Downregulation FSH

1580

BIOLOGY OF REPRODUCTION 70, 1580–1588 (2004)Published online before print 28 January 2004.DOI 10.1095/biolreprod.103.026898

Downregulation of Follicle-Stimulating Hormone (FSH)-Receptor Messenger RNALevels in the Hamster Ovary: Effect of the Endogenous and Exogenous FSH1

Yi-Ming Zhang3 and Shyamal K. Roy2,3,4

Departments of Physiology and Biophysics3 and Obstetrics and Gynecology,4 University of Nebraska Medical Center,Omaha, Nebraska 68198-4515

ABSTRACT

Although gonadotropins have been reported to downregulateFSH-receptor (FSHR) mRNA levels in the ovaries of female rats,the effect of the gonadotropin surge, particularly FSH, on ham-ster follicular FSHR mRNA levels warrants further examination.The objectives of the present study were to clone and determinethe complete FSHR cDNA sequence of the hamster and to de-lineate the effects of endogenous and exogenous FSH on thesteady-state levels of ovarian FSHR mRNA. Complete FSHRcDNA was derived from hamster ovarian total RNA by the strat-egy of 39- and 59-rapid amplification of cDNA ends. Ovarieswere obtained before and after the endogenous gonadotropinsurge or exogenous FSH administration, and the steady-state lev-els of FSHR mRNA were assessed by Northern blot hybridiza-tion. Cloned FSHR cDNA consists of a reading frame corre-sponding to exons 1–10 of the human FSHR gene and the 59-and 39-untranslated regions. The nucleic acid and amino acidsequences of the reading frame were at least 87% and 92%identical, respectively, to that of human, rat, and mouse FSHR.Furthermore, the amino acid sequence contained seven trans-membrane domains characteristic of the FSHR. The steady-statelevels of FSHR mRNA increased from estrus (Day 1) to reach apeak on proestrus (Day 4) noon; however, significant attenua-tion was noted following the gonadotropin surge, which wasblocked by phenobarbital. Exogenous FSH also downregulated,both dose- and time-dependently, ovarian FSHR mRNA levels.These data indicate that the nucleic acid sequence of hamsterFSHR has been identified and that FSH modulates FSHR mRNAlevels in the hamster ovary.

follicle-stimulating hormone, follicle-stimulating hormone recep-tor, granulosa cells, ovary

INTRODUCTION

Ovarian follicular development depends on FSH action[1]. Lack of FSH function because of a null mutation inthe FSHb [2], or FSH-receptor (FSHR) [3] gene, or hy-pophysectomy [1] in rodents results in impaired folliculardevelopment, leading to ovulation failure and infertility.Similarly, neutralization of serum FSH by an anti-FSH se-rum results in cessation of primordial follicle formation in

1Supported by grants HD28165 and HD38468 from the National Instituteof Child Health and Human Development (NICHHD, NIH) to S.K.R. Thefull-length cDNA sequence of the hamster FSH receptor has been depos-ited in the GenBank (accession no. AY509907).2Correspondence: Shyamal K. Roy, Departments of OB/GYN and Physi-ology and Biophysics, DRC 5013, University of Nebraska Medical Center,Omaha, NE 68198-4515. FAX: 402 559 6164; e-mail: [email protected]

Received: 22 December 2003.First decision: 20 January 2004.Accepted: 22 January 2004.Q 2004 by the Society for the Study of Reproduction, Inc.ISSN: 0006-3363. http://www.biolreprod.org

neonatal hamsters [4]. The action of FSH is mediated byits cognate receptor, which belongs to a G protein-coupledreceptor superfamily. The FSHRs contain a large extracel-lular domain that binds to the ligand [5, 6]. Heckert et al.[7] have demonstrated that the extracellular domain of ratFSHR is encoded by nine exons, whereas the entire trans-membrane and the C-terminal cytoplasmic domain is en-coded by exon 10. Furthermore, these exons are separatedby nine introns [5]. Similar structural organization has beenreported for the human FSHR gene [6]. The nucleic acidand corresponding amino acid sequences of FSHR havebeen determined for many species [5, 7–9], and this infor-mation has led to a significant understanding of the patho-physiology of FSHR gene expression [5, 10–12]. BecauseFSH has significant roles in hamster ovarian folliculoge-nesis [1, 4, 13, 14], identifying the nucleic acid and aminoacid sequences of hamster FSHR is essential to understand-ing the regulation of FSHR gene expression in hamster go-nads.

Whereas high doses of FSH have been shown to down-regulate FSHR mRNA levels both in vivo [15–17] and invitro [18], a small dose of FSH has been shown to maintainFSHR mRNA levels in cultured porcine granulosa cells[19]. Furthermore, ovulation induction in the rat with a re-combinant human FSH leads to suppression of FSHRmRNA levels in the granulosa cells [20]. Because granulosacells differentiate into LH-responsive, progesterone-produc-ing luteal cells after the preovulatory gonadotropin surge[21], decreases in the levels of FSHR mRNA and proteinfollowing the surge is expected. In the hamster, serum FSHlevels decline from the day of estrus to the morning ofproestrus, followed by a significant increase at 1500 h [22]and a surge at 1600 h in proestrus [23]. Despite this infor-mation, it is unclear whether the in vivo FSH surge influ-ences the levels of FSHR mRNA in the hamster ovary dur-ing the estrous cycle. Because FSH plays very importantroles during hamster follicular development [1, 13], it willbe important to know if FSH, whether endogenous or exo-genous, influences ovarian FSHR mRNA levels. This in-formation will be particularly relevant to understandinghow preantral follicles grow after the preovulatory gonad-otropin surge [24, 25]. The objectives of the present studywere to identify the complete cDNA sequence of the ham-ster FSHR, to determine the steady-state levels of FSHRmRNA in the ovary during the estrous cycle, and to eval-uate the effects of endogenous and exogenous FSH onFSHR mRNA levels.

MATERIALS AND METHODS

Materials

Adult golden female hamsters were purchased from Sasco (Madison,WI) and maintained in a climate-controlled room with a 14L:10D photo-

1581EFFECT OF GONADOTROPINS ON FSHR mRNA EXPRESSION

TABLE 1. Various primers used to clone hamster FSHR cDNA.

GeneRacer oligo (dT)GeneRacer 39-primerF (forward)R1 (reverse)R2 (reverse)GeneRacer 59-primerGeneRacer oligo RNA

59-CTGTCAACGATACGCTACGTAACGGCATGACAGTG(T)18-3959-GCTGTCAACGATACGCTACGTAACG-3959-TGTCATCACTGGCTGTGTCAT-3959-ATTTGGATGAAGTTCAGAGGTT-3959-ATCCTTGTGGGATGACTCG-39 (hamster)59-CGACTGGAGCACGAGGACACTGA-3959-CGACUGGAGCACGAGGACACUGACAUGGACUGAAGGAGUAGAAA-39

period and free access to food and water according to Institutional AnimalCare and Use Committee (IACUC) and U.S. Department of Agricultureguidelines. The use of hamsters for the present study was approved by theIACUC. Females with at least three consecutive estrous cycles were used.Ovine-FSH-20 and the FSH and LH RIA kits were kindly provided byDr. A.F. Parlow (Harbor-UCLA, Los Angeles, CA) at a subsidized price.Phenobarbital (65 mg/ml USP) for human injection was purchased fromthe University of Nebraska Medical Center pharmacy. The TRI reagentand RNAeasy kit for RNA extraction was from MRC, Inc. (Cincinnati,OH), and Qiagen (Valencia, CA), respectively. Nytran membrane forNorthern hybridization was from Schleicher and Schuell (Fisher ScientificCompany, Pittsburgh, PA), and 59- and 39-rapid amplification of cDNAends (RACE) kit and cDNA cloning kit (PCRII TOPO) were from Invi-trogen (Carlsbad, CA). All other chemicals were either from Sigma Chem-ical Company (St. Louis, MO) or Fisher Scientific.

59- And 39-RACE Cloning of Full-Length HamsterFSHR cDNA

Total ovarian RNA was extracted by TRI reagent according to themanufacturer’s protocol and as described previously [26, 27] and wasquantified by a spectrophotometer at 260 nm. The quality of RNA wasassured by two sharp ethidium bromide-stained bands of 18S and 28SrRNA without any obvious smear. Furthermore, no larger size correspond-ing to any putative DNA contamination was evident.

Two FSHR cDNAs, one containing a 39-untranslated region (UTR)plus a part of the poly A tail (henceforth called 39-FSHR) and the othercontaining a complete 59-UTR (henceforth called 59-FSHR), were gener-ated using GeneRacer Kit (Invitrogen) according to the manufacturer’sinstructions. Briefly, to synthesize a single-stranded 39-FSHR cDNA, 5 mgof total ovarian RNA were denatured at 658C for 5 min and then mixedwith 2.5 mM GeneRacer oligo (dT) (Table 1), 1 mM dNTPs, 40 U ofRNaseOUT, and 15 U of ThermoScript reverse transcriptase in a finalvolume of 20 ml and incubated at 608C for 60 min, followed by 5 min at858C. The RNA template was removed from the mixture by adding 2 Uof RNase H, followed by a 20-min incubation at 378C. Double-stranded39-hamster FSHR cDNA was synthesized by polymerase chain reaction(PCR) in a 50-ml reaction mixture containing 1 ml of reverse transcriptionproduct, 0.4 mM GeneRacer 39-primer (complementary to the last 25 basesof the GeneRacer oligo dT), 0.2 mg of a forward primer (F) (Table 1),2.75 mM MgCl2, 0.2 mM dNTPs, and 2.5 U of Taq polymerase. After a4-min denaturation step at 948C, PCR was continued for 30 cycles, fol-lowed by a final extension at 728C for 10 min. The PCR conditions were1.5 min at 948C, 1 min at 588C, and 2.5 min at 728C. To synthesize asingle-stranded 59-FSHR cDNA, a 59-oligo (GeneRacer RNA oligo [dT])(Table 1) was ligated to the 59-cap site of all mRNA in the total RNA andreverse transcribed in the presence of a FSHR gene-specific reverse primerR1 (Table 1). Double-stranded 59-hamster FSHR cDNA was synthesizedby PCR using R2 and GeneRacer 59-primers (Table 1). The PCR condi-tions were 4 min denaturation at 958C followed by 1 min at 948C, 1 minat 588C, and 1 min at 728C followed by a 10 min extension at 728C. Theforward (F) and reverse (R1) primers, which corresponded to 51–72 and851–873 bases, respectively, were designed from the conserved regions ofrat (GenBank accession no. L02842) [9], mouse (accession no. AF095642)[12], and human (accession no. AY429104) [8] FSHR sequences, whereasthe R2 primer was designed from a 800-base pair (bp) hamster FSHRcDNA, which was generated using F and R1 primers.

The PCR products were resolved in a 1% agarose gel, and cDNAs ofpredicted size were purified from the gel using Oligotex II kit (Qiagen),inserted into either pCRIITOPO (for 39-FSHR) or pCR4-TOPO (for 59-FSHR) vector (Invitrogen) and transformed in TOPO10 cells (Invitrogen).White colonies were selected under ampicillin, and after thorough screen-ing, a part of one white colony for each cDNA type was amplified byPCR and the presence of the insert verified by Southern blot analysis [26].The rest of the colony was grown overnight under ampicillin selection,and plasmid DNA was extracted and purified using Plasmid Maxi Kit

(Qiagen) according to the manufacturer’s instructions. The nucleotide se-quences of 39- and 59-FSHR cDNA fragments were determined by auto-mated DNA sequencing (Genomics Sequencing Service; Qiagen). The se-quences were overlapped to derive the full-length hamster FSHR cDNA,including the 59- and 39-UTRs. Analyses of nucleotide sequence and pre-dicted amino acid composition of hamster FSHR were carried out usingthe Vector NTI (Invitrogen) and GCG (Accelrys, San Diego, CA) pro-grams. To avoid PCR-derived sequence errors, positive alignment of atleast three independent sequences was requested for any given area of thehamster FSHR cDNA.

To synthesize a hamster FSHR cDNA probe for Northern hybridiza-tion, a 504-bp cDNA template, corresponding to 477–981 bases, was gen-erated by PCR from the 39-end hamster FSHR cDNA using F and R1primers (Table 1), cloned into pCRIITOPO vector, plasmid linearized withStuI restriction enzyme (New England Biolab, Beverly, MA), and used asa template for in vitro transcription of cRNA using Riboprobe kit (Pro-mega, Madison, WI).

Ovarian FSHR mRNA Levels Throughoutthe Estrous Cycle

Ovaries were collected from adult hamsters at 0900 h on Days 1–4(estrus through proestrus [24]) and at 1200, 1600, and 2000 h on Day 4.For negative tissue controls, kidney, liver, and adrenal glands were alsocollected. Trunk blood was collected to determine the serum levels of FSHand LH by specific RIAs.

Effect of Endogenous FSH on Ovarian FSHR mRNALevels

Hamsters were injected s.c. with phenobarbital (10 mg/100 g bodyweight) at 1300 h on Day 4 to block the preovulatory gonadotropin surges[22], and ovaries were collected at 2000 h on Day 4 and at 0900 h onDay 1.

In a parallel experiment, phenobarbitone was injected at 2000 h onDay 4 to block the second FSH surge [23, 28], and ovaries were collectedat 0900 h on Day 1. Ovaries from untreated hamsters were collected at2000 h on Day 4 and at 0900 h on Day 1. Trunk blood was collected todetermine the serum levels of FSH and LH by specific RIAs.

Effect of Exogenous FSH on Ovarian FSHR mRNA LevelsHamsters were injected s.c. with 1.64, 8.2, or 16.4 IU of ovine FSH-

20 in 0.1 ml of saline containing 0.5% bovine serum albumin at 0800 hon Day 4. A second group of hamsters received 8.2 IU of recombinanthuman FSH at the same time. Ovaries were collected 4 h after the injectionand used for RNA extraction.

Based on these results, in the next experiment hamsters were injecteds.c. with 16.4 IU of ovine FSH-20 at 0800 h on Day 4, and ovaries werecollected 1, 2, 4, 6, or 12 h after the injection. Hamsters allocated to the12-h group also received phenobarbitone (10 mg/100 g body weight) at1300 h to block the endogenous gonadotropin surge, which otherwisewould have overlapped with the exogenous FSH and made interpretationof the data difficult. Ovaries from untreated hamsters were collected at0800 h (0-h treatment group) or 4 h after the vehicle injection. Ovarieswere used for RNA extraction.

For all groups, tissues were removed immediately, quickly frozen inliquid N2, and stored at 2808C until extracted for total RNA for Northernhybridization detection of FSHR mRNA.

Measurements of Serum FSH and LHSerum levels of gonadotropins were determined by specific RIAs using

rat-rat RIA kits from the National Pituitary Hormone Program as describedpreviously [23, 25, 29]. All samples were assayed at the same time to

1582 ZHANG AND ROY

avoid any interassay variation. The coefficient of intraassay variation was7%.

Northern HybridizationNorthern hybridization analysis of ovarian FSHR mRNA was done as

described previously [26, 30]. Briefly, total RNA from the ovaries wereextracted using RNAeasy kit (Qiagen) according to the manufacturer’sinstructions, and 5 mg of denatured total RNA were fractionated in 1%agarose-formaldehyde gels, capillary transferred to Nytran membrane, ul-traviolet cross-linked, stained with 0.02% methylene blue in 0.3 M sodiumacetate (pH 5.5) to make the 28S and 18S ribosomal RNA bands visible.The signal intensity was digitized using a UVP bioimaging system (UVP,Upland, CA). After destaining and prehybridization, the membrane washybridized for 16–18 h in 10 ml of hybridization mixture containing 23107 cpm [32P]CTP-labeled FSHR cRNA probe, 40% formamide, 0.53Denhardt reagent, 7% SDS, and 10% dextran at 608C in a Hybaid hybrid-ization oven (Fisher Scientific). After rinsing, the membrane was exposedto phosphor screen, and the signal (digital light unit [DLU]) was digitizedin a Cyclone phosphorimager (Perkin-Elmer, Wellesley, MA). First, theDLU was normalized against the time of exposure. Then, the ratio of DLU(radioactive signal) to optical density (18S ribosomal RNA) was calculatedto correct for any variation caused by differences in sample loading. Fi-nally, values of all experimental groups were expressed as fold-changesrelative to respective controls.

Statistical AnalysisEach experiment was repeated at least three times, and the data were

analyzed by ANOVA with the Fisher protected least-significance-differencepost-hoc test using Statview software (SAS Institute, Cary, NC). The levelof significance was set at 5%.

RESULTS

Full-Length Cloning of Hamster FSHR cDNA

Because FSH influenced hamster follicular developmentfrom the primary stage onward and no information abouthamster FSHR cDNA was available, it was necessary toidentify the complete nucleic acid sequence of the hamsterFSHR, including the 59- and 39-UTRs for future examina-tion of FSHR mRNA expression in the granulosa cells. The59-RACE strategy allowed selection of only those mRNAmolecules that had an intact 59-cap structure and eliminatedall truncated mRNA and non-mRNA species, thus assuringthe identification of a complete 59-UTR, which started fromthe most probable transcription start site. Similarly, bindingof the oligo dT primer at the 39-end of the FSHR mRNAassured the identification of a complete 39-UTR. The RACEstrategy followed by sequencing revealed a 2392-bp cDNAcorresponding to FSHR mRNA, which contained a 109-bp59-UTR upstream of the translation start site, a 2082-bpopen reading frame, a stop codon, and a 198-bp 39-UTR(Fig. 1). The open reading frame was at least 90%, 87%,and 89% similar to human [6], rat [9], and mouse [12]FSHR cDNA, respectively. Furthermore, the open readingframe corresponded very well to the exon 1–10 sequencesof the human FSHR [6] sequence. The 59-UTR of the ham-ster FSHR cDNA had 85% and 80% sequence similaritieswith the corresponding 59-UTR sequences of the mouse(accession no. 487570) and rat (accession no. 581117)FSHR, respectively. The deduced amino acid sequence ofthe hamster FSHR corresponded to a peptide sequence of694 amino acids, which showed at least 92% similarity withrat, human, and mouse FSHR amino acid sequences [6, 9,12] (Fig. 2). Furthermore, sequence comparison showed allthe characteristics of FSHR, including a 17-amino-acid sig-nal peptide, seven transmembrane domains, four conservedputative sites for N-linked glycosylation, 11 cysteine resi-dues, and other putative posttranslational modification sites(Fig. 2 and Table 2).

Northern hybridization analysis of hamster ovarian RNArevealed a single band of FSHR mRNA of approximately2.5 kilobases (kb) (Fig. 3). Furthermore, as expected, noFSHR hybridization signal was detected for the kidney, ad-renal gland, or liver, indicating the specificity of the signal(Fig. 3). However, no alternately spliced FSHR transcriptcould be detected in the ovarian sample.

Ovarian FSHR mRNA Levels Throughoutthe Estrous Cycle

The steady-state levels of FSHR mRNA were relativelylow on the morning of estrus; however, a marked increaseoccurred by the morning of Day 2 (Fig. 4). Thereafter, ovar-ian FSHR mRNA levels continued increasing, to reach apeak level by 1200 h on Day 4 (Fig. 4). In the hamster [23,25, 31], serum levels of LH started increasing by 1400 hon Day 4, reaching a peak at 1600 h. The FSH surge occursat 1500 h on Day 4, reaching a peak at 1600 h. Coincidingwith the increase in serum levels of FSH and LH, ovarianFSHR mRNA levels decreased noticeably by 1600 h onDay 4 (Fig. 4), and these levels declined further by 2000 hto reach the levels of 0900 h at Day 1 (Fig. 4).

Effect of Endogenous FSH Surge on Ovarian FSHRmRNA Levels

To determine whether the decline in ovarian FSHRmRNA levels in the hamster was really induced by thepreovulatory gonadotropin surge, hamsters were injectedwith phenobarbitone at 1300 h to block the surge [25, 28],and the steady-state levels of ovarian FSHR mRNA wereexamined at 2000 h on Day 4 and at 0900 h on Day 1.Serum levels of LH and FSH increased significantly by1600 h on day 4, which was completely blocked by phe-nobarbitone pretreatment (data not shown) [22, 25]. Theblock in the hormone surge correlated well with high levelsof ovarian FSHR mRNA at 2000 h on Day 4, and the levelsremained high on Day 1, 0900 h (Fig. 5). Phenobarbitoneitself had no effect on ovarian FSHR mRNA levels (datanot shown).

Because a second FSH surge occurs in the hamster by2200 h on Day 4 [28], it was of interest to assess whetherthe persistent low levels of FSHR mRNA at 0900 h on Day1 were caused by the second FSH surge. To achieve thisgoal, the occurrence of the second FSH surge was blockedby phenobarbitone injection at 2000 h on Day 4, and FSHRmRNA levels in the ovary were determined at 0900 h onDay 1. Although phenobarbitone blocked the second rise inserum FSH (data not shown) [22], FSHR mRNA levels didnot show appreciable change by 0900 h on Day 1 (Fig. 5).

Effect of Exogenous FSH on Ovarian FSHR mRNA Levels

Because serum levels of both FSH and LH increaseduring the preovulatory gonadotropin surge [1], the ratio-nale for this experiment was to determine whether FSHalone, dose and time dependently, could mimic the effectof the gonadotropin surge on ovarian FSHR mRNA levels.Because the lowest levels of endogenous FSH were pre-sent on the morning of Day 4, when the injection wasmade, and started to increase only after approximately1400 h [23], the effect of exogenous FSH on expressionof its receptor was examined on the morning of Day 4.Ovine FSH at a dose level of 8.2 IU reduced the levels ofFSHR mRNA to 50% of the 0800-h control value (Fig.


FIG. 1. Nucleotide sequence of the ham-ster FSHR cDNA showing 59- and 39-un-translated regions and a single open read-ing frame identified by 59- and 39-RACEstrategy. Sequence corresponding to signalpeptide and stop codon are in bold. Be-ginning of putative exons 2–10 (based onhuman FSHR exon sequences) is markedby vertical lines and labeled Ex-2 to -10.Exon 1 sequence starts from the nucleo-tide 2109. 11, Putative translation initia-tion site.

6). A higher dose did not show additional suppression(Fig. 6). Furthermore, 8.2 IU of recombinant human FSHwas also able to significantly reduce FSHR mRNA levelsby 4 h, and the effect was comparable to that of a similardose of ovine FSH (Fig. 6), thus indicating that FSH, with-out the presence of LH, can downregulate FSHR mRNAlevels in the hamster ovary.

To compensate for the short systemic half-life of FSHand because a comparable effect of 8.2- and 16.4-IU dosesof FSH on FSHR mRNA levels was observed, a time-

course study was done using 16.4 IU of ovine FSH. Sig-nificant reduction of FSHR mRNA was evident by 2 h afterthe FSH administration (Fig. 7), which corresponded wellwith the surge of the estrous cycle. The efficacy of FSHsuppression of FSHR mRNA was maintained up to 6 h;however, FSHR mRNA levels showed an upward trend 12h after the hormone administration (Fig. 7). This durationof suppression correlated well with the duration of the ef-fect of the endogenous increase in FSH levels, which main-tained the decline up to the morning of Day 1.

1584 ZHANG AND ROY

FIG. 2. Deduced amino acid sequence of the hamster FSHR protein and its comparison with rat, mouse, and human FSHR amino acid sequences.Shaded areas represent the differences in the amino acid residues across species. Signal peptide sequence is underlined, and four putative N-linkedglycosylation sites are overlined. Seven putative transmembrane domains are boxed.

TABLE 2. Putative posttranslational modification sites.

Posttranslational modification Highly probable sites (amino acid sequence number)

N-glycosylationSulfationProtein kinase C phosphorylationCasein kinase II phosphorylationTyrosine phosphorylationMyristylation

191–194, 199–202, 293–296, 311–314322–336, 327–34126–28, 251–253, 346–348, 554–556, 595–597, 666–66878–81, 193–196, 263–266, 313–316, 330–333, 563–566323–32913–18, 64–69

DISCUSSION

To our knowledge, this is the first study documenting thecomplete nucleotide sequence and deduced amino acidcomposition of the hamster FSHR. Although informationabout the complete FSHR nucleotide sequence is availablefor many species [6, 12, 32, 33], information about thehamster FSHR nucleotide sequence has been lacking. Be-cause FSH plays an important regulatory role in hamsterpreantral folliculogenesis, regulation of FSHR mRNA lev-els in the granulosa cells during folliculogenesis will be animportant area of study. However, to address this issue, wefirst need to know the complete nucleotide sequence of thehamster FSHR. Our data show that the nucleotide and ami-no acid sequences of hamster FSHR are approximately 90–92% similar to those of other species. Comparison of boththe cDNA sequence and amino acid composition of the

FSHR reveals that differences in the sequences of at least10–15% are common across species [12], indicating thatthe observed differences in the nucleotide sequence or ami-no acid residues in the hamster FSHR are not caused byerrors introduced by PCR or sequencing. Besides, the sizeof the reading frame and the length of the deduced FSHRprotein corroborate well with those reported for other spe-cies. The presence of the 17-amino-acid N-terminal signalpeptide and seven putative transmembrane domains alsomatch very well with those of other species [12, 33]. Fur-thermore, three conserved putative sites for N-linked gly-cosylation and 11 cysteine residues in the extracellular do-main of the hamster FSHR correspond well with those ofthe rat [9], mouse [12], human [6], and other species [17,34] and have been shown to be essential for the tertiaryconformation of the binding domain of other G protein-


FIG. 3. Northern hybridization analysis of hamster FSHR mRNA. A)Phosphorimage showing a [32P]cRNA-labeled, approximately 2.5-kbFSHR mRNA in ovarian RNA. B) Corresponding methylene blue-stainedribosomal RNA bands. Ad, Adrenal gland; K, kidney; L, liver; OV, ovary.

FIG. 5. Effect of phenobarbitone (phenobarb) on the relative levels ofFSHR mRNA in the hamster ovary. Days of the cycle are as follows: 1,Estrus; 4, proestrus. Bars with the same letter are significantly (P , 0.05)different from each other.

FIG. 4. Relative levels of FSHR mRNA in the hamster ovary throughoutthe estrous cycle. Days of the cycle are as follows: 1, Estrus; 2, metestrus;3, diestrus; 4, proestrus. Bars with the same letter are significantly (P ,0.05) different from each other.

FIG. 6. Relative levels of FSHR mRNA in the hamster ovary on Day 4following in vivo administration of increasing doses of FSH. Saline or FSHwas injected at 0800 h, and ovaries were collected 4 h later. Bars withsame letter are significantly (P , 0.05) different from each other. oFSH,ovine FSH; rhFSH, recombinant human FSH.

coupled receptors, such as LH [35]. However, in contrastto other rodent species, hamster FSHR contains one moreputative N-linked glycosylation site at positions 311–314 inthe extracellular domain. The presence of a fourth N-linkedglycosylation site in the extracellular domain has been re-ported for human and monkey FSHR [17]. Therefore, ham-ster FSHR seems to be structurally similar to the primateand human FSHR.

This pattern of glycosylation appears to be important forproper folding and membrane trafficking of the FSHR [36].Besides these, domains formed by amino acids 26–47 and317–332 correspond well with those of other species andparticipate in ligand-receptor interaction in the rat [37, 38].Furthermore, the conserved cytoloop II has been suggestedto be a putative site for interaction with G proteins [39].Similar to the mouse FSHR, hamster FSHR also contains11 conserved cysteine residues and may participate in thedisulfide binding between exoloops I and II in other G pro-teins [12]. Several serine, threonine, and tyrosine phos-phorylation sites also correspond to those of other species[12]. All these features allow us to contend that the nucle-otide and deduced amino acid sequences truly representFSHR, and that the hamster FSHR mRNA is derived from

1586 ZHANG AND ROY

FIG. 7. Time course of FSH effect on the relative levels of FSHR mRNAin the hamster ovary. Saline or FSH was injected at 0800 h on Day 4,and ovaries were collected at indicated times. For the 12-h group, phe-nobarbitone was injected s.c. at 1300 h to block the endogenous gonad-otropin surge. Bars with same letter are significantly (P , 0.05) differentfrom each other.

10 exons, similar to that reported for other species, includ-ing human [33].

Because the 59-RACE strategy does not allow truncatedmRNA to be amplified but does allow the 59-oligo primerto be ligated at the 59-cap site, the putative transcriptionstart site of the hamster FSHR mRNA likely is located closeto nucleotide 2109, and the 59-UTR of the hamster FSHRmRNA is 109-bases long. Whereas sequencing of the 59-flanking region of the mouse FSHR exon 1 from a genomiclibrary reveals a 500-base 59-UTR [40], genomic cloningof the human FSHR exon 1 containing the 59-flanking re-gion has demonstrated five major transcription start sites,of which the major one is located close to nucleotide 299[41]. Genomic cloning of the 59-flanking region of the ham-ster FSHR will be needed to predict definitive transcriptionstart site(s); however, the hamster FSHR cDNA sequencefrom 21 to 2109 bases matches well with the correspond-ing sequences of the mouse and rat.

Analysis of the exon-intron structure of the FSHR geneindicates the possibility for the formation of alternatelyspliced transcripts. In fact, several alternately spliced FSHRtranscripts have been identified in all mammals [17, 33, 42,43], including the hamster (unpublished observation), byreverse transcription-PCR amplification of gonadal RNA.However, in the present study, Northern blot analysis re-veals only one transcript of approximately 2.5 kb, whichmatches well with the 2392-bp FSHR cDNA. Because theprimary transcript of FSHR has a relatively low abundancein all species, the failure to detect alternately spliced FSHRtranscripts in hamster ovarian RNA likely is caused by lowsensitivity of the Northern hybridization technique. Identi-fication of a single FSHR transcript in rat ovarian RNA byNorthern hybridization has also been reported [16]. It is,however, important to remember that only single functionalFSHRs, corresponding to the full-length cDNA, have so farbeen identified in the gonads of all species studied [17, 33].Therefore, identification of only the full-length FSHRcDNA in the hamster ovary by Northern hybridization ful-fills the objective of the present study. The size of the ham-

ster FSHR cDNA correlates well with those of the rat, pig,human, and rabbit [33].

The increases in the levels of FSHR mRNA with pro-gress of the estrous cycle correlate well with the develop-ment of antral follicles in the hamster ovary. In the hamster,atretic antral follicles remain in the ovary on Day 1, and anew cohort of antral follicles appears on Day 2 and growscontinuously to become the graafian follicles by the morn-ing of Day 4 (proestrus) [1]. Analysis of serum levels ofFSH in the cyclic hamsters reveals a sharp fall in the hor-mone levels by the second half of Day 1 to reach consis-tently low levels by Day 2 through the morning of Day 4[23], which correspond to follicular growth with a concur-rent increase in FSHR in the granulosa cells [44]. Dosageof eCG, which induces in vivo antral follicular developmentin the rat, also upregulates FSHR transcripts in the ovary[15, 15, 22, 45]. Besides, low doses of FSH increase thelevels of FSHR mRNA in cultured rat [46, 47] and porcine[19] granulosa cells. In hypophysectomized, estrogen-treatedrats, FSH administration results in increased levels ofFSHR mRNA and FSH binding [20]. All these lines ofevidence suggest that low levels of FSH may favor in-creased levels of FSHR mRNA during follicular growth,with a concurrent increase in FSHR in the granulosa cells.

The decline in ovarian FSHR mRNA levels followingthe gonadotropin surge suggest that in contrast to low lev-els, high levels of FSH suppress ovarian mRNA levels.Moreover, LH may also influence the decline, because itslevels also increase during the gonadotropin surge [1]. Sig-nificant reduction in levels of FSHR mRNA following thegonadotropin surge in the rat has been reported [45]. In-volvement of the gonadotropin surge in the down-regulation of FSHR mRNA levels on the evening of pro-estrus is further evident by the ability of phenobarbitone toblock the decline in ovarian FSHR mRNA levels. Pheno-barbitone blocks the preovulatory gonadotropin surge in thehamster [22]. Interestingly, whereas the second FSH surgehas been suggested to be important for follicular recruit-ment in the hamster and rat [28] and for inhibin productionin the rat [48], it may not be responsible for the loweredlevels of ovarian FSHR mRNA observed on estrus. How-ever, it is important to realize that very few healthy, largeantral follicles, which are the primary contributors of highlevels of FSHR mRNA in the ovary, exist on estrous morn-ing. Therefore, future studies will examine the role of thesecond FSH surge in regulating ovarian FSHR mRNA lev-els. Although LH has been shown to suppress FSHR tran-script levels in the rat [16], FSH alone may play a majorrole in this process. This is evident from the ability of high-ly pure ovine FSH as well as recombinant human FSH tomimic the effect of the gonadotropin surge. Similar resultshave also been reported for rat FSHR mRNA [17]. Fur-thermore, recombinant FSH is capable of inducing ovula-tion in hypophysectomized rats [20]. Collectively, all theselines of evidence suggest that FSH, by a biphasic modality,regulates FSHR mRNA levels in hamster ovarian follicles.Whereas low doses may induce transcription of the FSHRgene, mRNA stability, or both, high doses may stimulatethe rate of mRNA degradation as well. These biphasic ac-tions of FSH, in turn, may be regulated by cyclic AMPalong with other intraovarian paracrine factors [17, 33].

In summary, we have identified a complete cDNA se-quence of the hamster FSHR and obtained the deducedamino acid sequence, which show all the characteristics ofa complete FSHR molecule. We have also shown changesin the levels of ovarian FSHR mRNA during the hamster


estrous cycle and established a role for FSH in modulatingthe levels of its own receptor mRNA in the ovary. Theseresults will help in future studies aimed at understandingthe mechanisms and intraovarian factors that are involvedin regulation of FSHR mRNA levels in the hamster.

ACKNOWLEDGMENT

We thank Dr. A.F. Parlow (National Pituitary Program) for generouslyproviding the ovine FSH at a subsidized price.

REFERENCES

1. Greenwald GS, Roy SK. Follicular development and its control. In:Knobil E, Neill JD (eds.), The Physiology of Reproduction, vol. 1,2nd ed. New York: Raven Press; 1994:629–724.

2. Kumar TR, Low MJ, Matzuk MM. Genetic rescue of follicle-stimu-lating hormone b-deficient mice. Endocrinology 1998; 139:3289–3295.

3. Dierich A, Sairam MR, Monaco L, Fimia GM, Gansmuller A, LeMeurM, Sassone-Corsi P. Impairing follicle-stimulating hormone (FSH)signaling in vivo: targeted disruption of the FSH receptor leads toaberrant gametogenesis and hormonal imbalance. Proc Natl Acad SciU S A 1998; 95:13612–13617.

4. Roy SK, Albee L. Requirement for follicle-stimulating hormone ac-tion in the formation of primordial follicles during perinatal ovariandevelopment in the hamster. Endocrinology 2000; 141:4449–4456.

5. Misrahi M, Beau I, Meduri G, Bouvattier C, Atger M, Loosfelt H,Ghinea N, Hai MV, Bougneres PF, Milgrom E. Gonadotropin recep-tors and the control of gonadal steroidogenesis: physiology and pa-thology. Baillieres Clin Endocrinol Metab 1998; 12:35–66.

6. Gromoll J, Pekel E, Nieschlag E. The structure and organization ofthe human follicle-stimulating hormone receptor (FSHR) gene. Gen-omics 1996; 35:308–311.

7. Heckert LL, Daley IJ, Griswold MD. Structural organization of thefollicle-stimulating hormone receptor gene. Mol Endocrinol 1992; 6:70–80.

8. Minegishi T, Nakamura K, Takakura Y, Ibuki Y, Igarashi M, MinegishT. Cloning and sequencing of human FSH receptor cDNA. BiochemBiophys Res Commun 1991; 175:1125–1130.

9. Sprengel R, Braun T, Nikolics K, Segaloff DL, Seeburg PH. The tes-ticular receptor for follicle stimulating hormone: structure and func-tional expression of cloned cDNA. Mol Endocrinol 1990; 4:525–530.

10. Aittomaki K, Herva R, Stenman UH, Juntunen K, Ylostalo P, HovattaO, De la Chapelle A. Clinical features of primary ovarian failurecaused by a point mutation in the follicle-stimulating hormone recep-tor gene. J Clin Endocrinol Metab 1996; 81:3722–3726.

11. Danilovich N, Javeshghani D, Xing W, Sairam MR. Endocrine alter-ations and signaling changes associated with declining ovarian func-tion and advanced biological aging in follicle-stimulating hormonereceptor haploinsufficient mice. Biol Reprod 2002; 67:370–378.

12. Tena-Sempere M, Manna PR, Huhtaniemi I. Molecular cloning of themouse follicle-stimulating hormone receptor complementary deoxy-ribonucleic acid: functional expression of alternatively spliced variantsand receptor inactivation by a C566T transition in exon 7 of the cod-ing sequence. Biol Reprod 1999; 60:1515–1527.

13. Roy SK. Regulation of follicular development: beyond gonadotropins.In: Jay KP, Krishna A, Halder C (eds.), Comparative Endocrinologyand Reproduction. New Delhi: Narosa Publishing House; 1999:315–330.

14. Wang J, Roy SK. Growth differentiation factor-9 and stem cell factorpromote primordial follicle formation in the hamster: modulation byfollicle-stimulating hormone. Biol Reprod 2004; 70:577–585.

15. LaPolt PS, Tossy JL, Aihara T, Nishimori K, Hsueh AJW. Gonadotro-pin-induced up- and down-regulation of ovarian follicle-stimulatinghormone (FSH) receptor gene expression in immature rats: effects ofpregnant mare’s serum gonadotropin, human chorionic gonadotropin,and recombinant FSH. Endocrinology 1992; 130:1289–1295.

16. Nakamura K, Minegishi T, Takakura Y, Miyamoto K. Hormonal reg-ulation of gonadotropin receptor mRNA in rat ovary during folliculargrowth and luteinization. Mol Cell Endocrinol 1991; 82:259–263.

17. Simoni M, Gromoll J, Nieschlag E. The follicle-stimulating hormonereceptor: biochemistry, molecular biology, physiology, and patho-physiology. Endocr Rev 1997; 18:739–773.

18. Minegishi T, Tano M, Kishi H, Kameda T, Miyamoto K. Follicle-stimulating hormone regulation on its receptor messenger ribonucleic

acid levels in cultured rat granulosa cells. Biochim Biophys Acta MolCell Res 1997; 1359:165–173.

19. Sites CK, Patterson K, Jamison CS, Friezner Degen SJ, LaBarberaAR. Follicle-stimulating hormone (FSH) increases FSH receptor mes-senger ribonucleic acid while decreasing FSH binding in cultured por-cine granulosa cells. Endocrinology 1994; 134:411–417.

20. Galway AB, Lapolt PS, Tsafriri A, Dargan CM, Boime I, Hsueh AJ.Recombinant follicle-stimulating hormone induces ovulation and tis-sue plasminogen activator expression in hypophysectomized rats. En-docrinology 1990; 127:3023–3028.

21. Gore-Langton RE, Armstrong DT. Follicular steroidogenesis and itscontrol. In: Knobil E, Neill JD (eds.), Physiology of Reproduction,vol. 1, 2nd ed. New York: Raven Press; 1994:571–627.

22. Siegel HI, Bast JD, Greenwald GS. The effects of phenobarbital andgonadal steroids on periovulatory serum levels of luteinizing hormoneand follicle-stimulating hormone in the hamster. Endocrinology 1976;98:48–55.

23. Bast JD, Greenwald GS. Serum profiles of follicle-stimulating hor-mone, luteinizing hormone and prolactin during the estrous cycle ofthe hamster. Endocrinology 1974; 94:1295–1299.

24. Roy SK, Greenwald GS. Quantitative analysis of in vitro incorpora-tion of [3H]thymidine into hamster follicles during the estrous cycle.J Reprod Fertil 1986; 77:143–152.

25. Roy SK, Greenwald GS. Effects of FSH and LH on incorporation of[3H]thymidine into follicular DNA. J Reprod Fertil 1986; 78:201–209.

26. Roy SK. Regulation of transforming growth factor b-receptor type Iand type II messenger ribonucleic acid expression in the hamster ova-ry by gonadotropins and steroid hormones. Biol Reprod 2000; 62:1858–1865.

27. Yang P, Roy SK. Epidermal growth factor modulates transforminggrowth factor receptor messenger RNA and protein levels in hamsterpreantral follicles in vitro. Biol Reprod 2001; 65:847–854.

28. Greenwald GS, Siegel HI. Is the first or second periovulatory surgeof FSH responsible for follicular recruitment in the hamster? Proc SocExp Biol Med 1982; 170:225–230.

29. Terranova PF. Effects of phenobarbital-induced ovulatory delay on thefollicular population and serum levels of steroids and gonadotropinsin the hamster: a model for atresia. Biol Reprod 1980; 23:92–99.

30. Yang P, Kriatchko A, Roy SK. Expression of ER-a and ER-b in thehamster ovary: differential regulation by gonadotropins and ovariansteroid hormones. Endocrinology 2002; 143:2385–2398.

31. Terranova PF, Garza F. Relationship between the preovulatory lutein-izing hormone (LH) surge and androstenedione synthesis of preantralfollicles in the cyclic hamster: detection by in vitro responses to LH.Biol Reprod 1983; 29:630–636.

32. Hernandez ER. Regulation of the genes for insulin-like growth factor(IGF) I and II and their receptors by steroids and gonadotropins inthe ovary. J Steroid Biochem Mol Biol 1995; 53:219–221.

33. Findlay JK, Drummond AE. Regulation of the FSH receptor in theovary. Trends Endocrinol Metab 1999; 10:183–188.

34. Yarney TA, Fahmy MH, Sairam MR, Khan H, Macdonald EA. On-togeny of FSH receptor messenger ribonucleic acid transcripts in re-lation to FSH secretion and testicular function in sheep. J Mol En-docrinol 1997; 18:113–125.

35. Zhang R, Buczko E, Dufau ML. Requirement of cysteine residues inexons 1–6 of the extracellular domain of the luteinizing hormone re-ceptor for gonadotropin binding. J Biol Chem 1996; 271:5755–5760.

36. Davis D, Liu X, Segaloff DL. Identification of the sites of N-linkedglycosylation on the follicle-stimulating hormone (FSH) receptor andassessment of their role in FSH-receptor function. Mol Endocrinol1995; 9:159–170.

37. Dattatreyamurty B, Reichert LE, Jr. A synthetic peptide correspondingto amino acids 9–30 of the extracellular domain of the follitropin(FSH) receptor specifically binds FSH. Mol Cell Endocrinol 1992; 87:9–17.

38. Leng N, Dattatreyamurty B, Reichert LE Jr. Identification of aminoacid residues 300–315 of the rat FSH receptor as a hormone bindingdomain: evidence for its interaction with specific regions of FSH bsubunit. Biochem Biophys Res Commun 1995; 210:392–399.

39. Arora KK, Sakai A, Catt KJ. Effects of second intracellular loop mu-tations on signal transduction and internalization of the gonadotropin-releasing hormone receptor. J Biol Chem 1995; 270:22820–22826.

40. Huhtaniemi IT, Eskola V, Pakarinen P, Matikainen T, Sprengel R. Themurine luteinizing hormone and follicle-stimulating hormone receptorgenes: transcription initiation sites, putative promoter sequences andpromoter activity. Mol Cell Endocrinol 1992; 88:55–66.

41. Gromoll J, Dankbar B, Gudermann T. Characterization of the 59-flank-

1588 ZHANG AND ROY

ing region of the human follicle-stimulating hormone receptor gene.Mol Cell Endocrinol 1994; 102:93–102.

42. Tisdall DJ, Watanabe K, Hudson NL, Smith P, McNatty KP. FSH-receptor gene expression during ovarian follicle development in sheep.J Mol Endocrinol 1995; 15:273–281.

43. O’Shaughnessy PJ, Dudley K, Rajapaksha WRAK. Expression of fol-licle stimulating hormone-receptor mRNA during gonadal develop-ment. Mol Cell Endocrinol 1996; 125:169–175.

44. Oxberry BA, Greenwald GS. An autoradiographic study of the bind-ing of 125I-labeled follicle-stimulating hormone, human chorionic go-nadotropin, and prolactin to the hamster ovary throughout the estrouscycle. Biol Reprod 1982; 27:505–516.

45. Camp TA, Rahal JO, Mayo KE. Cellular localization and hormonalregulation of follicle-stimulating hormone and luteinizing hormone re-ceptor messenger RNAs in the rat ovary. Mol Endocrinol 1991; 5:1405–1417.

46. Tano M, Minegishi T, Kishi H, Kameda T, Abe Y, Miyamoto K. Theeffect of follicle-stimulating hormone (FSH) on the expression of FSHreceptor in cultured rat granulosa cells. Life Sci 1999; 64:1063–1069.

47. Tilly JL, LaPolt PS, Hsueh AJW. Hormonal regulation of follicle-stimulating hormone-receptor messenger ribonucleic acid levels incultured rat granulosa cells. Endocrinology 1992; 130:1296–1302.

48. Woodruff TK, D’Agostino J, Schwartz NB, Mayo KE. Dynamicchanges in inhibin messenger RNAs in rat ovarian follicles during thereproductive cycle. Science 1988; 239:1296–1299.

Downregulation FSH

Documents