BIOLOGY OF REPRODUCTION 53, 1417-1428 (1995)

Stress-Induced Genetic Expression of a Selective Corticotropin-Releasing Factor-ReceptorSubtype Within the Rat Ovaries: An Effect Dependent on the Ovulatory Cycle'

Rossella E. Nappi and Serge Rivest2

Laboratory of Molecular Endocrinology, CHUL Research Center and Laval UniversityQuebec, Canada G1 V 4G2

ABSTRACT

The identification of several sources of corticotropin-releasing factor (CRF) outside the brain, including the gonads, has suggested theintriguing possibility that CRF of systemic origin can influence gonadal function under both normal and stressful conditions. However, theexact sites of action and the type of cells targeted by this stress-related neuropeptide in the ovaries remain unknown. The present studytherefore investigated the effect of acute immobilization stress on the distribution of the recently cloned CRF receptor (CRF,, CRF,,, andCRF,,) genes in the ovaries of adult cycling female rats (200-250 g; 14 h light, lights-on at 0600 h). Reproductive stages were verified bydaily vaginal smears taken each morning for a minimum of 3-4 cycles prior to the experiment. Three hours after the start of a 90-minimmobilization session, rats were deeply anesthetized and transcardially perfused with a solution of 4% paraformaldehyde on the morning(1100 h) and on the afternoon (1700 h) of proestrus and diestrus-2. Frozen ovaries were mounted on a microtome, cut into 30-pm slices,and then processed for the detection of mRNAs encoding CRF1, CRF2 , or CRF2a receptors by in situ hybridization histochemistry using 35S-labeled riboprobes. Whereas the ovaries displayed barely detectable levels of CRF1 receptor mRNA in control and in stressed animals onthe morning of proestrus and the day of diestrus-2, a positive signal for this transcript was detected in stroma cells and in the thecasurrounding the ovulatory follicles during the afternoon of proestrus. Excluding the cumulus oophorus, which showed a light expressionof the mRNA encoding the type 1 CRF receptor, granulosa cells were completely devoid of transcript in Graafian follicles as well as in CL,regardless of the stage of maturation. Interestingly, immobilization stress induced a marked expression of CRF, receptor mRNA in thestroma cells in the afternoon of proestrus, suggesting that the ovaries may be sensitive to acute neurogenic challenge during the preovu-latory stage. On the other hand, CRF2, and CRF2a receptor mRNAs were undetectable both in control and stressed animals throughout theestrous cycle. These results provide clear evidence that the gene encoding the CRF, receptor but not the type 2 receptors can be finelyinduced in selective ovarian compartments in both control and stressful conditions during the gonadal life cycle. The temporal andanatomic selectivity of the ovarian periovulatory CRF, receptor gene expression may suggest a critical biological action of CRF during theovulatory process and suggests that the intraovarian environment may influence the stress-induced transcription of a selective CRFreceptor subtype within the ovary.

INTRODUCTION

The neuroanatomical distribution and regulation of thegene encoding corticotropin-releasing factor (CRF) in the ratbrain are consistent with the pivotal role of this neuropeptidein the stress response [1-4]. The antireproductive effect ofstress-related hormones has been extensively studied at allthree levels of the hypothalamic-pituitary-gonadal axis, andCRF is recognized as a potent mediator of the stress-relatedreproductive failure influencing LHRH-secreting neuronal ac-tivity [5-7]. Moreover, the concept of a strict bidirectionallinkage between activity of the hypothalamic-pituitary-adre-nal (HPA) axis and ovarian steroidogenesis is supported byvariations in the HPA axis response to stress during the es-trous cycle [8] and our recent report [9] indicating that theestrous cycle can interfere with the capacity of immobiliza-tion stress to stimulate the biosynthesis of CRF type 1 receptorin the hypothalamic paraventricular nucleus (PVN). The

Accepted August 11, 1995.Received May 4, 1995.1Supported by the Medical Research Council (MRC) of Canada. Rossella E. Nappi is a

postdoctoral fellow supported by IRCCS Policlinico S. Matteo, University of Pavia, Italy. SergeRivest is a Medical Research Council Scholar.

2 Correspondence: Serge Rivest, Laboratory of Molecular Endocrinology, CHUL ResearchCenter and Laval University, 2705, boul. Lauder, Quebec, Canada GIV 4G2. FAX: (418) 654-2761; e-mail: Serge.Rivest@crchul.ULaval.Ca.

identification of several systemic sources of CRF [10-14] alsosuggests that activation of the CRF system outside the braincan influence the ovulatory cycle and interfere locally withgonadal function under stressful conditions. Indeed, immu-noreactive CRF (CRF-ir) and CRF mRNA were identified inrat testes, and the fact that CRF secreted by Leydig cells neg-atively affects testicular steroidogenesis provided in vitro ev-idence that this peptide can act as an antireproductive hor-mone in the gonads [13-17].

A large variety of neuropeptides, most of which seem toact as putative intraovarian regulators, were localized in thefollicular unit by immunohistochemistry, in situ hybridiza-tion, and receptor studies in several species [18]. Recently,Mastorakos et al. [19] detected CRF-ir and CRF binding sitesin rat ovaries; positive immunostaining for CRF and auto-radiographic localization of CRF receptor were present pri-marily in the ovarian theca and stroma [19], the functionalcounterpart of the Leydig cells in the testis. No apparentdifferences were observed in CRF-ir staining and CRF bind-ing throughout the estrous cycle [19]. Likewise, CRF-ir wasfound in normal and polycystic human ovaries and in thefollicular fluid [20]. It is possible that CRF is involved as aproinflammatory agent within the ovary during the ovula-tory process, luteal development, luteolysis, and follicular

1417

NAPPI AND RIVEST

atresia because CRF is expressed in inflammatory sites[11, 21, 22], and the ovary is now considered a crucial siteof interaction between the immune and the endocrine sys-tems [23, 24]. White blood cells, mainly macrophages, con-stitute a major cellular component of the interstitial ovariancompartment [251, whereas migrating leukocytes are a po-tential source of cytokines involved in the inflammatory-likeovulation-associated phenomena [23, 24]. In particular,apart from the largely recognized effects on gonadal ste-roidogenesis in several species [26-28], cytokines promoteovulation in the perfused rat ovary and show cyclic varia-tions at various stages of the estrous cycle [29]. A similarphysiological involvement of CRF during the proestrusstage may thus be proposed. The possible importance ofovarian CRF in the local regulation of steroid biosynthesisis supported by the presence of CRF-ir in the human poly-cystic ovary, which may reflect a role in the occurrence ofhyperandrogenism [20]. Moreover, CRF inhibits in a dose-dependent fashion FSH-stimulated estradiol productionfrom rat granulosa cells in vitro [30].

The study of the ovarian distribution of the recentlycloned CRF receptor subtypes (types 1, 2a, and 2[) [31-34]is a powerful approach to investigating the potential role ofCRF and the exact sites of action of the peptide in the ova-ries during both normal and emergency situations for theorganism. Thus, in the present study we evaluated the effectof acute neurogenic stress immobilization on expression ofCRF1, CRF2,,, and CRF2, receptor transcripts in the ovaries ofadult female cycling rats.

MATERIALS AND METHODS

Animals

Adult female Sprague-Dawley rats (-200-250 g) wereacclimated to standard laboratory conditions (14L:10D cy-cle; lights-on at 0600 h) and given free access to rat chowand water. Reproductive stages were verified by daily vag-inal smears taken every morning (700-730 h) for a mini-mum of 3-4 cycles prior to the experiment. Three to sixrats were used for each group, and all protocols were ap-proved by the Laval University Animal Welfare Committee.This experiment was conducted twice.

Acute Immobilization Session

Rats were placed in individual immobilizers for 90 minand, at 1100 h and 1700 h on the morning and afternoon ofproestrus and diestrus-2, were killed 3 h after the beginningof the session. The time for killing the rats was chosen onthe basis of our previous results in male and female rats inwhich a strong hybridization signal for the CRF receptormRNA was observed in the hypothalamic PVN 90 min aftera 90-min stress session. Immobilization was accomplishedby use of adjustable restraining cages (Centrap Cages, 01-

282-10; Fisher Scientific Co., Pittsburgh, PA). Cages wereadjusted tightly so that the animals were unable to movewhen in the immobilizer. The rats were anesthetized withan i.p. injection of 0.3 ml of a mixture of ketamine hydro-chloride (80 mg/kg) and xylazine (10 mg/kg) and were rap-idly perfused transcardially with saline, followed by 4% par-aformaldehyde (PF) in 0.1 M borax buffer (pH 9.5 at 4°C).Ovaries were removed, cleaned of fat, postfixed, and thenplaced in a solution of 4% PF-10% sucrose overnight at 4°C.Frozen ovaries were mounted on a microtome (Reichert-Jung, Cambridge Instruments Company, Deerfield, IL) andcut into 30-gm sections. The slices were collected in a coldcryoprotectant solution (0.05 M sodium phosphate buffer,30% ethylene glycol, 20% glycerol) and stored at - 20°C.The mRNAs encoding each specific CRF receptor subtypewere assayed by in situ hybridization histochemistry.

In Situ Hybridization Histochemistry

Localization by histochemical hybridization of each tran-script (CRF1 , CRF2,,, and CRF2, receptor mRNAs) was carriedout in a 1-in-4 series (every fourth section) of ovary slices withuse of 35 S-labeled cRNA probes. Protocols for riboprobe syn-thesis, hybridization, and autoradiographic localization ofmRNA signal were adapted from Simmons et al. [35]. All so-lutions were treated with diethyl pyrocarbonate (DEPC) andsterilized to prevent RNA degradation. Tissue sectionsmounted onto poly-L-lysine-coated slides were desiccated un-der a vacuum overnight, fixed in 4% PF for 30 min, and di-gested by proteinase K (10 gg/ml in 100 mM Tris HC1, pH 8.0,and 50 mM EDTA) at 37°C for 25 min. Thereafter, the ovarysections were rinsed in sterile DEPC water followed by a so-lution of 0.1 M triethanolamine (TEA, pH 8.0), acetylated in0.25% acetic anhydride in 0.1 M TEA, and dehydrated throughgraded concentrations of alcohol (50, 70, 95, and 100%). Aftervacuum drying for a minimum of 2 h, 90 1l of a hybridizationmixture (10 7 cpm/ml) was spotted onto each slide, sealed un-der a coverslip, and incubated at 60°C overnight (-15-20 h)in a slide warmer. Coverslips were then removed, and theslides were rinsed in 4-strength SSC at room temperature. Thesections were digested by Ribonuclease A (20 pg/ml, 370C,30 min), rinsed in descending concentrations of SSC (double-strength, single-strength, 0.5-strength; single-strength SSC =0.15 M NaC, 15 mM trisodium citrate buffer, pH 7.0), washedin 0.1-strength SSC for 30 min at 65C, and dehydrated throughgraded concentrations of alcohol. After being dried for 2 hunder a vacuum, sections were exposed at 4°C to x-ray film(XAR5; Eastman Kodak, Rochester, NY) for 24-48 h, defattedin xylene, and dipped in NTB2 nuclear emulsion (Kodak; di-luted 1:1 with distilled water). Slides were exposed for 10-14 days, developed in D19 developer (Kodak) for 3.5 min at15°C, and fixed in rapid fixer (Kodak) for 5 min. Thereafterthe tissues were rinsed in running distilled water for 1-2 h,counterstained with thionin (0.25%), dehydrated through

1418

CRF RECEPTOR mRNA IN STRESSED RAT OVARIES

graded concentrations of alcohol, cleared in xylene, and cov-erslipped with DPX (Electron Microscopy Sciences, FortWashington, PA).

Complementary RNA Probe Synthesis and Preparation

Specific rat CRF receptor probe (1.3 kb) was generatedfrom the PstI-PstI fragment of rat prCRF PP1.3-BS cDNA (Dr.W. Vale, Peptide Biology Laboratory, The Salk Institute) sub-cloned into pBluescript II SK (Stratagene, La Jolla, CA) andlinearized with BamHI and HindIII (Pharmacia) for antisenseand sense probes, respectively [32]. The pBluescript (SK +)plasmids containing either a 275-bp insert of rat CRF2. recep-tor cDNA or a 200-bp insert of a rat CRF2, receptor cDNA(Dr. T. Lovenberg, Neurocrine Biosciences Inc., San Diego,CA [34]) were linearized with HindIII and BamHI to generateantisense and sense probes, respectively. These two probes(CRFz2 and CRFz2 ) have no overlap with one another andhave no similarity to the CRF receptor probe (T. Lovenberg,personal communication). Radioactive cRNA copies weresynthesized by incubation of 250 ng linearized plasmid in 6mM MgCl2, 40 mM Tris (pH 7.9), 2 mM spermidine, 10 mMNaC1, 10 mM dithiothreitol, 0.2 mM ATP/GTP/CTP (Fisher;Promega, Madison, WI), [a35S]UTP, 40 U RNAsin (Promega),and 20 U T7 (CRF, receptor antisense; CRF2a and CRF2 re-ceptor sense probes) and T3 (CRF receptor sense; CRF2. andCRF2, receptor antisense probes) RNA polymerase (Fisherand Promega) for 60 min at 37°C. Unincorporated nucleo-tides were removed by the ammonium-acetate method; 100gl. of DNAse solution (1 1ll DNAse, 5 Ill of 5 mg/ml tRNA, 94jpI of 10 mM Tris/10 mM MgCl2) was added, and 10 min lateran extraction was accomplished by use of a phenol-chloro-form solution. The cRNA was precipitated with 80 1l of 5 Mammonium acetate and 500 ill of 100% ethanol for 20 minon dry ice. The pellet was washed with 500 pil ethanol, driedin a vacuum centrifuge for 10 min, and resuspended in 100Il of 10 mM Tris/1 mM EDTA. A concentration of 107 -cpmprobe was mixed into 1 ml of hybridization solution (500 .llformamide, 60 ll 5 M NaCl, 10 Il 1 M Tris [pH 8.0], 2 tl 0.5M EDTA [pH 8.0], 50 pl 20-strength Denhardt's solution, 200$al 50% dextran sulfate, 50 ill 10 mg/ml tRNA, 10 ll 1 M di-thiothreitol (DTT), [118 il DEPC water-volume of probeused]). This solution was mixed and heated for 5 min at 65°Cbefore being spotted on slides.

RESULTS

Distribution of Type 1 and Type 2 (a and fi) CRF-ReceptormRNAs in the Rat Ovary

Figure 1 illustrates representative examples of CRF recep-tor gene expression in the ovaries of stressed (left) and con-trol (right) female rats on the morning and the afternoon ofproestrus and diestrus-2. Almost no positive hybridizationsignal for the mRNA encoding the type 1 CRF receptor was

observed on x-ray film in the ovaries of control or stressedanimals on the morning of proestrus or on the day of dies-trus-2. In the afternoon of proestrus, however, a low to mod-erate endogenous expression of CRF receptor transcript wasdetected in the stroma cells and theca surrounding the ovu-latory follicles of control rats (Pro-17, Control). Interestingly,immobilization stress caused a marked activation of CRF re-ceptor gene transcription in the stroma cells during this ovu-latory period without affecting the genetic expression of thisreceptor in other periods of the estrous cycle. Tissues hy-bridized with sense probe did not exhibit detectable signalin any of the sites that showed positive signal with antisenseprobe (Fig. 2, top). In contrast to CRF-R,, no positive signalwas hybridized with antisense probes to detect CRF2a andCRF2, receptor mRNAs in the ovaries of control (data notshown) and immobilized rats (Fig. 2, bottom panels). Thesefindings indicate that CRF of type 2 receptors are not presentin the rat ovary throughout the estrous cycle under both con-trol and stressful conditions.

Site-Specific Induction of CRF, Receptor Transcription inthe Ovaries of Stressed Female Rats during the Afternoonof Proestrus

Figure 3 shows an example of Graafian follicles of con-trol (right) and stressed (left) female rats in the afternoon ofproestrus. The mRNA encoding the type 1 receptor wasspontaneously expressed in the theca surrounding the ovu-latory follicles of control rats at this specific time of the cy-cle. No differences were found in such endogenous ex-pression of the CRF, receptor transcript in animals killed 3h after the beginning of the 90-min immobilization session.The selectivity of CRF receptor gene expression within thetheca surrounding ovulatory follicles in the afternoon ofproestrus in presented by Figure 4. Excluding the cumulusoophorus, which demonstrates a light expression of thegene encoding the CRF type 1 receptor, granulosa cellswere completely devoid of this transcript in the ovaries ofboth control and stressed animals.

Immobilization stress caused activation of CRF, receptortranscription in a defined ovarian compartment during theafternoon of proestrus (Fig. 5, left). Regardless of the stageof maturation, the CL did not express CRF receptor mRNA,whereas the interstitial compartment exhibited a positivehybridization signal 3 h after the acute immobilization ses-sion (Fig. 5, bottom left). In fact, strong levels of silver grainswere observed in the ovarian stroma (Fig. 6, right) and in-terstitial cells in anatomic proximity to blood vessels (Fig.6, left) 90 min after the end of the acute stress session ac-complished in the afternoon of proestrus.

DISCUSSION

The present study provides evidence that the gene encod-ing the type 1 but not the type 2 (a and 3) CRF receptors can

1419

420 NAPPI AND RIVEST

CRF RECEPTOR mRNA IN STRESSED RAT OVARIES

be finely induced in selective ovarian compartments in bothcontrol and stressful conditions during the gonadal life cycle.Indeed, whereas the ovary displayed barely detectable levelsof CRF1 receptor mRNA in control and stressed animals onthe morning of proestrus and the day of diestrus-2, a clearsignal for this transcript was detected in stroma cells and thetheca surrounding the ovulatory follicles during the after-noon of proestrus. Excluding the cumulus oophorus, whichshowed a light expression of CRF1 receptor mRNA, granulosacells were completely devoid of this particular transcript inGraafian follicles, as well as in CL, regardless of the stage ofmaturation. Interestingly, immobilization stress induced amarked expression of the gene encoding the type 1 CRF re-ceptor in the stroma cells in the afternoon of proestrus, sug-gesting that the ovary may be sensitive to acute neurogenicchallenge during the preovulatory stage.

The ovary has been demonstrated to be a peripheralsource of CRF [19, 20]. The chromatographic mobility of ratand human extracts can be superimposed on that of syntheticrat/human CRF ([19, 20] and personal observation), indicat-ing that ovarian CRF is similar to that produced by the ratand human hypothalamus, human placenta, rat Leydig cells,and peripheral inflammatory sites [2, 12, 14, 21, 36]. Also re-ported were CRF immunostaining in stroma and theca cellssurrounding follicles at different stages of maturation, in asubpopulation of cells within the CL, and in the oocytes ofthe antral follicles, as well as a sporadic positive staining ingranulosa cells of ovulatory follicles [19]. Although the pres-ence of CRF-ir in the ovaries of cycling rats is clearly estab-lished, the potential roles of such neuropeptides remain tobe elucidated. This picture is further complicated by the factthat we can only hypothesize about the possible site(s) ofCRF biosynthesis and release within the ovary. The patternof distribution of its own receptor and the present data show-ing a strong selective induction of the CRF1 receptor tran-scription in the thecal layer surrounding the ovulatory folli-cles and, more lightly, in the cumulus oophorus in theafternoon of proestrus may indicate that CRF constitutes anadditional neuropeptidergic signal involved in the gonadalreproductive cycle. Indeed, the striking temporal and ana-tomic selectivity of the ovarian periovulatory CRF1 receptorgene expression emerging from our study indicates that acritical biological action for this neuropeptide may be iden-tified mainly during the ovulatory process.

FIG. 1. Distribution of mRNA encoding type 1 CRF receptor in ovaries of stressedand control female rats on morning 1100 h) and afternoon (1700 h) of proestrus(PRO-11 and PRO-17) and diestrus-2 (DI-1 1 and D-17). These sections (30 pm) ofboth stressed and control cycling rat ovaries display positive signal on x-ray filmfor CRF, receptor mRNA selectively on afternoon of proestrus. A low to moderateendogenous expression of CRF, receptor transcript was detected in stroma cellsand theca surrounding ovulatory follicles of control rats on afternoon of proestrus.Note the marked induction of mRNA encoding CRF, receptor in stroma cells ofstressed ovaries during this ovulatory period (PRO-17, left).

One of the possible interpretations of the physiologicalrelevance of an up-regulation of the CRF1 receptor mRNAin the afternoon of proestrus is that CRF, through this par-ticular receptor subtype, may actively participate in the finalmaturation and rupture of the follicular unit and in the initialluteinization phenomena as well as exert a potential effecton steroid biosynthesis during ovulation. Hence, it is likelythat CRF may behave as a proinflammatory factor and bepart of the whole cascade of concomitant and/or coordi-nated periovulatory events [37]. This phenomenon includesa complex cohort of immune-related substances, namelycytokines, whose participation in the ovulatory process andin ovarian steroid hormone regulation has been stronglyinvestigated. Indeed, macrophage-derived cytokines canpromote ovulation in the perfused rat ovary [38] and stim-ulate the production of progesterone and androgens fromgranulosa and thecal layers, respectively, in preovulatoryfollicles [27, 39]. On the other hand, a gonadotropin-depen-dent preovulatory induction of interleukin-1 (IL-1) tran-script was reported in the rat as well as human ovary[28, 40], and LH seems to be the key factor leading to theproduction of cytokines and chemotactic stimuli in theovary. Expression of IL-13 mRNA in human peripheral leu-kocytes is sensitive to gonadal hormones [41], confirmingthat steroid-secreting cells may exert a modulatory influ-ence on immune-related peptides. Complex interplay be-tween the hormonal intraovarian environment and expres-sion of intragonadal regulators is a constant feature ofovarian physiology and the pivotal role of gonadotropins inthe promotion of follicular development. The demonstra-tion of a preovulatory-induced transcription of CRF, recep-tor could suggest that the acquisition of CRF responsivenessis central to the final maturative processes of the Graafianfollicle. An example of the reciprocal interactions existingbetween the hormonal intragonadal environment and theexpression and regulation of another neuropeptide oper-ating in the rat ovary is given by the GnRH system, whichincludes GnRH gene transcript, GnRH binding sites, andGnRH receptor mRNAs [42, 43]. The importance of ovarianGnRH during the periovulatory period is further supportedby a recent report indicating a marked decline in ovarianGnRH receptor gene expression 12 h after HCG injection ineCG-primed immature rats [44].

However, whether CRF participates in the biochemicalovulation-associated events more as a proinflammatory fac-tor or as a neuropeptidergic signal involved in the modu-lation of steroidogenetic activity is still an open question.Also remaining unknown is the nature of the modulatoryfactors influencing the transcription and stability of ovarianCRF, receptor mRNA and the biological activity of CRF. Onthe other hand, because CRF-ir is present throughout theentire estrous cycle, we cannot rule out the possibility thatthe peptide may be potentially involved in gonadal biolog-ical functions other than ovulation directly or via other me-

1421

NAPPI AND RIVEST1422

o0S C

CSo 'B-u

c

'>o

*a' o

, C

:- 0o .3

E

.=

o,

a) a)

oo

on

co E

cp-,o

a)E-,=

Xo

0 .C

E

0 a)

a )

C O

0< a)

C.) 0

CRF RECEPTOR mRNA IN STRESSED RAT OVARIES 1423

g E000

o 0

l.

- s °

Eor

0 ._

v c.

, 0 OZCL

E

ao,

_C _

W O Ca,

2 o-0 .X

CO

.2

aD . g

-c cr

O)OEZ 0

-o a

_r OM .0

0>0.

d -

0. o 2 X

NAPPI AND RIVEST1424

LC

c cB U

.o

o C

Es

c 'U

CL0-

E .--

0o.0 -

0r

ga,

ox

0-

.0s

" ->

cmCO

=D0

cmE

o ,

*_ -C

oCo0 =

( oc

0o. c a

O

;, 0

. o

CRF RECEPTOR mRNA IN STRESSED RAT OVARIES 1425

0E 3

o

0.o

- _

r0

o ._

. C

.U EO

o

ag

oc

Emo t0

E

.0

· o o

-o

O

- 00

o 0u-

L Tcr

u 0

0.y

Lfl0

.0 CL

1426 NAPPI AND RIVEST

0-0

X

.) o

0o

E

*:C

.

OEog

,0 0

o _

C)

0 0

e

0 C

E

0~ o

E Ooo

m Z:2Xm oe

e EE o

c

0.cc:

R- s

0) E02ou

000

uj 0

o0 B~jCin

CRF RECEPTOR mRNA IN STRESSED RAT OVARIES

diators, such as proopiomelanocortin (POMC)-related pep-tides. The ovarian content of immunoreactive J3-endorphin(3B-END-ir) and POMC mRNA increases after eCG adminis-tration [45], and granulosa cells located near the oocytesexhibit positive signals for -END-ir [46]. The presence ofCRF, receptor transcript and binding sites in the cumulusoophorus, together with evidence of CRF-ir in mature oo-cytes, provides anatomic evidence of a possible interactionbetween POMC-derived peptides and CRF within selectivecompartments of the ovaries during normal and emergencycircumstances in the female organism.

CRF mRNA and peptide were detected in peripheral leu-kocytes [11, 21, 22]. A large population of immune-relatedcell types is normally resident in the interstitial ovarian com-partment in anatomic vicinity to stroma blood vessels andperifollicular capillaries [23]. Whether selective cell types inthe ovaries contain the biosynthetic machinery to produceCRF is still under debate. Indeed, we (unpublished results)and others (C. Rivier, personal communication) were un-able to detect a positive hybridization signal for CRF mRNAin somatic ovarian cells. However, in situ hybridization his-tochemistry might not be sensitive enough to detect verylow levels of transcript that could be visible only by poly-merase chain reaction or other highly sensitive techniques.On the other hand, the massive ovarian infiltration by sev-eral types of white blood cells [47, 48] and the concomitantinflammatory-like phenomena occurring at the time of ovu-lation [37] could be responsible for delivering CRF locally,where it could exert, throughout interaction with its recep-tor, paracrine effects at the level of the different ovariancompartments. Our data showing a relative abundance ofCRF1 receptor transcripts in the loose connective tissue ar-eas, in the medullary regions surrounding the blood vessels,and in the thecal layer near perifollicular capillaries supportsuch a hypothesis. Moreover, the absence of a clear co-lo-calization between the cells expressing the CRF receptormRNA and the CRF-ir cells during the afternoon of proestrus(data not shown) seems to rule out the presence of auto-crine events in the ovarian CRF system.

The sympathetic nervous system could represent anotherpotential source of CRF delivery within the ovary [49]. Im-munocytochemical localization of CRF in the rat spinal cordhas been identified [50], and CRF may be released locallyby the fibers innervating blood vessels, interstitial tissues,and developing follicles [51]. A potential modulatory role ofadrenergic agents in follicular development, steroid secre-tion, and ovulation remains possible. Using isolated rat ova-ries, Ferruz et al. [52] observed a profile of norepinephrinerelease during the estrous cycle, and this phenomenon isinfluenced by gonadotropins, particularly during the day ofproestrus. However, little is known about the functional in-teractions between the activity of the sympathetic nervoussystem and the ovarian CRFergic system.

The fact that immobilization can stimulate transcription of

the gene encoding the type 1 receptor in the stroma cells inthe afternoon of proestrus, apparently without affecting thespontaneous CRF, receptor mRNA expression in the thecallayer around the ovulatory follicles, supports the concept ofstress-induced CRFergic activity within selective ovarian sub-divisions. These results also suggest that the ovary is moresusceptible to the local antireproductive influence of CRFduring a specific period of the ovulatory cycle in severelystressed animals. Indeed, the intraovarian environment maysomehow influence the gonadal responsiveness to the im-mobilization session. The exact physiological significance ofa selective reactivity of the ovaries limited to the afternoonof proestrus and the pathways involved in such stress-in-duced up-regulation of the CRF, receptor mRNA in the theca/interstitial compartment must be clarified. A possible involve-ment of local dynamic vascular and/or neural changesoccurring during stressful conditions may be proposed. Werecently reported that immobilization stress can selectivelyinduce the gene encoding the type 1 CRF receptor in theparvicellular division of the PVN and that such stress-inducedCRF-R5 transcription was significantly higher on the morningof proestrus [9]. Consequently, the gonadal steroid milieumay modulate the neuroendocrine CRFergic system in thefemale rat brain during stress. The observation that estradiolcan stimulate human CRF gene transcription [53] also sup-ports a direct influence of steroid hormones on the local ac-tivity of the ovarian CRFergic system. A better understandingof the relationship between the ovulatory cycle and the localinfluence of stressful stimuli will help to clarify the biologicalsignificance of the presence of CRF-ir cells in the gonadal lifecycle and the possible sites of biosynthesis and release of thepeptide within the ovary.

In conclusion, the mRNA encoding the type 1 CRF recep-tor can be finely induced in selective ovarian compartmentsunder both control and stressful conditions during the go-nadal life cycle. The temporal and anatomic selectivity of theovarian periovulatory CRF, receptor gene expression indi-cates that a critical biological action for CRF may be identifiedduring the ovulatory process and that the intraovarian envi-ronment may influence the stress-induced transcription ofthat particular CRF receptor subtype in rat ovaries.

ACKNOWLEDGMENTS

We thank Dr. Wylie W. Vale (PBL, The Salk Institute, La Jolla, CA) for the generous giftof rat CRF, receptor cDNA, Dr. Timothy W. Lovenberg (Department of Molecular Biology,Neurocrine Biosciences, Inc., San Diego, CA) for the gift of plasmids containing rat CRFRand CRF,2 receptor cDNAs, and Miss Nathalie Laflamme for invaluable technical assistance.

REFERENCES

1. Rivier CL, Plotsky PM. Mediation by corticotropin-releasing factor (CRF) of adenohy-pophysial hormone secretion. Annu Rev Physiol 1986; 48:475-494.

2. Vale W, Spiess J, Rivier C, Rivier J. Characterization of a 41-residue ovine hypothalamicpeptide that stimulates secretion of corticotropin and -endorphin. Science 1981;213:1394-1397.

1427

NAPPI AND RIVEST

3. Swanson LW Biochemical switching in hypothalamic circuits mediating responses tostress. In: Holstege G (ed.), Progress in Brain Research. Vol. 87. Amsterdam: ElsevierScience Publishers B.V. (Biochemical Division); 1991: 181-200.

4. Sawchenko PE, Imaki T, Potter E, Kovacs K, Vale W. The functional neuroanatomy ofcorticotropin-releasing factor. In: Chadwick DJ, Marsh J, Ackrill K (eds.), Conicotropin-Releasing Factor, Ciba Foundation Symposium 172. Vol. 1. Chichester: John Wiley &Sons Ltd.; 1993: 5-29.

5. Rivest S, Rivier C. The role of CRF and interleukin-1 in the regulation of neurons con-trolling reproductive functions. Endocr Rev 1995; 16:177-199.

6. Rivier C, Rivier J, Vale W. Stress-induced inhibition of reproductive functions: role ofendogenous corticotropin-releasing factor. Science 1986; 231:607-609.

7. Rivier C, Rivest S. Effect of stress on the activity of hypothalamic-pituitarv-gonadal axis:peripheral and central mechanisms. Biol Reprod 1991; 45:523-532.

8. Viau V, Meaney MJ. Variations in the hypothalamus-pituitary-adrenal response to stressduring the estrus cycle in the rat. Endocrinology 1991; 129:2503-2511.

9. Nappi RE, Rivest S. Ovulatory cycle influences the stimulatory effect of stress on theexpression of corticotropin-releasing factor receptor messenger ribonucleic acid in theparaventricular nucleus of the female rat hypothalamus. Endocrinology 1995; 136:4073-4083.

10. Hargreaves KM, Costello AH, JorisJC. Release from inflamed tissue of a substance withproperties similar to corticotropin-releasing factor. Neuroendocrinology 1989: 46:476-482.

11. Karalis K, Sano H, Redwine J, Listwak S, Wilder RL, Chrousos GP. Autocrine or paracrineinflammatory actions of corticotropin-releasing factor in vivo. Science 1991; 254:421-423.

12. Sasaki A, Tempst P, Liotta AS. Isolation and characterization of a corticotropin-releasing hor-mone-like peptide from human placenta. J Clin Endocrinol & Metab 1988; 67:768-773.

13. Yoon DJ, Sklar C, David R. Presence of immunoreactive corticotropin-releasing factor inrat testis. Endocrinology 1988; 122:759-761.

14. Fabbri A, TinajeroJC, Dufau ML. Corticotropin-releasing factor is produced by Leydig cellsand has a major antireproductive role in the testis. Endocrinology 1990 127:1541-1543.

15. Thompson RC, Seasholtz AF, Hebert E. Rat corticotropin releasing hormone gene: se-quence and tissue specific expression. Mol Endocnnol 1987 1:363-368.

16. Ulisse S, Fabbri A, Dufau ML. Corticotropin-releasing factor receptors and action in ratLeydig cells. J Biol Chem 1989; 264:2156-2163.

17. I)ufau ML, TinajeroJC, Fabbri A. Corticotropin-releasing factor: an antireproductive hormone of the testis. FASEB J 1993 7:299-307.

18. Tsafriri A, Adashi EY. Local non steroidal regulators of ovarian function. In: Knobil E,Neill JD (eds.), The Physiology of Reproduction. Vol. 1. New York: Raven Press Ltd.:1994: 817-860.

19. Mastorakos G, Webster EL, Friedman TC, Chrousos GP. Immunoreactive corticotropin-releasing hormone and its binding sites in the rat ovary. J Clin Invest 1993; 92:961-968.

20. Mastorakos G, Scopa CD, Vryonidu A, Friedman TC, Kattis D, Phenekos C, Merino MJ,Chrousos GP. Presence of immunoreactive corticotropin-releasing hormone in normaland polycystic human ovaries. J Clin Endocrinol & Metab 1994; 79:1191-1197.

21. Stephanou A, Jessop DS, Knight RS, Lightman SL. Corticotropin-releasing factor-like im-munoreactivity and mRNA in human leukocytes. Brain Behav Immun 1990; 4:67-73.

22. Crofford LJ, Sano H, Karalis K, Webster EL, Goldmuntz EA, Chrousos GP, Wilder RL.Local secretion of conicotropin-releasing hormone in the joints of Lewis rats with in-flammatory arthritis. J Clin Invest 1992; 90:2555-2564.

23. Adashi EY. The potential relevance of cytokines to ovarian physiology: the emerging roleof resident ovarian cells of the white blood cell series. Endocr Rev 1990; 11:454-464.

24. Norman J, Brannstrom M. White cells and the ovary-incidental invaders or essentialeffectors? J Endocrinol 1994; 140:333-336.

25. Dume DA, Halpin D, Charlton H, Gordon S. The mononuclear phagocyte system of themouse defined by immunohistochemical localization of antigen F4/80: macrophages ofendocrine organs. Proc Natl Acad Sci USA 1984; 81:4174-4177.

26. Fukuoka M, Mori T, Tail S, Yasuda K. Interleukin-1 inhibits luteinization of porcine granulosa cells in culture. Endocrinology 1988; 122:367-369.

27. Nakamura Y, Kato H, Terranova PF. Interleukin-lu increases thecal progesterone pro-duction of preovulatory follicles in cyclic hamster. Biol Reprod 1990; 43:169-173.

28. Hurwitz A, Ricciarelli E, Botero L, Rohan RM, Hernandez ET, Adashi EY. Endocrine- andautocrine-mediated regulation of rat ovarian (theca-interstitial) interleukin-l gene expression: gonadotropin-dependent preovulatory acquisition. Endocrinology 1991;129:3427-3429.

29. Brannstrom M, Norman RJ, Seamark RF. Robertson SA. Rat ovary produces cytokinesduring ovulation. Biol Reprod 1994: 50:88-94.

30. Calogero AE, Burrello N, Papale L, Palumbo M, Cianci A, Sanfilippo S., D'agata R. Cor-ticotropin-releasing hormone inhibits estrogen production from rat granulosa and hu-man granulosa-luteal cells in vitro. J Endocrinol Invest 1993; 16:5.

3]. Chen R, Lewis KA, Perrin MH, Vale WW. Expression cloning of a human corticotropin-releasing factor receptor. Proc Natl Acad Sci USA 1993; 90:8967-8971.

32. Perrin MH, Donaldson CJ, Chen R, Lewis KA, Vale WW. Cloning and functional expres-sion of a rat brain corticotropin-releasing factor (CRF) receptor. Endocrinology 1993;133:3058-3061.

33. Perrin M, Donaldson C, Chen R, Blount A, Berggren T, Bilezikjian L. Sawchenko P, ValeW. Identification of a second corticotropin-releasing factor receptor gene and characterization of a cDNA expressed in hearth. Proc Natl Acad Sci USA 1995; 92:2969-2973.

34. Lovemberg T, Liam C, Grigoriadis D, Clevenger W. Chalmers D, DeSouza E, OltersdorfT. Cloning and characterization of a functional distinct corticotropin-releasing factor receptor subtype from the rat brain. Proc Natl Acad Sci USA 1995; 92:836-840.

35. Simmons DM, Arriza JL, Swanson LW. A complete protocol for in situ hybridization ofmessenger RNAs in brain and other tissues with radio-labeled single-stranded RNAprobes. J Histotechnol 1989 12:169-181.

36. Suda T, Tomori N, Tozawa F, Demura H, Shizume K, Mouri T. Miura Y. Sasano N. Itnmu-noreactive corticotropin and corticotropin-releasing factor in human hypothalamus, adre-nal, lung cancer, and pheochromocytoma.J Clin Endocrinol & Metab 1984; 58:919-924.

37 Espey LL. Ovulation as an inflammatory reaction: a hypothesis. Biol Reprod 1980 22:73-106.

38. Brannstrom M, Wang L, Norman RJ. Ovulatory effect of interleukin-lp on the perfusedrat ovary. Endocrinology 1993: 132:399-404.

39. Roby KS, Terranova PF. Tumor necrosis factor a alters follicular steroidogenesis in vitro.Endocrinology 1988; 123:2952-2954.

40. Hurwitz A, Loukides J, Ricciarelli E, Botero L, Katz E, McAllister JM, Gercia JE, Rohan KRAdashi EY, Hernandez ER. Human intraovarian interleukin-1 (IL-I) system: highly com-partmentalized and hormonally dependent regulation of the genes encoding IL-I, itsreceptor, and its receptor antagonist. Clin Invest 1992; 89:1746-1754.

41. Polan ML, Loukides J, Nelson P, Carding S, Diamond M, Walsh A, Bottomley K. Progesterone and estradiol modulate IL-1 messenger ribonucleic acid levels in cultured humanperipheral monocytes. J Clin Endocrinol & Metab 1989: 69:1200-1206.

42. Hsueh AIW, Adashi EY. Extrapituitary actions of gonadotropin-releasing hormone. En-docr Rev 1981; 2:186-209.

43. Hsueh AJW, Adashi EY, Jones PBC, Welsh TH. Hormonal regulation of the differentiationof cultured granulosa cells. Endocr Rev 1984; 5:76-127.

44. Olofsson JI, Conti CC, Leung C, Krisinger J, Leung PCK. Tissue-specific regulation ofgonadotropin-releasing hormone (GnRH) receptor gene expression during the perio-vulatory period. EndocrJ 1994; 2:471-476.

45. Chen CLC, Chang CC, Krieger DT, Bardin CW. Expression and regulation of proopiomelanocortin-like gene in the ovary and placenta: comparison with the testis. Fndo-crinology 1986; 118:2382-2389.

46. Lolait SJ, Aurelitano DJ, Lim ATW, Smith Al, Toh H, FunderJW. Ovarian immnunoreactive[-endorphin and estrus cycle in the rat. Endocrinology 1985; 117:161-168.

47. Brannstrom M, Mayrhofer G., Robertson SA. Localization of leukocyte subsets in the ratovary during the periovulatory period. Biol Reprod 1993; 48:277-286.

48. Parr EL. Histological examination of the rat ovarian follicle wall prior to ovulation. BiolReprod 1974; 11:483-503.

49. Ojeda SR, Lara HE. The role of the sympathetic nervous system in the regulation ofovarian function. In: Pirke KM, Wuttke W, Schweiger U (eds.), The Menstrual Cycle iandIts Disorders. Berlin: Springer-Verlag; 1989: 26-33.

0. Merchenthaler , Hynes MA, Vigh S5, Schally AV, Petrusz P. Immunocytochem ical locali-zation of corticotropin-releasing factor (CRF) in the rat spinal cord. Brain Res 1983;275:373-377.

51. Burden HW. The adrenergic innervation of mammalian ovaries. In: Ben-Jonathan N.BaharJM, Weiner RI (eds.) Catecholamines as Hormone Regulators. New York: RavenPress; 1985: 279-292.

52. FerruzJ, Barria A, Galleguillos X, Lara HE. Release of norepinephrine from the rat ovary:local modulation by gonadotropins. Biol Reprod 1991; 45:592-597.

53 Vamvakopoulos NC, Chrousos GP. Evidence of direct estrogenic regulation of humancorticotropin releasing hormone gene expression: potential implications for the sexualdimorphism of the stress response and immune/inflammatory reaction. J Clin Invest1993; 92:1896-1902.

1428