(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
,
ALEXANDER PISARCHIK*,
BLAZEJ ZBYTEK
,
ELIZABETH A. LINTON
,
JOSEPH E. MAZURKIEWICZ and
EDWARD T. WEI**
* Department of Pathology, University of Tennessee Health Science Center, Memphis, Tennessee 38163, USA;
Department of Internal Medicine, Southern Illinois University, Springfield, Illinois 62701, USA;
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;
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
|
|---|
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
|
|---|
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
|
|---|
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 
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
|
|---|
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)
.
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 (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)
|
|---|
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 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.
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 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
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 and ß isoforms have been cloned in the rat,
mouse, and tree shrew (2
, 11
, 75
, 76
, 85
, 86)
, and the
CRH-R2-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 and CRH-R2ß are expressed in
peripheral tissues and the brain, whereas in rodents CRH-R2 is found
predominantly in the brain and CRH-R2ß in the periphery (cf. refs
45
, 76
, 77
). The CRH-R2 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-, 1-ß, 1-c, 1-d, and
2-) 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 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) (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 (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 (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--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--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- | |
TOP
ABSTRACT
INTRODUCTION
—
