FASEB J.
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 QUICK SEARCH:   [advanced]


     


This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by SLOMINSKI, A.
Right arrow Articles by WEI, E. T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by SLOMINSKI, A.
Right arrow Articles by WEI, E. T.
(The FASEB Journal. 2001;15:1678-1693.)
© 2001 FASEB

Cutaneous expression of corticotropin-releasing hormone (CRH), urocortin, and CRH receptors

ANDRZEJ SLOMINSKI*1, JACOBO WORTSMAN{dagger}, ALEXANDER PISARCHIK*, BLAZEJ ZBYTEK{ddagger}, ELIZABETH A. LINTON§, JOSEPH E. MAZURKIEWICZ{dagger}{dagger} and EDWARD T. WEI**

* Department of Pathology, University of Tennessee Health Science Center, Memphis, Tennessee 38163, USA;
{dagger} Department of Internal Medicine, Southern Illinois University, Springfield, Illinois 62701, USA;
{ddagger} Department of Histology and Immunology, Medical University of Gdansk, Gdansk, Poland;
§ Nuffield Department of Obstetrics and Gynaecology, John Radcliffe Hospital, University of Oxford, Oxford, OX3 9DU UK;
{dagger}{dagger} Center for Neuropharmacology and Neuroscience, Albany Medical College, Albany, New York 12208, USA; and
** School of Public Health, University of California, Berkeley, California 94720, USA

1Correspondence: Department of Pathology, 899 Madison Ave., Room 576M-BMH, 899 Madison, Ave., Memphis, TN 38163, USA. E-mail: aslominski{at}utmem.edu


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
OVERVIEW OF CORTICOTROPIN...
EXPRESSION OF CRH AND...
CRH RECEPTORS (CRH-Rs)
SUMMARY AND PERSPECTIVE
REFERENCES
 
Studies in mammalian skin have shown expression of the genes for corticotropin-releasing hormone (CRH) and the related urocortin peptide, with subsequent production of the respective peptides. Recent molecular and biochemical analyses have further revealed the presence of CRH receptors (CRH-Rs). These CRH-Rs are functional, responding to CRH and urocortin peptides (exogenous or produced locally) through activation of receptor(s)-mediated pathways to modify skin cell phenotype. Thus, when taken together with the previous findings of cutaneous expression of POMC and its receptors, these observations extend the range of regulatory elements of the hypothalamic-pituitary-adrenal axis expressed in mammalian skin. Overall, the cutaneous CRH/POMC expression is highly reactive to common stressors such as immune cytokines, ultraviolet radiation, cutaneous pathology, or even the physiological changes associated with the hair cycle phase. Therefore, similar to its central analog, the local expression and action of CRH/POMC elements appear to be highly organized and entrained, representing general mechanism of cutaneous response to stressful stimuli. In such a CRH/POMC system, the CRH-Rs may be a central element.—Slominski, A., Wortsman, J., Pisarchik, A., Zbytek, B., Linton, E. A., Mazurkiewicz, J., Wei, E. T. Cutaneous expression of corticotropin-releasing hormone (CRH), urocortin, and CRH receptors.


Key Words: skin • PKA • CRH gene • stress


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
OVERVIEW OF CORTICOTROPIN...
EXPRESSION OF CRH AND...
CRH RECEPTORS (CRH-Rs)
SUMMARY AND PERSPECTIVE
REFERENCES
 
THE SKIN, THE largest organ in the body, plays a critical role in maintaining internal homeostasis, serving as a barrier between the external environment and the internal milieu (1 , 2) . Because of its location, the skin is exposed continuously to a fluctuating environment with potentially noxious stimuli. It thus requires a precise and focused mechanism for the immediacy of its interactions with environmental stressors; preferably, such a mechanism should already be activated while cellular/tissue damage is still contained and of low magnitude. We have recently proposed that the skin possesses a high sensory capability for stressful stimuli that is tightly coupled to a local response system (1 , 2) . This response system would restrict tissue damage and restore local homeostasis. In analogy with the central response to stress, which involves predominantly the hypothalamic-pituitary-adrenal (HPA) axis, we proposed that the responses to stress of skin and the central nervous system (CNS) could share similar mediators (1 2 3 4) . This postulate was based on the known capabilities of mammalian skin to produce the POMC-derived ACTH and MSH peptides and to be a target for their actions through locally expressed MC and opioid receptors. In fact, the skin is also intrinsically capable of producing CRH and urocortin peptides, together with the corresponding receptors (1 , 4 5 6) . Both CRH and CRH receptors are recognized as central regulators of the response to chronic sustained stress. Therefore, we have suggested that this cutaneous CRH/POMC system is organized as a functional equivalent of the hypothalamic-pituitary-adrenal axis (1 , 3 , 4) .

CRH peptide and CRH receptors are expressed in many other tissues and organs besides CNS and skin; among them are gestational tissues, immune system, pancreas, liver, gastrointestinal tract, skeletal muscle, heart, lung, and endocrine organs (ovaries, testes, adrenals, and thyroid glands) (1 , 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27) . Biochemical studies have shown that CRH gene expression can be stimulated through pathways involving cAMP-dependent protein kinase A (PKA), calcium/calmodulin-dependent protein kinase (CaMK), diacylglycerol-dependent protein kinase C (PKC), and transcription factors associated with cytokine signaling (19 , 20 , 25 , 28 29 30 31 32) . Glucocorticoids inhibit CRH gene expression. CRH peptide production in peripheral tissues is enhanced by prostaglandins, epidermal growth factor (EGF), carbon monoxide (CO), and platelet-derived growth factor (PDGF) and is decreased by nitric oxide (NO) and progesterone (23 , 25 26 27 , 29 , 33) . CRH produced in peripheral tissues is thought to regulate local homeostasis; for example, CRH produced in the uterus and placenta may affect the progress of normal pregnancy (13 , 23 , 33 , 34) . CRH is also a potent immunomodulator with cellular target-dependent polarity, acting to inhibit or stimulate local immune function; it also acts on vascular function and is a local growth factor. This review will focus on the molecular, biochemical, and functional characterization of CRH, urocortin, and their CRH receptor in mammalian skin.


   OVERVIEW OF CORTICOTROPIN-RELEASING HORMONE (CRH) AND UROCORTIN EXPRESSION
TOP
ABSTRACT
INTRODUCTION
OVERVIEW OF CORTICOTROPIN...
EXPRESSION OF CRH AND...
CRH RECEPTORS (CRH-Rs)
SUMMARY AND PERSPECTIVE
REFERENCES
 
The CRH gene is composed of two exons separated by an intron (30 31 32 , 35 , 36) . The first exon encodes most of the 5'-untranslated region in the mRNA, whereas the second exon contains the prohormone sequence and the 3'-untranslated region. CRH transcripts in the rodent and human brain are ~1.4 and 1.5 kb long, respectively (30 , 31 , 36) . Translation of exon 2 generates the 196 amino acid (aa) long pre-proCRH, of which the first 26 aa represent the signal peptide. Pre-proCRH is cleaved in the rough endoplasmic reticulum to generate proCRH 27–196 with a molecular mass of 18 kDa, which increases to 23 kDa after post-translational modifications (29 , 37) . Pro-CRH undergoes endoproteolytic processing within the trans-Golgi network and secretory granules that appears to be controlled by the convertases PC1 and PC2 to generate the final 41 aa residue (4.7 kDa) CRH peptide (29 , 37 , 38) . Expression of the CRH gene is controlled by a promoter region located ~200 bp upstream that contains the TATA box and cyclic AMP-responsive element (CRE) (29 30 31) . CRE mediates activation of the CRH promoter by cAMP-dependent PKA and CaMK pathways (19 , 20 , 28 , 29) . The presence of multiple phorbol ester response elements in the 5'-flanking region of the CRH gene as well as the stimulation of CRH gene expression and secretion of the peptide by TPA indicate that the diacylglycerol-dependent PKC pathway and AP-1 are involved in this process. The cytokines interleukin 1 (IL-1), IL-6, and tumor necrosis factor {alpha} stimulate CRH production through the activation of the CRH promoter by transcription factors associated with cytokine signaling (19 , 20 , 29 , 39) . In the CNS, CRH production is stimulated by neurotransmitters and neuropeptides such as serotonin, acetylcholine, histamine, norepinephrine and epinephrine, arginine vasopressin, angiotensin II, neuropeptide Y, cholecystokinin, activin, and enkephalin (reviewed in refs 19 , 20 , 29 ). Leptin can also stimulate CRH gene expression and basal CRH release (40) . The well-known negative regulation of CRH production by glucocorticoids appears to result from direct interaction between the glucocorticoid receptor (GR) and either the GR binding site in the 5'-flanking CRH gene region or the CRE binding protein (19 , 20 , 28 , 29 , 41 , 42) . Estrogens and gamma-amino butyric acid as well as dynorphin, substance P, somatostatin, and galanin share the same inhibitory activity (19 , 20 , 29) .

In the brain CRH is produced predominantly in the paraventricular nucleus (PVN) of the hypothalamus for delivery into portal capillaries (28) , (29) . CRH produced by autonomic neurons of the PVN projecting to the brainstem and spinal cord regulates the sympathoadrenal system; CRH is also produced by neurons projecting to the pituitary that are involved in osmotic regulation (19 , 29) . Besides its central neurotransmitter and hormonal functions, CRH can act as a growth factor regulating corticotroph proliferation (43 , 44) . Extracranial CRH is produced in endometrium, placenta, uterine myometrium, cells of the immune system, gastrointestinal system, adrenal gland, and skin (reviewed in refs 1 , 13 , 17 , 19 , 20 , 24 25 26 27 , 33 , 45 , 46 ). Urocortin is a recently described member of the family of structurally related CRH-like peptides sharing a high degree of homology with CRH (47 48 49 50) . In humans, the final product of the urocortin gene is a peptide 40 amino acids long, with 45% homology to rat/human CRH, and more than 90% homology with rat, mouse, and hamster urocortin peptides (47 , 49 , 50) . Similar to CRH, the human, rodent, and pig urocortin gene is composed of one intron and two exons, with exon 2 encoding the urocortin peptide and a promoter containing negative and positive regulatory sequences (GenBank accession nos. AF038632, AF038633, AF093623, Y15169; refs 48 , 50 , 51 ). The latter sequence includes a TATA box, CRE, and Brn-2, C/EBP, and GATA binding sites (50) . A MyoD site has been found in the mouse promoter and an Ap1 site in the human promoter (50) . Expression of the urocortin gene is stimulated by cAMP. In rat brain, urocortin is particularly prominent in the Eddinger-Westphal nucleus, lateral superior olive, substantia nigra, ventral tegmental area, linear and dorsal raphe nuclei, and the hypothalamus (52 , 53) . In human brain, urocortin expression is more widespread, being found in every region tested, with the highest concentrations in the frontal cortex, temporal cortex, and hypothalamus. Urocortin has also been detected in the anterior pituitary gland (54 , 55) and peripherally in a variety of human and rat tissues that include placenta, uterus, immune system, stomach, small and large bowel, pancreas, adrenal gland, testis, and heart (7 , 16 17 18 , 22 , 56) .


   EXPRESSION OF CRH AND UROCORTIN IN THE SKIN
TOP
ABSTRACT
INTRODUCTION
OVERVIEW OF CORTICOTROPIN...
EXPRESSION OF CRH AND...
CRH RECEPTORS (CRH-Rs)
SUMMARY AND PERSPECTIVE
REFERENCES
 
CRH
Rodent skin
CRH was initially identified in mouse skin with the use of reverse-phase high-performance liquid chromatography (HPLC) combined with CRH radioimmunoassays (RIA) (4 , 57 ; Fig. 1A ). Despite several attempts with RT-PCR, CRH mRNA could not be detected in mouse skin (4 , 58 ; Fig. 1B ), although the levels of CRH peptide nevertheless were consistently detectable by RIA. In fact, mouse skin CRH reached concentrations significantly higher than in serum, and displayed hair cycle-dependent fluctuations (57) . Thus, actual skin CRH levels were 36 and 67 fmol/g of wet tissue in telogen and mid-anagen, respectively, whereas the corresponding serum levels were only 5.6 and 8.6 fmol/ml (57) . Using immunocytochemistry and indirect immunofluorescence, we found that CRH was localized to the pilosebaceous unit of the hair follicle, to keratinocytes of the outer root sheath (ORS), and to the matrix region of developing hair follicles and basal epidermis with expression in a hair cycle-dependent fashion (57 , 58) . Indirect immunofluorescence showed that CRH was also present in the nerve bundles and perifollicular neural network B throughout the entire hair cycle (57) . This finding supports the initial hypothesis that cutaneous CRH in the C57BL/6 mice originates through importation via descending nerves (4 , 58) . This purported mechanism could allow the precise targeting of specific domains for activation of CRH-dependent POMC production (cf. refs 1 , 4 ). In another rodent species—the rat—CRH was identified in dermal and subcutaneous (s.c.) compartments after experimentally induced inflammation; cells expressing CRH included macrophages, fibroblasts, and endothelial cells (10) .



View larger version (17K):
[in this window]
[in a new window]
 
Figure 1. C57 BL/6 mouse skin contains CRH but does not express CRH mRNA. A) RP-HPLC identification of CRH (CRF) in anagen III/IV ({square}{square}) and anagen VI (•—•) mouse skin. For details, see refs 4 , 57 . B) Southern blot hybridization with murine CRH cDNA of RT-PCR amplification products (35 cycles). CRH mRNA (arrow) is undetectable in telogen (lane 1), anagen II (lane 2), anagen IV (lane 3), early anagen VI (lane 4), late anagen VI (lane 5), or catagen skin (lane 6) despite its presence in the brain tissues used as positive control (lane 7). For details, see ref 58 .

Human skin
In contrast to mouse skin, CRH mRNA was detectable by RT-PCR in human skin, as shown by testing normal and pathological human skin biopsy specimens, as well as in cultured normal and malignant keratinocytes and melanocytes (Fig. 2 ; 59 60 61 ); CRH gene expression has also been detected by others in established melanoma cell lines, cultured junctional and dermal nevocytes, epidermal melanocytes, epidermis, and hair follicles (62 , 63) . Northern blot hybridization showed that the CRH mRNA transcript was 1.5 kb in melanoma and squamous cell carcinoma cells (59 , 61) . This gene expression was accompanied by the actual production of CRH peptide, as determined by RP-HPLC combined with CRH RIA (61) . Testing human skin with liquid chromatography-mass spectrometry (LC/MS) confirmed those results (Fig. 3 ; 6 ). The CRH antigen was localized to keratinocytes of the epidermis and hair follicle, dermal blood vessels, skeletal muscle, and nerve bundles of the human scalp (4) . Both CRH antigen and CRH mRNA were further detected in human pilosebaceous units and nevocytes by immunocytochemistry and in situ hybridization (63) , as well as by immunocytochemistry in melanoma specimens (62) . CRH peptide production was found to be regulated by UVB, forskolin, and dexamethasone in skin cells (59 , 61) . Thus, UVB markedly increased intracellular CRH immunoreactivity in normal melanocytes, albeit with lower release to the culture media cells (59) . In melanoma and squamous cell carcinoma cells, forskolin stimulated and dexamethasone inhibited CRH peptide production, although the corresponding CRH mRNA levels remained unchanged, suggesting post-transcriptional regulation of CRH expression (61) .



View larger version (22K):
[in this window]
[in a new window]
 
Figure 2. Expression of the CRH gene in human skin cells using primers directed to sequences in exon 2, as described in ref 64 . A cDNA fragment of 308 bp was obtained from human placenta (lanes 3, 4) and from skin cells that included human neonatal keratinocytes (lane 5), C1–4 squamous cell carcinoma cells (lane 6), SKMEL188 melanoma cells (lane 7), neonatal human melanocytes (land 8), and HaCaT keratinocytes (lane 9). DNA markers (lanes 1, 11); blank (lane 2). The cell lines have been described in ref 61 .



View larger version (24K):
[in this window]
[in a new window]
 
Figure 3. Identification of CRH in human skin by tandem LC/MS. Human skin extracts show a peptide (CRH) with mass spectrum (M+4H)+ ion at m/z 1189 (A) and retention time at 65–66 min (B) identical to the CRH standard (inserts). Reprinted with permission from the Endocrine Society; for details, see ref 6 .

The findings in skin are consistent with the wide expression of the CRH gene reported in extracranial tissues, but at levels much lower than in the brain. When peripheral processing of proCRH into CRH has been evaluated, it appears to be similar to that at the central level, at least in placenta, endometrium, uterus, and the immune system (13 , 15 , 34 , 64 , 65) . CRH production in peripheral tissue is enhanced by prostaglandins, CO, EGF, and PDGF and its release is decreased by NO and progesterone (23 , 26 , 27 , 33 , 34) . CRH produced in the uterus and placenta appears to play an important role in the normal progression of pregnancy (13 , 23 , 33 , 34) , whereas that produced by activated immune cells or released by peripheral nerve endings may regulate local inflammatory processes including vascular functions and tissue fluid distribution (8 , 10 , 21 , 24 , 65 66 67 68 69 70 71) .

Urocortin
Rodent skin
In clear contrast to the lack of expression of the CRH gene in mouse skin (4 , 58) , we have found expression of the urocortin gene together with accumulation of urocortin peptide (5) . Urocortin skin concentration was also related to hair cycle phase but in a pattern opposite to that of CRH, i.e., the highest concentration was found in telogen, with a subsequent decrease during anagen progression (5) . The significance of this reciprocal pattern is still unclear, although it suggests selective functions for the peptides in skin physiology and pathology. In another rodent species, the Syrian golden hamster (Mesocricetus auratus) urocortin was not detected in the skin despite being present in the brain (positive control). Thus, hamster skin may not produce urocortin under physiological conditions. However, since we have detected relatively high concentrations of urocortin immunoreactivity in transplantable Bomirski hamster MI hypomelanotic melanoma (at levels even higher than brain) and in an established amelanotic hamster melanoma line (AbC1) (5) , it is possible that hamster skin urocortin production potential is expressed only under pathological conditions, of which malignant melanoma is an example.

Human skin
Expression of the urocortin gene and its peptide product has been examined in human skin and in cultures of normal and malignant keratinocytes and melanoma cells using RIA, RP-HPLC, LC-MS, and RT-PCR with sequencing of the amplified cDNA fragment (5) . These studies have clearly shown expression of the urocortin gene and accumulation of the corresponding peptide in human skin cells (5) . The urocortin antigen has been localized to epidermal and follicular keratinocytes and sweat glands, nevocytes, and malignant melanocytes as well as blood vessels walls, dermal smooth muscle, mononuclear inflammatory cells, and dermal spindle cells. By itself, antigen detection at these sites may represent either receptor-mediated peptide internalization or urocortin production within the same cells. Definitive pinpointing of peptide production will require detailed testing with in situ hybridization techniques. Nevertheless, it is apparent that the skin represents a new extracranial source of urocortin with proven manufacturing capability similar to the human placenta, amnion, chorion, decidua, the gastrointestinal system, and lymphocytes (7 , 16 , 17 , 22 , 34 , 72) and to rodent cardiac myocytes, digestive system, pancreas, pituitary, adrenal, testis, and heart (17 , 18 , 56) .


   CRH RECEPTORS (CRH-Rs)
TOP
ABSTRACT
INTRODUCTION
OVERVIEW OF CORTICOTROPIN...
EXPRESSION OF CRH AND...
CRH RECEPTORS (CRH-Rs)
SUMMARY AND PERSPECTIVE
REFERENCES
 
The CRH-Rs represent a family with at least two distinct members (CRH-R1 and CRH-R2) encoded by separate genes (11 , 12 , 28 , 45 , 73 74 75 76 77 78) . CRH-R1 and CRH-R2 share high sequence homology (ca. 70%) and belong to the family of seven transmembrane receptor proteins coupled to the Gs signaling system. The genomic structure for the CRH-R1 has been established in humans (GeneBank accession nos. AF039510-AF3523; 79 ) and rats (GeneBank accession nos. U53486-U53498; ref 32 ), as has the genomic sequence for the CRH-R2 gene in humans (GeneBank accession no. AC004976) (Fig. 4 ). The genes for human and rat CRH-R1 contain 14 and 13 exons, respectively, whereas analysis of the recently deposited genomic sequence of the CRH-R2 (GeneBank accession no. AC004976) has shown that human CRH-R2 contains at least 15 coding exons (Fig. 4) . CRH-R1 is a protein with 98% sequence homology among different mammalian species and ~30% homology with receptors for the gut-brain family of neuropeptides (cf. refs 45 , 76 , 77 ). Four alternatively spliced transcripts have been identified in humans (Fig. 4) : CRH-R1{alpha} in which exon 6 (encoding a 29 aa sequence in the first intracellular loop) is spliced out, generating a 13 exon transcript that produces a 415 aa protein (GeneBank accession no.L23332; refs 73 , 78 ); CRH-R1ß, a longer variant that contains all 14 exons translating into a 444 aa protein (GeneBank accession no. L23333; ref 73 ); a CRH-R1c isoform where exon 3 (120 bp long coding for a 40 aa sequence from the extracellular amino-terminal domain) and exon 6 are spliced out to generate a 12 exon transcript producing a 375 aa protein (GeneBank accession no. U16273; ref 80 ); and an CRH-R1d isoform where 6 and 13 exons (coding for a 14 aa sequence within the seventh transmembrane domain) are missing, which translates into a 401 aa protein (GeneBank accession no. AF180301; ref 81 ). Three CRH-R1 isoforms have been identified in rats: isoform A spanning 1–13 exons; isoform B with spliced out exon 3; and isoform C with spliced out exons 7, 11, and 12 (32) . The CRH-R1 gene has also been cloned in the mouse, sheep, and cow (78 , 82) , but their genomic structures have not yet been reported.



View larger version (45K):
[in this window]
[in a new window]
 
Figure 4. Structure of CRH-R1 and CRH-R2 human genes products. A) Alternatively spliced isoforms {alpha}, ß, c, and d of CRH-R1. B) Alternatively spliced isoforms {alpha}, ß, {gamma}, and stomach variant of CRH-R2. C) Theoretical structure of the CRH-R1 protein coded by 14 exons. D) Theoretical structure of the CRH-R2 protein coded by 15 exons.

For CRH-R2, four differentially spliced variants have been identified that produce proteins differing in amino-terminal domains and anatomical distribution; all contain five potential N-glycosylation sites analogous to those found in CRH-R1 (cf. refs 11 , 12 , 28 , 45 , 77 , 83 ). To trace their gene source, we compared the cDNA sequences corresponding to different isoforms of CRH-R2 (GeneBank accession nos. U345871, AF011406, AF01938) with human genomic DNA (GenBank accession No AC004976) (Fig. 4) . This analysis showed that mRNA of the CRH-R2{alpha} isoform contains exons 4 through 15 (GenBank accession no. U34587; ref 12 ); the CRH-R2ß isoform comprises exons 1, 2, 5–15 (GenBank accession no. AF011406; ref 84 ); the CRH-R2{gamma} isoform contains exons 3, 5–15 (GenBank accession no. AF019381; ref 83 ); and the newest CRH-R2 isoform detected in the stomach contains exons 10–15 (GenBank accession no. E12750; Patent: JP199707289-A). In the latter variant, the translation starts from the methionine codon situated in exon 10 and produces only the carboxyl-terminal part of the CRH-R protein. Two additional aberrantly spliced ß1 and ß2 forms that respectively contain insertion of 227 and deletion of 94 nucleotides (GeneBank accession nos. Y10152 and Y10153 (84) have also been described. The CRH-R2{alpha} and ß isoforms have been cloned in the rat, mouse, and tree shrew (2 , 11 , 75 , 76 , 85 , 86) , and the CRH-R2{alpha}-tr isoform that contains unspliced introns 6 and 7 and encodes a 236 aa truncated protein has been cloned in rats (87) . The full genomic structure of rodent CRH-R2 has not as been published. In humans, CRH-R2{alpha} and CRH-R2ß are expressed in peripheral tissues and the brain, whereas in rodents CRH-R2{alpha} is found predominantly in the brain and CRH-R2ß in the periphery (cf. refs 45 , 76 , 77 ). The CRH-R2{gamma} variant has been found only in the brain (83) . CRH-R1 and CRH-R2 differ in pharmacological ligand profile and tissue distribution; e.g., CRH-R1 predominates in the pituitary and the brain, whereas CRH-R2 is present in heart, skeletal muscle, and brain (cf. refs 28 , 45 , 76 , 77 ).

Very recent work by Arai and colleagues has identified a third CRH receptor subtype in catfish that originates from a third gene, distinct from R1 and R2 receptor genes (88) . This R3 receptor shares significant sequence homology with catfish R1 and R2, but binds CRH with a fivefold greater affinity than other CRH-related peptides (88) . It remains to be determined whether an equivalent R3 gene exists in mammalian species.

At the central level, the binding of CRH to CRH-R activates adenylate cyclase with the production of cAMP and subsequent activation of PKA-dependent pathways; phospholipase C may also be activated with the production of inositol triphosphate (IP3), which in turn activates PKC-dependent and calcium-activated pathways (19 , 20 , 28 , 45 , 76 , 77) . Data also suggest that CRH signal transduction is directly coupled to calcium channels (4 , 19 , 45 , 89 90 91 92 93) and perhaps to potassium channels (92 , 93) . CRH-R1 is more efficient than CRH-R2 at transducing the CRH signal, with greater intracellular accumulation of cAMP (28 , 45 , 76 , 77) . Thus, whereas CRH and urocortin are more potent in activating adenylate cyclase through CRH-R1 than are sauvagine and urotensin I (both CRH-like peptides are found in frog skin and the urophysis of fish, respectively), urocortin, sauvagine, and urotensin I are more potent than CRH in cyclase activation via CRH-R2 (94) . The binding affinity of CRH to CRH-R1 is similar to that of urocortin and sauvagine whereas the binding affinity of CRH to CRH-R2 is significantly lower than of urocortin, sauvagine, and urotensin I (94) .

The extracranial tissues in which CRH receptors have been identified include the adrenal glands, testes, ovaries, prostate, kidney, liver, gut, spleen, circulating immune cells, synovium, heart, skeletal muscle, uterine myometrium, vascular endothelium, arterial smooth muscle, endometrium, placenta, lungs, and skin (9 , 11 , 14 , 19 , 20 , 23 , 24 , 34 , 45 , 46 , 65 , 74 , 76 , 77 , 81 , 84 , 95 96 97 98 99 100 101) . Molecular and pharmacologic characterization has shown that peripheral tissues express predominantly CRH-R2 (cf. refs 45 , 76 , 77 ). Nevertheless, CRH-R1 has also been detected in intrauterine tissues such as endometrium, myometrium, placenta, and fetal membranes, in ovary, in adrenal glands, and in immune cells (11 , 34 , 81 , 95 96 97 98 , 102) . The intracellular signaling pathway activated through peripheral CRH receptors involves the production of cAMP and subsequent activation of PKA (cf. refs 19 , 20 , 45 , 77 ). Involvement of calcium-activated pathways by phospholipase C or membrane-bound calcium channels has also been suggested (cf. refs 4 , 19 , 45 , 89 , 90 ). Recent data have demonstrated CRH receptor-mediated activation of MAP kinase signal transduction pathways (103) in several cell types, including normal and neoplastic corticotropes (104) , cardiac myocytes (105) , and PHM1–41 pregnant uterine myocyte cells (34) .

CRH receptors located in the heart and vasculature mediate the inotropic, vasodilatory, and anti-edema effects of CRH and/or CRH-related peptides (19 , 20 , 24 , 45 , 70 , 99) . In immune cells, the local effects of CRH are complex since it has been reported to both inhibit and stimulate production of the proinflammatory cytokines (IL-1 and IL-6) by peripheral blood mononuclear cells (21 , 25 , 65) . In general, systemically administered CRH may stimulate proinflammatory reactions in the periphery whereas direct application of CRH may produce an anti-inflammatory effect. An attempt to reconcile these apparently contradictory immunomodulatory actions has been made by Paez-Pereda and colleagues who showed that the up-regulation of IL-1 expression elicited by CRH in resting monocytes appears to be inhibited once monocytes are activated (21) . Their proposal that the activational state of leukocytes determines the direction of their responses to the peptide also explains the result in the endotoxin stimulation of leukocytes model, which invariably shows CRH to be immunoinhibitory (71) . In gonads and adrenal glands, locally produced CRH may regulate steroid production via a paracrine mechanism of action (19 , 20 , 106 107 108) . In the pregnant myometrium at term, five isoforms (1-{alpha}, 1-ß, 1-c, 1-d, and 2-{alpha}) of CRH receptors have been identified (96) . They are thought to be involved in maintaining quiescence throughout pregnancy and in the initiation and progression of labor, being activated by circulating placental CRH and by locally produced CRH or urocortin through paracrine or autocrine mechanisms (13 , 34 , 46 , 109) . The precise timing required for this action demands the developmentally programmed, tissue-restricted differential expression of CRH receptors in the placenta, decidua, fetal membranes, endometrium, myometrium, and uterine vasculature (13 , 23 , 26 , 33 , 34) .

CRH receptors in mammalian skin
Rodent skin
Molecular analyses have demonstrated the expression of both CRH-R1 and CRH-R2 genes in rodent skin (1 , 4 , 57 , 58) . Expression of the CRH-R1 gene has also been reported in rat mast cells (100) . Northern blot hybridization with murine CRH-R1 cDNA revealed that anagen skin restricted expression of a 2.7 kb transcript, corresponding to spliced CRH-R1 mRNA; this was not detected in telogen skin (Fig. 5A ; 58 ). Hamster melanoma showed a shorter CRH-R1 transcript (Fig. 5A ). Mouse skin, brain control, and hamster melanoma also expressed a 5 kb mRNA species that hybridized to mouse CRH-R1 cDNA, representing most likely a CRH-R1 mRNA precursor with nonspliced introns (Fig. 5A ). Using primer sequences corresponding to exons 3 and 7 of the CRH-R1 rat gene, RT-PCR identified in murine skin a 407 bp product representative of CRH-R1 mRNA, which was present throughout the entire hair cycle (Fig. 5B ; 58 ). Using primers from the coding exons 9 and 15 of the human and mouse CRH-R2 gene, amplification of the 615 bp fragment common to the {alpha} and ß variants revealed the expression of the corresponding gene in C57BL/6 mouse skin (1 , 4) .



View larger version (29K):
[in this window]
[in a new window]
 
Figure 5. Expression of the CRH R1 receptor gene in C57 BL/6 mouse skin and Bomirski hamster AbC-1 melanoma cells. A) Northern blot hybridization of mouse CRH-R1 cDNA to mRNA from anagen VI skin (lane 1), AbC-1 melanoma (lane 2), and mouse brain (lane 3). Left: RNA ladder (kb). Arrow, two arrowheads, and one arrowhead point to transcripts of ~5, 2.7, and, 2.3–2.5 kb, respectively. Conditions for Northern hybridization are described in ref 58 . B) RT-PCR identification of CRH-R1 mRNA in murine pituitary (lane 2), brain (lane 3), skin at anagens IV (lane 4) and VI (lane 5), and spleen (lane 6). DNA ladder (lane 1). Primer sequences and conditions for RT-PCR amplification are described in ref 58 .

The CRH-R1 protein was localized to the keratinocytes of the ORS, the hair matrix, dermal papilla of anagen hair follicles, the keratinocytes of inner root sheath, and the ORS of early catagen. The concentration of CRH-R1 protein was lowest in telogen skin (57) . Autoradiographic studies using 125I-[Tyr]-oCRH tracer showed a low density of CRH binding sites in epidermal and hair follicle keratinocytes and in dermal panniculus carnosus, with most binding sites localized to dermal muscles; curve analysis revealed a single high-affinity binding site with a dissociation constant (Kd) of 3.62 nM, similar to that found in the rat pituitary control (Kd of 1.53 nM), with a Bmax of 2.42 amol/ mm2 of scanned mouse skin section (4 , 57) . CRH binding sites were also found on blood vessels and epithelial cells of the rat skin (66) . In summary, CRH receptors are expressed in mouse and rat skin, and this expression is compartment dependent: CRH-R1 is the predominant form in keratinocytes and CRH-R2 in the panniculus carnosus.

Earlier studies performed on animal models of tissue injury (24 , 67 , 70 , 110 , 111) revealed that peptides of the CRH superfamily have anti-inflammatory effects (70) . For example, CRH decreased protein extravasation, edema, and swelling in the skin of the anesthetized rat paw after exposure to heat or extreme cold; in the tracheal mucosa of rats that had been exposed to formaldehyde; in skeletal muscle after a knife cut; and in brain cortex after freezing (111) . The anti-edema CRH action has been interpreted as arising from a ligand-induced inhibition of the negative interstitial fluid pressure generated in the extracellular matrix of traumatized tissues (112) . In a rabbit model of doxorubicin-induced eyelid inflammation, local injection of CRH decreased the severity of skin injury and enhanced wound healing (110) . CRH did not alter vascular permeability in this study, but reduced the expected acute influx of monocytes and macrophages. More recently, it was found that urocortin is significantly more potent than CRH at inhibiting heat-induced cutaneous edema in rats (70 , 113) . The anti-inflammatory actions of CRH are independent of its corticosteroid-releasing or hypotensive effects and may be due to direct modulation of interstitial fluid pressure after tissue injury (70 , 112 , 114 , 115) . CRH has been found to act as a functional antagonist of inflammatory mediators such as histamine and substance P (67) . Although CRH inhibits neurogenic inflammation, interactions with unmyelinated sensory neurons do not account for the wide range of its anti-inflammatory activities. Conversely, localized application of CRH prevented histamine-induced leaks in the hamster cheek pouch (114 , 115) , and displaceable binding sites to iodinated-CRH were found on blood vessels and on epithelial cells in close proximity to sites of vascular leakage (66) . Local administration of CRH also inhibits the response to nociceptive stimuli (24) , enhances wound healing (110) , and may affect angiogenesis (116) . CRH causes flushing and a sense of heat because of local vasodilatation, but also suppresses edema and is antinociceptive, perhaps, because of local release of POMC peptides (24 , 117 , 118) .

At relatively high concentrations (5 µg/ml), CRH (range 0.1–100 µM) or urocortin (0.01–100 µM) injected directly into the skin will exert proinflammatory actions (100 , 119) . Theoharides et al. found that intradermally injected CRH and urocortin induced local mast cell degranulation in rats and mice (100 , 119) . Since CRH-R1 mRNA has been detected in mast cells, these effects may be mediated by CRH-R1 (100) . CRH-induced degranulation of mast cells is accompanied by increased vascular permeability (reduced by treatment with H1-receptor antagonists), suggesting that this effect may enhance local vascular permeability (100) . In earlier studies (67) , however, it was noted that CRF (up to 2.5 µM) did not directly affect or influence alpha-IgE or compound 48/80-stimulated release of histamine from preparations of isolated peritoneal rat mast cells. Arbiser et al. (116) have suggested that CRH is a stimulator of angiogenesis and others showed that anti-CRH antibodies can decrease inflammation in vivo (65 , 120) . Proinflammatory CRH effects have also been suggested based on observations made in other experimental models (10 , 25 , 65 , 100) that include experimental arthritis and uveoretinitis, as well as on studies in transgenic immunodeficient mice (65 , 120 121 122) . Nevertheless, the CRH-R2 knockout mouse has enhanced susceptibility to thermal injury (123) . Most recently, Zbytek et al. (124) found that human keratinocytes respond to CRH stimulation with increased production of proinflammatory cytokine IL-6 and anti-inflammatory IL-11 and decreased production of IL-1ß. Thus, current evidence indicates that CRH and urocortin may either suppress or stimulate inflammation depending on the experimental model or local tissue environment and metabolic cell status, in agreement with the hypothesis of Paez-Pereda et al. (21) .

In our own studies with the C57BL/6 mouse, exogenous CRH was found to have variable effects on the skin depending on the specific cellular population being evaluated and phase of the hair cycle (4) . Thus, functional assays using a skin organ culture system showed that CRH selectively stimulated epidermal DNA synthesis in telogen and anagen IV skin while inhibiting epidermal DNA synthesis in anagen II, whereas dermal DNA synthesis was enhanced (4) . These variable effects could be related to differential expression of different forms of CRH receptors during hair cycling and/or to indirect effects through differential production of cytokines or POMC peptides by cutaneous cells (4) .

Studies of established rodent melanoma lines have clearly shown the presence of functional CRH receptors (4 , 89 , 125) . Specific binding sites were detected in hamster AbC1 melanoma cells and CRH produced dose-dependent increases in intracellular calcium readily detectable at ligand concentrations as low as 10-10 M (4 , 89) . Since this effect had a rapid onset (within seconds) and was inhibited by depletion of extracellular calcium, it was proposed that this CRH signal transduction was directly coupled to activation of calcium channels (4 , 89) . Since the concentrations of sauvagine and urocortin required to achieve the same effect were 1000-fold higher (89) and hamster melanoma do express CRH-R1 mRNA, we proposed that this effect was mediated by CRH-R1 (4) . Most recently, it has been found that CRH inhibits proliferation of Cloudman mouse melanoma cells, apparently acting through CRH-R1 (125) . The latter work also reported that an s.c. injection of CRH inhibits growth of B16 melanoma in vivo (125) . This finding agrees with the observation by Tjuvajev et al. (126) of CRH inhibition of W256 rat mammary carcinoma growth in vitro and in vivo. The antineoplastic effect of CRH was connected to inhibition of proliferation and stimulation of differentiation of tumor cells and to decreased local vascular permeability and angiogenesis (126) . However, other researchers have found that human epithelial 293 tumor cells engineered to secrete CRH and implanted s.c., stimulate angiogenesis, which in turn stimulates tumor growth (116) . Thus, the CRH effect on tumor growth in vivo can be complex and may depend on experimental models and functional activity of target cell populations.

Human skin
CRH-R1 mRNA was detected by RT-PCR combined with Southern hybridization in human skin biopsy specimens from normal scalp, compound nevus, basal cell carcinoma and perilesional facial skin, and normal and malignant melanocytes and keratinocytes (60 , 61 ; Fig. 6 ). However, retrospective analysis of the RT-PCR data in relation to genomic structure revealed that the primers used for amplification were located at exons 3 and 7 (cf. figure 3 from ref 60 ). Therefore, the 334 bp transcript identified in all the specimens above as well as in melanoma and squamous cell carcinoma cells, UVB-treated melanocytes, and UVB- and TPA-treated keratinocytes (cf. figure 3 from ref 60 ) represented a CRH-R1 variant with spliced-out exon 6 (i.e., CRH-R1{alpha}) (Fig. 7 ). An amplified fragment of 421 bp representing an additional variant, containing exon 6 (i.e., CRH-R1ß), was present only in the scalp; cf. figure 3 from ref 60 . The use of alternative primers based on sequences from exons 7 and 8 demonstrated a transcript of 380 bp corresponding to the predicted CRH-R1 mRNA fragment (61) . Using the same primers, Funasaka and colleagues found expression of CRH-R1 in all the lines of normal melanocytes and melanoma cells tested (62) . Hence, we have now revised our previous interpretation of lack of detection of CRH-R1 in nonstimulated (control) melanocytes (cf. figure 3 in ref 60 ) as being the result of alternative splicing of exon 3, which contained the primer sequence. In addition, UVB stimulates in melanoma cells expression of CRH-R1 mRNA isoform with a shorter coding sequence than the CRH-R1{alpha} (Fig. 7) . Immunocytochemistry studies have localized CRH-R1 immunoreactivity to epidermal and follicular keratinocytes in the human skin (127 ; Fig. 8 ). Kono et al. (63) detected CRH-R1 mRNA and protein in the pilosebaceous unit and nevocytes of normal human skin. In agreement with those studies, we also detected CRH-R1 antigen in biopsies of nevomelanocytic lesions and specimens of malignant melanoma (not shown).



View larger version (32K):
[in this window]
[in a new window]
 
Figure 6. Expression of the CRH R1 receptor gene in the human keratinocyte cell line HaCaT (4) using primers directed to sequences spanning exon 3/4 border and exon 10, as described by Rodriguez-Linares et al. (101) . Lane 1: DNA ladder; lanes 2 and 3: positive controls (AtT20 corticotrophic tumor cell line).



View larger version (16K):
[in this window]
[in a new window]
 
Figure 7. UVB induces alternative splicing of the CRH-R1 gene. A) Northern blot hybridization of SKMEL1888 melanoma mRNA with human CRH-R1 cDNA shows changes in expression of CRH-R1 mRNA transcripts of ~2.6–2.7 kb (arrow) and 2–2.2 kb (arrowhead). Left: RNA ladder (kb); control melanoma cells (lane 1); melanoma cells treated with 20 mj/cm2 of UVB (lane 2). The experiments were performed by Dr. Ashok Chakraborty and Dr. Gennady Ermak and the experimental conditions were described by Chakraborty et al. (137) and Slominski et al. (60) . B) Southern blot hybridization of human CRH-R1 cDNA with RT-PCR amplification products: CRH-R1 mRNA of 334 bp is present in pituitary (lane 1), neonatal melanocytes (lane 2), keratinocytes (lane 3), and control SKMEL188 melanoma cells (lane 4); it is absent in melanoma cells treated with 20 mj/cm2 of UVB (lane 5). Arrowhead points to a shorter transcript present only in UVB-treated melanoma cells. Left: DNA ladder (kb). The experiment was performed by Dr. Ashok Chakraborty and Dr. Gennady Ermak and the experimental conditions were described in Slominski et al. (60) .



View larger version (190K):
[in this window]
[in a new window]
 
Figure 8. Detection of CRH-R1 antigen (B) in epidermal (arrowhead) and hair follicle (arrow) keratinocytes. A) Negative control incubated with nonimmune goat serum. The immunostain was performed with commercially available anti-CRH-R1 antibody. From Quevedo et al. (127) , © 2001 by the Society for In vitro Biology (formerly the Tissue Culture Association). Reproduced with permission of the copyright owner.

Functional characterization of the CRH receptors has been performed recently in established lines of human melanoma, immortalized HaCaT keratinocytes, and neonatal keratinocytes. CRH stimulated increases in intracellular Ca2+ in human melanoma cells in a manner similar to that observed in the hamster AbC-1 melanoma line, i.e., the effect was rapid (within seconds), dose dependent, and inhibited by extracellular calcium depletion (4 , 89) . It was also CRH peptide specific, since significantly higher concentrations of the related sauvagine and urocortin peptides were required to achieve comparable effects (89) . Therefore, the CRH-induced intracellular Ca2+ accumulation in human melanoma appears to be mediated by direct coupling of the CRH-R1 to calcium channels (4 , 89) . Activation of CRH receptors resulted in the inhibition of DNA synthesis in melanoma cells (A. Slominski and B. Zbytek et al., unpublished results). CRH and urocortin also inhibited growth in HaCaT keratinocytes, but only at concentrations in the nanomolar range and not with micromolar levels (1 , 128) . Characterization of this CRH growth inhibitory effect showed that it was unaffected by supplementation of the culture medium with serum; CRH inhibitory potency was also greater than urocortin or urotensin I (1 , 128) . Flow cytometric analysis confirmed that the inhibitory effect of CRH occurred only at nanomolar concentrations and showed that it was due to the inhibition of the G1 to S transition of the cell cycle (Fig. 9 ). CRH has been shown to have growth inhibitory effects in the AT-t20 pituitary cell line (1 , 43 , 44) and in mammary cancer cells (126) . Testing with a specific CRH-R1 agonist, [D-Glu20]-CRH, revealed that HaCaT keratinocytes were inhibited. The dose-response relationship was bimodal, with lack of a CRH inhibitory effect at micromolar concentrations. This phenomenon may have resulted from the production of growth stimulators that counteract the growth-inhibitory action of the peptide. One such factor could be IL-6, since production of this cytokine is stimulated by micromolar concentrations of CRH (124) ; direct addition of IL-6 to the culture medium stimulates keratinocyte proliferation (129) , including HaCaT cells (B. Zbytek, unpublished results). Most recently, we have shown that CRH also inhibits the proliferation of normal neonatal keratinocytes and stimulates in a dose-dependent fashion the interferon-{gamma}-stimulated expression of hCAM and ICAM-1 adhesion molecules and of the HLA-DR antigen (127 ; Fig. 10 ). These findings led us to propose that the activation of CRH-Rs by high doses of the ligand may induce a shift in cellular energy metabolism away from proliferation activity, toward the up-regulation of immunoactivity in human neonatal keratinocytes. The result of this mobilization of the CRH/CRH-R1 system could be a cascade of pathogenic events addressed at enhancing the intensity of the stress response while restricting its field of activation.



View larger version (36K):
[in this window]
[in a new window]
 
Figure 9. CRH (10-10-10-9 M) inhibits transition from G1 to S phase of the cell cycle in HaCaT keratinocytes. Quadruplicate cultures were incubated in DMEM without or with CRH at the concentrations listed (CRH or vehicle was added every 6 h). After 16 h the cells were collected, washed with PBS, and fixed in ice-cold 70% ethanol. Flow cytometric analysis was performed using a Coulter Epics XL cytometer. The data from cytometer were analyzed using WinMDI 2.8 (freeware from Joe Trotter, Scripps Institute, La Jolla, CA). After exclusion of cellular debris in FS/SS and doublets in Aux/PI linear gates, appropriate PI linear histograms were created. The histograms were analyzed with Cylchred (freeware from Terry Hoy, UWCM, Cardiff, UK), which automatically estimated percentage of cells in distinct cell cycle phases. *Statistically significant differences (P<0.05, unpaired t test).



View larger version (92K):
[in this window]
[in a new window]
 
Figure 10. CRH stimulates INF-{gamma}-induced expression of cell adhesion molecules and HLA-DR in human neonatal keratinocytes. B, D, F) hCAM, HLA-DR, and ICAM-1 expression, respectively, in human keratinocytes pretreated with INF-