Sexual dimorphism of stress response and immune/ inflammatory reaction: the corticotropin releasing hormone perspective

This review higlghts key aspects of corticotropin releasing hormone (CRH) biology of potential relevance to the sexual dimorphism of the stress response and immune/inflammatory reaction, and introduces two important new concepts based on the regulatory potential of the human (h) CRH gene: (1) a proposed mechanism to account for the tissue-specific antithetical responses of hCRH gene expression to glucocorticolds, that may also explain the frequently observed antithetical effects of chronic glucocorticoid administration in clinical practice and (2) a heuristic diagram to illustrate the proposed modulation of the stress response and immune/ inflammatory reaction by steroid hormones, from the perspective of the CRH system.


Introduction
Experimental support fbr the hypothesis that adrenocorticotropin (ACTH) secretion was controlled by hypothalamic factors, was obtained in 1955. 2 '3 In 1981, a 41 amino acid C terminal amidated peptide from ovine hypothalami stimulating pituitary ACTH release in vitro was identified and characterized. 4 The biologically active form of this peptide, designated corticotropin releasing hormone (CRH), and also frequently referred to as corticotropin releasing factor (CRF), was synthesized and found to have potent ACTH-releasing actions in vivo. 5 CRH is the only permissive factor for the anterior pituitary release of ACTH known in man 5'6 and acts in synergy with arginine vasopressin (AVP) and, perhaps, other factors, to regulate pituitary ACTH secretion, and, therefore ultimately the activity of the pituitaryadrenal axis. [7][8][9] Since its discovery, it has become evident that CRH has roles which are much wider than initially thought. Thus, coordination of the behavioral and physical components of the stress response and regulation of the immune/inflammatory reaction were unravelled as major overall roles of this neuropeptide. ' In addition, this peptide was implicated in the pathophysiology of a large range of diseases associated with dysregulation of 12 the stress system and autoimmunity. Because of the central roles of CRH in homeostasis and pathogenesis of disease, knowledge of its gene (C) 1995 Rapid Communications of Oxford Ltd and its regulation would be essential for further progress. This brief review will outline the most critical aspects of CRH biology, will summarize the structure, function and regulation of the human (h) CRH gene by steroid hormones, and will introduce a tentative model to account for its tissue-specific antithetical hormonal responses. The implications of the regulatory potential of the hCRH gene for the sexual dimorphism of the stress response and immune/inflammatory reaction will also be discussed.
Overview of CRH biology CRH is synthesized as part of a prohormone, it is processed enzymatically, and in addition undergoes enzymatic modification to the amidated form. Mammalian CRH has homologies with non-mammalian vertebrate peptides x CRH and sauvagine 4 in amphibia (from frog brain/spleen and skin, respectively), urotensin-I in teleost fish 5 and the two diuretic peptides Mas-DPI and Mas-DPII from the tobacco hornworm Manduca sexta. 16 '17 The vertebrate homologues have been tested and found to possess potent mammalian and fish pituitary ACTHreleasing activity. In addition, they decrease peripheral vascular resistance and cause hypotension when injected into mammals. 15 '18'19 The amino terminus of CRH is not essential for binding to the receptor, whereas absence of the carboxy terminal amide abolishes CRH binding to its receptor. Oxidation of a methionine residue coordinate the overall stress response. 4 High abolishes the biological activity of CRH, and this doses of CRH cause behaviors characteristic of may be a mechanism for neutralization of the anxiety, suggesting that the behavioral effects of peptide in vivo. 5 CRH bioavailability is also regu-CRH are dose-dependent, with low doses prolated by binding to corticotrop_in releasing moting adaptation and high doses being malahormone binding protein (CRHBP), 2 with which daptive. 12 '42 it partially co-localizes in the rat central nervous There is a broad peripheral expression of CRH system (CNS) and other tissues. 21 The human and CRHR, including the peripheral nervous CRHBP gene has been assigned to 5q11.2system, lung, liver, gastrointestinal tract, immune q13.3. 22 In the pituitary, CRH acts by binding to cells and organs, gonads and placenta, m''4 The membrane receptors (CRHR) on corticotrophs, biological roles of extraneural CRH have not that couple to guanine nucleotide-binding pro-been fully elucidated as yet, although it is likely teins and stimulate the release of ACTH in the that it might participate in the auto/paracrine regpresence of Ca 2 + by a cAMP-dependent mechanulation of [-endorphin production and analgesia, ism. 5'23'24 CRH stimulation of cAMP production and that it may modulate immune/inflammatory increases in parallel with the secretion of ACTH responses and gonadal function. 12'41'44-46 The in rat pituitary corticotrophs 25 and human current consensus is that CRH produced in high corticotroph cells. 26 In addition to enhancing the amounts in inflammatory sites of both animals secretion of ACTH, CRH also stimulates the de and humans, designated immune CRH, is promotnovo biosynthesis of pro-opiomelanocortin ing inflammation by stimulating cytokine produc-(POMC). 25 Sequence analysis of hCRHR cDNAs isolated causes a prompt increase in the release of ACTH from cDNA libraries prepared from human corti-into the blood, followed by the secretion of corticotropinoma or total human brain mRNA, sol. The effect is specific for ACTH release and is revealed homology to the G-protein coupled inhibited by glucocorticoids. High cortisol levels receptor superfamily. ' The hCRHR cDNA reduce or abolish CRH action on the pituitary. sequences of the tumour and normal brain were CRH has been used as a diagnostic tool to differaligned and found to be identical. The hCRHR entiate causes of hypercortisolism and hypocortigene has been assigned to 17q12-qter. 2 The solism, but does not have an established sequences of mouse and rat CRHR cDNAs were therapeutic role. 24'54-56 The clinical applications also reported recently. ' Human/rodent CRHR of CRH were recently reviewed extensively. protein sequences differ primarily in their extra- The amidated active peptide form is stored cellular domains. In particular, positively charged within secretory granules. Stress stimulates a arginine amino acid(s) are present in the third variety of endogenous substances, which excite and fourth positions of the extracellular amino the CRH neuron in the PVN 12 and cause the terminal domain of the rodent but not the release of CRH into the portal system by the human CRHR peptide. This might be responsible classical mechanism of membrane fusion. 5r for the differential activity of the t-helical 9-41 Major intracellular signalling systems, such as the CRH antagonist 4 between rodents and primates cyclic adenosine monophosphate (cAMP)-depen-(C. Kalogeras, personal communication), dent protein kinase A (PKA) [cAMP/PKAI and Central sites of CRHR expression include the the diacyglycerol (DAG)-dependent protein hypothalamus, the cerebral cortex, the limbic kinase-C (PKC)[DAG/PKC] pathways, appear to system, the cerebellum and the spinal cord. 5 be involved in the regulation of CRH biosynthesis This may explain the broad range of neural and release. 5 Theoretically, hormonal regulation effects of intracerebroventricularly (i.c.v.) admi-of CRH biosynthesis and secretion, and CRHRnistered CRH, including arousal, increase of symmediated si0nal transduction may occur in any of pathetic system activity, elevations in systemic many steps. blood pressure, tachycardia, suppression of the hypothalamic component of gonadotropin reg-hCRH gene structure and chromosomal ulation (GnRH), suppression of growth and localization inhibition of feeding and sexual behaviours 3642 characteristic of emotional stress. Central The hCRH gene consists of two exons sepainjection of CRH in rats and monkeys thus rated by an intron in its 5' untranslated region induces complex physiological and behavioral (Fig. 1). 59 The rat and ovine CRH genes have a 60 61 responses, suggesting that central CRH pathways similar organization.' The hCRH gene has been mapped to the long arm of chromosome 8 Regulatory elements of the hCRH gene (8q13). 62-The 3' untranslated region of the hCRH gene contains several polyadenylation CRE (cAMP-responsive enhancer, CREB enhansites, which may be utilized differentially in a cer sequence/PKA activation). Activation of potentially tissue-specific manner. CRH mRNA adenylate cyclase by various effector systems polyA tail length is regulated by phorbol esters elevates intracellular cAMP and stimulates in the human hepatoma CRH-expressing cell line protein phosphorylation by cAMP-dependent NPLC, and this may have potential relevance for protein kinases and transactivation of genes with differential stability of CRH mRNA in various the consensus palindromic sequence 5'tissues in vivo. 63 Alignment of the human, rat TGACGTCA-3' in their 5' flanking regions. and ovine CRH gene sequences has allowed the cAMP-response element-binding protein (CREB), comparison of the relative degree of evolutionary is a member of the bZIP or leucine zipper conservation of their various segments. These family of transcription factors, that is phoscomparisons revealed that the 330 bp long phorylated by several protein kinases, including proximal segment of the 5' flanking region of the catalytic subunit of cAMP/PKA. CREB homothe hCRH gene had the highest degree of or heterodimerizes through its leucine zippers, homology (94%), suggesting that it may play a binds to DNA as a dimer, and modulates tranvery important role in CRH gene regulation scription of genes in response to hormonal stithroughout phylogeny, which is crucial for sur-mulation of the cAMP pathways.  The gene vival. A conserved polypurine sequence feature encoding CREB contains at least 11 exons spanof unknown biological significance is present at ning over 40 kb and produces multiple CREB -829 of hCRH,-801 of the ovine CRH gene, as isoforms by alternative splicing. 71'72 This is in well as in the -400 bp 5' flanking region of contrast to other bZIP members (e.g., c-jun 72 POMC,_ rat growth hormone and other hormone and C/EBPV3), which are encoded by intronless genes. 64 The sequence of a 3.7 kb stretch into genes. 74 The latter proteins can heterodimerize the 5' flanking region of the hCRH gene was with some members of the CREB/ATF family of also determined (Gene Bank accession no. proteins. 75'76 x67661). 65 A gene bank search for homologous CRH regulation by the PKA pathway is well sequences identified a segment at position documented. Administration of cAMP increases 2213-2580 with greater than 80% homology to CRH secretion from perfused rat hypothalami, members of the type O family of repetitive and forskolin, an activator of adenylate cyclase, elements, and another at -2835 to -2972 with increases CRH secretion and CRH mRNA levels in less homology to the 3' terminal half of the primary cultures of rat hypothalamic cells. 77 Reghuman AluI family of repetitive elements. The ulation of the hCRH gene by cAMP has also been rest of the sequence was found to be novel and demonstrated in the mouse tumorous anterior specific to the CRH gene. A computer program pituitary cell line AtT-20, stably or transiently was used to identify consensus recognition sites transfected with the hCRH gene. 7e8 The hCRH of DNA binding proteins involved in transcrip-5' flanking sequence contains a perfect consensus tional regulation in this region of the hCRH CRE element that is conserved in the rat and gene. The organization and spatial distribution sheep, and confers transcriptional activation to of several putative responsive elements of poten-chloramphenicol acetyltransferase (CAT) reportial relevance to CRH transcriptional regulation is ters in vivo, as demonstrated with both human 81 82 shown in Fig. 2 65 and rat CRH promoter-driven CAT constructs.  TPA response elements in their 5' flanking binding sites have been detected. 65 Their ability regions.8TRE or AP-l-binding proteins are the to mediate TPA directed enhancement of hCRH Jun and Fos families of the bZlP superfamily of gene expression has not been tested by contranscription factors. These proteins interact ventional reporter gene assays as yet. EGF, through their carboxy-terminal leucine zippers however, has been shown to stimulate ACTR and c-jun can form both homodimers with itself secretion in the primate and to directly stimulate and heterodimers with c-fos, while c-fos can CRH secretion by rat hypothalamic organ culheterodimerize with c-jun but cannot form tures in vitro. 87 homodimers. TPA and the activation of PKC induce c-jun. This is due to positive autoregula-GRE (glucocorticoid response elemenO. Glucotion mediated by binding of c-jun homodimers corticoids are the final effectors of the hypothaor c-jun:c-fos heterodimers to an AP-1 site within lamic-pituitary-adrenal (HPA) axis and partici- 84 85 the c-jun promoter. pate in the control of whole-body homeostasis TPA, an activator of PKC, stimulates CRH and the organism's adaptive response to mRNA levels and peptide secretion by 1.5to 2stress, m'l These hormones exert their effects fold in primary cultures of rat hypothalamic through their cytoplasmic receptors. When in the cells, v7 TPA also increases CRH mRNA levels by ligand-unbound/inactive state, these receptors are almost 16-fold and CRH mRNA poly-A tail length in the form of a hetero-oligomer with a hsp90 by about 100 nucleotides in the human hepadimer and other proteins, s The ligand-bound toma cell line NPLC. The proximal 0.9 kb 5' receptors dissociate from the hetero-oligomer, flanking the hCRH gene confers TPA inducibility homodimerize and translocate into the nucleus, to a CAT reporter in transient expression where they interact with DNA to transactivate assays. 8 In the absence of a clearly discernible appropriate hormone-responsive genes that perfect TRE in this region, it has been suggested contain the consensus sequence 5'-GGTA-CANNNTGTTCT-3', GRE, in their 5' flanking GGTCANNNTGACC-3') consensus sequences in regions. 89'9 The activated receptors also interact their 5' flanking regions, respectively. 9 The two at the protein level with the c-jun component of ERE binding domains are exact palindromes, and the AP-1 transcription factor, preventing this half palindromic EREs can also mediate E2 factor from exerting its effects on TRE/AP1enhancement of gene expression. 112 Unlike the responsive genes. 91'92 Consistent with the struc-GR, androgen and estrogen receptors (ERs) are tural features of its promoter, the glucocorticoid not constitutively expressed and have a narrower receptor (GR) gene is constitutively expressed in tissue distribution, n''4 ER, although not all or most tissues of the organism. 9-9 Cellular studied as well as GR, appears to have similar hsp90 content appears to be an important deter-hsp90-dependent activational characteristics. 1'5 minant of a cell's sensitivity to glucocorticoids It interacts with the c-jun and c-fos components and the interaction of hsp90 with the unliganded of the AP1 binding proteins and thus also regform of GR appears to be a modulator of the ulates gene expression from TRE sims in a negaglucocorticoid signal transduction cascade.  tive or positive fashion. 116'17 It is not known Glucocorticoids play a key regulatory role in whether 1/2 GREs may in some cases confer weak the biosynthesis and release of CRH. 'm Gluco-androgen receptor-mediated transactivation by corticoids down-regulate rat and ovine hypotha-analogy with the weak and delayed GR transactivalamic CRH peptide and mRNA levels.61'4 Stably tion by glucocorticoids via the 1/2 GRE enhanintroduced hCRH gene in AtT-20 cells is also cer.9'Some regions of the male brain possess subject to negative glucocorticoid regulation, m5 aromatase activity to convert androgens to estro-Adrenalectomy and dexamethasone administra-gens, which may then modulate CNS functions tion in the rat elicits differential CRH mRNA through the more commonly expressed ERs in responses in the PVN and the cerebral cortex, brain areas of both sexes. stimulating and suppressing it respectively in the Human female hypothalami have higher conformer, but not influencing it in the latter, m6 centrations of CRH than the male ones. 8 Glucocorticoids can also stimulate hCRH gene Chronic estradiol treatment of ovariectomized expression in other tissues such as the human rats stimulates PVN CRH peptide levels 19 and placenta mv and the central nucleus of the amygincreases ACTH and corticosterone secretion dala. m8 A construct containing the proximal 900 basally and in response to stress. 2 In addition, bp of the 5' flanking region of the hCRH gene rat PVN CRH mRNA levels increase in the aftercoupled to a CAT reporter has been transiently noon of proestrous, at the approximate time of expressed in COS cells and found to confer the Ei-induced preovulatory surge of Ell. 121 negative and positive glucocorticoid effects, These findings indicate that gonadal steroids may depending on the cotransfection of a GRcDNA have an effect on the CRH-secreting neuron and expression plasmid. 86 The molecular mechanism suggest bidirectional interactions between the by which glucocorticoids regulate hCRH gene HPA and HPG axes through their final effectors. expression is somewhat obscure. Glucocorticoid A direct E2 enhancement of the CAT reporter suppression of hCRH gene expression might be was found by using two overlapping hCRH 5' mediated by the inhibitory interaction of the acti-flanking region-driven CAT constructs in transient vated GR with the c-jun component of the AP-1 expression assays. Furthermore, the two perfect complex. On the other hand, glucocorticoid half palindromic EREs present in the common enhancement of hCRH gene expression might be area of both CRH constructs, bound specifically mediated by the potentially active half perfect to a synthetic peptide spanning the DNA binding 122 GREs present in the 5' flanking region of the domain of the human ER.
These findings gene. 86 pituitary-gonadal (HPG) axis. They control reprogenes that do not have a TATA box in their production and sexual dimorphic physiology and moter region are constitutively expressed and behavior. 1 Androgens and estrogens (E2) act by have multiple transcription initiation sites. 95 The specific receptor-mediated processes to modulate hCRH gene has two proximal TATA boxes the expression of genes with GRE or ERE (5'located at -30 and-195 in its 5' flanking region (Fig. 2). The promoter-like properties of the account for the obseeeed differences in placental more distal TATA box were studied in vitro after expression of the CRH gene between these deletion of the more proximal TATA box. 86

species and primates.
These studies demonstrated that the-195 TATA Differential distribution of short and long box was active in initiating transcription and was hCRH mRNA transcripts has been detected in responsive to cAMP, TPA and glucocorticoids, several tissues and under varying physiological The majority of hCRH transcripts in most conditions. 8'24 Tissue-specific and/or stresstissues and cell lines studied initiate at + 1. dependent differential utilization of the two However, transcripts starting at-163 and-130 hCRH promoters, may explain these observahave also been detected in some tissues and cell tions. Differential mRNA stability would then be a lines, suggesting that the-195 TATA box is tran-particularly important feature in CRH homeoscriptionally active in vivo in a variety of sites of stasis, primarily in conditions of chronic stress, CRH expression, contributing up to 30-40% of since in the latter case sustained production of the total pool of cytoplasmic hCRH mRNA. 80'124 eRR would be required and the long stable Additional upstream, potentially active start sites mRNAs produced by activation of the distal are present in this gene, and the ratio of tran-moter would be beneficial to the organism. 86pr scripts initiati,ng at such sites might also be tissue-specific. 6 Regarding the physiological sig-Harmony in antithesis. Hormonal nificance of the longer transcripts, it was proposed that these might have a higher degree of regulation of the hCRH gene secondary structure and might be more stable Apparently, depending on its site of expresand long-lived than the short ones. 86 sion, the hCRH gene responds antithetically to 65,77 107 108 TATA boxes have also been implicated in tran-glucocorticoids. The antithetical glucoscriptional regulation by the p53 growth supcorticoid effects on a hCRH promoter-drive CAT pressor gene. More specifically, p53 appears to construct may be explained by the stimulation of control cellular activity by suppressing the tran-the half GREs by high levels of ligand-bound GR, scription of genes with a TATA box in their prowhich presumably override the blockade of tranmoter region through direct interaction. 25 The scription exerted by lower levels of the activated region-1.7 to-3.3 kb flanking the hCRH gene GR through interaction with cjun-cfos, as dishas a total of eight additional TATA boxes, which cussed above. Is this mechanism physiologically might also interact with p53 and influence CRH relevant, particularly in view of the constitutive gene expression. 65 expression of the GR gene? Potentially yes, especially if one takes into account the heat shock Other elements: Tissue-specific and other eleproteins (hsp). ments, potentially unique to the CRH gene, may Mthough the GR content of various tissues is 101.113 be involved in the control of its expression, similar, their sensitivity to glucocorticoids Their identification will require detailed analysis may vary substantially, 125 suggesting that some by both conventional and transgenic functional other cellular factor(s) is the principal modulator assays, and their characterization will provide a of glucocorticoid effect. There is a high tissuebetter understanding of the control of the HPA specific fluctuation of hsp90 supporting a correaxis. lation between tissue hsp90 content and the sensitivity of this tissue to glucocorticoids. m For Tissue-specific and stress-related instance, tissues with high hsp90 content appear expression of the hCRH gene to be quite sensitive to glucocoricoids in contrast with tissues with a low hsp90 content, m'28 As discussed above, the CRH gene is widely and as hsp90 content increases one would expressed throughout the body, suggesting that expect a parallel increase of the effective conits product may have autocrine or paracrine centration of tissue GR. 97'] The relatively small actions. Strong evidence for the presence of fluctuation of hsp90 levels during immobilization tissue-specific enhancers in the human and stress, m2 on the other hand, suggests that the primate CRH gene is its expression in the pla-other proteins of the unbound GR hetero-oligo-126 G centa and decidua, auch expression is absent mer 99 might also participate in the control of the in the placenta of rodents, 127 suggesting that sensitivity of tissues to glucocorticoids during 01 these enhancers may be carried in a segment of stress. Very little is known about the tissuethe regulatory region of the gene potentially specific expression and the regulation of these subject to rearrangement in the non-expressing proteins during stress, inflammation or debilitat- 65 species. Alternatively, the presence of tissueing disease. A proposed model summarizing specific repressor sequences in rodents may these observations is shown in Fig. 3 In this model it is assumed that basal expression of the hCRH gene is under constant AP-1. enhancement. Elevation of ambient glucocorticoid levels may elicit activation of GR to different degrees, depending on the tissue hsp90 content. TM Stimulation of hCRH gene expression would be predicted in tissues with high hsp90 content, such as testis, thymus or pineal, mediated by excess activation of GR that bypasses the AP-1 block and enhances transcription through the 1/2 GRE element. Inhibition would be expected in tissues with a low hsp90 content such as liver, spleen or pituitary, mediated by suppression of AP-1 enhancement through inactive complex formation with activated GR. Immunophilins, hsp70 and other factors interacting with the GR/hsp90 complex 99 may also modulate the activation of GR.
introduces a hypothetical general mechanism to account for the differential sensitivity and direction of effects of various tissues to glucocorticoids. This mechanism involves genes that are regulated by both the growth-promoting AP1 factors and by the differentiation-promoting, antigrowth glucocorticoid hormones. Since the hCRH gene contains both types of enhancers in its promoter region, it may potentially respond as outlined in Fig. 3. The mechanism proposed in Fig. 3 may also explain the frequently observed antithetical effects of chronic glucocorticoid administration in clinical practice.
Potential implications of CRH gene regulation for the sexual dimorphism of the stress response and the immune/ inflammatory reaction Both the stress response and the immune/ inflammatory reaction are associated with sexual dimorphism, both being more robust in the female or castrated male than in the intact male.
The basis of this dimorphism may be gonadal steroid regulation of the components of the stress response. The demonstration of direct E2effects on hCRH gene expression, 122 implicates the CRH gene and, therefore, the HPA axis, as a potentially important target of ovarian steroids and a potential mediator of gender-related differences in the stress response and HPA axis activity. These effects of E2 on the CRH neuron suggest that the HPG axis, which is known to be inhibited by hormones of the HPA axis at the hypothalamic, pituitary, gonadal and sex steroid target tissue levels during stress, 12'129 also appears to influence the latter in a positive fashion, by slightly enhancing CRH gene transcription. Thus, these data support a mutual, bidirectional interaction between the HPG and HPA axes, as depicted heuristically in Fig. 4.
The slightly enhanced CRH neuron activation by estrogen may not only explain why normal women have a slightly higher ACTH response to oCRH than normal men, as well as a slightly decreased ability of the glucocorticoid negative ..... 13 which, unlike the PVN, lacks ERs. 132 The negative E 2 feedback might, thus, be exerted indirectly, via a subgroup of CRH neurons. CRH has been reported to suppress GnRH secretion through both a direct and an indirect, arcuate nucleus POMC/13-endorphinmediated path. 2' The immune/inflammatory reaction is greater in female than in male animals and humans, and in keeping with this, autoimmune inflammatory disease has a significantly higher prevalence in the female than the male sex of several species. 4-Estrogens, generally, have been shown to activate some components of the immune/inflammatory reaction, while androgens suppress it. l-s Markedly elevated secretion of immune CRH in various inflammatory sites has been demonstrated in the Lewis rat, 44'47 an animal model of increased susceptibility to autoimmune inflammatory disease, in which decreased hypothalamic CRH secretion and, hence, diminished glucocorticoid production and defective suppression of inflammation has been demonstrated. 19'4 Although the decreased production of CNS CRH and increased secretion of immune CRH are associated with the high susceptibility of this animal to autoimmune inflam- .ses, including expression of immune CRH in pergests that the interactions between these axes are not unidirec-ipheral inflammatory sites, are greater in the tional, with the HPA axis inhibiting the HPG axis at multiple female than the male. 37'38 E2-mediated enhancelevels, as reviewed, 12 but bidirectional, with estrogen potentially stimulating both the CRH neuron, and therefore, the HPA axis, ment of immune CRH secretion might be a the peripheral production of immune CRH, and, hence, the partial explanation for this sexual dimorphism in immune/inflammatory response. 122 Immune CRH has been the Lewis rat, as well as, albeit [0 a lesser extent, shown to exert proinflammatory effects in vitro and in vivo, including enhanced production of cytokines and other mediators in other rat strains or animal species. 141 of inflammation, which in turn stimulate hypothalamic produc-Inflammatory sites, such as the athritic joints of tion of CRH, pituitary production of ACTH and adrenal producpatients with rheumatoid arthritis, contain high tion of glucocorticoids. Although not included in the figure, gonadal steroid/cytokine interactions have been demonstrated levels of immunoreactive CRH in the synovial for IL-6 actions. 16 fluid and inside cultured synoviocytes. 48 Interestingly, patients with rheumatoid arthritis have been shown to have poor or deficient responses feedback to shut off the ACTH and cortisol of their HPA axes to the stress of major surgery, responses, 129 but may also provide an explana-when compared with patients with osteoarthritis 142 tion as to why various emotional disorders charhaving similar surgery. Rheumatoid arthritis acterized by elevated CRH secretion, 12 such as patients also have inappropriately normal or low depression and anxiety, have a higher incidence normal basal diurnal concentrations of plasma 142 143 in women than in men. 3 Also, the same findcortisol making them strictly analogous to ings may explain why puberty/adolescence, the the Lewis rat model of autoimmune/inflammatory 139 140 postpartum and the perimenopausal period, disease.
during all of which marked changes in estrogen The above studies suggest that homeostatic production take place, are characterized by regulation involves complex mutual interactions increased incidence of emotional disorders, between the reproductive axis, HPA axis and In addition to explaining the slightly increased, the immune system, in which t2 and CRH may basal and stress stimulated HPA axis function in play central roles (Fig. 4). Certainly, other molethe female gender, the Ei-induced enhancement cules involved in the regulation of these axes, such as several neurotransmitters, cytokines and lipid mediators, also participate in the above interactions and may contribute to their sexual dimorphism. 144 Thus, a neurotransmitter, serotonin, has been shown to stimulate both CRH and ACTH secretion. 145 These findings explain the original pioneering studies, in which CRH-, ACTHand glucocorticoid-releasing bioactivities were found in the serum or supernatants of stimulated immunocytes.155,158,159 Thus, immune CRH, by participating in the regulation of the immune response at the level of the leukocyte, may be also viewed as a peripheral coordinator of immune-neuroendocrine interactions.