This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Savino, W. Articles by Dardenne, M. Search for Related Content PubMed PubMed Citation Articles by Savino, W. Articles by Dardenne, M.

Neuroendocrine Control of Thymus Physiology¹

Laboratory on Thymus Research (W.S.), Department of Immunology, Oswaldo Cruz Institute, Oswaldo Cruz Foundation, 21045–900 Rio de Janeiro, Brazil; and CNRS UMR 8603 (M.D.), Université Paris V, Hôpital Necker, 75015 Paris, France

	Abstract

Top Abstract I. Introduction II. The Thymic Microenvironment... III. Neuroendocrine Control of... IV. Thymic Endocrine Function... V. Proliferation of Thymic... VI. Hormonal Modulation of... VII. Is Thymocyte Traffic... VIII. Expression of Receptors... IX. Intrathymic Expression of... X. Conclusions and Major... References

 
The thymus gland is a central lymphoid organ in which bone marrow-derived T cell precursors undergo differentiation, eventually leading to migration of positively selected thymocytes to the peripheral lymphoid organs. This differentiation occurs along with cell migration in the context of the thymic microenvironment, formed of epithelial cells, macrophages, dendritic cells, fibroblasts, and extracellular matrix components. Various interactions occurring between microenvironmental cells and differentiating thymocytes are under neuroendocrine control. In this review, we summarize data showing that thymus physiology is pleiotropically influenced by hormones and neuropeptides. These molecules modulate the expression of major histocompatibility complex gene products by microenvironmental cells and the extracellular matrix-mediated interactions, leading to enhanced thymocyte adhesion to thymic epithelial cells. Cytokine production and thymic endocrine function (herein exemplified by thymulin production) are also hormonally controlled, and, interestingly in this latter case, a bidirectional circuitry seems to exist since thymic-derived peptides also modulate hormonal production.

In addition to their role in thymic cell proliferation and apoptosis, hormones and neuropeptides also modulate intrathymic T cell differentiation, influencing the generation of the T cell repertoire.

Finally, neuroendocrine control of the thymus appears extremely complex, with possible influence of biological circuitry involving the intrathymic production of a variety of hormones and neuropeptides and the expression of their respective receptors by thymic cells.

I. Introduction

II. The Thymic Microenvironment and Its Role in T Cell Differentiation

A. Intrathymic T cell differentiation: general comments

B. Cellular interactions involving the thymic microenvironment

C. Heterogeneity of the thymic epithelium: the thymic nurse cell complex

III. Neuroendocrine Control of Membrane Interactions Between Thymocytes and Microenvironmental Cells

A. Is there a hormonal effect upon MHC expression by microenvironmental cells?

B. Extracellular matrix-mediated TEC-thymocyte interactions are hormonally modulated

C. Are inter-TEC gap junctions under neuroendocrine control?

IV. Thymic Endocrine Function and Cytokine Secretion by Microenvironmental Cells Are Controlled by Hormones and Neuropeptides: The Paradigm of Thymulin

A. Steroid and peptidic hormones influence in vitro cytokine secretion

B. Thyroid and pituitary hormone status modulates thymulin secretion

C. Effects of adrenalectomy and gonadectomy on thymulin levels

D. Is there an autocrine/paracrine circuitry controlling thymulin secretion?

E. Thymic hormones modulate endocrine glands and neuroendocrine circuits

V. Proliferation of Thymic Cells Is Hormonally Influenced

A. In vitro effects of pituitary hormones on TEC growth

B. Modulation of thymocyte proliferation by hormones and neuropeptides

VI. Hormonal Modulation of Intrathymic T Cell Differentiation

A. Thymocyte-derived cytokine profile

B. Changes in the T cell differentiation markers and TCR Vß repertoire

C. Hormone-mediated apoptosis in thymocytes

VII. Is Thymocyte Traffic Under Neuroendocrine Control?

A. Effects on the entrance of cell precursors into the thymus

B. Modulation of thymocyte traffic in TNCs

C. Control of thymocyte exit from the thymus

VIII. Expression of Receptors for Hormones and Neuropeptides by Thymic Cells

A. The steroid/thyroid hormone receptor family

B. Expression of PRL and GH receptors by thymocytes and TEC

C. Receptors for neuropeptides in thymic cells

IX. Intrathymic Production of Hormones and Neuropeptides

A. Intrathymic production of corticosterone: role in the shaping of the T cell repertoire

B. Expression of "classic" adenopituitary hormones by thymic cells

C. Is there a functional autocrine/paracrine IGF-I-mediated circuitry in the thymus?

D. Neuropeptide expression by the thymic microenvironment

X. Conclusions and Major Questions to Be Addressed

	I. Introduction

Top Abstract I. Introduction II. The Thymic Microenvironment... III. Neuroendocrine Control of... IV. Thymic Endocrine Function... V. Proliferation of Thymic... VI. Hormonal Modulation of... VII. Is Thymocyte Traffic... VIII. Expression of Receptors... IX. Intrathymic Expression of... X. Conclusions and Major... References

 
THE CROSS-TALK between the neuroendocrine and immune systems is now well demonstrated. These systems use similar ligands and receptors to establish a physiological intra- and intersystem communication circuitry that plays an important role in homeostasis. Increasing evidence has placed hormones and neuropeptides among potent immunomodulators, participating in various aspects of immune system function, in both health and disease (reviewed in Refs. 1, 2, 3, 4). More particularly, the physiology of the thymus is modulated by a variety of biological circuits including those mediated by steroid and polypeptidic hormones, as well as neuropeptides. Herein, we focus on both the microenvironmental and the lymphoid aspects of neuroendocrine control of the thymus. We will first briefly comment on the structure of the thymic microenvironment and its role in the complex process of intrathymic T cell differentiation.

	II. The Thymic Microenvironment and Its Role in T Cell Differentiation

Top Abstract I. Introduction II. The Thymic Microenvironment... III. Neuroendocrine Control of... IV. Thymic Endocrine Function... V. Proliferation of Thymic... VI. Hormonal Modulation of... VII. Is Thymocyte Traffic... VIII. Expression of Receptors... IX. Intrathymic Expression of... X. Conclusions and Major... References

 
The thymus gland is a central lymphoid organ in which bone marrow-derived T cell precursors undergo a complex process of maturation, eventually leading to migration of positively selected thymocytes to the T cell-dependent areas of peripheral lymphoid organs such as spleen, lymph nodes, Peyer’s patches, and tonsils (5). Such a differentiation process involves sequential expression of various membrane markers and rearrangements of the T cell receptor genes.

Importantly, although thymocyte proliferation and differentiation persist throughout life, they diminish with aging. Older thymuses are significantly atrophied and have fewer thymocytes than younger ones. However, even in humans, the adult thymus is still active in terms of delivering mature T lymphocytes to the periphery of the immune system (6, 7).

A. Intrathymic T cell differentiation: general comments
From the entrance of T cell precursors into the thymus to the exit of mature cells from the organ, a vast body of interactions promotes the complex process of T cell differentiation. This differentiation involves regulation of the expression of various membrane proteins. A key membrane protein is the T cell receptor (TCR), which in the cell membrane is physiologically coupled with a molecular complex, termed CD3. Additional accessory molecules, including CD4 and CD8, as well as CD25 and the proteoglycan CD44, are useful to define stages of intrathymic T cell differentiation.

The TCR is a heterodimer formed by an ß- or a -chain configuration. Although ⁺ thymocytes are the first to appear in the thymus with ontogeny of the organ, in the adult organ, around 99% of TCR⁺ thymocytes express TCRß and only 1% are T cells. A major point in intrathymic T cell differentiation is that, after gene rearrangements, any of the TCR peptide chains are generated, resulting in a large diversity of TCRs bearing distinct peptide specificities. This complex but well understood phenomenon is beyond the scope of this article but is reviewed in detail in recent publications (8, 9, 10).

The CD3 complex is an assembly of polypeptidic chains physically associated with the TCR. This association, together with the fact that CD3 bears cytoplasmic domains capable of phosphorylation, provides the intracellular signal transduction pathways necessary for TCR-driven T cell activation. Such activation follows ligation with a peptide presented by molecules of the major histocompatibility complex (MHC) expressed on the membranes of nonlymphoid cells. This ligation is favored by the accessory molecules CD4 and CD8, which are transmembrane glycoproteins that interact with class II and class I MHC molecules, respectively.

CD25 is the -chain of the interleukin 2 (IL-2) receptor, and when it is expressed together with the ß- and -chain, the receptor acquires high affinity for IL-2, thus favoring IL-2-driven thymocyte proliferation. The proteoglycan CD44 is a receptor for hyaluronic acid, and to a lesser extent for fibronectin and collagen. As seen below, it is associated with thymocyte migration events and is also considered a marker for T cell activation (11).

Cytofluorometric combined analysis of these markers proved to be useful in defining intrathymic T cell differentiation. The most immature thymocytes express neither the TCR/CD3 complex nor the accessory molecules CD4 or CD8 and thus are called double-negative thymocytes. Nevertheless, we can determine differentiation steps within the double-negative compartment by distinguishing the cells on the basis of their CD25 and CD44 expression. Thymocyte precursors that recently entered the thymus, in addition to being TCR/CD3^-CD4^-CD8^-, are also CD44⁺CD25^-. As they differentiate, these immature cells acquire CD25 on the cell membrane, becoming CD44⁺CD25⁺, and then sequentially lose CD44 and CD25. The whole double-negative compartment represents about 5% of total thymocytes. Thymocyte maturation then progresses with the acquisition of both CD4 and CD8 markers, generating the so-called CD4⁺CD8⁺ double-positive thymocytes. These cells are the most common in the thymus, comprising 80% of total thymocytes. In the double-positive stage, TCR genes are rearranged. In differentiation of TCRß-bearing cells, the ß-chain-related genes are rearranged first followed by the -chain genes. At this stage, TCR is expressed in low density on the cell membrane. Thymocytes that do not undergo a productive TCR gene rearrangement (i.e., that will not ultimately generate a peptide chain expressed on the cell membrane) die by default through apoptosis. By contrast, those expressing productive TCR will be able to react with peptides presented by molecules of the MHC, expressed on the membranes of nonlymphoid cells. This interaction will determine the positive and negative selection events, crucial for normal thymocyte differentiation. Positive selection allows the differentiation step whereby an immature, short-lived, CD4⁺CD8⁺ thymocyte escapes from programmed cell death and becomes a mature, long-lived, CD4⁺ or CD8⁺ single positive cell. This is a highly stringent process, sparing only a small proportion of the CD4⁺CD8⁺ population. Positive selection also coincides with lineage commitment: the decision to become a CD4⁺ or CD8⁺ single positive thymocyte, as a function of the class of MHC molecule with which the TCR can interact. Negative selection in the thymus is the screen for establishing self-tolerance in the T cell repertoire, promoting deletion of T cells that might potentially be autoreactive to self-proteins. As illustrated in Fig. 1, positive selection events begin earlier in CD3⁺ double-positive cells, whereas negative selection takes place in both double-positive and single-positive thymocytes (Reviewed in Ref. 8).

View larger version (93K): [in this window] [in a new window]   Figure 1. Stages of thymocyte differentiation. This scheme depicts the differentiation pathway for TCRß+ mouse thymocytes, as ascertained by the expression of various surface markers. When entering the thymus, T cell precursors bear the phenotype TCR/CD3-CD4-CD8-CD25-CD44+. As they differentiate, immature cells acquire CD25, become CD44+CD25+, and then gradually lose CD44 followed by loss of CD25. Thymocyte maturation then progresses with the acquisition of CD4 and CD8 markers, generating the CD4+CD8+ thymocytes. In the double-positive stage, TCR genes are rearranged, for sequential expression of ß- and -chains, respectively (illustrated by the intra-arrow wheels in the figure). Productive TCRs, complexed with CD3, are initially expressed in low density on the cell membranes. These cells are then exposed to positive and negative selection events. Positively selected thymocytes progress to the mature CD4+CD8- or CD4-CD8+ single positive stage. These mature thymocytes, which now express high densities of the TCR/CD3 complex, will ultimately leave the organ. Thymocytes that did not succeed in expressing TCR on the cell membrane, as well as those undergoing negative selection, will die by apoptosis. Localization of these various steps of thymocyte differentiation in the thymic lobule can be seen at the right side of the figure.

It is noteworthy that thymocyte differentiation occurs as cells migrate within the thymic lobules. As seen in Fig. 2, top panel, most of the immature thymocytes, including those bearing the phenotypes CD3^-CD4^-CD8^- and CD3⁺CD4⁺CD8⁺ are cortically located, whereas mature CD3⁺CD4⁺CD8^- and CD3⁺CD4^-CD8⁺ cells are found in the medulla, being those that will normally leave the organ to populate the T cell-dependent areas of peripheral lymphoid organs (5).

View larger version (87K): [in this window] [in a new window]   Figure 2. The thymic microenvironment and intrathymic T cell differentiation. A simplified model of thymocyte migration can be seen in the top panel. This panel depicts the common entrance site of precursor cells into the organ through blood vessels. Having entered the thymus, thymocytes migrate during differentiation to ultimately leave the organ (also by blood vessels), bearing the mature phenotypes of CD4+CD8- or CD4-CD8+ thymocytes. The bottom panel schematically depicts a thymic lobule, showing thymocytes intermingled with a heterogeneous cellular network representing the thymic microenvironment. This nonlymphoid compartment is composed of various cell types, including epithelial cells (yellow and orange), dendritic cells (red), macrophages (blue), and fibroblasts (green). Among TEC, morphological heterogeneity can be distinguished in subseptal/subcapsullary, cortical, and medullary regions. A particular cortically located lymphoepithelial complex, the thymic nurse cell, is seen.

B. Cellular interactions involving the thymic microenvironment
Several kinds of heterotypic interactions occur between differentiating thymocytes and microenvironmental cells. As mentioned above, one key cellular interaction involves the TCR/CD3 complex, expressed by differentiating thymocytes, with class I or class II MHC products on the microenvironmental cell membranes, complexed with a given endogenous peptide to be recognized, in the context of CD8 or CD4 molecules, respectively. The avidity of the resulting interaction is a determinant for positive vs. negative selection. Thymocytes with high avidity are negatively selected and are also deleted by apoptosis. This leads to the death of large numbers of potentially harmful autoreactive T cells. By contrast, a small percentage of thymocytes with intermediate avidity for recognition of MHC self-peptides appears to be rescued from death and is positively selected. Positive selection appears to be essentially conveyed by thymic epithelial cells (TEC) whereas negative selection can occur in the context of hematopoietic-derived DC, but also of TEC (5, 13).

In addition to the TCR/MHC-peptide interaction, the thymic microenvironment can influence the process of thymocyte migration/differentiation via other types of heterotypic membrane interactions. For example, TEC express classical membrane adhesion molecules such as ICAM-1 (intercellular adhesion molecule 1) and LFA-3, which respectively bind to LFA-1 and CD2 present on thymocytes (14, 15, 16). Moreover, TEC-thymocyte interactions can be mediated by ECM ligands such as fibronectin and laminin and their corresponding integrin receptors VLA-4/VLA-5 and VLA-6 (17, 18, 19). In fact, it is possible that ECM provides a complex macromolecular substrate onto which thymocytes migrate within the organ, following an ordered pattern as if on a conveyor belt (20). More recently, biochemical and functional evidence was provided that TEC communicate with each other by gap junctions, which are formed by proteins of the connexin family and allow direct passage of low molecular weight substances between adjacent cells (21). Microinjection of low molecular weight fluorochromes also revealed functional gap junction-mediated TEC-thymocyte interactions. Based on further findings that have appeared in the literature, we recently postulated that gap junctions may correspond to a novel route for cell-cell communication in the immune system (22).

Thymic microenvironmental cells can influence thymocyte differentiation and proliferation by means of soluble polypeptides. Both TEC and DC produce the cytokine IL-1, which stimulates thymocyte proliferation (23). Actually, various cytokines can be produced by thymic epithelium, including IL-3, IL-6, IL-7, IL-8, granulocyte colony-stimulating factor, granulocyte-macrophage colony-stimulating factor, transforming growth factor-, transforming growth factor-ß (TGF-ß), leukemia inhibitory factor, and stem cell factor (24, 25).

IL-7, in particular, has been proven to be crucial for thymocyte differentiation. For example, it was shown to promote rearrangement of the TCR genes by enhancing the production and activity of recombinases (26). In conjunction, IL-7^-/- as well as IL-7 receptor-deficient mice display a severe reduction in lymphoid development, whereas the transgene incorporation of IL-7 in nude mice induces T cell development (27). In addition to IL-7, SCF (also termed c-Kit ligand) is necessary for early thymocyte differentiation. When fetal thymuses of spontaneously SCF-deficient mice are grafted into normal wild-type recipients, the number of CD3^-CD4^-CD8^- is more than 10 times lower than in wild-type grafts (27).

In addition to classical cytokines, chemokines are also thymic microenvironment-derived secretory products important in thymus physiology. Chemokines correspond to a family of small polypeptidic molecules that control directional migration of leukocytes (reviewed in Ref. 28). Among others, one chemokine, the stromal cell-derived factor (SDF) is highly expressed in the thymus, being produced by stromal cells, particularly in the subcapsular region (29). In keeping with this topography, SDF preferentially attracts immature CD4^-CD8^- and CD4⁺CD8⁺ thymocytes. Conversely, another chemokine, MIP3ß, exerts chemoattraction for mature single positive thymocytes (30). This is in keeping with the differential expression of corresponding chemokine receptors in distinct CD4/CD8-defined stages of thymocyte differentiation.

This leads to the notion that several paracrine circuits involving TEC-derived factors are likely to have differentiating thymocytes as targets. In addition to producing cytokines, TEC secrete chemically defined thymic hormones, including thymosin-1, thymopoietin, and thymulin (31, 32, 33), that can also act upon the general process of thymocyte maturation (reviewed in Refs. 34, 35, 36). For instance, thymulin, a nonapeptide whose biological activity depends on its coupling to zinc (37, 38), is able to enhance thymocyte proliferation and to induce several T cell markers and functions (reviewed in Refs. 34, 35, 36, 39). The circulating levels of thymulin achieve maximal values early in postnatal life and decline with age (40). More recently, it has been shown that thymulin secretion follows a circadian rhythm, peaking during the night (41). The general characteristics of thymulin and its effects on the immune system are summarized in Table 1. For recent reviews on the various thymic hormones, see Refs. 34, 35, 36 .

View larger version (36K): [in this window] [in a new window]   Figure 3. Heterogeneity of interactions between TEC and thymocytes. Left panel shows types of TEC-thymocyte interactions. Among membrane interactions are those involving the MHC complex (in this case MHC class II molecules) expressed by TEC, binding endogenous peptide presented to the TCR/CD3 complex on thymocytes, in the context of accessory molecules (exemplified herein by the CD4, labeled in blue). Additional cell-cell interactions, comprising receptors and coreceptors, are represented by adhesion molecules, namely LFA-3/CD2 and ICAM-1/LFA-1. Moreover, TEC can interact with thymocytes via ECM ligands and receptors (ECM-R), or gap junctions. In addition to membrane interactions, bidirectional TEC-thymocyte exchange of biological signals can be conveyed by secretory products such as cytokines and thymic hormones. Right panel illustrates that homotypic TEC-TEC interactions can also take place, involving adjacent cell membranes as well as soluble molecules.

Macrophages and DCs are hematopoietic-derived cell types and quantitatively are a minor component of the thymic microenvironment. Dendritic cells are preferentially located in the medulla and at the corticomedullary junction of the thymic lobules, whereas macrophages are distributed throughout the lobule (5); they both express MHC class II molecules and can interact with differentiating thymocytes via membrane proteins and cytokines, including IL-1.

Concerning the constitution of the intrathymically generated T cell repertoire, there is strong evidence that dendritic cells of the thymic microenvironment are involved in determining negative selection of the T cell repertoire by thymocyte deletion (13). In contrast, a role for macrophages in negative selection remains to be conclusively demonstrated.

C. Heterogeneity of the thymic epithelium: the thymic nurse cell complex
The thymic epithelial network is a rather heterogeneous tissue in terms of morphology and phenotype, and cells in different locations within the thymic lobules may be responsible for influencing specific steps in T cell maturation (46). One cortically located lymphoepithelial complex, the thymic nurse cell (TNC), has been isolated in vitro. TNCs are lymphoepithelial multicellular structures formed by one TEC, which in mice can harbor 20–200 thymocytes (47), and are located in the cortical region of thymic lobules (48, 49). Most intra-TNC lymphocytes bear the CD4⁺CD8⁺ double-positive phenotype (50), although immature double-negative as well as mature single-positive cells can also be found. Interestingly, TNCs may create special microenvironmental conditions for thymocyte differentiation and/or proliferation, and within this complex distinct interactions apparently occur, including those mediated by soluble products, gap junctions, ECM, and MHC/TCR (reviewed in Ref. 51). Self-antigens appear to be presented to thymocytes within TNC (52), and intra-TNC lymphocyte apoptosis has recently been reported (53, 54). Once settled in culture, TNCs spontaneously release thymocytes, and TNC-derived epithelial cells can reconstitute lymphoepithelial complexes after being cocultured with fetal thymocytes (55). Thus, TNCs constitute an in vitro model of thymocyte migration within the TEC context (18, 19, 45, 51, 56).

Other experimental models have been used to dissect the sequence of acquisition/loss of differentiation markers, as well as their respective roles in intrathymic T cell differentiation. A significant contribution was the generation of genetically engineered mice (57). The in vitro model of fetal thymus organ cultures (FTOC) is also used to study intrathymic T cell differentiation. By day 14 of gestation, only immature CD4^-CD8^- thymocytes are seen, whereas after a 14-day culture of the thymic lobes, differentiation has progressed with the generation of CD4⁺ or CD8⁺ single-positive mature cells (58).

As detailed below, the various intrathymic cellular interactions as well as the in vivo and in vitro experimental models summarized above can be regarded as potential targets for control by hormones and neuropeptides.

	III. Neuroendocrine Control of Membrane Interactions Between Thymocytes and Microenvironmental Cells

Top Abstract I. Introduction II. The Thymic Microenvironment... III. Neuroendocrine Control of... IV. Thymic Endocrine Function... V. Proliferation of Thymic... VI. Hormonal Modulation of... VII. Is Thymocyte Traffic... VIII. Expression of Receptors... IX. Intrathymic Expression of... X. Conclusions and Major... References

 
A. Is there a hormonal effect upon MHC expression by microenvironmental cells?
Few data are so far available to determine whether hormones or neuropeptides alter the expression of MHC gene products by thymic microenvironmental cells. Recent findings suggest that this may be the case. When human peripheral blood leukocytes were subjected to a mixed lymphocyte culture, the levels of cell proliferation and cytotoxic activity were significantly higher in the presence of recombinant GH, respectively 4- and 7-fold compared with controls (59). This is in keeping with previous evidence that GH treatment in children after renal transplantation worsened allograft function (60). In addition, a rise in IFN- secretion and in the number of cells expressing the corresponding mRNA was observed in vitro (6- to 10-fold more cells in GH-treated patients than in controls). In the same study, it was demonstrated that such effects require the presence of autologous antigen-presenting cells. Given the enhancing effect of GH upon IFN- production, together with the cytokine-induced up-regulation of MHC class I and class II expression (42, 43, 44) and the role played by adhesion molecules of thymic microenvironmental cells in MHC-TCR interactions (45, 55), we suggest that the MHC-mediated influence on thymocyte differentiation may be modulated by fluctuations in GH. This suggestion remains a working hypothesis and should be tested in specifically designed experiments.

More direct evidence, recently reported by Sacedon and co-workers (61, 62, 63), showed an effect of glucocorticoid hormones. Thymic dendritic cells treated in vitro with dexamethasone exhibited a slight, yet consistent, increase in the membrane expression of MHC class I molecules. This seems to be specific to class I, since no effect was observed on the expression of MHC class II gene products.

The same research group showed earlier expression of MHC class I and II molecules during thymus ontogeny in rat fetuses whose mothers had been previously adrenalectomized (52). It should be noted that before full development of the hypothalamus-pituitary-adrenal axis, circulating glucocorticoids in the fetus are strictly of maternal origin. Thus, what these experiments tell us is that absence of circulating glucocorticoids in early fetal life accelerates intrathymic MHC expression. Although in this particular work double labeling for simultaneous detection of MHC and cytokeratin was not reported, the micrographs, showing MHC labeling in the whole microenvironmental network, led us to think that the thymic epithelial network may include MHC expression. This idea is supported by the authors’ finding of earlier detection of other known markers for TEC differentiation in fetal thymuses derived from adrenalectomized mothers.

Thus, the findings discussed above argue in favor of hormonal regulation of MHC expression by the thymic microenvironment. Yet, formal demonstration of a hormonal regulation of MHC expression by TEC is still lacking. Nor has the possible influence of hormones and neuropeptides on intrathymic MHC expression been approached in terms of its consequences on MHC-TCR interactions.

B. Extracellular matrix-mediated TEC-thymocyte interactions are hormonally modulated
Initial studies revealed that the intrathymic production of ECM proteins, including fibronectin, laminin, and type IV collagen, was enhanced in vivo in mice injected with hydrocortisone; thickened ECM-containing fibrils were observed in both cortical and medullary regions of the thymic lobules as early as 24 h posttreatment. In the protocol of a single dose injection, this effect was transient, being progressively reversed in parallel with thymocyte expansion. Additionally, augmented amounts of such ECM components were detected in mouse TEC cultures treated with glucocorticoid hormone (64), indicating that the effect of hydrocortisone enhancing ECM in the thymus represents direct activity on the thymic epithelium. Similar results were obtained with sex steroids (64). At variance with these data, however, it was reported that the levels of fibronectin [measured by enzyme-linked immunosorbent assay (ELISA)] in human TEC culture supernatants were not altered after hydrocortisone treatment (65). Considering that this steroid hormone was also shown to enhance ECM receptor on the TEC membranes (18), it is possible that the lack of modulation in the supernatant derives from the augmented levels of the ECM bound on the TEC surface stimulated by hydrocortisone, which would then mask the levels in the culture supernatants.

In a further study, long-term treatment (30 days) with T₃ in mice also yielded changes in the intrathymic distribution profile of ECM proteins, with an increase in thin ECM fibrils (thus differing from the thick fiber pattern seen after glucocorticoid injection), particularly in the cortical region of thymic lobules. Again, such an effect seems to be direct upon TEC, since enhanced ECM production was also seen when T₃ was added to cultures of a murine TEC line and to TNC-derived TEC preparations (66). More recently, similar results were obtained in vitro when various TEC cultures were subjected to PRL, GH, or insulin-like growth factor I (IGF-I) (56). It should be pointed out that, regarding the in vitro models mentioned above, not only were the amounts of fibronectin and laminin enhanced by various hormone treatments, but also was the expression of their corresponding receptors, VLA-5 and VLA-6 (56, 66).

Since thymocyte-TEC adhesion is at least partially mediated by ECM ligands and receptors, we also tested the various hormones cited above, corticosteroid, thyroid, and pituitary hormones, for their ability to modulate such heterotypic cellular interaction. All enhanced the degree of thymocyte adhesion to cultured TEC. Furthermore, regarding pituitary hormones (56), the hormone-induced enhancement of TEC-thymocyte adhesion was abrogated by monoclonal antibodies specific for each hormone or its corresponding receptor, and also by various anti-ECM or anti-ECM receptor antibodies (Fig. 4).

View larger version (36K): [in this window] [in a new window]   Figure 4. ECM-mediated, GH-induced enhancement of thymocyte adhesion to TEC. Freshly isolated mouse thymocytes are left to adhere onto cultures of a mouse TEC line in the absence or presence of GH. Additionally, some GH-treated cultures are incubated with antifibronectin (-FN), anti-VLA5 (-VLA5), antilaminin (-LN), or anti-VLA6 (-VLA-6) monoclonal antibodies before GH treatment. The GH-induced enhancement of thymocyte-TEC adhesion is clearly abrogated by anti-ECM or anti-ECM receptor antibodies. **, P < 0.01. [Adapted with permission from V. de Mello-Coelho et al.: J Neuroimmunol 76:39–49, 1997 (56 ) © Elsevier Science.]

C. Are inter-TEC gap junctions under neuroendocrine control?
Very recent data indicate that gap junctions mediating communication between adjacent TEC can also be under hormonal control. Decreased cell coupling between adjacent TEC (ascertained by diffusion of intracellularly injected low molecular weight fluorochrome) was seen when cultures were treated by the sex steroids testosterone, progesterone, and estrogen, as well as by the pituitary hormones ACTH and GH, and the neuropeptide calcitonin gene-related peptide (CGRP) and substance Y (67, 68). Conversely, results from our laboratory indicate that vasointestinal peptide (VIP), a neuropeptide that increases intracellular cAMP, enhances inter-TEC cell coupling (69). Similar enhancement was seen when cultures were treated with glucocorticoid hormones, an effect that was significantly abrogated by the use of the glucocorticoid receptor antagonist RU486 (70). Nevertheless, a systematic survey to determine which hormones and neuropeptides modulate gap junction opening between adjacent TEC and between TEC and thymocytes has not yet been made. Nor are data so far available concerning putative neuroendocrine control of connexin expression by TEC.

	IV. Thymic Endocrine Function and Cytokine Secretion by Microenvironmental Cells Are Controlled by Hormones and Neuropeptides: The Paradigm of Thymulin

Top Abstract I. Introduction II. The Thymic Microenvironment... III. Neuroendocrine Control of... IV. Thymic Endocrine Function... V. Proliferation of Thymic... VI. Hormonal Modulation of... VII. Is Thymocyte Traffic... VIII. Expression of Receptors... IX. Intrathymic Expression of... X. Conclusions and Major... References

 
A. Steroid and peptidic hormones influence in vitro cytokine secretion
In addition to acting upon membrane interactions, hormones and neuropeptides can modulate the production of cytokines and hormones by thymic microenvironmental cells. Exogenous IL-1ß enhances IL-6 and IL-8 production by cultured human TEC and induces granulocyte macrophage colony stimulating factor (GM-CSF) as well as leukemia inhibitory factor (LIF) production by these cells. Interestingly, hydrocortisone is able to selectively block the latter effect on GM-CSF and LIF, as ascertained by ELISA assessment of cytokines in TEC culture supernatants (65). In a mouse TEC line, retinoic acid promoted enhanced expression of IL-1, IL-6, and IL-7 (71). The production of IL-1 and IL-1ß by bovine nonepithelial thymic microenvironmental cells in vitro was increased by exogenous GH and by PRL (72). It was further shown that secretion of IL-6 is also up-regulated by GH or by PRL treatment, an effect that could be abrogated by the use of the IL-1 receptor antagonist. This indicates that hormonal effects on microenvironmental cell-derived IL-6 secretion are at least partially exerted through the IL-1 production pathway. Autocrine/paracrine control of cytokine production by the thymic microenvironment appears to involve TEC-derived neuropeptides as well. Martens and colleagues (73) demonstrated that the constitutive production of both IL-6 and LIF (but not IL-1ß) by primary cultures of human TEC was enhanced when monoclonal antibodies to oxytocin were added to the cultures. This strongly suggests that the secretion of these TEC-derived cytokines is partially under the control of oxytocin.

Together, the findings discussed above strongly favor the notion that cytokine production by the thymic microenvironment is under neuroendocrine control. Nonetheless, such a notion should be cautiously viewed since corresponding in vivo data are still lacking. In this respect, evaluation of cytokine production by the microenvironmental cells in mice genetically engineered for hyperproduction or lack of hormones or neuropeptides will be certainly worthwhile. The same is true for more precise analysis of IL-7 and of SCF production by thymic microenvironmental cells under various hormonal fluctuations.

B. Thyroid and pituitary hormone status modulates thymulin secretion
The concept of neuroendocrine control of thymic secretory substances has been particularly developed with regard to the production of the thymic hormone thymulin. In a first series of studies we demonstrated that in vivo treatment with T₃ enhanced thymulin secretion by mouse TEC, and that an opposite effect was seen if the animals were treated with propiothiouracil, an inhibitor of thyroid hormone synthesis (74). These results were confirmed by others (75). In aging mice, injection of T₄ increased thymulin serum levels to values found in young individuals (75). In humans, it was shown that patients with hyperthyroidism exhibit higher levels of circulating thymulin, whereas in hypothyroidism the opposite was observed. In both situations, adequate therapy brought thymulin serum levels within normal range (76).

Experiments using TEC cultures showed that the stimulatory effect of thyroid hormones upon thymulin secretion was due to a direct action of the hormone on the epithelial cells (66) and depended on de novo synthesis of thymulin, since it could be abrogated by cycloheximide (77).

Pituitary hormones were also shown to be potent up-regulators of thymulin secretion. For instance, experimental hyperprolactinemia induced by repeated PRL injections increased thymulin levels in both young and old animals. Conversely, administration of bromocriptine, an agonist of the dopamine receptor, which inhibits PRL biosynthesis, promoted a consistent, dose-dependent decrease in thymulin production (78). These results, obtained in mice, are in keeping with data derived from hyperprolactinemic patients bearing pituitary adenomas, who present abnormally high thymulin serum levels (79).

Fluctuations in GH levels also modulate thymulin secretion. Initial studies revealed that dwarf mice exhibit a precocious decay in thymulin levels (80). This is in keeping with more recent data showing that hypophysectomy in rats yielded a profound, although transient, decrease in thymulin serum levels (81). Similarly, low thymulin levels accompanied deficient GH production in children, whereas GH treatment consistently restored this thymic endocrine function, as early as 24 h after injection (82, 83). Moreover, GH treatment induced an increase in thymulin serum levels that correlated with the amounts of circulating IGF-I. In acromegalic patients, high thymulin serum titers also correlated with high IGF-I serum levels (84).

It is noteworthy that the enhancing effects of PRL and GH on thymulin secretion were directly obtained by treating murine or human TEC cultures and were abrogated by corresponding antihormone antibodies (78, 84). In vitro GH effects were abrogated by anti-IGF-I or anti-IGF-I receptor antibodies, thus incriminating TEC-derived IGF-I as a mediator of the GH effects upon TEC (84). This was further supported by findings that both anti-IGF-I and anti-IGF-I receptor antibodies were also able to block GH-dependent enhancement of TEC-thymocyte adhesion (56).

In addition to the effects observed with classical peptidic hormones, we found that endogenous opioids, namely ß-endorphin and Leu-enkephalin, can up-regulate thymulin secretion by cultured TEC (85). In vivo experiments to confirm this notion are still lacking, however. Nor are data available to define whether other neuropeptides influence thymulin secretion.

C. Effects of adrenalectomy and gonadectomy on thymulin levels
The effects of adrenal and gonadal steroids on thymulin secretion appear to be rather more complex. One of the experimental strategies to approach this issue is ablation of the adrenals and/or gonads. When a single surgery (adrenalectomy or gonadectomy) was performed, we observed in both male and female mice a transient fall in thymulin serum levels that lasted 1 month, peaking 1 week postsurgery, and then progressively returning to normal. More impressive, adrenalectomy + castration resulted in a long-term decrease in the levels of circulating thymulin that persisted until 3 months postsurgery, being followed by gradual restoration to normal levels by 6 months after surgery (86). Interestingly, in both single and double surgical procedures, an increase in the intrathymic content of thymulin was seen. Concomitant to the sustained low levels of the thymic hormone in these experimental conditions, an endogenous low molecular weight thymulin inhibitor was transiently detected in the mouse sera, with concentrations peaking when thymulin levels were lowest. Although the biochemical nature of such a thymulin inhibitor was not defined, its appearance was thymus dependent, since it was not found in mice undergoing thymectomy before adrenalectomy + gonadectomy (86). This series of experiments, although not conclusive, pointed to a rather complex mechanism involved in in vivo steroid hormone influence upon thymic hormone production, possibly comprising other biological circuits, including the hypothalamus-pituitary axis.

This possibility led us to study the influence of steroid hormones on murine and human TEC cultures, and we observed that physiological concentrations of glucocorticoid hormones, estradiol, progesterone, or testosterone, enhanced thymulin release into the culture supernatants. This effect was abrogated when TEC were simultaneously incubated with a given steroid hormone plus the specific antagonist of the corresponding hormone receptor (87).

In view of these findings, it is possible that the transient in vivo increase in the intrathymic contents of thymulin observed after adrenalectomy and/or gonadectomy corresponds to a TEC response to the fall in circulating levels of biologically active thymulin, secondary to the appearance of its natural inhibitor. Such a feedback circuit with increase in thymulin production had been previously shown in mice treated with antithymulin monoclonal antibodies (88).

D. Is there an autocrine/paracrine circuitry controlling thymulin secretion?
That in vivo treatment of mice with antithymulin monoclonal antibodies resulted in an increase in the intrathymic content of the hormone suggested that the level of circulating thymulin could influence its rate of secretion. Similarly, in keeping with these findings, we noted that incubation of cultured TEC with antithymulin monoclonal antibodies resulted in increased numbers of thymulin-containing cells (89). Conversely, thymulin release into TEC culture supernatants can be down-regulated by exogenous addition of the hormone itself.

Interestingly, IL-1, which is also produced in vivo and in vitro by TEC, is able to stimulate in vivo zinc uptake by the thymus (likely to be due to the increase in metallothionein biosynthesis by the epithelial cells), thus up-regulating thymulin secretion in vitro (90, 91).

Altogether, the data discussed above clearly indicate that the general control of thymulin secretion may be very complex, involving distinct biological circuits whose overall balance will dictate the amounts of thymulin to be secreted at a given moment. Due to this apparent complexity, it is predictable that compensatory loops may be triggered when one or more thymulin-controlling axes are disturbed. This would explain why in some experimental situations, fluctuations in thymulin serum levels are transient.

E. Thymic hormones modulate endocrine glands and neuroendocrine circuits
The concept of bidirectionality between the neuroendocrine and immune systems can also be applied to analysis of thymic hormones, since these substances modulate the production of hormones and neuropeptides of the hypothalamus-pituitary axis and some of their target endocrine glands.

Initial experiments revealed that neonatal thymectomy promotes developmental atrophy of female sexual organs (92). One might argue that such an effect could reflect an autoimmune process rather than a direct action of thymic hormones on the neuroendocrine system. This assumption is based on the fact that perinatal thymectomy in BALB/c mice induces autoimmune disease (93). Nevertheless, in the experiments supporting this view, thymectomy was performed on day 3 postnatally, when the thymus has already released a significant amount of thymocytes (initiated on the day of birth). By contrast, in the former experiments, thymectomy was carried out at birth, thus before colonization of peripheral lymphoid organs by T cells. Moreover, as detailed below, it has been shown that production of sex steroids is enhanced in vivo and in vitro by a single thymic hormone, thymulin.

In addition to the action on sexual organs, thymectomy at birth promoted a decrease in the number of secretory granules in acidophilic cells of the adenopituitary (94). This is in keeping with data showing that athymic nude mice exhibit significantly low levels of various pituitary hormones, including PRL, GH, LH, and FSH (95).

Regarding the effects of thymic peptides, it was shown that thymosin-ß4, when perfused intraventricularly, stimulates in vivo LH and its hypothalamic-releasing hormone LHRH (96). A similar stimulation of LH release was obtained with thymulin in perifused or fragmented pituitary preparations (97, 98). Another thymosin component, the MB-35 peptide, also enhanced PRL and GH production (99). In vivo studies in children showed that administration of thymopoietin increases GH and cortisol serum levels (100). Moreover, thymopentin (the synthetic biologically active peptide of thymopoietin) enhanced in vitro the production of POMC derivatives such as ACTH, ß-endorphin, and ß-lipotropin (101). Thymulin exhibited a similar in vitro stimulatory effect on perifused rat pituitaries, enhancing the release of GH, PRL, and, to a lesser extent, TSH (102). With regard to its effect on GH release, it has been shown to be age-dependent, being less efficient in pituitary cell cultures derived from senescent animals (103). The same study showed that this effect of thymulin is mediated by calcium influx, as well as cAMP and inositol phosphate (103). However, contrasting results were reported using short-term cultures of pituitary fragments: a consistent increase in ACTH secretion after in vitro thymulin treatment, with no changes in GH levels but a significant inhibition of PRL release (98).

Thymosin-1 was apparently able to down-regulate TSH, ACTH, and PRL secretion in vivo, although effects on GH levels were not detected (104). Interestingly, these inhibitory effects seem to occur through hypothalamic pathways, since production of corresponding releasing hormones by hypothalamic neurons was also decreased after in vitro treatment of medial basal hypothalamic fragments with thymosin-1 (105).

In addition to affecting the hypothalamus-pituitary axis, thymic hormones may act directly on its target endocrine glands (Fig. 5). In vitro experiments showed that thymulin can modulate gonadal tissues. Proliferation of oogonia from fetal rat ovaries, as well as gonocytes from fetal rat testicles, was consistently increased in the presence of thymulin (106, 107, 108). At least regarding the expansion of male germ cells in the same culture system, thymulin-inducing proliferative effects were largely prevented by TGF-ß1 (109).

View larger version (26K): [in this window] [in a new window]   Figure 5. Effects of thymulin on endocrine glands, as exemplified by the stimulatory effects on PRL, GH, TSH, and LH secretion by pituitary cells, as well as gonocyte proliferation.

It should be pointed out that the existence of direct effects of thymic hormones upon other endocrine glands that are physiological targets of the hypothalamus-pituitary axis, such as thyroid and adrenals, has not yet been studied.

Recent work has shown that thymulin can also modulate some peripheral nervous sensory functions, such as those related to pain sensitivity. In vivo injections of thymulin at high doses significantly reduced the hyperalgesia (related to both mechanical and thermal nociceptors) induced by intraplantar injection of endotoxin in rats and mice (113). Interestingly, when applied at much lower doses, this peptide instead generated hyperalgesia, an effect paralleled by a significant enhancement in the intrahepatic production of IL-1ß (114). Such paradoxical effects are in keeping with previous data showing that low doses of thymulin enhance IL-1ß secretion by peripheral blood cells, whereas, at high concentrations, thymulin suppresses its release as well as that of IL-2, IL-6, and TNF- (115). More recently, the cellular and molecular mechanisms involved in the thymulin-induced hyperalgesic effect have been further investigated. In the peripheral nervous system, the involvement of capsaicin-sensitive primary afferent neurons has been revealed. Intraperitoneal injection of capsaicin (known to destroy afferent nervous fibers) significantly abrogated the stimulatory thymulin effect on pain (116). Additionally, spinal cord neurons appear to be involved, since thymulin induces sustained expression of c-fos (a marker of spinal cord neuron activation) in those neurons known to be involved in nociception (117). It should be noted, however, that, in spite of these convincing data, the molecular basis for thymulin action on neurons is not complete, since thymulin receptors in neurons have not yet been determined.

Some data also indicate that other thymic hormones may exert a modulatory role in the central nervous tissue, including an effect on behavioral functions. In vivo injection of a thymopentin analog in rats was shown to counteract the stress response to experimentally induced social defeat (118), as measured in the elevated plus-maze apparatus, a recognized animal model of anxiety (119). Although the mechanism(s) involved in this thymopentin-mediated event were not characterized, direct action of the thymic hormone analog on the cholinergic innervation of the hypothalamus with consequent inhibition of CRF release was suggested (118). Such a putative direct anxiolytic effect of the thymopentin analog is further supported by previous data showing that injections of thymopentin normalize the numbers of benzodiazepine and -aminoisobutyric acid receptors in the hippocampus after stress (120). It is also noteworthy that neonatal thymectomy modulates the densities of nicotinic cholinergic receptors in skeletal muscle and brain (121).

For further details on the relationship between thymic hormones and the neuroendocrine system, including behavioral adaptive responses, see Refs. 122, 123 .

	V. Proliferation of Thymic Cells Is Hormonally Influenced

Top Abstract I. Introduction II. The Thymic Microenvironment... III. Neuroendocrine Control of... IV. Thymic Endocrine Function... V. Proliferation of Thymic... VI. Hormonal Modulation of... VII. Is Thymocyte Traffic... VIII. Expression of Receptors... IX. Intrathymic Expression of... X. Conclusions and Major... References

 
A. In vitro effects of pituitary hormones on TEC growth
One potentially relevant factor for better understanding thymus ontogeny is whether the expansion of TEC can be under neuroendocrine control. Pituitary hormones, such as PRL and GH, as well as IGF-I (78, 84), consistently enhance TEC proliferation in vitro. Similar results were obtained by such various methodological strategies as direct cell counting, [³H]thymidine incorporation, and immunofluorescence analysis of cultured TEC after bromodeoxyuridine uptake (Fig. 5, top panel). The proliferative effects of GH and IGF-I were recently confirmed using TEC lines derived from normal and from thymomatous rat thymus (124). Interestingly, the GH releasing inhibitor somatostatin, as well as its analog octreotide, significantly blocks human cultured TEC proliferation, as ascertained by [³H]thymidine incorporation (125).

A recent work suggested that TEC growth might also be under the control of thymic hormones. It was shown that thymopentin induces DNA synthesis in human TEC lines, either alone or in conjunction with FCS (126). Further studies using the complete thymopoietin molecule, in conjunction with similar assays using other thymic hormones, are necessary, however, to establish the concept of thymic hormone control of TEC growth.

Data are also scarce concerning the effects of steroid hormones on TEC proliferation. An apparent decrease in the proliferation rate was seen in a rat TEC line after treatment with progesterone, estrogens, or androgens, an effect probably mediated by protein kinase C (127). A similar inhibitory effect was seen with cortisol (128).

Much less is known about the in vivo effect of these hormones on TEC growth. However, it was shown that injections of metaclopramide, which promotes hyperprolactinemia, increased the number of solid epithelial islands in adult rat thymuses (129).

In vivo, Scheiff and co-workers (130) provided morphometric evidence that thyroid hormones were also able to induce TEC proliferation. Nevertheless, we and others observed no significant in vitro growth effect of T₃ using the model of a murine TEC line (66, 131). The reason for these apparently contrasting results is likely to lie in the distinct evaluation methodologies as well as the in vivo vs. in vitro situations.

B. Modulation of thymocyte proliferation by hormones and neuropeptides
In vitro thymocyte proliferation can be stimulated by supernatants of TEC cultures, whereas supernatants derived from fibroblast cultures have no effect. This thymocyte proliferative activity of TEC-derived supernatants was almost completely abolished by the presence of antithymulin monoclonal antibodies, but was enhanced when TEC were treated with T₃. By contrast, in the same study, T₃ did not directly influence thymocyte proliferation, as ascertained by short-term thymidine incorporation (132).

More recently, similar proliferative effects were seen using supernatants from PRL- or GH- treated TEC cultures (133, 134). Additionally, as depicted in Fig. 6, bottom panel, GH itself synergized with anti-CD3 in its stimulatory effect on thymocyte proliferation (135). This is in keeping with recent data showing that transgenic mice that overexpress GH or GH releasing hormone exhibit overgrowth of the thymus (136).

View larger version (56K): [in this window] [in a new window]   Figure 6. In vitro modulation of TEC and thymocyte proliferation by hormones. In the top panel, the stimulatory effect of PRL on human TEC growth is ascertained by direct cell counting, [3H]thymidine incorporation, and bromodeoxyuridine uptake, revealed by immunocytochemistry. The comitogenic effect of GH on thymocyte proliferation is seen in the bottom panel. In this experiment, mitogenic response was induced with anti-CD3 monoclonal antibody, and GH was added at various molar concentrations. Asterisks indicate statistically higher thymocyte expansion induced by anti-CD3 in the presence of GH. [Derived from Refs. 78 and 135.]

Further hormonal mediators are involved in the general control of thymocyte proliferation. For example, mitogen-induced human thymocyte expansion was blocked in the presence of anti-LH-RH serum, suggesting that intrathymically produced LH-RH may act as a costimulatory factor for thymocyte growth (140). At least in some experimental conditions, PRL also appears to be effective in directly stimulating thymocyte proliferation by enhancing IL-2 production and IL-2 receptor expression (141). This is keeping with data suggesting that PRL may play the role of a T cell growth factor, since it induces gene expression of cyclins D2 and D3 in the rat thymic lymphoma cell line Nb2 (142).

Proliferation of cultured human thymocytes was also directly stimulated in vitro by met-enkephalin, whereas VIP promoted an inhibitory effect, as compared with control untreated cultures, in both conditions of spontaneous and of phytohemagglutinin-induced mitogenesis (143). Similarly, somatostatin prevented concanavalin A-induced rat thymocyte proliferation (144). Another neuropeptide, CGRP, also inhibited mitogen-induced thymocyte proliferation, an effect that was abrogated in the presence of its antagonist CGRP8–37 (145).

Using the model of fetal thymus organ cultures, we recently demonstrated that expansion of thymocytes could be stimulated by insulin (146). Yet, since this is a complex heterocellular model, it remains to be determined whether the proliferative insulin effects are direct on the thymocytes or are mediated through microenvironmental cells. Tables 2 and 3 summarize the vast series of in vitro experiments concerning modulation of TEC or thymocyte proliferation by hormones and neuropeptides.

Injection of T₃ also promoted a consistent increase in thymocyte numbers with enhancement of spontaneous ex vivo [H³]thymidine incorporation (66).

Administration of synthetic TRH was also shown to enhance bromodeoxyuridine uptake by thymic cell suspensions (152). More recently, the same group showed in the rat model that such an effect could be seen in vivo, provided TRH was injected in the morning. In addition to confirming the previous data, these results reflect a time of day dependency for a physiological effect of TRH on thymocyte proliferation (153). Female rats continuously treated with leuprolide, an LH-RH agonist, exhibited a consistent increase in thymus weight (154), while in vivo injections of met-enkephalin in mice also enhanced thymus weight and cellularity (155). The latter finding is in keeping with the above in vitro stimulatory effect of this opioid on thymocyte proliferation.

Finally, intrathymic implants of the pineal gland derived from young mice, into age-matched recipients, led to a remarkable long-term maintenance of thymic size and the cortico-medullary architecture in the latter, thus preventing physiological age-related thymus atrophy (156). This suggested an effect of melatonin, either favoring proliferation and/or partially preventing apoptosis. In keeping with these findings, melatonin injection partially prevented stress-induced thymic atrophy (157). More recently, the mechanism by which melatonin could exert such an antiapoptotic role was revealed, with the demonstration that melatonin down-regulates the expression of glucocorticoid receptor (158).

By contrast, hormones such as sex steroids appear to provide an inhibitory tonus on thymocyte proliferation (reviewed in Ref. 159). Castration in young adult male mice promoted a rapid wave of thymocyte proliferation in vivo, particularly in cortically located cells bearing the immature phenotypes CD4^-CD8^- and CD4⁺CD8⁺ (160). This is in keeping with previous findings that castration in old rats led to enhancement of thymus weight, an effect that was abrogated by androgen treatment of castrated animals (161).

Taken together, the data summarized above clearly indicate that distinct hormones and neuropeptides can convey positive and negative signals for thymocyte proliferation.

	VI. Hormonal Modulation of Intrathymic T Cell Differentiation

Top Abstract I. Introduction II. The Thymic Microenvironment... III. Neuroendocrine Control of... IV. Thymic Endocrine Function... V. Proliferation of Thymic... VI. Hormonal Modulation of... VII. Is Thymocyte Traffic... VIII. Expression of Receptors... IX. Intrathymic Expression of... X. Conclusions and Major... References

 
A. Thymocyte-derived cytokine profile
One thymocyte function influenced by neuroendocrine circuits is the cytokine profile produced by these cells. For instance, concanavalin-A mitogenic response and IL-6 production were enhanced in thymocytes from GH-treated aging animals (162), and PRL induced IL-2 production by thymocytes (141). Vasopressin can replace IFN- in its role of inducing IL-2 production (163), and substance P stimulated IL-2 synthesis by the mouse thymic lymphoma cell line EL-4 (164). By contrast, VIP, PACAP27, and PACAP38 exerted an in vitro inhibitory action on the production of some thymocyte-derived cytokines, including IL-2, IL-4, and IL-10 (165, 166). Altogether, these findings provide consistent evidence for a hormone/neuropeptide balance in the control of cytokine production by thymocytes.

B. Changes in the T cell differentiation markers and TCR Vß repertoire
As detailed above, hormones and neuropeptides can affect thymic functions related to thymocyte differentiation. Thus, such differentiation could also be a target for neuroendocrine control. Implants of GH3 pituitary cells in aging rats increased total thymocyte numbers and the percentage of CD3-bearing cells, with a parallel decrease in the CD4^-CD8^- double-negative thymocytes, which normally accumulate in the aging rat thymus (147, 167). The role of GH in thymus development was further supported by findings in GH-deficient dwarf mice. In addition to the precocious decline in thymulin serum values (80), there was progressive thymic hypoplasia with decreased numbers of CD4⁺CD8⁺ double-positive thymocytes. Such defects could be restored by prolonged treatment with GH (168).

Injections of T₃ promoted an increase, both in relative and absolute numbers of thymocytes bearing the CD44 marker (66), which is an ECM receptor of the proteoglycan family, with specificities for hyaluronate and to a lesser extent fibronectin and collagen. By contrast, administration of high doses of glucocorticoid hormones yielded a profound decrease in the percentages of CD4⁺CD8⁺ thymocytes, with a relative increase in CD4^-CD8^- as well as CD4⁺CD8^- and CD4^-CD8⁺ cells (169, 170). In vivo treatment of mice with estradiol also promoted a depletion in the absolute numbers of CD4⁺CD8⁺, CD4⁺CD8^-, and CD4^-CD8⁺ thymocytes, with a decrease in the proportion of double-positive cells and an increase in the percentage of double-negative as well as single-positive mature cells (171). A striking loss of the very immature CD3^-CD4^-CD8^- cells was further demonstrated (172). Interestingly, the estrogen-induced thymic involution appeared to occur independently of glucocorticoids, since it was seen in adrenalectomized animals (170). In this same study, the proportions of CD4⁺CD8⁺ thymocytes in adrenalectomized estrogen-treated mice, although significantly lower than in control animals, were higher than those in hormone-treated mice not subjected to adrenalectomy.

Although considerable work is available concerning the neuroendocrine control of T cell differentiation markers, the influence of hormones and neuropeptides on shaping the intrathymically generated T cell repertoire remains poorly studied. However, the few data available point to such an influence. Mice treated with estradiol exhibited a selective increase in the percentages of CD4^-CD8^- TCR⁺ thymocytes expressing Vß6, Vß8, or Vß11 but not Vß3 gene products (173), thus promoting an imbalance in the generation of the TCR repertoire of the double-negative TCR⁺ cell lineage. Interestingly, an enhancement of IL-1 mRNA was seen in parallel with the increase in Vß8⁺ cells, in keeping with previous data showing that this cytokine exerts a mitogenic effect on Vß8⁺ thymocytes (173). More recently, release of autoreactive T cells bearing the Vß3 or Vß11 phenotypes, with autoreactivity to hepatocytes, was seen in mice injected with a single dose of estradiol (174).

In rat lymphoma-derived Nb2 cells, PRL induced in vitro gene expression of the TCR chain, whereas the TCR ß chain gene was suppressed (175), suggesting that the intrathymic PRL content may drive T-cell differentiation pathways. In keeping with this hypothesis, it has been demonstrated, not only in the Nb2 cell line but also in human thymocytes, that in the absence of TCR ligation PRL induces rapid phosphorylation of multiple TCR/CD3 complex proteins, including the CD3 -chain and the ZAP-70 tyrosine kinase, both essential for TCR function (176). However, recent data argue against the above hypothesis, since PRL receptor knockout mice apparently develop a normal thymocyte differentiation pathway, at least in terms of CD3, CD4, and CD8 markers (177, 178).

As detailed below, very elegant data strongly suggest that intrathymically produced glucocorticoids influence the generation of the T cell repertoire in the thymus by modulating positive and negative selection of thymocytes.

C. Hormone-mediated apoptosis in thymocytes
One of the best studied effects concerning the hormonal control of intrathymic cell death is that mediated through glucocorticoid hormones. When applied in relatively high doses, these substances trigger an apoptotic cell machinery in differentiating thymocytes, particularly cells bearing the immature CD4⁺CD8⁺ phenotype. The molecular basis of such a glucocorticoid hormone-induced biological response is now known in detail and is used as a paradigm for studies of apoptosis in various cell types (reviewed in Ref. 179). Briefly, glucocorticoids activate calcium-dependent endonucleases that eventually cleave DNA, with the formation of oligo-nucleosomes. A series of findings suggest that such events are potentiated by cAMP (180) and depend on the recently isolated transcription factor SRG3 (181). They do not require p53 tumo