BIOLOGY OF REPRODUCTION 57, 865-872 (1997)
Expression and Localization of Prolactin Receptor Messenger Ribonucleic Acid in Red Deer Ovary during the Estrous Cycle and Pregnancy' Luka A. Clarke,3 D. Claire Wathes,
and Henry N. abbour 2,5
Institute of Zoology, 3 The Zoological Society of London, Regent's Park, London NWI 4RY, United Kingdom Department of Farm Animal and Equine Medicine and Surgery, 4 Royal Veterinary College, Potters Bar, Hertfordshire EN6 NB, United Kingdom MRC Reproductive Biology Unit, 5 Centre for Reproductive Biology, Edinburgh EH3 9EW, United Kingdom ABSTRACT In this study, expression of the prolactin receptor (PRL-R) gene in the ovaries of cycling and pregnant red deer (Cervus elaphus) hinds was investigated. A 1.9-kilobase (kb) cDNA encoding the cervine long-form PRL-R was amplified by reverse transcriptase polymerase chain reaction from corpus luteum (CL) and liver poly(A) + RNA. Northern hybridization revealed a major mRNA transcript of 3.5 kb in both tissues. PRL-R mRNA transcripts were localized by in situ hybridization in 15-1.m frozen sections of red deer ovaries, collected during the estrous cycle and early pregnancy, with homologous 45-mer [35S]dATPlabeled sense and antisense oligonucleotide probes. Specific hybridization was assessed by measurement of autoradiograph optical density (OD) in CL, follicles, and stroma. PRL-R mRNA expression was higher (p < 0.001) in the CL (OD = 22.2 3.77, n = 11 CL) than in follicles (OD = 2.8 0.10, n = 224 follicles) and was undetectable in the stroma (OD < 1, limit of detection). No differences in abundance of PRL-R mRNA were observed between follicles divided on the basis of size, health vs. atresia, or stage of estrous cycle or pregnancy, or between CL from pregnant and nonpregnant hinds. In the follicles, PRL-R mRNA was localized to the theca layer. These results suggest a direct role for PRL in red deer ovarian physiology during the estrous cycle and pregnancy. INTRODUCTION The red deer (Cervus elaphus) is a seasonally breeding species with a marked circannual rhythm of reproduction and a well-defined annual cycle in secretion of prolactin (PRL; ). Unmated red deer hinds will undergo estrous cycles of 18-19 days throughout the autumn and winter while pituitary PRL secretion is low  and enter a prolonged period of anestrus when PRL secretion rises in the spring. Suppression of this elevation in PRL secretion by bromocriptine administration has been shown to delay significantly the onset of anestrus in red deer hinds , suggesting that the breeding season is terminated partly by the effect of seasonal hyperprolactinemia. PRL may exert this effect directly at the level of the ovary through modulation of LH receptor (LH-R) expression . Alternatively, elevated secretion of PRL may affect the GnRH axis through interaction with PRL receptors (PRL-R) in the brain . PRL binds within the hypothalamus of many mammalian species , and PRL-R mRNA has been localized by in situ hybridization in cells of the preoptic, supraoptic, and
rostral arcuate nuclei and choroid plexus . In addition, PRL has been shown to inhibit the release of GnRH from both cultured GT1 GnRH-secreting cells  and explanted rat hypothalamic tissue . During the breeding season, however, PRL secreted at basal concentrations may play an essential role in the regulation of ovarian physiology. PRL is known to be luteotrophic in sheep ; administration of PRL to hysterectomized, hypophysectomized ewes is essential for maintenance of corpus luteum (CL) weight and progesterone production. In the rodent and rabbit ovary, PRL can enhance luteal and follicular LH-R expression [11-13], stimulate progesterone production [14, 15], and inhibit ovulation either by reduction of follicular plasmin generation  or by stimulation of metalloproteinase inhibitor activity . Furthermore, PRL induces structural luteolysis in the rat . These multiple effects may be mediated directly via the ovarian PRL-R, of which multiple forms are detected in the ovarian tissues of rats [19, 20], mice , and sheep . Recently, a study of mice expressing a null mutation of the PRL-R gene has confirmed the significant role of the PRL-R in mediating ovarian function . We have previously cloned and sequenced a cDNA encoding the PRL-R long form in red deer liver  and demonstrated its expression in a number of tissues including the testis, where data also indicated the expression of a second, shorter form . The aim of the present study was to characterize the type and localize the site of expression of the PRL-R in the red deer ovary during the estrous cycle and pregnancy. Expression of the PRL-R gene in red deer ovarian tissue would imply a direct role for the basal concentrations of PRL secreted during the breeding season on reproductive function in seasonal mammals. This would be in contrast to the antigonadal effects of PRL observed during seasonal anestrus. MATERIALS AND METHODS Animals Ovaries were collected from 22 red deer hinds kept in Great Britain under natural light conditions and fed ad libitum. Collections took place during the early breeding season in October at four stages of the estrous cycle (late follicular, n = 2 hinds; early luteal, n = 3 hinds; late luteal, n = 3 hinds; early follicular, n = 3 hinds) from 11 synchronized hinds (; hinds were synchronized by application of progesterone-releasing controlled internal drugreleasing [CIDR] devices for 2 wk followed by administration of 200 IU eCG upon CIDR removal to induce ovulation). A further collection from nonpregnant hinds took place in January (n = 4 hinds; in Britain nonpregnant hinds normally cycle until March ). Ovaries were also collect-
Accepted June 2, 1997. Received February 6, 1997. 'This work was supported by a BBSRC grant to H.N.J. L.A.C. is the recipient of a postgraduate studentship from the Biotechnology and Biological Sciences Research Council. 2Correspondence: Henry Jabbour, MRC Reproductive Biology Unit, Centre for Reproductive Biology, 37 Chalmers Street, Edinburgh EH3 9EW, UK. FAX: 44 131 228 5571; e-mail: [email protected]
ed at 56 (n = 3 hinds), 100 (n = 2 hinds), and 120 (n =
CLARKE ET AL.
2 hinds) days of pregnancy. Whole ovaries were frozen directly upon collection in isopentane (BDH, Poole, UK) in liquid nitrogen and stored at -70 0 C until use. For RNA extraction, liver tissue and excised CL were snap frozen directly in liquid nitrogen. RNA Extraction Total RNA was extracted from CL and liver using a modification of the guanidinium thiocyanate method described by Chomczynski and Sacchi . RNA yields and purity were estimated by spectrophotometry (DU-65; Beckman, Fullerton, CA) at 260 nm and 280 nm. The poly(A) RNA-enriched fraction was isolated by passing total RNA through columns containing oligo(dT)-cellulose (Sigma, Poole, UK). Northern Hybridization Samples of 10 Rig poly(A)+ RNA isolated from tissues (pregnant CL, n = 2; nonpregnant CL, n = 2; liver, positive control), as described above, were size separated by electrophoresis in a 1% agarose/7% formaldehyde gel alongside a 0.24-9.5-kb RNA size ladder (Gibco BRL, Paisley, UK). Gels were blotted overnight with 20-strength SSC (singlestrength SSC is 0.015 M NaCl, 0.015 M Na 3C6H 50 7, pH 7.0) to Hybond-N nylon filters (Amersham International plc, Little Chalfont, UK). Blots were fixed by 3-min UV transillumination, followed by prehybridization for 3-4 h at 42°C in a solution containing 50% deionized formamide, 5-strength Denhardt's reagent, 5-strength saline-sodium phosphate-EDTA buffer, 0.1 mg/ml sheared herring sperm DNA, 1 [ig/ml polyadenylic acid, and 6% dextran sulfate. A 562-base pair (bp) cDNA corresponding to a fragment of the extracellular domain of the red deer PRL-R  was labeled with [- 32P]dCTP by random priming  and hybridized to the blots at 42°C overnight in the same buffer. Filters were washed twice in single-strength SSC and 0.1% SDS at room temperature for 5 min and once at 55°C for 20 min, once in 0.1-strength SSC and 0.1% SDS at 55°C for 20 min, and twice in 0.1-strength SSC at room temperature for 5 min. Filters were autoradiographed at -70°C with Fuji (Tokyo, Japan) RX film for approximately I wk. Polymerase Chain Reaction and Southern Hybridization Single-stranded cDNA was generated from 5 Lg poly(A) + RNA from CL by reverse transcription. The RNA was denatured at 65°C for 10 min with 2.5 M oligo(dT), cooled on ice, and then reverse transcribed for 1 h at 37°C in a total volume of 20 Ril in a solution containing 50 mM Tris (pH 8.3), 75 mM KC1, 10 mM dithiothreitol, 3 mM MgCI 2, 0.5 mM of each dNTP, 40 U RNase inhibitor, and 1 1 Superscript RT (Gibco BRL). One microliter (250 ng) of the resultant cDNA was then amplified by polymerase chain reaction in a volume of 50 1 in a reaction mixture that included 0.25 Il Taq polymerase and 25 pmol each of oligonucleotide primers designed from the 5' and 3' untranslated regions of the red deer PRL-R gene sequence at base pair positions -105 to -86 (5'-GCT-AAA-GAACGC-TTC-TGT-TC-3': forward primer) and 1750 to 1770 (5'-GGC-CAG-GTC-AGC-CTC-GGC-TGG-3': reverse primer). These primers flank the entire red deer long-form PRL-R gene coding region and were designed to amplify a product of 1875 bp. Liver cDNA and double-distilled H2 0 were used as positive and negative controls according to the same procedures. The amplification conditions were
as follows: 94°C for 3 min followed by 35 cycles of 94°C for 30 sec, 55°C for 1 min, and 72°C for 2 min. This was followed by extension at 72°C for 10 min. Polymerase chain reaction products were run alongside a DNA size ladder (I/X; Pharmacia Biotech, St. Albans, UK) on a 1% agarose gel and blotted overnight  in 0.4 M NaOH to a Hybond-N nylon filter, which was then hybridized overnight at 55°C in single-strength Church and Gilbert (0.2M NaH 2PO4, 1% BSA, 7% SDS, mM EDTA, 15% formamide) buffer using the [- 32P]dCTP-labeled 562-bp homologous probe. Washing and autoradiography of Southern blots were carried out using the Northern blot protocol described above. In Situ Hybridization The in situ hybridization technique used was that reported by Stevenson et al. . Unless otherwise stated, all chemicals were purchased from Sigma or BDH (both Poole, UK). The 15-jim cryostat sections were thaw mounted onto poly-L-lysine (1 mg/ml; Mr > 300 000)-coated baked glass slides and fixed in 4% (w:v) paraformaldehyde made up in 0.01 M PBS for 5 min; this was followed by three 0.01 M PBS washes (2 min each) and dehydration in 70% and 95% ethanol (5 min each). The slides were stored in 95% ethanol until needed. The PRL-R 45-mer oligonucleotide antisense (AS) and sense (S) probes designed for in situ hybridization corresponded to nucleotides 1324-1368 of the cervine PRL-R gene ; the AS sequence was 5'-TTG-GTC-CAG-CAAAGT-GGC-TGT-GGT-ACC-TGC-CAT-GCC-CAG-GGCCAG-3'. The probes were end-labeled at the 3' end with [3 5S]dATP (New England Nuclear Research Bioproducts, DuPont de Nemours, Dreiech, Germany) using terminal deoxynucleotidyl transferase (calf thymus, fast protein liquid chromatography pure; Pharmacia LKB Biotechnology, Milton Keynes, UK) to a specific activity of 6-9 x 108 dpm/Lig. The sections were hybridized overnight at 42°C in the presence of 100 tlI hybridization buffer containing 110 000 dpm probe in 50% deionized formamide, 4-strength SSC, 10% dextran sulfate, 5-strength Denhardt's solution, 0.2 mg/ml single-stranded sheared salmon sperm DNA, 0.2 mg/ml polyadenylic acid, 0.12 mg/ml heparin, 0.25 M sodium phosphate (pH 7.0), and 0.001 M sodium pyrophosphate. After hybridization the slides were washed under low-stringency conditions (30 min at room temperature in single-strength SSC and 0.2% [w:v] sodium thiosulfate pentahydrate), then at a higher stringency (60 min at 55°C in fresh single-strength SSC and 0.2% [w:v] sodium thiosulfate pentahydrate); they were then rinsed and dehydrated (single-strength SSC, 0.1-strength SSC, 70% ethanol, and 95% ethanol for 20 sec each). The slides were then exposed to Hyperfilm 3-max (Amersham International) for 5 wk in autoradiographic cassettes. Measurements of OD (arbitrary units on a linear scale of 0-2.1) were undertaken using the Seescan image analysis package (Seescan plc, Cambridge, UK). OD values were multiplied by 102 for clarity of pre102) was taken to be the sentation, and OD = 1 (0.01 lower limit of signal detection. Multiple readings (5 readings per section, 3 sections per ovary for both AS and S probes) were made in each tissue region (CL, individual follicles, and stroma) of each autoradiograph. Mean AS and S values were thus obtained for each tissue type in each ovary, and the S means were subtracted from the AS means to give comparable hybridization values. For more specific localization of the message, the slides were coated with
PRL-R mRNA EXPRESSION IN RED DEER OVARY
FIG. 2. RT-PCR of poly(A)* RNA from red deer liver and CL using primers flanking the coding region of the PRL-R cDNA. Identity of the 1.9-kb PCR products was confirmed by Southern blot (SB) hybridization with an 2 &a P-labeled cDNA probe from the extracellular domain of the PRL-R.
RESULTS FIG. 1. Northern blot analysis of PRL-R gene expression in liver and CL of red deer using a cDNA probe from the extracellular domain of the red deer PRL-R. A major transcript of 3.5 kb and minor transcripts of 7.1 and 12.2 kb are detected in both tissues.
photographic emulsion according to the manufacturer's instructions (K5; Ilford Photography Co., London, UK), stored for 5 wk at 40C, developed, and counterstained with hematoxylin and eosin. Data Analysis All follicles from all ovarian sections (n = 224 follicles) were examined under lightfield and classified as either large (> 4-mm diameter, n = 16 follicles), medium (2-4 mm, n = 68 follicles), or small (< 2 mm, n = 140 follicles). Health or atresia of all follicles (healthy, n = 98 follicles; atretic, n = 126 follicles) was determined according to the description of Carson et al. ; follicles were judged healthy unless they exhibited signs of class III to class V atresia, i.e., visibly pyknotic nuclei, marked perforation or discontinuity of the granulosa cell layer, and possible detachment of the granulosa cell layer from the theca interna. OD data from all follicles were log transformed to achieve homogeneity of variance and were analyzed using an unbalanced ANOVA design with individual deer as covariate and size of follicle, health of follicle, and stage of collection as factors. As no significant effects were noted, data from all follicles on each ovary having a CL were combined. These follicle means were then compared by paired t-test to the OD data from CL present on the same ovaries (n = 11). A comparison was also made by Student's t-test between OD data from CL collected from nonpregnant (n = 5 CL) and pregnant (n = 6 CL) hinds.
Northern Hybridization Northern blot analysis of poly(A) + RNA extracted from both CL and liver suggested that in both tissues the PRL-R is encoded by a major mRNA transcript of 3.5 kb (Fig. 1). A similar pattern of larger transcripts (12.2 and 7.1 kb) was also observed in each tissue, and the same pattern was observed in CL and livers from both pregnant and nonpregnant animals (blot from nonpregnant hind shown). Reverse Transcriptase Polymerase Chain Reaction (RTPCR) and Southern Blot The RT-PCR using primers flanking the coding region of the full PRL-R cDNA generated the expected product of approximately 1.9 kb in poly(A) + RNA from both the CL and liver. The identity of these PCR products was confirmed with Southern blotting and hybridization using a homologous cDNA probe coding for the extracellular domain of the red deer PRL-R (Fig. 2). No PCR product was generated that might correspond to a short-form receptor mRNA transcript. By contrast, with use of the same primers, a PCR product of 1.65 kb encoding a short PRL-R isoform has been amplified from red deer testis (unpublished results). Localization of PRL-R mRNA by In Situ Hybridization Hybridization of the PRL-R oligonucleotide probes to 15-tpm sections of the red deer ovary was assessed by OD measurements of absorbance (expressed as mean OD units - SEM; values multiplied by 102 for clarity of presentation). There was no detectable hybridization to the ovarian stroma (OD < 1, limit of detection). The highest expression was observed in the CL (22.2 + 3.77; n = 11), with significantly (p < 0.001) lower expression in follicle walls (2.8 + 0.10; n = 224). No differences in PRL-R expression were detected in follicles with regard to either size (small
CLARKE ET AL.
868 FIG. 3. PRL-R mRNA abundance during the red deer estrous cycle and early pregnancy, measured as OD from autoradiographs. There was no difference (p > 0.05) in PRL-R mRNA expression between small (S; n = 140), medium (M; n = 68), and large (L; n = 16) follicles (A); healthy (n = 98) and atretic (n = 126) follicles (B); follicles collected at different stages of the estrous cycle (C1: October, late follicular phase, n = 19; C2: October, early luteal phase, n = 28; C3: October, late luteal phase, n = 47; C4: October, early follicular phase, n = 57; C5: January, luteal phase, n = 35), and pregnancy (P1: 56 days, n = 19, P2: 100 days, n = 9, P3: 120 days of pregnancy, n = 10) (C); or CL collected from nonpregnant (n = 5) and pregnant (n = 6) hinds (D; note difference in scale). All values are OD means + SEM.
z E CG 1 -
E cca:1 r/ a.
K d K
o_ n_ II
KK K K K KK