase biomass production and grain yield (see Chapter IV.2). Consequently, we performed large-scale screening of GA-, BR-,

1 Bioscience and Biotechnology Center, Nagoya University, Nagoya, Aichi 464-8601, Japan e-mail: [email protected]; [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

121

122

A. Nakamura and M. Matsuoka

and auxin-related dwarf mutants and used them to analyze the biosynthetic and signal transduction pathways of these hormones. Here, we summarize recent knowledge regarding hormone signaling, based mainly on our own observations.

2 2.1

GA Signaling in Rice DELLA Protein Is a Key Regulator in GA Signaling

The DELLA-subfamily proteins of the GRAS super-family play an important role in the negative control of GA signaling. Proteins of the GRAS super-family are thought to function as transcription factors, although there is currently no direct evidence for this. Members of the DELLA subfamily contain the conserved amino acid motifs DELLA and TVHYNP in the N-terminal region. DELLA proteins are highly conserved in Arabidopsis and in several crop plants, including rice (SLR1; Ikeda et al. 2001), wheat (Rht; Peng et al. 1999), barley (SLN1; Chandler et al. 2002), and maize (d8; Peng et al. 1999). Gain-of-function mutations in this gene family result in dwarfism and reduced GA response, whereas loss-of-function mutations result in the GA-constitutive response phenotype, even in the presence of GA-biosynthesis inhibitors (Fig. 1). For instance, loss-of-function mutants of rice SLR1 show a slender phenotype with elongated leaf and stem (Ikeda et al. 2001). Further, GA-dependent α-amylase induction occurs in embryo half seeds in the absence of GA application. These slr1 phenotypes are similar to those of rice plants treated with exogenous GA, although the levels of endogenous GA are lower than in wild-type plants. Moreover, the GA-overdose phenotype of slr1 is not affected by the GA-biosynthesis inhibitor uniconazol (Ikeda et al. 2001). These observations indicate that DELLA proteins function as negative regulators in GA signaling. In contrast to the GA-constitutive phenotype of loss-of-function mutations in DELLA proteins, dominant alleles in genes encoding DELLA proteins confer a GA-insensitive phenotype with characteristic dwarfism. These dominant alleles have in-frame deletions in their conserved N-terminal domains such as DELLA and TVHYNP, resulting in constitutive DELLA protein function. Similarly, transgenic rice plants that produce an SLR1 protein truncated in the DELLA or TVHYNP domain have a dominant dwarf phenotype (Itoh et al. 2002). All of these mutants and transgenic plants show GA-insensitive characters, suggesting that the N-terminal region containing the DELLA and TVHYNP domains functions in the perception of upstream GA signals. We performed domain analysis of the rice DELLA protein SLR1 to investigate the biological function of each conserved domain in DELLA proteins (Itoh et al. 2002). This analysis showed that the C-terminal region containing the VHIID, PFYRE, and SAW domains, shared by all GRAS-family proteins, is involved in the suppressive function of DELLA proteins against GA action (Fig. 1). The proteins also contain Ser/Thr-rich residues, which may be involved in the regulation of their suppressive activity. It has been proposed that the activity of DELLA proteins is regulated by O-GlcNAc modification or phosphorylation

II.1 Hormonal Signal Transduction in Rice

DELLA

123

poly TVHYNP S/T/V LZ NLS VHIID

PFYRE

SAW

SLR1 in frame deletion

Leu289 one base deletion slr1-1

Trp561 stop codon slr1-2

Trp609 stop codon slr1-3

Trp620 stop codon slr1-4

Fig. 1 Above Schematic structure of the rice DELLA protein SLR1. Conserved domains located at the N-terminal portion are involved in perception of GA signaling; in-frame deletion of these domains causes constitutive suppression of GA signaling. Domains located in the C-terminal region, which are shared with other GRAS proteins, are involved in the suppressive function of the DELLA protein. Below Phenotypes of the SLR1 loss- and gain-of-function mutants. The gain-of-function mutation of SLR1 causes a dwarf phenotype via constitutive suppression of GA signaling (left), whereas its loss-of-function mutation causes a slender phenotype, as in rice plants treated with GA (right)

via the action of SPINDLY (SPY), which is another negative regulator of GA signaling, or kinase, with the Ser/Thr residues as the target site (Thornton et al. 1999; Shimada et al. 2006).

124

2.2

A. Nakamura and M. Matsuoka

GID2-dependent Degradation of SLR1 Is a Key Event in GA Signaling in Rice

The model of the DELLA protein degradation by 26S proteasome-mediated proteolysis was first suggested by the observation that the level of the barley SLN1 DELLA protein increases in the presence of 26S proteasome inhibitors (Fu et al. 2002). This model was greatly substantiated by the cloning of F-box genes from rice (OsGID2) and Arabidopsis (AtSLY1; McGinnis et al. 2003; Sasaki et al. 2003). The loss-of-function mutation of rice OsGID2 and Arabidopsis AtSLY1 results in GA-insensitive phenotypes. Positional cloning of the mutated gene revealed that OsGID2 encodes an F-box domain containing a protein that is a component of the SCF complex, which is named for its Skp1, cullin, and F-box protein subunits. The SCF complex catalyzes the transfer of ubiquitin from E2 to the target protein (Gagne et al. 2002). The addition of a polyubiquitin chain to the target protein by SCF induces the degradation of the target protein by the 26S proteasome. OsGID2 contains an F-box domain at the N terminus, as do other F-box proteins, but it lacks known protein–protein interaction domains at the C terminus. However, OsGID2 and its orthologous protein in Arabidopsis, AtSLY1, share conserved amino acid sequences not only at their N termini, but also within their C-terminal regions (McGinnis et al. 2003). Deletions of the conserved C-terminal regions cause a loss of function (Gomi et al. 2004), indicating that the C-terminal region is also important for their function. Yeast two-hybrid assays and in vivo immunoprecipitation experiments confirmed that OsGID2 is a component of the SCF complex through interaction with one of the rice Skp1-like proteins, OsSkp15 (Gomi et al. 2004). Ambient evidence supports the notion that the target of SCFGID2 is the DELLA protein SLR1: first, high levels of SLR1 accumulation are observed in gid2 mutants; and second, the rice double mutant carrying gid2-1/slr1-1 shows the slr1 phenotype (Sasaki et al. 2003; Dill et al. 2004). This suggests that the GA-insensitive phenotype of gid2 depends on the function of SLR1. Although direct interaction between Arabidopsis DELLA proteins and the Arabidopsis F-box protein AtSLY1 was observed in yeast cells (Dill et al. 2004), OsGID2 does not interact directly with SLR1 in yeast cells (Ueguchi-Tanaka, unpublished data). The recombinant GID2 protein produced in Escherichia coli interacts with SLR1 in rice crude extracts in vitro (Itoh et al. 2005), indicating that additional components are required for GID2–SLR1 interaction in rice cells.

2.3

GID1 Is a Soluble GA Receptor in Rice

Recently, we isolated a soluble GA receptor, GID1. The GID1 gene encoding a GA receptor was confirmed by the following observations (Ueguchi-Tanaka et al. 2005). First, the rice gid1 recessive mutant shows a typical GA-insensitive dwarf phenotype that is more severe than that of gid2 (Fig. 2). Second, the gid1-1/slr1-1

II.1 Hormonal Signal Transduction in Rice

125

Fig. 2 Phenotypic comparison of dwarfism among the wild-type (WT), gid2-2, and four gid1 alleles. All gid1 mutants show severe dwarfism in comparison with the null allele of the gid2 mutant

double mutant exhibits the slr1 phenotype, and GA treatment does not diminish the amount of SLR1 in gid1-1 plants, indicating that SLR1 is epistatic to GID1. The gid1 phenotype was similar to that of the GA-deficient mutant cps, indicating that GID1 functions upstream to SLR1 in the GA signaling pathway, but not in SLR1 degradation. Third, the most conclusive evidence for the GID1 gene encoding a GA receptor is that GID1 protein fused with a GST tag (GST–GID1) binds to 16,17-dihydro-GA4 with a reasonable dissociation constant (Kd) of 1.4×10−6 M. The ligand specificity of GST–GID1 for various GAs in vitro is generally consistent with the physiological activity of GAs. Namely, biologically active GAs generally have high binding affinity, whereas biologically inactive GAs have low affinity. Finally, the GA-perception activity of GID1 in vivo was confirmed by the GA-hypersensitive phenotype of transgenic rice plants that overproduce GID1. The GID1 protein is similar to those of the hormone-sensitive lipase (HSL) family, including the conserved HSL motifs HGG and GXSXG (Osterlund 2001). The severe phenotype of gid1-1 carrying a single amino acid exchange of the first G for D in the motif confirmed the importance of this GXSXG motif. Furthermore,

126

A. Nakamura and M. Matsuoka

GA binding is completely abolished when the GID 1 regions that are shared with the HSL family are deleted, indicating that the entire regions that are conserved between GID1 and HSL are essential for GA binding. Although GID1 shares some conserved regions with the HSL family, GID1 may not have lipase activity because it shares only two of the three conserved amino acid residues essential for HSL activity; the third residue, H, which is essential for the formation of the catalytic triad in the HSL family, is replaced by V. Further, recombinant GID1 does not hydrolyze an artificial substrate of HSL. GID1 interacts with SLR1 in a GA-dependent manner in yeast two-hybrid assays (Ueguchi-Tanaka et al. 2005), indicating that the GA–GID1 complex interacts directly with SLR1 and probably transduces the GA signal to SLR1. Further, the GA-binding activity of GID1 is increased about threefold by the presence of SLR1 (Ueguchi-Tanaka, unpublished data). This enhanced GA binding is caused by the decreased dissociation rate between GID1 and GA, whereas the association rate is not affected by SLR1 (Ueguchi-Tanaka, unpublished data). Thus, SLR1 stabilizes the interaction between GID1 and GA probably by a covering effect of SLR1 on the GA-interaction site of GID1. Domain analysis of SLR1 has revealed that the DELLA and TVHYNP domains are essential for its GA-dependent interaction with GID1 (Ueguchi-Tanaka, unpublished data). In contrast, deletion of the Leu-heptad domain, or the VHIID, PFYRE, and SAW domains, located on the C-terminal side, does not result in the complete loss of GID1 interaction. These observations indicate that the N-terminal portion of SLR1 is essential and sufficient for the GA-dependent interaction between GID1 and SLR1. Based on these observations, a model of GA perception mediated by GID1 has been proposed (Fig. 3). When GID1 binds with GA, the GA–GID1 complex can interact with SLR1 probably via conformational change. The DELLA/TVHYNP domains of SLR1 and the conserved HSL regions of GID1 are essential for the interaction between GID1 and SLR1. The association and dissociation of GID1 and GA occurs in the absence of SLR1, but when the GID1–GA complex interacts with SLR1, it is greatly stabilized. The stabilized tricomplex consisting of GA, GID1, and SLR1 may be a target of GID2, leading to SLR1 degradation by 26S proteasomes via ubiquitination of the SCFGID2 complex.

3 3.1

Brassinosteroid Signaling in Rice Characterization of BRASSINOSTEROID INSENSITIVE 1 Gene in Rice

BRASSINOSTEROID INSENSITIVE 1 (BRI1) was first isolated from Arabidopsis (Li and Chory 1997), and a recent molecular study has shown that BRI1 is a receptor for BRs (Wang et al. 2001). The first BR-related rice mutant was d61, which is defective in a gene homologous with Arabidopsis BRI1 (Yamamuro et al. 2000). To date, we have isolated ten alleles of the d61 mutant, d61-1 to d61-10. Mutants with mild alleles are

II.1 Hormonal Signal Transduction in Rice

127

GA

cytoplasm

nucleus

Ub

GA

GID1 DELLA/TVHYNP

SLR

SLR

SAW

SAW

GA action

GA action

Ub Ub

Rbx1

Cul1

Ub b U

GID1

GID1 DELLA/TVHYNP

GA

SLR

Skp1 F-box GID2

SCFGID2

degradation

GA action

Fig. 3 Model of GA signaling in rice. GID1 is a soluble GA receptor. When GID1 binds with GA, the GA–GID1 complex interacts with SLR1. The region containing the DELLA/TVHYNP domains of SLR1 is essential for the interaction between GID1 and SLR1. Association and dissociation of GID1 and GA occur rapidly without SLR1, but in the presence of SLR1, the GID1–GA complex is greatly stabilized. The stabilized complex consisting of GA, GID1, and SLR1 is targeted by GID2, leading to degradation of SLR1 by 26S proteasomes through ubiquitination of the SCFGID2 complex

fertile and grow to 80–90% of the height of the wild type, whereas the mutant with the severest allele is sterile and only reaches 5 cm in height 6 months after sowing (Fig. 4; Nakamura et al. 2006a). The Arabidopsis BRI1 protein consists of several domains, such as a putative signal peptide, two cysteine pairs, a leucine-rich repeat (LRR) domain, a transmembrane domain, and a kinase domain (Li and Chory 1997). The OsBRI1 protein also contains these domains, although the number of LRRs in Arabidopsis is 25, whereas that in rice is 22 (Yamamuro et al. 2000). The bri1 mutants of pea, Arabidopsis, rice, barley, and tomato greatly accumulate biologically active BRs such as castasterone (CS) and/or brassinolide (BL) because of loss of feedback control of BR biosynthesis (Chono et al. 2003; Fujioka and Yokota 2003). Using high-level accumulation of bioactive BRs in bri1 mutants, researchers have identified that dicotyledonous plants such as Arabidopsis and tomato use BL as a dominant bioactive BR, whereas monocotyledonous plants such as rice and barley use castasterone (CS), but not BL, as a dominant BR (Yamamuro et al. 2000; Chono et al. 2003; Nakamura et al. 2006). Whether the number of LRRs is related to the affinity with active BR is not yet

128

A. Nakamura and M. Matsuoka

Fig. 4 Phenotypes of d61 mutants. Left Gross morphology of the wild-type (WT), d61 mild (d61-1), and d61 intermediate (d61-2) alleles (bar = 20 cm). Right Gross morphology of the wild-type (WT) and severe allele (d61-4) 2 months after sowing (bar = 5 cm)

clear. The variety of d61 mutants is a good resource with which to clarify the detailed function of BRI1 in comparison with the series of Arabidopsis bri1 mutants.

3.2

OsBRL1 and OsBRL3, Which Are Homologous to OsBRI1, May Participate in BR Signaling

In rice, the most severe BR-biosynthetic mutant is brd1-1, which lacks the C-6 oxidase that catalyzes the final step of BR biosynthesis. We compared this BR-deficient mutant and the severest allele of bri1, d61-4, and found some differences between these BR-related mutants. In d61-4, the height of the shoot is less than 5 cm, whereas that of brd1-1 is more than 10 cm. In contrast, the length of the root of d61-4 is more than 10 cm, whereas that of brd1-1 is less than 5 cm (Fig. 5; Nakamura et al. 2006a). These observations are consistent with the fact that the shoot of d61-4 accumulates about 30 times the amount of bioactive CS as the wild type, whereas a high level of CS accumulation is not observed in its root (Nakamura et al. 2006a). In rice, as in Arabidopsis, there are three homologous genes for OsBRI1, the OsBRLs. Phylogenetic analyses of rice and Arabidopsis BRI1 and BRLs show that OsBRL1 and OsBRL3 are closely related to Arabidopsis BRL1 and BRL3 (Nakamura et al. 2006a). These two OsBRL genes are preferentially expressed in the roots. Thus, the two OsBRLs may rescue the OsBRI1-deficient phenotype of d61-4 in roots, and this may result in the difference in phenotype between brd1-1 and d61-4. Based on this hypothesis, OsBRL1 and OsBRL3 are at least partly involved in BR perception and rescue the loss-of-function of OsBRI1 in roots. The transcription regulation of OsBRl1 and OsBRL3 also supports this hypothesis; that is, their mRNA is increased in d61-4 and decreased in brd1-1. Because the BRI1 mRNA level is not regulated by the BR level in Arabidopsis, it is possible that Arabidopsis and rice possess different mechanisms of BRI1 gene expression.

II.1 Hormonal Signal Transduction in Rice

129

Fig. 5 Phenotypic comparison of d61-4 and brd1-1. Left Gross morphology of the shoots of 1-month-old d61-4 and brd1-1 plants. Right Gross morphology of the roots of 2-month-old wildtype (WT), d61-4, and brd1-1 plants (bar = 5 cm)

4 4.1

Auxin Signaling in Rice Early Auxin-responsive Genes

Aux/IAA (Auxin/INDOLE ACETIC ACID), GH3, and SAUR (Small Auxin-up RNA) genes are the most rapidly inducible genes in response to auxin. Molecular studies using Arabidopsis strongly suggest that these early auxin-inducible genes are involved in auxin signaling (Yang and Poovaiah 2000; Leyser 2002; Staswick et al. 2005). Homologous genes for Aux/IAA, GH3, and SAUR have been found in rice, and some members of these families are rapidly increased by auxin treatment. Another important family related to auxin signaling, ARF, has also been identified in rice. From in silico analyses, 31 OsAux/IAA, 12 OsGH3, 56 OsSAUR, and 11 OsARF genes have been found in the rice genome thus far (Sato et al. 2001; Jain et al. 2006a, 2006b; Terol et al. 2006). A novel P450, CYP87A3, has also been reported as an early auxinresponsive gene in rice (Chaban et al. 2003). CYP87A3, which encodes a P450 protein similar to BR biosynthetic enzymes, is induced within 1 h by auxin treatment, after which it rapidly decreases. This expression pattern in response to auxin is very similar to that of the Aux/IAA, GH3, and SAUR genes, implying that CYP87A3 is a novel factor involved in auxin signaling in rice.

4.2

Molecular Mechanism of Auxin Signaling in Rice

Aux/IAAs are short-lived nuclear proteins containing four conserved domains: I, II, III, and IV (Abel et al. 1995). Domain II, which consists of 13 conserved amino acids, is responsible for the rapid degradation of these Aux/IAA proteins, mediated by 26S proteasome (Gray et al. 2001; Ramos et al. 2001). By exchanging the conserved amino acid in domain II, Aux/IAA proteins become resistant to proteolytic degradation, even in the presence of auxin, and maintain ARFs in an inactive state; this leads to the inhibition of auxin signaling. In a yeast two-hybrid analysis, rice Aux/IAA proteins were found to interact with themselves and with ARFs, as observed in Arabidopsis (Kim et al. 1997; Ulmasov et al. 1997; Nakamura et al. 2006b). Rice Aux/IAA proteins are unstable and their degradation is inhibited by

130

A. Nakamura and M. Matsuoka

treatment with the proteasome inhibitor, MG132 (Thakur et al. 2005; Nakamura et al. 2006b). Recently, it has been reported that the F-box protein TIR1 is the auxin receptor in Arabidopsis (Kepinski and Leyser 2004; Dharmasiri et al. 2005). There are five homologous TIR1 genes in Arabidopsis, whereas there are four orthologous genes for TIR1 in the rice genome, i.e., LOC_Os02g52230, LOC_Os04g32460, LOC_Os05g05800, and LOC_Os11g31620 (TIGR database). In Arabidopsis, ASK1 and CULLIN1 (CUL1) form a complex with TIR1, which regulates auxin signaling. Orthologous genes for ASK1 and CUL1 are also found in the rice genome and are designated OsSKP15 and OsCUL1, respectively (Gomi et al. 2004). These observations suggest that rice uses a mechanism for auxin signaling similar to that of Arabidopsis, mediated by Aux/IAA and ARF proteins. We produced auxin-insensitive transgenic rice using mOsIAA3-GR, which overexpresses the mutant rice IAA protein in a dexamethasone (DEX)-dependent manner (Nakamura et al. 2006b). The auxin-insensitive transgenic rice was insensitive to gravitropic stimuli and exhibited short leaf blades, reduced crown and lateral root formation, and abnormal leaf formation (Nakamura et al. 2006b). The auxininsensitive rice mutant crl1/arl1, which is defective in the formation of crown (adventitious) roots, has been identified by our and other groups (Fig. 6; Inukai et al. 2005; Liu et al. 2005). CRL1/ARL1 encodes a nuclear protein containing an ASYMMETRIC LEAVES 2 (AS2)/LATERAL ORGAN BOUNDARIES (LOB) domain (Inukai et al. 2005; Liu et al. 2005). Exogenous auxin treatment induces CRL1/ARL1 gene expression without de novo biosynthesis, and this induction is not observed in DEX-treated mOsIAA3-GR transgenic plants (Inukai et al. 2005). CRL1/ARL1 has an auxin-responsive element in its promoter region that specifically interacts with ARF (Inukai et al. 2005). Thus, CRL1/ARL1 is the first transcription factor that functions just downstream of the Aux/IAA–ARF auxin signaling pathway, and it controls crown root development (Fig. 7). Recently, the novel mutants lrt1 (lateral rootless 1) and lrt2 were isolated in screening for 2,4-dichlorophenoxyacetic acid (2,4-D) resistance. These mutants fail to form lateral roots and exhibit altered root response to gravity (Chhun et al. 2003; Wang et al. 2006), and their phenotypes are similar to those of DEX-treated mOsIAA3-GR transgenic rice. Mapping of the LRT2 gene indicates that it is localized in a 10.8-cM interval

Fig. 6 Phenotypic comparison of mOsIAA3-GR and crown rootless1 (cr1l) roots. Cross sections of 2-week-old (left) wild-type, (middle ) DEX-treated mOsIAA3-GR, and (right) crl1 seedlings. The number of crown root primordia was decreased both in DEX-treated mOsIAA3-GR and crl1 seedlings (bar = 100 µm). CR Crown root

II.1 Hormonal Signal Transduction in Rice

131

IAA

ARF Degradation Dimarization

ARF

Aux/IAA, SAUR, GH3 CRL1

ARF Aux/IAA

Growth Auxin responses Fig. 7 Auxin signaling pathway in rice. Aux/IAAs function as negative regulators in auxin signaling to prevent the function of ARFs by direct interaction. Auxin treatment promotes the degradation of Aux/IAA to release ARF. The released ARFs interact with the auxin response element in the early auxin-inducible gene promoter to trigger its transcription, resulting in promotion of plant growth and auxin response

of chromosome 2, where some homologous genes for known auxin signaling genes are encoded (Wang et al. 2006). The abundance of rice mutants showing auxinrelated phenotypes suggests that there are many auxin-related mutants of rice, although most have not yet been characterized using molecular analyses. Consequently, we think that these uncharacterized mutants will contribute to extending our understanding of auxin signaling and biosynthesis, similar to the case of GA biosynthesis and signaling.

5

Future Perspectives

As mentioned in the Introduction, rice has a compact genome structure, and therefore has been selected as a model for monocot plants. Another advantage in using rice in terms of the study of hormone biosynthetic and signal transduction pathways is the absence of redundant hormone signaling genes. In another model plant, Arabidopsis, GA signaling genes are often redundant, making it difficult to isolate knockout mutants. Five genes encode the GA-suppressive DELLA protein in the Arabidopsis genome: GAI, RGA, RGL1, RGL2, and RGL3. In contrast, rice has only one gene encoding the DELLA protein, SLR1. Whereas Arabidopsis has three GA

132

A. Nakamura and M. Matsuoka

receptors, the rice genome has only one, GID1. The absence of redundant GA-related genes in rice provides an advantage in studying GA signaling; this may also hold true for other hormones. In addition, the phenotypic severity of rice mutants is much more variable than that of Arabidopsis mutants. For example, even the severest bri1 mutant of Arabidopsis, which lacks the primary BR receptor, has some flowers. In contrast, mild BRI1-deficient mutants of rice have erect leaves and form flowers, whereas severe mutants have malformed leaves and do not form flowers. Thus, the importance of an amino acid or a domain in a gene can be evaluated more precisely in rice than in Arabidopsis. These examples teach us that study of hormonal signal transduction in rice is important not only for expanding but also for deepening our knowledge in this field. Some important findings will be provided by forward and reverse genetics studies on the rice plant. Acknowledgement We thank Dr. Y. Inukai for kindly providing a picture of crl1 stem section.

References Abel S, Nguyen MD, Theologis A (1995) The PS-IAA4/5-like family of early auxin-inducible messenger- RNAs in Arabidopsis thaliana. J Mol Biol 251:533–549 An G, Lee S, Kim S-H, Kim S-R (2005) Molecular genetics using T-DNA in rice. Plant Cell Physiol 46:14–22 Ashikari M, Matsuoka M (2002) Application of rice genomics to plant biology and breeding. Bot Bull Acad Sin 43:1–11 Ashikari M, Sasaki A, Ueguchi-Tanaka M, et al. (2002) Loss-of-function of a rice gibberellin biosynthetic gene, GA20 oxidase (GA20ox-2), led to the rice ‘green revolution’. Breed Scie 52:143–150 Chaban C, Waller F, Furuya M, Nick P (2003) Auxin responsiveness of a novel cytochrome P450 in rice coleoptiles. Plant Physiol 133:2000–2009 Chandler PM, Marion-Poll A, Ellis M, Bubler F (2002) Mutants at the Slender1 locus of barley cv Himalaya. Molecular and physiological characterization. Plant Physiol. 129:181–190 Chhun T, Taketa S, Tsurumi S, Ichii M (2003) The effects of auxin on lateral root initiation and root gravitropism in a lateral rootless mutant Lrt1 of rice (Oryza sativa L.). Plant Growth Reg 39:161–170 Chono M, Honda I, Zeniya H, et al. (2003) A semidwarf phenotype of barley uzu results from a nucleotide substitution in the gene encoding a putative brassinosteroid receptor. Plant Physiol 133:1209–1219 Devos KM (2005) Updating the ‘crop circle’. Curr Opin Plant Biol 8:155–162 Dharmasiri N, Dharmasiri S, Estelle M (2005) The F-box protein TIR1 is an auxin receptor. Nature 435:441–445 Dill A, Thomas SG, Hu J, Steber CM, Sun T-p (2004) The Arabidopsis F-box protein SLEEPY1 targets gibberellin signaling repressors for gibberellin-induced degradation. Plant Cell 16:1392–1405 Fu X, Richards DE, Ait-ali T, et al. (2002) Gibberellin-mediated proteasome-dependent degradation of the barley DELLA protein SLN1 repressor. Plant Cell 14:3191–3200 Fujioka S, Yokota T (2003) Biosynthesis and metabolism of brassinosteroids. Annu Rev Plant Biol 54:137–164 Gagne JM, Downes BP, Shin-Han S, Durski AM, Vierstra RD (2002) The F-box subunit of the SCF E3 complex is encoded by a diverse superfamily of genes in Arabidopsis. Proc Natl Acad Sci USA 99:11519–11524

II.1 Hormonal Signal Transduction in Rice

133

Gomi K, Sasaki A, Itoh H, et al. (2004) GID2, an F-box subunit of the SCF E3 complex, specifically interacts with phosphorylated SLR1 protein and regulates the gibberellin-dependent degradation of SLR1 in rice. Plant J 37:626–634 Gray WM, Kepinski S, Rouse D, Leyser O, Estelle M (2001) Auxin regulates SCFTIR1-dependent degradation of AUX/IAA proteins. Nature 414:271–276 Ikeda A, Ueguchi-Tanaka M, Sonoda Y, et al. (2001) Slender rice, a constitutive gibberellin response mutant is caused by a null mutation of the SLR1 gene, an ortholog of the heightregulating gene GAI/RGA/RHT/D8. Plant Cell 13:999–1010 Inukai Y, Sakamoto T, Ueguchi-Tanaka M, et al. (2005) The Crown rootless1 gene in rice is essential for crown root formation and is a target of AUXIN RESPONSE FACTOR in auxin signaling. Plant Cell 17:1387–1396 Itoh H, Ueguchi-Tanaka M, Sato Y, Ashikari M, Matsuoka M (2002) The gibberellin signaling pathway is regulated by the appearance and disappearance of SLENDER RICE1 in nuclear. Plant Cell 14:57–70 Itoh H, Sasaki A, Ueguchi-Tanaka M, et al. (2005) Dissection of the phosphorylation of rice DELLA protein, SLENDER RICE1. Plant Cell Physiol 46:1392–1399 Jain M, Kaur N, Tyagi AK, Khurana JP (2006a) The auxin-responsive GH3 gene family in rice (Oryza sativa). Funct Integr Genomics 6:36–46 Jain M, Tyagi AK, Khurana JP (2006b) Genome-wide analysis, evolutionary expansion, and expression of early auxin-responsive SAUR gene family in rice (Oryza sativa). Genomics 88:360–371 Kepinski S, Leyser O (2004) Auxin-induced SCFTIR1-Aux/IAA interaction involves stable modification of the SCFTIR1 complex. Proc Natl Acad Sci USA 101:12381–12386 Khush GS (1999) Green revolution: preparing for the 21st century. Genome 42:645–655 Kim J, Harter K, Theologis A (1997) Protein–protein interactions among the Aux/IAA proteins. Proc Natl Acad Sci USA 94:11786–11791 Leyser O (2002) Molecular genetics of auxin signaling. Annu Rev Plant Biol 53:37–398 Li JM, Chory J (1997) A putative leucine-rich repeat receptor kinase involved in brassinosteroid signal transduction. Cell 90:929–938 Liu H, Wang S, Yu X, et al. (2005) ARL1, a LOB-domain protein required for adventitious root formation in rice. Plant J 43:47–56 McGinnis KM, Thomas SG, Soule JD, et al. (2003) The Arabidopsis SLEEPY1 gene encodes a putative F-box subunit of an SCF E3 ubiquitin ligase. Plant Cell 15:1120–1130 Nakamura A, Fujioka S, Sunohara H, et al. (2006a) The role of OsBRI1 and its homologous genes, OsBRL1 and OsBRL3, in rice. Plant Physiol 140:580 Nakamura A, Umemura I, Gomi K, et al. (2006b) Production and characterization of auxin-insensitive rice by overexpression of a mutagenized rice IAA protein. Plant J 46:297–306 Osterlund T (2001) Structure-function relationships of hormone-sensitive lipase. Eur J Biochem 268:1899–1907 Peng J, Richards DE, Hartley NM, et al. (1999) ‘Green revolution’ genes encode mutant gibberellin response modulators. Nature 400:256–261 Ramos JA, Zenser N, Leyser O, Callis J (2001) Rapid degradation of Auxin/Indoleacetic acid proteins requires conserved amino acids of domain II and is proteasome dependent. Plant Cell 13:2349–2360 Sasaki A, Ashikari M, Ueguchi-Tanaka M, et al. (2002) Green revolution: a mutant gibberellinsynthesis gene in rice. Nature 416:701–702 Sasaki A, Itoh H, Gomi K, et al. (2003) Accumulation of phosphorylated repressor for gibberellin signaling in an F-box mutant. Science 299:1896–1898 Sasaki T, Matsumoto T, Antonio BA, Nagamura Y (2005) From mapping to sequencing, postsequencing and beyond. Plant Cell Physiol 46:3–13 Sato Y, Nishimura A, Ito M, Ashikari M, Hirano H-Y, Matsuoka M (2001) Auxin response factor family in rice. Genes Genet Syst 76:373–380

134

A. Nakamura and M. Matsuoka

Shimada A, Ueguchi-Tanaka M, Sakamoto T, et al. (2006) The rice SPINDLY gene functions as a negative regulator of gibberellin signaling by controlling the suppressive function of the DELLA protein, SLR1, and modulating brassinosteroid synthesis. Plant J 48:390–402 Staswick PE, Serban B, Rowe M, et al. (2005) Characterization of an Arabidopsis enzyme family that conjugates amino acids to indole-3-acetic acid. Plant Cell 17:616–627 Terol J, Domingo C, Talon M (2006) The GH3 family in plants: genome wide analysis in rice and evolutionary history based on EST analysis. Gene 371:279–290 Thakur JK, Jain M, Tyagi AK, Khurana JP (2005) Exogenous auxin enhances the degradation of a light down-regulated and nuclear-localized OsiIAA1, an Aux/IAA protein from rice, via proteasome. Biochim Biophys Acta 1730:196–205 Thornton TM, Swain SM, Olszewski NE (1999) Gibberellin signal transduction presents…the SPY who O-GlcNAc’d me. Trends Plant Sci 4:424–428 Ueguchi-Tanaka M, Ashikari M, Nakajima M, et al. (2005) GIBBERELLIN INSENSITIVE DWARF1 encodes a soluble receptor for gibberellin. Nature 437:1008–1019 Ulmasov T, Hagen G, Guilfoyle TJ (1997) ARF1, a transcription factor that binds to auxin response elements. Science 276:1865–1868 Wang H, Taketa S, Miyao A, Hirochika H, Ichii M (2006) Isolation of a novel lateral-rootless mutant in rice (Oryza sativa L.) with reduced sensitivity to auxin. Plant Sci 170:70–77 Wang Z-Y, Seto H, Fujioka S, Yoshida S, Chory J (2001) BRI1 is a critical component of a plasma-membrane receptor for plant steroids. Nature 410:380–383 Yamamuro C, Ihara Y, Wu X, et al. (2000) Loss of function of a rice brassinosteroid insensitive1 homolog prevents internode elongation and bending of the lamina joint. Plant Cell 12:1591–1605 Yang T, Poovaiah BW (2000) Molecular and biochemical evidence for the involvement of calcium/Calmodulin in auxin action. J Biol Chem 275:3137–3143

II.2

Rice Heterotrimeric G Protein Signaling Yukimoto Iwasaki1(* ü ), Hisaharu Kato1, Yukiko Fujisawa1, 1 and Katsuyuki Oki

1

Introduction

Signaling pathways regulated by heterotrimeric G proteins are well known to be essential in eukaryotic organisms. In the last decade, it has been shown that plants also have genes for the canonical subunits of heterotrimeric G proteins. The genomes of rice and Arabidopsis thaliana have only a single copy gene for each of the α, β and γ1 subunits. This situation is in contrast to mammalian genomes that contain multiple genes for each of the 23 α, 6 β and 12 γ subunits. This is a striking characteristic of higher plant heterotrimeric G proteins. In this chapter, we will summarize the subunits of rice heterotrimeric G proteins and discuss the possible functions of the G proteins in rice and Arabidopsis.

2

2.1

Genes and Translation Products for Rice Heterotrimeric G Proteins Genes for the Subunits of Rice Heterotrimeric G Proteins

The genes for the α subunit (Rice heterotrimeric G protein α subunit: RGA1) (Ishikawa et al. 1995), the β subunit (RGB1) and the γ1 subunit (RGG1) (Kato et al. 2004) are present in the rice genome as a single copy for each subunit. The rice α subunit, β subunit and γ1 subunit are considered to be the canonical subunits, based on the presence of motifs conserved among mammals, yeast and plants (Jones and Assmann 2004). The α subunit is the active subunit catalyzing GTP hydrolysis. The motifs, such as the βγ binding sites, GTP binding sites and a receptor- interacting region, are well conserved among mammals, yeast and plants (Ma et al. 1990; Simon

1

Department of Bioscience, Fukui Prefectural University, 4-1-1 Kenjyojima, Matsuoka, Eiheiji-cho, Yoshida-gun, Fukui 910–1195, Japan e-mail: [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

135

136

Y. Iwasaki et al.

et al. 1991; Lambright et al. 1994; Ishikawa et al. 1995). The β subunit has a typical seven-repeat WD-40 motif, which is found in the subunits of mammals, yeast and plants (Neer et al. 1994; Weiss et al. 1994; Ishikawa et al. 1996). The sequence of the rice γ1 subunit is also homologous to the sequences of the γ subunits of mammals, yeast and plants (Mason and Botella 2000; Kato et al. 2004). Some characteristic repeated sequences are found in the upstream regions of the α-subunit gene in rice (Fig. 1A). These sequences are homologous to the sequence from the promoter to exon 1 in the α-subunit gene and are designated as the promoter-like sequence, ProL. Four ProLs, ProL1–ProL4, are found in the upstream region of the α-subunit gene. These promoter regions may be important for understanding the expression of the α-subunit gene. It remains to be ascertained whether ProL1–ProL4 have promoter activity or are involved in a new gene. A sequence highly homologous to that from exon 2 to exon 8 in the α-subunit gene is present in a region about 380-kb upstream and is designated RGA2 (Fig. 1A).

Fig. 1 Genome structure of genes of the α, β and γ1 subunit of rice heterotrimeric G proteins. A Location of the α-subunit gene of rice heterotrimeric G proteins (RGA1; black arrow). Sequences homologous to the sequence from the promoter to intron 1 of the RGA1 gene are shown by gray arrows and designated as ProL1, 2, 3 and 4. The box with oblique lines in the gene is the promoter region. Boxes shaded black and gray in the gene show exon and intron, respectively. RGA2 is a sequence highly homologous to that from exon 2 to exon 8 of the RGA1 gene. B Schematic diagrams of the structure of genes of the α, β (RGB1) and γ1 (RGG1) subunit of rice heterotrimeric G proteins. Boxes shaded black and gray show exon and intron, respectively. The accession numbers of RGA1, RGB1 and RGG1 are D38232, X89737 and AB120662, respectively.

II.2 Rice Heterotrimeric G Protein Signaling

137

A transcript of the RGA2 gene has been detected although the RGA2 gene is considered to be a pseudogene, because of the many stop codons present. The distribution of exons and introns in the genes for the α, β and γ1 subunits is illustrated in Fig. 1B.

2.2

Subunits Related to the Canonical Subunits of Rice Heterotrimeric G Proteins

In addition to the canonical subunits, four extra-large Gαs (XLG1–XLG4) and one extra-large γ subunit (RGG2) (Kato et al. 2004) have been found in the rice genome (Fig. 2). The extra-large γ subunit was synonymous with γ2 subunit in rice. These proteins contain extra regions at the N-terminus when compared to the canonical subunits. The extra-large Gαs and an extra-large γ subunit are also found in Arabidopsis and the amino acid sequences of these proteins are well conserved between monocot and dicot.

Fig. 2 Schematic diagrams of subunits related to canonical heterotrimeric G protein subunits. A Rice has four kinds of extra-large Gα (XLG1–XLG4) in addition to the canonical α subunit (RGA1). Extra parts are shown as gray boxes. Conserved amino acid sequences are found in six parts and shown as circled numbers. All six parts are GTP-binding regions. The accession numbers of XLG1, XLG2, XLG3 and XLG4 are AK069838, AK069964, NP_919599 and AK067863, respectively. B Rice has an extra-large Gγ (RGG2) in addition to the canonical γ subunit (RGG1). RGG2 protein has an extra N-terminal region containing 57 amino acids and no CAAX motif at the C terminal as found in RGG1. The accession number of RGG2 is AP008208

138

Y. Iwasaki et al.

The functions of these proteins with the extra regions remain unclear. If the four extralarge Gαs (XLG1–XLG4) and one extra-large γ subunit (RGG2) have functions similar to canonical subunits then higher plant G protein signaling will be complex beyond our expectations.

2.3

Intracellular Localization and Subunit Composition of Rice Heterotrimeric G Proteins

The rice α, β, γ1 and γ2 subunits were synthesized as recombinant proteins with tags such as histidine, thioredoxin and glutathione S-transferase in E. coli (Kato et al. 2004). Antibodies against the recombinant α, β, γ1 and γ2 subunits were then prepared. Western blot analysis with the antibodies demonstrated all subunits to be localized mainly in the plasma membrane. Examination of whether the heterotrimeric G proteins are actually heterotrimeric in rice has also been conducted (Kato et al. 2004). After solubilization of plasma membrane proteins with cholate, the solubilized proteins were separated by gel filtration. Using this technique, large complexes containing α, β, γ1 and γ2 subunits and large amounts of free βγ1 and βγ2 dimers were detected. We are interested in the subunit composition of the large complexes as well as elucidating the function of the free βγ dimers.

3

Mutants with Defects in the Subunit genes Encoding Plant Heterotrimeric G Proteins

A rice mutant with a defect in the α-subunit gene, d1, has been isolated (Ashikari et al. 1999; Fujisawa et al. 1999). Defective mutants for the α- and β- subunit genes, gpa1 (Ullah et al. 2001) and agb1 (Lease et al. 2001) have also been isolated from Arabidopsis. The results of analysis of the signaling pathways using these mutants will be described in Section 65.

4

Analysis of Signaling Mediated by Rice Heterotrimeric G Proteins Using a Constitutively Active Form of the a Subunit

In order to study the function of rice heterotrimeric G proteins, we compared wildtype (WT) rice (a line with normal α-subunit gene) with d1 (a knock-out mutant of the α-subunit gene) or with constitutively active rice, QL/d1 (a transformant expressing a constitutively active form of the α subunit, Q223L protein). The d1 plants show abnormal phenotypes such as dwarfism, dark green leaves and small round seeds. The QL/d1 plants bear 20% larger seeds than WT but show WT-like phenotypes with regard to their plant sizes and leaf colors (Oki et al. 2005). Thus,

II.2 Rice Heterotrimeric G Protein Signaling

139

the effect of the constitutively active form of the α subunit is restricted to seed size. The Q223L protein is localized in the plasma membrane in a monomeric form, although the wild-type α subunit is present as a subunit member in a large complex in the plasma membrane. This indicates that the Q223L protein may activate the G protein signaling pathway in rice plants in the absence of external signals. It has been suggested that plant heterotrimeric G proteins are involved in gibberellin (GA) signaling. There is a GA-related recessive mutant, slender rice (slr), which shows a GA constitutive response phenotype including the slender characteristics of pale green and elongated leaves (Ikeda et al. 2001). To determine the epistatic relationship between d1 and slr, the two mutants were crossed. The double mutant showed a slender phenotype (Ueguchi-Tanaka et al. 2000). The result demonstrates that slr may be epistatic to d1. However, the G protein does not seem to play a major role in gibberellin signaling, because QL/d1 does not show a slender phenotype (Oki et al. 2005).

5

Molecules Interacting with Arabidopsis Ga

Seven proteins that interact with Arabidopsis Gα have been identified by multiple biochemical approaches including bimolecular fluorescence complementation assay, co-immunoprecipitation assay, forster resonance energy transfer analysis, pull down assay, surface plasmon resonance assay, yeast two-hybrid assay and yeast split-ubiquitin system (Table 1).

5.1

Arabidopsis Putative G Protein-coupled Receptor 1 (GCR1)

GCR1 is considered to be a receptor coupled to heterotrimeric G proteins because it has the seven membrane spanning domain structure. To our knowledge, signal molecules received by GCR1 have not yet been identified. GCR1 is a negative regulator of abscisic acid (ABA) signaling in root growth (Pandey and Assmann 2004) and in stomata (Pandey and Assmann 2004). In contrast, GCR1 is a positive regulator of signaling of gibberellic acid (GA) (Chen et al. 2004a), brassinosteroid (BR) (Chen et al. 2004a) and blue light (Warpeha et al. 2006). GCR1 may also be involved in DNA synthesis (Apone et al. 2003).

5.2

Arabidopsis G Protein-coupled Receptor 2 (GCR2)

GCR2 has been identified as an ABA receptor in the plasma membrane (Liu et al. 2007). The specific binding of ABA to purified GCR2 has been studied and the equilibrium dissociation constant for the GCR2–ABA complex was found to be 20.1 nM. GCR2 is a positive regulator of almost all ABA signaling.

140

Y. Iwasaki et al.

Table 1 Molecules interacting with Gα and their mutants in Arabidopsis Categories

Molecules

Methods

Mutants

References

GCR1

YSU, PD, IP

gcr1

GCR2

YSU, SPR, IP, gcr2 BF

Pandey and Assmann 2004 Liu et al. 2007

AtRGS1

YSU, PD, IP

G protein-coupled receptor

Modulator Atrgs1

SPR

Effector

AtPirin1

YTH, PD

atpirin1

AtPLDα1

PD, YTH

pldα1

PD1

YTH, PD

pd1

THF1

YTH, PD, IP, FRET

thf1

Chen et al. 2003 Willard and Siderovski 2004 Lapik and Kaufman 2003 Mishra et al. 2006 Warpeha et al. 2006 Huang et al. 2006

Abbreviations: AtPirin1, Arabidopsis pirin 1; AtPLDα1, Arabidopsis phospholipase Dα1; AtRGS1, Arabidopsis regulator of G protein signaling 1; BF, bimolecular fluorescence complementation test; FRET, Forster resonance energy transfer; GCR1, Arabidopsis putative G protein-coupled receptor 1; GCR2, Arabidopsis G protein-coupled receptor 2; IP, Co-immunoprecipitation assay; PD, Pull down assay; PD1, Arabidopsis prephenate dehydrogenase 1; SPR, surface plasmon resonance; THF1, Arabidopsis thylakoid formation 1; YSU, Yeast split-ubiquitin system; YTH, yeast two-hybrid assay

5.3

Arabidopsis Regulator of G Protein Signaling 1 (AtRGS1)

AtRGS1 has been shown to possess the seven membrane spanning domain and GTPase activating domain (Chen et al. 2003). AtRGS1 antagonizes the activation of Arabidopsis Gα and functions as a negative regulator. AtRGS1 has roles in sugar signaling and the regulation of cell number.

5.4

Arabidopsis Pirin 1 (AtPirin1)

AtPirin1 is a protein containing a cupin-domain found in the pirin protein. The pirin interacts with a transcription factor that recognizes a CCAAT box (Lapik and Kaufman 2003). AtPirin1 has been shown to function as a negative regulator of ABA signaling in early seedling development (Warpeha et al. 2007).

II.2 Rice Heterotrimeric G Protein Signaling

5.5

141

Arabidopsis Phospholipase Da 1 (AtPLDa 1)

AtPLDα1 interacts with Arabidopsis Gα through the DRY motif (Zhao and Wang 2004). AtPLDα1 functions as a positive regulator of ABA signaling in the stomatal response through production of phosphatidic acid (PA) (Mishra et al. 2006).

5.6

Prephenate Dehydrogenase 1 (PD1)

PD1 is a key enzyme in the production of phenylalanine. Blue light is necessary for phenylalanine production and the activation of Arabidopsis Gα. PD1 is a positive regulator of G-protein-mediated blue light signaling.

5.7

Arabidopsis Thylakoid Formation 1 (THF1)

THF1 is a plastid protein. Contact between THF1 and Arabidopsis Gα appears to take place at sites where the plastid membrane abuts on the plasma membrane. THF1 is a negative regulator of sugar signaling and its protein levels are rapidly decreased by D-glucose.

6

Signaling Pathways Mediated by Plant Heterotrimeric G Proteins

The candidates for signaling pathways in which plant heterotrimeric G proteins are involved are pathogen, ABA, GA, BR, blue light, extracellular calmodulin and sugar (Table 2).

6.1

Pathogen

The induction of pathogenesis-related genes after treatment with pathogen, elicitor and sphingolipid elicitor (SE) was delayed in d1 compared with WT (Suharsono et al. 2002; Iwata et al. 2003; Komatsu et al. 2004). The production of H2O2 induced by SE is also reduced in d1 compared to WT (Suharsono et al. 2002). The production of reactive oxygen species was also impaired in gpa1, an Arabidopsis mutant of the Gα gene (Joo et al. 2005). Judging from these observations, plant heterotrimeric G proteins seem to function as positive regulators in the disease resistance system.

Bethke et al. 2006 Ueguchi-Tanaka et al. 2000 Ullah et al. 2002; Chen et al. 2004a

GA - Gα - Enhancement of seed germination (Seed germination is not enhanced in gpa1)

Pandey and Assmann 2004

Liu et al. 2007 Liu et al. 2007 Ullah et al. 2002; Lapik and Kaufman 2003 Lapik and Kaufman 2003

Pandey and Assmann 2004 Liu et al. 2007

Wang et al. 2001 Coursol et al. 2003 Mishra et al. 2006

Wang et al. 2001

Suharsono et al. 2002 Suharsono et al. 2002 Suharsono et al. 2002 Iwata et al. 2003; Komatsu et al. 2004 Joo et al. 2005 Viehweger et al. 2006

References

1. Pathogen Elicitor - Gα - OsRac1 - PR gene expression (PR gene expression is delayed in d1) Sphingolipid - Gα - OsRac1 - PR gene expression (PR gene expression is not induced in d1) Sphingolipid - Gα - OsRac1 - H2O2 production (H2O2 production is reduced in d1) Probenazol - Gα - PBZ1 accumulation (PBZ1 gene expression is delayed in d1) Ozone - Gα–ROS production (ROS production is reduced in gpa1) Elicitor - Gα - PLA2 − - H+ - Induction of phytoalexin biosynthesis in California poppy 2. Abscisic acid (ABA) and sphingosine-1-phospate (S1P) ABA - Gα - Inhibition of stomatal inwardly rectifying K+ channel - Stomatal closure (ABA inhibition of stomatal opening is impaired in gpa1) ABA - Gα - Activation of stomatal anion channel - Stomatal closure (Anion channel is not activated in gpa1) ABA - S1P - Gα - Stomatal closure (gpa1 does not respond to SIP) ABA - PLDa1 - PA - Gα - Stomatal closure (gpa1 impairs ABA inhibition of stomatal opening and pldα1 impairs both ABA promotion of stomatal closure and ABA inhibition of stomatal opening) ABA - S1P - (GCR1?) - Stomatal response (Stomata show hypersensitive response to ABA and S1P in gcr1) ABA - GCR2 - Gα - Inhibition of stomatal response (ABA inhibition of stomatal opening and closure is impaired in gcr2) ABA - GCR2 - Gα - Inhibition of seed germination (ABA inhibition is impaired in seed germination in gcr2) ABA - GCR2 - Gα - Gene expression (ABA marker genes do not respond to ABA in gcr2) ABA - Gα - Inhibition of seed germination (ABA inhibition in seed germination is hypersensitive in gpa1) ABA - AtPirin1 - Inhibition of seed germination (ABA inhibition in seed germination is hypersensitive in atpirin1) ABA - GCR1 - Inhibition of root growth (ABA inhibition in root growth is hypersensitive in gcr1) 3. Gibberellin (GA) GA - Gα - Increase of transcripts (Many transcripts are reduced in d1) GA - Gα - Increase of Ca2+ ATPase and GAmyb transcripts (These transcripts are reduced in d1)

Signals and downstream events in plant heterotrimeric G-protein-mediated signaling

Table 2 Pathways in which plant heterotrimeric G proteins may be involved. Candidate signals transmitted to heterotrimeric G proteins are indicated in bold. The differences in phenotypes of mutants of G-protein-mediated signaling compared to WT are described in parentheses

142 Y. Iwasaki et al.

GA - GCR1 - Enhancement of seed germination (Seed germination is not enhanced in gcr1) 4. Brassinosteroid (BR) BR - Gα - Enhancement of seed germination (Seed germination is not enhanced in gpa1) BR - GCR1 - Enhancement of seed germination (Seed germination is not enhanced in gcr1) BR (ABA) - GCR1 - Gα - LHCb gene expression (LHCb mRNA is reduced in gcr1, gpa1 and pirin1) 5. Blue light (BL) BL - Activation of Gα BL - GCR1 - Gα - PD1 - Phenylalanine (Phe) accumulation (Phe is not accumulated in gpa1, gcr1 and pd1) 6. Extracellular Calmodulin (ExtCaM) ExtCaM - Gα - Promotion of stomatal closure (Stomata do not respond to ExtCaM in gpa1) ExtCaM - Gα - H2O2 production and the elevation of cytosolic Ca2+ (These are reduced in gpa1) 7. Sugar D-glucose - Gα - Growth inhibition (Growth inhibition by sugar is accelerated or hypersensitive in gpa1and thf1) (Growth inhibition by sugar is not observed when constitutively active Gα and THF1 are overexpressed) D-glucose - AtRGS1 (Sugar-inhibition is not observed in Atrgs1) 8. Unknown signals (a) Cell proliferation Gα overexpression - Increase of cell proliferation GCR1 overexpression - Gα - PLC and IP3–Increase of DNA synthesis AtRGS1 (Cell number increases in Atrgs1) (b)Others Gα - Accumulation of OsMAPK6 protein (OsMAPK6 protein dose not accumulate in d1) Gα - Activation of PLDα activity Gα - Activation of Ca2+ channels Lieberherr et al. 2005 Munnik et al. 1995 Aharon et al. 1998

Ullah et al. 2001 Apone et al. 2003 Chen et al. 2003

Chen and Jones 2004

Huang et al. 2006

Chen et al. 2004b Chen et al. 2004b

Warpeha et al. 1991 Warpeha et al. 2006

Chen et al. 2004a Chen et al. 2004a Warpeha et al. 2007

Chen et al. 2004a

II.2 Rice Heterotrimeric G Protein Signaling 143

144

6.2

Y. Iwasaki et al.

Abscisic Acid (ABA) and Sphingosine-1 Phosphate (S1P)

The importance of heterotrimeric G proteins in higher plants has been noted in studies of stomatal responses to ABA treatment. In stomata, Arabidopsis Gα is considered as a positive regulator of ABA signaling (Wang et al. 2001; Coursol et al. 2003; Mishra et al. 2006). GCR1 is a negative regulator of ABA signaling in stomata (Pandey and Assmann 2004). GCR2 is a positive regulator of ABA signaling in stomata, because ABA inhibition of stomatal opening and closure was impaired in the gcr2 mutant (Liu et al. 2007). In seed germination, Arabidopsis Gα serves as a negative regulator of ABA signaling in seed germination (Ullah et al. 2002). Other researchers have reported that Arabidopsis Gα and AtPirin1 also function as negative regulators of ABA signaling in seed germination, because the inhibition of seed germination by ABA is accelerated in gpa1 and atpirin1 compared to WT (Lapik and Kaufman 2003). GCR1 functions as a negative regulator of ABA signaling in root growth (Pandey and Assmann 2004). Thus, GCR1 and Gα co-operate as negative regulators of ABA signaling in root growth. GCR2 functions as a positive regulator of ABA signaling in seed germination and in expression of ABA marker genes (Liu et al. 2007). In ABA signaling, Arabidopsis Gα may serve as a negative regulator in seeds and roots and as a positive regulator in stomata. GCR1 and GCR2 seem to be negative and positive regulators of ABA signaling, respectively, in all tissues examined.

6.3

Gibberellin (GA)

The heterotrimeric G proteins appear to function as positive regulators in rice (Ueguchi-Tanaka et al. 2000; Bethke et al. 2006) and Arabidopsis (Ullah et al. 2002; Chen et al. 2004a) in GA signaling. GCR1 acts as a positive regulator of GA signaling independent of Arabidopsis Gα in some aspects of seed germination.

6.4

Brassinosteroid (BR)

The heterotrimeric G proteins seem to function as positive regulators of BR signaling in Arabidopsis (Chen et al. 2004a). GCR1 acts as a positive regulator of BR signaling independent of Arabidopsis Gα in some aspects of seed germination.

6.5

Blue Light

The accumulation of phenylalanine in response to blue light irradiation decreases in gcr1, gpa1 and pd1 mutants compared to WT (Warpeha et al. 2006). Arabidopsis Gα functions as a positive regulator of blue light signaling in cooperation with GCR1 and PD1.

II.2 Rice Heterotrimeric G Protein Signaling

6.6

145

Extracellular Calmodulin (ExtCaM)

Stomata do not respond to ExtCaM in the gpa1 mutant, and H2O2 production by ExtCaM is reduced in the gpa1 mutant (Chen et al. 2004b). ExtCaM may be involved in ABA signaling.

6.7

Sugar

AtRGS1 functions as a negative regulator of heterotrimeric G proteins (Chen et al. 2003). A mutation of the AtRGS1 protein in Arabidopsis which results in losing its ability to hydrolyze GTP bound to Gα produces a constitutively active form of the Gα. As a result, Atrgs1 would be expected to show responses opposite to those displayed by the gpa1 mutant after addition of signals. As expected, Atrgs1 shows an insensitive response for growth inhibition by glucose, whereas gpa1 shows a hypersensitive response for growth inhibition by glucose, compared to WT. Both Arabidopsis Gα and THF1 seem to be negative regulators in growth inhibition by sugar (Huang et al. 2006). Growth inhibition by addition of sugar is enhanced in thf1 compared to WT and it is little observed in transformants overexpressing THF1.

6.8

Unknown Signals

There is a tendency for cell number to increase in transformants overexpressing Gα (Ullah et al. 2001) and GCR1 (Apone et al. 2003). Such a tendency is also observed in Atrgs1 (Chen et al. 2003). Consequently, plant heterotrimeric G proteins may be involved in the regulation of cell division. These have been shown to regulate a mitogen-activated protein kinase at the posttranslational level (Lieberherr et al. 2005). They have also been observed to regulate PLD activity (Munnik et al. 1995) and Ca2+ channels (Aharon et al. 1998).

7

Further Perspectives

The number of molecular species of canonical heterotrimeric G proteins is very limited in plants in contrast to the G proteins of mammals. Nevertheless, a large number of signaling pathways have been proposed to be regulated by the plant G proteins. Recently, seven molecules interacting with Gα have been identified in Arabidopsis. The findings strongly support the idea that plant heterotrimeric G proteins serve many signaling systems as signal transmitters. The details of G-protein-mediated signaling will be elucidated by further analysis of molecules interacting with Gα.

146

Y. Iwasaki et al.

References Aharon GS, Gelli A, Snedden WA, Blumwald E (1998) Activation of a plant plasma membrane Ca2+ channel by TGα1, a heterotrimeric G protein α-subunit homologue. FEBS Lett 424:17–21 Apone F, Alyeshmerni N, Wiens K, Chalmers D, Chrispeels MJ, Colucci G (2003) The G-proteincoupled receptor GCR1 regulates DNA synthesis through activation of phosphatidylinositolspecific phospholipase C. Plant Physiol 133:571–579 Ashikari M, Wu J, Yano M, Sasaki T, Yoshimura A (1999) Rice gibberellin-insensitive dwarf mutant gene Dwarf 1 encodes the α-subunit of GTP-binding protein. Proc Natl Acad Sci USA 96:10284–10289 Bethke PC, Hwang YS, Zhu T, Jones RL (2006) Global patterns of gene expression in the aleurone of wild-type and dwarf1 mutant rice. Plant Physiol 140:484–498 Chen JG, Jones AM (2004) AtRGS1 function in Arabidopsis thaliana. Methods Enzymol 389:338–350 Chen JG, Willard FS, Huang J, et al. (2003) A seven-transmembrane RGS protein that modulates plant cell proliferation. Science 301:1728–1731 Chen JG, Pandey S, Huang J, et al. (2004a) GCR1 can act independently of heterotrimeric Gprotein in response to brassinosteroids and gibberellins in Arabidopsis seed germination. Plant Physiol 135:907–915 Chen YL, Huang R, Xiao YM, Lu P, Chen J, Wang XC (2004b) Extracellular calmodulin-induced stomatal closure is mediated by heterotrimeric G protein and H2O2. Plant Physiol 136:4096–4103 Coursol S, Fan LM, Le Stunff H, Spiegel S, Gilroy S, Assmann SM (2003) Sphingolipid signalling in Arabidopsis guard cells involves heterotrimeric G proteins. Nature 423:651–654 Fujisawa Y, Kato T, Ohki S, et al. (1999) Suppression of the heterotrimeric G protein causes abnormal morphology, including dwarfism, in rice. Proc Natl Acad Sci USA 96:7575–7580 Huang J, Taylor JP, Chen JG, et al. (2006) The plastid protein THYLAKOID FORMATION1 and the plasma membrane G-protein GPA1 interact in a novel sugar-signaling mechanism in Arabidopsis. The Plant Cell 18:1226–1238 Ikeda A, Ueguchi-Tanaka M, Sonoda Y, et al. (2001) slender rice, a constitutive gibberellin response mutant, is caused by a null mutation of the SLR1 gene, an ortholog of the heightregulating gene GAI/RGA/RHT/D8. The Plant Cell 13:999–1010 Ishikawa A, Tsubouchi H, Iwasaki Y, Asahi T (1995) Molecular cloning and characterization of a cDNA for the α subunit of a G protein from rice. Plant Cell Physiol 36:353–359 Ishikawa A, Iwasaki Y, Asahi T (1996) Molecular cloning and characterization of a cDNA for the β subunit of a G protein from rice. Plant Cell Physiol 37:223–228 Iwata M, Umemura K, Teraoka T, Usami H, Fujisawa Y, Iwasaki Y (2003) Role of the α subunit of heterotrimeric G-protein in probenazol-inducing defense signaling in rice. J Gen Plant Pathol 69:83–86 Jones AM, Assmann SM (2004) Plants: the latest model system for G-protein research. EMBO Rep 5:572–578 Joo JH, Wang S, Chen JG, Jones AM, Fedoroff NV (2005) Different signaling and cell death roles of heterotrimeric G protein α and β subunits in the Arabidopsis oxidative stress response to ozone. The Plant Cell 17:957–970 Kato C, Mizutani T, Tamaki H, et al. (2004) Characterization of heterotrimeric G protein complexes in rice plasma membrane. Plant J 38:320–331 Komatsu S, Yang G, Hayashi N, Kaku H, Umemura K, Iwasaki Y (2004) Alterations by a defect in a rice G protein α subunit in probenazole and pathogen-induced responses. Plant Cell Environ 27:947–957 Lambright DG, Noel JP, Hamm HE, Sigler PB (1994) Structural determinants for activation of the α-subunit of a heterotrimeric G protein. Nature 369:621–628

II.2 Rice Heterotrimeric G Protein Signaling

147

Lapik YR, Kaufman LS (2003) The Arabidopsis cupin domain protein AtPirin1 interacts with the G protein α-subunit GPA1 and regulates seed germination and early seedling development. The Plant Cell 15:1578–1590 Lease KA, Wen J, Li J, Doke JT, Liscum E, Walker JC (2001) A mutant Arabidopsis heterotrimeric G-protein β subunit affects leaf, flower, and fruit development. The Plant Cell 13:2631–2641 Lieberherr D, Thao NP, Nakashima A, Umemura K, Kawasaki T, Shimamoto K (2005) A sphingolipid elicitor-inducible mitogen-activated protein kinase is regulated by the small GTPase OsRac1 and heterotrimeric G-protein in rice 1[W]. Plant Physiol 138:1644–1652 Liu X, Yue Y, Li B, et al. (2007) A G protein-coupled receptor is a plasma membrane receptor for the plant hormone abscisic acid. Science 315:1712–1716 Ma H, Yanofsky MF, Meyerowitz EM (1990) Molecular cloning and characterization of GPA1, a G protein α subunit gene from Arabidopsis thaliana. Proc Natl Acad Sci USA 87:3821–3825 Mason MG, Botella JR (2000) Completing the heterotrimer: isolation and characterization of an Arabidopsis thaliana G protein γ-subunit cDNA. Proc Natl Acad Sci USA 97:14784–14788 Mishra G, Zhang W, Deng F, Zhao J, Wang X (2006) A bifurcating pathway directs abscisic acid effects on stomatal closure and opening in Arabidopsis. Science 312:264–266 Munnik T, Arisz SA, De Vrije T, Musgrave A (1995) G Protein activation stimulates phospholipase D signaling in plants. The Plant Cell 7:2197–2210 Neer EJ, Schmidt CJ, Nambudripad R, Smith TF (1994) The ancient regulatory-protein family of WD-repeat proteins. Nature 371:297–300 Oki K, Fujisawa Y, Kato H, Iwasaki Y (2005) Study of the constitutively active form of the α subunit of rice heterotrimeric G proteins. Plant Cell Physiol 46:381–386 Pandey S, Assmann SM (2004) The Arabidopsis putative G protein-coupled receptor GCR1 interacts with the G protein α subunit GPA1 and regulates abscisic acid signaling. The Plant Cell 16:1616–1632 Simon MI, Strathmann MP, Gautam N (1991). Diversity of G proteins in signal transduction. Science 252:802–808 Suharsono U, Fujisawa Y, Kawasaki T, Iwasaki Y, Satoh H, Shimamoto K (2002) The heterotrimeric G protein α subunit acts upstream of the small GTPase Rac in disease resistance of rice. Proc Natl Acad Sci USA 99:13307–13312 Ueguchi-Tanaka M, Fujisawa Y, Kobayashi M, et al. (2000) Rice dwarf mutant d1, which is defective in the α subunit of the heterotrimeric G protein, affects gibberellin signal transduction. Proc Natl Acad Sci USA 97:11638–11643 Ullah H, Chen JG, Young JC, Im KH, Sussman MR, Jones AM (2001) Modulation of cell proliferation by heterotrimeric G protein in Arabidopsis. Science 292:2066–2069 Ullah H, Chen JG, Wang S, Jones AM (2002) Role of a heterotrimeric G protein in regulation of Arabidopsis seed germination. Plant Physiol 129:897–907 Viehweger K, Schwartze W, Schumann B, Lein W, Roos W (2006) The Gα protein controls a pHdependent signal path to the induction of phytoalexin biosynthesis in Eschscholzia californica. The Plant Cell 18:1510–1523 Wang XQ, Ullah H, Jones AM, Assmann SM (2001) G protein regulation of ion channels and abscisic acid signaling in Arabidopsis guard cells. Science 292:2070–2072 Warpeha KM, Hamm HE, Rasenick MM, Kaufman LS (1991) A blue-light-activated GTP-binding protein in the plasma membranes of etiolated peas. Proc Natl Acad Sci USA 88:8925–8929 Warpeha KM, Lateef SS, Lapik Y, Anderson M, Lee BS, Kaufman LS (2006) G-protein-coupled receptor 1, G-protein Gα-subunit 1, and prephenate dehydratase 1 are required for blue lightinduced production of phenylalanine in etiolated Arabidopsis. Plant Physiol 140:844–855 Warpeha KM, Upadhyay S, Yeh J, et al. (2007) The GCR1, GPA1, PRN1, NF-Y signal chain mediates both blue light and abscisic acid responses in Arabidopsis. Plant Physiol 143:1590–1600

148

Y. Iwasaki et al.

Weiss CA, Garnaat CW, Mukai K, Hu Y, Ma H (1994) Isolation of cDNAs encoding guanine nucleotide-binding protein β-subunit homologues from maize (ZGB1) and Arabidopsis (AGB1). Proc Natl Acad Sci USA 91:9554–9558 Willard FS, Siderovski DP (2004) Purification and in vitro functional analysis of the Arabidopsis thaliana regulator of G-protein signaling-1. Methods Enzymol 389:320–338 Zhao J, Wang X (2004) Arabidopsis phospholipase Dα1 interacts with the heterotrimeric G-protein α-subunit through a motif analogous to the DRY motif in G-protein-coupled receptors. J Biol Chem 279:1794–1800

II.3

Genetic Control of Embryogenesis in Rice Yutaka Sato1(* ü)

1

Introduction

The fundamental body organization in higher plants is established during embryogenesis, although most morphogenetic events occur after embryogenesis (Jürgens et al. 1994; Meinke 1995). Like all of the sexually reproducing organisms, vascular plants begin their existence as a single cell, the fertilized egg or zygote. This cell proliferates to become an embryo with differentiating organs and tissues. During the early stages of embryogenesis, several body axes, which form the basis for apical–basal and radial patterns, are formed. In monocotyledonous plants such as rice, the shape of the embryo is not radially symmetrical (Fig. 1). This leads to the existence of the third axis, the dorsiventral axis. In the embryo, the region where the shoot develops is defined as the ventral side, while the opposite side then becomes the dorsal side. The shoot apical meristem (SAM) is a center of morphogenesis in plants, as it produces most of the above-ground parts, including the leaves, stems, and axillary buds; the other type of meristematic tissue, the root apical meristem (RAM), generates the below-ground parts (Steeves and Sussex 1989). The SAM and RAM are first formed during early embryogenesis at fixed positions based on positional information defined by the three polarized axes, which determine the basic body organization. In comparison with other taxa, grasses produce a morphologically well-developed embryo. For example, in contrast to the embryo of Arabidopsis that does not develop foliage leaves during embryogenesis, the shoot structure of grasses develops in the mature embryo. That is, the SAM in the grass embryo produces primordia of several foliage leaves during embryogenesis. In the mature rice embryo, there are three primordia of foliage leaves, whereas in maize there are five. Thus, the grass embryo has incorporated vegetative growth within embryogenesis (Freeling 1992; Asai et al. 2002). This heterochoronic character may have been introduced by a mutation

1

Graduate School of Bioagricultural Sciences, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8601, Japan e-mail: [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

149

150

Y. Sato

Fig. 1 Morphology of a mature rice embryo. a Median longitudinal section through a mature rice embryo at 9 days after pollination (DAP). P1, youngest primordium; P2, second youngest primordium; RAM, root apical meristem; SAM, shoot apical meristem. The ventral side of the embryo faces towards left side of image. b Ventral side view of a mature rice embryo at 9 DAP obtained by means of scanning electron microscopy. Bars indicate 100 µm

during evolution of the grass family. There are another characteristics unique to grass embryos. For example, the SAM forms at a lateral position of the scutellum, which corresponds to a single cotyledon, whereas the SAM in dicotyledonous plants develops between two cotyledons. The outermost cell layer of the scutellum is called the scutellar epithelium and is responsible for the transportation of nutrients from the endosperm to the embryo. The coleoptile is the first leaf-like organ that forms during embryogenesis, and surrounds the embryonic shoot. The coleoptile is believed to originate from a primordium common to that which produces the scutellum, but the origin of the coleoptile remains controversial. The epiblast, which develops on the opposite side of the embryo from the scutellum and below the coleoptile, is unique to the embryos of some grass species such as rice and wheat, but is not seen in that of maize, and the function of the epiblast is not known (Fig. 1). Embryogenesis in rice proceeds much more quickly than in other cereals such as maize and barley. Most of the morphogenetic events are completed by 9 days after pollination (DAP) under normal conditions. The first cell division of the fertilized egg occurs asymmetrically within 12 h after fertilization, producing apical and basal cells. Unlike embryogenesis in Arabidopsis, the plane of the first cell division of the rice zygote is not always perpendicular to the apical–basal axis, but rather it occurs at an oblique angle. The apical and basal cells develop into the embryo and a suspensor, respectively. In Arabidopsis, the pattern of cell division during the

Fig. 2 Developmental course of embryogenesis in wild-type rice. Embryo age (days after pollination, DAP): a 1 DAP, b 2 DAP, c 4 DAP, d 5 DAP, e 7 DAP. Bars in a and b indicate 50 µm. Bars in c and d indicate 100 µm. Bar in e indicates 200 µm. Arrow in c indicates position of developing coleoptile, and white arrowhead indicates position of SAM. Black arrowheads in d and e indicate position of RAM. f A confocal image of the embryo at 2 DAP. Fixed ovules were stained with propidium iodide. After dehydration in a graded ethanol series, the ovule was transferred to methyl salicylate to make its tissues translucent and was then observed by confocal scanning microscopy. Embryo ages: g)3 DAP, h 4 DAP, i 6 DAP. In g to i, fixed ovules were dehydrated in a graded ethanol series, then transferred to methyl salicylate and observed by light microscopy through Nomarsky optics. j Schematic representation of the expression patterns of OSH1, ROC1, OsSCR and RAmy1A

152

Y. Sato

early stages of embryogenesis is stereotyped. In rice, however, the pattern seems irregular, and the direction of cell division is difficult to predict. Nonetheless, the appearance of various organs and tissues during embryogenesis follows a predictable pattern. After fertilization, the rice embryo takes on a globular shape through irregular cell divisions (Fig. 2a, b, f, g). In the early globular stage (2 DAP), embryos are composed of about 100 cells and are about 50 µm in diameter (Fig. 2b, g). The globular stage lasts until nearly 3 DAP. Although organogenesis is not observed before 3 to 4 DAP, a dorsiventral polarity is evident, judging from the gradual change in cell size during the late globular stage, from 2 to 3 DAP. The first morphological differentiation is observed at the ventral side as a protrusion of the coleoptile primordium at 4 DAP (Fig. 2c), when the embryo reaches 100 to 110 µm in length along the apical–basal axis and comprises 800 to 900 cells, then the SAM becomes visible just beneath the coleoptile primordium (Fig. 2c, h). The shoot and radicle apices are first observed at 4 DAP. Primordia of the first to third foliage leaves form successively between 5 and 9 DAP, and exhibit alternate phyllotaxis. Differentiation of the vascular tissues connecting the shoot, radicle, and scutellum is also visible after 5 DAP (Fig. 2d, i). By 9 DAP, most morphogenetic events in the embryo are complete and the embryo prepares for dormancy by activating genes required for seed maturation or dormancy. A developmental staging system for the course of rice embryogenesis has been proposed in the review written by Itoh et al. (2005).

2

Molecular Markers Used in the Analysis of Rice Embryogenesis

The developmental course of rice embryogenesis has been precisely documented using molecular markers or marker genes that can identify specific organs or tissues before their morphological differentiation (that is, before they become morphologically visible). These markers are also useful in determining the activity of cells with a specific function or the activity of cell division. One of the most useful markers is the OSH1 (Oryza sativa homeobox1) gene (Matsuoka et al. 1993; Sato et al. 1996). OSH1 encodes a KNOTTED1-like homeobox gene. Since OSH1 is expressed in cells of indeterminate nature in the region around the SAM, OSH1 is used as a marker of indeterminate cells in the SAM. It is also expressed in the SAM region during embryogenesis, and the onset of its expression during embryogenesis has been detected in the presumptive SAM region at the globular stage that exists at 2 DAP. Thus, OSH1 is useful in determining whether SAM cells with indeterminate character are becoming differentiated. Similarly, the differentiation of the RAM during embryogenesis can be identified by the expression of the QHB (quiescentcenter-specific homeobox) gene. QHB encodes a WUSCHEL-type homeobox gene and is expressed at the quiescent center of the RAM. Its expression during embryogenesis is observed at 2 to 3 DAP at the position where RAM subsequently develops (Kamiya et al. 2003a).

II.3 Genetic Control of Embryogenesis in Rice

153

In addition to markers that identify the differentiation of cells in particular organs, some markers are used to visualize regional identities or specific cell types in the embryo. The ROC1 gene (rice outermost cell layer–specific gene), which encodes a glabra2-type homeobox gene, is expressed in the protoderm cells during embryogenesis (Ito et al. 2002). The differentiation of a morphologically distinguishable protoderm with a layered structure is visible at around 2 to 3 DAP, but ROC1 expression begins just after fertilization. Thus, ROC1 is a good molecular marker for protoderm differentiation during embryogenesis. Another molecular marker that demarcates the specific domain of the epidermal layer is RAmy1A (rice a-amylase 1 gene) (Sugimoto et al. 1998; Miyoshi et al. 1999). RAmy1A is expressed at the dorsal side of the scutellar epithelium, facing the endosperm tissue. This gene is thus useful in visualizing the formation of the dorsiventral axis and the development and functionality of the scutellum. Although not many molecular markers are available to demarcate the cellular differentiation or tissue types of the inner part of the embryo, OsSCR, which encodes a homologue of the SCARECROW gene in Arabidopsis, is a useful marker to visualize the differentiation of tissues along with the radial axis at the early stage of embryogenesis (Kamiya e al. 2003b). Expression patterns of OSH1, ROC1, RAmy1A and OsSCR during embryogenesis are summarized in Fig. 2j. In addition to the markers mentioned above, a few more molecular markers such as OSH15 and OsPNH are used in the analysis of rice embryogenesis.

3

Analysis of Organogenesis During Embryogenesis

Understanding the mechanisms responsible for organogenesis is one of the major topics in biology. Organogenesis in plants occurs both during and after embryogenesis. The SAM is a center of organogenesis during post-embryonic development, as it produces most of the plant’s above-ground parts, including leaves, stems, and axillary buds (Steeves and Sussex 1989). The SAM continuously differentiates organs while maintaining itself in an indeterminate state. Thus, the SAM behaves as a population of indeterminate stem cells in plants. In Arabidopsis, key regulators for organizing stem cells, such as WUSCHEL (Mayer et al. 1998) and PLETHORA (Aida et al. 2004), have been identified. These regulators establish the functional SAM and RAM and confirm their pluripotent nature. The SAM forms first during early embryogenesis, at fixed positions based on positional information provided by the polarized axes, which specify the basic organization of the plant’s body. Embryonic SAM formation is an important developmental process that ensures post-embryonic development of the shoot architecture, and is the initial process that provides the local cellular environment for maintenance of stem cells. Thus, understanding the mechanism of embryonic SAM formation is a primary issue in the developmental biology of plants, and doing so provides a unique system for studying stem cell establishment.

154

Y. Sato

Genetic and molecular mechanisms governing embryonic SAM formation have been investigated extensively, and several key genetic components have been revealed. Class 1 KNOX genes are involved in embryonic SAM formation and maintenance (Reiser et al. 2000; Takada and Tasaka 2002). In several species, class 1 KNOX genes such as KNOTTED1 (KN1) in maize (Smith et al. 1995), OSH1 in rice (Sato et al. 1996), and SHOOT MERISTEMLESS (STM) in Arabidopsis (Long et al. 1996) are expressed in the presumptive SAM region of the early embryo, and their expression is maintained in indeterminate cells of the post-embryonic SAM (Jackson et al. 1994; Long et al. 1996; Sentoku et al. 1999). A strong mutant allele of the STM gene causes an embryonic phenotype that lacks SAM. Weak stm alleles can produce post-embryonic SAM but produce several defects in organ initiation (Clark et al. 1996; Endrizzi et al. 1996). Thus, STM is required for both embryonic SAM initiation and post-embryonic SAM maintenance in Arabidopsis. Genetic analysis of the functionally redundant genes CUP-SHAPED COTYLEDON1 (CUC1) and CUC2 in Arabidopsis has revealed that they are essential for embryonic SAM formation and act upstream of STM (Aida et al. 1997, 1999). These CUC genes encode NAC-domain proteins, such as the NO APICAL MERISTEM gene in petunia. The cuc1 and cuc2 double mutants and the nam mutant fail to form the embryonic SAM in both plants (Souer et al. 1996; Aida et al. 1997). In monocotyledonous species, several genes associated with embryonic SAM formation have been reported (Satoh et al. 1999; Pilu et al. 2002). In rice, the shootless (shl) mutants derived from four loci (SHL1 to SHL4) were originally identified as embryonic mutants that completely lack a SAM (Satoh et al. 1999). Except in shl3, a normal scutellum and radicle differentiate in the shl mutants, indicating that SHL genes function specifically in the establishment of the SAM-associated region but are not involved in embryonic axialization or the development of other embryonic organs (Satoh et al. 1999). Further analysis has provided some insights into the functions of the SHL genes. A series of alleles of shl1, shl2 and shl4 produce various degrees of SAM-deficient phenotype (Fig. 3a). The strongest alleles are associated with complete loss of SAM in the mature embryo, whereas in mutants with weaker alleles, the SAM is frequently formed and produces morphologically abnormal leaf primordia (Satoh et al. 2003). It is notable that the embryonic phenotypes of the weak shl1 and shl2 alleles are very similar to those of the shoot organization (sho) mutants, which were initially identified as a mutant with defects in the maintenance of SAM organization and the development of leaf primordia (Itoh et al. 2000; Satoh et al. 2003). This phenotypic resemblance suggests that SHL and SHO genes act in both the initiation and the maintenance of SAM. However, the molecular functions of these genes have not been specified. REVOLUTA (REV), PHABULOSA (PHB) and PHAVOLUTA (PHV), which encode class III homeodomain-leucine zipper (HD-ZIPIII) proteins in Arabidopsis, also play a role in embryonic SAM formation. Although a single loss-of-function mutation in any one of these genes permits a normal embryonic SAM, the rev phb double and the rev phb phv triple mutants are defective in SAM formation (Emery et al. 2003; Prigge et al. 2005). Thus, these genes may function redundantly in SAM initiation during embryogenesis. In addition, they are involved in the development of

II.3 Genetic Control of Embryogenesis in Rice

155

Fig. 3 Morphology of mutant rice embryos with defective organ differentiation. a A mature embryo of the shootless 4 (shl4) mutant. =Bar indicates 100 µm. b, d Mature embryos of the globular embryo 1 (gle1) mutant. Bar indicates 50 µm. c, e Mature embryos from the gle4 mutant. Bar indicates 100 µm

organ polarity and in vascular development. For example, rev phb phv triple mutants produce a single abaxialized radial cotyledon. Conversely, a gain-of-function mutation in any of these genes results in increased adaxial identities in the abaxial region of lateral organs and defects in vascular polarity (McConnell et al. 2001; Emery et al. 2003; Zhong and Ye 2004). Expressions of HD-ZIPIII genes are controlled by a class of microRNA (miRNA), miR165/ miR166 (Rhoades et al. 2002). The miRNAs are endogenous 20- to 22-nucleotide single-stranded RNAs that are important regulators of animal and plant development (Kidner and Martienssen 2005a; Plasterk 2006). The miRNAs are processed from the double-stranded hairpin structure of their precursors by DICER-like ribonuclease III (DCL) and are incorporated into an RNA-induced silencing complex, leading to cleavage of the target mRNA or repression of protein translation by the activities of ARGONAUTE (AGO) (Bartel 2004). Mutations in DCL1 and AGO1 of Arabidopsis cause wide-ranging developmental defects and result in polarity defects in lateral organs (Bohmert et al. 1998; Jacobsen et al. 1999; Schauer et al. 2002; Hunter and Poethig 2003; Kidner and Martienssen 2005b). This phenotype is partially explained by misregulation of HD-ZIPIII genes through an miR165/miR166-mediated process (Kidner and Martienssen 2004). In addition, single-nucleotide substitutions in the miR165/miR166 recognition sequence of REV, PHB and PHV of Arabidopsis and ROLLED LEAF1 (RLD1) of maize allow their mRNAs to escape microRNA-directed degradation, leading to

156

Y. Sato

organ adaxialization (McConnell et al. 2001; Bowman et al. 2002; Emery et al. 2003; Juarez et al. 2004; Zhong and Ye 2004). These studies have revealed that regulation by miR165/miR166 is necessary for the proper expression of HD-ZIPIII genes, which confer adaxial identities on lateral organs. However, the contribution of miR165/miR166 to the regulation of HD-ZIPIII genes in SAM formation during embryogenesis has not been clarified. In multicellular organisms, precise control of the spatio-temporal expression of regulatory genes is required for each developmental event, including organ formation, tissue differentiation and the determination of cell fate. Recent studies on the control of the expression of these genes have revealed that they are regulated in part by post-transcriptional control through miRNAs (Dugas and Bartel 2004). However, the regulatory mechanism of miRNA expression is still unknown in both plants and animals. Although several genes such as AGO1, SERRATE (SER), and ASYMMETRIC LEAVES1 (AS1) and AS2 affect the localization or accumulation level of miRNA (Kidner and Martienssen 2004; Grigg et al. 2005; Li et al. 2005; Xu et al. 2006), how these genes regulate the localization or amount of miRNA has not been elucidated. Because the expression of both HD-ZIPIII genes and miR166 in shl mutants of rice are misregulated, it is possible that SHL genes act upstream of the networks of transcriptional regulators, and that their post-transcriptional regulation specifies the initiation of SAM during embryogenesis (Nagasaki et al. 2007). Map-based cloning of SHO1, SHL2 and SHL4 genes revealed that they encode homologues of DICER, RNA-dependent RNA POLYMERASE and ARGONAUTE, respectively. Because these genes encode proteins related to the RNAi interference (RNAi) pathway, it is strongly suggested that an RNAi-like mechanism operates during the embryogenesis in rice, and that small RNAs such as short interfering RNA (siRNA) or microRNA (miRNA) may work as a signaling molecule to induce SAM initiation through the regulation of target genes, such as miR166 and HDZIPIII genes. Thus, SHL/ SHO genes are the upstream regulators of transcriptional regulators that regulate SAM initiation. So far, several mutations in genes related to the RNAi pathway have been reported in Arabidopsis. Some of them showed developmental defects, but their phenotypes are not related to SAM initiation during embryogenesis, indicating that the process of siRNA- or miRNA-dependent shoot development operates in rice, but not in Arabidopsis. Thus, the analysis of rice shl/ sho mutants has elucidated a new layer of regulation of HD-ZIPIII genes in SAM initiation during embryogenesis (Nagasaki et al. 2007). The mechanism of SAM development during embryogenesis has been analyzed by examining a series of mutations that affect SAM initiation, maintenance or function. Most of these mutants have specific defects in the development of SAM, although pleiotropic effects have been observed in some cases. On the other hand, certain mutations specifically delete the radicle during embryogenesis in rice. These mutants are designated radicleless (ral), and there are at least two loci, ral1 and ral2 (Hong et al. 1995). Thus, organogenesis of two of the essential parts that guarantee post-embryonic development is regulated, at least in part, by independent genetic mechanisms during embryogenesis. However, other mutations affect the development of both SAM and RAM. These mutations include globular embryo

II.3 Genetic Control of Embryogenesis in Rice

157

(gle) mutants (Hong et al. 1995), in which embryogenesis is arrested at the globular stage or later. In the latter case, the embryo continues to grow in a globular shape and does not undergo the differentiation of discrete organs (Fig. 3b–e). This type of mutation is relatively frequently observed among embryogenesis-defective mutants in rice, suggesting that the organogenesis is regulated by multiple loci. The globular-shaped embryo could be produced by mutation in genes regulating both developmental processes and cellular viability. To analyze these mutants, however, it is important to distinguish the genes essential for early embryogenesis from the rest of genes. In some mutants, it is possible that embryogenesis is arrested because of mutations in housekeeping genes that are not directly involved in organogenesis. Because of difficulties in distinguishing between these two possibilities, mutations that would result in arrested embryogenesis at the globular stage have not been extensively analyzed in Arabidopsis. In rice, however, the situation is different. Rice seeds contain an embryo and a well-developed endosperm. If mutations occur in housekeeping genes, normal development should be disturbed in both the embryo and the endosperm. Thus, collecting mutants that specifically show an embryo phenotype but no endosperm phenotype excludes the possibility of mutations in housekeeping genes and suggests that the causal genes are specifically involved in embryo development. There are at least four gle loci in the rice genome, and all mutations at these loci result in globular embryos with normal endosperm development, suggesting that the causal genes are not housekeeping genes but rather genes required for the initiation or progression of embryogenesis. Among the gle loci, gle4 has been most extensively analyzed (Kamiya et al. 2003b). A gle4 mutant embryo (Fig. 3c, e) continues to grow in a globular shape without generating SAM and RAM. Thus, the mature gle4 embryo becomes much larger than an embryo at the globular stage in wild-type rice. This observation also supports a hypothesis that GLE4 is not involved in fundamental cellular activities, such as cell division or the metabolism mediated by housekeeping genes, but rather is involved in developmental regulation. The analysis of organ differentiation in gle4 embryos using several molecular markers such as ROC1 and RAmy1A, has revealed that the epidermal layer, including the scutellar epithelium, is differentiated. However, the expression of OsSCR is perturbed, suggesting that the differentiation of tissues along with the radial axis is defective. Thus, gle4 mutants seem to be defective in the establishment of the radial axis. Recently, our research group has succeeded in identifying gle4 by means of transposon tagging (unpublished data). The gene was found to encode a member of the mitogen-activated protein kinase (MAPK) group. Because MAPKs are often involved in the transduction of signals from extracellular spaces, thereby regulating gene expression, it is possible that cell–cell communication plays an important role in the establishment of the radial axis, and that GLE4 (MAPK) mediates the transduction of the information provided by the extracellular signals into the nucleus and modulates gene expression. Thus far, it is known that the auxin group of phytohormones play an essential role in the establishment of the apical–basal axis as a signaling molecule (Willemsen and Scheres 2004). However, the mechanism of the establishment of the radial axis is not well understood. Therefore, the understanding

158

Y. Sato

of a signaling pathway that involves GLE4 MAPK may provide important information to elucidate the mechanism responsible for establishment of the radial axis.

4

Future Perspectives

The mechanisms that govern embryogenesis are a major topic in the field of plant developmental biology. Since about 1970, genetic approaches to dissecting the process of embryogenesis based on the isolation and study of mutants with defective embryogenesis have been used in various plant species, including Arabidopsis, maize and rice (Clark and Sheridan 1991; Mayer et al. 1991; Patton et al. 1991; Hong et al. 1995). Recent advances in molecular biology and molecular genetics have elucidated the genetic regulatory systems of part of this process, as discussed in this chapter. In Arabidopsis, a series of mutations that affect responses to and the transport of auxin have revealed that this phytohormone plays an essential role in the development of an apical–basal polarity during embryogenesis. However, little is known about the mechanism or mechanisms that govern the development of a radial pattern during embryogenesis in Arabidopsis. By contrast, in rice, the gle4 mutants have provided a means to study the genetic components of the formation of this radial pattern. Thus, analysis of the mutants with defective embryogenesis from various plant species will help us to gather the pieces involved in the genetic regulatory systems that operate during plant embryogenesis and will lead to a more comprehensive model of the initiation of plant development. In rice, an extensive collection of mutants with defective embryogenesis has been made (Hong et al. 1995). Thus, future analysis of these mutants will reveal the novel genetic components or pathways that operate during embryogenesis.

References Aida M, Ishida T, Fukaki H, Fujisawa H, Tasaka M (1997) Genes involved in organ separation in Arabidopsis: an analysis of the cup-shaped cotyledon mutant. Plant Cell 9:841–857 Aida M, Ishida T, Tasaka M (1999) Shoot apical meristem and cotyledon formation during Arabidopsis embryogenesis: interaction among the CUP-SHAPED COTYLEDON and SHOOT MERISTEMLESS genes. Development 126:1563–1570 Aida M, Beis D, Heidstra R, et al. (2004) The PLETHORA genes mediate patterning of the Arabidopsis root stem cell niche. Cell 119:109 Asai K, Satoh N, Sasaki H, Satoh H, Nagato Y (2002) A rice heterochronic mutant, mori1, is defective in the juvenile–adult phase change. Development 129:265–273 Bartel DP (2004) MicroRNAs: genomics, biogenesis, mechanism, and function. Cell 116:281–297 Bohmert K, Camus I, Bellini C, Bouchez D, Caboche M, Benning C (1998) AGO1 defines a novel locus of Arabidopsis controlling leaf development. EMBO J 17:170–180 Bowman JL, Eshed Y, Baum SF (2002) Establishment of polarity in angiosperm lateral organs. Trends Genet 18:134–141

II.3 Genetic Control of Embryogenesis in Rice

159

Clark JK, Sheridan WF (1991) Isolation and characterization of 51 embryo-specific mutations of maize. Plant Cell 3:935–951 Clark SE, Jacobsen SE, Levin JZ, Meyerowitz EM (1996) The CLAVATA and SHOOT MERISTEMLESS loci competitively regulate meristem activity in Arabidopsis. Development 122:1567–1575 Dugas DV, Bartel B (2004) MicroRNA regulation of gene expression in plants. Curr Opin Plant Biol 7:512–520 Emery JF, Floyd SK, Alvarez J, et al. (2003) Radial patterning of Arabidopsis shoots by class III HD-ZIP and KANADI genes. Curr Biol 13:1768–1774 Endrizzi K, Moussian B, Haecker A, Levin JZ, Laux T (1996) The SHOOT MERISTEMLESS gene is required for maintenance of undifferentiated cells in Arabidopsis shoot and floral meristems and acts at a different regulatory level than the meristem genes WUSCHEL and ZWILLE. Plant J 10:967–979 Freeling M (1992) A conceptual framework for maize leaf development. Dev Biol 153:44–58 Grigg SP, Canales C, Hay A, Tsiantis M (2005) SERRATE coordinates shoot meristem function and leaf axial patterning in Arabidopsis. Nature 437:1022–1026 Hong SK, Aoki T, Kitano H, Satoh H, Nagato Y (1995) Phenotypic diversity of 188 rice embryo mutants. Dev Genet 16:298–310 Hunter C, Poethig RS (2003) miSSING LINKS: miRNAs and plant development. Curr Opin Genet Dev 13:372–378 Ito M, Sentoku N, Nishimura A, Hong SK, Sato Y, Matsuoka M (2002) Position dependent expression of GL2-type homeobox gene, Roc1: significance for protoderm differentiation and radial pattern formation in early rice embryogenesis. Plant J 29:497–507 Itoh JI, Kitano H, Matsuoka M, Nagato Y (2000) SHOOT ORGANIZATION genes regulate shoot apical meristem organization and the pattern of leaf primordium initiation in rice. Plant Cell 12:2161–2174 Itoh JI, Nonomura KI, Ikeda K, et al. (2005) Rice plant development: from zygote to spikelet. Plant Cell Physiol 46:23–47 Jackson D, Veit B, Hake S (1994) Expression of maize KNOTTED1 related homeobox genes in the shoot apical meristem predicts patterns of morphogenesis in the vegetative shoot. Development 120:405–413 Jacobsen SE, Running MP, Meyerowitz EM (1999) Disruption of an RNA helicase/RNAse III gene in Arabidopsis causes unregulated cell division in floral meristems. Development 126:5231–5243 Juarez MT, Kui JS, Thomas J, Heller BA, Timmermans MC (2004) microRNA-mediated repression of rolled leaf1 specifies maize leaf polarity. Nature 428:84–88 Jürgens G, Torres-Ruiz RA, Berleth T (1994) Embryonic pattern formation in flowering plants. Annu Rev Genet 28:351–371 Kamiya N, Nagasaki H, Morikami A, Sato Y, Matsuoka M (2003a) Isolation and characterization of a rice WUSCHEL-type homeobox gene that is specifically expressed in the central cells of a quiescent center in the root apical meristem. Plant J 35:429–441 Kamiya N, Nishimura A, Sentoku N, et al. (2003b) Rice globular embryo 4 (gle4) mutant is defective in radial pattern formation during embryogenesis. Plant J 44:875–883 Kidner CA, Martienssen RA (2004) Spatially restricted microRNA directs leaf polarity through ARGONAUTE1. Nature 428:81–84 Kidner CA, Martienssen RA (2005a) The developmental role of microRNA in plants. Curr Opin Plant Biol 8:38–44 Kidner CA, Martienssen RA (2005b) The role of ARGONAUTE1 (AGO1) in meristem formation and identity. Dev Biol 280:504–517 Li H, Xu L, Wang H, et al. (2005) The putative RNA-dependent RNA polymerase RDR6 acts synergistically with ASYMMETRIC LEAVES1 and 2 to repress BREVIPEDICELLUS and microRNA165/166 in Arabidopsis leaf development. Plant Cell 17:2157–2171 Long JA, Moan EI, Medford JI, Barton MK (1996) A member of the KNOTTED class of homeodomain proteins encoded by the STM gene of Arabidopsis. Nature 379:66–69

160

Y. Sato

Matsuoka M, Ichikawa H, Saito A, Tada Y, Fujimura T, Kano-Murakami Y (1993) Expression of a rice homeobox gene cause altered morphology of transgenic plants. Plant Cell 5:1039–1048 Mayer KF, Schoof H, Haecker A, Lenhard M, Jürgens G, Laux T (1998) Role of WUSCHEL in regulating stem cell fate in the Arabidopsis shoot meristem. Cell 95:805–815 Mayer U, Torres-Ruis R, Berleth T, Misera S, Jürgens G (1991) Mutations affecting body organization in the Arabidopsis embryo. Nature 353:402–407 McConnell JR, Emery J, Eshed Y, Bao N, Bowman J, Barton MK (2001) Role of PHABULOSA and PHAVOLUTA in determining radial patterning in shoots. Nature 411:709–713 Meinke DW (1995) Molecular genetics of plant embryogenesis. Annu Rev Plant Phys Plant Mol Biol 46:369–394 Miyoshi K, Nakata E, Nagato Y, Hattori T (1999) Differential in situ expression of three ABAregulated genes of rice, RAB16A, REG2 and OSBZ8, during seed development. Plant Cell Physiol 40:443–447 Nagasaki H, Itoh JI, Hayashi K, et al. (2007) The small interfering RNA production pathway is required for shoot meristem initiation in rice. Proc Natl Acad Sci (in press) Patton DA, Franzmann LH, Meinke DW (1991) Mapping genes essential for embryo development in Arabidopsis thaliana. Mol Gen Genet 227:337–347 Pilu R, Consonni G, Busti E, et al. (2002) Mutations in two independent genes lead to suppression of the shoot apical meristem in maize. Plant Physiol 128:502–511 Plasterk RH (2006) Micro RNAs in animal development. Cell 124:877–881 Prigge MJ, Otsuga D, Alonso JM, Ecker JR, Drews GN, Clark SE (2005) Class III homeodomainleucine zipper gene family members have overlapping, antagonistic, and distinct roles in Arabidopsis development. Plant Cell 17:61–76 Reiser L, Sánchez-Baracaldo P, Hake S (2000) Knots in the family tree: evolutionary relationships and functions of knox homeobox genes. Plant Mol Biol 42:151–166 Rhoades MW, Reinhart BJ, Lim LP, Burge CB, Bartel B, Bartel DP (2002) Prediction of plant microRNA targets. Cell 110:513–520 Sato Y, Hong SK, Tagiri A, et al. (1996) A rice homeobox gene, OSH1, is expressed before organ differentiation in a specific region during early embryogenesis. Proc Natl Acad Sci USA 93:8117–8122 Satoh N, Hong SK, Nishimura A, Matsuoka M, Kitano H, Nagato Y (1999) Initiation of shoot apical meristem in rice: characterization of four SHOOTLESS genes. Development 126:3629–3636 Satoh N, Itoh J, Nagato Y (2003) The SHOOTLESS2 and SHOOTLESS1 genes are involved in both initiation and maintenance of the shoot apical meristem through regulating the number of indeterminate cells. Genetics 164:335–346 Schauer SE, Jacobsen SE, Meinke DW, Ray A (2002) DICER-LIKE1: blind men and elephants in Arabidopsis development. Trends Plant Sci 7:487–491 Sentoku N, Sato Y, Kurata N, Ito Y, Kitano H, Matsuoka M (1999) Regional expression of the rice KN1-type homeobox gene family during embryo, shoot, and flower development. Plant Cell 11:1651–1664 Smith LG, Jackson D, Hake S (1995) Expression of knotted1 marks shoot meristem formation during maize embryogenesis. Dev Genet 16:344–348 Souer E, van Houwelingen A, Kloos D, Mol J, Koes R (1996) The no apical meristem gene of Petunia is required for pattern formation in embryos and flowers and is expressed at meristem and primordia boundaries. Cell 85:159–170 Steeves TA, Sussex IM (1989) Patterns in plant development. Cambridge University Press, Cambridge, Massachusetts Sugimoto N, Takeda G, Nagato Y, Yamaguchi J (1998) Temporal and spatial expression of the α-amylase gene during seed germination in rice and barley. Plant Cell Physiol 39:323–333 Takada S, Tasaka M (2002) Embryonic shoot apical meristem formation in higher plants. J Plant Res 115:411–417

II.3 Genetic Control of Embryogenesis in Rice

161

Willemsen V, Scheres B (2004) Mechanisms of pattern formation in plant embryogenesis. Annu Rev Genet 38:587–614 Xu L, Yang L, Pi L, et al. (2006) Genetic interaction between the AS1–AS2 and RDR6–SGS3– AGO7 pathways for leaf morphogenesis. Plant Cell Physiol 47:853–863 Zhong R, Ye ZH (2004) amphivasal vascular bundle 1, a gain-of-function mutation of the IFL1/ REV gene, is associated with alterations in the polarity of leaves, stems and carpels. Plant Cell Physiol 45:369–385

II.4

Photoperiodic Flowering in Rice Takeshi Izawa1(* ü)

1

Introduction

Rice (Oryza sativa L.) is a staple food for humans, especially in Asian countries, and it feeds about half of the world’s population. During domestication and recent breeding programs, flowering-time responses have diversified in rice due to adaptation to cultivation styles in local areas. For instance, early-flowering and photoperiod-insensitive cultivars have been developed for cultivation in northern Japan. In contrast, to prolong vegetative phases and increase yields, late-flowering cultivars with weak photoperiod sensitivity are preferred in some tropical areas, such as Taiwan. Garner and Allard (1920) discovered that many plants can recognize day-length in a given environment to determine the best flowering time to produce offspring efficiently. Recent progress in plant molecular biology has revealed molecular mechanisms of floral regulation in the photoperiodic flowering of Arabidopsis thaliana, a model organism of long-day plants. For instance, a small protein encoded by the FT floral-switch gene (Kardailsky et al. 1999; Kobayashi et al. 1999) was shown to be a long-distance mobile signal, so-called florigen (Corbesier et al. 2007). First, a group in Europe proposed that mRNA of FT may be the molecule underlying the florigen (Huang et al. 2005). Recently this work has been retracted by the principal investigator. Instead, it has been shown that FT protein moves to the apex, indicating that FT protein acts as a long-distance signal that induces flowering (Corbesier et al. 2007). In addition, some A. thaliana accessions exhibit a clear response to vernalization treatments. Recent work on vernalization has revealed that epigenetic regulation of a strong floral repressor gene termed FLOWERING-TIME LOCUS C (FLC) is a key regulator of vernalization in A. thaliana (Sung and Amasino 2005). Thus, progress is being made in studies of the molecular mechanisms underlying floral transition in A. thaliana.

1 National Institute of Agrobiological Sciences, Tsukuba, Japan, e-mail: [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

163

This page intentionally blank

164

T. Izawa

Rice is a short-day plant that may serve as a second model for photoperiodic flowering. Molecular genetic studies, especially on the genetics of natural variations detected by quantitative trait locus (QTL) analyses in rice cultivars (Yano 2001; Izawa et al. 2003), have revealed a diversity of molecules controlling floral transition in rice. In this chapter, I first summarize the research on the molecular mechanisms underlying photoperiodic flowering in rice in comparison with those revealed in A. thaliana. Second, I discuss some novel aspects that explain the diversity of flowering regulation at the molecular level.

2

The Importance of Photoreception for Photoperiodic Flowering in Rice

The photoperiod sensitivity 5 (se5) mutant was originally isolated from an M2 population of γ-ray-treated seeds (M0) of the japonica cultivar ‘Norin 8’ (Yokoo and Okuno 1993; Izawa et al. 2000). The se5 mutant plants exhibited markedly earlier flowering and complete photoperiod insensitivity. The brighter-green color of se5 plants compared to wild-type plants led us to analyze the photomorphogenesis of se5, because photoreception seemed to affect its pigment biosynthesis. Coleoptile (first juvenile leaf) elongation in response to various light pulses revealed that the se5 mutant was a phytochrome-deficient mutant and might have some defects in chromophore biosynthesis (Izawa et al. 2000). At that moment, Dr. T. Kohchi (at Kyoto University) and colleagues had just cloned the HY1 gene from A. thaliana, which encodes a key enzyme for chromophore biosynthesis (Muramoto et al. 1999), and kindly provided the sequence data on the HY1 gene. Based on these data, we were able to skip the mapping process and succeeded in cloning the se5 gene; our research demonstrated that phytochromes conferred the photoperiodic control of flowering in rice (Izawa et al. 2000). Many physiological experiments have been performed to examine how light quality and intensity affect photoperiodic responses (see reviews in Thomas and Vince-Prue 1997). The results generally suggested that phytochromes are photoperiodic photoreceptors, as reversible responses of phytochromes between red and far-red light treatments were evident on flowering-time. In rare cases, however, blue light may play a role in controlling flowering time, as shown in some Brassica species, although the photoreceptors for blue light were not identified at that time (see also reviews in Thomas and Vince-Prue 1997). Molecular cloning and characterization of the flowering-time gene FHA in A. thaliana, a plant closely related to Brassica, has revealed that a blue light receptor gene, termed CRYPTOCHROME 2 (CRY2 = FHA), plays an important role in the photoperiodic control of flowering in A. thaliana (Guo et al. 1998) and transmits the light signal before dusk under long-day conditions to promote floral induction (Valverde et al. 2004). In contrast, phytochrome B (phyB) is not required for the long-day promotion of flowering in A. thaliana but instead plays a role in floral inhibition under short-day conditions. This finding is consistent with the data on

II.4 Photoperiodic Flowering in Rice

165

Brassica species described above. The phyA photoreceptor in A. thaliana, which mainly plays a role in far-red light reception, can also promote flowering of A. thaliana under long-day conditions. There are five phytochrome genes in A. thaliana. Molecular genetics has revealed that phyC, phyD, and phyE function redundantly with phyB to inhibit flowering in A. thaliana, whereas phyA can promote flowering (Casal et al. 2003; Franklin et al. 2005). A recent study has demonstrated that a single phyA mutation did not change flowering time in rice, whereas a phyB mutant exhibited an early flowering phenotype under long-day conditions (Takano et al. 2005). Like the se5 mutant, a phyAphyB double mutant in rice exhibited markedly earlier flowering and no apparent response to photoperiod. These results suggest that there may be at least two independent pathways in the control of flowering: one that requires only phyB functions and the other that needs both phyA and phyB functions. It is also possible that the differences among single and double phytochrome mutants in rice just contributed to the variations in the amount and sensitivity of phytochrome. In A. thaliana, another possible light receptor, termed FKF1, is involved in long-day floral promotion (Imaizumi et al. 2003). FKF1 possesses two domains, an LOV (a type of PAS domain) and an F-box. The LOV domain is a blue-light photoreception domain containing a flavin as a chromophore, whereas the F-box is involved in target specificity of protein degradation via ubiquitination. The FKF1 protein expression pattern is controlled by the circadian clock and FKF1 is a dusk-specific factor for protein degradation (Imaizumi et al. 2003). Recently, a protein that interacts with FKF1 was identified as the CDF1 transcription factor in A. thaliana (Imaizumi et al. 2005). Furthermore, CDF1 can suppress the transcription of a key transcription factor, termed CONSTANS (CO; see below). Therefore, FKF1 can control CO mRNA expression at dusk only under long-day conditions. Phylogenetic analyses with rice genome data have revealed that one or two FKF1 orthologs exist in the rice genome. Therefore, regulation of flowering time similar to that of A. thaliana may exist in rice, although no genetic evidence has yet been reported. In rice, phytochromes function as a strong repressor of transcription of floral-switch genes, such as Hd3a (Izawa et al. 2002), a rice FT ortholog. Recent studies have shown that FT and Hd3a protein move from the leaf to shoot apical meristem region upon floral transition and thus may act as a florigen in A. thaliana and rice (Corbesier et al. 2007; Tamaki et al. 2007). However, it is still unclear how phytochromes control the expression of floral-switch genes in rice.

3

Quantitative Trait Loci Analyses Reveal the Conserved Floral Pathway

The mapping of genes using associated DNA markers is a powerful tool to identify a key mutation in a gene that confers a certain biological phenotype. Natural variations between two parent plants, however, sometimes disturb mapping of the target gene due to endogenous quantitative trait loci (QTLs). Therefore, QTL analysis is

166

T. Izawa

a prerequisite for some traits with many natural variations, such as flowering-time genes. Several important flowering-time genes in rice were isolated by QTL analysis and subsequent fine mapping using advanced backcross progeny (Yano 2001). The Heading date 1 (Hd1) gene was cloned by Yano et al. (2000). They first performed QTL analysis using an F2 population from a cross between the cultivars ‘Nipponbare’ and ‘Kasalath’ and identified several QTLs including Hd1 (Yano et al. 1997). The group then identified a series of QTLs using advanced backcross populations, with 15 QTLs having been identified so far (Yano 2001). Cloning of Hd1 has revealed that this gene encodes a putative transcription factor similar to Arabidopsis CO (Yano et al. 2000). The CO gene is the sole key regulator in the photoperiodic control of flowering in A. thaliana. CO mRNA is mainly controlled by circadian clocks and is expressed during the circadian night (Suarez-Lopez et al. 2001). The CO protein controls a major downstream gene, FT (Kardailsky et al. 1999; Kobayashi et al. 1999). Cloning of another QTL, Hd3a, has revealed that Hd3a is a rice ortholog of the FT gene (Kojima et al. 2002). These results clearly indicate that a homologous floral-transition system exists between rice and A. thaliana, which diversified evolutionarily around 200 million years ago. We have further demonstrated that Hd1 promotes Hd3a mRNA expression under short-day conditions and represses it under long-day conditions (Izawa et al. 2002). In contrast, CO promotes FT mRNA expression only under long-day conditions (Suarez-Lopez et al. 2001). Therefore, the difference in transcriptional regulation of Hd3a/FT by Hd1/CO is a major difference between these short-day and long-day plants. Further QTL cloning in rice has revealed that the Hd6 gene, a QTL between ‘Nipponbare’ and ‘Kasalath’, encodes a casein kinase II (CK II) gene. The ‘Nipponbare’ allele contains a frame-shift mutation that results in a premature stop codon (Takahashi et al. 2001). When researchers first began to perform QTL analyses, most believed that the functional nucleotide polymorphisms for QTLs were subtle mutations that changed gene functions only slightly or were secondary or tertiary causes of the change in biological responses. The map-based cloning of several QTLs in rice, however, revealed that many natural variations in flowering-time genes represented null or severe mutations. Various combinations of these QTLs likely produce the phenotypic variation of flowering time in rice. Thus, although it may not be easy to clone the target genes corresponding to minor QTLs, these results clearly demonstrate the ability of QTL analysis to reveal the molecular mechanism underlying flowering-time regulation in rice.

4

A Unique Floral Pathway Exists in Rice

A QTL analysis using a different combination of cultivars has revealed a novel flowering-time gene in rice. Dr. K. Doi (at Kyushu University, Japan) and colleagues performed a QTL analysis using an F2 population between the japonica

II.4 Photoperiodic Flowering in Rice

167

cultivar ‘Taichung 65’ (T65) and a cultivar of Oryza glaberrima, and found a novel major QTL in chromosome 10, termed Early heading date 1 (Ehd1). Ehd1 functions as a floral activator gene mainly under short-day conditions. In collaboration with the group at Kyushu University, we cloned the Ehd1 gene and revealed that it encodes a B-type response regulator (Doi et al. 2004). The T65 allele of Ehd1 contained a natural mutation in the DNA binding domain called GARP, which caused an amino acid change resulting in severe loss of its DNA binding activity. The B-type response regulator is part of the two-component system (a typical bacteria-type signal transduction system) and contains a receiver domain and a GARP domain to control its downstream gene. More than 10 B-type response regulators have been identified in A. thaliana, but all of the examined genes are involved in the cytokinin signal transduction pathway (Heyl and Schmülling 2003). A phylogenetic analysis revealed that Ehd1 may have evolved uniquely among other B-type response regulators, as there is no proper ortholog of Ehd1 in the Arabidopsis genome. Ehd1 mRNA is induced under short-day conditions, indicating that Ehd1 functions at a downstream position in the photoperiodic flowering pathway in rice after plants recognize the day-length. In addition, we have found that T65 contains severe mutations in both the Ehd1 and Hd1 genes (Doi et al. 2004), which would account for it being a relatively late-flowering cultivar with a weak photoperiodic response. Because only a single functional Ehd1 allele into T65 can confer a clear short-day induction of flowering, Ehd1 appears to be a key member in a photoperiodic flowering mechanism unique to rice, which can exist independently of the Hd1-dependent flowering pathway. Furthermore, identification of downstream genes of Ehd1 as some FT-like genes and MADS-box genes indicates that the unique Ehd1 pathway is integrated within the evolutionarily conserved Hd1-Hd3a floral pathway (Doi et al. 2004).

5

The Importance of Floral Inhibition Control Under Long-day Conditions

Whether summer or winter annuals, long-day plants generally flower according to day-length in the spring or summer, when day-length increases gradually. When long-day plants are mature in spring enough to respond to the floral signal from the leaf and are subjected to long-day conditions, they flower. Thus, longer photoperiods simply may strengthen the floral signals in long-day plants. In contrast, the floral signals in short-day plants are rather complicated, as short-day plants may exhibit floral transition at various timings, sometimes even under apparent longday conditions. This flexibility in short-day plants allows rice to adapt to given environmental conditions in order to produce more offspring or higher yields as agricultural crops (Fig. 1). The relationship between Hd1 and Ehd1 may provide a comprehensive explanation for the flexibility of flower timing at the molecular level in rice.

168

T. Izawa 16:00 43.1 8(Sapporo) 35.7 8(Tokyo) 26.2 8 (Naha)

15:00 Photoperiod

14:00 13:00 12:00 11:00 10:00 9:00 30

Temperature(8C)

25 20

188C 15 10 5 O 12

-5 Jan

Mar

May

July

Sep

Nov

Fig. 1 Relationship between photoperiods and ambient temperature. Photoperiods and ambient average temperatures are plotted for three cities in Japan with different latitudes. Horizontal arrows indicate time spans with temperatures exceeding 18°C in the corresponding local areas. For proper rice development, ambient temperatures above 18°C are required for some developmental stages from floral transition to seed maturation, which may take at least 2 months. Therefore, floral transitions in Sapporo should occur under the longest day-length conditions. Even in Tokyo, floral transition under about 14-h photoperiods is preferred. In Naha city of Okinawa, rice may be grown in all months except January and February, allowing cultivation of two or three crops per year in Okinawa

When a functional allele of either Hd1 or Ehd1 was transformed into T65, the plants flowered significantly earlier than T65 under short-day conditions (9.5 h light/14.5 h dark) (Doi et al. 2004). This clearly demonstrated that both Hd1 and Ehd1 can function as a floral activator in rice under short-day conditions. In contrast, under long-day conditions (14.5 h light/9.5 h dark), T65 plants with Hd1 did not flower even at 180 days after sowing, because Hd1 functions as a strong floral repressor under long-day conditions. Ehd1 functions as a floral activator even under long-day conditions. Furthermore, T65 with both Hd1 and Ehd1 flowered at about 50 days and 120 days under short-day and long-day conditions, respectively (T. Higashi et al., unpublished data). These results clearly demonstrated that the

II.4 Photoperiodic Flowering in Rice

169

promoter activity of Ehd1 may compete with the repressor activity of Hd1 to determine flowering time under long-day conditions. This regulation is likely to be critical for the breeding of rice plants adapted to local areas in higher-latitude temperate regions. For instance, without the Ehd1 function, T65 with the Hd1 allele may not flower at such high latitudes until the day-length shortens beyond a critical photoperiod, under which Hd1 can promote flowering in rice. Such a flowering time at high latitudes may be too late to produce seeds, however, because the low ambient temperatures upon floral transition may not be appropriate for proper meiosis during pollen formation and proper embryogenesis after fertilization (Fig. 1). An Hd1-deficient near-isogenic cultivar flowered earlier than the wild-type cultivar in mainland Japan at temperate latitudes, but flowered later than the wild type in the southern islands of the Okinawa region at tropical latitudes (I. Ando and M. Yano, personal communication; Fig. 1). This result clearly demonstrated that Hd1 functions as a floral repressor in many parts of Japan. In addition, the T65 cultivar originated in Taiwan, which is located in a low-latitude tropical region. Therefore, mutations in both Hd1 and Ehd1 genes are required for the rice cultivar to obtain enough biomass before flowering. Taken together, these results clearly demonstrated that controlling the competition between Hd1 and Ehd1 under longday conditions is critical when breeding rice cultivars that are adaptable to broader cultivating regions.

6

Night-break Experiments in Rice

Night-break experiments, which examine the effects of light pulses in night-time on floral responses, were first conducted in 1936 (see Thomas 1998). The experiments suggested that many plants, especially typical short-day plants, use the nightlength rather than the day-length to determine the timing of floral transitions in the photoperiodic control of flowering. Subsequent research, however, revealed that day-length recognition in photoperiodic flowering is due to the interaction between acute light signals and circadian clocks. Thus, night-break treatments may have two effects: acting as an acute light signal (i.e., a light-on signal) and as an entrainment signal (i.e., a synchronizing signal). These two distinct effects were clearly shown using Pharbitis (Ipomoea) nil subjected to two light pulses: one for the stimuli and the other as a probe (Lumsden and Furuya 1986). The earlier night-break experiments revealed that phytochromes may be the photoreceptors responsible for the night-break responses. Recently, night-break experiments were conducted on rice plants (Ishikawa et al. 2005), representing the first study to investigate the night-break response of flowering at the molecular level. The experiments revealed that phyB is a major photoreceptor for night-break responses in rice (Ishikawa et al. 2005). The phytochrome signals in response to the light pulses repressed the Hd3a mRNA expression that is induced under short-day conditions but did not affect the Hd1 mRNA patterns. These results are comparable

170

T. Izawa

to those of Izawa et al. (2002), who examined phytochrome actions in the photoperiodic control of flowering in rice. Therefore, the night-break responses are likely to be closely related to day-length recognition, even at the molecular level. Only a single light pulse was required to repress Hd3a mRNA at dawn of the subsequent day (Ishikawa et al. 2005). This finding clearly indicates that day-length recognition is a daily event, whereas floral induction requires competence for reception of floral signals, which could be a function of both the developmental stage and the integrated sums of repetition and strength of daily Hd3a expression.

7

The Roles of Circadian Clocks in Floral Induction in Rice Are Still Unknown

Circadian clocks are deeply involved in the photoperiodic control of flowering in higher plants. Pittendrigh and Minis (1964) proposed the external coincidence model, in which the interaction between the acute light signal and the circadian clock determines the floral induction. This model is a refinement of Bünning’s hypothesis proposed in the 1930s (see Bünning 1960). At the molecular level, the CO function as a floral activator could be limited to dusk under long-day conditions, due to circadian clocks and light signals. In fact, CO mRNA is regulated mainly by circadian clocks (Suarez-Lopez et al. 2001) and fine-tuned by the FKF1 regulation described above (Imaizumi et al. 2003). However, no clear evidence has yet been reported on how circadian clocks regulate flowering time in rice. It is clear that circadian clocks exist in rice, as experiments have demonstrated that some gene expression was clearly regulated by circadian clocks using cab1R::luc as a reporter gene (Sugiyama et al. 2001). Furthermore, orthologs corresponding to major circadian-clock components in A. thaliana were identified in the rice genome (Izawa et al. 2003), and five pseudoresponse regulator (PRR) genes of rice were shown to be expressed in a circadianclock-dependent manner (Murakami et al. 2003). In contrast, no circadian-clock-related gene has been identified as a flowering-time gene in rice. Recently, a putative functional nucleotide polymorphism was identified in a PRR gene, OsPRR37, which is a possible circadian-clock gene in rice (Murakami et al. 2005). The QTL Hd2 exists around OsPRR37, although direct evidence that Hd2 encodes OsPRR37 has not yet been reported. In barley, an OsPRR37 ortholog was recently reported as a flowering-time gene (Turner et al. 2005). In a study using transgenic rice plants and the RNAi technique, mRNA expression of Os GIGANTEA (OsGI; Hayama et al. 2002), an ortholog of GIGANTEA (GI) in A. thaliana, was clearly reduced. Because these plants tended to flower early under long-day conditions, OsGI appears to be a floral repressor gene in rice (Hayama et al. 2003). As GI is required for CO mRNA expression in A. thaliana, Hayama et al. (2003) suggested that OsGI is required for Hd1 expression in rice. Although GI was identified as a circadian-clock component, the function of OsGI in circadian clocks was not examined in that work.

II.4 Photoperiodic Flowering in Rice

8

171

Differences Between Long-day and Short-day Plants

Molecular genetic studies in rice have clearly demonstrated that an evolutionarily conserved floral pathway exists in rice and A. thaliana with distinct transcriptional regulation, as described above. Compared with extensive studies using A. thaliana, however, there are apparent missing links in the rice pathway that require further genetic analysis. Because both the rice and Arabidopsis genomes have been completely sequenced, phylogenetic analysis allows us to infer these missing components (Izawa et al. 2003). Both OsGI and GI are single genes in the rice and Arabidopsis genomes. Both Hd1 and CO genes are members of the same clade within a small gene family that contains two B-box domains and a CCT motif (Izawa et al. 2003). The Arabidopsis genes FT and TSF belong to the same clade as the rice gene Hd3a. More than 10 predicted genes formed a single clade with FT, TSF, and Hd3a (Izawa et al. 2002). Among them, FTL and RFT1 are positioned near FT and Hd3a, suggesting they may play a role in floral induction in rice. The many FT-like genes may provide various signals in rice development. FT gene products in the meristems may interact with a bZIP-type transcriptional factor termed FD in A. thaliana (Abe et al. 2005; Wigge et al. 2005). Phylogenetic analysis of FD has suggested possible FD orthologs in the rice genome (Fig. 2), although these FD-like genes have not yet been analyzed. Recently, analysis of the maize mutant delayed flowering (dlf1) revealed that the gene corresponding to dlf1 encodes an ortholog of FD (Muszynski et al. 2006). Furthermore, analysis of OsMADS50, a rice ortholog of the Arabidopsis SOC1 gene, revealed that both OsMADS50 and SOC1 play a role in floral promotion (Lee et al. 2004). SOC1 is a major target gene of the key floral repressor FLC after treatment of vernalization in A. thaliana. In rice, there is no such MADS-box gene belonging to the same clade as FLC, and no vernalization-related response was observed. Nonhomologous regions between these orthologous genes may suggest some restrictions of biological functions of these otherwise unnecessary regions in the gene products. For instance, FD and FD-like genes do not show much homology, except the bZIP domains, suggesting that the amino acid restriction for the FD function is not severe or the function of FD-like genes in rice differs from that of the original FD. FD can induce the downstream MADS-box genes such as FUL/ AP1/SOC1 at the meristems in A. thaliana, whereas one of the FUL-like genes in rice, OsMADS14 (= RAP1B), was mainly expressed in the leaf. The homology itself does not explain how the photoperiod response differs between rice and A. thaliana at the molecular level. Transcriptional regulation of the Hd3a/FT gene by the Hd1/CO transcriptional factor is a primary difference in flowering-time regulation between rice and A. thaliana (Izawa et al. 2002, 2003; Fig. 3). In A. thaliana, CO activity was regulated by light signals through phyB and phyA/CRY2 (Yanovsky and Kay 2002; Valverde et al. 2004). This regulation of CO activity may be based on the control of protein stability through ubiquitination. In this regulation, the amount of CO protein is very

172

T. Izawa

Fig. 2 A phylogenetic tree of FD-like bZIP domains in Arabidopsis thaliana and rice. Only data for well-aligned bZIP domain regions were used to construct this neighbor-joining tree. It is clearly shown that two rice annotations (RAP-DB; http://rapdb.lab.nig.ac.jp/) are in the same clade with FD and its paralog FDP in A. thaliana

low at dawn, and Valverde et al. (2004) showed that CO protein becomes unstable in darkness due to an unknown degradation system. This suggests a clear difference between rice and A. thaliana. Because Hd1 can function as an activator of Hd3a transcription at dawn under short-day conditions, degradation of Hd1 protein in darkness should not occur in rice. In addition, Hd1 repression under long-day

II.4 Photoperiodic Flowering in Rice

173 Rice

Arabidopsis Blue

FKF1 protein FR Blue CDF1 protein PHYA CRY2

Long days Clock

CO mRNA

Long days

Short days Red

PHYB

Clock CO mRNA

Red

Short days

Clock

PHYB

Darkness

Clock

Light?

Red

Light?

Darkness

X?

X?

SE5

CO protein

CO protein Darkness

Hd1 protein Darkness

Hd1 protein Ehd1 mRNA

Ehd1 mRNA

FT mRNA

FT mRNA

Hd3a mRNA

Hd3a mRNA

FT protein

FT protein

Hd3a protein

Hd3a protein

Early flowering

Late flowering

Late flowering

Early flowering

Fig. 3 Molecular models of photoperiodic flowering in rice and Arabidopsis thaliana. Comparative schemes are presented to explain photoperiodic flowering of a long-day plant, A. thaliana, and a short-day plant, rice (O. sativa). Font sizes correspond to the amounts of each mRNA or gene product of genes. In rice, degradation of Hd1 protein under darkness should not occur, but for ease of comparison, it is indicated in gray

conditions requires the light signal transmitted by phytochromes. Although phyB m akes CO protein unstable during the daytime in A. thaliana, this degradation promoted by phytochrome signaling is not enough to explain Hd1 repression under long-day conditions. Long-day-specific repressor or modifier genes may be involved in Hd1 functions. The Hd6 CK II gene (Takahashi et al. 2001) confers long-day repression of flowering in rice and is a candidate for such long-day-specific modification, although the CK II gene is believed to be a circadian-clock component in A. thaliana (Daniel et al. 2004). In addition to differences found in the conserved floral regulation pathway, some unique flowering systems such as the Ehd1 pathway clearly exist in rice and regulate flowering time, allowing rice to become adapted to various local conditions. It is likely that additional critical differences in flowering-time regulation exist between rice and A. thaliana.

9

Future Perspectives

Based on several extensive studies using A. thaliana, the molecular mechanisms underlying day-length measurement in long-day plants can be explained well by the interaction between circadian clocks and acute light signals, such that CO mRNA and proteins are regulated by circadian clocks, FKF1 activity, and CRY2/phyA activity. These regulations, however, do not explain the short-day promotion and long-day repression pathways found in rice. In addition, most short-day plants have critical thresholds for night-lengths to induce floral transitions, and the molecular

174

T. Izawa

mechanisms underlying the setting of critical night-lengths are still unclear. Because the ancestor of rice diverged from that of A. thaliana around 200 million years ago, comparing the differences in photoperiodic flowering between rice and A. thaliana will help to reveal the molecular diversity in higher plants.

References Abe M, Kobayashi Y, Yamamoto S, et al. (2005) FD, a bZIP protein mediating signals from the floral pathway integrator FT at the shoot apex. Science 309:1052–1056 Bünning E (1960) Circadian rhythms and the time measurement in photoperiodism. Cold Spring Harbor Symp Quant Biol 25:249–256 Casal JJ, Luccioni LG, Oliverino KA, Boccalandro HE (2003) Light, phytochrome signalling and photomorphogenesis in Arabidopsis. Photochem Photobiol Sci 2:625–636 Corbesier L, Vincent C, Jang S, et al. (2007) FT protein movement contributes to long-distance signaling in floral induction of Arabidopsis. Science 10.1126/science.1141752 Daniel X, Sugano S, Tobin EM (2004) CK2 phosphorylation of CCA1 is necessary for its circadian oscillator function in Arabidopsis. Proc Natl Acad Sci USA 101:3292–3297 Doi K, Izawa T, Fuse T, et al. (2004) Ehd1, a B-type response regulator in rice, confers short-day promotion of flowering and controls FT-like gene expression independently of Hd1. Genes Dev 18:926–936 Franklin KA, Larner VS, Whitelam GC (2005) The signal transducing photoreceptors of plants. Int J Dev Biol 49:653–664 Garner WW, Allard HA (1920) Effect of the relative length of day and night and other factors of the environment on growth and reproduction in plants. J Agr Res 18:553–606 Guo H, Yang H, Mockler TC, Lin C (1998) Regulation of flowering time by Arabidopsis photoreceptors. Science 279:1360–1363 Hayama R, Izawa T, Shimamoto K (2002) Isolation of rice genes possibly involved in the photoperiodic control of flowering by a differential display method. Plant Cell Physiol 43:494–504 Hayama R, Yokoi S, Tamaki S, Yano M, Shimamoto K (2003) Adaptation of photoperiodic control of pathways produces short-day flowering in rice. Nature 422:719–722 Heyl A, Schmülling T (2003) Cytokinin signal perception and transduction. Curr Opin Plant Biol 6:480–488 Huang T, Bohlenius H, Eriksson S, Parcy F, Nilsson O (2005) The mRNA of the Arabidopsis gene FT moves from leaf to shoot apex and induces flowering. Science 309:1694–1696 Imaizumi T, Tran HG, Swartz TE, Briggs WR, Kay SA (2003) FKF1 is essential for photoperiodicspecific light signalling in Arabidopsis. Nature 426:302–306 Imaizumi T, Schultz TF, Harmon FG, Ho LA, Kay SA (2005) FKF1 F-box protein mediates cyclic degradation of a repressor of CONSTANS in Arabidopsis. Science 309:293–297 Ishikawa R, Tamaki S, Yokoi S, et al. (2005) Suppression of the floral activator Hd3a is the principle cause of the night break effect in rice. Plant Cell 17:3326–3336 Izawa T, Oikawa T, Tokutomi S, Okuno K, Shimamoto K (2000) Phytochromes confer the photoperiodic control of flowering in rice (a short-day plant). Plant J 22:391–399 Izawa T, Oikawa T, Sugiyama N, Tanisaka T, Yano M, Shimamoto K (2002) Phytochromes mediate the external light signal to repress FT orthologs in photoperiodic flowering of rice. Genes Dev 16:2003–2020 Izawa T, Takahashi Y, Yano M (2003) Comparative biology has come to bloom: genomic and genetic comparison of flowering pathways in rice and Arabidopsis. Curr Opin Plant Biol 6:113–120

II.4 Photoperiodic Flowering in Rice

175

Kardailsky I, Shukla VK, Ahn JH, et al. (1999) Activation tagging of the floral inducer FT. Science 286:1962–1965 Kobayashi Y, Kaya H, Goto K, Iwabuchi M, Araki T (1999) A pair of related genes with antagonistic roles in mediating flowering signals. Science 286:1960–1962 Kojima S, Takahashi Y, Kobayashi Y, et al. (2002) Hd3a, a rice ortholog of the Arabidopsis FT gene, promotes transition to flowering downstream of Hd1 under short-day conditions. Plant Cell Physiol 43:1096–1105 Lee S, Kim J, Han JJ, Han JM, An G (2004) Functional analyses of the flowering time gene OsMADS50, the putative SUPPRESSOR OF OVEREXPRESSION OF CO1/AGAMOUS-LIKE 20 (SOC1/AGL20) ortholog in rice. Plant J 38:754–764 Lumsden PJ, Furuya M (1986) Evidence for two actions of light in the photoperiodic induction of flowering in Pharbitis nil. Plant Cell Physiol 27:1541–1551 Mizuno T, Nakamichi N (2005) Pseudo-response regulators (PRRs) or true oscillator components (TOCs). Plant Cell Physiol 46:677–685 Murakami M, Ashikari M, Miura K, Yamashino T, Mizuno T (2003) The evolutionarily conserved OsPRR quintet: rice pseudo-response regulators implicated in circadian rhythm. Plant Cell Physiol 44:1229–1236 Murakami M, Matsushika A, Ashikari M, Yamashino T, Mizuno T (2005) Circadian-associated rice pseudo response regulators (OsPRRs): insight into the control of flowering-time. Biosci Biotechnol Biochem 69:410–414 Muramoto T, Kohchi T, Yokota A, Hwang I, Goodman HM (1999) The Arabidopsis photomorphogenic mutant hy1 is deficient in phytochrome chromophore biosynthesis as a result of a mutation in a plastid heme oxygenase. Plant Cell 11:335–348 Muszynski MG, Dam T, Li B, et al. (2006) delayed flowering 1 encodes a basic leucine zipper protein that mediates floral inductive signals at the shoot apex in maize. Plant Physiol DOI:10.1104/pp.106.088815 Pittendrigh CS, Minis DH (1964) The entrainment of circadian oscillations by light and their role as photoperiodic clocks. Am Nat 108:261–293 Suarez-Lopez P, Wheatley K, Robson F, Onouchi H, Valverde F, Coupland G (2001) CONSTANS mediates between the circadian clock and the control of flowering in Arabidopsis. Nature 410:1116–1120 Sugiyama N, Izawa T, Oikawa T, Shimamoto K (2001) Light regulation of circadian clockcontrolled gene expression in rice. Plant J 26:607–615 Sung S, Amasino RM (2005) Remembering winter: toward a molecular understanding of vernalization. Annu Rev Plant Biol. 56:491–508 Takahashi Y, Shomura A, Sasaki T, Yano M (2001) Hd6, a rice quantitative trait locus involved in photoperiod sensitivity, encodes the alpha subunit of protein kinase CK2. Proc Natl Acad Sci USA 98:7922–7927 Takano M, Inagaki N, Xie X, et al. (2005) Distinct and cooperative functions of phytochromes. A, B, and C in the control of deetiolation and flowering in rice. Plant Cell 17:3311–3325 Tamaki S, Matsuo S, Wong H-L, Yokoi S, Shimamoto K (2007) Hd3a protein is a mobile flowering signal in rice. Science 10.1126/science.1141753 Thomas B (1998) Photoperiodism: an overview. In: Lumsden, PJ, Millar, AJ (eds) Biological rhythms and photoperiodism in plants. Bios Scientific Publishers, Oxford, pp 151–181 Thomas B, Vince-Prue D (1997) Photoperiodism in plants. Academic Press, London Turner A, Beales J, Faure S, Dunford RP, Laurie DA (2005) The pseudo-response regulator Ppd-H1 provides adaptation to photoperiod in barley. Science 310:1031–1034 Valverde F, Mouradov A, Soppe W, Ravenscroft D, Samach A, Coupland G (2004) Photoreceptor regulation of CONSTANS protein in photoperiodic flowering. Science 303:1003–1006 Wigge PA, Kim MC, Jaeger KE, et al. (2005) Integration of spatial and temporal information during floral induction in Arabidopsis. Science 309:1056–1059 Yano M (2001) Genetic and molecular dissection of naturally occurring variation. Curr Opin Plant Biol 4:130–135

176

T. Izawa

Yano M, Harushima Y, Nagamura Y, Kurata N, Minobe Y, Sasaki T (1997) Identification of quantitative trait loci controlling heading date in rice using a high-density linkage map. Theor Appl Genet 95:1025–1032 Yano M, Katayose Y, Ashikari M, et al. (2000) Hd1, a major photoperiod sensitivity quantitative trait locus in rice, is closely related to the Arabidopsis flowering time gene CONSTANS. Plant Cell 12:2473–2483 Yanovsky MJ, Kay SA (2002) Molecular basis of seasonal time measurement in Arabidopsis. Nature 419:308–312 Yokoo M, Okuno K (1993) Genetic analysis of earliness mutations induced in the rice cultivar Norin 8. Jpn J Breed 43:1–11

II.5

Genetic Regulation of Meristem Maintenance and Organ Specification in Rice Flower Development Hiro-Yuki Hirano1(* ü)

1

Introduction

Molecular genetic studies in Arabidopsis thaliana over the last 15 years have produced a wealth of information that has considerably advanced our understanding of plant development at the molecular level. Among these advances, the ABC model of floral organ specification and the negative feedback model of meristem maintenance constitute milestones in this field (Coen and Meyerowitz 1991; reviewed by Lohmann and Weigel 2002; Bäurle and Laux 2003; Carles and Fletcher 2003; Jack 2004). Since establishment of the ABC model, many genes orthologous to Arabidopsis ABC genes have been isolated in other plant species, and it has been shown that the functions of the ABC genes are conserved in eudicots (Theissen et al. 2000). It is now of great interest to determine whether the genetic regulation of floral organ specification and of meristem maintenance is conserved throughout angiosperms, including monocots and basal angiosperms. Rice (Oryza sativa) is a member of the grass family, which contains about 10,000 species. The grass family is one of the large clades in monocots, which is distantly related to eudicots, and the divergence of monocots and eudicots is estimated to be over 150 million yeas ago. The flowers and inflorescences in the grass family are distinct from those in eudicots and monocots other than the grasses (Bommert et al. 2005b; Itoh et al. 2005). The grass inflorescence is composed of unique units such as spikelets and florets. The flower lacks obvious petals and sepals but instead has distinct structures called the palea and the lemma. A more detailed description of the inflorescence and floral structures of rice is provided in the following sections. Rice has many advantages for molecular genetic studies: the genomic sequences are completely determined; genetic transformation is relatively easy; many transposonmediated mutation lines are available to identify specific knockout mutants; and molecular tools such as microarray analysis are available. Recently, many genes that

1 Graduate School of Science, The University of Tokyo, Hongo, Tokyo 113-8654, Japan e-mail: [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

177

178

H.-Y. Hirano

play essential roles in rice development have been isolated by positional cloning and homology screening, and functional analyses of these genes have been carried out (for reviews, see Bommert et al. 2005b; Kater et al. 2006; Yamaguchi and Hirano 2006). In this chapter, we review recent molecular studies of the genetic regulation of meristem maintenance and floral organ specification in rice. These studies may provide important information for understanding not only developmental mechanisms in rice per se but also the evolution of the mechanisms and genes that regulate the body plan in angiosperms (Evo-devo studies).

2

Structure of the Inflorescence and Flower in Rice

After the transition from the vegetative phase to the reproductive phase, the inflorescence meristem at the shoot apex produces primary rachis branches in a spiral phyllotaxy (Bommert et al. 2005b; Itoh et al. 2005). The primary branch meristems give rise to the secondary branch meristems and spikelet meristems that form the glumes and the floret meristems. Since the branch meristems are both self-maintaining and produce spikelet meristems, they correspond to the inflorescence meristem. The rice floret is composed of a bisexual flower, which contains one pistil, six stamens and two lodicules, and the grass-specific structures, the palea and the lemma (Fig. 1). The floret is flanked by two sterile lemmas (or empty glumes), which are considered to be vestiges of two lower florets. There are two rudimentary glumes, which are highly reduced, outside of the sterile lemma. Thus, the rice spikelet contains one fertile floret, two strongly reduced sterile florets, and two rudimentary glumes. Unlike maize, in which the floret meristems are clearly distinguished from spikelet meristems, the boundary between the spikelet meristem and the floret meristem is ambiguous in rice. Genetic analyses have revealed that two genes, FRIZZY PANICLE and SUPERNUMERARY BRACT, seem to be involved in the transition from spikelet meristem to floral meristem in rice (Komatsu et al. 2003; Lee et al. 2007).

Fig. 1 Schematic representation of the wild-type rice flower and its constituent structures

II.5 Genetic Regulation of Meristem Maintenance

179

In this chapter, we use the terms floral (flower) meristem without distinguishing the spikelet and floret meristems, and inflorescence meristem instead of primary and secondary branch meristems. We refer to the regions where lodicules, stamens, and a pistil develop in rice and maize as whorl 2, whorl 3, and whorl 4, respectively, in line with the definition of these regions in Arabidopsis. [As the identities of the petal and lemma are controversial at present (see below), we avoid defining whorl 1 here.]

3 3.1

Floral Meristem Regulation of Meristem Maintenance and Floral Organ Number

The aerial meristem of plants contains a group of pluripotent undifferentiated cells (stem cells) in a central zone. The stem cells are self-maintaining and also produce daughter cells that initiate formation of lateral organs, such as leaf and floral organ primordia, in the peripheral zone. If the activities of genes that control meristem maintenance are lost, the size of the meristem may be enlarged or diminished, resulting in a disturbance of normal plant development. In Arabidopsis, meristem maintenance is regulated by a feedback loop involving the CLAVATA (CLV) signaling pathway and WUSCHEL (WUS) gene activity (for reviews, see Bäurle and Laux 2003; Carles and Fletcher 2003). WUS is essential for stem cell identity and encodes a protein for a transcription factor with a novel type of homeodomain. Loss-of-function mutations of WUS cause premature termination of the meristem. The CLV signaling pathway contains three major genes, CLV1, CLV2, and CLV3. CLV1 encodes an LRR-type receptor kinase, CLV2 encodes a similar LRR protein but which lacks a cytoplasmic kinase domain, and CLV3 encodes a small secreted protein with a CLE domain. A 12-amino acid peptide derived from the CLE domain of the CLV3 protein by an unknown processing mechanism is thought to act as the ligand for a receptor complex containing CLV1 and CLV2 (Kondo et al. 2006). WUS activity is negatively regulated by the CLV signaling pathway. Therefore, mutations in the CLV genes give rise to overproliferation of the stem cells, resulting in enlargement of the shoot apical meristems at both vegetative and reproductive phases and of floral meristems. CLV3 is expressed in the stem cells and is positively regulated by WUS. In wild type, therefore, meristems are maintained by the balance of WUS gene activity and its negative regulation by CLV signaling. Mutations in the two FLORAL ORGAN NUMBER genes (FON1 and FON2) give rise to an enlargement of the floral meristem, resulting in an increase in the number of floral organs (Nagasawa et al. 1996; Suzaki et al. 2004, 2006). The increases in floral organ number are especially prominent for the inner whorl organs: carpels in particular are repetitiously formed in plants carrying severe mutations. In contrast to the floral meristem, the vegetative and inflorescence meristems are not affected by mutations of these genes, and vegetative organs and inflorescences

180

H.-Y. Hirano

develop normally. The double mutant of fon1 and fon2 is indistinguishable from the fon1 or fon2 single mutants, suggesting that the two genes act in the same genetic pathway (Suzaki et al. 2006). Molecular cloning has revealed that FON1 and FON2 encode an LRR-type receptor kinase and a small secreted protein, respectively (Suzaki et al. 2004, 2006). FON1 is orthologous to CLV1 in Arabidopsis and FON2 is closely related to CLV3. FON1 is expressed throughout the meristem in rice, whereas CLV1 is expressed mainly in the L3 layer of the central zone in the meristem in Arabidopsis. Like CLV3, FON2 is expressed in the apical region of the meristem, which is probably composed of the stem cells. Thus, the spatial expression patterns of these orthologous genes in rice and Arabidopsis are similar. Functional analyses in Arabidopsis indicate that FON2 rescues the clv3 mutation, and ectopic expression of FON2 causes immature termination of the meristem in a similar manner to CLV3. These findings suggest that the genetic mechanism of meristem maintenance is largely conserved in rice and Arabidopsis, and presumably throughout monocots and eudicots. In addition to evolutionarily conserved mechanisms, rice also has unique aspects of genetic regulation of floral meristem maintenance (Suzaki et al. 2006). Mutations of either FON1 or FON2 affect only the floral meristem and not the vegetative shoot apical meristem (SAM) or inflorescence meristems, whereas clv mutations cause enlargement of all aerial meristems. Constitutive expression of FON2 in rice gives rise to abnormal inflorescences with fewer branches and spikelets, suggesting that the sizes of inflorescence meristems may be reduced. This observation appears inconsistent with the mutant phenotypes of fon1 or fon2, which show no abnormality in the inflorescences. To explain this apparent inconsistency, we proposed a model for a regulatory network that maintains the three types of aerial meristem in rice (Suzaki et al. 2006). We hypothesized the existence of a second pathway that acts in parallel to the FON2–FON1 pathway (Fig. 2). Both pathways are normally active in the regulation of inflorescence meristem development. Therefore, mutation of either FON1 or FON2 would not affect meristem phenotypes, but constitutive expression of FON2 would give rise to a reduction in inflorescence meristem size. In contrast, the second pathway is hypothesized to be inactive in the floral meristem because the enlargement of this meristem occurs in fon1 and fon2 mutants. FON1 may not function in the vegetative SAM because no effect is observed in rice lines that constitutively express FON2. Thus, each aerial meristem seems to be regulated independently by a distinct genetic pathway in rice. If this model is correct, then the regulation of meristem maintenance by parallel genetic pathways in rice contrasts with that of Arabidopsis, which is regulated by a single CLV pathway (Fig. 2). Loss of function of the LONELY GUY (LOG) gene causes premature termination of the floral meristems, resulting in reduction in floral organ number and rachis branches (Kurakawa et al. 2007). This mutant phenotype resembles that of constitutive expression of FON2. LOG encodes a novel cytokinin-activating enzyme that converts inactive cytokinins into biologically active forms, suggesting the importance of this phytohormone in meristem function. There is considerable interest in determining how cytokinin acts in the negative feedback loop of the CLV–WUS signaling pathway.

II.5 Genetic Regulation of Meristem Maintenance

181

Rice

Arabidopsis

signal molecules

FON2

X

CLV2

putative receptors

FON1

Y

CLV1/CLV2

stem cell identity

stem cell identity

Meristem Vegetative Inflorescence Flower Fig. 2 Models for meristem maintenance in rice and Arabidopsis. Existence of the X–Y pathway is postulated from observation of the effect of constitutive expression of FON2

3.2

Determinacy of the Floral Meristem

The vegetative SAM develops leaf primordia and axillary meristems, which become SAMs later. The inflorescence meristem continues to be self-maintaining before eventually converting into the floral meristem. Thus, both vegetative SAM and the inflorescence meristem are self-maintaining and, therefore, indeterminate in nature. In contrast, the floral meristem is determinate. After development of the floral organs, the stem cells in the floral meristem are finally consumed by carpel primordia and no further meristematic activity occurs. There are several genes that regulate meristem determinacy in rice. Knockdown of the class C MADS-box gene OsMADS58 in the floral meristem results in a severe disruption of determinacy (Yamaguchi et al. 2006). In OsMADS58 knockdown lines, a set of floral organs consisting of lodicules, stamens, and carpel-like organs is produced repeatedly (Fig. 3E). This repetition can form more than 50 floral organs. The class I homeobox gene OSH1, a marker of undifferentiated cells, is expressed in the central region of the meristem in OsMADS58 knockdown lines even after many floral organs are produced in the near mature flower, whereas expression disappears

182

H.-Y. Hirano

Fig. 3 Models for floral organ specification in rice and Arabidopsis. Schematic phenotypes and functional domains of each gene are shown for wild-type flowers in rice (A) and Arabidopsis and for mutant flowers in rice (B–E)

after carpels are initiated in the wild type. In Arabidopsis, mutation of the class C gene AGAMOUS (AG) results in loss-of-determinacy of the floral meristem. In this mutant, a set of floral organs (sepal–petal–petal) is repeatedly formed (for reviews, see Lohmann and Weigel 2002; Jack 2004). Thus, regulation of floral meristem determinacy by class C genes is conserved in rice and Arabidopsis. The other class C gene, OsMADS3, however, has little effect on meristem determinacy. OsMADS58 and OsMADS3 are believed to have evolved by duplication from the same ancestral gene during the evolution of grasses (Yamaguchi et al. 2006). However, OsMADS58 appears to have become more important in the regulation of meristem determinacy than OsMADS3. LEAFY HULL STERILE1 (LHS1), which belongs to the class E MADS-box genes (OsMADS1), acts in the determination of the floral meristem in addition to the specification of the palea and lemma. In a loss-of-function mutant of LHS, internal flowers are formed in a spikelet and, in extreme cases, the palea/lemmalike structures are indeterminately generated (Jeon et al. 2000; Agrawal et al. 2005). Conversely, the constitutive expression of LHS produces undeveloped floral organs,

II.5 Genetic Regulation of Meristem Maintenance

183

often lacking carpels, suggesting that promotion of LHS function accelerates the premature termination of the floral meristem (Prasad et al. 2001). Floral meristem determinacy is also partially regulated by genes such as FON1, FON2, DROOPING LEAF (DL), and ABERRANT PANICLE ORGANIZATION1 (APO1). In mutants of fon1, fon2, or apo1, the central organ carpels are repetitiously formed (Suzaki et al. 2004, 2006; Ikeda et al. 2005). The loss-of-determinacy of the floral meristem is accelerated in the double mutants fon1 apo1 and fon2 apo1. The major function of DL is as a floral homeotic gene (see below). Mutation of this gene causes production of many stamens instead of carpels in the central whorl (Yamaguchi et al. 2004). In the fon and dl mutants, morphologically undifferentiated cells remain in the central region of the meristem, surrounded by carpels or stamens that form repetitiously in each mutant flower. In these mutants, OSH1 expression is maintained in the central region of the meristem.

4 4.1

Floral Organ Development Carpel Specification

The DL gene has a crucial function in the specification of carpel identity in rice (Nagasawa et al. 2003; Yamaguchi et al. 2004). Initially, the dl mutant was identified because of its abnormal drooping leaf phenotype caused by the failure of midrib formation. In addition to the drooping leaf phenotype, severe mutations of DL show homeotic conversion of carpels into stamens (Fig. 3B). In weak and intermediate alleles, carpel identities are partially disturbed and may result in production of multiple carpels and/or undifferentiated cell clumps. Analysis of the DL gene shows that it encodes a protein for a putative transcription factor with zinc finger and YABBY domains (Yamaguchi et al. 2004). DL is expressed in the presumptive regions of carpel initiation and in carpel primordia. Thus, DL is unique as a floral homeotic gene in that it belongs to the YABBY and not the MADS-box gene family. The latter family includes all other floral homeotic genes so far reported except for APETALA2 (for reviews, see Lohmann and Weigel 2002; Jack 2004). Furthermore, DL function is confined to the central whorl, whereas the ABC class of floral homeotic genes acts in two adjacent whorls. DL is closely related to CRABS CLAW (CRC) of Arabidopsis. In crc mutants, carpel identity is partially lost; for example, carpels may not be fused at the top (Alvarez and Smyth 1999; Bowman and Smyth 1999). Thus, both DL and CRC are involved in carpel development. Although there is some indication that the ancestral function of the two genes is conserved in rice and Arabidopsis, their actual contributions to carpel development differ in the two species. DL may have acquired critical new functions in the specification of carpel identity during grass evolution. In eudicots such as Arabidopsis and Antirrhinum majus, carpel specification is regulated by the class C genes, AGAMOUS and PLENA, respectively (for reviews,

184

H.-Y. Hirano

lodcule

lodicule

lemma-side

palea-side

whorl 2

stamen

carpel specification)

whorl 3

carpel morphology

meristem determinacy

whorl 4

Fig. 4 A model for the genetic network regulating flower development in rice. Thickness of arrows indicates the extent of contribution of gene function

see Lohmann and Weigel 2002; Jack 2004). Rice has two class C genes, OsMADS3 and OsMADS58, as described above (Yamaguchi et al. 2006). A knockout mutant of OsMADS3 develops carpels with almost normal morphology, but the number of carpels increases (Fig. 3D). In rice lines with RNAi knockdown of OsMADS58 expression, carpel-like structures with altered morphology are formed in the flowers, which reiteratively develop floral organs (Fig. 3E). Even in double knockdown lines of OsMADS3 (weak allele of transposon insertion line) and OsMADS58 (RNAi lines), carpel-like structures are produced. DL is specifically expressed in the primordia of carpel-like structures in the double knockdown lines. These findings suggest that the contribution of class C genes in carpel development is either very weak or nonexistent in rice and that DL plays an essential role in carpel specification (Fig. 4).

4.2

Stamen Specification

Stamens are specified by the combined activities of class B and class C genes in Arabidopsis (for reviews, see Lohmann and Weigel 2002; Jack 2004). Loss-of-function mutation in either of the class B genes APETALA3 (AP3) or PISTILATA (PI) causes homeotic transformation of stamens into carpels and petals into sepals, whereas severe mutation of AG gives rise to transformation of stamens into petals and reiterative flowers consisting of sepals, petals, and petals in whorls 1, 2, and 3, respectively. In the superwoman1 (spw1) mutant of rice, stamens are homeotically converted to carpels, and lodicules homologous to petals are transformed into palea/lemma-like

II.5 Genetic Regulation of Meristem Maintenance

185

or slender leaf-like structures (Fig. 3C; Nagasawa et al. 2003). SPW1 encodes a MADS-box gene, known as OsMADS16, and is orthologous to Arabidopsis AP3. Suppression of endogenous SPW1 by RNAi generates a phenotype similar to spw1 (Xiao et al. 2003). The rice genome has two genes, OsMADS2 and OsMADS4, that belong to the PI clade (Chung et al. 1995; Lee et al. 2003; Yamaguchi et al. 2006). The reduction in the expression level of OsMADS4 by antisense RNA results in transformation of lodicules and carpels into palea/lemma-like structures and stamens (Kang et al. 1998). SPW1 and OsMADS4 are expressed in the primordia of lodicules and stamens (Chung et al. 1995; Nagasawa et al. 2003; Yamaguchi et al., unpublished data). Protein interaction between SPW1 and OsMADS4 has been demonstrated using the yeast two-hybrid system. Enhancement of OsMADS4 expression is observed in ectopic expression lines of OsMADS16 (SPW1) and reduction in RNAi lines (Lee et al. 2003; Xiao et al. 2003). These results suggest that both SPW1 and OsMADS4 form heterodimers and self-regulate expression in an autoregulatory circuit in rice similar to that of AP3 and PI in Arabidopsis (for reviews, see Lohmann and Weigel 2002; Jack 2004). Overall, the functions of these class B genes are highly similar to those of Arabidopsis and are required for stamen specification (Fig. 4). In contrast to the conserved function of OsMADS4, OsMADS2 may not be involved in stamen development but rather participates predominantly in lodicule specification (Prasad and Vijayraghavan 2003; see below). Thus, class B genes in the PI clade seem to be functionally divergent in rice. Reciprocal homeotic conversion of stamens and carpels in dl and spw1 mutants suggests that there is ectopic expression of class B genes and the DL gene in each mutant. This hypothesis is supported by the spatial expression pattern of both genes in the mutants; that is, DL and SPW1 are ectopically expressed in whorl 3 of spw1 and in whorl 4 of dl, respectively (Nagasawa et al. 2003; Yamaguchi et al. 2004). In addition, ectopic expression of OsMADS16 (SPW1) causes a transformation of carpels into stamens, or the loss of carpel development in whorl 4, suggesting that DL is repressed by the ectopic expression of OsMADS16 in whorl 4 (Lee et al. 2003). Taken together, these results indicate that DL and SPW1 antagonistically regulate each other’s expression (Fig. 3A–C, Fig. 4). The contribution of class C genes to stamen specification also appears to differ between OsMADS3 and OsMADS58, as in the case of floral meristem determinacy (Yamaguchi et al. 2006). Complete conversion of stamens to lodicules is observed in an OsMADS3 knockout line (Fig. 3D). Ectopic expression of OsMADS3 results in complete or partial transformation of lodicules into stamens (Kyozuka and Shimamoto 2002). These results strongly indicate that OsMADS3 plays a critical function in stamen specification. In contrast, even in severe knockdown lines of OsMADS58, normal stamens develop, although their number is reduced (Fig. 3E). Overall, the evidence indicates that stamen specification is regulated by the combined activities of class B and C genes in rice, and especially by SPW1/OsMADS4 and OsMADS3 (Fig. 4). Stamens are partially replaced by lodicules in apo1 mutants (Ikeda et al. 2005). Although SPW1 is normally expressed, OsMADS3 expression is downregulated in the apo1 mutant. These observations suggest that partial loss-of-identity of stamens

186

H.-Y. Hirano

in apo1 may be caused by the reduced activity of OsMADS3. If the other class C gene OsMADS58 is also downregulated in apo1, then the mutant phenotype, i.e., loss-of-meristem determinacy and homeotic transformation of stamens, might be the result of pleiotropy as a consequence of the reduction of expression of the two class C genes. Stamen identity also seems to be regulated by PISTILLOID STAMEN (PS): in mutants, stamens are partially transformed into pistils and chimeric organs of the stamen and pistil (Luo et al. 2006).

4.3

Lodicule Specification and Asymmetric Arrangement

Lodicules are floral organs specific to the grasses and are thought to be homologous to petals. Petal specification is regulated by the combination of class A and class B genes in Arabidopsis (for reviews, see Lohmann and Weigel 2002; Jack 2004). In a similar manner to that in class B mutants in Arabidopsis, lodicule identity is lost in the spw1 mutant in rice, and the lodicules are replaced by palea/lemma-like organs or leaf-like structures (Fig. 3C; Nagasawa et al. 2003). OsMADS2 knockdown lines show loss of lodicule identity (Prasad and Vijayraghavan 2003), suggesting that both AP3 and PI orthologs (SPW1 and OsMADS2, respectively) are involved in lodicule specification in rice (Fig. 4). In situ analyses show that class A genes, such as OsMADS15 (RAP1A) and OsMADS18 (FDRMADS7) (Jia et al. 2000; Kyozuka et al. 2000; Fornara et al. 2004), are expressed in the lodicules. To date, it is unclear if OsMADS14 (RAP1B) is expressed in the lodicule. As functional analysis of these genes is as yet incomplete, it is not known whether class A genes are required for lodicule specification. OsMADS14 and OsMADS18 are expressed in stamens and carpels, suggesting that these two genes are not repressed by class C genes in rice. Lodicules develop asymmetrically in the wild-type flower: two lodicules are initiated on the lemma side but not on the palea side (Fig. 1). Loss of function of OsMADS3 or OsMADS58 induces ectopic development of lodicules on the palea side in whorl 2, in addition to homeotic conversion of stamens into lodicules in whorl 3 (Fig. 3D,E; Yamaguchi et al. 2006). Both class C genes are expressed only on the palea side in whorl 2 in the wild-type flower. These findings suggest that these class C genes repress lodicule initiation on the palea side and are involved in the asymmetric arrangement of lodicules (Fig. 4).

5

Future Perspectives

The molecular genetic studies in rice described here indicate that the genetic mechanisms for regulating meristem maintenance and floral organ specification show considerable conservation between rice and Arabidopsis. It is likely that this conservation of gene function will also apply to other grass species, because the

II.5 Genetic Regulation of Meristem Maintenance

187

genes that regulate meristem maintenance and floral organ specification are common to those in maize (Bommert et al. 2005a, 2005b). In addition to conserved mechanisms, however, there is also evidence of genetic regulation that is unique to rice. With regard to meristem maintenance, an important research goal must be to isolate the genes responsible for the alternative pathway to the FON2–FON1 pathway. Another essential issue for understanding meristem maintenance in rice is identification of the key regulator that promotes stem cell identity, because the rice WUS ortholog is assumed to be unrelated to meristem maintenance on the basis of its expression pattern (Nardmann and Werr 2006). A complete understanding of rice flower development also requires the identification of the genes that act upstream and downstream of the floral homeotic genes. Finally, it will also be of interest to isolate the genes that specify grass-specific floral organs such as lemma, palea, and sterile lemmas. This would present a clear answer to the origin of these organs. Acknowledgements In this chapter, we have cited only the reviews for Arabidopsis studies due to space limitations. We apologize to those whose original work could not be included in this chapter. We thank the members of our laboratory for helpful discussions and Drs. Yamaguchi and Suzaki for remarkable accomplishments in the developmental studies of the rice flower.

References Agrawal GK, Abe K, Yamazaki M, Miyao A, Hirochika H (2005) Conservation of the E-function for floral organ identity in rice revealed by the analysis of tissue culture-induced loss-of-function mutants of the OsMADS1 gene. Plant Mol Biol 59:125–135 Alvarez J, Smyth DR (1999) CRABS CLAW and SPATULA, two Arabidopsis genes that control carpel development in parallel with AGAMOUS. Development 126:2377–2386 Bäurle I, Laux T (2003) Apical meristems: the plant’s fountain of youth. Bioessays 25:961–970 Bommert P, Lunde C, Nardmann J, et al. (2005a) thick tassel dwarf1 encodes a putative maize ortholog of the Arabidopsis CLAVATA1 leucine-rich repeat receptor-like kinase. Development 132:1235–1245 Bommert P, Satoh-Nagasawa N, Jackson D, Hirano H-Y (2005b) Genetics and evolution of inflorescence and flower development in grasses. Plant Cell Physiol 46:69–78 Bowman J, Smyth D (1999) CRABS CLAW, a gene that regulates carpel and nectary development in Arabidopsis, encodes a novel protein with zinc finger and helix—loop–helix domains. Development 126:2387–2396 Carles CC, Fletcher JC (2003) Shoot apical meristem maintenance: the art of a dynamic balance. Trends Plant Sci 8:394–401 Chung Y-Y, Kim SR, Kang HG, et al. (1995) Characterization of two rice MADS box genes homologous to GLOBOSA. Plant Sci 109:45–56 Coen ES, Meyerowitz EM (1991) The war of the whorls: genetic interactions controlling flower development. Nature 353:31–37 Fornara F, Parenicova L, Falasca G, et al. (2004) Functional characterization of OsMADS18, a member of the AP1/SQUA subfamily of MADS box genes. Plant Physiol 135:2207–2219 Ikeda K, Nagasawa N, Nagato Y (2005) Aberrant panicle organization 1 temporally regulates meristem identity in rice. Dev Biol 282:349–360 Itoh J-I, Nonomura K-I, Ikeda K, et al. (2005) Rice plant development: from zygote to spikelet. Plant Cell Physiol 46:23–47 Jack T (2004) Molecular and genetic mechanisms of floral control. Plant Cell 16:S1–S17

188

H.-Y. Hirano

Jeon JS, Jang S, Lee S, et al. (2000) Leafy hull sterile1 is a homeotic mutation in a rice MADS box gene affecting rice flower development. Plant Cell 12:871–884 Jia HW, Chen R, Cong B, Cao KM, Sun CR, Luo D (2000) Characterization and transcriptional profiles of two rice MADS-box genes. Plant Sci 155:115–122 Kang H-G, Jeon J-S, Lee S, An G (1998) Identification of class B and class C floral organ identity genes from rice plants. Plant Mol Biol 38:1021–1029 Kater MM, Dreni L, Colombo L (2006) Functional conservation of MADS-box factors controlling floral organ identity in rice and Arabidopsis. J Exp Bot 57:3433–3444 Komatsu M, Chujo A, Nagato Y, Shimamoto K, Kyozuka J (2003) FRIZZY PANICLE is required to prevent the formation of axillary meristems and to establish floral meristem identity in rice spikelets. Development 130:3841–3850 Kondo T, Sawa S, Kinoshita A, et al. (2006) A plant peptide encoded by CLV3 identified by in situ MALDI-TOF MS analysis. Science 313:845–848 Kurakawa T, Ueda N, Maekawa M, et al. (2007) Direct control of shoot meristem activity by a cytokinin-activating enzyme. Nature 445:652–655 Kyozuka J, Shimamoto K (2002) Ectopic expression of OsMADS3, a rice ortholog of AGAMOUS, caused a homeotic transformation of lodicules to stamens in transgenic rice plants. Plant Cell Physiol 43:130–135 Kyozuka J, Kobayashi T, Morita M, Shimamoto K (2000) Spatially and temporally regulated expression of rice MADS box genes with similarity to Arabidopsis class A, B and C genes. Plant Cell Physiol 41:710–718 Lee S, Kim J, Son J-S, et al. (2003) Systematic reverse genetic screening of T-DNA tagged genes in rice for functional genomic analyses: MADS-box genes as a test case. Plant Cell Physiol 44:1403–1411 Lee DY, Lee J, Moon S, Park SY, An G (2007) The rice heterochronic gene SUPERNUMERARY BRACT regulates the transition from spikelet meristem to floral meristem. Plant J 49:64–78 Lohmann JU, Weigel D (2002) Building beauty: the genetic control of floral patterning. Dev Cell 2:135–142 Luo H, Li Y, Yang Z, et al. (2006) Fine mapping of a pistilloid-stamen (PS) gene on the short arm of chromosome 1 in rice. Genome 49:1016–1022 Nagasawa N, Miyoshi M, Kitano H, Satoh H, Nagato Y (1996) Mutations associated with floral organ number in rice. Planta 198:627–633 Nagasawa N, Miyoshi M, Sano Y, et al. (2003) SUPERWOMAN 1 and DROOPING LEAF genes control floral organ identity in rice. Development 130:705–718 Nardmann J, Werr W (2006) The shoot stem cell niche in angiosperms: expression patterns of WUS orthologues in rice and maize imply major modifications in the course of mono- and dicot evolution. Mol Biol Evol 23:2492–2504 Prasad K, Vijayraghavan U (2003) Double-stranded RNA interference of a rice PI/GLO paralog, OsMADS2, uncovers its second-whorl-specific function in floral organ patterning. Genetics 165:2301–2305 Prasad K, Sriram P, Kumar CS, Kushalappa K, Vijayraghavan U (2001) Ectopic expression of rice OsMADS1 reveals a role in specifying the lemma and palea, grass floral organs analogous to sepals. Dev Genes Evol 211:281–290 Suzaki T, Sato M, Ashikari M, Miyoshi M, Nagato Y, Hirano H-Y (2004) The gene FLORAL ORGAN NUMBER1 regulates floral meristem size in rice and encodes a leucine-rich repeat receptor kinase orthologous to Arabidopsis CLAVATA1. Development 131:5649–5657 Suzaki T, Toriba T, Fujimoto M, Tsutsumi N, Kitano H, Hirano H-Y (2006) Conservation and diversification of meristem maintenance mechanism in Oryza sativa: function of the FLORAL ORGAN NUMBER2 gene. Plant Cell Physiol 47:1591–1602 Theissen G, Becker A, Di Rosa A, et al. (2000) A short history of MADS-box genes in plants. Plant Mol Biol 42:115–149 Xiao H, Wang Y, Liu DF, et al. (2003) Functional analysis of the rice AP3 homologue OsMADS16 by RNA interference. Plant Mol Biol 52:957–966

II.5 Genetic Regulation of Meristem Maintenance

189

Yamaguchi T, Hirano HY (2006) Function and diversification of MADS-box genes in rice. TSW Dev Embryol 1:99–108 Yamaguchi T, Nagasawa N, Kawasaki S, Matsuoka M, Nagato Y, Hirano H-Y (2004) The YABBY gene DROOPING LEAF regulates carpel specification and midrib development in Oryza sativa. Plant Cell 16:500–509 Yamaguchi T, Lee YD, Miyao A, Hirochika H, An G, Hirano H-Y (2006) Functional diversification of the two C-class genes, OSMADS3 and OSMADS58, in Oryza sativa. Plant Cell 18:15–28

II.6

Genetic Dissection of Sexual Reproduction in Rice (Oryza sativa L.) Ken-Ichi Nonomura1,2(* ü ) and Shinichiro Yamaki1

1

Introduction

Sexual reproduction is the most important step in increasing the genetic diversity of offspring. It is defined by two major events: meiosis and fertilization. Meiosis is a crucial event to form haploid spores and gametes, and is characterized by a single round of premeiotic DNA replication followed by two continuous rounds of chromosome segregation. Homologous recombination, an essential feature of meiosis, results in generating new haplotypes by shuffling alleles. Fertilization is achieved by the fusion of two gametes and produces new genotypes of diploid cell or zygote. In contrast, reproductive isolation contributes to establishing genetic stability of species, rather than genetic diversity, in most eukaryotes, and is achieved by various mechanisms, for example, the differential fitness of the gametophyte or zygote (Dobzhansky 1951; Stebbins 1958). Biological species are generally defined as groups of interbreeding populations that are reproductively isolated (Mayr 1942). Studies of sexual reproduction of crop and pasture plants often leads to advances in applied breeding. This is because a breeding system depends on the use of reproductive mechanisms to create genetic diversity. In this case, reproductive isolation becomes an obstacle to introducing useful genes into cultivars. Thus, analyses of genes controlling sexual reproduction are required for the establishment of efficient breeding programs by interspecific hybridization. In this chapter, we will focus on genetic dissection of plant sexual reproduction mainly using artificially induced mutations of rice, a model cereal. This research field has just begun in rice, so its situation is far from the advanced state of Arabidopsis research. However, rice genetics has revealed some reproductive systems of monocot plants that differ from

1 Experimental Farm/Plant Genetics Lab., National Institute of Genetics, Yata 1111, Mishima, Shizuoka 411-8540, Japan e-mail: [email protected] 2 Department of Life Science, Graduate University for Advanced Studies/Sokendai, Yata1111, Mishima, Shizuoka 411-8540, Japan

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

191

This page intentionally blank

192

K.-I. Nonomura and S. Yamaki

dicots. We hope that both monocot and dicot model plants will equally contribute to elucidating the genetic networks controlling plant sexual reproduction.

2

Reproductive Organ Development

The class I KNOX gene OSH1 (ORYZA SATIVA HOMEOBOX1), a molecular marker of meristematic indeterminate cells in rice (Sato et al. 1996), is expressed in the floral meristem and receptacle in the early stages of floral development. The expressed area of OSH1 becomes smaller during floral organ development, in which down-regulation of OSH1 occurs at first in stamen primordia and subsequently in carpel and ovule primordia (Yamaguchi et al. 2004). This indicates that development of male organs and germ cells starts earlier than that of female. After floral organs have acquired their identity, the archesporial cells, representing primordial germ cells in plants, differentiate just beneath the epidermis of male and female organs (Fig. 1). In cross sections of the stamen, the archesporial cells differentiate at the four corners of hypodermal cells (Raghavan 1988). Each archesporial cell divides periclinally to give rise to primary parietal cells (PPC) towards the exterior and a larger primary sporogenous cell (PSC). The PSC divides several times and finally matures into microgametocytes or pollen mother cells (PMC). PPCs further divide periclinally into two layers of secondary parietal cells (SPC) and finally form hypodermal anther walls composed of the endothecium, middle layer, and tapetum layer. In most monocot plants, the middle layer and tapetum are formed from an inner SPC, whereas they are formed from an outer SPC in most dicots (Davis 1966). The cell lineage in rice ovule development is less understood than in anther development. Soon after the stamen primordia protrude, a carpel primordium arises from the lemma side of the floral meristem (Fig. 1E), and morphological observations suggest that the remaining part of the meristem becomes an ovule primordium as a terminal organ. The ring-shaped outer and inner integuments emerge from chalaza (Fig. 1F), and the nucellus is finally enclosed with integuments in the ovule (Fig. 1G,H). Though it is difficult to conclude that the ovule primordium is directly differentiated from the floral meristem in rice, histological analyses of several mutants give some evidence to support it. The fon1 ( floral organ number1) mutant, which loses floral meristem determinacy, often forms malformed outer integument (Nagasawa et al. 2003; Suzaki et al. 2004; Fig. 2A). The gym (gypsy embryo) mutant, which exhibits incomplete determinacy of floral meristem, fails to separate the integument primordium into inner and outer integuments (Yamaki et al. 2005; Fig. 2B). These results indicate that the mutation in floral meristem affects ovule morphogenesis. In addition, the DL (DROOPING LEAF) gene, which determines carpel identity, is not expressed in ovule primordium (Yamaguchi et al. 2004; Fig. 2C), suggesting that the ovule primordium is differentiated from the tissue distinct from the carpel. This observation is inconsistent

II.6 Genetic Dissection of Sexual Reproduction in Rice (Oryza sativa L.)

193

Fig. 1 Histological features of male and female reproductive organ development in the wild-type flower. A–D Transverse sections of an anther locule. A Archesporial cells (AC) develop just beneath the epidermis (EP). B AC periclinally divide into primary parietal cells (PPC) and primary sporogenous cells (PSC). PPC further divide periclinally into secondary parietal cells (SPC) (arrow). C PPC periclinally divide into two layers of SPC, the inner of which further divides into tapetum cells (TA) and middle layer cells (ML). The outer SPC becomes endothecium (EN). D Pollen mother cells (PMC) undergo meiosis. E, F Transverse sections of the ovule primodium. E Carpel primordium (CA) is differentiated from the lemma side of the floral meristem (FM). The rest of the FM becomes ovule primordium. F During development of ovule primordium, the integument primordium separates into inner and outer integuments. G A single megaspore mother cell (MMC, colored in red) becomes enlarged just beneath the nucellar epidermis. H Completion of the inner integment (colored in yellow) enclosing the nucellus is a sign of female meiosis completion. The charazal-most megaspore (arrow) survives and three spores at the micropylar end (arrowheads) degenerate. I Cell lineage of male and female germ cells in rice

with the proposal in Arabidopsis, in which the ovule primordia are differentiated from the placenta which lies within the carpel margin (Skinner et al. 2004). In the developing ovule, a hypodermal female archespore is surrounded by nucellar cells and differentiates into a megaspore mother cell (MMC) (Fig. 1G). It is still unclear whether the nucellar cells adjacent to the MMCs differentiate to bear functions as nurse cells, like tapetum cells in the anther. Though the OsMADS13

194

K.-I. Nonomura and S. Yamaki

Fig. 2 Rice mutations and genes affecting ovule development. A The fon1 mutant flower often forms a malformed outer integument (arrow). B The gym mutant does not undergo separation of the integment primordium, and instead develops a single thick integument (asterisk). C The DL gene is expressed in carpel (CA) but not in ovule (OV) primordium. D OsMADS13 is expressed specifically in the ovule

gene is known to express specifically in the rice ovule (Lopez-Dee et al. 1999; Fig. 2D), further analyses using flower mutants and cell markers will be necessary to ascertain details of cell lineage and genetic regulation in ovule and MMC development in rice.

II.6 Genetic Dissection of Sexual Reproduction in Rice (Oryza sativa L.)

3 3.1

195

Sporogenesis The Receptor Kinase MSP1 (MULTIPLE SPOROCYTES 1) Determines Nursery Cell Fate

Only a few genetic processes are known in initiation and development of archesporial cells in plants. The rice gene MSP1 determines the fate of nursery cells adjacent to germ cells (Nonomura et al. 2003). A Tos17 insertional mutation in MSP1 gives rise to nursery cells converted into germ cells, and results in an increase in the number of sporocytes in both male and female organs, despite a lack of normal development of the inner anther walls including a tapetum layer. MSP1 encodes a serine/threonine kinase with a leucine-rich repeat receptor and expresses specifically in inner wall cells of anthers including the tapetum, but not in sporocytes. These results indicate that in anthers, PPC and their derivatives originally possess the potential to become germ cells, but differentiate to nursery cells due to inhibition by MSP1 function. In Arabidopsis, the EMS1 (EXCESS MICROSPOROCYTES1)/ EXS (EXTRA SPOROGENOUS CELLS) kinase is highly homologous to rice MSP1 and its mutation leads to disorganization of the anther wall, like the msp1 mutation, but no phenotypic aberration is observed in female organs (Sorensen et al. 2002; Zhao et al. 2002). A bisexual function of rice MSP1 may reflect a difference in female organ development between monocot and dicot plants.

3.2

Upstream and Downstream Factors of the MSP1 Signaling Pathway

The SPL (SPOROCYTELESS) gene of Arabidopsis encodes a putative transcription factor regulating sporogenesis and its mutation leads to a lack of sporocyte formation and SPC division (Yang et al. 1999). Recently, Ito et al. (2004) revealed that transcription of the SPL gene is promoted by a MADS-box transcription factor, AGAMOUS, in Arabidopsis, though the rice ortholog of SPL has not yet been identified. In msp1 and ems1/exs mutants, sporocytes are formed and enter into meiosis, in contrast to the spl mutant, indicating that SPL function is located upstream of the MSP1 signaling pathway. Recently, a germ cell-specific gene, MEL1 (MEIOSIS ARRESTED AT LEPTOTENE), was identified in rice, and a Tos17 insertional mutation bisexually affects premeiotic mitosis of PSCs and meiotic chromosome condensation (Nonomura et al. 2007). The MEL1 gene encodes an ARGONAUTE family protein, which is a key component of small RNA-mediated gene silencing (Vaucheret 2006). In situ expression of the MEL1 gene marked archesporial initial cells and PSCs. As a downstream factor of anther wall development, the rice UDT1 (UNDEVELOPED TAPETUM1) gene encodes a transcription factor with a basic helix—loop–helix structure (Jung et al. 2005). T-DNA or Tos17 insertional mutants of the UDT1 gene lack a mature tapetum layer.

196

K.-I. Nonomura and S. Yamaki

The gibberellin signaling also relates to anther development in rice. The OsGAMYB is a positive transcriptional regulator of gibberellin signaling, and its loss-of-function mutation impairs tapetum and pollen development (Kaneko et al. 2004). The OsGAMYB expresses in stamen primordia and tapetum cells, and is negatively regulated by the microRNA miR159 in flowers (Tsuji et al. 2006). This fact indicates that RNA interfering machinery is required not only for vegetative organ development (Nishimura et al. 2002), but also for reproductive organ development in rice.

4 4.1

Meiosis Overview of Rice Meiosis

Male meiosis of rice starts after four-layered anther wall formation is completed. PMCs in the same anther locule enter synchronously into premeiotic DNA replication, and rapidly shift to the meiotic process (Nonomura et al. 2004a). On the other hand, it is a sign of female meiosis completion that the inner integuments completely enclose the nucellus, in which female meiosis tends to be loosely synchronized with the male one (Itoh et al. 2005). Zhang and Zhu (1987) estimated a total time for all meiotic processes of about 30 h (Table 1). Leptotene and zygotene, the stages in which meiotic chromosomes start meiosis-specific condensation and synapsis, make up half of the entire meiotic process. The second longest stage is pachytene (5 h), the stage in which homologous chromosome synapsis is completed. This indicates that preparation for homologous chromosome synapsis and recombination are important events to undergo precise meiosis. For details of rice meiotic stages, refer to Itoh et al. (2005).

Table 1 The absolute time taken for completing each meiotic stage in pollen mother cells of rice [Zhang and Zhu (1987) with minor modifications] Stage Time (min) Number of cells observed Leptotene/zygotene Pachytene Diplotene Diakinesis Metaphase I Anaphase I Telophase I Prophase II Metaphase II Anaphase II Telophase II Total

965 300 23 123 154 19 27 77 96 38 38 1860

251 78 6 32 40 5 7 20 25 10 10 484

II.6 Genetic Dissection of Sexual Reproduction in Rice (Oryza sativa L.)

4.2

197

Forward Genetic Approach

The first knowledge of molecular genetics in rice meiosis was gleaned by identifying the PAIR1 (HOMOLOGOUS PAIRING ABERRATION IN RICE MEIOSIS1) and PAIR2 genes. The PAIR1 gene encodes a novel protein with putative DNA binding and coiled-coil motifs, and is expressed in both male and female meiocytes (Nonomura et al. 2004b). A Tos17 insertional mutation in the PAIR1 gene caused a failure not only in homologous chromosome synapsis but also in the typical condensation of meiotic chromosomes. Interestingly, despite a failure of meiotic condensation at the pachytene stage, meiotic chromosomes in pair1 mutants achieve full condensation at metaphase I. This indicates that a meiosis-specific condensation mechanism exists independent of the mitotic machinery and is promoted by PAIR1 function. The PAIR2 gene encodes a HORMA (Hop1, Rev7, Mad2)-domain protein homologous to yeast HOP1 (Hollingsworth et al. 1990) and Arabidopsis ASY1 (ASYNAPSIS1) (Caryl et al. 2000), and a Tos17 mutation of the gene displays complete lack of homologous chromosome synapsis (Nonomura et al. 2004a). The PAIR2 protein may play an important role in establishing the synaptonemal

Fig. 3 The meiotic protein PAIR2 associates with axial elements of meiotic chromosomes. After homologous chromosome synapsis is complete, the PAIR2 filamentous signal (arrow) gradually becomes discontinuous and faint (arrowhead)

198

K.-I. Nonomura and S. Yamaki

complex, by associating transiently with chromosomal axial elements (Fig. 3) and recruiting transverse filaments and/or central elements between homologous pairs of axial elements (Nonomura et al. 2006).

4.3

Reverse Genetic and Biochemical Approaches

The rice OsRAD21-4 protein is an ortholog of yeast REC8, which is a meiosisspecific component of the cohesin complex (Zhang et al. 2006), essential for establishing sister chromatid cohesion and reductional division of meiotic chromosomes (Klein et al. 1999; Watanabe and Nurse 1999). A knockdown line of OsRAD21-4 by RNA interference (RNAi) displays hypercondensation, precocious segregation of homologous pairs and sister chromatids, and segmentation of meiotic chromosomes (Zhang et al. 2006). This result suggests that the meiotic cohesin complex of rice plays some roles in homologous synapsis as well as in sister chromatid cohesion. DMC1, a meiosis-specific homologue of E. coli RecA family proteins, is essential for DNA strand exchange during homologous recombination in many organisms (Bishop et al. 1992; Masson and West 2001). The rice genome contains two paralogs, OsDMC1A and OsDMC1B (Kathiresan et al. 2002). A biochemical study has revealed that OsDMC1A catalyses DNA strand exchange coupled with DNAdependent ATP hydrolysis (Kant et al. 2005; Rajanikant et al. 2006).

4.4

Homologous Pairing Is Not Necessary for Chromosome Alignment at the Spindle Midzone

When chromosomes align at the spindle midzone in metaphase I, a balance of bipolar forces generated by motors associated with kinetochore microtubules is responsible for the establishment of the metaphase plate (Mitchison 1989). In most eukaryotes undergoing prereductional division, including rice, sister chromatids remain associated with each other through their centromeres by the cohesin function, and move together to the same pole in meiosis I (Watanabe and Nurse 1999). In fission yeast, REC8 localization at the inner centromere during premeiotic S phase leads to monopolar orientation of the sister kinetochores (Watanabe et al. 2001). In rice pair1 and pair2 mutants, unpaired univalents can align, but not precisely, at the spindle midzone, whereas several unpaired univalents lag in division to either pole in anaphase I (Nonomura et al. 2004a,b). Alignment of univalents at the midzone has also been reported in meiotic mutants dys1 and dys2 of maize (Chan and Cande 1998). Nonomura et al. (2006) revealed that sister kinetochores of most unpaired univalents in pair2 meiocytes are captured by bipolar spindle microtubules, as in mitosis, in anaphase I (Fig. 4A, B). Sister centromeres still adhere to each other, probably due to function of the cohesin complex including OsRad21-4,

II.6 Genetic Dissection of Sexual Reproduction in Rice (Oryza sativa L.)

199

Fig. 4 Bipolar attachment of the kinetochore to spindle microtubules is required for reliable disjunction of meiotic chromosomes in rice meiosis I. A–D A pair2 mutant meiocyte at metaphase I/anaphase I transition, in which unpaired univalents, centromere (OsCenH3, centromere-specific histone H3 variant) and spindle microtubules are pseudo-colored in blue, magenta and green, respectively. A Overview of the metaphase I/anaphase I transition stage. B Bipolar spindle microtubules capture a single kinetochore (centromere) on an unpaired univalent. C A centromere region on a univalent is elongated by the pulling force from bipolar microtubules. D One of the bipolar microtubules is finally detached at random

and the pulling force of kinetochore microtubules elongates centromere structure toward a bipolar orientation in unpaired univalents (Fig. 4C). Either of the bipolar microtubules capturing a univalent kinetochore is finally detached at random (Fig. 4D), resulting in nondisjunction of chromosomes. This finding indicates that homologous pairing is not necessarily required for chromosome alignment at the spindle midzone in plant meiosis I, but is required for reliable disjunction with precise positioning of the metaphase plate.

200

4.5

K.-I. Nonomura and S. Yamaki

Meiosis and Nursery Cells

The tapetum layer is thought to play an important role in pollen development by secreting nutritive materials and other substances such as a β-glucanase into the anther locule (Chaubal et al. 2000). In msp1 mutant anthers, male meiosis is arrested at very early stages, from leptotene to diakinesis, whereas meiotic events such as homologous chromosome synapsis and condensation are processed normally (Nonomura et al. 2003). This may be attributable to tapetum cell layers being completely absent in msp1 anthers. In contrast, however, the male meiocytes of the Arabidopsis mutant ems1/exs, orthologous to rice msp1 mutation, can complete meiosis and form tetrad spores (Sorensen et al. 2002; Zhao et al. 2002). The MMC of rice gym mutant carried out normal meiosis even in the ovule with underdeveloped inner integuments (Yamaki et al. 2005). In contrast, the proper elongation of inner integument is necessary to accomplish the MMC meiosis in Arabidopsis (Robinson-Beers et al. 1992; Elliott et al. 1996; Klucher et al. 1996). These may indicate the existance of different regulatory system of nursery-cell development between rice and Arabidopsis.

5

Gametogenesis

After male meiosis, each microspore of rice enlarges and undergoes asymmetric mitosis (Fig. 5A–C), in which one of the daughter nuclei is adjacent to the wall and finally becomes a generative cell nested in a vegetative cell (Fig. 5D). The generative cell divides into two sperm cells before pollen tube germination (Fig. 5E), and a three-celled pollen grain, a male gametophyte, is formed (Fig. 5F). Few genes have been cloned to date that are involved in male gametogenesis of rice. In contrast, in Arabidopsis, many gametophytic genes have been identified and cloned by using T-DNA tagged lines and antibiotic marker selection (reviewed in McCormick 2004). In rice, Moritoh et al. (2005) reported that RNAi knockdown plants of OsGEN-L, a member of the RAD2/XPG nuclease family, exhibit a defect in early microsporogenesis, resulting in male sterility. The OsGEN-L protein possesses flap endonuclease activity and DNA binding activity, indicating that rice microsporogenesis requires DNA metabolism. Embryo sac development in rice is monosporic and conforms to the Polygonum type (Maheshwari 1950; Davis 1966). Of the four spores derived from a single MMC in the ovule, only the chalazal-most megaspore survives (Fig. 1H). It undergoes three mitotic divisions and cellularization, giving rise to the seven-celled embryo sac, which consists of one egg cell, two synergids, one central cell and three antipodal cells (Itoh et al. 2005). The antipodal cells further divide and multicellularize at the charazal pole, as do those in maize (Huang and Sheridan 1994). Moreover, few genes involved in female gametogenesis have been isolated in rice. The msp1 mutant, which generates supernumerary germ cells, undergoes aberrant gametogenesis (Nonomura et al. 2003). Plural megaspores in a mutant ovule often

II.6 Genetic Dissection of Sexual Reproduction in Rice (Oryza sativa L.)

201

Fig. 5 Microgametogenesis in rice. A Tetrad spores. B A unicellular microspore. C First pollen mitosis. D A bicellular-stage cell in which a generative cell is nested in a vegetative cell. E Second pollen mitosis of generative cells. F A three-celled pollen grain, a male gametophyte

result in a single but aberrant embryo sac with supernumerary nuclei and many cell wall partitions. Interestingly, despite aberrant cell composition, more than 60% of msp1 flowers generate a functional embryo sac, in which about 40% of antipodal tissues develop normally at the charazal side. This suggests that the ovule may provide cues for embryo sac formation via intercellar contacts in rice. However, even in Arabidopsis, it is still unclear whether embryo sac formation is directed in a cell-autonomous and/or non-cell-autonomous fashion (reviewed by Reiser and Fischer 1993). More genetic analyses are also needed to elucidate female gametogenesis in rice.

6

Future Perspectives on Studies in Rice Sexual Reproduction

The knowledge described in this chapter is not easy to apply at the present time in breeding programs for cultivated rice. However, elucidating the genetic network of the reproductive process is a first step toward overcoming reproductive isolation between cultivars and wild species. The genus Oryza consists of more than 20 wild and two cultivated species (Vaughan and Morishima 2003). The cultivars O. sativa and O. glaberrima and five wild species are classified into the AA genome type. The rest of the wild species other than those with the AA genome are classified into

202

K.-I. Nonomura and S. Yamaki

diploid genome types BB, CC, EE, FF and GG, and tetraploid BBCC, CCDD and HHJJ types (see Chap.III.4 for details). These species distant from domesticated rice carry many genes resistant to insects and diseases (Brar and Khush 1997), but have been rarely used for rice breeding programs so far, because it is difficult to obtain fertile offspring. This is due to the lower affinity of the AA chromosomes to the non-AA chromosomes in meiosis of F1 plants. This situation suggests the possibility that the manipulation of meiotic genes promotes the introgression of chromosomal fragments from wild relatives into cultivars. Comparing genomic approaches among Oryza species is necessary to understand differentiation within the genus Oryza, not only for meiotic genes but also for all genes related to reproduction. All researches on plant sexual reproduction lead to breeding innovations. Acknowledgments We thank Drs. Y. Yamagata and A. Yoshimura (Kyushu Univ., Japan) for kindly providing photos of rice microgametegenesis.

References Bishop DK, Park D, Xu L, Kleckner N (1992) DMC1: a meiosis-specific yeast homolog of E. coli recA required for recombination, synaptonemal complex formation, and cell cycle progression. Cell 69:439–456 Brar DS, Khush GS (1997) Alien introgression in rice. Plant Mol Biol 35:35–47 Caryl AP, Armstrong SJ, Jones GH, Franklin FC (2000) A homologue of the yeast HOP1 gene is inactivated in the Arabidopsis meiotic mutant asy1. Chromosoma 109:62–71 Chan A, Cande WZ (1998) Maize meiotic spindles assemble around chromatin and do not require paired chromosomes. J Cell Sci 111:3507–3515 Chaubal R, Zanella C, Trimnell MR, Fox TW, Albertsen MC, Bedinger P (2000) Two male-sterile mutants of Zea mays (Poaceae) with an extra cell division in the anther wall. Am J Bot 87:1193–1201 Davis GL (1966) Systematic embryology of the angiosperms. Wiley, New York Dobzhansky T (1951) Genetics and the origin of species, 3rd edn. Columbia University Press, New York Elliott RC, Betzner AS, Huttner E, et al. (1996) AINTEGUMENTA, an APETALA2-like gene of Arabidopsis with pleiotropic roles in ovule development and floral organ growth. Plant cell 8:155–168 Hollingsworth NM, Goetsch L, Byers B (1990) The HOP1 gene encodes a meiosis-specific component of yeast chromosomes. Cell 61:73–84 Huang BQ, Sheridan WF (1994) Female gametophyte development in maize: microtubular organization and embryo sac polarity. Plant Cell 6:845–861 Ito T, Wellmer F, Yu H, et al. (2004) The homeotic protein AGAMOUS controls microsporogenesis by regulation of SPOROCYTELESS. Nature 430:356–360 Itoh J, Nonomura KI, Ikeda K, et al. (2005) Rice plant development: from zygote to spikelet. Plant Cell Physiol 46:23–47 Jung KH, Han MJ, Lee YS, et al. (2005) Rice Undeveloped Tapetum1 is a major regulator of early tapetum development. Plant Cell 17:2705–2722 Kaneko M, Inukai Y, Ueguchi-Tanaka M, et al. (2004) Loss-of-function mutations of the rice GAMYB gene impair alpha-amylase expression in aleurone and flower development. Plant Cell 16:33–44 Kant CR, Rao BJ, Sainis JK (2005) DNA binding and pairing activity of OsDmc1, a recombinase from rice. Plant Mol Biol 57:1–11

II.6 Genetic Dissection of Sexual Reproduction in Rice (Oryza sativa L.)

203

Kathiresan A, Khush GS, Bennett J (2002) Two rice DMC1 genes are differentially expressed during meiosis and during haploid and diploid mitosis. Sex Plant Reprod 14:257–267 Klein F, Mahr P, Galova M, et al. (1999) A central role for cohesins in sister chromatid cohesion, formation of axial elements, and recombination during yeast meiosis. Cell 98:91–103 Klucher KM, Chow H, Reiser L, Fischer RL (1996) The AINTEGUMENTA gene of Arabidopsis required for ovule and female gametophyte development is related to the floral homeotic gene APETALA2. Plant Cell 8:137–153 Lopez-Dee ZP, Wittich P, Pè ME, et al. (1999) OsMADS13, a novel rice MADS-box gene expressed during ovule development. Dev Genet 25:237–244 Maheshwari P (1950) An introduction to the embryology of angiosperms. McGraw-Hill, New York Masson JY, West SC (2001) The Rad51 and Dmc1 recombinases: a non-identical twin relationship. Trends Biochem Sci 26:131–136 Mayr E (1942) Systematics and the origin of species. Columbia University Press, New York McCormick S (2004) Control of male gametophyte development. Plant Cell 16 Suppl: S142–S153 Mitchison TJ (1989) Mitosis: basic concepts. Curr Opin Cell Biol 1:67–74 Moritoh S, Miki D, Akiyama M, et al. (2005) RNAi-mediated silencing of OsGEN-L (OsGEN-like), a new member of the RAD2/XPG nuclease family, causes male sterility by defect of microspore development in rice. Plant Cell Physiol 46:699–715 Nagasawa N, Miyoshi M, Sano Y, et al. (2003) SUPERWOMAN1 and DROOPING LEAF genes control floral organ identity in rice. Development 130:705–718 Nishimura A, Ito M, Kamiya N, Sato Y, Matsuoka M (2002) OsPNH1 regulates leaf development and maintenance of the shoot apical meristem in rice. Plant J 30:189–201 Nonomura KI, Miyoshi K, Eiguchi M, et al. (2003) The MSP1 gene is necessary to restrict the number of cells entering into male and female sporogenesis and to initiate anther wall formation in rice. Plant Cell 15:1728–1739 Nonomura KI, Makano M, Murata K, et al. (2004a) An insertional mutation in the rice PAIR2 gene, the ortholog of Arabidopsis ASY1, results in a defect in homologous chromosome pairing during meiosis. Mol Gen Genomics 271:121–129 Nonomura KI, Nakano M, Fukuda T, et al. (2004b) The novel gene HOMOLOGOUS PAIRING ABERRATION IN RICE MEIOSIS1 of rice encodes a putative coiled-coil protein required for homologous chromosome pairing in meiosis. Plant Cell 16:1008–1020 Nonomura KI, Nakano M, Eiguchi M, Suzuki T, Kurata N (2006) PAIR2 is essential for homologous chromosome synapsis in rice meiosis I. J Cell Sci 119:217–225 Nonomura KI, Morohoshi A, Eiguchi M, Miyao A, Hirochika H, Kurata N (2007) A germ cell-specific gene of the ARGONAUTE family is essential for the progression of premeiotic mitosis and meiosis during sporogenesis in rice. Plant Cell 19:2583–2594 Raghavan V (1988) Anther and pollen development in rice (Oryza sativa). Am J Bot 75:183–196 Rajanikant C, Kumbhakar M, Pal H, Rao BJ, Sainis JK (2006) DNA strand exchange activity of rice recombinase OsDmc1 monitored by fluorescence resonance energy transfer and the role of ATP hydrolysis. FEBS J 273:1497–1506 Reiser L, Fischer RL (1993) The ovule and the embryo sac. Plant Cell 5:1291–1301 Robinson-Beers K, Pruitt RE, Gasser CS (1992) Ovule development in wild-type Arabidopsis and two female-sterile mutants. Plant Cell 4:1237–1249 Sato Y, Hong SK, Tagiri A, et al. (1996) A rice homeobox gene, OSH1, is expressed before organ differentiation in a specific region during early embryogenesis. Proc Natl Acad Sci USA 93:8117–8122 Skinner DJ, Hill TA, Gasser CS (2004) Regulation of ovule development. Plant Cell 16 Suppl: S32–S45 Sorensen A, Guerineau F, Canales-Holzeis C, Dickinson HG, Scott RJ (2002) A novel extinction screen in Arabidopsis thaliana identifies mutant plants defective in early microsporangial development. Plant J 29:581–594

204

K.-I. Nonomura and S. Yamaki

Stebbins GL (1958) The inviability, weakness and sterility of interspecific hybrids. Adv Genet 9:147–215 Suzaki T, Sato M, Ashikari M, Miyoshi M, Nagato Y, Hirano HY (2004) The gene FLORAL ORGAN NUMBER1 regulates floral meristem size in rice and encodes a leucine-rich repeat receptor kinase orthologous to Arabidopsis CLAVATA1. Development 131:5649–5657 Tsuji H, Aya K, Ueguchi-Tanaka M, et al. (2006) GAMYB controls different sets of genes and is differentially regulated by microRNA in aleurone cells and anthers. Plant J 47:427–444 Vaucheret H (2006) Post-transcriptional small RNA pathways in plants: mechanisms and regulations. Genes Dev 20:759–771 Vaughan DA, Morishima H (2003) Biosystematics of the genus Oryza. In: Smith CW (ed) Rice: origin, history, technology, and production. Wiley, New York, pp. 27–65 Watanabe Y, Nurse P (1999) Cohesin Rec8 is required for reductional chromosome segregation at meiosis. Nature 400:461–464 Watanabe Y, Yokobayashi S, Yamamoto M, Nurse P (2001) Pre-meiotic S phase is linked to reductional chromosome segregation and recombination. Nature 409:359–363 Yamaguchi T, Nagasawa N, Kawasaki S, Matsuoka M, Nagato Y, Hirano HY (2004) The YABBY gene DROOPING LEAF regulates carpel specification and midrib development in Oryza sativa. Plant Cell 16:500–509 Yamaki S, Satoh H, Nagato Y (2005) Gypsy embryo specifies ovule curvature by regulating ovule/integument development in rice. Planta 222:408–417 Yang WC, Ye D, Xu J, Sundaresan V (1999) The SPOROCYTELESS gene of Arabidopsis is required for initiation of sporogenesis and encodes a novel nuclear protein. Genes Dev 13:2108–2117 Zhang L, Tao J, Wang S, Chong K, Wang T (2006) The rice OsRad21-4, an orthologue of yeast Rec8 protein, is required for efficient meiosis. Plant Mol Biol 60:533–554 Zhang TB, Zhu H (1987) Duration of different stages of meiosis in pollen mother cells of rice. Rice Genet Newsl 4:67 Zhao DZ, Wang GF, Speal B, Ma H (2002) The excess microsporocytes1 gene encodes a putative leucine-rich repeat receptor protein kinase that controls somatic and reproductive cell fates in the Arabidopsis anther. Genes Dev 16:2021–2031

II.7

Molecular Studies on Cytoplasmic Male Sterility-associated Genes and Restorer Genes in Rice Sota Fujii1, Tomohiko Kazama1, and Kinya Toriyama1(* ü)

1

Introduction

Cytoplasmic male sterility (CMS) is a maternally inherited trait that results in the inability to produce fertile pollen, and is widely known in higher plants. Recent studies have identified that an aberrant chimeric gene in mitochondria possibly causes CMS in various plant species. In some CMS lines, pollen fertility is recovered by a nuclear-encoded gene known as a fertility restorer gene (Rf ). Rf genes are known to normalize the ectopic mRNAs or proteins derived from a chimeric gene (Hanson and Bentolila 2004). Therefore, CMS is assumed to be a phenomenon of nuclear-mitochondrial incompatibility. Fertility restorer genes for CMS of petunia (Petunia hybrida L.), radish (Raphanus sativus L.) and rice (Oryza sativa L.) have been recently cloned. Here we describe molecular studies on the CMS-associated mitochondrial chimeric genes and fertility restorer genes in rice, with special reference to the BT-CMS/Rf1 system.

2

Types of CMS in Rice

CMS in rice was first discovered by Katsuo and Mizushima (1958), by introducing the nuclear genome of a japonica cultivar, Fujisaka 5, into a Chinese wild rice strain, W1 (Oryza rufipogon Griff.). Since then, about 20 different types of CMS cytoplasm have been found in rice. For example, BT-CMS originating from Chinsurah boro II, LD-CMS from Lead Rice, WA-CMS from a wild abortive line and HL-CMS from Hong Lian wild rice have been identified (see Kinoshita 1997 for a review; Gramene website http://www.gramene.org/index.html). The gene symbol for representative cytoplasm is shown in Table 1.

1 Graduate School of Agricultural Science, Tohoku University, Sendai 981-8555, Japan e-mail: [email protected]; [email protected]; [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

205

206

S. Fujii et al.

Table 1 Examples of cytoplasmic male sterility (CMS) and fertility restorer (Rf) genes in rice Type Gene symbol Cytoplasm Rf Mode of restoration BT-CMS LD-CMS HL-CMS CW-CMS WA-CMS

[cms-bo] [cms-ld] [cms-HL] [cms-CW] [cms-WA]

Chinsurah BoroII Lead rice Hong Lian Chinese wild rice Wild abortive

Rf1 (Chr.10) Rf2 (Chr.2) Rf5 (Chr.10), Rf6 (Chr.10) Rf17 (Chr.4) Rf3 (Chr.1), Rf4(Chr.10)

Gametophytic Gametophytic Gametophytic Gametophytic Sporophytic

Table 2 Mode of fertility restoration and theoretical genotypes and phenotype of F2 plants derived from F1 plant with Rf rf A. Sporophytic restoration B. Gametophytic restoration Male gamete Rf Female gamete Rf rf

Male gamete

rf

Rf

rf

Rf Rf (Fertile) Rf rf (Fertile)

-

Female gamete Rf Rf (Fertile) Rf rf (Fertile)

Rf rf (Fertile) rf rf (Sterile)

Rf rf

The most studied CMS/Rf system in rice is the BT-CMS or ms-bo-type CMS and a fertility restorer gene, Rf1 (Shinjyo 1969). Rf1 has been shown to be located on chromosome 10 (Shinjyo 1975), and molecular cloning of Rf1 has been recently achieved (Kazama and Toriyama 2003; Akagi et al. 2004; Komori et al. 2004; Wang et al. 2006). Rf 2 for LD-CMS has been shown to lie on chromosome 2 (Shinjyo and Sato 1994). Rf3 and Rf4 for WA-CMS have been mapped on chromosomes 1 and 10, respectively (Yao et al. 1997; Zhang et al. 1997). Two fertility restorer genes, Rf5 and Rf6(t) for HL-CMS, have been shown to be located on chromosome 10 (Liu et al. 2004). Rf17 for CW-CMS has been mapped on chromosome 4 (Fujii and Toriyama 2005). We previously designated the fertility restorer gene for CW-CMS as Rfcw (Fujii and Toriyama 2005). In this chapter, according to the nomenclature of the Gramene website, the fertility restorer gene is called Rf17. The modes of fertility restoration are categorized into two types: gametophytic type and sporophytic type. When fertility restoration behaves in a sporophytic manner due to the action of a single dominant Rf gene, the pollen fertility of the F1 plant between a CMS line and a restorer line is determined by the genotype of the sporophyte, meaning that all the pollen with or without a restorer gene in the F1 plant is fertile and participates in the fertilization. Thus, 25% of F2 plants are presumed to be sterile (Table 2A). If the fertility restoration behaves in a gametophytic manner, the genotype of the individual pollen grain determines its own fertility. Therefore, pollen without a restorer gene in the F1 plant is unable to germinate on a stigma, and all the F2 plants are expected to be fertile (Table 2B). Most fertility restorer genes, except those for WA-CMS, are known to function

II.7 Molecular Studies on Cytoplasmic Male Sterility-associated Genes

207

Fig. 1 Pollen grains at flowering stage of WA-CMS (A), BT-CMS (B), LD-CMS (C), CW-CMS (D), wild-type (E), and pollen germination of a wild-type (F–H) and a CW-CMS line (I–K). In F and I, localization of starch grains is observed by staining pollen with I2-KI. In G and J, pollen grains are germinated on in vitro germination medium. In H and K, pollen tubes are visualized by staining stigma after pollination. Pollen of CW-CMS lacks germination ability. Bar = 50 mm

208

S. Fujii et al.

gametophytically on fertility restoration. On the contrary, the fertility restorer genes Rf3 and Rf4 for WA-CMS function in a sporophytic manner (Table 1; see Kinoshita 1997 for a review). The morphology of pollen grains at the flowering stage differs between CMS types. The anthers of WA-CMS are slender and milky white, and empty of mature pollen grains. The pollen grains of WA-CMS are empty of starch and unstainable by an I2-KI solution (Fig. 1A). The pollen grains of the BT-CMS line are globular, smaller in size and slightly stainable by an I2-KI solution, as shown in Fig. 1B. The starch accumulation of the pollen grains of LD-CMS is slightly more abundant than that of BT-CMS (Fig. 1C), but less than that of wild-type (Fig. 1E). The mature pollen of CW-CMS is particularly interesting. It is morphologically normal under optical microscopy (Fig. 1D), but lacks germination ability (Fig. 1F–K). After anthesis, localization of starch grains to the polar site of the germpore is observed in wild-type pollen (Fig. 1F), but not in CW-CMS pollen (Fig. 1I). The pollen grains of CW-CMS do not germinate on an in vitro germination medium (Fig. 1J) or on the stigma after pollination in vivo (Fig. 1K), while wild-type pollen grains do (Fig. 1G and H).

3

Mitochondrial Chimeric Genes

Mitochondrial chimeric genes possibly involved in CMS of rice were first discovered in BT-CMS. In the BT-CMS line, pollen abortion is initiated after pollen mitosis. However, pollen fertility can be restored gametophytically by the gene product of a single dominant nuclear gene, Rf1. The mitochondrial genome of the BT-cytoplasm contains two duplicated copies of the atp6 gene encoding subunit 6 of the ATPase complex (Fig. 2A). One copy is completely identical to Nipponbare atp6 and designated as N-atp6. Another copy contains an altered 3′ UTR sequence, which is designated as B-atp6. A unique sequence (orf 79), which encodes 79 amino acid open reading frames (ORFs), is located downstream from the B-atp6 genes (Iwabuchi et al. 1993; Akagi et al. 1994). The orf 79 encodes a putative transmembrane protein. The N-terminal 29-amino acid sequences show 61% identity to rice mitochondrial cytochrome oxidase subunit I (Akagi et al. 1994). In the BT-CMS line, B-atp6 is transcribed as a 2.0-kb mRNA consisting of B-atp6 and the orf 79 sequences, whereas two discontinuous mRNAs of 1.5 kb encoding for atp6 and 0.45 kb for orf 79 are generated from the 2.0-kb mRNA by RNA processing in the presence of Rf1 (Fig. 2B). The findings of B-atp6 and orf 79 were a step toward understanding the nature of rice CMS. An HL-CMS line was found to contain a gene structure similar to B-atp6-orf 79, in which orfH79 is located subsequent to atp6 (Yi et al. 2002). However, the relationships between orfH79 and fertility restorer genes Rf5 and Rf6 for HL-type CMS have not yet been elucidated. Interestingly, we found that an LD-CMS line also carries B-atp6–orf 79 homologous genes L-atp6–orf 79, but lacks N-atp6. Similar to the case of B-atp6–orf 79, Rf1 processes L-atp6–orf 79 mRNA into 1.5- and 0.45-kb fragments, and restores the

II.7 Molecular Studies on Cytoplasmic Male Sterility-associated Genes

(A)

+1

209

+1014

N-atp6

N-atp6

100% Identical region +1014 +1 +236

+1 B-atp6

B-atp6-orf79

orf79

(B)

Rf1Rf1

rf1rf1

Processing site

(kb)

atp6-orf79

2.0

atp6

1.5

orf79

0.45

Fig. 2 Structure of N-atp6, B-atp6 and orf 79 in BT-CMS (A) and northern blot analysis of B-atp6–orf 79 probed with B-atp6–/orf 79 region (B). In the BT-CMS line (rf1 rf1), B-atp6 is transcribed as a 2.0-kb mRNA consisting of B-atp6 and the orf 79 sequence, whereas two discontinuous mRNAs of 1.5 kb for B-atp6 and 0.45 kb for orf 79 are generated from the 2.0-kb mRNA by RNA processing in the presence of Rf1 (Rf1 Rf1)

fertility of an LD-CMS line. However, the original fertility restorer gene of the LD-CMS line, Rf2, does not promote the processing of L-atp6–orf 79 mRNA. Rf 2 acts as the partial fertility restorer for BT-CMS, but the processing of B-atp6–orf 79 mRNA does not occur in the presence of Rf2. We concluded that Rf1 and Rf 2 possess distinct fertility restorer functions. It is notable that similar orf 79-like genes are present in different rice CMS cytoplasms. Crossing over the species, an orf similar to orf 79 has also been found in A3-type CMS sorghum line IS1112C (Tang et al. 1996). It is named orf107. The C-terminal sequences are highly similar to the C terminus of ORF79 of BT-CMS rice; the C-terminal 49 residues are 57% identical and 80% similar to the C terminus of ORF79. The orf109 in sorghum A3-type CMS cytoplasm is located downstream

210

S. Fujii et al.

of the atp6 gene (Tang et al. 1999). The intergenic region between atp6 and orf109 is spliced in the presence of the fertility restorer gene Rf3 in sorghum A3-type CMS. Despite the progress in understanding the nature of BT-CMS rice, the factors involved in other rice CMS are largely unknown. Some studies using RFLP and RAPD analysis to reveal the mitochondrial CMS-associated genes have been conducted in WA-cytoplasm (Seth et al. 1996). Although genomic polymorphisms were detected among WA, BT and normal cytoplasm, the authors found it difficult to determine which polymorphism is truly responsible for CMS induction.

4

Molecular Cloning of Rf1 and Molecular Mapping of Other Rf Genes

Fine-resolution mapping and molecular cloning of Rf1 have been accomplished by several researchers (Kazama and Toriyama 2003; Akagi et al. 2004; Komori et al. 2004; Wang et al. 2006). Based on the sequence of Rf1, Rf1 protein was shown to be a pentatricopeptide repeat (PPR) protein. PPR proteins are characterized by the presence of tandem arrays of degenerated 35 amino acids, and therefore are termed as a pentatricopeptide repeat (Small and Peeters 2000). Rf1 protein consists of 791 amino acids and contains18 repeats of PPR motif and 26 amino acids of the N-terminal sequence for mitochondrial targeting, but it does not contain any other known motif (Fig. 3). In Arabidopsis thaliana, several

Fig. 3 Amino acids sequence of Rf1. A mitochondrial targeting signal is underlined. Consensus sequence of Rf1 and that of 1303 PPR proteins (Small and Peeters 2000) are also shown below the Rf1 sequence

II.7 Molecular Studies on Cytoplasmic Male Sterility-associated Genes

211

genes encoding PPR proteins are shown to be related in the regulation of organelle gene expression. For example, PPR protein HCF152 of Arabidopsis thaliana is required in the processing of plastid psbB-psbT-psbH-petB-petD operon (Meierhoff et al. 2003). CRR4 is involved in the editing of a specific base in Arabidopsis thaliana plastid ndhD gene (Kotera et al. 2005). PPR genes comprise a large family, consisting of about 450 members in Arabidopsis thaliana and 480 members in rice, and they are considered to be involved in the modification of organelle mRNAs (Small and Peeters 2000; Lurin et al. 2004). We followed a positional cloning strategy to identify the Rf1 gene. After mapping of the Rf1 locus, we searched for ORFs containing a mitochondrial targeting signal and PPR motif, because restorer genes had been previously cloned in petunia (Bentolila et al. 2002) and in Kosena radish (Koizuka et al. 2003) and had been shown to encode the PPR protein. The identification of Rf genes in petunia and radish as PPR-containing genes suggested that searching for PPR motif genes near known restorer loci would be a useful strategy to identify Rf genes in other species, including rice. Kazama and Toriyama (2003) found three PPR genes with a mitochondrial targeting sequence, PPR8-1, PPR8-2 and PPR8-3, in a fertility restorer line Milyang 23. They were quite similar to each other and existed in a tandem array. The nucleotide sequence of PPR8-1 shows 93.2% identity to PPR8-2 and 93.7% identity to PPR8-3. Each PPR gene was introduced into the BT-CMS line, and the PPR8-1 was shown to have the ability to process the intergenic region of B-atp6-orf 79 mRNA in the transgenic calli and to restore fertility (Kazama and Toriyama 2003, 2005). Komori et al. (2004) identified four PPR genes, PPR794, PPR683, PPR762 and PPR791, present in the Rf1 region, in which PPR791 was revealed to be Rf1, in a fertility restorer line IR24. PPR762 is absent in other restorer lines such as Milyang 23 and MTC-10R (Akagi et al. 2004). Rf1 is identical to PPR8-1 (Kazama and Toriyama 2003), PPR791 (Komori et al. 2004) and Rf1-A (Akagi et al. 2004). The recessive allele rf1 in Nipponbare contains a one-nucleotide deletion, which causes the creation of a stop codon in the fifth PPR and a 574-nucleotide deletion in the middle of the Rf1 gene. Recently, Wang et al. (2006) carried out fine-resolution mapping of Rf1, and cloned two distinct fertility restorer genes, Rf1a and Rf1b, from a 105-kb region of the classical Rf1 locus of an elite fertility restorer line Minghui 63. Rf1a is identical to PPR8-1/PPR791/Rf1-A and encodes 18 PPR repeats. Rf1b contains 11 PPR repeats and is sufficient to restore fertility to BT-CMS. Consistent with the study of Kazama and Toriyama (2003), RF1a possessed the ability to process the intergenic region of dicistronic B-atp6–orf 79 mRNA. Interestingly, the transgenic BT-CMS line carrying Rf1b lost B-atp6 and orf 79 mRNAs, meaning that RF1B might promote degradation of whole B-atp6–orf 79 mRNA rather than just processing the intergenic sequences. Wang et al. (2006) also proposed that Rf1a is epistatic to Rf1b, based on the result that the B-atp6 transcription pattern of a line containing both Rf1a and Rf1b is similar to that of a line carrying only Rf1a. They concluded that once B-atp6–orf 79 mRNA is processed by RF1A, the processed mRNA is resistant to specific degradation by RF1B.

212

S. Fujii et al.

Molecular mapping toward cloning of other fertility restorer genes of rice is now in progress by several research groups (Fig. 4). Fertility restoration in WA-CMS is conferred by one of two loci: Rf 3 on chromosome 1 and Rf4 on chromosome 10 (Yao et al. 1997; Zhang et al. 1997; Tan et al. 1998). Rf4 is shown to be tightly linked to a DNA marker close to the Rf1 locus. From a recent study, Liu et al. (2004) mapped two HL-CMS fertility restorers, Rf5 and Rf6(t), both to chromosome 10. Rf5 was identified in the region approximately 10 cM from the Rf1 locus, and Rf6(t) was tightly close to the Rf1 locus. Interestingly, a study on Dian-type 1 CMS (Tan et al. 2004) identified that the fertility restorer gene Rf-D1(t) is located in the region similar to Rf1, Rf4, Rf5 and Rf6(t). This may not be considered as a random coincidence, but possibly originates from the co-evolution of a CMS cytoplasm and a fertility restorer locus. The presence of a PPR gene cluster found in the Rf1 region indicates a trace of frequent recombination and duplication within this region. Apart from the fertility restorer genes in chromosome 10, Rf 2 for LD-CMS is mapped on chromosome 2 and Rf17 for CW-CMS is on chromosome 4 (Fujii and Toriyama 2005), as shown in Fig. 4. By understanding the functions of these fertility restorer proteins, one might find an independent CMS restoration mechanism from the Rf1-mediated case.

cM

Chr. 1

Chr. 2

Chr.4

Chr.10

S11148

0

Rf1a

Rf17

60

RM5373

Rf4

Rf2

40

Rf6(t)

20

100

S10019

120

RM1108 Rf3

180 190

Cloned fertility restorers Mapped fertility restorer loci Centromere

Rf-D1

140 160

Rf5

Rf1b

80

RM228

Fig. 4 Map position of cloned Rf genes and mapped Rf loci. Rf1a and Rf1b are Rf genes for BT-CMS, Rf2 for LD-CMS, Rf3 and Rf4 for WA-CMS, Rf5 and Rf6(t) for HL-CMS, Rf17 for CW-CMS and Rf-D1 for D1-CMS

II.7 Molecular Studies on Cytoplasmic Male Sterility-associated Genes

5

213

Approaches Toward Understanding the Mechanism of CMS Induction

Although extensive studies have been reported for the BT-CMS/Rf1 system, the question “How is B-atp6–orf 79 involved in CMS?” has been asked for more than a decade and still remains to be clarified. Some of the answers to the question have been recently provided by Wang et al. (2006). They demonstrated that expression of orf 79 in transgenic rice plants with a normal cytoplasm caused gametophytic male sterility. Using IPTG-inducible promoter, they also found that orf 79 causes E. coli lethality, and the C terminus of ORF79 is important for the toxicity. They raised the antibody against recombinant ORF79, and showed that ORF79 protein accumulates specifically in the microspores of a BT-CMS line, and its production is suppressed by RF1A and/or RF1B by distinct mechanisms: endonucleolytic cleavage for RF1A and degradation of the dicistronic B-atp6– orf 79 mRNA for RF1B. Since neither RF1A nor RF1B contain a motif for any enzymes, RF1A and RF1B are considered to recognize and bind a certain sequence of B-atp6–orf 79 mRNA and to function with other proteins involved in RNA cleavage and degradation. They also demonstrated that RF1A functions to promote the editing of atp6 mRNA, while RF1B has no such effect. It is unlikely that the unedited atp6 is translated into altered polypeptide and causes toxicity in microspore development by inhibiting the assembly of functional ATPase proteins. Wang et al. (2006) concluded that orf 79 encodes a cytotoxic peptide and is a cause of BT-CMS. However, it is still unknown how ORF79 induces cytotoxicity in rice pollen. In order to reveal the nuclear genes that are involved in pollen abortion, a comprehensive analysis of the transcript profile in mature anthers between normal and CW-CMS cytoplasm has been carried out using 22 k rice oligoarray (Fujii et al. 2007). We detected 58 genes that were up-regulated more than three-fold in the CW-CMS line, as well as 82 genes that were down-regulated. Out of 20 genes further investigated in other organs, five genes, including genes for alternative oxidase, were found to be preferentially expressed in a CW-CMS line but not in a fertility-restored CW-CMS line with Rf17. Such CW-CMS-specific gene expression was only observed in mature anthers, not in leaves, stems or roots, indicating the presence of anther-specific mitochondrial retrograde regulation of nuclear gene expression. Our study demonstrated that many nuclear genes are regulated by mitochondria, and proper gene expression requires correct nucleus–mitochondrial interaction. We can conclude that the mitochondrion plays a special role in normal pollen development. The next step toward understanding CMS is to clarify if these incorrectly regulated genes in a CW-CMS line are related to pollen dysfunction. This kind of study on the nuclear gene expression should be carried out for the other types of CMS as well.

214

6

S. Fujii et al.

Conclusion

CMS is utilized for hybrid rice breeding in southeastern Asia. Apart from its commercial benefit, research on CMS has been conducted on the topic of nuclear– mitochondrial cross-talk. Since rice is one of the most studied crops and abundant genomic information is available, we propose that rice is a model plant for CMS study. Molecular cloning of a couple of restorer genes is now in progress. Comparison of Rf genes among several types of CMS in rice will facilitate the understanding of the whole mechanism of the CMS/Rf system not only in rice but also in other plant species.

References Akagi H, Sakamoto M, Shinjyo C, Shimada H, Fujimura T (1994) A unique sequence located downstream from the rice mitochondrial atp6 may cause male sterility. Curr Genet 25:52–58 Akagi H, Nakamura A, Yokozeki-Misono Y, et al. (2004) Positional cloning of the rice Rf-1 gene, a restorer of BT-type cytoplasmic male sterility that encodes a mitochondria-targeting PPR protein. Theor Appl Genet 108:1449–1457 Bentolila S, Alfonso AA, Hanson MR (2002) A pentatricopeptide repeat-containing gene restores fertility to cytoplasmic male-sterile plants. Proc Natl Acad Sci USA 99:10887–10892 Fujii S, Toriyama K (2005) Molecular mapping of the fertility restorer gene for ms-CW-type cytoplasmic male sterility of rice. Theor Appl Genet 111:696–701 Fujii S, Komatsu S, Toriyama K (2007) Retrograde regulation of nuclear gene expression in CWCMS of rice. Plant Mol Biol 63:405–417 Hanson MR, Bentolila S (2004) Interactions of mitochondrial and nuclear genes that affect male gametophyte development. Plant Cell 16 (Suppl):S154–S169 Iwabuchi M, Kyozuka J, Shimamoto K (1993) Processing followed by complete editing of an altered mitochondrial atp6 RNA restores fertility of cytoplasmic male sterile rice. EMBO J 12:1437–1446 Katsuo K, Mizushima U (1958) Studies on the cytoplasmic difference among rice varieties, Oryza sativa L. 1. On the fertility of hybrids obtained reciprocally between cultivated and wild varieties. Jpn J Breed 8:1–5 Kazama T, Toriyama K (2003) A pentatricopeptide repeat-containing gene that promotes the processing of aberrant atp6 RNA of cytoplasmic male-sterile rice. FEBS Lett 544:99–102 Kazama T, Toriyama K (2005) Genetic evolution of Rf1 locus for the fertility restorer gene of BTtypr CMS rice. In: Toriyama K, Heong KL, Hardy B (eds) Rice is life: scientific perspectives for the 21st century. Proc World Rice Research Conf, Tokyo and Tsukuba, 4–7 November 2004. International Rice Research Institute, Los Banos, and Japan International Research Center for Agricultural Sciences, Tsukuba Kinoshita T (1997) Gene symbols and information on male sterility. Rice Genet Newsl 14:13–22 Koizuka N, Imai R, Fujimoto H, et al. (2003) Genetic characterization of a pentatricopeptide repeat protein gene, orf687, that restores fertility in the cytoplasmic male-sterile Kosena radish. Plant J 34:407–415 Komori T, Ohta S, Murai N, et al. (2004) Map-based cloning of a fertility restorer gene, Rf-1, in rice (Oryza sativa L.). Plant J 37:315–325 Kotera E, Tasaka M, Shikanai T (2005) A pentatricopeptide repeat protein is essential for RNA editing in chloroplasts. Nature 433:326–330

II.7 Molecular Studies on Cytoplasmic Male Sterility-associated Genes

215

Liu XQ, Xu X, Tan YP, et al. (2004) Inheritance and molecular mapping of two fertility-restoring loci for Honglian gametophytic cytoplasmic male sterility in rice (Oryza sativa L.). Mol Genet Genomics 271:586–594 Lurin C, Andres C, Aubourg S, et al. (2004) Genome-wide analysis of Arabidopsis pentatricopeptide repeat proteins reveals their essential role in organelle biogenesis. Plant Cell 16:2089–2103 Meierhoff K, Felder S, Nakamura T, Bechtold N, Schuster G (2003) HCF152, an Arabidopsis RNA binding pentatricopeptide repeat protein involved in the processing of chloroplast psbBpsbT-psbH-petB-petD RNAs. Plant Cell 15:1480–1495 Seth P, Sane AP, Nath P, Sane PV (1996) Molecular characterization of mitochondrial genomes of rice lines containing wild abortive (WA) male sterile and fertile cytoplasms. J Plant Biochem Biotech 5:75–82 Shinjyo C (1969) Cytoplasmic-genetic male sterility in cultivated rice, Oryza sativa L. Jpn J Genet 44:149–156 Shinjyo C (1975) Genetical studies of cytoplasmic male sterility and fertility restoration in rice, Oryza sativa L. Sci Bull Coll Agr Univ Ryukyus 22:1–57 Shinjyo C, Sato S (1994) Chromosomal location of a fertility restoring gene Rf-2. Rice Genet Newsl 11:93–95 Small ID, Peeters N (2000) The PPR motif – a TPR-related motif prevalent in plant organellar proteins. Trends Biochem Sci 25:46–47 Tan XL, Vanavichit A, Amornsilpa S, Trangoonrung S (1998) Genetic analysis of rice CMS-WA fertility restoration based on QTL mapping. Theor Appl Genet 97:994–999 Tan XL, Tan YL, Zhao YH, et al. (2004) Identification of the Rf gene conferring fertility restoration of the CMS Dian-type 1 in rice by using simple sequence repeat markers and advanced inbred lines of restorer and maintainer. Plant Breed 123:338–341 Tang HV, Pring DR, Shaw LC, et al. (1996) Transcript processing internal to a mitochondrial open reading frame is correlated with fertility restoration in male-sterile sorghum. Plant J 10:123–133 Tang HV, Chen W, Pring DR (1999) Mitochondrial orf107 transcription, editing, and nucleolytic cleavage conferred by the gene Rf3 are expressed in sorghum pollen. Sex Plant Reprod 12:53–59 Wang Z, Zou Y, Li X, et al. (2006) Cytoplasmic male sterility of rice with boro II cytoplasm is caused by a cytotoxic peptide and is restored by two related PPR motif genes via distinct modes of mRNA silencing. Plant Cell 18:676–687 Yao F, Xu C, Yu S, et al. (1997) Mapping and genetic analysis of two fertility restorer loci in wildabortive cytoplasmic male sterility system of rice. Euphytica 98:183–187 Yi P, Wang L, Sun Q, Zhu Y (2002) Discovery of mitochondrial chimeric-gene associated with cytoplasmic male sterility of HL-rice. Chinese Sci Bull 47:744–747 Zhang G, Bharaji T, Lu Y, Virmani S, Huang N (1997) Mapping of the Rf-3 nuclear fertilityrestoring gene for WA cytoplasmic male sterility in rice using RAPD and RFLP markers. Theor Appl Genet 94:27–33

Section III

Evolution and Ecology

This page intentionally blank

III.1

Phylogeny and Biogeography of the Genus Oryza Duncan A. Vaughan1(* ü ), Song Ge2, Akito Kaga1, and Norihiko Tomooka1

1

Introduction

Plants with characteristics of species belonging to the tribe Oryzeae were present in India more than 60 million years ago (Ma), early in the history of grasses (Prasad et al. 2005). This tribe is now represented by 11 genera that are found in tropical and temperate regions of the world. Among genera in the tribe Oryzeae, the genus Oryza, with about 23 species, has been remarkably successful in evolutionary terms. Species of Oryza with the A and C genomes have a pan-tropical distribution. Rice has been domesticated from wild A genome wild Oryza several times and is the world’s most important staple food. Among cereals, rice has a small genome and is considered as a model for genome studies. The genus Oryza is of particular interest not only because it is the genus of rice but also because of what the genus can tell us about other grasses. The genus Oryza is subject to increased research attention now that the rice genome has been fully sequenced. There are now BAC libraries for species representing most of the genomes of the genus Oryza (Ammiraju et al. 2006). The objective of this chapter is to consider the genus Oryza from phylogenetic and biogeographic perspectives, with emphasis on recent research.

1 National Institute of Agrobiological Sciences, 2-1-2 Kannondai, Tsukuba, Ibaraki 305-8602, Japan e-mail: [email protected]; [email protected]; [email protected] 2

Key State Laboratory of Systematic and Evolutionary Botany, Institute of Botany, Chinese Academy of Sciences, Beijing 100093, China e-mail: [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

219

220

D.A. Vaughan et al.

Table 1 Dates associated with the rise of angiosperms and Oryza Event Date Origin of the angiosperms Origin of monocots Separation of Africa from other Gondwanan areas Origin of Poales Origin of Poaceae Evidence of Oyzeae First grass pollen Divergence of BEP/PACC clades Northern hemisphere BEP/PACC clade evidence Divergence of Oryzeae and Pooideae Breakup of the Antarctic connection between South America and Australia Divergence of subtribes Oryzinae and Zizaniinae Divergence of Leersia and Oryza Divergence of O. meridionalis and O. longistaminata from other AA genome Oryza Divergence of O. glaberrima and O. sativa Divergence of indica and japonica subspecies

2 2.1

Basis

145–208 Ma ≈134 Ma ≈105Ma >100Ma ≈85 Ma ≈65 Ma 70? 60–55 Ma ≈50 Ma? ≈34 Ma

Molecular clock Phytolith Fossil Phytolith

≈35 Ma ≈35 Ma

Molecular clock

≈20 Ma

Molecular clock

≈14 Ma ≈2 Ma

Molecular clock Molecular clock

≈0.7 or 0.64 ≈0.44, 0.4 or > 0.2 Ma

Molecular clock Molecular clock

Phylogeny Before Grasses

The origin of the angiosperms is believed to have been a complex process that involved various phases of rapid diversification, reflecting the interactive effects of biological traits and the environment, along different lineages (Davies et al. 2004). Fossil evidence is lacking for the time when angiosperms are thought to have diverged from other plants, but all more recent estimates based mainly on sequence data point to a time for this divergence in the Jurassic (145–208 Ma) (Table 1; Sanderson et al. 2004). Analyses of several multigene phylogenies have enabled the relationship in the angiosperms of monocots to other plant groups to be elucidated (Soltis and Soltis 2004). Within angiosperms, the monocots represent an ancient lineage and monocots are believed to have emerged before 100 Ma (Chase 2004). The herbaceous, aquatic family Ceratophyllaceae is one of the closest families to the monocots among dicots (Barkman et al. 2000; Kuzoff and Gasser 2000).

2.2

Poaceae Phylogeny

International taxonomic collaboration is being conducted in order to understand the relationships among the Poaceae using morphological, anatomical, cytological and

III.1 Phylogeny and Biogeography of the Genus Oryza

221

Fig. 1 Phylogeny of the Poaceae based on molecular and morphological characteristics, showing the main points of divergence. Species numbers are from Clayton and Renvoize (1986). Thick lines indicate C4 photosynthesis lineages. (From Vaughan et al. 2005)

genetic data (GPWG 2000, 2001). This has enabled us to determine the fundamental relationships among the major groups of grasses based on analysis of these traits (Fig. 1). Oryza in the current grass phylogeny is in a group paraphyletic with the ‘core’ bamboos (Bambusoideae) and is placed in the subfamily Ehrhartoideae (Kellogg 1998; Zhang and Clark 2000; GPWG 2001). The Bambusoideae sensu lacto (including Oryza) is a polyphyletic assemblage (Fig. 1; Zhang and Clark 2000). Among the major cereals, only rice is found in the ancient lineage of Bambusoideae (Gaut 2002). The tribe Oryzeae is the largest tribe in the subfamily Ehrhartoideae and includes 11 genera, all but Leersia, Luziola and Oryza consisting of five or fewer extant species (Table 2). Molecular analysis has reinforced earlier taxonomic studies that show that the tribe Oryzeae consists of two groups of genera that have been named subtribes Oryzinae and Zizaniinae (Ge et al. 2002; Guo and Ge 2005).

222

D.A. Vaughan et al.

Table 2 Genera of the tribe Oryzeae and their characteristics (updated from Vaughan 2003) Subtribe Species Chromosome number Spikelet structure Genus (no.) Distribution (2n)a Oryzinae 23 Pantropical (T) 24, 48 Bisexual Oryzab Leersia 17 Worldwide (t + T) 24, 28, 48, 72, 96 Bisexual Zizaniinae Chikusichloa 3 China, Japan (t) 24 Bisexual Hygroryza 1 Asia (t + T) 24 Bisexual Zizania 4 Europe, Asia and North 20, 30, 34 Unisexual America (t + T) Luziola 11 North and South 24 Unisexual America (t + T) Zizaniopsis 5 North and South 24 Unisexual America (t + T) Rhynchoryza 1 South America (T) 24 Bisexual Maltebrunia 5 Tropical and southern Unknown Bisexual Africa (T) Prosphytochloa 1 Southern Africa (t) 24 Bisexual Potomophila 1 Australia (t + T) 24 Unisexual and bisexual a From the Missouri Botanical Garden’s Index to Plant Chromosome Numbers database (http://mobot.mobot.org/W3T/Search/ipcn.html) b Recently the genus Porteresia has been merged with Oryza (Lu and Ge 2005)

2.3

Evolution of Grasses

Coprolites believed to be from dinosaur titanosaur sauropods have been found in India that contain diverse grass phytoliths (Prasad et al. 2005). Among the diversity of grass phytoliths found in these coprolites are bilobate and cross-shaped phytoliths characteristic of present-day genera of Oryzeae (Prasad et al. 2005). These coprolites are dated to the late Cretaceous, about 65 Ma. Other evidence of early grasses includes presumed fossil grass pollen dated to between 60 and 70 Ma and the earliest grass flower dated at approximately 55 Ma (Stromberg 2005). Molecular data suggest a date of about 83 Ma for the origin of the crown group of Poaceae (Bremer 2002). Thus early grasses, including those of the tribe Oryzeae, were present when some parts of Gondwana had not separated. The most primitive extant grasses are found in South America (Bremer 2002). Thus Gondwana vicariance may be a factor that explains the distribution of grasses in South America and Australia, since migration between these continents via Antarctica was possible until about 38 Ma. However, the break of Africa from the rest of Gondwana was much earlier, at about 105 Ma (McLoughlin 2001). The majority of grasses fall into two groups known by the acronyms of their constituent subfamilies: the BEP [Bambusoideae s. str., Ehrhartoideae (= Oryzoideae) and Pooideae] and PACC [Panicoideae, Arundinoideae s. str., Chloridoideae s. l., Centothecoideae] clades. Within the BEP clade are only grasses with C3 photosynthetic systems, including rice and wheat. The PACC clade includes grasses with C3 and C4 photosynthetic systems, including maize and millets (Fig. 1).

III.1 Phylogeny and Biogeography of the Genus Oryza

223

Phytolith evidence from North America indicates that both the BEP and PACC clades were present about 34 Ma (Stromberg 2005). Molecular clock data suggest a possible date of about 50 Ma for the divergence of these two clades (Bremer 2002). A much earlier date for the emergence of the PACC clade based on phytoliths (Prasad et al. 2005) is ambiguous at present (Piperno and Sues 2005). While the origin of grasses is now considered to be the late Cretaceous and diversification occurred early in their evolution, the rise in ecosystem dominance of grasses was slow. The spread of grass-dominated ecosystems was linked with global climate change, cooling and drying, in the late Tertiary (Chapman 1996; Stromberg 2005).

2.4

An Emerging Scenario for the Origin of Oryza

Based on molecular clock data, an estimate of 20–22 Ma has been given for the split between the subtribes of Oryzeae, Oryzinae and Zizaniinae, and 14.2 Ma for the separation of Oryza and Leersia (Guo and Ge 2005). The genus Oryza has been divided into three sections: Padia, Brachyantha and Oryza (Lu 1999). Of these, the basal section is thought to be section Padia which consists of the forest-dwelling Oryza, O. schlechteri, the O. ridleyi complex and the O. granulata complex (Sharma 2003). Oryza schlechteri is a diminutive tetraploid species that grows on unstable forested river banks. It exhibits prolific stolon development which enables this species to spread horizontally and climb over other vegetation. It has characteristic nodal hairs that seem to be an adaptation to absorbing moisture from the air. The species of the O. ridleyi complex also have stolons and are adapted to swampy forest habitats. In contrast, species of the O. granulata complex are adapted to non-flooded forests, often in highland areas (Table 3). The habitats of species in Table 3 The genomes, usual habitats and life cycles of species in the four Oryza species complexes O. granulata O. ridleyi O. officinalis O. sativa Characteristic complex complex complex complex Genome(s)

GG

HHJJ

BB, CC, BBCC, CCDD, EE

AA

Usual habitat conditions: (a) Elevation Tropical uplands Tropical lowlands Tropical lowlands Tropical and lowlands lowlands (b) Associated Tropical deciduous Tropical rainTropical deciduous Lakes, rivers forests forests forests and and seasonal main vegetation grasslands pools types (c) Basic water Dryland Seasonal wetlands Seasonal wetlands Seasonal and requirements permanent wetlands (d) Light Full to partial Full to partial Full or partial shade Full sun shade shade or full sun requirements Life cycle Perennial Perennial Perennial Perennial to annual

224

D.A. Vaughan et al.

section Padia are similar to the type of habitats that early grasses are thought to have evolved in (Feild et al. 2003). Recently molecular clock data have suggested that the O. granulata complex diverged from other Oryza about 8 Ma (Guo and Ge 2005). Diploid species that contributed the H, J and K genomes to the tetraploid O. schlechteri and O. ridleyi complex species are unknown. If these diploid species exist they will most likely be found in New Guinea and would reveal much about Oryza evolution. Oryza brachyantha is the only species in section Brachyantha. This species is widely distributed across Africa, in a specific habitat of iron-pan rock pools. It is morphologically the most similar Oryza species to Leersia (Launert 1965). Section Oryza consists of two species complexes, the O. officinalis complex with the B, C, D and E genomes and O. sativa complex with the A genome. The genetically most diverged of the species of the O. officinalis complex is the little-studied species O. australiensis with the EE genome. Species with the CC genome are the most widely distributed and are adapted to a wide spectrum of environments, from semi-shade of forest clearings to open savannah (Vaughan 2003). Hybrids between different CC genome species from different areas show high sterility (Ogawa 2003). Among CC genome species, O. eichingeri appears to be basal (Shcherban et al. 2000, 2001). Like the O. officinalis complex, the O. sativa complex has attained a pan-tropical distribution. However, it contrasts with the O. officinalis complex in that all species of this complex are diploid and have a single genome, the A genome. There are several reports of similar dates for divergence of AA genome Oryza species. These suggest the first divergence was between ancestors of the Australian AA genome species, O. meridionalis, and other AA genome Oryza about 2 Ma (Zhu and Ge 2005). Divergence between African and Asian AA genome species appears to have occurred twice: the first time between ancestors of O. longistaminata and Asian AA genome Oryza and a second time between ancestors of the annual African Oryza, O. barthii and O. glaberrima, and Asian AA genome Oryza. Dates for the first divergence of Asian–African AA genome species of 2–3 Ma have been suggested (Vitte and Panaud 2003). This would correspond to a time similar to that given for the divergence of O. longistaminata, the perennial African AA genome species, and O. meridionalis (Zhu and Ge 2005). Based on nuclear DNA sequence data, Ma and Bennetzen (2004) estimated a date for the second divergence between Asian and African A genome species of about 0.64 Ma by looking at the divergence between ancestors of O. glaberrima and O. sativa. Zhu and Ge (2005), using intron sequence data from three nuclear genes, reported very similar dates for this divergence (approximately 0.7 Ma). These molecular clock dates and various molecular studies do not support the hypothesis that O. longistaminata is the immediate perennial ancestor of O. barthii and O. glaberrima (Khush 1997). The date for divergence of O. longistaminata from Asian AA genome Oryza (2–3 Ma) and the later date for the divergence of the African-cultivated gene pool (O. glaberrima and O. barthii) (0.6–0.7 Ma) from the AA Oryza gene pool suggest that they represented different lineages from the Asian AA genome gene pool. If this were correct, O. longistaminata would not be the direct perennial ancestor of O. glaberrima. The studies of the catalase gene (Iwamoto et al. 1999), RFLP (Wang et al. 1992), AFLP (Aggarwal et al. 1999) and

III.1 Phylogeny and Biogeography of the Genus Oryza

225

p-SINE1 (Cheng et al. 2003) could all be interpreted to support the hypothesis that O. longistaminata is not the direct perennial ancestor of O. glaberrima. Based on a study of twin microsatellites, Akagi et al. (1998) concluded that O. barthii has a closer affinity with O. rufipogon than with O. longistaminata. Several studies have reported molecular divergence dates for the two subspecies of cultivated Asian rice, indica and japonica (Ma and Bennetzen 2004; Vitte et al. 2004; Zhu and Ge 2005). These studies point to divergence of 0.2–0.44 Ma that is unequivocally before the domestication of rice. While most accumulated evidence tends to support dual or multiple domestication of rice (Fuller 2003b; Londo et al. 2006), some recent data do not support this (Lin et al. 2007). Thus the story of the evolution of cultivated rice in Asia is still unclear (Sweeney and McCouch 2007; Tao and Ge 2007).

3

Oryza Biogeography

Extant genera of Oryzeae are found almost equally in the Americas (5 genera), Africa (4) and Asia (5). Thus it is not possible to suggest where this tribe arose. Comparison of the three Oryzeae genera, Leersia, Luziola and Oryza, with more than 10 species shows that each has greatest species diversity in different regions of the world. Oryza is most diverse in Australasia (Vaughan et al. 2005), Leersia in Africa (Table 4), and Luziola is confined to the Americas. That these closely related genera have diversified in different regions is probably a reflection of the antiquity of the tribe Oryzeae.

Table 4 Leersia species (based on Launert 1965; Pyrah 1969). (Chromosome numbers from http://mobot.mobot.org/W3T/Search/ipcn.html) Species Chromosome number (2n) Distribution L. monandra L. ligularis L. stipitate L. japonica L. hexandra L. lenticularia L. oryzoides L. virginica L. triandra L. nematostachya L. friesii L. perrieri L. tisserantii L. drepanothrix L. oncothrix L. angustifolia L. denudata

48

96 48,72 48 28,48 48

24 24

Caribbean, Texas, Florida and Mexico Mexico to northern Argentina and Paraguay Thailand East Africa Wet tropics and subtropics worldwide Eastern USA USA, Europe to Japan Eastern USA and Brazil Sierra Leone and Liberia Angola and Zambia Angola, Congo, Tanzania, Zambia Madagascar Congo, Kenya and Zambia West Africa, Sudan and Uganda Zambia Congo and Sudan Kenya and Zimbabwe

226

3.1

D.A. Vaughan et al.

The Distribution of Oryza Species and Species Complexes

As discussed above, Oryza has been divided into three taxonomic sections. Species of section Padia are only found in Asia, Australia, New Caledonia and Papua New Guinea. The monospecific section Brachyantha is only found in Africa. Section Oryza has a pan-tropical distribution but is most diverse in Asia and Australia. From ecological, genetic and morphological perspectives all but three Oryza species, O. brachyantha, O. coarctata and O. schlechteri, fall into four clearly defined species complexes. These species complexes are O. sativa, O. officinalis, O. ridleyi and O. granulata (Table 3; Tateoka 1962a). Species complexes refer to a group of species for which there are a lack of good taxonomic key characters; hence whether the taxonomic rank of species or subspecies is used is rather arbitrary. This is best illustrated with the O. granulata complex, for which five species names have been validly published: O. abromeitiana, O. granulata, O. indandamanica, O. meyeriana and O. neocaledonica. However, the main distinguishing criteria are spikelet size and shape and this has been shown to vary continuously when many samples are studied (Tateoka 1962b). Generally the two widely distributed species, O. granulata and O. meyeriana, are recognised and the other three species, represented by one or a very few known populations, are considered local variants of these two species. The phylogenetic distribution of functional traits suggests that the ecological setting for early angiosperms was shaded, disturbed-forest understory and/or shady stream sides (Feild et al. 2003). This is precisely the habitat where O. schlechteri grows. The type locality of O. schlechteri is the Finisterre Mountains of Papua New Guinea. This young mountain range is characterised by frequent landslips (Hovius et al. 1998). Oryza schlechteri grows in shade along streams on unstable soil and rocks. A visit to one population of this species by the author found part of it buried beneath a natural landslide (D.A.V., pers. observ. 2005). Distribution, morphology, habit and habitat all suggest that this species is likely to be descended from an early lineage in the genus Oryza. Early monocots are thought to have been aquatic (Doyle 1998) and living in forested areas or forest margins (Kellogg 2001). Among Oryza species complexes, the tetraploid O. ridleyi complex that shares the HH genome with O. schlechteri is found in seasonally inundated forests beside lakes and rivers. This complex is most common and diverse in New Guinea. Oryza ridleyi complex species share the trait of stolon formation with O. schlechteri. If diploid species with the genome of the O. ridleyi complex and O. schlechteri are still extant, they would be expected to be found in the vast tracks of wet forested lowlands or lower elevations in the mountains of New Guinea. Early diploid aquatic or semi-aquatic Oryza of wet forests may have adapted to dryer deciduous woodland habitats as climates changed in some regions, giving rise to the O. granulata complex species. The O. ridleyi complex is distributed across Southeast Asia and New Guinea (Fig. 2). The O. granulata complex is also most diverse across insular Southeast

III.1 Phylogeny and Biogeography of the Genus Oryza

227

Fig. 2 Distribution of the O. ridleyi complex. (From Vaughan et al. 2005)

Asia, particularly Java, Indonesia, from which it likely spread to continental Asia (Fig. 3). The O. ridleyi and O. granulata complexes belong to the ‘primitive’ or ‘ancestral’ section of the genus, section Padia (Sharma 2003). They are distributed where many of the closest relatives of grasses (Flagellariaceae, Joinvilleaceae and Ecdeiocoleaceae) also occur, the Australasian region (Watson and Dallwitz 1992). These data suggest that this is the region where Oryza arose. New Guinea has greater Oryza genome diversity than any other geographic region (Vaughan 1991). On the basis of current Oryza diversity and the distribution of grass relatives, the Australasian zone appears to be the most likely region where Oryza first evolved and from where they spread to other tropical regions (Fig. 4). The antiquity of Oryza may be reflected by the fact that all four Oryza complexes (O. granulata, O. ridleyi, O. sativa and O. officinalis) are distributed on both sides of the biogeographical boundary called the Wallace Line. In the case of the O. granulata complex, one species reported east of the Wallace Line is Oryza neocaledonica Morat. This species grows on the island where the monotypic genus Amborella is found. Amborella trichopoda is a sister outgroup for angiosperm phylogenetic studies and testifies to New Caledonia as a special place in the evolutionary history of plants (Friedman 2006). Section Oryza with the O. officinalis and O. sativa complexes appears to be the most recent lineage within Oryza (Sharma 2003). With regard to the O. officinalis complex, current centres of diversity are in East Africa and southern South Asia. Studies of AA genome species from a wide geographic area suggest that the Australasian species O. meridionalis is genetically most diverged (Doi et al. 2000; Zhu and Ge 2005). Due to the domestication of rice, AA genome wild species appear most diverse in mainland Asia and West Africa. Haplotype studies of O. rufipogon have suggested that this species is most ancient in India and Indochina (Londo et al. 2006). However, in New Guinea, O. rufipogon and O. meridionalis

228

D.A. Vaughan et al.

Fig. 3 Distribution of the O. granulata complex. (From Vaughan et al. 2005)

Fig. 4 The geographic center of diversity of the genus Oryza and possible routes of spread from this core center for the O. sativa (s), O. officinalis (o) and O. granulata (g) complexes. (From Vaughan et al. 2005)

lineages are found, resulting in higher genetic diversity for the A genome there than in other regions (Doi et al. 2000; D.A.V., unpublished data). While the origin of the genus Oryza may have been in the Australasian region, the O. sativa and O. officinalis complex ancestors spread and diversified to other

III.1 Phylogeny and Biogeography of the Genus Oryza

229

regions. All Africa Oryza species can be found in East Africa. Climate change and isolation after distribution have resulted in distinct ecogeographic races within the O. sativa and O. officinalis complexes (Second 1991). The distribution of Oryza in the New World probably occurred later than in other areas, with introductions of the O. sativa complex evolving into regionally distinct Latin American forms, now usually called O. glumaepatula (Akimoto et al. 1997). Most studies suggest a greater affinity of O. glumaepatula to African members of the O. sativa complex than to Asian members (Vaughan and Morishima 2003).

3.2

Animals, Humans and Birds and the Biogeography of Oryza Species

The hypothesis that Oryza arose on Gondwanaland and that continental drift can explain the distribution of Oryza species (Khush 1997) is no longer in accordance with known facts concerning the evolution of grasses and Oryza. Therefore Oryza biogeography must be explained by long-distance dispersal. The distribution of Oryza species can best be explained in relation to the movement of animals, including humans and birds. Biotic dispersal is common in angiosperms, with more than 50% of species being dispersed by biotic or biotic and abiotic means (Tiffney 2004). Many animal groups are distributed in both Asia and Africa and some of them, such as elephants and buffalo, are intimately associated with Oryza species. In Africa, forest elephants are believed to have adapted to savannah habitats about 2.6 Ma (Roca et al. 2005). African elephants (Loxodontia spp.) and Asian elephants (Elephas maximus) are believed to have diverged about 5–7 Ma or earlier (Debruyne 2001). Elephants eat wild rice and their role in seed dispersal has been documented (Ridley 1930). Phylogenetic studies suggest the African buffalo (genus Syncerus) and Asian water buffalo (genus Bubalus) diverged more recently than elephants, at less than 4 Ma (Ritz et al. 2000; Buntjer et al. 2002). Buffalo and other bovine species also eat wild rice, so their role in past seed dispersal between Asia and Africa may help explain the rather recent dates (less than 1 Ma) for divergence of annual forms of African and Asian wild AA Oryza species (Barbier et al. 1991; Ma and Bennetzen 2004; Zhu and Ge 2005). The phytogeographic link between Africa and Asia today looks tenuous because of the Sahara Desert and deserts of the Middle East which form a distinct barrier to plant distribution. However, these deserts were green as recently as 6000 years ago and at that time the area where the Sahara Desert is now situated had a hospitable environment for hippopotamuses, which need year-round water (Kutzbach et al. 1996; Kuper and Kröpelin 2006). In addition the archaeological record clearly shows that African plant domesticates appeared very early in South Asia (Fuller 2003a). Over the time scale that has been suggested for the divergence between Asian AA genome Oryza and progenitors of African rice (O. glaberrima) of

230

D.A. Vaughan et al.

approximately 0.64 Ma, a hospitable environment for animals and prehistoric humans to pass between Africa and Asia existed. Isozyme analysis of wild and primitive cultivated varieties of rice from western India has suggested a link between Asian and African rice (Lolo and Second 1988). The southern movement of the African monsoon caused the Sahara region to dry up about 5000 years ago. The distribution of O. barthii, the presumed annual ancestor of O. glaberrima, is widely distributed along the entire southern fringe of the Sahara Desert from Ethiopia to Senegal (Vaughan 1994). Elephants, buffaloes and other megafauna are possible vectors involved in Oryza distribution. Prior to the Pleistocene extinction (50,000–10,000 years ago) there were many more large animals than there are today that could have moved plant seeds between continents (Barnosky et al. 2004). Wild Oryza species are important forage for cattle today in some countries (Vaughan and Sitch 1991). In addition, the role of humans in prehistoric times as a vector of Oryza species is also possible. Oryza species are likely to have been gathered as food by humans long before agriculture in a manner similar to that observed today in various parts of the world (Vaughan and Sitch 1991). Another agent that may explain the disjunctive distribution of some Oryza species is migrating birds. For example, O. eichingeri is present in East Africa and Sri Lanka. There are birds, including many water-loving species, that migrate from Africa across the Indian Ocean to South Asia and back (Ackerman 2004). Since many aquatic birds are migratory and most Oryza species are found in aquatic habitats the role of birds in distributing rice seeds seems probable. Birds could have distributed Oryza species, particularly small-seeded forms, during their annual migration. In Sri Lanka a population of O. eichingeri grows at about 500 m on Ritigala Mountain as well as at the foot of the mountain (A.H.M. Jayasuriya, Plant Genetic Resources Center, Sri Lanka, pers. comm. 2004). It seems probable that birds migrating across the Indian Ocean from Africa landed on this rocky outcrop. The disjunct distribution of Oryza minuta in the Philippines and Papua New Guinea is also on a migratory route for birds from Australia to East Asia (Ackerman 2004). Migration of birds could also explain the introduction into Australia and the New World of species of the O. sativa and O. officinalis complexes prior to human migration. More difficult to explain based on biotic long-distance dispersal is the apparent diversification over the last 2 Ma of the Australasian species O. meridionalis and Africa species O. longistaminata that both have the A genome. Other genera have disjunction between Africa and Australia. The wild relative of mungbean, Vigna radiata var. sublobata, is found in Africa and Australia, as are some Sorghum and Gossypium species. In the tribe Oryzeae, Duisteramaat (1987) proposed that the African genera Prosphytochloa and Maltebrunia were part of the Australian genus Potomophila. While this is not generally accepted, a multiple gene phylogenetic study has revealed close affinity between Prosphytochloa and Potomophila (Guo and Ge 2005). These evidences point to the phytogeographic link between Africa and Australia. Perhaps the best explanation may be that these two basal AA genome Oryza species evolved from a now extinct widely distributed unknown proto AA genome Oryza species.

III.1 Phylogeny and Biogeography of the Genus Oryza

231

In historic times human migration has introduced not only the cultivated rices into new areas but also probably at the same time Asian wild rice as contaminants of rice. This would explain the variation found in wild AA genome germplasm in Australia and Latin America. A Cuban accession commonly used in research on ‘wild’ Oryza from Cuba (IRRI 100961, NIG 1169/70) repeatedly has shown genetic similarity to Asian rice rather than the Latin American AA genome species O. glumaepatula (e.g. Juliano et al. 1998). Thus this accession collected from a hybrid swarm population could be an historic introduction to Cuba from Asia. Humans may well have carried other wild Oryza on their voyages. The presence of O. longistaminata in Martinique, the Caribbean, can best be explained by an accidental human introduction (Vaughan 1994). The hypothesis presented here that the current biogeography of Oryza species can be explained by the spread across contiguous land masses by animals, such as elephant and buffalo, is in accordance with dates when animals roamed between continents and when ancestors of current Oryza species are believed to have evolved. The disjunct distribution of Oryza species on land masses separated by oceans may best be explained by their introduction by birds and, in the historic past, by humans.

References Aggarwal RK, Brar DS, Nandi S, Huang N, Khush GS (1999) Phylogenetic relationships among Oryza species revealed by AFLP markers. Theor Appl Genet 98:1320–1328 Akagi H, Yokozeki Y, Inagaki A, Fujimura T (1998) Origin and evolution of twin microsatellites in the genus Oryza. Heredity 81:187–197 Ackerman J (2004) Cranes. National Geographic, http://magma.nationalgeographic.com/ ngm/0404/feature2/index.html Akimoto M, Shimamoto Y, Morishima H (1997) Genetic differentiation in O. glumaepatula and its phylogenetic relationships with other AA genome species. Rice Genet Newsl 14:37–39 Ammiraju JSS, Luo M, Goicoechea JL, et al. (2006) The Oryza bacterial artificial chromosome library resource: construction and analysis of 12 deep-coverage large insert BAC libraries that represent the 10 genome types of the genus Oryza. Genome Res 16:140–147 Barbier P, Morishima H, Ishihama A (1991) Phylogenetic relationships of annual and perennial wild rice: probing by direct DNA sequencing. Theor Appl Genet 81:693–702 Barkman, TJ, Chenery G, McNeal JR, et al. (2000) Independent and combined analyses of sequences from all three genomic compartments converge on the root of flowering plant phylogeny. Proc Natl Acad Sci USA 97:13166–13171 Barnosky AD, Koch PL, Feranec RS, Wing SL (2004) Assessing the causes of late Pleistocene extinctions on the continents. Science 306:70–75 Bremer K (2002) Early Cretaceous lineages of monocot flowering plants. Proc Natl Acad Sci USA 97:4707–4711 Buntjer, JB, Otsen M, Nijman IJ, Kuiper MTR, Lenstra JA (2002) Phylogeny of bovine species based on AFLP finger printing. Heredity 88:46–51 Chapman GP (1996) The biology of grasses. Cambridge University Press, Cambridge Chase MW (2004) Monocot relationships: an overview. Am J Bot 91:1645–1655 Cheng C, Motohashi R, Tsuchimoto S, Fukata Y, Ohtsubo H, Ohtsubo E (2003) Polyphyletic origin of cultivated rice: based on the interspersion pattern of SINEs. Mol Biol Evol 20:67–75

232

D.A. Vaughan et al.

Clayton WD, Renvoize SA (1986) Genera Graminum, grasses of the world. Kew Bull 13 Davies JT, Barraclough TG, Chase MW, et al. (2004) Darwin’s abominable mystery: insights from a supertree of the angiosperms. Proc Natl Acad Sci USA 101:1904–1909 Debruyne R (2001) Molecular phylogeny of living elephants and discussion on infraspecific systematics of Loxodonta africana and Elephas maximus. Proc Int Congr, The World of Elephants, Rome, pp. 628–629 Doi K, Nonomura M, Yoshimura A, Iwata N, Vaughan DA (2000) RFLP relationships of A-genome species in the genus Oryza. J Fac Agr Kyushu Univ 45:83–98 Doyle JA (1998) Phylogeny of vascular plants. Annu Rev Ecol Syst 29:567–599 Duistermaat H (1987) A revision of Oryza (Graminae) in Malesia and Australia. Blumea 32:157–193 Feild, TS, Arens NC, Dawson TE (2003) The ancestral ecology of angiosperms: emerging perspectives from extant basal lineages. Int J Plant Sci 164 (Suppl 3):S129–S142 Friedman WE (2006) Embryological evidence for developmental lability during early angiosperm evolution. Nature 441:337–340 Fuller DQ (2003a) African crops in prehistoric South Asia: a critical review. In: Neumann K, Butler A, Kahlheber S (eds) Food, fuel and fields, progress in African archaeobotany. Africa Praehistorica 15, Heinrich-Barth-Institute, Cologne, pp. 239–271 Fuller DQ (2003b) An agricultural perspective on Dravidian historical linguistics: archaeological crop packages, livestock and Dravidian crop vocabulary. In: Bellwood P, Renfrew C (eds) Examining the farming/language dispersal hypothesis. McDonalds Institute Monographs, Cambridge, pp. 191–213 Gaut BS (2002) Evolutionary dynamics of grass genomes. New Phytologist 154:15–28 Ge S, Li A, Lu BR, Zhang SZ, Hong DY (2002) A phylogeny of the rice tribe Oryzeae (Poaceae) based on matK sequence data. Am J Bot 89:1967–1972 GPWG (Grass Phylogeny Working Group) (2000) A phylogeny of the grass family (Poaceae), as inferred from eight character sets. In: Jacobs SWL, Everett JE (eds) Grasses: systematics and evolution. Commonwealth Scientific and Industrial Research Organization, Collingwood, Victoria, pp. 3–7 GPWG (Grass Phylogeny Working Group) (2001) Phylogeny and subfamilial classification of the grasses (Poaceae). Ann MO Bot Gard 88:373–457 Guo YL, Ge S (2005) Molecular phylogeny of Oryzeae (Poaceae) based on DNA sequences from chloroplast, mitochondrial, and nuclear genomes. Am J Bot 92:1548–1558 Hovius N, Stark CP, Tutton MA, Abbott LD (1998) Landslide-driven drainage network evolution in a pre-steady-state mountain belt; Finisterre Mountains, Papua New Guinea. Geology 26:1071–1074 Iwamoto M, Nagashima H, Nagamine T, Higo H, Higo K (1999) A Tourist element in the 5′-flanking region of the catalase gene CatA reveals evolutionary relationships among Oryza species with various genome types. Mol Gen Genet 262:493–500 Juliano AB, Naredo MEB, Jackson MT (1998) Taxonomic status of Oryza glumaepatula Steud. I. Comparative morphological studies of New World diploids and Asian AA genome species. Genet Resour Crop Evol 45:197–203 Kellogg EA (1998) Relationships of cereal crops and other grasses. Proc Natl Acad Sci USA 95:2005–2010 Kellogg EA (2001) Evolutionary history of the grasses. Plant Physiol 125:1198–1205 Khush GS (1997) Origin, dispersal, cultivation and variation of rice. Plant Mol Biol 35:25–34 Kuper R, Kröpelin S (2006) Climate-controlled Holocene occupation in the Sahara: motor of Africa’s evolution. Science 313:803–807 Kutzbach J, Bonan G, Foley J, Harrison SP (1996) Vegetation and soil feedbacks on the response of the African monsoon to orbital forcing in the early to middle Holocene. Nature 384:623–626 Kuzoff RK, Gasser CS (2000) Recent progress in reconstructing angiosperm phylogeny. Trends Plant Sci 5:330–336 Launert E (1965) A survey of the genus Leersia in Africa. Senckenb Biol 46:29–153

III.1 Phylogeny and Biogeography of the Genus Oryza

233

Lin Z, Griffith ME, Li X, et al. (2007) Origin of seed shattering in rice (Oryza sativa L.). Planta 226:11–20 Lolo OM, Second G (1988) Peculiar genetic characteristics of O. rufipogon from western India. Rice Genet Newsl 5:67–70 Londo JP, Chiang YC, Hung KH, Chiang TY, Schaal B (2006) Phylogeography of Asian wild rice, Oryza rufipogon, reveals multiple independent domestications of cultivated rice, Oryza sativa. Proc Natl Acad Sci 103:9578–9583 Lu BR (1999) Taxonomy of the genus Oryza (Poaceae): historical perspectives and current status. IRRI Newsl 24:4–8 Lu BR, Ge S (2005) Oryza coarctata: the name that best reflects the relationships of Porteresia coarctata (Poaceae: Oryzeae). Nord J Bot 23:555–558 Ma J, Bennetzen JL (2004) Rapid recent growth and divergence of rice nuclear genomes. Proc Natl Acad Sci USA 101:12404–12410 McLoughlin S (2001) The breakup of Gondwana and its impact on pre-Cenozoic floristic provincialism. Aust J Bot 49:271–300 Ogawa T (2003) Genome research in genus Oryza. In: Nanda JS, Sharma SD (eds) Monograph on genus Oryza. Science Publishers, Enfield, New Hampshire, pp. 171–212 Piperno DR, Sues HD (2005) Dinosaurs dined on grass. Science 310:1126–1128 Prasad V, Stromberg CAE, Alimohammadian H, Sahni A (2005) Dinosaur coprolites and the early evolution of grasses and grazers. Science 310:1177–1180 Pyrah GL (1969) Taxonomic and distributional studies in Leersia (Graminae). Iowa State Univ J Sci 44:215–258 Ridley HN (1930) Dispersal of plants throughout the world. L. Reeve and Company, Kent Ritz LR, Glowatzki-Mullis ML, MacHugh DE, Gaillard C (2000) Phylogenetic analysis of the tribe Bovini using microsatellites. Anim Genet 31:178–185 Roca AL, Georgiadis N, O’Brien SJ (2005) Cytonuclear genomic dissociation in African elephant species. Nature Genet 37:96–100 Sanderson MJ, Thorne JL, Wikstrom N, Bremer K (2004) Molecular evidence on plant divergence times. Am J Bot 91:1656–1665 Second G (1991) Molecular markers in rice systematics and the evaluation of genetic resources. In: Bajaj YPS (ed) Biotechnology for Agriculture and Forestry, 14: Rice. Springer, Berlin Heidelberg New York, pp 468–494 Sharma SD (2003) Species of the genus Oryza and their interrelationships. In: Nanda JS, Sharma SD (eds) Monograph on the genus Oryza. Science Publishers, Enfield, New Hampshire, pp 73–111 Shcherban AB, Vaughan DA, Tomooka N (2000) Isolation of a new retrotransposon-like DNA and its use in analysis of diversity within the Oryza officinalis complex. Genetica 108:145–154 Shcherban AB, Vaughan, DA, Tomooka N, Kaga A (2001) Diversity in the integrase coding domain of a gypsy-like retrotransposon among wild relatives of rice in the Oryza officinalis complex. Genetica 110:43–53 Soltis PS, Soltis DE (2004) The origin and diversification of angiosperms. Am J Bot 91:1614–1626 Stromberg CAE (2005) Decoupling taxonomic radiation and ecological expansion of open-habitat grasses in the Cenozoic of North America. Proc Natl Acad Sci USA 102:11980–11984 Sweeney M, McCouch S (2007)The complex history of the rice domestication. Ann Bot (in press) Sang T, Ge S (2007) The puzzle of rice domestication. J Integrative Plant Biol 49:760–768 Tateoka T (1962a) Taxonomic studies of Oryza I. O. latifolia complex. Bot Mag Tokyo 75:418–427 Tateoka T (1962b) Taxonomic studies of Oryza II. Several species complexes. Bot Mag Tokyo 76:165–173 Tiffney BH (2004) Vertebrate dispersal of seed plants through time. Ann Rev Ecol Evol Syst 35:1–29 Vaughan DA (1991) Biogeography of the genus Oryza across the Malay Archipelago. Rice Genet Newsl 8:73–75

234

D.A. Vaughan et al.

Vaughan DA (1994) The wild relatives of rice; a genetic resources handbook. International Rice Research Institute, Los Banos Vaughan DA (2003) Genepools of the genus Oryza. In: Nanda JS, Sharma SD (eds) Monograph on genus Oryza. Science Publishers, Enfield, New Hampshire, pp. 113–138 Vaughan DA, Morishima H (2003) Biosystematics of the genus Oryza. In: Wayne Smith C, Dilday RH (eds) Rice: origin, history, technology, and production. Wiley, Hoboken, New Jersey, pp 27–65 Vaughan DA, Sitch LA (1991) Gene flow from the jungle to farmers. BioScience 41:22–28 Vaughan DA, Kadowaki K, Kaga A, Tomooka N (2005) On the phylogeny and biogeography of the genus Oryza. Breed Sci 55:113–122 Vitte C, Panaud O (2003) Formation of Solo-LTRs through unequal homologous recombination counterbalances amplifications of LTR retrotransposons in rice, Oryza sativa L. Mol Biol Evol 20:528–540 Vitte C, Ishii T, Lamy F, Brar D, Panaud O (2004) Genomic paleontology provides evidence for two distinct origins of Asian rice (Oryza sativa L.). Mol Gen Genomics 272:504–511 Wang ZY, Second G, Tanksley SD (1992) Polymorphism and phylogenetic relationships among species in the genus Oryza determined by analysis of nuclear RFLPs. Theor Appl Genet 83:565–581 Watson L, Dallwitz MJ (1992 onwards) The families of flowering plants: descriptions, illustrations, identification, and information retrieval. Version 2, June 2006, http://delta-intkey.com Zhang W, Clark LG (2000) Phylogeny and classification of the Bambusoideae (Poaceae). In: Jacobs SWL, Everett J (eds) Grasses: systematics and evolution. CSIRO, Melbourne, pp. 35–42 Zhu Q, Ge S (2005) Phylogenetic relationships among A-genome species of the genus Oryza revealed by intron sequences of four nuclear genes. New Phytol 167:249–265

III.2

Chromosome and Genome Evolution in Rice Nori Kurata1

1

Introduction

Genomic and species divergences in rice offer a major advantage in genetic and evolutional studies, as well as for practical breeding purposes. The genus Oryza is classified into nine genomes comprising 23 species (two cultivated and 21 wild). Classification of the genome has been made based on the ability of a known tester genome to form paired bivalent chromosomes during meiosis in an F1 hybrid with an accession of unknown genome. Varied levels of partial sterility have been observed in the crosses between a variety of accessions, both across species as well as within a species, suggesting continuous evolution of the genome. The pattern of genome and chromosomal evolution of the genus Oryza and of the family Poaceae has been resolved progressively over the last decades. The observed results, possible relationships and impact on genome and chromosome evolution are reviewed in this chapter.

2

Chromosomes in the Different Genomes

Twelve chromosome complements of Oryza sativa were identified in meiotic pachytene by Khush et al. (1984) and in mitotic pro-metaphase by Kurata and Omura (1978), with few discrepancies between the studies. The misclassifications in both nomenclature systems were resolved and the relationship between chromosome numbers and linkage groups was finally established in 1990 (reported by Khush). Chromosome analysis was also carried out with various wild species of BB, CC, EE, FF and GG genomes. Hu (1961) first reported the differences in chromosome size among different genomes. Analysis of pachytene and pro-metaphase chromosomes

1 National Institute of Genetics, Yata 1111, Mishima, Japan 411-8540 e-mail: [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

235

236

N. Kurata

of the wild species revealed very similar features in all 12 chromosomes among genomes with very little morphological variations (Kurata 1985), while sizes of chromosomes and genome varied among several different genome species (Kurata and Fukui 2003; Miyabayashi et al. 2007). The results indicate that EE, CC and GG genome species have larger and the FF genome has smaller chromosome complements, as shown in Fig. 1. Genome size of Oryza species (Table 1) corresponded very well to the enlargement of chromosomes in each species. However, in spite of more than a twofold difference in genome sizes, the chromosome complements of all species showed very similar morphology. This seemed to be caused by the proportional expansion or reduction of DNA content along the whole chromosome length. Evidence for the dispersion of some repetitive sequences was obtained by chromosome painting analysis using fluorescent in situ hybridization (FISH) of chromosomes from a unique genome. Experiments revealed that the total genomic DNA from O. officinalis of the CC genome could paint all 12 chromosomes of the CC genome evenly but not those of the BB genome in the cells of O. punctata of BBCC genome species (Fukui et al. 1997). The results indicated that CC genome-specific repetitive sequences were hybridized with chromosomes on dispersed regions of the CC genome and that the BB genome has less or lacks similar repetitive sequences. Clear molecular evidence of whole genome expansion of retrotransposon sequences in EE, GG and HHJJ genomes has also been presented, as shown in the Section 5

Fig. 1 Mitotic prometaphase chromosomes of Oryza species. A O. sativa of AA genome. B O. punctata of BB genome. C O. officinalis of CC genome. D O. australiensis of EE genome. E O. brachyantha of FF genome. F O. granulata of GG genome. Bars represent 10 µm. [Photos are combined from Kurata and Fukui (2003), Miyabayashi et al. (2007) and from original figure of N.K.]

III.2 Chromosome and Genome Evolution in Rice

237

Table 1 Genome size of Oryza species Our studya

Previous studies

Genome size DNA content DNA content Oryza species

Genome type Mb/1C c

pg/2C

Referenceb

pg/2C

d

O. sativa AA 390 0.91 0.91 1, 2, 3 O. rufipogon AA 377 0.88 0.91, 0.95 1, 2 O. barthii AA 403 0.94 O. longistaminata AA 390 0.91 0.81 2 O. glumaepatula AA 433 1.01 0.99 2 O. meridionalis AA 381 0.89 1.02 2 O. punctata diplo. BB 369 0.86 0.88, 1.11 1, 2 O. punctata tetra. BBCC 840 1.96 O. minuta BBCC 772 1.80 2.33 3 O. officinalis CC 549 1.28 1.35, 1.45, 1.14 1, 2, 3 O. eichingeri CC 596 1.39 1.47, 1.17 2, 3 O. rhizomatis CC 823 1.92 O. alta CCDD 866 2.02 2.09 1 O. latifolia CCDD 806 1.88 2.32 3 O. grandiglumis CCDD 891 2.08 1.99 3 O. australiensis EE 823 1.92 2.00, 1.96, 1.99 1, 2, 3 O. brachyantha FF 261 0.61 0.75, 0.72 1, 2 O. granulata GG 1020 2.38 1.83 1 O. meyeriana GG 999 2.33 O. ridleyi HHJJ 1076 2.51 2.66, 1.93 1, 3 O. longiglumis HHJJ 1209 2.82 a The latest results from Miyabayashi et al. (2007) b 1 Ammiraju et al. (2006); 2 Uozu et al. (1997); 3 Martinez et al. (1994) c Genome size of Nipponbare estimated at 390 Mb (Sasaki et al. 2005) was adapted d DNA content of O. sativa ssp. japonica cv. Nipponbare (0.91 pg/2C) was used as a standard

of this chapter. Regarding the small morphological variations on chromosomes detected by cytology, recent advances in FISH analysis on rice pachytene chromosomes (Cheng et al. 2001) might be able to show evidence of structural changes. Comparisons of the BAC contig physical maps between species currently identifies structural variations such as indels, duplications and inversions on many chromosome regions among species, as mentioned below briefly. Some small morphological variations of chromomere number detected on pachytene chromosomes of wild species (Kurata 1985) may correspond to such structural changes.

3

Classification of Genomes Based on Chromosome Pairing

The ‘genome’ is the minimum genetic set necessary for an organism to live with its own distinct characteristics and to propagate itself. The entity of the genome is regarded as being embodied as the haploid chromosome complement seen in the gamete. Many species in the genus Oryza have been classified into nine genome

238

N. Kurata

types. The method used for the classification of each genome type is called ‘genome analysis’, which consists of the examination of chromosome pairing in meiosis of the F1 hybrid between the tester parent of known genome type and the parent with unknown genome. Pairing status can vary from 12 bivalents to 24 univalents (in the case of a cross between diploids) or 12 bivalents plus 12 to 36 univalents (in the case of a cross between diploid and tetraploid plants). The frequency of bivalent formation indicates levels of genome similarity between the two species and can classify the unknown genome species as the same or different genome. Genome analyses have been carried out in many cross combinations by several cytogeneticists (reviewed by Katayama 1990; Aggarwal et al. 1997) and those results are summarized in Fig. 2 and Table 2. However, there have been a fair number of inconsistencies among those results in the bivalent numbers observed in the hybrids. It might be that the discrepancies in some of the results came from the differences in cross combinations where parents are derived from divergent accessions in the same species. Although there were several unclear pairing features in the genome analysis, 23 species and nine genome types have been identified to date in the genus Oryza (Table 2).

O.longistaminata AA O.glumaepatula AA

O.sativa AA 24 I

36 I (30~26I+3~5II) O.officinalis CC

O.minuta BBCC

O.eichingeri CC

O.punctata BB

O.punctata BBCC

O.meridionalis AA

12II

O.rufipogon AA

12II +12 I

O.barthii AA

O.glaberrima AA

12II +12I

O.latifolia CCDD

36 I 24 I

24 I

36 I

36 I

36 I

36 I

36 I

36 I

O.glandiglumis CCDD O.alta CCDD

O.granulata O.meyeriana GG GG 12 II 36 I O.australiensis EE

36 I 36 I

36 I O.brachyantha FF

24 I

O.ridleyi HHJJ

O.longiglumis HHJJ 24 II

Fig. 2 Summary of genome analysis in Oryza performed by observing bivalent (II) and univalent (I) chromosome number of hybrids between two species connected with solid lines. Numbers of bivalent and univalent chromosomes in F1 hybrids are indicated beside each line. Results of many studies were collected from the reviews of Katayama (1990) and Aggarwal et al. (1997). Broken lines indicate rare hybridization of genomic DNA of connected species with O. granulata total genomic DNA probes which suggested low homology between them

III.2 Chromosome and Genome Evolution in Rice

239

Table 2 Genomes and species belonging to the genus Oryza Species Chr. no. Genome Oryza sativa complex Oryza sativa L. O. rufipogon sensu lato O. glaberrima Steud. O. barthii A. Chev. O. longistaminata Chev. et Roehr. O. meridionalis Ng O. glumaepatula Steud. O. officinalis complex O. officinalis Wall ex Watt O. minuta J.S. Presl. ex C.B. Presl. O. rhizomatis Vaughan O. eichingeri Peter O. punctata Kotschy ex Steud. O. latifolia Desv. O. alta Swallen O. grandiglumis (Doell) Prod. O. australiensis Domin O. ridleyi complex O. ridleyi Hook. O. longiglumis Jansen O. granulata complex O. granulata Nees et Arn. ex Watt O. meyeriana (Zoll. et Mor. ex Steud.) Baill. Others O. brachyantha Chev. et Roehr. O. schlechteri Pilger

Geographical distribution

24 24 24 24 24 24 24

AA AA AA AA AA AA AA

All over the world Asia, Oceania West Africa Africa Africa Australia Central and South America

24 48 24 24 24,48 48 48 48 24

CC BBCC CC CC BB,BBCC CCDD CCDD CCDD EE

Asia Philippines Sri Lanka Africa, Sri Lanka Africa Central and South America Central and South America South America Australia

48 48

HHJJ HHJJ

Asia, New Guinea New Guinea

24 24

GG GG

Asia Asia

24 48

FF Unknown

Africa New Guinea

How have these 23 species been formed in the evolution of Oryza? In the following sections, events in the evolution of the rice genome are reviewed at three separate levels: first at large-scale chromosome rearrangement, second at genomewide expansion of retrotransposons, and third at the level of nucleotide variation among species.

4

Genome Duplication in the Genus Oryza

Two large-scale genome duplication events that greatly changed the structure of the rice genome have been revealed in recent genome studies. The last and regional duplication in the genome was evident at the end of chromosomes 11 and 12 (Wu et al. 1998). Along about 12 Mb length of the ends of these two chromosomes, highly homologous genes are aligned in the same order, suggesting relatively recent occurrence of the duplication. The presence of the same duplicated regions in chromosomes 11 and 12 in all species of AA genomes (Wing 2006b) indicates that the

240

N. Kurata

duplication occurred at some time before divergence of the AA genome. Meanwhile, the sequence duplicated at the ends of chromosomes 11 and 12 was not seen in other cereal genomes other than Oryza. Therefore the duplication could be estimated to have occurred at 21 million years ago (Mya), postdating grass divergence (Yu et al. 2005). The genome-wide similarity search using O. sativa indica and japonica genome sequences revealed that the Oryza genome contains 19 chromosome segments that show long-range similarity to other chromosome regions (Yu et al. 2005). The homology was high enough to detect nucleotide similarity between respective shared regions even though a lot of variations were formed between them. These duplicated regions occupied 65.7% of the whole genome. Guo et al. (2006) compared the duplicated segments in detail for three pairs of regions spanning over 30 Mb. The results indicated multiple patterns of segmental evolution accompanied by up to 50% reduction or 70% increase in size between the duplicated regions caused by various deletions and insertions. Comparison of nucleotide divergence between duplicated regions indicated the age of duplication of the ancestor genome to be post-date to that of monocot eudicot divergence at 170–235 Mya (Yu et al. 2005).

5

Genome-wide Evolution Among Oryza Species: Genome Size, Retrotransposon and Centromere

In the genus Oryza, the genome size of the 9 genomes and 23 species varies, as reported by several research groups (Table 1). The genome size of cultivar Nipponbare, a standard strain of the genome sequencing project of O. sativa, AA genome, was estimated to be 390 Mb (Sasaki et al. 2005). Eight AA and one BB genome species have an almost similar genome size of around 380 to −430 Mb. Three CC genome species of O. officinalis, O. eichingeri and O. rhizomatis have larger genomes than O. sativa by 41–111%. Six tetraploid species of O. punctata and O. minuta assigned to the BBCC genome, O. alta, and O. grandiglumis with the CCDD genome, and O. ridleyi and O. longiglumis with the HHJJ genome showed almost twofold the DNA content of O. sativa, AA genome. This suggests that the DNA content of DD, HH and JJ genomes could be estimated as almost the same as that of the AA genome, although diploid species of these three genomes have not been identified to date. Three species, O. australiensis with the EE genome and O. glanulata and O. meyeriana with the GG genome were revealed to contain about double the DNA in their genomes compared to the AA genome. On the contrary, the FF genome to which only one species of O. brachyantha belongs showed the smallest genome size in the genus Oryza, possessing about 70% of that of the AA genome. Although the genome sizes are different among species, the morphology of all 12 chromosomes of these species carrying larger or smaller genomes resembled very much the chromosome complement of O. sativa with the AA genome. From these observations, the increase and decrease in genome size might be attributed to

III.2 Chromosome and Genome Evolution in Rice

241

the evenly occurring expansion or reduction in DNA content along whole chromosomes. Piegu et al. (2006) reported that the O. australiensis of the EE genome contains three highly expanded retrotransposons occupying about 60% of the EE genome. Three retrotransposons, TY1/Copia type RIRE1and TY3/Gypsy types Kangourou and Wallabi, were estimated to have expanded their copies at independent periods of evolution after O. australiensis diversification. Three retrotransposons were detected in 29% of BAC clones of an O. australiensis genomic library, suggesting a moderately scattered distribution of them along the whole O. australiensis genome. Aggarwal et al. (1997) also indicated expansion of genome-specific repetitive sequences in O. granulata and O. meyeriana of the GG genome, and O. ridleyi and O. longiglumis of the HHJJ genome through genomic DNA hybridization to all genome species. Total genomic DNA from GG and HHJJ genomes almost exclusively hybridized to only the DNA of their own genomes. Although the cause of genome size reduction in the FF genome has not yet been clarified, it might have been related to the deletion of some repetitive sequences in the genome. The centromere is one of the key structures of the chromosome. This also has unique characteristics in relation to the evolution of chromosome structure and function. Centromeres of plant organisms consist of a huge number of tandem repeat sequences of about 130–180 bp long repeat units (Kurata et al. 2002). The genome-specific centromere repeats have been isolated as 155 bp CentO repeats from O. sativa of the AA genome (Cheng et al. 2002), as other genome-specific centromere repeats from O. officinalis and O. rhizomatis of the CC genome (Bao et al. 2006) and from O. brachyantha of the FF genome (Lee et al. 2005). The interesting characteristics of the core centromere repeats are as follows: (1) the sequence of the repeat unit is unique to individual genomes but varies in number among chromosomes, (2) core repeats in the genome are common in most of the 12 chromosomes, but some are mixed with other centromere repeat units, and (3) the tandem repeats form long-range blocks from scores of kilobases to several megabases as core centromere sequences and are surrounded by pericentromeric transposon/retrotransposon sequences of various kinds and numbers. These results indicate the rapid replacement of unique centromere repeats in multiple chromosomes and that the frequency of replacement of the centromere repeats may reflect the evolutional distance from one genome to another.

6

Regional Genome Evolution Among Oryza Species

Whole genome structure comparison between the AA and other genomes has been undertaken by aligning BAC contiguous clones of all genomes on the whole sequences of the 12 chromosomes of O. sativa, cultivar Nipponbare, within the Oryza Map Alignment Project (OMAP) by Wing et al. (2006a). The present view of comparative maps between genomes indicates a huge amount of small-scale

242

N. Kurata

insertions, deletions, duplications, inversions and translocations, as well as a small number of large-scale events which have occurred after Oryza species diversification (Wing et al. 2005; Hass-Jacobus et al. 2006). Within the AA genome, variations in the coding regions between two major subspecies of O. sativa, japonica and indica, are characterized by an excess amount of deleterious mutations, which is thought to have accumulated during domestication (Lu et al. 2006), and also a large amount of intron loss from the duplicated segments (Lin et al. 2006). In regard to specific chromosomal regions, repetitive units are largely related to their structural evolution. A large-scale retrotransposon invasion has been clarified as expanding genome size, as described above. Another key repeat component driving chromosome evolution is that of centromere tandem satellite repeats. Rapid changes in repeat number and the sequences of centromere repeat units became apparent among genomes from several studies, as discussed above. To reveal centromere evolution in detail, the structure of CEN8, which is the shortest centromere in rice harboring 68.5-Kb CentO repeats, was extensively analyzed and variations in both pericentromeric numerous retrotransposons and centromeric core repeats have been detected. The recent analysis of the 1.97-Mb sequences including CEN8 between japonica and indica rice and also within the japonica genome showed specific features of centromere evolution (Ma and Bennetzen 2006). The centromere regions showed about twofold lower nucleotide substitution rate than other euchromatic regions and higher sequence identity of CentO satellite repeats than other repeat units, suggesting recent rearrangements of centromere sequences. Thus, all results obtained to date have shown rapid evolution and rearrangement of centromere sequences, at least in CEN8. More detailed molecular characterization of the genome structures of all genomes and species is expected to clarify the evolutionary processes of the genus Oryza, including specific chromosome regions.

7

Genome Evolution in the Poaceae Family

The high syntenic relationship detected amongst the grass genomes of rice, maize, wheat, barley and sorghum was one of the epoch-making works in cereal genome research (Devos and Gale 1997). Complete sequencing of the rice genome and accumulation of other cereal genome sequences gave more opportunities to compare orthologous regions in detail across grass species. The comparison revealed multiple rearrangements by insertions, deletions and translocations that disrupt the colinearity of grass genomes (Ilic et al. 2003). However, the grass genomes in the Poaceae family still show high syntenic relationships to the Oryza genome, despite many rearrangements both on the duplicated regions within the genome and between genomes (Devos 2005). This and other studies together indicated polyploid and common ancestor origins of all the major cereal plants and that the genome-wide duplication predates the grass divergence (Paterson et al. 2004). In regard to the

III.2 Chromosome and Genome Evolution in Rice

243

Fig. 3 A hypothetical ancestor of the Oryza genome composed of either a set of five chromosomes (left: chromosomes 1, 2, 3, 8 and 11) or seven chromosomes (right: chromosomes 4, 5, 6, 7, 9, 10 and 12). The present 12 chromosomes can be divided into two chromosome sets with almost equal genome constitution, but are not derived from a mixed-up configuration of duplicated segments. Filled blocks represent chromosomes. Numbers in the blocks are chromosome number. Chromosome segments showing similarities with each other are connected by wide lines based on the duplicated regions within the genome (Yu et al. 2005)

common origin of the grass family, a composite genome consisting of 19 chromosome segments was once proposed as a common origin of cereals (Moore et al. 1997). However, taking together the syntenic relationship of grass species and that genome duplication predates grass divergence, we may hypothesize that the common ancestor of grasses was composed of five or seven chromosomes (Fig. 3). This corresponds well with the long-unsolved problem that the original number of rice chromosomes could be five or seven, suggested from the observation of secondary association between paired bivalent chromosomes in the meiosis of diploid rice plants (Sakai 1935; Nandi 1936). The old observation may well agree with the molecular evidence of basic chromosome structure in rice and the common origin of grass families.

References Aggarwal RK, Brar DS, Khush GS (1997) Two new genomes in the Oryza complex identified on the basis of molecular divergence analysis using total genomic DNA hybridization. Mol Gen Genet 254:1–12

244

N. Kurata

Ammiraju JSS, Luo M, Goicoechea JL, et al. (2006) The Oryza bacterial artificial chromosome library resource: construction and analysis of 12 deep-coverage large-insert BAC libraries that represent the 10 genome types of the genus Oryza. Genome Res 16:140–147 Bao W, Zhang W, Yang Q, Zhang Y, Han B, Gu M (2006) Diversity of centromeric repeats in two closely related wild rice species, Oryza officinalis and Oryza rhizomatis. Mol Gen Genet 275:421–430 Cheng Z, Presting GG, Buell CR, Wing RA, Jiang J (2001) High-resolution pachytene chromosome mapping of bacterial artificial chromosomes anchored by genetic markers reveals the centromere location and the distribution of genetic recombination along chromosome 10 of rice. Genetics 157:1749–1757 Cheng Z, Dong F, Langdon T, et al. (2002) Functional centromeres are marked by a satellite repeat and a centromere-specific retrotransposon. Plant Cell 14:1691–1704 Devos KM (2005) Updating the ‘crop circle’. Curr Opinion Plant Biol 8:155–162 Devos KM, Gale, MD (1997) Comparative genetics in the grasses. Plant Mol Biol 35:3–15 Fukui K, Shishido R, Kinoshita T (1997) Identification of the rice D-genome chromosomes by genomic in situ hybridization. Theor Appl Genet 95:1239–1245 Guo X, Xu G, Zhang Y, Wen X, Hu W, Fan L (2006) Incongruent evolution of chromosomal size in rice. Genet Mol Res 30:373–389 Hass-Jacobus B, Futrell-Griggs M, Abernathy B, et al. (2006) Integration of hybridization-based markers (overgoes) into physical maps for comparative and evolutionary explorations in the genus Oryza and in sorghum. BMC Genomics 7:199–214 Hu CH (1961) Comparative karyological studies of wild and cultivated species of Oryza. Taiwan Provincial College of Agriculture, Taichung Ilic K, SanMigiuel PJ, Bennetzen JL (2003) A complex history of rearrangement in an orthologous region of the maize, sorghum, and rice genomes. Proc Natl Acad Sci USA 100:12265–12270 Katayama T (1990) Relationships between chromosome numbers and genomic constitutions in genus Oryza. In: Matsuo T, Futsuhara Y, Kikuchi F, Yamaguchi H (eds) Science of the rice plant, vol 3, Genetics, pp 39–48. Food and Agriculture Policy Research Center Press, Tokyo Khush GS (1990) Report of meetings to discuss chromosome numbering system in rice. Rice Genet Newslett 7:12–15 Khush GS, Singh RJ, Sur SC, Librojo AL (1984) Primary trisomics of rice. Origin, morphology, cytology and use in linkage mapping. Genetics 107:141–163 Kurata N (1985) Chromosome analysis of meiosis and mitosis in rice. Rice Genetics I:143—152. International Rice Research Institute, Manila Kurata N, Fukui K (2003) Chromosome research in genus Oryza. In: Nanda JS, Sharma SD (eds) Monograph on genus Oryza, pp 213—261. Science Publishers, Enfield, USA, Plymouth UK Kurata N, Omura T (1978) Karyotype analysis in rice. I. A new method for identifying all chromosome pairs. Jpn J Genet 53:251–255 Kurata N, Nonomura K, Harushima Y (2002) Rice genome organization: the centromere and genome interactions. Annals Botany 90:427–435 Lee H-R, Zhang W, Langdon T, et al. (2005) Chromatin immunoprecipitation cloning reveals rapid evolutionary patterns of centromeric DNA in Oryza species. Proc Natl Acad Sci USA 102:11793–11798 Lin H, Zhu W, Silva JC, Gu X, Buell CR (2006) Intron gain and loss in segmentally duplicated genes in rice. Genome Biol 7:R41 Lu J, Tang H, Huang J, Shi S, Wu C-I (2006) The accumulation of deleterious mutations in rice genomes: a hypothesis on the cost of domestication. Trends Genet 22:126–131 Ma J, Bennetzen JL (2006) Recombination, rearrangement, reshuffling and divergence in a centromeric region of rice. Proc Natl Acad Sci USA 103:383–388 Miyabayashi T, Nonomura K-I, Kurata N (2007) Genome size determination of twenty wild Oryza species by flow cytometric and chromosome analyses. Breed Sci 57:73–78 Moore G, Aragon-Alcaide L, Roberts M, Reader S, Miller T, Foote T (1997) Are rice chromosomes components of a holocetric chromosome ancestor? Plant Mol Biol 35:17–23

III.2 Chromosome and Genome Evolution in Rice

245

Nandi HK (1936) The chromosome morphology, secondary association and origin of cultivated rice. J Genet 33:315–336 Paterson AH, Bowers JE, Chapman BA (2004) Ancient polyploidization predating divergence of the cereals, and its consequences for comparative genomics. Proc Natl Acad Sci USA 101:9903–9908 Piegu B, Guyot R, Picault N, et al. (2006) Doubling genome size without polyploidization: dynamics of retrotransposition-driven genomic expansions in Oryza australiensis, a wild relative of rice. Genome Res 16:1262–1269 Sakai K (1935) Study of rice chromosomes. 1. Secondary pairing of meiotic chromosomes. Jpn J Genet 11:145–156 (in Japanese) Sasaki T, Matsumoto T, Antonio BA, Nagamura Y (2005) From mapping to sequencing, postsequencing and beyond. Plant Cell Physiol 46:3–13 Uozu S, Ikehashi H, Ohmido N, Ohtsubo H, Ohtsubo E, Fukui K (1997) Repetitive sequences: cause for variation in genome size and chromosome morphology in the genus Oryza. Plant Mol Biol 35:791–799 Wing R, Ammiraju JSS, Luo M, et al. (2005) The Oryza map alignment project: the golden path to unlocking the genetic potential of wild rice species. Plant Mol Biol 59:53–62 Wing R, Luo M, Goicoechea JL, et al. (2006a) The Oryza bacterial artificial chromosome library resource: construction and analysis of 12 deep-coverage large-insert BAC libraries that represent the 10 genome types of the genus Oryza. Genome Res 16:140–147 Wing R, Kim HR, Goicoechea JL, et al. (2006b) The Oryza map aligment project: a new resource for comparative genomic studies within Oryza. Proc 4th Int Rice Funct Genomics Symp 11 Wu J, Kurata N, Tanoue H, et al. (1998) Physical mapping of duplicated genomic regions of two chromosome ends in rice. Genetics 150:1595–1603 Yu J, Wang J, Lin W, et al. (2005) The genomes of Oryza sativa: a history of duplication. Plos Biol 3:266–281

III.3

Genetics of Speciation in Rice Yohei Koide1, Kazumitsu Onishi1, Akira Kanazawa2, and Yoshio Sano1(* ü)

1

Introduction

One of the central issues of evolutionary biology is the origin of the species, although the definition of species is an endlessly debated issue (Coyne and Orr 2004). According to the biological species concept (BSC), a species is a group of an actually or potentially interbreeding natural population, which is reproductively isolated from other such groups (Mayr 1942). No concept of speciation could be complete without a genetic interpretation of the rise of isolating mechanisms. Fitness reduction can range from maladaptation to inviability or sterility. The loci that underlie such reduction in fitness might be considered ‘speciation genes’, which are important in driving the nascent species to become independent genetic entities (Wu and Ting 2004). Therefore, we can analyze the genetic basis of speciation as a more tractable problem by focusing on the genetic basis for reproductive isolation. Recent work on reproductive isolation in Drosophila has advanced our understanding of many fundamental questions about speciation (see review in Coyne and Orr 2004). The BSC can be favorably adopted regarding domesticated plants, and the concept of gene pools based on the degree of their sexual affinities is useful for their classification (Harlan 1975). Any good species are by no means completely isolated. Wild and cultivated complexes in crops are taxonomically distinct but phylogenetically conspecific. Their genetic differentiation is maintained through disruptive selection associated with habitat adaptation, indicating that domestication proceeds at the intra-specific level under human influence. Genetic discontinuities between species are maintained by a variety of isolating barriers which are classified into pre-mating and post-mating barriers. Pre-mating

1 Laboratory of Plant Breeding, Graduate School of Agriculture, Hokkaido University, Kita-9, Nishi-9, Kita-ku, Sapporo 060-8589, Japan e-mail: [email protected]; [email protected]; [email protected] 2

Laboratory of Cell Biology and Manipulation, Graduate School of Agriculture, Hokkaido University, Kita-9, Nishi-9, Kita-ku, Sapporo 060-8589, Japan e-mail: [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

247

248

Y. Koide et al.

barriers include divergences in spatial and ecological habitats, flowering time, floral organs, and reproductive modes (autogamy and apomixis), while post-mating barriers include cross-incompatibility, hybrid inviability (or weakness), hybrid sterility, and hybrid breakdown. Several isolating barriers may act together to prevent gene flow. Pre-mating barriers are highly effective for reproductive isolation, as reported in Mimulus (Ramsey et al. 2003); however, their present importance may distort our assessment of their historical importance during speciation. Pre-mating barriers are sometimes incomplete, and species must have internal barriers that protect their own gene pools from destructive effects after hybridization. Two cultivated rice species (Oryza sativa and O. glaberrima) and their five wild relatives share the same genome A, since their hybrids show no significant disturbance in chromosome pairing. According to the classification of the gene pool, Asian rice (O. sativa) and its wild progenitor (O. rufipogon) belong to the same biological species, forming a primary gene pool, while African rice (O. glaberrima) and its wild progenitor (O. barthii) form another primary gene pool. A crossing barrier is found only between O. longistaminata and its relatives, but those AA genome species are reproductively isolated from each other by F1 sterility barriers (Chu et al. 1969; Morishima 1969). Following the studies of Kato et al. (1928), it has been observed that an F1 sterility barrier is developed even between ssp. indica and ssp. japonica in Asian rice. The two subspecies are well differentiated in molecular markers as well as morphological and physiological characteristics (Garris et al. 2005); however, the F1 sterility relationship is unable to clearly distinguish them in comparison with other characteristics (Oka 1988). Thus, the subspecies status reflects the ambiguity in their species status. Will they diverge or fuse? A continuous degree of isolating barriers is observed at intra- and inter-specific levels in rice, which may enable comparative analysis regarding the time-course of the development of isolating barriers. A flood of studies on isolating barriers have been reported in domesticated plants because interbreeding is extensively carried out for their improvement. In this chapter we review recently accumulated information on the genetic basis of reproductive isolation in plants, and discuss the implications for evolutionary dynamics, focusing on rice.

2 2.1

Genetic Mechanisms of Post-mating Barriers Pre-zygotic Isolation

Reproductive isolation after pollination is classified into pre- and post-zygotic barriers in plants. The former results mainly from pollen–pistil interactions, and the latter from an arrest of the development of young zygotes. Regarding pre-zygotic barriers, in intergeneric hybridizations among bread wheat, rye, and Hordeum bulbosum, cross-incompatibility is regulated by three Kr loci that cause an arrest of pollen tube growth at the base of the stigma, thereby preventing the subsequent penetration of the ovary wall (Snape et al. 1980). This type of cross-incompatibility

This page intentionally blank

III.3 Genetics of Speciation in Rice

249

is rare in rice, as mentioned above. On the other hand, numerous loci for gametophyte factor (ga) have been reported in maize. A Ga allele is well known to show a pronounced advantage in effecting fertilization through competition among pollen grains, giving rise to transmission ratio distortion (TRD) in later generations. The extreme case could be involved in cross-incompatibility when alien pollen grains are a weak competitor. Numerous ga loci are also known in rice; however, it is unclear how they are related to isolating barriers.

2.2

Post-zygotic Isolation

Post-zygotic barriers include hybrid inviability, hybrid sterility, and hybrid breakdown. The genes responsible for post-zygotic barriers reported in plants are listed in Table 1.

2.2.1

Hybrid Inviability

F1 inviability (or weakness) has frequently been described in plants. The products of intra- and inter-specific hybridization may fail to reach reproductive maturity due to a failure of seed development or an aberrant vegetative growth. Abortion in the early stage of seed formation is found as cross-incompatibility after fertilization, which frequently results from an arrest of endosperm development in plants. The failure of endosperm development is known in hybrids between O. sativa and O. longistaminata, in which zygotic lethality is caused by an interaction between Da and Db genes through dosage effects in the triploid endosperm, as shown in Table 1. It is known that seed development is regulated by the balance of maternal and paternal genomes in the endosperm in plants, as demonstrated in interploidy crosses in maize and Arabidopsis, suggesting that the two parental genomes are not equivalent (Lin 1984; Scott et al. 1998). To explain the failure of endosperm formation, it has been proposed that normal development requires a proper balance of effective ploidy between the female and male genome in Avena (Nishiyama and Yabuno 1979) and Solanum (Johnston et al. 1980), although the genetic basis for this is unknown. In rice, unidirectional cross-incompatibility after fertilization is detected between O. sativa and O. rufipogon (Matsubara et al. 2003). The cross incompatibility is controlled by Cif in the female and by cim in the male, indicating a conflict between the two sexes. In addition, a dominant suppressor, Su-Cif, changes the reaction in the female. All three genes act sporophytically, suggesting that the phenomenon might be explained by mechanisms such as a transmission of some products and signals from gametes (Browning and Strome 1996) or epigenetic modifications (Grossniklaus et al. 1998; Kinoshita et al. 1999). Aberrant developmental syndromes, such as necrosis or chlorosis, have been described in hybrids of plants. Such an F1 inviability or weakness is also controlled by a set of complementary genes in rice and wheat, and by the heterozygous state at a single locus in cotton and Crepis (Table 1).

250

Y. Koide et al.

Table 1 The genes responsible for post-zygotic barriers reported in plants. Note: gametophyte genes are not included since their gene action is unclear in most cases (see text) Gene No. of Gene or factor (ref.) loci Phenotypea actionb Species F1 hybrid inviability and weakness 1 I (plant) S Crepis capillaris and C. tec- L (1) torum 1 I (plant) S Gossypium hirsutum and ck (2) G. barbadense 2 I (seed) S Oryza sativa and O. longis- Da/Db (3) taminata 2 I (seed) S O. sativa and O. rufipogon Cif/cim (3) 2 I (plant) S Within Secale cereale Ner1/Ner2 (4) 2 I (plant) S Within Triticum aestivun Ne1/Ne2 (5) 2 I (plant) S T. aestivum and T. macha, Ch1/Ch2 (6) T. dicoccum, T. dicocoides, T. durum 2 I (plant) S T. aestivum and S. cereale Chr1 (7) 2 I (plant) S T. dicocoides and Cs1/Cs2 (8) T. timopheevi 2 W (plant) S O. glaberrima and O. barthii hwb1/hwb2 (3) 2 W (plant) S Within O. sativa Hwa1/Hwa2 (3), Hwc1/Hwc2 (3) ? I (plant) S Nicotiana tabacum and Q-chromosome (9) N. suaveolens F1 hybrid sterility 1

S (pollen)

G

1

S (pollen)

G

1 1

S (pollen) S (pollen)

G G

1

S (pollen)

G

1 1

S (pollen) S (pollen)

G S

1

S (pollen)

?

1 1

S (pollen) S (pollen)

? ?

1

S (seed)

G

1 1 1

S (seed) S (seed) S (seed)

G G G

N. tabacum and N. plumbag- Kl (10) inifolia Within O. sativa S7 (3), S8 (3), S9 (3), S14 (3), Sa (3), Sb (3),Sc (3), Sd (3), Se (3), Sf (3), ga11 (3), ga14 (3) O. sativa and O. glaberrima S3 (3), S29 (3), S34 (3) O. sativa and O. glumaeS12 (3), S22 (3), S27 (3), S28 (3) patula O. sativa and O. longistaS13 (3) minata Within Triticum aestivum Ki (11) T. aestivum and Hordeum Shw (12) vulgare Within O. sativa S24 (3), S25 (3), S26 (3), f5 (13), pf5 (3), pf1 (3) O. sativa and O. glaberrima S18 (3), S19 (3), S20 (3), S21 (3) O. sativa and O. glumaeS23 (3) patula Within O. sativa S5 (3), S10 (3), S16 (3), S17 (3), S29 (3) O. sativa and weedy rice S30 (14) T. aestivum and Agropyron Agropyron chromosome (15) Tripsacum chromosome (16) Zea mays and Tripsacum dactyloides (continued)

III.3 Genetics of Speciation in Rice Table 1 (continued) Gene No. of loci Phenotypea actionb 1 S (seed) ?

1 1 1 1 ? 2 2

S (pollen, seed) S (pollen, seed) S (pollen, seed) S (pollen, seed) S (anther) S (anther) S (pollen)

G

Species Within O. sativa

Gene or factor (ref.) S31 (3), S32 (3),qESA-12-1 (3), qESA-12-2 (3), qESA-6 (3), f (13), f1 (13), f6 (13), f8 (13), Ef6 (3), spf5 (3), spf6 (3), spf8 (3) Ge (17)

G

Lycopersicon esculentum and L. pimpinellifolium O. sativa and O. glaberrima S1 (3), S2 (3), S33 (3)

G

O. sativa and O. rufipogon

S6 (3)

G

T. aestivum and Aegilops

Gc (18)

S S G

O. sativa and O. glaberrima Rfj (3) Within O. sativa SA/SB (3) Within O. sativa sa1/sa2 (3), sd1/sd (3), se1/se2 (3), sc1/sc2 (3), SA1/SA (3), SB1/SB (3) X/Y chromosome (19) Within Silene latifolia Within Hordeum vulgare sfg1/sfg2 (20) Within O. sativa d60/gal (3)

2 2 2

S (pollen) G S (seed) G S (pollen, G seed) F2 hybrid breakdown 2

I or W (plant)

S

Within O. sativa

2 2

S (pollen) S (pollen)

S S

2 3 ?

S (seed) S (seed) W (plant)

S S/G S

?

S (seed)

?

Within Brassica napus Mimulus guttatus and M. nastus Within O. sativa Within O. sativa O. sativa and O. glumaepatula Within O. sativa

a

251

hca1/hca2 (3), hwd1/hwd2 (3), hwe1/hwe2 (3), hwg1/hwg2 (3) Bnms1/Bnms2 (21) hms1/hms2 (22) fes1/fes2 (3), s (3) hsa1/hsa2/hsa3 (3) hwf1 (3) qSS-2 (23), qSS-6 (23), qSS-8 (23)

I, W, and S indicate inviability, weakness, and sterility, respectively S and G indicate sporophytic and gametophytic types, respectively References: 1 Hollingshead 1930; 2 Stephens 1946; 3 Rice Genetics Newsletters and Oryzabase (http://www.shigen.nig.ac.jp/rice/oryzabase/top/top.jsp); 4 Ren and Lelley 1990; 5 Tsunewaki 1960; 6 Tsunewaki 1966; 7 Tomar and Singh 1998; 8 Tsunewaki 1992; 9 Tezuka and Marubashi 2006; 10 Cameron and Moav 1957; 11 Loegering and Sears 1963; 12 Taketa et al. 2002; 13 Wang et al. 1998; 14 Zhu et al. 2005; 15 Scoles and Kibirge-Sebunya 1983; 16 Maguire 1963; 17 Rick 1966; 18 Endo and Tsunewaki 1975; 19 Taylor and Ingvarsson 2003; 20 Fukushima and Konishi 1994; 21 Yi et al. 2006; 22 Sweigart et al. 2006; 23 Wang et al. 2005

b

252

2.2.2

Y. Koide et al.

Hybrid Sterility

The genetic mechanism of hybrid sterility has been extensively studied in rice. Hybrid sterility is defined as genic, chromosomal, or cytoplasmic (Grant 1981). Chromosomal rearrangements frequently cause infertility in hybrids, as indicated by abnormal chromosomal pairing, formation of multivalents, and other abnormalities at meiosis. Cryptic differences in chromosomes are regarded as a major cause of hybrid sterility in plants. Although the chromosome doubling test can be used to evaluate the presence of chromosomal rearrangements, there are discrepancies regarding the involvement of cryptic differences in the subspecific crosses of Asian rice (Oka 1988). Thus, cryptic differences in chromosomes cannot always be distinguished from genic differences. Hybrid sterility is often caused by cytoplasmic differences. Cytoplasmic male sterility (CMS) has also been extensively studied in rice (Li and Zho 1986; Virmani and Shinjo 1988). The cytoplasm from O. rufipogon frequently induced male sterility (Shinjo 1984). Restoration of cytoplasmic sterility can be controlled by gametophytic or sporophytic restorers, which mostly act as dominant genes. Since sporophytic restorers are prevalent in plants, cytoplasmic sterility works as hybrid breakdown rather than as an F1 sterility barrier, except for gametophytic restorers. As reviewed by Fujii et al. (Chap. II.7, this volume), a gametophytic restorer (Rf1) was recently identified as PPR motifs, although it remains to be studied whether it is responsible for reproductive isolation. Genes causing hybrid sterility are either gametophytic or sporophytic in action. Assuming one or two loci, four genetic models for F1 hybrid sterility are possible. These genetic systems could affect the development of gametes produced by the male or female, or both. In the one-locus model, both gametophytic and sporophytic sterility genes induce infertility through an allelic interaction. It should be noticed in this one-locus model that sporophytic genes give a similar segregation pattern to interchanges (interstitial translocation), although no such gene is known in any plant species. A hybrid sterility gene (S) induces abortion of gametes carrying its opposite allele (Sa) in the heterozygotes (S/Sa) in a one-locus model. Both homozygotes (S/S and Sa/Sa) are fertile so that infertility only appears following hybridization between them, showing the selfish nature of S. This type of hybrid sterility is detected in various plant species, including Nicotiana, Triticum, Zea, Lycopersicon, Aegilops, and Agropyron, as well as rice species (Table 1). In these species, a gamete eliminator causes dysfunction in both male and female gametophytes, showing the strongest effect on fertility in hybrids of the plants. Oryza sativa and O. glaberrima each carries a gamete eliminator (S1 or S2), which induces abortion of gametes with alleles from its opposite parent (Sano et al. 1979). The list of the reported genes suggests that gamete eliminators are not always found between distantly related taxa, and are detected in crosses between plants even within a species (Morishima et al. 1992). At the Gamete eliminator (Ge) locus in Lycopersicon, three alleles (Gep, Gec, and Gen) are present, in which the distorter allele Gep might have arisen from the neutral allele Gen that gave no abortion with the other two alleles (Rick 1966, 1971). In addition, a dominant suppressor for

III.3 Genetics of Speciation in Rice

253

gametocidal gene (Gc) is found in wheat (Tsujimoto and Tsunewaki 1985). Such a suppressor could act as a neutral allele when it linked to Gc. Instead of gamete eliminators, the causal alleles frequently induce abortion of only megaspores or microspores in various plants. Male sterility seems to be more prevalent than female sterility; however, female sterility is often reported in crosses between the two subspecies of Asian rice (Ikehashi and Araki 1986; Zhu et al. 2004). In these loci, neutral alleles conferring wide compatibility are present and some of them have been finely mapped (Ji et al. 2005; Qiu et al. 2005; Wang et al. 2006). However, it remains to be determined whether these neutral alleles act specifically on the alleles at their own locus or on various sterility genes at different loci. Regarding the two-locus model, sporophytic sterility genes have only been reported to explain anther indehiscence found within O. sativa and between O. sativa and O. glaberrima, whereas sterility genes with gametophytic action are prevalent (Table 1). In the case of gametophytic genes, the genetic model assumes that the genotypes of parents are s1s1+1+1 and +1+1 s2s2, respectively, and that the recombinant gametes having s1s2 become inviable in the hybrid. This was first reported in a varietal cross of O. sativa and was called the duplicate gametic lethal model (Oka 1988). A similar type of sterility genes is frequently found between ssp. indica and ssp. japonica, within ssp. japonica and in Hordeum, as listed in Table 1, although the affected organs (male or female, or both) vary in these crosses. Duplicate gametic lethal genes have been detected only within species, but this might result from the fact that their effects are relatively small, since only 25% of gametophytes are sterile in the plants of +1s1+2s2. It should be mentioned that such an interlocus epistasis can restrict recombination between genes on different chromosomes (Oka 1988). In addition, the duplicate gametic lethal model is interesting with regard to the evolutionary role of duplicated genes. Gene duplication is a source of material for the origin of evolutionary novelties, but it is still unclear how frequently they result in new functions. Duplicate gametic lethal genes might support the hypothesis that stochastic silencing of duplicate genes plays a significant role in the passive origin of new species (Lynch and Conery 2000), although gene transposition possibly contributes to this phenomenon, as reported in Drosophila (Masly et al. 2006).

2.2.3

Hybrid Breakdown

Even when F1 hybrids are vigorous and fertile, severe infertility or a high rate of mortality is often reported in their progenies. This phenomenon has been described in many plant species and is known as hybrid breakdown. In rice, hybrid breakdown is frequently found in varietal crosses and it is mostly attributable to a set of complementary genes, either of which is necessary for normal growth (Table 1). In contrast to F1 inviability, developmental abnormalities are observed in the F2 and its later generations. The occurrence of abnormalities in F1 or F2 depends on their degree of dominance. In most cases, F2 breakdown appears to be caused by sporophytic (or complementary recessive) genes (Li et al. 1997; Fukuoka et al. 1998). However, an

254

Y. Koide et al.

epistatic interaction between sporophytic and gametophytic genes has been reported to explain F2 sterility (Kubo and Yoshimura 2005). In Mimulus, interlocus epistasis is considered to induce nearly complete male sterility and partial female sterility (Sweigart et al. 2006). Further, CMS gives rise to F2 breakdown, as mentioned above, although CMS and its restorers are not included in the list of Table 1.

3 3.1

Evolution of Reproductive Isolation Fixation of ‘Speciation Genes’

One of the debated issues in studies of speciation is how ‘speciation genes’ could be fixed in a species despite the fact that deleterious genes reducing fitness are eliminated within a population. A simple genetic mechanism for the origin of reproductive isolation was proposed by Dobzhansky and Muller (the DM model; Dobzhansky 1970; Coyne and Orr 2004). Assuming two loci for simplicity, one daughter species becomes fixed for an allele at one locus, whereas the other daughter species becomes fixed for a second allele at another locus. Hybrid incompatibility would be established without reducing fitness if both of these mutations are neutral (or advantageous) within the population in which they arose; however, they cause hybrid incompatibility when expressed together in the hybrid. In plants, the DM model is also adopted for the two-locus model for hybrid inviability, hybrid sterility, and hybrid breakdown. Another possible explanation is stepwise mutations allowing the development of isolating barriers at a single locus without a reduction in fitness (Nei et al. 1983). This model assumed that there are multi-allelic loci, including intermediate alleles which are compatible with alleles causing hybrid incompatibilities (Nei et al. 1983). An example is that an allele (Gep) of gamete eliminator has arisen from the neutral allele Gen in Licopersicon, as mentioned above.

3.2

The Selfish Nature of Hybrid Sterility Genes

The DM or stepwise-mutation model means that the evolution of reproductive isolation need not be opposed by natural selection; however, it does not indicate how such genes can spread and become fixed in the population. These two models simply assume that the alleles ultimately causing hybrid incompatibility have little or no effect on fitness in their normal species background and randomly drift to fixation, although this is supported by little empirical data. The alternative assumption is that some form of positive selection, such as the selfish genetic elements, plays an important role in the origin of post-zygotic isolation (Hurst and Werren 2001; Coyne and Orr 2004).

III.3 Genetics of Speciation in Rice

255

Meiotic drive (MD) or the TRD system is one of the selfish genetic elements that violates Mendel’s rules by the preferential transmission of a particular chromosome or allele at the expense of its partner. MD was first defined to explain such a force that is a consequence of the mechanics of the meiotic divisions (Sandler and Novitski 1957), while well-known MD systems such as t-haplotype in mouse and Segregation Distorter (SD) in Drosophila result from post-meiotic dysfunction of spores or gametes, as frequently found in F1 sterility genes of plants. The selfish behavior will drastically alter the frequencies of alleles in a population, thereby affecting the genomic constitution; however, TRD occurs through various factors, including preferential transmission of chromosomes (or the true MD system; Fishman and Willis 2005). Segregation distortion is often observed in intra- and inter-specific crosses in plants. Gametophyte genes causing certation (competition among pollen grains) result in TRD in rice (Nakagahra 1972; Oka 1988). The development of highdensity molecular linkage maps significantly contributes to understanding their genome-wide distribution (Xu et al. 1997; Harushima et al. 2001). These findings are meaningful because the concept of reproductive isolation is fundamentally a whole-genome concept. A number of causal factors for segregation distortion have been reported throughout the genome, although certation due to a gametophyte gene is not easily distinguished from post-meiotic dysfunction of gametes in many cases. Between maize and teosinte, for example, cross-incompatibility is controlled by a series of alleles at the ga1 locus, which was originally detected as a gametophyte gene causing segregation distortion due to certation (Evans and Kermicle 2001). Theoretical studies indicate that sterility barriers could be evolved by a mutual imbalance between MD systems (in a broad sense) (Frank 1991; Hurst and Pomiankowski 1991). The MD genes can spread throughout populations due to their selfish nature, and two populations might evolve different MD factors. Recent empirical work in Drosophila supports the contribution of the MD system to the rise of reproductive isolation (Coyne and Orr 2004). In plants, male-sterility could be maintained within a population, especially in perennials, since partially male-sterile plants show less effects on the fecundity. A vast number of haploid genomes can be screened by intensified gametophytic selection of males, and genetic changes are achieved at relatively little cost (Mulcahy 1979; Sano 1983). Self-pollinated plants make it possible to yield homozygotes, allowing the fixation of pollen killer and gamete eliminator, and thus minimizing heterozygote disadvantages in the population. It is also postulated that sterility genes in rice have a selfish nature (Oka 1974). The duplicated gametic lethal model is regarded as a typical example of the DM model. It is, however, postulated that the gametes carrying s1+2 or +1s 2 show a higher fertilization capacity due to the presumed certation than those carrying +1+2. When s 1 and s 2 are present together with + 1 and + 2 within a population, reproductively isolated plants (s1 s1+2+2 and + 1+ 1 s2 s2) appear automatically through certation.

256

4

Y. Koide et al.

Conclusions and Perspectives

As far as the reported genes in plants are concerned, similar genes for reproductive isolation are found in intra- and inter-specific hybrids. This implies that genes for partially isolating barriers rapidly accumulate during speciation. The mechanism of such a rapid divergence is still mysterious even though ‘genetic revolution’ or a peak shift model (Carson 1975), ‘genome shock’ (McClintock 1984), and epigenetic remodeling (Comai et al. 2000) have been proposed to explain genome reorganization. Regarding isolating barriers in rice, it seems that male sterility is highly polygenic and complex, involving interlocus epistasis. Incompatibilities influencing hybrid male sterility appear to evolve most readily, such that male sterility factors are much more prevalent than female sterility or hybrid inviability factors, as indicated in Drosophila (Wu 2001). On the other hand, hybrid male sterility is frequently controlled by gametophytic genes in plants, while male sterility is frequently controlled by sporophytic genes in Drosophila, supporting the notion that the genetics of speciation may be distinct between plants and animals (Doebley et al. 1997; Walbot and Evans 2003). Reproductive isolation is not produced directly by natural selection, in contrast to adaptive divergence, showing that it is a by-product of the genetic divergence of isolated gene pools. We know now that numerous genes affecting reproductive isolation are involved among species; however, we know little regarding why so many hybrid incompatibility genes arise during speciation, and not under the direct effect of natural selection. The genetics of speciation reveals that naturally occurring speciation is a process whereby populations will not lose their divergence upon contact and, as a result, will continue to diverge. In crops, artificial selection and wide hybridization greatly accelerates the rate of evolution in response to changing environments, although our present knowledge is limited as to the possible risks to the sustainability of crops. The evolutionary trajectory (fusion vs. speciation) of populations in a biological species depends on how strongly genes are coadapted within each subdivided population and how often the coadapted gene complexes are broken up. Both the genetic architecture of differentiation and the extent of gene flow between nascent species are crucial for their continuing evolution. The genetics of speciation in crops should be revisited after we have a comprehensive understanding of ‘speciation genes’ and their role, which will undoubtedly give more insight into how to manage crops and their diversity in our changing world.

References Browning H, Strome S (1996) A sperm-supplied factor required for embryogenesis in C. elegans. Development 122:391–404 Cameron DR, Moav R (1957) Inheritance in Nicotiana tabacum XXVII. Pollen killer, an alien genetic locus inducing abortion of microspores not carrying it. Genetics 42:326–335 Carson HL (1975) The genetics of speciation at the diploid level. Am Nat 109:83–92

III.3 Genetics of Speciation in Rice

257

Chu YE, Morishima H, Oka HI (1969) Reproductive barriers distributed in cultivated rice species and their wild relatives. Jpn J Genet 44:207–223 Comai L, Tyagi AP, Winter K, et al. (2000) Phenotypic instability and rapid gene silencing in newly formed Arabidopsis allotetraploids. Plant Cell 12:1551–1568 Coyne JA, Orr HA (2004) Speciation. Sinauer Associates, Sunderland, Massachusetts Dobzhansky T (1970) Genetics of the evolutionary process. Columbia University Press, New York Doebley J, Stec A, Hubbard L (1997) The evolution of apical dominance in maize. Nature 386:485–488 Endo TR, Tsunewaki K (1975) Sterility of common wheat with Aegilops triuncialis cytoplasm. J Hered 66:13–18 Evans MMS, Kermicle JL (2001) Teosinte crossing barrier1, a locus governing hybridization of teosinte with maize. Theor Appl Genet 103:259–265 Fishman L, Willis JH (2005) A novel meiotic drive locus almost completely distorts segregation in Mimulus (monkeyflower) hybrids. Genetics 169:347–353 Frank S (1991) Divergence of meiotic drive-suppression system as an explanation for sex-biased hybrid sterility and inviability. Evolution 45:262–267 Fukuoka S, Namai H, Okuno K (1998) RFLP mapping of the genes controlling hybrid breakdown in rice (Oryza sativa L.). Theor Appl Genet 97:446–449 Fukushima Y, Konishi T (1994) Genetic studies on hybrid sterility in barley. Jpn J Genet 69:719–726 Garris AJ, Tai TH, Coburn J, Kresovich S, McCouch S (2005) Genetic structure and diversity in Oryza sativa L. Genetics 169:1631–1638 Grant V (1981) Plant speciation, 2nd edn. Columbia University Press, New York Grossniklaus U, Vielle-Calzada JP, Hoeppner MA, Gagliano WB (1998) Maternal control of embryogenesis by MEDEA, a polycomb group gene in Arabidopsis. Science 280:446–450 Harlan JR (1975) Crop and man. American Society of Agronomy, Madison, Wisconsin Harushima Y, Nakagahra M, Yano M, Sasaki T, Kurata N (2001) A genome-wide survey of reproductive barriers in an intraspecific hybrid. Genetics 159:883–892 Hollingshead L (1930) A lethal factor in Crepis effective only in an interspecific hybrid. Genetics 15:114–140 Hurst GDD, Werren JH (2001) The role of selfish genetic elements in eukaryotic evolution. Nat Rev Genet 2:597–606 Hurst LD, Pomiankowski A (1991) Causes of sex ratio bias may account for unisexual sterility in hybrid: a new explanation of Haldane’s rule and related phenomena. Genetics 128:841–858 Ikehashi H, Araki H (1986) Genetics of F1 sterility in remote crosses of rice. In: International Rice Research Institute (ed) Rice genetics. IRRI, Los Baños Ji Q, Lu J, Chao Q, Gu M, Xu M (2005) Delimiting a rice wide-compatibility gene S5n to a 50 kb region. Theor Appl Genet 111:1495–1503 Johnston SA, den Nijs TPM, Peloquin SJ, Hanneman RE (1980) The significance of genic balance to endosperm development in interspecific crosses. Theor Appl Genet 57:5–9 Kato S, Kosaka H, Hara S (1928) On the affinity of rice varieties as shown by fertility of hybrid plants. Bull Sci Fac Agric Kyushu Univ 3:132–147 Kinoshita T, Yadegari R, Harada JJ, Goldberg RB, Fischer RL (1999) Imprinting of the MEDEA polycomb gene in the Arabidopsis endosperm. Plant Cell 11:1945–1952 Kubo T, Yoshimura A (2005) Epistasis underlying female sterility detected in hybrid breakdown in a japonica–indica cross on rice (Oryza sativa L.). Theor Appl Genet 110:346–355 Li Z, Zho Y (1986) Rice male-sterile cytoplasm and fertility-restoration. In: International Rice Research Institute (ed) Hybrid rice. IRRI, Manila, pp 85–102 Li Z, Pinson SRM, Paterson AH, Park WD, Stansel JW (1997) Genetics of hybrid sterility and hybrid breakdown in an intersubspecific rice (Oryza sativa L.) population. Genetics 145:1139–1148 Lin BY (1984) Ploidy barrier to endosperm development in maize. Genetics 107:103–115 Loegering WQ, Sears ER (1963) Distorted inheritance of stem-rust resistance of Timstein wheat caused by a pollen-killing gene. Can J Genet 5:65–72

258

Y. Koide et al.

Lynch M, Conery JS (2000) The evolutionary fate and consequences of duplicate genes. Science 290:1151–1155 Maguire MP (1963) High transmission frequency of a Tripsacum chromosome in corn. Genetics 48:1185–1194 Masly JP, Jones CD, Noor MAF, Locke J, Orr HA (2006) Gene transposition as a cause of hybrid sterility in Drosophila. Science 313:1448–1450 Matsubara K, Khin-Thidar, Sano Y (2003) A gene block causing cross-incompatibility hidden in wild and cultivated rice. Genetics 165:343–352 Mayr E (1942) Systematics and the origin of species. Columbia University Press, New York McClintock B (1984) The significance of responses of the genome to challenge. Science 226:792–801 Morishima H (1969) Phenetic similarity and phylogenetic relationships among strains of Oryza perennis, estimated by methods of numerical taxonomy. Evolution 23:429–443 Morishima H, Sano Y, Oka HI (1992) Evolutionary studies in cultivated rice and its wild relatives. Oxford Surv Evol Biol 8:135–184 Mulcahy DL (1979) The rise of the angiosperms: a genecological factor. Science 206:20–23 Nakagahra M (1972) Genetic mechanism on the distorted segregation of marker genes belonging to the eleventh linkage group in cultivated rice. Jpn J Breed 22:232–238 Nei M, Maruyama T, Wu CI (1983) Models of evolution of reproductive isolation. Genetics 103:557–579 Nishiyama I, Yabuno T (1979) Triple fusion of the primary endosperm nucleus, a cause of interspecific cross-incompatibility in Avena. Euphytica 28:57–65 Oka HI (1974) Analysis of genes controlling F1 sterility in rice by the use of isogenic lines. Genetics 77:521–534 Oka HI (1988) Origin of cultivated rice. Elsevier, Tokyo Qiu SQ, Liu K, Jiang JX, et al. (2005) Delimitation of the rice wide compatibility gene S5n to a 40-kb DNA fragment. Theor Appl Genet 111:1080–1086 Ramsey J, Bradshaw HD, Schemske DW (2003) Components of reproductive isolation between the monkeyflowers Mimulus lewisii and M. cardinalis (Scrophulariaceae). Evolution 57:1520–1534 Ren ZL, Lelley T (1990) Chromosomal localization of genes in the R genome causing hybrid necrosis in rye and triticale. Genome 33:40–43 Rick CM (1966) Abortion of male and female gametes in the tomato determined by allelic interaction. Genetics 53:85–96 Rick CM (1971) The tomato Ge locus: linkage relations and geographic distribution of alleles. Genetics 67:75–85 Sandler L, Novitski E (1957) Meiotic drive as an evolutionary force. Am Nat 91:105–110 Sano Y (1983) A new gene controlling sterility in F1 hybrids of two cultivated rice species. J Hered 74:435–439 Sano Y, Chu YE, Oka HI (1979) Genetic studies of speciation in cultivated rice, I. Genic analysis for the F1 sterility between O. sativa L. and O. glaberrima Steud. Jpn J Genet 54:121–132 Scoles GJ, Kibirge-Sebunya IN (1983) Preferential abortion of gametes in wheat induced by an Agropyron chromosome. Can J Cytol 25:1–6 Scott RJ, Spielman M, Bailey J, Dickinson HG (1998) Parent-of-origin effect on seed development in Arabidopsis thalina. Development 125:3329–3341 Shinjo C (1984) Cytoplasmic male sterility and fertility restoration in rice having genome A. In: Tsunoda S, Takahashi N (eds) Biology of rice. Elsevier, Amsterdam, pp 321–338 Snape JW, Bennett MD, Simpson E (1980) Post-pollination events in crosses of hexaploid wheat with tetraploid Hordeum bulbosum. Z Pflanzenzüchtg 85:200–204 Stephens SG (1946) The genetics of “corky” I. The New World alleles and their possible role as an interspecific isolating mechanism. J Genet 47:150–161 Sweigart AL, Fishman L, Willis JH (2006) A simple genetic incompatibility causes hybrid male sterility in Mimulus. Genetics 172:2465–2479

III.3 Genetics of Speciation in Rice

259

Taketa S, Choda M, Ohashi R, Ichii M, Takeda K (2002) Molecular and physical mapping of a barley gene on chromosome arm 1HL that causes sterility in hybrids with wheat. Genome 45:617–625 Taylor DR, Ingvarsson PK (2003) Common features of segregation distortion in plants and animals. Genetica 117:27–35 Tezuka T, Marubashi W (2006) Hybrid lethality in interspecific hybrids between Nicotiana tabacum and N. suaveolens: evidence that the Q chromosome causes hybrid lethality based on Q-chromosome-specific DNA markers. Theor Appl Genet 112:1172–1178 Tomar SMS, Singh B (1998) Hybrid chlorosis in wheat × rye crosses. Euphytica 99:1–4 Tsujimoto H, Tsunewaki K (1985) Gametocidal genes in wheat and its relatives. II. Suppressor of chromosome 3C gametocidal gene of Aegilops triuncialis. Can J Genet Cytol 27:178–185 Tsunewaki K (1960) Monosomic and conventional gene analysis in common wheat. III. Lethality. Jpn J Genet 35:71–75 Tsunewaki K (1966) Gene analysis on chlorosis of the hybrid, Triticum aestivum var. Chinese Spring × T. macha var. Subletschchumicum, and its bearing on the genetic basis of necrosis and chrolosis. Jpn J Genet 41:413–426 Tsunewaki K (1992) Aneuploid analysis of hybrid necrosis and hybrid chlorosis in tetraploid wheats using the D genome chromosome substitution lines of durum wheat. Genome 35:594–601 Virmani SS, Shinjo C (1988) Current status of analysis and symbols for male-sterile cytoplasms and fertility-restoring genes. Rice Genet Newsl 5:9–15 Walbot V, Evans MS (2003) Unique features of the plant life cycle and their consequences. Nat Rev Genet 4:369–379 Wang C, Zhu C, Zhai H, Wan J (2005) Mapping segregation distortion loci and quantitative trait loci for spikelet sterility in rice (Oryza sativa L.). Genet Res 86:97–106 Wang GW, He YQ, Xu CG, Zhang Q (2006) Fine mapping of f5-Du, a gene conferring widecompatibility for pollen fertility in inter-subspecific hybrids of rice (Oryza sativa L.). Theor Appl Genet 112:382–387 Wang J, Liu KD, Xu CG, Li XH, Zhang Q (1998) The high level of wide-compatibility of variety ‘Dular’ has a complex genetic basis. Theor Appl Genet 97:407–412 Wu CI (2001) The genic view of the process of speciation. J Evol Biol 14:851–864 Wu CI, Ting CT (2004) Genes and speciation. Nat Rev Genet 5:247–257 Xu Y, Zhu L, Xiao J, Huang N, McCouch SR (1997) Chromosomal regions associated with segregation distortion of molecular markers in F2, backcross, doubled haploid, and recombinant inbred populations in rice (Oryza sativa L.). Mol Gen Genet 253:535–545 Yi B, Chen Y, Lei S, Tu J, Fu T (2006) Fine mapping of the recessive genic male-sterile gene (Bnms1) in Brassica napus L. Theor Appl Genet 113:643–650 Zhu S, Jiang L, Wang C, Zhai H, Li D, Wan J (2005) The origin of weedy rice Ludao in China deduced by genome wide analysis of its hybrid sterility genes. Breed Sci 55:409–414 Zhu SS, Wang CM, Zheng TQ, Zhao ZG, Wan JM, Ikehashi H (2004) A novel gene causing hybrid sterility in a remote cross of rice (Oryza sativa L.). Rice Genet Newsl 21:44–45

III.4

Genetic Diversity in Wild Relatives of Rice and Domestication Events Hong-Wei Cai1, Masahiro Akimoto2, and Hiroko Morishima3(* ü)

1

Introduction

Two cultivated rice species, Oryza sativa and O. glaberrima, are classified as the O. sativa complex, together with five related wild species. These wild taxa are O. rufipogon (Asia and Oceania), O. barthii (Africa), O. longistaminata (Africa), O. glumaepatula (Latin America) and O. meridionalis (Oceania). They are diploid and share the AA genome in common. According to the gene pool concept (Harlan 1975), O. sativa of Asiatic origin forms a primary gene pool with its wild progenitor O. rufipogon, and O. glaberrima of African origin forms a second primary gene pool with its wild progenitor O. barthii. During the past decade, a wealth of data, obtained by molecular tools, on AA-genome Oryza species have significantly increased our knowledge of genetic diversity among and within species of this taxonomic group. However, speciation mechanisms are still not fully elucidated. In this chapter, we intend to draw an evolutionary perspective of the O. sativa complex from the viewpoints of genetics and ecology, with special emphasis on the evolutionary dynamics of O. rufipogon and domestication that occurred within the species.

1

College of Agronomy and Biotechnology, China Agricultural University, Yuanmingyuan West Road, Beijing 100094, China e-mail: [email protected] 2 School of Agriculture, Obihiro University of Agriculture and Veterinary Medicine, Inada-cho, Obihiro 080-8555, Japan e-mail: [email protected] 3 Saiwai-cho 15-2, Hiratsuka 254-0804, Japan e:mail: [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

261

262

2 2.1

H.-W. Cai et al.

Genetic Diversity of AA-genome Wild Taxa Variation in Phenotypic Characters and Habitat Conditions

An extensive study on phenotypic variation among and within AA-genome wild species, based on 22 morphophysiological traits of 160 accessions, indicated that a major trend of variation is related to differentiation in their life history (fecundity and survivorship schedules) and mating system (Akimoto et al. 1998a). Members of this plant group are basically aquatic and sun-loving plants. They show wide adaptability to diverse environments differentiating various reproductive strategies. O. rufipogon shows large intra-species diversity and can be divided into perennial and annual (often referred as O. nivara) ecotypes, with a continuum of intermediate types as reviewed by Oka (1988) and Morishima et al. (1992). The perennial-type populations are found in deep swamps and characterized by high vegetative-propagating ability, low seed productivity, tall stature, long anthers and late flowering. On the other hand, the annual-type populations are found in frequently disturbed temporary swamps, which are parched in the dry season, and characterized by high seed propagating ability, high seed productivity, short stature, short anthers, high seed dispersability and early flowering. O. glumaepatula also shows large intra-species diversity, and different geographical groups seem to be associated with ecotype differentiation (Akimoto et al. 1998b). The first group, populations distributed in Central America and northern South America, have characteristics similar to those of perennial O. rufipogon. This similarity often provokes controversy about the taxonomic status of O. glumaepatula. The second group, populations growing in the Amazon River, exhibit a unique adaptive strategy to survive an annual oscillation of water depth of as much as 10 m. Confronted by the rapid increase in river water, their culms are easily broken at the basal part, allowing the released plants to continue their growth on the water surface as “floating meadow”. They are considered to have an annual life cycle. Based on population genetic study, Akimoto et al. (1998b) showed that probably the Amazonian ecotype is predominantly inbreeding, while populations of other regions seem to have a mixed-mating system. The third group, populations growing in the Pantanal, a huge swampy area in southern Brazil, can be distinguished from other members of this species by their tall stature and large panicles. Judging from vigorous ratooning at the post-reproductive stage, the Pantanal populations are regarded as perennial, even though they produce many seeds. Their spikelets shed on the water can float due to air occupying about 40% of the inner space of their hulls. This may be an adaptive structure of the spikelets for seed dispersal on the water surface. O. meridionalis and O. barthii are completely annual species and phenotypic diversity within species is relatively small. Both species share a set of characteristics such as short anthers, high seed fecundity and weak vegetative propagation ability. They commonly inhabit unstable and occasionally disturbed environments,

This page intentionally blank

III.4 Genetic Diversity in Wild Relatives of Rice

263

such as a shallow swamp that is parched during the dry season. Seed propagation is plausibly effective for maintaining the populations in such an environment. These two species probably have a relatively high rate of inbreeding. On the other hand, O. longistaminata is a strongly perennial species propagating by rhizomes. It has large anthers and seems to invest considerable resources in pollen production. Partial self-incompatibility found in O. longistaminata suggests that this species is predominantly outbreeding. Thus, vegetative propagating ability and outcrossing rate are positively correlated among species as well as among ecotypes within species. This association has a great impact on the genetic structure of populations, and consequently evolutionary fate.

2.2

Intra-species Genetic Diversity and Species Relationships

AA-genome wild species are distributed on different continents respectively (except for two species in Africa) and reproductively isolated from one another by F1 pollen sterility. Molecular genetic studies have revealed distinct genetic differences among these species (Second 1985; Dally and Second 1990; Wang et al. 1992; Doi et al. 2000; Kanazawa et al. 2000; Cheng et al. 2002; Zhu and Ge 2005), indicating they have followed independent evolutionary pathways. Phylogenetic relationships inferred in those studies showed largely congruous results, so far as nuclear DNA is concerned, but with some inconsistencies. Judging from our nuclear RFLP study, based on 183 accessions of five species, O. rufipogon, which contains a large amount of genetic variation, is allied with O. glumaepatula and O. barthii (Fig. 1). Within O. rufipogon, geographical differentiation into South Asia, Southeast Asia, China and Oceania was recognized. Regarding genetic difference between perennial and annual ecotypes in O. rufipogon, DNA sequencing analysis of three introns in the phytochrome gene revealed that two ecotypes growing in close proximity tended to show similarity with each other and are clearly differentiated from the accessions collected in remote regions (Barbier et al. 1991). Nucleotide diversity study in Adh1 and Adh2 regions of O. rufipogon was conducted to elucidate the maintenance mechanism of DNA variation (Yoshida and Miyashita 2005). Their study showed that level of polymorphism in these regions, particularly in replacement sites possibly responsible for tolerance to anaerobiosis, was quite low, indicating purifying selection. The phylogenetic relationship between O. glumaepatula and other Old World species is somewhat problematic. Several studies based on nuclear genome DNA, including our own study (Fig. 1), have revealed its close relation with O. barthii (Wang et al. 1992; Doi et al. 2000; Cheng et al. 2002). Regarding variation within O. glumaepatula, we did not find significant differentiation in nuclear genome level, but a complicated geographical pattern was detected in organellar DNA. Populations distributed in Central America and northern South America share common band profiles with O. longistaminata, whereas Amazonian populations

264

H.-W. Cai et al.

O. glumaepatula O. rufipogon O. meridionalis O. longistaminata O. barthii

South Asia China Southeast Asia Oceania

Fig. 1 Diagramatic representation of the phylogenetic tree of AA-genome species estimated according to nuclear DNA RFLP. The size of each species (or sub-group) box roughly represents the number of accessions examined

showed similarity with O. barthii. It is assumed that founders of O. glumaepatula with different organellar components migrated from Africa, presumably several times, and introgressed with each other during the process of eco-geographic differentiation. Admixture of O. glumaepatula and O. rufipogon in Latin America, which is often argued, is probably due to introgression of the Asian AA gene pool by some unknown routes. O. meridionalis is considered to have diverged from others in the early stage of its evolution. Genetic variation within species is relatively small. Geographical isolation of Oceania might have hampered gene flow with other members. Its inbreeding habit associated with an annual life cycle might have helped to reinforce genetic uniformity and remoteness from others. Gene flow between O. meridionalis and O. rufipogon is suspected and remains to be further investigated. The situation of annual and perennial species in Africa contrasts with that of annual and perennial ecotypes in Asia, in the sense that O. barthii and O. longistaminata are inter-sterile and sometimes co-occur in the same habitat, while two ecotypes of O. rufipogon are inter-fertile and allopatric. Many studies have consistently indicated that distinct genetic differences exist between O. barthii and O. longistaminata (Fig. 1). Most probably, speciation of these two African species from the AA-genome gene pool occurred independently at different times (Zhu and Ge 2005). An AFLP study of O. longistaminata indicated that genetic diversity is

III.4 Genetic Diversity in Wild Relatives of Rice

265

larger within than between populations, and larger between populations within countries than among countries (Kiambi et al. 2005). This suggests that highly outbreeding O. longistaminata evolved as metapopulations that were connected by gene flow.

3 3.1

Differentiation Within O. rufipogon Life-history Strategy and Mating System

Populations of O. rufipogon exhibit a wide variation in life history, showing a tendency toward differentiation into perennial and annual ecotypes. Component traits related to life history are intercorrelated with each other (Sano and Morishima 1982). Particular associations among alternative states of each trait (as mentioned in Sect. 1.1 of this chapter) underlie the adaptive strategies of perennial and annual ecotypes growing in contrasting habitat environments, respectively. O. rufipogon varies also in mating system within species. As reviewed by Oka (1988), outcrossing rates estimated by various methods ranged from 0.10–0.30 for annuals and from 0.30–0.55 for perennials. Adaptive floral structure and function are differentiated; the outbreeders have a larger anther and stigma and longer time interval from flowering to pollen emission than the inbreeders. Inbreeding depression, acting as an opposing factor to the autonomous increase of selfing genes through transmission advantage of alleles, may be the main selective force behind the evolution of mating system. Ritland (1990) estimated inbreeding depression in Mimulus species using the change in inbreeding coefficient (F) during the life-cycle stage. This method was applied to our allozyme-based study, in which F values obtained from seed population and from field population in the same site were assumed to represent the genetic structure before and after selection against selfed progeny, respectively. Inbreeding depressions thus estimated were 0.28–0.09 for predominantly selfing annual populations and 0.42–0.68 for mixed-mating perennial populations. This supports the idea that prolonged selfing decreases inbreeding depression by purging deleterious genes. It is a general trend in plants that annuals have lower outcrossing rates than perennials. Among O. rufpogon accessions, reproductive allocation (the percentage of seed weight to total dry weight) that signifies annual habit is negatively correlated with outcrossing rate (r = −0.71). This association between longevity and mating system has an important consequence for genetic population structure. The perennial populations, which usually propagate by clone but occasionally by seeds, persist in stable habitats and accumulate genetic diversity within populations through outcrossing. By contrast, annual populations follow repeated extinction/recolonization events under habitat disturbance, and tend to have small intra-population and large inter-population genetic diversity through inbreeding followed by selection (Barbier 1989). The genetic basis of

266

H.-W. Cai et al.

this adaptive differentiation is not fully elucidated, though some fragmental approaches have been attempted (Morishima 1991; Kohn et al. 1997). The annual type most probably differentiated from the perennial type in order to adapt to drier and/or disturbed habitats. In fact, geographical distribution of the annual type is limited to tropical continental Asia in which a severe dry season prevails, while the perennial type is distributed in broader areas in Asia and Oceania. In the areas where both types are grown, the two types are allopatric, showing their differential habitat preference. Since no reproductive barriers exist among perennial and annual ecotypes and O. sativa, when encountered they freely hybridize among each other, deriving intermediate or weedy types. It was observed that perennial and annual types occupied the deep center and shallow fringe of a small pond, respectively (Morishima et al. 1984). This fact implies that two inter-fertile ecotypes can be isolated by a strong disruptive selection. Many natural populations of O. rufipogon are now exposed to the risk of extinction by habitat destruction as well as by gene flow from cultivars. In-situ conservation of natural populations is an urgent issue in addition to ex-situ conservation. One Thai population that is becoming extinct (Akimoto et al. 1999) and one Chinese population that is well isolated from cultivated rice (Gao et al. 2001) have been extensively investigated regarding intra-population structure in relation to their habitat environments.

3.2

Phylogeographic Differentiation

In contrast to phenotypic traits exhibiting ecotype differentiation, molecular variation studies in O. rufipogon carried out by various researchers have consistently revealed geographical differentiation. Our isozyme study, shown in Fig. 2, suggests that expansion of geographical distribution of O. rufipogon led to genetic divergence that is a joint product of adaptive and non-adaptive processes. It seems that genetic differentiation proceeded around South Asia towards Southeast Asia, China and western India. The Oceanian group that is not included in this figure represents another direction of differentiation, as shown in Fig. 1. Populations of O. rufipogon in China were not well known, but recently surveyed extensively using many molecular markers, and large genetic divergence among populations was reported (Sun et al. 2001; Zhou et al. 2003). Populations distributed on the west coast of India, i.e. the western fringe of the distribution of O. rufipogon, consisting of annual to perennial types, are a particularly interesting group, in that they show some similarity with O. barthii in allozyme (Lolo and Second 1988). Geographical patterns of genetic variation have been shaped by the demographic history of populations (neutral process) as well as by selective forces for regional and ecological adaptation. Effects of population history act across the entire genome, while effects of selection are specific to loci or chromosomal region. Based on this premise, distribution of Fst (parameter for inter-population differentiation) of individual loci obtained from a multi-locus survey on multiple populations is effectively

III.4 Genetic Diversity in Wild Relatives of Rice

267

used in various organisms to identify candidate loci targeted by selection (Beaumont 2005). In our polymorphism survey on seven populations of O. rufipogon collected from various sites in China and South/Southeast Asia, a remarkable heterogeneity of Fst values was found among individual loci (Cai et al. 2004). Figure 3 shows the relationship between Fst and intra-population genetic diversity (Hs) of individual gene loci. Among 46 isozyme and RFLP loci examined, some ten distributed in the upper left in the figure with relatively high Fst and small Hs appear to deviate from the distribution zone estimated from the neutrality model and could be candidates of selected regions for local adaptation. In fact, those loci show significant difference in allele frequencies between China and other areas. On the other hand, some loci in the lower left of the figure with small Fst as well as small Hs are likely to represent the genomic regions, which involve the targets of purifying selection. An isozyme locus Est10 is involved in this group, at which one specific allele among six alleles is fixed in most wild populations, though highly polymorphic among different cultivar groups (Cai et al. 2003). Probably geographical variation of O. rufipogon can be largely explained by the “isolation by distance model” of neutral genes, but some loci could be candidate targets of divergent or purifying selection. To confirm evidence of selection, a more precise genomic scan and a functional study of candidate genes are desired. It should be noted that biologically significant variation revealed by molecular markers always explains only a part of the entire variation involved in this species. For example, in Fig. 2 the first two variation axes of PCA (principal component

China Southeast Asia South Asia Western India

Fig. 2 Geographical variation of Asian strains of O. rufipogon. Strains are scattered by the first (Y axis) and second (X axis) component scores of PCA obtained from polymorphism data of 29 isozyme loci

268

H.-W. Cai et al. 1 Isozyme RFLP

0.9 0.8

FST

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

Hs Fig. 3 Relationships between FST and HS of 46 isozyme and RFLP loci computed from polymorphism data of seven populations of O. rufipogon (original data are in Cai et al. 2004)

analysis) extracted only 16% of the total variation. This contrasts with quantitative trait variation, in which PCA extracted more than 60% of the total variation by the first two axes. AMOVA (analysis of molecular variance) applied to allozyme data obtained from a number of Asian accessions showed that among-region variance component relative to the total variation was 13.6% and between two ecotypes was 10.6%. The extent of non-random association between alleles at different isozyme loci (gametic disequilibrium) was much lower in O. rufipogon (R2 = 0.022) than in O. sativa (R2 = 0.322) (Cai et al. 1995). Furthermore, association between nuclear genome type and organellar genome type in the context of Japonica and Indica is also weaker in O. rufipogon than in O. sativa (Sun et al. 2002). These facts all imply that genetic variation in O. rufipogon at the molecular level is not so clearly structured as in O. sativa, in which a major part of the variation can be attributed to Japonica vs. Indica differentiation.

4 4.1

Genetics of Domestication The Genetic Basis of Domestication-related Traits

As with many other cereal crops, domestication produced clear differences in O. sativa compared with its wild progenitor, O. rufipogon, namely reduced seed shattering, loss of seed dormancy, synchronous growth, and several morphological

III.4 Genetic Diversity in Wild Relatives of Rice

269

features related to cultivation practice and seed yield. In rice, life history and mating system also changed along with the domestication process; perennial to annual habit and allogamy to autogamy.

4.1.1

Allelic Diversity and Selection in Domestication Genes

Phenotypic diversity was dramatically increased in domesticates. On the other hand, investigation of DNA polymorphisms has revealed less genetic variation in domesticates than in their wild progenitors, presumably due to the joint effect of population bottleneck and strong selection during and after domestication. Molecular dissection of several domestication-related genes, which were identified by early classical genetic studies, provided us the clues concerning the level and pattern of allelic diversity and selection by man. For instance, the Waxy gene that controls divergent starch qualities was extensively studied by several researchers. Based on the sequence analysis of a 500-kb genome region centered on the Waxy gene, Olsen et al. (2006) found that strong selection (selection sweep > 250 kb) for the Wx splice donor mutation had a profound impact on this genomic region. Sweeney et al. (2006) cloned the red pericarp gene (Rc) and found that there is a 14-bp deletion in the recessive white allele compared with the dominant Rc allele. Current QTL mapping has revealed that a relatively small number of genes are responsible for the important trait variations such as seed shattering. Li et al. (2006a) cloned the major shattering gene sh4 from O. rufipogon, which accounted for 69% of the variance for shattering in a cross with O. sativa. They concluded that a single amino acid change is principally responsible for the loss of shattering. Konishi et al. (2006) cloned another shattering QTL (qSH1) from cultivated rice, the results showing that a single nucleotide change in cultivated rice destroys a cis-regulatory element required for the expression of qSH. All the above findings indicate that modest genetic changes in single genes could induce dramatic changes in phenotype during and after domestication.

4.1.2

Clusters of Domestication-related QTLs

Many domestication-related traits are of quantitative nature. QTL mapping conducted in various crop species, including rice, consistently identified clusters of genes representing “domestication syndrome”, a phrase coined by Harlan (1975) (rice; Xiong et al. 1999; Cai and Morishima 2002; Lee et al. 2005; Li et al. 2006b). Using the RILs population derived from a cross between O. sativa (Indica cultivar) and O. rufipogon (Chinese perennial strain), we analyzed QTLs for 37 traits. QTLs for seed dormancy, seed shattering, anther length and awn length tended to be concentrated on particular regions of chromosomes 1, 3, 6, 8, 9, 11 and 12 (Cai and Morishima 2002). Figure 4 illustrates our linkage maps of four chromosomes,

270

H.-W. Cai et al. Chr.1

R1613

(1) (2) (3) (4) ANL

CDO328 G165b

P/T BVP

KCL

SHD ANL

G270 R2635 G302 CDO470 RG345

AWL

SHD* DOR

(1) (2) (3) (4)

CDO20 DOR RZ329 BCD454a G144

DOR AWL

C746

[ KCL MSL

APH GMS DOR

SWD

BCD454

BCD828

ANL

SHD*

SHD*

SLW

GMS

DOR*

DOR

G249

CDO118 RB1 G393 Est10 OP

Chr.3 C25

KCL GMS ANL PTB BVP SHD

C12 G232 Pgi1 SHD*

gf CDO345

[

R1927 CDO122

DOR ANL

APH

PFR DOR APH

[ DOR KCL

AWL MSL

ANL

ANL GMS MSL

ANL DOR

SHD

BCD450 Mal3 AWL

RZ393

GMS MSL

DOR*

G370 G54

C112

Chr.6 G30

(1) (2) (3) (4) G200 Est2 Pgi2

Chr.8 (1) (2) (3) (4)

RZ143

Cat1 Amp3 Amp7

KCL GMS

APH DOR [SFR

APH

C400a RZ588

C400b G278

R2171 RZ144 RZ144H

G104 R1394A

SHD SWD

G2132 G187 RG181 Amp2

[ SHD

G1314c

DOR AWL

R2549 G12

LTR

DOR [ APH

KCL GMS

AWL AWL DOR SHD SHD

DOR [ANL

ANL SHD

GMS MSL AWL SFR KCL

SHD*

DOR

G122 G1314a

PTB

SFR

AWL DOR

NT

DOR ANL

RGF DTH BVP

G1314b

Fig. 4 Linkage maps of chromosomes 1, 3, 6 and 8 and locations of QTLs (Cai and Morishima 2002). QTLs of four domestication-related traits are shown in bold (DOR seed dormancy; SHD seed shattering; AWL awn length; ANL anther length), and compared with those detected in other studies: (1) Xiong et al. 1999; (2) Gu et al. 2005; (3) Li et al. 2006b; (4) Lee et al. 2005. Asterisks indicate the region in which QTLs of yield-related morpholgy are clustered

together with QTLs of the above-mentioned four traits reported by other researchers, though precise correspondence of chromosomal locations are difficult due to a lack of common markers. As shown in Fig. 4, irrespective of ecotype or variety of parental lines, domestication QTLs tend to cluster on specific common regions of several

III.4 Genetic Diversity in Wild Relatives of Rice

271

chromosomes. QTLs for yield-related morphology identified by other studies, which were not analyzed in our cross, tended to join some of our clusters (marked by asterisks). To explain linkages of QTLs for co-adapted but developmentally unrelated traits such as domestication syndrome, several alternative inferences are possible. It may be more plausible to assume that “multifactorial linkages” (Grant 1981) could be inherited from the wild progenitor than to assume unlinked genes were brought together during the domestication process by some unknown mechanisms. Probably, favorable mutations that occurred within the linked block were accumulated to form a final domestication syndrome, through unconscious and conscious selection by man. Closely linked coadapted genes are not limited to domesticates, but are predominant in well-adapted natural populations (Grant 1981). In our mapping study, QTLs for several Indica–Japonica diagnostic traits were unexpectedly found to cluster near domestication-related blocks. This may reflect an adaptive gene block in the wild parent of this cross (Chinese O. rufipogon), which lives near the northern periphery of the geographical distribution of this species and carries Japonica-specific characters and alleles. Further, we found evidence for apparent pleiotropy of several traits in this cross (Cai and Morishima 2002).

4.2

Progenitors of Japonica and Indica Cultivars

Japonica and Indica types are widely recognized as two major variety groups (or subspecies) in Asian rice. In addition, intermediate or atypical minor variety groups are known. Clear genetic differences were detected between the typical Japonica and Indica types at various levels ranging from phenotypic characters (Oka 1988) to molecular variation in nuclear as well as organellar genome (e.g. Wang and Tanksley 1989; Zhang et al. 1992; Garris et al. 2005). Tendency to non-random association between characters or between alleles of different loci (gametic disequilibrium) in O. sativa reflects mainly this differentiation. There has been a long debate on the origin of Japonica and Indica types. The core issue to be discussed was whether Japonica–Indica differentiation existed in the pre-domestication era. As mentioned in the previous section, genetic variation in O. rufipogon is not so strongly structured, but molecular phylogenetic trees constructed based on combined data of wild and cultivated rice consistently revealed the existence of wild accessions allied with Japonica and Indica cultivars, respectively (Doi et al. 2000; Cheng et al. 2003; Londo et al. 2006). These two accession groups of O. rufipogon consist of collections from mainly China and South/ Southeast Asia, respectively. Further, comparative analysis of genome sequence of typical Japonica and Indica cultivars brought us an estimate of 0.2–0.4 my as their divergence time (e.g. Zhu and Ge 2005), which largely predates rice domestication. Thus the diphyletic hypothesis that Japonica and Indica types have originated in China and in tropical Asia from different wild ancestors, respectively, became currently accepted.

272

H.-W. Cai et al.

According to the historical records, wild rice in China, which was more widely distributed further north in ancient days than at present, became gradually extinct due to development of intensive agriculture or climatic change, and receded south (You 1986). Therefore, incipient domesticates in China probably separated from their wild ancestors relatively early. Presumably, primitive Japonica, which spread mainly to high latitude and highland in lower latitude where wild rice did not grow, tended to retain its original genetic identity after the domestication bottleneck as a monophyletic clade. On the other hand, Indica cultivars diverged into various land races. Their large genetic diversity might be caused partly by gene flow from neighboring O. rufipogon undergoing a “hybridization and differentiation cycle”, as proposed by Harlan (1975). Another problem is whether the ancestor of O. sativa was a perennial or annual type. Japonica-related Chinese wild populations are clonal plants, but seed propagation also occurs when they encounter drought or habitat disturbance (Xie et al. 2001). Previously, we postulated that perennial–annual intermediate populations are most probably the ancestors of cultivated rice because they are able to accumulate genetic variability within populations and release a variety of variation when seed propagation is imposed (Sano et al. 1980). In contrast, most of the extant Indica-related accessions of O. rufipogon are annual or intermediate types, though whether or not they are truly direct ancestors remains unsolved. Indicas certainly inherited some degree of annual habit from wild ancestors directly or through gene flow. It might be true that ancestral polymorphism corresponding to “Japonica” and “Indica” types existed before rice domestication. Tang et al. (2006), based on the extensive search of highly polymorphic regions in rice genome, found evidence indicating that two distinctly divergent haplotypes in strong linkage disequilibrium characterizing Japonica and Indica types, respectively, are widely distributed in other AA-genome species in addition to O. rufipogon. While the allelic diversity pattern recently found in a shattering locus sh4 posed an interesting problem, six single nucleotide polymorphisms including one functional SNP located within this locus are completely identical among all O. sativa tested, including both Japonica and Indica cultivars, though O. rufipogon populations are polymorphic (Li et al. 2006a). This evidence suggests that the problem of origin of Japonica and Indica cultivars is still an open question. How ancestral polymorphism and the initial birth of domesticated rice were interrelated may not be so simple a story. Regarding the origins of several minor variety groups, the possibility cannot be ruled out that they were independently derived from the diverse gene pool of O. rufipogon. For instance, a particular variety group of deep-water rice grown in a certain area of Bangladesh (Rayada and some Aman varieties) has unique genotypes carrying specific alleles that are rarely found in cultivars from other regions but are common in O. rufipogon (Cai and Morishima 2000). In conclusion, the incipient use of Asian rice by man is most probably a set of multiple events that occurred in different times and in different places. It is not an easy task to pinpoint the time and place of the initial birth of cultivated rice. When phylogeography of many domestication genes and life-history evolution of

III.4 Genetic Diversity in Wild Relatives of Rice

273

O. rufipogon are more deeply elucidated and linked together, the scenario of rice domestication in Asia will be more clearly illuminated. Lastly, we would like to add a cautionary note. Many of the “wild relatives of rice” existing at present in their natural habitats are more or less hybrid derivatives of wild types and domesticates. This is not a recent phenomenon but has continually occurred since the initial appearance of rice. Even gene-bank stocks are not free from natural hybridization before and after collection. In addition, genetic erosion is inevitable during seed multiplication for preservation and lowers the quality of highly polymorphic wild rice populations. Such understanding is essential for wild rice scientists.

References Akimoto M, Shimamoto Y, Morishima H (1998a) Comparison between phenotype variation and isozyme diversity within and between AA genome wild rice species. Rice Genet Newsl 15:78–80 Akimoto M, Shimamoto Y, Morishima H (1998b) Genetic population structure of wild rice Oryza glumaepatula distributed in the Amazon flood area influenced by its life-history traits. Mol Ecol 7:1371–1381 Akimoto M, Shimamoto Y, Morishima H (1999) The extinction of genetic resources of Asian wild rice, Oryza rufipogon Griff.: a case study in Thailand. Genet Res Crop Evol 46:419–425 Barbier P (1989) Genetic variation and ecotypic differentiation in the wild rice species Oryza rufipogon. II Influence of the mating system and life-history traits on the genetic structure of populations. Jpn J Genet 64:273–285 Barbier P, Morishima H, Ishihama A (1991) Phylogenetic relationships of annual and perennial wild rice: probing by direct DNA sequencing. Theor Appl Genet 81:693–702 Beaumont MA (2005) Adaptation and speciation: what can Fst tell us? Trends Ecol Evol 20:435–440 Cai HW, Morishima H (2000) Diversity of rice varieties and cropping system in Bangladesh deepwater areas. Jpn Aric Res Q 34:225–231 Cai HW, Morishima H (2002) QTL clusters reflect character associations in wild and cultivated rice. Theor Appl Genet 104:1217–1228 Cai HW, Wang XK, Morishima H (1995) Isozyme variation in Asian common wild rice, Oryza rufipogon. Rice Genet Newsl 12:178–180 Cai HW, Zhang Y, Xu J, Zhu L, Cheng K, Wang X (2003) Polymorphism at the esterase isozyme locus Est10 associated with phylogenetic differentiation in rice. Genes Genet Syst 78:285–290 Cai HW, Wang XK, Morishima H (2004) Comparison of population genetic structures of common wild rice (Oryza rufipogon Griff.), as revealed by analyses of quantitative traits, allozymes, and RFLPs. Heredity 92:409–417 Cheng C, Tsuchimoto S, Ohtsubo H, Ohtsubo E (2002) Evolutionary relationships among rice species with AA genome based on SINE insertion analysis. Genes Genet Syst 77:323–334 Cheng C, Motohashi R, Tsuchimoto S, Fukuta Y, Ohtsubo H, Ohtsubo E (2003) Polyphyletic origin of cultivated rice: based on the interspersion pattern of SINEs. Mol Evol Biol 20:67–75 Dally AM, Second G (1990) Chloroplast DNA diversity in wild and cultivated species of rice (Genus Oryza, section Oryza). Cladistic-mutation and genetic distance analysis. Theor Appl Genet 80:209–222 Doi K, Nakano M, Yoshimura A, Iwata N, Vaughan DA (2000) RFLP relationships of A-genome species in the genus Oryza. J Fac Agr Kyushu Univ 45:83–98

274

H.-W. Cai et al.

Gao LZ, Wei CL, Yang QJ, Hong DY, Ge S (2001) Intra-population genetic structure of Oryza rufipogon Griff. in Yunnan, China. J Plant Res 114:107–113 Garris AJ, Tai TH, Coburn J, Kresovich S, McCouch S (2005) Genetic structure and diversity in Oryza sativa L. Genetics 169:1631–1638 Grant V (1981) Plant speciation. Columbia University Press, New York Gu XY, Kianian SF, Foley ME (2005) Phenotypic selection for dormancy introduced a set of adaptive haplotypes from weedy into cultivated rice. Genetics 171:695–704 Harlan JR (1975) Crops and man. American Society of Agronomy, Crop Science Society of America, Madison, Wisconsin Kanazawa A, Akimoto M, Morishima H, Shimamoto Y (2000) Inter- and intra-specific distribution of Stowaway transposable elements in AA genome species of wild rice. Theor Appl Genet 101:327–335 Kiambi DK, Newbury HJ, Ford-Lloyd B, Dawson I (2005) Contrasting genetic diversity among Oryza longistaminata (A. Chev. et Roehr) populations from different geographic origins using AFLP. Afr J Biotech 4:308–317 Kohn JR, Leyva N, Dossey R, Sobral B, Morishima H (1997) Quantitative trait locus analysis of trait variation among annual and perennial ecotypes of Oryza rufipogon. Int Rice Res Notes 22(2):4–5 Konishi S, Izawa T, Lin SY, et al. (2006) An SNP caused loss of seed shattering during rice domestication. Science 312:1392–1396 Lee SJ, Oh CS, Suh JP, McCouch SR, Ahn SN (2005) Identification of QTLs for domesticationrelated and agronomic traits in an Oryza sativa × O. rufipogon BC1F7 population. Plant Breed 124:209–219 Li CG, Zhou AL, Sang T (2006a) Rice domestication by reducing shattering. Science 311:1936–1939 Li CG, Zhou AL, Sang T (2006b) Genetic analysis of rice domestication syndrome with the wild annual species, Oryza nivara. New Phytologist 170:185–194 Lolo OM, Second G (1988) Peculiar genetic characteristics of O. rufipogon from western India. Rice Genet Newsl 5:67–70 Londo JP, Chiang YC, Hung KH, Chiang TY, Schaal BA (2006) Phylogeography of Asian wild rice, Oryza rufipogon, reveals multiple independent domestications of cultivated rice, Oryza sativa. Proc Natl Acad Sci USA 103:9578–9583 Morishima H (1991) Association between Pox-1 variation and seed productivity potential in wild rice. In: Rice Genetics II. International Rice Research Institute, Manila, pp1–9 Morishima H, Sano Y, Oka HI (1984) Differentiation of perennial and annual types due to habitat conditions in the wild rice Oryza perennis. Plant Syst Evol 144:119–135 Morishima H, Sano Y, Oka HI (1992) Evolutionary studies in cultivated rice and its wild relatives. Oxford Surv Evol Biol 8:135–184 Oka HI (1988) Origin of cultivated rice. Japan Scientific Societies Press, Elsevier, Tokyo Olsen KM, Caicedo AL, Polato N, McClung A, McCouch S, Purugganan MD (2006) Selection under domestication: evidence for a sweep in the rice Waxy genomic region. Genetics 173:975–983 Ritland K (1990) Inference about inbreeding depression based on changes of the inbreeding coefficient. Evolution 44:1130–1240 SanoY, Morishima H (1982) Variation in resource allocation and adaptive strategy of a wild rice, Oryza perennis Moench. Bot Gaz 143:518–523 Sano Y, Morishima H, Oka HI (1980) Intermediate perennial-annual populations of Oryza perennis found in Thailand and their evolutionary significance. Bot Mag Tokyo 93:291–305 Second G (1985) Evolutionary relationships in the Sativa group of Oryza based on isozyme data. Genet Sel Evol 17:89–114 Sun CQ, Wang XK, Li ZC, Yoshimura A (2001) Comparison of the genetic diversity of common wild rice (Oryza rufipogon Griff.) and cultivated rice (O. sativa L.) using RFLP markers. Theor Appl Genet 102:157–162

III.4 Genetic Diversity in Wild Relatives of Rice

275

Sun CQ, Wang XK, Yoshimura A, Doi K (2002) Genetic differentiation for nuclear, mitochondorial and chloroplast genomes in common wild rice (Oryza rufipogon Griff.) and cultivated rice (O. sativa L.). Theor Appl Genet 104:1335–1345 Sweeney MT, Thomson MJ, Pfeil BE, McCouch S (2006) Caught red-handed: Rc encodes a basic helix-loop-helix protein conditioning red pericarp in rice. Plant Cell 18:183–294 Tang T, Lu J, Huang J, et al. (2006) Genomic variation in rice: genesis of highly polymorphic linkage blocks during domestication. PLOS Genetics 2(11) e199:1824–1833 Wang ZY, Tanksley SD (1989) Restriction fragment length polymorphism in Oryza sativa L. Genome 32:1113–1118 Wang ZY, Second G, Tanksley SD (1992) Polymorphism and phylogenetic relationships among species in the genus Oryza as determined by analysis of nuclear RFLPs. Theor Appl Genet 83:565–581 Xie ZW, Lu YQ, Ge S, Hong DY, Li FZ (2001) Clonality in wild rice (Oryza rufipogon, Poaceae) and its implications for conservation management. Am J Bot 88:1058–1064 Xiong LZ, Liu KD, Dai XK, XU CG, Zhang Q (1999) Identification of genetic factors controlling domestication-related traits of rice using an F2 population of a cross between Oryza sativa and O. rufipogon. Theor Appl Genet 98:243–251 Yoshida K, Miyashita NT (2005) Nucleotide polymorphism in the Adh2 region of the wild rice Oryza rufipogon. (2005) Theor Appl Genet 111:1215–1228 You XL (1986) Origin of rice culture around Lake Taihu, its spread and development. Chinese Agric Hist 1:8–19 (in Chinese) Zhang QF, Saghai Maroof MA, Lu TY, Shen,BZ (1992) Genetic diversity and differentiation of indica and japonica rice detected by RFLP analysis. Theor Appl Genet 83:495–499 Zhou HF, Xie ZW, Ge S (2003) Microsatellite analysis of genetic diversity and population genetic structure of a wild rice (Oryza rufipogon Griff.) in China. Theor Appl Genet 107:332–339 Zhu Q, Ge S (2005) Phylogenetic relationships among A-genome species of the genus Oryza revealed by intron sequences of four nuclear genes. New Phytol 167:249–265

III.5

Rice Retroposon, p-SINE, and Its Use for Classification and Identification of Oryza Species Hisako Ohtsubo(* ü )1, Suguru Tsuchimoto1, Jian-Hong Xu1, Chaoyang 1 Cheng , Marcia Y. Koudo1, Nori Kurata2, and Eiichi Ohtsubo1

1

Introduction

Short interspersed elements (SINEs) are retroelements found in a wide variety of eukaryote genomes. They are 70- to 500-bp repetitive DNA sequences and each has an internal promoter for RNA polymerase III, an A-rich or T-rich tail of various size at the 3' end, flanked by the direct repeat DNA sequences of several nucleotides, indicating that they have proliferated via retrotransposition. Many SINEs appear to be related to tRNAs, whereas a few, such as the primate Alu and rodent B1 family elements, are related to 7SL RNA, and zebrafish SINE3 is related to 5S rRNA. SINEs have been used as markers for phylogenetic studies owing to their specific characters; once inserted, SINE remains in that locus and have never been excised, and the probability of independent retroposition at the same site on the chromosome is virtually zero (Batzer and Deininger 1991; for a review, see Shedlock and Okada 2000). In particular, the presence or absence of a SINE inserted at a locus is easy to assay by PCR, and thus SINEs have found increasing use as phylogenetic markers to study relationships among the species of humans (Batzer et al. 1994), primates (Bailey and Shen 1993; Hamdi et al. 1999), whales and even-toed ungulates (Shimamura et al. 1997; Nikaido et al. 1999), salmonid fish and cichlids (Murata et al. 1993; Takahashi et al. 1998), old world pond turtles (Sasaki et al. 2006), and two kinds of plants, Oryza species (rice) and Brassicaceae species (Mochizuki et al. 1993; Deragon et al. 1994; Motohashi et al. 1997; Tatout et al. 1999; Deragon and Zhang 2006). In this chapter, we summarize our work on rice SINE, p-SINE, and discuss such use for the classification and identification of varieties of Oryza species.

1

Institute of Molecular and Cellular Biosciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-0032, Japan e-mail: [email protected]; [email protected]; [email protected]

2

National Institute of Genetics, Mishima, Shizuoka 411-8540, Japan e-mail: [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

277

278

2 2.1

H. Ohtsubo et al.

p-SINE Families in the Oryza Genus Identification and Distribution of p-SINE1 Members Among Various Rice Strains with the AA Genome

The rice genus, Oryza, consists of 22 species, distributed in Asia, Africa, Australia and Central and South America. Among them, two are cultivated species and the rest are wild ones. They are classified into six diploid genome types (AA, BB, CC, EE, FF and GG) and four tetraploid genome types (BBCC, CCDD, HHJJ and HHKK), based on interspecific crossing, subsequent cytogenetic analysis and genomic DNA hybridization (Khush 1997; Ge et al. 1999). Two cultivated rice species, Oryza sativa and Oryza glaberrima, and five wild species belong to the AA genome type. p-SINE1 is the first plant SINE (Short INterspersed Element) identified in the Waxy gene in O. sativa. It is a short insertion element of 122 to 125 bp in length, and is dispersed on the rice genome with high copy numbers. p-SINE1 is estimated to be present in 6500 and 3000 copies per haploid genome of O. sativa and O. glaberrima, respectively (Motohashi et al. 1997). At the beginning of this research, many p-SINE1 members were isolated from the O. sativa genome by inverse PCR or genomic library screening (Mochizuki et al. 1992; Motohashi et al. 1997). Most of the members are present in all the species with the AA genome, while some show “interspecific insertion polymorphism” among the strains of species with the AA genome. These p-SINE1 members with insertion polymorphism might have undergone retrotransposition into the respective loci during the divergence of the rice species with the AA genome (Mochizuki et al. 1992; Hirano et al. 1994). Thus, we found they were suitable for classifying rice species, as well as for inferring their phylogenetic relationships among Oryza species (Cheng et al. 2003; Ohtsubo et al. 2004; Xu et al. 2007).

2.2

p-SINE2 and p-SINE3, the Related Elements of p-SINE1

During the course of the identification and characterization of many p-SINE1 members, we found other p-SINE elements, named p-SINE2 and p-SINE3, from rice. They were screened among rice strains of species with different genome types (Xu et al. 2005). Consensus sequences of each p-SINE1 family revealed that their 5'-end regions with the RNA polymerase III promoter show significant homology with the 5'-end region of p-SINE1, but not with the 3'-end region of p-SINE1 (Fig. 1A). Despite the three elements sharing minimal homology in their 3'-end region, the deduced RNA secondary structure of these elements was similar, suggesting the structure may have an important role in p-SINE retroposition. These findings suggest that the three p-SINE elements originated from a common ancestor. The p-SINE2 members were present at respective loci not only in the strains of the species with the AA genome, but also in those of other species with the BB, CC, DD or EE

This page intentionally blank

III.5 Rice Retroposon, p-SINE, and Its Use for Classification and Identification

279

Fig. 1 Comparison of consensus sequences of p-SINE elements. A Schematic structures of p-SINE1, p-SINE2 and p-SINE3. Alignments of consensus sequences of these elements are also shown below. B Consensus sequences of the p-SINE1 family: RA subfamily, RAα group and RAβ group members, respectively. Bars denote identical nucleotides to those in the p-SINE1 family consensus sequence

genome. The p-SINE3 members were, however, only present in strains of species with the AA genome. These findings suggest that p-SINE2 has originated in an ancestral species with the AA, BB, CC, DD and EE genomes, like p-SINE1, whereas p-SINE3 has originated in an ancestral strain of the species with the AA genome. Indeed, some p-SINE2 members were found to be right markers for the classification of non-AA types of wild rice species (Xu 2004). The p-SINE1 members are more diverged in the nucleotide sequence than p-SINE2 or p-SINE3, indicating that p-SINE1 is older than p-SINE2 and p-SINE3.

2.3

The RA (Recently Amplified) Subfamily Members of p-SINE1 and Their Diagnostic Three Mutations

By the end of 2001, just before completing the rice genome sequencing projects of O. sativa Nipponbare (japonica) and 93-11 (indica), we had collected 47 p-SINE1 members at different loci on the chromosomes of O. sativa japonica and indica strains, by using shot-gun cloning, IPCR (inverse PCR) and ADL-PCR (adaptorligation PCR) (Motohashi et al. 1997; Cheng et al. 2003). Species-specific p-SINE1 members were also isolated from other wild rice species with the AA genome, and some p-SINE1 members were found to be polymorphic among the rice strains of O. sativa and its close relative, O. rufipogon. Alignment of the nucleotide sequences of the polymorphic p-SINE1 members showed that they share three diagnostic substitution mutations (T at nucleotide position 7, A at 63, and A at 114) (Fig. 1B). These three mutations were also found in many p-SINE1 transcripts identified by 5' RACE-PCR from O. sativa, a sign of recently amplified SINE members (Tsuchimoto,

280

H. Ohtsubo et al.

unpublished results). These observations strongly suggest that the members with these three diagnostic mutations have been amplified recently during rice evolution. We therefore designated them as RA (recently amplified) subfamily members. RA subfamily members show two mutations in the A- and B-box sequences of the internal pol III promoter, and another mutation in the distal end region. The discovery of RA-subfamily members of p-SINE1 elements was an epochmaking milestone for our SINE research, because we could preferentially obtain polymorphic p-SINE1 members by employing ADL-PCR with specific primers reflecting the diagnostic mutations described above (Cheng et al. 2003). To date, 84 RA-subfamily members have been identified from O. sativa. Some members contained three additional common substitution mutations (C at 77, T at 113, and T at 117), and other members contained two additional common substitution mutations (T at 10 and T at 117). These suggest that RA-subfamily members consist of three groups, referred to as RA, RAα and RAß. (Fig. 1B).

3

3.1

Phylogenetic Analysis of Oryza Species Based on SINE Insertion Analysis Phylogenetic Analysis of Rice Species with the AA Genome

Phylogenetic analysis of the AA genome type rice strains was carried out by constructing a data matrix based on the presence or absence of p-SINE1 members at the respective loci in each rice strain. A phylogenetic tree constructed by the neighbor-joining (N-J) method showed that all the strains are divided into five groups (Fig. 2; Cheng et al. 2002). Strains of Oryza longistaminata and Oryza meridionalis are distantly related to those of the other five species, O. rufipogon, Oryza barthii and Oryza glumaepatula. The latter five species are closely related to one another (Fig. 2). Note that, in the cluster of the species O. satva/O. rufipogon in Fig. 2, the strains seem to be further subdivided into several groups, suggesting that the relationship among those strains could be further analyzed by using the right and specific p-SINE1 markers for that population. As described above, we noticed that some p-SINE1 members showed polymorphism among the strains of O. sativa and O. rufipogon, indicating that they have been amplified after the divergence of O. sativa and O. rufipogon. This finding gave us the idea of analyzing the evolutionary relationship among the subspecies belonging to the O. sativa complex, especially focusing on the origin of two cultivated rice subspecies, japonica and indica.

3.2

Phylogenetic Analysis of O. sativa and O. rufipogon Strains

O. rufipogon is the species closest to O. sativa and is generally thought to be its progenitor (Oka 1988; Morishima et al. 1992; Khush 1997). O. sativa and O. rufipogon,

III.5 Rice Retroposon, p-SINE, and Its Use for Classification and Identification

281

Fig. 2 A phylogenetic tree showing relationships among strains with the AA genome. The tree was constructed by the N-J method based on the pattern of presence or absence of p-SINE1 members and three other transposable elements (RILN20, Tnr8 and RIRE11)

however, show high intraspecific variation. Morphologically, O. sativa strains are classified into two subspecies, indica and japonica (Kato et al. 1928). A third subspecies, javanica, is thought to be a tropical component of a single japonica group (Oka 1958). O. rufipogon strains are classified as two major ecotypes, perennial and annual, which differ markedly in life-history traits and habitat preference (Morishima et al. 1992). An intermediate type has been noted for some O. rufipogon strains (Sano et al. 1980). The intraspecific variation of O. sativa has been investigated on RFLP analysis and by isozyme analysis, whereas few extensive molecular investigations have been conducted on variation in O. rufipogon. The origin of O. sativa has been questioned for a long time. Perennial type O. rufipogon is considered to be a progenitor, because the high genetic variability in the perennial population would provide higher evolutionary potential than that in the annual population. Annual type O. rufipogon is also considered to be a possible progenitor of O. sativa, because annual strains have high seed productivity and certain other characters similar to O. sativa. Some O. rufipogon strains have been reported to be capable of evolving into both the indica and japonica types of O. sativa. Biochemical and molecular studies, however, show that the indica and japonica strains are closer to different O. rufipogon strains, respectively, in some characters than they are to each other (Wang et al. 1992; Mochizuki et al. 1993; Hirano et al. 1994), which may support the idea that the two subspecies of O. sativa strains originated diphyletically. We have carried out phylogenetic analyses twice for both O. rufipogon strains and O. sativa strains. The first time, we used a total of 25 p-SINE1 members isolated from a strain of only O. sativa, including a japonica (Nipponbare) strain and

282

H. Ohtsubo et al.

two indica strains (IR36, C5924) (Cheng et al. 2003). The second time, an additional 26 members were isolated from five strains of O. rufipogon representing perennial, annual and intermediate ecotypes (Xu et al. 2007). Of the total of 51 p-SINE1 members, 44 contained three diagnostic substitution mutations, as described in the previous section for the RA subfamily members (Xu et al. 2007). A total of 108 rice strains (68 O. sativa, 35 O. rufipogon and five other rice species with the AA genome) were examined for the presence or absence of p-SINE1 members, by PCR with primer pairs that hybridize to the regions flanking each p-SINE1 member. To determine the relationships of the rice strains examined, they were bar-coded on the basis of the presence or absence of p-SINE1 members at the respective loci, and a phylogenetic tree of the strains was constructed using the bar codes assigned to each of the strains with the N-J method. According to the phylogenetic tree obtained, the strains examined were classified into three groups, namely I, II and III. (Fig. 3; Xu et al. 2007).

3.2.1

Origin of japonica Strains from the Chinese O. rufipogon Perennial Population

Group I of the N-J tree consisted of O. rufipogon perennial strains and O. sativa japonica strains. This result supports the idea that the O. rufipogon perennial strains and O. sativa ssp. japonica strains originated from a common ancestor. Note that Intermediate

Group II

Other species

Perennial

Group III Hypothetical ancestor

Indica strains made by breeders

II

III

Perennial (China)

I

Group I Indica Japonica 0.1

Annual & Indica

Japonica from insular Southeast Asia (Javanica)

Fig. 3 A phylogenetic tree showing relationships among O. sativa and O. rufipogon strains. The tree was constructed on the basis of p-SINE1-insertion polymorphism by the N-J method

III.5 Rice Retroposon, p-SINE, and Its Use for Classification and Identification

283

five of the six O. rufipogon perennial strains are from China. This fact suggests that japonica rice strains originated in the O. rufipogon perennial population in China (Cheng et al. 2003; Xu et al. 2007), if there were no introgressions from cultivated rice into these O. rufipogon strains. Japonica strains in Group I include those from the temperate area of East Asia and those from the tropical area of Southeast Asia. Note that the strains from the tropical area of insular Southeast Asia (such as Indonesia, the Philippines and Malaysia) were clustered together in the phylogenetic tree, forming a branch that is distinct from those of the strains from the temperate area of East Asia and mainland Southeast Asia (Fig. 3). This cluster was observed only when we used 26 additional p-SINE1 members derived from O. rufipogon, but was not noted previously in the first analysis by Cheng et al. (2003). Most of the strains from the tropical area of insular Southeast Asia have been classified as the strains of the third subspecies, javanica (H. Morishima and N. Kurata, personal communication). Consequently, these findings suggest that javanica strains distributed in insular Southeast Asia can be distinguished from the other japonica strains by our SINE method (Xu et al. 2007).

3.2.2

Polyphyletic Origin of O. sativa Strains from O. rufipogon

Group II includes O. sativa indica strains, which are clustered with O. rufipogon annual strains. This shows that O. rufipogon annual strains and O. sativa indica strains are closely related to each other. Of indica strains, the majority formed a branch clearly distinct from that of the annual strains of O. rufipogon, indicating that both types originated from a common ancestor, and this ancestor is most likely O. rufipogon of the perennial type (Xu et al. 2007). In the phylogenetic tree, six indica strains, including IR36, IR24 and Milyang 23, are clustered to form one subgroup shown as indica strains in Fig. 3. Such a subgroup was not observed in the first phylogenetic tree constructed by Cheng et al. (2003), probably because no p-SINE1 members from O. rufipogon were employed in the first phylogenetic analysis of O. sativa and O. rufipogon (Cheng et al. 2003). These indica strains are the progenies, made by breeders in different institutes in different countries through hybridization with various indica strains, as well as a few japonica strains. The SINE insertion analyses suggest that japonica strains are closely related to the perennial types of O. rufipogon strains, and indica strains are closely related to annual types of O. rufipogon strains. This supports the idea that O. sativa originated polyphyletically from O. rufipogon (Cheng et al. 2003; Xu et al. 2007).

3.2.3

The Perennial Ecotype of O. rufipogon Strains Might Be Derived from Intermediate Ecotype Strains

Group III consists of O. rufipogon strains of perennial and intermediate ecotypes (Fig. 3). The hypothetical ancestor with no p-SINE1 insertion at any respective loci

284

H. Ohtsubo et al.

was placed with the strains of group III in the phylogenetic tree. The representative strains of the other species with the AA genome appeared close to the hypothetical ancestor in the phylogenetic tree. Note that O. rufipogon strains of the intermediate ecotype were clustered together, forming a subgroup that is the closest to the hypothetical ancestor among O. rufipogon strains. This cluster was not observed in the previous analysis (Cheng et al. 2003). Morishima et al. (1984) suggested that the intermediate plants are ecologically unstable and when the selective pressure is strong enough, they can be differentiated into perennial and annual types. Therefore it is likely that the ancestral population of O. rufipogon had the intermediate characters between perennial and annual types. The p-SINE1 insertion analysis above showed that O. rufipogon intermediate strains forming a subgroup are closer to the hypothetical ancestor. This leads us to suggest that the perennial and intermediate O. rufipogon strains originated from the common ancestral O. rufipogon population, which may have consisted of intermediate-type strains.

3.3

Some O. rufipogon Perennial Strains Show Mixed Features of the Different Groups by Our Present SINE Method

During the course of the phylogenetic analysis, we noticed that bootstrap values of the three groups were relatively low. The tree without O. rufipogon strains showed high bootstrap values, whereas the values of the tree without O. sativa strains stayed low, suggesting that the low bootstrap values were caused by the presence of O. rufipogon strains. We found that the bootstrap values of the three groups (I, II and III) were relatively high when several perennial-type strains were excluded from the analysis (Xu et al. 2007). These facts suggest that the O. rufipogon perennial-type strains have more or less mixed features of the three groups (I, II and III), which might be partly the result of natural hybridizations between different groups. We also constructed a phylogenetic tree with the unweighted pair group method with arithmetic mean (UPGMA), using the same dataset as the N-J tree, and found that the strains were also classified into three groups, which contain the same members as were observed in the groups I, II and III deduced by the N-J method, except that several O. rufipogon perennial strains of group III were clustered together with those of group II. We also inferred population structure of the strains with the “STRUCTURE” program (Pritchard et al. 2000), and obtained results consistent with those from the N-J method (Xu et al. 2007). The newly appeared subgroups observed when using N-J methods, such as the sub-cluster of javanica strains from insular Southeast Asia, or that of O. rufipogon intermediate strains, were also clustered, respectively, in the UPGMA tree. This supports the idea that the phylogenetic tree obtained by the p-SINE1 insertion polymorphism could explain the relationship among the strains of cultivated rice species, O. sativa, and wild rice species, O. rufipogon, even though some O. rufipogon strains show mixed features.

III.5 Rice Retroposon, p-SINE, and Its Use for Classification and Identification

4

4.1

285

“SINE Code” Is an Excellent Tool for Identification of Oryza Species What Is the “SINE Code”?

Until now, more than 400 p-SINE1 members have been identified from various species with the AA genome as well as the non-AA genome. Extensive PCR analysis with many strains and loci has revealed that some p-SINE1 members were present only in strains of one species, whereas the others were present in strains of two or more species, showing interspecific insertion polymorphism of the members among the rice species. In addition to the “interspecific” insertion polymorphism, we have also identified p-SINE members that show genome-type-specific, subspecies-specific or ecotype-specific insertion polymorphism. Based on the insertion polymorphism of p-SINE1 members described above, we summarize the useful p-SINE members in Fig. 4. They are useful to distinguish the five rice species with the AA genome (Fig. 4A), or the six different genome type strains (Fig. 4B), three different subspecies of O. sativa strains (Fig. 4C), or three different ecotypes of O. rufipogon strains (Fig. 4D). We have developed the system named “SINE code”, that is the (1, 0) code, to classify Oryza strains. The insertion polymorphism of p-SINE1 markers at the several loci is encoded to a numeric sequence of 1 (presence; plus) and 0 (absence; minus) for each strain. The “AA genome code” is a six-digit SINE code that begins with “AA” to categorize the AA genome-type strains into species (Fig. 4A). It stands for the presence or absence of a set of the species-specific six p-SINE1 members, r1, r502, r601, r801, r904 and r705, at the corresponding loci on the rice chromosomes with the AA genome. p-SINE1-r1, present in all the AA genome type species, was added as a kind of positive control for PCR analysis. These SINE codes are shown in the Oryzabase at the National Institute of Genetics (NIG), Mishima, Japan (http://www.shigen.nig.ac.jp/rice/oryzabase/wild/psinecodeAbout.jsp).

4.2

The “AA Genome Code” Is Useful for Classification and Identification of Rice Species with the AA Genome in Oceania

The AA genome code was helpful to distinguish O. meridionalis and O. rufipogon from northern Australia. O. meridionalis and O. rufipogon are species that can be found at the same locations in northern Australia. Obviously, such characteristics can lead to difficulties in distinguishing and classifying them. We have found that several O. rufipogon strains of Australia showed the SINE code (= AA genome code) for O. meridionalis. Extensive survey of the presence or absence of the O. rufipogon-specific (or O. meridionalis-specific) p-SINE1 members with 239 individuals of O. rufipogon and O. meridionalis strains was carried out on the wild rice stocks in the Oryzabase at NIG. The wild rice resources analyzed had been

286

H. Ohtsubo et al.

A.

r63 r70 r3003* r3004*

r1 r502 r601 r801 r904 r705

r502 O. rufipogon / O. sativa O. barthii / r801 O. glaberrima

r904

-

-

- -

AA110000

+ -

+

-

- -

AA101000

+ -

+

+

- -

AA101100

+ -

-

-

+ -

AA100010

+ -

-

-

- -

AA100001

r824

r902 r906

r1

AA genome code

+ +

r601

O. glumaepatula

r806 O. meridionalis

O. longistaminata * p-SINE3 members

r705

B.

Genome species type r441 r452,

GT code

AA

+

+

-

-

GT1100

BB

O.punctata

+

-

+

-

GT1010

BBCC

O.punctata O.minuta

+

-

CC

O.officinalis O.eichingeri O.rhizomatis

+

-

-

+

GT1001

CCDD

O.grandiglumis O.alta O.latifolia

+

-

-

+/-

GT1002

O.australiensis

+

-

-

-

GT1000

r301 r302 r305 r2001* r2005*

r2003 r441 r301 r307

O..sativa O.rufipogon O.glaberrima O.barthii O.glumaepatula O.longistaminata O.meridionalis

r445, r2014*

+/- +/-

GT1022

r2003*

r24* r2002* r2007*

r307

DD EE

* p-SINE2 members

C.

species O. sativa

D.

subspecies

r1

r502

Japonica

+

+

r55

Javanica

+

+

indica

+

+

+

species

ecotypes

r1

r502

r503

O. rufipogon

perennial

+

+

+

O. rufipogon

intermediate

+

+

O. rufipogon

annual

+

+

r507

Os code

+

Os 1101 Os 1100 Os 1110

r215

Or code Or 1110 Or 1100

+

+

Or 1111

Fig. 4 SINE codes. A AA genome code. Tree at left shows relationships among species with the AA genome. Species-specific p-SINE members are also shown. Six p-SINE1 members used for the AA genome code are boxed. Presence or absence pattern of the six p-SINE1 members in each species is shown. AA genome codes corresponding to each species in the phylogenetic tree are shown right. B GT code. Tree at left shows relationships among genome types. Genome-typespecific p-SINE members are also shown. Four p-SINE members used for the GT code are boxed. Presence or absence patterns of the four p-SINE members in each genome type are shown. GT codes corresponding to each genome type in the phylogenetic tree are shown right. C Os code. Typical presence or absence patterns of the four p-SINE1 members in each subspecies. Member r502 was used to confirm that the strains belong to the AA-genome type species. Os codes are also shown right. Note that, in this table, we define “javanica” strains as those from insular Southeast Asia. D Or code. Typical presence or absense patterns of the four p-SINE1 members in each ecotype are shown. Tentative Or codes are also shown right

III.5 Rice Retroposon, p-SINE, and Its Use for Classification and Identification

287

Fig. 5 SINE code (the AA genome code) for wild rices (Core collection) in Oryzabase (http:// www.shigen.nig.ac.jp/rice/oryzabase/wild/ coreCollection.jsp). Other SINE codes will be determined in the near future

collected from Oceania (Australia and New Guinea) and insular Southeast Asia (Indonesia and the Philippines). The results indicated that some O. rufipogon strains are classified as O. meridionalis by our method. The results of the analysis are shown in the Oryzabase (Fig. 5; http://www.shigen.nig.ac.jp/rice/oryzabase/ wild/psineGenomeCode.jsp).

4.3

Other Types of SINE Codes

Other types of SINE codes have been issued to determine the genome type of the distantly related rice accessions, as well as to classify ecotypes and subspecies within a specific species, such as O. rufipogon. The “GT (= Genome type) code” has been developed to classify wild rice species with the genome type of AA to EE. It is a four-digit SINE code that begins with GT (Fig. 4B). Recently, BAC-end sequences of 12 wild rice species, representing 10 genome types, have been determined and deposited in GenBank (Ammiraju et al. 2006). The GT code could be extended to other genome types such as FF, GG, HHJJ and HHKK in future.

288

H. Ohtsubo et al.

The “Os code” has been developed to distinguish various strains of indica and japonica subspecies among O. sativa strains (Fig. 4C). Up till now, we have employed only four p-SINE1 members to set the Os code, and found it works well in categorizing O. sativa strains into subspecies. Among 68 O. sativa strains surveyed, 67 strains were categorized into respective groups, as determined previously. We also have many p-SINE1 members that will further divide those subspecies into smaller groups. Those members are listed in Xu et al. (2007). The “Or code” still needs refining and is a tentative method used to distinguish the strains of three ecotypes of O. rufipogon. It is a four-digit SINE code that begins with “Or” to categorize the accessions into perennial, annual and intermediate ecotypes (Fig. 4D). In any case of the SINE codes described above, we assigned basic sets of p-SINE members as well as alternative ones in order to classify each accession without being misleading. Alternative members are useful and important to obtain the right classification by SINE codes whenever PCR is unsuccessful at any corresponding loci on the chromosomes of the strains analyzed. Note that the AA genome code and GT code are designed to distinguish “species” of the rice strains under study. On the other hand, the Or code and Os code are designed to categorize the strains within species into ecotypes or subspecies. Greater numbers of “ecotypes or subspecies-specific” p-SINE members would identify a more accurate relationship among those strains in future.

5

Final Remarks

Recently, Mark Batzer, one of the founders of the SINE method in human and primate systems, described SINE as a “nearly perfect character”. He and his colleagues have reviewed the extensive body of data available for more than 11,000 primate-specific Alu SINEs in order to evaluate the misleading events based on parallel insertions, precise excisions and lineage sorting artifacts. Their extensive survey shows that there are very low levels of each potentially confounding factor in phylogenetic analysis. For example, “near-parallel insertion” events, which are the most common misleading events and easily resolved by sequencing efforts, are only noted in 41 of 11,000 (or more) cases (Ray et al. 2006). In our rice SINE system, we avoided such misleading data by checking almost all the ambiguous PCR results by sequencing analysis. Our concern is rather to do with the mixed features of the O. rufipogon strains themselves, which might be partly due to natural hybridizations inter- or intra-species of the wild and cultivated rice. In this chapter, we have described how phylogenetic analysis based on p-SINE insertion polymorphism is a powerful tool to classify the strains of cultivated and wild rice species and to determine their evolutionary relationships. We constructed a phylogenetic tree of O. sativa and O. rufipogon strains using p-SINE1 RA subfamily members. It is important to use p-SINE1 members from O. sativa as well as O. rufipogon strains for the phylogenetic analysis, in order to avoid ascertainment bias, which gives ambiguity in the relationships of the strains of distantly related O.

III.5 Rice Retroposon, p-SINE, and Its Use for Classification and Identification

289

rufipogon species to O. sativa or vice versa. We found that the phylogenetic tree constructed by using 26 additional p-SINE1 members isolated from O. rufipogon was basically similar to but not the same as that constructed previously by Cheng et al. (2003). Typically we observed a distinct cluster consisting of javanica strains in insular Southeast Asia when we employed polymorphic p-SINE1 members isolated from O. rufipogon and O. sativa. It might be helpful to isolate more p-SINE1 members from several O. rufipogon strains in order to more clearly understand the intra- and inter-species relationships among those strains. The most important conclusion to our work is that the SINE method is useful to reveal the polyphyletic (at least diphyletic) origin of domesticated rice, O. sativa japonica and indica. Recently, similar findings have been accumulating using different methods such as SSR polymorphism analysis (Garris et al. 2005). A phylogeographic approach to O. rufipogon populations has also revealed the multiple independent domestications of these two major cultivated rice species (Londo et al. 2006). We described here four SINE codes, the AA genome code, GT code, Or code and Os code, which are useful to identify species, genome type, ecotypes and subspecies of any rice strains of unknown origin and their relationships to the rice strains that have been previously analyzed.

6

Future Perspectives

In conclusion, complete taxon sampling and the collection of sufficient numbers of informative insertion events are basic requirements for SINE-based phylogenetic analysis. In that sense, it might be worthwhile to identify more ecotype-specific p-SINE members from O. rufipogon in order to clarify the phylogenetic relationship among them. Another interesting research would be to find new SINE or other types of retroelements which will be helpful to distinguish Gramineae plants. p-SINE is not useful for this purpose, as its distribution is limited to Oryza species. Retroelements distributed within Oryza and other Gramineae species could be a new tool for the analysis of the evolutionary relationship among Gramineae plants. Acknowledgments We acknowledge Ms. T. Miyabayashi for providing wild rice strains and DNA, and Dr. Y. Fukuta for DNA used in this work. We also thank Prof. H. Morishima, Prof. Y. Sano and Prof. M. Batzer for their invaluable discussions regarding phylogenetic analysis. This work was partly supported by a grant from the Ministry of Agriculture, Forestry and Fisheries of Japan (Rice Genome Project GD2007 to H.O.). J-H.X. and C.C. were the recipients of a Japanese Government (Monbukagakusho) Scholarship.

References Ammiraju JSS, Luo M, Goicoechea JL, et al. (2006) The Oryza bacterial artificial chromosome library resource: construction and analysis of 12 deep-coverage large-insert BAC libraries that represent the 10 genome types of the genus Oryza. Genome Res 16:140–147

290

H. Ohtsubo et al.

Bailey AD, Shen CKJ (1993) Sequential insertion of Alu family repeats into specific genomic sites of higher primates. Proc Natl Acad Sci USA 90:7205–7209 Batzer MA, Deininger PL (1991) A human-specific subfamily of Alu sequences. Genomics 9:481–487 Batzer MA, Stoneking M, Alegria-Hartman M, et al. (1994) African origin of human-specific polymorphic Alu insertions. Proc Natl Acad Sci USA 91:12288–12292 Cheng C, Tsuchimoto S, Ohtsubo H, Ohtsubo E (2002) Evolutionary relationships among rice species with the AA genome based on SINE insertion analysis. Genes Genet Syst 77:323–334 Cheng C, Motohashi R, Tsuchimoto S, Ohtsubo H, Ohtsubo E (2003) Polyphyletic origin of cultivated rice: based on the interspersion pattern of SINEs. Mol Biol Evol 20:67–75 Deragon JM, Zhang X (2006) Short interspersed elements (SINEs) in plants: origin, classification, and use as phylogenetic markers. Syst Biol 55:949–956 Deragon JM, Landry BS,Pelissier T, Tutois S, Tourmente S, Picard G (1994) An analysis of retroposition in plants based on a familyof SINEs from Brassica napus. J Mol Evol 39:378–386 Garris AJ, Tai TH, Coburn J, Kresovich S, McCouch S (2005) Genetic structure and diversity in Oryza sativa L. Genetics 169:1631–1638 Ge S, Sang T, Lu BR, Hong DY (1999) Phylogeny of rice genome with emphasis on origins of allotetraploid species. Proc Natl Acad Sci USA 96:14400–14405 Hamdi H, Nishio H, Zielinski R, Dugaiczyk A (1999) Origin and phylogenetic distribution of Alu DNA repeats: irreversible events in the evolution of primates. J Mol Biol 289:861–871 Hirano HY, Mochizuki K, Umeda M, Ohtsubo H, Ohtsubo E, Sano Y (1994) Retrotransposition of a plant SINE into the wx locus during evolution of rice. J Mol Evol 38:132–137 Kato S, Kosaka H, Hara S (1928) On the affinity of rice varieties as shown by fertility of hybrid plants. Bull Sci Fac Agric Kyushu Univ 3:132–147 (in Japanese) Khush G S (1997) Origin, dispersal, cultivation and variation of rice. Plant Mol Biol 35:25–34 Londo JP, Chiang YC, Hung KH, Chiang TY, Schaal BA (2006) Phylogeography of Asian wild rice, Oryza rufipogon, reveals multiple independent domestications of cultivated rice, Oryza sativa. Proc Natl Acad Sci USA 103:9578–9583 Mochizuki K, Umeda M, Ohtsubo H, Ohtsubo E (1992) Characterization of a plant SINE, p-SINE1, in rice genomes. Jpn J Genet 67:155–166 Mochizuki K, Ohtsubo H, Hirano H, Sano Y, Ohtsubo E (1993) Classification and relationships of rice strains with AA genome by identification of transposable elements at nine loci. Jpn J Genet 68:205–217 Morishima H, Sano Y, Oka H I (1984) Differentiation of perennial and annual types due to habitat conditions in the wild rice Oryza perennis. Plant Syst Evol 144:119–135 Morishima H, Sano Y, Oka H I (1992) Evolutionary studies in cultivated rice and its wild relatives. Oxford Surv Evol Biol 8:135–184 Motohashi R, Mochizuki K, Ohtsubo H, Ohtsubo E (1997) Structures and distribution of p-SINE1 members in rice genomes. Theor Appl Genet 95:359–368 Murata S, Takasaki N, Saitoh M, Okada N (1993) Determination of the phylogenetic relationships among Pacific salmonids by using short interspersed elements (SINEs) as temporal landmarks of evolution. Proc Natl Acad Sci USA 90:6995–6999 Nikaido M, Rooney A P, Okada N (1999) Phylogenetic relationships among cetartiodactyls based on insertions of short and long interspersed elements: hippopotamuses are the closest extant relatives of whales. Proc Natl Acad Sci USA 96:10261–10266 Ohtsubo H, Cheng C, Tsuchimoto S, Ohtsubo E (2004) Rice retroposon p-SINE1 and origin of the cultivated rice. Breed Sci 58:1–11 Oka H I (1958) Intervarietal variation and classification of cultivated rice. Indian J Genet Plant Breed 18:79–89 Oka H I (1988) Origin of cultivated rice. Elsevier, Tokyo Oka H I, Morishima H (1967) Variation in the breeding systems of a wild rice, Oryza perennis. Evolution 21:249–258

III.5 Rice Retroposon, p-SINE, and Its Use for Classification and Identification

291

Pritchard J K, Stephens M, Donnelly P (2000) Inference of population structure using multilocus genotype data. Genetics 155:945–959 Ray DA, Xing J, Salem A-H, Batzer MA (2006) Sines of a nearly perfect character. Syst Biol 55:928–935 Sano Y, Morishima H, Oka HI (1980) Intermediate perennial–annual populations of Oryza perennis found in Thailand and their evolutionary significance. Bot Mag Tokyo 93:291–305 Sasaki T, Yasukawa Y, Takahashi K, Miura S, Shedlock AM, Okada N (2006) Extensive morphological convergence and rapid radiation in the evolutionary history of the family Geoemydidae (old world pond turtles) revealed by SINE insertion analysis. Syst Biol 55:912–927 Shedlock AM, Okada N (2000) SINE insertions: powerful tools for molecular systematics. BioEssays 22:148–160 Shimamura M, Yasue H, Ohshima K, et al. (1997) Molecular evidence from retroposons that whales form a clad within even-toed ungulates. Nature 388:666–670 Takahashi K, Terai Y, Nishida M, Okada N (1998) A novel family of short interspersed repetitive elements (SINEs) from cichlids: the pattern of insertion of SINEs at orthologous loci support the proposed monophyl of four major groups of cichlid fishes in Lake Tanganyika. Mol Biol Evol 15:391–407 Tatout C, Warwick S, Lenoir A, Deragon JM (1999) SINE insertions as clade markers for wild crucifers species. Mol Biol Evol 16:1614–1621 Vaughan DA, Morishima H, Kadowaki K (2003) Diversity in the Oryza genus. Curr Opin Plant Biol 6:139–146 Wang ZY, Second G, Tanksley SD (1992) Polymorphism and phylogenetic relationships among species in the genus Oryza as determined by analysis of nuclear RFLPs. Theor Appl Genet 83:565–581 Xu J-H (2004) Phylogenetic analysis of rice strains of the Oryza genus by insertion polymorphism of SINEs. PhD Thesis, University of Tokyo Xu J-H, Kurata N, Akimoto M, Ohtsubo H, Ohtsubo E (2005) Identification and characterization of Australian wild rice strains of Oryza meridionalis and Oryza rufipogon by SINE insertion polymorphism. Genes Genet Syst 80:129–134 Xu J-H, Cheng C, Tsuchimoto S, Ohtsubo H, Ohtsubo E (2007) Phylogenetic analysis of Oryza rufipogon strains and their relations to Oryza sativa strains by insertion polymorphism of rice SINEs. Genes Genet Syst 82:217–229

Section IV

Improvement of Rice

IV.1

Detection and Molecular Cloning of Genes Underlying Quantitative Phenotypic Variations in Rice Toshio Yamamoto1 and Masahiro Yano1(* ü)

1

Introduction

Sequencing of the entire rice genome has made remarkable progress (IRGSP 2005). This sequence information has provided new tools for genetics and has created a new paradigm of plant breeding. Many phenotypic traits of economic interest are controlled by multiple genes and often show complex and quantitative inheritance. Recent progress in rice genomics has had a great impact on the genetic dissection of such traits into single genetic factors, or quantitative trait loci (QTLs) (Tanksley 1993; Yano and Sasaki 1997). Such genetic factors can subsequently be identified at the molecular level by map-based strategies (Yano 2001). Many QTL mapping studies in rice have been conducted during the last decade. Information on individual QTLs is collected and summarized in a cereal genome database, Gramene (http://www.gramene.org/Oryza_sativa/). It is difficult to review all progress due to the tremendous amount of QTL information in this database. Thus, in this chapter, we summarize QTLs with relatively large effects of economic or agronomic interest. Some of them have already been cloned at the molecular level (Table 1). In addition, we describe the platform for use in the systematic exploitation of natural variations and QTLs and in further analyses of QTLs, such as molecular cloning and marker assisted selection (MAS) for the biological study and breeding of rice.

2 2.1

Molecular Cloning of Major QTLs with Agronomic Values Heading Date

Heading date is a key determinant for adaptation of rice to different cultivation areas and cropping seasons. Therefore, control of heading date is one of the leading objectives in rice breeding. Many genetic studies have been conducted for QTL 1 National Institute of Agrobiological Sciences, Tsukuba, Ibaraki 305-8602, Japan e-mail: [email protected]; [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

295

This page intentionally blank

4

13,000 1,850 2,973 10,388

1,505 2,207 2,807 2,500 4,022 6.3 27 7.4 0.61

12 20 26 16 182

Reference

Li et al. (2006b) Nishimura et al. (2005)

Not determined

Ashikari et al. (2005) Ueda et al. (2005) Ren et al. (2005) Konishi et al. (2006)

Yano et al. (2000) Kojima et al. (2002) Takahashi et al. (2001) Doi et al. (2004) Xu et al. (2006)

Not determined Amino acid substitution Amino acid substitution Single nucleotide substitution in regulatory region Amino acid substitution

Premature stop codon Not determined Premature stop codon Amino acid substitution Gene deletion

Population sizea Candidate (no. of plants) regionb (kb) FNPc

Myb3 DNA binding domain 12,000 1.7 and NLS Regeneration PSR1 1 Ferredoxin-nitrate 3,800 50.8 ability reductase a No. of plants used in large-scale linkage mapping b Physical distance of candidate genomic region defined by mapping c Functional nucleotide polymorphism causing alleleic difference between parental lines

Seed shattering Sh4

Cytokinin oxidase/dehydrogenase CPD photolyase HKT-type transporter BEL-type homeodomain protein

Gn1a 1 qUVR10 10 SKC1 1 qSH1 1

Protein and function

Zn-finger domain and CCT motif FT-like protein Casein kinase 2 alpha B-type response regulator Putative ethylene response factor

Chr.

6 6 3 10 9

QTL

Hd1 Hd3a Hd6 Ehd1 SubA1

Trait

Heading date Heading date Heading date Heading date Submergence tolerance Seed number UVB resistance Salt tolerance Seed shattering

Table 1 QTLs with major effects on agronomic traits in rice

296 T. Yamamoto and M. Yano

IV.1 Detection and Molecular Cloning of Genes Underlying Quantitative

297

mapping of heading date using advanced backcross progeny (Yano et al. 2001). We have genetically identified 15 QTLs (Hd1, Hd2, Hd3a, Hd3b, Hd4–Hd14) using several kinds of progeny from a cross between japonica cultivar Nipponbare and indica cultivar Kasalath (Yano et al. 2001). Among them, nine QTLs – Hd1, Hd2, Hd3a, Hd3b, Hd4, Hd5, Hd6, Hd8, and Hd9 –, were mapped as single Mendelian factors (Yano et al. 2001; Lin et al. 2002, 2003). Detection of QTLs for heading date has allowed further genetic analyses, such as the development of nearly isogenic lines (NILs), analysis of epistatic interaction among QTLs, and map-based cloning. Hd1 has been found to encode a protein with zinc finger and CCT motifs and to be an ortholog of Arabidopsis CONSTANS (Table 1; Yano et al. 2000). Hd6 and Hd3a were found to encode a casein kinase 2 alpha and an Arabidopsis FT-like protein (Table 1; Takahashi et al. 2001; Kojima et al. 2002). A major QTL, Early heading date 1 (Ehd1), for heading date on chromosome 10 has been detected by using a BC1F1 population derived from a cross between cultivar T65 and an accession of another cultivated species, Oryza glaberrima (Doi et al. 1998). Further analysis revealed that Ehd1 encodes a B-type response regulator (Table 1; Doi et al. 2004). In all cases of QTL cloning for heading date, large-scale linkage mapping was required to narrow the candidate genomic region for the QTLs (Table 1). These efforts led us to identify functional nucleotide polymorphisms in Hd1, Hd6, and Ehd1. These studies have definitely contributed to our understanding of heading date in rice (Yano et al. 2001). Furthermore, they have had an impact on comparative biology in plant flowering as well as rice breeding (Izawa et al. 2003).

2.2

Seed Shattering

Seed shattering is another important trait for the improvement of rice cultivars. The degree of shattering should be matched to the threshing system, such as by hand or machine. Thus, it is very important to reveal the molecular basis of seed shattering in rice in order to modulate the degree of shattering. Recently two QTLs, qSH1 (QTL for seed shattering on chromosome 1) and Sh4 (Shattering 4), have been isolated by a map-based strategy (Table 1; Konishi et al. 2006; Li et al. 2006b). Loss of function of qSH1 resulted in complete loss of shattering in japonica cultivars, and loss of function of Sh4 resulted in partial loss of shattering in indica cultivars. Konishi et al. (2006) revealed that the qSH1 gene, a major QTL of seed shattering in rice, encodes a BEL-type homeodomain protein. On the basis of a large-scale linkage analysis using more than 10,000 plants, a candidate genomic region for the functional difference was delimited as less than 612 bp long, and a single nucleotide polymorphism (SNP) was detected. This SNP, lying in the 5' regulatory region of qSH1, caused loss of seed shattering owing to the absence of abscission layer formation. In-situ hybridization results revealed that the SNP resulted in loss of mRNA transcription in the abscission layer in Nipponbare, but not in other tissues, indicating pleiotropic functions of qSH1

298

T. Yamamoto and M. Yano

in spikelet and flower development. Thus, it was speculated that this unique Nipponbare allele could only have been selected by humans and survived during rice domestication (Konishi et al. 2006). Li et al. (2006a) identified three QTLs for seed shattering in an F2 population derived from a cross between an indica cultivar and the wild annual species O. nivara. Of these QTLs, Sh4 explained about 70% of total phenotypic variance. Further analysis clearly demonstrated that Sh4 encoded a putative Myb DNA-binding-domain protein (Li et al. 2006b).

2.3

Salt Tolerance

Soil salinity is a major stress in rice and other crop plants. Extensive genetic analyses of salinity tolerance have been carried out. Lin et al. (2004) detected eight QTLs for salinity tolerance by using an F2/F3 population derived from a cross between a salt-tolerant cultivar, Nona Bokra, and an intolerant cultivar, Koshihikari. Among them, SKC1, on chromosome 1, had a major effect and was a target for map-based cloning (Lin et al. 2004). A large-scale linkage analysis in an advanced backcross population delimited a 7.4-kb genomic region as a candidate for SKC1 (Table 1; Ren et al. 2005). Complementation analysis and other physiological analyses revealed that SKC1 encodes an HKT-type transporter (Ren et al. 2005). It was suggested that the functional difference between the Nona Bokra and Koshihikari alleles might be due to four nucleotide polymorphisms, which result in an amino acid substitution. This study also demonstrated that SKC1 plays a function in the regulation of K+/Na+ homeostasis under stress conditions (Ren et al. 2005). This finding is useful for rice breeding and will be a trigger for the artificial control of salt tolerance in rice.

2.4

Submergence Tolerance

Submergence by deep water causes severe stress to rice in south-east Asia, where flooding occurs during the monsoon season. QTL analysis of tolerance to submergence has been carried out using a highly tolerant indica cultivar, FR13A, and an intolerant japonica cultivar, M-202 (Xu and Mackill 1996). A major QTL, Submergence 1 (Sub1), was detected near the centromere of chromosome 9. Further fine genetic mapping using a large-scale mapping population (4,022 plants) revealed a genomic region containing a cluster of three genes encoding putative ethylene response factors as a candidate for Sub1 (Table 1; Xu et al. 2000, 2006). Most intolerant cultivars analyzed contained Sub1C or Sub1B but not Sub1A. This observation suggests that Sub1A is the most probable candidate for Sub1, on the basis of the sequence variability of several rice accessions. Overexpression of Sub1A resulted in enhancement of tolerance (Xu et al. 2006). The Sub1A locus was introgressed by MAS into an elite cultivar widely grown in Asia, and resultant lines showed promising agronomic performance in yield and other agronomic traits, as well as tolerance to submergence (Xu et al. 2006).

IV.1 Detection and Molecular Cloning of Genes Underlying Quantitative

2.5

299

Regeneration Ability

Regeneration ability is an important trait for generating transgenic rice plants. Cultivars show a wide range of natural variation in regeneration ability. Many QTL analyses have been performed to identify genetic factors controlling such ability. Nishimura et al. (2005) detected four QTLs for regeneration ability using BC1F1/ BC1F2 lines derived from a cross between Koshihikari, a low-regeneration cultivar, and Kasalath, a high-regeneration cultivar. Of the four QTLs identified, Promoter of Shoot Regeneration 1 (PSR1), on chromosome 1, has been subjected to mapbased cloning. High-resolution mapping using about 3,800 plants delimited a region of less than 50.8 kb as a candidate genomic region for PSR1 (Table 1). Sequencing analysis and genetic complementation tests showed that PSR1 encodes ferredoxin-nitrate reductase. It was also suggested that PSR1 can be used as a selection marker for rice transformation (Nishimura et al. 2005).

2.6

UVB Resistance

A wide range of natural variation in ultraviolet-B (UVB) resistance is seen among rice cultivars. Three QTLs for UVB resistance have been mapped on chromosomes 1, 7, and 10 by using backcross inbred lines (BILs) from a cross between Nipponbare and Kasalath (Sato et al. 2003). Among them, qUVR10 (QTL for ultraviolet-B resistance on chromosome 10) showed the largest effect. Highresolution mapping using 1,850 F2 plants delimited qUVR10 to less than 27 kb (Table 1). We identified a gene encoding the cyclobutyl pyrimidine dimer (CPD)specific photolyase in this region, and complementation analysis showed that qUVR10 encodes CPD photolyase (Ueda et al. 2004, 2005). Comparison of qUVR10 sequences between Nipponbare and Kasalath alleles revealed one probable candidate for the functional nucleotide polymorphism. A single-base substitution in the CPD photolyase gene resulted in an amino acid substitution that altered the CPD photorepair activity and UVB resistance. Development of an NIL for qUVR10 will be required to prove whether this functional difference in qUVR10 results in phenotypic differences, such as growth and yield potential.

3 3.1

Identification of Major QTLs with Agronomic Value Partial or Field Resistance to Rice Blast

Most disease resistance reactions are controlled by single major genes that have been used extensively in past rice breeding programs. However, new cultivars with major genes frequently show a rapid breakdown of resistance to rice blast, which is

300

T. Yamamoto and M. Yano

caused by Magnaporthe grisea. On the other hand, Japanese upland rice cultivars often exhibit a high level of blast resistance. A major objective of breeding for rice blast resistance is to introduce partial, non-race-specific resistance into modern rice cultivars in temperate regions, such as Japan and Korea. This resistance is often termed field resistance and is thought to be durable. Its inheritance is controlled by several genes, and the mode of inheritance is complex. So far, QTL analyses have been used to dissect genetic factors controlling such resistance (Fukuoka and Okuno 2001; Miyamoto et al. 2001; Zenbayashi et al. 2002). Four QTLs for partial resistance to rice blast have been detected by using F2/F3 populations derived from a cross between a Japanese lowland cultivar, Nipponbare, and an upland cultivar, Owarihatamochi. Of those QTLs, pi21, on chromosome 4, explained about 45% of total phenotypic variation in the F3 generation. Map-based cloning of pi21 is progressing, and the gene has recently been identified at the molecular level (Fukuoka et al., unpublished data). Several genetic studies have been performed to detect other QTLs for field resistance to rice blast. One major QTL, qBFR4-1, has been identified on chromosome 4 using F2/F3 populations derived from a cross between a Japanese upland rice cultivar, Kahei (high level of resistance), and a lowland cultivar, Koshihikari (susceptible), and explains about 60% of total phenotypic variance (Miyamoto et al. 2001). Zenbayashi et al. (2002) detected a major QTL on chromosome 11 in the analysis of progeny derived from rice cultivars Chubu 32 (high level of resistance) and Norin 29 (susceptible). This QTL explained about 45% of total phenotypic variance in F3 lines. These two QTLs could be targets for map-based cloning.

3.2

Pre-harvest Sprouting (Seed Dormancy)

Seed dormancy is an important trait in the breeding of rice and other cereal crop species, because it is associated with pre-harvest sprouting. Pre-harvest sprouting often occurs in hot, humid conditions at maturity, resulting in a reduction in grain quality. The degree of seed dormancy is a typical complex trait determined by a series of QTLs. Several genetic analyses by QTL mapping of seed dormancy or pre-harvest sprouting in rice have been reported (Lin et al. 1998; Cai and Morishima 2000; Takeuchi et al. 2003; Gu et al. 2005). Lin et al. (1998) identified five QTLs for seed dormancy using 98 BILs derived from a cross between Nipponbare and Kasalath. Further genetic analysis using advanced backcross progeny demonstrated that Seed dormancy 1 (Sdr1), on chromosome 3, could be mapped as a single Mendelian factor (Takeuchi et al. 2003). Furthermore, Sdr4, on chromosome 7, has been targeted for map-based cloning (Sugimoto, Takeuchi, and Yano, unpublished data). In addition, it was recently been reported that five QTLs are involved in variation in pre-harvest sprouting of BC1 progeny derived from a cross between weedy strain SS18-2 and the breeding line EM93-1 (Gu et al. 2005). Among them, qSD1,

IV.1 Detection and Molecular Cloning of Genes Underlying Quantitative

301

qSD7-1, and qSD12 could be targets for molecular cloning, owing to their large allelic differences (Gu et al. 2006).

3.3

Cold Tolerance at Booting Stage

Low temperature at booting stage is the most serious abiotic constraint on rice production in temperate regions. Low temperature causes abnormal pollen development and thus seed sterility. Several QTL analyses have been performed on cool temperature tolerance (Saito et al. 1995; Takeuchi et al. 2001; Andaya and Mackill 2003; Dai et al. 2004). Saito et al. (1995) demonstrated that two putative regions of chromosomes 3 and 4 were responsible for the cold tolerance of Norin-PL8, which harbors several introgressed chromosome segments from a tropical japonica cultivar, Silewah. Further genetic analysis revealed two closely linked genes, Ctb1 and Ctb2, in the QTL on chromosome 4 (Saito et al. 2001). Recently, a putative candidate gene for Ctb1 has been identified by fine mapping (Saito et al. 2004). Three QTLs for cool temperature tolerance have been detected using doubled haploid lines (DHLs) derived from a cross between temperate japonica cultivars (Takeuchi et al. 2001). Among them, qCT7, on chromosome 7, showed a major phenotypic effect and explained about 22% of total phenotypic variance in the DHLs. In addition, nine QTLs have been detected by using F2/F3 populations derived from a cross between a Chinese cultivar, Kunmingxiaobaigu (cool temperature tolerant), and a Japanese cultivar, Towada (intolerant). Among them, two QTLs on chromosomes 7 and 10 explained about 20% and 37%, respectively, of total phenotypic variance. These major QTLs are currently being subjected to map-based cloning.

3.4

Low-temperature Germinability

Low temperature at germination is a major constraint on the stable establishment of seedlings sown by the direct seeding method. Two studies have been performed on the QTL analysis of low-temperature germinability (Miura et al. 2001; Fujino et al. 2004). Fujino et al. (2004) identified three QTLs using 122 BILs derived from a cross between two temperate japonica cultivars, Italica Livorno (vigorous germination at low temperature), and Hayamasari (low germinability). One of them, qLTG3-1, was mapped on chromosome 3 and explained about 35% of the total phenotypic variance. This QTL is under study. Five QTLs for low-temperature germination located on chromosomes 2, 4, 5, and 11 were detected in a cross between Nipponbare and Kasalath (Miura et al. 2001). Even though the effects of these QTLs are relatively small, large phenotypic differences have been observed between advanced backcross lines for two QTLs, qLTG4-1 and qLTG11, and the recurrent parent, Nipponbare. These two QTLs could be targets for map-based cloning.

302

3.5

T. Yamamoto and M. Yano

Seed Longevity

Seed longevity affects the regeneration cycle of accessions stored in gene banks. Rapid seed deterioration in humid tropical climates is a serious problem in rice production throughout monsoon Asia. QTLs controlling seed longevity have been identified using a mapping population derived from a cross between japonica and indica cultivars (Miura et al. 2002; Sasaki et al. 2005). Miura et al. (2002) identified three putative QTLs for seed longevity using 98 BILs derived from a cross between Nipponbare and Kasalath. The QTL with the largest effect, qLG-9, on chromosome 9, explained about 60% of total phenotypic variance. The effect of the Kasalath allele at this QTL was verified by using chromosome segment substitution lines (CSSLs), and the QTL is a promising candidate for map-based cloning.

4

4.1

Systematic Genetic and Molecular Analysis and Utilization of Natural Variations Development of Novel Mapping Populations

Although sequence information and molecular tools have already been rapidly accumulated, plant materials for genetic analysis are still being developed. A lack of genetic materials can limit our comprehensive understanding of quantitative traits. In fact, as mentioned above, advanced backcross progeny must be developed to clone QTLs. Without such novel plant materials, it is very difficult to understand the genetic basis of quantitative traits. However, since material development usually requires a long time and much labor, most QTL analyses have been done on primary mapping populations, such as F2 and recombinant inbred lines (RILs). Primary QTL mapping usually provides only an approximate chromosomal location. In some cases, QTLs detected in primary mapping populations might be false positives. To resolve these problems, novel mapping populations will be required in order to facilitate further genetic analyses, including molecular cloning of target QTLs. Recently, novel mapping populations, such as introgression lines and CSSLs in rice (Kubo et al. 2002; Ebitani et al. 2005), have been developed. In these lines, a particular chromosomal segment from a donor line is substituted into the genetic background of the recurrent line. The substituted segments cover all chromosomes in a whole set of lines. To enhance the potential of the CSSL platform, we are now developing CSSLs using a wide range of cross combinations, with an elite Japanese cultivar, Koshihikari, as the recurrent parent. Donor parental lines are indica and japonica cultivars, including some core collections (Kojima et al. 2005).

IV.1 Detection and Molecular Cloning of Genes Underlying Quantitative

4.2

303

Exploration and Cloning of Useful Alleles Using CSSLs

The potential of CSSLs in QTL detection has been demonstrated in many ways. We have established a systematic research flow for exploration and cloning of useful genes (Fig. 1). For example, CSSLs can be used in genetic analysis to associate QTLs with particular chromosomal regions and to quickly develop NILs of target regions containing QTLs of interest. In general, when an association is detected between a chromosomal region and a trait, it is often difficult to validate the QTLs, especially those with very small genetic effects. In such a case, NILs are required in order to analyze genetic effects in detail (Miura et al. 2001; Sato et al. 2003; Ueda et al. 2004). Because CSSLs normally have one chromosomal region substituted, they can be used as NILs themselves or as starting material to develop NILs. Such NILs enable us to combine two or three QTLs in one genetic background in

Fig. 1 Systematic research flow for the exploration and utilization of natural variations in rice

304

T. Yamamoto and M. Yano

order to clarify the epistatic interaction among them (Lin et al. 2000, 2003; Yamamoto et al. 2000). Furthermore, once we detect significant differences between the CSSLs and the recurrent parental line, comparison of the size of the substituted segments enables us to delineate candidate chromosomal regions of QTLs (substitution mapping). If a significant difference is found between a particular CSSL and the recurrent parent, a large mapping population can be easily produced by a simple crossing of the CSSL with the recurrent parent. Map-based cloning of the QTLs detected can be started quickly by using such plant materials. There is a disadvantage in detecting QTLs by CSSLs compared with primary mapping populations. It may be difficult to detect phenotypic differences generated by a combination of two or more chromosomal regions. In this case, the use of RILs may be an alternative way to detect such phenotypic differences. The mapping resolution for QTLs in CSSLs may be lower than that in primary mapping populations, because it depends on the size of the substituted chromosome segments in CSSLs. However, this disadvantage can be easily overcome by fine mapping of putative QTLs using the CSSLs as base materials.

4.3

Implications of Use of CSSLs in Rice Breeding

When an elite cultivar is used as the recurrent parent in the development of CSSLs, introgression breeding can be systematically and rapidly achieved. Once favorable alleles from the donor cultivar in CSSLs are identified, NILs of the elite cultivar with improved characteristics can be developed within two or three years. In addition, such NILs could be used as new cultivars and as promising parental lines in future rice breeding. For example, tiny chromosomal segments for submergence tolerance and heading date have been introgressed by using closely linked DNA markers (Xu et al. 2006; Takeuchi et al. 2006). Four QTLs for rice heading date – Hd6, Hd1, Hd4, and Hd5 – were introgressed from Kasalath into Koshihikari by MAS in order to enhance the cropping potential of Koshihikari, one of the leading cultivars in Japan (Takeuchi et al. 2006). As a result, NILs of Koshihikari with early and late heading dates have been successfully developed. The size of the introgressed chromosomal segments in those lines was very small: 300 to 600 kb in NILs for Hd1, Hd6, and Hd5. Precise information on the chromosomal locations of the genes allowed the breeders to minimize the length of the substituted chromosome segments containing the target QTLs. This study clearly demonstrated the potential power of MAS in rice breeding. Furthermore, by using various NILs and DNA markers, we may introduce genes derived from different donor cultivars into elite cultivars (QTL/gene pyramiding) (Ashikari and Matsuoka 2006). For example, two genes, semidwarf 1 (sd1) and Gn1 (Grain number 1), were successfully combined into elite Japanese cultivar Koshihikari (Ashikari et al. 2005). This strategy has not yet been adopted for phenotype-based selection in rice breeding. New series of CSSLs from different elite

IV.1 Detection and Molecular Cloning of Genes Underlying Quantitative

305

cultivars will be necessary to enhance variations of economic interest, and the development of these materials should facilitate the improvement of rice cultivars.

5

Future Prospects

As mentioned above, recent progress in rice genomics has facilitated our understanding of quantitative traits with complex inheritance. Our deep understanding of traits with agronomic interest has created a new paradigm of rice breeding (Fig. 1). In addition, past and present efforts to develop novel mapping populations will enhance the performance of the identification of novel alleles in natural variation in terms of plant biology and rice breeding. To exploit the potential use of natural variations, unique phenotyping methods are crucial. Minimizing experimental errors caused by environmental conditions and artificial factors is essential to obtaining reliable information on QTLs. It will be very important to pay attention to the development of new phenotyping methods. Furthermore, gene and environment interaction is an important aspect of gene expression. This is an important factor in the breeding of rice cultivars adapted to regional environmental conditions. Therefore, to enhance the exploitation and utilization of novel alleles in rice breeding, novel mapping populations such as BILs and CSSLs should be tested under different environmental conditions. In this chapter, we focus only on QTLs with major effects. For the last decade, major QTLs have been isolated mainly by map-based strategy. However, many QTL studies have demonstrated that not only QTLs with major effects but also those with minor effects generate natural variations among cultivars in traits with agronomic importance, such as yield potential and drought tolerance. It is difficult to apply the map-based strategy to the cloning of minor QTLs. It will be necessary to establish another way to understand such highly complex traits controlled by minor QTLs. In this regard, artificial mutations will be an option for verification of the function of target QTLs in conjunction with map-based cloning. Even in major QTLs, it is sometimes difficult to delimit a candidate genomic region of less than 50 kb. In such cases, several candidate genes should be subjected to functional validation, for example through genetic complementation by transgenic analysis. However, in general, it is difficult to perform such large-scale experiments. In these cases, mutants of candidate genes will provide crucial evidence for the molecular identification of QTLs. In rice, several types of mutant panels have been generated by Tos17 or T-DNA insertion, exposure to chemicals, and gamma irradiation (Wu et al. 2005; Miyao et al. 2007). Mutants of target genes can be systematically screened by using Tos17 sequences and T-DNA and by tiling (Raghavan et al. 2007). Once a particular mutant of interest is obtained, morphological and physiological analyses of it will be helpful for the functional validation of candidate genes. In terms of rice breeding, QTL pyramiding can be used with traits controlled by a few major QTLs, but not with traits controlled by many QTLs with minor effects.

306

T. Yamamoto and M. Yano

Even with major QTLs, developing NILs for each takes a long time and much labor. Thus, it will be necessary to develop other methods to combine several QTLs with minor effects in rice breeding.

References Andaya VC, Mackill DJ (2003) QTLs conferring cold tolerance at the booting stage of rice using recombinant inbred lines from a japonica and indica cross. Theor Appl Genet 106:1084–1090 Ashikari M, Matsuoka M (2006) Identification, isolation and pyramiding of quantitative trait loci for rice breeding. Trends Plant Sci 11:344–350 Ashikari M, Sakakibara H, Lin SY, et al. (2005) Cytokinin oxidase regulates rice grain production. Science 309:741–745 Cai HW, Morishima H (2000) Genomic regions affecting seed shattering and seed dormancy in rice. Theor Appl Genet 100:840–846 Dai L, Lin X, Ye C, et al. (2004) Identification of quantitative trait loci controlling cold tolerance at the reproductive stage in Yunnan landrace of rice, Kunmingxiaobaigu. Breed Sci 54:253–258 Doi K, Yoshimura A, Iwata N (1998) RFLP mapping and QTL analysis of heading date and pollen sterility using backcross populations between Oryza sativa L. and Oryza glaberrima Steud. Breed Sci 48:395–399 Doi K, Izawa T, Fuse T, et al. (2004) Ehd1, a B-type response regulator in rice, confers short-day promotion of flowering and controls FT-like gene expression independently of Hd1. Genes Dev 18:926–936 Ebitani T, Takeuchi Y, Nonoue Y, Yamamoto T, Takeuchi K, Yano M (2005) Construction and evaluation of chromosome segment substitution lines carrying overlapping chromosome segments of indica rice cultivar ‘Kasalath’ in a genetic background of japonica elite cultivar ‘Koshihikari’. Breed Sci 55:65–73 Fujino K, Sekiguchi H, Sato T, et al. (2004) Mapping of quantitative trait loci controlling lowtemperature germinability in rice (Oryza sativa L.). Theor Appl Genet 108:794–799 Fukuoka S, Okuno K (2001) QTL analysis and mapping of pi21, a recessive gene for field resistance to rice blast in Japanese upland rice. Theor Appl Genet 103:185–190 Gu XY, Kianian SF, Hareland GA, Hoffer BL, Foley ME (2005) Genetic analysis of adaptive syndromes interrelated with seed dormancy in weedy rice (Oryza sativa). Theor Appl Genet 110:1108–1118 Gu XY, Kianian SF, Foley ME (2006) Isolation of three dormancy QTLs as Mendelian factors in rice. Heredity 96:93–99 IRGSP (International Rice Genome Sequencing Project) (2005) The map-based sequence of the rice genome. Nature 436:793–800 Izawa T, Takahashi Y, Yano M (2003) Comparative biology comes to bloom: genomic and genetic comparison of flowering pathways in rice and Arabidopsis. Curr Opin Plant Biol 6:113–120 Kojima S, Takahashi Y, Kobayashi Y, et al. (2002) Hd3a, a rice ortholog of the Arabidopsis FT gene, promotes transition to flowering downstream of Hd1 under short-day condition. Plant Cell Physiol 43:1096–1105 Kojima Y, Ebana K, Fukuoka S, Nagamine T, Kawase M (2005) Development of an RFLP-based rice diversity research set of germplasm. Breed Sci 55:431–440 Konishi S, Izawa T, Lin SY, et al. (2006) An SNP caused loss of seed shattering during rice domestication. Science 312:1392–1396

IV.1 Detection and Molecular Cloning of Genes Underlying Quantitative

307

Kubo T, Aida Y, Nakamura K, Tsunematsu H, Doi K, Yoshimura A (2002) Reciprocal chromosome segment substitution series derived from japonica and indica cross of rice (Oryza sativa L.). Breed Sci 52:319–325 Li C, Zhou A, Sang T (2006a) Genetic analysis of rice domestication syndrome with the wild annual species, Oryza nivara. New Phytol 170:185–194 Li C, Zhou A, Sang T (2006b) Rice domestication by reducing shattering. Science 311:1936–1939 Lin HX, Yamamoto T, Sasaki T, Yano M (2000) Characterization and detection of epistatic interactions of 3 QTLs, Hd1, Hd2, and Hd3, controlling heading date in rice using nearly isogenic lines. Theor Appl Genet 101:1021–1028 Lin HX, Ashikari M, Yamanouchi U, Sasaki T, Yano M (2002) Identification and characterization of a quantitative trait locus, Hd9, controlling heading date in rice. Breed Sci 52:35–41 Lin HX, Liang ZW, Sasaki T, Yano M (2003) Identification and characterization of a quantitative trait locus, Hd4 and Hd5, controlling heading date in rice. Breed Sci 53:51–59 Lin HX, Zhu MZ, Yano M, et al. (2004) QTLs for Na+ and K+ uptake of the shoots and roots controlling rice salt tolerance. Theor Appl Genet 108:253–260 Lin SY, Sasaki T, Yano M (1998) Mapping quantitative trait loci controlling seed dormancy and heading date in rice, Oryza sativa L., using backcross inbred lines. Theor Appl Genet 96:997–1003 Miura K, Lin SY, Yano M, Nagamine T (2001) Mapping quantitative trait loci controlling low temperature germinability in rice (Oryza sativa L.). Breed Sci 51:293–299 Miura K, Lin SY, Yano M, Nagamine T (2002) Mapping quantitative trait loci controlling seed longevity in rice (Oryza sativa L.). Theor Appl Genet 104:981–986 Miyamoto M, Yano M, Hirasawa H (2001) Mapping of quantitative trait loci conferring blast field resistance in the Japanese upland rice variety Kahei. Breed Sci 51:257–261 Miyao A, Iwasaki Y, Kitano H, et al. (2007) A large-scale collection of phenotypic data describing an insertional mutant population to facilitate functional analysis of rice genes. Plant Mol Biol 63:625–635 Nishimura A, Ashikari M, Lin SY, et al. (2005) Isolation of a rice regeneration quantitative trait loci gene and its application to transformation systems. Proc Natl Acad Sci USA 102:11940–11944 Raghavan C, Naredo MEB, Wang H, et al. (2007) Rapid method for detecting SNPs on agarose gels and its application in candidate gene mapping. Mol Breed 19:87–101 Ren ZH, Gao JP, Li LG, et al. (2005) A rice quantitative trait locus for salt tolerance encodes a sodium transporter. Nat Genet 37:1141–1146 Saito K, Miura K, Nagano K, et al. (1995) Chromosomal location of quantitative trait loci for cool tolerance at the booting stage in rice variety ‘Norin-PL8.’ Breed Sci 45:337–340 Saito K, Miura K, Nagano K, Hayano-Saito Y, Araki H, Kato A (2001) Identification of two closely linked quantitative trait loci for cold tolerance on chromosome 4 of rice and their association with anther length. Theor Appl Genet 103:862–868 Saito K, Hayano-Saito Y, Maruyama-Funatsuki W, Sato Y, Kato A (2004) Physical mapping and putative candidate gene identification of a quantitative trait locus Ctb1 for cold tolerance at the booting stage of rice. Theor Appl Genet 109:515–522 Sasaki K, Fukuta Y, Sato T (2005) Mapping of quantitative trait loci controlling seed longevity of rice (Oryza sativa L.) after various periods of seed storage. Plant Breed 124:361–366 Sato T, Ueda T, Fukuta Y, Kumagai T, Yano M (2003) Mapping of quantitative trait loci associated with ultraviolet-B resistance in rice (Oryza sativa L.). Theor Appl Genet 107:1003–1008 Takahashi Y, Shomura A, Sasaki T, Yano M (2001) Hd6, a rice quantitative trait locus involved in photoperiod sensitivity, encodes the α subunit of protein kinase CK2. Proc Natl Acad Sci USA 98:7922–7927 Takeuchi Y, Hayasaka H, Chiba B, et al. (2001) Mapping quantitative trait loci controlling cooltemperature tolerance at booting stage in temperate japonica rice. Breed Sci 51:191–197 Takeuchi Y, Lin SY, Sasaki T, Yano M (2003) Fine linkage mapping enables dissection of closely linked quantitative trait loci for seed dormancy and heading in rice. Theor Appl Genet 107:1174–1180

308

T. Yamamoto and M. Yano

Takeuchi Y, Ebitani T, Yamamoto T, et al. (2006) Development of isogenic lines of rice cultivar Koshihikari with early and late heading by marker-assisted selection. Breed Sci 56:405–413 Tanksley SD (1993) Mapping polygenes. Annu Rev Genet 27:205–233 Ueda T, Sato T, Numa H, Yano M (2004) Delimitation of chromosomal region for a quantitative trait locus, qUVR-10, conferring resistance to ultraviolet-B radiation in rice (Oryza sativa L.). Theor Appl Genet 108:385–391 Ueda T, Sato T, Hidema J, et al. (2005) qUVR-10, a major quantitative trait locus for ultraviolet-B resistance in rice, encodes cyclobutane pyrimidine dimer photolyase. Genetics 171:1941–1950 Wu JL, Wu C, Lei C, et al. (2005) Chemical- and irradiation-induced mutants of indica rice IR64 for forward and reverse genetics Plant Mol Biol 59:85–97 Xu, K, Mackill DJ (1996) A major locus for submergence tolerance mapped on rice chromosome 9. Mol Breed 2:219–224 Xu K, Xu X, Ronald PC, Mackill DJ (2000) A high-resolution linkage map of the vicinity of the rice submergence tolerance locus Sub1. Mol Gen Genet 263:681–689 Xu K, Xu X, Fukao T, et al. (2006) Sub1A is an ethylene-response-factor-like gene that confers submergence tolerance to rice. Nature 442:705–708 Yamamoto T, Lin HX, Sasaki T, Yano M (2000) Identification of heading date quantitative trait locus Hd6, and characterization of its epistatic interaction with Hd2 in rice using advanced backcross progeny. Genetics 154:885–891 Yano M (2001) Genetic and molecular dissection of naturally occurring variations. Curr Opin Plant Biol 4:130–135 Yano M, Sasaki T (1997) Genetic and molecular dissection of quantitative traits in rice. Plant Mol Biol 35:145–153 Yano M, Katayose Y, Ashikari M, et al. (2000) Hd1, a major photoperiod sensitivity quantitative trait locus in rice, is closely related to the Arabidopsis flowering time gene CONSTANS. Plant Cell 12:2473–2484 Yano M, Kojima S, Takahashi Y, Lin HX, Sasaki T (2001) Genetic control of flowering time in rice, a short-day plant. Plant Physiol 127:1425–1429 Zenbayashi K, Ashikawa T, Tani T, Koizumi S (2002) Mapping of the QTL (quantitative trait locus) conferring partial resistance to leaf blast in rice cultivar Chubu 32. Theor Appl Genet 104:547–552

IV.2

Rice Yielding and Plant Hormones Motoyuki Ashikari1(* ü ) and Tomoaki Sakamoto2

1

Introduction

Food shortage is a serious global problem in this century. According to FAO estimates, 852 million people worldwide were undernourished in 2000–2002 (FAO 2004). The global population, now at 6.4 billion, is still growing rapidly and is supposed to reach 8.9 billion by 2050 (UNFPA 2004). Cereals are an important source of calories for humans, both by direct intake and as the main feed for livestock. Approximately 50% of the calories consumed by the world population originate from three cereal species: rice (23%), wheat (17%), and maize (10%) (Khush 2003). However, the rate of world population growth currently exceeds the rate of growth in food production. To meet the expanding food demands, crop grain production needs to be increased by another 50% by 2025 (Khush 2001). Rice (Oryza sativa L.) is one of the most important staple foods; 50% of the human population depends on rice as their main source of nutrition (White 1994). In particular, it is the dominating crop in the monsoon areas of Asia where it has a long history of cultivation; it is deeply ingrained in the daily lives of Asian peoples. As populations are rapidly growing in Asia, increases in rice production will be required to prevent widespread food scarcity. For many years, rice has been the subject of breeding studies aiming at higher yields and better tasting cultivars. These projects were crucial in avoiding famines in the 20th century. However, in the 21st century, food problems will be more severe. The recent decades have brought dramatic advancements in the field of plant genomics (genome science). Numerous rice genome projects have been launched and are providing useful information for plant biology and plant breeding. This

1 Bioscience and Biotechnology Center, Nagoya University, Furoucho, Chikusaku, Nagoya, Aichi 464-8601, Japan e-mail: [email protected] 2

Institute for Advanced Research, Nagoya University, Furoucho, Chikusaku, Nagoya, Aichi 464-8601, Japan e-mail: [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

309

310

M. Ashikari and T. Sakamoto

information now should be fed back to create practical benefits by enabling the generation of new varieties with increased yield. In this chapter we review the control of rice yield by phytohormones in the light of recent discoveries.

2

Breeding of Semi-dwarf Rice Varieties Led to Dramatically Increased Grain Production

In the 1960s, the acceleration of the world population growth rate and the decrease in arable lands raised fears that food production would not meet the growing demand, leading to a global food crisis. In an attempt to solve the problem, the International Rice Research Institute (IRRI) in 1966 bred a semi-dwarf high-yielding variety, IR8, also known as “miracle rice”. IR8 originated from crossing the Taiwanese native semi-dwarf variety Dee-geo-woo-gen which carries the semi dwarf 1 (sd1) gene, and the Indonesian good-tasting variety Peta (Hargrove and Cabanilla 1979; Dalrymple 1986; Khush 1999). In general, nitrogen fertilization is essential to increasing grain production in rice, but it also induces culm elongation, resulting in an overall increase in plant height. Such tall plants are easily lodged by wind and rain and, consequently, yield losses occur. The IR8 semidwarf rice variety resolved this problem because it responded to fertilizer inputs by increased grain yield while culm elongation was inhibited due to the activity of the sd1 allele. The widespread adoption of IR8 led to major increases in rice grain production, giving rise to the so-called Green Revolution. Like IR8, the highyielding varieties Taichung Native 1 in Taiwan and Tongil in Korea, which also contained the sd1 allele from Dee-geo-woo-gen, contributed to food security in those countries (Aquino and Jennings 1966; Suh and Heu 1978). Similarly, the Japanese native semi-dwarf variety Jikkoku (Kikuchi et al. 1985) and the Xray-induced variety Reimei (Futsuhara et al. 1967), as well as the X-ray-induced variety Calrose 76 in the United States (Foster and Rutger 1978), carried different sd1 alleles and were widely used in rice breeding programs. The fact that numerous sd1 alleles have been utilized in numerous rice breeding programs for both indica and japonica varieties demonstrates that the sd1 locus is suitable for controlling plant height in rice.

3

Gibberellins Regulate Plant Height

The breeding of high-yielding rice varieties carrying sd1 alleles doubtlessly is the greatest success in the history of rice breeding. Insights gained from analyses of sd1 mutants have been applied in rice breeding programs (Futsuhara et al. 1967; Suge 1975; Kikuchi et al. 1985). However, for many years it has been puzzling breeders and scientists alike what the molecular basis might be of the semi-dwarf stature of SD1 mutants.

IV.2 Rice Yielding and Plant Hormones

311

Recently, both wheat (Triticum aestivum L.) and rice Green Revolution genes have been identified. The wheat Reduced height1 (Rht1) alleles are caused by semidominant gain-of-function mutations; the Rht1 gene encodes a negative regulator of gibberellin (GA) responses (Peng et al. 1999). Like Rht1, the rice sd1 is related to GA, but it is recessive and GA-responsive. The SD1 gene encodes a rate-limiting enzyme of the GA biosynthetic pathway, GA 20-oxidase (GA20ox) (Fig. 1; Ashikari et al. 2002; Sasaki et al. 2002). sd1-1 was caused by a 383-bp deletion which induced a frame-shift and created a premature stop codon, whereas the other 3 sd1 alleles (sd1-2, sd1-3, and sd1-4) had single nucleotide substitutions which induced amino acid changes (Ashikari et al. 2002; Sasaki et al. 2002). The levels of GA44, GA19, GA20, GA1, GA29, and GA8 in sd1-1 were lower than in the original strain, whereas the amount of GA53 was slightly higher (Ashikari et al. 2002; Sakamoto et al. 2004). These results indicated that the activity of GA20ox, which catalyzes the steps from GA53 to GA20 via GA44 and GA19, was weaker in sd1 than in wild-type plants. The recombinant SD1 protein expressed in E. coli cells catalyzed the conversion of GA53 to GA20 (Ashikari et al. 2002; Sasaki et al. 2002), which confirmed that SD1 encodes an active GA20ox. Four GA20ox genes were identified in the rice genome (Sakamoto et al. 2004). All rice GA20ox genes were expressed in immature and mature panicles at different levels. In addition, OsGA20ox1 was expressed in all vegetative organs including leaf sheaths, leaf blades, stems, and roots. The OsGA20ox2/SD1 transcript was

Fig. 1 The sd1 mutant lacks the enzymatic activity of GA 20-oxidase. A Plant morphology. Left Wild type; right sd1 mutant. B GA biosynthetic pathway. CPS ent-copalyl diphosphate synthase; KS ent-kaurene synthase; KO ent-kaurene oxidase; KAO ent-kaurenoic acid oxidase; GA20ox GA20 oxidase; GA3ox GA3 oxidase. C Mutation site of sd1 in the GA20ox gene, OsGA20ox2

312

M. Ashikari and T. Sakamoto

highly accumulated in stems and other vegetative organs. OsGA20ox3 was not expressed in any vegetative organ, whereas OsGA20ox4 expression was moderate in leaf sheaths, leaf blades, and stems, and weak in vegetative shoot apices and roots (Toyomasu et al. 1997; Ashikari et al. 2002; Sasaki et al. 2002; Sakamoto et al. 2004). These expression profiles suggest that OsGA20ox2/SD1 is the dominant GA20ox gene in stems, and that OsGA20ox1 and OsGA20ox4 might also be involved in GA biosynthesis in vegetative organs. Functional redundancy in GA20ox explains why the null mutations in OsGA20ox2, sd1, exhibited a semi-dwarf phenotype. In addition to OsGA20ox2, OsGA20ox1 and OsGA20ox4 were expressed in all vegetative organs of rice, and all OsGA20ox genes were expressed in the reproductive organs. This overlapping expression pattern, accompanied by the feedback up-regulation of other GA biosynthetic enzymes by the homeostatic system (Hedden and Phillips 2000a), compensates the defect in OsGA20ox2 function in shoot elongation. Consequently, the defect in OsGA20ox2/SD1 induces semi-dwarf phenotype suitable for rice breeding. To date, many GA-related dwarf rice mutants have been isolated, and several genes involved in GA biosynthesis and signal transduction have been characterized (Ashikari et al. 1999, 2002; Ueguchi-Tanaka et al. 2000, 2005; Ikeda et al. 2001; Itoh et al. 2001, 2002b, 2004; Sasaki et al. 2002, 2003; Sakamoto et al. 2004). Though there are various causes of dwarf phenotypes in plants, the characterization of the Green Revolution genes has taught us that controlling GA metabolism or sensing is a highly promising target in attempts to create high-yield semi-dwarf cultivars by means of traditional crop breeding. Today, genetic engineering approaches are at our disposal to create more desirable plant architectures. Two approaches to reduce endogenous GA content have been established: suppression of GA biosynthesis and enhancement of GA catabolism. An example of the first approach involves antisense expression of GA biosynthetic enzyme genes. Itoh et al. (2002a) generated antisense transformants of the rice GA 3-oxidase (GA3ox) gene, OsGA3ox2. However, only a few plants showed a semi-dwarf phenotype; moreover, this phenotype was not inheritable. Similar results were observed in antisense transformants of Arabidopsis (Arabidopsis thaliana L.) GA20ox genes (Coles et al. 1999). These instabilities of phenotypes may have resulted from GA homeostasis, as reduced GA contents stimulate up-regulation of other GA biosynthetic enzyme genes (Coles et al. 1999; Itoh et al. 2002a). The second approach is exemplified by ectopic expression of the gene encoding the GA-catabolic enzyme GA 2-oxidase (GA2ox). Overexpression of the rice GA2ox gene, OsGA2ox1, under the control of a constitutive actin promoter caused drastic decreases in the bioactive GA level, and resulted in a severely dwarfed phenotype (Fig. 2A; Sakamoto et al. 2001). Similar results were observed in transgenic Arabidopsis and wheat carrying the bean (Phaseolus coccineus L.) GA2ox gene, PcGA2ox1 (Hedden and Phillips 2000a,b). Rice transformants also showed severe defects in flower and grain development, indicating that GA is essential for reproductive organ development. As such defects are not suitable for breeding, it is essential to employ organ- or tissue-specific promoters when GA2ox overproduction is employed to produce dwarf phenotypes of any agronomic significance.

IV.2 Rice Yielding and Plant Hormones

313

One solution to this problem is to use the promoter of one of the rice GA3ox genes (Sakamoto et al. 2003). GA3ox catalyzes the final step of GA biosynthesis. Rice has two GA3ox genes: OsGA3ox1, which is specifically expressed in reproductive organs, and OsGA3ox2, which is active in vegetative organs. The loss-offunction mutant of OsGA3ox2, d18, is severely impaired in shoot elongation but not in flower or grain development (Itoh et al. 2001). Thus, when OsGA2ox1 is ectopically expressed in transgenic rice under the control of the D18 promoter, inhibition of GA synthesis should be restricted to leaves and stems. In fact, D18::OsGA2ox1 rice transformants exhibit only moderate dwarfing and form normal flowers and grains, in contrast to the over-expressing transformants described above (Fig. 2B). Using the OsGA3ox2/D18 promoter has the additional benefit of stabilizing the inheritance of the semi-dwarf trait. GA homeostasis is achieved by means of negative feedback regulation of biosynthetic enzyme genes and positive feed-forward regulation of catabolic enzyme genes. This homeostatic regulation may cause unexpected phenotypes when one attempts to modify GA levels by genetic manipulation. However, when the OsGA3ox2/D18 promoter is used to control expression of GA2ox, the resulting reduced levels of GA will up-regulate expression of OsGA3ox2 (endogenous biosynthetic enzyme gene) and that of the introduced OsGA2ox1 (catabolic enzyme gene). Consequently, one can overcome the homeostatic regulation

Fig. 2 Generating semi-dwarf rice by ectopic expression of GA catabolic genes. A Phenotype of Act::OsGA2ox1 transgenic rice. B Phenotype of D18::OsGA2ox1 transgenic rice

314

M. Ashikari and T. Sakamoto

of GA. The complexity of GA biosynthesis and signaling allows for other targets for manipulation; the negative regulators of GA signaling, such as Arabidopsis GAI, wheat Rht, and rice SLR, are good candidates (Peng et al. 1999; Fu et al. 2001; Ikeda et al. 2001). Even though different approaches probably will be necessary for different species, the strategy outlined above enables the introduction of a single dominant dwarfing gene that generates semi-dwarf plants and avoids conventional long-term breeding programs.

4

Brassinosteroids Control Leaf Angle and Biomass Production

Since the harvest index [ratio of grain to (grain plus straw)] for rice is approaching a maximum, future increases in yield will have to be based on biomass increases; that is, there will have to be more net photosynthesis (Mann 1999). It is well known that phenotypes with erect leaves can capture more light for photosynthesis and can be planted more densely, with a higher leaf area index (one-sided leaf area per unit of land area), both of which contribute to increased yields (Sinclair and Sheehy 1999). In rice, the contribution of the lower leaves to photosynthesis is significant even though their photosynthetic capacity is considerably lower than that of the upper leaves (Horton 2000). Erect leaves allow greater penetration of light to the lower leaves, thereby optimizing canopy photosynthesis (Sinclair and Sheehy 1999). Though the erect-leaf phenotype is the most common among all brassinosteroid (BR)-related rice mutants, it has not been utilized in traditional rice breeding, probably because most BR-related mutants have small grains or decreased fertility (Hong et al. 2002, 2003, 2005; Tanabe et al. 2005). However, the weakest mutant recently identified, osdwarf4-1, showed erect leaves without abnormal flower or grain morphology (Fig. 3A; Sakamoto et al. 2006). This phenotype is a loss-offunction mutant of OsDWARF4 which encodes a BR biosynthetic cytochrome P450. In rice, two different cytochrome P450s, CYP90B2/OsDWARF4 and CYP724B1/D11, redundantly catalyze the C-22 hydroxylation in BR biosynthesis. C-22 hydroxylation is catalyzed by a single enzyme, CYP90B1/DWARF4 (the ortholog of OsDWARF4), in Arabidopsis; in rice, CYP724B1/D11 plays a dominant role in catalyzing this step, as judged from the severity of mutant phenotypes (Sakamoto et al. 2006). The mutant phenotypes indicate that CYP724B1/D11 controls the synthesis of bioactive BR for the regulation of shoot elongation and reproductive development, whereas OsDWARF4 controls BR synthesis for the regulation of leaf inclination but not reproductive development. This is the reason why osdwarf4-1 shows slight dwarfism and erect leaves but no abnormal leaf, flower, or grain morphology, and thus lends itself to further breeding efforts. Under dense planting conditions, erect leaves greatly increase the aboveground biomass and grain yield of osdwarf4-1 mutant rice, even without extra fertilizer application (Fig. 3B). These results suggest a strategy for genetic improvement of crop production by modulation of BR biosynthesis. Morinaka et al. (2006) have

IV.2 Rice Yielding and Plant Hormones

315

Fig. 3 Erect leaf phenotype increases rice grain yield. A Erect leaf phenotype of osdwarf4-1. Degree of bending between leaf blade and sheath of osdwarf4-1 (right) is less than that of the wild type (left). lb Leaf blade; ls leaf sheath. B Comparison of estimated grain yields between wild type (W; white bar) and osdwarf4-1 (M; black bar). Density indicates conventional planting (22.2 plants m−2) or dense planting (44.4 plants m−2). Fertilizer indicates the level of nitrogen fertilizer; the conventional level is ×1 (6 g m−2), and the two increased levels are × 1.5 and × 2 (9 and 12 g m−2 respectively). Bars labeled with different letters are significantly different according to Tukey’s HSD test ( p < 0.01)

generated transgenic rice that partially suppresses the expression of the rice BR receptor gene OsBRI1. These transformants also exhibited a moderately dwarfed phenotype with erect leaves and thus have the potential to increase grain yield. Although further studies are needed to confirm increased yields in these transgenic plants at high planting densities in paddy fields, the results show that erect-leaf plants without defects in reproductive development can be generated by modifying the metabolism and sensing of BR. A single dominant transgene might be sufficient to generate plants with erect leaves and increased grain yield, avoiding the need for conventional long-term breeding programs and without the negative environmental effects caused by nitrogen fertilizers.

5

Cytokinins Regulate Grain Number

In contrast to monogenic characteristics, such as disease and insect resistance, many important agronomic traits including yield, heading date, culm length, grain quality, and stress tolerance show continuous phenotypic variation (Yano 2001). These complex traits usually are governed by a number of genes known as quantitative trait loci (QTLs) derived from natural variations. Polygenic characteristics including QTLs were previously very difficult to analyze using traditional plant breeding methods. Recent progress in the rice genome project has greatly facilitated the detection of QTLs.

316

M. Ashikari and T. Sakamoto

Grain number is an important trait that directly contributes to grain productivity. During the past decade, many attempts have been made to characterize QTLs for grain number (Gn). To identify genes of QTLs for Gn, Ashikari et al. (2005) chose an indica rice variety, ‘Habataki’, and a japonica variety, ‘Koshihikari’, not only because they exhibit large variations in grain number, but also because numerous molecular markers are available. A choice of parental lines that show wide phenotypic variation in the targeted traits is necessary for QTL analysis because QTL detection is based on natural allelic differences between parental lines. QTL analysis detected five QTLs for increasing Gn (Gn1–5). The most effective QTL, Gn1, was chosen as the target for cloning. In QTL cloning, producing nearly isogenic lines (NILs) carrying only one target QTL helps to eliminate the effects of other QTLs. Consequently, the QTL of interest in the NIL can be considered a single Mendelian factor. We produced the NIL-Gn1 carrying the Gn1 region from ‘Habataki’ in the ‘Koshihikari’ background and used it for cloning. Fine mapping of QTL-Gn1 showed that Gn1 actually consisted of two QTLs (QTL-Gn1a and QTL-Gn1b). Positional cloning demonstrated that Gn1a corresponded to cytokinin oxidase/dehydrogenase, OsCKX2, an enzyme that degrades the phytohormone cytokinin (CK). OsCKX2 of ‘Koshihikari’ and ‘Habataki’ consists of four exons and three introns, and encodes proteins of 565 and 563 amino acids, respectively. Comparison of the DNA sequences between the cultivars revealed several nucleotide changes, including a 16-bp deletion in the 5′-untranslated region, a 6-bp deletion in the first exon, and three nucleotide changes resulting in amino acid variation in the first and fourth exon of the ‘Habataki’ allele. Both OsCKX2 alleles encode proteins which enzymatically degrade CK. However, expression levels of OsCKX2 in Habataki are reduced and cause CK accumulation in inflorescence meristems. CK was first characterized as a plant hormone that promotes cell division. It is now known to influence various aspects of plant growth and development including seed germination, apical dominance, leaf expansion, reproductive development, and delay of senescence. In tobacco (Nicotiana tabacum L.) and Arabidopsis, overexpression of CKX results in reduced levels of endogenous CK and lowered meristem activity (Werner et al. 2001, 2003). In transgenic Arabidopsis, overexpression of AtCKX3, the allele showing the highest similarity to OsCKX2 of the seven AtCKXs, reduced flower number due to a decreased rate of primordia formation in the shoot apical meristem (Werner et al. 2003). These findings are in agreement with the results from rice described above. Obviously, CK regulates flower and grain number in plants; its regulation and sensing are promising targets for future breeding programs.

6

Future Perspectives: Accumulation of Yielding-related Genes for Breeding

Plant hormones are directly and indirectly associated with crop yielding. The molecular studies mentioned above have demonstrated that CK regulates the number of grains, GA is a key player in controlling crop height, and BR regulates

IV.2 Rice Yielding and Plant Hormones

317

leaf angles. These results underline the value of research into the genetic control of plant hormones for crop breeding. QTL pyramiding is also an efficient strategy for crop improvement. This is based on the combination of desirable QTLs through conventional crossing with the use of molecular marker. Once desirable QTLs are identified, production of NILs harboring only one target QTL is necessary for QTL pyramiding. The NILQTLs are constructed by repeated backcrossing. NILs allow an evaluation of the effects of single QTLs, because other QTL positions were substituted by the recurrent parents’ genomic background. If desirable traits were detected in NIL-QTLs, the NIL-QTLs were crossed to pyramid the QTLs in the same genome. For example, Ashikari et al. (2005) crossed two NILs, NIL-Gn1 and NIL-sd1 carrying the Gn1 and sd1 regions, respectively, derived from ‘Habataki’ in the ‘Koshihikari’ genome background. The resulting QTL pyramiding line, NIL-sd1+Gn1, shows increased grain number and a semi-dwarf phenotype, and the grain yield was significantly improved. Interspecific crosses between O. sativa and wild relatives could lead to the discovery of useful QTLs from a range of allelic variations much wider than that present in cultivated lines. Furthermore, wild rice species are likely to provide access to QTLs for higher yield (Brar and Khush 1997; Tanksley and McCouch 1997). Discovering useful genes, improving agricultural traits hidden in the plant genome, and applying these findings to crop breeding will pave the way for a new Green Revolution. The QTL pyramiding approach is an efficient tool for crop breeding, but it is not the only one. Genetic engineering approaches with transgenic technology also hold great potential for crop improvement. Today, the cloning of genes and production of transgenic plants referred to as genetically modified organisms (GMOs) are routine technologies. We think that both approaches are necessary to meet future challenges. The growth rate of the world population has once again exceeded the rate of growth in food production (Khush 1999), and prompt measures and action for the second Green Revolution have been called for to avoid widespread food scarcity in the future. To achieve this level of productivity, highly efficient breeding with very close collaboration between genomic scientists and breeders is necessary. Thanks to the growing infrastructure of plant genomics, more genes coding important agronomic traits will be identified, and these scientific discoveries and tools will feedback to practical breeding programs. In addition, the rice genome is very similar to those of other major cereal crops such as maize, barley, and wheat. Therefore, rice genomics has significant implications not only for rice breeding but also for other cereal breeding programs. We hope that the potential of rice genomics will be fully exploited by scientists and breeders, so that it will contribute to the welfare of humans.

318

M. Ashikari and T. Sakamoto

References Aquino RC, Jennings PR (1966) Inheritance and significance of dwarfism in an indica variety. Crop Sci 6:551–554 Ashikari M, Wu J, Yano M, Sasaki T, Yoshimura A (1999) Rice gibberellin-insensitive dwarf mutant gene Dwarf 1 encodes the alpha-subunit of GTP-binding protein. Proc Natl Acad Sci USA 96:10284–10289 Ashikari M, Sasaki A, Ueguchi-Tanaka M, et al. (2002) Loss-of-function of a rice gibberellin biosynthetic gene, GA20 oxidase (GA20ox-2), led to the rice ‘Green revolution’. Breed Sci 52:143–150 Ashikari M, Sakakibara H, Lin S, et al. (2005) Cytokinin oxidase regulates rice grain production. Science 309:741–745 Brar DS, Khush GS (1997) Alien introgression in rice. Plant Mol Biol 35:35–47 Coles JP, Phillips AL, Croker SJ, Garcia-Lepe R, Lewis MJ, Hedden P (1999) Modification of gibberellin production and plant development in Arabidopsis by sense and antisense expression of gibberellin 20-oxidase genes. Plant J 17:547–556 Dalrymple DG (1986) Development and spread of high-yielding rice varieties in developing countries. Bureau for Science and Technology, Agency for International Development, Washington, DC FAO (2004) The state of food insecurity in the world. Food and Agriculture Organization of the United Nations, http://www.fao.org/sof/sofi/index_en.htm Foster KW, Rutger JN (1978) Inheritance of semi-dwarfism in rice, Oryza sativa L. Genetics 88:559–574 Fu X, Sudhakar D, Peng J, Richards DE, Christou P, Harberd NP (2001) Expression of Arabidopsis GAI in transgenic rice represses multiple gibberellin responses. Plant Cell 13:1791–1802 Futsuhara Y, Toriyama K, Tsunoda K (1967) Breeding of a new rice variety “Reimei” by gammaray irradiation. Jpn J Breed 17:85–90 Hargrove TR, Cabanilla VL (1979) The impact of semi-dwarf varieties on Asian rice-breeding program. BioScience 29:731–735 Hedden P, Phillips AL (2000a) Gibberellin metabolism: new insights revealed by the genes. Trends Plant Sci 5:523–530 Hedden P, Phillips AL (2000b) Manipulation of hormone biosynthetic genes in transgenic plants. Curr Opin Biotech 11:130–137 Hong Z, Ueguchi-Tanaka M, Shimizu-Sato S, et al. (2002) Loss-of-function of a rice brassinosteroid biosynthetic enzyme, C-6 oxidase, prevents the organized arrangement and polar elongation of cells in the leaves and stem. Plant J 32:495–508 Hong Z, Ueguchi-Tanaka M, Umemura K, et al. (2003) A rice brassinosteroid-deficient mutant, ebisu dwarf (d2), is caused by a loss of function of a new member of cytochrome P450. Plant Cell 15:2900–2910 Hong Z, Ueguchi-Tanaka M, Fujioka S, et al. (2005) The rice brassinosteroid-deficient dwarf 2 mutant, defective in the rice homolog of Arabidopsis DIMINUTO/DWARF1, is rescued by the endogenously accumulated alternative bioactive brassinosteroid, dolichosterone. Plant Cell 17:2243–2254 Horton P (2000) Prospects for crop improvement through the genetic manipulation of photosynthesis: morphological and biochemical aspects of light capture. J Exp Bot 51:475–485 Ikeda A, Ueguchi-Tanaka M, Sonoda Y, et al. (2001) Slender rice, a constitutive gibberellin response mutant, is caused by a null mutation of the SLR1 gene, an ortholog of the heightregulating gene GAI/RGA/RHT/D8. Plant Cell 13:999–1010 Itoh H, Ueguchi-Tanaka M, Sentoku N, Kitano H, Matsuoka M, Kobayashi M (2001) Cloning and functional analysis of gibberellin 3β-hydroxylase genes that are differently expressed during the growth of rice. Proc Natl Acad Sci USA 98:8909–8914

IV.2 Rice Yielding and Plant Hormones

319

Itoh, H, Ueguchi-Tanaka M, Sakamoto T, et al. (2002a) Modification of rice plant height by suppressing the height-controlling gene, D18, in rice. Breed Sci 52:215–218 Itoh H, Ueguchi-Tanaka M, Sato Y, Ashikari M, Matsuoka M (2002b) The gibberellin signaling pathway is regulated by the appearance and disappearance of SLENDER RICE1 in nuclei. Plant Cell 14:57–70 Itoh H, Tatsumi T, Sakamoto T, et al. (2004) A rice semi-dwarf gene, Tan-Ginbozu (D35), encodes the gibberellin biosynthesis enzyme, ent-kaurene oxidase. Plant Mol Biol 54:533–547 Khush GS (1999) Green revolution: preparing for the 21st century. Genome 42:646–655 Khush GS (2001) Challenges for meeting the global food and nutrient needs in the new millennium. Proc Nutr Soc 60:15–26 Khush GS (2003) Productivity improvement in rice. Nutr Rev 61:114–116 Kikuchi F, Itakura N, Ikehashi H, Yokoo M, Nakane A, Maruyama K (1985) Genetic analysis of semi-dwarfism in high yielding rice varieties in Japan. Bull Nat Inst Agr Sci Ser D 36:125–145 Mann CC (1999) Crop scientists seek a new revolution. Science 283:310–314 Morinaka Y, Sakamoto T, Inukai Y, et al. (2006) Morphological alteration caused by brassinosteroid insensitivity increases the biomass and grain production of rice. Plant Physiol 141:924–931 Peng J, Richards DE, Hartley NM, et al. (1999) ‘Green revolution’ genes encode mutant gibberellin response modulators. Nature 400:256–261 Sakamoto T, Kobayashi M, Itoh H, et al. (2001) Expression of a gibberellin 2-oxidase gene around the shoot apex is related to phase transition in rice. Plant Physiol 125:1508–1516 Sakamoto T, Morinaka Y, Ishiyama K, et al. (2003) Genetic manipulation of gibberellin metabolism in transgenic rice. Nature Biotech 21:909–913 Sakamoto T, Miura K, Itoh H, et al. (2004) An overview of gibberellin metabolism enzyme genes and their related mutants in rice. Plant Physiol 134:1642–1653 Sakamoto T, Morinaka Y, Ohnishi T, et al. (2006) Erect leaves caused by brassinosteroid deficiency increase biomass production and grain yield in rice. Nature Biotech 24:105–109 Sasaki A, Ashikari M, Ueguchi-Tanaka M, et al. (2002) Green revolution: a mutant gibberellinsynthesis gene in rice. Nature 416:701–702 Sasaki A, Itoh H, Gomi K, et al. (2003) Accumulation of phosphorylated repressor for gibberellin signaling in an F-box mutant. Science 299:1896–1898 Sinclair TR, Sheehy JE (1999) Erect leaves and photosynthesis in rice. Science 283:1456–1457 Suge H (1975) Complementary genes for height inheritance in relation to gibberellin production in rice plants. Jpn J Genet 50:121–131 Suh HS, Heu HM (1978) The segregation mode of plant height in the cross of rice varieties. II. Linkage analysis of the semi-dwarfness of rice variety “Tongil”. Korean J Breed 10:1–6 Tanabe S, Ashikari M, Fujioka S, et al. (2005) A novel cytochrome P450 is implicated in brassinosteroid biosynthesis via the characterization of a rice dwarf mutant, dwarf11, with reduced seed length. Plant Cell 17:776–790 Tanksley SD, McCouch SR (1997) Seed banks and molecular maps: unlocking genetic potential from the wild. Science 277:1063–1066 Toyomasu T, Kawaide H, Sekimoto H, et al. (1997) Cloning and characterization of a cDNA encoding gibberellin 20-oxidase from rice (Oryza sativa) seedlings. Physiol Plant 99:111–118 Ueguchi-Tanaka M, Fujisawa Y, Kobayashi M, et al. (2000) Rice dwarf mutant, d1, which is defective in the α subunit of the heterotrimeric G protein, affects gibberellin signal transduction. Proc Natl Acad Sci USA 97:11638–11643 Ueguchi-Tanaka M, Ashikari M, Nakajima M, et al. (2005) GIBBERELLIN INSENSITIVE DWARF1 encodes a soluble receptor for gibberellin. Nature 437:693–698

320

M. Ashikari and T. Sakamoto

UNFPA (2004) State of the world population. United Nations Population Fund, http://www.unfpa. org/swp/2004/english/ch1/index.htm Werner T, Motyka V, Strnad M, Schmulling T (2001) Regulation of plant growth by cytokinin. Proc Natl Acad Sci USA 98:10487–10492 Werner T, Motyka V, Laucou V, et al. (2003) Cytokinin-deficient transgenic Arabidopsis plants show multiple developmental alterations indicating opposite functions of cytokinins in the regulation of shoot and root meristem activity. Plant Cell 15:2532–2550 White PT (1994) Rice: the essential harvest. Natl Geogr 185:48–79 Yano M (2001) Genetic and molecular dissection of naturally occurring variation. Curr Opin Plant Biol 4:130–135

IV.3

Regulation of Iron and Zinc Uptake and Translocation in Rice Takanori Kobayashi1 and Naoko K. Nishizawa2(* ü)

1

1.1

Iron (Fe) and Zinc (Zn) Uptake from the Rhizosphere Mediated by Mugineic Acid Family Phytosiderophores (MAs) Contribution of MAs to Fe and Zn Uptake

Higher plants take up essential nutrients from the rhizosphere, in which several nutrients tend to be insoluble, thus limiting their availability. Deficiencies of the micronutrients Fe and Zn constitute major factors in low crop yield. Based on their mechanisms of Fe acquisition from the soil, higher plants can be grouped into two categories: Strategy-I and Strategy-II plants (Römheld and Marschner 1986). Plants in the second group, graminaceous plants, secrete mugineic acid family phytosiderophores (MAs), which solubilize Fe(III) in the rhizosphere, and the resulting Fe(III)-MA complexes are taken up by roots through a specific transporter in the plasma membrane (Takagi 1976; Curie et al. 2001). In addition to Fe(III), MAs chelate various divalent cations, including Zn(II) (Murakami et al. 1989). Although Zn has been thought to be taken up mainly as free Zn2+ ions, the uptake of Zn(II)-MAs complexes has also been proposed (Welch 1995; von Wirén et al. 1996). Recently, Suzuki et al. (2006a) showed increased expression of the genes that participate in the biosynthesis and increased secretion of MAs in Zn-deficient barley. Moreover, an analysis using the positron-emitting tracer imaging system (PETIS) confirmed that Zn-deficient barley absorbs both Zn(II)-MAs and Zn2+ ions, with higher uptake rates observed for Zn(II)-MAs

1 Graduate School of Agricultural and Life Sciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan, e-mail: [email protected] (N.K. Nishizawa); [email protected] (T. Kobayashi) 2 Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Corporation, Japan

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

321

322

T. Kobayashi and N.K. Nishizawa

(Suzuki et al. 2006a). However, in rice, the secretion of MAs decreases under Zn deficiency (Suzuki et al. 2005). A PETIS experiment also showed that Zn-deficient rice plants absorb lower amounts of Zn when supplied with MAs, suggesting a relatively low contribution of MAs to Zn uptake in rice (Suzuki et al. 2006b).

1.2

Biosynthesis of MAs

The biosynthetic pathway of MAs has been identified through extensive biochemical and physiological studies (Mori and Nishizawa 1987; Shojima et al. 1990; Ma and Nomoto 1993; Ma et al. 1999; Mori 1999), and the genes for all of the enzymes that participate in MAs biosynthesis have been identified (Table 1). Methionine is the primary precursor of MAs (Mori and Nishizawa 1987). S-adenosyl- l-methionine synthetase (SAMS) converts methionine to S-adenosyl-lmethionine (SAM). Nicotianamine synthase (NAS) subsequently catalyzes the trimerization of SAM molecules, to produce nicotianamine (NA). NAS genes were first isolated in barley (HvNAS1-7) through the purification of enzymes from Fe-deficient barley roots (Higuchi et al. 1999). Rice and maize each possess three NAS genes (OsNAS1-3; Higuchi et al. 2001; ZmNAS1-3; Mizuno et al. 2003). Northern and promoter-β-glucuronidase (GUS) analyses of the rice NAS genes revealed that OsNAS1 and OsNAS2, but not OsNAS3, are strongly induced by Fe deficiency in the root epidermal, exodermal, and cortical cells (Higuchi et al. 2001; Inoue et al. 2003), and are therefore expected to be involved in MAs biosynthesis, contributing to the uptake of Fe from the rhizosphere. In addition, all of the three rice NAS genes (OsNAS1-3) are expressed in the pericycle and companion cells in the roots, the companion cells in the leaves, and reproductive organs, including flowers and maturing seeds (Inoue et al. 2003; Takahashi et al. 2004), and are thought to be involved in Fe translocation within the plant body. Under Zn deficiency, OsNAS1 and OsNAS2 are downregulated in roots, whereas OsNAS3 is upregulated in roots and shoots (Suzuki et al. 2005). NAS and NA are also present in nongraminaceous plants (Noma and Noguchi 1976; Ling et al. 1999), in which NA serves as a common metal chelator involved in the internal transport of various micronutrients, including Fe and Zn (Hell and Stephan 2003; Takahashi et al. 2003). NA aminotransferase (NAAT) catalyzes the first step specific to graminaceous plants: transamination of NA to produce the 3˝-oxo intermediate. NAAT genes were also first isolated from barley (HvNAAT-A and HvNAAT-B; Takahashi et al. 1999), and then from rice (OsNAAT1-5; Inoue et al. 2004a). Among the five rice NAAT genes, only OsNAAT1 is induced by Fe deficiency (Inoue et al. 2004a). The spatial pattern of OsNAAT1 expression is very similar to those of OsNAS1 and OsNAS2 (Inoue et al. 2004a; Takahashi et al. 2004). OsNAAT1 expression is also induced by Zn deficiency in roots and shoots (Suzuki et al. 2005), possibly producing MAs, which would enhance Zn translocation inside the plant body (Suzuki et al. 2006b). The 3″-oxo intermediate produced by NAAT is then reduced to 2′-deoxymugineic acid (DMA) by DMA synthase (DMAS). The DMAS gene of rice, OsDMAS1, was identified through microarray analysis as a member of the aldo-keto reductase superfamily, which is upregulated in Fe-deficient roots (Bashir et al. 2006). Functional

Coding protein

Methylthioriburose-1-phosphate isomerase Ribose-5-phosphate isomerase Dehydorase-enolase-phosphatase Aspartate/tyrosine/aromatic aminotransferase Formate dehydrogenase Adenine phosphoribosyltransferase

OsIDI2 RPI DEP OsIDI4

OsFDH OsAPT1

OsMTK1

S-adenosylmethionine synthetase S-adenosylmethionine synthetase Methylthioadenosine/S-adenosyl homocysteine nucleosidase Methylthioribose kinase

OsSAMS1 OsSAMS2 MTN IDE1, IDE2

-S (R, S); -Fe (R, S) b

-Fe (R, S) -Fe (R, S)

-Fe (R, S) -Fe (R, S) -Fe (R, S) -Fe (R, S)

IRO2

IDE1, IDE2, IRO2 IDE2, IRO2

IDE1

IDE1

IDE2 IDE1, IDE2

IDE1

IDE1, IDE2 IRO2

IDE2, IRO2

Cis-elementd

-Fe (R, S) -Fe (R, S); submergence, ethylene (R, S) -Fe (R) -Fe (R, S) -Fe (R)

-Fe (R, S); -Zn (R, S)

Nicotianamine aminotransferase

OsNAAT1

Deoxymugineic acid synthase OsDMAS1 OsIDI1/OsARD1 2-keto-methylthiobutylic acid synthase

-Fe (R, S) -Fe (R); -Zn (R, S)

Nicotianamine synthase Nicotianamine synthase

-Fe (R, S)

Inductiona

OsNAS2 OsNAS3

MAs and NA biosynthesis Nicotianamine synthase OsNAS1

Gene name

Table 1 Rice genes related to the uptake and translocation of Fe and Zn

Kobayashi et al. (2005) Kobayashi et al. (2005) (continued)

Sauter et al. (2004); Kobayashi et al. (2005) Kobayashi et al. (2005) Kobayashi et al. (2005) Kobayashi et al. (2005) Kobayashi et al. (2005)

Kobayashi et al. (2005) Kobayashi et al. (2005) Kobayashi et al. (2005)

Bashir et al. (2006) Kobayashi et al. (2005) Sauter et al. (2005)

Higuchi et al. (2001); Inoue et al. (2003) Inoue et al. (2003) Inoue et al. (2003); Suzuki et al. (2005) Inoue et al. (2004a); Suzuki et al. (2005)

Reference

IV.3 Regulation of Iron and Zinc Uptake and Translocation in Rice 323

Coding protien

Koike et al. (2004) Ishimaru et al. (2005) Ogo et al. (2006)

-Fe (S) -Zn (R, S) -Fe (R, S)

IDE2

Ramesh et al. (2003) Ramesh et al. (2003)

-Zn (R, S)c -Zn (R, S)c

Bughio et al. (2002) Ishimaru et al. (2006)

-Fe (R) -Fe (R)

Reference Koike et al. (2004); unpublished data Nishizawa (2006); unpublished data

IDE1, IRO2

Cis-elementd

-Fe (R)

Inductiona

R Root; S shoot b Sauter et al. (2004) reported an unchanged expression level during Fe deficiency, whereas Kobayashi et al. (2005) reported induction by Fe deficiency c Ramesh et al. (2003) reported Zn deficiency-induced expression, but Ishimaru et al. (2005) detected no induction by Zn deficiency d IDE1, IDE2, and IRO2 in this column represent the presence of IDE1-like, IDE2-like, and IRO2-binding sequences, respectively, within 1.5-kb upstream regions

a

Fe(III)-MAs uptake and translocation Fe(III)-MAs transporter OsYSL15 Fe(III)-MAs transporter OsYSL18 Fe2+ uptake and translocation Fe2+ transporter OsIRT1 Fe2+ transporter OsIRT2 2+ Zn uptake and translocation Zn2+ transporter OsZIP1 Zn2+ transporter OsZIP3 Fe(II)-NA translocation Fe(II)-NA and Mn(II)-NA transporter OsYSL2 Zn2+ translocation Zn2+ transporter OsZIP4 Fe deficiency-related gene regulation bHLH transcription factor OsIRO2

Gene name

Table 1 (continued)

324 T. Kobayashi and N.K. Nishizawa

IV.3 Regulation of Iron and Zinc Uptake and Translocation in Rice

325

orthologues of OsDMAS1 have also been isolated from barley (HvDMAS1), wheat (TaDMAS1), and maize (ZmDMAS1). The expression of these DMAS1 genes is induced by Fe deficiency. The spatial pattern of OsDMAS1 expression is also similar to those of OsNAS1, OsNAS2, and OsNAAT1 (Bashir et al. 2006). All of the MAs share the same biosynthetic pathway from methionine to DMA, which is then converted into other MAs in some species and cultivars (Mori and Nishizawa 1987; Shojima et al. 1990). Two barley dioxygenase genes, IDS2 and IDS3, are responsible for the hydroxylation of MAs at the 3- and 2′-positions, respectively (Nakanishi et al. 2000; Kobayashi et al. 2001). Rice lacks close homologues of IDS2 and IDS3, and usually secretes only DMA (Nakanishi et al. 2000; Kobayashi et al. 2001). To maintain the constant biosynthesis of MAs in Fe-deficient roots, the methionine cycle works vigorously to meet the increased demand for methionine (Ma et al. 1995). Recent progress in the sequencing of the rice genome has enabled us to identify genes that encode putative enzymes of the methionine cycle (Kobayashi et al. 2005; Suzuki et al. 2006a; Table 1). A microarray analysis revealed that the genes involved in all of the predicted steps of the methionine cycle are upregulated under Fe deficiency (Kobayashi et al. 2005; Table 1). Cross-species microarray analyses showed that barley homologues of these genes are strongly induced by Fe deficiency and are moderately induced by Zn deficiency in roots (Negishi et al. 2002; Suzuki et al. 2006a).

1.3

Secretion of MAs

In contrast to the biosynthetic pathway of MAs, the molecular components involved in the secretion of MAs remain unclear. The secretion of MAs from barley follows a distinct diurnal rhythm (Takagi et al. 1984). In parallel, certain vesicles in the root cells of Fe-deficient barley change in shape (Nishizawa and Mori 1987). These vesicles are thought to be the site of DMA synthesis (Negishi et al. 2002). The diurnal rhythm of the secretion of MAs in rice has not been fully characterized. A microarray analysis identified barley genes that may be related to vesicle transport, including the ras-related GTP-binding protein RIC1 and the ADP-ribosylation factor 1 (ARF1) (Negishi et al. 2002). Northern analysis showed that in rice, homologues of RIC1 and ARF1, as well as the OsNAS and OsNAAT genes, also exhibit diurnal changes in expression (Nozoye et al. 2004). These results suggest that the secretion of MAs from graminaceous roots is under the control of vesicular transport.

1.4

Genes Encoding the Fe(III)-MAs Transporter

The maize yellow stripe 1 (ys1) mutant is defective in the uptake of Fe(III)-MAs and also exhibits decreased uptake of Zn(II)-MAs (von Wirén et al. 1994, 1996).

326

T. Kobayashi and N.K. Nishizawa

YS1 was isolated from maize by Curie et al. (2001), and the gene complements the Fe-uptake defect of a mutant yeast strain when supplied with Fe(III)-DMA. Maize YS1 expression increases in both roots and shoots under Fe deficiency, but is not strongly affected by Zn or Cu deficiency (Curie et al. 2001; Roberts et al. 2004). YS1 functions as a proton-coupled symporter for various DMA-bound metals, including Fe(III), Zn(II), Cu(II), and Ni(II) (Schaaf et al. 2004). YS1 also transports NA-chelated Ni(II), Fe(II), and Fe(III) complexes. In contrast, the recently identified barley HvYS1 transporter has been reported to be highly specific for Fe(III)-MAs, with a rather low transport activity for MAs chelated with Zn(II), Cu(II), Ni(II), or Co(II) (Murata et al. 2006). HvYS1 specifically localizes to the plasma membrane of epidermal cells in Fe-deficient roots. Our search for YS1 homologues in the rice genome database identified 18 putative OsYSL (Oryza sativa YS1-like) genes (Koike et al. 2004). Northern analysis detected the transcripts of OsYSL2, 6, 13, 14, 15, and 16 in roots or leaves of Fe-deficient or Fe-sufficient rice plants. The expression of OsYSL15 and 16 was induced in Fe-deficient roots. Recent electrophysiological and promoter-GUS analyses have shown that OsYSL15 is an Fe(III)-MAs transporter that is expressed in the root epidermis and is responsible for Fe uptake from the rhizosphere (Nishizawa 2006; Inoue et al., unpublished). We are currently analyzing the transport properties and expression patterns of other OsYSL genes.

2 2.1

Uptake of Fe2+ and Zn2+ Ions from the Rhizosphere Genes Encoding Fe2+ Transporters

In non-graminaceous plants, Fe uptake from the rhizosphere is mediated by Fe2+-ion transporters. Eide et al. (1996) isolated the Arabidopsis IRT1 gene, which is the dominant Fe2+ transporter in the Fe-uptake process. Homologues of IRT1, which are present in a wide range of plants, animals, protists, fungi, and bacteria, were named zinc-regulated transporter, iron-regulated transporter-like protein (ZIP) family (Grotz et al. 1998). In spite of being a Strategy-II plant, rice possesses homologues of the Arabidopsis IRT1 gene, OsIRT1 and OsIRT2, the Fe2+-transport capacity of which was demonstrated by functional complementation in yeast (Bughio et al. 2002; Ishimaru et al. 2006). OsIRT1 and OsIRT2 localize to the plasma membrane when transiently expressed in onion epidermal cells. OsIRT1 expression is strongly induced in Fe-deficient roots, and OsIRT2 is expressed similarly with a lower expression level. Promoter-GUS analysis showed that OsIRT1 is mainly expressed in the epidermis, exodermis, and inner layer of the cortex in Fe-deficient roots, as well as in companion cells of shoots. Moreover, a PETIS analysis revealed that rice is able to take up both Fe(III)-DMA and Fe2+. Thus, rice plants possess a system other than the MAs-based Strategy II for Fe uptake (Ishimaru et al. 2006). In contrast to their Fe2+-transporting ability, Fe-deficient rice roots do not induce Fe(III)-chelate reductase activity (Ishimaru et al. 2006), which is a hallmark of the Strategy-I

IV.3 Regulation of Iron and Zinc Uptake and Translocation in Rice

327

response. Such a system to take up Fe2+ directly, without reducing the Fe(III) chelates, seems advantageous for growth in submerged conditions, in which the dominant form of soil Fe is Fe2+ (Bughio et al. 2002; Ishimaru et al. 2006).

2.2

Genes Encoding Zn2+ Transporters

In addition to Fe2+ uptake, the ZIP family transporters are thought to be responsible for the uptake of other divalent metals, including Zn, into the roots of both graminaceous and non-graminaceous plants. Arabidopsis possesses 15 ZIP transporters, among which IRT1 transports Zn2+ in addition to Fe2+ (Korshunova et al. 1999). ZIP1 and ZIP3 transport Zn2+, and the expression of their genes is induced in Zn-deficient roots (Grotz et al. 1998). Rice possesses 12 ZIP transporters (Ishimaru et al. 2005, 2006), of which OsZIP1, OsZIP3, and OsZIP4 transport Zn2+ (Ramesh et al. 2003; Ishimaru et al. 2005). An in situ hybridization analysis detected OsZIP1 and OsZIP3 transcripts in the root epidermis, suggesting their involvement in Zn2+ uptake from the rhizosphere. The expression of OsZIP1 and OsZIP3 was induced from 1 to 4 days after Zn-deficiency treatment (Ramesh et al. 2003), but no induction was reported 2 weeks after Zn-deficiency treatment (Ishimaru et al. 2005). OsZIP4 is strongly upregulated under Zn deficiency, and is thought to be responsible for Zn translocation rather than Zn uptake because of its expression pattern (Ishimaru et al. 2005; Sect. 3.2).

3 3.1

Fe and Zn Translocation Inside the Plant Body Chelator-mediated Translocation of Fe and Zn

The translocation of minerals inside the plant body involves a sequence of processes that require various metal chelators and transporters (Hell and Stephan 2003; Fig. 1). Fe has specific chemical properties, including poor solubility and high reactivity, that require plants to use suitable chelators for its control inside the plant body (Marschner 1995). Zn is less reactive than Fe, because it exists only as Zn(II) and does not participate in oxidoreduction reactions. Nevertheless, plants also utilize Zn chelators to maintain the chemical status of this element. Physiological studies have implicated many types of mineral chelators inside the plant body, including NA, MAs, and organic acids. In all higher plants, NA is thought to be one of the principal metal chelators, including Fe and Zn. The expression of the OsNAS genes in the vascular bundle, including companion cells, strongly suggests that NA plays a role in the long-distance transport of Fe (Inoue et al. 2003). Recent characterization of one OsYSL gene, OsYSL2, further supports the functioning of NA in Fe transport (Koike et al. 2004). OsYSL2 transports Fe(II)-NA and Mn(II)-NA, but does not transport Zn(II)-NA,

328

T. Kobayashi and N.K. Nishizawa

Apoplast / Xylem

Soil Particle

MAs

?

Zn

MAs NA

Companion cell / Phloem

Fe

?

?

Fe

Zn2+

?

?

Zn2+

Fe Met

YSL?

Zn(II)YSL? MAs/NA?

Zn(II)-MAs

YSL?

Fe(III)-MAs

YSL

YSL

Fe2+

IRT

IRT

Fe2+

IRT

ZIP

Zn2+

ZIP

Zn2+

Rhizosphere

ZIP

Root Epidermis / Cortex / Endodermis

Fe(III)-MAs YSL Fe(II)-NA

Vascular cylinder

Fig. 1 Schematic representation of rice components that participate in the uptake and translocation of Fe and Zn. Both Fe and Zn are taken up from the rhizosphere either as chelates with mugineic acid family phytosiderophores (MAs) or as free ions. Black ovals represent transporters responsible for the membrane transport of each chemical form. These influx transporters also serve for Fe and Zn translocation inside the plant body, typically in phloem loading and xylem unloading in the vascular cylinder. In addition to MAs, nicotianamine (NA) serves as an important chelator of Fe and Zn for the internal transport. Xylem loading and phloem unloading of Fe and Zn would require efflux-type transporters, which are little characterized in graminaceous plants. Possible transporters for the secretion of MAs to the rhizosphere, as well as those responsible for intracellular trafficking of Fe and Zn, are yet to be reported in graminaceous plants

Fe(III)-DMA, or Mn(II)-DMA. OsYSL2 localizes to the plasma membrane. An OsYSL2 promoter-GUS analysis showed that OsYSL2 is expressed in root companion cells, phloem cells of leaves and leaf sheaths, and other tissues of Fe-deficient leaves. Strong expression of OsYSL2 was also observed in the vascular bundles of flowers and in developing seeds. These results strongly suggest that OsYSL2 functions as an Fe(II)-NA transporter that is responsible for the phloem transport of Fe to sink tissues, including the grain (Koike et al. 2004). The requirement for NA in the transport of Fe, including phloem loading/ unloading and the distribution of Fe from cells adjacent to the veins to the leaf lamina, has also been demonstrated in non-graminaceous plants through examination of the tomato mutant chloronerva, which lacks NA as a result of a point mutation in the NAS gene (Ling et al. 1999; Hell and Stephan 2003). Moreover, transgenic tobacco plants overexpressing the HvNAAT-A gene exhibit phenotypes similar to that of the chloronerva mutant, owing to the exhaustion of endogenous NA (Takahashi et al. 2003). Characterization of the transformants suggested the essential function of NA in the intercellular and intracellular transport of Fe, Zn, and Cu in both vegetative and reproductive organs.

IV.3 Regulation of Iron and Zinc Uptake and Translocation in Rice

329

In addition to the central role of NA, MAs synthesized by graminaceous plants are thought to be involved in Fe and Zn translocation. The expression of the OsNAAT1 and OsDMAS1 genes in vascular bundles supports this idea (Inoue et al. 2004a; Bashir et al. 2006). Recent studies of the OsYSL family have clarified that OsYSL15 and OsYSL18 encode functional Fe(III)-MAs transporters that translocate Fe (Nishizawa 2006; Aoyama et al., unpublished; Inoue et al., unpublished). In addition, PETIS results suggest roles for MAs in the translocation of Fe and Zn to new leaves and grains in rice (Suzuki et al. 2006a, 2006b; Tsukamoto et al., unpublished). Large amounts of DMA are present in rice phloem sap, in which Fe would be chelated by DMA (Mori et al. 1991). Organic acids are also common metal chelators inside the plant body. In particular, Fe(III)-citrate has long been believed to be the dominant form of Fe in the xylem sap (Tiffin 1966; Brown and Chaney 1971). The Arabidopsis FRD3 gene encodes a multidrug and toxin efflux (MATE)-family transporter that may cause the efflux of an organic acid, most probably citrate, resulting in xylem loading (Durrett et al. 2006). The rice FRD3 homologue OsFRDL1 is also expressed in cells that participate in long-distance transport (Inoue et al. 2004b).

3.2

Free-ion Translocation of Zn and Fe

Unlike Fe, Zn is thought to be easily mobilized inside the plant body as free Zn2+ ions. In rice, the most characterized family involved in the free-ion transport of Zn is the ZIP transporter family. In situ hybridization has revealed OsZIP1 and OsZIP3 transcripts in vascular bundles (Ramesh et al. 2003). We found that OsZIP4 transcripts are strongly induced in Zn-deficient roots and leaves, and are more abundant than those of OsZIP1 and OsZIP3 (Ishimaru et al. 2005). OsZIP4 transcripts were detected in Zn-deficient plants, in particular in vascular bundles and leaf mesophyll cells, but were not detected in root epidermal or exodermal cells, suggesting the role of OsZIP4 in the long-distance transport of Zn. Interestingly, OsZIP4 is also strongly expressed in apical meristems, where the strong requirement of Zn in cell division would cause OsZIP4 expression (Ishimaru et al. 2005). Fe2+ is also transported across the plasma membrane inside the plant body, in harmony with chelating and dechelating events, preventing cells from oxidation stress. A promoter-GUS analysis of OsIRT1 revealed that the gene is expressed in companion cells and the root epidermis (Ishimaru et al. 2006). In general, rice ZIP transporters may possess a stricter selectivity for transporting metals than nongraminaceous ZIP transporters (Ishimaru et al. 2005). OsZIP1, OsZIP3, and OsZIP4 transport Zn2+ but not Fe2+ (Ramesh et al. 2003; Ishimaru et al. 2005). OsIRT1 transports Fe2+ but not Zn2+, Mn2+, or Cu2+ (Ishimaru et al. 2006). OsIRT1 and OsIRT2 also transport Cd2+, which may cause the accumulation of Cd in rice plants under aerobic conditions (Nakanishi et al. 2006). Other transporter families known to transport Fe2+ and/or Zn2+ in higher plants include the natural resistance-associated macrophage protein (NRAMP) family, the

330

T. Kobayashi and N.K. Nishizawa

heavy metal ATPase (HMA) family, and the cation diffusion facilitator (CDF) family (Colangelo and Guerinot 2006), none of which have been well characterized in graminaceous plants, including rice.

4 4.1

Regulation of Expression of Fe Uptake-related Genes General Aspects of Gene Regulation

In order to maintain mineral homeostasis, plants must regulate mineral uptake and translocation through various steps, possibly including the sensing of nutritional status, signal transduction, transcriptional regulation, and protein activation and turnover. The mechanisms that regulate micronutrient metabolism in higher plants are poorly understood. Recently, two basic helix-loop-helix (bHLH) transcription regulators, tomato FER and its Arabidopsis homologue FIT1/FRU/AtbHLH29, were reported to regulate the genes for the Strategy-I response under Fe-deficient conditions (Ling et al. 2002; Colangelo and Guerinot 2006). However, the regulation mechanisms by which FER and FIT1/FRU/AtbHLH29 function in the Fe-deficiency response remain unclear. Rice seems to lack close FER homologues (Ogo et al. 2006). The expression properties of the genes related to the biosynthesis of MAs and various transporters mentioned above suggest that plants contain a finely tuned regulation mechanism for maintaining metal homeostasis. The conditions that trigger the most dramatic expression changes in the expression of the Fe uptake-related genes were thought to be the Fe-deficiency response and spatially regulated expression. Until recently, no Fe deficiency-responsive cis-acting elements had been identified in higher plants. Therefore, we decided to identify such elements.

4.2

Iron Deficiency-responsive Elements Regulate the Genes Involved in Fe Uptake

We analyzed the promoter region of the barley Fe deficiency-responsive IDS2 gene. Transgenic tobacco plants carrying the IDS2 promoter fused to the GUS gene showed strong GUS activity in roots in response to Fe deficiency (Yoshihara et al. 2003). The expression was dominant in the pericycle and endodermis, closely resembling the endogenous expression of the IDS2 gene in native barley (Yoshihara et al. 2003), suggesting that the mechanism that regulates the IDS2 gene is conserved beyond Strategy-II plants. Through intensive deletion and mutation analysis, we identified two novel cis-acting elements, iron deficiency-responsive elements 1 (IDE1) and IDE2 (Kobayashi et al. 2003a); these were the first identified elements related to micronutrient deficiencies in plants. IDE1 and IDE2 synergistically induce Fe deficiency-responsive expression in roots.

IV.3 Regulation of Iron and Zinc Uptake and Translocation in Rice

331

We also found that IDS2 promoter fragments containing IDE1 and IDE2 are able to drive Fe-deficiency-induced expression in roots and leaves when introduced into rice (Kobayashi et al. 2004). Interestingly, several promoters of barley genes responsible for the biosynthesis of MAs, those of HvNAS1, HvNAAT-A, IDS2, and IDS3, which drive nearly root-specific expression in response to Fe deficiency in native barley (Nakanishi et al. 1993, 2000; Okumura et al. 1994; Higuchi et al. 1999; Takahashi et al. 1999), drive Fe deficiency-responsive expression in both roots and leaves when introduced into rice (Higuchi et al. 2001; Kobayashi et al. 2001, 2004; Takahashi et al. 2001). Histochemical analysis of IDS2 and IDS3 expression in rice showed pronounced expression in the root exodermis and vascular tissues of roots and leaves (Kobayashi et al. 2004), strongly resembling the expression patterns of OsNAS1, OsNAS2, OsNAAT1, and OsDMAS1 (Inoue et al. 2003, 2004a; Bashir et al. 2006). These results suggest the conservation of IDE-mediated expression among Fe uptake-related genes. In fact, the promoter regions of the barley HvNAS1, HvNAAT-A, HvNAAT-B, and IDS3 genes and the rice OsNAS1, OsNAS2, OsNAAT1, and OsDMAS1 genes possess sequences homologous to IDE1 and/or IDE2 (Kobayashi et al. 2003a, 2005; Table 1). We also searched for IDE-like sequences in the promoter regions of Fe deficiency-upregulated genes identified by a rice microarray analysis. IDE1- and IDE2-like sequences were overrepresented among the promoters of the genes related to Fe uptake, as compared to other Fe deficiency-induced genes and genes that do not respond to Fe deficiency, suggesting the importance of the IDE-mediated pathway in Fe uptake (Kobayashi et al. 2005).

4.3

Other Elements and Factors Regulating Micronutrient Homeostasis

To understand the molecular mechanisms that regulate Fe acquisition, we also characterized Fe deficiency-induced transcription factors. Microarray analyses revealed the upregulation of several transcription factor genes in barley and rice (Negishi et al. 2002; Ogo et al. 2006), among which a putative bHLH transcription factor gene, IRO2, is of particular interest because of its pronounced transcriptional upregulation by Fe deficiency in shoots and roots of barley and rice (Ogo et al. 2006). IRO2 is highly conserved only among graminaceous plants, and is distinct from FER orthologues in non-graminaceous plants. The core sequence for OsIRO2 binding was experimentally determined to be CACGTGG, a sequence that is overrepresented among Fe deficiency-inducible gene promoters in rice (Ogo et al. 2006). The promoter region of OsIRO2 possesses IDE-like sequences. Thus, rice Fe deficiency-responsive genes appear to be regulated via several pathways. Plants also respond to Fe overload, to prevent oxidative stress to cells. Fe overload induces an accumulation of ferritin, a ubiquitous protein involved in Fe storage in plants, animals, fungi, and bacteria. Plant ferritin (phytoferritin) expression is mainly regulated at the transcriptional level by Fe abundance, and is also affected by oxidative stress, abscisic acid, and developmental control. Through transient

332

T. Kobayashi and N.K. Nishizawa

assays of promoter deletions and gel-retardation assays, Petit et al. (2001) identified a 14-bp cis-acting sequence, iron-dependent regulatory sequence (IDRS), which derepresses the expression of phytoferritin genes via Fe loading. The IDRS-mediated gene expression system seems to be conserved in the maize and Arabidopsis ferritin genes, and is distinct from those of Fe deficiency-induced genes. The expression of Zn deficiency-responsive genes is less well understood than that of Fe deficiency-responsive genes. Although the barley IDS2 gene is induced by Zn or Mn deficiency (Okumura et al. 1994; Suzuki et al. 2006a), this response does not occur in tobacco (Kobayashi et al. 2003b). Also, IDE1 and IDE2 introduced into rice as a pair do not respond to Zn or Mn deficiency (unpublished result), indicating that the Zn-deficiency (and Mn-deficiency) response is mediated through mechanisms distinct from the Fe-deficiency response.

5

Future Perspectives

In this era of genomics, many components that are important in plant mineral nutrition have been identified and characterized. In particular, molecular biological approaches have helped reveal complicated mechanisms involved in Fe and Zn uptake and translocation in rice, but a physiological outlook will still be a prerequisite to observe phenomena in planta. Graminaceous plants possess mechanisms to maintain the homeostasis of Fe and Zn that are distinct from those of other plant species. Further characterization of these mechanisms should pave the way for providing sustainable food production in adverse soils, as well as developing Fe- and Zn-enriched crops for human health.

References Bashir K, Inoue H, Nagasaka S, et al. (2006) Cloning and characterization of deoxymugineic acid synthase genes from graminaceous plants. J Biol Chem 281:32395–32402 Brown JC, Chaney RL (1971) Effect of iron on the transport of citrate into the xylem of soybean and tomatoes. Plant Physiol 47:836–840 Bughio N, Yamaguchi H, Nishizawa NK, Nakanishi H, Mori S (2002) Cloning of iron-regulated metal transporter from rice. J Exp Bot 53:1677–1682 Colangelo EP, Guerinot ML (2006) Put the metal to petal: metal uptake and transport throughout plants. Curr Opin Plant Biol 9:322–330 Curie C, Panaviene Z, Loulergue C, Dellaporta SL, Briat JF, Walker EL (2001) Maize yellow stripe 1 encodes a membrane protein directly involved in Fe(III) uptake. Nature 409:346–349 Durrett T, Gassmann D, Rogers E (2006) Functional characterization of FRD3, a novel organic acid effluxer involved in iron homeostasis. In: Abstracts 13th Int Symp on Iron Nutrition and Interactions in Plants, Montpellier, p. 43 Eide D, Broderius M, Fett J, Guerinot ML (1996) A novel iron-regulated metal transporter from plants identified by functional expression in yeast. Proc Natl Acad Sci USA 93:5624–5628 Grotz N, Fox T, Connolly E, Park W, Guerinot ML, Eide D (1998) Identification of a family of zinc transporter genes from Arabidopsis that respond to zinc deficiency. Proc Natl Acad Sci USA 95:7220–7224

IV.3 Regulation of Iron and Zinc Uptake and Translocation in Rice

333

Hell R, Stephan UW (2003) Iron uptake, trafficking and homeostasis in plants. Planta 216:541–551 Higuchi K, Suzuki K, Nakanishi H, Yamaguchi H, Nishizawa NK, Mori S (1999) Cloning of nicotianamine synthase genes, novel genes involved in the biosynthesis of phytosiderophores. Plant Physiol 119:471–479 Higuchi K, Watanabe S, Takahashi M, et al. (2001) Nicotianamine synthase gene expression differs in barley and rice under Fe-deficient conditions. Plant J 25:159–167 Inoue H, Higuchi K, Takahashi M, Nakanishi H, Mori S, Nishizawa NK (2003) Three rice nicotianamine synthase genes, OsNAS1, OsNAS2, and OsNAS3 are expressed in cells involved in long-distance transport of iron and differentially regulated by iron. Plant J 36:366–381 Inoue H, Suzuki M, Takahashi M, Nakanishi H, Mori S, Nishizawa NK (2004a) Rice nicotianamine aminotransferase gene (NAAT1) is expressed in cells involved in long-distance transport of iron. In: Abstracts 12th Int Symp on Iron Nutrition and Interactions in Plants, Tokyo, p 204 Inoue H, Suzuki M, Takahashi M, Nakanishi H, Mori S, Nishizawa NK (2004b) A rice FRD3-like (OsFRDL1) gene is expressed in the cells involved in long-distance transport. Soil Sci Plant Nutr 50:1133–1140 Ishimaru Y, Suzuki M, Kobayashi T, et al. (2005) OsZIP4, a novel zinc-regulated zinc transporter in rice. J Exp Bot 56:3207–3214 Ishimaru Y, Suzuki M, Tsukamoto T, et al. (2006) Rice plants take up iron as an Fe3+-phytosiderophore and as Fe2+. Plant J 45:335–346 Kobayashi T, Nakanishi H, Takahashi M, Kawasaki S, Nishizawa NK, Mori S (2001) In vivo evidence that Ids3 from Hordeum vulgare encodes a dioxygenase that converts 2 -deoxymugineic acid to mugineic acid in transgenic rice. Planta 212:864–871 Kobayashi T, Nakayama Y, Itai RN, et al. (2003a) Identification of novel cis-acting elements, IDE1 and IDE2, of the barley IDS2 gene promoter conferring iron-deficiency-inducible, rootspecific expression in heterogeneous tobacco plants. Plant J 36:780–793 Kobayashi T, Yoshihara T, Jiang T, et al. (2003b) Combined deficiency of iron and other divalent cations mitigates the symptoms of iron deficiency in tobacco plants. Physiol Plant 119:400–408 Kobayashi T, Nakayama Y, Takahashi M, et al. (2004) Construction of artificial promoters highly responsive to iron deficiency. Soil Sci Plant Nutr 50:1167–1175 Kobayashi T, Suzuki M, Inoue H, et al. (2005) Expression of iron-acquisition-related genes in iron-deficient rice is co-ordinately induced by partially conserved iron-deficiency-responsive elements. J Exp Bot 56:1305–1316 Koike S, Inoue H, Mizuno D, et al. (2004) OsYSL2 is a rice metal-nicotianamine transporter that is regulated by iron and expressed in the phloem. Plant J 39:415–424 Korshunova YO, Eide D, Clark WG, Guerinot ML, Pakrasi HB (1999) The IRT1 protein from Arabidopsis thaliana is a metal transporter with a broad substrate range. Plant Mol Biol 40:37–44 Ling HQ, Koch G, Bäumlein H, Ganal MW (1999) Map-based cloning of chloronerva, a gene involved in iron uptake of higher plants encoding nicotianamine synthase. Proc Natl Acad Sci USA 96:7098–7103 Ling HQ, Bauer P, Bereczky Z, Keller B, Ganal M (2002) The tomato fer gene encoding a bHLH protein controls iron-uptake responses in roots. Proc Natl Acad Sci USA 99:13938–13943 Ma JF, Nomoto K (1993) Two related biosynthetic pathways of mugineic acids in Gramineous plants. Plant Physiol 102:373–378 Ma JF, Shinada T, Matsuda C, Nomoto K (1995) Biosynthesis of phytosiderophores, mugineic acids, associated with methionine cycling. J Biol Chem 270:16549–16554 Ma JF, Taketa S, Chang YC, et al. (1999) Genes controlling hydroxylations of phytosiderophores are located on different chromosomes in barley (Hordeum vulgare L.). Planta 207:590–596 Marschner H (1995) Mineral nutrition of higher plants, 2nd edn. Academic Press, London Mizuno D, Higuchi K, Sakamoto T, Nakanishi H, Mori S, Nishizawa NK (2003) Three nicotianamine synthase genes isolated from maize are differentially regulated by iron nutritional status. Plant Physiol 132:1989–1997

334

T. Kobayashi and N.K. Nishizawa

Mori S (1999) Iron acquisition by plants. Curr Opin Plant Biol 2:250–253 Mori S, Nishizawa N (1987) Methionine as a dominant precursor of phytosiderophores in Graminaceae plants. Plant Cell Physiol 28:1081–1092 Mori S, Nishizawa N, Hayashi H, Chino M, Yoshimura E, Ishihara J (1991) Why are young rice plants highly susceptible to iron deficiency? Plant Soil 130:143–156 Murakami T, Ise K, Hayakawa M, Kamei S, Takagi S (1989) Stabilities of metal complexes of mugineic acids and their specific affinities for iron(III). Chem Lett 12:2137–2140 Murata Y, Ma JF, Yamaji N, Ueno D, Nomoto K, Iwashita T (2006) A specific transporter for iron(III)-phytosiderophore in barley roots. Plant J 46:563–572 Nakanishi H, Okumura N, Umehara Y, Nishizawa NK, Chino M, Mori S (1993) Expression of a gene specific for iron deficiency (Ids3) in the roots of Hordeum vulgare. Plant Cell Physiol 34:401–410 Nakanishi H, Yamaguchi H, Sasakuma T, Nishizawa NK, Mori S (2000) Two dioxygenase genes, Ids3 and Ids2, from Hordeum vulgare are involved in the biosynthesis of mugineic acid family phytosiderophores. Plant Mol Biol 44:199–207 Nakanishi H, Ogawa I, Ishimaru Y, Mori S, Nishizawa NK (2006) Iron deficiency enhances cadmium uptake and translocation mediated by the Fe2+ transporters OsIRT1 and OsIRT2 in rice. Soil Sci Plant Nutr 52:464–469 Negishi T, Nakanishi H, Yazaki J, et al. (2002) cDNA microarray analysis of gene expression during Fe-deficiency stress in barley suggests that polar transport of vesicles is implicated in phytosiderophore secretion in Fe-deficient barley roots. Plant J 30:83–94 Nishizawa NK (2006) OsYSL family transporters involved in the uptake and translocation of iron in rice plants. In: Abstracts 13th Int Symp on Iron Nutrition and Interactions in Plants, Montpellier, p 37 Nishizawa N, Mori S (1987) The particular vesicle appearing in barley root cells and its relation to mugineic acid secretion. J. Plant Nutr 10:1013–1020 Noma M, Noguchi M (1976) Occurrence of nicotianamine in higher plants. Phytochemistry 15:1701–1702 Nozoye T, Itai RN, Nagasaka S, et al. (2004) Diurnal changes in the expression of genes that participate in phytosiderophore synthesis in rice. Soil Sci Plant Nutr 50:1125–1131 Ogo Y, Itai RN, Nakanishi H, et al. (2006) Isolation and characterization of IRO2, a novel ironregulated bHLH transcription factor in graminaceous plants. J Exp Bot 57:2867–2878 Okumura N, Nishizawa NK, Umehara Y, et al. (1994) A dioxygenase gene (Ids2) expressed under iron deficiency conditions in the roots of Hordeum vulgare. Plant Mol Biol 25:705–719 Petit JM, van Wuytswinkel O, Briat JF, Lobréaux S (2001) Characterization of an iron-dependent regulatory sequence involved in the transcriptional control of Atfer1 and Zmfer1 plant ferritin genes by iron. J Biol Chem 276:5584–5590 Ramesh SA, Shin R, Eide DJ, Schachtman DP (2003) Differential metal selectivity and gene expression of two zinc transporters from rice. Plant Physiol 133:126–134 Roberts LA, Pierson AJ, Panaviene Z, Walker EL (2004) Yellow stripe 1. Expanded roles for the maize iron-phytosiderophore transporter. Plant Physiol 135:112–120 Römheld V, Marschner H (1986) Evidence for a specific uptake system for iron phytosiderophore in roots of grasses. Plant Physiol 80:175–180 Sauter M, Cornell KA, Beszteri S, Rzewuski G (2004) Functional analysis of methylthiorivose kinase genes in plans. Plant Physiol 136:4061–4071 Sauter M, Lorbiecke R, Yang BQ, Pochapsky TC, Rzewuski G (2005) The immediate-early ethylene response gene OsARD1 encodes an acireductone dioxygenase involved in recycling of the ethylene precursor S-adenosylmethionine. Plant J 44:718–729 Schaaf G, Ludewig U, Erenoglu BE, Mori S, Kitahara T, von Wirén N (2004) ZmYS1 functions as a proton-coupled symporter for phytosiderophore- and nicotianamine-chelated metals. J Biol Chem 279:9091–9096 Shojima S, Nishizawa NK, Fushiya S, Nozoe S, Irifune T, Mori S (1990) Biosynthesis of phytosiderophores. In-vitro biosynthesis of 2 -deoxymugineic acid from l-methionine and nicotianamine. Plant Physiol 93:1497–1503

IV.3 Regulation of Iron and Zinc Uptake and Translocation in Rice

335

Suzuki M, Ishimaru Y, Inoue H, et al. (2005) 22k microarray analysis of Zn-deficient rice. In: Abstracts Symp on Plant Nutrition for Food Security, Human Health and Environmental Protection, Beijing, pp. 134–135 Suzuki M, Takahashi M, Tsukamoto T, et al. (2006a) Biosynthesis and secretion of mugineic acid family phytosiderophores in zinc-deficient barley. Plant J 48:85–97 Suzuki M, Tsukamoto T, Takahashi M, Nakanishi H, Mori S, Nishizawa NK (2006b) The contribution of mugineic acids in transport and absorption of Zn in graminaceous plants. Plant Cell Physiol 47:s156 Takagi S (1976) Naturally occurring iron-chelating compounds in oat- and rice-root washing. I. Activity measurement and preliminary characterization. Soil Sci Plant Nutr 22:423–433 Takagi S, Nomoto K, Takemoto S (1984) Physiological aspect of mugineic acid, a possible phytosiderophore of graminaceous plants. J Plant Nutr 7:469–477 Takahashi M, Yamaguchi H, Nakanishi H, Shioiri T, Nishizawa NK, Mori S (1999) Cloning two genes for nicotianamine aminotransferase, a critical enzyme in iron acquisition (Strategy II) in graminaceous plants. Plant Physiol 121:947–956 Takahashi M, Nakanishi H, Kawasaki S, Nishizawa NK, Mori S (2001) Enhanced tolerance of rice to low iron availability in alkaline soils using barley nicotianamine aminotransferase genes. Nature Biotech 19:466–469 Takahashi M, Terada Y, Nakai I, et al. (2003) Role of nicotianamine in the intracellular delivery of metals and plant reproductive development. Plant Cell 15:1263–1280 Takahashi M, Inoue H, Ushio Y, Nakanishi H, Mori S, Nishizawa NK (2004) Role of nicotianamine and deoxymugineic acid in plant reproductive development. In: Abstracts 12th Int Symp on Iron Nutrition and Interactions in Plants, Tokyo, p. 220 Tiffin LO (1966) Iron translocation: II. Citrate/iron ratios in plant stem exudates. Plant Physiol 41:515–518 von Wirén N, Mori S, Marschner H, Römheld V (1994) Iron inefficiency in maize mutant ys1 (Zea mays L. cv yellow-stripe) is caused by a defect in uptake of iron phytosiderophores. Plant Physiol 106:71–77 von Wirén N, Marschner H, Römheld V (1996) Roots of iron-efficient maize also absorb phytosiderophore-chelated zinc. Plant Physiol 111:1119–1125 Welch RM (1995) Micronutrient nutrition of plants. Crit Rev Plant Sci 14:49–82 Yoshihara T, Kobayashi T, Goto F, et al. (2003) Regulation of the iron-deficiency responsive gene, Ids2, of barley in tobacco. Plant Biotech 20:33–41

IV.4

Abiotic Stress Takayuki Ohnishi1, Mikio Nakazono1, and Nobuhiro Tsutsumi1(* ü)

1

Introduction

Agriculturally and environmentally important plants are subject to various abiotic stresses, resulting in significant damage to plant growth in the field. Abiotic stresses such as drought, high salinity, and low temperature not only damage current crop species, but also act as a barrier to introducing crop plants into areas not currently being used for agriculture. Plants are distinguished from animals by their inability to escape the surrounding environment. Plants have developed numerous physical and biochemical strategies to cope with adverse conditions. Much progress has been made in the identification and characterization of the mechanisms to perceive external signals and to manifest adaptive responses with proper physical changes. Plants exhibit various responses to these stresses at the molecular, cellular, and whole plant levels. At the molecular level, the perception of environmental stimuli and the subsequent activation of defense responses require a complex interplay of signaling cascades (Fig. 1).

2 2.1

Signal Transduction Sensors

Receptors are the molecules that first perceive stress stimulus and then relay the signal to downstream molecules to initiate the signal transduction pathway. However, it is very difficult to find receptors for respective stresses; so far no

1 Graduate School of Agricultural and Life Sciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan e-mail: [email protected], [email protected]; [email protected]

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

337

338

T. Ohnishi et al.

Fig. 1 Generic pathway for the abiotic stress transduction signal in plants. Examples of signaling components in each of the steps are also shown. Abbreviations are defined in the text. Sn indicates section numbers of this chapter in which corresponding topics are described

receptor component has been confirmed to be a stress sensor in plants. A major difficulty in screening for stress sensors is that each type of stress may cause many effects by a multiplicity of physical and chemical stimuli rather than by an individual stimulus. For example, low temperature may immediately result in mechanical constraints, changes in activities of molecules, and reduced cellular osmotic potential. On the basis of this multiplicity, it is unlikely that there is only one sensor that perceives those stress stimuli and controls all subsequent signaling. There may be multiple primary sensors that perceive the initial stress stimulus. In yeast, a well-known type of membrane protein sensor for osmotic stress perception could be a two-component histidine kinase. Evidence suggests that the

IV.4 Abiotic Stress

339

yeast histidine kinases SLN1 and SHO1 are osmotic stress sensors that regulate desaturase gene expression in response to water potential downshifts. In the genome of Arabidopsis, several putative two-component histidine kinases have been identified, although no evidence has been reported for any of these histidine kinases as osmosensors. An Arabidopsis histidine kinase, AtHK1, can complement mutations in the yeast two-component histidine kinase sensor SLN1, and therefore may be involved in osmotic stress signal transduction in plants. Understanding the in vivo function of AtHK1 and other putative histidine kinases and their relationship to osmotic stressactivated phosphorylation pathways will certainly shed light on osmotic stress signal transduction. In rice, the element of two-component systems concerning the function of stress transduction remains unclear. Doi and colleagues (Doi et al. 2004) showed that an Early heading date 1 (Ehd1) gene codes canonical B type response regulator, which serves as one of the elements in two-component systems, and contributes to flowering promotion. Interestingly, although Ehd1 has a classical receiver domain and a functional GARP DNA-binding motif, it very likely has no ortholog in Arabidopsis. These findings indicate that a novel two-component signaling cascade, which is different from the system in Arabidopsis, is integrated into the pathway of photoperiodic control of flowering in rice. So far no gene or mechanism has been shown to function as an osmotic stress sensor in plants. An interesting question is whether two-component systems involving these genes regulate stress signal transductions in plants in the same way that they do in yeast.

2.2

Second Messengers

A second messenger is any small molecule or ion in the cytoplasm of a cell that is generated in response to a signal received by a cell-surface sensor, and that activates various kinases regulating the activities of other enzymes. Several intracellular signaling molecules are involved in stress signal transduction. These include reactive oxygen species, lipid phosphate-derived signals, and cyclic nucleotide-related signals.

2.2.1

Phospholipid

Membrane phospholipids not only serve structural roles, but also generate a multitude of signal molecules during stress responses. It is suggested that phospholipids, like reactive oxygen species, activate downstream adaptive responses. Three phospholipases, phospholipase C, D, and A2, are capable of generating lipids that act as second messengers. The phospholipase C (PLC) pathway has been best characterized, particularly in nonplant organisms. PLC catalyzes the hydrolysis of phosphatidylinositol 4,5-

This page intentionally blank

340

T. Ohnishi et al.

bisphosphate (PIP2), generating the second messengers inositol 1,4,5-trisphosphate (IP3) and diacylglycerol (DAG). IP3 releases Ca2+ from internal stores, whereas DAG activates protein kinase C. Several studies have shown that in various plant species this pathway may be activated rapidly in response to hyperosmotic stress or exogenous ABA. A rice gene for phosphoinositide-specific-PLC, OsPI-PLC1, is activated in systemic-acquired resistance. A connection between phosphoinositides/phospholipids and stress gene expression has also been demonstrated genetically. The Arabidopsis fry1 mutant, the FRY1 gene, encodes an inositol polyphosphate-1-phosphatase, and results in elevated IP3 accumulation and enhanced induction of gene transcription by osmotic stress and ABA. In transgenic plants expressing an antisense PLC gene or overexpressing an Ins(1,4,5)P3 5-phosphatase gene, IP3 levels decreased and the plants were less sensitive to osmotic stress or ABA in germination and in gene induction. It is interesting that, despite increased IP3 levels and enhanced stress gene expression in fry1, the mutant plants were more susceptible to damage by salt, drought, and freezing stress. Similar compensatory increases in detoxification responses occur in the sos1 mutant where high levels of proline were found. Although the increased stress gene expression or proline level did not complement the stress-sensitive phenotypes of the respective mutants, it is possible that without these compensatory responses, the mutants would be even more susceptible to stress damage.

2.2.2

Abscisic Acid (ABA)

Abiotic stresses are well known to induce abscisic acid (ABA) accumulation, by both activation of synthesis of ABA and inhibition of its degradation. To date, identification of ABA metabolic genes has been revealed in Arabidopsis. In rice, whether and how respective abiotic stress regulates ABA metabolic genes is not clear. The limiting step of ABA biosynthesis is the reaction catalyzed by 9-cis-epoxycarotenoid dioxygenase (NCED). NCED converts 9′-cis-neoxanthin and 9-cisviolaxanthin to xanthoxin. The NCED is also postulated to have a regulatory role in ABA biosynthesis in non-photosynthetic organs (Nambara and MarionPoll 2005). The rice genome has three putative NCED genes (OsNCEDs). Indeed, the NCED gene and its homologous genes are upregulated by abiotic stress, such as drought and salt stress. Other ABA biosynthesis genes such as ABA1, AAO3, and ABA3 are also upregulated by osmotic stress. It is evident that ABA biosynthesis is subjected to abiotic stress regulation at multiple steps. In the case of ABA catabolism, the major pathway is thought to be hydroxylation to 8′-hydroxy ABA by cytochrome P450 followed by spontaneous isomerization into phaseic acid (PA) and reduction to dihydrophaseic acid (DPA). On the other hand, ABA is inactivated by conjugation with sugars. ABA glucose ester (ABA-GE) is one of the predominant ABA conjugates. It is becoming evident that

IV.4 Abiotic Stress

341

in Arabidopsis, ABA catabolism is also required to determine the ABA level in response to environmental conditions. Current efforts to uncover ABA perception mechanisms are mainly focused on putative receptor-linked components or those putative receptor molecules that are regulated by stress or ABA. Previous studies on barley aleurone protoplasts, utilizing microinjected and cell-impermeable ABA, consistently pointed to the existence of a plasma-membrane receptor for ABA, with the possible existence of an additional cytosolic receptor for ABA. Biochemical approaches provide a way to isolate ABA receptors by the identification of ABA-binding proteins that are putative ABA receptors. Recently, the RNA-binding protein FCA, a homologue of an ABA-binding protein ABAP1, was identified as an ABA receptor in the regulation of flowering time in Arabidopsis. Furthermore, Arabidopsis ABAR/CHLH was shown to be an ABA receptor that regulates seed development, post-germination growth, and stomatal aperture. The discovery of these receptors also challenges conventional views of plant hormone signaling and raises intriguing questions regarding the nature of ABA perception and the initiation of their signaling pathway.

2.2.3

Reactive Oxygen Species (ROS)

In recent years, reactive oxygen species (ROS) have been shown to have various roles in the control and regulation of biological processes, such as growth, cell cycle, programmed cell death, hormone signaling, biotic and abiotic stress responses, and development. Being toxic molecules, they are also major second messengers in stress signal transduction. In Arabidopsis, a network of at least 152 genes is involved in managing the level of ROS. This network is highly dynamic and redundant, and encodes ROS-scavenging and ROS-producing proteins. Perturbed ROS levels are perceived by different proteins, enzymes or receptors and modulate different developmental, metabolic, and defense pathways. ROS can be generated by various enzymatic activities, of which the best studied are NADPH oxidases, and are removed by an array of ROS-scavenging enzymes. The reactive oxygen gene network therefore modulates the steady-state level of ROS in different cellular compartments for signaling purposes as well as for protection against oxidative damage. Although ROS receptors are unknown at present, plant cells sense ROS by unidentified receptor proteins and redox-sensitive transcription factors, and also by direct inhibition of phosphatases by ROS. Downstream signaling events associated with ROS sensing involve Ca2+ and Ca2+-binding proteins, such as calmodulin, the activation of G-protein, and the activation of phospholipid signaling, which results in the accumulation of phosphatidic acid. It is possible that the localization of ROS signals in specific cellular sites is similar to that of Ca2+ signals in response to stimuli. It is clear that ROS contributes to stress damage, as evidenced by observations that transgenic plants overexpressing ROS scavengers or mutants with higher ROS scavenging ability show increased tolerance to environmental stresses. Many questions

342

T. Ohnishi et al.

about the ROS network related to its mode of regulation, its protective roles, and its modulation of signaling networks that control stress response remain unanswered.

2.2.4

Calcium

Calcium functions as a versatile messenger in mediating responses to hormones, stress signals, and a variety of developmental cues in plants. Studies using Ca2+ imaging techniques have demonstrated that a wide variety of stimuli rapidly alter cytosolic free Ca2+ ([Ca2+]cyt) levels in plant cells. The fact that almost all signals induce [Ca2+]cyt changes in plants suggests that Ca2+ is widely used in signal transduction mechanisms as compared to other messengers. Therefore, much attention has been given to deciphering the role of Ca2+ and its regulation of downstream components in Ca2+-mediated signal transduction cascades that couple signal to cellular processes in plants. The central dogma of Ca2+ signaling consists of three major “nodes” (generation of Ca2+ changes, recognition of these changes, and transduction) that produce a specific cellular response to a Ca2+ message. Any given Ca2+-mediated cellular process begins with the generation of a signal-specific “Ca2+-signature” in the cytoplasm (or nucleus) by the synchronized activity of channels, pumps, and transporters. The signature acts as a cellular “chemical switch”. Changes in [Ca2+]cyt levels are sensed by a specific set of protein(s) with or without EF-hand motifs termed as “Ca2+-sensors”. Upon activation by Ca2+ binding, Ca2+ sensors regulate the activity/ function of specific protein “targets” in the cell that are involved in producing a signal-specific cellular/physiological/developmental response.

2.3

Phosphorylation

Protein phosphorylation is such a central theme in cell signaling that its involvement in stress adaptation was predicted a long time ago. In fact, several plant protein kinases have now been found to be activated by stress. Although plants accumulate compatible osmolytes for osmotic adjustment, it is unclear whether they use similar membrane proteins, a two-component histidine kinase, and mitogen-activated protein (MAP) kinase cascades, which lead to increased synthesis and accumulation of osmolytes. Because osmotic stress elicits calcium signaling, calcium-dependent protein kinases (CDPKs) are prime candidates that link the calcium signal to downstream responses. A dominant negative form of a CDPK was able to block stress or ABA induction of a reporter gene. Recent studies have identified several CDPKs that are activated by osmotic stress in rice. In rice plants, a membrane-associated CDPK was activated by cold treatment. In addition, overexpression of OsCDPK7 resulted in increased cold and osmotic stress tolerance in rice. It is unclear how CDPK is related to MAPK pathways that are also activated during signal cascades. Results from previous studies in animals and yeast also

IV.4 Abiotic Stress

343

failed to show a clear connection between Ca2+-binding protein/calmodulin and MAPK pathways, though MAPK pathways are responsible for the production of compatible osmolytes and antioxidants. In addition to MAPK pathways, other protein kinases are involved in osmotic stress signal transduction. For example, the Snf1-related protein kinase (SnRK2) family of Arabidopsis is involved in osmotic signaling. In fact, protoplast transient expression assays demonstrated that hyperosmotic and saline stresses activated all SnRK2 proteins, except SnRK2.9. Using the same approach, a recent study identified 10 rice SnRK2s that were activated by salt stress in protoplasts. In addition, all members of the rice SnRK2 family proteins were shown to be activated in response to hyperosmotic stress via phosphorylation. Only some of the osmotically activated SnRK2 proteins were activated by ABA both in Arabidopsis and rice, suggesting the involvement of ABA-dependent and ABA-independent signaling pathways.

2.4

Transcription Factors

To study the regulation of stress-responsive genes, it may be of value to consider them as either “early-response genes” or “delayed-response genes”. Early-response genes are induced within minutes and often transiently. In contrast, delayedresponse genes, which account for the majority of the stress-responsive genes, are activated by stress more slowly (several hours later), and their expression is often sustained. Examples of early-response genes in salt, drought, cold, and ABA regulation include transcription factors, such as the DREB/CBF gene family that activate downstream delayed-response genes. A major research effort is needed to identify transcription factors that bind to the cis-elements in these genes. The upstream transcription factors are typically constitutively expressed and are regulated by stress at the posttranscriptional level and/or by phosphorylation change. Although the signaling pathways are largely unknown, transcriptional activation of some stress-responsive genes is well understood, owing to studies on the RD29A/COR78/LTI78 (responsive to dehydration/cold-regulated/low-temperature-induced) gene of Arabidopsis. The promoter of this gene contains both an ABRE (abscisic acid responsive element) and a DRE/CRT (dehydration-responsive element/C-repeat). ABRE and DRE/CRT are cis-elements that function in ABA-dependent and ABAindependent gene expression in response to stress, respectively (Fig. 2). Transcription factors that bind to DRE/CRT were isolated and named DREB1A/ CBF3 (DRE-binding protein/C-repeat-binding factor), DREB1B/CBF1, DREB1C/ CBF2, DREB2A, and DREB2B in Arabidopsis. Orthologous genes of DREB1/CBF have been isolated in many plant species including rice. Overexpression of DREB1/ CBF was found to activate the DRE/CRT class of genes and also to increase the tolerance of transgenic plants to cold, drought, and salt stresses. In rice, four DREB1/CBF homologous genes and one DREB2 homologous gene, OsDREB1A, OsDREB1B, OsDREB1C, and OsDREB1D, and OsDREB2A, respectively, have

344

T. Ohnishi et al.

been isolated. Overexpression of OsDREB1A in transgenic Arabidopsis resulted in improved salt and freezing stress tolerance. Furthermore, Oh and colleagues reported that constitutive overexpression of DREB1A using the cauliflower mosaic virus (CaMV) 35S promoter in transgenic rice resulted in increased stress tolerance to drought and salt stresses. These observations suggest that similar regulatory systems are conserved in monocots as well as dicots, and that transferring the DRE/ DREB regulon to crop plants can improve their tolerance of drought, salt, and cold stresses. The galactinol synthase gene (AtGolS) is one of the target genes of DREB1A/CBF3. Transgenic Arabidopsis plants overexpressing the AtGolS2 gene accumulated galactinol and raffinose, showed a reduced transpiration rate, and were more tolerant to drought stress than control plants. Several transcription factors that can bind to ABRE have been isolated in Arabidopsis, namely AREB1/ABF2, AREB2/ABF4, AREB2/ABF4, AREB3, ABF1, and ABF3. Overexpression of ABF3 or AREB2/ABF4 in Arabidopsis resulted in ABA hypersensitivity, reduced the transpiration rate, and enhanced drought tolerance. For ABA-responsive transcription, a single copy of ABRE is not sufficient. ABRE and coupling elements such as CE1 and CE3 constitute an ABAresponsive complex in the regulation of wheat HVA1 and HVA22 genes. Two ABRE sequences are necessary for the expression of Arabidopsis RD29B in seeds and for the ABA-responsive expression of RD29B in vegetative tissue. One of these ABRE sequences might function as a coupling element. Either additional copies of the ABRE or coupling elements are necessary for ABA-responsive gene expression. Recently, ABA-activated SnRK2 protein kinases were shown to phosphorylate the conserved regions of AREB/ABFs. These kinases might phosphorylate and activate the AREB/ABF-type proteins both in Arabidopsis and in rice. The rice ABAactivated SnRK2s can phosphorylate a rice ABRE-binding factor, TRAB1. In rice, gene expression profiling using cDNA microarrays or gene chips is a useful approach for analyzing the expression patterns of genes under conditions of drought, cold, and salt stresses. Many transcription factors for genes induced by drought, cold, and salt stress have been found, suggesting that various transcriptional regulatory mechanisms function in these stress signal transduction pathways either cooperatively or separately. For example, the OsNAC6 gene, a member of the NAC transcription factor family, is activated by both biotic and abiotic stresses including cold, salt, drought, ABA, wounding, and jasmonic acide (JA) treatment (Ohnishi et al. 2005).

2.5

Stress-responsive Genes

One way to make sense of the large number of stress-responsive genes is to group them functionally. For salt stress, many of the induced genes function in ionic homeostasis; these include plasma membrane Na+/H+ antiporters for Na+ extrusion, vacuolar Na+/H+ antiporters for Na+ compartmentation in the vacuole, and

IV.4 Abiotic Stress

345

high-affinity K+ transporters for K+ acquisition. By increasing the ion concentration in the vacuole, vacuolar Na+/H+ antiporters also function in osmotic homeostasis. Other salt- or drought-induced genes for osmotic homeostasis include those coding for aquaporins and enzymes in osmolyte biosynthesis. In addition to having a possible role in osmotic adjustment, organic compatible osmolytes seem to function in detoxification or damage prevention or repair. In fact, the majority of salt- and drought-induced genes appear to function in damage limitation or repair. These include a large number of osmolyte biosynthesis genes, LEA/dehydrin-type genes, detoxification enzymes, chaperones, proteases, and ubiquitination-related enzymes.

3

Particular Stresses

3.1

Salt and Drought Stress

Salt and drought stress consists of both ionic and osmotic aspects. In Arabidopsis the ionic aspect caused by salt stress activates the SOS (salt overly sensitive) pathway. Osmotic stress activates several protein kinases including mitogen-activated protein kinases, which may mediate osmotic homeostasis and/or detoxification responses. ABA biosynthesis is regulated by osmotic stress at multiple steps. Both ABA-dependent and ABA-independent osmotic stress signaling first modify constitutively expressed transcription factors, leading to the expression of early response transcriptional regulators, which then activate downstream stress tolerance effector genes (Fig. 2).

3.1.1

The SOS Pathway

When under salt stress, plants maintain a high concentration of K+ and a low concentration of Na+ in the cytosol. They do this by regulating the expression and activity of K+ and Na+ transporters and of H+ pumps that generate the driving force for transport. The SOS pathway plays a central role in coordinating the activities of several of the transport systems (Fig. 3). The SOS pathway was recently determined through a series of genetic, molecular, and biochemical analyses. Salt stress elicits a cytosolic calcium signal. How the calcium signal is different from that triggered by drought, cold, or other stimuli remains a mystery. A myristoylated calcium-binding protein encoded by SOS3 presumably senses the salt-elicited calcium signal and translates it to downstream responses. SOS3 interacts with and activates SOS2, a serine/threonine protein kinase. SOS2 and SOS3 regulate the expression level of SOS1, a salt tolerance effector gene encoding a plasma membrane Na+/H+ antiporter. SOS2 and SOS3 are required for the activation of SOS1 transport activity.

346

T. Ohnishi et al.

Fig. 2 Transcription factors controlling stress-responsive gene expression are shown in ellipses. Cis-elements involved in stress-responsive transcription are shown in straight boxes. Broken arrows indicate putative pathways

Fig. 3 Signaling pathways that regulate the expression and activities of ion transporters to maintain and allow cytoplasmic concentration of Na+ under salt stress and osmotic stress. Excessive Na+ and hyperosmolarity are each perceived by unknown sensors. The Ca2+-responsive SOS3– SOS2 protein kinase pathway mediates Na+ regulation of the expression and activities of Na+ transporters. Hyperosmolarity is proposed to induce the synthesis of ABA, which in turn upregulates the transcription of AtNHX1, an ion-transporter gene

IV.4 Abiotic Stress

347

Vacuolar sequestration of Na + not only lowers Na + concentration in the cytoplasm but also contributes to osmotic adjustment to maintain water uptake from saline solutions. Other organelles, such as plastids and mitochondria, may also accumulate some Na+ and thus contribute to the overall subcellular compartmentation of Na+. In Arabidopsis, the AtNHX family of Na+/H+ antiporters functions in Na compartmentation. AtNHX1 and AtNHX2 are localized in the tonoplast membrane, and their transcript levels are upregulated by ABA or osmotic stress. The transcript levels of vacuolar H+-ATPase components also increase in response to salt stress. Salt tolerance has been shown to be substantially enhanced by overexpression of AtNHX1 in various plants, by overexpression of an Atriplex homologue of AtNHX1 in rice, and by overexpression of the vacuolar H+-pyrophosphatase in Arabidopsis. The SOS pathway is also conserved in rice. The OsSOS1/OsNHA1 shares high similarity to the Na+/H+ antiporter coding gene SOS1 from Arabidopsis and does not have significant similarity to any other genes. Therefore, SOS1 is a single copy gene in rice. The expression of OsSOS1/OsNHA1 was upregulated in rice seedlings under salt stress, whereas it was not induced in rice seedlings treated to drought stress. This result suggests that OsSOS1/OsNHA1 expression might be specially induced by Na+ ion toxicity, not by osmotic stress. OsSOS1/OsNHA1 was phosphorylated and activated by the Arabidopsis SOS2/SOS3 proteins, which strongly suggests that the mechanistic details of the biochemical regulation of SOS1 protein are conserved among these species and that, as a consequence, functional homologues of the SOS2 and SOS3 proteins should exist in rice. SOS2-like protein kinase genes and genes encoding SOS3-like calcium-binding protein are also found in rice. The rice genome has up to 30 CIPK/PKS kinases of the SnRK3 family and 10 CBL/ SCaBP interacting calcium sensors. OsCIPK24 and OsCBL4 acted cooperatively to activate OsSOS1 in yeast cells. They can also be exchanged with their Arabidopsis counterpart to form heterologous protein kinase modules, i.e., both OsSOS1 and AtSOS1. Such heterologous modules suppressed the salt sensitivity of sos2 and sos3 mutants of Arabidopsis. These results demonstrate that the SOS pathway operates in cereals and that SOS proteins from dicots and monocots have a high degree of structural similarity.

3.2

Osmotic Stress

The accumulation of compatible solutes, such as proline and betaines, is a widespread response that may protect against osmotic stress. Although an adaptive role for these compounds in mediating osmotic adjustment and protecting subcellular structure has become a central dogma in stress physiology, the evidence in favor of this hypothesis is largely indirect. Betaines are quaternary ammonium compounds in which the nitrogen atom is fully methylated. One of the most common betaines in plants is glycine betaine (GB). GB is widely distributed as a kind of osmolyte in a number of species of

348

T. Ohnishi et al.

plants and is synthesized in many plant species at elevated rates in response to osmotic stress. Whereas several taxonomically distant species are accumulators of GB, others such as Arabidopsis, rice, and tobacco are considered to be nonaccumulators. In higher plants, GB is synthesized in chloroplasts from betaine aldehyde by betaine aldehyde dehydrogenase (BADH). BADH gene expression is induced by salt, drought, cold, and ABA treatment. This study suggested that the accumulation of GB is possibly affected by ABA. Transgenic plants of various species, which accumulate GB at various levels, exhibit enhanced tolerance of several types of stress. Several GB-producing transgenic rice lines were generated in which the Arthrobacter pascens choline oxidase gene, fused to a chloroplast targeting sequence, was expressed under the control of an ABA-inducible promoter. In this transgenic rice, statistically greater levels of stress tolerance were found. Not only plants but also many organisms produce proline. Proline has functions in many processes, such as osmotic regulation, the protection of cell membranes, and detoxification of ROS. Transgenic plants that have elevated proline levels were found to have improved osmotic stress tolerance. An enzyme called ∆1-pyrroline5-carboxylate synthetase (P5CS) plays a key role in the production of proline. The induction of OsP5CS mRNA also preceded the accumulation of proline under highsalt conditions in rice as in Arabidopsis. In comparing rice cultivars under high salt conditions, P5CS transcript and proline levels steadily increased in a salt-tolerant cultivar, Dee-geo-woo-gen (DGWG), whereas they increased only slightly in a saltsensitive cultivar, IR28. Rice plants transformed with P5CS under the control of a stress-inducible promoter overproduced P5CS enzyme and proline under stress, and showed improved growth under salt-stress and water-stress conditions. Osmotic stress-induced genes include not only osmolyte biosynthesis genes, but also a large number of LEA/dehydrin-type genes, detoxification enzymes, chaperones, proteases, and ubiquitination-related enzymes. These genes all contribute to osmotic stress tolerance.

3.3

Cold Stress

The role of ABA in cold responses is still unclear. Only a few years ago, ABA was thought to have a major role in cold responses. Lang and colleagues found transiently increased ABA accumulation in response to chilling treatment of Arabidopsis. Exogenous ABA application increased the freezing tolerance of plants. Furthermore, cold and ABA induce a common set of genes. However, it is currently thought that cold stress induction also has an ABA-independent pathway (Fig. 2). DRE, which is an essential cis-element for ABA-independent pathways, is also found in the promoter regions of many cold-inducible genes of Arabidopsis. Expression of the DREB1/CBF genes is induced by cold, but not by dehydration and salt stresses. By contrast, expression of the DREB2 gene is induced by dehydration and salt stresses but not by cold stress. Indeed, transgenic Arabidopsis plants overexpressing DREB1B/CBF1 or DREB1A/CBF3 under control of the CaMV

IV.4 Abiotic Stress

349

35S promoter showed a high tolerance to freezing stress. More than 40 genes downstream of DREB1/CBF have been identified through the use of both cDNA and GeneChip microarray. A gene for a transcription factor, ICE1 (inducer of CBF expression 1), was identified through map-based cloning of the Arabidopsis ice1 mutation. ICE1 encodes a MYC-like bHLH protein that regulates the expression of DREB1A/CBF3 but not that of other DREB1/CBF genes, indicating that there are different expression mechanisms among the three DREB1/CBF genes (Fig. 2).

3.4

Submergence

Oxygen deprivation induced by submergence, flooding, and waterlogging is an environmental stress that affects the growth of plants and production of crops. Rice is one of the few crops that can survive under complete submergence for an extended period of time because it can rapidly change its metabolic pathways, such as from oxidative phosphorylation to glycolysis and ethanolic fermentation. Two of the ecotypes of rice, deepwater rice and lowland rice, use different strategies to overcome submergence. When deepwater rice varieties are submerged, their internodes elongate rapidly to allow the leaf tips to reach the water surface where they can take up oxygen and thus avoid complete submergence. On the other hand, lowland rice is cultivated in rain-fed areas that are often subjected to flash flooding. Some lowland rice cultivars, such as FR13A, can survive submergence by suppressing shoot elongation, because excess shoot elongation in lowland rice not only uses up the energy supply but also results in lodging and less resistance to pests after recovery from submergence. Submergence-1 (Sub1), which is derived from the cultivar FR13A, is a major quantitative trait locus contributing to great submergence tolerance and is located on chromosome 9. Xu et al. (2006) identified three sequentially arrayed genes (designated Sub1A, Sub1B, and Sub1C) that encode a putative ethylene response factor (ERF) at the Sub1 locus. Among the three genes, Sub1B and Sub1C are invariably present in the Sub1 locus in both submergencetolerant and submergence-intolerant rice cultivars. In contrast, the presence of Sub1A is variable: submergence-tolerant indica cultivars have one variant (Sub1A-1), submergence-intolerant indica cultivars have another variant (Sub1A-2), and japonica cultivars have no Sub1A gene. Furthermore, overexpression of Sub1A-1 in a submergence-intolerant japonica cultivar was shown to enhance tolerance to submergence by showing repressed leaf elongation, chlorophyll degradation, and carbohydrate consumption. These results indicate that Sub1A-1 is a primary determinant of submergence tolerance in the Sub1 locus. In addition to the abilities of tolerance during submergence, abilities of avoidance and alleviation of post-submergence injury are important for the survival of plants subjected to anaerobic stresses. Post-submergence injury appears to be caused by ROS and acetaldehyde. Examples of ROS are the superoxide radical (O2 .–), the hydroxyl radical (OH.), and hydrogen peroxide (H2O2). The production of ROS

350

T. Ohnishi et al.

is immediately induced upon exposure of submerged plant tissues to normal oxygen tension and, as a result, proteins, nucleic acids, and membranes can undergo severe peroxidation. Plants have some defense mechanisms for scavenging ROS after re-aeration. Superoxide dismutase, ascorbate peroxidase, and catalase (CAT) are mainly engaged in the detoxification of ROS in plants. Re-aeration also induces the production of acetaldehyde, as a result of the oxidation of ethanol, which is produced and accumulated by ethanolic fermentation under anaerobic conditions; Fig. 4. Ethanol is assumed to be rapidly oxidized to acetaldehyde by alcohol dehydrogenase (ADH) and/or CAT. CAT probably oxidizes ethanol through its reduction of H2O2 that is produced during re-aeration. Boamfa et al. (2005) reported that the submergence-tolerant cultivar FR13A probably utilized more H2O2 (for conversion of ethanol to acetaldehyde) than the submergence-intolerant cultivar CT6241 under micro-aerobic conditions (i.e., 0.05% O2) and following re-aeration, thereby diverting some of the superoxide radical (O2 .–) away from lipid peroxidation (Fig. 4). As a result, FR13A produces more acetaldehyde during and after micro-aerobic exposure than does CT6241. Thus, mechanisms to metabolize acetaldehyde to lesstoxic compounds may also help to avoid and alleviate post-submergence injury. One of the mechanisms is the conversion of acetaldehyde to less toxic acetate by aldehyde dehydrogenase (ALDH) (Fig. 4). Rice has two mitochondrial ALDHs, ALDH2a and ALDH2b, which are responsible for this conversion. Under submerged

Fig. 4 Proposed metabolic pathways of rice under submerged conditions and following re-aeration. Under submerged conditions (left), pyruvate, which is produced by pathways such as glycolysis, is converted to acetaldehyde by pyruvate decarboxylase (PDC). At the same time, acetaldehyde is converted to ethanol by alcohol dehydrogenase (ADH) and to acetate by aldehyde dehydrogenase (ALDH). When the submerged plants are transferred to aerobic conditions (right), the anaerobically accumulated ethanol is rapidly oxidized to acetaldehyde by the reverse reaction of ADH and/or peroxidation of ethanol by catalase (CAT) during the conversion of H2O2 to H2O. The pathway catalyzed by CAT probably enhances the superoxide dismutase (SOD)-mediated diversion of the superoxide radical (O2 .–) from lipid peroxidation. Moreover, our evidence suggests that mitochondrial ALDH rapidly increases under re-aeration to detoxify acetaldehyde

IV.4 Abiotic Stress

351

conditions, the amount of ALDH2a transcripts increased whereas the amount of ALDH2b transcripts decreased. Interestingly, re-aerated rice plants showed an intense induction of ALDH2a protein despite a decline of ALDH2a mRNA. During re-aeration of rice, acetaldehyde-oxidizing ALDH activity increased along with the increase of ALDH2a protein, thereby causing the acetaldehyde content to decrease. These findings suggest that rice ALDH2a mRNA is accumulated in order to quickly metabolize acetaldehyde that is produced upon re-aeration, and that mitochondrial ALDH is involved in the alleviation of post-anoxic injury induced by acetaldehyde.

4

Future Perspectives

As stress signal transduction is one of the most active research fields in plant biology, stress response mechanisms are being elucidated on a daily basis. Recent studies have suggested that not only ABA but also other phytohormones are included in stress response. A number of unknown factors, such as receptors, kinases, transcription factors, and phytohormones, are also thought to be involved. In the future, such new factors and pathways will be identified and the relationships between novel pathways and well-known pathways will be uncovered. As research progresses, it is expected that many new findings about signal transduction in plants will facilitate progress in agronomy. Many transgenic crops are now commercially available. For example, some crops have transgenes that provide resistance to the herbicide Roundup (chemical name: glyphosate). Transgenic approaches to improving abiotic stress tolerance are considered feasible. Previously, Arabidopsis, a model plant, played a leading role in revealing stress tolerance pathways. However, it is hoped that other crops including rice will be utilized as a material for studies. Because transgenic Arabidopsis has been found to have improved survival under stress conditions, by using crops for the research material we will be able to improve crop production under such conditions based on understanding stress signal transduction. In addition to previous studies, Arabidopsis continues to provide paradigms for testing in rice in order to assess function across taxonomic divisions and in a crop species. For example, the SOS salt tolerance pathway identified in Arabidopsis operates in cereals, and a high degree of structural conservation among the SOS proteins from dicots and monocots is shown. As plants are usually affected by simultaneous stresses, a common goal is to produce a multistress-tolerant phenotype. This chapter focuses only on abiotic stress, but plants undergo continuous exposure to many kinds of biotic stresses in the field. Plants have evolved intricate mechanisms for responding to biotic and abiotic stresses. We need to exploit and improve these mechanisms to allow plants, mainly crops, to survive under multistressed environments.

352

T. Ohnishi et al.

References Apel K, Hirt H (2004) Reactive oxygen species: metabolism, oxidative stress, and signal transduction. Annu Rev Plant Biol 55:373–399 Apse MP, Blumwald E (2002) Engineering salt tolerance in plants. Curr Opin Biotechnol 13:146–150 Apse MP, Aharon GS, Snedden WA, Blumwald E (1999) Salt tolerance conferred by overexpression of a vacuolar Na+/H+ antiport in Arabidopsis. Science 285:1256–1258 Blokhina OB, Fagerstedt KV, Chirkova TV (1999) Relationships between lipid peroxidation and anoxia tolerance in a range of species during post-anoxic reaeration. Physiol Plantarum 105:625–632 Boamfa EI, Veres AH, Ram PC, Jackson MB, Reuss J, Harren FJM (2005) Kinetics of ethanol and acetaldehyde release suggest a role for acetaldehyde production in tolerance of rice seedlings to micro-aerobic conditions. Annals Botany 96:727–736 Boudsocq M, Barbier-Brygoo H, Lauriere C (2004) Identification of nine sucrose nonfermenting 1-related protein kinases 2 activated by hyperosmotic and saline stresses in Arabidopsis thaliana. J Biol Chem 279:41758–41766 Chen HH, Li PH, Brenner ML (1983) Involvement of abscisic acid in potato cold acclimation. Plant Physiol 71:362–365 Chen TH, Murata N (2002) Enhancement of tolerance of abiotic stress by metabolic engineering of betaines and other compatible solutes. Curr Opin Plant Biol 5:250–257 Coelho SM, Taylor AR, Ryan KP, Sousa-Pinto I, Brown MT, Brownlee C (2002) Spatiotemporal patterning of reactive oxygen production and Ca(2+) wave propagation in fucus rhizoid cells. Plant Cell 14:2369–2381 Cutler AJ, Krochko JE (1999) Formation and breakdown of ABA. Trends Plant Sci 4:472–478 Dietz KJ, Tavakoli N, Kluge C, et al. (2001) Significance of the V-type ATPase for the adaptation to stressful growth conditions and its regulation on the molecular and biochemical level. J Exp Bot 52:1969–1980 Doi K, Izawa T, Fuse T, et al. (2004) Ehd1, a B-type response regulator in rice, confers short-day promotion of flowering and controls FT-like gene expression independently of Hd1. Genes Dev 18:926–936 Dubouzet JG, Sakuma Y, Ito Y, et al. (2003) OsDREB genes in rice, Oryza sativa L., encode transcription activators that function in drought-, high-salt- and cold-responsive gene expression. Plant J 33:751–763 Fowler S, Thomashow MF (2002) Arabidopsis transcriptome profiling indicates that multiple regulatory pathways are activated during cold acclimation in addition to the CBF cold response pathway. Plant Cell 14:1675–1690 Fukao T, Xu KN, Ronald PC, Bailey-Serres J (2006) A variable cluster of ethylene response factor-like genes regulates metabolic and developmental acclimation responses to submergence in rice(W). Plant Cell 18:2021–2034 Igarashi Y, Yoshiba Y, Sanada Y, Yamaguchi-Shinozaki K, Wada K, Shinozaki K (1997) Characterization of the gene for delta 1-pyrroline-5-carboxylate synthetase and correlation between the expression of the gene and salt tolerance in Oryza sativa L. Plant Mol Biol 33:857–865 Ishitani M, Nakamura T, Han SY, Takabe T (1995) Expression of the betaine aldehyde dehydrogenase gene in barley in response to osmotic stress and abscisic acid. Plant Mol Biol 27:307–315 Jackson MB, Ram PC (2003) Physiological and molecular basis of susceptibility and tolerance of rice plants to complete submergence. Annals Botany 91:227–241 Kende H, van der Knaap E, Cho HT (1998) Deepwater rice: a model plant to study stem elongation. Plant Physiol 118:1105–1110

IV.4 Abiotic Stress

353

Kim KN, Lee JS, Han H, Choi SA, Go SJ, Yoon IS (2003) Isolation and characterization of a novel rice Ca2+-regulated protein kinase gene involved in responses to diverse signals including cold, light, cytokinins, sugars and salts. Plant Mol Biol 52:1191–1202 Knight H, Trewavas AJ, Knight MR (1997) Calcium signalling in Arabidopsis thaliana responding to drought and salinity. Plant J 12:1067–1078 Kobayashi Y, Yamamoto S, Minami H, Kagaya Y, Hattori T (2004) Differential activation of the rice sucrose nonfermenting1-related protein kinase2 family by hyperosmotic stress and abscisic acid. Plant Cell 16:1163–1177 Kobayashi Y, Murata M, Minami H, et al. (2005) Abscisic acid-activated SNRK2 protein kinases function in the gene-regulation pathway of ABA signal transduction by phosphorylating ABA response element-binding factors. Plant J 44:939–949 Kolukisaoglu U, Weinl S, Blazevic D, Batistic O, Kudla J (2004) Calcium sensors and their interacting protein kinases: genomics of the Arabidopsis and rice CBL-CIPK signaling networks. Plant Physiol 134:43–58 Kreuzwieser J, Harren FJM, Laarhoven LJJ, et al. (2001) Acetaldehyde emission by the leaves of trees – correlation with physiological and environmental parameters. Physiol Plantarum 113:41–49 Lang V, Mantyla E, Welin B, Sundberg B, Palva ET (1994) Alterations in water status, endogenous abscisic acid content, and expression of rab18 gene during the development of freezing tolerance in Arabidopsis thaliana. Plant Physiol 104:1341–1349 Liu J, Zhu JK (1997) Proline accumulation and salt-stress-induced gene expression in a salthypersensitive mutant of Arabidopsis. Plant Physiol 114:591–596 Liu Q, Kasuga M, Sakuma Y, et al. (1998) Two transcription factors, DREB1 and DREB2, with an EREBP/AP2 DNA binding domain separate two cellular signal transduction pathways in drought- and low-temperature-responsive gene expression, respectively, in Arabidopsis. Plant Cell 10:1391–1406 Lovegrove A, Hooley R (2000) Gibberellin and abscisic acid signalling in aleurone. Trends Plant Sci 5:102–110 Martin ML, Busconi L (2001) A rice membrane-bound calcium-dependent protein kinase is activated in response to low temperature. Plant Physiol 125:1442–1449 Martinez-Atienza J, Jiang X, Garciadeblas B, et al. (2006) Conservation of the SOS salt tolerance pathway in rice. Plant Physiol Maruyama K, Sakuma Y, Kasuga M, et al. (2004) Identification of cold-inducible downstream genes of the Arabidopsis DREB1A/CBF3 transcriptional factor using two microarray systems. Plant J 38:982–993 Mittler R, Vanderauwera S, Gollery M, Van Breusegem F (2004) Reactive oxygen gene network of plants. Trends Plant Sci 9:490–498 Monk LS, Braendle R, Crawford RMM (1987) Catalase activity and postanoxic injury in monocotyledonous species. J Exp Botany 38:233–246 Nakashima K, Shinwari ZK, Sakuma Y, et al. (2000) Organization and expression of two Arabidopsis DREB2 genes encoding DRE-binding proteins involved in dehydration- and high-salinity-responsive gene expression. Plant Mol Biol 42:657–665 Nakazono M, Tsuji H, Li YH, et al. (2000) Expression of a gene encoding mitochondrial aldehyde dehydrogenase in rice increases under submerged conditions. Plant Physiol 124:587–598 Nambara E, Marion-Poll A (2005) Abscisic acid biosynthesis and catabolism. Annu Rev Plant Biol 56:165–185 Nanjo T, Kobayashi M, Yoshiba Y, Kakubari Y, Yamaguchi-Shinozaki K, Shinozaki K (1999) Antisense suppression of proline degradation improves tolerance to freezing and salinity in Arabidopsis thaliana. FEBS Lett 461:205–210 Nuccio ML, Rhodes D, McNeil SD, Hanson AD (1999) Metabolic engineering of plants for osmotic stress resistance. Curr Opin Plant Biol 2:128–134 Oh SJ, Song SI, Kim YS, et al. (2005) Arabidopsis CBF3/DREB1A and ABF3 in transgenic rice increased tolerance to abiotic stress without stunting growth. Plant Physiol 138:341–351

354

T. Ohnishi et al.

Ohnishi T, Sugahara S, Yamada T, et al. (2005) OsNAC6, a member of the NAC gene family, is induced by various stresses in rice. Genes Genet Syst 80:135–139 Ohta M, Hayashi Y, Nakashima A, et al. (2002) Introduction of a Na+/H+ antiporter gene from Atriplex gmelini confers salt tolerance to rice. FEBS Lett 532:279–282 Pavelic D, Arpagaus S, Rawyler A, Brandle R (2000) Impact of post-anoxia stress on membrane lipids of anoxia-pretreated potato cells. A re-appraisal. Plant Physiol 124:1285–1292 Pfistersieber M, Brandle R (1994) Aspects of plant behavior under anoxia and postanoxia. Proc R Soc Edinb Sect B – Biol Sci 102:313–324 Rabbani MA, Maruyama K, Abe H, et al. (2003) Monitoring expression profiles of rice genes under cold, drought, and high-salinity stresses and abscisic acid application using cDNA microarray and RNA gel-blot analyses. Plant Physiol 133:1755–1767 Razem FA, El-Kereamy A, Abrams SR, Hill RD (2006) The RNA-binding protein FCA is an abscisic acid receptor. Nature 439:290–294 Razem FA, Luo M, Liu JH, Abrams SR, Hill RD (2004) Purification and characterization of a barley aleurone abscisic acid-binding protein. J Biol Chem 279:9922–9929 Razem FA, El-Kereamy A, Abrams SR, Hill RD (2006) The RNA-binding protein FCA is an abscisic acid receptor. Nature 439:290–294 Reddy VS, Reddy AS (2004) Proteomics of calcium-signaling components in plants. Phytochemistry 65:1745–1776 Reiser V, Raitt DC, Saito H (2003) Yeast osmosensor Sln1 and plant cytokinin receptor Cre1 respond to changes in turgor pressure. J Cell Biol 161:1035–1040 Saijo Y, Hata S, Kyozuka J, Shimamoto K, Izui K (2000) Over-expression of a single Ca 2+-dependent protein kinase confers both cold and salt/drought tolerance on rice plants. Plant J 23:319–327 Sakamoto A, Murata N (2002) The role of glycine betaine in the protection of plants from stress: clues from transgenic plants. Plant Cell Environ 25:163–171 Sanchez JP, Chua NH (2001) Arabidopsis PLC1 is required for secondary responses to abscisic acid signals. Plant Cell 13:1143–1154 Sauter M (2000) Rice in deep water: “how to take heed against a sea of troubles”. Naturwissenschaften 87:289–303 Scrase-Field SA, Knight MR (2003) Calcium: just a chemical switch? Curr Opin Plant Biol 6:500–506 Shen Q, Zhang P, Ho TH (1996) Modular nature of abscisic acid (ABA) response complexes: composite promoter units that are necessary and sufficient for ABA induction of gene expression in barley. Plant Cell 8:1107–1119 Shen YY, Wang XF, Wu FQ, et al. (2006) The Mg-chelatase H subunit is an abscisic acid receptor. Nature 443:823–826 Shinozaki K, Yamaguchi-Shinozaki K (2000) Molecular responses to dehydration and low temperature: differences and cross-talk between two stress signaling pathways. Curr Opin Plant Biol 3:217–223 Song FM, Goodman RM (2002) Molecular cloning and characterization of a rice phosphoinositidespecific phospholipase C gene, OsPI-PLC1, that is activated in systemic acquired resistance. Physiol Molec Plant Pathol 61:31–40 Su J, Hirji R, Zhang L, He C, Selvaraj G, Wu R (2006) Evaluation of the stress-inducible production of choline oxidase in transgenic rice as a strategy for producing the stress-protectant glycine betaine. J Exp Bot 57:1129–1135 Taji T, Ohsumi C, Iuchi S, et al. (2002) Important roles of drought- and cold-inducible genes for galactinol synthase in stress tolerance in Arabidopsis thaliana. Plant J 29:417–426 Tan BC, Joseph LM, Deng WT, et al. (2003) Molecular characterization of the Arabidopsis 9-cis epoxycarotenoid dioxygenase gene family. Plant J 35:44–56 Thomashow MF (1999) PLANT COLD ACCLIMATION: freezing tolerance genes and regulatory mechanisms. Annu Rev Plant Physiol Plant Mol Biol 50:571–599

IV.4 Abiotic Stress

355

Tsuji H, Meguro N, Suzuki Y, Tsutsumi N, Hirai A, Nakazono M (2003a) Induction of mitochondrial aldehyde dehydrogenase by submergence facilitates oxidation of acetaldehyde during re-aeration in rice. FEBS Lett 546:369–373 Tsuji H, Tsutsumi N, Sasaki T, Hirai A, Nakazono M (2003b) Organ-specific expressions and chromosomal locations of two mitochondrial aldehyde dehydrogenase genes from rice (Oryza sativa L.), ALDH2a and ALDH2b. Gene 305:195–204 Uno Y, Furihata T, Abe H, Yoshida R, Shinozaki K, Yamaguchi-Shinozaki K (2000) Arabidopsis basic leucine zipper transcription factors involved in an abscisic acid-dependent signal transduction pathway under drought and high-salinity conditions. Proc Natl Acad Sci USA 97:11632–11637 Urao T, Miyata S, Yamaguchi-Shinozaki K, Shinozaki K (2000) Possible His to Asp phospho relay signaling in an Arabidopsis two-component system. FEBS Lett 478:227–232 Xiong L, Lee B, Ishitani M, Lee H, Zhang C, Zhu JK (2001) FIERY1 encoding an inositol polyphosphate 1-phosphatase is a negative regulator of abscisic acid and stress signaling in Arabidopsis. Genes Dev 15:1971–1984 Xiong L, Schumaker KS, Zhu JK (2002) Cell signaling during cold, drought, and salt stress. Plant Cell 14 Suppl:S165–S183 Xu K, Xu X, Fukao T, et al. (2006) Sub1A is an ethylene-response-factor-like gene that confers submergence tolerance to rice. Nature 442:705–708 Xu KN, Mackill DJ (1996) A major locus for submergence tolerance mapped on rice chromosome 9. Molec Breed 2:219–224 Yamaguchi-Shinozaki K, Shinozaki K (2006) Transcriptional regulatory networks in cellular responses and tolerance to dehydration and cold stresses. Annu Rev Plant Biol 57:781–803 Yokoi S, Quintero FJ, Cubero B, et al. (2002) Differential expression and function of Arabidopsis thaliana NHX Na+/H+ antiporters in the salt stress response. Plant J 30:529–539 Zhang DP, Wu ZY, Li XY, Zhao ZX (2002) Purification and identification of a 42-kilodalton abscisic acid-specific-binding protein from epidermis of broad bean leaves. Plant Physiol 128:714–725 Zhou GA, Jiang Y, Yang Q, Wang JF, Huang J, Zhang HS (2006) Isolation and characterization of a new Na+/H+ antiporter gene OsNHA1 from rice (Oryza sativa L.). DNA Seq 17:24–30 Zhu BC, Su J, Chan MC, Verma DPS, Fan YL, Wu R (1998) Overexpression of a delta(1)-pyrroline-5-carboxylate synthetase gene and analysis of tolerance to water- and salt-stress in transgenic rice. Plant Sci 139:41–48 Zhu JK (2001) Plant salt tolerance. Trends Plant Sci 6:66–71 Zhu JK (2003) Regulation of ion homeostasis under salt stress. Curr Opin Plant Biol 6:441–445 Zuckermann H, Harren FJM, Reuss J, Parker DH (1997) Dynamics of acetaldehyde production during anoxia and post-anoxia in red bell pepper studied by photoacoustic techniques. Plant Physiol 113:925–932

IV.5

Health-promoting Transgenic Rice: Application of Rice Seeds as a Direct Delivery System for Bioactive Peptides in Human Health Fumio Takaiwa1(* ü ), Lijun Yang1, and Hiroshi Yasuda1,2

1

Introduction

The frequency of lifestyle-related diseases such as diabetes, hypertension and obesity has increased over the last few decades in the developed industrial nations. These diseases can lead to further serious conditions, such as cardiovascular and cerebrovascular diseases. The incidence of allergic diseases such as hay fever, asthma and allergic dermatitis has also been increasing, so that about 30% of the developed world’s population is now affected. The causes for increased disease incidence are likely to involve a complex interaction of genetics and the living environment with lifestyle choices such as the quality and quantity of food and level of physical activity. Prevention of lifestyle-related and allergic diseases has become increasingly important, since they seriously affect quality of life, and their economic costs have become a burden on even the most developed economies. Thus, there is a strong societal demand for effective prophylactic and mitigating therapies. Many foods are known to have qualities that reduce the risk of lifestylerelated diseases, some of which are due to bioactive peptides. Transgenic plants can thus be generated with components that have activity against lifestyle-related or allergic diseases. High-value-added crops, such as genetically engineered rice, that promote human health would have many benefits for consumers and producers alike, and should be socially acceptable irrespective of their genetic modification status by virtue of a net positive in any risk–benefit analysis.

1 Transgenic Crop Research and Development Center, National Institute of Agrobiological Sciences, Kannondai 2-1-2, Tsukuba Ibaraki 305-8602, Japan e-mail: [email protected]; [email protected]; [email protected] 2 National Agriculture Research Center for Hokkaido Region, Hitsujigaoka 1, Toyohira-ku, Sapporo Hokkaido 062-8555, Japan

H.-Y. Hirano et al. (eds.), Rice Biology in the Genomics Era. Biotechnology in Agriculture and Forestry 62. © Springer-Verlag Berlin Heidelberg 2008

357

358

2

F. Takaiwa et al.

Plant Production System as a Bioreactor for Foreign Recombinant Proteins

Plant production of therapeutic peptides has several advantages over conventional recombinant E. coli, yeast fermentative production systems or transgenic animals. The benefits of plant-based production include but are not limited to economy, scalability, product safety (freedom from contamination by animal-derived pathogens) and ease of storage and transportation (Fischer and Emans 2000; Ma et al. 2003; Twyman et al. 2003), but in some applications may also include direct administration without the need for extraction or further purification. The possibility of mass or distributed production in response to demand is a characteristic of plant-based molecular farming. High-value pharmaceutical plant products can be produced at the same cost as non-transgenic plant grains. Recombinant proteins produced in seeds are stably accumulated at high levels without the complication of toxicity to the host plant, which is frequently observed in transgenic plants using the constitutive CaMV 35S, ubiquitin or actin promoters for expression in vegetative tissues. Sequestration of an engineered protein in seed storage tissues, even at high expression levels, thus prevents the inhibition of vegetative growth or yield associated with other production systems. The problem of target protein accumulation at less than 1% of total soluble protein, which has been a weakness of plant-based production platforms, can also be overcome by using seed-specific promoters (Daniell et al. 2001; Stoger et al. 2005). Seed can accumulate higher amounts of recombinant proteins compared with other vegetative tissues, because seed is a natural storage organ for the proteins, starch and lipids that serve as nitrogen and energy sources for the germinating embryo. Production levels of pharmaceutical peptides have exceeded 25% of total soluble protein or 0.5% of rice grain net weight using an endosperm-specific promoter (Nandi et al. 2002). It has recently been shown that endosperm tissue provides a superior production platform for recombinant proteins, since foreign recombinant products can be accumulated in endosperm even if they are not, or cannot be, accumulated in other tissues (Takaiwa et al. 2007), and bioactive peptides expressed in seeds do not lose activity even when stored at room temperature for more than a year (Fischer et al. 2004; Stoger et al. 2005).

3

Characteristics of the Rice Seed Production System as a Bioreactor

Rice is the most important grain crop in the world because it is consumed as a staple food by more than half the world’s population, especially in Asia, Africa and South America, and it provides the bulk of the world’s caloric requirements. Recent demonstrations that rice is an excellent production platform for pharmaceuticals indicate that it has tremendous potential as a bioreactor for the production of bioactive

IV.5 Health-promoting Transgenic Rice

359

peptides or recombinant proteins that promote human health. Rice presents extraordinary advantages as a direct production and delivery system without the need for processing outside of the normal food production chain, particularly where cooked rice makes up a major part of the daily diet. The post-harvest processing of recombinant products is the most expensive part of the production process, accounting for up to 80% of overall production costs (Hood et al. 2002). Rice seed-based pharmaceutical peptide production has a number of important advantages: high expression levels have been achieved in seed, engineered peptides can be produced and sequestered exclusively in specific tissues within the rice grain, rice has a high grain yield per unit biomass (5000—6000 kg grain per hectare), a selection marker-free genomic transformation system has been established, the complete genomic sequence of one variety of rice and the partial sequences of several other agronomically important varieties are now available, and rice is selfpollinating, thus reducing the risk of out-crossing with non-transgenic rice which may be cultivated nearby to minimal levels. In addition, cultivation, harvesting, processing and storage systems have been established across the rice-growing regions that are appropriate for local conditions.

4

4.1

Production Tools Required for Expression of Recombinant Proteins or Peptides in Rice Seed Transcriptional Regulation Promoters

Several molecular tools are available to provide high accumulation levels of recombinant proteins in rice seed. Accumulation of foreign recombinant protein gene products is largely determined at the transcriptional level. About a score of promoters from genes that are highly expressed in developing rice seed have been isolated based on data obtained from EST tags and SAGE technology (Wu et al. 1998; Qu and Takaiwa 2004). Qu and Takaiwa (2004) have transcriptionally fused these promoters to the β-glucuronidase (GUS) reporter gene, and then introduced the chimeric constructs into the rice genome by Agrobacterium-mediated transformation. When promoter activity was examined in stable transgenic rice plants, expression patterns could be separated into several groups based on spatial and temporal specificities, i.e. inner starchy endosperm-specific (26-kDa globulin, 14- to 16-kDa allergen), subaleurone-specific (glutelins), peripheral-specific (prolamins), aleuroneand embryo-specific (oleosine, embryo globulin), scutellum and whole endosperm (PPDK, AlaAT). The relative strengths of these isolated gene promoters were also characterized. Rice seed storage protein glutelin GluB family, 26-kDa globulin, and the 10- and 16-kDa prolamin promoters were identified as strong endospermspecific promoters (Qu and Takaiwa 2004). Endosperm-specific expression observed in these seed storage protein genes is essentially regulated by combinatorial interactions of RISBZ, RPBF and Myb transcription factors, and binding to the

This page intentionally blank

360

F. Takaiwa et al.

GCN4 motif (TGAG/CTCA), prolamin box (AAAG) and AACA motif cis-elements found in their promoters (Onodera et al. 2001; Yamamoto et al. 2006). The GCN4 motif acts as a key cis-regulatory element conferring endosperm-specific expression, because the repeated GCN4 motif, with a minimum of three repeats, confers endosperm-specific expression when fused to the −46 CaMV core promoter (Wu et al. 1998).

4.2

Post-translational Regulation

4.2.1

Intracellular Localization

Post-transcriptional events can also have a considerable impact on the accumulation of recombinant proteins. Subcellular targeting of recombinant proteins can be used as a general method to increase the yield of recombinant proteins, because the subcellular destination influences folding, assembly and post-translational modifications such as glycosylation (Twyman et al. 2003). The endomembrane system of developing seeds has a high capacity for protein processing owing to its abundance of molecular chaperones. A variety of N-terminal signal peptides with lengths ranging from 18–30 amino acids is involved in the initial delivery to the endoplasmic reticulum (ER) lumen. The addition of signal peptides to recombinant proteins is thus a requirement for accumulation in seeds. The absence or removal of signal peptides results in undetectable levels of accumulation despite high levels of transcription (Sharma et al. 2000; Yang et al. 2002; Takagi et al. 2005b). Furthermore, addition of an H/KDEL ER retention signal, which is involved in ER retrieval from cisternal Golgi, to the C terminus of a recombinant protein results in accumulation levels two to ten times that of identical proteins lacking the H/KDEL signal (Wandelt et al. 1992; Schouten et al. 1996; Conrad and Fiedler 1998; Takagi et al. 2005b). Recombinant proteins with signal peptides and KDEL ER retention signals at the N and C termini expressed in the endosperm are predominantly accumulated in ER-derived protein bodies or as aggregates within the ER lumen, whereas levels of ER-resident recombinant proteins are low (Tores et al. 2001; Takagi et al. 2005b). The apoplast has been proposed as a suitable targeting site for the accumulation of recombinant proteins in vegetative tissues such as leaves and fruits. However, the addition of a secretory signal peptide from α-amylase or chitinase that directs products into extracellular spaces (apoplast) suppressed the accumulation of recombinant proteins in rice seeds (Yasuda et al. 2006). The strategy of secreting recombinant proteins into the apoplast with a signal peptide may therefore not be suitable for expression in rice endosperm. There are two types of protein bodies in the same cells of rice endosperm: ERderived protein bodies, called PB-I, and protein storage vacuole bodies, designated PB-II. Major rice seed storage protein glutelins and 26-kDa globulin are accumulated in the PB-II, whereas prolamins are located in the PB-I. Trafficking to the

IV.5 Health-promoting Transgenic Rice

361

PB-II occurs via the Golgi network, or precursor-accumulating (PAC) vesicles as an alternative route from the ER to protein storage vacuoles (PSVs) (Crofts et al. 2004). Currently, there is little information about vacuolar sorting determinants of glutelin into the PB-II. In the case of 26-kDa globulins that are accumulated in the PB-II, the N-terminal region is required for PB-II targeting (Kawagoe et al. 2005). However, it is noteworthy that consensus cysteine residues in the conserved ABC regions of the cereal prolamin and dicot 2S albumin families are responsible for protein body sorting, because disruption of the cysteine residues results in mis-targeting of a GFP reporter from PB-II to PB-I (Kawagoe et al. 2005). Rice seed storage protein mRNA localization is intimately involved in subsequent trafficking to protein bodies (Hamada et al. 2003). That is, prolamin and glutelin mRNAs are transported and localized in the PB-ER or cisternal ER through their specific targeting cis-element sequences, respectively, and then the prolamins and glutelins translated within the ER lumen are directed into separate endomembrane compartments leading to either PB-I or PB-II. One reliable method of designating the accumulation of a recombinant peptide or protein into a particular PB is to express the product as a fusion to a specific rice storage protein. Up to an 80 amino acid recombinant protein can be effectively accumulated in rice seed when it is inserted into the highly variable regions of glutelin (Wakasa et al. 2006). Longer recombinant proteins can be targeted to the PB-I as a fusion at the C terminus of the 13-kDa prolamin protein (Yasuda et al. 2006). The relationship between intracellular localization and accumulation levels has recently been examined using a bioassay with the insulin-stimulating glucagon-like peptide (GLP-1). (For a general model of direct delivery of bioactive peptides, see Fig. 1.) GLP-1 is a peptide of only 30 amino acids. When the codon-optimized synthetic gene coding for modified GLP-1 (mGLP-1) was directly expressed under the control of an endosperm-specific promoter, it was not accumulated, likely due to the silencing that could be expected because of its small size (Yasuda et al. 2005). Therefore, mGLP-1 was inserted into the variable regions of rice storage proteins and expressed as fusion proteins, resulting in targeting to the appropriate protein body. That is, when mGLP-1 was fused to the glutelin or 26-kDa globulin, it was targeted to PB-II, whereas linkage to prolamin targeted the chimera to PB-I (Sugita et al. 2005; Yasuda et al. 2006). Suppression of small peptide synthesis was avoided by constructing mGLP-1 tandem repeats of at least three times. A fused pentamer of mGLP-1 (5×mGLP-1) was directly expressed with the glutelin signal peptide, resulting in high accumulation (approximately 150 µg/grain). Indirect immunohistochemical analysis showed that mGLP-1 was mainly localized in the ER. Attachment of the KDEL ER retention signal decreased accumulation slightly, contrary to expectations. Targeting to the apoplast by the attachment of a rice chitinase signal or to protein bodies by fusion to storage proteins also depressed accumulation levels compared with targeting to the ER. The designation of an optimal intracellular accumulation site will likely depend upon the particular properties of the fused protein.

362

F. Takaiwa et al.

(A)

(B)

(C) 1

SP (GluB)

2

SP (chi)

3

SP (GluA)

4

SP (GluB) 13kDa prolamin

1

mGLPx5

relative amount

mGLPx5 GluA

mGLPx5

2

1

2

3

4

17.5

1.0

1.5

7.5

Apo.

PBII

PBI

localization ER GluA mGLPx6

(D) apoplast protein body II (PBII)

protein body I (PBI)

3

4 rough ER

Golgi body

AAAAA AAAAA membrane-attached ribosome

Bars = 10 µm

Fig. 1 Relationship between localization and level of accumulation of transgene products in rice endosperm cells. A Four transgenes for expression of mGLP-1 peptide in rice endosperm cells. All transgenes were controlled by the glutelin GluB1 promoter and GluB1 terminator. SP Signal peptide; chi chitinase; GluA glutelin GluA2 gene. B Subcellular localization of mGLP-1 peptide in transgenic rice seed cells. Green and red signals indicate localization of mGLP-1 peptide and PB-I, respectively. The mGLP-1 peptide is localized in ER (construct 1), in apoplastic (extracellular) spaces (construct 2), in PB-II (construct 3) and in PB-I (construct 4) of seed cells of rice transformants. C Summary of the relationship between localization and expression level of mGLP-1 peptide in seed cells of rice transformants. In the case of mGLP-1 peptide, the most abundant accumulation of the peptide was obtained from construct 1, which expressed the peptide in ER. D Trafficking of transgene products to each destination site. When the peptide has no signal (except for signal peptide), it was localized in the apoplast or remained in ER. Fusion to glutelin or prolamin targeted the peptide to PB-II or PB-I, respectively, according to the seed storage protein property

4.2.2

Genetic Background

Accumulation levels of recombinant proteins can also be affected by the host plant’s genetic background. Rice mutants deficient in glutelins, prolamins or 26-kDa globulin have been isolated which can provide a larger capacity for the accumulation of foreign gene products (Iida 1995). The inability of the plant to synthesize these storage proteins provides a condition in which there are sufficient pools of amino acids and transcription factors for replacement proteins. Furthermore, there would be little competition between native storage proteins and foreign

IV.5 Health-promoting Transgenic Rice

363

recombinant proteins for ER subdomains or other trafficking factors, including chaperones such as binding protein (BiP) or protein disulphide isomerase (PDI). When the soybean storage protein glycinin gene was introduced into low glutelin mutants such as LGC-1 and a123 and expressed under the control of the glutelin promoter, accumulation levels were about double that in the wild-type background (Tada et al. 2003).

4.2.3

Codon Optimization and Gene Copy Number

There is a positive correlation between codon-usage bias and the level of gene expression. Compatibility between the codon bias of frequently used codons in rice storage protein and the transgene may be critical for high levels of expression. Rare codons in a recombinant protein gene sequence have to be replaced by high-frequency codons, because they appear to destabilize transcripts due to ribosomal pausing. Rare codons also affect translation rates because they may not reflect the composition of the pool of available tRNAs. Optimized codon usage genes expressed in endosperm tissue have been shown to be significantly enhanced compared with native forms. Furthermore, high levels of expression can be achieved by modifying the 5′ and 3′ untranslated sequences and polyadenylation signals that are responsible for mRNA stability. Expression levels are also affected by position on the chromosome and by transgene copy number. Since site-specific introduction by homologous recombination has not yet been established in higher plants at a practical level, repeated transformation followed by genetic crosses of high expression lines are required to obtain stable high accumulation lines. Therefore, we have developed a simple vector construction system in which up to three gene expression cassettes can be simultaneously introduced into one binary vector (Wakasa et al. 2006). Using this system, transgenic rice seed that accumulates large amounts of recombinant products can be efficiently produced from a construct containing multiple gene expression cassettes. One example of a successful application of this system is the expression of the hypocholesterol bioactive peptide lactostatin (IIAEK).

5

5.1

Creation of Functional Foods that Contribute to Human Health Bioactive Peptides Derived from Food Proteins

Dietary proteins have functions beyond providing energy and amino acids as nutrients and physiochemical and sensory properties. They also have specific biological properties (Yoshikawa et al. 2000; Arai et al. 2001; Kitts and Weiler 2003;

364

F. Takaiwa et al.

Korhonen and Pihlanto 2003). Many kinds of bioactive peptides such as opioid, immunostimulating and vasoactive peptides have been isolated from enzymatic digests of various food proteins. These peptides can act as ligands for receptors, enzyme inhibitors, regulators of intestinal absorption, antimicrobials or antioxidative peptides. These bioactive peptides are usually 3 to 20 amino acids in length, making them somewhat resistant to gastrointestinal proteases. They are easily absorbed into the blood via the small intestine. Thus, some bioactive peptides derived from food proteins are effective even after oral administration. However, it should be noted that the bioactivity of ingested proteins is normally much lower than most endogenous bioactive peptides. Therefore, if food-derived bioactive peptides were functionally activated by substitutive mutations and could be accumulated at sufficiently high levels in the edible parts of foods, they would offer a promising approach for the prevention, control and treatment of diseases through a regulated diet or direct oral delivery (Fig. 2).

Endosperm-specific promoter

Modified storageprotein gene containing functional peptide + 1 Construction of transgene

Control of hypertension Control of diabetes Control of allergic disease

2 Gene introduction into rice

Bioactive peptides Rice 3 Transgenic rice highly accumulating bioactive peptides

4 Specific cleavage by digestion enzymes

Fig. 2 Strategy for creating functional transgenic rice seeds to promote human health. Bioactive peptides or proteins are modified for optimal expression in rice, cloned as a fusion protein with a seed storage protein under a suitable set of promoters and other regulatory and targeting factors, and transformed into an appropriate genetic background host plant. Upon ingestion, the active peptide is released by digestive enzymes and absorbed through the gut into the bloodstream

IV.5 Health-promoting Transgenic Rice

5.2

365

Development of Transgenic Rice Seeds with the Hypocholesterolaemic Activity Peptide Lactostatin (IIAEK)

Nagaoka et al. (2001) reported that the peptide Ile-Ile-Ala-Glu-Lys (IIAEK, lactostatin) derived from bovine β-lactoglobulin of whey protein powerfully influences serum cholesterol levels. Lactostatin has greater hypocholesterolaemic activity in rats than casein, soybean protein or the medicine β-sitosterol. After the introduction of a DNA sequence encoding multiple copies of lactostatin into the variable regions of soybean glycinin A1aB1b subunit, the recombinant protein was expressed and accumulated in E. coli cells (Prak et al. 2006). Approximately 80% of the lactostatin peptides were released from the recombinant A1aB1b subunit by in vitro digestion with trypsin. Thus, lactostatin is considered a good candidate for plant-based hypocholesterolaemic peptide production. To develop transgenic rice plants that accumulate high levels of lactostatin in endosperm, tandem multimers of the nucleotide sequence encoding 12×IIAEK were substituted into the variable regions of rice glutelin storage proteins GluA, GluB and GluC, which were driven by strong endosperm-specific 2.3-kb GluB1, 16-kDa prolamin and 10-kDa prolamin promoters, respectively (Wakasa et al. 2006). Three types of binary vectors were constructed into which single-, doubleand triple-gene expression cassettes were inserted by MultiSite Gateway LR clonase reactions. After introduction of the binary vectors into rice plants via Agrobacterium-mediated transformation, transgenic rice plants specifically accumulated glutelin-containing lactostatin in seeds. Immuno-quantitative dot blot analysis indicated that transgenic rice had accumulation levels twice as high in the two-cassette construct as the single-cassette, whereas the three-cassette construct only gave 13% more than the two-cassette insert. Accumulation levels in transgenic rice with the double cassette may be nearly maximized because of limitations not directly related to expression per se, although the highest accumulation of lactostatin was obtained from transgenic seed containing triple gene inserts (Wakasa et al. 2006). Moreover, Western blot analysis showed that modified fusion proteins derived from single-, double- and triple-gene expression cassettes were processed properly into mature acidic and basic subunits, indicating that modified fusion glutelins were localized accurately in PB-II. The obvious next step is to study hypocholesterolaemic activity of the transgenic lactostatin rice in an animal system.

5.3

Development of Transgenic Rice Seeds with Hypertensive Control Activity

Hypertension is an important global public health challenge because of its high prevalence and role in increasing the risk of related diseases (Kearney et al. 2004; Nicolson et al. 2004). Hypertension involves complex interactions between genetic, dietary and

366

F. Takaiwa et al.

environmental factors. A diet rich in foods with anti-hypertensive activity can reduce the risk of developing hypertension or disease severity, as a supplement to drug therapy. Ovokinin (2–7) is one of the anti-hypertensive peptides obtained from egg protein ovalbumin after enzymatic hydrolysis (Matoba et al. 1999). RPLKPW, a modified sequence based on the structure of ovokinin (2–7), showed much greater anti-hypertensive activity in spontaneously hypertensive rats (SHRs) (Yamada et al. 2002). After introducing RPLKPW or RRPLKPWQ into the α′-subunit of soybean β-conglycinin by base substitution mutation, the systolic blood pressure of SHRs was significantly reduced with doses of 10 and 2.5 mg/kg, respectively (Matoba et al. 2001; Onishi et al. 2004). The four-fold increase in activity associated with the addition of the R-Q spacer indicates a more efficient release of the RPLKPW peptide in vivo by gastrointestinal proteases. To develop transgenic rice plants with increased peptide accumulation, a DNA sequence encoding a dimer of RRPLKPWQ was substituted into the variable regions of the major rice glutelin storage proteins GluA, GluB and GluC under the control of strong rice endosperm-specific promoters (Yang et al. 2006). The modified glutelins accumulated to levels around 10% of total seed protein. Oral administration of the crude fusion protein in SHRs significantly lowered systolic blood pressure up to –28 ± 7 mmHg in SHRs at 4 h at a dose equivalent to 140 µg RPLKPW peptide/kg (Fig. 3). A half-dose (equivalent to 70 µg/kg) gave a maximum decrease at 2 h after oral administration with a proportional reduction (–12 ± 2.2 mmHg) (Yang et al. 2006). Furthermore, pulverized transgenic rice seeds reduced the systolic blood pressure of SHRs an average of –15.6 ± 4.8 mmHg 2 h after oral administration at a dose of 1 g/kg (corresponding to 470 µg of RPLKPW peptide/kg).

5.4

Development of Transgenic Rice Seed for Prevention of Type II Diabetes

Diabetes is one of the most serious lifestyle-related diseases, and the number of diabetics in Japan is estimated at 7.4 million, with up to 16.4 million having pre-diabetic symptoms. The latest WHO estimate is that 177 million people have diabetes worldwide, with a projected increase to at least 300 million by 2025. Untreated diabetes can lead to the complications of blindness, gangrene, neurodegenerative symptoms and organ failure. GLP-1 is a peptide hormone consisting of 30 amino acids that induces insulin synthesis and secretion from pancreatic β-cells in response to increases in blood glucose concentrations (Drucker et al. 1987; Weir et al. 1989). This peptide is an attractive candidate as a therapeutic agent of type II diabetes, because of its glucose-dependent insulin stimulation. As a first step toward producing rice seeds with sufficient and active GLP-1, the peptide was modified for resistance to trypsin digestion without loss of bioactive function. In order to accumulate the mGLP-1 peptide in endosperm tissue, the codon-optimized mGLP-1 gene was expressed as a fusion protein with rice 26-kDa globulin by insertion into its variable region. The boundary site amino acids between

IV.5 Health-promoting Transgenic Rice

367

A

LB

pGluB1-HRP

GluB-1 P Hind III

pGluA2-HRP/GluC-HRP

Systolic blood pressure (mmHg)

Sac 1

GluC cDNA

GluB-4 T

Eco RI

GluB-4 P

GluA2 cDNA

GluB-2 P

Xba I

NosT

Sac I

Bam HI

C

B 5 0 -5 -10 -15 -20 -25 -30 -35 -40

RB

Eco R1

Bam HI

Sac I

Bam HI CaMV35S P

hpt

GluB-1 T

GluB-1 cDNA Nco 1

pAg7

** ** **

0

2

*

***

4

6

0

2

4

6

Time after administration (h)

Fig. 3 Anti-hypertensive activities of transgenic rice that accumulates modified glutelins containing multimers of the RPLKPW peptide. A Two expression plasmids used for transformation. hpt Hygromycin phosphotransferase; pAg7 agropine synthase poly(A) signals; RB right border; LB left border; GluA-2, GluB-1, GluB-2, GluB-4 and GluC rice glutelins A2, B1, B2, B4 and C, respectively; Nos T nopaline synthase poly(A) signals; GluB-1 P, GluB-2 P and GluB-4 P glutelin B1, B2 and B4 promoters, and GluB-1 T and GluB-4 T glutelin B1 and B4 3′ untranslated terminal regions, respectively; blank triangle RRPLKPWQ peptide substitution site in variable region of the basic subunit of glutelin B1; solid triangles dimer peptide RRPLKPWQRRPLKPWQ substitution sites in variable regions of acidic subunits of glutelins. B and C Oral administration of 30 mg/kg of a crude glutelin fraction of GluB1-HRP and 1 g/kg of unpolished pulverized GluA2-HRP/GluC-HRP rice seeds, respectively. Solid squares GluB1-HRP glutelin fraction or GluA2-HRP/GluC-HRP seeds; blank squares glutelin fraction or seeds from rice variety Kitaake (non-transformant); * P<0.05; ** P<0.01; *** P<0.001

globulin and mGLP-1 were replaced with lysine and arginine residues to release mGLP-1 after trypsin digestion in the small intestine. Transgenic rice seeds were generated using the marker-free MAT-vector system. The highest level of mGLP-1 accumulation was estimated at about 50 µg/grain. Seed extracts treated with trypsin stimulated insulin synthesis and secretion from a cultured mouse pancreatic β-cell line (Sugita et al. 2005). As an alternative approach, a gene coding for tandem penta-repeated mGLP-1 peptide (mGLP×5) was directly expressed under the control of the 2.3-kb GluB1 promoter. The linking sites of individual peptides were flanked by arginine residues for trypsin digestion. The transgenic seeds with the highest expression of mGLP × 5 accumulated about 150 µg/grain (Yasuda et al. 2006). Oral administration of these

368

F. Takaiwa et al.

transgenic seeds in mice significantly suppressed elevated blood glucose levels compared with non-transgenic rice seeds.

5.5

Peptide Immunotherapy for Allergic Disease Using Rice Seed-based Edible Vaccines

Allergen-specific immunotherapy is the only current treatment that cures allergic diseases. It does so by inducing tolerance through the basic immune response mechanisms. Conventional immunotherapy by administering gradually increasing amounts of allergen extracts by subcutaneous injection has been successful, but this type of therapy is associated with risk of local reactions, systemic anaphylaxis, injection discomfort and treatment for 3 to 5 years (Frew 2003). These drawbacks emphasize the need for a more effective, safe and easy immunotherapy. Peptide immunotherapy using dominant T-cell epitopes has been proposed in order to avoid side-effects and to shorten the treatment term by high dose administration (Wallner and Gefter 1996; Larche and Wraith 2005). Oral immunotherapy is expected to offer an effective, safe and simple route of vaccine administration, since both mucosal and systemic immune responses can be induced by the delivery of a plantderived vaccine to a mucosal tissue (Mayer and Shao 2004). However, administration of higher doses of oral vaccine is required for delivery to the mucosal immune system, because epitope antigens are subjected to enzymatic digestion in the gastrointestinal tract. It has been recently demonstrated that oral administration of T-cell epitopes with an immune adjuvant such as cholera toxin could effectively induce immune tolerance by suppressing allergen-specific T-cell-specific IgE production and release of histamine. As a new immunotherapy based on the principle of peptide immunotherapy, rice seeds that accumulate major T-cell epitopes derived from pollen allergens can be used as a vehicle to deliver tolerogens to the mucosal immune system (Hiroi and Takaiwa 2006). Rice seed-based peptide vaccines have a number of striking benefits. Vaccine produced in plant cells may be protected from gastrointestinal enzymes due to packaging by cell walls and compartmentalization in organelles, thus offering a good delivery system to the mucosal immune system (Walmsley and Arntzen 2000). The feasibility of using a rice seed-based edible peptide vaccine has been recently examined for treatment of Japanese cedar pollen allergic disease (pollinosis, allergic rhinitis). Japanese cedar pollinosis is the most prevalent allergic disease in Japan. About 20% of the Japanese population is afflicted with this pollinosis. Cry j 1 and Cry j 2 are the major Japanese cedar pollen allergens, and their major T-cell epitopes have been characterized for mice and humans (Hirahara et al. 2001; Yoshitomi et al. 2002). To confirm the oral tolerogenic efficacy of a rice seed-based edible vaccine, transgenic rice seed containing only the BALB/c mouse major T-cell epitope peptides Cry j 1 p277–290 and Cry j 2 p246–259 was created (Takagi et al. 2005a). These short peptides were introduced into variable regions of the soybean seed storage protein glycinin A1aB1b, and then specifically expressed as

Stimulation index of specific T cell

Daily oral feeding

Systemic challenge of cedar pollen proteins

Histamine(ng/ml)

Allergen specific IgE levels (reciprocal log2 titer)

Analysis

Fig. 4 Oral immune suppression induced by feeding transgenic rice seeds that accumulate glycinin A1aB1b-containing major mouse T-cell epitopes derived from Cry j 1 and Cry j 2 allergens. A Schematic presentation of the binary vector used for expression of modified glycinin A1aB1b in transgenic rice seed. B Timeline of feeding schedule and challenge with cedar pollen allergens. C Inhibition of allergen-specific serum IgE, serum histamine and number of sneezes following oral administration of transgenic rice seeds compared with non-transgenic rice seeds. GluB-1 Pro Rice glutelin GluB-1 promoter; 35S CaMV Pro cauliflower mosaic virus 35S promoter; hpt hygromycin phosphotransferase gene; pAg7 agropine synthase polyadenylation signal; RB right border; LB left border

C

B

Sneeze number/5min

A

IV.5 Health-promoting Transgenic Rice 369

370

F. Takaiwa et al.

a fusion protein in the endosperm tissue of transgenic rice seeds under the control of the rice major seed storage protein glutelin GluB-1 promoter (Takagi et al. 2005a). The modified glycinin accumulated at a level of 7 µg/grain (0.5% of total seed protein) in mature rice seeds. Oral daily administration of the transgenic rice seeds to mice for 4 weeks prior to systemic challenge with total cedar pollen protein inhibited the development of allergen-specific serum IgE and IgG antibodies and CD4+ T-cell proliferative responses, when compared with control mice fed with non-transgenic rice seeds (Fig. 4). The levels of allergen-specific CD4+ T-cellderived allergy-associated Th2 cytokine production of IL-4, IL-5 and IL-13, and histamine release in serum were also significantly decreased (Takagi et al. 2005a). Moreover, inhibition of the pollen-induced clinical symptoms of nasal sneezing was observed in an experimental sneezing mouse model. These results demonstrated that oral immune tolerance was induced by the administration of transgenic rice seeds containing the major T-cell epitopes, thus indicating that rice-based peptide vaccines can provide a new therapeutic tool for allergic diseases. Based on the feasibility of an oral peptide immunotherapy strategy against pollen allergy using rice seed-based vaccines in mice, a human version of the transgenic rice seed-based edible vaccine has been generated. An artificial hybrid peptide 7Crp gene encoding a 96 amino acid protein of seven linked human dominant T-cell epitope peptides derived from Cry j 1 and Cry j 2 allergens was synthesized using codons optimized for expression in rice seed (Takagi et al. 2005b). In order to enhance accumulation of the 7Crp peptide, the GluB-1 signal peptide and the KDER ER retention signal were ligated to the N and C termini of 7Crp. When the 7Crp gene was specifically expressed under the control of the GluB-1 promoter, it was accumulated in mature seeds at a maximum level of about 60 µg/grain (4% of total seed protein). Observation by immuno-electron microscopy indicated that the 7Crp peptide was predominantly deposited in protein bodies in the rice endosperm in spite of the attachment of the KDEL ER retention signal (Takagi et al. 2005b). When transgenic rice seeds containing 7Crp were orally administered to B10.S mice, which recognize only one epitope derived from Cry j 1, and then nasally challenged with intact Cry j 1 allergen, both the T-cell proliferative responses and serum IgE levels against Cry j 1 were depressed compared to control mice fed with nontransgenic rice seeds (Takagi et al. 2005b). The 7Crp peptide is not glycosylated and its expression has little effect on accumulation of endogenous allergen proteins. T-cell proliferative activity (i.e. immunogenicity) is retained even after boiling 7Crp rice for 20 min at 100°C or autoclaving for 20 min, indicating that oral immune tolerance would be effective with steamed or cooked rice. Taken together, these results indicate that the transgenic immunotolergenic rice is safe, and may provide a new peptide immunotherapy for Japanese cedar pollinosis. The food safety of 7Crp seeds, or of any subsequently developed rice seeds containing immuno-active constituents, must be carefully examined prior to its use. Regulatory standards based on substantial equivalence have been used to determine the food safety of transgenic crops. In this method, a comparative assessment of the toxicology and nutrient composition is made between the newly developed transgenic

IV.5 Health-promoting Transgenic Rice

371

food and a conventional counterpart. When a number of macro- and micro-nutrients were compared between the cognate parental non-transgenic and transgenic 7Crp seeds, the amino acid, lipid and mineral compositions were found to be essentially identical (Takagi et al. 2006). Acknowledgements This work was supported by research grants (‘Research for the utilization and industrialization of agricultural biotechnology’, ‘Functional analysis of genes relevant to agriculturally important traits in rice genome’ and ‘Development and demonstration of crop genomic breeding technology’) to F. Takaiwa.

References Arai S, Osawa T, Ohigashi H, et al. (2001) A mainstay of functional food science in Japan – history, present status, and future outlook. Biosci Biotechnol Biochem 65:1–13 Conrad U, Fiedler U (1998) Compartment-specific accumulation of recombinant immunoglobulins in plant cells: an essential tool for antibody production and immunomodulation of physiological functions and pathogen activity. Plant Mol Biol 38:101–109 Croft AJ, Washida H, Okita TW, Ogawa M, Kumamaru T, Satoh H (2004) Targeting of proteins to endoplasmic reticulum-derived compartments in plants. The importance of RNA localization. Plant Physiol 136:3414–3419 Daniell H, Streatfield SJ, Wycoff K (2001) Medical molecular farming: production of antibodies, biopharmaceuticals and edible vaccines in plants. Trends Plant Sci 6:219–226 Drucker DJ, Philippe J, Mojsov S, Chick WL, Habener JF (1987) Glucogon-like peptide 1 stimulates insulin gene expression and increases cyclic AMP levels in a rat islet cell line. Proc Natl Acad Sci USA 84:3434–3438 Fischer R, Emans N (2000) Molecular farming of pharmaceutical proteins. Transgenic Res 9:279–299 Fischer R, Stoger E, Schillberg S, Christou P, Twyman RM (2004) Plant-based production of biopharmaceuticals. Curr Opin Plant Biol 7:152–158 Frew AJ (2003) Immunotherapy of allergic disease. J Allergy Clin Immunol 111:S712–S719 Hamada S, Ishiyama K, Sakulsingharoj C, et al. (2003) Dual regulated RNA transport pathways to the cortical region in developing rice endosperm. Plant Cell 15:2265–2272 Hirahara K, Tatsuta T, Takatori T, et al. (2001) Preclinical evaluation of an immunotherapeutic peptide comprising 7 T-cell determinats of Cry j 1 and Cry j 2, the major Japanese cedar pollen allergens. J Allergy Clin Immunol 108:94–100 Hiroi T, Takaiwa F (2006) Peptide immunotherapy for allergeic diseases using a rice-based edible vaccine. Curr Opin Allergy Clin Immunol 6:455–460 Hood EE, Woodard SL, Horn ME (2002) Monoclonal antibody manufacturing in transgenic plants – myths and realities. Curr Opin Biotechnol 13:630–635 Iida S (1995) Screening and genetic analysis of rice seed storage protein mutants. Gamma Field Symp 34:31–48 Kawagoe Y, Suzuki K, Tasaki M, et al. (2005) The critical role of disulfide bond formation in protein sorting in the endosperm of rice. Plant Cell 17:1141–1153 Kearney PM, Whelton M, Reynolds K, Whelton PK, He J (2004) Worldwide prevalence of hypertension: a systematic review. J Hypertens 22:11–19 Kitts DD, Weiler K (2003) Bioactive proteins and peptides from food sources. Applications of bioprocesses used in isolation and recovery. Curr Pharm Des 9:1309–1323 Korhonen H, Pihlanto A (2003) Food-derived bioactive peptides – opportunities for designing future foods. Curr Pharm Des 9:1297–1308

372

F. Takaiwa et al.

Larche M, Wraith DC (2005) Peptide-based therapeutic vaccines for allergic and autoimmune diseases. Nat Med 11:S69–S76 Ma JKC, Pascal MW, Drake PW, Christou P (2003) The production of recombinant pharmaceutical proteins in plants. Nat Rev Genet 4:794–805 Matoba N, Usui H, Fujita H, Yoshikawa M (1999) A novel anti-hypertensive peptide derived from ovalbumin induces nitric oxide-mediated vasorelaxation in an isolated SHR mesenteric artery. FEBS Lett 452:181–184 Matoba N, Doyama N, Yamada Y, Maruyama N, Utsumi S Yoshikawa M (2001) Design and production of genetically modified soybean protein with anti-hypertensive activity by incorporating potent analogue of ovokinin (2–7). FEBS Lett 497:50–54 Mayer L, Shao L (2004) Therapeutic potential of oral tolerance. Nature Rev Immunol 4:407–419 Nagaoka S, Futamura Y, Miwa K, et al. (2001) Identification of novel hypocholesterolemic peptides derived from bovine milk beta-lactoglobulin. Biochem Biophys Res Commun 281:11–17 Nandi S, Suzuki YA, Huang J, et al. (2002) Expression of human lactoferrin in transgenic rice grains for the application in infant formula. Plant Sci 163:713–722 Nicolson DJ, Dikinson HO, Campbell F, Mason JM (2004) Lifestyle interventions or drugs for patients with essential hypertension: a systematic review. J Hypertens 22:2043–2048 Onishi K, Matoba N, Yamada Y, et al. (2004) Optimal designing of beta-conglycinin to genetically incorporate RPLKPW, a potent anti-hypertensive peptide. Peptides 25:37–43 Onodera Y, Suzuki A, Wu CY, Washida H, Takaiwa F (2001) A rice functional transcriptional activator, RISBZ1, responsible for endosperm-specific expression of storage protein genes through GCN4 motif. J Biol Chem 276:14139–14152 Prak K, Maruyama Y, Maruyama N, Utsumi S (2006) Design of genetically modified soybean proglycinin A1aB1b with multiple copies of bioactive peptide sequences. Peptide 27:1179–1186 Qu LQ, Takaiwa F (2004) Evaluation of tissue specificity and expression strength of rice seed component gene promoters in transgenic rice. Plant Biotech J 2:113–125 Rutherfurd-Markwick KJ, Moughan PJ (2005) Bioactive peptides derived from food. J AOAC Int 88:955–966 Schouten A, Roosien J, van Engelen FA, et al. (1996) The C-terminal KDEL sequence increases the expression level of a single-chain antibody designed to be targeted to both the cytosol and the secretory pathway in transgenic tobacco. Plant Mol Biol 30:781–793 Sharma A, Sharma R, Imamura M, Yamakawa M, Machii H (2000) Transgenic expression of cecropin B, an antibacterial peptide from Bombyx mori, confers enhanced resistance to bacterial leaf blight in rice. FEBS Lett 484:7–11 Stoger E, Ma JKC, Fischer R, Christou P (2005) Sowing the seeds of success: pharmaceutical proteins from plants. Curr Opin Biotech 16:167–173 Sugita K, Endo-Kasahara S, Tada Y, et al. (2005) Genetically modified rice seeds accumulating GLP-1 analogue stimulate insulin secretion from a mouse pancreatic beta-cell line. FEBS Lett 579:1085–1088 Tada Y, Utsumi S, Takaiwa F (2003) Foreign gene products can be enhanced by introduction into low storage protein mutants. Plant Biotech J 1:411–422 Takagi H, Hiroi T, Yang L, et al. (2005a) A rice-based edible vaccine expressing multiple T cell epitopes induces oral tolerance for inhibition of Th2-mediated IgE responses. Proc Natl Acad Sci USA 102:17525–17530 Takagi H, Saito S, Yang L, Nagasaka S, Nishizawa N, Takaiwa F (2005b) Oral immunotherapy against a pollen allergy using a seed-based peptide vaccine. Plant Biotech J 3:521–533 Takagi H, Hirose S, Yasuda H, Takaiwa F (2006) Biochemical safety evaluation of transgenic rice seeds expressing T cell epitopes of Japanese cedar pollen allergens. J Agric Food Chem 54:9901–9905 Takaiwa F, Takagi H, Hirose S, Wakasa Y (2007) Endosperm tissue is good production platform for artificial recombinant proteins in transgenic rice. Plant Biotech J 5:84–92

IV.5 Health-promoting Transgenic Rice

373

Torres E, Gonzalez-Melendi P, Stoger E, et al. (2001) Native and artificial reticuloplasmins coaccumulate in distinct domains of the endoplasmic reticulum and in post-endoplasmic reticulum compartments. Plant Physiol 127:1212–1223 Twyman RM, Stoger E, Schillberg S, Christou P, Fischer R (2003) Molecular farming in plants: host systems and expression technology. Trends Biotech 21:570–578 Wakasa Y, Yasuda H, Takaiwa F (2006) High accumulation of bioactive peptide in transgenic rice seeds by expression of introduced multiple genes. Plant Biotech J 4:499–510 Wallner BP, Gefter ML (1996) Peptide therapy for treatment of allergic diseases. Clin Immunol Immunopathol 80:105–109 Walmsley AM, Arntzen CJ (2000) Plants for delivery of edible vaccines. Curr Opin Biotech 11:126–129 Wandelt CI, Khan MRI, Craig S, Schroeder HE, Spencer D, Higgins TJV (1992) Vicillin with carboxy-terminal KDEL is retained in the endoplasmic reticulum and accumulates to high levels in the leaves of transgenic plants. Plant J 2:181–192 Weir GC, Mojsov S, Hendrick GK, Hebener JF (1989) Glucagon-like peptide 1 (7–37) actions on endocrine pancreas. Diabetes 38:338–342 Wu CY, Suzuki A, Washida H, Takaiwa F (1998) The GCN4 motif in a rice glutelin gene is essential for endosperm-specific gene expression and is activated by Opaque-2 in transgenic rice plants. Plant J 14:673–683 Yamada Y, Matoba N, Usui H, Onishi K, Yoshikawa M (2002) Design of a highly potent antihypertensive peptide based on ovokinin (2–7). Biosci Biotechnol Biochem 66:1213–1217 Yamamoto MP, Onodera Y, Touno SM, Takaiwa F (2006) Synergism between RPBF Dof and RISBZ1 bZIP activators in the regulation of rice seed expression genes. Plant Physiol 141:1694–1707 Yang LJ, Tada Y, Yamamoto MP, Zhao H, Yoshikawa M, Takaiwa F (2006) A transgenic rice seed accumulating an antihypertensive peptide reduces the blood pressure of spontaneously hypertensive rats. FEBS Lett 580:3315–3320 Yang SH, Moran DL, Jia HW, Bicar EH, Lee M, Scott MP (2002) Expression of a synthetic porcine α-lactoalbumin gene in the kernels of transgenic maize. Trangenic Res 11:11–20 Yasuda H, Tada Y, Hayashi Y, Jomori T, Takaiwa F (2005) Expression of the small peptide GLP-1 in transgenic plants. Trangenic Res 14:677–684 Yasuda H, Hayashi Y, Jomori T, Takaiwa F (2006) The correlation between expression and localization of a foreign gene product in rice endosperm. Plant Cell Physiol 47:756–763 Yoshikawa M, Fujita H, Matoba N, et al. (2000) Bioactive peptides derived from food proteins preventing lifestyle-related diseases. BioFactors 12:143–146 Yoshitomi T, Hirahara K, Kawaguchi J, et al. (2002) Three T-cell determinants of Cry j 1 and Cry j 2, the major Japanese cedar pollen antigens, retain their immunogenicity and tolerogenicity in a linked peptide. Immunology 107:517–522

Index

A Ab initio method, 14–17 Abscisic acid (ABA), 141, 142, 144, 340–348, 351 ABAR/CHLH, 341 dihydrophaseic acid (DPA), 340 FCA, 341 phaseic acid (PA), 340 Activation tagging, 95–98 Alternative splicing, 56 Alternative terminated transcripts, 28 Alternatively terminated isoforms, 24 Amino acid-auxin permease (AAAP) family, 64 Amino acid-polyamine-organocation (APC) superfamily, 64 Annexin, 63 Anther, 192, 193, 195, 196, 200 Antisense transcript(s), 24, 27–29, 56 Arabidopsis Gα., 139–141, 144, 145 gpa1, 141–145 ATP-binding cassette (ABC) proteins, 63 Autonomous element, see Transposable element Auxin, 121, 129–131

B Backcross inbred lines (BILs), 305 Bacterial blight, 31 Beijing Genomics Institute (BGI), 54, 56, 57 Bioactive peptide, 357, 358, 361, 363, 364 Biogeography, 225–231 Biological species concept (BSC), 247 Biological species, 247, 248, 256 Blast resistance, 300 BLAST, 16, 17, 19

BLASTN, 61 BLASTX, 61 Bodymap technology, 54 Brachypodium distachyon, 10 Brassinosteroid, 121, 126–129, 132, 141, 143, 144, 314, 315 BT-CMS, 205–213

C Ca2+ storage proteins, 64 Ca2+/phospholipid binding protein, 64 Cap-trapper method, 55, 56 CentO, 5, 8 Centromere, 5, 240–242 evolution, 242 region, 242 repeat, 241–242 Chikusichloa, 222 Chromosome, 235–243 bivalent, 235, 243 complement, 235–237, 240 pairing, 237–238 rearrangement, 239 size, 235 Chromosome segment substitution lines (CSSLs), 302–305 Circadian clocks GI, 170, 171 OsGI, 170 OsPRR37, 170 Class I and class II transposons, 5 Coding region, 54 Codon optimization, 363 codon usage, 363 Comparative method, 13, 14 Complementary DNAs (cDNAs), 14, 17, 19, 53–55 Cool temperature tolerance, 301 375

376 Core collection, 7, 302 Cross-hybridization, 55, 61 Cross-incompatibility, 248, 249, 255 cv 93-11, 23 CW-CMS, 206–208, 212, 213 Cytokinin, 315, 316 Cytoplasmic male sterility (CMS), 205, 206, 208–210, 212–214, 252, 254

D Deletion, 240–242 Diabetes, 357, 364, 366 DM model, 254, 255 DNA Data Bank of Japan (DDBJ), 56 DNA pool, 99–100, 103 Domestication genes, 269, 272 related traits, 268–271 syndrome, 269, 271 Duplication, 237, 239, 240, 242, 243

E Ecotype annual, 281, 282, 286, 288 intermediate, 282–284, 286, 288 perennial, 281–283, 286, 288 EF-hand, 64 End-sequence, 6, 8 Enhancer-trap, 97, 99 Entrapment tagging, 98–99 ER retention signal, 360, 361, 370 H/KDEL, 360 Evolution, 235, 239–242 genome, 241–243 Expressed sequence tags (ESTs), 15, 17, 19, 53–55, 64

F F1 hybrid, 24, 29–31 F1 hybrid rice, 30 FASTA, 62 Fertility restorer (Rf) gene, 205, 206, 208–212, 214 FISH, 236, 237 Flanking sequence tag (FST), 99–103, 266–268 Flowering genes CDF1, 165, 173 CO, 165, 166, 170, 171, 173 Ehd1, 167–169, 173, 339

Index FD, 171, 172 FKF1, 165 FLC, 163, 171 FT, 163, 165, 166, 171 GI, 170, 171 Hd1, 166–173 Hd3a, 165, 166, 169–172 Hd6, 166, 173 HY1, 164 OsGI, 170, 171 OsMADS14, 171 OsMADS50, 171 OsPRR37, 170 RFT1, 171 se5, 164, 165 SOC1, 171 TSF, 171 FON genes FON1, 179–181, 183 FON2, 179–181, 183 Foundation for the Advancement of International Science (FAIS), 55 Full-length cDNA (FL-cDNA), 53–61, 63–65 Fusion protein, 361, 364–366, 370

G Gamete eliminator (GE), 251–255 Gametophyte genes, 255 Gamma-carboxyglutamic acid, 64 GenBank, 54, 61, 62 Gene complexity analysis, 54 duplication, 253 flow, 264–266, 272 pool, 247, 248, 256 tagging, 81–84, 89–92 therapy, 88 Gene-finding, 13–14, 16 Gene ontology (GO), 16, 17, 19 Gene targeting, 81–84, 86–89, 91, 92 ectopic gene targeting (EGT), 85, 86 one-sided invasion (OSI), 85, 86 true gene targeting (TGT), 82, 83, 85, 86 Genetic revolution, 256 Genome, 3–11, 235–243 analysis, 238 divergence, 240 duplication, 239–240, 243 size, 236, 237, 240–242 Germ cell, 192, 193, 195, 200

Index GFP, 98, 99 Gibberellin, 121–127, 131, 132, 141, 142, 144, 310–314 Glucagon-like peptide (GLP-1), 361, 366 b-glucuronidase (GUS), 98, 99 G proteins, 135–139 β subunit, 135–138 α subunit, 135–139 constitutively active form, 138, 139, 145 d1, 138–139 extra-large Gα, 137 extra-large Gγ, 137 γ1 subunit, 135–138 γ2 subunit, 137, 138 G protein-coupled receptor GCR1, 139, 140, 142–145 GCR2, 139, 140, 142, 144 G-protein-mediated signaling, 142–143, 145 abscisic acid, 141, 142, 144 blue light, 141, 143, 144 brassinosteroid, 141, 143, 144 extracellular calmodulin, 141, 143, 145 gibberellin, 141, 142, 144 pathogen, 141, 142 sugar, 141, 143, 145

H HD-ZIPIII, 154–156 PHB, 154, 155 PHV, 154, 155 REV, 154, 155 RLD, 155 Heading date, 295–297, 304 Heterosis, 29–31 High-quality map-based sequence, 4 Histidine kinase, 338, 339, 342 AtHK1, 339 early heading date 1 (Ehd1), 167–169, 173, 339 SHO1, 339 SLN1, 339 Homologous recombination (HR), 81, 82, 84–89, 91, 92 Homology-based method, 14, 15 Host-pathogen interactions, 32 Hybrid breakdown, 248, 249, 251–254 inviability, 248–250, 254, 256 peptide 7Crp, 370 sterility, 248–250, 252–254 Hybridization and differentiation cycle, 272 Hygroryza, 222

377 Hypertension, 357, 364–366 Hypocholesterolaemic activity, 365

I IDE, 331 Immunotherapy, 368, 370 peptide vaccine, 368, 370 T-cell epitope, 368–370 Inbreeding, 262–265 Inbreeding depressions, 265 indica, 6, 53, 54, 56–58, 62 Inositol 1,4,5-trisphosphate (IP3), 340 FRY1 / fry1, 340 Insertion, 240, 242 Insertion polymorphism ecotype-specific, 285 genome-type specific, 285 interspecific, 278, 285 subspecies specific, 285 Institute of Genome Research (TIGR), 56–59, 62 Institute of Physical and Chemical Research (RIKEN), 55 Intergenic transcripts, 24 International Nucleotide Sequence Databases, 17 International Rice Genome Sequencing Project (IRGSP), 4, 13, 17, 19, 20, 55–57, 61, 62 InterPro, 17, 19 IRO2, 323, 324, 331 Isolating barriers, 247–249, 254, 256 Isolation by distance model, 267

J Japanese cedar pollen allergen, 368 Cry j 1, 368–370 Cry j 2, 368–370 Japanese Rice Genome Research Program (RGP), 54, 55 japonica, 4, 6, 53, 54, 56, 58, 62 Japonica and Indica, 268, 271–272

K Knockin, 87, 88 Knockout, 82, 86–88, 95, 96, 99, 100, 103 gene disruption, 82 Knowledge-based Oryza Molecular Biological Encyclopedia (KOME), 56, 57, 60, 61, 64

This page intentionally blank

378 L Lactostatin, 363, 365 LD-CMS, 205–209, 212 Leersia L. angustifolia, 225 L. denudate, 225 L. drepanothrix, 225 L. friesii, 225 L. hexandra, 225 L. japonica, 220, 225 L. lenticularia, 225 L. ligularis, 225 L. monandra, 225 L. nematostachya, 225 L. oncothrix, 225 L. oryzoides, 225 L. perrieri, 225 L. stipitate, 225 L. tisserantii, 225 L. triandra, 225 L. virginica, 225 Lipid-containing proteins, 64 Low glutelin mutant, 363 Low temperature germinablity, 301 Luziola, 221, 222, 225

M Magnaporthe transcripts, 32 Major facilitator superfamily (MFS) Major intrinsic protein (MIP), 63 Maltebrunia, 222, 230 Map-based cloning, 297, 298, 300, 301, 304, 305 Mapping population, 298, 302, 304, 305 Marker assisted selection (MAS), 295, 298, 304 MAs, 321–331 DMA, 322, 325, 329 Meiosis, 191, 193, 195–200, 202 Meiotic drive (MD), 255 Meristem branch meristem, 178, 179 floral (flower) meristem, 179–183, 185 floret meristem, 178, 179 inflorescence meristem, 178–181 spikelet meristem, 178 Microarray, 37–49, 53–55, 61–63, 65 expression array, 38–41, 46 SuperSAGE Array, 40, 45 tiling array, 38, 40, 42–44, 46, 49 MicroRNAs (miRNAs), 24, 25, 28, 32–34, 65 miR165, 155, 156 miR166, 155, 156

Index Microsynteny, 9 Minimum information about a micro array experiment (MIAME), 65 MITE, 71, 72, 76, 77 mPing, 83 stowaway, 71 Stowaway-Os3, 71 tourist, 70, 71 wanderer, 72 Mitochondria, 107, 109, 110, 112, 114 Mitochondrial DNA (mtDNA), 109, 110 Mitochondrial genome, 107–110, 112–114 Mitochondrion, 108, 109, 111, 113 Molecular clock, 220, 222–224 Monovalent cation-proton antiporter 1 (CPA1) family, 64 Monovalent cation-proton antiporter 2 (CPA2) family, 64 MPSS, 23–34 hits=1, 26 reliable, 26, 27 TPM, 24, 25, 29 TPQ, 25 mRNA, 24–28, 31, 34 MultiSite Gateway LR clonase reaction, 365

N NA, 322, 323, 327–329 NADPH oxidase, 341 National Center for Biotechnology Information (NCBI), 54, 62, 64 Natural antisense transcripts (NATs), 28, 29 Natural hybridization, 273 variation, 295, 299, 302, 303, 305 NCBI Gene Expression Omnibus (GEO), 62 nDart, 81, 83, 84, 89–92 aDart, 83, 84, 89, 90 nDart1-0, 83, 89, 90, 92 Nearly isogenic lines (NILs), 297, 299, 303, 304, 306 NIAS GeneBank, 7 Nipponbare, 23, 29–31 Non-autonomous element, see Transposable element Nonhomologous end-joining (NHEJ), 85–87, 91 Non-random association, 268, 271 Nuclear genome, 107, 108, 110–114 Nucleus, 107, 108, 111–114

Index O Oryza rufipogon annual, 281–284, 286, 288 intermediate, 281–284, 286, 288 perennial, 281–284, 286, 288 Oryza sativa, 261, 266, 268, 269, 271, 272 indica, 279–283, 286, 288, 289 japonica, 279–282, 286, 288, 289 javanica, 281, 286 Oligocapping method, 55, 56 Open reading frame (ORF), 54 orf79, 209 Oryza, 4, 6–8 O. abromeitiana, 226 O. barthii, 224, 225, 230, 261–264, 266 O. brachyantha, 223, 224, 226 O. coarctata, 226 O. eichingeri, 224, 230 O. glaberrima, 220, 224, 225, 229, 230 O. glumaepatula, 229, 231, 261–264 O. granulate, 223, 224, 226–228 O. indandamanica, 226 O.longistaminata, 261, 263–265 O. meridionalis, 220, 224, 227, 230, 261, 262, 264 O. meyeriana, 226 O. minuta, 230 O. neocaledonica, 226, 227 O.nivara, 262 O. officinalis, 223, 224, 226–230 O. ridleyi, 223, 224, 226, 227 O.rufipogon, 225, 227, 261–269, 271–273 O. sativa, 220, 223, 224, 226–230 O. schlechteri, 223, 224, 226 Oryza Map Alignment Project (OMAP), 8 Oryzeae, 219–223, 225, 230 OsMADS genes OsMADS1, 182 OsMADS2, 185, 186 OsMADS3, 182, 184–186 OsMADS4, 185 OsMADS14, 171, 186 OsMADS15, 186 OsMADS16, 185 OsMADS18, 186 OsMADS50, 171 OsMADS58, 181, 182, 184–186 Outbreeding, 263, 265 Ovokinin (2-7), 366 Ovule, 192–194, 200, 201

379 P Pentatricopeptide repeat (PPR), 210–212 Perennial and annual ecotypes, 263, 265, 266 Phospholipase Dα 1 AtPLDa1, 140, 141 plda1, 140, 142 Phosphorylated intermediate form (P-ATPases), 63 Photoreception (photoreceptor) phytochromes, 164, 165, 169, 170, 173 Photoreceptor (Photoreception) CRY2, 164 FKF1, 165 phyA, 165 phyB, 169 Phylogenetic relationships, 263 Phylogeny, 219–225 Phylogeographic differentiation, 266–268 Pi-9 R-gene, 31 Pirin 1 AtPirin1, 140, 142, 144 atprin1, 140, 142, 144 Plant-based production, 358 Plant hormone, 121 Plastid, 107–114 Plastid DNA (ptDNA), 109, 110 Plastid genome, 107–110, 113, 114 Poaceae, 9, 10 Positive-negative selection, 86, 87, 91 Potomophila, 222, 230 Pre-harvest sprouting, 300–301 Prephenate dehydrogenase 1 PD1, 140, 141, 143, 144 pd1, 140, 143, 144 Probe, 54, 55, 61, 62, 64, 65 Promiscuous DNA, 109 Prosphytochloa, 222, 230 Protein bodies, 360, 361, 370 PB-I, 360–362 PB-II, 360–362, 365 Proton pump F0F1 ATPases (F-ATPases), 63 Proton pyrophosphatase (H+-PPase), 63 Pseudogenes, 27 Pseudomolecule, 8, 56, 57 p-SINE1 RA subfamily, 279–280 RAα, 279, 280 RAβ, 279, 280

380 Q qSH1, 269 Quantitative trait loci (QTLs), 295–306 domestication-related QTL, 269–271 linkage of QTL, 270, 271 mapping, 269 Quantitative traits, 302, 305

R RAP database (RAP-DB), 61, 62 rdr2, 25, 33, 34 Reactive oxygen species (ROS), 339, 341–342, 348–350 Recombinant inbred lines (RILs), 302, 304 Regeneration ability, 296, 299 Regulator of G protein signaling AtRGS1, 140, 143, 145 Atrgs1, 140, 143, 145 Repetitive sequence, 236, 241 Reproduction, 191, 192, 201, 202 Restriction fragment length polymorphism (RFLP), 7, 54 Retrograde regulation, 213 Retrotransposon, 114, 236, 239–242 Reverse genetics, 95, 99, 100, 103 tool, 99 Reverse transcriptase polymerase chain reaction (RT-PCR), 65 Rf1, 206, 208–212 Rhynchoryza, 222 Rice Annotation Program 1 (RAP1), 61 Rice Annotation Project (RAP), 14, 15, 17–19 Rice Annotation Project Database (RAP-DB), 14, 16–20 Rice blast, 31, 299, 300 Rice Cyc initiative, 63 Rice Expression Database (RED), 55 Rice Genome Resource Center (RGRC), 62 Rice Microarray Opening Site (RMOS), 54, 55 RNAi, 95, 103

S 35S enhancers, 97, 98 Salinity tolerance, 298 Seed dormancy, 300–301 longevity, 302 shattering, 296–298

Index Seed-specific promoter, 358 globulin, 359 glutelin, 359 prolamin, 359, 360 Shattering gene sh4, 269 Shootless mutant, 154–156 shl, 154, 156 shl1, 154 shl2, 154 shl3, 154 shl4, 154, 155 Signal transduction, 121, 122, 131, 132 SINE p-SINE1, 278–286, 288, 289 p-SINE2, 278, 279, 286 p-SINE3, 278, 279, 286 SINE code AA genome code, 285–289 GT code, 286–289 Or code, 286, 288, 289 Os code, 286, 288, 289 Single nucleotide polymorphism (SNP), 23, 30, 297 Small interfering RNAs (siRNAs), 24, 25, 32–34, 65 Small RNA, 24, 25, 28, 32–34 SOS pathway, 345–347 AtNHX, 346, 347 Speciation genes, 247, 254, 256 Species, 235–243 Stem cell, 153 RAM, 153 SAM, 153 STRUCTURE, 284 Submergence, 296, 298, 304 Subspecies, 53 Superfamily, see Transposable element Superhybrid cultivar, 54

T TD, 72, 74, 75 T-DNA, 82–87, 95–103 Thylakoid formation 1 THF1, 140, 141, 143, 145 thf1, 140, 143, 145 TIGR v4.0, 23, 24, 28 TIGR’s rice genome annotation database (Osa1), 57 Tos17, 100–102 Transcription factor, 359, 362 Transcription unit (TU), 56, 61

Index Transcriptome, 37, 39–41, 44, 45, 48, 53, 59, 62 Transmission ratio distortion (TRD), 249, 255 Transposable elements (TEs), 27, 34, 57, 59, 65, 82, 83, 89, 114 aDart, 76 CACTA, 70, 72 dDart, 76 hAT, 70, 76 Helitron, 9 iDart, 76 mPing, 70, 72–77, 83 Mutator, 70, 71 Mutator-like elements (MULEs), 27 nDart, 76, 77, 89, 91, 92 PIF/Harbinger, 70, 71 Ping, 72–74, 77 Pong, 72, 73 Rim 2/Hipa, 72 Tc1/mariner, 70, 71 Tnr1, 70, 71 Tnr2, 70 Tnr3, 70 Tnr5, 70 Transposon see Transposable element

U Ultraviolet-B resistance, 296, 299 UniProt Knowledgebase (UniProtKB)

381 V Voltage-gated ion channel (VIC), 63

W WA-CMS, 205–208, 212 Wide compatibility, 253

X Xa21, 7 Xa21 resistance gene (R-gene), 31 Xanthomonas, 7

Y Yeast artificial chromosome (YAC), 54 Yield and hormone brassinosteroid, 314–315 cytokinin, 315–316 gibberellin, 310–314 YS1, 325, 326 YSL, 328

Z ZIP, 326–329 IRT, 328 Zizania, 222 Zizaniopsis, 222

Rice biology in the genomics era (Biotechnology in Agriculture and Forestry, Vol. 62)

Rice Improvement in the Genomics Era

Rice Improvement in the Genomics Era

Genetic Modification of Plants: Agriculture, Horticulture and Forestry (Biotechnology in Agriculture and Forestry)

Genetic Modification of Plants: Agriculture, Horticulture and Forestry (Biotechnology in Agriculture and Forestry)

Transgenic Crops VI (Biotechnology in Agriculture and Forestry)

Biotechnology in Agriculture and Forestry. Transgenic Crops V

Cotton: Biotechnological Advances (Biotechnology in Agriculture and Forestry, 65)

Allelopathy in Sustainable Agriculture and Forestry

Plant Biotechnology for Sustainable Production of Energy and Co-products (Biotechnology in Agriculture and Forestry, 66)

Microorganisms in Sustainable Agriculture and Biotechnology

Biotechnology, Genomics and Bioinformatics

MYCORRHIZAE SUSTAINABLE AGRICULTURE AND FORESTRY

Agricultural Biotechnology in International Development (Biotechnology in Agriculture Series)

Genomics: Applications in Human Biology

Genomics: Applications in Human Biology

Genomics: Applications in Human Biology

Genomics: Applications in Human Biology

Pharmaceutical Biotechnology (Advances in Experimental Medicine and Biology Vol 655)

Molecular Genetic Approaches to Maize Improvement (Biotechnology in Agriculture and Forestry, Volume 63)

Molecular Biology and Genomics

Molecular biology and genomics

Tobacco BY-2 Cells: From Cellular Dynamics to Omics (Biotechnology in Agriculture and Forestry)

Plant Functional Genomics (Methods in Molecular Biology Vol 236)

Bacteria in biology, biotechnology, and medicine

Bacteria in Biology, Biotechnology and Medicine

Microengineering in Biotechnology (Methods in Molecular Biology)

Life As Surplus: Biotechnology and Capitalism in the Neoliberal Era

Molecular Biology and Biotechnology

Molecular Biology and Biotechnology

Advances in Genetics, Genomics and Control of Rice Blast Disease

Rice biology in the genomics era (Biotechnology in Agriculture and Forestry, Vol. 62)

Rice Improvement in the Genomics Era

Rice Improvement in the Genomics Era

Genetic Modification of Plants: Agriculture, Horticulture and Forestry (Biotechnology in Agriculture and Forestry)

Genetic Modification of Plants: Agriculture, Horticulture and Forestry (Biotechnology in Agriculture and Forestry)

Transgenic Crops VI (Biotechnology in Agriculture and Forestry)

Biotechnology in Agriculture and Forestry. Transgenic Crops V

Cotton: Biotechnological Advances (Biotechnology in Agriculture and Forestry, 65)

Allelopathy in Sustainable Agriculture and Forestry

Plant Biotechnology for Sustainable Production of Energy and Co-products (Biotechnology in Agriculture and Forestry, 66)

Microorganisms in Sustainable Agriculture and Biotechnology

Biotechnology, Genomics and Bioinformatics

MYCORRHIZAE SUSTAINABLE AGRICULTURE AND FORESTRY

Agricultural Biotechnology in International Development (Biotechnology in Agriculture Series)

Genomics: Applications in Human Biology

Genomics: Applications in Human Biology

Genomics: Applications in Human Biology

Genomics: Applications in Human Biology

Pharmaceutical Biotechnology (Advances in Experimental Medicine and Biology Vol 655)

Molecular Genetic Approaches to Maize Improvement (Biotechnology in Agriculture and Forestry, Volume 63)

Molecular Biology and Genomics

Molecular biology and genomics

Tobacco BY-2 Cells: From Cellular Dynamics to Omics (Biotechnology in Agriculture and Forestry)

Plant Functional Genomics (Methods in Molecular Biology Vol 236)

Bacteria in biology, biotechnology, and medicine

Bacteria in Biology, Biotechnology and Medicine

Microengineering in Biotechnology (Methods in Molecular Biology)

Life As Surplus: Biotechnology and Capitalism in the Neoliberal Era

Molecular Biology and Biotechnology

Molecular Biology and Biotechnology

Advances in Genetics, Genomics and Control of Rice Blast Disease

Recommend Documents