Neuroendocrine Control of the Thymus (Neuroimmunomodulation, 1-2)

Neuroimmunomodulation 1999;6:5 Introduction Previously the journal published original articles, short reviews and occa...

Author: W. Savino

185 downloads 892 Views 8MB 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

Neuroimmunomodulation 1999;6:5

Introduction

Previously the journal published original articles, short reviews and occasionally abstracts of reports presented at selected meetings. We have instituted a new policy of publishing not only full papers from selected neuroimmunomodulation meetings but also issues devoted to specific topics within the discipline. This is the first of the latter type of issues: an issue devoted to the role of thymus gland in neuroimmunomodulation. Dr. Wilson Savino is the Guest Editor of this series of excellent papers. We thank him for his efforts in developing an excellent contribution to the journal Neuroimmunomodulation. The Editors S.M. McCann J.M. Lipton

ABC Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

© 1999 S. Karger AG, Basel

Accessible online at: http://BioMedNet.com/karger

5

Neuroimmunomodulation 1999;6:6–22

Use of Cultured Thymic Tissues for the Regeneration of the Thymus G.V. Jones

C.A. Botham

M.D. Kendall

The Thymus Laboratory, The Babraham Institute, Cambridge, UK

Key Words Transplantation W Thymus W Thymic epithelial cells W Culturing thymic cells

Abstract An account of research conducted on the transplantation of thymic cells and tissues in order to restore the functional activity of the thymus is reviewed, and discussed in the context of current concepts. Although most rodent work has been conducted on the transplantation of cultured fragments under the kidney capsule, human transplantation studies and models have used other sites or other species such as the severe combined immunodeficient mouse as hosts. The factors affecting the growth of thymic cells in culture is considered in detail since the methodology can strongly influence the cells favoured under culture conditions. An extension of this work to characterize both thymic fragments implanted under the kidney capsule of rats and cultured thymic cells has recently led to the appreciation that the adult thymus must contain a small number of neural crest-like cells. These cells have a high level of proliferation in the implanted fragments, expand in culture, and belong to a family of cytokeratin-positive cells exhibiting immunoreactivity for a wide range of neuropeptides and transmitters. Thus primary cultures of thymus can contain a wide range of glia-like cells.This can be explained by the fact that the thymus, in addition to having a mesenchymal neural crest component, is probably derived from car-

ABC

© 1999 S. Karger AG, Basel 1021–7401/99/0062–0006$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: http://BioMedNet.com/karger

diac neural crest. Closely associated neural crest also has glia-like properties (the supporting cells of the enteric nervous system). These finding can account for the large number of reports of epithelial cells of the adult thymus being immunoreactive to antibodies raised to neuroendocrine and neurotransmitters. Neuroimmune interactions in the thymus are more fundamental than previous work had suggested.

Introduction

With the increased awareness that a lack of CD4 cells in acquired immunodeficiency syndrome (AIDS) patients can lead to death, there is a rekindling of interest in the functional capacity of the adult thymus, and the possibility of regenerating thymic tissue in immunocompromised patients. These are not new ideas, but research was not actively pursued in these areas until comparatively recently. Now, animal work has led to a better understanding of the transplantation of thymic tissue within a species and as a result, transplantation across the species barrier is now possible. In this review we firstly document the steps that have taken place in the history of thymic transplantation work, then consider factors and techniques that influence the culturing of cells for implantation and conclude with considering the implications for successful transplantation, of our own most recent work documenting neural crest-like

Prof. Marion D. Kendall The Thymus Laboratory, The Babraham Institute Cambridge, CB2 4ET (UK) Tel. +44 1223 832312, Fax +44 1223 837912 E-Mail [email protected]

One of the first reports of thymic tissue being successfully transplanted was made in 1963 [1]. They implanted millipore diffusion chambers containing either embryonic or neonatal thymic tissue intraperitoneally into neonatally thymectomized mice. The capacity of these animals to reject skin grafts, to form IgG and IgM antibodies to sheep erythrocytes, and to repopulate lymphoid organs was demonstrated. Further work [2–4] supported these findings in the case of ‘restricted’ thymic implants, even to the point of being able to maintain growth in thymectomized animals [2]. In all of these cases the donor tissue was syngeneic in origin. Early attempts to reconstitute neonatally thymectomized mice by the use of xenogeneic neonatal thymic tissue from rat were not successful in developing immunological competence [5], but were ‘successful’ in that the graft developed [5, 6]. Later work by Law [7] demonstrated the restoration of thymic function in neonatally thymectomized mice bearing xenogeneic thymic grafts. The conclusion of this report was that the restoration of an immunologic function was the result of a humoral factor which depended on the presence of surviving epithelial-reticular elements of the xenogeneic thymic graft. This humoral factor was then capable of inducing the maturation of residual host lymphoid cells and of re-establishing lymphopoiesis. It was suggested that this humorally mediated mechanism is not species specific. Further supporting evidence for the ability of xenogeneic (and isogeneic) thymic grafts to restore an immune response in neonatally thymectomized mice was provided in 1969 [8]. In this instance both free grafts and those contained within cell impermeable millipore chambers were successful in inducing neonatally thymectomized mice to reject a rat carcinosarcoma. No other isogeneic tissues from sources other than the thymus were effective in restoring this response. At this time the work on thymic transplantation was limited, due to reliance upon the use of neonatally thy-

mectomized animals as recipients. This in itself created a number of problems, especially that complete thymectomy was not always guaranteed as alluded to by a number of workers [e.g. 1, 3]. It was not until the appearance of that most useful of biological tools, the nude mouse/rat, that progress in the area of thymic transplantation was extended. In the mid-1970s Richard Hong and his colleagues began to fully exploit the nude mouse in their studies on thymic fragment and epithelial cell transplantation. It had already been shown by others, using different tissues, that culturing tissues and organs in vitro prior to transplantation into an allogeneic host enabled acceptance of transplanted ovaries [9] and thyroid [10]. It was suggested that the acceptance of these allografts were due to the removal of ‘passenger leucocytes’ that had been trapped in the graft. These ‘passengers’ are essential for the sensitization of the recipients’ T lymphocytes. During in vitro culture of the organ, the leucocytes are lost and removed from the culture system. In the thymus, the most likely source of these passenger leucocytes were suggested to be derived from the macrophage population. Hong et al [11] initially transplanted thymic epithelium cultured for 18 days into a 10-month-old female suffering from severe combined immunodeficiency disease (SCID). Prior to transplantation, the recipient’s IgG count was around 35 mg/dl and IgA and IgM counts were undetectable. After combined intraperitoneal and intramuscular transplants of cultured thymic epithelial cells, the IgM counts rose to a maximum of 870 mg/dl, IgA to a maximum of 1300 mg/dl and IgG to 11,000 mg/dl. Following the success of this treatment a further preliminary study was performed [12], whereupon the transplantation of cultured thymic tissue was investigated across a species barrier namely BalbC mice and rabbits. The recipient mice were able to accept the xenogeneic transplants and after a 10-week period there was repopulation of the xenograft involving lymphocytic and corticomedullary differentiation. There subsequently followed a series of papers by Richard Hong and co-workers which investigated in greater detail the processes involved in the transplantation of cultured thymic fragments. It was demonstrated [13] that there was a progressive loss of lymphoid cells during the cultivation of thymic fragments and that by 18 days most free cells were gone, although the epithelial cells were still well developed and contained prominent nuclei. Two percent of the cells that had left the fragment by 20 days were macrophages. Upon transplantation using cultured thymic fragments (CTF) from syngeneic, allogeneic and xenogeneic donors, there

Transplantation of Thymic Tissues

Neuroimmunomodulation 1999;6:6–22

cells within the thymus. The finding of such cells in the adult rodent thymus should lead to better success in culturing cells or fragments prior to manipulation, and this opens the possibility of autotransplants that should eliminate, or at least simplify, any potential rejection processes.

Past and Present Perspectives of Thymic Transplantation

7

was lymphoid repopulation in the majority of animals and in the case of some there was evidence of Hassall’s corpuscles. In contrast, noncultured fragments did not develop. It is suggested that the acceptance of thymic implants are due to the fact that macrophages (which are present in relatively high numbers) are lost during the cultivation process. It is further argued that the lymphocytes in the fragment represent repopulation as opposed to chronic or subacute rejection. Best results were consistently achieved using syngeneic neonatal donors [14]. Then investigations [15] were concentrated upon the ability of CTF to reconstitute nude mice using both syngeneic and allogeneic transplants. From this it is concluded that CTF consisting primarily of epithelial elements could effectively repair the thymic deficiency of nude mice, and that syngeneic fragments had no advantage over allogeneic fragments in the restoration procedure. Further work demonstrated [16] that nude mice receiving CTF from two mouse strains tolerated the H-2 of the donor strains as assessed by skin graft acceptance. Notably, when CTF from the two separate strains were implanted simultaneously under the kidney capsule of the nude mice, skin grafts from the donor source were not accepted. In a continuation of this study [17] it was demonstrated that restoration of IgA, IgE, haemagglutinating responses and mitogenic proliferation in nude mice were dependent upon the amount of thymic tissue being transplanted, and that the greater the mass of tissue, the greater the response elicited. This is seen to correlate well with the observed heterogeneity of the Di George syndrome and the amount of thymic tissue available [18] and in human CTF transplantation studies [19]. It is now quite remarkable that after the elapse of nearly 30 years since CTF transplantation was first attempted as a cure for complete Di George syndrome sufferers, it is only in 1997 [20] that the first successful report of the use of CTF to treat complete Di George syndrome was recorded. As the authors concede, their work does not irrevocably confirm that CTF transplantation was causal in the recipient’s recovery since spontaneous recovery cannot be ruled out. Considering the need for good innervation it is rather surprising that these researchers obtained such encouraging results given that the CTF were implanted into the recipients’ quadriceps muscles. The next investigation in the series [21] examined the ability of xenogeneic (rat) CTF to reconstitute the T-cell function of the nude mouse. They showed that mice given CTF from rat survived longer than control nude mice, and that these mice with implants could reject allogeneic skin grafts (but did not reject donor skin grafts). Further-

8

Neuroimmunomodulation 1999;6:6–22

more, their results suggested that the tolerance demonstrated to the mixed lymphocyte response (MLR) is specific to the major histocompatibility complex (MHC) of the thymus donor. In the same study, experiments to ascertain the ability of xenogeneic CTF to cause antibody production was inconclusive, with only partial restoration of the antigenic response. The authors’ overall conclusion was therefore tempered to acknowledging that xenogeneic transplantation of CTF was only partially successful in restoring immune function to nude mice. Much other work was conducted during this period [22–24] and indicated a lack of consistency. It appears that thymocytes in the thymus can select for the entire or partial population of, or none of the T-cell self-receptors [25]. In light of their previous work, Markert et al. [20] investigated whether partial H-2 disparity between donor and recipient affected allotolerance. Unlike others [26] it was shown [25] that allotolerance to class I antigens was achieved regardless of the extent of the H-2 disparity between donor and host. These differences can be explained mainly in terms of the fact that there are intrinsic differences between tolerance in neonates and adults. With the onset of AIDS, research in the field of thymic transplantation turned towards investigating ways of reconstituting an immunocompromised immune system. Early work on this aspect included the use of cultured human thymic epithelial cells to induce CD4 expression on mammalian cells from AIDS patients in vitro [27]. It is well documented that one of the most clear-cut characteristics of immunity in AIDS is the inability of peripheral blood T lymphocytes to express CD4 (T4, Leu –3). It is concluded that when mononuclear cells from an AIDS patient were incubated for 2 h on thymic epithelial monolayers, there was an increase in CD4 expression. When this work was performed in AIDS patients (in vivo), success was only marginal as the development of the disease was only delayed. It should be noted that at this time, CTF were only being implanted into the deltoid muscles of patients [27]. Further work [28] reinforced the previously described work which showed that transplantation of CTF into nude mice and rats could potentially lead to a fully developed cellular immune system. They progressed to demonstrate that in the rat, the haplotype (of the donor), e.g. PVG (RT1c), RP (RT1p(u,1)) or DA (RT1a), bears no relevance in the generation of the alloantigen recognition repertoire. A follow-up study [29] proved that CTF after implantation under the kidney capsule of athymic nude rats become populated with both lymphocytes and RT1 class II-positive cells of the recipient. These fragments prior to implantation did not con-

Jones/Botham/Kendall

Table 1. Immunohistochemical identification of the cellular diversity present in thymic organ fragments, pre- and postimplantation

Tissue/cell types present

Prior to implantation

Two weeks after implantation

Three to four weeks after implantation

Elongated epithelial cells

Yes

Yes

Yes, abundant

Yes, found ubiquitously

HIS 37+ve

Yes, surrounding central macrophages Yes, present in the centre of implant Yes, majority of epithelium

Yes, majority of epithelium

HIS 38+ve HIS 39+ve Blood vessels

A few individual cells Yes, majority of epithelium None described

A few individual cells Yes, majority of epithelium Yes, some newly formed

Nerve fibres

None described

None described

Mast cells Lymphocytes Eosinophilic granulocytes Fibroblasts Collagenous matrix

None described None described None described None described Yes, surrounding central tissue None described

Groups found deep in implant A few often close to blood vessels Mature cells None described Yes, interstitial

Yes, mid and deep cortex, a few medullary Yes, cortical epithelium Yes, subcapsular and medullary regions Yes, corticomedullary junction and connective tissue Yes, mainly in association with blood vessels and connective tissue Yes, mainly in the collagenous matrix Yes Yes Yes, mainly in the collagenous matrix Yes, as connective tissue between the lobules Yes, multinucleated, often associated with fibroblasts

Macrophages

Giant cells

Yes, multinucleated

tain dendritic cells, however after 2–4 weeks’ implantation, these cells were present by immunohistological techniques. The appearance of the dendritic cells coincided with the influx of lymphocytes into the graft and the subsequent restoration of thymic architecture. This supports the previously suggested role for ‘passenger leucocytes’ in the generation of T cells within the thymic microenvironment. Further investigative work on the composition of the thymic microenvironment was performed in association with this laboratory [30]. Thymic fragment morphology was examined both ultrastructurally and at the light microscope level as well as immunohistologically, prior to implantation, 2 weeks after implantation and 3–4 weeks after implantation. The diversity of cell types present during the development of the implant is summarized in table 1. The effects of implanting CTF from two different strains under the kidney capsule was examined [31] in what was essentially a repetition of other work by Hong and Klopp [16]. The authors also found that double implantation resulted in the rejection of skin grafts from either of the CTF donor strains. This implies that this allogeneic rejection feature may be related to the chimaeric state of the allogeneic CTF after regeneration of the thymic microenvironment. They proposed that this was due to a peripheral suppression mechanism which in vivo

reduced the alloreactivity of the response. It is suggested that this peripheral suppression mechanism (if present) is not sufficient to suppress the alloreactivity generated in a CTF which expresses MHC antigens irrelevant to the alloreaction reducing determinants. Work on the mouse had previously demonstrated that it was possible to transplant embryonic CTF into euthymic mice, successfully crossing the MHC barrier [32]. This was achieved by incorporating deoxyguanosine (dGuo) into the culture medium into which the fragments were grown. The authors speculated that the dGuo caused the removal of the dendritic cells (macrophages?) from the donor thymic fragment. In this study, the loss of the dendritic cells was only inferred, not demonstrated. In another attempt [33] to repeat this experiment in the euthymic rat using adult thymus instead of foetal, and in addition to using dGuo, the authors also used cyclosporin A which had been previously demonstrated to deplete the thymus of dendritic cells. Whilst the syngeneic and allogeneic CTF grafted into athymic rats were successful, only syngeneic CTF were successfully transplanted into euthymic rats. The allogeneic transplants into the euthymic rats were rejected. It is concluded that the difference between rat and mouse is due to the inability of cyclosporin A or dGuo to remove dendritic cells from CTF.

Transplantation of Thymic Tissues

Neuroimmunomodulation 1999;6:6–22

9

A further conclusion that must therefore be drawn from this particular study is that cyclosporin A has completely different effects on the dendritic cell populations of rat thymic tissue, depending on whether it is used in vivo or in vitro. Cyclosporin A, it is suggested, only removes macrophages in vivo, not in vitro. The involvement of nerves in the neuroimmunological thymus has been well reviewed [34, 35]. The innervation demonstrated is both noradrenergic and peptidergic. Thymus cells have also been shown to be positive for a large number of neuroendocrine markers [33]. It had also been demonstrated in 1987 [36] that autonomic innervation is a fundamental feature of thymic tissue transplanted into nude mice. Their work identified both myelinated and nonmyelinated nerve fibres in transplanted CTF. They showed that nerves into the parenchyma were unmyelinated but myelinated in the interlobular parenchyma of the transplanted thymic tissue. These authors also observed a network of subcapsular nerves terminating amongst the thymocyte population, whereas intrathymic nerves entered the parenchyma in association with the vasculature. The nerves detected were acetylcholine esterase positive in the subcapsular cortex at the corticomedullary boundaries. Later in development catecholaminergic (CA) innervation was noted, with CA-positive fibres entering the transplant along the vasculature, the septum and the interface between the gland and the kidney. These fibres then formed perivascular plexuses at the corticomedullary junctions. As previously mentioned, the presence of nerves in growing implants was also noted by Kendall et al. [30]. Nerves were found in bundles near large blood vessels. The majority of nerves were unmyelinated with frequent Schwann cells visible. Other small unmyelinated nerves and their Schwann cells were noted running in the connective tissue; these nerves were generally free in the tissue, and were not associated with blood vessels. Further investigations into the neuroendocrine component of the rat thymus [37] studied CTF before and after transplantation into athymic and euthymic rats. As intimated previously [30] there appears to be an unusual type of epithelial cell present in the CTF prior to and after implantation which is not often seen in adult thymus. This undifferentiated ‘epithelial’ cell has been investigated using a range of neural and pituitary hormone-related markers. The results they subsequently obtained indicated the presence of epithelial-like cells that were positive to protein gene product (PGP) 9.5 in syngeneic transplants prior to, and 1 week after implantation. Furthermore, considerable tyrosine hydroxylase (TH) staining of nerve fibres and reticu-

10

Neuroimmunomodulation 1999;6:6–22

lar structures were observed in relation to cortex-like areas, connective tissues, and those areas associated with blood vessels. It is concluded that there are many neuroendocrine cells in normal thymic tissue, and these are lost during the culturing process, resulting in an epithelial precursor type of cell, that upon subsequent transplantation develops into the heterogenous thymic epithelial cell population normally seen in the adult thymus. The redevelopment of heterogeneity is coincident with a high immunoreactivity for pituitary hormones. This may therefore indicate a potential role of neuroendocrine cells of the epithelial microenvironment in restoring thymic architecture. Recent work at our laboratory has clearly demonstrated that there is a small population of cells that persist in the thymus, are present in cultured thymic fragments and proliferate in the transplanted fragments, slowly diminishing as the fragment develops. These particular cells are positive to not only neural crest cell markers such as HNK-1 and A2B5, but are also positive for a number of neural markers and neural cell adhesion molecules including L1, TH, PGP 9.5, S-100, CGRP. The findings are summarized in table 2. It is our conclusion that these cells are multipotent cells of neural crest lineage, which eventually have the ability to develop into either neural or epithelial type cells. Work in this decade has concentrated in the main on the transplantation of cultured and nonthymic tissue across a discordant or xenogeneic transplantation barrier. The successful use of noncultured human thymus/liver grafts reconstituted T-cell and IgG parameters in nude BNX mice and CB-17 SCID mice [38]. In BNX mice in particular, elevated levels of mouse IgG parameters were observed and a specific immune response to KLH was induced. A chimaeric population of mouse and human thymocytes was also noted immunohistochemically in the thymus with the murine thymocyte being located within discrete areas within the human thymic graft. Others [39] also transplanted embryonic pig thymus cells into thymectomized T and NK cell deplete C57BL/10 mice. They were able to demonstrate mature mouse CD4+ cells in the pig thymus that subsequently migrated to the mouse’s peripheral system. Furthermore no anti-pig IgG response was produced. More recently it has also been shown that the thymus is essential in ensuring that mouse CD4+ cells are developed in nude mice. These grafted nude mice were also able to accept donor SLA-matched (paternal) skin, whilst rejecting allogeneic mouse skin and SLA-mismatched skin. The interest in the use of thymic fragments has now changed emphasis in contemporary studies. In-

Jones/Botham/Kendall

Table 2. Comparison of neuromarker

immunoreactivities between postnatal thymic tissues

Antibody

Adulta

Implantb

Fragc

Comment

L1 TH HNK-1 ChromA PGP 9.5 S-100 NGF GAD CGRP Thymulin A2B5 Cytoker

+b +b +b NTd + + NT – + + + +

+ + + + + + + + + – + +

+ + + NT + NT + + + + NT +

Co-localized with polyTH Less TH+ve than L1+ve Co-localized with polyTH Co-localized with TH Co-localised with L1 Co-localized with L1 Co-localized with L1 Subcap/med. epith. (adults) Medullary epith. (adults) Weak in early fragments

Cell morphologies: a = epithelial/reticular (up to 120 Ìm long); b = rounded cells (aver. diam. 8.7 B 0.6 Ìm); c = shapes – rounded, neural, epithelial; d = not tested. + = Immunoreactive cells present (not quantitative); – = no immunoreactive cells. Taken from Jones et al. [132].

stead of using thymic fragments to reconstitute immune response, a number of research groups have investigated the use of thymic fragments to induce a degree of tolerance to xenograft tissue. This approach has now been taken a step further by others [40] who have started to investigate the potential for using thymic autografts to create a new type of organ to facilitate the tolerance of xenotransplantation. They autotransplanted thymic tissue under the kidney capsule of donor piglets; these transplants subsequently (after 20 weeks) developed to express all the characteristics of a normal thymic microenvironment. If successful this ‘thymo-kidney’ could be exploited as a new source of tissue for xenotransplantation into humans displaying end-stage renal failure.

Culturing Thymic Epithelial Cells

The objective of isolating cells and their subsequent growth in vitro enables the study of cellular interactions and subsequent lineage diversification in the absence of other cell types, whose presence may influence the cells in vivo, via the production of hormones and cytokines. Manipulation of the cellular environment, in particular media components and nutrients, can result in the expression and production of hormone and cytokines or neuronal phenotypes. Raising heterogeneous and homogeneous cell lines from the thymic microenvironment has enabled

Transplantation of Thymic Tissues

researchers to uncover the complexities of the neuroendocrine-immune function of this gland, in particular the identity of biomolecules involved in the induction of plasticity and the expression of cytokines, neurotransmitters/ peptides and hormones and the role these play in the proliferation and education of thymocytes. Cell culture has also been instrumental in discovering cellular lineage and differentiation, as well as identifying different cell types and their role within this intricate system. Characterization of cell lines can be described genetically, by morphology, structure and expression of biomolecules. Both lymphoid and nonlymphoid cells from the thymus have been isolated and cultured over differing periods of time, ranging from several weeks to several years. Cell culture techniques have been established for each stromal cell type which includes six different subsets of epithelial cells [41– 51], macrophages and dendritic cells [52–54] and thymic nurse cells [55, 56] Techniques have also been established for thymic stromal cultures, which encompass all cell types and maintain heterogeneity [57, 58].

Raising Epithelial Cell Cultures

The initial harvesting of thymic stromal cells can be achieved by various methods. The two most common methods are based on mechanical then total enzymic dissociation, using trypsin or collagenase, or by allowing growth from a thymic fragment, attached to a coated cul-

Neuroimmunomodulation 1999;6:6–22

11

ture dish. The former method seems most effective when used in conjunction with other methods such as separation via a Percoll gradient or seeding the dissociated cells upon feeder or seeding cell layers. However, the use of the enzyme trypsin can cause severe damage to the cell membranes whereas collagenase is much more gentle. Certainly, only collagenase is necessary for the dissociation of embryonic thymus. In absence of these other methods, growing cells from thymic fragments pretreated with enzymes provide a culture of ready-dividing cells, growing out in a flat monolayer of cells. The use of stainless steel wire mesh to obtain epithelial cells has only been reported as being successful in the murine foetal thymus, however the yield was low [59]. Either embryonic, neonatal, postnatal or adult thymus can be used, depending on the experimental requirements of the cell culture. The proportion of epithelial cells in the murine thymus has been reported as varying with age [41, 60], which can directly effect the proportion obtained in culture. The thymus of younger animals consists of mainly cortical cells, with only 20% of the epithelial content pertaining to be medullary. As the animal ages or undergoes treatment (e.g. with cortisone), the proportion of medullary to cortical epithelial increases, due to involution. Foetal thymus, depending on the age, consists of undifferentiated epithelial cells and contains less lymphocytes, and therefore is useful for determining cell lineage and diversification/differentiation. Thymus taken from animals up to 8 weeks of age allows the study of the vast heterogeneity of cells found in the adult thymus, however the use of postnatal thymus after this time period results in poor epithelial growth. This can be rectified by the use of triiodothyronine.

Media

The media components for raising thymic epithelial cell (TEC) lines have been the subject of much research, as a tool to manipulate cell cultures, either to encourage the proliferation of one cell type or discourage another. Though previous authors had successfully cultured TEC, the cultures had been adulterated with other cell types. The need for a pure cell culture, consisting of epithelial cells only, triggered the search for the ‘ideal media’. An in-depth study was conducted [43] into the correct method for thymic dissociation plus culture media for human thymic epithelial cells. They concluded that to obtain TEC free from fibroblasts and to prevent differentiation, a serum-free media with epithelial growth factor and calcium were deemed to be essential. Similarly, Small et al.

12

Neuroimmunomodulation 1999;6:6–22

[41] found that adding cortisone to the media of murine thymic stromal cell cultures created positive selection for medullary cells and caused the cortical cells to rapidly deplete in number. The growth of thymic stromal cells from mice was dependent on transferrin, insulin, hydrocortisone, epidermal growth factor (EGF), high-density lipoproteins, selenium and 3,5,3)-triiodothyronine, in the absence of serum, as reported by Eshel et al. [46]. The elimination of any of these factors or change in the extracellular matrix resulted in significant changes in the stromal populations of the monolayers. The use of cholera toxin is also an essential component in cellular growth as is the use of serum-free media to establish a pure epithelial monolayer [48]. Five growth factors in serum-free media were tested for their respective positive and negative effects on cell proliferation and production of cytokines in a human TEC [61]. They concluded that transferrin and cholera toxin were superfluous as they impeded proliferation. Hydrocortisone, while stimulating proliferation, suppressed cytokine production and differentiation. The final two, EGF and insulin, were deemed essential for prevention of cell death and enhancing differentiation and production of cytokines, respectively. These findings highlight the different nutritional needs of cells from different species. Basic fibroblast growth factor can also have mitotic-inducing effects on human thymic epithelial cells, following its purification from bovine pituitary extract [62]. However, serum-free media can only be used in the short-term culturing of thymic stromal cells, with minimal passaging and requires a mixture of growth factors to instigate proliferation. This is ideal for the preservation of the morphologic and functional characteristics of primary stromal cultures and their subsequent interactions with thymocytes, but will not provide an endless supply of homogeneous material for research purposes. Serum is essential for the long-term culturing techniques. Use of foetal calf serum (FCS) or foetal bovine serum (FBS) dismisses the need for growth factors but increases the risk of viral contamination, in addition to the adulteration of the media with unknown quantities of hormones, cytokines and other macromolecules, which could not only alter the growth pattern or instigate lineage diversification of the cell type under scrutiny, but also encourage the growth of undesirable cell types. To address this problem, researchers have had to find alternative methods for raising long-term TEC cultures. Rat TEC cultures which expressed IL-1 and IL-6 were raised using RPMI media with added insulin, EGF, dexamethasone and 15% FCS [63]. Cells were grown on poly-L-lysine. Purity of this cell line

Jones/Botham/Kendall

was obtained by using a weak trypsin/EDTA buffer solution to instigate the removal of cellular impurities. Longterm epithelial cultures from rat thymus were established with the use of donor horse serum and manipulation of plating techniques and biochemical components, resulting in an IL-6-producing cell type with 95% epithelial purity [51]. It could be surmised from the authors quoted so far that cross-species thymic epithelial cells possess different nutritional requirements. Human TEC seem to have a higher dependency on the addition of selenium and transferrin to serum-free media compared to murine cultures, which depend more on EGF and hydrocortisone, also in serum-free conditions [64]. However, it was noted [46] that mesenchymal cells derived from murine thymus also required selenium as an important nutritional component, indicating that maybe different thymic stromal cells within one species possess different nutritional needs and can therefore be selected or deselected by media components. The differing requirements for supporting proliferation of these cells cross-species could be due to different epithelial cell types rather than species difference. Additives to media can be classed into several groups. The growth factors increase mitosis, depending on cell type, but some will suppress cytokine and hormone production. Other additives such as insulin, cortisone, somatostatin, 3,5,3)-triiodothyronine, cholera toxin and transferrin result in differentiation and plasticity. Serum, as well as a source for the above-mentioned additives, also supplies cholesterol. Substitution of essential amino acid for cell growth, L-valine with D-valine suppresses the growth of fibroblasts in culture [65]. The lowering of calcium ion content in media is essential for the growth of thymic epithelia. High calcium content results in the cessation of proliferation [55]. The choice of media components by researchers should reflect the experimental needs of their research and requires careful consideration, so as not to restrict the biological activities of the cells under examination.

diated mammary tumour cells (LA7), a cell line with epithelial origins. With this system of using ‘support cells’, in the presence of FBS, colonies of TEC can be established from a single cell. Other similar techniques include the use of fibroblast ‘feeder’ layers, whereby epithelial growth is stimulated by production of unidentified molecules. This seems to highlight the density dependence of certain TEC in culture.

Secretion Potential of Cultured Thymic Cells

Another method for isolating and growing TEC is via co-culture with other cell types. Co-plating of irradiated filler fibroblasts (3T3) with freshly dissociated murine thymus following fractionation, to establish a long-term epithelial cell line in the presence of FBS, was achieved in a low calcium ion concentration media [47]. A similar technique was successful [45], whereby long-term proliferating murine TEC were co-plated with lethally irra-

It has long been recognized that the normal functioning of the thymus relies on a three-way mediated loop [66], known as the neuroendocrine-immunological loop [67]. This consists of neuropeptides and transmitters, hormones, growth factors and cytokines. Both lymphoid and nonlymphoid cells are capable of excreting and expressing receptors for growth factors and cytokines, but research suggests that the majority of the expression of these biomolecules is via medullary and subcapsular epithelial cells [61, 63]. The culture of these cells has enabled researchers to identify a small number of pathways which trigger Tcell education/proliferation, or epithelial regeneration. Cytokines and growth factors identified in vitro have several biological roles. This includes the induction and modulation of other cytokines (IL-1, TGF-ß, IL-6), proliferation and differentiation of T cells (IL-1, IL-2, IL-3, G-CSF, M-SCF, IL-6, IL-7, IL-12, IL-15, SCF, TNF-·, IFN-Á, IRAP-2, lymphotoxin), proliferation of TEC (IL-1, IL-6) and inhibition of proliferation of T cells (TGF-ß) [68, and unpubl. data]. Lymphotoxin and IFN-Á are produced exclusively in T cells, CSFs and IRAP-2 are expressed by stromal cells only, the remainder of cytokines being produced by T cells and stromal cells, creating an overlap situation [69]. Several researchers have reported that many of these factors rely on a synergistic effect between themselves [70] and other factors, such as EGF and TGF-ß, inducing IL-1 and IL-6 biological activity and mRNA levels for IL-1·, IL-1ß and IL-6 genes [71, 72]. IL-1· and IL-1ß induce TEC to produce molecules which cause thymocytes to proliferate [73]. It has recently been reported that TGF-ß can modulate the induced effects of LIF, IL-6, IL-1·, IL-1ß in thymic epithelial cell cultures, suggesting that this growth factor has a regulatory role in cytokine production by TEC and T-cell proliferation [74]. Certainly, growth factors incorporated in culture media also exert an induction or inhibitory effect on the expression of cytokines. What has become apparent from this research is the fact that developing T cells play a

Transplantation of Thymic Tissues

Neuroimmunomodulation 1999;6:6–22

Co-Culturing

13

vital role in their own education via feedback of cytokines from themselves to the stromal cells [75]. In addition to the production of growth factors and cytokines, it has also been well documented that the epithelial cells in culture are capable of expressing a number of thymic hormones. These hormones are known to play an important role in neuroendocrine immunomodulation and maturation of T cells and possess restorative capabilities in situations where immunocompetence has been compromised [76]. Thymic epithelial cells in vitro can express thymosin-·1, thymopoietin and thymulin, localized in cytoplasmic granules of epithelial cells [77]. Many authors have reported that the expression of these hormones is regulated by extrathymic biomolecules including pituitary hormones [78], steroids [79] and thyroid hormones [80]. Intrathymic biomolecules also associated with modulation of thymic hormones include growth hormone and insulin-like growth factor-I [81], IL-1 [82] and neuropeptides [83]. It has also been reported that thymic hormones can exert an effect on gonadotrophin expression in pituitary cells in culture, illustrating that a feedback link exists between the two [84]. In our own laboratory, we observed an increased expression of the thymic hormone thymulin in stromal cells following dissociation, which generally decreased once the cultures had reached confluence. Expression also varied between cell morphological types, with small highly mitotic cells being much more immunoreactive. These observations, coupled with the fact that immunoreactivity to thymulin has been witnessed in the glial cells of the cerebellum [85], has led to speculation whether thymulin could also be linked to a neuronal role. The third set of biomolecules in the mediated threeway loop of the neuroendocrine-immune system are the neuropeptides and transmitters. Those observed in thymic epithelial culture include oxytocin, vasopressin, neurophysin, somatostatin, neurotensin, PGP 9.5, calcitonin gene-related peptide (CGRP), TH, neuropeptide-Y, metenkephalin, substance P and neurokinin-A [86, and unpubl. data]. In vitro research has verified the existence of receptors for certain neuropeptides on T cells and a link between the sympathetic nervous system and thymocytes has been established [87]. The presence of these substances indicates the extent and the importance to which the normal functioning of the thymic microenvironment is associated with sympathetic innervation. It has already been stated that CGRP and acetylcholinesterase has a suppressive effect on thymulin production in vitro [83]. Several authors have reported that neuropeptides induce excretion of growth factors and vice versa. Nerve growth

14

Neuroimmunomodulation 1999;6:6–22

factor enhances IL-6 gene expression and these two are synergistic in enhancing neurofilament and synapsin expression [88]. Substance P stimulates the production of IL-2 [89], whereas the addition of EGF to TEC cultures triggers neurofilament expression in addition to cytokine expression [90]. The presence of neuropeptides, neurotransmitters, hormones and cytokines are seemingly interrelated to the modulation of one another, to the direct and indirect maturation of thymocytes and the establishment of innervation following disruption to the tissue or involution. In summary, the majority of in vitro research conducted into the neuroendocrine-immunological functioning of the thymus has been based around its role in the education and maturation of T cells and in this respect, knowledge gained is highly fractionated, revealing a highly complex feedback system between the endocrine, neural and immune systems, both intra- and extrathymically, possibly linked by parasympathethic innervation. Many of the growth factors used in culture media elicit a specific response: EGF will induce the expression of cytokines, which in turn will induce other cytokines or effect proliferation of T cells. This response to EGF is modulated by TGF-·. The control of thymic hormones, whose roles are multifarious, seems to be, in part, controlled by neuropeptides, growth factors and IL-1. EGF will also induce certain neuropeptides, involved in T-cell maturation. Thus, it can be seen that the up-regulation of one compartment of the thymic neuroendocrine immunomodulation has an effect on the other two compartments either directly or indirectly.

Culturing Methods of Thymic Organ Fragments

The culturing of thymic organ fragments has been implemented as a method for the preparation of donor thymus prior to implantation. Culturing was found necessary as a preventative measure in averting rejection, where the recipient still possessed partial immunocompetence. Culturing methods between researchers have differed only slightly in the last 30 years, when compared to the changes in culture methods for thymic stromal cells. Donor thymus can be obtained from either foetal or postnatal subjects, foetal usually providing the better material. However, due to the short supply of foetal thymic tissue, postnatal thymus was deemed the more approachable method. Thymus glands are extracted from the donor, the thymic capsule is removed and the remaining tissue is cut into fragments of size approximately 1 mm3. The culture

Jones/Botham/Kendall

Table 3. Results of immunoreactive staining in preimplantation fragments and primary TEC cultures

Primary antibody

Whole thymus

Cultured organ fragment

Primary cell culture

Glial fibrillary acidic protein

Medulla and cortical clusters

Staining on periphery and throughout on stellate, elongate and round cells

Mainly spherical and elongate cells, with a few polygonal cells

S-100

Medulla and cortical clusters

Staining on periphery; stellate, elongate and round

Spherical and elongate cells only

Nerve growth factor

Medulla and cortical clusters

Strongest on periphery; staining on all cell types

Staining apparent on all cell types

A2B5

Medulla and cortical clusters

Staining on periphery; stellate, elongate and round and cobblestone

Staining apparent on all cell morphologies

L1/NCAM

Medulla and cortical clusters

Staining on periphery; stellate, elongate and round and cobblestone

Spherical and elongate cells only

His 37

Cortex and cortical-medullary junction

Peripheral staining and on small clusters throughout cortex

Stellate, elongate and round and cobblestone

His 39

Medulla and cortical-medullary junction

Peripheral staining and throughout cortex on stellate and spherical cells

Stellate, elongate and round and cobblestone

Laminin

Medulla and cortical clusters

Staining apparent throughout the fragment

Staining cobble-shaped cells

media consists of buffered media with added antibiotics and varying concentrations of FBS. The purpose of culturing is solely to initiate removal of macrophages and thymocytes, therefore the culture media is simply to ensure survival of the tissue over the culturing period. The use of FBS will supply the growth factors necessary for the nutrition of the stromal cells, however, it will also encourage the growth of fibroblasts and macrophages and therefore should be used in low concentrations. To facilitate the removal of lymphoid cells, deoxyguanosine can be added to the media and has been shown to be instrumental in successful transplantation into histoincompatible subjects [32, 91]. It has also been reported [41] that deoxyguanosine will stimulate medullary cells and inhibit cortical epithelium, much the same as cortisone. Other researchers have found that at least 40% of the implant has to consist of medullary thymic epithelium, to ensure successful regeneration of the thymic fragment, postimplantation [14, 92]. Therefore the use of deoxyguanosine to diminish T cells from fragments may also help establishment by stimulating proliferation of medullary cells. A buffered media, supplemented with growth factors which would favour the growth of thymic epithelial cells, plus a lower concentration of FBS and addition of deoxyguanosine have been suggested by Hong and Moore [93], which would restrict the growth of macrophages.

The length of culturing time depends on the immunological and genetic status of the recipient. Where donor and recipient are genetically identical, less culturing time is required to prevent rejection. A culture period of 6–8 days was sufficient for implantation into allogeneic rats, according to Schuurman et al. [29]. Restorative implantation into complete immunoincompetent recipients can be assumed without any culturing. After 8 h postremoval, Kollmann et al. [38] found that implantation of human foetal thymus into nude and SCID mice was successful. Incompatible recipients require their implants to undergo either longer culturing times or treatment with deoxyguanosine. In genetically different mice, Schulte-Wissermann et al. [13] found that a culturing period of not less than 18 days was required. It has also been reported that the presence of neuropeptides and thymic hormones play an important role in establishing a fully functional thymus. The presence of PGP 9.5 and pituitary hormones, both pre- and postimplantation, was reported by MartinFontecha et al. [33]. In our own laboratory, we took this characterization a step further, with the use of a large antibody panel and comparison with epithelial cells in culture. The results can be seen in table 3. Fragments were cultured for 9 days prior to implantation under the kidney capsule in syngeneic rats. Preimplantation characterization was conducted on day 9. We found the presence of

Transplantation of Thymic Tissues

Neuroimmunomodulation 1999;6:6–22

15

cells immunoreactive to S-100, PGP 9.5, L1/NCAM, laminin, A2B5 and thymulin, indicating glial-like cells, derived from the neural crest, possessing neuroendocrinal properties. The presence of these cells in our TEC cultures was observed as small, highly mitotic cells, capable of several different morphologies. We concluded that these cells were responsible for the initiation of innervation in implanted thymic fragments, leading to a fully functioning thymus.

Implantation of Culture Thymic Stromal Cells

The problems encountered so far in the implantation of thymic material, such as rejection and the low availability of foetal thymic tissue, has led researchers to speculate whether incomplete immunoincompetence (such as AIDS sufferers) would benefit more from implanted thymic epithelial/stromal cell cultures. Many authors have shown that different subsets of epithelial cells play different roles in thymocyte maturation, and are capable of doing so in vitro. In vitro experiments established that thymic epithelium was capable of inducing T-cell maturation in the absence of thymic lymphoid tissue [94]. Following this, in 1978, Willis-Carr et al. [95] went on to show that human epithelial monolayers and in certain cases the supernatant from these cultures could induce Tlymphocyte maturation, in peripheral blood taken from immunodeficient patients. However, in vivo studies were not so successful. Injection of pure epithelial cell suspensions has resulted in a limited success, whereas, since the late 1960s, implantation of cultured thymic organ had resulted in a fully functional thymus. Several researchers [96, 97] tried a similar method of injecting epithelial cell suspensions into muscle. In the first trial, cultured epithelial cells were injected into AIDS sufferers, resulting in only 1 patient having postoperative life extension for 11 months. The second trial involved injection of cells into a juvenile, born immunodeficient. Survival was extended for 13 months. In both these trials, any immunocompetence restored was transitory and could be related to the lack of parasympathetic innervation. Several researchers have proven that the thymus gland is heavily reliant on parasympathetic innervation for normal neuroendocrineimmunological functioning [36]. After several weeks without reinnervation occurring, the injected cells would have perished. However, the limited success of these trials would suggest that immune restoration is possible, using cultured cells. It has since been established that nonepithelial cells, i.e. cells from a mesenchymal origin, also play

16

Neuroimmunomodulation 1999;6:6–22

an important role in the initial maturation steps [61]. To this end, Patel et al. [98], by using an artificial capillary system, have recently co-cultured thymic fibroblasts and epithelial cells, resulting in a biological structure capable of supporting the development of CD1a-positive cells in vitro. Our own research has focused on the growth of primary stromal cultures in a three-dimensional format, which can be easily implanted. Comparative studies between preimplanted cultured thymic organ fragments and primary cell cultures were undertaken, to established whether the cellular subpopulations were similar. Fragments from cell monolayers were formed by dissociating the monolayers into sheets of cells in the media and allowing these sheets to round naturally into ‘fragments’, ready for implantation. Preliminary results look promising, as after 7 days in culture, the cells are still viable, as indicated by their immunoreactivity to several antibodies. In the future, this method could produce an inexhaustible supply of tissue for implantation.

Plasticity, Maturation and Diversification of Cultured Cells

From the previous sections it is clear that cells in culture can be directed along certain differentiation pathways by variations in technique. However, more fundamental were the findings of Röpke et al. [99] who suggested a common stem cell for both cortical and medullary epithelial cells, after positively identifying cells in murine foetal thymic epithelial cultures, which double-labelled for medullary and cortical markers. Since it has been observed that such ‘double-positive’ epithelial cells were, in many cases, the primary component of thymic tumours, this conclusion had previously been hypothesized by Lampert and Ritter [100]. Such cells are commonly present in rat TEC cultures in our laboratories [51, 83] and indeed we have observed the growth of all types TEC from fragments composed of such cells when they were implanted under the kidney capsule of nude rats [30]. These cells are probably the undifferentiated type 5 epithelial cells found occasionally in human thymus [101]. Fascinatingly, the underdeveloped SCID mouse has a thymus composed primarily of epithelium with the ultrastructural characteristics of these type 5 cells, and they do not differentiate into the other TEC types until the SCID mouse is reconstituted with bone marrow-derived cells [102]. The concept of plasticity as applied to TEC in culture is illustrated by the derivation of TEC from thymic nurse

Jones/Botham/Kendall

cell cultures [103], and the demonstration that this plasticity works both ways, when nurse cell attributes were induced in an established TEC line [104]. In our laboratory we have observed cultures that contain predominantly cobblestone-shaped, rapidly mitosing cells in serum-added media, that are capable of forming medulla-like structures within the cellular monolayer, as described previously [55]. However, when these same cells were placed in serum-free media, their morphology changed into larger polygonal-shaped cells, with a low mitotic rate. On replacing the serum, they reverted to their original form. Several other authors have remarked upon the changing ratios of cell types in culture [46, 105] and this can be reflected in functional changes. Cells from postnatal murine thymus had low cortical epithelial immunoreactivity in serum-free cultures, but on addition of serum the cells reacted with cortical markers [106]. However, it must always be borne in mind in such studies that cultures may contain small numbers of cells of one type that may not easily be noticed under some culture conditions, yet such cells may proliferative with, for example, the addition of serum. However, to overcome this problem, Nieburgs et al. [107] cultured murine thymus and by selection obtained and maintained two clearly defined, morphologically different cell types, which were designated small and large (TECS and TECL respectively). A high incidence of small epithelial cells in cultures has been observed by others. Certainly in our TEC cultures, the smaller cells were greater in number, highly proliferative and were not contact inhibited, compared to the larger polygonal cells, which were contact inhibited and were low in numbers, though some large cells were multinucleate. On further investigation, Nieburgs et al. [108] showed that only TECS possessed the receptor for EGF and gave functional responses to challenge with EGF. In addition, and more significant, is that plasticity appears to extend beyond the development of clearly distinguishable epithelial cell types also studied The secretion of IL-6 in response to EGF stimulation was studied by Screpanti et al. [90]. They noted that EGF promoted the expression of a neuronal phenotype in TEC culture, in addition to enhancing cells that express cytokines. This cytokine is essential for thymocyte proliferation, suggesting that EGF plays an important role in the up-regulation of neuronal expression in cells derived from the neural crest [90]. Similar effects using EGF and TGF-· have also been reported [71]. Immunocytochemical and morphological characterization of primary thymic epithelial cultures for up to 6 weeks in our laboratory shows several categories of cell:

epithelial, neural and glial [unpubl. data]. By carefully controlled conditions these cells have been further subcultured to generate epithelial cell lines that are 98% pure after the 3–4th subculture [51, 83] and secrete thymulin like the type 1 subcapsular/perivascular and some of the type 6 medullary epithelial cells. During primary and early cultures many of the epithelial (cytokeratin-positive) cells also express a range of neuropeptides and transmitters (fig. 1). In particular they are positive in double immunostaining protocols to antibodies raised against laminin, glial fibrillary acidic protein (GFAP), L1/NCAM and A2B5, as is characteristic of neural crest cells. The adult rat thymus also contains a small population of cells that are immunoreactive to L1/NCAM, HNK-1/NCAM and A2B5 (single immunostaining observations). Epithelial cells immunopositive to antibodies raised against TH and CGRP epithelial cells have been demonstrated at both the light and electron microscope levels in humans and rats [109], and acetylcholinesterase-positive cells (in addition to nerves) were found in the rat [30]. Epitheliallike cells with immunoreactivity for S-100 [110], A2B5 [111], GFAP [unpubl. data], PGP 9.5 [112], myelin-associated glycoprotein [113] and NSE [pers. observations of M.D.K.] are also present. These immunoreactivities (especially HNK-1/NCAM) suggest cells of a neural crest lineage. We thus propose that the adult rat thymus contains a small number of neural crest-derived cells that can be expanded in primary culture when they mitose rapidly and express typical neural crest immunoreactivity. Whilst this degree of plasticity might appear surprising, biomolecular manipulations and clinical data support an embryological rationale. Numerous immunocytochemical and functional similarities both in vivo and in vitro suggest an astroglia/thymic epithelia link [unpubl. data]. Firstly, during embryonic development, the neural crest arises from a stem cell population that also produces neurons and glia [114]. Secondly, the thymus is dependent upon neural crest for its development. Evidence for this comes from embryological studies [115–117], clinical conditions such as cri-du-chat and the Di George syndrome [118], and the defects in animals with targetted disruption or point mutations of genes involved in neural crest patterning [119, 120]. The similarities in function (predominantly the cytokine profile) of glia and thymic epithelium have not been fully expanded above and form the basis of a paper now in preparation [unpubl. data]. However, a strong link from our own studies are the small rounded cells of TEC cultures and the antibody to L1 (a 200-kD neural adhesion molecule of the Ig superfamily found on glial cells and nerves). The L1 molecule binds

Transplantation of Thymic Tissues

Neuroimmunomodulation 1999;6:6–22

17

a

b

c

d

e

f Fig. 1. Morphology and immunoreactivity in primary cell cultures and cultured organ fragments. a Confluent cells in culture, showing medulla-like structures in the monolayer (M-LS). !124. Bar = 100 Ìm. b Confluent cells in culture showing several different morphologies. !224. Bar = 50 Ìm. c Confluent cells immunoreactive to pan-cytokeratin. !224. Bar = 50 Ìm. d Primary thymic cells stained with 4ß (anti-thymulin) antibody. !206. Bar = 50 Ìm. e GFAP staining in day 9 culture fragment. !224. Bar = 50 Ìm. f PGP 9.5 staining in day 9 culture fragment. !124. Bar =

100 Ìm.

homeophilically, or to NCAM, and to a site on the HNK-1 molecule. L1 has a regulatory motif in its gene that binds transcription factors encoded by Hox and Pax genes and thereby affects NCAM expression [121]. NGF also upregulates L1 expression [122]. The concept that many epithelial cells are derived from a neural crest origin and possess the ability to function in

18

Neuroimmunomodulation 1999;6:6–22

a similar way to glial cells partially explains the previously inexplicable findings of many, that thymic epithelium appears to have a strange ability to be immunoreactive to many neurotransmitters, peptides and neurohormones [67, 90, 112, 123–131]. Neuroimmune interactions are fundamental to thymic function.

Jones/Botham/Kendall

Future Potential of Implanting Cultured Thymic Cells and Fragments

There is a serious clinical interest in the regeneration of thymic function, not only for patients who are immunocompromised, but also for the elderly who have reduced T-cell functions and immune responses. Together these two groups form a large and costly spectrum of patients who could benefit from enhanced T-cell immunity. Care and attention to the preparation of cells prior to implantation should now result in a high success rate in thymic growth for implants. Although regenerated thymic tissue has been shown to be functional in numerous studies with animals, such work remains to be explored for

humans. Also with the newer approaches to immunomodulation being used in, for example, cancer therapy, where cells can be genetically modified before injection into patients, such manipulations could be applied to cells in culture. Thus the functional capabilities of any growing thymic tissue could also be directed to ensure a better immune response after treatment. The future is exciting.

Acknowledgments The authors are pleased to thank the Volkswagen Stiftung for funding the project, and in addition M.D.K. would like to thank the Welton Foundation for additional support.

References 1 Osoba D, Miller JFAP: Evidence for a humoral thymus factor responsible for the maturation of immunological faculty. Nature 1963;199:653– 654. 2 Levey RH, Trainin N, Law LWJ: Evidence for function of thymic tissue in diffusion chambers implanted in neonatally thymectomized mice. Preliminary report. J Natl Cancer Inst 1963;31: 199. 3 Law LW, Trainin N, Levey RH, Barth WF: Humoral thymic factor in mice: Further evidence. Science 1964;143:1049–1051. 4 Osoba D, Miller JFAP: The lymphoid tissues and immune responses of neonatally thymectomized mice bearing thymus tissue in millipore diffusion chambers. J Exp Med 1964;119: 177–209. 5 Yunis EJ, Martinez C, Good RA: Failure to reconstitute neonatally thymectomized mice by ‘successful’ rat thymus transplantation. Nature 1964;204:664–666. 6 Dalmasso AP, Martinez C, Sjodin K, Good RA: Studies on the role of the thymus in immunology. J Exp Med 1963;118:1089. 7 Law LW: Restoration of thymic function in neonatally thymectomized mice bearing xenogeneic thymic grafts. Nature 1966;210:1118– 1120. 8 Hallenbeck GA, Kubista TP, Shorter RG: Restoration of immunologic competence of neonatally thymectomized mice by isogeneic and xenogeneic thymic grafts. Proc Soc Exp Biol Med 1969;130:1142–1146. 9 Jacobs DM: Effects of concanavalin A on the in vitro responses of mouse spleen cells to Tdependent and T-independent antigens. J Immunol 1975;114:365–370. 10 Lafferty KJ, Cooley MA, Woodnough J, Walker KZ: Thyroid allograft immunogenicity is reduced after a period in organ culture. Science 1975;188:259–261.

Transplantation of Thymic Tissues

11 Hong R, Santosham M, Schulte-Wissermann H, Horowitz S, Hsu SH, Winkelstein JA: Reconstitution of B and T lymphocyte function in severe combined immunodeficiency disease after transplantation with thymic epithelium. Lancet 1976;ii:1270–1272. 12 Schulte-Wisserman HS, Manning D, Hong R: Cultured thymic epithelium and immunological reconstitution. Fed Proc 1977;36:1210. 13 Schulte-Wissermann H, Manning D, Hong R: Transplantation of cultured thymic fragments. I. Morphological and technical considerations. Scand J Immunol 1978;8:387–396. 14 Loor F, Hägg LB: The restoration of the T lymphoid system of nude mice: Lower efficiency of non-lymphoid epithelial thymus grafts. Cell Immunol 1977;26:29. 15 Hong R, Schulte-Wissermann H, Jarrett-Toth E, Horowitz SD, Manning DD: Transplantation of cultured thymic fragments. II. Results in nude mice. J Exp Med 1979;149:398–415 16 Hong R, Klopp R: Transplantation of cultured thymus fragments. III. Induction of allotolerance. Thymus 1982;4:91–106. 17 Manning JK, Hong R: Transplantation of cultured thymic fragments: Results in nude mice. IV. Effect of amount of thymic tissue. Thymus 1983;5:407–417. 18 Lischner HW: DiGeorge syndrome(s). J Pediatr 1972;81:1042. 19 Hong R, Horowitz SD, Borcherding W: Diminished reactivity to alloantigen following transplantation of cultured thymic fragments. Thymus 1982;4:155–161. 20 Markert ML, Kostyu DD, Ward FE, McLaughlin TM, Watson TJ, Buckley RH, Schiff SE, Ungerleider RM, Gaynor JW, Oldham KT, Mahaffey SM, Ballow M, Driscoll DA, Hale LP, Haynes BF: Successful formation of a chimeric human thymus allograft following transplantation of cultured postnatal human thymus. J Immunol 1997:158:998–1005.

21 Manning JK, Hong R: Transplantation of cultured thymic fragments: Results in nude mice. Scand J Immunol 1984;19:403–410. 22 Kruisbeek AM, Sharrow SO, Mathieson BJ, Singer A: The H-2 phenotype of the thymus dictates the self-specificity expressed by thymic but not splenic cytotoxic T lymphocyte precursors in thymus-engrafted nude mice. J Immunol 1981;127:2168–2176. 23 Zinkernagel RM, Althage A, Waterfield E, Kindred B, Welsh RM, Callahan G, Pincetl P: Restriction specificities, alloreactivity, and allotolerance expressed by T cells from nude mice reconstituted with H-2-compatible or -incompatible thymus grafts. J Exp Med 1980; 151:376–399. 24 Lake JP, Andrew ME, Pierce CW, Braciale TJ: Sendai virus-specific, H-2-restricted cytotoxic T lymphocyte responses of nude mice grafted with allogeneic or semi-allogeneic thymus glands. J Exp Med 1980;152:1805–1810. 25 Jenski LJ, Hong R: Transplantation of cultured thymic fragments. VI. H-2 recombinant donors. J Immunol 1985;135:947–953. 26 Streilein JW: Neonatal tolerance: Towards an immunogenetic definition of self. Immunol Rev 1979;46:125–146. 27 Danner SA, Schuurman H-J, Joep MA, Lange JMA, Frits HJ, Gmelig Meyling FHJ, Schellekens PTA, Huber J, Kater L: Implantation of cultured thymic fragments in patients with acquired immunodeficiency syndrome. Arch Intern Med 1986;146:1133–1136. 28 Schuurman H-J, Vaessen LMB, Vos JG, Hertogh A, Geertzeema JGN, Brandt CJWM, Rozing J: Implantation of cultured thymic fragments in congenitally athymic nude rats: Ignorance of thymic epithelial haplotype in generation of alloreactivity. J Immunol 1986;137: 2440–2447.

Neuroimmunomodulation 1999;6:6–22

19

29 Schuurman H-J, Vaessen LMB, Broekhuizen R, Brandt CJWM, Holewijn MC, Vos JG, Rozing J: Implantation of cultured thymic fragments in congenitally athymic (nude) rats. Scand J Immunol 1987;26:129–139. 30 Kendall MD, Schuurman H-J, Fenton J, Broekhuizen R, Kampinga J: Implantation of cultured thymic fragments in congenitally athymic (nude) rats. Cell Tissue Res 1988;254: 283–294. 31 Schuurman H-J, Rozing J, Broekhuizen R, Tielen F, Van De Bergh P: Implantation of cultured thymic fragments in athymic nude rats: Studies on tolerance towards donor haplotype. Thymus 1989;13:123–128. 32 Ready AR, Jenkinson EJ, Kingston R, Owen JJT: Successful transplantation across major histocompatibility barrier of deoxyguanosinetreated embryonic thymus expressing class II antigens. Nature 1984;310:231–233. 33 Martin-Fontecha A, Broekhuizen R, De Heer C, Zapata A, Schuurman H-J: Transplantation of cultured thymic fragments in congenitally athymic and euthymic rats. Scand J Immunol 1992;35:575–587. 34 Bulloch K: Neuroanatomy of lymphoid tissue: A review; in Guillemin R (ed): Neural Modulation of Immunity. New York, Raven Press, 1985, pp 111–141. 35 Bellinger DL, Ackerman KD, Felten SY, Lorton D, Felten DL: Noradrenergic sympathetic innervation of thymus, spleen, and lymph nodes: Aspects of development, aging and plasticity in neural immune interaction; in Hadden JW, Masck K, Nistico G (eds): Interactions among CNS, Neuroendocrine and Immune Systems. Rome, Pythagora, 1989, chap 4, pp 35–66. 36 Bulloch K, Cullen MR, Schwartz RH, Long DL: Development of innervation within syngeneic thymus tissue transplanted under the kidney capsule of the nude mouse: A light and electron microscope study. J Neurosci Res 1987;18:16–27. 37 Martin-Fontecha A, Broekhuizen R, De Heer C, Zapata A, Schuurman H-J: The neuroendocrine component of the rat thymus: Studies on cultured thymic fragments before and after transplantation in congenitally athymic and euthymic rats. Brain Behav Immun 1993;7:1–15. 38 Kollmann TR, Goldstein MM, Goldstein H: The concurrent maturation of mouse and human thymocytes in human foetal thymus implanted in NIH-beige-nude-xid mice is associated with the reconstitution of the murine immune system. J Exp Med 1993;177:821–832. 39 Lee LA, Gritsch HA, Sergio JJ, Arn JS, Glaser RM, Sablinski T, Sachs DH, Sykes M: Specific tolerance across a discordant xenogeneic transplantation barrier. Proc Natl Acad Sci USA 1994;91:10864–10867. 40 Lambrigts D, Franssen C, Martens H, Van Calster P, Meurisse M, Geenen V, CharletRenard C, Dewaele A, Coignoul F, Lamy M, Alexandre GPJ: Development of thymus autografts under the kidney capsule in the pig: A new ‘organ’ for xenotransplantation. Xenotransplantation 1996;3:296–303.

20

41 Small M, Van Ewijk W, Gown AM, Rouse RV: Identification of subpopulations of mouse thymic epithelial cells in culture. Immunology 1989;68:371–377. ˇ olic´ M, Gasˇic´ S, Stojanovic´ N, Popovic´ LJ, 42 C Dujic´ A: Phenotypic and ultrastructural characterisation of an epithelial cell line established from rat thymic cultures. Immunology 1992; 77:201–207. 43 Christensson B, Biberfeld P, Graftström R, Matell G: In vitro culture of thymic epithelial cells in serum-free media. APMIS 1989;97: 926–934. 44 Piltch A, Zhang F, Hayashi J: Culture and characterisation of thymic epithelium from autoimmune NZB and NZB/W mice. Cell Immunol 1990;131:325–337. 45 Ehmann UK, Shiurba RA, Peterson WD: Long-term proliferation of mouse thymic epithelial cells in culture. In Vitro Cell Dev Biol 1986;22:738–747. 46 Eshel I, Savion N, Shoham J: Analysis of thymic stromal cell subpopulations grown in vitro on extracellular matrix in defined medium. I. Growth conditions and morphology of murine thymic epithelial and mesenchymal cells. J Immunol 1990;144:1554–1562. 47 Farr AG, Eisenhardt DJ, Anderson SK: Isolation of murine thymic epithelium and an improved method for its propagation in vitro. Anat Rec 1986;216:85–94. 48 Röpke C, von Deurs B, Petersen OW: Shortterm cultivation of murine thymic epithelial cells in a serum-free medium. In Vitro Cell Dev Biol 1990;26:671–681. 49 Mizuochi T, Kasai M, Kokuho T, Kakiuchi T, Hirokawa K: Medullary but not cortical epithelial cells present soluble antigens to helper Tcells. J Exp Med 1992;175:1601–1605. 50 Cirne-Lima EO, van Ewijk W, Savino W: Cortical and medullary phenotypes within a mouse thymic epithelial cell line. In Vitro Cell Dev Biol 1993;29:443–445. 51 Kurz B, von Gaudecker B, Krisch B, Mentlien R: Rat thymic epithelial cells in vitro and in situ: Characterisation by immunocytochemistry and morphology. Cell Tissue Res 1996;283: 221–229. 52 Wong TW, Klinkert WEF, Bowers WE: Immunological properties of thymus-cell subpopulations – rat thymic dendritic cells are potent accessory cells and stimulators in a mixed leucocyte culture. Immunobiology 1982;160:413– 423. 53 Pelletier M, Tautu C, Landry D, Montplaisir S, Chartrand C, Perreault C: Characterisation of human thymic dendritic cells in culture. Immunology 1986;58:263–270. 54 Lafontaine M, Landry D, Montplaisir S: The human thymic dendritic cell phenotype and its modification in culture. Cell Immunology 1992;142:238–251. 55 Toussaint-Demylle D, Scheiff J-M, Haumont S: Thymic nurse cells in culture: Morphological and antigenic characterisation. Cell Tissue Res 1993;272:343–354.

Neuroimmunomodulation 1999;6:6–22

56 Philip D, Pezzano M, Li Y, Omene C, Boto W, Guyden J: The binding, internalisation and release of thymocytes by thymic nurse cells. Cell Immunol 1993;148:301–315. 57 Brown KM, Spirito S, Basch RS: Thymic stromal cells in culture. 1. Establishment and characterization of a line which is cytotoxic for normal thymocytes and produces hematopoietic growth factor(s). Cell Immunol 1991;134:442– 457. 58 Imaizumi A, Goldschneider I, Yoshida T: Reproducible procedures for establishing mouse stromal cell lines. Cell Immunol 1993;150:81– 89. 59 Loor F: Mouse thymus reticulo-epithelial cells in vitro: Isolation, cultivation and preliminary characterisation. Immunoogy 1979;37:157– 177. 60 Small M, Barr-Nea L, Aronson M: Culture of thymic epithelial cells from mice and age-related studies on the growing cells. Eur J Immunol 1984;14:936. 61 Andersen A, Pedersen H, Bendtzen K, Röpke C: Effects of growth factors on cytokine production in serum-free cultures of human thymic epithelial cells. Scand J Immunol 1993;38: 233–238. 62 Galy A, Jolivet M, Jolivet-Reynaud C, Hadden J: Fibroblast growth factor (FGF) and an FGFlike molecule in pituitary extracts stimulate thymic epithelial cell proliferation. Thymus 1990;15:199–211. ˇ olic´ M, Pejnovic´ N, Kataranovski M, Stoja63 C novic´ N, Terzic´ T, Dujic´ A: Rat thymic epithelial cells constitutively secrete IL-1 and IL-6. Int Immunol 1991;3:1165–1174. 64 Schreiber L, Eshel I, Meilin A, Sharabi Y, Shoham J: Analysis of thymic stromal cell subpopulations grown in vitro on extracellular matrix in defined medium. III. Growth conditions of human thymic epithelial cells and immunoregulatory activities in their culture supernatant. Immunology 1991;74:621–629 . 65 Gilbert SF, Migeon BR: D-Valine as a selective agent for normal human and rodent epithelial cells in culture. Cell 1975;5:11–17. 66 Blalock EJ, Smith EM: A complete regulatory loop between the immune and neuroendocrine systems. Fed Proc 1985 ;44:108–111. 67 Marchetti B, Morale MC, Gallo F, Batticane N, Farinella Z, Cioni M: Neuroendocrine-immunology at the turn of the century – Towards a molecular understanding of basic mechanisms and implication for reproductive physiopathology. Endocrinology 1995;3:845–861. 68 Galy AHM, Hadden EM, Touraine J-L, Hadden JW: Effects of cytokines on human thymic epithelial cells in culture. IL-1 induces thymic epithelial cell proliferation and change in morphology. Cell Immunol 1989;124:13–27. 69 Wolf SS, Cohen A: Expression of cytokines and their receptors by human thymocytes and thymic stromal cells. Immunology 1992;77: 362–328. 70 Cohen-Kaminsky S, Delattre R-M, Devergne O, Rouet P, Gimond D, Berrih-Aknin S, Galanaud P: Synergistic induction of interleukin-6 and gene expression in human thymic cells by LPS and cytokines. Cell Immunol 1991;138:19–93.

Jones/Botham/Kendall

71 Le PT, Lazorick S, Whichard LP, Haynes BF, Singer KH: Regulation of cytokine production in the human thymus: Epidermal growth factor and transforming growth factor-· regulate mRNA levels of interleukin-1· (IL-1·), IL-1ß, and IL-6 in human thymic epithelial cells at a post-transcriptional level. J Exp Med 1991; 174:1147–1157. 72 Le PT, Singer KH: Human thymic epithelial cells: Adhesion molecules and cytokine production. Int J Clin Lab Res 1993;23:56–60. 73 Galy AHM, Dinarello CA, Kupper TS, Kameda A, Hadden EM: Effects of cytokines on human epithelial cells in culture. II. Recombinant IL-1 stimulates thymic epithelial cells to produce IL-6 and GM-CSF. Cell Immunol 1990;129:161–175. 74 Schluns KS, Cook JE, Le PT: TGF-ß differentially modulates epidermal growth factor-mediated increases in leukemia-inhibitory factor, IL-6 and IL-1· in human thymic epithelial cells. J Immunol 1997;158:2704–2712. 75 Galy AHM, Spits H: IL-1, IL-4 and IFN-Á differentially regulate cytokine production and cell surface molecule expression in cultured human thymic epithelial cells. J Immunol 1991; 147:3823–3830. 76 Safieh-Garabedian B, Kendall MD, Khamashta MA, Hughes GRV: Thymulin and its role in immunomodulation. J Autoimmun 1992;5: 547–555. 77 Fabien N, Auger C, Monier J-C: Immunolocalisation of thymosin-·1, thymopoietin and thymulin in mouse thymic epithelial cells at different stages of culture: A light and electron microscope study. Immunology 1988;63:721– 727. 78 Dardenne M, Savino W, Gagnerault MC, Itoh T, Bach J-F: Neuroendocrine control of thymic hormonal production. 1. Prolactin stimulates in vivo and in vitro the production of thymulin by human and murine thymic epithelial cells. Endocrinology 1989;125:3–12. 79 Savino W, Bartoccioni E, Homodelarche F, Gagnerault MC, Itoh T, Dardenne M: Thymic hormone containing cells. 9. Steroids in vitro modulate thymulin secretion by human and murine thymic epithelial cells. J Steroid Biochem 1988;30:479–484. 80 Dardenne M, Savino W, Bach J-F: Modulation of thymic endocrine function by thyroid and steroid hormones. Int J Neurosci 1988;39:325– 334. 81 Timsit J, Savino W, Safieh B, Chanson P, Gagnerault MC, Bach JF, Dardenne M: Growth hormone and insulin-like growth factor-1 stimulate hormonal function and proliferation of thymic epithelial cells. J Clin Endocrinol Metab 1992;71:183–188. 82 Coto JA, Hadden EM, Sauro M, Zorn N, Hadden JW: Interleukin-1 regulates secretion of zinc-thymulin by human epithelial cells and its action on T-lymphocyte proliferation and nuclear protein kinase-c. Proc Natl Acad Sci USA 1992;89:7752–7756.

Transplantation of Thymic Tissues

83 Head GM, Mentlein R, Kranz A, Downing JEG, Kendall MD: Modulation of dye-coupling and proliferation in cultured rat thymic epithelium by factors involved in thymulin secretion. J Anat 1997;191:355–365. 84 Mendoza ME, Martin D, Candelaria PG, Romano MC: Evidence that secretory products of the reticulo-epithelial cells of the rat thymus modulate the secretions of gonadotrophins by rat pituitary cells in culture. J Reprod Immunol 1995;28:203–215. 85 Kendall MD, Stebbings RJ: The endocrine thymus. Endocr J 1994;2:333–339. 86 Screpanti I, Modesti A, Gulino A: Heterogeneity of thymic stromal cells and thymocyte differentiation: A cell culture approach. J Cell Sci 1993;105:601–606. 87 Vizi ES, Orso E, Osipenko ON, Hasko G, Elenkov IJ: Neurochemical, electrophysiological, and immunocytochemical evidence for a noradrenergic link between the sympathetic nervous system and thymocytes. Neuroscience 1995;68:1263–1276. 88 Screpanti I, Meco D, Scarpa S, Morrone S, Frati L, Gulino A, Modesti A: Neuromodulatory loop mediated by nerve growth factor and interleukin-6 in thymic stromal cells. Proc Natl Acad Sci USA 1992;89:3209–3212. 89 Rameshwar P, Gascon P, Ganea D: Immunoregulatory effects of neuropeptides. Stimulation of interleukin-2 production by Substance P. J Neurobiol 1992;37:65–74. 90 Screpanti I, Scarpa S, Meco D, Bellavia D, Stuppia L, Frati L, Modesti A, Gulino A: Epidermal growth factor promotes a neural phenotype in thymic epithelial cells and enhances neuropoietic cytokine expression. J Cell Biol 1995;130:183–192. 91 Benson MT, Buckley G, Jenkinson EJ, Owen JJT: Survival of deoxyguanosine-treated fetal thymus allografts is prevented by priming with dendritic cells. Immunology 1987;60:593–596. 92 Mandel T, Russel PJ: Differentiation of foetal mouse thymus. Ultrastructure of organ cultures and subcapsular grafts. Immunology 1971;21: 659–674. 93 Hong R, Moore AL: Organ culture for thymus transplantation. Transplantation 1996;61: 444–448. 94 Willis JI: The restorative effects of thymic epithelial monolayers on lymphoid cells from neonatally thymectomised rats. Anat Rec 1975; 181:510. 95 Willis-Carr JI, Ochs HD, Wedgwood RJ: Induction of T-lymphocyte differentiation by thymic epithelial cell monolayers. Clin Immunol Immunopathol 1978;10:315–324. 96 Dupuy J-M, Gilmore N, Goldman H, Tsoukas C, Pekovic JP, Chausseau R, Duperval M, Joly M, Pelletier L, Thibaudeau Y: Thymic epithelial cell transplantation in patients with acquired immunodeficiency syndrome. Thymus 1991;17:205–258. 97 Herrod HG, Wheeler WB, Hanissian AS, Ochs HD, Willis-Carr J, Handorf CR: Differentiation of lymphocyte function in vivo following transplantation of thymic epithelial monolayers. Clin Immunol Immunopathol 1981;18: 322–333.

98 Patel DD, Whichard LP, Miralles GD, Sundy JS, Haynes BF: In vitro formation of the human thymic microenvironment: Similarity to SCID thymus. J Allergy Clin Immunol 1997;99:1602. 99 Röpke C, Van Soest P, Platenburg PP, Van Ewijk W: A common stem cell for murine cortical and medullary thymic epithelial cells? Dev Immunol 1995;4:149–156. 100 Lampert IA, Ritter MA: The origin of the diverse epithelial cells of the thymus: Is there a common stem cell? Thymus Update 1988;1: 5–25. 101 Wijngaert FP van de, Kendall MD, Schuurman H-J, Rademakers LHMP, Kater L: Heterogeneity of human thymus epithelial cells at the ultrastructural level. Cell Tissue Res 1984;237:227–237. 102 Mitchell B, Kendall M, Adam E, Schumacher U: Innervation in the thymus in normal and bone marrow reconstituted severe combined immunodeficient (SCID) mice. J Neuroimmunol 1997;75:19–27. 103 Vakharia DD, Mitchison NA: Thymic epithelial cells derived from the cultures of thymic nurse cells. Immunol Lett 1984;7:261–266. 104 Hiramine C, Hojo K, Koseto M, Nakagawa T, Mukasa A: Establishment of a murine thymic epithelial cell line capable of inducing both thymic nurse cell formation and thymocyte apoptosis. Lab Invest 1990;62:41–54. 105 Renata Z, Nerina G, Noa ML, Enzo M, Daniella S, Zvi V, Emmapaola S: Ras-grf, the activator of ras, is expressed preferentially in mature neurons of the central nervous system. Mol Brain Res 1997;48:140–144. 106 Singer KH, Harden EA, Robertson AL, Lobach DF, Haynes BF: In vitro growth and phenotypic characterization of mesodermalderived and epithelial components of normal and abnormal human thymus. Hum Immunol 1985;13:161–176. 107 Nieburgs AC, Picciano PT, Korn JH, McCalister T, Allred C, Cohen S: In vitro growth and maintenance of two morphologiaclly distinct populations of thymic epithelial cells. Cell Immunol 1985;90:439–450. 108 Nieburgs AC, Korn JH, Picciano PT, Cohen S: Thymic epithelium in vitro. IV. Regulation of growth and mediator production by epidermal growth factor. Cell Immunol 1987;108: 396–404. 109 Kranz A, Kendall MD, Gaudecker B von: Studies on rat and human thymus to demonstrate immunoreactivity of calcitonin generelated peptide, tyrosine hydroxylase and neuropeptide Y. J Anat 1997;191:441–450. 110 Ushiki T, Iwanaga T, Masuda T, Takahashi Y, Fujita T: Distribution and ultrastructure of S-100 immunoreactive cells in the human thymus. Cell Tissue Res 1984;235:509–514. 111 Haynes BF, Shimizu K, Eisenbarth GS: Identification of human and rodent thymic epithelium using tetanus toxin and monoclonal antibody A2B5. J Clin Invest 1983;71:9–14.

Neuroimmunomodulation 1999;6:6–22

21

112 De Leeuw FE, Jansen GH, Batanero E, Vanwichen DF, Huber J, Schuurman HJ: The neural and neuro-endocrine component of the human thymus. 1. Nerve-like structures. Brain Behav Immun 1992;6:234–248. 113 Kuramoto H, Hozumi I, Inuzuka T, Sato S: Occurrence of myelin-associated glycoprotein (MAG)-like immunoreactivity in some nervous, endocrine and immune-related cells of the rat. Mol Chem Neuropathol 1997;31:85– 94. 114 Stemple DL, Anderson DJ: Isolation of a stem cell for neurons and glia from the mammalian neural crest. Cell 1992;71:973–985. 115 Le Douarin NM, Jotereau FV: Tracing of cells of the avian thymus through embryonic life in interspecific chimeras. J Exp Med 1975;142:17–40. 116 Le Douarin NM, Jotereau FV: The ontogeny of the thymus. Thymus Update 1981;1:133– 155. 117 Bockman DE, Kirby M: Dependence of thymic development on derivatives of the neural crest. Science 1984;223:498–500. 118 Couly G, Lagrue A, Griscelli C: Le syndrome de Di George, neurocristopathie rhombencéphalique exemplaire. Rev Stomatol Chir Maxillofac 1983;84:103–108.

22

119 Manley NR, Capecchi MR: The role of Hoxa3 in mouse thymus and thyroid development. Development 1995;121:1989–2003. 120 Chisaka O, Capecchi M: Regionally restricted developmental defects resulting from targeted disruption of the mouse homeobox gene Hox1.5. Nature 1991;355:473–479. 121 Edelman GM, Jones FS: Developmental control of N-CAM expression by Hox and Pax gene products. Phil Trans R Soc Lond 1995; 349:305–312. 122 Salton SRJ, Shelanski ML, Greene LA: Biochemical properties of the nerve growth factor-inducible large external (NILE) glycoprotein. J Neurosci 1983;3:2420–2430. 123 Geenen V, Legros J-J, Franchimont P, Baudrihaye M, Defresne M-P, Boniver J: The neuroendocrine thymus: Co-existence of oxytocin and neurophysin in the human thymus. Science 1986;222:508–510. 124 Fuller PJ, Verity K: Somatostatin gene expression in the thymus gland. J Immunol 1989;143:1015–1017. 125 Gomariz RP, Lorenzo MJ, Cacicedo L, Vicente A, Zapata AG: Demonstration of immunoreactive vasoactive-intestinal-peptide (ir-VIP) and somatostatin (Ir-SOM) in rat thymus. Brain Behav Immun 1990;4:151– 161. 126 Piantelli I, Maggiano N, Larocca LM, Ricci R, Ranelletti FO, Lauriola L, Capelli A: Neuropeptide-immunoreactive cells in human thymus. Brain Behav Immun 1990;4:189– 197.

Neuroimmunomodulation 1999;6:6–22

127 Batanero E, De Leeuw FE, Jansen GH, Van Wichen DF, Huber J, Schuurman HJ: The neural and neuro-endocrine component of the human thymus. 2. Hormone immunoreactivity. Brain Behav Immun 1992;6:249– 264. 128 Dardenne M, Savino W: Control of thymus physiology by peptidic hormones and neuropeptides. Immunol Today 1994;15:518–523. 129 Bulloch K, Hausman J, Radojcic T, Short S: Calcitonin gene-related peptide in the developing and aging thymus – An immunocytochemical study. Ann NY Acad Sci 1991;621: 218–228. 130 Møll UM: Functional histology of the neuroendocrine thymus. Microsc Res Tech 1997; 38:300–310. 131 Atoji Y, Yamamoto Y, Komatsu T, Suzuki Y: Localization of neuropeptides in endocrine cells of the chicken thymus. J Vet Med Sci 1997;59:601–603. 132 Jones GV, Botham CA, Clarke AG, Kendall MD: Immunoreactivity of neural crest-derived cells in thymic tissue developing under the rat kidney capsule. Brain Behav Immun, in press.

Jones/Botham/Kendall

Neuroimmunomodulation 1999;6:23–30

Accelerated Maturation of the Thymic Stroma in the Progeny of Adrenalectomized Pregnant Rats Rosa Sacedo´n a Alberto Varas a Eva Jiménez a Juan José Mu´ñoz a Angeles Vicente b Agustı´n G. Zapata a a Department of Cell Biology, Faculty of Biology and b Department of Cell Biology, Faculty of Medicine, Complutense University, Madrid, Spain

Key Words Glucocorticoids W Thymic epithelium W Laminin W Macrophages W Dendritic cells

Abstract The possible role played by glucocorticoids (GCs) in the development of thymic stromal cell components has been studied in the progeny of adrenalectomized pregnant rats (FAdx), an experimental model which ensures the absence of GCs until the establishment of the fetal hypothalamus-pituitary gland-adrenal gland axis. As previously demonstrated for thymocytes, the lack of GCs early in ontogeny results in an accelerated maturation of the thymic stromal elements. Early expression of specific cell markers for thymic epithelial cell subsets and appearance of a well-established cytokeratin-positive epithelial cytoreticulum confirmed the ultrastructural evidence of a faster maturation of the thymic epithelium in FAdx than in FSham. A similar faster and stronger pattern of both class I and class II molecule expression on the epithelial cells occurred in the former fetuses than in control ones. Changes in the pattern of expression of laminin, but not that of fibronectin, throughout thymic maturation also reflected accelerated maturation. Immunohistochemically identified thymic macrophages ap-

ABC

© 1999 S. Karger AG, Basel 1021–7401/99/0062–0023$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: http://BioMedNet.com/karger

peared late in both FSham and FAdx but in higher numbers in these latter indirectly demonstrating their faster development. Finally, the maturation and turnover of thymic dendritic cells showed a remarkable acceleration in the FAdx. In 15- to 16-day-old FAdx thymuses there was a high number of dendritic cells which sharply decreased in the following days suggesting a massive migration to the periphery and/or in situ cell death. In parallel a new wave of dendritic cell progenitors began to differentiate in the FAdx thymuses but not in the FSham ones. The results are discussed from the view of close relationships known to occur between thymocytes and the stromal components, although a direct effect of GCs cannot be discarded.

Introduction

Apart from endogenous mechanisms which control the functioning of the immune system, this is homeostatically modulated by various endocrine factors, including steroids. The thymus gland plays a key role in this process [1–3]. Thus, adrenalectomy provokes a profound thymic hypertrophy while glucocorticoid (GC) administration induces atrophy of the gland [4, 5]. The origin of these mod-

Agustı´n G. Zapata Department of Cell Biology Faculty of Biology, Complutense University E–28040 Madrid (Spain) Tel. +34 9 394 49 79, Fax +34 9 394 49 81, E-Mail [email protected]

ifications are, however, largely unknown. In addition, information on the role played by these hormones in the immunity maturation is very limited and contradictory. Morale et al. [6], in a transgenic mouse unable to respond to GCs, described increased thymic cell content and cell proliferation as well as higher numbers of both peripheral CD4+CD8– and CD4+CD8+ lymphocytes compared to control mice. On the contrary, in a similar experimental model, the hyporesponsiveness of thymocyte to GCs was accompanied by a reduction in thymic size largely due, in turn, to a decrease in the number of DP (CD4+CD8+) cells [7]. Recently, we analyzed the T-cell development in the thymus of FAdx [1]. In this experimental model, which ensures the absence of GCs during the earliest stages of development until the establishment of the fetal hypothalamus-pituitary gland-adrenal gland axis (HPA) around day 18 of embryonic life, we found an accelerated maturation of T cells presumably due to the influence of GCs in determining both the numbers and the proliferative capacity of cell progenitors which colonize the thymus gland early in ontogeny. Because, until our information, only a few studies have correlated GC levels and changes in thymic epithelial cells [8, 9] or extracellular matrix (ECM) components [10] and since a correlation between thymocyte development and thymic nonlymphoid cell maturation has been emphasized in the last years [11–13], we have used this experimental model again to study the ontogeny of thymic nonlymphoid cells (dendritic cells, macrophages, epithelial cells, ECM components) in FAdx. Our results confirm a role for GCs in the maturation of rat thymus gland.

Methods Animals and Treatment Wistar rats inbred in the laboratory facilities were either adrenalectomized or sham adrenalectomized on the first day of pregnancy, considering as day 0 of gestation the first day on which spermatozoa appeared in the vaginal smear. From this moment, the pregnant rats were transferred to individual cages. Bilateral adrenalectomy (Adx) or sham adrenalectomy (Sham) was performed using the dorsal approach under ether anesthesia. The Adx mothers received 0.9% NaCl to drink instead of water until sacrifice. Thymi of the progeny were removed aseptically on days 15–16 of fetal life to day of birth. Corticosterone Levels The blood samples were collected in nonheparinized tubes and after 4 h at room temperature centrifuged at 2,200 rpm for 15 min at 4 ° C. Sera were stored at –70 ° C until assayed. Steroid extraction was performed before testing using methylene chloride. A double antibody commercial RIA kit (Gamma-B-125I-Corticosterone RIA, IDS,

24

Neuroimmunomodulation 1999;6:23–30

UK), which provides a highly sensitive method (0.04 ng/ml), was used for the determination of serum corticosterone levels from both mothers and fetuses. Immunohistochemistry Aseptically removed fetal thymic tissue was immediately frozen and stored at –70 ° C. Cryosections were fixed in acetone at room temperature for 10 min and incubated for 1 h with specific mouse/ rabbit antibodies for cortical (His-37) and medullary/subcapsulary (His-39) rat thymic epithelial cells (Dr. Kampinga, University of Groningen, The Netherlands), cytokeratin (pan) (NCL-PAN-CK, Seikagaku Corp., Tokyo, Japan), rat macrophages (ED-1, Dr. Dijkstra, Free University, Amsterdam, The Netherlands; R-MC42, Dr. Colic, Medical Faculty, Belgrade, Yugoslavia), fibronectin (Fn-15) and laminin (Lam-89) (Sigma Chemicals, St. Louis, Mo., USA), as well as against both MHC class I (OX18) and II (OX6) molecules (Serotec, Oxford, UK). After blocking endogenous peroxidase activity (methanol 1% H202 for 5 min), tissue sections were incubated for 1 h with 1% normal rabbit/goat serum in PBS solution of peroxidaseconjugated rabbit anti-mouse/goat anti-rabbit Ig in PBS (Dako, Co., Denmark/Jackson ImmunoResearch Laboratories, Inc., USA). The peroxidase reaction was developed with 0.05% 3,3)diaminobenzidine (Sigma Chemicals) in PBS with 0.1% H2O2 for 10 min. The acid phosphatase activity was evaluated according to Barka and Anderson [14] using ·-naftil AS-biphosphate-specific substrate. Negative controls were carried out omitting the substrate. The slides were counterstained with methylene blue. Electron Microscopy Thirteen-day-old fetuses aseptically isolated from either control Sham or Adx pregnant rats were fixed by immersion in 2.5% glutaraldehyde, buffered to pH 7.3 with Millonig’s fluid, postfixed in 1% osmium tetroxide in the same buffer, and dehydrated in acetone for embedding in AralditeTM. Semithin sections stained with an alkaline solution of toluidine blue were used to identify and isolate the thymic primordia. Ultrathin sections of the selected areas were obtained with a Reichert OM-U3 ultratome, double stained with uranyl acetate and lead citrate and examined with a Jeol 1010 electron microscope of the Servicio Comu´n de Investigacio´n (Complutense University of Madrid). DC Enrichment Thymic cell suspensions obtained from 17 to 22-day-old fetuses were enriched on DCs as previously described [15]. Briefly, 0.5–5 ! 106 cells were incubated with an anti-CD2 mAb for 1 h in PBS–5% FCS at 4 ° C. After two washes, cells were resuspended in 0.1 ml of PBS–5% FCS containing rabbit anti-mouse IgG-coated magnetic beads (6:1 bead-to-cell ratio; Dynabeads, Dynal, Oslo, Norway). After 30 min in ice with intermittent shaking, 3 ml of PBS–2% FCS was added and the rosettes formed (CD2+ cells) were separated from the nonrosetting cells by placing the tube on a magnetic particle concentrator (Dynal) for 3 min. The enriched CD2– cell suspensions were immunocytochemically analyzed on cytospin preparations by an indirect immunoperoxidase labelling assay (as described above) using the mAb OX6 (Serotec) specific for rat MHC class II antigens. The analysis was performed scoring at least 500–1,000 cells from three different slides of each group of animals.

Sacedo´n/Varas/Jiménez/Mu´ñoz/Vicente/ Zapata

Results

Statistically significant lower values of circulating GCs, as measured by RIA, occurred until fetal day 18 in the serum of FAdx compared to control FSham (table 1). In the following days, the significant differences disappeared, although in all stages studied the levels of GCs were lower in FAdx than in FSham. An earlier maturation of the thymic epithelium as well as of ECM was evidenced both immunohistochemically and ultrastructurally in FAdx compared to FSham. Thus the mAb His-37, which recognizes a specific determinant expressed on thymic cortical epithelial cells in adult rats (but in the medullary region early in thymic ontogeny), identified epithelial cells in the thymus of 16-day-old FAdx but these were not evident until day 19 in FSham (fig. 1a, b). In addition, although cytokeratin-positive epithelial cells appeared on day 16 in the thymus of both FAdx and FSham, a cortical cytoreticulum similar to that occurring in the nonmanipulated adult thymus was already observed in the thymus of 17-day-old FAdx but not in that of FSham. Likewise, MHC class I- and class IIexpressing cells appeared in both groups of rats early in ontogeny (day 16), but, at this stage, whereas in the FSham OX6+ cells constituted a compact reticulum in the incipient thymic medulla, in FAdx the pattern was reticular and similar to that found in control animals on day 19. At that stage, the class II expression already showed an adult pattern in FAdx but not in FSham (fig. 1c,d). The ultrastructural study confirmed the early maturation of thymic epithelium in the FAdx. On day 13 of gestation, the thymic primordium of control FSham consisted of a homogenous meshwork of primitive epithelial cells, as previously reported [16]. On the contrary, the 13day-old FAdx thymus was largely invaded by lymphoid progenitors and contained numerous, more differentiated epithelial cells which exhibited prominent nucleoli, few condensed chromatin, long profiles of rough endoplasmic reticulum and electron-lucent mitochondria (fig. 1e,f). Well-developed desmosomes and bundles of cytoplasmic microfilaments also occurred in these cells (fig. 1e,f). Two important components of the thymic ECM, fibronectin and laminin, were also immunohistochemically studied. In control FSham, as previously reported for hamster [8], mouse [10] and rats [17], fibronectin and laminin constitute a discontinuous network during the early ontogenetical stages. Afterwards, the expression of both fibronectin and laminin was restricted to the perivascular, subcapsular and subtrabecular areas associated

Glucocorticoids and Thymus Development

Table 1. Effects of maternal adrenalectomy on circulating corticoste-

rone levels of the progeny: data represented are the average of 3–4 experiments B SD Day

Corticosterone, ng/ml FSham

16 17 18 19 20 21 22

70B10 97.7B16.2 229B92 260B51 220B37 194.5B7.8 250B20

FAdx 11B3.8*** 24.7B7.67** 104B27.5* 192B35.7 178B23.4 166B59.5 220B50

* p ^ 0.05; ** p ^ 0.01; *** p ^ 0.001.

to components of the basement membrane. In the FAdx, the pattern of expression of fibronectin was similar to that exhibited in the FSham, but the evolution of the laminin was strikingly different (fig. 2a,b). In the thymus of 16day-old FAdx the expression of laminin was similar to that exhibited in control fetuses of the same age. One day later, however, the expression in FAdx thymus resembled that found in the thymus of FSham at the end of fetal life, with unstained areas in the central region of thymic primordium (fig. 2a,b). In the following days in which according to our own results on T-cell maturation in FAdx [1] the second wave of cell progenitors started to differentiate in the FAdx thymus and, therefore, mature thymocytes were absent, these unstained regions disappeared. Accordingly, the establishment of an adult pattern of laminin expression underwent a slight delay in FAdx compared to FSham. The appearance and maturation of thymic macrophages in both FAdx and FSham was evaluated by detection of acid phosphatase reactivity and the appearance of ED-1+ or R-MC42+ cells. The first cells showing acid phosphatase activity occurred in control, Sham thymus on day 17 of gestation but 1 day earlier in the FAdx. As previously demonstrated [18], expression of specific markers for rat macrophages occurs late in thymic ontogeny. In both FAdx and FSham thymus, the time of appearance of either ED-1+ or R-MC42+ cells was the same (around days 19–20 of gestation), but the numbers of positive cells were higher in the former, suggesting a faster maturation of the thymic macrophages in the FAdx (fig. 2c,d). Three types of morphological and phenotypically different thymic dendritic cells have been identified in rat

Neuroimmunomodulation 1999;6:23–30

25

Fig. 1. His-37-positive cells (arrowheads) occur in the thymus of 16-day-old FAdx (b) but not in that of FSham rats (a). !95. A positive MHC class II cytoreticulum is well established in the thymus of 19-day-old FAdx rats (d) but is incipiently developed in FSham of the same age (c). !95. Thymic primordium of 13-day-old FSham (e) and FAdx rats (f). In the FSham the organ consists of homogeneous meshwork devoid of lymphoid cells and formed by primitive epithelial cells (Ep) joined together by incipient, poorly developed desmosomes (insert e). Epithelial cells (Ep) of Adx thymus are well-developed elements with mature desmosomes (insert f), which organize a network housing

numerous lymphoid progenitors (L). Nu = Nucleus; arrowheads = basement membrane; Fb = fibroblast. !6,600.

26

Neuroimmunomodulation 1999;6:23–30

Sacedo´n/Varas/Jiménez/Mu´ñoz/Vicente/ Zapata

Fig. 2. Laminin expression in the thymus either of 17-day-old FAdx (b) or FSham rats (a). Note the existence of central areas de-

void of staining (stars) in the Adx thymus. !84. Higher numbers of ED-1-positive macrophages in the thymus of 20-day-old FAdx thymus (d) compared to those found in control, FSham of the same age (c). !42.

thymus ontogeny [15]. Presumptive DC precursors (type I) were MHC class IIlo monocyte-like cells; irregular, mature MHC class IIhi DCs were named type II, while a third subset (type III) showed higher MHC class II expression than mature type II cells, as well as very long thin cell processes. A remarkably high number of DCs (around 9% of total thymic cell population) occurred in the thymus of 15- to 16-day-old FAdx (table 2). Twenty-five percent of these cells corresponded to mature type II DCs whereas type III ones appeared 1 day later (data not shown). On the contrary, in FSham thymuses the first mature DCs appeared on day 17, representing less than 9% of the thymic DC population and the type III subpopulation was not found until day 18 of gestation. In the following days, the absolute numbers of total DCs gradually grew in FSham thymuses, while in FAdx they sharply decreased (table 2). At that time 70% of these cells were type I DCs, suggesting that a new wave of DC precursors had arrived into the organ and was differentiating. The absolute numbers of thymic DCs duplicated again in the 21- to 22-day-old FAdx (table 2).

Table 2. Both percentage and absolute numbers (in parentheses) of dendritic cells in the thymus of either FSham or FAdx from day 15 to 22 of gestation

Day

Total DCs Sham

Adx

15

0.031B0.015 (0.0002B0.0001)

9.47B3.37 (0.041B0.015)

16

1.32B0.09 (0.045B0.0006)

6.56B0.3 (0.13B0.04)

17

0.685B0.23 (0.11B0.03)

0.74B0.14 (0.059B0.01)

18

0.59B0.034 (0.395B0.05)

0.27B0.05 (0.081B0.03)

19

0.22B0.04 (0.74B0.08)

0.337B0.069 (0.69B0.075)

20

0.31B0.05 (1.8B0.2)

0.471B0.13 (2.26B0.09)

21

0.29B0.07 (2.18B0.29)

0.432B0.0178 (2.27B0.4)

22

0.324B0.044 (2.59B0.45)

0.456B0.033 (4.79B0.7)

Despite results which emphasize GCs as being essential for an adequate fetal development [19], there is very little information on the possible role played by these hor-

mones in the maturation of the immune system. Maternal GCs are the only source of circulating hormone in the fetuses until establishment of the HPA axis [1, 20] because the recently reported ACTH-dependent intrathymic production of GCs [21] has not been shown early

Glucocorticoids and Thymus Development

Neuroimmunomodulation 1999;6:23–30

Discussion

27

in ontogeny before the appearance of circulating ACTH. Accordingly, analysis of the development of the immune system in the progeny of adrenalectomized pregnant rats is a good model to study the possible influence of these hormones in the process. We had previously used this experimental model to demonstrate that the lack of GCs early in ontogeny profoundly affects the thymic cellularity, thymocyte differentiation and T-cell repertoire [1]. Our current results extend this influence of GCs to the maturation of thymic stromal cell components. In agreement, a preliminary study on the thymus gland of fetal, adult and aged transgenic mice unable to respond to GCs [6] demonstrates profound alterations in the pattern of expression of MHC class II molecules, cytokeratin, laminin and thymic macrophages [Sacedo´n et al., in preparation]. The earlier appearance of both keratin-positive epithelial cytoreticulum and specific cell markers for adult thymic epithelial cell subsets as well as, according to our ultrastructural analysis, differentiated epithelial cells in the thymic primordium of 13-day-old FAdx but not FSham, suggest an accelerated maturation of the thymic epithelial cells in the former fetuses, as reported above for T-cell development [1]. In support, a stronger expression of both class I and class II MHC molecules in the epithelial cells and an earlier acquisition of their adult, reticular pattern occurs in the thymus of FAdx compared to that of control ones. Although a direct effect of the absence of GCs on the thymic epithelial cells cannot be discarded, since GCs stimulate the expression of certain cytokeratins in these cells [8, 22], the parallel differentiation of thymic epithelium and thymocytes seems to suggest an indirect influence of GCs on epithelial cell development through thymocytes. In this regard, the re-expression on thymic cortical epithelial cells of specific cell markers characteristic of the embryonic periods after hydrocortisone treatment has been related to the arrival of primitive cell precursors to recover the thymic lymphoid cell population rather than to a direct effect of GCs on the epithelial cells [9]. On the other hand, mutual bidirectional influences between the thymic stromal cell components, principally the epithelial cells, and the thymocytes have been claimed [11–13]. Some recent data have demonstrated, however, that the maturation (but not the appearance of thymic epithelium) is exclusively due to the expression of the whn gene, independently of the occurrence or not of thymocytes [23]. Evidence demonstrating down-regulated expression of MHC molecules after GC administration [24, 25] sug-

28

Neuroimmunomodulation 1999;6:23–30

gests a possible direct effect of the lack of GCs on the accelerated expression of class I and class II molecules in the FAdx thymus. However, a close relationship between lymphoid colonization and expression of MHC molecules in the thymic primordium of both mice [26] and rats [16] has been evidenced, although some authors claimed a role for IGF-I rather than for the lymphoid colonization in the process [27]. This again supports the relevance of the mutual influences between thymocytes and thymic epithelial cells for explaining the accelerated maturation of both thymic cell components. Likewise, the pattern of expression of laminin, but not that of fibronectin, reflects changes in the appearance/disappearance of mature thymocytes during the ontogeny of FAdx thymus. Reported results on the effects of GCs on the ECM are controversial and no information is available, to our knowledge, on the condition in the absence of steroids. The administration of GCs stimulates the synthesis of both fibronectin and laminin in the murine and hamster thymus [8, 10] but GCs negatively regulate the production of specific mRNA for ß chain laminin, tenascin [28] and other components of the extracellular matrix of bone marrow cultures [29]. It seems possible to conclude that the changes observed in the pattern of maturation of the thymic stromal cell components of FAdx are an indirect effect of the accelerated differentation of thymocytes which occurs in this group of embryos. Although the appearance of macrophages occurs late during ontogeny in both FAdx and FSham thymuses, the higher numbers of positive cells observed in the former and the earlier finding of acid phosphatase activity in the FAdx thymus also suggests a faster maturation of the thymic macrophages, as previously described for both thymocytes and epithelial cells in these embryos. Despite evidence on the capability of GCs to inhibit monocytemacrophage lineage differentiation [30], there are no data, to our knowledge, on the effects in the opposite condition (lack of GCs). Presumably, the accelerated maturation of thymic macrophages is associated, like those of both thymocytes and DCs (see below), with an earlier colonization of the thymic primordium by macrophage progenitors in FAdx than in FSham. In this respect, our ultrastructural study of the thymic primordium of 13-dayold FAdx demonstrated the existence of numerous primitive progenitor cells. On the contrary, at the same developmental stage, the thymic primordium of FSham only consists of a homogenous mass of primitive epithelial cells devoid of any precursor cells. The maturation and turnover of thymic DCs, like other components of the thymic stroma, exhibit an important

Sacedo´n/Varas/Jiménez/Mu´ñoz/Vicente/ Zapata

acceleration in FAdx. Because different evidences support the occurrence of a common cell precursor for lymphocytes, thymic DCs and NK cells [31–33], the lack of GCs in the FAdx could be modifying the pattern of differentiation of these progenitors resulting in an accelerated maturation of both thymocytes and thymic DCs. Alternatively, the lack of GCs could induce increased production of growth factor(s) involved in thymic DC development. Remarkably, IL-7 treatment of rat FTOCs provokes increased production of DCs similar to that observed in the FAdx thymuses [34]. On the other hand, the significance of the sharp decrease in absolute numbers of thymic DCs between 16 and 17 days, another feature of the FAdx thymuses, which occurs in parallel with both the migration of thymocytes to spleen and the increased thymocyte apoptosis [1], remains unclear. Likewise, DC activation after LPS administration results 2 days later in their disappearance from the spleen [35]. Periodical renewal of thymic DCs occurs in adult rat thymus [36] and the absolute numbers

of DCs gradually decrease after intrathymic injection of cell progenitors in lethally irradiated mice [37]. Alternatively, since in FAdx the decrease in thymic DCs occurs after a slight increase of GCs, it could be associated to the known function of GCs to mobilize distinct cell population, including DCs [38]. Taking these result together, we can conclude that, as previously demonstrated for thymocytes, the lack of GCs in early ontogeny of the progeny of adrenalectomized rats results in an accelerated maturation of the nonlymphoid cell components of thymic stroma, including epithelium, macrophages, DCs and elements of the ECM.

Acknowledgments This work was supported in part by PR181/96-6824, from the UCM and CAYCIT grants PB91-0374 and PB94-0332 from the Spanish Ministry of Education and Science. R.S., E.J. and J.J.M. are recipients of a fellowship from the Spanish Ministry of Education and Science.

References 1 Sacedo´n R: Ontogenia de las poblaciones T en la progenie de ratas Wistar adrenalectomizadas; Master thesis, Complutense University, 1995. 2 Hadden JW: Thymic endocrinology. Int J Immunopharmacol 1992;14:345–352. 3 Homo-Delarche F, Durant S: Hormones, neurotransmitters and neuropeptides as modulators of lymphocyte functions; in Rola-Pleszczynsky M (ed): The Handbook of Immunopharmacology. Immunopharmacology of Lymphocytes. London, Academic Press, 1994, pp 169–240. 4 Screpanti I, Morone S, Meco D, Santoni A, Gulino A, Paolini R, Crisanti A, Mathieson BJ, Franti L: Steroid sensitivity of thymocyte subpopulations during intrathymic differentiation. Effects of 17ß-estradiol and dexamethasone on subsets expressing T cell antigen receptor or IL2 receptor. J Immunol 1989;142:3378–3383. 5 Tsuchida M, Konishi M, Jojima K, Naito K, Fujikura Y, Fukumoto T: Analysis of cell surface antigens on glucocorticoid-treated rat thymocytes with monoclonal antibodies. Immunol Lett 1994;39:209–217. 6 Morale MC, Nunzio B, Gallo F, Barden N, Marchetti B: Disruption of hypothalamic-pituitary-adrenocortical system in transgenic mice expressing type II glucocorticoid receptor antisense ribonucleic acid permanently impairs T cell function: Effects on T cell trafficking and T cell responsiveness during postnatal development. Endocrinology 1995;136:3949–3960.

Glucocorticoids and Thymus Development

7 King LB, Vacchio MS, Dixon K, Hunziker R, Margulies DH, Ashwell JD: A targeted glucocorticoid receptor antisense transgene increases thymocyte apoptosis and alters thymocytes development. Immunity 1995;3:647– 656. 8 De Souza LRM, Savino W: Modulation of cytokeratin expression in the hamster thymus: Evidence for a plasticity of the thymic epithelium. Dev Immunol 1993;3:137–146. 9 Stojanovic S, Dragojevic-Simic V, Colic M: Corticosteroid treatment and X-ray irradiation in adult rats induce the re-expression of fetal markers in cortical epithelial cells. Thymus 1995;24:1–7. 10 Lannes-Vieira J, Dardenne M, Savino W: Extracellular matrix components of the mouse thymus microenvironment: Ontogenetic studies and modulation by glucocorticoid hormones. J Histochem Cytochem 1991;39:1539– 1546. 11 Ritter MA, Boyd RL: Development in the thymus: It takes two to tango. Immunol Today 1993;14:462–469. 12 Van Ewijk W, Shores EW, Singer A: Crosstalk in the mouse thymus. Immunol Today 1994; 15:214–217. 13 Holländer GA, Wang B, Nichogiannopoulou A, Platenburg PP, van Ewijk W, Burakoff SJ, Gutierrez-Ramos JC, Terhorst C: Developmental control points in induction of thymic cortex regulated by a subpopulation of prothymocytes. Nature 1995;373:350–353.

14 Barka T, Anderson PJ: Histochemical methods for acid phosphatase using hexazonium pararosanilin as coupler. J Histochem Cytochem 1962;10:741–746. 15 Vicente A, Varas A, Alonso L, Go´mez del Moral M, Zapata AG: Ontogeny of rat thymic dendritic cells. Immunology 1994;82:75–81. 16 Vicente A, Varas A, Sacedo´n R, Zapata AG: Histogenesis of the epithelial component of rat thymus. An ultrastructural and immunohistological analysis. Anat Rec 1996;244:506–519. 17 Vicente A: Ana´lisis de la diferenciacio´n de las células T durante la ontogenia del timo de rata; thesis, Complutense University, 1994. 18 Vicente A, Varas A, Moreno J, Sacedo´n R, Jiménez E, Zapata AG: Ontogeny of rat thymic macrophages. Phenotypic characterization and possible relationships between different cell subsets. Immunology 1995;85:99–105. 19 Muglia L, Jacobson L, Dikkes P, Majzoub JA: Corticotropin-releasing hormone deficiency reveals major fetal but not adult glucocorticoid need. Nature 1995;373:427–432. 20 Milkovic S, Milkovic K, Paunovic J: The initiation of fetal adrenocorticotrophic activity in the rat. Endocrinology 1973;92:380–384. 21 Vacchio MS, Papadopoulos V, Ashwell JD: Steroid production in the thymus: Implications for thymocyte selection. J Exp Med 1994;179: 1835–1846.

Neuroimmunomodulation 1999;6:23–30

29

22 Savino W, Cirne-Lima EO, Soares JFT, Leitede Moares MC, Ono IPC, Dardenne M: Hydrocortisone increases the numbers of KL1 cells, a discrete thymic epithelial cell subset characterized by high molecular weight cytokeratin expression. Endocrinology 1988;123: 2557–2564. 23 Nehls M, Kyewski B, Messerle M, Waldschütz R, Schüddekopf K, Smith AJH, Boehm T: Two genetically separable steps in the differentiation of thymic epithelium. Science 1996;272: 886–889. 24 Fercht D, Schoenberg DR, Germani RN, Tou JYL, Vogel SN: Induction of macrophage Ia antigen expression by rIFNÁ and down-regulation by IFN ·/ß and dexamethasone are mediated by changes in steady-state levels of Ia mRNA. J Immunol 1987;139:244–249. 25 Van Rees EP, Van Der Ende MB, Sminia Y: Influence of prenatal administration of dexamethasone on the developing immune system; in Imhof BA, Berrih-Aknin S, Ezine S (eds): Lymphatic Tissues and in vivo Immune Responses. New York, Dekker, 1991, pp 239– 244. 26 Jenkinson EJ, Owen JJT, Aspinall R: Lymphocyte differentiation and major histocompatibility complex antigen expression in the embryonic thymus. Nature 1980;284:177–179.

30

27 Shinohora T, Honjo T: Studies in vitro on the mechanism of the epithelial/mesenchymal interaction in the early fetal thymus. Eur J Immunol 1997;27:522–529. 28 Ekblom M, Fassler R, Tomasini-Johansson B, Nilsson K: Downregulation of tenascin expression by glucocorticoids in the bone marrow stromal cells and in fibroblasts. J Cell Biol 1993;123:1037–1045. 29 Hemesalth TJ, Stefansson K: Expression of tenascin in thymus and thymic nonlymphoid cells. J Immunol 1994;152:422–428. 30 Baybutt HN, Holsboer F: Inhibition of macrophage differentiation and function by cortisol. Endocrinology 1990;127:476–480. 31 Ardavin C: Thymic dendritic cells. Immunol Today 1997;18:350–361. 32 Wu L, Li C, Shortman K: Thymic dendritic cell precursors: Relationships to the T lymphocyte lineage and phenotype of the dendritic cell progeny. J Exp Med 1996;184:903–911. 33 Wu L, Nichogiannopoulou A, Shortman K, Georgopoulos K: Cell-autonomous defects in dendritic cell populations of Ikaros mutant mice point to developmental relationship with the lymphoid lineage. Immunity 1997;7:483– 492.

Neuroimmunomodulation 1999;6:23–30

34 Varas A, Vicente A, Sacedo´n R, Zapata AG: Interleukin 7 influences the development of thymic dendritic cells. Blood 1998;92:93–100. 35 De Smedt T, Pajak B, Muraille E, Lespagnard L, Heinen E, De Baetselier P, Urbain J, Leo O, Moser M: Regulation of dendritic cell numbers and maturation by polysaccharide in vivo. J Exp Med 1996;184:1413–1424. 36 Kampinga J, Nieuwenhuis P, Roser B, Aspinal R: Differences in turnover between thymic medullary dendritic cells and a subset of cortical macrophages. J Immunol 1990;138:1659– 1663. 37 Wu L, Vremec D, Ardavı´n C, Winkel K, Suss G, Georgiou H, Maraskovsky E, Cook W, Shortman K: Mouse thymus dendritic cells: Kinetics of development and changes in surface markers during maturation. Eur J Immunol 1995;25:418–425. 38 Moser M, De Smedt T, Sornasse T, Tielemans F, Chentoufi AA, Muraille E, Van Mechelen M, Urbain J, Oberdan L: Glucocorticoids down-regulate dendritic cell function in vitro and in vivo. Eur J Immunol 1995;25:2818– 2824.

Sacedo´n/Varas/Jiménez/Mu´ñoz/Vicente/ Zapata

Neuroimmunomodulation 1999;6:31–38

Rapid Progesterone Actions on Thymulin-Secreting Epithelial Cells Cultured from Rat Thymus Gail M. Head a James E.G. Downing a Cosima Brucker b Rolf Mentlein c Marion D. Kendall d a Department

of Biology, Imperial College, London, UK; b Frauenklinik, Klinikum Innenstadt der LMU, München, and Institut, Kiel, Germany; d Thymus Laboratory, The Babraham Institute, Cambridge, UK

c Anatomisches

Key Words Ion channels W Calcium W Acrosome reaction W Secretion W Thymic epithelial cells

Abstract Many soluble factors of neural, endocrine, paracrine and autocrine origin are present in the thymus and modulate its function. Long-term effects of sex steroids have been documented for thymocytes and cells of the thymic microenvironment. In this report we examine rapid actions of progesterone upon aspects of epithelial cell physiology. Progesterone (0.1–10 ÌM ) was applied to cultured thymulin-secreting thymic epithelial cells (TSTEC) and changes in transmembrane potential, transmembrane current, intracellular calcium levels and thymulin secretion were assessed. Rapid changes in electrophysiology and intracellular calcium provide evidence for a membrane-bound progesterone receptor in these cells, in addition to classical cytoplasmic receptors. Application of progesterone to TS-TEC caused electrophysiological changes in 56% of cells (n = 40), activating an inward current (–24 B 9 pA at 1 ÌM, n = 7, p ! 0.02) and dose-dependent depolarization (7.1 B 1.8 mV at 1 ÌM, n = 19, p ! 0.01). Intracellular calcium levels, monitored by the ratiometric fluorescent calcium indicator fura-2, increased within seconds of progesterone (1 ÌM ) application. Progesterone (1 ÌM ) increased thymulin levels in

ABC

© 1999 S. Karger AG, Basel 1021–7401/99/0062–0031$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: http://BioMedNet.com/karger

supernatant, as measured by ELISA, above the levels in the preapplication period (142 B 16% of the preapplication period, n = 3, p ! 0.02). This effect was reduced in the presence of cobalt chloride which blocks voltage-dependent calcium channels. In addition, TS-TEC in culture were immunoreactive to antibody AG7. This antibody was raised to a membrane-bound antigen involved in calcium influx subsequent to progesterone binding in sperm. Thus we suggest that progesterone acts upon many aspects of TS-TEC physiology through both cytoplasmic and membrane-bound receptors.

Introduction

Many authors have shown that sex steroid levels alter structure and function of the thymic microenvironment [1–8]. Because of the complexity of steroid interactions and catabolism, the effects of progesterone are not clear. Although all of the sex steroids cause involution of the thymus similar to that seen in pregnancy [reviewed in 9], thymic involution at this time is correlated with serum progesterone levels [10]. However, in general it appears that for most parameters so far studied, oestrogens have a greater effect than progesterone [11–14], appearing to arrest thymocyte development at the earliest stage [15].

Prof. M.D. Kendall The Thymus Laboratory, The Babraham Institute Cambridge, CB2 4AT (UK) Tel. +44 1223 832312, Fax +44 1223 837912, E-Mail [email protected]

Cytoplasmic receptors for progesterone and oestrogen have been demonstrated in the thymus. Levels of cytoplasmic receptors for these hormones are much lower in lymphoid cells of the thymus than in the stromal cells [16– 18]. Epithelial steroid receptors were found in the medullary region, especially at the corticomedullary junction with a few positive cells in the subcapsular area [18]. Some of these cells were thymulin immunopositive [19]. Synthesis of cytoplasmic progesterone receptors is increased by oestrogens [17], decreased by progesterone [16, 20], and upregulated in pregnancy [21]. Classically, steroid receptors are considered to be localized in the cytoplasm and to act in a genomic manner. As early as 1942, Selye [4] noted immediate anaesthetic effects of progesterone that are incompatible with a genomic action. However, such effects were slow to be recognized [22]. A clear example of this type of nongenomic action has been supplied by the work on the acrosome reaction whereby rapid direct action of progesterone at the cell membrane initiates calcium influx and capacitates sperm [23]. Membrane-bound progesterone receptors have been recently identified in various cells [24– 26]. Within thymulin-secreting thymic epithelial cells (TSTEC), the genomic action of sex hormones via the nuclear receptor complexes are indicated by studies that co-localized sex hormone receptors and thymulin in TEC [19] and reports showing long-term alterations in production and expression of thymulin by sex hormone treatment [27]. Steroids also altered prostoglandin production by TEC [28]. Nongenomic mechanisms have been suggested for sex steroid-induced changes in TEC proliferation subsequent to alterations in protein kinase C levels [29]. In vitro studies of steroid action on thymic epithelial cells demonstrated increased thymulin secretion from cultures of rat and human TEC after long-term incubations with progesterone, testosterone and 17ß-estradiol [27]. The steroids also increased the percentage of thymulinsecreting cells in these cultures suggesting an effect on proliferation. Long-term co-incubation with the progesterone antagonist RU-486 abrogated the progesterone effect. Steroid-induced increases in TS-TEC proliferation were confirmed by Head et al. [30] who also showed decreased intercellular coupling in TS-TEC after incubations with steroids. This paper investigates rapid actions of progesterone upon a range of physiological parameters of TS-TEC to determine if nongenomic actions of progesterone are present and to clarify that progesterone has direct actions upon these thymic epithelial cells.

32

Neuroimmunomodulation 1999;6:31–38

Materials and Methods Cell Culture Thymic epithelial cell cultures were produced, characterized and passaged according to published methods [31]. Three culture media (two with additions from Sigma) were used in the study. A basal medium was used for cell equilibration comprising DMEM/F12 (Gibco) alone. A defined, serum-free medium was used for control incubations, comprising DMEM/F12 plus the following additions: 2 ÌM l-glutamine, 100 U/ml penicillin/streptomycin and 25 g/ml transferrin). The standard culture medium was used for routine culture of the cells, comprising the defined medium plus 10% horse serum, 5 g/ml insulin, 10 ng/ml cholera toxin and epidermal growth factor (100 ng/ml). All experiments were carried between passages 20–30 when the culture characteristics are stable [30]. Estimation of Thymulin in Supernatants Cultured cells were equilibrated for 1 h in basal medium. They were then incubated for 2 h each in the defined medium, or defined medium plus either 1 ÌM progesterone, 2.5 mM cobalt chloride or 1 ÌM progesterone with 2.5 mM cobalt chloride. After this incubation the cells were returned to the defined medium. The experiment was replicated 3 times. The supernatant from each incubation was reserved and stored at –70 ° C before estimation of thymulin content by ELISA as previously described [30]. Briefly, each supernatant sample was mixed with an optimal dilution of primary antibody RMK4 and allowed to equilibrate for 18 h at 4 ° C. Free thymulin in the sample bound excess antithymulin antibody in the liquid phase. In the second step, 120 ml of each sample/RMK4 mixture was transferred to a high binding plate, previously coated with synthetic thymulin, to which any remaining free antibody bound. Thus the amount of free antibody that reacted in this step correlated inversely with the thymulin concentration in the initial sample. The antibody bound to the thymulin, and hence the plate, was reacted with peroxidase-conjugated secondary antibodies, that catalysed colour development of 3,3),5,5)-tetramethylbenzidine in the presence of H2O2. Colour development was stopped by the addition of H2SO4 to produce a static endpoint, read as optical density at 450 nm on a spectrophotometer and compared against a calibration curve (from the same plate) of known concentrations of synthetic thymulin. Electrophysiology TEC cultures were grown in the standard medium in 35-mm dishes for 3 days after passage. The cells were rounded up for recording by a 5-min incubation in enzyme-free dissociation fluid (Sigma) after which they were equilibrated for 15 min in a modified Ringer’s solution containing 144 mM NaCl, 5.4 mM KCl, 1 mM MgCl, 2.5 mM CaCl2, 5.6 mM D-glucose, 5 mM HEPES and 0.001% w/v phenol red. Intercellular recording solutions contained 140 mM KCl, 5 mM NaCl, 0.5 mM CaCl2, 3 mM Mg-ATP, 5 mM EGTA and 10 mM HEPES. During control and recovery periods the modified Ringer’s solution was perfused across the surface of the cell through a gravity feed, Ytube assembly. Differing concentrations of progesterone (0.1, 1 and 10 ÌM ) were made up in modified Ringer’s solution and administered through the perfusion apparatus which was quickly interchangeable between control and test solutions. Recordings were made in whole cell current clamp and voltage clamp at –100 mV to investigate the currents and membrane potential changes induced by progesterone. Glass micropipette electrodes of resistance between 0.2 and 10 M were applied to the cell at 400! magnification on an Olympus

Head/Downing/Brucker/Mentlein/Kendall

Fluorescent Calcium Imaging TEC cultures were grown in the standard medium upon sterile glass coverslips previously coated with poly-d-lysine (1–10% v/v) for 3 days after passage. Cells were loaded with the ratiometric calcium indicator fura-2 either by incubation for 15–60 min, at 37 ° C with 1–10 ÌM of the AM ester of fura-2 after being washed in serum-free medium, or by a 30-min incubation with liposomes (composed of phosphatidylserine, phosphatidylcholine, and cholesterol in a 2:20:10 ratio) containing fura-free acid or fura-2 dextran conjugate, both with the zinc chelating compound N,N,N),N)-tetrakis(2-pyridylmethyl)ethylenediamine (TPEN) which was also applied to the cultures for between 30 min to 2 h prior to liposome application. After fura-2 incorporation the cultures were washed with modified Ringer’s solution (140 mM NaCl, 5.4 mM KCl, 1 mM MgCl, 10 mM CaCl2, 5.6 mM D-glucose and 5 mM HEPES) in which they were then equilibrated for 10 min. The coverslip was placed in a recording chamber that was sealed with silicon grease (RS components). The holding vessel was then placed into a heated stage on the microscope and maintained at 37 ° C. The cells were observed using a Nikon inverting microscope with a 40! objective. Intracellular calcium changes were monitored with a dynamic video imaging system (Applied Imaging MagiCal system Newcastle, UK [32]). Each cell was calibrated at the end of the recording by application of 10 ÌM calcium ionophore (A23187, Sigma) in modified Ringer’s and in calcium-free Ringer’s (140 mM NaCl, 5.4 mM KCl, 1 mM MgCl, 10 mM EGTA, 5.6 mM D-glucose and 5 mM HEPES). Cells which did not produce a swift, reversable change in fluorescent ratio to the calibration protocol were excluded from the experiment. To quantify changes in intracellular calcium in response to progesterone (1 ÌM ), a control wash with modified Ringer’s solution only was applied to the recording bath, then the same solution with added progesterone was applied in excess, after an application of 30 s to 1 min the compound was washed away by flushing the recording chamber with the control Ringer.

Percentage change in supernatant thymulin

(IMT-2) inverting microscope under phase-contrast illumination. Seals were formed by the application of suction through the microelectrode. Seals with resistance ! 1 Gø were discarded. Whole cell current and voltage recordings were obtained using a RK400 patchclamp amplifier (Biologic). DC offsets resulting from junction potentials were corrected. Errors due to series resistance and capacitance were minimized using the compensation circuits of the amplifier. Raw data was displayed on a Schlumberger Enertec 5072 storage oscilloscope. The signal was by, and command pulses routed via, a 1401 interface (Cambridge Electronic Design). Signals were filtered at 1 kHz through a 5-pole Tchebicheff filter, digitized at between 0.1 and 10 kHz using a 1401 interface (Cambridge Electronic Design) and recorded on the hard drive of an IBM compatible 386 computer. Voltage and current commands from the amplifier were remotely controlled from the computer, and analysis performed, using Vclamp software (Cambridge Electronic Design, version 5).

180

]

160 140 120 ]

100 80 60 40 20 0

Control

Prog 1 mM

Prog + Cobalt

Cobalt 2.5 mM

Fig. 1. Mean B SEM change in thymulin level compared to a prein-

cubation period of similar duration. Supernatant thymulin levels were determined by ELISA. The known thymulin secretagogue progesterone (Prog) significantly increased the levels of supernatant thymulin. Co-application of cobalt with progesterone prevented the progesterone-induced increase in thymulin levels, a significant reduction compared to the effect of progesterone. Cobalt also reduced basal secretion (82 B 32%, n = 3) but this change was not significant. * p ! 0.05.

up in phosphate-buffered saline containing 1% Triton X (Sigma). A solution containing 0.1% casein and 0.1% bovine serum albumin (both from Sigma) was applied to the cells to block nonspecific binding. Immunocytochemistry was performed conventionally using optimally diluted monoclonal antibodies: 4ß (raised to synthetic thymulin by Mary Ritter, London); AG7 (raised to sperm acrosome antigen-1 by Cosima Brucker, Munich) and pan-cytokeratin (Sigma). For negative controls the primary antibody was omitted. The reaction was visualized by enhanced diaminobenzidine and viewed under bright-field illumination. Statistics Data is expressed as mean B SEM, and significance analysed using Student’s t tests.

Results Thymulin Determination

Immunocytochemistry TEC cultures were grown in the standard medium for 3 days after passage upon sterile glass coverslips, previously coated with poly-dlysine (1–10% v/v). The cells were fixed in acetone for 2 min and air dried, after which they could be stored, wrapped in cling film at – 30 ° C for up to 1 week, being defrosted at room temperature, unwrapped and air dried before processing. All reagents were made

Supernatant thymulin levels after 2 h were increased by 1 ÌM progesterone (fig. 1) to 142 B 16% of the preapplication period (p ! 0.02, n = 3). This increase was not seen in the presence of 2.5 mM cobalt chloride (82 B 15% of controls, n = 3). Cobalt chloride alone caused a slight, but not significant, reduction of thymulin levels (82 B 32% of controls, n = 3).

Progesterone and Thymic Epithelia

Neuroimmunomodulation 1999;6:31–38

33

Electrophysiology Progesterone induced membrane depolarizations (¢Em) and inward currents (¢I) in 56% of TEC (n = 40). Illustrations of these responses are shown in figure 2A and B respectively. Responses were desensitizing, a second sequential application having a lower effect (36 B 12% of 1st application, n = 7, p ! 0.05, pooled Em and I) as shown in figure 2B. The mean inward current induced by 1 ÌM progesterone was –24 B 9 pA (n = 7, p ! 0.02). Depolarizations of up to 22 mV resulted from the application of progesterone (mean ¢Em = 7.1 B 1.8 mV at 1 ÌM, n = 19, p ! 0.01). These sometimes took the cell to positive membrane voltages (fig. 2A). Depolarizations were dose-dependent (100 nM, ¢Em = –1.1 B 1.5 mV; 1 ÌM, ¢Em = 4.6 B 1.3 mV, p ! 0.05; 10 ÌM, ¢Em = 11.9 B 3.7 mV, p ! 0.05; all n = 7, paired recordings; fig. 2C). The reversal potential of the current was at +41 B 14 mV (n = 7). In addition, depolarizing responses became hyperpolarizing after substitution of extracellular sodium with choline in 5 cells (¢Em = +4 B 1 mV to ¢Em = –5 B 2 mV, n = 5, p ! 0.02), implicating sodium as a major permeant ion. Fluorescent Calcium Imaging In cells with adequate ionomycin calibration intracellular calcium was increased following application of 1 ÌM progesterone (n = 10/30, p ! 0.001). The temporal and spatial response characteristics of a progesterone-evoked calcium signal are shown in figure 3. The peak intracellu-

Fig. 3. This figure shows the change in nominal intracellular calcium

Fig. 2. Electrophysiological effects of progesterone. A Application of

progesterone caused a depolarization of the TEC membrane potential. In this case 1 mM progesterone caused the cell to depolarize from –55 to +15 mV. Most depolarizations were of a smaller magnitude although other depolarizations to positive membrane voltages were seen. B Application of progesterone induced an inward current. Here, repeated applications of 1 mM progesterone are shown illustrating both the inward current and the desensitization seen. C Depolarization of the cell in response to progesterone was dose-dependent. The mean B SEM change in membrane potential of 5–9 cells in response to progesterone at 0.1, 1.0 and 10 mM is shown. ** p ! 0.01.

34

Neuroimmunomodulation 1999;6:31–38

(y-axis) with time (x-axis) in the cell shown as an inset. The pseudocolour scale within the y-axis shows the relationship between colour and calcium concentration used to produce the pseudocolour image of the calcium concentration within the cell (inset) during the response to 1 ÌM progesterone (frame position is indicated by the white star). The progesterone application, to wash out the control ringer, took place during the period marked by the white bar. Progesterone caused a rapid, reversible increase in intracellular calcium of this TEC. Progesterone remained in the chamber until the end of this recording, after which a response to calcium ionophore was obtained (not shown). Fig. 4. Immunoreactivity of TEC, 3 days after passage, for antibodies to AG7 (A), thymulin (B), and cytokeratin (C). The scale is given by the scale bar in A which represents 10 Ìm. A Immunoreactivity to AG7 was observed in all cells and was distributed fairly evenly. Small cells and those near the edge of the monolayer were in general more intensely stained. B Thymulin staining of TEC. All cells showed a positive reaction. C Cytokeratin was seen in all cells, forming a network through the cell body. Cells at the edges of monolayers were more intensely stained than those in the centre. Single cells were invariably darkly stained.

Head/Downing/Brucker/Mentlein/Kendall

3

4

Progesterone and Thymic Epithelia

Neuroimmunomodulation 1999;6:31–38

35

lar calcium levels were achieved 4–6 min after application of 1 ÌM progesterone. Changes in intracellular calcium were not evenly distributed over the cell. The highest levels were seen in some areas of the cell border and in the nucleus. No changes were seen with vehicle control (29 cells). Immunocytochemistry Cultured cells were immunoreactive for cytokeratin, thymulin and AG7 (fig. 4). No unstained cells were observed with these antibodies. No negative controls were stained.

Discussion

TEC are potential targets of neuroendocrine ligands present in the circulation [33–35] and of input from the thymic innervation [36–39]. Ionic mechanisms of TEC have previously been implicated in thymulin secretion [40, 41]. The present study and our previous work show that TS-TEC possess ligand-induced adrenergic and steroidal mechanisms [42, 43]. Here we demonstrate direct actions of progesterone upon various physiological responses of TS-TEC. Classically, progesterone action involves cytoplasmic steroid receptors that induce changes in DNA transcription. These receptors are present on TEC and steroid-induced changes in thymic physiology and thymulin secretion have mainly been ascribed to genomic action [18, 27]. Some of the progesterone-induced responses previously demonstrated by us in these cells [30], namely a reduction in intercellular coupling and an increase in proliferation, may be partly due to such genomic responses. However, the electrophysiological changes and calcium flux initiated by progesterone in this study occur too rapidly to be genomic effects. The inhibition of progesterone-evoked secretion by cobalt also suggests that short-term (2 h) changes in thymulin release are a direct effect of calciumdependent exocytosis rather than of increased thymulin expression. Both electrophysiological recordings and calcium imaging suggest that nongenomic responses may not occur in all cells. This could be a reflection of the morphological heterogeneity of the cultures [30] and their variable expression of ion channels [44]. Electrophysiological responses of TEC were observed in response to the application of progesterone. These responses comprised the activation of transmembrane currents and changes in transmembrane voltages that

36

Neuroimmunomodulation 1999;6:31–38

were desensitizing and dose-dependent. These changes were consistent with the activation of a poorly selective cation current. This is a novel action of progesterone in somatic cells. It is not consistent with established actions, whereby progesterone and/or its metabolites increase and prolong the Á-aminobutyric acid-A chloride current [45] or inhibit cation currents stimulated by cholinergic agonists at nicotinic acetylcholine receptors [46]. Poorly selective cation currents activated by progesterone have also been reported in sperm which result in an influx of sodium and calcium activating the acrosome reaction [47]. A membrane-bound receptor initiating a nongenomic signal transduction pathway has been implicated in this effect. The membrane-bound sperm acrosome antigen-1 (SAA-1), that is internalized after progesterone binding, forms part of this pathway [48]. SAA-1 is recognized and inhibited by the antibody AG7 [1]. SAA-1 could be a progesterone receptor, a separate ion channel or an accessory protein [48]. The positive immunoreactivity to AG7 suggests that a similar mechanism may be present in TS-TEC. Furthermore, the characteristics of the current induced in sperm by progesterone are comparable to those seen in TEC, and both result in rapid calcium influx. The nongenomic progesterone receptor in TS-TEC has not yet been characterized and no details of its structure or molecular weight are known. Functionally we have demonstrated that in TS-TEC, activation of this progesterone receptor leads to ion channel activation. Although progesterone membrane-binding proteins have been found in a number of cell types [26], none of them have been shown to activate ion channels. Three hypothetical mechanisms, whereby the progesterone receptor of TEC may open an ion channel, are presented in figure 5. Firstly, the receptor may be metabotropic, coupling to a separate ion channel via second messengers (fig. 5A). Secondly, the receptor may be a membrane-bound protein that aggregates as a subunit or component of an ion channel (fig. 5B). Thirdly, the receptor may incorporate an ion channel, allowing it to be directly ionotropic (fig. 5C). Any of these models may allow calcium influx through the channel, although in such nonselective cation channel the proportion of calcium is likely to form only a small proportion of the current. The large calcium influx seen suggests the presence of voltage-gated calcium channels that open in response to the progesterone-induced depolarization. This hypothesis is supported by the inhibition of secretion seen with cobalt. A functional consequence of an influx of calcium is the secretion of thymulin [40, 41]. Thus calcium influx could

Head/Downing/Brucker/Mentlein/Kendall

Fig. 5. Models of nongenomic progesterone signal transduction in TS-TEC. A Progester-

one binds to a novel membrane-bound receptor which activates an ion channel via second messengers. B Progesterone binds to a novel membrane-bound receptor which activates an ion channel by associating with the ion channel molecule. C Progesterone binds to a novel ionotropic receptor. D Although calcium influx may occur through the initial channel (A) it may be augmented through cell depolarization and the subsequent opening of voltage-dependent calcium channels. Additional signals initiated by progesterone are mediated by cytoplasmic receptors and genomic action.

explain the rapid progesterone-mediated secretion of thymulin through non-genomic pathways. These are in addition to the documented long-term effects of progesterone upon thymulin synthesis and cell proliferation [27, 30].

Conclusions

The thymus, as one of the primary lymphoid organs, plays a crucial role in immunological defence and its output can be altered by neural and endocrine stimuli. These stimuli can act upon immature thymocytes but also affect the stromal cells of the gland which create the thymic microenvironment and provide developmental cues for the thymocytes. The stromal cells of the thymus do not appear to be a static, structural framework, carrying out predetermined steps in thymocyte differentiation. The results of this and our previous studies show that TS-TEC form a plastic network that is highly responsive to a wide range of inflammatory, neuroendocrine and neural signals. The range of ligands to which these cells respond suggest that ligands have coordinate actions upon these cells. The large number of inputs may suggest that TS-TEC are involved in signal recognition and processing for the thymic-immune axis. This paper confirms progesterone to be one of the factors capable of stimulating TS-TEC to modulate aspects of their physiology. The ligand-induced responses of TSTEC are diverse. Progesterone acting through a combina-

Progesterone and Thymic Epithelia

tion of genomic and nongenomic actions influences proliferation, secretion, the degree of communication between epithelial cells and hence modulates interactions with thymocytes.

Acknowledgments G.M.H. was supported by the Medical Research Council, and J.E.G.D. by the Royal Society, with additional support provided by the University of London Central Research Fund and Smith Kline (1982) Foundation. R.M. and M.D.K. are funded by the Volkswagen Stiftung, and M.D.K. would like to thank the Welton Foundation for additional financial assistance.

Neuroimmunomodulation 1999;6:31–38

37

References 1 Brucker C, Kassner G, Loser C, Hinrichsen M, Lipford GB: Progesterone-induced acrosome reaction – Potential role for sperm acrosome antigen-1 in fertilization. Hum Reprod 1994;9: 1897–1902. 2 Persike EC: Involution of the thymus during pregnancy in young mice. Proc Soc Exp Biol Med 1940;45:315–317. 3 Grégoire C: Sur le méchanisme de l’atrophie thymique déclenchée par les hormones sexuelles. Arch Intern Pharmacodyn 1945;70:45– 77. 4 Selye H: The general adaption syndrome and the disease of adaption. J Clin Endocrinol 1942;6:117–230. 5 Dougherty TF: Effect of hormones on lymphatic tissue. Physiol Rev 1952;32:379–401. 6 Stimson WH, Hunter IC: Oestrogen-induced immunoregulation mediated through the thymus. J Clin Lab Invest 1980;4:27–33. 7 Grossman CJ: Regulation of the immune system by sex steroids. Endocr Revs 1984;5:435– 455. 8 Greenstein BD, Fitzpatrick FTA, Adcock IM, Kendall MD, Wheeler MJ: Reappearance of the thymus in old male rats after orchidectomy: Inhibition of regeneration by testosterone. J Endocrinol 1986;110:417–422. 9 Kendall MD, Clarke AG: The female thymus and reproduction in mammals. Oxf Rev Reprod Biol 1994;16:165–213. 10 Chambers SP, Clarke AG: Measurement of thymus weight, lumbar node weight and progesterone levels in syngeneically pregnant, allogeneically pregnant and pseudopregnant mice. J Reprod Fertil 1979;55:309–315. 11 Bimes C, de Graeve P, Guilhem A, Amiel S: La cytologie thymique sous l’action des hormones genitales chez le cobaye. C R Séanc Soc Biol Fil 1975;169:233–238. 12 Hellig HR, Gerneke WH: A histological study of the effect of cortisol and some six steroids on the immune response to sheep erythrocytes by the mouse. Onderstepoort J Vet Res 1975;42: 53–62. 13 Stimson WH, Crilly PJ: Effects of steroids on the secretion of immunoregulatory factors by thymic epithelial cell cultures. Immunology 1981;44:401–407. 14 Fitzpatrick FT, Greenstein BD: Effects of various steroids on the thymus, spleen, ventral prostate and seminal vesicles in old orchidectomized rats. J Endocrinol 1987;113:51–55. 15 Rijhsinghani AG, Thompson K, Bhatia SK, Waldschmitt TJ: Estrogen blocks early T-cell development in the thymus. Am J Reprod Immunol 1996;36:269–277. 16 Pearce PT, Khalid BAK, Funder JW: Progesterone receptors in rat thymus. Endocrinology 1983;113:1287–1291. 17 Fujii-Hanamoto H, Seiki K, Sakabe K, Ogawa H: Progestin receptor in the thymus of ovariectomized immature rats. J Endocrinol 1985; 107:223–229. 18 Sakabe K, Seiki K, Fujiihanamoto H: Histochemical localization of progestin receptor cells in the rat thymus. Thymus 1986;8:97–107.

38

19 Kawashima I, Sakabe K, Seiki K, Fujiihanamoto H, Akatsuka A, Tsukamoto H: Localization of sex steroid-receptor cells, with special reference to thymulin (FTS)- producing cells in female rat thymus. Thymus 1991;18:79–93. 20 Pasqualini JR, Gulino A, Sumida C, Screpanti I: Anti-oestrogens in fetal and newborn target tissues. J Steroid Biochem 1984;20:121–128. 21 Pearce P, Funder JW: Cytosol and nuclear levels of thymic progesterone receptors in pregnant, pseudopregnant and steroid-treated rats. J Steroid Biochem 1986; 25:65–69. 22 Wehling M: Specific, nongenomic actions of steroid hormones. Annu Rev Physiol 1997;59: 365–393. 23 Meizel S, Turner KO: Progesterone acts at the plasma-membrane of human sperm. Mol Cell Endocrinol 1991;77:R1–R5. 24 Meyer C, Schmid R, Scriba PC, Wehling M: Purification and partial sequencing of highaffinity progesterone-binding site(s) from porcine liver membranes. Eur J Biochem 1996; 239:726–731. 25 Falkenstein E, Meyer C, Eisen C, Scriba PC, Wehling M: Full-length cDNA sequence of a progesterone membrane-binding protein from porcine vascular smooth muscle cells. Biochem Biophys Res Commun 1996;229:86–89. 26 Eisen C, Meyer C, Wehling M: Characterization of progesterone membrane binding sites from porcine liver probed with a novel azidoprogesterone radioligand. Cell Mol Biol 1997; 43:165–173. 27 Savino W, Bartoccioni E, Homo-Delarche F, Gagnerault M, Itoh T, Dardenne M: Thymic hormone containing cells. IX. Steroids in vitro modulate thymulin secretion by human and murine thymic epithelial cells. J Steroid Biochem 1988;30:1–6. 28 Homo F, Russo-Marie F, Papiernik M: Prostaglandin secretion by human thymic epithelium: In vitro effects of steroids. Prostaglandins 1981;22:377–385. 29 Sakabe K, Kawashima I, Urano R, Seiki K, Itoh T: Effects of sex steroids on the proliferation of thymic epithelial cells in a culture model – A role of protein kinase C. Immunol Cell Biol 1994;72:193–199. 30 Head GM, Mentlein R, Kranz A, Downing JEG, Kendall MD: Modulation of dye-coupling and proliferation in cultured rat thymic epithelium by factors involved in thymulin secretion. J Anat 1997;191:355–365. 31 Kurz B, von Gaudecker B, Kranz A, Krisch B, Mentlein R: Rat thymic epithelial cells in vitro and in situ – Characterization by immunocytochemistry and morphology. Cell Tissue Res 1996;283:221–229. 32 Mason WT, Hoyland J, Neylon CB, Kato M, Ackerman S, Bunting R, Tregear RT, Zorec R: Dynamic, real-time imaging of fluorescent probes of biological activity in living cells. J Med Lab Sci 1991;5:41–52. 33 Dardenne M, Savino W: Neuroendocrine control of the thymic epithelium: Modulation of thymic endocrine function, cytokine expression and cell proliferation by hormones and peptides. PNEI 1990;3:18–25.

Neuroimmunomodulation 1999;6:31–38

34 Millington G, Buckingham JC: Thymic peptides and neuroendocrine-immune communication. J Endocrinol 1992;133:163–168. 35 Kendall MD, Stebbings RJ: The endocrine thymus. Endocr J 1994;2:333–339. 36 Kendall MD, Al-Shawaf A: Innervation of the rat thymus gland. Brain Behav Immun 1991;5: 9–28. 37 Kendall MD, Atkinson BA, Munoz F, De La Riva C, Clarke A, von Gaudecker B: The noradrenergic innervation of the thymus during pregnancy and in the post-partum period. J Anat 1994;185:617–625. 38 Kurz B, Feindt J, von Gaudecker B, Kranz A, Loppnow H, Mentlein R: ß-Adrenoceptor-mediated effects in rat cultured thymic epithelial cells. Br J Pharmacol 1997;120:1401–1408. 39 Head GM, von Patay B, Mentlein R, Downing JEG, Kendall MD: Neuropeptides exert direct effects on rat thymic epithelial cells in culture. Dev Immunol 1998;6:95–104. 40 Buckingham JC, Safieh B, Singh S, Arduino LA, Cover PO, Kendall MD: Interactions between the hypothalamo-pituitary-adrenal axis and the thymus in the rat: A role for corticotrophin in the control of thymulin release. J Neuroendocrinol 1992;4:295–301. 41 Head GM, Kendall MD, Downing JEG: Intrinsic ionic mechanisms contribute to the control of thymulin release from thymic epithelia of rat in vitro. J Physiol 1998, in press. 42 Head GM, Kendall MD, Downing JEG: Adrenergic modulation of thymulin release and electrophysiological changes in cultured epithelium from rat thymus (abstract). 3rd Annu Meet of the Brain-Immune Network Group (BING), London, September 1997. 43 Head GM, Kendall MD, Kranz A, Brucker C, Downing JEG: Modulation of rat thymic epithelial cell physiologies in vitro by progesterone may involve non-genomic mediation through sperm acrosome antigen-1 (abstract). 3rd Annu Meet of the Brain-Immune Network Group (BING), London, September 1997. 44 Head GM, Downing JEG: Anionic leak conductance and macroscopic outward currents of cultured secretory epithelia from rat thymus. J Physiol 1996;495P:108P. 45 Lambert JJ, Belelli D, Hill-Venning C, Peters JA: Neurosteroids and GABAA receptor function. Trends Pharmacol Sci 1995;16:295–303. 46 Bullock AE, Clark AL, Grady SR, Robinson SF, Slobe BS, Marks MJ, Collins AC: Neurosteroids modulate nicotinic receptor function in mouse striatal and thalamic synaptosomes. J Neurochem 1997;68:2412–2423. 47 Garcia MA, Meizel S: Importance of sodium ion to the progesterone-initiated acrosome reaction in human sperm. Mol Reprod Dev 1996;45:513–520. 48 Brucker C, Sandow BA, Blackmore PF, Lipford GB, Hodgen GD: Monoclonal antibodyAG7 inhibits fertilization post sperm-zona binding. Mol Reprod Dev 1992;33:451–462.

Head/Downing/Brucker/Mentlein/Kendall

Neuroimmunomodulation 1999;6:39–44

Cytokine-Mediated or Direct Effects of Thymulin on the Nervous System as Assessed by Pain-Related Behavior Bared Safieh-Garabedian a Salim A. Kanaan a Suhayl J. Jabbur b Nayef E. Saadé b, c a Department

of Biology, Faculty of Arts and Sciences; Departments of b Physiology and c Human Morphology, Faculty of Medicine, American University of Beirut, Lebanon

Key Words Thymulin W Hyperalgesia W Cytokines W Nerve growth factor W Prostaglandin-E2 W Neuroimmunology

Abstract Thymulin is a thymic hormone with known immunomodulatory activities. Recent evidence has indicated a signaling role for this peptide in the interaction between the immune, endocrine and the nervous system. In this report, we review recent experimental findings on the analgesic actions of thymulin (high doses) in rats with endotoxin-induced localized inflammation and the hyperalgesic actions (low doses) of this peptide in intact animals. These actions involve both proinflammatory cytokines and PGE2. The possibility of a dual role played by thymulin as a hormone that might also involve a direct effect on the nervous system is discussed.

ABC

© 1999 S. Karger AG, Basel 1021–7401/99/0062–0039$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: http://BioMedNet.com/karger

Introduction

Thymulin, a nanopeptide [1], which requires zinc for its biological action [2], is secreted primarily by thymic subcapsular and medullary epithelial cells [3]. Recently however, immunohistochemical techniques have shown the presence of this peptide in astrocytic glial cells in the brain and skin basal keratinocytes [4]. In the early days of thymic hormone research, thymulin together with several other thymic peptides were shown to be involved mainly in immunomodulation [5]. As illustration, thymulin induces, in vitro, the expression of differentiation markers on bone marrow-derived thymocyte precursor cells and on both T and B lymphocytes [6, 7]. Also in the peripheral circulation, thymulin appears to interact preferentially with the suppressor subset of T cells as well as enhance natural killer cell activity [8, 9]. More recent evidence points to an important role for thymulin as a signaling molecule for interaction between the immune, endocrine and the nervous system. Several studies have described a bidirectional interaction between the thymic epithelium and the hypothalamus-pituitary axis [reviewed in 10]. As example, thymulin has been shown to be released into the blood with a circadian rhythm coinciding with the activity of the hypothalamus-pituitary axis [11, and unpubl.

Dr. Bared Safieh-Garabedian Department of Biology, Faculty of Arts and Sciences American University of Beirut, PO Box 11-0236 Beirut (Lebanon) Tel. +961 3 791313/ext 3913, Fax +961 1 351706, E-Mail [email protected]

data]. Moreover, ACTH at physiological levels injected into rats elevates plasma thymulin, whereas ACTH, administered in vitro, stimulates thymulin release from cultured thymic fragments [12]. Other hormones including prolactin, growth and thyroid hormones also exert regulatory effects on thymulin release [10, 13, 14]. In this report, data from recent experiments are presented in an attempt to answer the following two questions: (1) Is thymulin involved in the immune-endocrine feedback loop to the central nervous system (CNS)? (2) What are the effects of high and low doses of thymulin injections on pain-related behavior in intact animals and animals subjected to local inflammation?

Effects of High Doses of Thymulin

It is now increasingly evident that during inflammation, interactions between the immune and the nervous systems involve shared ligands and receptors [15]. Previous work has indicated that thymulin in high doses inhibits the production of the cytokine interleukin-1ß (IL-1ß) by peripheral blood mononuclear cells [16]. Various cytokines have been shown to have an important role in the sensory hypersensitivity associated with inflammation, either directly, by acting on receptors found on neurons, or indirectly, by stimulating the release of agents that act on neurons. Such a role has already been described for IL-1ß [17–19]. In a recent set of experiments, carried out in our laboratories, the possible effects of thymulin on pain behavior in normal animals and in animals with localized inflammation were investigated. The classical tests for mechanical (paw pressure (PP) test) and thermal (hot plate (HP) and tail flick (TF) tests) hyperalgesia were utilized. Different groups of rats and mice were subjected to pain tests for 3 consecutive days before and 1 day after intraperitoneal (i.p.) thymulin injections (0.5, 1 and 2 Ìg in 100 Ìl saline). Baseline values of the various pain tests in thymulin-injected animals were comparable to those of saline-injected rats. In a second group of experiments, the animals also received i.p. injections of thymulin (0.5, 1 and 2 Ìg in 100 Ìl saline) prior to intraplantar (i.pl.) endotoxin (ET) injections (1.25 Ìg). At this dose level, ET has been shown to induce local inflammation and hyperalgesia which peaks at 9 h in rats and 24 h in mice and subsequently recovers by 24 and 48 h respectively [20]. Thymulin injections reduced, in a dose-dependent manner, the ET-induced mechanical and thermal hyperalgesia (fig. 1A–C) [21]. It was thus concluded that thymulin in supraphysiological doses can reverse inflam-

40

Neuroimmunomodulation 1999;6:39–44

matory hyperalgesia without altering pain thresholds in intact animals. Further experiments were carried out to investigate whether this analgesic effect of thymulin correlates with changes in the levels of the proinflammatory cytokines IL-1ß and nerve growth factor (NGF), specially in view of the recent demonstrations of a major role for the neurotrophin NGF in mediating inflammatory hyperalgesia [22–24]. During ET-induced hyperalgesia there was a significant increase in the levels of IL-1ß and NGF and pretreatment with steroidal and nonsteroidal antiinflammatory drugs reversed the ET-induced hyperalgesia and down-regulated the levels of IL-1ß and NGF [25]. Pretreatment of rats with thymulin (1 Ìg in 100 Ìl saline, i.p.) prior to ET injection, down-regulated IL-1ß and NGF levels (fig. 1D, E) [26] in a manner comparable to that observed with the anti-inflammatory drugs [25]. These findings strongly suggest the attenuation of hyperalgesia by high doses of thymulin involve the down-regulation of the proinflammatory cytokines.

Effects of Low Doses of Thymulin

Thymulin in low concentrations can up-regulate certain cytokines [16] and enhance immune responses [27]. Based on these findings, we investigated the effect of low doses of thymulin injections on pain-related behavior and cytokine production in intact animals. Various doses of thymulin were injected either i.p. (5, 20, 50 and 150 ng) or i.pl. (0.5, 1, 5 and 10 ng) and their effects on pain-related behavior and cytokine levels were assessed. Intraperitoneal injection of low doses of thymulin resulted in both mechanical (PP test) and thermal (HP and TF tests) hyperalgesia [28]. The effect was apparent at doses ranging between 20 and 150 ng (fig. 2). Thymulin also up-regulated the levels of IL-1ß, NGF and PGE2, as measured in the liver [28, and unpubl. data]. Similar results were obtained with i.pl. thymulin injection, which resulted in a significant reduction in nociceptive thresholds, as assessed by the different pain tests (fig. 3A), localized to the injected hind paw of the rats [29]. A concentration of thymulin as low as 0.5 ng of thymulin was effective in eliciting hyperalgesia in the injected paws. There was also a significant elevation in the levels of IL-1ß and NGF, in the injected paw as compared with the noninjected paw. Thymulin-induced hyperalgesia was reversed by pretreatment with lys-D-pro-val (fig. 3B). This (·-MSH-related tripeptide has been shown to antagonize both IL-1ß- and PGE2-induced hyperalgesia [30]. On the other hand, lysD-pro-thr, a tripeptide which antagonizes IL-1ß-induced

Safieh-Garabedian/Kanaan/Jabbur/Saadé

Fig. 1. Effects of pretreatment with high dose of thymulin (1 Ìg, i.p.) on ET-induced (1.25 Ìg, i.pl.) mechanical (PP) (A), thermal (HP and TF) (B and C) hyperalgesia at 9 h and on ET-induced elevation in IL-1ß (D) and NGF (E) levels at 4 h.

hyperalgesia only [17], reduced partially thymulin-induced hyperalgesia. These results indicate that PGE2 has an important role in mediating thymulin-induced hyperalgesia. This is further supported by studies showing that thymulin in low concentrations induces PGE2 synthesis in cultured mononuclear cells [31] and thymocytes [32]. A role for IL-1ß and NGF is further substantiated by the

partial reversal of thymulin-induced hyperalgesia following the injection of either interleukin-1 receptor antagonist (IL-1ra) or a polyclonal anti-NGF antiserum, administered 30 min before i.pl. thymulin (fig. 3B) [29]. Therefore, thymulin injections, either i.p. or i.pl. in intact rats, produced hyperalgesia and up-regulation in cytokine levels, without any visible signs of inflammation.

Thymulin and Pain-Related Behavior

Neuroimmunomodulation 1999;6:39–44

41

Fig. 2. Effect of low-dose thymulin (50 ng, i.p.) on mechanical (PP)

Fig. 3. Effect of low-dose thymulin (5 ng, i.pl.) on mechanical (PP)

and thermal (HP and TF) nociceptive thresholds in intact rats as compared to saline-treated animals at 6 h (A) and on changes in cytokine levels in the liver as compared to saline-treated animals 4 h (B).

and thermal (HP and TF) nociceptive thresholds in intact rats as compared to saline-treated animals, measured at 6 h (A). In B, the effect of different antagonists and antiserum on thymulin-induced mechanical (PP) and thermal (HP and TF) hyperalgesia, measured at 6 h, is given.

42

Neuroimmunomodulation 1999;6:39–44

Safieh-Garabedian/Kanaan/Jabbur/Saadé

Possible Mechanisms of Thymulin Actions and Concluding Remarks

Based on the correlation of circadian variations of ACTH with thymulin levels it is possible to assume that high doses of this peptide can act in a similar manner to that of the hypothalamus-pituitary-adrenal axis in stress conditions by down-regulating the immune systems, as manifested by its inhibitory effect on cytokine production. In a recent study of collagen-induced arthritis in rats, Aono et al. [33] showed that thymulin at high concentrations resulted in a significant decrease in the symptoms of this autoimmune model, as well as a significant reduction in the levels of anti-type II collagen antibodies. It is interesting to note here that in a model of parasitic infection, in mice, induced by inoculation with Leishmania major, the hyperalgesia obtained was reduced by pretreatment with high doses of thymulin, but without any effect on the elevated levels of IL-1ß and NGF [Kanaan et al., unpubl. data]. Therefore, thymulin may either act independently or through these cytokines on the nervous system. Furthermore, thymulin effects, at high doses, were only manifested during inflammatory reactions, but did not affect pain thresholds in intact animals.

Even though low doses of thymulin resulted in significant increases in the level of several cytokines as well as hyperalgesia in intact animals, treatment with anti-NGF antiserum or IL-1ra did not reverse the thymulin-induced hyperalgesia. This indicates a possible direct action of thymulin on either the central or the peripheral nervous system. Preliminary data from our laboratory has shown that thymulin-induced hyperalgesia (i.p.) is mediated via the vagus nerve, especially through the capsaicin-sensitive primary afferent fibers [34]. In conclusion, data from our experiments, with high and low doses of thymulin, suggest a dual effect of this peptide. First, thymulin can act as a hormone with immunoregulatory function, especially during chronic stress and inflammatory conditions leading to a down-regulation of cytokines. Second, thymulin in low doses can act as a signaling molecule with a role in the feedback mechanism affecting CNS functions, as evidenced by the hyperalgesia and the elevated NGF and PGE2 levels.

Acknowledgments The authors would like to thank the technical assistance of John J. Haddad and Raffy H. Jalakhian. Funding for research was obtained in part from the University Research Board.

References 1 Bach JF, Dardenne M, Pleau JM, Rosa J: Biochemical characterization of a serum thymic factor. Nature 1976;266:55–56. 2 Dardenne M, Pleau JM, Nabarra B, Lefrancier P, Derrien M, Choay J, Bach JF: Contribution of zinc and other metals to the biological activity of the serum thymic factor. Proc Natl Acad Sci USA 1982;79:5370–5373. 3 von Gaudecker B, Kendall MD, Ritter MA: Immuno-electron microscopy of the thymic microenvironment. Microsc Res Tech 1997; 38:237–249. 4 Kendall MD, Marsh JA: Neuroendocrine activity of the thymus; in Marsh JA, Kendall MD (eds): The Physiology of Immunity. Boca Raton, CRC Press, 1996, pp 171–182. 5 Safieh-Garabedian B, Kendall MD, Khamashta MA, Hughes GRV: Thymulin and its role in immunomodulation. J Autoimmun 1992;5: 547–555. 6 Dardenne M, Charreire J, Bach JF: Alterations in thymocyte surface markers after in vivo treatment by serum thymic factor. Cell Immunol 1978;39:47–54. 7 Incefy GS, Mertelsmann R, Dardenne M, Bach JF, Good RA: Induction of differentiation in human marrow T cell precursors by the synthetic serum thymic factor (FTS). Clin Exp Immunol 1980;40:396–400.

Thymulin and Pain-Related Behavior

8 Kaiserlian D, Dardenne M: Studies on the mechanisms of the inhibitory effects of serum thymic factor on murine allograft immunity. Cell Immunol 1982;66:360–371. 9 Muzioli M, Mocchegiani E, Bressani N, Bevilacqua P, Fabris N: In vitro restoration by thymulin of NK activity of cells from old mice. Int J Immunopharmacol 1992;14:57–61. 10 Dardenne M, Savino W: Control of thymus physiology by peptidic hormones and neuropeptides. Immunol Today 1994;15:518–523. 11 Safieh B, Venn GE, Ritter M, Singh S, Buckingham JC, Kendall MD: Plasma thymulin concentrations in cardiac patients: Involvement with the hypothalamo-pituitary-adrenal axis. J Physiol 1991;438:45P. 12 Buckingham JC, Safieh B, Singh S, Arduino LA, Cover PO, Kendall MD: Interactions between the hypothalamo-pituitary-adrenal axis and the thymus in the rat: A role for corticotrophin in the control of thymulin release. J Neuroendocrinol 1992;4:295–301. 13 Timsit J, Savino W, Safieh B, Chanson P, Bach JF, Dardenne M: Growth hormone and insulin-like growth factor-1 stimulate hormonal function and proliferation of thymic epithelial cells. J Clin Endocrinol Metab 1992;75:183– 188.

14 Fabris N, Mocchegiani E, Mariotti S, Pacini F, Pinchera A: Thyroid function modulates thymic endocrine activity. J Clin Endocrinol Metab 1986;62:474–478. 15 Owens T, Renno T, Taupin V, Krakowski M: Inflammatory cytokines in the brain: Does the CNS shape immune responses? Immunol Today 1994;15:566–571. 16 Safieh-Garabedian B, Khalid A, Khamashta MA, Hughes GRV: Thymulin modulates cytokine release by peripheral blood mononuclear cells: A comparison between healthy volunteers and patients with systemic lupus erythematosus. Int Arch Allergy Immunol 1993;101:126– 131. 17 Ferreira SH, Lorenzetti BB, Bristow AF, Poole S: Interleukin-1-beta as a potent hyperalgesic agent antagonized by a tripeptide analogue. Nature 1988;334:698–700. 18 Watkins LR, Goehler LE, Relton J, Brewer MT, Maier SF: Mechanisms of tumor necrosis factor-alpha hyperalgesia. Brain Res 1995;692: 244–250. 19 Safieh-Garabedian B, Poole S, Allchorne A, Winter J, Woolf CJ: Contribution of interleukin-1ß to the inflammation-induced increase in nerve growth factor levels and inflammatory hyperalgesia. Br J Pharmacol 1995;115:1265– 1275.

Neuroimmunomodulation 1999;6:39–44

43

20 Kanaan SA, Saadé NE, Haddad JJ, Abdelnoor AM, Jabbur SJ, Safieh-Garabedian B: Endotoxin-induced local inflammation and hyperalgesia in rats and mice: A new model for inflammatory pain. Pain 1996;66:373–379. 21 Safieh-Garabedian B, Jalakhian RH, Saadé NE, Haddad JJ, Jabbur SJ, Kanaan SA: Thymulin reduces hyperalgesia induced by peripheral endotoxin injection in rats and mice. Brain Res 1996;717:179–183. 22 Lewin GR, Ritter AM, Mendell LM: Nerve growth factor-induced hyperalgesia in the neonatal and adult rat. J Neurosci 1993;13:2136– 2148. 23 Donnerer J, Schuligoi R, Stein C: Increased content and transport of substance P and calcitonin gene-related peptide in sensory nerves innervating inflamed tissue: Evidence for a regulatory function of nerve growth factor in vivo. Neuroscience 1992;49:693–698. 24 Woolf CJ, Safieh-Garabedian B, Ma Q-P, Crilly P, Winter J: Nerve growth factor contributes to the generation of inflammatory sensory hypersensitivity. Neuroscience 1994;62:327– 331.

44

25 Safieh-Garabedian B, Kanaan SA, Haddad JJ, Abou Jaoude P, Jabbur SJ, Saadé NE: Involvement of interleukin-1ß, nerve growth factor and prostaglandin-E2 in endotoxin-induced localized inflammatory hyperalgesia. Br J Pharmacol 1997;121:1619–1626. 26 Safieh-Garabedian B, Jalakhian RH, Jabbur SJ, Saadé NE, Kanaan SA: Thymulin at high doses reduces endotoxin-induced hyperalgesia by reducing interleukin-1ß and nerve growth factor levels in the hind paw of rats; in Apkarian AV, Ayrapetian S (eds): Pain Mechanisms and Management. Ohmsha, IOS Press, 1998, pp 131–138. 27 Ritter MA, Crispe IN: The thymic microenvironment; in Male D (ed): The Thymus. New York, Oxford University Press, 1992, pp 57– 72. 28 Safieh-Garabedian B, Kanaan SA, Jalakhian RH, Poole S, Jabbur SJ, Saadé NE: Hyperalgesia induced by low doses of thymulin injections: Possible involvement of prostaglandin E2. J Neuroimmunol 1997;73:162–168. 29 Safieh-Garabedian B, Kanaan SA, Jalakhian RH, Jabbur SJ, Saadé NE: Involvement of interleukin-1ß, nerve growth factor, and prostaglandin-E2 in the hyperalgesia induced by intraplantar injections of low doses of thymulin. Brain Behav Immun 1997;11:185–200.

Neuroimmunomodulation 1999;6:39–44

30 Poole S, Bristow AF, Lorenzetti BB, Gaines Das RE, Smith TW, Ferreira SH: Peripheral analgesic activities of peptides related to ·melanocyte-stimulating hormone and interleukin-1ß 193–195. Br J Pharmacol 1992;106: 489–492. 31 Gualde N, Rigaud M, Bach JF: Stimulation of prostaglandin synthesis by the serum factor (FTS). Cell Immunol 1982;70:362–366. 32 Rinaldi-Garaci C, Jezzi T, Baldassare AM, Dardenne M, Bach JF, Garaci E: Effect of thymulin on intracellular cyclic nucleotides and prostaglandin E2. Eur J Immunol 1985;15: 548–552. 33 Aono H, Morishita M, Sasano M, Okamoto M, Okahara A, Nakata K, Mita S: Amelioration of type II collagen-induced arthritis in rats by treatment with thymulin. J Rheumatol 1997; 24:1564–1569. 34 Safieh-Garabedian B, Major SC, Jabbur SJ, Atweh SF, Saadé NE: Thymulin-induced hyperalgesia is mediated through capsaicin-sensitive primary afferent fibers. 3rd Annual Meeting Brain-Immune Network Group, London, Sept 4–5, 1997.

Safieh-Garabedian/Kanaan/Jabbur/Saadé

Neuroimmunomodulation 1999;6:45–50

Effect of Transmitters and Co-Transmitters of the Sympathetic Nervous System on Interleukin-6 Synthesis in Thymic Epithelial Cells Birte von Patay

Bodo Kurz Rolf Mentlein

Universität Kiel, Anatomisches Institut, Kiel, Deutschland

Key Words Thymic epithelial cells W Interleukin-6 W Catecholamines W Purines W Noradrenaline W Adenosine triphosphate W Adrenoceptors

Abstract In the thymus, sympathetic nerves run in septa in close connection to subcapsular/perivascular thymic epithelial cells (TEC). Since TEC are supposed to create a microenvironment of cytokines necessary for the development of thymocytes to T cells, we investigated the influence of sympathetic transmitters and co-transmitters on interleukin-6 (IL-6) synthesis in cultivated rat TEC that express markers of perivascular/subcapsular TEC. Noradrenaline and ATP stimulated IL-6 production in the culture supernatants 14- and 23-fold over basal values after 24 h. Costimulation with noradrenaline and ATP yielded an additive effect. Synthesis of IL-6 was concentration-dependent upon ATP and appeared to be mediated by P2 purinoceptors. During 24 h stimulation with 1 mM ATP, two thirds of the ligand was degraded mainly to ADP, production of AMP and adenosine was minor or negligible. Thus, in TEC, transmitters and co-transmitters of the sympathetic nervous system have a co-stimulatory effect on synthesis of IL-6 that is an important factor for thymocyte differentiation and proliferation.

ABC

© 1999 S. Karger AG, Basel 1021–7401/99/0062–0045$15.00/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: http://BioMedNet.com/karger

Introduction

Thymocytes maturate and differentiate into T cells in the thymus under the influence of a microenvironment that is created by a number of supporting cells. Thymic epithelial cells (TEC) are supposed to be the major cell type involved, mast cells, fibroblasts and cells of the monocytic-macrophage lineage may be of additional importance [1, 2]. As key factors of the thymic microenvironment, intercellular signal substances like cytokines or other paracrine peptides as well as surface molecules are regarded. TEC have been shown to synthesize the interleukins (IL)-1, IL-3, IL-6 and IL-7, leukemia inhibitory factor, granulocyte-macrophage- as well as granulocyteand macrophage-colony-stimulating factors [3, 4], neuropeptides like calcitonin gene-related peptide [5], oxytocin and vasopressin and special thymic peptides like thymulin [6]. However, the factors regulating the release of these bioactive peptides are largely unknown. TEC, in particular subcapsular and medullary TEC, are exposed to circulating hormones as well as to direct innervation. In previous studies we and others have shown that catecholaminergic nerve fibers innervating the thymus run mainly in the septa in close connection to subcapsular/perivascular TEC and normally do not enter the thymus parenchyma [7, 8]. Noradrenaline has been identified as the main catecholamine in the thymus [9],

Prof. Dr. Rolf Mentlein Universität Kiel, Anatomisches Institut Olshausenstrasse 40 D–24098 Kiel (Germany) Tel. +49 431 8802460, Fax +49 431 8801557, E-Mail [email protected]

and TEC have been shown to express functionally active ß1- and ß2-adrenoceptors [7]. In sympathetic nerves, adenosine triphosphate (ATP) is co-released together with noradrenaline. ATP itself or adenosine, the product of its enzymatic hydrolysis, bind to special purine receptors [10, 11] and induce different intracellular responses depending on the receptor subtype. Apart from effects on their own, purines as co-transmitters of the sympathetic nervous system may also modulate the (nor)adrenergic response in their target cells. In order to investigate the possible neuromodulatory influence of catecholamines and purines in the thymus, we therefore evaluated in vitro whether noradrenaline (or ß-adrenergic agonists) and ATP, alone or in combination, might regulate the synthesis and release of the cytokine IL-6 that is an important component of the thymus microenvironment. The studies were performed with cultured rat TEC which had been shown to express subcapsular/perivascular markers [12] and thus may be potentially involved in the formation of the thymic microenvironment.

Materials and Methods TEC Cultures TEC were prepared from thymus of 6- to 8-week-old female Wistar rats (strain Han:WIST) by collagenase dissociation. Cultures were purified and freed from fibroblasts by combination of a preplating method, differential trypsin treatment and cultivation in the presence of horse serum described in detail elsewhere [12]. TEC were cultured in Dulbecco’s modified Eagle’s medium (DMEM)/F12 (1:1) medium containing 2 mM L-glutamine, 25 Ìg/ml transferrin, 5 Ìg/ ml insulin, 1 ng/ml cholera toxin, 100 U/ml penicillin and streptomycin and 10% horse serum, and used after 3–25 passages. Immunocytochemistry Purity of the cultures was routinely checked by immunostaining with anti-cytokeratin as described previously [12]. In brief, cells cultivated on poly-D-lysine-coated coverslips were fixed with cold acetone (–20 ° C) for 10 min, rinsed in Tris-buffered saline (TBS; 0.14 M NaCl 20 mM Tris/HCl, pH 7.4), incubated with 10% horse serum in TBS, stained overnight at 4 ° C with monoclonal anti-pan-cytokeratin (1:50 in TBS; M821 from Dako, Glostrup, Denmark), washed with TBS, incubated fluorescein-isothiocyanate-conjugated horse antimouse IgG (1:50 in TBS; FI-2000 from Vector, Burlingame, Calif., USA), embedded in Citifluor and inspected microscopically at 390– 420 nm excitation and 510–560 nm emission. Stimulation Experiments and Cytokine Measurements TEC were cultured in 35-mm culture dishes (200,000 cells/dish) until confluence and incubated in DMEM/F12 (1:1) containing 10% horse serum with the agonists at the concentrations indicated. Solutions of catecholamines and purines (purchased from Sigma, Munich, Germany) were freshly prepared for each experiment. After

46

Neuroimmunomodulation 1999;6:45–50

24 h the supernatants were harvested, stored at –20 ° C or directly used for cytokine assay. IL-6 was determined using a commercial ELISA for rat IL-6 (Biosource, Camarillo, Calif., USA) as described by the manufacturer. Cell numbers were determined (a) by trypsination and counting cells in a Neubauer-chamber, and (b) after lysis of the cells by fluorometric determination of the DNA content with ethidium bromide [13]. A suspension of 500,000 cells obtained by trypsination and counting served as reference (5 pg DNA/cell; in accordance to published values for rat cells [cf. 13]). All data are expressed as mean values B SEM for n tested samples. Statistical significance was tested with the t-test. Degradation of ATP TEC (20,000 cells) in 24-well plates were grown to confluence (5 days) as above. Then, the medium was replaced by a synthetic incubation medium that consisted of 145 mM NaCl, 5.4 mM KCl, 20 mM glucose, 20 mM Hepes buffer, pH 7.4. After addition of 1 mM ATP, cells were incubated at 37 ° C (without CO2 atmosphere) and aliquots of the supernatants were withdrawn after 0, 6, 12 and 24 h. ATP and its degradation products were analyzed by ion exchange chromatography. For this, aliquots were diluted 4-fold in 20 mM Tris/HCl buffer, pH 8.0, and 50 Ìl diluted sample applied onto a Mini Q PC3.2/3 ion exchange column that was eluted by 5 ml of a linear gradient of 0–0.25 M NaCl in 20 mM Tris/HCl buffer, pH 8.0, at a flow rate of 400 Ìl/min. Purines were detected in the eluate by their absorbance at 259 nm. 10 nmol ATP, ADP, AMP or adenosine (Sigma) served as standards. As control, 1 mM ATP was incubated in the above buffer without cells and potential degradation monitored as above.

Results Noradrenaline and ATP Stimulate IL-6 Production in TEC

Cultured rat TEC served as an in vitro model to study the influence of noradrenaline and ATP on IL-6 synthesis. The cultures used were more than 95% immunopositive for cytokeratin, a general epithelial marker (fig. 1), and devoid of contaminating thymocytes, fibroblasts and macrophages as evidenced by immunostaining with the cell-specific markers CD5, vimentin and CD68 [cf. 12]. The cultured TEC expressed several markers of subcapsular and medullary TEC, a subset (5–10%) was positive for cortical markers as described previously in detail [12]. When applied alone for 24 h on TEC, noradrenaline (at optimal concentration of 10 ÌM) and ATP induced IL-6 synthesis about 14- or 23-fold as determined by a specific ELISA (fig. 2). The higher stimulatory effect of ATP over noradrenaline could be repeated in several independent experiments with different individual cultures.

von Patay/Kurz/Mentlein

Co-Stimulation of Noradrenaline and ATP Increases IL-6 Synthesis Co-stimulation of TEC with noradrenaline and ATP increased IL-6 synthesis about 32-fold (fig. 2). The nearly additive effect of the transmitter and co-transmitter was constantly observed in several experiments.

Stimulation of IL-6 Synthesis by ATP Is Dose-Dependent To show that the effect of ATP on IL-6 synthesis is mediated via specific, saturable receptors, we investigated its dependency from the ATP concentration (fig. 5). A half-maximum stimulation was obtained with about 0.3 ÌM ATP, the maximal effect with about 1 mM. However, the partial degradation of the ligand during the stimulation should be considered.

Discussion

Fig. 1. Immunocytochemical characterization of cultured rat TEC

(20th subculture) with cytokeratin antibodies. Immunofluorescence staining for the epithelial cell marker cytokeratin shows tonofilaments that form a network in the cytoplasm and converge to desmosomes connecting adjacent cells to each other.

400

300 IL-6 (pg/106 cells)

ATP Is Partly Degraded during the Incubations Since ATP is a relatively labile compound in several cellular systems, we investigated its stability in our TEC cultures. Supernatants of cultures incubated parallel to the stimulation experiments were analyzed for ATP degradation to less phosphorylated products after different times by high-pressure ion exchange chromatography (fig. 3). Under conditions comparable to the above assays, a considerable part of ATP is degraded mainly to ADP, whereas AMP and adenosine are only formed in trace amounts (fig. 4). However, even after 24 h about one third of ATP is intact, and the initial degradation rate of about 55 nmol W h –1 W 106 cells –1 is relatively low as compared to other cultivated cells (e.g. microglia cells analyzed in our laboratory).

200

100

Numerous studies have demonstrated interactions between the nervous/endocrine and the immune system. We evaluated the potential influence of noradrenaline, a transmitter of the sympathetic nervous system, and its cotransmitter, ATP, on the thymus microenvironment using a cell culture model. Immunocytochemically defined cultured rat TEC were exposed to stimulation of both transmitters alone or in combination and the synthesis of IL-6, an important cytokine of the microenvironment, was measured in the culture supernatants. In previous studies we have already shown that the basal synthesis of IL-6 in our cultures is very low and that nearly all IL-6 produced after stimulation is released into the medium [14]. We could also show that noradrenaline,

Modulation of Cytokine Synthesis in TEC

0 None

NA

ATP

NA + ATP

Fig. 2. Effect of noradrenaline (NA, 10 ÌM) and adenosine triphosphate (ATP, 1 mM ) alone or in combination on IL-6 synthesis by TEC after 24 h at 37 ° C. Rat IL-6 was determined by a specific ELISA, cell numbers after the experiment by DNA measurement (subculture 22, n = 3, B SEM). Statistical significance: none versus all others: p ! 0.01, combination NA + ATP versus NA p ! 0.01 and versus ATP p ! 0.05.

Neuroimmunomodulation 1999;6:45–50

47

3

48

Neuroimmunomodulation 1999;6:45–50

von Patay/Kurz/Mentlein

100

0h

12 h

90

6h

24 h

Peak area (% of total)

80 70 60 50 40 30 20 10 0

ATP

ADP

AMP

Adenosine

Fig. 4. ATP metabolites in TEC culture supernatants after different incubation times. Quan-

titative evaluation of the peak areas after ion exchange chromatography shown in figure 3. ATP is degraded to ADP, from which AMP and adenosine are formed in small amounts after longer incubation times (n = 3).

adrenaline or the ß-adrenergic agonist isoproterenol dosedependently raised IL-6 synthesis to similar maximal values. This effect was mediated via ß-adrenoceptors that have been identified on our cultured TEC on the mRNA, protein and functional level [7]. We now demonstrate that also ATP via saturable receptors increases – even higher than noradrenaline – IL-6 synthesis in TEC. Since part of the ATP is degraded only to ADP during the incubation and adenosine is formed only in trace amounts, the ATP receptors responsible for this effect should belong to the P2-type of purinoceptors that are activated by ATP or ADP, but not by

Fig. 5. Dependency of IL-6 synthesis in TEC from extracellular ATP concentrations. Cells were stimulated and assayed for IL-6 production as described in figure 2.

Fig. 3. Analysis of ATP degradation in culture supernatants by highpressure ion exchange chromatography. TEC (bold line) under conditions parallel to the IL-6 measurements or blank culture dishes (thin lines) were incubated with 1 mM ATP, aliquots of the culture supernatants withdrawn, diluted and applied on a Mini Q ion exchange column that was eluted by 5 ml of a linear gradient of 0–25 M NaCl in 20 mM Tris/HCl, pH 8.0 (see Methods). ATP, ADP, AMP and adenosine (A) detected in the eluate by absorption at 259 nm were identified and quantified by their retention times and peak areas obtained with standards; * unknown product. Doubling of purine peaks results from the different ionic forms of the purines. In the absence of cells, ATP is practically not degraded, TEC produce ADP as the main metabolite.

adenosine [10, 11]. The subtype involved has to be further clarified by using more specific agonists or antagonists. An atypical ATP receptor (not classified in the present nomenclature) has been described in the rat thymic epithelial cell line TEA3A1 [15]. Since the effect of noradrenaline and ATP are additive, we suggest that they are induced by different signal transduction pathways. Activation of ß-adrenoceptors results in an elevation of intracellular cAMP levels in TEC [7], P2-purinoceptors are (as far known) linked to elevation of inositol triphosphate/diglyceride or open ion channels [11]. The promoter region of IL-6 which controls its expression contains a cAMP response element, an AP-1 site, a glucocorticoid response element, and an NF-ÎB response element that can be activated either by protein kinases A (activated by cAMP), C (activated by diglycerides) or other kinases [16, 17]. Thus, the IL-6 promoter contains different targets for the signal transduction of ßadrenoceptors and P2-purinoceptors that control the cellspecific expression of IL-6 in TEC. Regulation of IL-6 expression (or that of other cytokines like IL-1, IL-7 or IL-15) in TEC by neurotransmitters or hormones is of particular interest because cytokines have been implicated in T-cell development [18]. IL-6 in combination with interferon-Á and IL-2 induces

Modulation of Cytokine Synthesis in TEC

Neuroimmunomodulation 1999;6:45–50

49

the differentiation of cytotoxic T cells from immature thymocytes [19]. IL-6 together with IL-2 or IL-4 enhances the proliferation of CD4+CD8- or CD4-CD8+ single positive mouse thymocytes [20, 21]. IL-6 induces the expression of the p55 IL-2 receptor [22] and serves as a second signal for the production of IL-2 by CD4+ thymocytes activated by a submitogenic concentration of mitogen or by T-cell receptor cross-linking [23]. In addition to its effects on thymocytes, IL-6 has been shown to act as an autocrine (or thymocyte-derived) growth factor for TEC themselves [24, 25]. A thymic hyperplasia found in patients with myasthenia gravis may be caused by a de-regulated IL-6 production in TEC resulting in an increased TEC proliferation [26].

These findings show that IL-6 is an important factor of the thymic microenvironment produced by TEC. The costimulatory effects of catecholamines and ATP on IL-6 synthesis in TEC in vitro should – when relevant in vivo – promote especially the differentiation and proliferation of single positive thymocytes. It may be speculated that adrenergic stimulation might augment the production of T cells.

Acknowledgments We thank Dagmar Freier for her excellent technical assistance. This work was supported by the Stiftung Volkswagenwerk in the project ‘Neuroimmunology, Mood and Behaviour’.

References 1 Boyd RL, Tucek CL, Godfrey DI, Izon DJ, Wilson TJ, Davidson DJ, Bean AGD, Ladyman HM, Ritter MA, Hugo P: The thymic microenvironment. Immunol Today 1993;14: 445–459. 2 Kendall MD: Functional anatomy of the thymus microenviroment. J Anat 1991;177:1–29. 3 Le PT, Singer KH: Human thymic epithelial cells: Adhesion molecules and cytokine production. Int J Clin Lab Res 1987;23:56–60. 4 Wolf SS, Cohen A: Expression of cytokines and their receptors by human thymocytes and thymic stromal cells. Immunology 1992;77: 362–368. 5 Kurz B, von Gaudecker B, Kranz A, Krisch B, Mentlein R: Calcitonin gene-related peptide and its receptor in the thymus. Peptides 1995; 16:1497–1503. 6 Kendall MD: The thymus: New views of an old gland. Endeavour 1992;16:158–163. 7 Kurz B, Feindt J, von Gaudecker B, Kranz A, Loppnow H, Mentlein R: ß-Adrenoceptor-mediated effects in rat cultured thymic epithelial cells. Br J Pharmacol 1997;120:1401–1408. 8 Kendall MD, Al-Shawaf H, Zaidi SA: The cholinergic and adrenergic innervation of the rat thymus. Adv Exp Med Biol 1998;237:255– 261. 9 Kendall MD, Atkinson BA, Mun´oz FJ, de la Riva C, Clarke A, von Gaudecker B: The noradrenergic innervation of the rat thymus during pregnancy and in the post partum period. J Anat 1994;185:617–625. 10 Linden J: Purinergic systems; in Siegel GJ, Agranoff BW, Albers RW, Molinoff PB (eds): Basic Neurochemistry. New York, Raven Press, 1993, pp 401–416.

50

11 Malec D: Purinergic receptors. Pol J Pharmacol 1996;48:457–465. 12 Kurz B, von Gaudecker B, Krisch B, Mentlein R: Rat thymic epithelial cells in vitro and in situ: Characterisation by immunocytochemistry and morphology. Cell Tissue Res 1996;283: 221–229. 13 Karsten U, Wollenberger A: Improvements of the ethidium bromide method for direct fluorometric estimation of DNA and RNA in cell and tissue homogenates. Anal Biochem 1977;77: 464–470. 14 von Patay B, Loppnow H, Feindt J, Kurz B, Mentlein R: Catecholamines and lipopolysaccharide synergistically induce the release of interleukin-6 from thymic epithelial cells. J Neuroimmunol 1998;86:182–189. 15 Liu P, Wen M, Hayashi J: Characterization of ATP receptor responsible for the activation of phospholipase A2 and stimulation of prostaglandin E2 production in thymic epithelial cells. Biochem J 1995;308:399–404. 16 Tanabe OS, Akira T, Kamiya GG, Wong T, Hirano T, Kishimoto T: Genomic structure of the murine IL-6 gene: High degree of conservation of potential regulatory sequences between mouse and human. J Immunol 1988;141: 3875–3881. 17 Kishimoto T: The biology of interleukin-6. Blood 1989;74:1–10. 18 Carding SR, Hayday AC, Bottomly K: Cytokines in T-cell development. Immunol Today 1991;12:239–245. 19 Takai Y, Wong GG, Clark SC, Burakoff SJ, Herman SH: B-cell stimulatory factor-2 is involved in the differentiation of cytotoxic Tlymphocytes. J Immunol 1987;40:508–513.

Neuroimmunomodulation 1999;6:45–50

20 Chen W-F, Fischer M, Frank G, Zlotnik A: Distinct patterns of lymphokine requirement for the proliferation of various subpopulations of activated thymocytes in single cell assay. J Immunol 1989;143:1598–1605. 21 Suda T, Murray R, Guidos C, Zlotnik A: Growth-promoting activity of IL-1·, IL-6, and tumor necrosis factor-· in combination with IL-2, IL-4, or IL-7 on murine thymocytes. J Immunol 1990;144:3039–3045. 22 Le J, Frederick G, Reis LFL, Diamantstein T, Hirano T, Kishimoto T, Vileck J: Interleukin2-dependent and interleukin-2-independent pathways of regulation of thymocyte function by interleukin 6. Proc Natl Acad Sci USA 1988; 85:8643–8647. 23 Garmann RD, Jacobs KA, Clark SC, Raulet DH: B-cell stimulating factor 2 (ß2-interferon) functions as a second signal for interleukin-2 production by mature murine T cells. Proc Natl Acad Sci USA 1987;84:7629–7633. 24 Colic M, Pejnovic N, Kataranovski M, Popovic L, Gasic S, Dujic A: Interferon gamma alters the phenotype of rat thymic epithelial cells in culture and increases interleukin-6 production. Dev Immunol 1992;2:151–160. 25 Meilin A, Shoham J, Schreiber L, Sharabi Y: The role of thymocytes in regulating thymic epithelial cell growth and function. Scand J Immunol 1995;42:185–190. 26 Cohen-Kaminsky S, Devergne O, Delattre RM, Klingel-Schmitt I, Emilie D, Galanaud P, Berrih-Akinin S: Interleukin-6 overproduction by cultured thymic epithelial cells from patients with myasthenia gravis is potentially involved in thymic hyperplasia. Eur Cytokine Network 1993;4:121–132.

von Patay/Kurz/Mentlein

Neuroimmunomodulation 1999;6:51–55

A Possible Role for Acetylcholine in the Dialogue between Thymocytes and Thymic Stroma I. Rinner a A. Globerson b K. Kawashima c W. Korsatko d K. Schauenstein a a Institute

of General and Experimental Pathology, University of Graz, Austria; b Department of Cell Biology, The Weizmann Institute of Science, Rehovot, Israel; c Kyoritsu College of Pharmacy, Tokyo, Japan; d Institute of Pharmaceutical Technology, University of Graz, Austria

Key Words Thymus W Thymic epithelial cells W Apoptosis W Nicotinic acetylcholine receptors W Acetylcholine production

Abstract In this article we will review data suggesting that acetylcholine takes part in the mutual interplay between developing T cells and thymic epithelium, and thereby may influence the generation of the T-cell repertoire. In the first part we will recapitulate our findings according to which cholinergic agonists affect thymocyte apoptosis via a nicotinergic effect on thymic epithelial cells. In the second part we will present evidence that acetylcholine within the thymus is mainly derived from the thymocytes themselves, and that the production and release of this neurotransmitter is dependent on activation of thymic lymphocytes.

Introduction

The selection processes of developing T cells in the thymus are not only based on cellular recognition events with antigen presenting cells and/or thymic epithelial cells [1], but are also regulated by soluble mediators produced by stromal cells [2], whereby the developing T cells in turn appear to provide maturation and survival signals to

ABC

© 1999 S. Karger AG, Basel 1021–7401/99/0062–0051$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: http://BioMedNet.com/karger

maintain the thymic microenvironment [3, 4]. Furthermore, endocrine and neuroendocrine signals may have a significant impact on thymic functions. Glucocorticoids exert strong apoptotic effects on immature and mature lymphocytes, and it has recently been shown that these hormones are locally produced within the thymus and that the receptor antagonist RU-486 affects spontaneous apoptosis of thymic lymphocytes [5, 6]. Our own interest focuses on the relevance of the autonomic nervous system in this respect. The thymus is strongly innervated with adrenergic fibers [7] and there exist older data by Singh et al. [8] suggesting that sympathetic denervation results in enhanced development of fetal thymic lobes in nude mice. Very recently, evidence was reported that catecholamines induce apoptosis of thymocytes in vitro [9], and this is in line with our own observations in vivo [Rinner et al., unpubl. data]. Much less is known about a cholinergic regulation of thymic functions. Whether the thymus receives cholinergic innervation is still a matter of debate [10, 11]. Cholinergic as well as adrenergic receptors were recently detected on thymic lymphocytes [12, 13] and epithelial cells [14–16]. These findings prompted us to investigate more in depth the possible role of acetylcholine in the regulatory dialogue between thymic lymphocytes and stromal cells. More specifically, the following two questions were addressed: (1) What is the effect of cholinergic stimulation on the survival of developing T cells, and (2) What is the source of acetylcholine within the thymic microenvironment?

I. Rinner Institute of General and Experimental Pathology University of Graz Heinrichstrasse 31a A–8010 Graz (Austria)

Detection of choline acetyltransferase and acetylcholine: The determination of choline acetyltransferase activity in homogenates of thymus cell suspensions was performed according to the method of Fonnum [22]. The preparation of the acetylcholine extracts of thymus cells or their culture supernatants was done by acid precipitation. Acetylcholine content of the extracts was determined by a radioimmunoassay described by Kawashima et al. [23].

Results Effect of Cholinergic Treatment in vivo on Spontaneous Apoptosis of Rat Thymocytes

Fig. 1. Cholinergic protection of thymocytes against apoptosis in the rat. Chronic cholinergic treatment was performed for 7 days as described. AOlow cells: acridine orange low binding (apoptotic) cells. The results are means B SEM of 6 animals.

Materials and Methods Male Sprague-Dawley rats were implanted with constant-release pellets containing physostigmine with an output rate of 12 Ìg/day, or atropine (1.2 mg output rate/day). After 7 days thymus cells were prepared and the percentage of apoptotic cells was determined. In vitro studies: A coculture system of cortical or medullary murine (C57BL/6) thymic epithelial cell (TEC) lines together with fetal thymus lobes (FTLs) or thymocytes in suspension was used. TEC lines were prepared from thymus lobes as described [17]. The identification of the cortical (TEC 1.4) or medullary (TEC 2.3) origin of the TEC lines was performed using monoclonal antibodies reacting specifically with cortical [18] medullary TECs [19]. 5 ! 104 cells of either TEC line were seeded on Nucleopore filters resting on gelatin sponges as described previously [20]. After 24 h, four FTLs were placed on top of each TEC layer or on control sponges without TECs. Cultures were incubated in RPMI 1640 medium supplemented with 10% FCS. Carbachol at concentrations from 10 –6 to 10 –8 M was added each day to the cultures alone or together with the blocking agents (atropine, d-tubocurarine, 10 –5 M ). After 7 days the thymocytes were collected for flow cytometry. For the coculture with single thymus cell suspensions, 5 ! 104 TEC line cells were seeded into 24-well Costar plates and incubated for 24 h at culture conditions. Thereafter the adherent cells were washed and freshly prepared thymic lymphocytes (1 ! 106) from young and adult and the cholinergic drugs were added as indicated in the tables. The percentage of apoptotic and of dead cells was determined 24 h later. Flow cytometry: Analysis of CD4/CD8 subsets was performed with a Becton-Dickinson FACScan cytometer using anti-Lyt2 (CD8) FITC and anti-L3T4 (CD4) phycoerythrin conjugates. Apoptotic cells were identified by DNA staining with acridine orange and determination of the percentage of the cells with low, subdiploid binding of the dye by flow cytometry according to Compton et al. [21].

52

Neuroimmunomodulation 1999;6:51–55

Five days’ treatment of rats with the cholinergic antagonist atropine leads to a pronounced increase in the percentage of apoptotic thymocytes, whereas physostigmine, which enhances the endogenous cholinergic tonus by inhibition of acetylcholinesterase, has no effect. Thus, it appears that endogenously produced acetylcholine protects from apoptosis in thymocytes in vivo (fig. 1). Cholinergic Effect on Murine Thymocytes Cultured with TEC Lines TEC 1.4 and TEC 2.3 enhanced apoptosis in cells of cocultivated FTLs, the effect being more pronounced with the cortical TEC line 1.4 (table 1). In vitro treatment with carbachol at a concentration of 10 –7 M suppressed significantly the apoptosis-promoting effect of TEC 1.4, but not of TEC 2.3 (table 1). Other concentrations were ineffective (not shown). No effect on apoptosis was observed on FTL cultures in the absence of TEC lines. TEC 1.4 exhibited also the effect of decreasing the proportion of CD4+CD8+ double positive (DP) cells in cocultured thymus lobes (table 2). Daily addition of 10–7 M carbachol significantly antagonized this effect of TEC line 1.4 also. The CD4–CD8– double negative (DN) subset was influenced in the opposite way (not shown). Other concentrations of carbachol were again ineffective. In contrast to the effects on the FTLs, both TEC lines had the same promoting effect on apoptosis of thymus cells in suspension (table 3). This appears to be a thymusspecific phenomenon, as adherent fibroblasts had no effect on cocultured thymus cell suspensions, and the viability of spleen cells was not affected by coculture with either TEC line under investigation. The apoptosis-promoting effect of TEC 1.4 and of TEC 2.3 is not dependent on physical contact of the TECs with the thymus cells, since the conditioned supernatants of both TECs induced apoptosis in freshly prepared thymus cells (not shown). The active factor(s) is resistant to dialysis and has a molecular weight of 130 kD. Addition of carbachol

Rinner/Globerson/Kawashima/Korsatko/ Schauenstein

Table 1. Effect of in vitro treatment with carbachol on the percentage of apoptotic cells in murine fetal thymus cocultured with cortical (TEC 1.4) or medullary (TEC 2.3) epithelial cell lines (results are means B SEM of three independent experiments)

Control 10 –8 M carbachol 10 –7 M carbachol 10 –6 M carbachol

No TEC

TEC 1.4

100.0B1.8 95.8B16.4 101.2B14.5 115.2B10.3

307.3B1.2+++ 165.5B19.4 257.6B28.4+++ 112.1B26.1 207.9B9.1**, +++ 105.5B24.8 267.3B12.2+++ 98.2B12.1

Table 2. Effect of in vitro treatment with carbachol on the percentage of CD4+CD8+ cells in murine thymus cell suspensions cocultured with cortical (TEC 1.4) or medullary (TEC 2.3) epithelial cell lines; data are presented as % of controls (results are means B SEM of three independent experiments)

TEC 2.3

** Significantly different (p ! 0.01) from the respective control without drug. +++ Significantly different (p ! 0.001) from lobes cultured without TEC according to ANOVA and Student-Newman-Keul’s post-hoc analysis.

Control 10 –8 M carbachol 10 –7 M carbachol 10 –6 M carbachol

No TEC

TEC 1.4

TEC 2.3

83.0B1.5 85.3B2.7 86.7B2.9 89.1B2.9

63.1B1.5+++ 66.2B0.9+++ 73.3B1.3*, ++ 66.2B2.1+++

81.2B1.3 77.3B2.0 79.7B1.5 81.6B2.4

* Significantly different (p ! 0.05) from the respective control without drug. ++, +++ Significantly different (p ! 0.01 and 0.001 respectively) from cultures without TEC, according to ANOVA and Student-Newman-Keul’s post-hoc analysis.

Table 3. Effect of in vitro treatment with carbachol and carbachol in combination with dtubocurarine (10 –5 M ) or with atropine (10 –5 M ) on the percentage apoptotic cells in murine thymus cell suspensions cocultured with cortical (TEC 1.4) or medullary (TEC 2.3) epithelial cell lines; data are presented as % of controls – apoptotic cells in cultures without TEC, i.e. 45% of all cells (results are means B SEM of three independent experiments)

Control 10 –8 M carbachol 10 –7 M carbachol 10 –6 M carbachol 10 –7 M carbachol + 10 –5 M d-tubocurarine 10 –7 M carbachol + 10 –5 M atropine

No TEC

TEC 1.4

TEC 2.3

100.0B0.1 101.6B4.4 95.6B2.2 101.2B2

160.0B0.2+++ 151.6B1.6+++ 128.2B2.0*, ++ 140.6B2.7+++

173.3B3.8+++ 181.6B7.3+++ 168.9B7.9+++ 164.4B5.9+++

102.2B4.4

153.3B8.9+++

177.7B9.7+++

126.6B11.1*, +++

173.3B9.5+++

88.9B4.4

* Significantly different (p ! 0.05) from the respective control without drug. ++, +++ Significantly different (p ! 0.01 and 0.001 respectively) from cultures without TEC, according to ANOVA and Student-Newman-Keul’s post-hoc analysis.

(10 –7 M) again attenuated the effect of the cortical TEC 1.4, but not of the medullary TEC 2.3 (table 2). Spontaneous apoptosis of the thymus cells in culture without TEC was not affected by cholinergic drugs. Since d-tubocurarine, but not atropine, counteracted this carbachol effect, it can be concluded that the effect of carbachol is mediated via nicotinic cholinergic receptors on the TEC 1.4 line. Source of Thymic Acetylcholine In view of the questionable data on cholinergic innervation of the thymus, we have hypothesized that immune

Acetylcholine-Epithelial Interactions in the Thymus Gland

cells may produce and secrete their own acetylcholine. Figure 2 shows the time course of the activity of the acetylcholine-producing enzyme, choline acetyltransferase, in thymus cells. We also confirmed the expression of the enzyme message in thymocytes by semiquantitative RTPCR (fig. 3). Direct determination of acetylcholine in homogenates of thymus cells showed that thymus cells contain acetylcholine. The amount of acetylcholine released in the culture supernatant is significantly increased by stimulation of the cells with PHA (fig. 4).

Neuroimmunomodulation 1999;6:51–55

53

✽✽

Fig. 2. Choline acetyltransferase activity radioenzymatically determined in rat thymus cells in dependence on time. Each point represents the mean B SEM of four independent experiments.

Fig. 4. Influence of mitogenic stimulation with optimal concentrations of PHA on acetylcholine production. Acetylcholine levels are determined by radioimmunoassay in the cellular homogenate (left) and in the culture supernatants (right) of rat thymus cells. Results are expressed as mean B SEM of six experiments.

Fig. 3. RT-PCR to detect choline acetyl-

transferase mRNA using four different primer pairs annealing in different sections of exon 1 of the rat choline acetyltransferase gene or ß-actin (lanes 5 and 10). Lanes 1–5, mRNA from thymus cells; lanes 6–10, mRNA from rat spinal cord.

Discussion

Our results suggest that cholinergic stimulation in vivo and in vitro regulates survival and differentiation of thymic lymphocytes. In rats, systemic in vivo application of the muscarinergic antagonist atropine increases the percentage of apoptotic cells spontaneously occurring in the thymus. We interpret this as a protective effect of the endogenous cholinergic tonus on developing thymocytes. These results appear in line with observations of Lindenboim et al. [24], according to which activation of muscarinic receptors mediates inhibition of cell cycle progression and apoptosis in nonlymphoid (PC12) cells. This in vivo effect, as observed in the rat being clearly mediated

54

Neuroimmunomodulation 1999;6:51–55

by muscarinic receptors, is to be distinguished from indirect protective effects of cholinergic stimulation via nicotinergic receptors of TECs, as detected in cocultures with fetal or adult murine thymocytes. Thus, albeit that the muscarinergic in vivo effect clearly needs further in-depth investigations, it appears that the cholinergic protection against thymocyte apoptosis is secured by different, possibly redundant mechanisms. Muscarinic and nicotinic cholinergic receptors have been identified on thymocytes and TECs [13, 14]. Since thymic cholinergic innervation, if it exists, is restricted to the perivascular space, the source of acetylcholine within the thymic parenchyma was obscure. We have been able to show that thymocytes express mRNA for choline ace-

Rinner/Globerson/Kawashima/Korsatko/ Schauenstein

tyltransferase, and that cellular homogenates exhibit measurable activities of this enzyme. It is noteworthy that the activities measured with isolated thymocytes were identical to those reported by others [25] in tissue homogenates, which were taken as evidence for a cholinergic innervation of the thymus. Thus, our data would suggest that thymic acetylcholine is indeed derived to the major part from lymphocytes, rather than from cholinergic nerve endings. The release of acetylcholine into the culture supernatants of thymic lymphocytes was significantly enhanced by mitogenic stimulation, which indicates that the production and secretion of this neurotransmitter is associated with cell activation. Furthermore, the production of acetylcholine is obviously not restricted to thymus cells, but if found to variable degrees also with peripheral T and B lymphocytes [26]. Although cholinergic stimulation has been reported to lead to changes in functions of peripheral lymphocytes [27] the possible roles of acetylcholine in

peripheral immunoregulation remain still to be defined. Also the mechanisms of acetylcholine release from immune cells are still enigmatic, as lymphocytes are not supposed to contain secretory vesicles. Very recent preliminary RT-PCR data, however, suggest the vesicular acetylcholine transporter protein to be expressed in rat lymphocytes [Rinner, unpubl. data]. In conclusion, our data suggest that acetylcholine has an impact on T-cell development in the thymus, and that thymocytes produce acetylcholine as part of the interplay with stroma cells providing the appropriate microenvironment for selection and maturation of the T-cell repertoire.

Acknowledgments Supported by the Austrian Research Council, project No. 10422 MED and by the Jubiläumsfonds der Österreichischen Nationalbank, project No. 4349.

References 1 Kappler JW, Roehm N, Marrack P: T-cell tolerance by clonal deletion. Cell 1987;49:273– 280. 2 Boyd LR, Tucek CL, Godfrey DI, Izon DJ, Wilson TJ, Davidson NJ, Bean AGD, Ladyman HM, Ritter MA, Hugo P: Rat thymic epithelial cells in culture constitutively secrete IL-1 and IL-6. Int Immunol 1991;3:1165– 1174. 3 Colic M, Vucevic D, Pavlovic MD, Lukic T, Milinkovic M, Popovic L, Popovic P, Dujic A: Adhesion molecules in the thymic microenvironment: Interactions between thymocytes and cloned thymic epithelial cell; in Immunoregulation in Health and Disease. New York, Academic Press, 1997, pp 13–33. 4 Ritter MA, Boyd RL: Development in the thymus: It takes two for a tango. Immunol Today 1993;14:462–469. 5 Vacchio MS, Ashwell J: Thymus-derived glucocorticoids regulate antigen-specific positive selection. J Exp Med 1997;185:2033–2038. 6 Vacchio MS, Papadopoulos V, Ashwell JD: Steroid production in the thymus: Implications for thymocyte selection. J Exp Med 1994;179: 1835–1846. 7 Felten SY, Felten DL, Bellinger DL, Carlson SL, Ackerman KD, Madden KS, Olschowka JA, Livnat S: Noradrenergic sympathetic innervation of lymphoid organs. Prog Allergy 1988;22:3–13. 8 Singh U, Fatani J, Mehta L, Mohajir AM: Implantation of fetal thymus and sympathetic ganglion within the anterior eye chamber in mice, to study neuro-immune interaction in thymic development. Acta Anat Basel 1990; 137:54–58.

Acetylcholine-Epithelial Interactions in the Thymus Gland

9 del Rey A, Besedowsky H: Noradrenaline induces apoptosis in mouse thymocytes. Symposium Volkswagenstiftung 1997, abstract 3. 10 Bulloch K, Moore RY: Innervation of the thymus gland by brain stem and spinal cord in mouse and rat. Am J Anat 1981;162:157–166. 11 Nance DM, Hopkins DA, Bieger DA: Reexamination of the innervation of the thymus gland in mice and rats. Brain Behav Immun 1987;1: 134–147. 12 Singh U, Millson DS, Smith PA, Owen JJ: Identification of beta-adrenoceptors during thymocyte ontogeny in mice. Eur J Immunol 1979;9:31–35. 13 Rinner I, Porta S, Schauenstein K: Characterization of 3H-methylscopolamine binding to intact rat thymocytes. Endocrinol Exp 1990;24: 125–132. 14 Engel EK, Trotter JL, McFarlin DE, McIntosh CL: Thymic epithelial cell contains acetylcholine receptor. Lancet 1977;i:1310–1311. 15 Kurtz B, Feindt J, von Gaudecker B, Kranz A, Loppnow H, Mentlein R: Beta-adrenoceptormediated effects in rat cultured thymic epithelial cells. Br J Pharmacol 1997;120:1401– 1408. 16 Wakkach A, Guyon T, Bruand C, Tzartos T, Cohen-Kaminsky S, Berrih-Aknin S: Expression of acetylcholine receptor genes in human thymic epithelial cells: Implications for myasthenia gravis. J Immunol 1996;157:3752– 3760. 17 Kasai M, Hirokawa K: A novel cofactor produced by a thymic epithelial cell line, promotion of immature thymic lymphocytes by the presence of interleukin and various mitogens. Cell Immunol 1991;132:377–390. 18 Mizuochi T, Kasai M, Kokuho T, Kakiuchi T, Hirokawa K: Medullary but not cortical epithe-

19

20

21

22

23

24

25

26

27

lial cells present soluble antigens to helper T cells. J Exp Med 1992;175:1601–1605. Hirokawa K, Utsuyama M, Moriizumi E, Handa S: Analysis of the thymic microenvironment by monoclonal antibodies with specific reference to the nurse cells. Thymus 1986;8:349– 360. Eren R, Zharhary D, Abel L, Globerson A: Agerelated changes in the capacity of bone marrow cells to differentiate in thymic organ cultures. Cell Immunol 1988;112:449–455. Compton MM, Haskill JS, Cidlowski JA: Analysis of glucocorticoid actions on rat thymocyte deoxyribonucleic acid by fluorescence-activated flow cytometry. Endocrinology 1988; 122:2158–2164. Fonnum FA: Rapid radiochemical method for the determination of choline acetyltransferase. J Neurochem 1975;24:407–409. Kawashima K, Ishikawa H, Mochizuki M: Radioimmunoassay for acetylcholine in the rat brain. J Pharmacol Methods 1980;3:115–123. Lindenboim L, Pinkas-Kramarski R, Sokolovsky M, Stein R: Activation of muscarinic receptors inhibits apoptosis in PC12M1 cells. J Neurochem 1995;64:2491–2499. Tria MA, Vantini G, Fiori MG, Rossi A: Choline acetyltransferase activity in murine thymus. J Neurosci Res 1992;31:380–386. Rinner I, Kawashima K, Schauenstein K: Rat lymphocytes produce and secrete acetylcholine in dependence of differentiation and activation. J Neuroimmunol 1998;81:31–37. Rossi A, Tria MA, Baschieri S, Doria G, Frasca D: Cholinergic agonists selectively induce proliferative response in the mature subpopulations of murine thymocytes. J Neurosci Res 1989;24:369–372.

Neuroimmunomodulation 1999;6:51–55

55

Neuroimmunomodulation 1999;6:56–68

The Immune-Endocrine Loop during Aging: Role of Growth Hormone and Insulin-Like Growth Factor-I William Burgess a Qiang Liu a Jian-Hua Zhou a Qingsong Tang a Akihito Ozawa a Roger VanHoy a Sean Arkins b Robert Dantzer c Keith W. Kelley a a Laboratory

of Immunophysiology, Department of Animal Sciences, University of Illinois, Urbana, Ill., and of Biology, Illinois State University, Normal, Ill., USA; c Neurobiologie Integrative, INRA-INSERM U394, Bordeaux, France

b Department

Key Words Apoptosis W Phosphatidylinositol 3)-kinase W IGF-I receptor W IL-4 receptor W Insulin receptor substrate-1 W Bcl-2 W Hematopoiesis

Abstract Why a primary lymphoid organ such as the thymus involutes during aging remains a fundamental question in immunology. Aging is associated with a decrease in plasma growth hormone (somatotropin) and IGF-I, and this somatopause of aging suggests a connection between the neuroendocrine and immune systems. Several investigators have demonstrated that treatment with either growth hormone or IGF-I restores architecture of the involuted thymus gland by reversing the loss of immature cortical thymocytes and preventing the decline in thymulin synthesis that occurs in old or GH-deficient animals and humans. The proliferation, differentiation and functions of other components of the immune system, including T and B cells, macrophages and neutrophils, also demonstrate age-associated decrements that can be restored by IGF-I. Knowledge of the mechanism by which cytokines and hormones influence hematopoietic cells is critical to improving the health of aged

ABC Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

© 1999 S. Karger AG, Basel

Accessible online at: http://BioMedNet.com/karger

individuals. Our laboratory has recently demonstrated that IGF-I prevents apoptosis in promyeloid cells, which subsequently permits these cells to differentiate into neutrophils. We also demonstrated that IL-4 acts much like IGF-I to promote survival of promyeloid cells and to activate the enzyme phosphatidylinositol 3)-kinase (PI 3kinase). However, the receptors for IGF-I and IL-4 are completely different, with the intracellular ß chains of the IGF receptor possessing intrinsic tyrosine kinase activity and the · and Ác subunit of the heterodimeric IL-4 receptor utilizing the Janus kinase family of nonreceptor protein kinases to tyrosine phosphorylate downstream targets. Both receptors share many of the components of the PI 3-kinase signal transduction pathway, converging at the level of insulin receptor substrate-1 or insulin receptor subtrate-2 (formally known as 4PS, or IL-4 Phosphorylated Substrate). Our investigations with IGF-I and IL-4 suggest that PI 3-kinase inhibits apoptosis by maintaining high levels of the anti-apoptotic protein Bcl2. The sharing of common activation molecules, despite vastly different protein structures of their receptors, forms a molecular explanation for the possibility of cross talk between IL-4 and IGF-I in regulating many of the events associated with hematopoietic differentiation, proliferation and survival.

Keith W. Kelley Laboratory of Immunophysiology, University of Illinois 207 Edward R. Madigan Laboratory 1201 West Gregory Drive, Urbana, IL 61801 (USA) Tel. +1 217 333 5141, Fax +1 217 244 5617, E-Mail [email protected]

The classic studies that first showed that growth hormone is thymotropic were published over 30 years ago [5]. These early experiments, published by Walter Pierpaoli, Nicola Fabris, and Ernst Sorkin, showed that mice injected with antiserum to growth hormone develop thymic atrophy and body wasting and that growth hormone-deficient mice have proportionately smaller thymus glands and spleens. When adequate amounts of relatively purified growth hormone became available, it could readily be established that growth hormone is clearly thymotropic [6]. Istvan Berczi’s [7, 8] pioneering studies subsequently confirmed that the thymus gland and spleen are involuted in hypophysectomized rats and that certain aspects of the both humoral and cellular immune responses are defective in these animals. All of these immunological changes

are reversed by injection of growth hormone, and the increase in proliferation of thymocytes and splenocytes caused by growth hormone [9] is accompanied by an increase in the expression of the proto-oncogene c-myc [10]. Professor Berczi [11] was also one of the first investigators to recognize that growth hormone increases the number of bone marrow cells. We established that the pituitary gland is required for adequate protection against lethal infections caused by Salmonella typhimurium, and that injections of either pituitary-derived or recombinant growth hormone significantly protect these rats from death caused by this bacterium [12–14]. Thymulin is a zinc-bound nonapeptide that is secreted by thymic epithelial cells. A number of experiments published by Mireille Dardenne in France, Wilson Savino in Brazil and Nicola Fabris in Italy have clearly established, both in vivo and in vitro in human and rodent studies, that growth hormone stimulates the synthesis of thymulin [15–17]. This may occur by growth hormone increasing the production of IGF-I, which also increases the amount of thymulin secretion, or by growth hormone or IGF-I increasing attachment of thymocytes to thymic epithelial cells by augmenting expression of extracellular matrix ligands [18]. These results are consistent with the findings that growth hormone-deficient children and adults have low plasma levels of thymulin, whereas acromegalics have increased plasma thymulin. The secretion of thymulin declines significantly with age, which is correlated with the reduction in synthesis of both growth hormone and IGF-I. Injections of growth hormone into aged mice or humans, particularly in the presence of adequate zinc, increases synthesis of thymulin [19]. It is likely that there is a feedback system between thymulin derived from the thymus and growth hormone synthesized from either the pituitary gland or possibly leukocytes [20–22], including subpopulations of human [23] and rat [24] thymocytes. These data have been very well reviewed recently by Weigent and Blalock [25, 26]. Most of the early information on the positive relationship between the thymus and pituitary glands emphasized growth hormone. However, following the elucidation of the amino acid structure of the somatomedins and their new assignment to the insulin-like growth factor nomenclature in 1978 [27], investigators began to determine the direct effects of IGF-I, rather than growth hormone, on leukocytes. The concept that IGF-I is also thymotropic was initially supported by several investigations demonstrating the effects of IGF-I on primary and secondary lymphoid organs. For example, IGF-I treatment reverses atrophy of the thymus and increases the capability of thy-

Immune-Endocrine Axis during Aging

Neuroimmunomodulation 1999;6:56–68

Introduction

The thymus gland, particularly the cortex, is well known to atrophy during aging, but the functional and molecular explanation for involution of the thymus gland remains an enigma. The emerging view is that the answer to this question lies in understanding how thymic development is regulated by the neuroendocrine system. This conclusion is based upon the findings that plasma levels of several hormones, including growth hormone (somatotropin) and insulin-like growth factor-I (IGF-I), decline significantly in aging rodents and man [1, 2]. This ‘somatopause’ has been considered similar in principle to the menopause and andropause of aging [3]. Although estrogen/progestin replacement therapy for menopasual women is widely used, the safety of similar replacement strategies for increasing plasma levels of growth hormone or IGF-I to those levels that occur during mid-adult life are not yet satisfactorily understood. However, it is well accepted that systemic injections of either of these hormones increase the size and cellularity of both primary and secondary lymphoid organs in rodents [4]. Significant clinical advances in improving the dysregulated immune system of the aged, involving cells of both the lymphoid and myeloid lineages, requires that the molecular basis for reciprocal communication systems between the immune and central nervous systems be understood. These investigations are also likely to yield important benefits in understanding neurodegenerative diseases of the aged, such as Alzheimer’s disease.

Growth Hormone: A Link between the Thymus and Anterior Pituitary Gland

57

mocytes to proliferate in diabetic rats [28]. The increased growth of the thymus and spleen in response to IGF-I infusion was attributed to increased cellular proliferation in hypophysectomized rats [29]. Similar increases in the size of the thymus and spleen were observed in normal adult mice treated with IGF-I [30]. These data suggest that the thymotropic effects of growth hormone can be mediated by IGF-I, especially since thymic epithelial cells have been shown to secrete IGF-I in response to growth hormone stimulation [31]. Recent information on the role of growth hormone and IGF-I on specific aspects of immunity have been reviewed in detail by Kooijman et al. [32]. We [33–35] and others [36] have also recently reviewed this topic. The immunological effects of the closely related molecule to growth hormone, prolactin, with particular emphasis on signaling via the prolactin receptor in T lymphocytes, have also been recently reviewed [37].

Deficient Myeloid Cell Activation in Aged Subjects: Reversal by Growth Hormone

Substantial evidence has accumulated showing that T cells are affected during aging. The result is a reduction in the proliferation of T cells, a reduction in the synthesis of IL-2 and a change in the types of circulating peripheral T cells in the direction of more memory and fewer naı¨ve T lymphocytes [38, 39]. More recent studies using highly purified, sorted populations of CD4+ or CD8+ T lymphocytes have failed to detect a reduction in IL-2 production by naı¨ve T cells (CD44low) or an increase in IL-4 or interferon-Á synthesis by memory cells (CD44high) in aged mice [40]. However, an age-associated reduction in IL-2 synthesis and increase in interferon-Á production could be readily detected in unseparated populations of splenic T cells. Similarly, we have reported that it is easier to detect age-associated reductions in proliferative responses to Tcell mitogens in splenocytes than in enriched populations of T cells [41, 42]. Although several explanations are possible [40], one hypothesis is that macrophages or their products that are present in unsorted populations act to suppress T-cell proliferation from aged rats [42]. If this were to occur, growth hormone or IGF-I could act indirectly on T cells from aged subjects by affecting the activity of macrophages. We have shown that important targets for the action of growth hormone are both neutrophils and macrophages, and this may be critical in aged subjects. Indeed, the ability of neutrophils from aged rats to kill serum-opsonized

58

Neuroimmunomodulation 1999;6:56–68

Escherichia coli is reduced by 50% compared to the killing ability of neutrophils from young animals, and this defect can be significantly reversed in vivo by growth hormone [43]. This is probably caused by an inability of neutrophils and macrophages [44, 45] from aged rats to generate optimal amounts of superoxide anion in response to low doses of interferon-Á, a defect that is reversed by treatment with growth hormone [43, 44]. This newly defined synergism between growth hormone and a classic macrophage-activating factor, interferon-Á, can be detected in neutrophils from aged but not young subjects. This synergism is likely to be important clinically in aged humans in light of new data that show dysregulation of myeloid cells in growth hormone-deficient and -overexpressing (acromegalics) humans. Ten years ago we established that growth hormone primes macrophages from animals for enhanced free radical secretion [46]. These results have been extended to human monocytes and neutrophils [47, 48] and confirmed by other laboratories in species ranging from fish to cows [33, 49]. Two separate reports have established that phagocytosis of E. coli by peripheral blood leukocytes [50] and killing of Mycobacterium avium by human monocytes is significantly enhanced in acromegalic patients [51]. Conversely, there is a substantial reduction in the capability of both monocytes and neutrophils from children who are clinically deficient in growth hormone to phagocytize E. coli [52]. After treatment of these growth hormone-deficient children for 6 months with recombinant human growth hormone, the phagocytic activity of both neutrophils and monocytes returned to the level of control children, and this was accompanied by the expected increase in height velocity of the deficient children. As we reported earlier [53, 54], IGF-I also increases the oxidative burst and phagocytic ability of neutrophils from clinically normal adults [55]. These results have led us to investigate the role of growth hormone and IGF-I in the development of cells of the myeloid as well as the lymphoid lineages. It is not clear from the in vivo clinical studies if the increase in phagocytic activity is the result of either growth hormone or IGF-I acting directly on myeloid cells or if it indirectly results from growth hormone increasing the synthesis of interferon-Á by T lymphocytes. For example, pretreatment of rats with growth hormone, but not IGF-I, increases plasma interferon-Á following injection of endotoxin [56]. When human peripheral blood mononuclear cells [57] or murine splenocytes [58] are treated with growth hormone in vitro, there is also an increase in interferon-Á following activation with either allogeneic mononuclear cells or S. typhimurium, respectively.

Burgess/Liu/Zhou/Tang/Ozawa/VanHoy/ Arkins/Dantzer/Kelley

In a healthy adult, approximately 2 ! 1011 red blood cells, 1011 white blood cells and 1–2 ! 1011 platelets are produced every day to replace the loss of aging cells, which are cleared by phagocytes. The body needs an efficient system to support the replenishment of this many blood cells, and this process begins with the hematopoietic stem cell (HSC) [59]. For a long time there was a question whether different types of blood cells developed from different ancestors or were derived from common initiating cells. The discovery of colony-forming units in the spleen by Till and McCulloch [60] ended this argument. These workers found that bone marrow or spleen cells injected into the circulatory system form cell colonies in the spleen of lethally irradiated mice. Each of these colonies is derived from a single cell and is composed of erythroblasts, granulocytes, megakaryocyte and B lymphocytes. The capacity to develop into different lineages is called multipotentiality and is the basic feature of HSCs. They do not differentiate directly into terminal blood cells, but instead pass through two major types of progenitor cells: lymphoid and myeloid progenitors. Lymphoid progenitors give rise ultimately to T cells and B cells (including natural killer cells). Myeloid progenitor cells can form at least six different types of mature cells, including erythrocytes, megakaryocytes, monocytes and the polymorphonuclear granulated cells (neutrophils, eosinophils and basophils (including mast cells)). Another feature of HSCs is their capacity for self-renewal. HSCs make up only 0.05% of the bone marrow cell population and continually give rise to all lineages of blood cells. Human HSCs are cells that express a particular phenotype that is defined by a panel of antibodies against various leukocyte antigens (e.g., Thy-loLin – CD34+). Human HSCs differentiate into myelomonocytic and B-lymphoid lineages in long term in vitro cultures on bone marrow stromal cells and into T lymphoid cells when transplanted into SCID mice [61]. Recently, Osawa et al. [62] used a monoclonal antibody to mouse CD34 to purify murine HSCs. Unlike in humans, murine HSCs are found in CD34lo/– populations. The lymphohematopoietic systems of 21% of lethally irradiated mice were reconstituted by a single CD34lo/–c-kit+Sca-1+Lin – murine HSC. Understanding the process of differentiation of HSCs has progressed rapidly with the discovery of various hematopoietic cytokines and the use of colony-forming assays. Colonies of all hematopoietic cell types, with the exception of both T and B lymphocytes, can form from progenitors isolated from bone marrow and peripheral

blood by the addition of only cytokines. These progenitors are named according to the lineages into which they develop, such as colony-forming units erythroid (CFU-E), colony-forming units granulocyte/macrophage (CFU-GM), and colony-forming units megakaryocyte (CFU/Meg). Similar terminology is used for progenitors of basophils and eosinophils that are derived from a common myeloid pluripotent stem cell. Some cytokines induce the commitment of a designated progenitor, such as erythropoietin (e.g., CFU-E) and colony-stimulating factors (e.g., CFUM). The others, like stem cell factor (SCF), IL-3 and IL-6, have multilineage effects and synergize with lineage-specific cytokines. The progress made in understanding hematopoietic progenitors and cytokines has already made it possible to expand various types of blood cells in vitro from HSCs or progenitors isolated from patients [63, 64]. Compared with myeloid cells, the study of lymphopoiesis is more difficult because of the problems associated with in vitro cultivation. It was greatly advanced by the development of long-term cultures of B lymphocytes and their precursors on feeder layers of adherent bone marrow cells [65]. The question of whether T and B lymphocytes develop from common progenitors or from unipotential progenitors was not completely answered until the identification of a clonogenic common lymphoid progenitor (CLP) [66]. An IL-7 receptor-positive (IL-7R+) population expressing the phenotype Lin – IL-7R+Thy-1.1 – Sca-lockitlo isolated from bone marrow fractions possesses the capacity of lymphoid-restricted (T, B and natural killer cells) reconstitution. This population of cells appears to represent the true common lymphoid progenitor cells that develops from HSC and is formed prior to the development of T- and B-lineage precursors. IL-7 is essential for the survival of T-cell precursors and promotes B-cell proliferation and rearrangement of immunoglobulin heavy chain genes in B-cell precursors [67]. The thymus provides a microenvironment producing a large variety of cytokines, which play an important role in T-cell differentiation. In addition to IL-3, IL-6 and IL-7, SCF and Flk-2/ Flt-3 ligand are two critical cytokines that have different effects on CD4lo T-cell progenitors and pro-T cells which are in the next stage of development. Flt-3 ligand appears to have renewal and/or proliferation/expansion effects on CD4lo cells, but shows no significant effect on pro-T cells. In contrast, SCF accelerates the differentiation of CD4lo cells and the repopulation of pro-T cells [68]. The CD4lo cells may include a small fraction of CLPs that have immigrated to the thymus. CD4lo cells are also capable of forming B cells, natural killer cells and dendritic cells, as well as T cells, when injected into irradiated recipients [69].

Immune-Endocrine Axis during Aging

Neuroimmunomodulation 1999;6:56–68

Overview of Hematopoiesis

59

IGF-I in Hematopoiesis

There is substantial evidence that IGF-I is synthesized in hematopoietic tissues such as the thymus [32] yolk sac endoderm and mesoderm, fetal liver [70], bone marrow stromal cells [70–72] and differentiating macrophages [73, 74]. These reports establish that IGF-I is found in areas of hematopoiesis and can stimulate hematopoietic cell proliferation and differentiation. One report has concluded that normal human CD34+ cells do not express the IGF-I receptor [75], although the functional studies in this report used 10% bovine calf serum that contains substantial amounts of active IGF-I. We also have found that human KG-1 cells do not express detectable binding to an anti-IGF-I receptor antibody [76], and these cells express substantial amounts of CD34. Interestingly, normal human CD34 cells have been reported to express abundant amounts of receptors for other hormones, such as prolactin [77, 78]. Furthermore, some of the earliest detectable promyeloid (e.g., human HL-60 and murine FDCP-1 cells) and pre-T lymphocytes (PXTL) express easily detectable receptors for IGF-I (see below). If indeed HSCs do not express detectable amounts of the IGF-I receptor, we interpret these results to suggest that CSFs may induce expression of the IGF-I receptor as the HSCs mature and reduce surface expression of CD34. T-Cell Development Although growth hormone increases the number of CD3 thymocytes when injected into aged mice [79] and the number of CD4+CD8+ double positive thymocytes when injected into dwarf mice [80, 81], it is unknown if these effects are mediated by IGF-I. The bulk of the literature indicates that IGF-I appears to increase the number of thymocyte progenitors that enter the thymus, as well as the proliferation and differentiation of T cells once they enter the thymus. For example, IGF-I increases thymocyte binding of peanut agglutinin (PNA) in vitro [30]. Since PNA binding is a marker for immature thymocytes, this finding suggests a role for IGF-I as a chemoattractant in the recruitment of precursor cells to the thymus. Growth hormone may act similarly on bone marrow stromal cells. When depleted fetal thymus lobes are colonized with bone marrow cells and then subsequently stimulated with growth hormone only during the initial phase of thymic colonization, there is an increase in the number of T cells derived from the donor cells [79] suggesting that growth hormone affects thymocyte progenitors homing to the thymus. Primary X-ray-induced thymic lymphoma (PXTL) cells have a phenotype very similar to pre-T cells

60

Neuroimmunomodulation 1999;6:56–68

found in the thymus. Receptors for IGF-I are easily detected on these cells and IGF-I increases proliferation of these cells in vitro [82]. Several studies have demonstrated increased proliferation of thymocytes by IGF-I in vitro [23, 83]. When injected in vivo, IGF- I increases the proportion of CD4+CD8+ double positive cells [28]. This is consistent with the expression of IGF-I receptors on double positive thymocytes as well as on their precursor CD4–CD8– double negative progenitor cells [84]. IGF-I is therefore likely to play a role in the maturation of double negative to double positive thymocytes. B-Cell Development IGF-I has been found to exert its effects at various stages of B-cell development. IGF-I derived from S17 bone marrow stromal cells increases the expression of cytoplasmic immunoglobulin Ì heavy chain, indicating an induced shift from a pro-B to a pre-B cell phenotype [85]. IGF-I also enhances IL-7 and S10 stromal cell proliferation of pro-B cells from day 14 fetal mouse liver [86]. Similarly, IGF-I increases spleen size and splenocyte populations in mice [87] and growth hormone-deficient Snell dwarf mice [88]. Irradiation followed by bone marrow transplantation and IGF-I treatment increases the proliferation of all lineages of B cells in mice [89]. IGF-I increases the number of mature IgG-bearing cells and the ability of these cells to produce antibody in response to antigenic stimulation [87]. IGF-I knockout mice have less absolute pre-B and B cells than their normal littermate controls, suggesting a role for IGF-I on B-cell lymphopoiesis [90]. Clearly, IGF-I is able to promote B-cell proliferation and differentiation in young mice. In aged mice there are fewer pro-B cells available, and the ability of these pro-B cells to respond to the combination of IGF-I and IL-7 by differentiating into pre-B cells declines 10-fold with age [91]. This could be due to a defect in responsiveness of either the IL-7 receptor or the IGF-I receptor. In contrast, both IGF-I and stem cell factor are able to synergize with IL-7 to enhance the proliferation of pro-B cells, and this synergy remains and is even enhanced in pro-B cells from aged mice [67]. In contrast to the findings by Merchant et al. [91] with the differentiation of pre-B cells, these data would suggest that there is no reduction in the responsiveness of the IGF-I receptor to potentiate IL-7induced proliferation in pro-B cells from aged mice. However, pro-B cells from aged mice have a diminished ability to proliferate in response to IL-7, even though there is no reduction in expression of either the common Á chain or specific · chain of the IL-7 receptor on pro-B cells from aged mice [67]. This defect in signaling through the IL-7

Burgess/Liu/Zhou/Tang/Ozawa/VanHoy/ Arkins/Dantzer/Kelley

receptor is likely to be important in T cells as well. Indeed, it has been recently demonstrated that the thymic atrophy of aging is associated with an inability to rearrange the ß chain of the T-cell receptor [92], and it was speculated that this defect was caused by inadequate activity of IL-7. Myelopoiesis The possibility that IGF-I can increase the development of myeloid cells has also been investigated. Both IGF-I and growth hormone increase the number of colonies formed by human bone marrow myeloid progenitors by 41 and 38%, respectively. An anti-IGF-I receptor antibody (·IR3) inhibits the increases produced by both growth hormone and IGF-I, suggesting the effects of growth hormone are mediated by IGF-I [93]. These workers found that an adherent cell in the marrow is the source of IGF-I, and this cell is likely to be a monocyte [93]. Consistent with this possibility, we have found that macrophages are the major type of hematopoietic cells that express IGF-I mRNA transcripts and protein [73, 74]. Furthermore, colony-stimulating factor-1 (CSF-1) stimulates expression of IGF-I mRNA by 50- to 75-fold in freshly isolated bone marrow cells. The CSF-1-induced proliferation of myeloid precursors is inhibited by IGF-I binding protein-3, suggesting a role for IGF-I as a differentiation and proliferation agent in myeloid precursor cells [94]. Since CSF-1 stimulates the development of monocytes/macrophages, this finding is in accord with the idea that macrophages are the primary source of IGF-I derived from hematopoietic cells. Even in the human thymus gland, macrophages appear to be the primary source of IGF-I, whereas thymic epithelial cells synthesize IGFII [95]. Promyeloid cells express receptors for IGF-I and proliferate in response to this peptide. For example, both human (HL-60 cells [76]) and murine (FDCP [96]) promyeloid cells and some myeloma cell lines [97] express the IGF-I receptor. All these cells respond to IGF-I by increasing their rate of proliferation. Human HL-60 promyeloid cells also proliferate in response to IGF-II and the IGF-I analog desIGF-I. Interestingly an anti-IGF-I receptor antibody inhibits DNA synthesis by HL-60 cells in response to not only IGF-I and desIGF-I, but also by IGF-II. Since IGF-II also has been shown to affect a number of immune events in both the thymus and spleen [98– 100], it could be that these effects of IGF-II are mediated by the IGF-I receptor. Finally, IGF binding protein-3 (IGFBP-3) also inhibits IGF-I and IGF-II, but not desIGF-I, induced proliferation of HL-60 cells [76]. This is expected since desIGF-I binds poorly to IGFBP-3. How-

Immune-Endocrine Axis during Aging

ever, we could not detect the synthesis of any IGF binding protein by human promyeloid cells, even though both human (U-937) and murine (TMC) mature macrophages synthesize abundant IGFBP-4 [101]. These investigations demonstrate that IGF-I stimulates proliferation of promyeloid cells, and they suggest that IGF binding proteins from fully developed macrophages can inhibit this process. Erythropoiesis IGF-I stimulates proliferation of early and late stage erythroid progenitor cells. IGF-I treatment increases in vitro colony formation in both erythroid blast-forming units (BFU-E) and the more mature colony-forming units (CFU-E) by 62 and 121%, respectively [102]. The increased proliferation observed in CFU-E and BFU-E cells by IGF-I stimulation is inhibited by an anti-IGF-I receptor antibody, suggesting a direct role for IGF-I on erythropoiesis [103]. Increased heme synthesis is found in CFUE treated with IGF-I [104]. Clinical studies demonstrate the ability of IGF-I and growth hormone to stimulate erythropoiesis by increasing proliferation and differentiation of erythroid progenitor cells in adults [105] and hemoglobin levels in children [106] with growth hormone deficiencies. IGF-I is clearly a potent stimulator of erythropoiesis.

IGF-I Inhibits Apoptosis in Hematopoietic Cells

In addition to its role as a stimulator of proliferation in hematopoietic cells, IGF-I promotes hematopoietic development by preventing apoptosis. Apoptosis is a form of programmed cell death that is characterized by the cleavage of cellular DNA into fragments of approximately 200 base pairs. Apoptosis plays a critical role in embryonic development, tumor suppression and negative selection of hematopoietic cells [107]. Hematopoietic cells deprived of CSFs will die by an apoptotic mechanism as demonstrated by the DNA fragmentation that occurs when IL-3 is removed from FDCP-mix and FDCP-1 myeloid progenitor cells [108]. IGF-I has been found to prevent apoptosis induced by CSF withdrawal in IL-3dependent Baf-3 pro-B cells and bone marrow-derived FDCP-mix cells [109]. Our laboratory has investigated the mechanism responsible for the IGF-I-induced prevention of apoptosis [96]. IGF-I inhibits apoptosis in cytokine-deprived IL-3dependent FDCP-1 promyeloid cells. IL-3 and IGF-I stimulate similar increases in phosphatidylinositol 3)-

Neuroimmunomodulation 1999;6:56–68

61

kinase (PI 3-kinase) activity in these cells. PI 3-kinase, as its name implies, is a lipid/serine kinase that produces phosphorylated lipids in cell membranes that can act as second messengers in signal transduction. Wortmannin, a PI 3-kinase inhibitor, prevents PI 3-kinase activation by both IGF-I and IL-3. However, wortmannin inhibits only the IGF-I, but not the IL-3, induced cell survival. We interpret these results to indicate that there must be at least two pathways to promote cell survival: one signaling pathway in which PI 3-kinase is required and another pathway that uses an alternative signal transduction cascade. Several other investigations support a PI 3-kinasedependent cell survival pathway, including PC-12 rat pheochromocytoma cells [110], cerebellar neurons [111] and fibroblasts [112]. It is therefore likely that progress made in understanding the mechanisms of action of IGFI on the survival of hematopoietic progenitor cells will contribute significantly to delineating the mechanism by which IGF-I also promotes the survival of neurons. This could have important implications in aging research. The downstream molecule in the PI 3-kinase signal transduction pathway that is responsible for promoting cell survival remains to be elucidated. One candidate is the cellular oncogene Bcl-2 that has been shown to inhibit apoptosis in both neuronal and hematopoietic systems. One investigation showed that transfection of Bcl-2 into IL-3-dependent FDCP-mix cells prevents apoptosis in the absence of cytokine [113]. IGF-II activation of the IGF-I receptor also stimulates PI 3-kinase, maintains Bcl-2 levels and prevents apoptosis in neuroblastoma cells [114]. The Bcl-2 (B Cell Lymphoma) family of proteins contains members that both promote (Bax) and inhibit (Bcl-2, BclXL) apoptosis. Bcl-2 family members form both homoand heterodimers with one another. An increase in Bax homodimers stimulates apoptosis. It is believed that Bcl-2 binds to Bax and forms a heterodimer that is less capable of initiating apoptosis [107]. The relationship between Bcl-2, Bax and apoptosis was investigated by our laboratory. The withdrawal of IL-3 produces a time-dependent decrease in Bcl-2 expression in conjunction with an increase in apoptosis, while Bax levels remain the same [115]. IGF-I, IL-4 and IL-3 maintain not only Bcl-2 expression but also the amount of Bax that co-precipitates with Bcl-2. We believe that IGF-I and IL-4 prevent apoptosis by a signal transduction pathway that converges prior to the activation of PI 3-kinase. This convergence point is likely to be at the level of the insulin receptor substrate docking proteins (see below). This idea is supported by the finding that IL-4 maintains Bcl-2 levels in T cells [116].

62

Neuroimmunomodulation 1999;6:56–68

IGF-I Promotes Hematopoietic Cell Differentiation

Prevention of apoptosis in hematopoietic progenitor cells has been suggested to allow these cells to progress along their intrinsic preprogrammed maturation process [113]. IGF-I not only maintains the survival of murine bone marrow progenitors [96, 115], but it may also permit the completion of the differentiation process in hematopoietic precursor cells by inhibiting apoptotic cell death. We tested this possibility in human progenitor hematopoietic cells induced to differentiate with retinoic acid [117]. Human HL-60 promyelocytes undergo granulocytic differentiation upon treatment with retinoic acid, as assessed by expression of the mature leukocyte surface antigen CD11b. Almost all the publications that have investigated the differentiation of HL-60 cells into granulocytes utilized a culture medium that contained at least 5% fetal bovine serum. In the absence of fetal bovine serum, granulocytic differentiation is substantially impaired. We demonstrated that these serum-deprived progenitor cells undergo apoptosis, as measured by flow cytometric analysis, following induction of differentiation with retinoic acid. However, addition of IGF-I rescues the cells from apoptotic cell death and therefore allows completion of granulocytic differentiation initiated by retinoic acid. This IGF-I-enhanced granulocytic differentiation is specifically abrogated by treatment with a monoclonal antibody directed against the IGF type I receptor, indicating that IGF-I signals specifically through this tyrosine kinase receptor [117]. In accordance with other reports in PC-12 pheochromocytoma cells [110] and neurons [111, 118], IGF-Iinduced activation of PI-3 kinase appears to be a key upstream mechanism for preventing apoptosis in these human myeloid precursor cells. Abrogation of IGF-Iinduced PI 3-kinase by its irreversible inhibitor, wortmannin, results in re-appearance of an apoptotic cell population and impairment of cell differentiation. These findings strongly suggest that a likely role for IGF-I is to maintain the survival of hematopoietic progenitor cells, which subsequently permits these promyeloid cells to complete the differentiation program that is induced by retinoic acid. Promyeloid cells also have the capability to differentiate into macrophages when induced by other stimuli, such as vitamin D3. In contrast to granulocytic differentiation induced by retinoic acid, we found that vitamin D3 does not cause the apoptotic death of HL-60 cells when cultured in serum-free medium [119]. This finding is con-

Burgess/Liu/Zhou/Tang/Ozawa/VanHoy/ Arkins/Dantzer/Kelley

sistent with other reports where IGF-I has been shown to directly promote the differentiation process in proadipocytes [120], myocytes [121] and nerve cells [122], independent of preventing cells from apoptotic cell death. We therefore explored the possibility that IGF-I augments myeloid cell maturation towards the macrophage/monocytic lineage. By using a defined serum-free system, we demonstrated that IGF-I increases expression of the myeloid differentiation marker, CD11b, as promyeloid progenitors develop into macrophages in the presence of vitamin D3 [119]. In the absence of IGF-I, vitamin D3 induces macrophage differentiation in approximately 25% of the promyelocyte population, whereas the proportion of vitamin D3-induced differentiated cells increases to roughly 75% in the presence of 100 ng/ml IGF-I. Since we cannot detect apoptotic cells induced by vitamin D3, this demonstrates that IGF-I acts to promote rather than to permit the differentiation process into macrophages. In addition, both a major serum IGF binding protein (IGFBP-3) and a monoclonal antibody against the IGF type I receptor profoundly inhibit vitamin D3-induced CD11b expression in serum-containing medium, further suggesting that IGF-I is a critical peptide in serum that promotes macrophage maturation. Protein kinase C-˙ (PKC-˙) is an atypical protein kinase C that does not depend upon increases in intracellular Ca2+ for activation. Overexpression of PKC-˙ in human promonocytic cells stimulates the phenotypic expression of monocytic maturation markers [123]. However, the mechanism by which this serine/threonine kinase activates macrophage differentiation in vitamin D3-treated promyeloid cells has only recently been studied [119]. The promotion of macrophage differentiation by IGF-I occurs concomitantly with activation of PKC-˙. More importantly, suppression of PI 3-kinase by two pharmacologic inhibitors, wortmannin and LY294002, abrogates IGF-I-induced activation of PKC-˙ and subsequently impairs macrophage differentiation. These data strongly suggest that IGF-I promotes macrophage differentiation induced by vitamin D3 by activating PI 3-kinase and its putative intracellular downstream target enzyme, PKC-˙. IGF-I has been classically viewed as a cell progression factor, acting late and the G1 exit point of the cell cycle. Our recent data establish that IGF-I also plays a key role in myeloid cell differentiation towards both the macrophage and granulocytic lineages. IGF-I indirectly enhances granulocytic cell differentiation by preventing apoptotic cell death. In contrast, IGF-I can directly act on the macrophage differentiation machinery via a mechanism that is dependent upon activation of PI 3-kinase and PKC-˙.

Although our experiments established that IGF-I, IL-4 and IL-3 each activate PI 3-kinase and increase expression of Bcl-2 in promyeloid progenitor cells, the mechanism by which these diverse receptors promote an increase in Bcl-2 expression is unknown. It is unlikely that the increase in Bcl-2 caused by IL-3 is mediated by PI 3-kinase, because inhibition of the activity of this enzyme does not affect survival of myeloid progenitors [96]. However, since the survival-promoting effects of IGF-I are mediated by PI 3-kinase, and both IGF-I and IL-4 lead to the tyrosine phosphorylation of common intracellular proteins known as insulin receptor substrates-1 and -2 (IRS-1 and IRS-2/4PS), we speculate that PI 3-kinase is critically involved in cell survival promoted by both IGFI and IL-4. The question is how these totally different receptors can increase the expression of a survival protein such as Bcl-2. The structure and mode of activation of the IGF-I and IL-4 receptors are very different (fig. 1). The IGF-IR consists of two · and ß chain dimers linked together by disulfide bonds. The specificity of the extracellular · chain determines ligand binding whereas the cytoplasmic ß chain contains intrinsic tyrosine kinase activity [124, 125]. Ligand binding induces autophosphorylation of three critical tyrosines (PY) on the ß chain as well as phosphorylation of other proteins containing tyrosine phosphorylation sites [124]. The IL-4 receptor (IL-4R) is a member of the hematopoietin superfamily of receptors. Although it is also a heterodimer, it consists of a 140-kD · subunit (IL-4R·) and a common Á (Ác) chain. In contrast to the IGF-I receptor, the IL-4R has no intrinsic tyrosine kinase activity. Although the IL-4 receptor complex does not contain any consensus sequences that encode either serine/threonine or tyrosine kinases, binding of ligand does activate the nonreceptor protein tyrosine kinases Jak-1 and Jak-3 of the Janus kinase family. Co-immunoprecipitation studies have shown that IL-4R· associates with Jak-1, which leads to tyrosine phosphorylation of the · chain, and Ác associates with Jak-3 [126]. Jak family members are now recognized to be important in signaling the activation of Signal Transducers and Activators of Transcription (STAT). STAT 6 is tyrosine phosphorylated and translocated to the nucleus following engagement of the IL-4R· chain [127]. Activation of Jak-1 leads to the tyrosine phosphorylation of both STAT 6 and IRS-1 [128].

Immune-Endocrine Axis during Aging

Neuroimmunomodulation 1999;6:56–68

A Molecular Mechanism for Intracellular Cross Talk between IGF-I and IL-4 Receptors: A Postulate

63

Fig. 1. A proposed molecular mechanism for intracellular cross talk between IGF-I and IL-4 receptors. Ligand binding to the IGF-I receptor · chain leads to autophosphorylation of three critical tyrosine sites on the intracellular ß chain. The activated receptor phosphorylates tyrosine residues on either IRS-1 or IRS-2 (IRS-1/2), which serve as docking sites for SH2 domains on the p85-kD regulatory subunit of PI 3-kinase. Regulatory subunit binding localizes the p110-kD catalytic subunit to the proximity of the plasma membrane and activates this PI 3-kinase subunit, promoting the phosphorylation of the D3 position on the inositol ring of PI 4-phosphate and PI 4,5-bisphosphate. These activated lipid products act as second messengers that interact with downstream proteins, such as PKC-˙, p70 S6 kinase and Akt, to promote the anti-apoptotic effects of Bcl-2 at both the mitochondrial and nuclear membranes. The IGF-I receptor and IL-4 receptor signal transduction pathways meet at IRS-1/2, but the IL-4 receptor does not involve intrinsic tyrosine kinase activity of the IL-4 receptor. IL-4 binds to the · chain of the IL-4 receptor, leading to activation of nonreceptor protein tyrosine kinases of the Janus kinase family (Jak 1 and Jak 3). Following dimerization of the · chains after ligand stimulation and the subsequent recruitment of the Ác chains, Jak-1 and Jak-3 transactivate one another. Jak family members are now recognized to be important in signaling the activation of Signal Transducers and Activators of Transcription (STAT). STAT 6 (not shown) is tyrosine phosphorylated and translocated to the nucleus following engagement of the IL-4 receptor · chain. Of the two Jak kinases activated by IL-4, Jak-1 is required for IL-4 signaling because it causes the tyrosine phosphorylation of not only the · chain of the IL-4 receptor but also the tyrosine phosphorylation of IRS-1/2. The activation of PI 3-kinase and the formation of 3)-phosphorylated derivatives of phosphatidylinositol and the subsequent increase in Bcl-2 protein at the mitochondrial and nuclear membrane inhibit apoptosis and promote survival of hematopoietic progenitor cells.

One of the primary phosphorylation targets for the IGF-I and IL-4 receptors is IRS-1, a large 185-kD protein that contains up to 21 tyrosine phosphorylation sites that serve as docking sites for proteins that possess src homology 2 (SH2) and SH3 domains [129]. One of the proteins

64

Neuroimmunomodulation 1999;6:56–68

that binds phosphorylated IRS-1 is the lipid and serine kinase PI 3-kinase [129]. PI 3-kinase is the most widely studied signaling molecule that associates with IRS-1. PI 3-kinase consists of two domains, a p85-kD regulatory subunit which mediates IRS-1 binding via two SH2 sites, and a p110-kD catalytic subunit that catalyzes the phosphorylation of position 3 on the inositol ring of phosphatidylinositol 4-phosphate and phosphatidylinositol 4,5-bisphosphate (fig. 1). The mechanism by which these lipid products stimulate further signal transduction is currently being actively investigated by many laboratories. The receptors for IGF-I, IL-4 and insulin all share a common motif on their cytoplasmic domain. Using truncated mutants of the IL-4R, Keegan et al. [130] found that the region of the IL-4R that shares homology with both the insulin and IGF-I receptors is critical for IRS-1 activation and inhibition of survival. More importantly, it has been demonstrated that IL-4 and IGF-I stimulate the phosphorylation of another downstream docking molecule known 4PS (IL-4 Phosphorylated Substrate) in hematopoietic cells [131, 132]. Cloning of 4PS revealed it was very similar to IRS-1, which led to its designation as IRS-2 [133]. The results of these investigations indicate that the IGF-I and IL-4 signaling pathways converge at the levels of IRS-1 or IRS-2, which subsequently lead to the activation of PI 3-kinase and its recruitment to the membrane. Since PI 3-kinase activity is required for IGF-I and probably IL-4 to protect myeloid progenitor cells from apoptosis, we speculate that downstream products that are activated by PI 3-kinase are responsible for maintaining elevated expression of Bcl-2 in these cells. Therefore, regardless of whether the very different IGF-I or IL-4 receptors are stimulated in myeloid progenitor cells, the end result is that the large IRS docking proteins are tyrosine phosphorylated. This is accomplished by the intrinsic tyrosine kinase activity of the ß chains of the IGF-I receptor or by activation of Jak-1 following stimulation with IL-4. In either case, the p85 subunit of PI 3-kinase is recruited via its SH2 domains, which localizes the p110 subunit to the membrane where it can act upon the substrates phosphatidylinositol 4-phosphate and phosphatidylinositol 4,5bisphosphate to generate 3)-phosphorylated lipid second messengers. In this scenerio, the common intracellular mechanism by which both IGF-I and IL-4 promote survival occurs by activation of at least two common proteins, IRS-1 or IRS-2 and PI 3-kinase. It is likely that other common downstream effectors will be discovered that ultimately lead to heightened expression of Bcl-2 and inhibition of apoptosis. However, the information to date

Burgess/Liu/Zhou/Tang/Ozawa/VanHoy/ Arkins/Dantzer/Kelley

makes it clear that IGF-I and IL-4 share components in the PI 3-kinase signal transduction pathway. This sharing of activation molecules forms a molecular explanation for the possibility of cross talk between IL-4 and IGF-I in regulating many of the events associated with hematopoietic differentiation, proliferation and survival.

Acknowledgments Preparation of the manuscript was supported by grants to K.W.K. from the National Institutes of Health (AG-06246, DK49311 and MH-51569) and the Pioneering Research Project in Biotechnology financed by the Japanese Ministry of Agriculture, Forestry and Fisheries.

References 1 Corpas ES, Harman M, Blackman MR: Human growth hormone and human aging. Endocr Rev 1993;14:20–39. 2 De Boer H, Blok GJ, van der Veen EA: Clinical aspects of growth hormone deficiency in adults. Endocr Rev 1995;16:63–86. 3 Lamberts SWJ, van den Beld AW, van der Lely AJ: The endocrinology of aging. Science 1997; 278:419–424. 4 Kelley KW, Arkins S, Li YM: Growth hormone, prolactin and insulin-like growth factorI: New jobs for old players. Brain Behav Immun 1992;6:317–326. 5 Kelley KW: Growth hormone, lymphocytes and macrophages. Biochem Pharmacol 1989; 38:705–713. 6 Van Buul-Offers S, van den Brande J: The growth of different organs of normal and dwarfed Snell mice, before and during growth hormone therapy. Endocrinologica 1981;96: 46–58. 7 Berczi I: The role of the growth and lactogenic hormone family in immune function. Neuroimmunomodulation 1994;1:201–216. 8 Berczi I: Pituitary hormones and immune function. Acta Paediatr Suppl 1997;423:70–75. 9 Postel-Vinay MC, de Mello Coelho V, Gagnerault MC, Dardenne M: Growth hormone stimulates the proliferation of activated mouse T lymphocytes. Endocrinology 1997;138:1816– 1820. 10 Berczi I, Nagy E, de Toledo SM, Matusik RJ, Friesen HG: Pituitary hormones regulate cmyc and DNA synthesis in lymphoid tissue. J Immunol 1991;146:2201–2206. 11 Nagy E, Berczi I: Pituitary dependence of bone marrow function. Br J Haematol 1989;71:457– 462. 12 Edwards CK III, Yunger LM, Lorence RM, Dantzer R, Kelley KW: The pituitary gland is required for protection against lethal effects of Salmonella typhimurium. Proc Natl Acad Sci USA 1991;88:2274–2277. 13 Edwards CK III, Arkins S, Yunger LM, Blum A, Dantzer R, Kelley KW: The macrophageactivating properties of growth hormone. Cell Mol Neurobiol 1992;12:499–510. 14 Edwards CK III, Ghiasuddin SM, Yunger LM, Lorence RM, Arkins S, Dantzer R, Kelley KW: In vivo administration of recombinant growth hormone or interferon-Á activates macrophages: Enhanced resistance to experimental Salmonella typhimurium infection is correlated with the generation of reactive oxygen

Immune-Endocrine Axis during Aging

15

16

17

18

19

20

21

22

23

24

25

26

intermediates. Infect Immun 1992;60:2514– 2521. Savino W, Dardenne M: Immune-neuroendocrine interactions. Immunol Today 1995;16: 318–322. Dardenne M, Savino W: Interdependence of the endocrine and immune systems. Adv Neuroimmunol 1996;6:297–307. Fabris N, Mocchegiani E, Provinciali M: Plasticity of neuroendocrine-thymus interactions during aging. Exp Gerontol 1997;32:415–429. De Mello-Coelho V, Vila-Verde DM, Dardenne M, Savino W: Pituitary hormones modulate cell-cell interactions between thymocytes and thymic epithelial cells. J Neuroimmunol 1997;76:39–49. Mocchegiani E, Sartorio A, Santarelli L, Ferrero S, Fabris N: Thymulin, zinc and insulin-like growth factor-I (IGF-I) activity before and during recombinant growth hormone (rec-GH) therapy in children and adults with GH deficiency. J Endocrinol Invest 1996;19:630–637. Kooijman R, Berus D, Malur A, Delhase M, Hooghe-Peters EL: Human neutrophils express GH-N gene transcripts and the pituitary transcription factor Pit-1b. Endocrinology 1997; 138:4481–4484. Kooijman R, Malur A, van Buul-Offers SC, Hooghe-Peters EL: Growth hormone expression in murine bone marrow cells is independent of the pituitary transcription factor Pit-1. Endocrinology 1997;138:3949–3955. Weigent DA, Baxter JB, Blalock JE: The production of growth hormone and insulin-like growth factor-I by the same subpopulation of rat mononuclear leukocytes. Brain Behav Immun 1992;6:365–376. Sabharwal P, Varma S: Growth hormone synthesized and secreted by human thymocytes acts via insulin-like growth factor I as an autocrine and paracrine growth factor. J Clin Endocrinol Metab 1996;81:2663–2669. Binder G, Revskoy S, Gupta D: In vivo growth hormone gene expression in neonatal rat thymus and bone marrow. J Endocrinol 1994;140: 137–143. Weigent DA, Blalock JE: Associations between the neuroendocrine and immune systems. J Leuk Biol 1995;58:137–150. Weigent DA, Blalock JE: Production of peptide hormones and neurotransmitters by the immune system. Chem Immunol 1997;69:1–30.

27 Rinderknecht E, Humbel RE: The amino acid sequence of human insulin-like growth factor I and its structural homology with proinsulin. J Biol Chem 1978;253:2769–2776. 28 Binz K, Joller P, Froesch P, Binz H, Zapf J, Froesch ER: Repopulation of the atrophied thymus in diabetic rats by insulin-like growth factor I. Proc Natl Acad Sci USA 1990;87: 3690–3694. 29 Guler HP, Zapf J, Scheiwiller E, Froesch ER: Recombinant human insulin-like growth factor-I stimulates growth and has distinct effects on organ size in hypophysectomized rats. Proc Natl Acad Sci USA 1988;85:4889–4893. 30 Clark R, Strasser J, McCabe S, Robbins K, Jardieu P: Insulin-like growth factor-1 stimulation of lymphopoiesis. J Clin Invest 1993;92:540– 548. 31 Lin B, Kinoshita Y, Hato F, Tsuji Y: Enhancement of DNA synthetic activity of thymic lymphocytes by the culture supernatant of thymus epithelial cells stimulated by growth hormone. Cell Mol Biol 1997;43:351–359. 32 Kooijman R, Hooghe-Peters EL, Hooghe R: Prolactin, growth hormone, and insulin-like growth factor-I in the immune system. Adv Immunol 1996;63:377–454. 33 Minshall C, Liu Q, Arkins S, Kelley KW: Growth hormone and immunology; in Torosian MH (ed): Growth Hormone in Critical Illness – Research and Clinical Studies. Austin, Landes Co, 1996, pp 161–186. 34 Johnson RW, Arkins S, Dantzer R, Kelley KW: Hormones, lymphohemopoietic cytokines and the neuroimmune axis. Comp Biochem Physiol 1997;116:183–201. 35 Kelley KW, Meier WA, Minshall C, Schacher DH, Liu Q, VanHoy R, Burgess W, Dantzer R: Insulin-like growth factor-I inhibits apoptosis in hematopoietic progenitor cells: Implications in thymic aging; in Sternberg EM (ed): Proceedings of the 3rd International Congress of the International Society of Neuroimmunomodulation. New York, New York, Academy of Sciences, 1998, p 840. 36 Weigent DA: Immunoregulatory properties of growth hormone and prolactin. Pharmacol Ther 1996;69:237–257. 37 Yu-Lee LY: Molecular actions of prolactin in the immune system. PSEBM 1997;215:35–52. 38 Hodes RJ: Molecular alterations in the aging immune system. J Exp Med 1995;182:1–3.

Neuroimmunomodulation 1999;6:56–68

65

39 Miller RA: The aging immune system: Primer and prospectus. Science 1996;273:70–74. 40 Engwerda CR, Fox BS, Handwerger BS: Cytokine production by T lymphocytes from young and aged mice. J Immunol 1996;156:3621– 3630. 41 Franklin RA, Li YM, Arkins S, Kelley KW: Glutathione augments in vitro proliferative responses of lymphocytes to concanavalin A to a greater degree in old than in young rats. J Nutr 1990;120:1710–1717. 42 Franklin RA, Arkins S, Li YM, Kelley KW: Macrophages suppress lectin-induced proliferation of lymphocytes from aged rats. Mech Aging Dev 1993;67:33–46. 43 Fu YK, Arkins S, Li YM, Dantzer R, Kelley KW: Reduction of superoxide anion secretion and bactericidal activity of neutrophils from aged rats: Reversal by the combination of gamma interferon and growth hormone. Infect Immun 1994;62:1–8. 44 Davila DR, Edwards CK III, Arkins S, Simon J, Kelley KW: Interferon-Á-induced priming for secretion of superoxide anion and tumor necrosis factor-· declines in macrophages from aged rats. FASEB J 1990;4:2906–2911. 45 Ding A, Hwang S, Schwab R: Effect of aging on murine macrophages: Diminished response to IFN-Á for enhanced oxidative metabolism. J Immunol 1994;153:2146–2152. 46 Edwards CK III, Ghiasuddin SM, Schepper JM, Yunger JM, Kelley KW: A newly defined property of somatotropin: Priming of macrophages for production of superoxide anion. Science 1988;239:769–771. 47 Warwick-Davies J, Lowrie DB, Cole PJ: Growth hormone is a human macrophage activating factor: Priming of human monocytes for enhanced release of H2O2. J Immunol 1995; 154:1909–1918. 48 Ryu H, Jeong S-M, Jun C-D, Lee J-H, Kim JD, Lee B-S, Chung H-T: Involvement of intracellular Ca2+ during growth hormone-induced priming of human neutrophils. Brain Behav Immun 1997;11:39–46. 49 Arkins S, Dantzer R, Kelley KW: Somatolactogens, somatomedins, and immunity. J Dairy Sci 1993;76:2437–2450. 50 Kotzman H, Köller S, Czernin M, Clodi T, Svoboda M, Riedl G, Boltz-Nitulescu C, Zielinski C, Luger C: Effect of elevated growth hormone concentrations on the phenotype and functions of human lymphocytes and natural killer cells. Neuroendocrinology 1994;60:618– 625. 51 Sabharwal P, Zwilling B, Glaser R, Malarkey WB: Cellular immunity in patients with acromegaly and prolactinomas. Prog Neuroendocrinimmunol 1992;5:120–125. 52 Manfredi R, Tumietto F, Azzaroli L, Zucchini A, Chiodo F, Manfredi G: Growth hormone (GH) and the immune system: Impaired phagocytic function in children with idiopathic GH deficiency is corrected by treatment with biosynthetic GH. J Pediatr Endocrinol 1994;7: 245–251.

66

53 Fu YK, Arkins S, Wang BS, Kelley KW: A novel role of growth hormone and insulin-like growth factor-I: Priming neutrophils for superoxide anion secretion. J Immunol 1991;146: 1602–1608. 54 Fu YK, Arkins S, Fuh G, Cunningham BC, Wells JA, Fong S, Cronin MJ, Dantzer R, Kelley KW: Growth hormone augments superoxide anion secretion of human neutrophils by binding to the prolactin receptor. J Clin Invest 1992;89:451–457. 55 Bjerknes R, Aarskog D: Priming of human polymorphonuclear neutrophilic leukocytes by insulin-like growth factor I: Increased phagocytic capacity, complement receptor expression, degranulation, and oxidative burst. J Clin Endocrinol Metab 1995;80:1948–1955. 56 Liao W, Rudling M, Angelin B: Contrasting effects of growth hormone and insulin-like growth factor I on the biological activities of endotoxin in the rat. Endocrinology 1997;138: 289–295. 57 Benfield MR, Vail A, Bucy RP, Weigent DA: Growth hormone induces interferon gamma production and may play a role in presentation of alloantigens in vitro. Neuroimmunomodulation 1997;4:19–27. 58 Sommese L, Donnarumma G, de l’Ero C, Marcatili A, Vitiello M, Galdiero M: Growth hormone modulates IL-alpha and IFN-gamma release by murine splenocytes activated by LPS or porins of Salmonella typhimurium. J Med Microbiol 1996;45:40–47. 59 Ogawa M: Differentiation and proliferation of hematopoietic stem cells. Blood 1993;81: 2844–2853. 60 Till JE, McCulloch EA: A direct measurement of the radiation sensitivity of normal mouse bone marrow cells. Radiat Res 1961;14:213– 222. 61 Baum CM, Weisman IL, Tsukamoto AS, Buckle AM, Peault B: Isolation of a candidate human hematopoietic stem-cell population. Proc Natl Acad Sci USA 1992;89:2804–2808. 62 Osawa M, Hanada K, Hamada H, Nakauchi H: Long-term lymphohematopoietic reconstitution by a single CD34-low/negative hematopoietic stem cell. Science 1996;273:242–245. 63 Sachs L, Lotem J: The network of hematopoietic cytokines. PSEBM 1994;206:170–175. 64 To LB, Haylock DN, Simmons PJ, Juttner CA: The biological and clinical uses of blood stem cells. Blood 1997;89:2233–2258. 65 Whitlock CA, Witte ON: Long-term culture of B lymphocytes and their precursors from murine bone marrow. Proc Natl Acad Sci USA 1982;79:3608–3612. 66 Kondo M, Weissman IL, Akashi K: Identification of clonogenic common lymphoid progenitors in mouse bone marrow. Cell 1997;91:661– 672. 67 Stephan RP, Lill-Elghanian DA, Witte PL: Development of B cells in aged mice: Decline in the ability of pro-B cells to respond to IL-7 but not to other growth factors. J Immunol 1997; 158:1598–1609.

Neuroimmunomodulation 1999;6:56–68

68 Moore TA, Zlotnik A: Differential effects of Flk-2/Flt-3 ligand and stem cell factor on murine thymic progenitor cells. J Immunol 1997; 158:4187–4192. 69 Wu LM, Antica GR, Johnson R, Scollay R, Shortman K: Developmental potential of the earliest precursor cells from the adult mouse thymus. J Exp Med 1991;174:1617–1627. 70 Clawson TF, Lee WH, Yoder MC: Differential expression of insulin-like growth factor binding proteins in murine hematopoietic stromal cell lines. Mol Cell Endocrinol 1996;120:59–66. 71 Abboud SL, Bethel CR, Aron DC: Secretion of insulin-like growth factor I and insulin-like growth factor binding proteins by murine bone marrow stromal cells. J Clin Invest 1991;88: 470–475. 72 Witte PL, Frantsve LM, Hergott M, Rahbe SM: Cytokine production and heterogeneity of primal stromal cells that support B lymphopoiesis. Eur J Immunol 1993;23:1809–1817. 73 Arkins S, Rebeiz R, Birgayn A, Reese DL, Kelley KW: Murine macrophages express abundant insulin-like growth factor-I class I Ea and Eb transcripts. Endocrinology 1993;133:2334– 2343. 74 Arkins S, Rebeiz N, Brunke-Reese DL, Birgayn A, Kelley KW: Interferon-gamma inhibits macrophage insulin-like growth factor-I synthesis at the transcriptional level. Mol Endocrinol 1995;9:350–360. 75 Ratajczak MZ, Kuczynski WI, Onodera K, Moore J, Ratajczak J, Kregenow DA, DeRiel K, Gewirtz AM: A reappraisal of the role of insulin-like growth factor I in the regulation of human hematopoiesis. J Clin Invest 1994;94: 320–327. 76 Li YM, Schacher DH, Liu Q, Arkins S, Rebeiz N, McCusker R, Dantzer R, Kelley KW: Regulation of myeloid growth and differentiation by the insulin-like growth factor-I receptor. Endocrinology 1997;138:362–368. 77 Bellone G, Geuna M, Carbone A, Silvestri S, Foa R, Emanuelli G, Matera L: Regulatory action of prolactin on the in vitro growth of CD34+ve human hemopoietic progenitor cells. J Cell Physiol 1995;163:221–231. 78 Bellone G, Astarita P, Artusio E, Silvestri S, Mareschi K, Turletti A, Buttiglieri S, Emanuelli G, Matera L: Bone marrow stroma-derived prolactin is involved in basal and platelet-activating factor-stimulated in vitro erythropoiesis. Blood 1997;90:21–27. 79 Knyszynski A, Adler-Kunin S, Globerson A: Effects of growth hormone on thymocyte development from progenitor cells in the bone marrow. Brain Behav Immun 1992;6:327–340. 80 Murphy WJ, Durum SK, Longo DL: Role of neuroendocrine hormones in murine T cell development: Growth hormone exerts thymopoietic effects in vivo. J Immunol 1992;149: 3851–3857. 81 Murphy WJ, Durum SK, Longo DL: Differential effects of growth hormone and prolactin on murine T cell development and function. J Exp Med 1993;178:231–236.

Burgess/Liu/Zhou/Tang/Ozawa/VanHoy/ Arkins/Dantzer/Kelley

82 Gjerset RA, Yeargin J, Volkman SK, Vila V, Arya J, Haas M: Insulin-like growth factor-I supports proliferation of autocrine thymic lymphoma cells with a pre-T cell phenotype. J Immunol 1990;145:3497–3501. 83 Yamada M, Hato F, Kinoshita Y, Tominaga K, Tsuji Y: The indirect participation of growth hormone in the thymocyte proliferation system. Cell Mol Biol 1994;40:111–121. 84 Kooijman R, Scholtens LE, Rijkers GT, Zegers BJM: Type I insulin-like growth factor receptor expression in developmental stages of human thymocytes. J Endocrinol 1995;147:203–209. 85 Landreth KS, Narayanan R, Dorshkind K: Insulin-like growth factor-I regulates pro-B cell differentiation. Blood 1992;80:1207–1212. 86 Gibson L, Piktel D, Landreth K: Insulin-like growth factor-I potentiates expansion of interleukin-7-dependent pro-B cells. Blood 1993;82: 3005–3011. 87 Robbins K, McCabe S, Scheiner T, Strasser J, Clark R, Jardieu P: Immunological effects of insulin-like growth factor-I enhancement of immunoglobulin synthesis. Clin Exp Immunol 1994;95:337–342. 88 Montecino-Rodriguez E, Johnson A, Dorshkind K: Thymic stromal cells can support B cell differentiation from intrathymic precursors. J Immunol 1996;156:963–967. 89 Jardieu P, Clark R, Mortensen D, Dorshkind K: In vivo administration of insulin-like growth factor-I stimulates primary B lymphopoiesis and enhances lymphocyte recovery after bone marrow transplantation. J Immunol 1994;152:4320–4327. 90 Montecino-Rodriguez E, Clark RG, PowellBraxton L, Dorshkind K: Primary B cell development is impaired in mice with defects of the pituitary/thyroid axis. J Immunol 1997;159: 2712–2719. 91 Merchant MS, Garvey BA, Riley RL: B220 – bone marrow progenitor cells from New Zealand black autoimmune mice exhibit an ageassociated decline in pre-B and B-cell generation. Blood 1995;7:1850–1857. 92 Aspinall R: Age-associated thymic atrophy in the mouse is due to a deficiency affecting rearrangement of the TCR during intrathymic T cell development. J Immunol 1997;158:3037– 3045. 93 Merchav S, Tatarsky I, Hochberg Z: Enhancement of human granulopoiesis in vitro by biosynthetic insulin-like growth factor I/somatomedin C and human growth hormone. J Clin Invest 1988;81:791–797. 94 Arkins S, Rebeiz N, Brunke-Reese DL, Minshall C, Kelley KW: The colony-stimulating factors induce expression of insulin-like growth factor-I messenger ribonucleic acid during hematopoiesis. Endocrinology 1995;136:1153– 1160. 95 Geenen V, Achour I, Robert F, Vandersmissen E, Sodoyez JC, Defresne MP, Boniver J, Lefebvre PJ, Franchimont P: Evidence that insulin-like growth factor 2 (IGF2) is the dominant thymic peptide of the insulin superfamily. Thymus 1993;21:115–127.

Immune-Endocrine Axis during Aging

96 Minshall C, Arkins S, Freund GG, Kelley KW: Requirement for phosphatidylinositol 3)-kinase to protect hemopoietic progenitors against apoptosis depends upon the extracellular survival factor. J Immunol 1996;156: 939–947. 97 Freund GG, Kulas DT, Mooney RA: Insulin and IGF-I increase mitogenesis and glucose metabolism in the multiple myeloma cell line, RPMI 8226. J Immunol 1993;151:1811– 1820. 98 Van Buul-Offers SC, de Haan K, ReijnenGresnigt MG, Meinsma D, Jansen M, Oei SL, Bonte EJ, Sussenbach JS, Van den Brande: Overexpression of human insulin-like growth factor-II in transgenic mice causes increased growth of the thymus. J Endocrinol 1995;144: 491–502. 99 Kooijman R, van Buul-Offers SC, Scholtens LE, Reijnen-Gresnigt RG, Zegers BJ: T and B cell development in pituitary deficient insulin-like growth factor-II transgenic dwarf mice. J Endocrinol 1997;155:165–170. 100 Van der Ven LT, Roholl PJ, Reijnen-Gresnigt MG, Bloeman RJ, van Buul-Offers SC: Expression of insulin-like growth factor-II (IGFII) and histological changes in the thymus and spleen of transgenic mice overexpressing IGF-II. Histochem Cell Biol 1997;107:193– 203. 101 Li YM, Arkins S, McCusker RH Jr, Donovan SM, Liu Q, Jayaraman S, Dantzer R, Kelley KW: Macrophages synthesize and secrete a 25-kilodalton protein that binds insulin-like growth factor-I. J Immunol 1996;156:64–72. 102 Merchav S, Tatarsky I, Hochberg Z: Enhancement of erythropoiesis in vitro by human growth hormone is mediated by insulin-like growth factor I. Br J Haematol 1988;70:267– 271. 103 Merchav S, Silvian-Drachsler I, Tatarsky I, Lake M, Skottner A: Comparative studies of the erythroid- potentiating effects of biosynthetic human insulin-like growth factors I and II. J Clin Endocrinol Metab 1992;73:447– 452. 104 Muta K, Krantz SB, Bondurant MC, Wickrema A: Distinct roles of erythropoietin, insulin-like growth factor I, and stem cell factor in the development of erythroid progenitor cells. J Clin invest 1994;94:34–43. 105 Kotzmann H, Reidl M, Clodi M, Barnas U, Kaider A, Hocker P, Luger A: The influence of growth hormone substitution therapy on erythroid and myeloid progenitor cells in adult patients with growth hormone deficiency. Eur J Clin Invest 1996;26:1175–1181. 106 Vihervouri E, Virtanen M, Koistinen H, Koistinen R, Seppala M, Siimes MA: Hemoglobin level is linked to growth hormonedependent proteins in short children. Blood 1996;87:2075–2081. 107 Cory S: Regulation of lymphocyte survival by the Bcl-2 gene family. Annu Rev Immunol 1995;13:513–543. 108 Williams GT, Smith CA, Spooncer E, Dexter TM, Taylor DR: Haemopoietic colony stimulating factors promote cell survival by suppressing apoptosis. Nature 1990;343:76–79.1

109 Rodriguez-Tarduchy G, Collins MKL, Garcia I, Lopez-Rivas A: Insulin-like growth factor-I inhibits apoptosis in IL-3-dependent hemopoietic cells. J Immunol 1992;149:535– 450. 110 Yao R, Cooper GM: Requirement for phosphatidylinositol 3-kinase in the prevention of apoptosis by nerve growth factor. Science 1995;267:2003–2006. 111 Dudek H, Datta SR, Franke TF, Birnbaum MJ, Yao R, Copper GM, Segal RA, Kaplan DR, Greenberg ME: Regulation of neuronal survival by the serine-threonine protein kinase Akt. Science 1997;275:661–665. 112 Kulik G, Klippel A, Weber MJ: Antiapoptotic signalling by the insulin-like growth factor I receptor, phosphatidylinositol 3-kinase, and Akt. Mol Cell Biol 1997;17:1595–1606. 113 Fairbairn LJ, Cowling GJ, Reipert BM, Dexter TM: Suppression of apoptosis allows differentiation and development of a multipotent hemopoietic cell line in the absence of added growth factors. Cell 1993;74:823–832. 114 Singleton JR, Dixit VM, Feldman EL: Type I insulin-like growth factor receptor activation regulates apoptotic proteins. J Biol Chem 1996;271:31791–31794. 115 Minshall C, Arkins S, Straza J, Connors J, Dantzer R, Freund GG, Kelley KW: IL-4 and insulin-like growth factor-I inhibit the decline in Bcl-2 and promote the survival of IL-3deprived myeloid progenitors. J Immunol 1997;159:1225–1232. 116 Vella A, Teague TK, Ihle J, Kappler J, Marrack P: Interleukin-4 (IL-4) or IL-7 prevents the death of resting T cells: Stat 6 is probably not required for the effects of IL-4. J Exp Med 1997;186:325–330. 117 Liu Q, Schacher D, Hurth C, Freund GG, Dantzer R, Kelley KW: Activation of PI 3kinase by insulin-like growth factor-I rescues promyeloid cells from apoptosis and permits their differentiation into granulocytes. J Immunol 1997;159:829–837. 118 Pa´rrizas M, Saltiel AR, LeRoith D: Insulinlike growth factor 1 inhibits apoptosis using the phosphatidylinositol 3'-kinase and mitogen-activated protein kinase pathways. J Biol Chem 1997;272:154–161. 119 Liu Q, Ning W, Dantzer R, Freund GG, Kelley KW: Activation of protein kinase C-˙ and phosphatidylinositol 3)-kinase and promotion of macrophage differentiation by insulinlike growth factor-I. J Immunol 1998;160: 1393–1401. 120 Smith PJ, Wise LS, Berkowitz R, Wan C, Rubin CS: Insulin-like growth factor-I is an essential regulator of the differentiation of 3T3-L1 adipocytes. J Biol Chem 1988;263: 9402–9408. 121 Rosenthal SM, Cheng ZQ: Opposing early and late effects of insulin-like growth factor-I on differentiation and cell cycle regulatory retinoblastoma protein in skeletal myoblasts. Proc Natl Acad Sci USA 1995;92:10307– 10311.

Neuroimmunomodulation 1999;6:56–68

67

122 Pahlman S, Meyerson G, Lindgren E, Schalling M, Johansson I: Insulin-like growth factor I shifts from promoting cell division to potentiating maturation during neuronal differentiation. Proc Natl Acad Sci USA 1991;88: 9994–9998. 123 Ways DK, Posekany K, de Vente J, Garris T, Chen J, Hooker J, Qin W, Cook P, Fletcher D, Parker P: Overexpression of protein kinase C-˙ stimulates leukemic cell differentiation. Cell Growth Differ 1994;5:1195–1203. 124 Jones JI, Clemmons DR: Insulin-like growth factors and their binding proteins: Biological actions. Endocr Rev 1995;16:3–34. 125 Yenush L, White MF: The IRS-signalling system during insulin and cytokine action. Bioessays 1997;19:491–500.

68

126 Miyazaki T, Kawahara A, Fujii H, Nakagawa Y, Liu ZJ, Oishi I, Silvennoinen O, Witthuhn BA, Ihle JN: Functional activation of Jak1 and Jak3 by selective association with IL-2 receptor subunits. Science 1994;266:1045– 1047. 127 Leaman DW, Leung S, Li X, Stark GR: Regulation of STAT-dependent pathways by growth factors and cytokines. FASEB J 1996; 10:1578–1588. 128 Chen XH, Bharvin KR, Wang LM, Frankel M, Ellmore N, Flavell RA, LaRochelle WJ, Pierce JH: Jak1 expression is required for mediating interleukin-4-induced tyrosine phosphorylation of insulin receptor substrate and Stat6 signaling molecules. J Biol Chem 1997;272:6556–6560. 129 White MF, Kahn CR: The insulin signaling system. J Biol Chem 1994;269:1–4. 130 Keegan AD, Nelms K, White M, Wang LM, Pierce JH, Paul WE: An IL-4 receptor region containing an insulin receptor motif is important for IL-4 mediated IRS-1 phosphorylation and cell growth. Cell 1994;76:811–820.

Neuroimmunomodulation 1999;6:56–68

131 Wang LM, Keegan AD, Li W, Lienhard GE, Pacini S, Gutkind JS, Myers MG Jr, Sun XJ, White MF, Aaronson SA, Paul WE, Pierce JH: Common elements in interleukin-4 and insulin signaling pathways in factor-dependent hematopoietic cells. Proc Natl Acad Sci USA 1993;90:4032–4036. 132 Welham MJ, Bone H, Levings M, Learmouth L, Wang LM, Leslie KB, Pierce JH, Schrader JW: Insulin receptor substrate-2 is the major 170-kDa protein phosphorylated on tyrosine in response to cytokines in murine lymphohematopoietic cells. J Biol Chem 1997;272: 1377–1381. 133 Patti ME, Sun XJ, Bruenings JC, Araki E, Lipes MA, White MF, Kahn CR: 4PS/insulin receptor substrate (IRS)-2 is the alternative substrate of the insulin receptor in IRS-1-deficient mice. J Biol Chem 1995;270:24670– 24673.

Burgess/Liu/Zhou/Tang/Ozawa/VanHoy/ Arkins/Dantzer/Kelley

Neuroimmunomodulation 1999;6:69–80

Use of Neuroendocrine Hormones to Promote Reconstitution after Bone Marrow Transplantation Mary A. Woody a Lisbeth A. Welniak b Susan Richards b Dennis D. Taub c Zhi-Gang Tian e Rui Sun e Dan L. Longo c William J. Murphy d a Laboratory of Leukocyte Biology, National Cancer Institute-Frederick Cancer Research and Development Center, Frederick, Md.; b Genzyme Corporation, Framingham, Mass.; c National Institute on Aging, Baltimore, Md.; d IRSP, SAIC-Frederick, National Cancer Institute, Frederick Cancer Research and Development Center, Frederick, Md., USA; e Shandong Academy of Medical Sciences, Jinan, PR China

Key Words Prolactin W Growth hormone W Immune parameters W Hematopoiesis W Transplantation

Abstract A survey of the previous literature and the data shown here indicate that neuroendocrine hormones such as growth hormone and prolactin may be of potential clinical use after bone marrow transplantation (BMT) to promote hematopoietic and immune recovery. The amounts of hormones used in our model do not promote weight gain suggesting that their lymphohematopoietic actions were independent of their anabolic effects. While the hormones may not produce the same extent of immune/hematopoietic effects when compared to conventional hematopoietic and immune stimulating cytokines (i.e. IL-2 or G-CSF), their pleiotropic effects and limited toxicity after systemic administration makes them attractive to test in the post-BMT setting. However, more work needs to be performed to understand the mechanism(s) of their action, particularly with regard to T-cell function and development.

ABC

© 1999 S. Karger AG, Basel 1021–7401/99/0062–0069$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: http://BioMedNet.com/karger

Introduction

Bone marrow transplantation (BMT) is currently used to treat a variety of disease states ranging from cancer to aplastic anemia to immune deficiency disorders. However, a significant obstacle limiting the efficacy of BMT is the occurrence of opportunistic infections, which result due to the severe immunosuppression from the transplant-conditioning regimen. The length of this window of immunosuppression is determined by the rate of recovery of the patient’s immune system. Besides being a clinical problem, the subsequent longer hospital stay is a main factor in the expense of BMT. Agents that would accelerate the process of immune reconstitution would therefore be advantageous, not only in increasing resistance to opportunistic infections but also in allowing for shorter hospital stays. Cytokines, such as granulocyte-colony stimulating factor (G-CSF) and granulocyte/monocytecolony stimulating factor (GM-CSF) are currently used to accelerate neutrophil recovery post-BMT [1]. Interleukin (IL)-2 has also been given post-BMT with variable effects on immune parameters [1]. However, due to the pleiotropic effects of cytokines, their systemic administration is also associated with significant toxicities. Agents that would accelerate T-cell reconstitution would be very useful from the standpoint of promoting resistance to viral

William J. Murphy, PhD IRSP, SAIC-Frederick, NCI-FCRDC, Bldg 567, Rm 210 Frederick, MD 21702-1201 (USA) Tel. +1 301 846 5443, Fax +1 301 846 6107 E-Mail [email protected]

pathogens (i.e., cytomegalovirus) and increasing the antitumor efficacy of BMT when BMT is used for the treatment of cancer. However, there have been few clinical studies examining such agents in BMT, in part due to the severe myelosuppression and fragility of the patient postBMT and in part related to the many drugs such patients receive to prevent infection and control graft-versus-host disease. Neuroendocrine hormones may be advantageous post-BMT as they have a variety of potentially useful biological effects and may be better tolerated than cytokines. This review will attempt to survey the previous literature concerning the immunologic and hematopoietic effects of neuroendocrine hormones, focusing on growth hormone (GH) and prolactin (PRL). We shall first examine the effects of GH and PRL on hematopoiesis. This is important for several reasons. First, any agent given post-BMT to promote immune reconstitution should not have a deleterious effect on myeloid reconstitution. Second, if the agent promoted myeloid reconstitution, particularly neutrophils, it would increase resistance to bacterial pathogens and possibly bypass the need for other cytokines. Third, one potential mechanism by which a neuroendocrine hormone might promote T-cell recovery would be by exerting effects in the bone marrow on the earliest Tcell progenitors. We will then examine the roles of GH and PRL on T-cell function to ascertain if they may be of use to promote T-cell recovery after BMT.

Neuroendocrine Hormones and Hematopoiesis

The neuroendocrine hormones PRL and GH have been reported to stimulate hematopoietic activity both in vivo and in vitro. These activities have been investigated in a variety of models including hypophysectomized animals, pit-1-deficient (dwarf) mice, exogenous administration of hormone as well as in drug-induced myelosuppression. More recent models have included the use of knockout and transgenic animals to understand the specific contributions of the individual hormones. The receptors for PRL and GH are members of the hematopoietin receptor family that includes receptors for erythropoietin, leukemia inhibitory factor (LIF), IL-2, IL3, IL-4, IL-5, GM-CSF and G-CSF [2]. Activation of receptors by ligand binding results in activation of the Jak2/Stat5 pathway among others [3]. The PRL and GH receptors are expressed on a number of hematopoietic cell types. PRL receptors are expressed on 95% of bone marrow-derived cells. These cell types include B, T and natural killer (NK) cells, monocytes, macrophages, polymor-

70

Neuroimmunomodulation 1999;6:69–80

phonuclear cells and CD34+ stem cells in humans [4, 5]. GH receptors are expressed at variable levels on macrophages, T and B cells within the mouse bone marrow, spleen and on peripheral blood leukocytes [6]. PRL and GH activity in the bone marrow or splenic microenvironment is not only due to systemic circulation of hormone secreted from the anterior pituitary (AP) but also localized production of these same hormones in hematopoietic lymphoid tissue [4, 7, 8]. Regulation of the local release may not be under the same control as GH and PRL release in the AP especially because some, but not all, production of hematopoietic/splenic PRL and GH is regulated by pit-1 (a transcription factor required for production of PRL, GH and TSH in the AP) [8–12]. Additionally, GH can exert indirect effects via stimulation of the secondary mediator insulin-like growth factor 1 (IGF-1), which is produced by many cell types in response to GH [13].

Background on Hematopoietic Studies

To study the putative hematopoietic effects of a given agent, investigators can study effects of the loss or addition of the agent in vivo. Though direct effects of the agent cannot be assessed definitely in these studies, the overall effect of the agent can be observed. Peripheral blood cell counts and the cell counts or cellularities of the bone marrow (and spleen in the mouse) are useful endpoints for hematopoietic function. However, restoration of blood counts is not the only important endpoint. Normal individuals require sufficient hematopoiesis to respond to stress. Thus, a complete evaluation includes an assessment of physiologic reserve. These studies can include analysis of the precursor populations in the resting animal or the response of the hematopoietic system to a particular stress, for example, exposure to a myelotoxic agent. How the myelorestorative agent under investigation affects hematopoiesis under these conditions often reveals hematopoietic effects that were not apparent in the unperturbed animal. Another wrinkle in the studies of hematopoietic activity is determining whether observed effects are direct or secondary. This can be dissected in cell cultures at three levels in clonogenic assays of precursor cells: in co-culture with pre-established bone marrow-derived stromal cell layers, in unfractionated bone marrow or spleen cells, or at the single cell level. Bone marrowderived stromal cells are the nonhematopoietic cells of the bone marrow that support and maintain hematopoiesis. Studies in co-cultures remove systemic influences but still

Woody/Welniak/Richards/Taub/Tian/Sun/ Longo/Murphy

maintain the complex local cellular interactions. Bone marrow and spleen cell preparations contain many cell types that can influence hematopoietic cell growth in both a positive and negative manner. Clonogenic assays (CFUc) performed on unfractionated cell preparations or samples depleted of single or multiple lineages limit the number of variables affecting hematopoietic growth and differentiation but still do not address direct effects of the agent under investigation. Single cell analysis of highly purified precursors is the only method to determine direct effects without paracrine interactions during the first cell division.

Growth Hormone and Hematopoiesis

Hematopoietic activity of GH has been investigated in both in vivo and in vitro experiments. Studies in DW/J dwarf mice [14] and hypophysectomized rats [15] indicate that loss of neuroendocrine hormones depresses hematological parameters [14]. Several hormones are deficient in these models and the requirement for a single factor cannot be assumed. With few exceptions, most cytokines and hormones with demonstrated activity on hematopoiesis are redundant and effects in gene knockout mice are mild at the least. In fact, the recent generation of a PRL-deficient mouse reveals no effect on steady-state hematopoiesis [16]. Administration of human GH to dwarf mice restores hematopoietic function except for Blymphopoiesis [14], which is corrected by the administration of thyroxine [17, 18]. The result with the knockout mouse indicates that either PRL plays little role in the homeostasis of normal hematopoiesis or that other factors can compensate for the absence of PRL. Unlike mice, in humans with GH deficiencies there are no clinically significant hematologic deficiencies. However, in a study of 11 patients with hypopituitarism receiving replacement therapy with thyroxine, cortisol and sex steroids, but not GH, hematopoietic progenitor populations were in the low normal range. Significant increases in erythroid and myeloid progenitor colonies were observed following GH replacement treatment without significant effects on peripheral blood cell counts [19]. Under physiologic conditions, the increase in progenitor pools may be a stress response, generating adequate stores of precursors for rapid expansion. Similarly in normal or transgenic mice, GH administration or overexpression does not change the peripheral blood cell counts but hematopoietic progenitor cells and DNA synthesis increase and splenic hyperplasia occurs [14, 20, 21]. When

Hormones and Bone Marrow Transplantation

hematologic parameters are decreased due to myelosuppression induced by AZT, GH or a secondary mediator, IGF-1 can stimulate both CFU-C and peripheral blood count recovery [14, 22]. The use of ovine GH, which binds only the GH receptor and not the PRL receptor as opposed to human GH, which binds both receptors, improves erythroid but not leukocyte counts in myelosuppression induced by AZT [23]. These finding support a role for GH on erythropoiesis and a role for PRL alone or in combination with GH on myelopoiesis. In vitro studies with GH have shown direct enhancement of erythroid cell growth [20] and indirect stimulation of granulopoiesis via IGF-1 production [24]. IGF-1 enhancement of hematopoietic, especially erythroid, cell growth appears to be an indirect phenomenon. Human CD34+ cells do not express IGF-1 receptors and IGF-1 does not enhance progenitor cell growth. Bone marrow-derived IGF-1 receptor-positive cells restore the erythropoietic activity in cocultures with CD34+ cells [25]. However, exogenous administration of IGF-1 has been shown to promote hematopoiesis in mice and reverse the myelosuppression induced by AZT [23]. Thus, both GH and IGF-1 can exert hematopoietic growth-promoting effects in vivo and in vitro.

Prolactin and Hematopoiesis

The differences in hematopoietic activity seen with administration of ovine GH or human GH suggest that engagement of the PRL receptor will have hematological consequences. PRL, like GH, restores bone marrow function in hypophysectomized rats [20]. If instead residual PRL is depleted from these animals with anti-PRL antibodies, they become severely anemic and die of hematological failure [15]. In the DW/J dwarf mouse model, administration of PRL also increases bone marrow cellularity [18]. In vitro, PRL directly enhances the effect of IL-3, GM-CSF and erythropoietin on CFU-G and BFU-E from purified CD34+ cells [5]. In addition to pituitary sources of PRL, the bone marrow microenvironment is a local source of PRL [26]. Basal levels are produced in vitro that give rise to a slight enhancement of erythropoiesis but stimulation with platelet-activating factor (PAF) increases production of PRL from the stromal cells with concomitant enhancement of erythroid and myeloid progenitor cell growth. Taken together, the studies on the hematopoietic activities of the neuroendocrine hormones GH and PRL suggest only minor roles for either hormone alone in basal or

Neuroimmunomodulation 1999;6:69–80

71

Table 1. Effects of r-hGH and r-hPRL on hematopoietic progenitor

content in mice after syngeneic BMT Organ

Treatment

CFU-C

Spleen

Control r-hGH r-hPRL

13.8B3.8 25.9B2.5* 29.9B1.3+

Bone marrow

Control r-hGH r-hPRL

9.4B1.3 21.6B2.0+ 28.1B3.1+

Balb/c mice received 850 cGy and 1 ! 106 syngeneic BMC i.v. The mice then received PBS, 10 Ìg r-hGH, or 10 Ìg r-hPRL i.p. every other day until end of assay. CFU-C was assessed 21 days postBMT. There were 4 mice per group. Data are presented as means B SEM. * p ! 0.05; + p ! 0.01, relative to PBS-treated controls.

steady-state hematopoiesis. The only demonstrable activities are limited effects on erythropoiesis and hematopoietic progenitor pools. However, deficiencies in multiple hormones reveal redundant hematopoietic functions and a role in homeostasis. Additionally, individually each hormone can stimulate enhanced hematopoietic recovery from myelosuppression. The role these hormones play in hematopoiesis and under what physiologic conditions remain to be seen. However, the findings are consistent with a co-stimulatory role. It is possible that the conventional colony-stimulating factors are more active in precursor cells conditioned by GH or PRL stimulation. Bone marrow or hematopoietic cell transplantation can be used as a tool to further dissect the effect of the agents on the hematopoietic system as well as the functional reconstitution of the immune system. In transplant models, the agent can be administered to the donor before the graft is harvested or to the recipient before, concurrent to or following the transplant. Depending on the manipulation of the transplant protocols, information can be generated on the short- and long-term marrow-repopulating ability of the grafts and the rate of recovery induced by the agent compared to controls. We have found that after lethal irradiation and reconstitution with syngeneic bone marrow cells, administration of either GH or PRL can result in significant increases in hematopoietic progenitor content in the spleens and bone marrow of mice (table 1). The hormones were given at doses that did not promote weight gain. Consideration to the local concentration of the biological agent attained by exogenous administration should be taken into account. Achievement of concentra-

72

Neuroimmunomodulation 1999;6:69–80

tions in the range of physiological levels can yield information into the activity of the agent when it is increased in pathological conditions. Supraphysiological levels, while not as relevant to studies of normal biological activities, can reveal clinically exploitable properties. These studies therefore suggest that both GH and PRL have significant hematopoietic growth-promoting effects and may therefore accelerate recovery post-BMT. These hematopoietic effects may not be limited to myeloid lineages as lymphocytic reconstitution may also be promoted.

Effects of Neuroendocrine Hormones on Immune Parameters

A number of studies in animal models that are deficient in pituitary hormones suggested a link between neuroendocrine hormones and immune function or development, particularly T cells. It was recognized that injection of antihypophysis [27] or anti-GH [28] sera into mice induced a wasting syndrome that was characterized by thymic and peripheral lymphoid tissue atrophy and resembled the effects of neonatal thymectomy. Destruction or removal of the hypophysis of animals to induce pituitary hormone deficiency or pituitary-deficient dwarf mice have been two major models that have been used to examine the effects of neuroendocrine hormones on immune development and function. As examples of studies with the former model, rodents with hypothalamic lesions or hypophysectomies that decreased GH and PRL secretion exhibited suppressed NK maturation and activity [29–31]. Antibody titers to SRBC [32, 33] and contact sensitivity to dinitrochlorobenzene [33], reduced in hypophysectomized rats, were partly restored by treatment with GH, PRL [33], or syngeneic pituitary implants [32]. These and other studies that associated decreased NK numbers or activity and hypophysectomy are discussed in the section on the effects of the hormones on NK cells. Much use has been made of the various pituitary-deficient dwarf mouse models to examine the effects of neuroendocrine hormones on immune function. The phenotypes of the Ames, Snell-Bagg, and Jackson strains of dwarf mice are essentially indistinguishable: all lack lactotrophs, somatotrophs, and thyrotrophs, and have hypoplastic anterior pituitaries [34]. It is important to note that these mouse strains have multiple hormonal deficiencies. The Snell-Bagg and Jackson strains have allelic mutations in the Pit-1 gene, which is a transcription factor that activates the PRL and GH genes. Both the Snell-Bagg strain [17, 35–38] and Ames strain [39] are characterized

Woody/Welniak/Richards/Taub/Tian/Sun/ Longo/Murphy

by reduced body weights relative to normal littermates, involution of the thymus and peripheral immune system accompanied by lymphopenia, and early death from a wasting disease. This phenotype becomes apparent 2–3 weeks after birth. A progressive loss of the thymic cortex occurs in the Snell-Bagg mouse, and the effect of the hormonal deficiency on thymic and peripheral immune system development closely resembles that of neonatal thymectomy [37] or administration of anti-GH serum in normal 2- and 3-week-old mice [28]. Some have reported much less profound effects in the thymuses of these animals [40], but the comparison with normal littermates was made at 4 months, by which time normal mouse thymuses have begun involution. Delayed weaning of SnellDwarf mice tends to prevent the progressive loss of CD4+CD8+ thymocytes characteristic of these animals (presumably because of hormone ingestion through mother’s milk) [41]. Treatment of the dwarf mice with GH alone [35], or in combination with thyroxine [17] or normal lymph node lymphocytes [38], restored thymic morphology and function. However, other groups did not detect this loss of T-cell progenitors in these mice [16]. This may be due to differences in housing conditions. We have found that these mice are extremely susceptible to stress. In fact, they need to be caged with normal littermates, as they tend to suffer from hypothermia when housed alone. This, in addition to differences in weaning and time of assay, may reconcile why some groups, including ours, reported T-cell deficiencies in these mice [41]. Nonetheless, as stress does play a significant role in thymocyte survival, this does give insight into a potential role of these hormones in thymus function. More recent analysis of the effects of neuroendocrine hormones in dwarf mice has included phenotypic analysis of immune cell populations and finer dissection of the roles of GH and PRL. Thymic size increased and the CD4+CD8+ thymocyte population reappeared in SnellBagg dwarf mice after treatment with recombinant human GH [42]. Because recombinant human GH also has PRL-like activity, ovine GH and PRL were used to define the role of each hormone on thymic reconstitution in dwarf animals: ovine GH increased thymic cellularity and restored the CD4+CD8+ thymic population, whereas PRL further depleted these cells [43]. Thus, GH may have greater thymopoietic potential than PRL. When we administered GH to mice post-BMT we found that these mice presented with significantly greater thymocyte numbers compared with their controls suggesting that GH was capable of promoting T-cell reconstitution after syngeneic BMT (table 2).

Table 2. Effects of ovine GH or ovine PRL on thymic cellularity

Hormones and Bone Marrow Transplantation

Neuroimmunomodulation 1999;6:69–80

after syngeneic BMT in mice Day of assay

Treatment

Cellularity (!106)

Day 8

PBS Ovine PRL Ovine GH

0.24B0.03 0.77B0.16* 0.24B0.06

Day 14

PBS Ovine PRL Ovine GH

4.90B0.4 7.53B1.04 15.03B4.6

Day 21

PBS Ovine PRL Ovine GH

28.7B1.5 42.0B1.4+ 30.9B2.2

Day 28

PBS Ovine PRL Ovine GH

54.9B1.6 75.7B5.16 74.1B6.8

Balb/c mice received 850 cGy and 1 ! 106 BMC i.v. The mice then received either PBS, 10 Ìg ovine PRL or 10 mg ovine GH i.p. every other day. There were 2–4 mice per group. * p ! 0.05; + p ! 0.01, relative to PBS treated controls.

The immunomodulatory activities of GH have been studied in other animal models. GH- and PRL-secreting pituitary adenoma implants restored thymic function in aged rats, increasing numbers of cortical thymocytes [44]. GH can mediate its effects through IGF-1, and IGF-1 treatment-repopulated thymuses of diabetic rats [45], increased thymic weights in normal mice [46], and augmented proliferation of thymic lymphoma cells [47]. In summary, in various animal models GH has positive effects on thymopoiesis. The restorative effects of the neuroendocrine hormones on peripheral immune function will be discussed below. The relevance of the animal models just discussed to humans is unclear because there does not seem to be any consistent correlation between neuroendocrine deficiencies and extent or specific type of immune impairment in humans. Normal proportions of T and B cells have been reported in pituitary dwarf patients and these cells frequently seem to be functionally normal, as determined by serum immunoglobulin levels and proliferation in response to phytohemagglutinin. However, NK cell activity or number is often reduced in these patients, as will be discussed later. Administration of human GH to patients often, but not always, improves immunological parameters. Peripheral lymphocytes isolated from GH-deficient children responded similarly to cells from age-matched

73

controls to phytohemagglutinin and in the autologous mixed lymphocyte reaction, but 3 of 4 patients studied responded poorly in allogeneic mixed lymphocyte reactions [48]. Proportions of total T cells and CD4+ cells were similar to controls, whereas B and CD8+ cells were, in some patients, more numerous than in controls. No consistent differences in T-cell populations were found in GH-deficient patients, and treatment with GH-releasing hormone did not alter these levels [49]. Thus, these studies suggest no strong link between pituitary hormone deficiency and immunological impairment in humans, but obviously a complete understanding of the importance of neuroendocrine hormones to human immune development and function is hampered by small populations in clinical studies and the impossibility of obtaining lymphoid tissue samples from patients that are readily harvested from rodents. However, animal model data do suggest that these hormones may affect T-cell development, although the mechanism underlying these effects remains to be deciphered.

Effects of Neuroendocrine Hormones on Trafficking of T-Cell Progenitors

Another means by which GH increases thymic cellularity could be by increasing trafficking of T-cell progenitors to the thymus. Evidence for this effect has come through studies in the huPBL-SCID mouse model [50], in which human peripheral blood lymphocytes (huPBL) are engrafted into mice with severe combined immunodeficiency (SCID) to permit study of human lymphocytic function and possible therapeutic agents in vivo. This model has been particularly useful in evaluation of the effects of human GH on trafficking of human T cells, owing to an unusual and fortuitous observation that engrafted human T cells could re-enter the thymus, which suggested that GH had potential for augmentation of long-term peripheral T-cell engraftment [42]. Human GH enhanced adhesion of activated and resting human T cells to various human and murine substrates, including ICAM-1, through induction of ß2 and ß1 integrin expression, indicating that GH could increase adherence to endothelial tissue [51]. The GH also promoted T-cell migration. Interestingly, human but not ovine GH augmented engraftment of human T cells in the SCID thymus, further supporting the conclusion that one mechanism by which GH augments thymopoiesis is by increasing the trafficking of T cells.

74

Neuroimmunomodulation 1999;6:69–80

Effects of Neuroendocrine Hormones Directly on the Thymus

Another possible mechanism of the neuroendocrine hormones includes their effects on thymic epithelial cells. Expression of receptors, discussed in the next section, provides evidence that these hormones act directly on immune cells, but neuroendocrine hormones can stimulate secretion of additional hormones that ultimately influence cellular function and development. Thymulin production by human and rodent thymic epithelial cells increased following treatment with GH [52], IGF-1 [52], or PRL [53]. IGF-1 and PRL also caused proliferation of thymic epithelial cells. IGF-1 was demonstrated to mediate the effects of GH, as antibodies to IGF-1 abrogated the effects of GH. The hormones may also play a role in the protection of the thymocyte from apoptosis. Apoptosis plays a critical role in thymocyte selection during Tcell development [54]. There are data to suggest, at least with hematopoietic progenitor cells, that IGF-1 protects them from apoptosis [55]. More work needs to be performed to understand the potential role of GH and PRL on thymocyte apoptosis.

Effects of Hormones on the Peripheral Immune System

Another means by which hormones could promote Tcell reconstitution post-BMT would be by their effects on the mature cells in the periphery. Recently, evidence has been presented that immune cells have receptors for both PRL and GH. The numerous data suggesting that PRL plays a role in immune regulation are supported by the variety of immune cells in both humans and rodents that express a receptor for this hormone: human adult peripheral blood and cord blood lymphocytes [56]; human adult peripheral blood mononuclear cells [57]; human, murine [58], and rat [59] T and B lymphocytes; murine thymocytes; murine macrophages [58]; and human large granular lymphocytes, which includes the NK cell population [60]. However, surface PRL receptors have not been detected on freshly isolated rat NK cells, but could be induced by treatment of the animals with bromocriptine [61]. PRL receptors are found on stromal cells of lymphoid tissues, and T-cell activation upregulates expression of the PRL receptor [62]. Binding of 125I-PRL to human lymphocytes was competitively inhibited by hGH as effectively as with unlabeled hPRL, indicating that the hGH also binds the PRL receptor [56]. Specific receptors

Woody/Welniak/Richards/Taub/Tian/Sun/ Longo/Murphy

for PRL are found on Daudi cells, a human B-cell lymphoma line; K562, a human erythroleukemia line; EL-4, a murine thymoma line, and SP 2/0-Ag 14, murine plasmacytoma line [57], and PRL-dependent Nb2 cells, which express a mutant form that binds PRL with higher affinity than the wild-type long form [63]. Interestingly, 10 –6 M of cyclosporin A, which is in the therapeutic range for suppression of transplant rejection, completely inhibited binding of PRL by human splenic T and B cells suggesting that receptor expression is under the influence of NFAT [57]. Similar observations on the competition of cyclosporin A and PRL for a common binding site [60, 64] suggest a mechanism for immunosuppression by cyclosporin A as PRL appears to be necessary for maintenance of immunocompetence of T cells. Dysregulation of PRL levels and receptor expression in NZB mice is suggested to occur in autoimmune disease. However, a recent report has argued against an essential role for PRL in regulation of immune responses, as no immune defect was observed in mice with a targeted disruption of the PRL gene [16]. Murine thymocytes [65], human peripheral blood B cells, and to a lesser extent, T and NK cells express receptors for GH [66]. Both activated and resting human T cells express receptors for IGF-I [67]. Dwarf mouse strains also have deficiencies in their peripheral immune systems that include decreased numbers of peripheral white blood cells, reduced spleen and lymph node weights, prolonged rejection of skin allografts and GVH reactions of reduced intensity, and delayed antibody response to SRBC relative to normal littermates [37]. Decreased peripheral blood counts have been detected in all lineages (erythroid, lymphoid, myeloid, and platelet), and spleens are hypoplastic in Snell-Bagg mice. T cells of the immature CD4+CD8+ phenotype appeared in lymph nodes [42]. Treatment of the dwarf animals with GH partially restored these immune parameters to those of normal mice, which suggests that GH deficiency incompletely accounts for the observed immune defects of the dwarf mice [28]. GH acted synergistically with thyroxin [17] and normal lymphocytes [38] in restoration of antibody titers to SRBC and lymph node cell populations. Studies with ovine PRL and GH, which are specific for their respective receptors, have indicated that GH has little effect on the peripheral immune system, but PRL augments the number and function of mitogen-stimulated or antigen-specific peripheral T cells in dwarf mice following immunization [43]. Implantation of normal syngeneic thymus and pituitary glands into nude mice resulted in elevated PRL in serum and increased the proportion of CD4+ cells in lymph nodes, suggesting that PRL preferen-

tially stimulated the development of T-helper cells [68]. Therefore, PRL is apparently more important regarding development and maintenance of peripheral T-cell function than GH. PRL augments function of the peripheral immune system and primarily exerts these effects by direct action on T cells. Nb2 cells, a T-cell-like rat lymphoma, proliferate in response to mitogens when treated with PRL [69]. PRL in the physiological range increases human T-cell response to polyclonal stimuli, but is inhibitory when in pathological concentrations [70]. Lymphokine-active killers derived from T cells are activated with PRL [71]. The hormone can induce switching of T-cell receptor expression from the · chain to the Á chain [72]. That PRL stimulates peripheral T-cell function is also demonstrated in studies in which animals are depleted of the hormone by treatment with bromocriptine (BRC), an agonist of dopamine that inhibits PRL secretion, or by in vitro addition of PRL-specific antisera. BRC treatment of rodents resulted in inhibition of a number of T-celldependent reactions: contact sensitivity to DNCB, adjuvant arthritis, skin allograft rejection, antibody production to the T-dependent antigen SRBC, experimental allergic encephalitis [73], response to concanavalin A [74], localized graft-versus-host reactions, and mixed lymphocyte reactions [64]. Exogenous PRL [64, 74] or sygeneic pituitary grafts [73] reversed the suppressive effects of BRC. Binding of cyclosporin A, an immunosuppressive drug, to human T lymphocytes could be competitively inhibited with PRL, suggesting that the immunosuppressive effects of cyclosporin A could be mediated in part through competition with PRL for a common binding site [64]. Interestingly, BRC treatment of mice resulted in loss of binding sites for cyclosporin A on splenocytes [65]. Prolactin-specific antisera decrease the proliferation of murine lymphocytes stimulated with concanavalin A [74] or IL-2 [75] and of human peripheral blood mononuclear cells stimulated with concanavalin A or PHA [76]. It has been reported that human peripheral blood lymphocytes express mRNA related to the PRL transcript and secrete a peptide that is immunoreactive with polyclonal anti-PRL serum when stimulated with concanavalin A that may have autocrine function [76]. We have found that administration of recombinant human (r-h) PRL after syngeneic BMT in mice leads to an increased ability to mount a primary immune response. Mice received PRL post-BMT and were then challenged with an antigen, keyhole limpet hemocyanin (KLH) antigen with or without PRL. After 14 days the ability of the lymph node T cells to respond to the antigen were

Hormones and Bone Marrow Transplantation

Neuroimmunomodulation 1999;6:69–80

75

Table 3. Effects of r-hPRL on splenic T-cell mitogen responsiveness after syngeneic BMT

Group

PBS r-hPRL

Concanavalin A concentration, Ìg/ml 0 Ìg

1 Ìg

10 Ìg

1.0 1.0

3.29B0.45 9.74B3.53*

21.34B1.97 43.17B13.22

Balb/c mice received 850 cGy total body irradiation and then received 1 ! 106 syngeneic BMC i.v. The mice then received either PBS or 10 Ìg r-hPRL i.p. every other day. Concanavalin A responsiveness was assessed 14 days after BMT by 3H-thymidine incorporation, and stimulation indices were calculated based on the 3H-thymidine incorporation for 0 Ì/ml concanavalin A (227.3 B 41.1 and 245.7 B 13.5 cpm for PBS and r-hPRL, respectively). The data presented are the mean stimulation indices of 45 animals B SEM. * p ! 0.05, relative to stimulation index of PBS-treated animals.

assessed. It was found that treatment with PRL resulted in significant increases in proliferation [manuscript in preparation]. Similar effects on T-cell function were seen with regard to mitogen responsiveness of splenic to cells after BMT (table 3). These results indicate that PRL may be of use post-BMT to promote T-cell function. The effects of PRL on T cells may be indirectly mediated through modulation of secretion of thymic hormones. PRL can increase the synthesis and secretion of thymulin by thymic epithelial cells in primary mouse or human cells and augments thymic epithelial cell proliferation in vivo [53]. However, it is difficult to see how these thymic effects are related to the peripheral T-cell effects. Treatment of mice with BRC decreased, whereas human GH increased thymulin secretion. GH has less pronounced effects on the peripheral immune system than PRL [77], but may have subtle consequences on the differentiation of T cells. GH stimulated proliferation of human normal and hairy T-cell leukemia [78] and of murine T cells activated by anti-CD3 or concanavalin A [79], acted as an adjuvant for an inactivated tick-borne encephalitis vaccine [80], and acted synergistically with PRL [79]. The hormone augmented formation of cytotoxic lymphocytes, but not proliferation, when incubated with a mixed population of murine lymphocytes [81]. T-helper cells from bovine GH transgenic mice secreted significantly less IL-4 than cells from normal animals following injection with Staphylococcus aureus enterotoxin B [82]. The immunoglobulin isotype pattern of the transgenic mice was switched from IgG2a from the

76

Neuroimmunomodulation 1999;6:69–80

IgG1 of normal mice following immunization, and death from septic shock was associated with T-cell impairment in the transgenic animals. Thus, it appears that GH could influence T-cell differentiation. Although treatment of mice with IGF-1 increased spleen weights [46], response to bovine serum albumin, a T-cell-dependent antigen, was similar in diabetic and normal rats (diabetic animals have decreased IGF-1 levels), indicating that the diabetic animals still retained normal peripheral T-cell function, and suggesting that IGF-1 has little effect on the function of these cells [45]. IGF-1 did not enhance proliferation of murine T cells [79].

Neuroendocrine Effects on NK Cells

NK cells play a pivotal role in innate immunity and have been shown to be closely linked with T cells developmentally [50]. NK cells, also known as large granular lymphocytes based on their morphologic features, are CD3– CD16+CD56+ lymphocytes that spontaneously mediate the major histocompatibility complex-unrestricted killing of tumor and virally infected cells. Because of their cytotoxic activity and the variety of cytokines these cells produce, NK cells are an essential component of the innate immune response and may play a role in specific immunity. Studies in rodents with hypothalamic lesions or hypophysectomies suggested that decreased GH and PRL secretion suppress NK maturation and activity [29–31]. Destruction of the hypothalamus in mice resulted in increased tumor growth concurrent with decreased NK activity [30]. Treatment of hypophysectomized mice with GH restored NK function [31]. GH-deficient patients have reduced NK cell numbers and activity, compared to healthy controls [49, 83]. Short-term treatment with either GH or GH-releasing hormone did not restore NK function in patients with hypothalamic deficiencies [49]. A possible reduction in numbers of NK cells in GH-deficient children has been reported, but NK cells were not directly identified and enumerated [48]. Studies on the effects of PRL on NK cells are difficult to interpret and contradictory. NK cells have few specific surface markers [50], and because of this, data often have been gathered from experiments on heterogeneous mixtures of cells, sometimes predominantly T cells. Presented here is a discussion of the effects of PRL on NK cells (or activity), as distinguished from the lymphokine-activated killer (LAK) cells, which also include some IL-2-activated cytotoxic T cells.

Woody/Welniak/Richards/Taub/Tian/Sun/ Longo/Murphy

Current evidence indicates that the effects of PRL on NK cells are dependent on dose of the hormone. Suboptimal concentrations of the hormone synergize with IL-2 to augment in vitro activation [71, 84], but concentrations 5–10 times higher than physiological levels are inhibitory [70]. Physiological concentrations of PRL stimulate the activity and increase DNA, RNA and protein synthesis in purified NK cells, whereas higher levels of the hormone have been associated with decreased NK activity [85]. It is unclear whether hyperprolactinemia, frequently observed in some autoimmune diseases, is associated with defects in NK cell function [86–88]. NK cell activity was significantly reduced in untreated hyperprolactinemic women compared to age-matched controls, but was restored in patients that were treated with BRC [86]. Decreased NK activity was observed in cancer patients who had elevated serum PRL levels following treatment with morphine [88]. In contrast, Matera et al. [87] found no differences in NK activity between hyperprolactinemic patients and normal age-matched controls. The discrepancy among these reports may be partly accounted for by differences in preparations (and definition) of NK cells from PBL and the hormone preparations (i.e., purified vs. recombinant). The observation that NK cell activity, derived from normal PBL, was not significantly enhanced or suppressed after 18 h incubation with PRL (1–1,000 ng/ml) [87] suggests that other endocrine factors suppressed NK cell activity and numbers in hyperprolactinemic patients in the earlier study. There is an intriguing tentative link of depressed NK activity, dysregulation of PRL levels, and growth of oral carcinomas, the progression of endometriosis [89], or aging [89–91], but the causal relationship between aberrant PRL production and decreased NK activity to disease remains to be demonstrated.

Neuroendocrine Hormones and Function of Monocytes and Macrophages

[95]. The effects of neuroendocrine hormones on macrophages have been studied in infectious disease models in which macrophage function is crucial to clearance of the infection and survival of the host. PRL increased survival of Salmonella typhimurium-infected mice [95], and GH prolonged survival in Salmonella-infected hypophysectomized rats [96]. Possibly, increased survival could have resulted from suppression of TNF-· production by monocytes or macrophages, because human GH has been reported to decrease TNF-· production by human monocytes that have been stimulated with LPS [97]. PRL in combination with IFN-Á or TNF-· increased survival of mice infected with Toxoplasma gondii, and the hormone together with TNF-· decreased occurrence of brain cysts [98]. PRL restored release of IL-1ß and IL-6 by macrophages and increased survival of sepsis after hemorrhage [99]. The hormone or metoclopramide, a dopamine antagonist that increases circulating levels of PRL, counteracted the suppressive effects of sepsis on expression of mRNA of the proinflammatory cytokines IL-1ß, IL-6, and TNF-· by macrophages [100]. Phagocytosis of Candida albicans by macrophages is increased by PRL treatment [92]. The action of PRL on macrophage function may be mediated indirectly through T cells that release cytokines that in turn activate macrophages [101]. Production of IFN-Á by T cells is impaired in BRC-treated mice, leading to reduced macrophage activity against Listeria monocytogenes and Mycobacterium bovis. These data therefore indicate that neuroendocrine hormones such as GH and PRL may promote T-cell responses through their effects on monocytes/macrophages and vice versa.

Acknowledgments We are extremely grateful to Ms. Laura Knott for excellent secretarial services. We also thank Mr. Steve Stull for technical services.

Neuroendocrine hormones have been reported to augment macrophage activity. As macrophages play a key role in antigen presentation and the generation of primary immune responses, another means by which GH and PRL may promote antigen-specific T-cell responses may be through their effects on macrophages. Superoxide anion production by isolated macrophages is increased after in vivo administration of PRL [92] or in vitro treatment with GH [93], and GH is a potent macrophage chemoattractant [94]. PRL increases chemotaxis of granulocytes

Hormones and Bone Marrow Transplantation

Neuroimmunomodulation 1999;6:69–80

77

References 1 Murphy W, Longo D: The potential role of NK cells in the separation of graft-versus-tumor effects from graft-versus-host disease after allogeneic bone marrow transplantation. Immunol Rev 1997;157:167–176. 2 Bazan J: Haemopoietic receptors and helical cytokines. Immunol Today 1990;11:350–354. 3 Gouilleux F, Pallard C, Dusanter-Fourt I, Wakao H, Haldosen L, Norstedt D, Groner B: Prolactin, growth hormone, erythropoietin and granulocyte-macrophage colony stimulating factor induce MGF-Stat5 DNA binding activity. EMBO J 1995;14:2005–2013. 4 Pellegrini I, Lebrun J, Ali S, Kelly P: Expression of prolactin and its receptor in human lymphoid cells. Mol Endocrinol 1992;6:1023– 1031. 5 Bellone G, Geuna M, Carbone A, Silvestri S, Foa R, Emaneulli G, Matera L: Regulatory action of prolactin on the in vitro growth of CD34+ve human hemopoietic progenitor cells. J Cell Physiol 1995;163:221–231. 6 Gagnerault M-C, Postel-Vinay M, Dardenne M: Expression of growth hormone receptors in murine lymphoid cells analyzed by flow cytofluorometry. Endocrinology 1996;137:1719– 1726. 7 Gala R: Prolactin and growth hormone in the regulation of the immune system. Proc Soc Exp Biol Med 1991;198:513–527. 8 Kooijman R, Malur A, Van Buul-Offers S, Hooghe-Peters E: Growth hormone expression in murine bone marrow cells is independent of the pituitary transcription factor pit-1. Endocrinology 1997;138:3949–3955. 9 Delhase M, Vergani P, Malur A, Hooghe-Peters E, Hooghe R: The transcription factor pit1/ghf-1 is expressed in hemopoietic and lymphoid tissues. Eur J Immunol 1993;23:951– 955. 10 Gellersen B, Kempf R, Telgmann R, DiMattia G: Nonpituitary human prolactin gene transcription is independent of pit-1 and differentially controlled in lymphocytes and in endometrial stroma. Mol Endocrinol 1994;8:356– 373. 11 Gala R: The influence of thyroxine, growth hormone and prolactin alone and in combination on the production of prolactin-like activity by splenocytes from Snell dwarf mice. Life Sci 1995;57:113–122. 12 Melen L, Hennen G, Dullaart R, Heinen E, Igout A: Both pituitary and placental growth hormone transcripts are expressed in human peripheral blood mononuclear cells. Clin Exp Immunol 1997;111:336–340. 13 Han V, D’Ercole A, Lund P: Cellular localization of somatomedin (insulin-like growth factor) messenger RNA in the human fetus. Science 1987;236:193–197. 14 Murphy W, Durum S, Anver M, Longo D: Immunologic and hematologic effects of neuroendocrine hormones. Studies on DW/J dwarf mice. J Immunol 1992;148:3799–3805.

78

15 Nagy E, Berczi I: Hypophysectomized rats depend on residual prolactin for survival. Endocrinology 1991;128:2776–2784. 16 Horseman N, Zhao W, Montecino-Rodriquez E, Tanaka M, Nakashima K, Engle S, Smith F, Markoff E, Dorshkind K: Defective mammopoiesis, but normal hematopoiesis, in mice with a targeted disruption of the prolactin gene. EMBO J 1997;16:6926–6935. 17 Baroni C, Fabris N, Bertoli G: Effects of hormones on development and function of lymphoid tissues. Synergistic action of thyroxin and somatotropic hormone in pituitary dwarf mice. Immunology 1969;17:303–314. 18 Montecino-Rodriquez E, Clark R, Johnson A, Collins L, Dorshkind K: Defective B cell development in Snell dwarf (dw/dw) mice can be corrected by thyroxine treatment. J Immunol 1996;157:3334–3340. 19 Kotzmann H, Riedl M, Clodi M, Barnas U, Kaider A, Hocker P, Luger A: The influence of growth hormone substitution therapy on erythroid and myeloid progenitor cells and on peripheral blood cells in adult patients with growth hormone deficiency. Eur J Clin Invest 1996;26:1175–1181. 20 Nagy E, Berczi I: Pituitary dependence of bone marrow function. Br J Haematol 1989;71:457– 462. 21 Blazar B, Brennan C, Broxmeyer H, Shultz L, Vallera D: Transgenic mice expressing bovine growth hormone (bGH) or human GH releasing hormone (hGRH) have increased splenic progenitor cell colony formation and DNA synthesis in vitro and in vivo. Exp Hematol 1995; 23:1397–1406. 22 Tsarfaty G, Longo D, Murphy W: Human insulin-like growth factor I exerts hematopoietic growth-promoting effects after in vivo administration. Exp Hematol 1994;22:1273–1277. 23 Murphy W, Tsarfaty G, Longo D: Growth hormone exerts hematopoietic growth-promoting effects in vivo and partially counteracts the myelosuppressive effects of azidothymidine. Blood 1992;80:1443–1447. 24 Merchav S, Tatarsky I, Hochberg Z: Enhancement of human granulopoiesis in vitro by biosynthetic insulin-like growth factor I/somatomedin C and human growth hormone. J Clin Invest 1988;81:791–797. 25 Ratajczak M, Kuczynski W, Onodera K, Moore J, Ratajczak J, Kregenow D, DeRiel K, Gewirtz A: A reappraisal of the role of insulinlike growth factor I in the regulation of human hematopoiesis. J Clin Invest 1994;94:320– 327. 26 Bellone G, Astarita P, Artusio E, Silvestri S, Mareschi K, Turletti A, Buttiglieri S, Emanuelli G, Matera L: Bone marrow stroma-derived prolactin is involved in basal and platelet-activating factor-stimulated in vitro erythropoiesis. Blood 1997;90:21–27. 27 Pierpaoli W, Sorkin E: Relationship between thymus and hypophysis. Nature 1967;215: 834–837.

Neuroimmunomodulation 1999;6:69–80

28 Pierpaoli W, Sorkin E: Hormones and immunologic capacity. I. Effect of heterologous antigrowth hormone (ASTH) antiserum on thymus and peripheral lymphatic tissue in mice. Induction of a wasting syndrome. J Immunol 1968; 101:1036–1043. 29 Cross R, Markesbery W, Brooks W, Roszman T: Hypothalamic-immune interactions: Neuromodulation of natural killer activity by lesioning of the anterior hypothalamus. Immunology 1984;51:399–405. 30 Forni G, Bindoni M, Santoni A, Belluardo N, Marchese A, Giovarelli M: Radiofrequency destruction of the tuberoinfundibular region of the hypothalamus permanently abrogates NK cell activity in mice. Nature 1983;306:181– 184. 31 Saxena Q, Saxena R, Adler W: Regulation of natural killer activity in vivo. III. Effect of hypophysectomy and growth hormone treatment on the natural killer activity of the mouse spleen cell population. Int Arch Allergy Appl Immunol 1982;67:169–174. 32 Berczi I, Nagy E, Kovacs K, Horvath E: Regulation of humoral immunity in rats by pituitary hormones. Acta Endocrinol (Copenh) 1981;98: 506–513. 33 Nagy E, Berczi I, Friesen H: Regulation of immunity in rats by lactogenic and growth hormones. Acta Endocrinol (Copenh) 1983;102: 351–357. 34 Voss J, Rosenfeld M: Anterior pituitary development: Short tales from dwarf mice. Cell 1992;70:527–530. 35 Pierpaoli W, Baroni C, Fabris N, Sorkin E: Hormones and immunological capacity. II. Reconstitution of antibody production in hormonally deficient mice by somatotropic hormone, thyrotropic hormone and thyroxin. Immunology 1969;16:217–230. 36 Baroni C: Thymus, peripheral lymphoid tissues and immunological responsiveness of the pituitary dwarf mouse. Experientia 1967;23:282– 283. 37 Fabris N, Pierpaoli W, Sorkin E: Hormones and the immunological capacity. III. The immunodeficiency disease of the hypopituitary Snell-Bagg dwarf mouse. Clin Exp Immunol 1971;9:209–225. 38 Fabris N, Pierpaoli W, Sorkin E: Hormones and the immunological capacity. IV. Restorative effects of developmental hormones or of lymphocytes on the immunodeficiency syndrome of the dwarf mouse. Clin Exp Immunol 1971;9:227–240. 39 Duquesnoy R: Immunodeficiency of the thymus-dependent system of the Ames dwarf mouse. J Immunol 1972;108:1578–1590. 40 Dumont F, Robert F, Bischoff P: T and B lymphocytes in pituitary dwarf Snell-Bagg mice. Immunology 1979;38:23–31. 41 Cross R, Bryson J, Roszman T: Immunological disparity in the hypopituitary dwarf mouse. J Immunol 1992;148:1347–1352.

Woody/Welniak/Richards/Taub/Tian/Sun/ Longo/Murphy

42 Murphy W, Durum S, Longo D: Role of neuroendocrine hormones in murine T cell development. Growth hormone exerts thymopoietic effects in vivo. J Immunol 1992;149:3851– 3857. 43 Murphy W, Durum S, Longo D: Differential effects of growth hormone and prolactin on murine T cell development and function. J Exp Med 1993;178:231–236. 44 Kelley K, Brief S, Westly H, Novakofski J, Bechtel P, Simon J, Walker E: GH3 pituitary adenoma cells can reverse thymic aging in rats. Proc Natl Acad Sci USA 1986;83:5663–5667. 45 Binz K, Joller P, Froesch P, Binz H, Zapf J, Froesch E: Repopulation of the atrophied thymus in diabetic rats by insulin-like growth factor I. Proc Natl Acad Sci USA 1990;87:3690– 3694. 46 Clark R, Strasser J, McCabe S, Robbins K, Jardieu P: Insulin-like growth factor-1 stimulation of lymphopoiesis. J Clin Invest 1993;92:540– 548. 47 Gjerset R, Yeargin J, Volkman S, Vila V, Arya J, Haas M: Insulin-like growth factor-I supports proliferation of autocrine thymic lymphoma cells with a pre-T cell phenotype. J Immunol 1990;145:3497–3501. 48 Gupta S, Fikrig S, Noval M: Immunological studies in patients with isolated growth hormone deficiency. Clin Exp Immunol 1983;54: 87–90. 49 Kiess W, Malozowski S, Gelato M, Butenand O, Doerr H, Crisp B, Eisl E, Maluish A, Belohradsky B: Lymphocyte subset distribution and natural killer activity growth hormone deficiency before and during short-term treatment with growth hormone releasing hormone. Clin Immunol Immunopathol 1988;48:85–94. 50 Murphy W, Taub D, Longo D: The huPBLSCID mouse as a means to examine human immune function in vivo. Semin Immunol 1996;8:233–241. 51 Taub D, Tsarfaty G, Lloyd A, Durum S, Longo D, Murphy W: Growth hormone promotes human T cell adhesion and migration to both human and murine matrix proteins in vitro and directly promotes xenogeneic engraftment. J Clin Invest 1994;94:293–300. 52 Timsit J, Savino W, Safieh B, Chanson P, Gagnerault M-C, Bach J-F, Dardenne M: Growth hormone and insulin-like growth factor-I stimulate hormonal function and proliferation of thymic epithelial cells. J Clin Endocrinol Metab 1992;75:183–188. 53 Dardenne M, Savino W, Gagnerault M-C, Itoh T, Bach J-F: Neuroendocrine control thymic hormonal production. I. Prolactin stimulates in vivo and in vitro the production of thymulin by human and murine thymic epithelial cells. Endocrinology 1989;125:3–12. 54 Connoy L, Alexander D: The role of intracellular signalling pathways regulating thymocyte and leukemic T cell apoptosis. Leukemia 1996; 10:1422–1435. 55 Minshall C, Arkins S, Freund G, Kelley K: Requirement for phosphatidyl 3)-kinase to protect hemopoietic progenitors against apoptosis depends upon the extracellular survival factor. J Immunol 1996;156:939–947.

Hormones and Bone Marrow Transplantation

56 Bellussi G, Muccioli G, Ghè C, Di Carlo R: Prolactin binding sites in human erythrocytes and lymphocytes. Life Sci 1987;41:951–959. 57 Russell D, Kibler R, Matrisian L, Larson D, Poulos B, Magun B: Prolactin receptors on human T and B lymphocytes: Antagonism of prolactin binding by cyclosporine. J Immunol 1985;134:3027–3031. 58 Gala R, Shevach E: Identification by analytical flow cytometry of prolactin receptors on immunocompetent cell populations in the mouse. Endocrinology 1993;133:1617–1623. 59 Viselli S, Mastro A: Prolactin receptors are found on heterogeneous subpopulations of rat splenocytes. Endocrinology 1993;132:571– 576. 60 Matera L, Muccioli G, Cesano A, Bellussi G, Genazzani E: Prolactin receptors on large granular lymphocytes: Dual regulation by cyclosporine A. Brain Behav Immun 1988;2:1–10. 61 Chambers W, Amoscato A, Smith M, Kenniston T, Herberman R, Appasamy P: Prolactin receptor expression by rat NK cells. Nat Immun 1995;14:145–156. 62 Savino W, Dardenne M: Immune-neuroendocrine interactions. Immunol Today 1995;16: 318–321. 63 Ali S, Pellegrini I, Kelly P: A prolactin-dependent immune cell line (Nb2) expresses a mutant form of prolactin receptor. J Biol Chem 1991;266:20110–20117. 64 Hiestand P, Mekler P, Nordmann R, Grieder A, Permmongkol C: Prolactin as a modulator of lymphocyte responsiveness provides a possible mechanism of action of cyclosporin. Proc Natl Acad Sci USA 1986;83:2599–2603. 65 Arrenbrecht S: Specific binding of growth hormone to thymocytes. Nature 1974;252:255– 257. 66 Badolato R, Bond H, Valerio G, Petrella A, Morrone G, Waters M, Venuta S, Tenore A: Differential expression of surface membrane growth hormone receptor on human peripheral blood lymphocytes detected by dual fluorochrome flow cytometry. J Clin Endocrinol Metab 1994;79:984–990. 67 Tapson V, Boni-Schnetzler M, Pilch P, Center D, Berman J: Structural and functional characterization of the human T lymphocyte receptor for insulin-like growth factor I in vitro. J Clin Invest 1988;82:950–957. 68 Gaufo G, Diamond M: Prolactin increases CD4/CD8 cell ratio in thymus-grafted congenitally athymic nude mice. Proc Natl Acad Sci USA 1996;93:4165–4169. 69 Bates L, Grove D, Mastro A: Mechanisms of activation and suppression in rat Nb 2 lymphoma cells: A model for interactions between prolactin and the immune system. Exp Cell Res 1995;218:567–572. 70 Matera L, Cesano A, Bellone G, Oberholtzer E: Modulatory activity of prolactin on the resting and mitogen-induced activity of T, B, and NK lymphocytes. Brain Behav Immun 1992;6: 409–417.

71 Cesano A, Oberholtzer E, Contarini, Geuna M, Bellone G, Matera L: Independent and synergistic effects of interleukin-2 and prolactin on development of T- and NK-derived LAK effectors. Immunopharmacology 1994;28:67–75. 72 Hosokawa Y, Yang M, Kaneko S, Tanaka M, Nakashima K: Prolactin induces switching of T-cell receptor gene expression from a to g in rat Nb2 pre-T lymphoma cells. Biochem Biophys Res Commun 1996;220:958–962. 73 Nagy E, Berczi I, Wren G, Asa S, Kovacs K: Immunomodulation by bromocriptine. Immunopharmacology 1983;6:231–243. 74 Bernton E, Hartmann D, Holaday J: Antibody to mouse prolactin inhibits murine lymphocyte responses to T-cell growth factors (TCGF). J Leukoc Biol 1987;42:335. 75 Clevenger C, Russell D, Appasamy P, Prystowsky M: Regulation of interleukin 2-driven T-lymphocyte proliferation by prolactin. Proc Natl Acad Sci USA 1990;87:6460–6464. 76 Sabharwal P, Glaser R, Lafuse W, Varma S, Liu Q, Arkins S, Kooijman R, Kutz L, Kelley K, Malarkey W: Prolactin synthesized and secreted by human peripheral blood mononuclear cells: An autocrine growth factor for lymphoproliferation. Proc Natl Acad Sci USA 1992;89:7713–7716. 77 Murphy W, Rui H, Longo D: Effects of growth hormone and prolactin immune development and function. Life Sci 1995;57:1–14. 78 Mercola K, Cline M, Golde D: Growth hormone stimulation of normal and leukemic human T-lymphocyte proliferation in vitro. Blood 1981;58:337–340. 79 Postel-Vinay M-C, de Mello Coelho V, Gagnerault M-C, Dardenne M: Growth hormone stimulates the proliferation of activated mouse T lymphocytes. Endocrinology 1997;138:1816– 1820. 80 Stephenson J, Lee J, Bailey N, Shepherd A, Melling J: Adjuvant effect of human growth hormone with an inactivated flavivirus vaccine. J Infect Dis 1991;164:188–191. 81 Snow E, Feldbush T, Oaks J: The effect of growth hormone and insulin upon MLC responses and the generation of cytotoxic lymphocytes. J Immunol 1981;126:161–164. 82 Gonzalo J, Mazuchelli R, Mellado M, Frade J, Carrera A, von Kobbe C, Merida I, Martinez AC: Enterotoxin shock protection and deficient T-helper 2 cytokine production in growth hormone transgenic mice. J Immunol 1996;157: 3298–3304. 83 Span J, Pieters G, Smals A, Koopmans P, Kloppenborg P: Number and percentage of NK cells are decreased in growth hormone-deficient adults. Clin Immunol Immunopathol 1996;78:90–92. 84 Matera L, Bellone G, Lebrun J, Kelly P, Hooghe Peters E, Di Celle P, Foa R, Contarini M, Avanzi G, Asnaghi V: Role of prolactin in the in vitro development of interleukin-2-driven anti-tumoural lymphokine-activated killer cells. Immunology 1996;89:619–626. 85 Matera L, Bellone G, Cesano A: Prolactin and the neuroimmune network. Adv Neuroimmunol 1991;1:158–172.

Neuroimmunomodulation 1999;6:69–80

79

86 Gerli R, Rambotti P, Nicoletti I, Orlandi S, Migliorati G, Riccardi C: Reduced number of natural killer cells in patients with pathological hyperprolactinemia. Clin Exp Immunol 1986; 64:399–406. 87 Matera L, Ciccarelli E, Cesano A, Veglia F, Miola C, Camanni F: Natural killer activity in hyperprolactinemic patients. Immunopharmacology 1989;18:143–146. 88 Provinciali M, Di Stefano G, Stronati S, Raffaeli W, Pari G, Fabris N: Role of prolactin in the modulation of NK and LAK cell activity after short- or long-term morphine administration in neoplastic patients. Int J Immunopharmacol 1996;18:577–586. 89 Provinciali M, Di Stefano G, Muzzioli M, Garzetti G, Ciavattini A, Fabris N: Relationship between 17ß-estradiol and prolactin in the regulation of natural killer cell activity during progression of endometriosis. J Endocrinol Invest 1995;18:645–652. 90 Chakraborty A, Chakraborty N, Chattophdhyay U: Prolactin response of NK cells, but not of LAK cells, is deficient in patients with carcinoma of oral cavity and during aging. Int J Cancer 1996;66:65–69.

80

91 Muzzioli M, Mocchegiani E, Bressani N, Bevilacqua P, Fabris N: In vitro restoration by thymulin of NK actvity of cells from old mice. Int J Immunopharmacol 1992;14:57–61. 92 Chen Y, Johnson A: In vivo activation of macrophages by prolactin from young and aging mice. Int J Immunopharmacol 1993;15:39– 45. 93 Edwards C III, Ghiasuddin S, Schepper J, Yunger L, Kelley K: A newly defined property of somatotropin: Priming of macrophages for production of superoxide anion. Science 1988;239: 769–771. 94 Wiedermann C, Reinisch N, Brausteiner H: Stimulation of monocyte chemotaxis by human growth hormone and its deactivation by somatostatin. Blood 1993;82:954–960. 95 Di Carlo R, Meli R, Galdiero M, Nuzzo I, Bentivoglio C, Romano Carratelli C: Prolactin protection against lethal effects of Salmonella typhimurium. Life Sci 1993;53:981–989. 96 Edwards C III, Yunger L, Lorence R, Dantzer R, Kelley K: The pituitary gland is required for protection against lethal effects of Salmonella typhimurium. Proc Natl Acad Sci USA 1991; 88:2274–2277.

Neuroimmunomodulation 1999;6:69–80

97 Haeffner A, Thieblemont N, Déas O, Marelli O, Charpentier B, Senik A, Wright S, Haeffner-Cavaillon N, Hirsch F: Inhibitory effect of growth hormone on TNF-· secretion and nuclear factor-ÎB translocation in lipopolysaccharide-stimulated human monocytes. J Immunol 1997;158:1310–1314. 98 Benedetto N, Folgore A, Galdiero M, Meli R, Di Carlo R: Effect of prolactin, rIFN-Á or rTNF-· in murine toxoplasmosis. Pathol Biol 1995;43:395–400. 99 Zellweger R, Zhu X-H, Wichmann M, Ayala A, DeMaso C, Chaudry I: Prolactin administration following hemorrhagic shock improves mortality from subsequent sepsis. J Immunol 1996;157:5748–5754. 100 Zhu X-H, Zellweger R, Wichmann M, Ayala A, Chaudry I: Effects of prolactin and metoclopramide on macrophage gene expression in late sepsis. Cytokine 1977;9:437–446. 101 Bernton E, Meltzer M, Holaday J: Suppression of macrophage activation and T-lymphocyte function in hypoprolactinemic mice. Science 1988;239:401–404.

Woody/Welniak/Richards/Taub/Tian/Sun/ Longo/Murphy

Neuroimmunomodulation 1999;6:81–89

Extracellular ATP: A Further Modulator in Neuroendocrine Control of the Thymus Luiz A. Alves a Robson Coutinho-Silva b Wilson Savino a a Laborato ´ rio de Pesquisas sobre o Timo, Departamento de Imunologia, Instituto Oswaldo Cruz, Fundaça˜o Oswaldo Cruz, Rio de Janeiro, e b Laborato´rio de Imunobiofı´sica, Instituto de Biofı´sica Carlos Chagas Filho, Universidade Federal do Rio de Janeiro, Brazil

Key Words ATP W Purinergic receptors W Thymic epithelial cells W Thymic nurse cells

Abstract It is well established that the process of thymocyte differentiation and maturation occurs in the thymus, where cell-to-cell communication is essential for providing the messages to T-cell precursors. At least two pathways are important for such communication: one via membrane surface molecules and the other via soluble mediators such as cytokines and some hormones. Recently, the presence of receptors for extracellular ATP has been demonstrated on thymocytes and microenvironment cells, and putative functions for this molecule have been proposed. Herein we focus on the recent evidence which supports the view of extracellular ATP and some related nucleotides as novel intrathymic signal molecules. In addition, we discuss the possible physiological implications of such purinergic receptors for the physiology of the thymus.

ABC

© 1999 S. Karger AG, Basel 1021–7401/99/0062–0081$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: http://BioMedNet.com/karger

Introduction

The main function of the thymus is to convey the essential signals for thymocyte differentiation, so that mature T lymphocytes exiting the organ will be able to colonize specific (T-dependent) areas of peripheral lymphoid organs [1]. These signals are provided by the thymic microenvironment, a tridimensional network composed of distinct cell types, such as epithelial cells, macrophages and dendritic cells, as well as extracellular matrix [3,4]. Cell-to-cell communication is essential for transmitting the signals for T-lymphocyte differentiation. The communication of the thymic microenvironment with thymocytes, which is bidirectional, takes place through at least two pathways: one through classical adhesion molecules and the major histocompatibility complex gene products, and the other through soluble factors such as cytokines and thymic hormones [2–4]. Recently, we and others have demonstrated the presence of receptors for extracellular ATP on thymocytes and microenvironment cells [5–8]. Extracellular ATP and some related molecules may act as new signaling molecules in the thymus, working as a neurotransmitter or paracrine/autocrine extracellular messenger. It may appear rather surprising that the main intracellular source of

Luiz A. Alves Laborato´rio de Pesquisas sobre o Timo Departamento de Imunologia – Instituto Oswaldo Cruz, Fundaça˜o Oswaldo Cruz Av. Brasil 4365 – Manguinhos, Rio de Janeiro 21045-900 (Brazil) E-Mail [email protected]

Table 1. Main characteristics of purinoceptors

Receptor subtype

Main agonist

Antagonist

Species

Signal transduction

Tissue and cellular distribution

Ref.

P2X1

2-MeSATP

PPADS Suramin

Rat

Ligand-gated channel

Smooth muscle

67

P2X2

ATP = 2-MeSATP = ATPgS

PPADS Suramin

Rat

Ligand-gated channel

Brain, adrenal medulla, pituitary

68

P2X3

2-MeSATP = ATP

PPADS Suramin

Rat

Ligand-gated channel

Sensory ganglia

69, 70

P2X4

ATP

None

Rat

Ligand-gated channel

Wide

71, 72, 73, 74

P2X5

ATP

PPADS Suramin

Rat

Ligand-gated channel

Sensory ganglia, cervical ganglion

75

P2X6

ATP

None

Rat

Ligand-gated channel

Wide

76

P2X7 (P2z)

BzATP

Ox-ATP

Rat, human

Ligand-gated channel

Immune system; some tumor cells

77, 78

P2Y1 (P2y)

2-MeSATP

Reactive blue 2

Chicken, rat, mouse, G-protein coupled bovine, human

Wide

79

P2Y2 (P2u)

ATP = UTP

NT

Mouse, rat, human

G-protein coupled

Very wide

80, 81

P2Y3

ADP

NT

Chicken

G-protein coupled

Brain, spleen, kidney, 82, 83 lung, spinal cord

P2Y4

UTP 11 ATP

NT

Rat

G-protein coupled

Placenta

84, 85

P2Y5

ATP

NT

Chicken

G-protein coupled

T lymphocyte

86

P2Y6

UDP

Reactive blue 2, Rat Suramin

G-protein coupled

Wide

87, 88

P2Y7

ATP = 2-Cl-ATP = ATPgS

Suramin

Human

G-protein coupled

Wide

89

P2Y8

ATP = UTP = CTP = GTP = ITP

Suramin

Frog

G-protein coupled

Neural plate, tailbud 90

·,ß-meATP, ·,ß-methyleno-ATP; 2-MeSATP = 2-methylthio-ATP; BzATP = 2),3)-(4-benzoyl)benzoyl-ATP; Ox-ATP = oxidized ATP; PPADS = pyridoxol phosphate-6-azophenyl-2),4)-disulfonic acid; NT = not tested. The old classification of P2 receptors is indicated in parenthesis.

energy can also function as a ‘local hormone’. However, there are other internal metabolites that have similar functions, such as glutamate and glycine. Receptors for nucleotides were found in all mammalian tissues thus far investigated, suggesting an ubiquitous presence and a relevant role in cell physiology [8,9]. Interestingly, some unicellular organisms and invertebrate species express ATP and AMP nucleotide receptors, illustrating that the use of nucleotides as signal molecules may be a very ancient biological issue [10, 11]. In vertebrate species the effects of ATP and other analogues are mediated by a family of so-called purinergic

82

Neuroimmunomodulation 1999;6:81–89

receptors, that are currently classified as P1 and P2 [8, 9, 12]. The former is a receptor for adenosine and has four subtypes that have been cloned so far [12], and are also named as A1, A2a, A2b and A3. The latter is a receptor for phosphorylated nucleotides that is subdivided into at least five receptors: P2X and P2Z, receptors that are directly linked to ion channels, and P2U, P2T and P2Y, which are linked to a G-protein pathway [8, 9]. Recently, these receptors have been cloned, and a new classification based on sequence homology has emerged that is divided into two classes – P2X and P2Y (see table 1 for comparison). Seven P2X and eight P2Y receptors have been

Alves/Coutinho-Silva/Savino

cloned from vertebrate species. The P2X receptors are ionotropic receptors, i.e, ligand-gated ion channels, and P2Y are metabotropic receptors, i.e., G-protein coupled. The common feature shared by P2 receptors is the rise of intracellular calcium after ligand binding. An important difference between these receptors is the fact that the P2X7 receptor is associated with a large pore with low selectivity, allowing passage of molecules, that range from 300 to 900 daltons in different cell types of the immune system and some tumor cell lines [8, 9]. Even though several of these receptors have been cloned, it is useful to classify them using characteristics such as agonist potency, electrophysiological properties, tissue distribution and signal transduction mechanisms. It is important to stress that similar phenomena occur with other receptors due to a large amount of data generated by molecular biology. Then it is recommended that a new classification be based on structural and pharmacological information, such as the potency rank of several agonists [13]. Activation of ATP receptors increases the secretion of several hormones such as insulin, glucacon, luteinizing hormone, catecholamines and steroids, and also yields several effects on the cardiovascular system [8, 14, 15]. In this review we will focus on P2 purinoreceptors in the thymus, and the potential role of extracellular ATP (ATPe) as a further neuroendocrine signal molecule.

Pharmacological and Functional Characterization of Purinoreceptors in Thymic Microenvironmental Cells

The first evidence for ATPe receptors was recently obtained in one non-transformed rat thymic epithelial cell line, which has properties of normal epithelial cells of the subcapsular and medullary region [16]. In this cell line, ATPe increases the production of prostaglandin E2. However, this receptor is antagonized by GTP, which is not an antagonist for known P2 purinoreceptors. Thus, the ATP receptor expressed by this TEC line remains to be fully characterized. More recently, we have characterized a rise of intracellular calcium triggered by ATPe in several TEC lines as well as primary culture of mouse TEC (fig. 1). According to a pharmacological classification, the receptor responsible for the rise of intracellular calcium in most of the studied cells is the P2Y2 (also named P2u) receptor, since the following rank of potency was obtained: ATP = UTP 1 ADP 11 adenosine and cAMP [Alves et al., in preparation]. Apparently, the P2Y2 receptor subtype is the most

Extracellular ATP in the Thymus

Fig. 1. Effect of ATP on intracellular calcium in thymic epithelial cells. a, d Representative profiles after more than five distinct exper-

iments. Intracellular Ca2+ levels are clearly up-regulated when the various TEC preparations are exposed to ATPe. a and c are derived from two mouse TEC lines (respectively named 2BH4 and IT76M1), b is the IT45-R1 rat TEC line, whereas d corresponds to primary TEC cultures derived from thymic nurse cell complexes. Arrows indicate the time when cells were exposed to ATP at a final concentration of 5 ÌM. Bars indicate 100 s.

frequently expressed in TEC, although some TEC lines appear to express other purinoreceptors, which are presently under investigation. The other thymic microenvironmental cell type which expresses P2 purinergic receptors is the phagocytic cell of the thymic reticulum (P-TR). This cell shares characteristics with macrophages and dendritic cells such as: inter-

Neuroimmunomodulation 1999;6:81–89

83

leukin (IL)-1 and IL-2 receptors, CR3 complement receptor, phagocytosis of IgG-opsonized sheep erythrocytes, few lysosomes, dendritic shape, high levels of class II MHC molecules on their surface, and positive staining for peroxidase. P-TR express at least one P2 purinergic receptor named P2X7 (also named P2z) [5], which, as previously mentioned, is associated with large membrane pores. This receptor was identified by electrophysiological experiments and dye uptake assays, and has characteristics very similar to the P2X7 purinoceptors found in peritoneal macrophages and spleen dendritic cells [5]. Yet, since peritoneal macrophages have at least three different P2 subtypes [23], it is possible that P-TR also express other P2 purinergic receptors. In the same vein, it is also likely that thymic macrophages and dendritic cells express P2 purinergic receptors. Nevertheless, to our knowledge, no study has been done in this respect.

Characterization and Effects of ATP in Lymphocytes

ATPe has different effects on thymocytes and peripheral lymphocytes. Depending on the dose, ATPe may induce blastogenesis of PMA-treated medullary thymocytes [17], be involved in mitogenic stimulation and down-regulation of L-selectin in human peripheral T cells [18, 19], and have lytic effects on thymocytes, and mouse and human T and B peripheral lymphocytes [5,18, 20– 24]. The permeabilizing effect in these cells was observed by measuring changes in the intracellular calcium concentrations, depolarization of the plasma membrane, and membrane permeabilization to low-molecular-weight fluorescence dyes such as TMA-DPH (Mr 290) and ethidium bromide (Mr 314) [20, 25]. The depolarization of the plasma membrane was associated with the ATPe-activated nonselective cation ion channels [22, 26, 27] and the permeabilizing effect of ATPe has been claimed as a consequence of the ion channel/pore opening [8, 22]. Recently, the depolarizing current was resolved at the single-channel level using the patch-clamp technique in outside-out configuration in macrophages and phagocytic cells of the thymic reticulum [6] as well as in peripheral lymphocytes [28]. In these cell types, a ligand-gated ion channel induced by ATPe, with a unitary conductance in the range of 5–9 pS (picosiemens) was detected. These channels are too small to allow the transport of larger molecules such as Tris and NMDG and, therefore, cannot be directly responsible for the transport of molecules bearing molecular mass higher than 900 Da in macrophages [8] or

84

Neuroimmunomodulation 1999;6:81–89

400 Da in lymphocytes [29]. This difference in the molecular mass cut-off may reflect the existence of subtypes of the purinergic receptor related to the response, or a differential modulation of the membrane permeability depending on the cell type. It is also possible that fast-activating transmembrane cation currents and membrane permeabilization (pore formation) are indeed separate phenomena [23, 30–32]. In agreement with this hypothesis, we have recently described an ATPe-induced channel with a conductance around 400 pS in macrophages [33]. These channels are permeable to both large cations and anions, such as Tris, NMDG and glutamate. These pores display several properties of the P2-purinoceptor-associated permeabilization phenomenon. Under some experimental situations, we observed a subconductance state of the pore, or alternatively, a second channel with a conductance around 250 pS, with the same characteristics as the large pore. The existence of two ATPe-induced pores or opening states is consistent with data showing that, although macrophages are permeable to molecules of up to 900 Da, some lymphocytes seem to have a molecular weight cut-off of !400 Da.

Modulation of Purinoceptors during Thymocyte Differentiation: Possible Link with Apoptosis

Along with differentiation, thymocytes display a differential susceptibility to ATPe. It was shown that mouse medullary thymocytes exhibit a greater permeability to propidium iodide after ATPe treatment than mixed populations [34]. This result is in agreement with the fact that CD4/CD8-defined single positive thymocytes express more P2X7 purinoceptors than double negative or double positive thymocytes [35]. These findings were recently corroborated by an elegant study measuring variations of intracellular calcium (Ca2+) induced by ATPe in single cells (using cell calcium image techniques) [36]. The authors characterized the kind of purinoceptor present (P2X7) and the phenotype of the responding cells, performing simultaneously immunofluorescence and calcium cell imaging: both immature (CD4-CD8- and CD4+CD8+) and mature (CD4+CD8- and CD4-CD8+) thymocyte populations respond to ATP with a rise in intracellular calcium. Further separation of the double positive population by size revealed that the ATPemediated (Ca2+) response was much more pronounced in large (actively dividing) than in small (terminally divided) CD4+CD8+ thymocytes. Taking these data into account, a new function for P2X7 purinoceptors as co-stimulators

Alves/Coutinho-Silva/Savino

to drive thymocyte differentiation with TCR/MHC-stimulated signals, was proposed [36]. Recently, two additional P2 receptors were identified in thymocytes: P2X1 and P2Y2 [37, 38]. The former is up-regulated in rat thymocytes during dexamethasoneinduced apoptosis, and antagonists of this receptor protect thymocytes from cell death. In addition, it was showed that P2X1 mRNA and protein levels increase in thymocytes induced to die in vivo by the superantigen staphylococcal enterotoxin B. However, these findings may be species-specific since two other groups could not find any changes in P2X1 purinoceptor mRNA expression in mouse thymocytes during glucocorticoid treatment [38, 39]. Thus, at least in rats, the P2X1 purinoceptor may be involved in apoptosis. Additionally, it has been described a transient, protein synthesis-independent enhancement of mRNA expression of the P2Y2 purinoceptor in mouse thymocytes after addition of steroid hormone or T-cell receptor (TCR) cross-linking by anti-TCR mAB as well as agents such as ionomycin and PMA, that directly activate intracellular signaling pathways [38]. Moreover, the up-regulation of the P2Y2 purinoceptor mRNA was observed after longterm (up to 16 h) incubation of thymocytes in the absence of either growth factors or P2Y2 receptor ligands, a condition that led to ‘spontaneous’ apoptosis in thymocytes. The authors then proposed that the rapid increase of P2Y2 purinoceptor mRNA expression could be a common early event in T cells responding to different activating stimuli. In a second vein, it was recently demonstrated that A2a adenosine receptors may mediate the strong inhibition of TCR-triggered proliferation and up-regulation of IL-2 receptor · chain (CD25) in peripheral T lymphocytes [40]. Thus, this type of adenosine receptor might control lymphocyte expansion with negative feedback. In summary, purinergic receptors may modulate differentiation, proliferation and apoptosis in the thymus. In keeping with this idea is the enhancement of the calcium response with maturation, observed when thymocytes are stimulated by anti-CD3Â monoclonal antibody [41]. In addition, some researchers have showed that calcineurin (a calcium and calmodulin-dependent phosphatase) is important for positive selection [42, 43], possibly requiring a small calcium signal, which would be in contrast to the large calcium signal that may be used in the process of negative selection [44]. Future knockout mice for purinergic receptor will probably help to elucidate the physiological role of these receptors during thymocyte selection.

Extracellular ATP in the Thymus

The Mechanism of Lysis Induced by ATPe in Thymocytes and Peripheral Lymphocytes

The intracellular mechanisms whereby ATPe triggers cell lysis remains an open question. Based on electron microscopy studies, it was firstly proposed that ATPe induces necrosis and programmed cell death in thymocytes and certain tumor cell lines, in keeping with the fact that ATP-mediated cell lysis is accompanied by fragmentation of the target cell DNA [45]. The intracellular mechanism whereby ATPe triggers DNA fragmentation and cell lysis is still unknown, although a role for an increase in the cytosolic free Ca2+ concentration has been proposed. Furthermore, it was recently shown that phospholipase D activation by ATPe in human lymphocytes is dependent on bivalent cation influx [46] and that both PTKs and PTPases are involved in ATPe-mediated apoptotic cell death [47]. It has also been suggested that the activation of caspases, a novel family of cytoplasmic proteases related to IL-1ß converting enzyme, may be involved in apoptosis induced by ATPe [48–50].

Putative Sources of ATPe in the Thymus

The concentration of ATPe may increase by different ways and the general sources of ATPe have been reviewed elsewhere [8, 51]. In fact, one of the most obvious is the cytosolic ATP (the concentration of intracellular ATP is 2–5 mM in most cells), which could be released after transient cell injury or loss of cell viability. Two additional and physiologically relevant sources of extracellular nucleotides could be (1) exocytotic release of ATP specifically concentrated in secretory granules on vesicles, as in the case of platelets, neurons, mast cells, and chromaffin cells of the adrenal gland [51], and (2) the release of cytosolic ATP via intrinsic plasma membrane channels or pores in the absence of irreversible cytolysis: some membrane transporters such as P-glycoprotein and the CFTR protein have been proposed to transport ATP as well as smaller ions [52–54]. This mode of ATP release has received strong support by recent findings showing that microglial cells, lymphocytes and macrophages release ATP via a nonlytic mechanism [55–57]. In addition, ATP is released from cytotoxic T lymphocytes during antigen presentation [58]. In the thymus, at least three further sources of ATP might be conceived (fig. 2). First, this organ displays cholinergic and adrenergic innervation, and ATP may be coreleased with epinephrine or acetylcholine, similar to

Neuroimmunomodulation 1999;6:81–89

85

Fig. 2. Hypothetical sources of ATP in the thymus. a, c ATPe derived from synaptic vesicles can paracrinally reach specific P2X and/or P2Y receptors expressed on TEC and thymocytes. b TEC is being stimulated by ATPe secreted by thymocytes. By contrast, c illustrates the possibility that ATPe can be released by apoptotic thymocytes, either

through a nonspecific permeability mechanism or through an ion channel permeable to ATP.

what has been found in other autonomic sites [59–61]. In this regard, adrenergic innervation has been found in a large scale in the thymic cortex [59], and adrenoreceptors have been detected in both thymocytes and TEC [62, 63]. By contrast, it remains to be definitely established whether the thymic microenvironment is actually innervated by cholinergic terminals or if such an innervation remains restricted to vascular muscle cells [64]. Second, apoptotic thymocytes may display a transient increase in the plasma membrane permeability, allowing passage of low-molecular-weight molecules [65]. Finally, ATP might be released by thymocytes during MHC-dependent presentation of endogenous peptides, that occurs during thymus selection [56].

86

Neuroimmunomodulation 1999;6:81–89

Possible Physiological Roles of Purinoceptors in the Thymic Microenvironment

The physiological meaning of ATPe/purinoceptor interactions in thymic microenvironmental cells is still poorly understood. One possibility is that intrathymically-released ATP stimulates cytokine or thymic hormone secretion, as occurs in macrophages where ATPe increases IL-1· maturation and secretion [49]. Likewise, ATP may modify the density of membrane adhesion molecules. The modulation of adhesion molecule expression has been convincingly shown for some leukocytes, hence, ATP released in injury sites stimulates the migration of neutrophils, increasing the expression of some adhesion molecules [66]. One can, therefore, imagine that some apoptotic thymocytes release ATP, which might change

Alves/Coutinho-Silva/Savino

the expression of adhesion molecules, and facilitate the migration of thymocytes which undergo thymic selection. Lastly, ATP released by innervation or apoptotic thymocytes could stimulate endocytosis by TEC, macrophages and/or dendritic cells. At least regarding thymocytes, this activity would in turn enhance the rate of dead thymocyte uptake. This possibility is based on the observation that ATP can stimulate phagocytosis in macrophages and macropynocytosis in some TEC and dendritic cells [Nihei et al., in preparation]. In conclusion, although initial steps to elucidate the role of ATP and other nucleotides in the neuroendocrine

control of the thymus are accomplished, much work remains to be done to understand the physiological role of extracellular nucleotides in the thymus. Acknowledgments The authors thank H.N.M. Diniz for computer drawings, R.C. Bisaggio for helping with figures, and Drs. David M. Ojcius, V. Cotta de Almeida and O.K. Nihei for critical reading of the manuscript. This work was financially supported by Conselho Nacional de Desenvolvimento Cientı´fico e Tecnolo´gico (CNPq), Financiadora de Estudos e Projetos (FINEP), Fundaça˜o de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ) and PRONEX/CNPq.

References 1 Robey EA, Fowlkes B: Selective events in T cell development. Annu Rev Immunol 1994;12: 675–705. 2 Dardenne M, Savino W: Control of thymus physiology by peptide hormones and neuropeptides. Immunol Today 1994;15:518–523. 3 van Ewijk W: T-cell differentiation is influenced by thymic microenvironments. Annu Rev Immunol 1991;9:591–615. 4 Boyd RL, Tucek CL, Godfrey DI, Izon DJ, Wilson TJ, Davidson NJ, Bean AGD, Ladyman HM, Ritter MA, Hugo P: The thymic microenvironment. Immunol Today 1993;14: 445–459. 5 Coutinho-Silva R, Alves LA, Campos-de-Carvalho AC, Savino W, Persechini PM: Characterization of P2Z purinergic receptors on phagocytic cells of the thymic reticulum in culture. Biochim Biophys Acta 1996;1280:217– 222. 6 Coutinho-Silva R, Alves LA, Savino W, Persechini PM: A cation non-selective channel induced by extracellular ATP in macrophages and phagocytic cells of thymic reticulum. Biochim Biophys Acta 1996;1278:125–130. 7 Apasov S, Koshiba M, Redegeld F, Sitkovsky MV: Role of extracellular ATP and P1 and P2 classes of purinergic receptors in T-cell development and cytotoxic T lymphocyte effector functions. Immunol Rev 1995;146:5–19. 8 Dubyak GR, El-Moatassim C: Signal transduction via P2-purinergic receptors for extracellular ATP and other nucleotides. Am J Physiol 1993;265:C577-C606. 9 Brake AJ, Julius D: Signaling by extracellular nucleotides. Annu Rev Cell Dev Biol 1996;12: 19–541. 10 Carr WES, Gleeson RA, Ache BW, Milstead ML: Olfactory receptors of the spiny lobster: ATP-sensitive cells similarities to P2-type purinoreceptors of vertebrates. J Comp Physiol A 1986;158:331–338. 11 Devreotes PN, Zigmond SH: Chemotaxis in eukaryotic cells: A focus on leukocytes and Dictyostelium. Annu Rev Cell Biol 1988;4:649– 686.

Extracellular ATP in the Thymus

12 Olah ME, Stiles GL: Adenosine receptor subtypes: Characterization and therapeutic regulation. Annu Rev Pharmacol Toxicol 1995;35: 581–597. 13 Fredholm BB, et al: Sixth Nomenclature and Classification of Purinoceptors. Pharmacol Rev 1994;46:143–156. 14 Chen ZP, Levy A, Lightman SL: Nucleotides as extracellular signalling molecules. J Neuroendocrinol 1995;7:83–96. 15 Bertrand G, Chapal J, Loubatieres-Mariani MM: Potentiating synergism between adenosine diphosphate or triphosphate and acetylcholine on insulin secretion. Am J Physiol 1986;251:E416-E421. 16 Liu P, Wen M, Hayashi J: Characterization of ATP receptor responsible for the activation of phospholipase A2 and stimulation of prostaglandin E2 production in thymic epithelial cells. Biochem J 1995;308:399–404. 17 El-Moatassim C, Dornand J, Mani JC: Extracellular ATP increases cytosolic free calcium in thymocytes and initiates the blastogenesis of the phorbol 12-myristate 13-acetate-treated medullary population. Biochim Biophys Acta 1987;927:437–444. 18 Baricordi OR, Ferrari D, Melchiorri L, Chiozzi P, Hanau S, Chiari E, Rubini M, DiVirgilio F: An ATP-activated channel is involved in mitogenic stimulation of human T lymphocytes. Blood 1996;87:682–690. 19 Jamieson GP, Snook MB, Thurlow PJ, Wiley JS: Extracellular ATP causes loss of L-selectin from human lymphocytes via occupancy of P2Z purinoceptors. J Cell Physiol 1996;166: 637–642. 20 El-Moatassim C, Mani JC, Dornand J: Extracellular ATP4- permeabilizes thymocytes not only to cations but also to low-molecularweight solutes. Eur J Pharmacol 1990;181: 111–118.

21 DiVirgilio F, Bronte V, Collavo D, Zanovello P: Responses of mouse lymphocytes to extracellular adenosine 5)-triphosphate (ATP): Lymphocytes with cytotoxic activity are resistant to the permeabilizing effects of ATP. J Immunol 1989;143:1955–1960. 22 Pizzo P, Zanovello P, Bronte V, DiVirgilio F: Extracellular ATP causes lysis of mouse thymocytes and activates a plasma membrane ion channel. Biochem J 1991;274:139–144. 23 Persechini PM, Bisaggio RC, Alves-Neto JL, Coutinho-Silva R: Extracellular ATP in the lymphohematopoietic system: P2Z purinoceptors and membrane permeabilization. Braz J Med Biol Res 1998;31:25–34. 24 Wiley JS, Mayger W, Cragoe EJ, Jopson M: Extracellular ATP opens an amiloride-sensitive cation channel in human lymphocytes. Ann NY Acad Sci 1990;603:439–440. 25 Wiley JS, Chen R, Jamieson GP: The ATP4receptor-operated channel P2Z class of human lymphocytes allows Ba2+ and ethidium+ uptake: Inhibition of fluxes by suramin. Arch Biochem Biophys 1993;305:54–60. 26 Bretschneider F, Klapperstuck M, Lohn M, Markwardt F: Nonselective cationic currents elicited by extracellular ATP in human and Blymphocytes. Eur J Physiol 1995;429:691– 698. 27 Albuquerque C, Oliveira SM, Coutinho-Silva R, Oliveira-Castro GM, Persechini PM: ATPand UTP-induced currents in macrophages and macrophage-polykaryons. Am J Physiol 1993;265:C1663-C1673. 28 Markwardt F, Lohn M, Bohm T, Klapperstuck M: Purinoceptor-operated cationic channels in human b lymphocytes. J Physiol 1997;498: 143–151. 29 Valeva A, Weisser A, Walker B, Kehoe M, Bayley H, Bhakdi S, Palmer M: Molecular architecture of a toxin pore – a 15-residue sequence lines the transmembrane channel of staphylococcal alpha-toxin. EMBO J 1996;15:1857– 1864.

Neuroimmunomodulation 1999;6:81–89

87

30 Blanchard DK, Hoffman SL, Djeu JY: Inhibition of extracellular ATP-mediated lysis of human macrophages by calmodulin antagonists. J Cell Biochem 1995;57:452–464. 31 Nuttle LC, Dubyak GR: Differential activation of cation channels and non- selective pores by macrophage P–2Z purinergic receptors expressed in Xenopus oocytes. J Biol Chem 1994; 269:13988–13996. 32 Petrou S, Ugur M, Drummond RM, Singer JJ, Walsh JV: P2X7 purinoceptor expression in Xenopus oocytes is not sufficient to produce a pore-forming P2Z-like phenotype. FEBS Lett 1997;411:339–345. 33 Coutinho-Silva R, Persechini PM: P2Z purinoceptor-associated pores induced by extracellular ATP in macrophages and J774 cells. Am J Physiol 1997;273:C1793-C1800. 34 Nagy P, Panyi G, Jenei A, Bene L, Gaspar R, Matko J, Damjanovich S: Ion-channel activities regulate transmembrane signaling in thymocyte apoptosis and T- cell activation. Immunol Lett 1995;44:91–95. 35 Chused TM, Apasov S, Sitkovsky M: Murine T lymphocytes modulate activity of an ATP-activated P–2z-type purinoceptor during differentiation. J Immunol 1996;157:1371–1380. 36 Ross PE, Ehring GR, Cahalan MD: Dynamics of ATP-induced calcium signaling in single mouse thymocytes. J Cell Biol 1997;138:987– 998. 37 Chvatchko Y, Valera S, Aubry JP, Renno S, Buell G, Bonnefoy JY: The involvement of an ATP-gated ion channel, P2X1 in thymocyte apoptosis. Immunity 1996;5:275–283. 38 Koshiba M, Apasov S, Sverdlov V, Chen P, Erb L, Turner JT, Weisman GA, Sitkovsky MV: Transient up-regulation of P2Y2 nucleotide receptor mRNA expression is an immediate early gene response in activated thymocytes. Proc Natl Acad Sci USA 1997;94:831–836. 39 Jiang S, Kull B, Fredholm BB, Orrenius S: P2X purinoceptor is not important in thymocyte apoptosis. Immunol Lett 1996;49:197–201. 40 Huang S, Apasov S, Koshiba M, Sitkovsky M: Role of A2a extracellular adenosine receptormediated signaling in adenosine-mediated inhibition of T-cell activation and expansion. Blood 1997;90:1600–1610. 41 Eichmann KA: Signal strength hypothesis of thymic selection: Preliminary considerations. Immunol Lett 1995;44:87–90. 42 Wang CR, Hashimoto K, Kubo S, Yokochi T, Kubo M, Suzuki M, Suzuki K, Tada T, Nakayama T: T-Cell receptor-mediated signaling events in CD4+CD8+ thymocytes undergoing thymic selection: Requirement of calcineurin activation of thymic positive selection but not negative selection. J Exp Med 1995;181:927– 941. 43 Guidos JC: Positive selection of CD4+ and CD8 cells. Curr Opin Immunol 1996;8:225– 232. 44 Andjelic S, Jain N, Nikolic-Zugic J: Immature thymocytes become sensitive to calcium-mediated apoptosis with the onset of CD8, CD4, and the T cell receptor expression: A role for bcl–2? J Exp Med 1993;178:1745–1751.

88

45 Zheng LM, Zychlinsky A, Liu CC, Ojcius DM, Young JDE: Extracellular ATP as a trigger for apoptosis or programmed cell death. J Cell Biol 1991;112:279–288. 46 Gargett CE, Cornish EJ, Wiley JS: Phospholipase D activation by P–2Z-purinoceptor agonists in human lymphocytes is dependent on bivalent cation influx. Biochem J 1996;313: 529–535. 47 Bronte V, Macino B, Zambon A, Rosato A, Mandruzzato S, Zanovello P, Collavo D: Protein tyrosine kinases and phosphatases control apoptosis induced by extracellular adenosine 5)-triphosphate. Biochem Biophys Res Commun 1996;218:344–351. 48 Hogquist KA, Nett MA, Unanue ER, Chaplin DD: Interleukin-1 is processed and released during apoptosis. Proc Natl Acad Sci USA 1991;88:8485–8489. 49 Perregaux D, Gabel CA: Interleukin-1ß maturation and release in response to ATP and nigericin. J Biol Chem 1994;269:15195– 15203. 50 Cohen GM: Caspases: the executioners of apoptosis. Biochem J 1997;326:1-16. 51 Gordon JL: Extracellular ATP: Effects, sources and fate. Biochem J 1986;233:309–319. 52 Abraham EH, Prat AG, Gerweck L, Seneveratne T, Arceci RJ, Kramer R, Guidotti G, Cantiello HF: The multidrug resistance (mdr1) gene product functions as an ATP channel. Proc Natl Acad Sci USA 1993;90:312–316. 53 Reisin IL, Prat AG, Abraham EH, Amara JF, Gregory RJ, Ausiello DA, Cantiello HF: The cystic fibrosis transmembrane conductance regulator is a dual ATP and chloride channel. J Biol Chem 1994;269:20584–20591. 54 Schwiebert EM, Egan M, Hwang TH, Fulmer SB, Allen S, Cutting GR, Guggino WB: CFTR regulates outwardly rectifying chloride channels through an autocrine mechanism involving ATP. Cell 1995;81:1063–1073. 55 Ferrari D, Chiozzi P, Falzoni S, Hanau S, DiVirgilio F: Purinergic modulation of interleukin-1-beta release from microglial cells stimulated with bacterial endotoxin. J Exp Med 1997;185:579–582. 56 Blanchard DK, Wei S, Duan CN, Pericle F, Diaz JI, Djeu JY: Role of extracellular adenosine triphosphate in the cytotoxic T-lymphocyte-mediated lysis of antigen presenting cells. Blood 1995;85:3173–3182. 57 Buell G, Michel AD, Lewis C, Collo G, Humphrey PPA, Surprenant A: P2X1 receptor activation in HL60 cells. Blood 1996;87:2659– 2664. 58 Filippini A, Taffs RE, Sitkovsky MV: Extracellular ATP in T-lymphocyte activation: Possible role in effector functions. Proc Natl Acad Sci USA 1990;87:8267–8271. 59 Felten SY, Felten DL, Bellinger DL, Carlson SL, Ackerman KD, Madden KS, Olschowka JA, Livnat S: Noradrenergic sympathetic innervation of lymphoid organs. Prog Allergy 1988;43:14–36. 60 Kugelgen I, Klaus S: Noradrenaline-ATP cotransmission in the sympathetic nervous system. Trends Pharmacol Sci 1991;12:319–324.

Neuroimmunomodulation 1999;6:81–89

61 Bulloch K, Moore RY: Innervation of the thymus gland by brain stem and spinal cord in the mouse and rat. Am J Anat 1984;162:157–166. 62 Marchetti B, Morale MC, Paradis P, Bouvier M: Characterization, expression, and hormonal control of a thymic beta-2-adrenergic receptor. Am J Physiol 1994;267:E718-E731. 63 Kurz B, Feindt J, von Gaudecker B, Kranz A, Loppnow H, Mentlein R: Beta-adrenoceptormediated effects in rat cultured thymic epithelial cells. Br J Pharmacol 1997;120:1401– 1408. 64 Singh U, Fatani JA, Mohajir AM: Ontogeny of cholinergic innervation of thymus in mouse. Dev Comp Immunol 1987;11:627–635. 65 Lammas DA, Stober C, Harvey CJ, Kendrick N, Panchalingam S, Kumararatne DS: ATPinduced killing of mycobacteria by human macrophages is mediated by purinergic P2Z(P2X7) receptors. Immunity 1997;7:433– 444. 66 Fredholm BB: Purines and neutrophil leukocytes. Gen Pharmacol 1997;28:345–350. 67 Valera S, Hussy N, Evans RJ, Adami N, North RA, Surprenant A, Buell G: A new class of ligand-gated ion channel defined by P2X receptor for extracellular ATP. Nature 1994;317: 516–519 68 Brake AJ, Wagenbach MJ, Julius D: New structural motif for ligand-gated ion channels defined by an ionotropic ATP receptor. Nature 1994;317:519–523. 69 Lewis C, Neidhart S, Holy C, North RA, Buell G, Surprenant A: Coexpression of P2X2 and P2X3 receptor subunits can account for ATPgated currents in sensory neurons. Nature 1995;377:432–435. 70 Chen CC, Akoplan AN, Sivilloti L, Colquhoun D, Burnstock G, Wood JN: A P2X purinoceptor expressed by a subset of sensory neurons. Nature 1995;377:428–431. 71 Wang CZ, Namba N, Gonoi T, Inagaki N, Seino S: Cloning and pharmacological characterization of a fourth P2X receptor subtype widely expressed in brain and peripheral tissues including various endocrine tissues. Biochem Biophys Res Commun 1996;220:196–202. 72 Soto F, Garcia-Guzman M, Gomez-Hernandez JM, Hollmann M, Karschin C, Stuhmer W: P2X4: An ATP-activated ionotropic receptor cloned from rat brain. Proc Natl Acad Sci USA 1996;93:3684–3688. 73 Buell G, Lewis C, Collo G, North RA, Suprenant A: An antagonist-insensitive P2X receptor expressed in epithelia and brain. EMBO J 1996;15:55–62. 74 Bo X, Zhang Y, Nassar M, Burnstock G, Shoepfer R: A P2X purinoceptor cDNA conferring a novel pharmacological profile. FEBS Lett 1995:375:129–133. 75 Garcia-Guzman M, Soto F, Laube B, Stuhmer W: Molecular cloning and functional expression of a novel rat heart P2X purinoceptor. FEBS Lett 1996;388:123–127. 76 Soto F, Garcia-Guzman M, Karschin C, Stuhmer W: Cloning and tissue distribution of a novel P2X receptor from rat brain. Biochem Biophys Res Commun 1996;223:456–460.

Alves/Coutinho-Silva/Savino

77 Suprenant A, Rassendren F, Kawashima E, North RA, Buell G: The cytolytic P2Z receptor for extracellular ATP identified as a P2X receptor (P2X7). Science 1996;272:735–738. 78 Rassendren F, Buell GN, Virginio C, Collo G, North RA, Suprenant A: The permeabilizing ATP receptor, P2X7. J Biol Chem 1997;272: 5482–5486. 79 Tokuyama Y, Hara M, Jones EMC, Fan Z, Bell I: Cloning of rat and mouse P2Y purinoreceptors. Biochem Biophys Res Commun 1995; 211:211–218. 80 Parr CE, Sullivan DM, Paradiso AM, Lazorowski ER, Burch LH, Olsen JC, Erb L, Weisman GA, Boucher RC, Turner JT: Cloning and expression of a human P2U nucleotide receptor, a target for cystic fibrosis pharmacotherapy. Proc Natl Acad Sci USA 1994;91:3275– 3279. 81 Lustig KD, Shiau AK, Brake AJ, Julius D: Expression cloning of an ATP receptor from mouse neuroblastoma cells. Proc Natl Acad Sci USA 1993;90:5113–5117.

Extracellular ATP in the Thymus

82 Webb TE, Henderson D, King BF, Wang S, Simon J, Bateson AN, Burnstock G, Barnard EA: A novel G protein-coupled P2 purinoceptor (P2Y3) activated preferentially by nucleoside diphosphates. Mol Pharmacol 1996;50: 258–265. 83 Barnard EA, Burnstock G, Webb TEG: Protein-coupled receptors for ATP and other nucleotides: A new receptor family. Trends Pharmacol Sci 1994;15:67–70. 84 Nguyen T, Erb L, Weiman GA, Marchese A, Heng HHQ, Garrad RC, George SR, Turner JT, O’Dowd BF: Cloning, expression, and chromosomal localization of the human uridine nucleotide receptor gene. J Biol Chem 1995;270:30845–30848. 85 Communi D, Pirotton S, Parmentier M, Boeynaems JM: Cloning and functional expression of a human uridine nucleotide receptor. J Biol Chem 1995;270:30849–30852.

86 Webb TE, Kaplan MG, Barnard EA: Identification of 6H1 as a P2Y purinoreceptor: P2Y5. Biochem Biophys Res Commun 1996;219: 105–110. 87 Chang K, Hanaoka K, Kumada M, Takuwa Y: Molecular cloning and functional analysis of a novel P2 nucleotide receptor. J Biol Chem 1995;270:26152–26158. 88 Communi D, Pirotton S, Parmentier M, Boeynaems JM: Cloning, functional expression and tissue distribution of the human PY6 receptor. Biochem Biophys Res Commun 1996;222: 303–308. 89 Akbar GKM, Dasari VR, Webb TE, Ayyanathan K, Pillarisetti K, Sandhu AK, Athwal RS, Daniel JL, Ashby B, Banard EA, Kunapuli SP: Molecular cloning of a novel P2 purinoceptor from human erythroleukemia cells. J Biol Chem 1996;271:18363–18367. 90 Bogdanov YD, Dale L, King BF, Whittock N, Burnstock G: Early expression of a novel nucleotide receptor in neural plate of Xenopus embryos. J Biol Chem 1997;272:12583– 12590.

Neuroimmunomodulation 1999;6:81–89

89

Neuroimmunomodulation 1999;6:90–96

Thymus-Derived Glucocorticoids and the Regulation of Antigen-Specific T-Cell Development Eva Tolosa Jonathan D. Ashwell Laboratory of Immune Cell Biology, National Cancer Institute, National Institutes of Health, Bethesda, Md., USA

Key Words Steroids W Glucocorticoids W Thymocyte development W Thymocyte selection W Autoimmunity

Abstract Bidirectional interactions of both a stimulatory and inhibitory nature occur between the neuroendocrine and the immune systems, and these interactions play an important modulatory role during T-cell ontogeny. Specifically, glucocorticoids potently induce apoptosis in thymocytes and activated T cells, but can also rescue these cells from activation-induced cell death. The objective of this review is to discuss current data on the interactions of the immune system with steroid hormones in the thymus and to describe a model that includes glucocorticoids in the shaping of the peripheral T-cell antigen-specific repertoire and deals with their potential role in the generation of autoimmune disease.

Abbreviations DHEA, dehydroepiandrosterone; GR, glucocorticoid receptor; MHC, major histocompatibility complex; PNAr, peanut agglutinin receptor; SLE, systemic lupus erythematosus; TEC, thymic epithelial cells; TCR, T-cell receptor; TNC, thymic nurse cells.

ABC Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

© 1999 S. Karger AG, Basel

Accessible online at: http://BioMedNet.com/karger

The Immune Thymus Thymocytes

Early T-cell progenitors originating in the bone marrow migrate through the blood to the cortex of the thymus. Once established, thymocytes move from the outer cortex to the central medulla, in the process undergoing a series of differentiation events that are easily followed by assessing the cell surface phenotype. Very immature thymocytes express neither the TCR for antigen nor CD4 or CD8 (the latter called co-receptors because they function in concert with the TCR to deliver activating signals). Because these cells do not express CD4 or CD8, they are referred to as double negative cells. As thymocytes mature they rearrange the TCR ß gene locus, and cells that have undergone successful rearrangement express a ‘pre-TCR’ that consists of CD3 (the Á, ‰, and Â chains), a ˙ homodimer, TCR ß, and a nonpolymorphic pre-TCR · chain [1]. Although no ligand has yet been described for the preTCR, it seems likely that this receptor recognizes an endogenous nonpolymorphic (and not classical MHCencoded) ligand, because upon expression of the pre-TCR the thymocytes express CD4 and CD8 (double positive cells) and undergo a burst of proliferation. It is also at this time that thymocytes rearrange the TCR · locus, and productively rearranged TCR · chains replace the pre-· chain on the cell’s surface [2]. At this stage of differentiation thymocytes begin to express low levels of the ·ß TCR

Jonathan D. Ashwell Room 1B-40, Building 10 National Institutes of Health, Bethesda, MD 20892 (USA) Tel. +1 301 496 4931, Fax +1 301 402 4844, E-Mail [email protected]

(TCRloCD4+CD8+ cells) and recognition of endogenous ligands (self peptides bound to self MHC-encoded molecules) begins to play an essential role in dictating each cell’s fate. It is generally thought that thymocytes having TCRs with subthreshold avidity for self antigen/MHC die in the thymus, a default death pathway that has been termed ‘death by neglect’. Similarly, thymocytes bearing TCRs with high avidity for self antigen/MHC undergo apoptotic death (negative selection), ensuring that these potentially autoreactive cells cannot populate the periphery. Finally, thymocytes bearing TCRs with so-called lowto-moderate avidity for self antigen/MHC are rescued from the default death pathway, differentiate into CD4+CD8– or CD4–CD8+ (single positive) cells, and migrate to the periphery (positive selection). Positive selection is also accompanied by upregulation (approx. 10fold) of cell surface TCR expression to the levels found on mature peripheral T cells. These selection processes determine to a large extent the range of antigens to which mature peripheral T cells can respond (the antigen-specific repertoire). Stromal Cells In addition to the role that TCR occupancy has in determining thymocyte fate, antigen-presenting cells also play an important role in thymocyte homeostasis and selection [3]. This is accomplished by a variety of means, including their direct interaction with thymocytes (via TCR binding to antigen/MHC or by interactions of other activating/regulatory molecules) or their production of soluble biologically active factors (paracrine signaling). The dendritic configuration of thymic epithelial cells, with extended ‘arms’ amongst packed thymocytes, appears to be very favorable for both processes. TEC, macrophages, and dendritic cells constitute the antigen-presenting cell contingent that promotes negative and positive selection of thymocytes. Dendritic cells, found exclusively in the thymic medulla and the corticomedullary junction, together with macrophages and medullary TEC are responsible for the clonal elimination of autoreactive thymocytes [4]. These cell types express high levels of MHC class II molecules and B7. B7 is the ligand for CD28 and CTLA-4, T-cell transmembrane molecules that enhance or suppress activating stimuli, respectively. Cortical epithelial cells express a lower density of MHC class II molecules and do not express B7 [5]. Although not universally observed [6], the majority of data supports the notion that antigen presentation by cortical epithelial cells is responsible for the positive selection of thymocytes bearing TCRs with low-to-moderate avidi-

Regulation of Thymocyte Development by Steroids

ty for self [7–9]. One elegant study demonstrating this used transgenic mice in which MHC class II expression was targeted exclusively to thymic cortical epithelium. These mice exhibited normal positive selection but impaired negative selection, as evidenced by the presence of autoreactive cells in the mature T-cell population [10]. Possible explanations for the different outcomes of antigen presented by cortical versus medullary epithelial cells include different processing pathways in the two cell types [11] and/or different co-stimulatory capabilities. TNC represent an especially intriguing, if controversial, thymic component. TNC are MHC class I- and class II-expressing epithelial cells found in the outer cortical and subcapsular regions [12]. Since they express a number of neuropeptides, they are considered of neuroendocrine origin, and can be identified by the presence of complex gangliosides, particularly ganglioside GQ, recognized by the monoclonal antibody A2B5 and also expressed on neurons, and gangliosides GD and GT [13]. An unusual property of these cells is that they physically envelop thymocytes (up to 200 per TNC), and these thymocytes are of the TCRintermediateCD4+CD8+ phenotype, the stage at which antigen-specific T-cell selection and differentiation occur [14]. Since the thymocytes within TNC were found to be viable and proliferating, it has been suggested that TNC provide a stimulatory or ‘nursing’ environment for the process of positive selection [15]. This possibility was supported by the finding that CD4+CD8+ thymocytes engulfed by TNC have a cell surface phenotype indicative of positive selection in that they shift from the immature phenotype TCRlow and PNArhigh to the more mature TCRintermediate PNArlow phenotype, and have reduced levels of DNA fragmentation compared to the nonengulfed thymocytes [16].

Steroids

Steroids are lipophilic compounds derived from a common precursor, cholesterol, that have many essential biological functions, including the control of the rates of synthesis or degradation of proteins, carbohydrates, and lipids by inducing the synthesis of a large number of enzymes. Steroids circulate in a complex with carrier proteins, from which they must dissociate before entering cells. Once having diffused into cells, steroids bind either cytoplasmic (glucocorticoids) or nuclear-resident receptors, which are composed of functionally discrete domains including an N-terminal modulatory portion required for maximal hormone activity and tissue specifici-

Neuroimmunomodulation 1999;6:90–96

91

ty, a highly conserved central DNA-binding portion, and a C-terminal ligand-binding portion. Upon hormone binding, the receptors become activated (and in the case of the GR, translocate to the nucleus) and regulate transcription of susceptible gene(s) [17, 18]. The major mechanism by which steroid receptors activate gene transcription involves binding of the ligand-activated receptor as a dimer to specific sequences in response genes and recruitment of co-activators that can loosen the nucleosomal structure, allowing the formation of a stable regulatory complexes [19]. By contrast, binding of antagonists induces a different conformational change in the receptor so that even if they dimerize and bind to the response element the associated co-repressors cannot be displaced, resulting in a nonproductive interaction [20]. Steroid receptors can also repress transcriptional activity by physically interacting with other transcription factors, such as AP-1 [17]. The main steroidogenic organs are the adrenal cortex, responsible for the production of glucocorticoids, mineralocorticoids and adrenal androgens, and the gonads, responsible for the synthesis of sex steroids (androgens and estrogens). Besides their role as regulators of metabolic processes, steroids have been shown to affect multiple aspects of the immune response, a few of which are summarized below. Sex Steroids and the Immune Response Sex steroid receptors have been detected in immature thymocytes [21, 22], making these cells potential targets for direct actions of sex steroids. One particularly clear example of this is the dramatic thymic involution that occurs during pregnancy, due to the killing of cortical thymocytes by high levels of estrogens [reviewed in 23]. In vivo administration of estrogens, estradiol in particular, results in a decrease in the numbers and proportion of the CD4+CD8+ population, analogous to the effects of dexamethasone [24]. Another sex steroid, DHEA, is quantitatively the most abundant secreted steroid in mammals, with a wide variety of physiological effects, including protection against infections [25]. Two striking features of DHEA are its steady decline with age and its deficiency in patients with a number of major diseases, including several types of cancer [26, 27], atherosclerosis [28], and autoimmune disease [29]. At least two groups have shown that administration of DHEA protects thymocytes from glucocorticoid-induced killing [30, 31]. Thus, although the mechanisms have not yet been elucidated, thymus cellularity and, presumably, function are susceptible to the actions of sex hormones.

92

Neuroimmunomodulation 1999;6:90–96

Various aspects of immune function are enhanced in females compared to males, and at least in some cases this appears to be due to differences in sex steroids. For example, lymphocytes from immunized female mice respond more vigorously to soluble antigens such as KLH and OVA in vitro than lymphocytes from male animals or females treated with testosterone [32]. Female mice are also more resistant to a particular strain of encephalitis virus [33] and to infection with larvae of Brugia pahangi [34], mount a more efficient cytotoxic response against Coxsackie virus [35], and have a better antibody response to Candida albicans [36]. Also perhaps indicative of a more active immune response, the occurrence of most autoimmune diseases is higher in females, as evidenced by the female-to-male ratios of autoimmune thyroiditis (20:1), SLE (9:1) and rheumatoid arthritis (4:1) in humans, and by the higher susceptibility of murine females to induced autoimmune syndromes and spontaneous SLE [reviewed in 37] and diabetes mellitus [38]. An attractive explanation for this difference is direct modulation of the immune response by sex steroids, but thus far studies on the effects of sex hormones in immune responses have not yielded conclusive results. What is clear at this point is that androgens have an ameliorative effect in a variety of SLE models such as MRL-lpr/lpr [39] and NZB/NZW mice [40]. Androgens also inhibit the in vitro production of anti-DNA antibodies by peripheral blood cells of patients with SLE [41], have a protective effect in experimental autoimmune encephalitis [42] and collagen-induced arthritis [43], and prevent islet cell destruction in female nonobese diabetic mice [44]. Rheumatoid arthritis [29] and multiple sclerosis [45] remit in females during pregnancy, a period of increased levels of progesterone, corticosterone and estrogens, but both diseases recur in the postpartum period, coinciding with gonadal and adrenal steroid deficiency. Of note is the influence of sex steroids in extrathymic T-cell differentiation, where it has been shown that the thymic atrophy achieved with estrogen is followed by the initiation of T-cell differentiation in the liver [46]. Since extrathymic T cells that originate in liver are characteristically autoreactive [47], these findings may indicate an interesting link between estrogens and the preponderance of autoimmune disease in females. Glucocorticoids in Thymocyte Development It has long been known that adrenalectomy of adult animals leads to thymic hypertrophy [48], leading to the idea that physiological levels of glucocorticoids suppress thymocyte viability. All lymphocytes express functional

Tolosa/Ashwell

GR [49], and glucocorticoid stimulation results in the transcription of a large number steroid-responsive genes [50]. Immature thymocytes are strongly susceptible to glucocorticoid-induced apoptosis, and can be killed by concentrations of glucocorticoid elicited by stress [51, 52]. The mechanism by which glucocorticoids cause thymocyte death is not understood, but it is known to require new mRNA and protein synthesis [53] and is inhibited by depletion of ATP. Since the vast majority of immature thymocytes die in the thymus, it has been speculated that glucocorticoids play a role in the elimination of those thymocytes with insufficient TCR avidity to trigger positive selection (death by neglect) [54]. On the other hand, the finding that albino mutant mice deficient in their ability to upregulate gluconeogenic hepatic enzymes have abnormally small thymi [55, 56] suggests that glucocorticoids may also play a positive role in thymocyte viability. A critical clue to the significance of glucocorticoids in thymocyte development came from in vitro studies with murine T-cell hybridomas. While occupancy of the GR or the TCR in T-cell hybridomas induces apoptosis, the simultaneous occupancy of both receptors prevents cell death [57–59]. This phenomenon, which was termed mutual antagonism, prompted us to propose that cross-talk between the TCR and the GR might participate in the antigen-specific selection of thymocytes. This model holds that occupancy of the GR signals for apoptosis, and in the absence of TCR occupancy may account for the default death pathway of thymocyte development. Occupancy of the TCR counters glucocorticoid-induced apoptosis, and in the case of low-to-moderate avidity ligands the thymocyte is rescued from death (positive selection). Thymocytes bearing TCRs that recognize self antigen/ MHC with high avidity would receive a potent apoptotic stimulus from the TCR that would overcome the antagonism by glucocorticoids (negative selection). In this model, unlike most others of positive selection, there is no ‘positive’ signal delivered via the TCR by low-to-moderate avidity ligands – the signal generated by this event induces apoptosis. This signal is blocked by glucocorticoids, just as TCR signaling blocks the apoptotic signal delivered by glucocorticoids. The observation that adrenal production of glucocorticoids is very low in fetal and neonatal life prompted us to ask if the thymus itself is a source of steroid production. In initial experiments, cultured TEC were found to produce pregnenolone, the first intermediary in steroid biosynthesis [60]. Immunohistochemistry demonstrated that enzymes required for synthesis of steroids, and in particular glucocorticoids, are present in cytokeratin-positive epi-

thelial cells. The distribution of these cells was largely cortical and subcapsular, reminiscent of the distribution of TNC. Thymic production of pregnenolone was maximal during fetal and postnatal development and decreased thereafter, reaching a nadir at approximately 4 weeks of age (the inverse of glucocorticoid production by the adrenal cortex). Therefore, thymus-derived steroids are highest at a time of active thymocyte development. A thymic epithelial cell line has also been shown to induce TCRindependent apoptosis of double positive thymocytes, a process that was blocked by drugs that inhibited steroid production or glucocorticoid receptor signaling [61]. Evidence supporting the mutual antagonism model of thymocyte selection was provided in a number of distinct biological systems. Blockade of glucocorticoid biosynthesis in cultured fetal thymi resulted in increased thymocytes sensitivity to TCR-induced apoptosis [60]. Furthermore, using an antigen-specific TCR ·ß transgenic mouse model, it was shown that inhibition of thymic glucocorticoid biosynthesis in organ culture causes the apoptotic death of double positive thymocytes with low-to-moderate avidity for self antigen/MHC but not thymocytes bearing TCRs that do not recognize self [62]. Thus, it was possible to ‘turn positive selection into negative selection’ by inhibiting local corticosteroid production. To test the role of glucocorticoids in thymocyte development in vivo, transgenic mice were made in which GR antisense transcript expression was limited specifically to immature thymocytes [63]. In these transgenic animals (GR-TKO mice), in which double positive thymocytes were relatively insensitive to glucocorticoids, the number of thymocytes was 80–90% reduced due to a decrease in double positive and single positive cells (CD4–CD8– thymocytes were present in normal numbers). As with cultured fetal thymi in which corticosteroid production was inhibited, the double positive thymocytes were exquisitely sensitive to anti-TCR-mediated deletion.

Regulation of Thymocyte Development by Steroids

Neuroimmunomodulation 1999;6:90–96

Can Glucocorticoids Shape the T-Cell Repertoire? The presence of nearly normal numbers of peripheral T cells in adult GR-TKO mice, despite the reduced size of the thymus in young animals, allowed us to examine the antigen-specific repertoire of mice whose thymocytes are hyporesponsive to glucocorticoids. If the range of TCR avidities for self antigen/MHC are a continuum, the mutual antagonism model would predict that limiting glucocorticoid responsiveness would result in the elimination of cells that are normally at the high end of the low-tomoderate avidity spectrum. That is, as shown with TCR ·ß transgenic mice, ‘positive selection would be turned

93

into negative selection’. Thus, the peripheral repertoire of the GR-TKO mice should be ‘contracted’, with the cells surviving selection have relatively less avidity for self antigen/MHC than in wild-type animals. To test this, the GR antisense transgene was introduced into a spontaneously autoimmune mouse strain, MRL-lpr/lpr (referred to as lpr). lpr mice provide a model of spontaneous autoimmune disease that resembles human SLE, with autoantibodies, immune complex-mediated glomerulonephritis and arthritis, and a high incidence of mortality by 5 months of age [64, 65]. In addition, due to a defect in the fas gene that results in inefficient apoptosis of activated T cells, these animals develop progressive lymphadenopathy due to the accumulation of T cells with an unusual cell surface phenotype (TCR+Thy-1+CD4–CD8–B220+), so-called double negative T cells [66]. lpr mice transgenic for antisense GR (lpr.TKO mice) had near normal numbers of peripheral T cells that responded normally to TCR-mediated activation [67]. Strikingly, the lpr.TKO mice had retarded and substantially milder autoimmune disease than did their lpr nontransgenic counterparts. Anti-dsDNA and rheumatoid factor autoantibody titers were markedly decreased in the lpr.TKO mice, and histopathological analysis of renal disease indicated much less lymphocytic infiltration and evidence of glomerular disease. In agreement with these data, the 22-week survival rate in the lpr.TKO cohort was 93% versus 53% in lpr controls. In addition to ameliorated autoimmune disease, the lpr.TKO animals had a marked reduction in lymphoproliferative disease compared to nontransgenic lpr controls. At 5 months of age, for example, the lymphoid mass (combined weight of detectable lymph nodes, spleen, and thymus) of lpr females was increased 24-fold over the MRL non-lpr control mice, while the female lpr.TKO mice had an increase of only 6-fold. This was accounted for by less accumulation of the aberrant double negative T cells that are the hallmark of the lpr disease. These double negative T cells are thought to originate from activated T cells that fail to undergo normal deletion because they lack the apoptosisinducing Fas molecule [68, 69]. The decrease in their accumulation, therefore, is due to the lessened autoreactivity in these mice. Ideally, one would like to be able to analyze the antigen-specific repertoire of the lpr.TKO mice and demonstrate that it has been altered (away from high avidity anti-self) by the presence of the antisense GR transgene. This is not possible for lpr-associated autoimmunity, however, because there is no dominant antigenic epitope recognized by the autoreactive T cells. It was possible to address the question of whether the repertoire had

94

Neuroimmunomodulation 1999;6:90–96

been altered, however, by quantitating the use of particular Vß-bearing TCRs in mice with and without the antisense transgene. Certain Vß-bearing thymocytes are positively selected by particular MHC class I or class II molecules, presumably complexed with peptide antigen, a process that occurs independently of non-MHC-encoded genes and superantigens [70]. In the case of H-2k mice (the haplotype of the MRL-lpr/lpr strain), this results in overrepresentation of Vß4 and Vß10 in CD4+ cells and Vß2 and Vß6 and Vß14 in CD8+ cells. Although this was confirmed for lpr mice, in each instance the expression of the positively selected Vß was blunted (reduced to or near levels found in animals with nonselecting haplotypes) in lpr.TKO mice. This finding represents a physiologic example in which hyporesponsiveness to glucocorticoids has prevented positive selection of cells that recognize self antigen/MHC, presumably by causing their deletion. In addition to confirming the regulatory role that glucocorticoids play in thymocyte selection, these results raise the possibility that dysregulation of glucocorticoid responses in the thymus might influence the predisposition of an animal to autoimmune disease. For example, small changes in thymocyte GR number, thymic production of glucocorticoids, or the sensitivity of thymocytes to the antagonistic action of glucocorticoids could conceivably allow the survival of autoreactive cells that would otherwise undergo negative selection.

Conclusion

The observations reviewed here demonstrate that the immune system, and in particular the thymus, is subject to regulation by steroid hormones, and that this regulation contributes to both normal homeostasis and the development of antigen-specific immunity. The relationship between the prevalence of certain sex hormones and disease is intriguing, but at this time a mechanistic framework has not been developed. In contrast, it is becoming clear that glucocorticoids contribute to the shaping of the antigen-specific T-cell repertoire by controlling the signaling threshold that discriminates positive from negative selection. It is an intriguing possibility that aberrant regulation of glucocorticoid production in the thymus or thymocyte responsiveness to glucocorticoids might play a role in some forms of autoimmune disease.

Tolosa/Ashwell

References 1 van Oers NS, von Boehmer H, Weiss A: The pre-T-cell receptor (TCR) complex is functionally coupled to the TCR-zeta subunit. J Exp Med 1995;182:1585–1590. 2 Kisielow P, von Boehmer H: Development and selection of T cells: Facts and puzzles. Adv Immunol 1995;58:87–209. 3 van Ewijk W: T-cell differentiation is influenced by thymic microenvironments. Annu Rev Immunol 1991;9:591–615. 4 Hoffmann MW, Allison J, Miller JF: Tolerance induction by thymic medullary epithelium. Proc Natl Acad Sci USA 1992;89:2526–2530. 5 Schneider SC, Sercarz EE: Antigen processing differences among APC. Hum Immunol 1997; 54:148–158. 6 Volkmann A, Zal T, Stockinger B: Antigen-presenting cells in the thymus that can negatively select MHC class II-restricted T cells recognizing a circulating self antigen. J Immunol 1997; 158:693–706. 7 Hugo P, Kappler JW, Godfrey DI, Marrack PC: A cell line that can induce thymocyte positive selection. Nature 1992;360:679–682. 8 Vukmanovic S, Grandea AG III, Faas SJ, Knowles BB, Bevan MJ: Positive selection of T-lymphocytes induced by intrathymic injection of a thymic epithelial cell line. Nature 1992;359:729–732. 9 Anderson G, Owen JJ, Moore NC, Jenkinson EJ: Thymic epithelial cells provide unique signals for positive selection of CD4+CD8+ thymocytes in vitro. J Exp Med 1994;179:2027– 2031. 10 Abraham N, Miceli MC, Parnes JR, Veillette A, Abbot S: Enhancement of T-cell responsiveness by the lymphocyte-specific tyrosine protein kinase p56lck. Nature 1991;350:62–66. 11 Kasai M, Hirokawa K, Kajino K, Ogasawara K, Tatsumi M, Hermel E, Monaco JJ, Mizuochi T: Difference in antigen presentation pathways between cortical and medullary thymic epithelial cells. Eur J Immunol 1996;26:2101– 2107. 12 Wekerle H, Ketelsen UP: Thymic nurse cells – Ia-bearing epithelium involved in T-lymphocyte differentiation? Nature 1980;283:402– 404. 13 Haynes BF, Shimizu K, Eisenbarth GS: Identification of human and rodent thymic epithelium using tetanus toxin and monoclonal antibody A2B5. J Clin Invest 1983;71:9–14. 14 Li Y, Pezzano M, Philp D, Reid V, Guyden J: Thymic nurse cells exclusively bind and internalize CD4+CD8+ thymocytes. Cell Immunol 1992;140:495–506. 15 Wick G, Rieker T, Penninger J: Thymic nurse cells: A site for positive selection and differentiation of T cells. Curr Top Microbiol Immunol 1991;173:99–105. 16 Pezzano M, Li Y, Philp D, Omene C, Cantey M, Saunders G, Guyden JC: Thymic nurse cell rescue of early CD4+CD8+ thymocytes from apoptosis. Cell Mol Biol 1995;41:1099–1111.

Regulation of Thymocyte Development by Steroids

17 Tsai MJ, O’Malley BW: Molecular mechanisms of action of steroid/thyroid receptor superfamily members. Annu Rev Biochem 1994; 63:451–486. 18 Almawi WY, Beyhum HN, Rahme AA, Rieder MJ: Regulation of cytokine and cytokine receptor expression by glucocorticoids. J Leukoc Biol 1996;60:563–572. 19 Jenster G, Spencer TE, Burcin MM, Tsai SY, Tsai MJ, O’Malley BW: Steroid receptor induction of gene transcription: A two-step model. Proc Natl Acad Sci USA 1997;94:7879–7884. 20 Shibata H, Spencer TE, Onate SA, Jenster G, Tsai SY, Tsai MJ, O’Malley BW: Role of coactivators and co-repressors in the mechanism of steroid/thyroid receptor action. Recent Prog Horm Res 1997;52:141–164. 21 Bridges ED, Greenstein BD, Khamashta MA, Hughes GR: Specificity of estrogen receptors in rat thymus. Int J Immunopharmacol 1993;15: 927–932. 22 Viselli SM, Olsen NJ, Shults K, Steizer G, Kovacs WJ: Immunochemical and flow cytometric analysis of androgen receptor expression in thymocytes. Mol Cell Endocrinol 1995;109: 19–26. 23 Clarke AG, Kendall MD: The thymus in pregnancy: The interplay of neural, endocrine and immune influences. Immunol Today 1994;15: 545–551. 24 Screpanti I, Morrone S, Meco D, Santoni A, Gulino A, Paolini R, Crisanti A, Mathieson BJ, Frati L: Steroid sensitivity of thymocyte subpopulations during intrathymic differentiation. Effects of 17-beta-estradiol and dexamethasone on subsets expressing T cell antigen receptor or IL-2 receptor. J Immunol 1989;142: 3378–3383. 25 Ben-Nathan D, Lachmi B, Lustig S, Feuerstein G: Protection by dehydroepiandrosterone in mice infected with viral encephalitis. Arch Virol 1991;120:263–271. 26 Secreto G, Zumoff B: Abnormal production of androgens in women with breast cancer. Anticancer Res 1994;14:2113–2117. 27 Stahl F, Schnorr D, Pilz C, Dorner G: Dehydroepiandrosterone (DHEA) levels in patients with prostatic cancer, heart diseases and under surgery stress. Exp Clin Endocrinol 1992;99: 68–70. 28 LaCroix AZ, Yano K, Reed DM: Dehydroepiandrosterone sulfate, incidence of myocardial infarction, and extent of atherosclerosis in men. Circulation 1992;86:1529–1535. 29 Wilder RL: Adrenal and gonadal steroid hormone deficiency in the pathogenesis of rheumatoid arthritis. J Rheumatol Suppl 1996;44: 10–12. 30 Blauer KL, Poth M, Rogers WM, Bernton EW: Dehydroepiandrosterone antagonizes the suppressive effects of dexamethasone on lymphocyte proliferation. Endocrinology 1991;129: 3174–3179. 31 May M, Holmes E, Rogers W, Poth M: Protection from glucocorticoid-induced thymic involution by dehydroepiandrosterone. Life Sci 1990;46:1627–1631.

32 Weinstein Y, Ran S, Segal S: Sex-associated differences in the regulation of immune responses controlled by the MHC of the mouse. J Immunol 1984;132:656–661. 33 Andersen AA, Hanson RP: Influence of sex and age on natural resistance to St. Louis encephalitis virus infection in mice. Infect Immun 1974; 9:1123–1125. 34 Nakanishi H, Horii Y, Terashima K, Fujita K: Effect of testosterone on the susceptibility of C57BL/6 mice to infection with Brugia pahangi with reference to inflammatory cell response. J Parasitol 1989;75:455–460. 35 Huber SA, Job LP, Auld KR, Woodruff JF: Sex-related differences in the rapid production of cytotoxic spleen cells active against uninfected myofibers during Coxsackievirus B-3 infection. J Immunol 1981;126:1336–1340. 36 Rifkind D, Frey JA: Sex difference in antibody response of CFW mice to Candida albicans. Infect Immun 1972;5:695–698. 37 Ansar Ahmed S, Penhale WJ, Talal N: Sex hormones, immune responses, and autoimmune diseases. Mechanisms of sex hormone action. Am J Pathol 1985;121:531–551. 38 Fox CJ, Danska JS: IL-4 expression at the onset of islet inflammation predicts nondestructive insulitis in nonobese diabetic mice. J Immunol 1997;158:2414–2424. 39 Steinberg AD, Roths JB, Murphy ED, Steinberg RT, Raveche ES: Effects of thymectomy or androgen administration upon the autoimmune disease of MRL/Mp-lpr/lpr mice. J Immunol 1980;125:871–873. 40 Michalski JP, McCombs CC, Roubinian JR, Talal N: Effect of androgen therapy on survival and suppressor cell activity in aged NZB/ NZW-F1 hybrid mice. Clin Exp Immunol 1983;52:229–233. 41 Kanda N, Tsuchida T, Tamaki K: Testosterone suppresses anti-DNA antibody production in peripheral blood mononuclear cells from patients with systemic lupus erythematosus. Arthritis Rheum 1997;40:1703–1711. 42 Dalal M, Kim S, Voskuhl RR: Testosterone therapy ameliorates experimental autoimmune encephalomyelitis and induces a T helper 2 bias in the autoantigen-specific T lymphocyte response. J Immunol 1997;159:3–6. 43 Williams PJ, Jones RH, Rademacher TW: Reduction in the incidence and severity of collagen-induced arthritis in DBA/1 mice, using exogenous dehydroepiandrosterone. Arthritis Rheum 1997;40:907–911. 44 Fox HS: Androgen treatment prevents diabetes in nonobese diabetic mice. J Exp Med 1992; 175:1409–1412. 45 Damek DM, Shuster EA: Pregnancy and multiple sclerosis. Mayo Clin Proc 1997;72:977– 989. 46 Okuyama R, Abo T, Seki S, Ohteki T, Sugiura K, Kusumi A, Kumagai K: Estrogen administration activates extrathymic T cell differentiation in the liver. J Exp Med 1992;175:661– 669.

Neuroimmunomodulation 1999;6:90–96

95

47 Rocha B, Guy-Grand D, Vassalli P: Extrathymic T cell differentiation. Curr Opin Immunol 1995;7:235–242. 48 Jaffe HL: The influence of the supradrenal gland on the thymus. I. Regeneration of the thymus following double supradrenalectomy in the rat. J Exp Med 1924;40:325–341. 49 Plaut M: Lymphocyte hormone receptors. Annu Rev Immunol 1987;5:621–669. 50 Baughman G, Harrigan MT, Campbell NF, Nurrish SJ, Bourgeois S: Genes newly identified as regulated by glucocorticoids in murine thymocytes. Mol Endocrinol 1991;5:637–644. 51 Wyllie AH: Glucocorticoid-induced thymocyte apoptosis is associated with endogenous endonuclease activation. Nature 1980;284:555– 556. 52 Gruber J, Sgonc R, Hu YH, Beug H, Wick G: Thymocyte apoptosis induced by elevated endogenous corticosterone levels. Eur J Immunol 1994;24:1115–1121. 53 Wyllie AH, Morris RG, Smith AL, Dunlop D: Chromatin cleavage in apoptosis: Association with condensed chromatin morphology and dependence on macromolecular synthesis. J Pathol 1984;142:67–77. 54 Cohen JJ: Glucocorticoid-induced apoptosis in the thymus. Semin Immunol 1992;4:363–369. 55 DeFranco D, Bali D, Torres R, DePinho RA, Erickson RP, Gluecksohn-Waelsch S: The glucocorticoid hormone signal transduction pathway in mice homozygous for chromosomal deletions causing failure of cell type-specific inducible gene expression. Proc Natl Acad Sci USA 1991;88:5607–5610.

96

56 Erickson RP, Gluecksohn-Waelsch S, Cori CF: Glucose-6-phosphatase deficiency caused by radiation-induced alleles at the albino locus in the mouse. Proc Natl Acad Sci USA 1968;59: 437–444. 57 Zacharchuk CM, Mercep M, Chakraborti P, Simons SS Jr, Ashwell JD: Programmed T lymphocyte death: Cell activation- and steroidinduced pathways are mutually antagonistic. J Immunol 1990;145:4037–4045. 58 Zacharchuk CM, Mercep M, Ashwell JD: Thymocyte activation and death: A mechanism for molding the T-cell repertoire. Ann NY Acad Sci 1991;636:52–70. 59 Iwata M, Hanaoka S, Sato K: Rescue of thymocytes and T-cell hybridomas from glucocorticoid- induced apoptosis by stimulation via the T cell receptor/CD3 complex: A possible in vitro model for positive selection of the T cell repertoire. Eur J Immunol 1991;21:643–648. 60 Vacchio MS, Papadopoulos V, Ashwell JD: Steroid production in the thymus: Implications for thymocyte selection. J Exp Med 1994;179: 1835–1846. 61 Zilberman Y, Yefenof E, Oron E, Dorogin A, Guy R: T cell receptor-independent apoptosis of thymocyte clones induced by a thymic epithelial cell line is mediated by steroids. Cell Immunol 1996;170:78–84. 62 Vacchio MS, Ashwell JD: Thymus-derived glucocorticoids regulate antigen-specific positive selection. J Exp Med 1997;185:2033–2038.

Neuroimmunomodulation 1999;6:90–96

63 King LB, Vacchio MS, Dixon K, Hunziker R, Margulies DH, Ashwell JD: A targeted glucocorticoid receptor antisense transgene increases thymocyte apoptosis and alters thymocyte development. Immunity 1995;3:647–656. 64 Theofilopoulos AN, Dixon FJ: Murine models of systemic lupus erythematosus. Adv Immunol 1985;37:269–390. 65 Cohen PL, Eisenberg RA: Lpr and gld: Single gene models of systemic autoimmunity and lymphoproliferative disease. Annu Rev Immunol 1991;9:243–269. 66 Watanabe-Fukunaga R, Brannan CI, Copeland NG, Jenkins NA, Nagata S: Lymphoproliferation disorder in mice explained by defects in Fas antigen that mediates apoptosis. Nature 1992;356:314–317. 67 Tolosa E, King LB, Ashwell JD: Thymocyte glucocorticoid resistance alters positive selection and inhibits autoimmunity and lymphoproliferative disease in MRL-lpr/lpr mice. Immunity 1998;8:67–76. 68 Budd RC, Mixter PF: The origin of CD4– CD8–TCR·ß+ thymocytes: A model based on T-cell receptor avidity. Immunol Today 1995; 16:428–431. 69 Kotzin BL, Babcock SK, Herron LR: Deletion of potentially self-reactive T cell receptor specificities in L3T4–, Lyt–2– T cells of lpr mice. J Exp Med 1988;168:2221–2229. 70 Tomonari K, Fairchild S, Rosenwasser OA: Influence of viral superantigens on Vß- and V·specific positive and negative selection. Immunol Rev 1993;131:131–168.

Tolosa/Ashwell

Neuroimmunomodulation 1999;6:97–107

Vasoactive Intestinal Peptide in Thymus: Synthesis, Receptors and Biological Actions Mario Delgado Carmen Martinez Javier Leceta Rosa P. Gomariz Departamento de Biologı´a Celular, Facultad de Biologı´a, Universidad Complutense, Madrid, Spain

Key Words Vasoactive intestinal peptide W Thymocytes W Neuroimmunomodulation W T-cell differentiation W Apoptosis W Gene expression W Neuropeptide secretion W Cytokine production

Abstract Evidence summarized in this report indicates that thymocytes produce and secrete VIP. Moreover, different stimuli such as Con A, LPS and anti-TCR antibody induce a significant increase in VIP production by thymocytes. In addition, proinflammatory cytokines such as IL-1, IL-6 and TNF-·, but not IL-2, stimulate in a similar timedependent manner VIP production by lymphocytes. We also describe the expression of VIP1 receptor and VIP2 receptor mRNA in murine thymocytes. Thus, VIP released in thymus microenvironment may modulate immune functions through direct binding to VIP receptors on thymocytes. Our functional data support that VIP through the interaction with their specific receptors affect three important aspects of thymocytes function: cytokine production, mobility and apoptosis.

ABC

© 1999 S. Karger AG, Basel 1021–7401/99/0062–0097$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: http://BioMedNet.com/karger

Introduction

In the last few decades, as a result of the interaction between different areas of research, the new interdisciplinary and exciting field of neuroimmunology has emerged. Neuroimmunomodulation is defined as the interaction among the nervous, endocrine, and immune systems, mediated through cytokines, peptidic hormones, neuropeptides, and their corresponding receptors. Several immune-neuroendocrine interactions are known to occur both in primary and secondary lymphoid organs. By 1985, evidence from the literature could be advanced to provide a rational basis for a neuroendocrine function of the thymus. The importance of the thymus in the maturation and differentiation of T lymphocytes and the induction of self-tolerance has long been established. T-cell differentiation is a highly complex process which includes the progressive acquisition of different membrane markers and the arrangement of the genes coding for the antigen-specific T-cell receptor (TCR). It seems that, rather than following pure automatic genetically programmed mechanisms, the selection of T-cell repertoire and TCR gene rearrangements are controlled and probably induced by the thymic microenvironment. Moreover, T-cell growth and differentiation are also subjected to neuroendocrine regulation involving endocrine, paracrine and autocrine mechanisms [1]. By using different methodological approaches, both the intrathymic production of several neuropeptides and the expression of their specific

Rosa P. Gomariz Departamento de Biologı´a Celular Facultad de Biologı´a, Universidad Complutense E–28040 Madrid (Spain) Tel. +34 1 3944971, Fax +34 1 3944981, E-Mail [email protected]

receptors by thymic epithelial cells and thymocytes have been demonstrated. Three different sources of thymic neuropeptides have been described, i.e. the innervating fibers, the thymic epithelial cells and, more recently, the developing thymocytes. The thymic innervation includes the presence of substance P (SP), neuropeptides Y (NPY), somatostatin (SOM) and enkephalin (ENK) [2, 3]. Thymic epithelial cells are the source of SOM, growth hormone (GH), adrenocorticotropic hormone (ACTH), luteinizing hormone (LH), follicle-stimulating hormone (FSH), thyroid-stimulating hormone (TSH), ß-endorphin, neurotensin (NT), neurophysins, oxytocin (OT), and vasopressin (AVP) [1, 4–6]. Finally, thymocytes themselves produce different hormones and neuropeptides such as GH, prolactin (PRL), LH, luteinizing hormone-releasing hormone (LHRH), SOM, ENK, nerve growth factor (NGF) and atrial natriuretic peptide [1, 2, 7–11]. Furthermore, these hormones and neuropeptides exert their immune action through specific cell surface receptors on thymocytes, including GH, PRL, ACTH, OT, AVP, LHRH, growth hormone-releasing hormone (GHRH), and ß-endorphin receptors [1]. The aim of this review is to summarize the evidence accumulated with regard to the production of vasoactive intestinal peptide (VIP) by thymocytes and its possible functional implications.

Synthesis of VIP in Thymus

VIP, a 28-amino-acid neuropeptide has a broad tissue distribution first isolated from hog intestinal extracts [12], and later from the central nervous system [13]. VIP may act as a neurotransmitter, hormone and/or immunomodulator. In immune system, VIP affects various immune responses, such as lymphocyte adhesion and traffic, proliferation of peripheral T cells, immunoglobulin production, natural killer cell and macrophage activity and cytokine production [14–16]. Immunohistochemical methods have revealed VIPcontaining nerves that distribute to the lymphoid organs. In thymus, VIP-like immunoreactive (ir) fibers are most abundant in the capsule and extend into the interlobular septa. Varicose VIP-ir nerves in the capsular/interlobular septal system form linear arrays that come in close association with mast cells. Fine VIP+ fibers exit this plexus to enter the thymic cortex, particularly the deep regions, and surround thymic cells. Occasionally, VIP-ir fibers are found in the medulla [2, 3]. Surgical sympathectomy did not alter the thymic VIP content, and the innervation of this organ was not altered, suggesting an origin for the

98

Neuroimmunomodulation 1999;6:97–107

VIP+ fibers other than the sympathetic nervous system [17]. In addition, several studies suggest that VIP may be synthesized and secreted by different inflammatory and immune cells such as mast cells [18–20], mononuclear and polymorphonuclear leukocytes [21–24], and more recently, lymphocytes [6, 25–29]. The first evidence for the production of VIP by cells of the thymus was the identification of VIP immunoreactivity in histological sections from mouse and rat thymus [6, 25]. VIP+ thymocytes were numerous in the deep cortex and scattered in the medulla. The VIP immunoreactivity was localized in both small and large lymphocytes. These results were later confirmed in thymocyte cell suspensions by both light and electron microscopy [28] (fig. 1A). The VIP protein was distributed in small round or elongated vesicles 30–50 nm in diameter throughout the thymocyte cytoplasm. Flow cytometry analysis with a polyclonal anti-VIP antiserum showed that 11% of thymocytes were VIP+ [28]. However, recent studies performed in our laboratory using a recently developed monoclonal anti-VIP antibody [30], indicated that the percentage of VIP+ thymocytes is slightly higher (35%; fig. 1C) [unpubl. data]. The flow cytometry analysis indicated that using the polyclonal antibody VIP+ thymocytes overlapped strongly with negative controls [28]; however, the monoclonal anti-VIP antibody has two well-separated populations of thymocytes, VIP+ and VIP– thymocytes (fig. 1C). Indeed, the monoclonal antibody has a higher specificity than the polyclonal antibody, and it is presumable that the data obtained with the monoclonal antibody have more guarantees. The biochemical characterization of thymic VIP by reverse phase-high performance liquid chromatography (RP-HPLC) and radioimmunoassay showed that most of the VIP-ir material eluted at the same retention time as the synthetic VIP1-28. In addition, the presence of two additional peaks with longer retention times, probably corresponding to higher molecular weight precursors, suggested the possibility that VIP may be synthesized by thymocytes (fig. 1B) [25]. The amounts of VIP detected in our thymocyte suspensions (0.5–2.5 pmol/108 thymocytes) were about 3–10 times higher than those reported for peripheral blood mononuclear cells in human and pig by Lygren et al. [22] and comparable to that found in mast cells [18, 20]. Evidence for the production of VIP by thymocytes has also been obtained by demonstrating the presence of VIP-specific mRNA. In situ hybridization localized the VIP mRNA in cells with a recognizable lymphoid morphology in the corticomedullary and medullary regions of the rat

Delgado/Martinez/Leceta/Gomariz

Fig. 1. VIP synthesis by thymocytes. A Cytocentrifuge preparations showing VIP-ir thymocytes. ! 980. B RP-HPLC

profile of VIP-like immunoreactivity in lymphocyte extracts from mouse thymus. The arrow shows the retention time of synthetic VIP. C Histogram of flow cytometry analysis of thymus cells subjected to single immunofluorescent staining for VIP using a monoclonal antibody. Dashed plot: control staining. D Above: VIP-specific PCR analysis of cerebral cortex (lane 1, positive control), hepatocytes (lane 2, negative control), and thymocytes (lane 3). Below: Southern blot hybridization of the PCR products with a labeled VIP-specific probe.

thymus [26]. The precise distribution of thymocytes containing VIP mRNA suggests that thymocytes express VIP at specific, later differentiation stages. VIP synthesis by thymocytes was also confirmed by RT-PCR (fig. 1D) [27, 31, 32]. We have reported that the VIP-specific PCR product appears in the double positive (CD4+CD8+) and single positive (CD4+ and CD8+) subsets (table 1) [31]; however, Xin et al. [32] found VIP mRNA in all thymocyte subpopulations, including double negative thymocytes (CD4–CD8–). This discrepancy could be due to the

different methods used to isolate the thymocyte subsets. Northern blot analysis of mRNA from rat thymocytes showed two transcripts for VIP of approximately 1.7 and 1.0 kb, with a predominant 1.0 kb transcript [33]. Although the 1.7 kb mRNA is preferentially expressed in most tissues, the lowest molecular form predominates in the rat anterior pituitary gland [34]. This transcript is the result of the utilization of the proximal polyadenylation signal of the preproVIP/PHI gene [35]. This could also occur in the lymphoid cells.

VIP in Thymus

Neuroimmunomodulation 1999;6:97–107

99

Several different groups have reported an increase in neuropeptide production by immune cells after mitogen stimulation in culture [36, 37]. Recent experiments performed in our laboratory have investigated the effect of concanavalin A (Con A) and lipopolysaccharide (LPS) on the production of VIP by rat thymocytes [submitted for publication]. In these studies, thymocytes were cultured for 48 h with or without Con A or LPS, and at different times, the percentage of VIP+ thymocytes was determined by flow cytometry. The results showed that both mitogens significantly stimulate in a time-dependent manner the VIP production by thymocytes (fig. 2A). The increase in the VIP production after mitogen stimulation was confirmed by the fact that VIP mRNA is more abundant in Con A-stimulated cells as determined by Northern blot analysis and in situ hybridization in cytospin preparations [33, and unpubl. data]. Moreover, the fact that a higher amount of immunoreactive product for VIP was detected in lymphoblasts and that it was associated with the endomembrane system [28] suggests that thymocytes are able to secrete VIP. Although VIP amounts in extracellular thymic suspension (10.4 B 1.4 ng/thymus) are greater than those in intracellular thymic suspension (352.7 B 31.6 pg/thymus) [6], detected by ELISA and radioimmunoassay, respectively, the possibility is not excluded that extracellular VIP in the thymic microenvironment is due to the VIP-ergic nerve terminals or to the classical endocrine pathway through the blood vessels (serum VIP concentration 295 B 19 pg/ml) [unpubl. data]. In order to clarify this question, we cultured thymocytes in presence or absence of several stimuli and assayed at different times the VIP content in the supernatants by a specific ELISA. We observed a significant increase in the secretion of VIP when thymocytes were incubated in the presence of Con A, LPS, or anti-TCR antibody (fig. 2B).

Fig. 2. Stimulation of production and release of VIP by thymocytes.

Thymocytes were incubated in the presence or absence of Con A or LPS. At different time points, the percentage of VIP-like immunoreactive thymocytes (A) and the concentration of VIP secreted by thymocytes (B) was determined by flow cytometry analysis and a specific ELISA, respectively. Data represent the mean B SD of 6–10 experiments performed in duplicate. * p ! 0.001 (ANOVA) with respect to control cultures without any stimulus.

Table 1. VIP, VIP1R and VIP2R gene

Thymocyte subpopulations

expression in different thymocyte subpopulations VIP VIP1R VIP2R

Ref.

CD4–CD8–

CD4+CD8+

CD4+CD8–

CD4–CD8+

– – +

+ + +

+ + +

+ + +

31 31 32

Thymocyte subpopulations from rat and mouse were isolated by different methods. mRNA was extracted and subjected to RT-PCR using specific primers for VIP, VIP1R and VIP2R genes. – = No expression; + = expression.

100

Neuroimmunomodulation 1999;6:97–107

Delgado/Martinez/Leceta/Gomariz

Fig. 3. Factors and intracellular pathways implicated in the regulation of VIP production by thymocytes. PKA = cAMP-dependent protein kinase; PKC = protein kinase C; TCR = T-cell receptor; PMA = phorbol esters.

Interestingly, the maximal VIP secretion was observed at 24–36 h. This is correlated with the peak of intracellular immunoreactive VIP, 12–24 h before maximal secretion (fig. 2A). In addition, proinflammatory cytokines such as IL-1, IL-6 and TNF-·, but not IL-2, stimulated in a similar time-dependent manner the VIP production by thymocytes [submitted for publication]. The 5)-flanking region of the VIP gene has been reported to contain potential NF-IL-6 binding sites [38]. Therefore, LPS-induced VIP production in thymocytes could be mediated through the induction of IL-1, IL-6 and/or TNF-· production by thymic stromal cells [39–41]. Indeed, the addition of antibodies against these proinflammatory cytokines inhibit partially the LPS stimulatory effect on VIP production by thymocytes [submitted for publication]. It has been shown that the expression of the VIP gene is regulated by various agents that cause an activation of signal transduction pathways mediated by cAMP, Ca2+, or protein kinase C [42–46]. The expression of the VIP gene is also lineage-specific [45]. At least two different 5)-flanking regions appear to be important for the lineage-specific expression, i.e. the tissue-specific element (TSE) and the cAMP-responsive element (CRE) [45, 46]. Indeed, forskolin (an enhancer of cAMP production) and PMA (a stimulator of protein kinase C) stimulate in a dose-dependent manner the secretion of VIP by rat thymocytes [submitted for publication]. Since PMA and forskolin showed

an addititive effect on VIP production, the two intracellular transduction pathways are probably independent. Thus, both protein kinase A and protein kinase C are involved in the activation of the VIP gene in thymic lymphocytes. Finally, we observed that VIP regulates positively its own production in rat thymocytes (fig. 3).

VIP in Thymus

Neuroimmunomodulation 1999;6:97–107

VIP Receptors in Thymus

VIP released in the microenvironment of the thymus can modulate immune functions through direct binding to VIP receptors on lymphoid cells. The existence and characterization of specific VIP binding sites on immune cells and lymphocytic cell lines has been documented initially through binding studies [reviewed in 47]. Recently, three different VIP receptors have been cloned primarily from rat and human nonlymphoid cDNA libraries. The type 1 VIP receptor (VIP1R) cloned from rat lung and from human HT29 intestinal epithelial cells is expressed primarily in the cerebral cortex and hippocampus in the central nervous system and in various organs such as lung and intestine [48–50]. The type 2 VIP receptor (VIP2R) cloned from rat pituitary, mouse MIN-6 pancreatic cells, and human SUP-T1 lymphoblasts is expressed in the central nervous system in areas different from those expressing VIP1R, such as the olfactory bulb, hypothalamus and

101

Fig. 4. Expression of VIP1R and VIP2R in

the cellular surface of thymocytes. Histogram of flow cytometry analysis of thymocytes subjected to single immunofluorescent staining for VIP2R (A) and VIP1R (B) using two specific polyclonal antibodies. Dashed plot: control staining.

suprachiasmatic nuclei, and in various glands and organs such as the pituitary, pineal, pancreatic islets, stomach, colon, testes and ovary [50–53]. The two VIP receptors have different tissue distribution, but seem functionally analogous exhibiting similar peptide binding specificity and affinity [48–53]. In addition to these two receptors, a third class of VIP binding site is the PACAP receptor, that recognizes both forms of the pituitary adenylate cyclaseactivating polypeptide (PACAP27 and PACAP38), two neuropeptides structurally related with VIP, with the same affinity, but has a 300- to 1,000-fold lower affinity for VIP; this receptor is found in anterior pituitary, brain, astrocytes, neuroblastoma cells, adrenal cells, the rat pancreatic cancerous cell line AR 4-2J, and liver membranes [reviewed in 54]; although the expression and biochemical characterization of PACAP receptor in peritoneal macrophages have been reported [55], no expression has been demonstrated either in lymphocytes of peripheral lymphoid organs or in thymocytes [unpubl. data]. Initial binding studies indicate that murine thymocytes express less VIP binding sites than peripheral lymphocytes, as low as 150 sites per cell [56]. However, more recent binding and autoradiographic studies demonstrated the presence of both high- (Kd 1.12 nM) and low(Kd 88.5 nM) affinity VIP binding sites on avian thymocytes [57]. Moreover, binding studies indicated the presence of specific VIP receptors on T-cell lymphomas of thymic origin [58]. In addition, recent experiments performed in our laboratory, using polyclonal anti-VIP1R and anti-VIP2R antibodies [59, 60], indicate a significant difference between the expression of the two receptors in freshly isolated rat thymocytes. VIP2R is predominantly expressed (80% of thymocytes are VIP2R+ by flow cytometry) (fig. 4A), with VIP1R being more restricted (approximately 22% of thymocytes are VIP1R+ ) (fig. 4B). The VIP1R+ thymocytes were preferentially situated in the

102

Neuroimmunomodulation 1999;6:97–107

corticomedullary and medullary regions [unpubl. data]. At the present time, there are no data about the possible expression of the two VIP receptors in the same cell, although a definitive answer could be obtained using simultaneously antibodies against the two receptors. The expression and distribution of the two VIP receptors in thymus have recently been studied by different molecular techniques. Ishihara et al. [49] demonstrated a weak expression of VIP1R mRNA in rat thymus using Northern blot hybridization and PCR analysis. In addition, Usdin et al. [50] have found mRNA expression for both VIP1R and VIP2R in the murine thymus by in situ hybridization and RT-PCR. Labeling by the VIP1R probe was mostly within the cortical area, whereas both the cortex and medulla were labeled by the VIP2R probe [50]. We have recently described the expression of VIP1R and VIP2R mRNA in rat and mouse thymocytes [61–63]. Whereas VIP1R RNA appears to be constitutively expressed in cultured thymocytes, VIP2R mRNA is upon stimulation with anti-CD3 plus PMA or treatment with VIP [62, 63]. Studies regarding the distribution of the two VIP receptors in thymocyte subpopulations are very scarce. In rat thymocytes, we found that VIP1R mRNA was expressed in the double positive (CD4+CD8+) and single positive (CD4+ and CD8+) subsets, but not in the double negative (CD4–CD8–) cells (table 1) [31]. Xin et al. [32] have recently reported that in Balb/c mice only the single positive (CD4+ and CD8+) thymocytes express VIP1R mRNA, whereas all four subsets (CD4–CD8–, CD4+CD8+, CD4+CD8–, and CD4–CD8+) express VIP2R mRNA. The fact that the EL4.IL-2 cell line, representative of the immature double negative population, expresses only VIP2R [32, 62] is in agreement with the observation that double negative thymocytes express VIP2R but not VIP1R mRNA [32].

Delgado/Martinez/Leceta/Gomariz

Although there are many reports describing the role of VIP as a immunomodulator in the peripheral lymphoid organs, the effects on thymic cells have been scarcely studied. At the present time, VIP has been shown to affect three areas of thymocyte function: cytokine production, mobility and apoptosis. (1) Cytokine production: The immunomodulatory activities of VIP are mediated, at least partially, through the effects on the production of cytokines, such as IL-2, IL-4,

IL-5, IL-10 [reviewed in 16]. In thymus, the maturation of self-MHC restricted and self-tolerant T-cells requires multiple interactions with nonlymphoid cells such as thymic epithelial cells, fibroblasts, dendritic cells, and macrophages, either via direct cellular contacts or through soluble factors. Thymic IL-2, IL-4 and IL-10 production is limited to specific thymocyte subpopulations and is differentially expressed during development [65, 66]. In this context, the modulation of these cytokines by neuropeptides present in the thymic microenvironment could play a role in T-cell maturation. Previous reports indicated that VIP inhibits IL-2, IL-4 and IL-10 production in unfractionated thymocytes stimulated through the TCR/CD3 receptor complex [67, 68]. The inhibitory effect of VIP on IL-2 and IL-10 production occurs at the transcriptional level [68, 69], whereas the effect of IL-4 is posttranscriptional and indirect, being mediated through the effect on IL-2 [69]. Studies with isolated thymocyte subpopulations confirmed that VIP inhibits IL-2 production only when thymocytes were stimulated through the TCR [32], suggesting that VIP interferes with TCRengaged transduction pathways. In addition, comparative studies of the effect of VIP and the expression of VIP receptors in thymocyte subsets and T-cell lines indicated that the presence of VIP2R is sufficient to mediate the inhibitory effect on IL-2 [32]. This is in agreement with the observation that either VIP1R or VIP2R can mediate the inhibitory effect of VIP on IL-2 and IL-10 production in splenocytes [unpubl. data]. Although VIP has been shown to induce cAMP in thymocytes [70, 71], and cAMP-inducing agents such as forskolin were shown to inhibit IL-2 and IL-10 production [62, 68, 72, 73], the link between intracellular cAMP and the various transducers induced by TCR engagement is not known at the present time. (2) Mobility: The migration capacity is an important property of lymphocytes and the bidirectional migration or chemotaxis is a first and crucial event in the immune response. Experimental evidence suggests that VIP may influence the tissue distribution of lymphocytes [14, 15]. In thymus, VIP inhibits in a specific manner the in vitro chemotaxis and spontaneous mobility of thymocytes, elevating the intracellular cAMP levels by binding to VIP1R [71]. Although the physiological relevance of the inhibitory effect of VIP on thymocyte chemotaxis and mobility is not clear, this suggests a possible effect on thymocyte homing and/or thymic migration. (3) Apoptosis: VIP is a neuropeptide that has been implicated in mechanisms related to the proliferation and maturation of the nervous and immune system. VIP has

VIP in Thymus

Neuroimmunomodulation 1999;6:97–107

The physiological significance of the distinct distribution of the two VIP receptors in thymocyte subsets is unresolved at the present time. In most organs the VIP1R and VIP2R do not overlap anatomically, and in some cases, particularly in brain, the distribution seems to be complementary [50]. However, the issue of distinct VIP receptor distribution in thymus is complicated by the fact that various thymocyte subsets express various maturation stages. Therefore, the expression of particular VIP receptors could be developmentally regulated. Indeed, the expression of VIP2R appears to be developmentally regulated in several systems [50]. In addition, stimulation of thymocytes could also play a role in differential VIP receptor expression. For example, Metwali et al. [64] reported that only VIP1R is expressed in the thymus of schistosome-infected mice. We also showed that VIP2R mRNA is induced following stimulation of unfractionated normal thymocytes through the TCR or following treatment with VIP [62]. Similar studies with thymocyte subsets have yet to be performed. The correlation between VIP receptors, mRNA levels and receptor protein expression on the surface of thymocytes is difficult to ascertain at the present time. The expression of VIP receptors on the surface of thymocytes may be regulated both at transcriptional and posttranscriptional level. In splenocytes for example, following Con A stimulation there is an increase in the expression of VIP1R and VIP2R at the protein level at 24 h followed by a decrease to basal levels at 48 h [unpubl. data]. However, the VIP1R mRNA levels did not appear to change significantly after Con A stimulation, whereas VIP2R gene expression increased at both 24 and 48 h [62]. Although a definitive answer will be obtained only in studies with purified lymphocyte subpopulations, these data suggest that both transcriptional and posttranscriptional regulatory mechanisms play a role in the expression of VIP receptors.

Biological Actions of VIP in Thymus

103

Fig. 5. VIP protects thymocytes from gluco-

corticoid-induced apoptosis. Thymocytes were cultured with medium alone (controls, 1), with dexamethasone (2), or dexamethasone plus VIP (3). DNA was extracted and separated by agarose gel electrophoresis and cytocentrifugation was performed to visualize the fragmentation and condensation of nuclear chromatin characteristic of apoptosis.

been involved in the regulation and maintenance of different neuronal populations [72, 73], and as a differentiating agent for the human neuroblastoma cell line SK-N-SH [74]. One of the initial observations about the effect of VIP in the immune system was the inhibition of T-cell proliferation in response to mitogen stimulation. In this respect, VIP inhibits in a dose-dependent manner the mitogenesis of human and murine thymocytes stimulated with phytohemagglutinin and through the TCR/CD3 complex [67, 75, 76], suggesting a role in the differentiation, activation, and/or proliferation of T cells. In a recent study, we have described that VIP protects thymocytes from glucocorticoid-induced apoptosis [77]. VIP was shown to inhibit the DNA fragmentation characteristic of glucocorticoid-induced apoptosis and to increase the cell survival of thymocytes (fig. 5), through a VIP1R-mediated mechanism. Phenotypic analysis showed that VIP protects predominantly CD4+CD8+ thymocytes from glucocorticoid-induced apoptosis [77]. Similar results were obtained by Ernstrom et al. [78] with human and mouse thy-

104

Neuroimmunomodulation 1999;6:97–107

mocytes exposed to a cytolytic dose of prednisolone in vitro. These results suggest that the binding of VIP to VIP1R triggers a signaling pathway responsible for the acquisition of the resistance to glucocorticoids. In thymus, the T-cell precursors migrate in a directed manner from cortex into the medulla, a process accompanied by changes in receptors and adhesion molecules, and by positive and negative selection, which shape the T-cell repertoire. Thymic differentiation of T cells is associated with a complex and precisely regulated gene expression process, proceeding from CD4–8– cells to mature CD4+8– or CD4–8+ thymocytes. The intermediate stage CD4+8+TCRlow thymocytes, which represent approximately 80% of all thymocytes, are induced to differentiate upon interaction with MHC (positive selection). Cells incapable of recognizing self-MHC are eliminated, presumably through endogenous glucocorticoid-induced apoptosis. This type of thymocyte elimination is called death by neglect.

Delgado/Martinez/Leceta/Gomariz

Fig. 6. Summary of synthesis, receptors and biological actions of VIP on thymus. TSC = Thymic stromal cells.

Besides nonreactive thymocytes which die by neglect, the strong self-reactive potentially autoimmune cells are eliminated through negative selection. Most experimental results favor the affinity/avidity model for thymic selection, which predicts that low affinity/avidity interactions between the ·ß TCR and a MHC/peptide ligand results in positive selection, whereas high-affinity/avidity interactions result in negative selection. Similar to death by neglect, negative selection proceeds through apoptosis. However, whereas the former appears to be mediated by glucocorticoids, the latter is induced through TCR-mediated signals. These findings, together with evidence for the presence of VIP and the expression of VIP receptors during thymocyte development and the protective effect of VIP against glucocorticoid-induced apoptosis in the mouse, suggest that this neuropeptide may be involved in intrathymic T-cell maturation. A bidirectional relationship appears to exist between glucocorticoids and VIP/ VIP receptors. In tissues other than thymus, the expression of the VIP and VIP1R genes was shown to be regu-

lated by glucocorticoids [79, 80]. In agreement with these reports, we found that dexamethasone stimulates in a time- and dose-dependent manner VIP production in thymocytes [submitted for publication]. Finally, the VIP secreted from the peptidergic nerve endings found in thymus, may also play a role in intrathymic T-cell maturation. However, a recent study by Mitchell et al. [81] using severe combined immunodeficient mice suggests that the VIP+ innervation is more important for the establishment of the thymic microenvironment than for the maintenance of thymocyte differentiation. In conclusion, VIP synthesized and secreted by thymocytes themselves could regulate in an autocrine/paracrine manner the differentiation and maturation of T cells, modulating thymocyte homing, the positive selection, and the production of cytokines which participate as soluble factors in the thymic microenvironment. In the same way, several immunological factors, such as proinflammatory cytokines, bacterial components and glucocorticoids, could regulate indirectly these functions modulating the

VIP in Thymus

Neuroimmunomodulation 1999;6:97–107

105

production of VIP in the thymus (fig. 6). The present review provides an additional link between neuropeptides and immune system, and suggests that VIP can be included as a novel immunoregulatory peptide that can modulate in vivo physiology of the thymus.

Acknowledgments We thank Dr. Doina Ganea from Newark University, N.J., for improvement of the manuscript. This work is supported by Grant PB94-0310 from the Spanish Department of Education and Science.

References 1 Dardenne M, Savino W: Control of thymus physiology by peptidic hormones and neuropeptides. Immunol Today 1994;15:518–523. 2 Bellinger DL, Lorton D, Romano TD, Olschowka JA, Felten SY, Felten DL: Neuropeptides and immunopeptides. Ann NY Acad Sci 1990;594:17–33. 3 Felten DL, Felten SY, Carlson SL, Olschowka JA, Livnat S: Noradrenergic and peptidergic innervation of lymphoid tissue. J Immunol 1985;135:755S–765S. 4 Geenen V, Cormann-Goffin N, Martens H, Vardermissen E, Robert F, Benhida A, Legros JJ, Martial J, Franchimont P: Thymic neurohypophysial related peptides and T-cell selection. Regul Pept 1993;45:273–278. 5 Blalock JE: A molecular basis for bidirectional communication between the immune and neuroendocrine systems. Physiol Rev 1989;69:1– 32. 6 Gomariz RP, Lorenzo MJ, Cacicedo L, Vicente A, Zapata AG: Demonstration of immunoreactive vasoactive intestinal peptide and somatostatin in rat thymus. Brain Behav Immun 1990; 4:151–161. 7 O’Neil KD, Montgomery DW, Truong TM, Yu-Lee LY: Prolactin gene expression in human thymocytes. Mol Cell Endocrinol 1992; 87:R19–R23. 8 Rosen H, Behar O, Abramsky O, Ovadia D: Regulated expression of proenkephalin A in normal lymphocytes. J Immunol 1989;143: 3703–3707. 9 Fuller PJ, Verity K: Somatostatin gene expression in the thymus gland. J Immunol 1989;143: 1015–1017. 10 Santambrogio L, Benedetti M, Chao MV, Muzaffar R, Kulig K, Gabellini N, Hochwald G: Nerve growth factor production by lymphocytes. J Immunol 1994; 153:4488–4495. 11 Vollmar AK, Lang RE, Hanze A, Schulz R: The rat thymus: A site of atrial natriuretic peptide synthesis. Peptides 1990;11:33–37. 12 Said SI, Mutt V: Potent peripheral and splanchnic vasodilator peptide from normal gut. Nature 1970;225:863–864. 13 Carlquist M, Jornvall H, Tatemoto K, Mutt V: A porcine brain polypeptide is identical to the vasoactive intestinal polypeptide. Gastroenterology 1982;83:245–250. 14 De la Fuente M, Delgado M, Gomariz RP: VIP modulation of immune cell functions. Adv Neuroimmunol 1996;6:75–91. 15 Bellinger LD, Lorton D, Brouxhon S, Felten S, Felten DL: The significance of vasoactive intestinal polypeptide in immunomodulation. Adv Neuroimmunol 1996;6:5–27.

106

16 Ganea D: Regulatory effects of vasoactive intestinal peptide on cytokine production in central and peripheral lymphoid organs. Adv Neuroimmunol 1996;6:61–74. 17 Bellinger DL, Lorton D, Horn L, Brouxhon S, Felten SY, Felten DL: Vasoactive intestinal polypeptide innervation of rat spleen, thymus, and lymph nodes. Peptides 1997;18:1139– 1149. 18 Cutz E, Chan W, Track NS, Goth A, Said SI: Release of vasoactive intestinal peptide in mast cells by histamine liberators. Nature 1978;275: 661–662. 19 Goetzl EJ, Grotmol T, Van Dyke RW, Turck CW, Wershil B, Galli SJ, Sreedharan SP: Generation and recognition of vasoactive intestinal peptide by cells of the immune response. Ann NY Acad Sci 1990;131:34–44. 20 Wershil BK, Turck CW, Sreedharan SP, Yang J, An S, Galli SJ, Goeltz EJ: Variants of vasoactive intestinal peptide in mouse mast cells and rat basophilic leukemia cells. Cell Immunol 1993;151:369–378. 21 O’Dorisio MS, O’Dorisio T, Cotland S, Bakerzak SP: Vasoactive intestinal polypeptide as a biochemical marker for polymorphonuclear leukocytes. J Lab Clin Med 1980;96:666–672. 22 Lygren I, Revhaug A, Burgol PG, Giercksky KE, Jenssen TG: Vasoactive intestinal peptide and somatostatin in leukocytes. Scand J Clin Lab Invest 1984;44:347–351. 23 Turk CW, Aliakbari J, Sreedharan SP, Goeltz EJ: Distinctive vasoactive intestinal peptides from rat basophilic leukemia cells. FASEB J 1988;2:A1234. 24 Weinstock JV, Blum AM: Detection of vasoactive intestinal peptide and localization of its mRNA within granulomas of murine schistosomiasis. Cell Immunol 1990;125:291–300. 25 Gomariz RP, De la Fuente M, Hernanz A, Leceta J: Occurrence of vasoactive intestinal peptide in lymphoid organs from rat and mouse. Ann NY Acad Sci 1992;650:13–18. 26 Gomariz RP, Delgado M, Naranjo JR, Mellstrom B, Tormo A, Mata F, Leceta J: VIP gene expression in rat thymus and spleen. Brain Behav Immun 1993; 7:271–278. 27 Gomariz RP, Leceta J, Garrido E, Garrido T, Delgado M: Vasoactive intestinal peptide (VIP) mRNA expression in rat T and B lymphocytes. Regul Pept 1994; 50:177–184. 28 Leceta J, Martinez MC, Delgado M, Garrido E, Gomariz RP: Lymphoid cell subpopulations containing vasoactive intestinal peptide in the rat. Peptides 1994;15:791–797.

Neuroimmunomodulation 1999;6:97–107

29 Delgado M, Pozo D, Martinez C, Garrido E, Leceta J, Calvo JR, Gomariz RP: Characterization of gene expression of VIP and VIP1-receptor in rat peritoneal lymphocytes and macrophages. Regul Pept 1996;62:161–166. 30 Wong HC, Sternini C, Lloyd K, De Giorgio R, Walsh JH: Monoclonal antibody to VIP: Production, characterization, immunoneutralizing activity, and usefulness in cytochemical staining. Hybridoma 1996;15:133–139. 31 Delgado M, Martinez C, Leceta J, Garrido E, Gomariz RP: Differential VIP and VIP1 receptor gene expression in rat thymocyte subsets. Peptides 1996;17:803–807. 32 Xin Z, Jiang X, Wang HY, Denny TN, Dittel BN, Ganea D: Effect of vasoactive intestinal peptide (VIP) on cytokine production and expression of VIP receptors in thymocyte subsets. Regul Pept 1997;72:41–54. 33 Leceta J, Martinez C, Delgado M, Garrido E, Gomariz RP: Expression of vasoactive intestinal peptide in lymphocytes: A possible endogenous role in the regulation of the immune system. Adv Neuroimmunol 1996;6:29–36. 34 Lara JI, Lorenzo MJ, Cacicedo L, Tolon R, Balsa JA, Lopez-Fernandez J, Sanchez-Franco F: Induction of vasoactive intestinal peptide gene expression and prolactin secretion by insulinlike growth factor I in rat pituitary cells: Evidence for an autoparacrine regulatory system. Endocrinology 1994;135:2526–2532. 35 Chew LJ, Murphy D, Carter DA: Alternatively polyadenylated vasoactive intestinal peptide mRNAs are differentially regulated at the level of stability. Mol Endocrinol 1994;8:603–613. 36 Zurawski G, Benedik M, Kamb BJ, Abrams JS, Zurawski SM, Lee FD: Activation of mouse Thelper cells induces abundant preproenkephalin mRNA synthesis. Science 1986;232:772– 775. 37 Clarke BL, Gebhardt BM, Blalock JE: Mitogen-stimulated lymphocytes release biologically active corticotropin. Endocrinology 1993; 132:983–988. 38 Baumann H, Symes AJ, Comeau MR, Morella KK, Wang Y, Friend D, Ziegler SF, Fink JS, Gearing DP: Multiple regions within the cytoplasmic domains of the leukemia inhibitory factor receptor and gp130 cooperate in signal transduction in hepatic and neuronal cells. Mol Cell Biol 1994;14:138–146. 39 Carding SR, Hayday AC, Bottomly K: Cytokines in T-cell development. Immunol Today 1991;12:239–244.

Delgado/Martinez/Leceta/Gomariz

40 Zlotnik A, Moore TA: Cytokine production and requirements during T-cell development. Curr Opin Immunol 1995;7:206–213. 41 Deman J, Van Meurs M, Claassen E, Humblet C, Boniver J, Defresne MP: In vivo expression of interleukin-1ß (IL-1ß), IL-2, IL-4, IL-6, tumour necrosis factor-· and interferon-Á in fetal murine thymus. Immunology 1996;89:152–157. 42 Eiden LE, Hotchkiss AJ: Cyclic adenosine monophosphate regulates vasoactive intestinal polypeptide and enkephalin biosynthesis in cultured chromaffin cells. Neuropeptides 1983; 4:1–9. 43 Hayakawa Y, Obata KI, Itoh N, Yanaihara N, Okamoto H: Cyclic AMP regulation of provasoactive intestinal peptide/PHM-27 synthesis in human neuroblastoma cells. J Biol Chem 1984;259:9207–9211. 44 Tsukada T, Fink JS, Mandel G, Goodman RH: Identification of a region in the human vasoactive intestinal peptide gene responsible for regulation by cyclic AMP. J Biol Chem 1987;262: 8743–8747. 45 Waschek JA, Hsu CM, Eiden LE: Lineage-specific regulation of the vasoactive intestinal peptide gene in neuroblastoma cells is conferred by 5.2 kilobases of 5)-flanking sequence. Proc Natl Acad Sci USA 1988;85:9547–9551. 46 Fink JS, Verhave M, Walton K, Mandel G, Goodman RH: Cyclic AMP- and phorbol esterinduced transcriptional activation are mediated by the same enhancer element in the human vasoactive intestinal peptide gene. J Biol Chem 1991;266:3881–3887. 47 Calvo JR, Pozo D, Guerrero JM: Functional and molecular characterization of VIP receptors and signal transduction in human and rodent immune systems. Adv Neuroimmunol 1996;6:39–47. 48 Sreedharan SP, Huang JX, Cheung MC, Goetzl EJ: Structure, expression, and chromosomal localization of the type I human vasoactive intestinal peptide receptor gene. Proc Natl Acad Sci USA 1995;92:2939–2943. 49 Ishihara T, Shigemoto R, Mori K, Takahashi K, Nagata S: Functional expression and tissue distribution of a novel receptor for vasoactive intestinal peptide. Neuron 1992;8:811–819. 50 Usdin T, Bonner TI, Mezey E: Two receptors for vasoactive intestinal polypeptide with similar specificity and complementary distributions. Endocrinology 1994;135:2662–2680. 51 Svoboda M, Tastenoy M, Van Rampelbergh J, Goosens JF, De Neef P, Waelbroeck M, Robberecht P: Molecular cloning and functional characterization of a human VIP receptor from SUP-T1 lymphoblasts. Biochem Biophys Res Commun 1994;205:1617–1624. 52 Lutz EM, Sheward WJ, West KM, Morrow JA, Fink G, Harmar AJ: The VIP2 receptor: Molecular characterization of a cDNA encoding a novel receptor for vasoactive intestinal peptide. FEBS Lett 1993;334:3–8. 53 Inagaki N, Yoshida H, Mizuta M, Mizuno N, Fujii Y, Gonoi T, Miyazaki J, Susumu S: Cloning and functional characterization of a third PACAP receptor subtype expressed in insulinsecreting cells. Proc Natl Acad Sci USA 1994; 91:2679–2683.

VIP in Thymus

54 Christophe J: Type I receptors for PACAP (a neuropeptide even more important than VIP?). Biochim Biophys Acta 1993;1154:183–199. 55 Pozo D, Delgado M, Martinez C, Gomariz RP, Guerrero JM, Calvo JR: Functional characterization and mRNA expression of pituitary adenylate cyclase activating polypeptide (PACAP) type I receptors in rat peritoneal macrophages. Biochim Biophys Acta 1997;1359:250–262. 56 Ottaway CA, Greenberg GR: Interaction of vasoactive intestinal peptide with mouse lymphocytes: Specific binding and the modulation of mitogen response. J Immunol 1984;132:417– 423. 57 Lacey CB, Elde RP, Seybold VS: Localization of vasoactive intestinal peptide binding sites in the thymus and bursa of Fabricius of the chick. Peptides 1991;12:383–391. 58 Robberecht P, Abello J, Damien C, De Neef P, Vervisch E, Hooghe R, Christophe J: Variable stimulation of adenylate cyclase activity by vasoactive intestinal peptide-like peptides and beta-adrenergic agonists in murine T cell lymphomas of immature, helper and cytotoxic cells. Immunobiology 1989;179:422–431. 59 Ichikawa S, Sreedharan SP, Owen RL, Goetzl EJ: Immunolocalization of type I VIP receptor and NK-1-type substance P receptor in rat lung. Am J Physiol 1995;268:L584–L588. 60 Xia M, Sreedharan SP, Goetzl EJ: Predominant expression of type II vasoactive intestinal peptide receptors by human T lymphoblastoma cells: Transduction of both Ca2+ and cyclic AMP signals. J Clin Immunol 1996;16:21–30. 61 Gomariz RP, Garrido E, Leceta J, Martinez C, Abalo R, Delgado M: Gene expression of VIP receptor in rat lymphocytes. Biochem Biophys Res Commun 1994;203:1599–1604. 62 Delgado M, Martinez C, Johnson M, Gomariz RP, Ganea D: Differential expression of vasoactive intestinal peptide receptors 1 and 2 (VIP-R1 and VIP-R2) in murine lymphocytes. J Neuroimmunol 1996;68:27–38. 63 Johnson MC, McCormack RJ, Delgado M, Martinez C, Ganea D: Murine T lymphocytes express vasoactive intestinal peptide receptor 1 (VIP-R1) mRNA. J Neuroimmunol 1996;68: 109–119. 64 Metwali A, Elliot D, Blum AM, Sandor M, Weinstock JV: T-cell vasoactive intestinal peptide receptor subtype expression differs between granulomas and spleen of schistosomeinfected mice. J Immunol 1996;157:265–270. 65 Adkins B, Ghanei A, Hamilton K: Developmental regulation of IL-4, IL-2 and IFNÁ production by murine T lymphocytes. J Immunol 1993;151:6617–6626. 66 Lin T, Svetic Z, Ganea D, Rameshwar P, Gascon P, Gause W, Raveche E: Cytokines in NZB CD5+ B clones. Ann NY Acad Sci 1992;651: 581–589. 67 Xin Z, Tang H, Ganea D: VIP inhibits IL-2 and IL-4 production in murine thymocytes activated via TCR/CD3 complex. J Neuroimmunol 1994;54:59–68.

68 Martinez C, Delgado M, Gomariz RP, Ganea D: Vasoactive intestinal peptide and pituitary adenylate cyclase-activating polypeptide-38 inhibit IL-10 production in murine T lymphocytes. J Immunol 1996;156:4128–4136. 69 Wang HY, Xin Z, Tang H, Ganea D: Vasoactive intestinal peptide inhibits IL-4 production in murine T cells by a post-transcriptional mechanism. J Immunol 1996;156:3243–3253. 70 Jiang X, Wang HY, Yu J, Ganea D: VIP1 and VIP2 receptors but not PVR1 mediate the effect of VIP/PACAP on cytokine production in T lymphocytes. Ann NY Acad Sci 1998, in press. 71 Delgado M, De la Fuente M, Martinez M, Gomariz RP: Pituitary adenylate cyclase-activating polypeptide (PACAP27 and PACAP38) inhibit the mobility of murine thymocytes and splenic lymphocytes: Comparison with VIP and implication of cAMP. J Neuroimmunol 1995;62:137–146. 72 Pincus DW, DiCicco-Bloom E, Black IB: Vasoactive intestinal peptide regulates mitosis, differentiation and survival of cultured sympathetic neuroblasts. Nature 1990;343:564–566. 73 Brenneman DE, Westbrood GL, Firzgerald SP, Ennist DL, Elkins KL, Ruff MR, Pert CB: Neuronal cell killing by the envelope protein of HIV and its prevention by vasoactive intestinal peptide. Nature 1988;335:639–641. 74 Goosens JF, Manechez D, Pommery N, Formstecher P, Henichart JP: VIP potentiates retinoic-acid effect on tissue transglutaminase activity in human neuroblastoma, the SK-N-SH cells. Neuropeptides 1993;24:99–106. 75 Yiangou Y, Serrano R, Bloom SR, Peña J, Festenstein H: Effects of prepro-vasoactive intestinal peptide-derived peptides on the murine immune response. J Neuroimmunol 1990;29: 65–72. 76 Soder O, Hellstrom PM: Neuropeptide regulation of human thymocyte, guinea pig T lymphocyte, and rat B lymphocyte mitogenesis. Int Arch Allergy Appl Immunol 1987;84:205–211. 77 Delgado M, Garrido E, Martinez C, Leceta J, Gomariz RP: Vasoactive intestinal peptide and pituitary adenylate cyclase-activating polypeptides (PACAP27 and PACAP38) protect CD4+CD8+ thymocytes from glucocorticoidinduced apoptosis. Blood 1996;12:5152–5161. 78 Ernstrom U, Gafvelin G, Mutt V: Rescue of thymocytes from cell death by vasoactive intestinal peptide. Regul Pept 1995;57:99–104. 79 Chew LJ, Burke ZD, Morgan H, Gozes I, Murphy D, Carter DA: Transcription of the vasoactive intestinal peptide gene in response to glucocorticoids: Differential regulation of alternative transcripts is modulated by a labile protein in rat anterior pituitary. Mol Cell Endocrinol 1997;130:83–91. 80 Pei L: Identification of a negative glucocorticoid response element in the rat type 1 vasoactive intestinal peptide receptor gene. J Biol Chem 1996;271:20879–20884. 81 Mitchell B, Kendall M, Adam E, Schimacher U: Innervation of the thymus in normal and bone marrow reconstituted severe combined immunodeficient mice. J Neuroimmunol 1997;75:19–27.

Neuroimmunomodulation 1999;6:97–107

107

Neuroimmunomodulation 1999;6:108–114

Thymic Expression of the Pancreatic Endocrine Hormones Mark Throsby Jean-Marie Pleau Mireille Dardenne Françoise Homo-Delarche CNRS URA 1461, Université Paris V, Hôpital Necker, Paris, France

Key Words Preproinsulin W Proglucagon W Prosomatostatin W Propancreatic polypeptide W RT-PCR W Antigen-presenting cells W Mouse

Abstract The thymus plays a central role in the selection of T lymphocytes that are tolerant to ‘self’ antigens and responsive to foreign pathogens. We and others have reported the expression of the pancreatic endocrine hormones, preproinsulin, proglucagon, prosomatostatin and propancreatic polypeptide in the human and mouse thymus. While mRNA expression is very low there is evidence for the presence of the translated product. In addition, we have investigated the cell types responsible for expression. In the thymus, hormone expression is enriched in the antigen-presenting cell population. Interestingly, while proglucagon, prosomatostatin and propancreatic polypeptide appear to be expressed in a macrophage population, preproinsulin expression was restricted to dendritic cells which are more potent antigenpresenting cells. The functional significance of the endogenous expression of insulin in the thymus has been indirectly investigated using transgenic models in which the transgene is introduced by the rat insulin promoter. The data suggest that thymic expression of the transgene is critical in the induction of T-cell tolerance to the

ABC

© 1999 S. Karger AG, Basel 1021–7401/99/0062–0108$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: http://BioMedNet.com/karger

transgene in the periphery. Taken together, the evidence suggests that the low-level pancreatic hormone expression in the thymus may be involved in central tolerance to proteins of restricted expression.

Identification of Insulin in Thymus

The thymus, as a primary lymphoid organ, is responsible for promoting the proliferation and differentiation of a specific T-cell repertoire through which the immune system is able to distinguish ‘self’ antigen from invading pathogens and mutagens [1]. To facilitate this activity, the thymus depends on both cell-cell contact, between surface ligands such as the major histocompatibility complex (MHC) and the T-cell receptor (TCR), and humoral interaction [2, 3]. In addition to the many classical cytokines that exert their influence in an autocrine or paracrine manner, the thymus produces a number of unique humoral factors and many neuroendocrine and peripheral hormones [4]. The endocrine pancreatic hormones are a group of four polypeptides produced by different cell types that together constitute the islets of Langerhans [5]. Insulin, a 51amino-acid disulfide-bonded polypeptide, is synthesized by ß cells, which make up 70% of the islet mass in the adult. Its primary function is the maintenance of metabol-

Dr. Mark Throsby CNRS URA 1461, Université Paris V, Hôpital Necker 161, rue de Sèvres, F–75743 Paris Cedex 15 (France) Tel. +33 01 44 49 06 78, Fax +33 01 44 49 06 76, E-Mail [email protected]

Table 1. Thymic expression and functional significance of pancreatic hormones, pancreatic hormone transgene constructs and auto-

antigens Gene/ transgene

Major site of expression

Thymic expression

Thymic functional role

Type of tolerance

Site of assay

Ref.

Insulin Glucagon Somatostatin PP GAD67 Trypsin Amylase Elastase RIP-Tag Elastase-Tag RIP-H-Kd RIP-LMCV IPRBP S-Ag

pancreas pancreas pancreas pancreas pancreas pancreas pancreas pancreas pancreas pancreas pancreas pancreas eye eye

+ + + + ++ + + + + + + +/– +/– +/–

unknown unknown unknown unknown neurotransmitter ? unknown unknown unknown tolerance induction tolerance induction tolerance induction tolerance induction tolerance induction tolerance induction

– – – – – – – – CD4 (CD8+/–) CD4 only ND ND ND ND

– – – – – – – – in vitro/in vivo in vitro/in vivo in vivo in vivo in vivo in vivo

11, 14, 21 11, 14, 21 11, 14, 21, 27 11, 14, 21 11, 21 11, 21 11, 21 11, 21, 45 21 45 13 12 47 47

PP = Pancreatic polypeptide; GAD67 = glutamic acid decarboxylase, 67 kD; RIP = rat insulin promoter; Tag = SV40 T-cell antigen; LCMV = lymphocytic choriomeningitis virus; IPRBP = interphotoreceptor retinoid-binding protein.

ic homeostasis through the regulation of glucose uptake. A striking feature of insulin expression is its almost complete restriction to ß cells of the pancreatic islet in normal adult mammals [6]. During embryonic development, insulin is detected in the fetal yolk sac and transiently in the fetal brain. In the fetal pancreas, cells positive for insulin and other hormones are commonly found in putative endocrine progenitors from the pancreatic ducts; however, from birth, expression of insulin has been detected only in the ß cell [7]. This fact has been of interest to both developmental biologists studying mechanisms of gene restriction and also, more recently, to investigators generating transgenic mouse models. The latter have exploited the specificity of the rat insulin promoter (RIP) to direct the synthesis of transgenes in the ß cell [8]. This approach has been successful in studying the effects of a multitude of protein and peptide hormones and soluble factors. Immunologists studying peripheral tolerance have been particularly interested in this approach. Their studies are concerned with the pathological consequences of expressing a foreign protein in a host cell. Conventional theories of central tolerance suggest that, in this situation, transgene reactive T cells migrating through the thymus would not be deleted and attack the transgene-expressing ß cells [9, 10]. In such a study, Jolicoeur et al. [11] introduced the T-cell superantigen SV40 (Tag) as a transgene into ß cells of a normal

mouse. Surprisingly, they found that T cells were seemingly completely tolerant to this highly reactive antigen. When the investigators checked various organs for spurious expression of the transgene they found low but significant expression in the thymus. Expression of the transgene was not found in any other organ tested, including the spleen and various lymph nodes. Therefore the thymic expression of the transgene appeared to be functionally significant (table 1). This finding was interpreted in the light of two other studies that also identified thymic expression of insulin promoter transgenes and some degree of T-cell nonresponsiveness to the transgene product both in vitro and in vivo [12, 13]. To determine whether endogenous expression of the promoter exists in the thymus, further analysis with RTPCR for preproinsulin mRNA (ppIns) expression as well as several other pancreatic endocrine and exocrine hormones was carried out. This revealed low, but detectable expression for all hormones and enzymes but carboxypeptidase A and amylase [11]. We have confirmed many of these observations. Using RT-PCR, we find that the level of ppIns expression measured in the murine thymus is between four and five orders of magnitude lower than in the pancreas [14]. This suggests that either insulin is expressed at very low levels in cells or a very small number of cells are capable of hormone expression. In contrast to previous reports [11], we did not observe a decrease in

Thymic Expression of the Pancreatic Hormones

Neuroimmunomodulation 1999;6:108–114

109

thymic insulin expression until animals were 20 weeks of age. To investigate further the role of insulin in the thymus, we measured its expression in different subpopulations of cells. In mice 2 weeks of age, macrophages and dendritic cells were separated from epithelial cells and concentrated from thymocytes by density gradient separation [15]. Epithelial cells were prepared from deoxyguanosine-treated neonatal thymic lobes. We found expression of ppIns only in the low buoyant density population of thymic cells enriched for macrophages and dendritic cells, while the high buoyant density fraction and the 195% pure epithelial cell preparation were negative. In addition, we were able to demonstrate the presence of immunoreactive (ir) insulin exclusively in the low buoyant density population. However, the amount of ir insulin detected, although above the level of sensitivity of the assay, was very low. From this cell population, only those selected by FACs sorting for reactivity to the dendritic cell surface marker CD11c (N418) demonstrated expression of ppIns. The expression was observed in 3 ! 104 cells, which represents less than 0.05% of the total thymic population, demonstrating a very significant enrichment of transcript expression, but still below what might be expected if all dendritic cells constitutively synthesized the protein. Dendritic cells are present throughout the body, where they are critical in capturing and displaying antigen in specialized T-cell microenvironments to initiate immune responses [3, 16, 17]. In the thymus, dendritic cells are derived from early T-lymphocyte progenitors, whereas in the periphery they appear to derive from both myeloid and lymphoid lineages [18, 19]. It is noteworthy that hormone expression was not present in any other tissue investigated, whether lymphoid or nonlymphoid. This suggests either that the specific thymic microenvironment modulates insulin expression or that tissue-specific subpopulations of cells are responsible for their expression. Unfortunately, efforts to locate the site of insulin production in situ by immunohistochemical methods on cryostat or paraffin-embedded sections were unsuccessful. Similarly, another study in which a proinsulin antibody was used did not achieve definitive results [20]. However, Hanahan’s group [21], in a study specifically addressing the expression of the insulin-Tag hybrid in the thymus using his transgenic models, found an extremely small number of cells positive for the Tag antigen, estimated at approximately 100 cells per tissue. All positive cells were found in the medulla of the gland, the compartment containing dendritic cells and specialized epithelial cells involved in clonal deletion of T cells [3]. They also sepa-

110

Neuroimmunomodulation 1999;6:108–114

rated cells by density gradient and found enrichment in the low buoyant density population, confirming our result. Although they did not identify the cells directly by sorting, they reported an absence of hormone expression in cell populations after depletion of MHC II-bearing cells. This is again consistent with our observations since, in the thymus, MHC II is primarily expressed by dendritic cells and medullary epithelial cells and to a lesser extent macrophages [22, 23]. Their attempts to colocalize Tagpositive cells in the thymus with cell surface markers were inconclusive, but they were able to detect a few insulinpositive cells by immunohistochemistry. Thus, several lines of evidence now suggest that, in mice, insulin is expressed in antigen-presenting cells and, in theory, is available for expression on the cell surface in the context of MHC II molecules. In humans, two papers have also reported the expression of ppIns in the fetal and neonatal thymus [24, 25]. Levels of expression detected were again very low. Ir insulin levels, measured by sensitive RIAs, were reported to be 104- to 105-fold lower than in the pancreas, consistent with findings in the mouse. Using a specific RIA for the insulin precursor it was found that the level of ir proinsulin was 4 times higher than ir insulin and that ir insulin levels were not distinguishable from control tissues [25]. The investigators concluded that the major immunoreactive product in the thymus was proinsulin and that the ir insulin detected was probably derived from the circulation, suggesting that the specialized posttranslational processing present in the ß cell is not intrinsic to the thymus. No information is available at this time about the cell type or anatomical distribution of insulin expression in the human thymus.

Identification of Glucagon, Somatostatin and Pancreatic Polypeptide in the Thymus

Glucagon, somatostatin and pancreatic polypeptide (PP), like insulin, are synthesized by unique cell types in the islets of Langerhans, the ·, ‰, and PP cells, respectively [5]. They have various roles in metabolic and digestive physiology and are expressed in some tissues other than the pancreas. Glucagon is the product of a large precursor gene that also produces hormones in the gastrointestinal (GI) tract [26]. Somatostatin is a neuroinhibitory peptide expressed in many regions of the brain as well as in the GI tract [5]. PP is a member of a larger neuropeptide family that includes neuropeptide Y.

Throsby/Pleau/Dardenne/Homo-Delarche

In a previous report, Northern blotting of a poly A+ mRNA preparation from the rat thymus using a cRNA probe for somatostatin revealed a signal at the baseline of sensitivity [27]. More recently, somatostatin, along with the other pancreatic endocrine hormones, was measured in the thymus by the more sensitive RT-PCR method (table 1) [11, 14, 21]. Like insulin, there are several orders of magnitude difference between the thymic and pancreatic expression of somatostatin, glucagon and PP. However, these three hormones appear to be expressed at relatively higher levels than insulin in the thymus. After density gradient separation, all hormones showed the same pattern of expression in thymic subpopulations as insulin, but when separated on the basis of specific markers, surprisingly, expression of glucagon, somatostatin and PP were found in the F4/80 sorted cells (macrophages) and not the CD11c (dendritic cell) population. The level of expression relative to the serially diluted age-matched thymus tissue, included as a positive control, was roughly similar for all these hormones, in contrast to observations in the pancreas where there are large differences in their levels of expression. The macrophage belongs to the heterogeneous myeloid cell lineage whose members are present in most tissues of the body and the circulation [28]. The anatomical, morphological and functional characteristics of these cells vary greatly and depend on the environment in which they seed [29, 30]. In the thymus, macrophages have two major functional roles: phagocytosis of apoptotic lymphocytes (immature T cells which are unable to develop further) and to a lesser extent antigen presentation to developing lymphocytes [3, 31–33]. The majority of thymic macrophages are located in the cortex and corticomedullary region of the thymus. However, some are also present in the medulla. Smith et al. [21] were able to detect ir somatostatin, by immunohistochemistry, in the medulla of the thymus at about the same frequency as ir insulin and the Tag protein, but were not able to detect ir glucagon or ir PP. Thus, the immunohistochemical evidence is inconclusive in terms of both the cellular identity and whether the same or different cell types produce all hormones. Although a broad separation between insulin and the other hormones has been made by FACS analysis on the basis of two distinct surface markers, considering the heterogeneity of the myeloid cells in general and the still unresolved nature of the dendritic cell lineage both in the periphery and in the thymus, a more complete study using two or three color stainings is required. In addition, creation of chimeric animals by bone marrow transfer and thymic transplanta-

tion, as outlined by Smith et al. [21], will aid identification of these cells.

Thymic Expression of the Pancreatic Hormones

Neuroimmunomodulation 1999;6:108–114

Functional Significance of Pancreatic Hormone Expression in the Thymus

It has been known for some time that upon antigen activation T cells express the insulin receptor on their surface membrane [34, 35]. Various roles have been ascribed to this expression, including the maintenance of an activated state during submaximal antigen stimulation, modulation of cytotoxic T-lymphocyte response and regulation of metabolic factors within lymphocytes. However, under normal circumstances activated T cells are not present in the thymus and thus insulin receptor expression would be restricted to macrophages and possibly dendritic cells and epithelial cells [3]. What is unclear is how thymic insulin could be physiologically relevant against the normal concentration gradient of circulating insulin. It is possible that insulin might act in an autocrine or paracrine manner, during membrane-bound ligand interaction for instance, in such a way as to greatly increase its concentration at the receptor site. However, the evidence from fetal and neonatal human subjects suggests that the transcribed proinsulin is not further processed in the thymus [24, 25]. The insulin precursor has no known physiological activity and certainly does not react with insulin receptors [36]. Thus, a conventional functional role for the insulin product in the thymus appears unlikely. Somatostatin, which is more widely expressed than the other pancreatic endocrine genes, is not found in the circulation at physiologically relevant levels and acts in a strictly paracrine manner in both the CNS and the digestive system [5]. Receptors for somatostatin have been detected on T cells, B cells and macrophages of the spleen and lymph nodes but their expression has not been described in the thymus [35]. Somatostatin has been demonstrated to have various immunomodulatory roles in the periphery [4, 35]. Thus, it may have some functional role in the thymus. Little can be added for glucagon and PP in terms of their functional significance in the thymus.

Possible Relationship between Thymic Insulin Expression and Type I Diabetes

An unexpected finding in our study was the expression of insulin in a different population of cells than the other pancreatic endocrine hormones. Of particular interest are

111

the functional differences between these two cell types. As described above, the dendritic cell has a far greater potential than the macrophage to present antigen in the thymus for selection purposes. Insulin has been identified as a potential autoantigen in type I diabetes and in insulindependent diabetes mellitus (IDDM) [37]. Autoantibodies to insulin are found in most diabetic patients and relatives and insulin-reactive T-cell clones have been isolated from the spleen and pancreas of the NOD mouse, a spontaneous animal model of IDDM [37–39]. In addition, treatment of NOD mice with native or metabolically inactive insulin analogs protects against IDDM development [40], and transgenic expression of insulin within the thymus has also been reported to protect NOD mice [41]. Therefore, in theory, any alteration in the expression of insulin in the ‘professional’ antigen-presenting dendritic cells could influence the ability of the thymus to select against the generation of autoreactive T cells to insulinderived peptides. Two studies investigating the correlation between restriction length polymorphisms in the insulin promoter region and insulin expression in the fetal and neonatal human thymus lend support to this hypothesis [24, 25]. They focused on the variable number of tandem repeats (VNTR) region that is present 5) of the insulin promoter and inside the IDDM2 susceptibility locus that accounts for roughly 10% of diabetes [42]. Previously, it was demonstrated that low numbers of repeats (class I) predispose to diabetes while larger restriction fragments (class III) had a dominant protective effect [42]. Unexpectedly, it was found that the protective class III alleles were associated with 30% lower levels of insulin expression in the pancreas than class I alleles [43]. This paradox was partially resolved when it was demonstrated that in the thymus the inverse was true; class III alleles were correlated with higher levels of thymic insulin expression than class I alleles [24, 25]. The inference being that lower surface expression of insulin in class I bearing antigen-presenting cells might allow potentially autoreactive T cells to escape from the thymus. In preliminary studies, we compared the expression of insulin in the thymus of NOD mice and normal mouse strains and found consistently lower levels of ppIns expression in NOD mice. Since the levels of expression for the other pancreatic hormones were identical between strains, these results suggest that the result is specific to insulin. Interestingly, others have failed to detect ppIns expression in NOD mice [41]. These data can be interpreted in two ways: either there is a lower level of expression by individual cells or the number/maturation of the

112

Neuroimmunomodulation 1999;6:108–114

dendritic cells is abnormal in the NOD mouse. Evidence for the latter proposal has been described for macrophages and myeloid progenitors [44]. This is currently being investigated in our laboratory. In two different neonatal thymic transfer models, both using Tag under the control of different pancreatic promoters (the exocrine pancreatic gene elastase and RIP), thymic expression of a transgene was demonstrated to induce tolerance in its athymic recipient in contrast to nontransgenic thymuses (table 1) [21, 45]. Interestingly, while the induced tolerance in the CD4+ (Th) compartment was complete, cytotoxic T-cell responses (CTL) were only partially inhibited in both models. This is in contrast to transgenic animals that express Tag in both the thymus and ß cells and which display full tolerance, suggesting that both central and peripheral mechanisms operate in inducing cytotoxic T-cell tolerance. Of interest on this point is the study of Brocker et al. [46] who directed expression of the MHC II molecule I-E on an I-E-deficient background using the dendritic cell-specific CD11c promoter, the surface marker we used to select thymic dendritic cells. When assayed for tolerance to the I-E product, it was found that CD4+ cells were completely tolerized but the CTL was only partly tolerized, the same situation observed in the transfer experiments. Smith et al. [21] also described a transgenic line, RIP-Tag5, identical to other transgenic lines except for the absence of thymic expression of the transgene. These animals developed a ß-cell-specific autoimmune disease directed against the Tag antigen, presumably as a direct result of the lack of thymic Tag expression and by extension clonal deletion of Tag-reactive T cells.

Conclusion

Strong if indirect evidence now links the expression of insulin in the thymus with clonal deletion of insulin-reactive T cells. In the absence of any data to the contrary, it appears to be the sole purpose of insulin expression in the thymus. Insulin is of particular interest because of its restricted expression; however, its presence and that of the other pancreatic hormones in the thymus raises the question of how many peripheral proteins might be expressed there. We and others have found a wide variety of autoantigens involved in IDDM and other autoimmune diseases expressed in the thymus [11, 21, 45, 47, 48], many of which probably have functional roles. However, some like insulin may only be present to induce immune tolerance. If so, what are the factors controlling their

Throsby/Pleau/Dardenne/Homo-Delarche

expression? The lack of thymic and presence of pancreatic RIP-Tag expression in certain transgenic lines suggests that insulin expression is not a random event and may point to a cryptic promoter sequence in the gene. A more complete understanding of the cell type(s) producing these hormones and functional testing of their products will help to answer these questions.

Acknowledgments The authors would like to thank Anne Esling, Carmi Pelegri, Veronique Alves, Corrine Garcia and Isabelle Cisse for assistance with work described in the text and Doreen Bromer for reviewing the manuscript. M.T. was supported by the Fondation pour La Recherche Médicale and L’Association Claude-Bernard.

References 1 Mondino A, Khoruts A, Jenkins M: The anatomy of T-cell activation and tolerance. Proc Natl Acad Sci USA 1996;93:2245–2252. 2 Ernst B, Surh CD, Sprent J: Thymic selection and cell division. J Exp Med 1995;182:961– 971. 3 Anderson G, Moore NC, Owen JJ, Jenkinson EJ: Cellular interactions in thymocyte development. Annu Rev Immunol 1996;14:73–99. 4 Savino W, Dardenne M: Immune-neuroendocrine interactions. Immunol Today 1995;16: 318–322. 5 Slack JMW: Developmental biology of the pancreas. Development 1995;121:1569–1580. 6 Dumonteil E, Philippe J: Insulin gene: Organisation, expression and regulation. Diabetes Metab 1996;22:164–173. 7 Alpert S, Hanahan D, Teitelman G: Hybrid insulin genes reveal a developmental lineage for pancreatic endocrine cells and imply a relationship with neurons. Cell 1988;53:295–308. 8 Hanahan D: Transgenic mice as probes into complex systems. Science 1989;246:1265– 1275. 9 Kappler JW, Roehm M, Marrack P: T cell tolerance by clonal elimination in the thymus. Cell 1987;49:273. 10 Ashton-Rickardt PG, Bandeira A, Delaney JR, Van Kaer L, Pircher HP, Zinkernagel RM, Tonegawa S: Evidence for a differential avidity model of T cell selection in the thymus. Cell 1994;76:651–663. 11 Jolicoeur C, Hanahan D, Smith KM: T-cell tolerance toward a transgenic beta-cell antigen and transcription of endogenous pancreatic genes in thymus. Proc Natl Acad Sci USA 1994;91:6707–6711. 12 Vonherrath MG, Guerder S, Lewicki H, Flavell RA, Oldstone MBA: Coexpression of B7-1 and viral (‘self’) transgenes in pancreatic beta cells can break peripheral ignorance and lead to spontaneous autoimmune diabetes. Immunity 1995;3:727–738. 13 Heath W, Allison J, Hoffmann M, Schonrich G, Hammerling G, Arnold B, Miller J: Autoimmune diabetes as a consequence of locally produced interleukin-2. Nature 1992;359:547– 549. 14 Throsby M, Homo-Delarche F, Chevenne D, Goya R, Dardenne M, Pleau J-M: Pancreatic hormone expression in the murine thymus: Localization in dendritic cells and macrophages. Endocrinology 1998;139:2399–2406.

Thymic Expression of the Pancreatic Hormones

15 Vremec D, Zorbas M, Scollay R, Saunders D, Ardavin C, Wu L, Shortman K: The surface phenotype of dendritic cells purified from mouse thymus and spleen: Investigation of the CD8 expression by a subpopulation of dendritic cells. J Exp Med 1992;176:47–58. 16 Lanzavecchia A: Mechanisms of antigen uptake. Curr Opin Immunol 1996;8:348–354. 17 Steinman RM, Pack M, Inaba K: Dendritic cells in the T-cell areas of lymphoid organs. Immunol Rev 1997;156:25–37. 18 Ardavin C, Wu L, Li C, Shortman K: Thymic dendritic cells and T cells develop simultaneously in the thymus from a common precursor population. Nature 1993;362:761–763. 19 Kronin V, Vremec D, Winkel K, Classon B, Miller R, Mak T, Shortman K, Suss G: Are CD8+ dendritic cells (DC) veto cells? The role of CD8 on DC in DC development and in the regulation of CD4 and CD8 T cell responses. Int Immunol 1997;9:1061–1064. 20 Geenen V, Achour I, Robert F, Vandersmissen E, Sodoyez JC, Defresne MP, Boniver J, Lefebvre PJ, Franchimont P: Evidence that insulin-like growth factor 2 (IGF2) is the dominant thymic peptide of the insulin superfamily. Thymus 1993;21:115–127. 21 Smith KM, Olson DC, Hirose R, Hanahan D: Pancreatic gene expression in rare cells of thymic medulla: Evidence for functional contribution to T cell tolerance. Int Immunol 1997;9:1355–1365. 22 Ewijk Wv, Ron Y, Monaco J, Kappler J, Marrack P, Meur ML, Gerlinger P, Durand B, Benoist C, Mathis D: Compartmentalization of MHC class II gene expression in transgenic mice. Cell 1988;53:357–370. 23 Surh CD, Sprent J: T-cell apoptosis detected in situ during positive and negative selection in the thymus (see comments). Nature 1994;372: 100–103. 24 Vafiadis P, Bennett ST, Todd JA, Nadeau J, Grabs R, Goodyer CG, Wickramasinghe S, Colle E, Polychronakos C: Insulin expression in human thymus is modulated by INS VNTR alleles at the IDDM2 locus. Nat Genet 1997; 15:289–292.

25 Pugliese A, Zeller M, Fernandez A, Zalcberg LJ, Bartlett RJ, Ricordi C, Pietropaolo M, Eisenbarth GS, Bennett ST, Patel DD: The insulin gene is transcribed in the human thymus and transcription levels correlate with allelic variation at the INS VNTR-IDDM2 susceptibility locus for type 1 diabetes. Nat Genet 1997;15:293–297. 26 Philippe J: Structure and pancreatic expression of the insulin and glucagon genes. Endocr Rev 1991;12:1–20. 27 Fuller P, Verity K: Somatostatin gene expression in the thymus gland. J Immunol 1989;143: 1015–1017. 28 Hume D, Halpin D, Charlton H, Gordon S: The mononuclear phagocyte system of the mouse defined by immunohistochemical localization of antigen F4/80: Macrophages of endocrine organs. Proc Natl Acad Sci USA 1984;81: 4174–4177. 29 Vicente A, Varas A, Moreno J, Sacedon R, Jimenez E, Zapata A: Ontogeny of rat thymic macrophages. Phenotypic characterization and possible relationships between different cell subsets. Immunology 1995;85:99–105. 30 Castro A, Bono MR, Simon V, Vargas L, Rosemblatt M: Spleen-derived stromal cells. Adhesion molecules expression and lymphocyte adhesion to reticular cells. Eur J Cell Biol 1997; 74:321–328. 31 Robinson J: The ontogeny of thymic macrophages: Thymic macrophages express Ia from 15 days’ gestation onwards in the mouse. Cell Immunol 1984;84:422–426. 32 Soga H, Nakamura M, Yagi H, Kayaba S, Ishii T, Gotoh T, Itoh T: Heterogeneity of mouse thymic macrophages. 1. Immunohistochemical analysis. Arch Histol Cytol 1997;60:53–63. 33 Banuls MP, Alvarez A, Ferrero I, Zapata A, Ardavin C: Cell-surface marker analysis of rat thymic dendritic cells. Immunology 1993;79: 298–304. 34 Helderman JH, Ayuso R, Rosenstock J, Raskin P: Monocyte-T lymphocyte interactions for regulation of insulin receptors on the activated T lymphocyte. J Clin Invest 1987;79:566–571. 35 Homo-Delarche F, Durant S: Hormones, neurotransmitters and neuropeptides as modulators of lymphocyte functions; in Rola-Pleszczynski M (ed): Handbook of Immunopharmacology. London, Academic Press, 1994, pp 171–239.

Neuroimmunomodulation 1999;6:108–114

113

36 Halban PA: Structural domains and molecular lifestyles of insulin and its precursors in the pancreatic beta cell. Diabetologia 1991;34: 767–778. 37 Gottlieb PA, Eisenbarth GS: Mouse and man: Multiple genes and multiple autoantigens in the aetiology of type I DM and related autoimmune disorders. J Autoimmun 1996;9:277– 281. 38 Wegmann DR: The immune response to islets in experimental diabetes and insulin-dependent diabetes mellitus. Curr Opin Immunol 1996;8:860–864. 39 Wegmann DR, Norbury-Glaser M, Daniel D: Insulin-specific T cells are a predominant component of islet infiltrates in pre-diabetic NOD mice. Eur J Immunol 1994;24:1853–1857. 40 Karounos DG, Bryson JS, Cohen DA: Metabolically inactive insulin analog prevents type I diabetes in prediabetic NOD mice. J Clin Invest 1997;100:1344–1348.

114

41 French MB, Allison J, Cram DS, Thomas HE, Dempsey Collier M, Silva A, Georgiou HM, Kay TW, Harrison LC, Lew AM: Transgenic expression of mouse proinsulin II prevents diabetes in nonobese diabetic mice. Diabetes 1997;46:34–39. 42 Bennett ST, Todd JA: Human type 1 diabetes and the insulin gene: Principles of mapping polygenes. Annu Rev Genet 1996;30:343–370. 43 Bennett S, Wilson A, Cucca F, Nerup J, Pociot F, McKinney P, Barnett A, Bain S, Todd J: IDDM2-VNTR-encoded susceptibility to type 1 diabetes: Dominant protection and parental transmission of alleles of the insulin genelinked minisatellite locus. J Autoimmun 1996; 9:415–421. 44 Serreze D, Gaedeke J, Leiter E: Hematopoietic stem-cell defects underlying abnormal macrophage development and maturation in NOD/ Lt mice: Defective regulation of cytokine receptors and protein kinase C. Proc Natl Acad Sci USA 1993;90:9625–9629.

Neuroimmunomodulation 1999;6:108–114

45 Antonia SJ, Geiger T, Miller J, Flavell RA: Mechanisms of immune tolerance induction through the thymic expression of a peripheral tissue-specific protein. Int Immunol 1995;7: 715–725. 46 Brocker T, Riedinger M, Karjalainen K: Targeted expression of major histocompatibility complex class II molecules demonstrates that dendritic cells can induce negative but not positive selection of thymocytes in vivo. J Exp Med 1997;185:541–550. 47 Egwuagu CE, Charukamnoetkanok P, Gery I: Thymic expression of autoantigens correlates with resistance to autoimmune disease. J Immunol 1997;159:3109–3112. 48 Geenen V, Goxe B, Martens H, Vandersmissen E, Vanneste Y, Achour I, Kecha O, Lefebvre P: Cryptocrine signaling in the thymus network and T cell education to neuroendocrine selfantigens. J Mol Med 1995;73:449–455.

Throsby/Pleau/Dardenne/Homo-Delarche

Neuroimmunomodulation 1999;6:115–125

The Thymic Repertoire of Neuroendocrine-Related Self Antigens: Biological Role in T-Cell Selection and Pharmacological Implications Vincent Geenen Ouafae Kecha Fabienne Brilot Chantal Charlet-Renard Henri Martens Institute of Pathology CHU-B23, Laboratory of Radioimmunology and Neuroendocrine-Immunology, University of Liège, Belgium

Key Words Thymus W Neuropeptides W Self tolerance W Autoimmunity

Abstract Thymic epithelium, including nurse cells (TEC/TNC), as well as other thymic stromal cells (macrophages and dentritic cells), express a repertoire of polypeptide belonging to various neuroendocrine protein families (such as the neurophypophysial, tachykinin, neurotensin and insulin families). A hierarchy of dominance exists in the organization of the thymic repertoire of neuroendocrine precursors. Oxytocin (OT) is more expressed in the TEC/ TNC than vasopressin (VP); insulin-like growth factor 2 (IGF-2) thymic expression predominates over IGF-1, and much more over (pro)insulin. Thus, OT was proposed to be the self antigen of the neurohypophysial family, and IGF-2 the self antigen precursor of the insulin family. The dual role of the thymus in T-cell life and death is recapitulated at the level of the thymic neuroendocrine protein repertoire. Indeed, thymic polypeptides behave as accessory signals involved in T-cell development and positive selection according to the cryptocrine model of signaling. Moreover, thymic neuroendocrine polypeptides are the source of self antigens presented by thymic MHC molecules to developing pre-T cells. This presentation

ABC

© 1999 S. Karger AG, Basel 1021–7401/99/0062–0115$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: http://BioMedNet.com/karger

might induce the negative selection of T cells bearing a randomly rearranged antigen receptor (TCR) oriented against neuroendocrine families. Using an animal model of autoimmune type 1 diabetes (BB rat), we have shown a defect in intrathymic expression of the self antigen of the insulin family (IGF-2) and in IGF-2-mediated T-cell education to recognize and tolerate the insulin family. Altogether these studies have enlighted the crucial role played by the thymus in the induction of the central self tolerance of neuroendocrine families. The tolerogenic properties of thymic self peptides could be used in a novel type of vaccination for the prevention of autoimmune diseases.

Introduction

For a long time the thymus was considered to be an intrinsic component of the endocrine system though the endocrine model of cell-to-cell signalling had not been fully validated for this organ. With the discovery of its primary role in T-lymphopoiesis [1], the endocrine role of the thymus progressively vanished. In the past 15 years, the question of a neuroendocrine component in thymic physiology was reinvestigated. From a number of studies,

Vincent Geenen, MD, PhD Institute of Pathology CHU-B23 University of Liège-Sart Tilman B–4000 Liège 1-Sart Tilman (Belgium) Tel. +32 43 66 25 50, Fax +32 43 66 29 77, E-Mail [email protected]

it now appears that the thymus represents a crucial site for a cross-talk between the neuroendocrine and immune systems, in particular during fetal development. Thymic epithelial and stromal cells express a repertoire of neuroendocrine-related genes/protein precursors. It was shown that thymic polypeptides may serve as signals interacting with their cognate neuroendocrine receptors on developing pre-T lymphocytes. This cryptocrine form of cell-tocell signaling could play a role in T-cell development and maturation. On the other side, there is ample evidence that thymic neuroendocrine-related polypeptides also behave as a source of self antigens which are presented to pre-T cells and are thought to induce the negative selection of T cells bearing a randomly rearranged antigen receptor (TCR) oriented against endogenous neuroendocrine families (self-reactive T cells). The objective of this review is to expose most of the scientific arguments which support the important role of the thymus in the education of T lymphocytes to recognize and tolerate neuroendocrine functions.

The Establishment of Immunological Self Tolerance

The process of self tolerance induction involves a multilayered organization in which various tolerizing mechanisms are interconnected in series, from the early steps in immune cell ontogeny to an advanced stage in life [2]. With regard to the T-lymphocyte system, the primary tolerizing steps occur within the thymus, the primary lymphoid organ responsible for T-cell differentiation. To complete their differentiative program, immature T cells receive a series of signals from the thymic cellular microenvironment. Such activatory signals may be emitted by thymic stromal cells (like hormones or cytokines) or may result from direct interactions between cell adhesion molecules expressed on pre-T cells (thymocytes) and thymic stromal cells [3, 4]. Along their differentiation, immature T cells randomly rearranged the genes coding for the segments of their TCR. A lot of these random TCR combinations are oriented against self antigens which are expressed in the thymic microenvironment, then presented by proteins encoded in the major histocompatibility complex (MHC). The interaction of self-reactive T-cell clones with their cognate self antigens is thought to lead to their negative selection either by programmed cell death (apoptosis), or by their developmental arrest. This process of thymic clonal deletion was demonstrated with the use of MMTV (mouse mammary tumor virus)-encoded su-

116

Neuroimmunomodulation 1999;6:115–125

perantigens [5], and with transgenic mice expressing a TCR specific for the male antigen H-Y [6]. Since the thymus does not express all the components of the self structure, this organ does not delete all potential autoreactive T cells. Consequently, the existence of other tolerizing mechanisms (such as T-cell anergy) at the periphery was postulated, and they were effectively shown to intervene in the process of immunological self tolerance. Nevertheless, thymic clonal deletion of self-reactive T cells is by far the most important mechanism involved in self education of the immune system [7].

Thymic T-Cell Life and Death

Self peptides are not only involved in the induction of central T-cell self tolerance but they also intervene in the process of T-cell maturation or positive selection [8]. Thus, the thymus is the site for an important paradox of contemporary cell biology: How can T lymphocytes be both positively and negatively selected in the thymic microenvironment [9]? A first explanation proposed that thymic epithelial cells (TEC) were responsible for T-cellpositive selection, whereas other thymic bone-marrowderived stromal cells (macrophages and dendritic/interdigitating (IDC) cells) induced deletion of self-reactive Tcell clones [10]. However, this hypothesis is no longer supported by recent experiments which have established that TEC are able to delete self-reactive T cells (see after). The ‘avidity/affinity hypothesis’ has been proposed as another explanation of the thymic paradox [11, 12]. This hypothesis is based on experiments showing that transgenic TCRs (specific for defined antigens) do not mature in organ cultures of fetal thymuses from MHC class I-defective animals. However, they do so if peptides related to the cognate antigen of TCR are added in the cultures. So T lymphocytes are positively selected if their TCR is barely engaged with self peptide presented by MHC ligands, and they are deleted if their TCR is strongly engaged. However, since the usual affinity of a TCR for its cognate antigen is already rather low (10 –8 M at the maximum), one may question the importance of the biological effects mediated by a lower affinity. If the experiments mentioned above have convincingly shown that T-cell positive selection is peptide-specific and depends on ligand concentration, one may also question the nature and the amount of peptide/MHC combinations that contribute in vivo to positive selection of a particular TCR in a normal thymus. Because of their high polymorphism, thymic molecules derived from MHC cannot establish the dis-

Geenen/Kecha/Brilot/Charlet-Renard/ Martens

crimination between self and nonself antigens. Given the hazardous nature of the recombination of TCR gene segments, the precise identity of thymic peptides supporting T-cell positive and negative selection has become an important current question [13].

Thymic Neurohypophysial-Related Peptides

this hybrid protein, whereas neurophysin could bind OT for presentation to pre-T cells. In relationship with our studies, other authors have shown the translocation of a neurophysin-like material in the cell membranes of cancer cells. They also provided several arguments supporting the behavior of neurohypophysial-related peptides as candidate tumoral antigens [30, 31]. Thus, both in the hypothalamo-neurohypophysial axis and in the thymus, the neurophysin part of the OT precursor fulfills the same function: binding of the active nonapeptide OT and transport to the external limit of neurons or TEC/TNC. The tyrosine residue in position 2 of OT and VP plays an important role in their binding to neurophysin [32]. Interestingly, the residue tyrosine in the same position was found to play a crucial role in the binding of antigens to some MHC class I alleles for their presentation [33]. As another argument for an antigenic role of thymic OT, we have shown that the immunological recognition of OT by specific mAbs at the outer surface of human TEC plasma membrane induced a marked secretion of the cytokines interleukin(IL)-6 and leukemia inhibitory factor (LIF) in the supernatant of TEC cultures [34]. Given the nature of the epitopes recognized by anti-OT mAbs, we were able to conclude that thymic OT is fully processed at the level of the TEC plasma membrane. The absence of biological effects following the treatment of TEC cultures with anti-VP mAbs supports our hypothesis that thymic OT behaves as the self antigen of the neurohypophysial hormone family.

At the beginning of this century, Ott and Scott [14] described the galactogogue activity of thymic extracts after injection into the goat. At that time, oxytocin (OT) had not been identified as the primary mediator of the galactokinesis and the oxytocic activity of thymic extracts was not further characterized. TEC and thymic nurse cells (TNC) from different species synthesize polypeptide precursors of the neurohypophysial family, with a dominance of the OT lineage [15–21]. The oxytocic action of human fetal thymic extracts was also described [22]. TNC in the subcapsular and outer cortex of thymic lobules constitute an intimate neuroendocrine-immune microenvironment since their epithelial component (but not the TNC-engulfed pre-T cells) produces neurohypophysialrelated peptides and expresses the phenotype of neuroendocrine cell types [23]. Transcripts of proOT and proVP were detected both in human [17] and murine [24] thymic extracts, and the intrathymic expression of neurohypophysial genes during ontogeny is under current investigation. However, the synthesis of OT in TEC/TNC is not coupled with the secretion of the nonapeptide or its neurophysin in the supernatant of human TEC/TNC primary cultures. As shown in the murine thymus, immunoreactive (ir)-OT is not located in secretory granules but is diffuse in the cytosol, in vesicles of the endoplasmic reticulum, and associated with keratin filaments [25]. Interestingly, similar ultrastructural features were also reported for OT and VP expressed by murine spleen eosinophillike cells [26]. As discussed above, the thymic function is closely associated with the presentation of the self molecular structure to developing T cells. This action was long thought to be mediated by thymic macrophages and IDC only, but there is now large evidence that TEC/TNC are actively involved in the induction of central self tolerance [97, 98]. Using the appropriate methodology, we provided evidence that the presentation of thymic OT implicates a hybrid 55-kD protein probably bearing a neurophysin (10 kD) and a MHC class I heavy chain domain (45 kD) [29]. Following this putative explanation, the MHC class I domain would be implicated in membrane targeting of

The model of cell-to-cell cryptocrine signaling has been proposed by Funder [35] to characterize the direct membrane-to-membrane exchange of chemical information between large epithelial nursing cells and immature elements which migrate and differentiate at their contact. Besides its role as a self antigen, there is also evidence that OT mediates a cryptocrine-type signaling between TEC/ TNC and pre-T cells. Neurohypophysial peptide receptors have been detected in the rat thymus and on rat thymocytes [36, 37], on a murine pre-T cell line (RL12-NP) [38] and on murine cytotoxic T cells [38, 39]. Estrogens were shown to increase the affinity of OT receptors in the rat thymus [40]. Interestingly, the expression of the rat V1b (or V3) receptor was recently identified in tissues outside the anterior pituitary, including the thymus [41]. Already in 1969, the mitogenic effect of neurohypophysial peptides on rat thymocytes was described [42], where-

Thymic Neuroendocrine Self Peptides

Neuroimmunomodulation 1999;6:115–125

Intrathymic Cryptocrine Signaling

117

as OT was reported to stimulate glucose oxidation by rat thymocytes [43]. On the basis of antagonist effects, murine pre-T cells express a V1 (or V3) subtype of neurohypophysial receptor, while mature cytotoxic T cells harbor receptors of the OT type; this observation suggests that the neurohypophysial reception system expressed by T lymphocytes could ‘mature’ in parallel with their stage of differentiation. In both of T-cell types, neurohypophysial peptide receptors transduce OT and VP via the phosphoinositide pathway and neurohypophysial-related signals increase the incorporation of tritiated thymidine by freshly isolated murine thymocytes suggesting a mitogenic effet [38]. Western blots of RL12-NP-extracted proteins with anti-phosphotyrosine revealed a number of proteins the phosphorylation of which was stimulated either by OT or VP. Two of these proteins were precipitated with anti-focal adhesion (FAK) mAb 2A7 and were identified one as p125FAK and the other as a co-precipitating 130-kD protein (probably p130Cas). Another protein phosphorylated by OT in RL12-NP cells was identified as paxillin, a 68-kD protein located at focal adhesion sites and associated with p125FAK. Interestingly, OT was more potent than VP in inducing p125FAK phosphorylation and this OT effect was inhibited by a V1 receptor antagonist, confirming that immature T cells bear a V1-type neurohypophysial receptor [44]. Stimulation of focal adhesions could play an important role in promoting T-cell interactions with the thymic cellular microenvironment which are fundamental for the T-cell differentiation programme. There is thus large experimental evidence that thymic OT mediates a functional cryptocrine signaling that could serve as an accessory pathway in the positive selection of T cells. The existence of a functional signaling between thymic OT and neurohypophysial receptors expressed by immature and cytotoxic T cells raises the possibility of a pharmacological modulation of T-cell activity by OT receptor antagonists. Using the methodology of whole blood cell cultures, OT hexapeptide antagonists (developed by Merck Sharp & Dohme Research Laboratories) were shown to inhibit the production of IL-1ß and IL-6 elicited by human T-cell activation with anti-CD3 mAb [45]. Specific antagonists of OT receptors expressed by immune cells could offer a therapeutic benefit in circumstances during which an enhancement of the immune reactivity and a relapse of autoimmune diseases are observed (such as during the postpartum or during lactation).

118

Neuroimmunomodulation 1999;6:115–125

Application to Other Neuroendocrine Polypeptide Families

A number of neuroendocrine-related polypeptides have been detected and characterized in TEC and thymic stromal cells from different species (table 1). The wellcharacterized neuropeptides neurotensin (NT) and somatostatin have been extracted from the chicken thymus, especially after hatching, and were characterized both immunochemically and chromatographically [46]. We have shown the expression of ir-NT at the cell surface of human TEC. Cultured human TEC contain F5 ng ir-NT/ 106 cells, of which 5% is associated with plasma cell membranes. HPLC analysis of ir-NT present in human TEC revealed a major peak of ir-NT corresponding to intact NT1–13. Ir-NT was not detected in the supernatant of human TEC primary cultures. Using an affinity column prepared with an anti-MHC class I Ab, NT-related peptides were retained on the column and were eluted together with MHC class I proteins [47]. Neurokinin A (NKA) is the peptide of the tachykinin family encoded in human and rat TEC by the preprotachykinin A (PPT-A) gene [48]. Thymic PPT-A expression was shown to be glucocorticoid-dependent since adrenalectomy of Sprague-Dawley rats markedly enhanced the levels of thymic PPT-A (and NPY) mRNAs [Ericsson and Geenen, unpubl. data]. Interestingly, NKA exerts IL-1like mitogenic effects on murine thymocytes [49], and this effect suggests the expression of specific tachykinin receptors by immature T cells which could be implicated in another accessory pathway for T-cell maturation and positive selection. The amino-acid sequence of NKA shares the same C-terminal epitope with other members of the tachykinin family, and the leucine residue in position 9 could be used in the binding to some MHC class I alleles, thus making NKA the self antigen of the tachykinin family. The other tachykinin encoded by PPT-A, substance P (SP), is not detected in TEC but within sensory nerve fibers of the thymus [50]. Thymic-specific receptors for SP are associated with the vasculature in the medulla, where they could control local blood flow and vascular permeability [51]. The expression within the rat thymus of natriuretic peptides (ANP, BNP and CNP) has been well documented. ANP seems to be the dominant thymic peptide and is expressed by thymic macrophages, while ir-CNP has been detected in thymocytes. The different types of natriuretic peptide receptors were also detected by RTPCR [52, 53]. The treatment with ANP of murine fetal thymic organ cultures (FTOC) was shown to decrease the

Geenen/Kecha/Brilot/Charlet-Renard/ Martens

Table 1. The thymic repertoire of neuroendocrine self peptides

Neuroendocrine families

Physiological aspects

Thymic repertoire

Water metabolism Vasocontriction Reproduction (CNS and periphery)

OT 11 VP

Glucose metabolism Growth control Fetal development

IGF-2 1 IGF-1 11 Insulin

Parathormones Parathormone (PTH) Parathormone-related peptide (PTH-rP)

Calcium metabolism

PTH-rP 11 PTH [106]

Calcitonins Calcitonin (CT) Calcitonin gene-related peptide (CGRP)

Calcium metabolism

CGRP 11 CT [107]

Pain – inflammation Mastocyte degranulation

NKA 11 SP (NKB?)

Sodium metabolism

ANP?

Hypothermia Analgesia Pancreatic exosecretion Vasodilatation

NT

Neurohypophysial peptides Vasopressin (VP or ADH) Oxytocin (OT) Insulin family Insulin Insulin-like growth factor 1 (IGF-1) Insulin-like growth factor 2 (IGF-2)

Tachykinins Substance P (SP) Neurokinin A (NKA) Neurokinin B (NKB) Natriuretic peptides ANP BNP CNP Neuromedins Neurotensin (NT)

Neuromedin N

total thymocyte yield in FTOC, to increase the CD4–8– and to decrease the CD4+8+ thymocyte subpopulations [54]. It has to be mentioned that a series of anterior pituitary hormone immunoreactivities have been detected in different TEC subpopulations of the human thymus. These TEC populations were different from OT/VP/neurophysin-containing epithelial cells. However, it is not yet clear whether these hormones are locally synthesized or stored in TEC from peripheral blood [55], though they were identified in cultured rat thymic fragments [56]. Importantly, the expression of thymic hormones has been shown to be under the control of the neuroendocrine and steroid microenvironment [57], and this will be discussed in detail in another chapter.

Thymic Neuroendocrine Self Peptides

Thymic Expression of Insulin-Related Genes

In the line of our working model that central T-cell tolerance of neuroendocrine functions is induced by the thymic repertoire of neuroendocrine self antigens, a series of investigations were undertaken to identify the dominant member of the insulin family expressed in the thymic microenvironment. By immunocytochemistry with a panel of specific Abs directed against distinct epitopes of the insulin family, ir-IGF-2 was clearly identified as the dominant member of the insulin family expressed by TEC/TNC [58]. A mAb against proinsulin [59] revealed a slight labeling but outside thymic lobules, in the interstitial tissue of the thymic capsule and in interlobular septae. Thymic labeling was also negative with mAbs against the C-terminal part of the insulin B chain. A few IGF-1-posi-

Neuroimmunomodulation 1999;6:115–125

119

tive cells were also stained in thymic lobules but they were not epithelial; their distribution and morphology were similar to those of macrophages. Of interest, murine macrophages were shown to express IGF-1 Ea and Eb transcripts [60]. Interferon-Á inhibits macropage IGF-1 at the transcriptional level [61], whereas colony-stimulating factors induce IGF-1 mRNA [62]. Ir-IGF-2 was not detected in the supernatant of human TEC primary cultures and, with the use of confocal microscopy, a large part of irIGF-2 was found to be associated with the outer surface of TEC plasma membranes. This was not the case for either IGF-1 or insulin. In the human thymus, IGF-2, IGF-1 and (pro)insulin concentrations were respectively 96.7 B 10.6 ng/g, 42.9 B 5.0 ng/g, and !0.1 ng/g wet weight. IGF2 transcripts have been isolated from whole human thymic extracts, as well as from primary cultures of human TEC. With RT-PCR and different specific primers, the expression of IGF2 in the human thymus was found to be controlled by the same promoters as in other fetal and adult extrahepatic tissues [63, 64, and Kecha et al., submitted]. The effects of Igf2 overexpression under the control of the MHC H-2Kb promoter have been investigated by the generation of transgenic mice. The highest levels of transgene expression were found in thymus and spleen. Only the thymus showed a significant increase of weight in these transgenic mice, in agreement with the high mRNA expression within this organ [65]. Going back to the initial observations made by Pansky et al. [66], there is evidence that the thymic insulin-like reticular factor isolated by them corresponds in fact to IGF-2. The IGF-2 structure closely related to (pro)insulin explains a cross-reactivity with the polyclonal Abs directed against insulin that were used in 1965. The hypoglycemic properties of IGF-2 have been well described [67] and might explain the biological activity of thymic extracts on glucose metabolism. Moreover, the syndrome of hypoglycemia and lymphoid leukemia associated with thymic hyperplasia of some AKR female mice could in fact result from the overexpression of Igf2 in hyperplastic thymic epithelium, with a subsequent secretion of IGF-2 in the bloodstream, and a profoundly disturbed thymic T-cell lymphopoiesis. The hypothesis of a central T-cell tolerance of the insulin family and, secondarily, of the peripheral insulinsecreting pancreatic islet ß cells was further supported by the observation that transcripts of proinsulin and of 67kD isoform of glutamic acid decarboxylase (GAD) genes can be detected in the murine thymus with 30 cycles of RT-PCR [68]. Thymic insulin gene (INS) expression was highest in perinatal mice and persisted until 12 weeks of

120

Neuroimmunomodulation 1999;6:115–125

age. Two recent papers confirmed these findings and reported that INS transcripts, as well as (pro)insulin protein can be detected at very low levels (100–1,000 fmol/g wet weight) in the human fetal thymus [69, 70]. The question of an illegitimate INS trancription was ruled out by the RIA detection of ir-(pro)insulin within thymic tissues. Preliminary in situ hybridization studies have shown the presence of INS transcripts within murine thymic IDC, but the protein could not be evidenced by immunocytochemistry [Homo-Delarche, pers. commun.]. A functional signaling mediated in the thymus by IGFs and IGF receptors is highly plausible because type 1 and type 2 IGF receptors have been detected on rat thymocytes and murine thymoma cells [71], on human phytohemagglutinin A (PHA)-activated T cells and on anti-CD3activated human T lymphocytes [72, 73]. Kooijman et al. [74] have described a differential expression of type 1 IGF receptors in relation to the stage of activation and differentiation of human T lymphocytes. Interestingly, in Igf2 transgenic mice, the increased thymic cellularity is associated with a stimulated generation of phenotypically normal T cells, in particular CD4 T cells [75]. In our hands, specific type 2 IGF receptors were detected on a murine immature T-cell line (RL12-NP), as well as on Jurkat T cells [76]. By affinity cross-linking, the type 2 IGF receptor expressed by lymphocytes was found to have a molecular weight (B260 kD) similar to that found on other cells. We currently investigate the biological action of thymic IGFs on T-cell differentiation through the use of murine FTOC and with specific Abs against IGF receptors [Kecha et al., in preparation].

Some Principles and Advantages of Thymic T-Cell Education to Neuroendocrine Self Antigens

A model has been proposed according to which neuroendocrine-related thymic polypeptides engage two distinct types of interactions with pre-T cells depending on their involvement as self antigens of their family or as cryptocrine signals (table 2) [77]. The interaction of neuroendocrine self antigens with their corresponding TCR implies a binding of moderate affinity (10–6 to 10 –8 M), but with a high power of discrimination. Neuroendocrine self antigens usually correspond to peptide sequences of neuroendocrine precursors which have been highly conserved during the evolution of their related family. On the other hand, cryptocrine signaling between thymic neuroendocrine-related peptides and their cognate receptors

Geenen/Kecha/Brilot/Charlet-Renard/ Martens

expressed by pre-T cells implies a high-affinity binding (10 –10 to 10 –11 M ), with a low discrimination. Moreover, a hierarchy of dominance appears in the organization of the polypeptide repertoire expressed in the thymus (table 1). This is very significant since self tolerance primarily concerns self determinants that are dominant on self molecules [78–80]. Some selective advantages appear from this model of thymic neuroendocrine-related precursors of cryptocrine signals and self antigens in T-cell positive and negative selection, respectively. A first advantage is the absence of a tight allelic restriction in thymic T-cell education to neuroendocrine families. Such an allelic restriction of central T-cell tolerance of neuroendocrine families was hardly conceivable and our data seem to indicate that it is not the case in reality. Concerning the presentation of thymic OT for example, our data suggest that, though MHC class I molecules are of course involved in the process, it is the invariant neurophysin domain of the hybrid membrane 55-kD protein that binds OT for presentation to pre-T cells. Another selective advantage resides in the potential presentation to pre-T cells of the structure characteristic of the neurohypophysial family. With regard to the thymic presentation of NT, there is no physical constraint for a non covalent binding to MHC since this neuropeptide is a linear peptide (in contrast to cyclic OT and IGF-2). In addition, the C-terminal sequence of NT includes tyrosine, isoleucine and leucine residues which can all be used in the anchorage to most of the MHC class I alleles. Given these characteristics, it is logical to postulate that NT and NT-derived C-terminal fragments could behave as natural ligands for a majority (if not all) of MHC class I alleles. This hypothesis is also in agrement with the high degree of conservation of NT-related C-terminal region throughout evolution [81]. For IGFs, the role of binding and transport proteins is ensured by IGF-binding proteins (IGFBPs). IGFBPs have co-evolved with IGFs but they are not part of IGF precursors and are encoded by distinct genes. These proteins are thought to play a prominent role in regulating the bioavailability and distribution of IGFs [82, 83]. Interestingly, some IGFBPs are in close relationship with cell plasma membranes (through binding to integrins or the extracellular matrix), but their relationship with MHC as well as their potential implication in thymic IGF presentation to immature T cells deserve to be further investigated.

Table 2. The dual role of thymic neuroendocrine self peptides in T-cell differentiation

Thymic Neuroendocrine Self Peptides

Neuroimmunomodulation 1999;6:115–125

Cryptocrine signalling

Presentation of neuroendocrine self antigens

Physiology Accessory signal in T-cell development and activation (positive selection)

Physiology T-cell education to neuroendocrine families (negative selection of selfreactive T cells)

Pathophysiology Oversecretion in the bloodstream (paraneoplastic syndrome)

Pathophysiology Failure (or breakdown) of imunological tolerance to neuroendocrine families Autoimmune endocrine diseases

Involvement in the biology of T-cell lymphomas Pharmacology Immunomodulation by neuropeptide agonists or antagonists

Pharmacology ‘Reprogramming’ of the immunological tolerance to neuroendocrine families ‘Tolerogenic’ vaccination for prevention of autoimmune endocrine diseases

A Role Played by a Trouble in Thymic T-Cell Education in Autoimmunity?

The development of an autoimmune disease affecting the neuroendocrine system may be viewed as a failure to develop or maintain tolerance to cellular or molecular components which are constitutively expressed by neuroendocrine cells (i.e. autoantigens such as insulin or GAD). In this view, a large body of research recently focused on the physiological mechanisms which underline the establishment of self tolerance and about the potential factors leading to its breakdown in autoimmunity. Though the relationship between lymphoepithelial structures and autoimmunity was suspected in 1962 by Burnet and Mackay [84], the question of a defective thymic T-cell negative selection or self education in the pathophysiology of autoimmune diseases has not been intensively investigated. Nevertheless, it was shown that neonatal thymectomy prevents the emergence of diabetes in an animal model of autoimmune type 1 diabetes, the Bio-Breeding (BB) rat [85]. In clinical practice also, thymectomy usually induces an improvement of patients suffering from autoimmune myasthenia gravis, especially when a thymoma (hyperplasia of thymic epithelium) is associated [86]. In both cases, the benefit of thymectomy might in fact result

121

from the removal of the defective thymic censorship which is responsible for a continuous release and enrichment of the peripheral T-cell pool with intolerant and selfreactive lymphocytes. The development of diabetes is prevented by the transplantation of thymus from diabetesresistant (DR) to diabetes-prone (DP) BB rats [87]. The transplantation of the thymus from NOD mice to DR mouse strains was also shown to induce diabetes in the recipients [88]. While bone marrow transplantation is rather ineffective in preventing autoimmune diseases of MRL/+ mice, thymus transplantation is a crucial factor for their prevention [89]. A defective process of thymic T-cell negative selection has been suggested on the basis that the thymus of DR BB rats contains thymocytes predisposed to autoreactivity [90]. Another argument is the observation that grafts of pure thymic epithelium from NOD mouse embryos to newborn C57BL/6 athymic mice induced CD4 and CD8 T-cell-mediated insulitis and sialitis [91]. At the histological level, a defect in thymic function could be linked to a disorganization of the microenvironment such as the giant perivascular spaces observed in the NOD mouse thymus [92], and the epithelial defects of BB rat thymus [93]. Recently, we examined the elution profiles of ir-IGFs in the thymus from Wistar-Furth (WF) normal rats, DR and DP BB rats. A peak of ir-IGF-2 1 10 ng/ml was observed in the G75 profile of WF thymus extracts; a peak around 1.5 ng/ml was eluted from DR BB rat thymic extracts, while IGF-2 concentrations were almost undetectable in DP BB rats [94]. Altogether, these observations support the hypothesis that a defective thymic censorship or T-cell self education might well take an active part in the pathophysiology of autoimmune type 1 diabetes.

Neuroendocrine Self Antigens versus Autoantigens: Toward the Design of Tolerogenic Vaccines for the Prevention of Autoimmune Diseases?

Three types of factors are usually thought to be implicated in the pathogeny of autoimmune diseases. (1) The effector immune components are CD4- and CD8-autoreactive T cells which are specifically oriented against a given target cell or molecule. These autoreactive T cells result from a spontaneous breakdown of T-cell tolerance, either at the central thymic and/or the peripheral level. (2) A series of extra- and intra-MHC genes have been demonstrated to be related to different autoimmune diseases. Some of these genes could intervene in the presen-

122

Neuroimmunomodulation 1999;6:115–125

tation of target autoantigens to autoreactive T lymphocytes, but others certainly not. (3) Finally, an environmental factor is involved and could be implicated in establishing a link between the target autoantigens and autoreactive T cells. A molecular mimicry between target autoantigens and microorganisms might play at this level and this hypothesis recently received some experimental support [95]. The intervention of microbial superantigens has also been proposed to activate peripheral autoreactive T cells [96]. A preventive strategy of autoimmune diseases can hardly been designed on the basis of the genetic components of autoimmune disease or the hazardous environmental factors. Manipulation of autoreactive T cells seems to be a more promising way by which an efficient prevention of autoimmunity can be envisioned. In the neurohypophysial family, evidence has been presented that OT seems to be the neurohypophysial self antigen. A strong immunological tolerance protects the OT lineage, more than the VP one, from an autoimmune aggression. Indeed, some cases of idiopathic diabetes insipidus have been shown to result from an autoimmune hypothalamitis oriented toward VP-producing neurons [97, 98]. Given the implication of the OT lineage in the reproductive process, a stronger tolerance of this lineage is important for the preservation of the species. Thus, in the neurohypophysial family, while OT behaves as the self antigen, VP is suspected to be one target autoantigen of the autoimmune process. As discussed previously, this conclusion is also supported by the frequence and the titers of Abs induced by active immunization against neurohypophysial peptides [VP 11 OT]. An infiltration of the hypothalamo-neurohypophysial tract by inflammatory mononuclear cells has been observed repeatedly, both after active immunization against VP [99], and in spontaneous diabetes insipidus [98]. These observations suggest that hypothalamic magnocellular neurons express on their surface antigenic markers specific of their neurosecretory activity. There is now substantial evidence that insulin is one important among other autoantigens tackled by various autoreactive components of the immune system both in animal and human type 1 diabetes [100, 101]. Moreover insulin is the specific marker of the pancreatic islet endocrine ß cells. Oral, intranasal and parenteral administration of insulin or insulin-derived dominant autoantigens have been shown to inhibit the occurrence of diabetes in animal models of type 1 diabetes [102, 103]. However, one cannot exclude the risk of priming or triggering autoimmunity by peripheral administration of an autoantigen [104]. Reprogramming the immunological tolerance that

Geenen/Kecha/Brilot/Charlet-Renard/ Martens

is thought to be broken in autoimmunity represents an attractive strategy for the prevention of devastating autoimmune diseases such as multiple sclerosis, rheumatoid arthritis or type 1 diabetes. Such reprogramming could be based upon the natural tolerogenic properties of the thymic epithelium. Instead of a classical vaccination (with immune activation and induction of memory immunocompetent cells), a form of tolerogenic vaccination is proposed that would lead to the deletion or the anergy of peripheral autoreactive T lymphocytes. The induction of T-cell tolerance following peptide vaccination has already been obtained with synthetic peptides representing cytotoxic CD8 epitopes of T cells oriented against tumor antigens or viruses [105]. An efficient and nontoxic prevention of autoimmune diseases, perhaps even their erad-

ication, could depend upon the strategic choice that will be made between either the induction of tolerance to specific autoantigens following their parenteral administration, or the exploitation of the putative tolerogenic proporties of thymic self antigens. Acknowledgments V. Geenen is Senior Research Associate of the National Fund of Scientific Research (Belgium); H. Martens is supported by Télévie/ FRSM. Our studies have been performed with the financial help of the Juvenile Diabetes Foundation International, the Suzanne et Jean Pirart Fund of the Belgian Association of Diabetes, the National Fund of Scientific Research [Belgium], Télévie/FRSM, the Association contre le Cancer [Belgium], and the Foundation Léon Fredericq [Liège University Medical School].

References 1 Miller JFAP: Immunological function of the thymus: Lancet 1961;ii:748–749. 2 Geenen V, Kroemer G: Multiple ways to cellular immune tolerance. Immunol Today 1993; 14:573–575. 3 Dustin ML, Springer TA: Role of lymphocyte adhesion receptors in transient interactions and cell locomotion. Annu Rev Immunol 1991; 9:27–66. 4 Ritter MA, Rozing J, Schuurman HJ: The true function of the thymus? Immunol Today 1988; 9:189–193. 5 MacDonald HR, Glasebrook AL, Schneider R, Lees RL, Pircher H, Pedrazzini T, Kanagawa O, Nicolas JF, Howe RC, Zinkernagel RM, Hengartner H: T cell reactivity and tolerance to Mls·-encoded antigens. Immunol Rev 1989; 107:89–108. 6 Von Boehmer H: The developmental biology of T lymphocytes. Annu Rev Immunol 1989;6: 309–326. 7 Sprent J, Webb S: Can self/nonself discrimination be explained entirely by clonal deletion? Res Immunol 1992;143:285–287. 8 Hogquist KA, Jameson SC, Bevan MJ: The ligand for positive selection of T lymphocytes in the thymus. Curr Opin Immunol 1994;6: 273–278. 9 Von Boehmer H: Thymic selection: A matter of life and death. Immunol Today 1992;13:454– 458. 10 Sprent J, Tough DT: Lymphocyte life-span and memory. Science 1994;265:1395–1400. 11 Ashton-Rickardt PG, Tonegawa S: A differential avidity model for T-cell selection. Immunol Today 1994;15:362–366. 12 Sebzda E, Wallace VA, Mayer J, Young RSM, Mak T, Ohashi PS: Positive and negative thymocyte selection induced by different concentrations of a single peptide. Science 1994;263: 1615–1618.

Thymic Neuroendocrine Self Peptides

13 Allen PM: Peptides in positive and negative selection: A delicate balance. Cell 1994;76: 593–596. 14 Ott I, Scott JC: The galactogogue action of the thymus and corpus luteum. Proc Soc Exp Biol Med 1910;8:49. 15 Geenen V, Legros JJ, Franchimont P, Baudrihaye M, Defresne MP, Boniver J: The neuroendocrine thymus: Coexistence of oxytocin and neurophysin in the human thymus. Science 1986;232:508–511. 16 Markwick AJ, Lolait SJ, Funder JW: Immunoreactive arginine vasopressin in the rat thymus. Endocrinology 1986;119:1060–1064. 17 Geenen V, Legros JJ, Franchimont P, Defresne MP, Boniver J, Ivell R, Richter D: The thymus as a neuroendocrine organ. Synthesis of vasopressin and oxytocin in human thymic epithelium. Ann NY Acad Sci 1987;496:56–66. 18 Moll UM, Lane BL, Robert F, Geenen V, Legros JJ: The neuroendocrine thymus. Abundant occurrence of oxytocin-, vasopressin-, and neurophysin-like peptides in epithelial cells. Histochemistry 1988;89:385–390. 19 Argiolas A, Gessa GL, Melis MR, Stancampiano R, Vaccari A: Effects of neonatal and adult thyroid dysfunction on thymic oxytocin. Neuroendocrinology 1990;52:556–559. 20 Robert F, Geenen V, Schoenen J, Burgeon E, De Groote D, Defresne MP, Legros JJ, Franchimont P: Colocalization of immunoreactive oxytocin, vasopressin and interleukin-1 in human thymic epithelial neuroendocrine cells. Brain Behav Immun 1991;5:102–115. 21 Geenen V, Robert F, Martens H, Benhida A, Degiovanni G, Defresne MP, Boniver J, Legros JJ, Martial J, Franchimont P: At the cutting edge. Biosynthesis and paracrine/cryptocrine actions of ‘self’ neurohypophysial-related peptides in the thymus. Mol Cell Endocrinol 1991; 76: C27–C31.

22 Milan J, Barbijeri M, Kovacevic D, Arambasic M, Kartaljevic G, Natalic D, Pazin S: Identification of neuroendocrine oxytocic activity of the human fetal thymus. Thymus 1990;15: 181–185. 23 Geenen V, Defresne MP, Robert F, Legros JJ, Franchimont P, Boniver J: The neurohormonal thymic microenvironment: Immunocytochemical evidence that thymic nurse cells are neuroendocrine cells. Neuroendocrinology 1988; 47:365–368. 24 Geenen V, Vandersmissen E, Martens H, Goxe B, Kecha O, Legros JJ, Lefèbvre PJ, Benhida A, Rentier-Delrue F, Martial JA: Cellular and molecular aspects of thymic T-cell education to neurohypophysial principles; in Saito T, Kurokawa K, Yoshida S (eds): Neurohypophysis: Recent Progress of Vasopressin and Oxytocin Research. Amsterdam, Elsevier, 1995, pp 309– 319. 25 Wiemann M, Ehret G: Subcellular localization of immunoreactive oxytocin within thymic epithelial cells of the male mouse. Cell Tissue Res 1993; 273:79–87. 26 Kumamoto K, Matsuura T, Amagai T, Kawata M: Oxytocin-producing and vasopressin-producing eosinophils in the mouse spleen: Immunohistochemical, immuno-electron-microscopic and in situ hybridization studies. Cell Tissue Res 1995;281:1–10. 27 Webb SR, Spent J: Tolerogenicity of thymic epithelium. Eur J Immunol 1990;20:2525– 2528. 28 Lorenz RG, Allen PM: Thymic cortical epithelial cells can present self-antigens in vivo. Nature 1989;337:560–562. 29 Geenen V, Vandersmissen E, Cormann-Goffin N, Martens H, Legros JJ, Degiovanni G, Benhida A, Martial J, Franchimont P: Membrane translocation and relationship with MHC class I of a human thymic neurophysin-like protein. Thymus 1993;22:55–66.

Neuroimmunomodulation 1999;6:115–125

123

30 Rosenbaum LC, Neuwelt EA, Van Tol HHM, Peng Loh Y, Verbalis J, Hellström I, Hellström KE, Nilaver G: Expression of neurophysinrelated precursor in cell membranes of a smallcell lung carcinoma. Proc Natl Acad Sci USA 1990;87:9928–9932. 31 North WG, Yu X: Forms of neurohypophysial peptides generated by tumors, and factors regulating their expression. Regul Pept 1993;45: 209–216. 32 Griffin GH, Alazard R, Cohen P: Complex formation between bovine neurophysin-1 and oxytocin, vasopressin and tripeptide analogs of their NH2-terminal region. J Biol Chem 1973; 248:7975–7978. 33 Maryanski JL, Romero P, Van Pel A, Boon T, Salemme FR, Cerrottini JC, Corradin G: The identification of tyrosine as a common key residue in unrelated H-2Kd restricted antigenic peptides. Int Immunol 1991;3:1035–1042. 34 Martens H, Malgrance B, Robert F, Charlet C, De Groote D, Heymann D, Godard A, Soulillou JP, Moonen G, Geenen V: Cytokine production by human thymic epithelial cells: Control by the immune recognition of the neurohypophysial self-antigen. Regul Pept 1996;67:39– 45. 35 Funder JW: Paracrine, cryptocrine, acrocrine. Mol Cell Endocrinol 1990;70:C21–C24. 36 Elands J, Resink A, De Kloet ER: Oxytocin receptors in the rat thymic gland. Eur J Pharmacol 1988;151:345–351. 37 Elands J, Resink A, De Kloet ER: Neurohypophysial hormone receptors in the rat thymus, spleen and lymphocytes. Endocrinology 1990; 126:2703–2710. 38 Martens H, Robert F, Legros JJ, Geenen V, Franchimont P: Expression of functional neurohypophysial peptide receptors by immature and cytotoxic T-cell lines. Prog Neuroendocrinimmunol 1992;5:31–39. 39 Torres BA, Johnson HM: Arginine vasopressin (AVP) replacement of helper cell requirement in IFNg production. Evidence for a novel AVP receptor on mouse lymphocytes. J Immunol 1988;81:132–136. 40 Caldwell JD, Walker CA, Noonan LR, Jirikowski GF, Peterson G, Pedersen CA, Mason GA: Changes in thymic oxytocin receptors during early development and in steroid-treated adult rats. Prog NeuroendocrinImmunol 1991; 4:223–233. 41 Lolait SJ, O’Carroll AM, Mahan LC, Felder CC, Button DC, Young WS III, Mezey E, Brownstein MJ: Extrapituitary expression of the rat V1b vasopressin receptor gene. Proc Natl Acad Sci USA 1995;92:6783–6787. 42 Whitfield JF, Perris AD, Youdale T: The calcium-mediated promotion of mitotic activity in rat thymocyte populations by growth hormone, neurohormones, parathyroid hormone and prolactin. J Cell Physiol 1969;73:203– 209. 43 Goren HJ, Okabe T, Lederis K, Hollenberg MD: Oxytocin stimulates glucose oxidation in rat thymocytes. Proc West Pharmacol Soc 1984;27:461.

124

44 Martens H, Kecha O, Charlet-Renard C, Defresne MP, Geenen V: Neurohypophysial peptides stimulate the phosphorylation of pre-T cells focal adhesion kinases. Neuroendocrinology 1998;67:282–289. 45 Geenen V, Martens H, Robert F, VrindtsGevaert Y, De Groote D, Franchimont P: Immunomodulatory properties of cyclic hexapeptide oxytocin antagonists. Thymus 1992;20: 217–226. 46 Sundler F, Carraway RE, Hakanson R, Alumets J, Dubois MP: Immunoreactive neurotensin and somatostatin in the chicken thymus. A chemical and histochemical study. Cell Tissue Res 1978;194:367–376. 47 Vanneste Y, Ntodou Thome A, Vandersmissen E, Charlet C, Franchimont D, Martens H, Lhiaubet AM, Schimpff RM, Rostène W, Geenen V: Identification of neurotensin-related peptides in human thymic epithelial cell membranes and relationship with major histocompatibilty complex class I molecules. J Neuroimmunol 1997;76:161–166. 48 Ericsson A, Geenen V, Robert F, Legros JJ, Vrindts-Gevaert Y, Franchimont P, Brene S, Persson H: Expression of preprotachykinin A and neuropeptide-Y messenger RNA in the thymus. Mol Endocrinol 1990;4:1211–1218. 49 Söder O, Hellström PM: The tachykinins neurokinin A and physalaemin stimulate murine thymocyte proliferation. Int Arch Allergy Appl Immunol 1989;90:91–96. 50 Geppetti P, Theodorsson-Norheim E, Ballerini G, Alessandri M, Maggi CA, Santicioli P, Amenta F, Fanciullacci M: Capsaicin-sensitive tachykinin-like immu-noreactivity in the thymus of rats and guinea pigs. J Neuroimmunol 1988;19:3–9. 51 Shigematsu K, Saavedra JM, Kurihara M: Specific substance P binding sites in rat thymus and spleen: In vitro autoradiographic study. Regul Pept 1986;16:147–156. 52 Vollmar AM, Schulz R: Atrial natriuretic peptide is synthesized in the human thymus. Endocrinology 1990;126:2227–2281. 53 Vollmar AM, Wolf R, Schulz R: Co-expression of the natriuretic peptides (ANP, BNP, CNP) and their receptors in normal and acutely involuted rat thymus. J Neuroimmunol 1995;57: 117–127. 54 Vollmar AM: Influence of atrial natriuretic peptide on thymocyte development in fetal thymic organ culture. J Neuroimmunol 1997; 78:90–96. 55 Batanero E, De Leeuw FE, Jansen GH, Van Wicken DF, Huber J, Schuurman HJ: The neural and neuroendocrine component of the human thymus. Brain Behav Immun 1992;6:249– 264. 56 Martin-Fontecha A, Broekhuizen R, De Heer C, Zapata A, Schuurman HJ: The neuroendocrine component of the rat thymus: Studies on cultured thymic fragments before and after transplantation in congenitally athymic and euthymic rats. Brain Behav Immun 1993;7:1–15. 57 Savino W, Dardenne M: Immune-neuroendocrine interactions. Immunol Today 1995;16: 318–322.

Neuroimmunomodulation 1999;6:115–125

58 Geenen V, Achour I, Robert F, Vandersmissen E, Sodoyez JC, Defresne MP, Boniver J, Lefèbvre PJ, Franchimont P: Evidence that insulin-like growth factor 2 is the dominant thymic peptide of the insulin superfamily. Thymus 1993;21:115–27. 59 Sodoyez JC, Koch M, Lemaire I, Sodoyez-Goffaux F, Rapaille A, François-Gérard C, Sondag D: Influence of the affinity of antibodies upon their detection by liquid phase radiobinding assay and solid phase enzyme-linked immunosorbent assay: Demonstration using monoclonal antibodies raised against rDNA human proinsulin. Diabetologia 1991;34:463–468. 60 Arkins S, Rebeiz N, Biragyn A, Reese DL, Kelley KW: Murine macrophages express abundant insulin-like growth factor-I Ea and Eb transcripts. Endocrinology 1993;133:2334– 2343. 61 Arkins S, Rebeiz N, Brunke-Reese DL, Biragyn A, Kelley KW: Interferon-gamma inhibits macrophage insulin-like growth factor-I synthesis at the transcriptional level. Mol Endocrinol 1995;9:350–360. 62 Arkins S, Rebeiz N, Brunke-Reese DL, Minshall C, Kelley KW: The colony-stimulating factors induce expression of insulin-like growth factor-I messenger ribonucleic acid during hematopoiesis. Endocrinology 1995;136:1153– 1160. 63 Kecha O, Achour I, Hodzic D, Goxe B, Winkler R, Geenen V: IGF-II expression in the human thymus. Proc 10th Int Congr Endocrinol 1996, abstr P1-659. 64 Kecha O, Achour I, Martens H, Winkler R, Lefèbvre PJ, Geenen V: Characterization of the insulin-like growth factor (IGF) axis in the human thymus. Proc 79th Annu Meeting Endocr Soc, 1997, abstr P3-263. 65 van Buul-Offers SC, de Haan K, Reijnen-Gresnigt MG, Meinsma D, Jansen M, Oei SL, Bonte EJ, Sussenbach JS, Van den Brande JL: Overexpression of human insulin-like growth factor II in transgenic mice causes increased growth of the thymus. J Endocrinol 1995;144:491–502. 66 Pansky B, House EL, Cole LA: An insulin-like thymic factor. A preliminary report. Diabetes 1965;14:325–332. 67 Zapf J, Hauri C, Waldvogel M, Froesch ER: Acute metabolic effects and half-lives of intravenously administered insulin-like growth factors I and II in normal and hypophysectomized rats. J Clin Invest 1986;77:1768–1775. 68 Jolicœur C, Hanahan D, Smith KM: T-cell tolerance toward a transgenic ß-cell antigen and transcription of endogenous pancreatic genes in the thymus. Proc Natl Acad Sci USA 1994; 91:6707–6711. 69 Vafiadis P, Bennett ST, Todd JA, Nadeau J, Grabs R, Goodyear CG, Wickramasinghe S, Colle E, Polychronakos C: Insulin expression in human thymus is modulated by INS VNTR alleles at the IDDM2 locus. Nat Genet 1997;15: 289–292.

Geenen/Kecha/Brilot/Charlet-Renard/ Martens

70 Pugliese AA, Zeller M, Fernandez A Jr, Zalcberg LJ, Bartlett RJ, Ricordi C, Pietropaolo M, Eisenbarth GS, Bennett SJ, Patel DD: The insulin gene is transcribed in the human thymus and transcription levels correlate with allelic variation at INS VNTR-IDDM2 susceptibility locus for type 1 diabetes. Nature Genet 1997;15:293–297. 71 Verland S, Gammeltoft S: Functional receptors for insulin-like growth factors I and II in rat thymocytes and mouse thymoma cells. Mol Cell Endocrinol 1989;67:207–216. 72 Kozak RW, Hassell JF, Greenstein LA, Rechler MM, Waldmann TA, Nissley SP: Type 1 and 2 insulin-like growth factor receptors on human phytohemagglutinin-activated lymphocytes. Cell Immunol 1987;109:318–331. 73 Johnson EW, Jones LA, Kozak RW: Expression and function of insulin-like growth factor receptors on anti-CD3-activated human T lymphocytes. J Immunol 1992;148:63–71. 74 Kooijman R, Scholtens LE, Rijkers GT, Zegers BJM: Differential expression of type 1 insulinlike growth factor receptors in different stages of human T cells. Eur J Immunol 1995;25:931– 935. 75 Kooijman R, van Buul-Offers SC, Scholtens LE, Schuurman HJ, Van den Brande JL, Zegers BJM: T cell development in insulin-like growth factor-II transgenic mice. J Immunol 1995;154: 5736–5745. 76 Goxe B, Martens H, Vandersmissen E, Achour I, Kecha O, Geenen V: Interactions entre les cellules T et la famille insulinique: étude du récepteur de l’IGF-II. Ann Endocrinol 1995; 56:399. 77 Martens H, Goxe B, Geenen V: The thymic repertoire of neuroendocrine-related self-antigens: Physiological implications in T-cell life and death. Immunol Today 1996;17:312–317. 78 Gammon G, Sercarz E: How some T cells escape tolerance induction. Nature 1989;342: 183–185. 79 Adorini L, Appella E, Doria G, Nagy ZA: Mechanisms influencing the immunodominance of T cell determinants. J Exp Med 1988; 168:2091–2104. 80 Cabaniols JP, Cibotti R, Kourilsky P, Kosmatopoulos K, Kanellopoulos J: Dose-dependent T cell tolerance to an immunodominant self peptide. Eur J Immunol 1994;24:1743–1749. 81 Carraway RE, Ruane SE, Kim HR: Distribution and immunochemical character of neurotensin-like material in representative vertebrates and invertebrates: Apparent conservation of the COOH-terminal region during evolution. Peptides 1982;3:115–123.

Thymic Neuroendocrine Self Peptides

82 Clemmons DR, Busby WH, Arai T, Nam TJ, Clarke JB, Jones JI, Ankrapp DK: Role of insulin-like grow factor binding proteins in the control of IGF actions. Prog Growth Factor Res 1995;6:357–366. 83 Kelley KM, Oh Y, Gargosky SE, Gucev Z, Matsumoto T, Hwa V, Ng L, Simpson DM, Rosenfeld RG: Insulin-like growth factor-binding proteins and their regulatory dynamics. Int J Biochem Cell Biol 1996; 6:619–637. 84 Burnet FM, Mackay IR: Lymphoepithelial structures and autoimmune disease. Lancet 1962;ii:1030–1033. 85 Like AA, Kislaukis E, Williams RM, Rossini AA: Neonatal thymectomy prevents spontaneous diabetes mellitus in the BB:W rat. Science 1982;216:644–646. 86 Newsom-Davis J: Myasthenia gravis. Med Int 1987;48:1988–1991. 87 Georgiou HM, Bellgrau D: Thymus transplantation and disease prevention in the diabetesprone Bio-Breeding rat. J Immunol 1989;142: 3400–3405. 88 Georgiou HM, Mandel TE: Induction of insulitis in athymic (nude) mice. The effect of NOD thymus and pancreas transplantation. Diabetes 1995;44:49–59. 89 Hosaka N, Nose M, Kyogoku M, Nagata N, Miyashima S, Good RA, Ikehara S: Thymus transplantation, a critical factor for correction of autoimmune disease in aging MRL/+ mice. Proc Natl Acad Sci USA 1996;93:8558–8562. 90 Whalen BJ, Rossini AA, Mordes JP, Greiner DL: DR-BB rat thymus contains thymocyte populations predisposed to autoreactivity. Diabetes 1995;44:963–967. 91 Thomas-Vaslin V, Damotte D, Coltey M, Le Douarin NM, Coutinho A, Salaün J: Abnormal T cell selection on NOD thymic epithelium is sufficient to induce autoimmune manifestations in C57BL/6 athymic nude mice. Proc Natl Acad Sci USA 1997;94:4598–4603. 92 Savino W, Carnaud C, Luan JJ, Bach JF, Dardenne M: Characterization of the extracellular matrix-containing giant perivascular spaces in the NOD mouse thymus. Diabetes 1993;42: 134–140. 93 Doukas J, Mordes JP, Swymer C, Niedzwiecki D, Mason R, Rozing J, Rossini AA, Greiner DL: Thymic epithelial defects and predisposition to autoimmune diabetes in BB rats. Am J Pathol 1994;145:1517–1525. 94 Geenen V, Achour I, Kecha O, Greiner DL, Rossini AA, Lefèbvre PJ: Thymic insulin-like growth factors in man and in an animal model of autoimmune IDDM. Diabetologia 1996;39 (suppl 1):A15. 95 Atkinson MA, Maclaren NK: The pathogenesis of insulin-dependent diabetes mellitus. N Engl J Med 1994;331:1428–1436.

96 Conrad B, Weidmann E, Trucco G, Rudert WA, Behboo R, Ricordi C, Rodriquez-Rilo H, Finegold D, Trucco M: Evidence for superantigen involvement in insulin-dependent diabetes mellitus aetiology. Nature 1994;371: 351–355. 97 Scherbaum WA, Bottazzo GF: Autoantibodies to vasopressin cells in idiopathic diabetes insipidus: Evidence for an autoimmune variant. Lancet 1983;i:897–901. 98 Imura H, Nakao K, Shimatsu A, Ogawa Y, Sando T, Fujisawa I, Yamabe H: Lymphocytic infundibuloneurohypophysitis as a cause of central diabetes insipidus. N Engl J Med 1993;329:683–689. 99 Cau P, Rougon-Capuzzi G: Autoimmune alterations in the neurohypophysis of rabbits immunized against vasopressin. Brain Res 1979;177:265–271. 100 Simone EA, Yu L, Wegmann DR, Eisenbarth GS: T cell receptor gene polymorphisms associated with anti-insulin, autoimmune T cells in diabetes-prone NOD mice. J Autoimmun 1997;10:317–321. 101 Daniel D, Gill RG, Schloot N, Wegmann DR: Epitope specificity, cytokine production profile and diabetogenic activity of insulin-specific T cell clones isolated from NOD mice. Eur J Immunol 1995;25:1056–1062. 102 Zhang ZJ, Davidson L, Eisenbarth GS, Weiner HL: Suppression of diabetes in nonobese diabetic mice by oral administration of porcine insulin. Proc Natl Acad Sci USA 1991; 88:10252–10256. 103 Daniel D, Wegmann DR: Protection of nonobese diabetic mice from diabetes by intranasal or subcutaneous administration of insulin peptide B-(9–23). Proc Natl Acad Sci USA 1996;93:956–960. 104 Blanas E, Carbone FR, Allison J, Miller JFAP, Heath WR: Induction of auto-immune diabetes by oral administration of autoantigen. Science 1996;274:1707–1709. 105 Toes RM, Offringa R, Blom RJJ, Melief CJM, Kast WM: Peptide vaccination can lead to enhanced tumor growth through specific Tcell tolerance induction. Proc Natl Acad Sci USA 1996;93:7855–7860. 106 Kramer S, Reynolds FH Jr, Castillo M, Valenzuela DM, Thorikay M, Sorvillo JM: Immunological identification and distribution of parathyroid hormone-like protein polypeptide in normal and malignant tissues. Endocrinology 1991;128:1927–1937. 107 Bulloch K, McEwen BS, Diwa A, Baird S: Relationship between dehydro-epiandrosterone and calcitonin gene-related peptide in the mouse thymus. Am J Physiol 1995;268: E168–E173.

Neuroimmunomodulation 1999;6:115–125

125

Neuroimmunomodulation 1999;6:126–136

Immunoneuroendocrine Connectivity: The Paradigm of the Thymus-Hypothalamus/Pituitary Axis Wilson Savino a Eduardo Arzt b Mireille Dardenne c a Laboratory on Thymus Research, Department of Immunology, Institute Oswaldo Cruz, Foundation Oswaldo Cruz, Rio de Janeiro, Brazil; b Laboratory of Physiology and Molecular Biology, Department of Biology, University of Buenos Aires and CONICET, Buenos Aires, Argentina, and c Hôpital Necker, CNRS-URA 1461, Paris, France

Key Words Thymus W Thymic epithelial cells W Thymocytes W Hypothalamus W Pituitary gland W Cytokines Abstract It is now largely established that the immune and neuroendocrine systems cross-talk by using similar ligands and receptors. In this context, the thymus-hypothalamus/pituitary axis can be regarded as a paradigm of connectivity in both normal and pathological conditions. For example, cytokines and thymic hormones modulate hypothalamic-pituitary functions: (a) interleukin (IL)-1 seems to upregulate the production of corticotropinreleasing factor and by adrenocorticotropin by hypothalamic neurons and pituitary cells, respectively; (b) thymulin enhances LH secretion. Conversely, a great deal of data strongly indicate that the hypothalamic-pituitary axis plays a role in the control of thymus physiology. Growth hormone (GH) for example, enhances thymulin secretion by thymic epithelial cells (TEC), both in vivo and in vitro, also increasing extracellular matrix-mediated TEC/thymocyte interactions. Additionally, gap junction-mediated cell coupling among TEC is upregulated by ACTH. In a second vein, it was shown that GH injections in aging mice increased total thymocyte numbers and the percentage of CD3-bearing cells, as well concanavalin-A mitogenic response and IL-6 production. In addition to mutual effects, thymus-pituitary similari-

ABC

© 1999 S. Karger AG, Basel 1021–7401/99/0062–0126$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: http://BioMedNet.com/karger

ties for cytokine and hormone production have been demonstrated. Cytokines such as IL-1, IL-2, IL-6, interferon-Á, transforming growth factor-ß and others can be produced by hypothalamic and/or pituitary cells. Conversely, hormones including GH, PRL, LH, oxytocin, vasopressin and somatostatin can be produced intrathymically. Moreover, receptors for various cytokines and hormones are expressed in both the thymus and the hypothalamus/pituitary axis. Lastly, it is noteworthy that a thymus-pituitary connectivity can also be seen under pathological situations. In this regard, an altered HPA axis has been reported in AIDS, human falciparum malaria and murine rabies, that also show a severe thymic atrophy.

The Interdependence of the Neuroendocrine and Immune Systems

The cross-talk between the immune and neuroendocrine systems is now largely established. These systems use similar soluble ligands to create and maintain physiological intra- and inter-system communication circuitries, which play a relevant role in homeostasis [1–3]. Not only do the cells of the immune system produce classical hormones, but also the endocrine glands and nervous tissues synthesize and release a variety of cytokines. Moreover, specific receptors for such distinct molecular families

Wilson Savino Laboratory on Thymus Research – Department of Immunology Institute Oswaldo Cruz – Foundation Oswaldo Cruz Ave. Brasil 4365 Manguinhos, Rio de Janeiro, RJ 21045-900 (Brazil) Tel. +55 21 280 1486, Fax +55 21 280 1589, E-Mail [email protected]

Cytokines are now recognized as playing important roles in modulating the neuroendocrine system, particularly the hypothalamus-pituitary-adrenal (HPA) axis, and in this respect, interleukin (IL)-1 is a potent regulatory molecule. In spite of some discrepancies in the literature it seems to act at hypothalamic, pituitary and adrenal levels, inducing respectively corticotropin-releasing factor (CRH), adrenocorticotropin (ACTH) and cortisol production [4–6]. When the influence on hormone secretion was compared, IL-1· was less potent than IL-1ß [7, 8]. Specifically at the pituitary level, however, the effects of IL-1 are controversial. Growth hormone (GH), prolactin (PRL), luteinizing hormone (LH), and follicle-stimulating hormone (FSH) secretion by rat pituitary cells have been reported not to be influenced by IL-1· [9], whereas other reports have shown stimulation of GH, LH and thyroidstimulating hormone (TSH) secretion [10], together with inhibition of PRL release [11, 12]. In rat pituitary cell cultures, ACTH secretion was stimulated [7, 9, 10, 12–14] or not influenced [15, 16] by IL-1ß, whereas secretion of this pituitary hormone is enhanced by IL-1· in AtT20 cells [17–20] and in human corticotroph pituitary adenoma cell cultures [17]. Lastly, IL-1 further increases ACTH secretion by corticotrophs previously exposed to CRH [21]. Other inflammatory cytokines such as IL-6 and tumor necrosis factor-· (TNF-·) share, to a certain extent, the effects of IL-1. They act at three levels of the HPA axis, stimulating CRH production by the hypothalamus, but displaying particular stimulatory patterns [5, 22]. Concerning the pituitary gland, IL-6 stimulates the release of ACTH, PRL, GH, LH and FSH from normal rat anterior pituitary cells in vitro [22–24] and stimulates ACTH secretion from AtT20 cells [18]. In the same model, IL-6 inhibits stimulation of adenylate cyclase by vasoactive intestinal peptide (VIP) and suppresses the enhancement in inositol phosphate and intracellular calcium induced by thyrotropin-releasing hormone (TRH) [25]. In addition, IL-6 modulates proliferation of pituitary cells, inhibiting normal rat pituitary cell growth [26, 27] and stimu-

lating growth of GH3 and MtT/E rat pituitary tumor cell lines [26, 28]. In this regard, IL-6 was shown to enhance c-fos expression by pituitary gland but not hypothalamic cells [29, 30]. TNF-· stimulates ACTH, GH, TSH and PRL secretion of rat pituitary cells in vitro [31, 32]. In contrast, TNF-· was found to inhibit the release of ACTH and other hormones in response to hypothalamic factors, by acting directly on pituitary cells [33]. Additionally, leukemia inhibitory factor (LIF) stimulates proopiomelanocortin (POMC) expression and ACTH release in vitro by corticotroph cells [34, 35]. In keeping with ACTH stimulation in vitro, recent data revealed that mice carrying the LIF transgene associated to the GH promoter presented over 2-fold more corticotroph cells than wild controls, whereas the numbers of GH and PRL cells were reduced [36]. Classical T-cell-derived cytokines, such as IL-2 and interferon-Á also act on the HPA axis. When administered to human cancer patients, IL-2 increases ß-endorphin, ACTH and cortisol levels [37, 38]. Rat IL-2 enhances ACTH secretion in vivo [39], and induces corticosterone production by a direct action on rat adrenocortical cells [40] as well as CRH release from hypothalamic neurons [41]. In normal rat pituitaries, IL-2 has been shown to stimulate POMC mRNA expression [42, 43] as well as ACTH, PRL and TSH secretion, but to inhibit LH, FSH and GH release [44–47]. The release pattern of pituitary hormones after IL-2 stimulation closely mimics the alterations of pituitary hormone secretion in response to stress [44]. IL2-induced PRL secretion is blocked by dopamine [46]. IL2 enhances ACTH secretion in AtT20 cells [45] and stimulates PRL release from GH3 cells [48]; in the latter, antiestrogens inhibit IL-2-induced PRL secretion. Interferon-Á also stimulates PRL release in rat anterior pituitary cell cultures, probably acting via the release of IL-6 from folliculostellate cells [49]. Nevertheless, other studies showed that these cells mediate an inhibitory action of interferon-Á on PRL secretion, as well as on ACTH and GH release [50], possibly involving the release of nitric oxide by folliculostellate cells [51, 52]. Interestingly, another cytokine, the stem cell factor, when injected in vivo in rats, was able to enhance ACTH and PRL in a dose-dependent manner, as well as activate hypothalamic neurosecretory neurons, as ascertained by c-fos expression [53]. Cytokine effects on hypothalamic-pituitary hormonal functions are shown in table 1. In addition to cytokines, thymic hormones modulate the production of classic hypothalamic-pituitary hormones and neuropeptides. Initial experiments revealed

Immunoneuroendocrine Connectivity

Neuroimmunomodulation 1999;6:126–136

can be detected in both the neuroendocrine and immune systems. In this context, the thymus-hypothalamus/pituitary axis can be regarded as a paradigm to analyze the connectivity between these two systems in both normal and pathological conditions.

Cytokines and Thymic Hormones Modulate Hypothalamic-Pituitary Functions

127

Table 1. Cytokine effects upon

hypothalamus-pituitary hormone productiona

Cytokine

ACTH

PRL

GH

TSH

FSH

LH

CRH

IL-1 IL-2 IL-6 IFN-Á TNF-· LIF

NO or ; ; ; ' ; or ' ;

NO or ' ; ; ; or ' ; ND

NO or ; ' ; ' ; ND

; ; ND ND ; ND

NO ' ; ND ND ND

NO or ; ' ; ND ND ND

; ; ND ND ; ND

a Increase (;), decrease (') or no effect (NO) on hormone production. ND = Not determined.

that neonatal thymectomy promotes a decrease in the number of secretory granules in acidophilic cells of the adenopituitary [54]. In the same vein, athymic nude mice exhibit significantly low levels of various pituitary hormones, including PRL, GH and the gonadotropins LH and FSH [55]. Concerning thymic peptides, it was shown that thymosin-ß4, when perfused intraventricularly, stimulates LH and its hypothalamic releasing hormone LHRH [54]. A similar effect of LH release stimulation was obtained with thymulin, in perifused or fragmented pituitary preparations [56, 57]. Another thymosin component, the MB-35 peptide, enhances PRL and GH production [58]. In vivo studies in children showed that administration of thymopoietin (a further chemically-defined thymic hormone) increases GH and cortisol serum levels. Moreover, thymopentin (the synthetic biologically active peptide of thymopoietin) enhances in vitro the production of POMC derivatives such as ACTH, ß-endorphin and ·-lipotropin [59]. Lastly, thymulin exhibits an in vitro stimulatory effect on perifused rat pituitaries, enhancing GH, PRL, and to a lesser extent, TSH and LH release [60]. Moreover, using short-term cultures of pituitary fragments, a consistent increase in ACTH secretion was seen after in vitro thymulin treatment, with no changes in GH levels and a significant inhibition of PRL release [57]. More recently, a further thymosin peptide was isolated, which stimulates IL-6 release from rat glioma cells [61]. By contrast, thymosin-·1 is apparently able to downregulate TSH, ACTH and PRL secretion in vivo, although changes on GH levels were not detected [62]. These inhibitory effects seem to occur through hypothalamic pathways, since production of corresponding releasing hormones by hypothalamic neurons was also decreased following in vitro treatment of medial basal hypothalamic fragments with thymosin-·1 [63].

128

Neuroimmunomodulation 1999;6:126–136

Altogether, these findings point to a complex circuitry involving the action of distinct thymic peptides on the hypothalamus-pituitary axis.

Hypothalamic-Pituitary Control of Thymus Physiology

Studies on neuroendocrine control of the immune function have included the distinct levels of immune system organization, involving primary and secondary lymphoid organs, as well as sites of effector immunological activities. The thymus gland is one of the primary lymphoid organs and has been well studied in the recent years. Within this compartment, bone marrow-derived Tcell precursors undergo a complex process of maturation that includes selection of the T-cell repertoire, with positively selected cells eventually migrating to the T-dependent areas of peripheral lymphoid organs, where they further expand [64, 65]. Intrathymic T-cell differentiation is driven by interactions with the thymic microenvironment, a tridimensional network composed of various cell types including epithelial cells, dendritic cells and macrophages, as well as extracellular matrix (ECM) [66, 67]. The thymic microenvironment controls thymocyte migration and differentiation in distinct ways, including secretion of a variety of polypeptides such as thymic hormones and cytokines [68] and cell-cell contacts, such as the interactions occurring through classical adhesion molecules [69], and those involving peptide-bound major histocompatibility complex (MHC) class I and class II proteins expressed by thymic microenvironmental cells, interacting with the T-cell receptor in the context of CD8 or CD4 molecules respectively. Additionally, microenvironmental cells bind to and interact with maturing thymocytes via ECM ligands and receptors [67]. In fact, the intra-

Savino/Arzt/Dardenne

thymic ECM network may function as a substrate onto which thymocytes migrate in ordered fashion [70]. Lastly, gap junctions allow direct communication between adjacent thymic epithelial cells (TEC), and probably between TEC and thymocytes [71], possibly corresponding to a further pathway for cell-cell communication in thymus physiology [72]. One biological activity of TEC which is under neuroendocrine control is secretion of thymic hormones. Different laboratories have definitely demonstrated that secretion of thymulin, a zinc-containing nonapeptide produced exclusively by TEC [73, 74] is modulated by various pituitary hormones. Both in vivo and in vitro treatments with PRL upregulate thymulin secretion [75], an effect that can be obtained even in aging mice, normally having low levels of the circulating thymic hormone [76]. In the same vein, patients with prolactinomas exhibit high thymulin circulating levels, compared to normal agematched controls [77]. Conversely, administration of bromocriptine (an agonist of the dopamine receptor, widely used as an inhibitor of PRL synthesis by adenopituitary cells) decreases thymulin serum levels in mice, an effect which is reversed after exogenous administration of PRL [75]. Thymulin secretion is also enhanced by GH in various mammalian species including mice, rats, dogs and humans [78]. In GH-related pituitary hyperfunction such as acromegaly, the abnormally high levels of circulating thymulin are decreased after appropriate therapy [79]. Conversely, GH deficiency in children is accompanied by low thymulin levels, whereas GH treatment consistently restores this thymic endocrine function [80]. In keeping with these observations, treatment of murine and human TEC cultures with GH enhances the thymulin content of culture supernatants [79]. Lastly, exogenous GH enhances thymulin production in old animals [81]. It should be pointed out that the control of thymulin secretion by GH appears to be mediated by insulin-like growth factor 1 (IGF-1), since GH-induced enhancing of thymulin production in vitro can be prevented when TEC cultures are subjected to antibodies specific for IGF-1 or IGF-1 receptor [79]. Additionally, IGF-1 alone stimulates thymulin production by cultured TEC, and there is a clear-cut positive correlation between serum levels of thymulin and IGF-1 in patients with acromegaly [77]. An important concept regarding the pituitary control of TEC physiology is the pleiotropic nature of the effects. For example, PRL upregulates the expression of highmolecular-weight cytokeratins by medullary TEC [75]. Epithelial growth is also increased in vitro following PRL and GH treatments [75, 79] and ECM ligands and recep-

tors were shown to be enhanced by these pituitary hormones [82]. A further aspect of intrathymic T-cell differentiation concerns direct cell-cell interactions between thymocytes and thymic microenvironmental cells. We recently demonstrated that adhesion of thymocytes to cultured TEC can be enhanced by treating the latter with PRL, GH or IGF-1 [82]. Again, the effects of GH in this system could be abrogated by anti-IGF-1 or anti-IGF-1 receptor antibodies. We observed that these pituitary hormones enhance thymocyte release by cultured thymic nurse cells, a lymphoepithelial complex that partially supports thymocyte differentiation [83]. In further studies, it was shown that gap junction mediated cell coupling among TEC is also under neuroendocrine control, being upregulated by ACTH and VIP [68, 84]. Since hypothalamic/pituitary hormones affect functions of microenvironmental cells related to thymocyte differentiation, it is apparent that the latter process is also under neuroendocrine control. However, besides indirect influence mediated by the thymic microenvironment, direct effects have been reported. For example, synthetic TRH enhances bromodeoxyuridine uptake by thymic cell suspensions [86], an effect apparently shared by PRL, likely being mediated by enhancement of IL-2 production and IL-2 receptor expression [87]. In a second vein, GH was shown to be co-mitogenic for thymocyte proliferation [88, 89]. A series of in vivo experiments also evidenced that important changes in thymocyte differentiation occur under neuroendocrine influence. GH injections in aging mice increased total thymocyte numbers and the percentage of CD3-bearing cells [90], in keeping with our data showing an enhanced concanavalin-A mitogenic response as well as IL-6 production by thymocytes from GHtreated animals [81]. Similar findings were observed in animals treated with IGF-1 [91]. IGF-1 was also able to induce repopulation of the atrophic thymus from diabetic rats [92]. Moreover, mouse substrains selected for bearing high or low IGF-1 circulating levels exhibited differential thymus developmental patterns that positively correlated with IGF-1 levels [93]. The role of GH in thymus development is also shown by findings in GH-deficient dwarf mice. Besides the precocious decline in thymulin serum values [94], there is progressive thymic hypoplasia with decreased numbers of CD4/8 double-positive thymocytes. Such defects are largely restored by long-term treatment with GH [95, 96].

Immunoneuroendocrine Connectivity

Neuroimmunomodulation 1999;6:126–136

129

Thymus-Pituitary Similarities in Cytokine and Hormone Production

The thymus-pituitary cross-talk can be demonstrated by the common production of several soluble mediators. One site of origin of cytokines is the pituitary gland, but they are also produced intrathymically. IL-1 immunoreactive material and respective mRNA, found in rat pituitaries, increased after in vivo treatment with bacterial lipopolysaccharide [97]. In keeping with these findings, IL-1ß mRNA expression was detected by RT-PCR in a series of pituitary adenomas cultured in vitro [98]. Production of IL-6 by cells from rat anterior pituitary glands and expression of corresponding mRNA were also demonstrated [99–101]. They were stimulated, along with other cytokines, by IL-1 [102, 103]. Concerning the human anterior pituitary, IL-6 mRNA expression was detected in corticotrophic adenoma cell cultures [104] and IL-6 was found to be secreted from 7 out of 10 pituitary tumors cultured in vitro [105]. By immunocytochemistry, the presence of IL-6 was demonstrated in almost all pituitary adenomas tested [106]. The expression of IL-2 and its receptor (IL-2R) by pituitary cells of different species was also described [26, 27, 107]. In the mouse AtT20 pituitary tumor cell line, we detected IL-2 mRNA expression after stimulation with CRH or phorbol myristate acetate (PMA). In human corticotrophic adenoma cells, basal IL-2 mRNA expression as well as IL-2 secretion were further stimulated by PMA [107]. Other cytokines are also expressed in the pituitary gland. TNF-· gene expression has been demonstrated by RT-PCR in pituitary adenoma tissue and culture [108]. LIF has been shown to be secreted by bovine pituitary follicular cells in culture [109]. LIF protein and respective mRNA, as well as LIF binding sites, have been demonstrated in developing human fetal pituitary and in normal and adenomatous adult human tissue [110], as well as in mouse hypothalamus [111]. Macrophage-migration inhibitory factor (MIF), which plays a central role in the response to endotoxemia, is also expressed in the pituitary and this expression increases after LPS treatment [112]. Immunocytochemical studies performed with nontumorous human pituitaries revealed that MIF is consistently found in corticotrophs, whereas neurohypophysis was negative. Additionally, corticotroph adenomas were MIFimmunoreactive, but not somatotroph or lactotroph tumors [113]. Moreover, MIF was also detected by immunocytochemistry in tanycytes in the basomedial hypothalamus [114]. Recently, the expression of transforming

130

Neuroimmunomodulation 1999;6:126–136

growth factor-ß (TGF-ß) as well as the TGF-ß type II receptor, has been defined by various methods in both normal and neoplastic human pituitaries [115]. Intrathymic production of typical pituitary hormones has also been demonstrated by several groups. Specific immunoreactivity for PRL, GH, TSH, ACTH, FSH and LH, as well as oxytocin and vasopressin (together with respective neurophysins) were detected in thymic cells [reviewed in 116]. At least regarding GH, PRL, LH, oxytocin, vasopressin and somatostatin, specific messenger RNAs were also detected in the thymus [117–122]. In addition, typical hypothalamic releasing hormones, namely CRH and LHRH, as well as neuropeptides such as ß-endorphin, VIP and substance Y, were evidenced in the organ [123–126]. It should be pointed out that, similar to what is found in the hypothalamus-pituitary axis, there is a certain degree of cell type specificity regarding the production of different hormones. For example, oxytocin and vasopressin appear to be exclusively produced by TEC [127], whereas PRL expression is apparently restricted to thymocytes. GH, however can be produced by both thymocytes and epithelial cells [85]. Altogether, the data discussed above clearly show that the thymus and the hypothalamus/pituitary axis share common secretory products, as depicted in figure 1.

How Can Two Systems Be Endocrinally and Paracrinally Controlled by Similar Molecules?

Cytokines have been shown not only to act as lymphocyte-derived messengers but also to constitute autocrine and paracrine factors in the regulation of pituitary hormone secretion and pituitary cell growth. Similarly, hormones such as PRL and GH not only act as endocrine messengers from the pituitary gland, but also appear to represent autocrine and paracrine factors involved in the general regulation of the thymus. The molecular basis for these actions is provided by two complementary lines of evidence: (a) as discussed above, various cytokines and pituitary hormones are produced in both organs, and (b) receptors for these molecules are also expressed in distinct cells types in the pituitary and in the thymus. For example, IL-1 receptors and respective mRNA were characterized in mouse pituitary cells and AtT20 corticotrophs [128, 129]. Furthermore, IL-6 receptors are expressed in rat anterior pituitary cells [130].

Savino/Arzt/Dardenne

Fig. 1. Similarities between the thymus (left

side) and the H-P axis (right side) in cytokine and hormone production. This concept is illustrated by the various cytokines that can be produced both intrathymically and by cells of the hypothalamus/pituitary axis. Conversely, hypothalamic neurohormones and classic adenopituitary hormones can also be produced by thymic cells.

We also found detectable amounts of IL-2 and expression of the receptor on cell membranes in both adenoma and AtT20 cells [107]. A related protein was detected in AtT20 cells [45]. IL-2 receptor · chain is expressed in cultures of normal rat pituitary cells including ACTH, GH and PRL producing cells [26, 27]. Also, the IL-2 receptor ß chain has been detected and cloned from AtT20 cells [131]. The presence of the receptor in these cells is in agreement with previous studies showing the action of IL2 on the secretion of these hormones [44, 46]. Using a binding assay and a molecular biology approach, a TNF receptor has also been shown in the pituitary gland cells, such as the AtT20 corticotrophs and TtT/GF folliculostellate cells [132]. Intrathymic expression of pituitary hormone receptors has also been demonstrated by distinct approaches (summarized in table 2). Regarding PRL receptors, their presence was evidenced in human and murine TEC, by immunocytochemistry, immunoblotting and Northern blot [133], and PRL receptor gene was detected by PCR in thymocyte-derived extracts. We further demonstrated by tricolor cytofluorometry that PRL receptors are expressed in the various CD4-CD8 defined thymocyte subsets in both mice and humans [134, 135]. Interestingly, mitogenic stimulation by concanavalin A resulted in enhancement of PRL receptor expression in thymocytes, as ascertained by flow cytometry [134]. In addition to PRL receptor, GH binding sites were initially detected in murine and human TEC [136], and then by immunocytochemistry, in situ hybridization and

Immunoneuroendocrine Connectivity

Table 2. Intrathymic production of pituitary and hypothalamic hor-

mones, and expression of respective receptorsa, b Hormonec

GH PRL ACTH LH Oxytocin Vasopressin CRH LHRH

Hormone productionb

Receptor expressionb

thymocyte

TEC

thymocyte

TEC

+ + + + – – + +

+ – + ND + + + +

+ + + ND + + + +

+ + + ND + + + ND

a

Expression determined by techniques including ligand binding, immunocytochemistry, peptide sequencing, immunoblotting, northern blotting and/or PCR. b + = Positive; – = negative; ND = not determined. c GH = Growth hormone; PRL = prolactin; ACTH = corticotropin; LH = luteinizing hormone; CRH = corticotropin-releasing hormone; LHRH = gonadrotopin-releasing hormone.

PCR techniques [137, 138]. Very recently, we noted that only immature CD4–CD8– human thymocytes express GH receptor, suggesting that this differentiation stage is the main target for GH action [85]. The data discussed above should now be considered in the following context. The three criteria used to establish an autocrine or paracrine role for a substance are followed

Neuroimmunomodulation 1999;6:126–136

131

by cytokines at the level of the anterior pituitary and hypothalamus, and by hormones at the thymus level: (1) the cytokine (or the hormone) that modulates some aspect in the physiology of the hypothalamus-pituitary (H-P) axis, also modulates some aspect of thymus physiology; (2) the cytokine is produced in the H-P axis or the hormone is produced at the thymus, and (3) receptors for the cytokine or the hormone are expressed in the H-P axis or the thymus respectively. For example, IL-1, IL-2, IL-6 and LIF, as well as their respective receptors, are expressed in the pituitary, and are able to influence the growth and function of pituitary cells. Reciprocally, GH and PRL and their receptors are expressed in the thymus and these hormones regulate the physiology of this gland. By consequence, all these molecules fulfill the criteria for autocrine or paracrine regulators of pituitary and thymus functions.

The Thymus-Hypothalamus/Pituitary Connectivity under Pathological Situations

It is well recognized that the HPA axis is an essential component of the stress response generated by a variety of stimuli, including infectious agents. CRH acts as a key mediator of the stress response, and stress has been shown to be almost always associated with elevated levels of circulating glucocorticoids [139]. Moreover, the HPA axis is essential in the coupling of the stress response to the immune system, including those elicited by infections. CRH, ACTH and glucocorticoids affect lymphocyte activation and other immune functions, also regulating the expression and synthesis of cytokines [reviewed in 4, 5]. PRL, which is also elevated during stress, is essential for the T-cell response, acting as a co-mitogen of IL-2 [140]. Since the H-P axis and the thymus share common molecules, several questions arise concerning how their connectivity can be influenced not only in normal but also in pathological conditions. During the stress response some pituitary hormones like ACTH and PRL are elevated. In what way does this elevation influence the action and expression of the hormones in the thymus? Reciprocally, during stress, and as a consequence of glucocorticoid elevation, cytokines produced by lymphocytes are inhibited. How does this inhibition influence the action and expression of cytokines in the pituitary? In view of the multiple levels of connectivity between the thymus and the H-P axis, we anticipate that a corresponding response should occur during adaptation of the

132

Neuroimmunomodulation 1999;6:126–136

immunoneuroendocrine circuitry to any imbalance in homeostasis. Acute infectious diseases can be relevant examples to study this question. In particular, we previously showed that in acute experimental Chagas’ disease (a parasitic affection caused by the flagellate protozoan Trypanosoma cruzi), mice exhibit a progressive increase in corticosterone levels [141]. Moreover, a cortical thymocyte depletion occurs, in the context of an intrathymic infection by the parasite [142]. Interestingly, however, thymocyte depletion was still seen in adrenalectomized infected animals [141]. One important point is to evaluate in T. cruziinfected mice the recently reported intrathymic production of corticosterone [143, 144], in order to seek whether a putative ACTH-thymic corticosterone axis could be involved. An altered HPA axis has been reported in severe human falciparum malaria as well as in murine rabies [145, 146]. Interestingly, in experimental rabies virus infection, a rise in IL-1ß and TNF-· was detected in various hypothalamic regions including the supraoptic and paraventricular nuclei [147]. In a second vein, it was shown that the HIV-derived gp120 protein, when chronically microinfused intracerebroventricularly in rats, resulted in an increase in the production of cytokines by hypothalamic cells, including IL-1, TNF-· and TGF-ß, thus suggesting that an imbalance in the H-P cytokine network may occur in AIDS [148]. It is noteworthy that, in parallel with the H-P changes in these three infectious diseases, thymic atrophy was consistent [149, 150]. Another open question concerns hypothetical in vivo infection of the H-P axis, that may generate abnormal patterns of hormone secretion with consequences in the thymus. Also, it will be important to define whether the hypothalamus and hypophysis are targets for an autoimmune process in Chagas’ disease and AIDS, similar to what has been demonstrated for thymocytes and TEC [142, 150, 151]. In conclusion, a better understanding of the thymushypothalamus/pituitary axis in distinct pathological situations, particularly during infectious diseases, will be necessary for designing neuroendocrine-based therapy in such affections. However, much effort has to be made before such procedures become routine clinical practice.

Acknowledgments This work was partially funded with grants from PRONEX and CNPq (Brazil), the University of Buenos Aires (Argentina) and CNRS (France). The authors are indebted to Mrs. Martine Netter for the computer drawing.

Savino/Arzt/Dardenne

References 1 Blalock JE: The syntax of immune-neuroendocrine communication. Immunol Today 1994; 15:504–511. 2 Savino W, Dardenne M: Immunoneuroendocrine interactions. Immunol Today 1995;7: 318–322. 3 Dardenne M, Savino W: Interdependence of the endocrine and immune systems. Adv Neuroimmunol 1996;6:297–307. 4 Besedovsky H, Del Rey A: Immune-neuroendocrine interactions: Facts and hypotheses. Endocr Rev 1996;17:64–102. 5 Bateman A, Singh A, Kral T, Solomon S: The immune-hypothalamic-pituitary-adrenal axis. Endocr Rev 1989;10:92–112. 6 Hermus ARMM, Sweep CGJ: Cytokines and the hypothalamic-pituitary-adrenal axis. J Steroid Biochem Mol Biol 1990;37:867–871. 7 Rivier C, Vale W, Brown M: In the rat, interleukin-1· and -ß stimulate adrenocorticotropin and catecholamine release. Endocrinology 1989;125:3096–3102. 8 Matta SG, Linne KM, Sharp BM: Interleukin1· and interleukin-1ß stimulate adrenocorticotropin secretion in the rat through a similar hypothalamic receptor(s): Effects of interleukin-1 receptor antagonist protein. Neuroendocrinology 1993;57:14–22. 9 Uehara A, Gillis S, Arimura A: Effects of interleukin-1 on hormone release from normal rat pituitary cells in primary culture. Neuroendocrinology 1987;45:343–347. 10 Bernton EW, Beach JE, Holaday JW, Smallridge RC, Fein HG: Release of multiple hormones by a direct action of interleukin-1 on pituitary cells. Science 1987;238:519–521. 11 Florio T, Meucci O, Landolfi E, Grimaldi M, Ventra C, Scorziello A, Marino A, Schettini G: Interleukin-1 modulation of anterior pituitary function: Effect on hormone release and second messenger systems. Pharmacol Res 1989;21 (suppl 1):35–36. 12 Kehrer P, Turnill D, Dayer JM, Muller AF, Gaillard RC: Human recombinant interleukinbeta and -alpha, but not recombinant tumor necrosis factor-alpha stimulate ACTH release from rat anterior pituitary cells in vitro in a prostaglandin E2 cAMP independent manner. Neuroendocrinology 1988;48:160–166. 13 Cambronero JC, Rivas FJ, Borrel J, Guaza C, Interleukin-1· induces pituitary adrenocorticotropin secretion: Evidence for glucocorticoid modulation. Neuroendocrinology 1992;55: 648–654. 14 Beach JE, Smallridge RC, Kinzer CA, Bernton EW, Holaday JW, Fein HG: Rapid release of multiple hormones from rat pituitaries perifused with recombinant interleukin-1. Life Sci 1989;44:1–7. 15 Parsadaniantz SM, Lenoir V, Terlain B, Kerdelhue B: Lack of effect of interleukin-1· and ß, during in vitro perfusion, on anterior pituitary release of adrenocorticotropin hormone and ß-endorphin in the male rat. J Neurosci Res 1993;34:315–323.1

Immunoneuroendocrine Connectivity

16 Renner U, Newton CJ, Pagotto U, Sauer J, Artz E, Stalla GK: Involvement of interleukin1 and interleukin-1 receptor antagonist in rat pituitary cell growth regulation. Endocrinology 1995;136:3186–3193. 17 Malarkey WB, Zvara BJ: Interleukin-1· and other cytokines stimulate adrenocorticotropin release from cultured pituitary cells of patients with Cushing’s disease. J Clin Endocrinol Metab 1989;69:196–199. 18 Fukata J, Usui T, Naitoh Y, Nakai Y, Imura H: Effects of recombinant human interleukin-1·, -1ß, -2 and -6 on ACTH synthesis and release in the mouse pituitary tumor cell line AtT-20. J Endocrinol 1989;122:33–39. 19 Woloski BMRNJ, Smith EM, Meyer WJ, Fuller GM, Blalock JE: Corticotropin-releasing activity of monokines. Science 1985;230:1035– 1037. 20 Gwosdow AR, Spencer JA, O’Connell NA, Abou-Samra AB: Interleukin-1 activates protein kinase A and stimulates adrenocorticotropin hormone release from AtT20 cells. Endocrinology 1993;132:710–714. 21 Payne LC, Weigent DA, Blalock JE: Induction of pituitary sensitivity to interleukin-1: A new function for corticotropin-releasing hormone. Biochem Biophys Res Commun 1994;198: 480–484. 22 Spangelo BL, Judd AM, Isakson PC, MacLeod RM: Interleukin-6 stimulates anterior pituitary hormone release in vitro. Endocrinology 1989; 125:575–577. 23 Lyson K, McCann SM: The effect of interleukin-6 on pituitary hormone release in vivo and in vitro. Neuroendocrinology 1991;54:262– 266. 24 Yamaguchi M, Matsuzaki N, Hirota K, Miyake A, Tanizawa O: Interleukin-6 possibly induced by interleukin-1ß in the pituitary gland stimulates the release of gonadotropins and prolactin. Acta Endocrinol 1990;122:201–205. 25 Grimaldi M, Meucci O, Scorziello A, Florio T, Ventra C, De Mercato R: Interleukin-6 modulation of second messenger systems in anterior pituitary cells. Life Sci 1992;51:1243–1248. 26 Arzt E, Buric R, Stelzer G, Stalla J, Sauer J, Renner U, Stalla GK: Interleukin involvement in anterior pituitary cell growth regulation: Effects of IL-2 and IL-6. Endocrinology 1993; 132:459–467. 27 Arzt E, Sauer J, Buric R, Stalla J, Renner U, Stalla GK: Characterization of IL-2 receptor expression and action of IL-2 and IL-6 on normal anterior pituitary cell growth. Endocrine 1995;3:113–119. 28 Sawada T, Koike K, Kanda Y, Ikegami H, Jikihara T, Maeda T: Interleukin-6 stimulates cell proliferation of rat anterior pituitary clonal cell lines in vitro. J Endocrinol Invest 1995;18:83– 90. 29 Pereda MP, Goldberg V, Chervin A, Carrizo G, Molina A, Andrada J, Sauer J, Renner U, Stalla GK, Arzt E: IL-2 and IL-6 upregulate cfos protooncogene expression in human pituitary adenoma explants. Mol Cell Endocrinol 1996;124:33–42.

30 Callahan TA, Piekut DT: Differential Fos expression induced by IL-1ß and IL-6 in rat hypothalamus and pituitary gland. J Neuroimmunol 1997;73:207–211. 31 Milenkovic L, Rettori V, Snyder GD, Beutler B, McCann SM: Cachectin alters anterior pituitary hormone release by a direct action in vitro. Proc Natl Acad Sci USA 1989;86:2418– 2422. 32 Koike K, Hirota K, Ohmichi M, Kadowaki K, Ikegami H, Yamaguchi M: Tumor necrosis factor-· increases release of arachidonate and prolactin from rat anterior pituitary cells. Endocrinology 1991;128:2791–2798. 33 Gaillard RC, Turnill D, Sappino P, Muller AF: Tumor necrosis factor-· inhibits the hormonal response of the pituitary gland to hypothalamic releasing factors. Endocrinology 1990;127: 101–106. 34 Ray DW, Ren SG, Melmed S: Leukemia inhibitory factor stimulates proopiomelanocortin expression in a corticotroph cell line. J Clin Invest 1996;97:1852–1859. 35 Stefana B, Ray DW, Melmed S: Leukemia inhibitory factor induces differentiation of pituitary corticotroph function: An immuno-neuroendocrine phenotypic switch. Proc Natl Acad Sci USA 1996;93:12502–12506. 36 Akita S, Readhead C, Stefaneanu L, Fine J, Tampanaru-Sarmesiu A, Kovacs K, Melmed S: Pituitary-directed leukemia inhibitory factor transgene forms Rathke’s cleft cysts and impairs adult pituitary function. A model for human pituitary Rathke’s cysts. J Clin Invest 1997;99:2462–2469. 37 Lotze MT, Frana LW, Sharrow SO, Robb RJ, Rosenberg AS: In vivo administration of purified human interleukin-2. I. Half-life and immunologic effects of the Jurkat cell line-derived IL-2. J Immunol 1985;134:157–166. 38 Denicoff KD, Durkin TM, Lotze MT, Quinlan PE, Davis CL, Listwak SJ, Rosenberg SA, Rubinow DR: The neuroendocrine effects of interleukin-2 treatment. J Clin Endocrinol Metab 1989;69:402–410. 39 Naito Y, Fukata J, Tominaga T, Masui Y, Hirai Y, Murakami N, Tamai S, Mori K, Imura H: Adrenocorticotropic hormone-releasing activities of interleukins in a homologous in vivo system. Biochem Biophys Res Commun 1989; 164:1262–1267. 40 Tominaga T, Fukata J, Naito Y, Usui T, Murakami N, Fukushima M, Nakai Y, Hirai Y, Imura H: Prostaglandin-dependent in vitro stimulation of adrenocortical steroidogenesis by interleukins. Endocrinology 1991;128:526–531. 41 Cambronero JC, Rivas FJ, Borrell J, Guaza C: Interleukin-2 induces corticotropin-releasing hormone release from superfused rat hypothalami: Influence of glucocorticoids. Endocrinology 1992;131:677–683. 42 Brown SL, Smith LR, Blalock JE: Interleukin-1 and interleukin-2 enhance proopiomelanocortin gene expression in pituitary cells. J Immunol 1987;139:3181–3183.

Neuroimmunomodulation 1999;6:126–136

133

43 Harbuz MS, Stephanou A, Knight RA, ChoverGonzalez AJ, Lightman SL: Action of interleukin-2 and interleukin-4 on CRF mRNA in the hypothalamus and POMC mRNA in the anterior pituitary. Brain Behav Immun 1992;6: 214–222. 44 Karanth S, McCann SM: Anterior pituitary hormone control by interleukin-2. Proc Natl Acad Sci USA 1991;88:2961–2965. 45 Smith LR, Brown SL, Blalock JE: Interleukin-2 induction of ACTH secretion: Presence of an interleukin-2 receptor ·-chain-like molecule on pituitary cells. J Neuroimmunol 1989;21:249– 254. 46 Karanth S: The influence of dopamine on interleukin-2 induced release of prolactin, luteinizing hormone and follicle-stimulating hormone by the anterior pituitary (abstract). Program and Abstracts of the 73rd Annual Meeting of The Endocrine Society, Washington, DC, 1991, p 210. 47 Karanth S, McCann SM: Influence of dopamine on the altered release of prolactin, luteinizing hormone and follicle-stimulating hormone induced by interleukin-2 in vitro. Neuroendocrinology 1992;56:871–880. 48 Newton CJ, Arzt E, Stalla GK: Involvement of the estrogen receptor in the growth response of pituitary cells to interleukin-2. Biochem Biophys Res Commun 1994;205:1930–1937. 49 Yamaguchi M, Koike K, Matsuzaki N, Yoshimoto Y, Taniguchi T, Miyake A: The interferon family stimulates the secretions of prolactin and interleukin-6 by the pituitary gland in vitro. J Endocrinol Invest 1991;14:457–461. 50 Vankelecom H, Andries M, Billiau A, Denef C: Evidence that folliculo-stellate cells mediate the inhibitory effect of interferon-Á on hormone secretion in rat anterior pituitary cell cultures. Endocrinology 1992;130:3537–3546. 51 Vankelecom H, Matthys P, Denef C: Involvement of nitric oxide in the interferon-Á-induced inhibition of growth hormone and prolactin secretion in anterior cell cultures. Mol Cell Endocrinol 1997;129:157–167. 52 Vankelecom H, Matthys P, Denef C: Inducible nitric oxide synthase in the anterior pituitary gland: Induction by interferon-Á in a subpopulation of folliculostellate cells and in an unidentifiable population of non-hormone-secreting cells. J Histochem Cytochem 1997;45:847– 857. 53 Kovacs KJ, Foldes A, Vizi ES: C-kit ligand (stem cell factor) affects neuronal activity, stimulates pituitary-adrenal axis and prolactin secretion in rats. J Neuroimmunol 1996;65: 133–141. 54 Goya RG, Sosa YE, Console GM, Dardenne M: Altered thyrotropic and somatotropic responses to environmental challenges in congenitally athymic mice. Brain Behav Immun 1995; 9:79–86. 55 Daneva T, Spinedi E, Hadid R, Gaillard R: Impaired hypothalamo-pituitary-adrenal axis function in Swiss nude athymic mice. Neuroendocrinology 1995;62:79–87.

134

56 Zaidi SA, Kendall MD, Gillham B, Jones MT: The release of luteinizing hormone from pituitaries perifused with thymic extracts. Thymus 1988;12:253–264. 57 Hadley AJ, Rantle CM, Buckinham JC: Thymulin stimulates corticotropin release and cyclic nucleotide formation in the rat anterior pituitary gland. Neuroimmunomodulation 1997;4:62–69. 58 Badamchiam M, Spangelo BL, Damavandy T, Mac Leod RM, Goldstein AL: Complete amino acid sequence of a peptide isolated from the thymus that enhances release of growth hormone and prolactin. Endocrinology 1991;128: 1580–1588. 59 Malaise MG, Hazee-Hagelstein MT, Reuter AM, Vrinds-Gevaert Y, Goldstein G, Franchimont P: Thymopoietin and thymopentin enhance the levels of ACTH, beta-endorphin and beta-lipotropin from rat pituitary cells in vitro Acta Endocrinol 1987;115:455–462. 60 Goya RG, Sosa YE, Brown OA, Dardenne M: In vitro studies on the thymus-pituitary axis in young and old rats. Ann NY Acad Sci 1994; 741:108–114. 61 Tijerina M, Gorospe WC, Bowman KL, Badamchian M, Goldstein AL, Spangelo BL: A novel thymosin peptide stimulates interleukin6 release from rat C6 glioma cell in vitro. Neuroimmunomodulation 1997;4:163–170. 62 Milenkovic L, McCann SM: Effects of thymosin-alpha-1 on pituitary hormone release. Neuroendocrinology 1992;55:14–19. 63 Milenkovic L, Lyson K, Aguila MC, McCann SM: Effect of thymosin-alpha-1 on hypothalamic hormone release. Neuroendocrinology 1992;56:674–679. 64 Van Ewijk W: T-cell differentiation is influenced by thymic microenvironments. Annu Rev Immunol 1991;9:591–615. 65 Anderson G, Moore NC, Owen JJT, Jenkinson EJ: Cellular interactions in thymocyte development. Annu Rev Immunol 1996;14:73–99. 66 Boyd RL, Tucek CL, Godfrey DI, Izon DJ, Wilson TJ, Davidson NJ, Bean AGD, Ladyman HM, Ritter MA, Hugo P: The thymic microenvironment. Immunol Today 1993;14:445–459. 67 Savino W, Villa-Verde DMS, Lannes-Vieira J: Extracellular matrix proteins in intrathymic T cell migration and differentiation? Immunol Today 1993;14:158–161. 68 Savino W, Villa-Verde DMS, Alves LA, Dardenne M: Neuroendocrine control of the thymus. Ann NY Acad Sci 1998;840:470–479. 69 Patel DD, Haynes BF: Cell adhesion molecules involved in intrathymic T cell development. Semin Immunol 1993;5:283–292. 70 Savino W, Dardenne M, Carnaud C: Conveyor belt model for intrathymic cell migration. Immunol Today 1996;27:97–98. 71 Alves LA, Campos de Carvalho AC, CirneLima EO, Rocha e Souza CM, Dardenne M, Spray DC, Savino W: Functional gap junctions in thymic epithelial cells are formed by connexin 43. Eur J Immunol 1995;25:431–437. 72 Alves LA, Campos de Carvalho AC, Savino W: Gap junctions: A novel route for cell-cell communication in the immune system? Immunol Today 1998;19:269–275.

Neuroimmunomodulation 1999;6:126–136

73 Savino W, Dardenne M, Papiernik M, Bach JF: Thymic hormone containing cells. Characterization and localization of serum thymic factor in young mouse thymus studied by monoclonal antibodies. J Exp Med 1982;156:628–632. 74 Dardenne M, Savino W, Berrih S, Bach JF: Evidence for a zinc-dependent epitope on the molecule of thymulin, a thymic hormone. Proc Natl Acad Sci USA 1985;82:7035–7039. 75 Dardenne M, Savino W, Gagnerault MC, Itoh T, Bach JF: Neuroendocrine control of thymic hormonal production. I. Prolactin stimulates in vivo and in vitro the production of thymulin by human and murine thymic epithelial cells. Endocrinology 1989;125:3–12. 76 Savino W, Dardenne M, Bach JF: Thymic hormone containing cells. II. Evolution of cells containing the serum thymic factor (FTS or thymulin) in normal and autoimmune mice, as revealed by anti-FTS monoclonal antibodies. Relationship with Ia-bearing cells. Clin Exp Immunol 1983;52:1–7. 77 Timsit J, Safieh B, Gagnerault MC, Savino W, Lubetzki J, Bach JF, Dardenne M: Augmentation des taux circulants de thymuline au cours de l’hyperprolactinémie et de l’acromégalie. CR Acad Sci Paris 1989;310:7–10. 78 Mello-Coelho V, Savino W, Postel-Vinay MC, Dardenne M: Role of prolactin and growth hormone on thymus physiology. Dev Immunol 1998;6:317–323. 79 Timsit J, Savino W, Safieh B, Chanson P, Gagnerault MC, Bach JF, Dardenne M: Effects of growth hormone and insulin-like growth factor 1 in thymic hormonal function in man. J Clin Endocrinol Metabolism 1992;75:1251–1260. 80 Mocchegiani E, Fabris N, Travaglini P, Sartorio A, De Min C, Paolucci P: Thymic endocrine activity in children with idiopathic growth-hormone deficiency. Int J Neurosci 1991;59:151– 157. 81 Goya RG, Gagnerault MC, Leite de Moraes MC, Savino W, Dardenne M: In vivo effects of growth hormone on thymus function in aging mice. Brain Behav Immun 1992;6:341–354. 82 Mello-Coelho V, Villa-Verde DMS, Dardenne M, Savino W: Pituitary hormones modulate cell-cell interactions between thymocytes and thymic epithelial cells. J Neuroimmunol 1997; 76;39–49. 83 Villa Verde DMS, Mello-Coelho V, LagrotaCandido JM, Savino W: The thymic nurse cell complex: An in vitro model for extracellular matrix-mediated intrathymic T cell migration. Braz J Med Biol Res 1995;28:907–916. 84 Head GM, Mentlein R, Krans A, Downing JE, Kendall MD: Modulation of dye-coupling and proliferation of cultured thymic rat epithelium by factors involved in thymulin seretion. J Anat 1997;191:355–365. 85 Mello-Coelho V, Gagnerault MC, Souberbielle JC, Strasburger CJ, Savino W, Dardenne M, Postel-Vinay MC: Growth hormone and its receptor are expressed in human thymic cells. Endocrinology 1998;139:3837–3842. 86 Pawlikowski M, Zerek-Melen G, Winczyk K: Thyroliberin increases thymus cell proliferation in rats. Neuropeptides 1992;23:199–202.

Savino/Arzt/Dardenne

87 Viselli SM, Stanek EM, Mukherjee P, Hymer WC, Mastro AM: Prolactin-induced mitogenesis of lymphocytes from ovariectomized rats. Endocrinology 1991;129:983–990. 88 Sabharwal P, Varma S: Growth hormone synthesized and secreted by human thymocytes acts via insulin-like growth factor I as an autocrine and paracrine growth factor. J Clin Endocrinol Metab 1996;81:2663–2669. 89 Postel-Vinay MC, Mello-Coelho V, Gagnerault MC, Dardenne M: Growth hormone stimulates the proliferation of activated mouse T lymphocytes. Endocrinology 1997; 138:1816–1820. 90 Li YM, Brunke DL, Dantzer R, Kelley KW: Pituitary epithelial cell implants reverse the accumulation of CD4-CD8-lymphocytes in thymus glands of aged rats. Endocrinology 1992;130:2703–2709. 91 Clark R, Strasser J, McCabe S, Robbins K, Jardieu P: Insulin-like growth factor-I stimulation of lymphopoiesis. J Clin Invest 1993; 92:540–548. 92 Binz K, Joller P, Froesch P, Binz H, Zapf J, Froesch ER: Repopulation of atrophied thymus in diabetic rats by insulin-like growth factor-I. Proc Natl Acad Sci USA 1990;87: 3690–3694. 93 Siddiqui RA, McCutcheon SN, Blair HT, Mackenzie DD, Morel PC, Breier BH, Gluckman PD: Growth allometry of organs, muscles and bones in mice from lines divergently selected on the basis of plasma insulin-like growth factor-I. Growth Dev Aging 1992;56: 53–60. 94 Pelletier M, Montplaisir S, Dardenne M, Bach JF: Thymic hormone activity and spontaneous autoimmunity in dwarf mice and their littermates. Immunology 1976;30:783– 789. 95 Murphy WJ, Durum SK, Longo DL: Role of neuroendocrine hormones in murine T cell development. Growth hormone exerts thymopoietic effects in vivo. J Immunol 1992; 149:3851–3857. 96 Knyszynski A, Adler-Kunin S, Globerson A: Effects of growth hormone on thymocyte development from progenitor cells in the bone marrow. Brain Behav Immun 1992;6:327– 340. 97 Koenig JI, Snow K, Clark BD, Toni R, Cannon JG, Shaw AR, Dinarello CA, Reichlin S, Lee SL, Lechan RM: Intrinsic pituitary IL-1ß is induced by bacterial LPS. Endocrinology 1990;126:3053–3058. 98 Schneider JH, Hofman FM, Weiss MH, Hinton DR: Cytokine expression in pituitary adenomas (abstract). Int Congr on Pituitary Adenomas, Marina del Rey, 1993, MP-10. 99 Vankelecom H, Carmeliet P, Van Damme J, Billiau A, Denef C: Production of IL-6 by folliculo-stellate cells of the anterior pituitary gland in a histiotypic cell aggregate culture system. Neuroendocrinology 1989;49:102– 106. 100 Spangelo BL, MacLeod RM, Isakson PC: Production of interleukin-6 by anterior pituitary cells in vitro. Endocrinology 1990;126:582– 586.

Immunoneuroendocrine Connectivity

101 Spangelo BL, Judd AM, MacLeod RM, Goodman DW, Isakson PC: Endotoxin-induced release of interleukin-6 from rat medial basal hypothalami. Endocrinology 1990;127: 1779–1785. 102 Spangelo BL, Judd AM, Isakson PC, MacLeod RM: Interleukin-1 stimulates interleukin-6 release from rat anterior pituitary cells in vitro. Endocrinology 1991;128:2685– 2691. 103 Yamaguchi M, Matsuzaki N, Hirota K, Miyake A, Tanizawa O: Interleukin-6 possibly induced by interleukin-1ß in the pituitary gland stimulates the release of gonadotropins and prolactin. Acta Endocrinol Copenh 1990; 122:201–205. 104 Velkeniers B, D’Haens G, Smets G, Vergani P, Vanhaelst L, Hooghe-Peters EL: Expression of IL-6 mRNA in corticotroph cell adenomas (abstract). J Endocrinol Invest 1991; 14(suppl1):31. 105 Jones TH, Justice S, Price A, Chapman K: Interleukin-6 secreting human pituitary adenomas in vitro. J Clin Endocrinol Metab 1991;73:207–209. 106 Tsagarakis S, Kontogeorgeos G, Giannou P: Interleukin-6, a growth promoting cytokine, is present in human pituitary adenomas: An immuno-cytochemical study. Clin Endocrinol 1992;37:163–167. 107 Arzt E, Stelzer G, Renner U, Lange M, Müller OA, Stalla GK: Interleukin-2 and IL-2 receptor expression in human corticotrophic adenoma and murine pituitary cell cultures. J Clin Invest 1992;90:1944–1951. 108 Todd VL, Atkin SL, Speirs V, White MC: PCR expression of cytokines in anterior pituitary adenomas. J Endocrinol 1995;144 (suppl):P285–P286. 109 Ferrara N, Winer J, Henzel WJ: Pituitary follicular cells secrete an inhibitor of aortic endothelial cell growth: Identification as leukemia inhibitory factor. Proc Natl Acad Sci USA 1992;89:698–702. 110 Akita S, Webster J, Ren S-G, Takino H, Said J, Zand O, Melmed S: Human and pituitary expression of leukemia inhibitory factor. Novel intrapituitary regulation of adrenocorticotropin hormone synthesis and secretion. J Clin Invest 1995;95:1288–1298. 111 Wang Z, Ren SG, Melmed S: Hypothalamic and pituitary leukemia inhibitory factor gene expression in vivo: A novel endotoxin-inducible neuro-endocrine interface. Endocrinology 1996;137:2947–2953. 112 Bernhagen J, Calandra T, Mitchell RA, Martin SB, Tracey KL, Voelter W, Manogue KR, Cerami A, Bucala R: MIF is a pituitaryderived cytokine that potentiates lethal endotoxaemia. Nature 1993;365:756–759. 113 Tampanaru-Sarmesiu A, Stefaneanu L, Thapar K, Kovacs K, Donnelly T, Metz CN, Bucala R: Immunocytochemical localization of macrophage migration inhibitory factor in human hypophysis and pituitary adenomas. Arch Pathol Lab Med 1997;121:404–410.

114 Nishibori M, Kakaya N, Mori S, Saeki K: Immunohistochemical localization of macrophage migration inhibitory factor in tanycytes, subcommissural organ and choroid plexus in the rat brain. Brain Res 1997;758: 259–262. 115 Jin L, Qian X, Kulig E, Sanno N, Scheithauer BW, Kovacs K, Young WF Jr, Lloyd RV: Transforming growth factor-ß, transforming growth factor-ß receptor II, and p27Kip1 expression in nontumorous and neoplastic human pituitaries. Am J Pathol 1997;151:509– 519. 116 Dardenne M, Savino W: Neuroendocrine control of thymus physiology by peptidic hormones and neuropeptides. Immunol Today 1994;15:518–523. 117 De Leeuw FE, Jansen GH, Batanero E, van Wichen DF, Huber J, Schuurman HJ: The neural and neuroendocrine component of the human thymus. I. Nerve-like structures. Brain Behav Immun 1992;6:234–248. 118 Montgomery DW, Shen GK, Ulrich ED, Steiner LL, Parrish PR, Zukoski CF: Human thymocytes express a prolactin-like messenger ribonucleic acid and synthesize bioactive prolactin-like proteins. Endocrinology 1992; 131:3019–3026. 119 Wu H, Devi R, Malarkey WB: Expression and localization of prolactin messenger ribonucleic acid in the human immune system. Endocrinology 1996;137:349–353. 120 Maggiano N, Piantelli M, Ricci R, Larocca LM, Capelli A, Ranelletti FO: Detection of growth hormone-producing cells in human thymus by immunohistochemistry and nonradioactive in situ hybridization. J Histochem Cytochem 1994;42:1349–1354. 121 Sabharwal P, Varma S: Growth hormone synthesized and secreted by human thymocytes acts via insulin-like growth factor I as an autocrine and paracrine growth factor. J Clin Endocrinol Metab 1996;81:2663–2669. 122 Robert F, Geenen V, Schoenen J, Burgeon E, De Groote D, Defresne MP, Legros JJ, Franchimont P: Colocalization of immunoreactive oxytocin, vasopressin and interleukin–1 in human thymic epithelial neuroendocrine cells. Brain Behav Immun 1991;5:102–115. 123 Jessop DS, Renshaw D, Lightman SL, Harbuz MS: Changes in ACTH and beta-endorphin immunoreactivity in immune tissues during a chronic inflammatory stress are not correlated with changes in corticotropin-releasing hormone and arginine vasopressin. J Neuroimmunol 1995;60:29–35. 124 Gomariz RP, Lorenzo MJ, Cacicedo L, Vicente A, Zapata AG: Demonstration of immunoreactive vasoactive intestinal peptide and somatostatin in rat thymus. Brain Behav Immun 1990;4:151–161. 125 Al-Shawaf AA, Kendall MD, Cowen T: Identification of neural profiles containing vasoactive intestinal polypeptide, acetylcholinesterase and catecholamines in the rat thymus. J Anat 1991;174:131–143.

Neuroimmunomodulation 1999;6:126–136

135

126 Jessop D, Biswas S, D’Souza L, Chowdrey H, Lightman S: Neuropeptide Y immunoreactivity in the spleen and thymus of normal rats and following adjuvant-induced arthritis. Neuropeptides 1992;23:203–207. 127 Moll UM, Lane BL, Robert F, Geenen V, Legros JJ: The neuroendocrine thymus. Abundant occurrence of oxytocin-, vasopressin-, and neurophysin-like peptides in epithelial cells. Histochemistry 1988;89:385–390. 128 De Souza EB, Webster EL, Grigoriadis DE, Tracey DE: Corticotropin-releasing factor and interleukin-1 receptors in the brain-pituitary-immune axis, Psychopharmacol Bull 1989;25:299–305. 129 Bristulf J, Simoncsits A, Bartfai T: Characterization of a neuronal interleukin-1 receptor and the corresponding mRNA in the mouse anterior pituitary cell line AtT-20. Neurosci Lett 1991;128:173–176. 130 Ohmichi M, Hirota K, Koike K, Kurachi H, Ohtsuka S, Matsuzaki NM, Miyake A, Tanizawa O: Binding sites for interleukin-6 in the anterior pituitary gland. Neuroendocrinology 1992;55:199–203. 131 Petitto JM, Huang Z, Rinker CM, McCarthy DB: Isolation of IL-2ß cDNA clones from AtT-20 pituitary cells: Constitutive expression and role in signal transduction. Neuropsychopharmacology 1997;17:57–66. 132 Kobayashi H, Fukata J, Murakami N, Usui T, Ebisui O, Muro S, Hanaoka I, Inoue K, Imura H, Nakao K: Tumor necrosis factor receptors in the pituitary cells. Brain Res 1997;758:45–50. 133 Dardenne M, Kelly PA, Bach JF, Savino W: Identification and functional activity of prolactin receptors in thymic epithelial cells. Proc Natl Acad Sci USA 1991;88:9700– 9704. 134 Gagnerault MC, Touraine P, Savino W, Kelly PA, Dardenne M: Expression of prolactin receptors on murine lymphoid cells in normal and autoimmune conditions. J Immunol 1993;151:1–9.

136

135 Dardenne M, Leite de Moraes MC, Kelly PA, Gagnerault MC: Prolactin receptors expression in human hematopoietic tissues analysed by flow cytofluorometry. Endocrinology 1993;134:2108–2114. 136 Ban E, Gagnerault MC, Jammes H, PostelVinay MC, Haour F, Dardenne M: Specific binding sites for growth hormone in cultured mouse thymic epithelial cells. Life Sci 1991; 48:2141–2148. 137 Gagnerault MC, Postel-Vinay MC, Dardenne M: Expression of growth hormone receptors in murine lymphoid cells analyzed by flow cytofluorometry. Endocrinology 1996;137: 1719–1726. 138 Mertani HC, Delehaye-Zervas MC, Martini JF, Postel-Vinay MC, Morel G: Localization of growth hormone receptor messenger RNA in human tissues. Endocrine 1995;3:135– 142. 139 Reisine T, Affolter H-U, Rougon G, Barbet J: New insights into molecular mechanisms of stress. Trends Neurosci 1986;9:574–576. 140 Clevenger CV, Sillman AL, Hanley-Hyde J, Prystowsky MB: Requirement for prolactin during cell cycle regulated gene expression in cloned T-lymphocytes. Endocrinology 1992; 130:3216–3222. 141 Leite de Moraes MC, Hontebeyrie-Joskowicz M, Leboulanger F, Savino W, Dardenne M, Lepault F: Studies on the thymus in Chagas’ disease. II. Thymocyte subset fluctuations in Trypanosoma cruzi-infected mice: Relationship to stress. Scand J Immunol 1991;33: 267–275. 142 Savino W, Leite de Moraes MC, HontebeyrieJoskowicz M, Dardenne M: Studies on the thymus in Chagas’ disease. I. Changes in the thymic microenvironment in mice acutely infected with Trypanosoma cruzi. Eur J Immunol 1989;19:1727–1733.

Neuroimmunomodulation 1999;6:126–136

143 Vacchio MS, Papadopoulos V, Ashwell JD: Steroid production in the thymus: Implications for thymocyte selection. J Exp Med 1994;179:1835–1846. 144 Vacchio MS, Ashwell JD: Thymus-derived glucocorticoids regulate antigen-specific positive selection. J Exp Med 1997;185:2033– 2038. 145 Davis TM, Tran QB, Robertson K, Dyer JR, Phan TD, Meyer D, Beaman MH, Trinh KA: The hypothalamic-pituitary-adrenocortical axis in severe falciparum malaria: Effects of cytokines. J Clin Endocrinol Metab 1997;82: 3029–3033. 146 Torres-Anjel MJ, Volz D, Torres MJ, Turk M, Tshikuka JG: Failure to thrive, wasting syndrome, and immunodeficiency in rabies: A hypophyseal/hypothalamic/thymic axis effect of rabies infection. Rev Infect Dis 1988; 10(suppl 4):710–725. 147 Marquette C, vam Dam AM, Ceccaldi PE, Weber P, Haour F, Tsiang H: Induction of immunoreactive interleukin-1ß and tumor necrosis factor-· in the brains of rabies infected rats. J Neuroimmunol 1996;68:45–51. 148 Ilyin SE, Plata-Salaman CR: HIV-I Gp120 modulates hypothalamic cytokine mRNAs in vivo: Implications to cytokine feedback systems. Biochem Biophys Res Commun 1997; 231:514–518. 149 Cardenas-Palomo LF, de Souza Matos DC, Chaves-Leal E, Bertho AL, Markovistz R: Lymphocyte subsets and cell proliferation analysis in rabies-infected mice. J Clin Lab Immunol 1995;46:49–61. 150 Savino W, Dardenne M, Marche C, Trophylme D, Dupuis JM, Pekovic D, Bach JF: Thymic epithelium in AIDS: An immunohistologic study. Am J Pathol 1986;122:302– 307. 151 Savino W, Silva JS, Silva-Barbosa SD, Dardenne M, Ribeiro dos Santos R: Anti-thymic cell autoantibodies in human and murine chronic Chagas’ disease. EOS J Immunol Immunopharmacol 1990;10:204–205.

Savino/Arzt/Dardenne

Neuroimmunomodulation 1999;6:137–142

The Thymus-Pituitary Axis and Its Changes during Aging Rodolfo G. Goya Oscar A. Brown Federico Bolognani INIBIOLP-Histology B, Faculty of Medicine, National University of La Plata, Argentina

Key Words Aging W Immunoneuroendocrine network W Thymic hormones W Thymosin W Thymulin W Homeostatic thymus hormone W Pituitary desensitization

Abstract The pituitary-thymic axis constitutes a bidirectional circuit where the ascending feedback loop is effected by thymic factors of epithelial origin. The aim of the present article is to review the evidence demonstrating that aging brings about a progressive disruption in the integration of this network. In doing so, we briefly review the experimental evidence supporting the view that immune and neuroendocrine aging are interdependent processes. The advantages and limits of the nude mouse as a model of thymus-dependent accelerated aging is also discussed. Next, we review a number of studies which show that the endocrine thymus produces several bioactive molecules, generally called thymic hormones, which in addition to possessing immunoregulatory properties are also active on nervous and endocrine circuits. In particular, the reported activities of thymosin fraction 5 (TF5), thymosin ·-1 and thymosin ß-4 on ß-endorphin, ACTH, glucocorticoids, LHRH and LH secretion in different animal and cell models are reviewed. The known hypophysiotropic actions of other thymic hormones like thymulin, homeostatic thymus hormone (HTH) and thymus factor are summarized. Aging has a significant impact on pituitary responsiveness to thymic hormones.

ABC

© 1999 S. Karger AG, Basel 1021–7401/99/0062–0137$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: http://BioMedNet.com/karger

Thus, it has been reported that TF5 and HTH have thyrotropin-inhibiting activity in young but not in old rats. Furthermore, intravenous administration of HTH was also able to reduce plasma GH and increase corticosterone levels in both young and old rats, although these responses were much weaker in the old animals. Further evidence on this topic is discussed. It is proposed that in addition to its central role in the regulation of the immune function, the thymus gland may extend its influence to nonimmunologic components of the body, including the neuroendocrine system. The early onset of thymus involution might therefore act as a triggering event which would initiate the gradual decline in homeostatic potential that characterizes the aging process.

The Immune-Neuroendocrine Network from Development to Aging

It now seems clear that the thymus gland and the neuroendocrine system influence the maturation of each other during early ontogeny and perinatal life in mammals as indicated by the hormonal derangements caused by neonatal thymectomy or congenital athymia in some species (see below). There are also the established findings that in species in which neonatal thymectomy does not produce any evident impairment of the immune capacity [1], neuroendocrine functions are already highly developed at birth [2].

Rodolfo G. Goya, PhD INIBIOLP, Facultad de Medicina UNLP CC 455, La Plata 1900 (Argentina) Tel. +54 21 3 4833/25 6735, Fax +54 21 25 0924/25 8988 E-Mail [email protected]

In adult animals, clear changes in the electric activity of the hypothalamus have been shown to occur during the immune response [3, 4]. The immunologic activity also affects hormone levels. Thus, during the course of the immune response in vivo the serum levels of corticosterone increase [5, 6] whereas circulating thyroxine (T4) and triiodothyronine (T3) decline [6, 7]. These hormonal changes seem to play a role in the modulation of the immune response. Thus, adrenalectomy in rats abolishes the phenomenon of sequential antigenic competition [5] which is defined as the blockade of an immune response against one antigen by the administration of a second, unrelated antigen, a few days before the first one. It is believed that the corticosterone surge during the immune response contributes to prevent an exaggerated nonspecific production of antibodies. Interestingly, by 12 months of age, mice lose the ability to increase their circulating levels of corticosterone during the immune response, an occurrence that is accompanied by a parallel loss of their ability to show sequential antigenic competition [8]. It seems therefore possible that aging may bring about a progressive disruption in immune-endocrine integration which in turn could play a significant role in ageassociated immunopathologies, particularly autoimmunity. Although much attention has been devoted to the role of thymus as a pacemaker of immunologic aging, very little work has been done on the endocrine consequences (and nonimmune consequences, in general) of thymus decline during aging. Particularly significant in this area is the work of Fabris et al. [9]. These authors reported that injection of mature lymph node lymphocytes from normal littermates into Snell-Bag dwarf mice or reconstitution of their lymphoid system with growth hormone (GH) and T4 markedly prolonged the mean life span of these short-lived animals (mean life span is 5 months). Not only were the immune deficiencies corrected but also other nonimmune aging characteristics like graying and loss of hair, cutaneous and subcutaneous atrophy, bilateral cataracts and reduced cellular turnover, were prevented in 7-month-old animals. In other studies the same group has shown that grafting of neonatal thymus into old mice was able to correct their abnormal serum levels of T3 and insulin, as well as the decreased response of their submandibular glands to isoproterenol [10]. Neonatal thymus grafting into old mice has also been reported to reverse the agerelated decrease of ß1-adrenoceptor density in brain cortex [11] and to correct the age-associated increase in hepatocyte mean nuclear volume [12]. Other workers have reported that the accelerated aging and tumorigenesis of

138

Neuroimmunomodulation 1999;6:137–142

the ovaries that occur in neonatally thymectomized mice can be overcome by the transplant of an intact thymus or injection of T cells [13, 14]. The neuroendocrine system has a powerful influence on immune function [15]. Furthermore, there is compelling evidence that thymus involution, which begins shortly after puberty, is triggered by the neuroendocrine system. Thus, adrenal and gonadal steroids have a marked inhibitory effect on the thymus while adrenalectomy or castration are followed by hypertrophy of the thymus in adult animals [16]. Interestingly, castration in old male rats and mice is followed by a spectacular recovery of thymic mass and morphology, indicating that even the low levels of circulating testosterone in the old animals are exerting a significant inhibitory action on the thymus [17, 18]. Grafting of GH3 pituitary tumor cells, which secrete GH and prolactin (PRL), has been shown to be another effective intervention to restore thymus structure as well as T-cell proliferation and interleukin-2 synthesis in old rats [19].

The Nude Mouse as a Model of Accelerated Aging: Current View

The importance of the thymus in the homeostatic network became evident after the discovery of the congenitally athymic (nude) mouse in the mid-60s [20]. In the initial studies with nude mice it was noted that shortly after birth these mutants as well as normal mice neonatally thymectomized, developed degenerative alterations in the thyroid, adrenal [21] and pituitary glands [22, 23]. By 3 months of age these animals developed a senescent or ‘wasted’ external appearance [21]. In this mutant, the homozygous (nu/nu) females have severe deficiencies in reproductive function in comparison with their phenotypically normal heterozygous littermates (nu/+). The times of vaginal opening and first ovulation are delayed [24], fertility is reduced [25], and follicular atresia is increased such that premature ovarian failure results [26]. Similar abnormalities result from neonatal thymectomy of normal female mice [27, 28]. Nude female mice show significantly reduced levels of circulating and pituitary gonadotropins, a fact that seems to be causally related to the reproductive derangements in these mutants [25]. These and other early life dysfunctions led some investigators to consider the nude mouse as a suitable model of thymus-dependent accelerated aging [29]. It is now generally accepted that, when nude mice are kept under strict germ-free conditions, they are not shortlived. Furthermore, more recent studies in nude mice

Goya/Brown/Bolognani

have demonstrated that, although athymic animals do show altered pituitary hormone responses to certain stimuli, they have no morphologic abnormalities in their pituitaries [30, 31]. Therefore, we can conclude that although athymia brings about functional derangements in the immune-neuroendocrine network, adult nudes raised under specific pathogen-free conditions can compensate these deficits and live reasonably long lives.

Hypophysiotropic Activity of Thymic Hormones Thymosin Peptides as Immunotransmitters

In 1985, Goldstein and co-workers [32] proposed the term immunotransmitter to define those substances produced by the immune system which are active on neural and neuroendocrine circuits. A number of thymosin preparations fall within this category. Thymosin fraction 5 (TF5) is a partially purified thymic preparation which has been extensively characterized by Goldstein and co-workers [33, 34]. TF5 contains a number of biologically active peptides with molecular weights ranging from 1,000 to 15,000, several of which have been purified to homogeneity, sequenced and synthesized. The numerous biological activities of TF5 on immune cells are well documented [33, 34]. In addition to the immunologic activities of TF5, a number of studies indicate that this preparation and two of its component peptides, thymosin ·-1 (T·1) and thymosin ß-4 (Tß4) are capable of modifying neuroendocrine activity. Thus, it has been shown that TF5 and Tß4 are able to stimulate the secretion of luteinizing hormone-releasing hormone (LHRH) from superfused mediobasal hypothalami (MBH) as well as the release of luteinizing hormone (LH) from pituitary glands superfused in sequence with the hypothalami [35]. No release of LH was noted when only pituitaries were superfused with thymosin. It was also found that intracerebroventricular injection of Tß4, but not T·1 into the lateral ventricle of mice caused a significant increase in circulating LH [36]. It was also observed that T·1, but not Tß4, caused a significant increase in corticosterone levels when injected into the lateral ventricle. In other studies, intracerebroventricular administration of T·1 to conscious male rats weakly reduced plasma levels of PRL, thyrotropin (TSH) and adrenocorticotropic hormone (ACTH), whereas incubation of T·1 with rat hemipituitaries evoked a modest release of TSH and ACTH as well as a strong release of LH, with no effect on either PRL, GH or follicle-stimulating hormone (FSH) [37]. In superfused rat hypothalami, T·1 had an inhibito-

Thymic Hormones and Aging

ry effect on the release of thyrotropin-releasing hormone (TRH), corticotropin-releasing hormone (CRH) and somatostatin [38]. TF5 elevated ACTH, ß-endorphin and cortisol in a dose- and time-dependent manner when administered intravenously to prepubertal cynomolgus monkeys [39]. Also, administration of TF5 to rodents resulted in significant elevation of corticosterone in vivo [40]. It has been further demonstrated that TF5 stimulates CRH release from MBH in vitro [41] and ACTH release from cultured rat pituitary cells [42] but does not affect glucocorticoid release from isolated adrenal fasciculata cells [43]. Also, TF5 has been shown to stimulate the release of PRL and GH from rat pituitary cells in vitro [44] as well as immunoreactive ß-endorphin and ACTH from mouse corticotropic tumor cells [45–47]. Thymulin Thymulin is a thymic hormone involved in several aspects of intra- and extrathymic T-cell differentiation [48]. Thymulin, which is exclusively produced by the thymic epithelial cells (TEC), consists of a biologically inactive nonapeptide component (facteur thymique sérique or FTS) coupled in an equimolecular ratio to the ion zinc [49], which confers biological activity to this molecule [50]. Initial studies showed that thymulin exerts a controlling feedback effect on its own secretion both in vivo and in vitro [51, 52]. Additionally, thymulin production and secretion is influenced directly or indirectly by the neuroendocrine system [53]. There is increasing evidence that thymulin possesses hypophysiotropic activity in vitro. Thus, thymulin has been shown to stimulate LH release from perifused rat pituitaries [54] and ACTH from incubated rat pituitary fragments, the latter being an effect mediated by intracellular cyclic adenosine monophosphate (cAMP) and cyclic guanosine monophosphate (cGMP) accumulation [55]. Thymulin has been found to stimulate gonadotropins, PRL and GH in dispersed rat pituitary cells at doses from 10 –8 to 10 –3 M [56, and Brown et al., unpubl. data] whereas others have reported that thymulin doses of 10 –11 M stimulate LH, inhibit PRL release and have no effect on GH secretion in incubated rat pituitary fragments [55]. Thymulin also stimulates 3H-thymidine incorporation into gonocytes of newborn (2 days) rat testes [57]. Homeostatic Thymus Hormone (HTH) HTH is a chromatographically homogeneous thymic preparation of bovine origin [58]. Further purification of HTH preparations was shown to yield two polypeptide chains, HTH· and HTHß [59], whose primary structures

Neuroimmunomodulation 1999;6:137–142

139

were found to be identical to those of histones H2A and H2B, respectively [60]. Replacement therapy with HTH has been shown to prevent most, if not all, of the immune and endocrine consequences of perinatal thymectomy in guinea pigs [61]. In rats, the typical morphological alterations induced in lymphoid organs, hypophysis, thyroid, adrenal cortex and gonads by thymus removal can be prevented by substitutive treatment with HTH [62]. HTH also restores antibody production in thymectomized rats [63]. The effects of HTH on pituitary hormone secretion change with age (see below). Other Thymic Preparations with Hypophysiotropic Activity A 28-kD thymus factor (TF) has been partially purified from prepubertal (14–15 days) rats. It potentiates gonadotropin-releasing hormone (GnRH)-stimulated gonadotropin secretion in cultured rat pituitary cells [64]. This factor appears to be produced by the reticuloepithelial cells of the thymus [65]. TF also acts at gonadal level inhibiting the binding of human chorionic gonadotropin (hCG) to its receptor on rat testis [66] as well as testosterone production [67]. In rat ovarian dispersed cells, TF reduced the hCG-stimulated production of progesterone, estradiol and testosterone [68]. A pituitary hormone-releasing activity termed thymicneuroendocrine-releasing factor (TNRF) has been detected in conditioned medium from thymic reticular monolayers. This factor, whose molecular weight is 110 kD, potentiated TRH-stimulated PRL release and was additive to the effects of GH-releasing hormone (GHRH) on GH release in rat pituitary cells [69]. The 35-amino-acid histone H2A fragment called peptide MB-35, which was purified from TF5, was reported to stimulate PRL and GH release from cultured rat pituitary cells [70, 71]. This peptide also stimulates, although weakly, ACTH, TSH and gonadotropin release from dispersed rat pituitary cells [72–74].

ance of detectable levels of circulating thymulin [80]. Hypothalamic and pituitary extracts from young mice stimulate thymulin release from TEC cultures but this stimulation declines when the pituitary and hypothalamic extracts are obtained from old mice [81]. Conversely, there is evidence that the neuroendocrine system of old rodents becomes less responsive to thymic signals. It has been reported that TF5 and HTH have TSH-inhibiting activity in young but not in old rats [82–84]. Furthermore, intravenous administration of HTH was also able to reduce plasma GH and increase corticosterone levels in both young and old rats, although these responses were much weaker in the old animals [84, 85]. Peptide MB-35 stimulates ACTH release in dispersed pituitary cells from young but not mature rats [72]. In perifused anterior pituitary cells thymulin induces multiple hormone release, an effect that declines with the age of the pituitary donors [86, 87].

Concluding Remarks

Although further work is necessary to clarify the physiological significance of the data reviewed here, the emerging view is that in addition to its central role in the regulation of the immune function, the thymus gland may extend its influence to nonimmunologic components of the body, including the neuroendocrine system. This opens an interesting hypothesis in gerontology, namely that the early onset of thymus involution might act as a triggering event which would initiate the gradual decline in homeostatic potential that characterizes the aging process. Certainly, an age-dependent desensitization of the neuroendocrine system and other integrative centers of the body to thymic signals could act as one of the pacemakers of homeostatic decline.

Acknowledgments Thymic Hormones and Aging

Aging brings about a progressive reduction in serum levels of thymic hormones which is paralleled by a decline in circulating levels of various pituitary hormones [75]. Interestingly, it has been shown that treatment of aged animals with GH and T4, separately or in combination, can partly reverse their decreased thymic hormone levels [76–79]. Furthermore, treatment of old mice with hypothalamic extracts from young mice results in reappear-

140

Neuroimmunomodulation 1999;6:137–142

The authors are grateful to Ms. Yolanda E. Sosa for assistance with the bibliographic search and the typing of the manuscript. Part of our own work mentioned here was supported by grants from the Argentine Research Council (CONICET).

Goya/Brown/Bolognani

References 1 Solomon JB: Ontogeny of defined immunity in mammals; in Neuberger A, Tatum EL (eds): Fetal and Neonatal Immunology. Front Biol. New York, Elsevier, 1971, vol 20, pp 234–306. 2 Jost A: The extent of foetal and endocrine autonomy; in Wolstenholme GEW, O’Connor M (eds): Foetal Autonomy. Ciba Found Symp. London, Churchill, 1969, pp 79–94. 3 Besedovsky HO, Sorkin E: Network of immunoneuroendocrine interactions. Clin Exp Immunol 1977;27:1–12. 4 Saphier D, Abramsky O, Mor G, Ovadia H: A neurophysiological correlate of an immune response; in Jankovic BD, Markovic BM, Spector NH (eds): Neuroimmune Interactions. Proc Sec Int Workshop Neuroimmunomodulation. Ann NY Acad Sci 1987;496:354–359. 5 Besedovsky HO, Del Rey AE, Sorkin E: Antigenic competition between horse and sheep red blood cells as a hormone-dependent phenomenon. Clin Exp Immunol 1979;37:106–113. 6 Besedovsky HO, Sorkin E, Keller M, Müller J: Changes in blood hormone levels during the immune response. Proc Soc Exp Biol Med 1975;150:466–470. 7 Keast D, Ayre DJ: Antibody regulation in birds by thyroid hormones. Dev Comp Immunol 1980;4:323–330. 8 Tokuda S, Trujillo LC, Nofchissey RA: Hormonal regulation of the immune function; in Cooper EL (ed): Stress, Immunity and Aging. New York, Dekker, 1984, pp 141–155. 9 Fabris N, Pierpaoli W, Sorkin E: Lymphocytes, hormones and aging. Nature 1972;240:557– 559. 10 Piantanelli L, Basso A, Muzzioli M, Fabris N: Thymus-dependent reversibility of physiological and isoproterenol evoked age-related parameters in athymic (nude) and old normal mice. Mech Age Dev 1978;7:171–182. 11 Viticchi C, Gentile S, Piantanelli L: Ageing and thymus-induced differential regulation of ß1and ß2-adrenoceptors of mouse brain cortex. Arch Gerontol Geriatr 1989;8:13–20. 12 Pieri C, Giuli C, Del Moro M, Piantanelli L: Electron microscopic morphometric analysis of mouse liver. II. Effect of aging and thymus transplantation in old animals. Mech Age Dev 1980;13:275–283. 13 Sakakura T, Nishizuka Y: Thymic control mechanism in ovarian dysgenesis in thymectomized mice by replacement with thymic and other lymphoid tissues. Endocrinology 1972; 90:431–437. 14 Sakaguchi S, Takahashi T, Nishizuka Y: Study on cellular events in post-thymectomy autoimmunity oophoritis in mice. II. Requirement of Lyt-1 cells in normal female mice for the prevention of oophoritis. J Exp Med 1982;156: 1577–1586. 15 Besedovsky HO, Del Rey A: Immune-neuroendocrine interactions: Facts and hypotheses. Endocr Rev 1996;17:64–102.

Thymic Hormones and Aging

16 Comsa J: Hormonal interactions of the thymus; in Luckey TD (ed): Thymic Hormones. Baltimore, University Park Press, 1973, pp 59– 96. 17 Greenstein BD, Fitzpatrick FTA, Adcock IM, Kendall MD, Wheeler MJ: Reappearance of the thymus in old rats after orchidectomy: Inhibition of regeneration by testosterone. J Endocr 1986;110:417–422. 18 Utsuyama M, Hirokawa K: Hypertrophy of the thymus and restoration of immune functions in mice and rats by gonadectomy. Mech Age Dev 1989;47:175–185. 19 Kelley KW, Brief S, Westly HJ, Novakofski J, Bechtel PJ, Simon J, Walker EB: GH3 pituitary adenoma cells can reverse thymic aging in rats. Proc Natl Acad Sci USA 1986;83:5663–5667. 20 Flanagan SP: ‘Nude’ a new hairless gene with pleiotropic effects in the mouse. Genet Res 1966;8:295–309. 21 Pierpaoli W, Sorkin E: Alterations of adrenal cortex and thyroid in mice with congenital absence of the thymus. Nature 1972;238:282– 285. 22 Bianchi E, Pierpaoli W, Sorkin E: Cytological changes in the mouse anterior pituitary after neonatal thymectomy: A light and electron microscopical study. J Endocrinol 1971;51:1–6. 23 Pierpaoli W, Sorkin E: Cellular modifications in the hypophysis of neonatally thymectomized mice. Br J Exp Pathol 1967;48:627–631. 24 Besedovsky HO, Sorkin E: Thymus involvement in female sexual maturation. Nature 1974;249:356–358. 25 Rebar RW, Morandini IC, Erickson GF, Petze JE: The hormonal basis of reproductive defects in athymic mice. Endocrinology 1981;108: 120–126. 26 Lintern-Moore S, Pantelouris EM: Ovarian development in athymic nude mice. I. The size and composition of the follicle population. Mech Age Dev 1975;4:385–390. 27 Michael SD, Taguchi O, Nishizuka Y: Effects of neonatal thymectomy on ovarian development and plasma LH, FSH, GH and PRL in the mouse. Biol Reprod 1980;22:343–350. 28 Nishizuka Y, Sakakura T: Ovarian dysgenesis induced by neonatal thymectomy in the mouse. Endocrinology 1971;89:889–893. 29 Piantanelli L, Fabris N: Hypopituitary dwarf and athymic nude mice and the study of the relationships among thymus, hormones and aging. Genetic effects on aging. Birth Defects 1978;14:315–333. 30 Goya RG, Sosa YE, Co´nsole GM, Dardenne M: Altered thyrotropic and somatotropic responses to environmental challenges in congenitally athymic mice. Brain Behav Immun 1995; 9:79–86. 31 Goya RG, Sosa YE, Co´nsole GM, Dardenne M: Altered regulation of serum prolactin in nude mice. Med Sci Res 1996;24:279–280. 32 Hall NR, McGillis JP, Spangelo BL, Goldstein AL: Evidence that thymosin and other biological response modifiers can function as neuroactive immunotransmitters. J Immunol 1985; 135:806S–811S.

33 Goldstein AL, Low TLK, Thurman GB, Zatz MM, Hall NR, Chen J, Hu S-K, Naylor PB, McClure JE: Current status of thymosin and other hormones of the thymus gland. Recent Prog Horm Res 1981;37:369–415. 34 Spangelo BL, Hall NR, Goldstein AL: Biology and chemistry of thymosin peptides; in Jankovic BD, Markovic BM, Spector NH (eds): Neuroimmune Interactions. Proc Sec Int Workshop Neuroimmunomodulation. Ann NY Acad Sci 1987;496:196–204. 35 Rebar RW, Miyake A, Low TLK, Goldstein AL: Thymosin stimulates secretion of luteinizing hormone. Science 1981;214:669–671. 36 Hall NR, McGillis JP, Spangelo BL, Palaszynski E, Moody T, Goldstein AL: Evidence for a neuroendocrine-thymus axis mediated by thymosin polypeptides; in Serrou B, Rosenfeld C, Daniel JC, Saunders JP (eds): Current Concepts in Human Immunology and Cancer Immunomodulation. Amsterdam, Elsevier, 1982, pp 653–660. 37 Milenkovic L, McCann SM: Effects of thymosin alpha-1 on pituitary hormone release. Neuroendocrinology 1992;55:14–19. 38 Milenkovic L, Lyson K, Aguila MC, McCann SM: Effect of thymosin alpha-1 on hypothalamic hormone release. Neuroendocrinology 1992;56:674–679. 39 Healy DL, Hodgen GD, Schulte HM, Chrousos DL, Loriaux N, Hall NR, Goldstein AL: The thymus-adrenal connection: Thymosin has corticotropic-releasing activity in primates. Science 1983;222:1353–1355. 40 McGillis JP, Hall NR, Vahouny JV, Goldstein AL: Thymosin fraction five causes increased serum corticosterone in rodents in vivo. J Immunol 1985;134:3952–3955. 41 Spinedi EJ, Hadid R, Daneva T, Gaillard RC: Cytokines stimulate the CRH but not the vasopressin neuronal system: Evidence for the median eminence site of interleukin-6 action. Neuroendocrinology 1992;56:46–53. 42 McGillis JP, Hall NR, Goldstein AL: Thymosin fraction five stimulates secretion of ACTH from cultured rat pituitaries. Life Sci 1988;42: 2259–2268. 43 Vahouny GV, Kyeyune-Nyombi E, McGillis JP, Tare NS, Huang KY, Tombes R, Goldstein AL, Hall NR: Thymosin peptides and lymphokines do not directly stimulate adrenal corticosteroid production in vitro. J Immunol 1983; 130:791–794. 44 Spangelo BL, Judd AM, Ross PC, Login IS, Jarvis WD, Badamchian M, Goldstein AL, MacLeod RM: Thymosin fraction five stimulates PRL and GH release from anterior pituitary cells in vitro. Endocrinology 1987;121: 2035–2043. 45 Farah JR, Hall NR, Bishop JF, Goldstein AL, O’Donohue TL: Thymosin fraction five stimulates secretion of immunoreactive ß-endorphin in mouse corticotropic tumor cell. J Neurosci Res 1987;18:140–146.

Neuroimmunomodulation 1999;6:137–142

141

46 Goya RG, Castro MG, Hannah MJ, Sosa YE, Lowry PJ: Thymosin peptides stimulate corticotropin release by a calcium-dependent mechanism. Neuroendocrinology 1993;57:230–235. 47 Goya RG, Castro MG, Hannah MJ, Sosa YE, Lowry PJ: Corticotropin-releasing (CRH) activity of thymic peptides on CRH-insensitive corticotropic tumor cells. Medicina 1993;53: 108–112. 48 Bach JF: Thymulin (FTS-Zn). Clin Immunol Allergy 1983;3:133–156. 49 Gastinel LN, Dardenne M, Pléau JM, Bach JF: Studies on the zinc-binding site to the serum thymic factor. Biochim Biophys Acta 1984; 797:147–155. 50 Dardenne M, Nabarra B, Lefrancier P: Contribution of zinc and other metals to the biological activity of serum thymic factor FTS. Proc Natl Acad Sci USA 1982;79:5370–5373. 51 Savino W, Dardenne M, Bach JF: Thymic hormones containing cells. III. Evidence for a feedback regulation of the secretion of the serum thymic factor FTS by thymic epithelial cells. Clin Exp Immunol 1983;52:7–12. 52 Cohen S, Berrih S, Dardenne M, Bach JF: Feed-back regulation of the secretion of a thymic hormone (thymulin) by human thymic epithelial cells in culture. Thymus 1986;8:109– 119. 53 Savino W, Dardenne M: Immune-neuroendocrine interactions. Immunol Today 1995;16: 318–322. 54 Zaidi SAA, Kendall MD, Gillham B, Jones MT: The release of LH from pituitaries perifused with thymic extracts. Thymus 1988;12: 253–264. 55 Hadley AJ, Rantle CM, Buckingham JC: Thymulin stimulates corticotrophin release and cyclic nucleotide formation in the rat anterior pituitary gland. Neuroimmunomodulation 1997;4:62–69. 56 Goya RG, Sosa YE, Brown OA, Dardenne M: In vitro studies on the thymus-pituitary axis in young and old rats. Ann NY Acad Sci 1994; 741:108–114. 57 Prépin J, Le Vigouroux P, Dadoune JP: Effects of thymulin on in vitro incorporation of 3Hthymidine into gonocytes of newborn rat testes. Reprod Nutr Dev 1994;34:289–294. 58 Bernardi G, Comsa J: Purification chromatographique d’une préparation de thymus douée d’activité hormonale. Experientia 1965;21: 416–417. 59 Reichhart R, Zeppezauer M, Jörnvall H: Preparations of homeostatic thymus hormone consist predominantly of histones 2A and 2B and suggest additional histone functions. Proc Natl Acad Sci USA 1985;82:4871–4875.

142

60 Reichhart R, Jörnvall H, Carlquist H, Zeppezauer M: The primary structure of two polypeptide chains from preparations of homeostatic thymus hormone (HTH· and HTHß). FEBS Lett 1985;188:63–67. 61 Comsa J: Action of the purified thymus hormone in thymectomized guinea pigs. Am J Med Sci 1965;250:79–85. 62 Comsa J, Philipp EM, Leonhardt H: Effects of thymectomy on the endocrine glands of the rat. Israel J Med Sci 1977;13:354–362. 63 Comsa J, Schwarz JA, Neu H: Intreraction between thymic hormone and hypophyseal growth hormone on production of precipitating antibodies in the rat. Immunol Commun 1974; 3:11–18. 64 Mendoza ME, Romano MC: Prepubertal rat thymus secretes a factor that modulates gonadotropin secretion in cultured rat pituitary cells. Thymus 1989;14:233–242. 65 Mendoza ME, Martin D, Candelaria PG, Romano MC: Evidence that secretory products of the reticuloepithelial cells of the rat thymus modulate the secretion of gonadotrophins by rat pituitary cells in culture. J Reprod Immunol 1995;28:203–215. 66 Hiriart M, Romano MC: Human chorionic gonadotropin binding to rat testis receptors is inhibited by a thymus factor. Life Sci 1986;38: 789–795. 67 Reyes-Sparza JA, Romano MC: An age-dependent thymic secretion modulates testicular function. J Steroid Biochem 1989;34:541–545. 68 Aguilera G, Romano MC: Influence of the thymus on steroidogenesis by rat ovarian cells in vitro. J Endocrinol 1989;123:367–373. 69 Spangelo BL, Ross PC, Judd AM, MacLeod RM: Thymic stromal elements contain an anterior pituitary hormone stimulating activity. J Neuroimmunol 1989;25:37–46. 70 Badamchian M, Huang SS, Spangelo BL, Damavandy T, Goldstein AL: Chemical and biological characterization of MB-35: A thymic derived peptide that stimulates the release of GH and PRL from rat AP cells. PNEI 1990;3: 258–265. 71 Badamchian M, Spangelo BL, Damavandy T, MacLeod RM, Goldstein AL: Complete amino acid sequence analysis of a peptide isolated from the thymus that enhances release of GH and PRL. Endocrinology 1991;128:1580– 1588. 72 Goya RG, Castro MG, Saphier PW, Sosa YE, Lowry PJ: Thymus-pituitary interaction during ageing. Age Ageing 1993;22:S19–S25. 73 Brown OA, Sosa YE, Goya RG: Gonadotrophin-releasing activity of histones H2A and H2B. Cell Mol Life Sci 1998;54:288–294. 74 Brown OA, Sosa YE, Goya RG: Thyrotropinreleasing activity of histones H2A, H2B and peptide MB-35. Peptides 1997;18:1315–1319.

Neuroimmunomodulation 1999;6:137–142

75 Goya RG, Naylor PH, Goldstein AL, Meites J: Changes in circulating levels of neuroendocrine and thymic hormones during aging in rats: A correlation study. Exp Gerontol 1990;25:149– 157. 76 Goya RG, Gagnerault MC, Leite de Moraes MC, Savino W, Dardenne M: In vivo effects of growth hormone on thymus function in aging mice. Brain Behav Immun 1992;6:341–354. 77 Goya RG, Gagnerault MC, Sosa YE, Bevilacqua JA, Dardenne M: Effects of growth hormone and thyroxine on thymulin secretion in aging rats. Neuroendocrinology 1993;58:338– 343. 78 Fabris N, Mocchegiani E: Endocrine control of thymic serum factor production in young-adult and old mice. Cell Immunol 1985;91:325– 335. 79 Goff BL, Roth JA, Arp LH, Incefy GS: Growth hormone treatment stimulates thymulin production in aged dogs. Clin Exp Immunol 1987; 68:580–587. 80 Folch H, Eller G, Mena M, Esquivel P: Neuroendocrine regulation of thymus hormones: Hypothalamic dependence of FTS level. Cell Immunol 1986;102:211–216. 81 Goya RG, Gagnerault MC, Sosa YE, Dardenne M: Reduced ability of pituitary extracts from old mice to stimulate thymulin secretion in vitro. Mech Age Dev 1995;83:143–154. 82 Goya RG, Takahashi S, Quigley KL, Sosa YE, Goldstein AL, Meites J: Immune-neuroendocrine interactions during aging: Age-dependent thyrotropin-inhibiting activity of thymosin peptides. Mech Age Dev 1987;41:219–227. 83 Goya RG, Sosa YE, Quigley KL, Gottschall PE, Goldstein AL, Meites J: Differential activity of thymosin peptides (thymosin fraction 5) on plasma thyrotropin in female rats of different ages. Neuroendocrinology 1988;47:379– 383. 84 Goya RG, Quigley KL, Takahashi S, Reichhart R, Meites J: Differential effect of homeostatic thymus hormone on plasma thyrotropin and growth hormone in young and old rats. Mech Age Dev 1989;49:119–128. 85 Goya RG, Sosa YE, Quigley KL, Reichhart R, Meites J: Homeostatic thymus hormone stimulates corticosterone secretion in a dose- and age-dependent manner in rats. Neuroendocrinology 1990;51:59–63. 86 Brown OA, Sosa YE, Dardenne M, Pléau J-M, Goya RG: Growth hormone-releasing activity of thymulin: Effects of age. Neuroendocrinology, in press. 87 Brown OA, Sosa YE, Bolognani F, Goya RG: Thymulin stimulates prolactin and thyrotropin release in an age-related manner. Mech Age Dev, in press.

Goya/Brown/Bolognani

Author Index Vol. 6, No. 1–2, 1999

Alves, L.A. 81 Arkins, S. 56 Arzt, E. 126 Ashwell, J.D. 90 Bolognani, F. 137 Botham, C.A. 6 Brilot, F. 115 Brown, O.A. 137 Brucker, C. 31 Burgess, W. 56

Liu, Q. 56 Longo, D.L. 69 Martens, H. 115 Martinez, C. 97 McCann, S.M. 5 Mentlein, R. 31, 45 Mu´ñoz, J.J. 23 Murphy, W.J. 69 Ozawa, A. 56

Charlet-Renard, C. 115 Coutinho-Silva, R. 81

Patay, B. von 45 Pleau, J.-M. 108

Dantzer, R. 56 Dardenne, M. 108, 126 Delgado, M. 97 Downing, J.E.G. 31

Richards, S. 69 Rinner, I. 51

Geenen, V. 115 Globerson, A. 51 Gomariz, R.P. 97 Goya, R.G. 137 Head, G.M. 31 Homo-Delarche, F. 108 Jabbur, S.J. 39 Jiménez, E. 23 Jones, G.V. 6 Kanaan, S.A. 39 Kawashima, K. 51 Kecha, O. 115 Kelley, K.W. 56 Kendall, M.D. 6, 31 Korsatko, W. 51 Kurz, B. 45

Saadé, N.E. 39 Sacedo´n, R. 23 Safieh-Garabedian, B. 39 Savino, W. 81, 126 Schauenstein, K. 51 Sun, R. 69 Tang, Q. 56 Taub, D.D. 69 Throsby, M. 108 Tian, Z.-G. 69 Tolosa, E. 90 VanHoy, R. 56 Varas, A. 23 Vicente, A. 23 Welniak, L.A. 69 Woody, M.A. 69 Zapata, A.G. 23 Zhou, J.-H. 56

Leceta, J. 97 Lipton, J.M. 5

ABC Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

© 1999 S. Karger AG, Basel

Accessible online at: http://BioMedNet.com/karger

143

Subject Index Vol. 6, No. 1–2, 1999

Acetylcholine production 51 Acrosome reaction 31 Adenosine triphosphate 45 Adrenoceptors 45 Aging 137 Antigen-presenting cells 108 Apoptosis 51, 56, 97 ATP 81 Autoimmunity 90, 115 Bcl-2 56 Calcium 31 Catecholamines 45 Culturing thymic cells 6 Cytokine(s) 39, 126 – production 97 Dendritic cells 23 Gene expression 97 Glucocorticoids 23, 90 Growth hormone 69 Hematopoiesis 56, 69 Homeostatic thymus hormone 137 Hyperalgesia 39 Hypothalamus 126 IGF-I receptor 56 IL-4 receptor 56 Immune parameters 69 Immunoneuroendocrine network 137 Insulin receptor substrate-1 56 Interleukin-6 45 Ion channels 31 Laminin 23 Macrophages 23 Mouse 108

ABC Fax + 41 61 306 12 34 y-Mail [email protected] www.karger.com

© 1999 S. Karger AG, Basel

Accessible online at: http://BioMedNet.com/karger

Nerve growth factor 39 Neuroimmunology 39 Neuroimmunomodulation 97 Neuropeptide(s) 115 – secretion 97 Nicotinic acetylcholine receptors 51 Noradrenaline 45 Phosphatidylinositol 3)-kinase 56 Pituitary desensitization 137 Pituitary gland 126 Preproinsulin 108 Proglucagon 108 Prolactin 69 Propancreatic polypeptide 108 Prosomatostatin 108 Prostaglandin-E2 39 Purinergic receptors 81 Purines 45 RT-PCR 108 Secretion 31 Self tolerance 115 Steroids 90 T-cell differentiation 97 Thymic epithelial cells 6, 31, 45, 51, 81, 126 – epithelium 23 – hormones 137 – nurse cells 81 Thymocyte(s) 97, 126 – development 90 – selection 90 Thymosin 137 Thymulin 39, 137 Thymus 6, 51, 115, 126 Transplantation 6, 69 Vasoactive intestinal peptide 97