Glucocorticoid Production and Regulation in Thymus: Of Mice and Birds

The thymus is the key immunological organ for the maturation of T cells in mammals. Enlargement of the thymus gland in patients with Addison’s disease and in adrenalectomized rats was recognized a century ago (1,2). Elevation of glucocorticoids (GC) due to chronic stress or therapeutic administration causes involution of the thymus. T cells, especially immature thymocytes, are particularly sensitive to apoptosis induced by GC (3,4). GC act through the GC receptor (GR), a member of the steroid receptor superfamily of ligand-activated transcription factors. Upon binding to GC, GR located in the cytosol are translocated to the nucleus where they exert transcriptional effects. They also modulate gene transcription indirectly, without binding to DNA, by interacting with other transcription factors, including activation protein-1, nuclear factor-κB, and signal transducer and activator of transcription proteins (5,6). This type of protein-protein interaction mediates the cross talk important for the regulation of the immune system (5,7).

GC are strong inducers of apoptosis in thymocytes and play a significant role in the development, differentiation, homeostasis, and function of T cells (6,8). Immature double-negative thymocytes (CD4⁻ CD8⁻) proliferate and differentiate in the thymus, undergoing extensive genetic and phenotypic alteration to generate a double-positive (CD4⁺ CD8⁺) cell population. Most CD4⁺ CD8⁺ thymocytes undergo apoptosis; the surviving double-positive cells differentiate into single-positive CD4⁺ or CD8⁺ cells that populate the peripheral lymphoid tissues (4,9,10,11). According to the mutual antagonism model, thymic selection is controlled by the cross talk of activation-induced and GC-dependent cell death of double-positive thymocytes (4,11). Studies using transgenic and knockout (KO) models addressing GR function clearly demonstrate GR-induced apoptosis but have been equivocal in addressing the role of GC in T cell development (4,10,12).

GC are produced primarily in the adrenal gland but are also produced in other organs including the brain (13,14), intestinal tract (15), skin (16,17,18), and thymic epithelial and immune cells (19,20,21,22,23,24,25) and express the necessary steroidogenic enzymes for the synthesis of GC, which apparently act in an autocrine or paracrine fashion. The thymus has endocrine properties and expresses various hormones and receptors of the hypothalamic-pituitary-adrenal axis, corticotropin-releasing factor (26), ACTH (27), and ACTH receptors (28,29) including melanocortin receptor subtype 2 (MC2R) and MC5R (30) in thymus or T cells (28,31). Thymus epithelial cells and thymocytes express mRNA for all the necessary steroidogenic components including the steroidogenic acute regulator (StAR), CYP11A1, 3β-hydroxysteroid dehydrogenase, CYP21, and CYP11B1 enzymes and can synthesize GC (19,20,21,22,32,33,34,35,36,37). StAR, CYP11A1, and CYP11B1 are expressed in both thymus epithelial cells and thymocytes of 4-wk-old mice; at 14 wk, they are significantly increased in thymocytes but decreased in thymus epithelial cells. The CYP17 enzyme is expressed at very low levels in both the thymus and adrenal gland of mice (22). Measurement of individual enzymatic activity with exogenous substrate demonstrated that the enzymes are functional (19,21). However, because the availability of substrate relative to the kinetic requirements for optimal functioning of the enzymes and mRNA for steroidogenic enzymes in comparison with the adrenal are both very low, the inherent synthesis of GC by thymocytes was not certain. The de novo synthesis of GC was elegantly demonstrated using a reporter gene assay in COS cells transfected with the cDNAs for the GR and a GC response element-driven luciferase reporter coincubated with thymocytes. Use of inhibitors of the various enzymes of steroidogenesis trilostane or metyrapone clearly demonstrated that thymocytes synthesize GC using endogenous substrate (20,22,37). The functional significance of the synthesis of GC within the thymus in vivo has been difficult to demonstrate.

Studies of mice with a thymus-specific inducible GR transgene demonstrated that there was a significant reduction in thymocyte number in these mice after overexpression of GR was induced after adrenalectomy in comparison with noninduced adrenalectomized transgenic animals. The reduction in thymocyte number in the thymus GR-induced adrenalectomized mice was prevented by the administration of a GR antagonist (36). These studies clearly demonstrated the functional importance of locally produced GC.

The report by Qiao et al. (37) in the current issue addresses the regulation of the thymic GC synthesis in an innovative way. Adrenalectomy results in the lack of feedback inhibition of CRH and ACTH production by GC and a dramatic increase in plasma ACTH. One might predict that this increase in ACTH would stimulate thymocyte steroidogenesis, as it does in the adrenal gland and, if the system were functional, would cause thymus involution. However, as was known, thymus size increased significantly, suggesting that GC synthesis in the thymus does not have a significant regulatory role in thymocyte function. They found that stimulation with ACTH and cAMP, the second messenger in the ACTH signal cascade, produced a significant down-regulation of the CYP11B1 mRNA expression and GC synthesis in thymocytes. This effect was a direct result of ACTH, because thymocyte numbers were not increased by adrenalectomy in IL-1β/ IL18 double-KO mice in which the increase in plasma ACTH is significantly less than in wild-type mice. Administration of ACTH to IL-1β/IL-18 double-KO mice increased thymocyte numbers, indicating a direct effect of ACTH in thymocytes either mediated by the down-regulation of the steroidogenic enzymes or a direct effect of ACTH unrelated to steroidogenesis. In contrast, ACTH stimulates both the expression of steroidogenic enzymes and proliferation of the steroidogenic adrenal cells. The explanation of this paradox is not clear. Both MC2R and MC5R are expressed in adrenal steroidogenic cells and thymocytes. Although MC2R is the predominant ACTH receptor in the adrenal gland (30,31), the relative expression and function of each of these receptors in thymocytes is unknown.

Birds also have extraadrenal GC synthesis in organs of the immune system. The requisite enzymes for the de novo synthesis of corticosterone and cortisol are expressed in the thymus and bursa of Fabricius of birds and GC synthesis occurs in these organs as well as the adrenal (23,24,25,34). Corticosterone is the most abundant GC produced by the adrenal and in the circulation, although cortisol is also produced in small amounts. Cortisol concentrations in the thymus and the bursa of Fabricius of the captive zebra finch is greater than in plasma, with GC concentrations in the immune organs decreasing with age. This finding is not uniform, because concentrations of corticosterone and cortisol were very low in the thymus and bursa of wild European starlings in comparison with captive zebra finches (25). Whether this is related to species differences of stress is not clear.

In summary, the thymus of mice and thymus and bursa of birds express the same elements of steroidogenesis and steroidogenic control by the hypothalamic-pituitary axis that the adrenal zona fasciculata does, suggesting that local extraadrenal synthesis of GC modulates immune cell function. Additional studies will determine whether this occurs in other species, including humans, and whether the control by the local CRH-ACTH-thymic axis differs, allowing GC modulation of immune organ function some independence from the hypothalamic-pituitary-adrenal axis. The article published in this issue brings us closer to this goal and expands the puzzle (37).