General and Comparative Endocrinology 160 (2009) 236–242

Contents lists available at ScienceDirect

General and Comparative Endocrinology

journal homepage: www.elsevier .com/locate /ygcen

Gonadotrophin subunit and GnRH receptor gene expression in the pars distalisof the equine pituitary

Julie Townsend, Karen Westcott, Domingo J. Tortonese *

Department of Anatomy, University of Bristol, Southwell Street, Bristol BS2 8EJ, UK

a r t i c l e i n f o a b s t r a c t

Article history:Received 11 August 2008Revised 10 October 2008Accepted 19 November 2008Available online 10 December 2008

Keywords:Seasonal breedingGonadotrophinGnRH-RmRNA

0016-6480/$ - see front matter � 2008 Elsevier Inc. Adoi:10.1016/j.ygcen.2008.11.029

* Corresponding author.E-mail address: d.tortonese@bristol.ac.uk (D.J. Tor

In the horse, pronounced changes in fertility occur annually in response to photoperiod. However, themechanisms regulating gonadotrophin synthesis and release in this species remain unclear. Here, weinvestigated the expression of gonadotrophin subunits and GnRH receptor (GnRH-R) mRNA in the pitu-itary glands of Thoroughbred horses during the breeding (BS) and non-breeding (NBS) season. Seasonaleffects on the prevalence of gonadotrophs in the pars distalis were also examined. GnRH-R and commona-, LHb- and FSHb-subunit mRNA contents were determined by Northern analysis and the prevalence ofLH–gonadotrophs assessed by immunohistochemistry in pituitaries from sexually active females (mares)in the BS, and sexually inactive mares in the NBS. These variables were then measured in castrated malehorses (geldings). In mares, pituitary content of FSHb mRNA was significantly higher in the NBS (P < 0.01).Conversely, the content of common a-subunit mRNA was significantly higher during the BS (P < 0.05). Incontrast, GnRH-R and LHb mRNA abundance were unaffected by season. Interestingly, whereas no sea-sonal effects were apparent on the number of LH–gonadotrophs/field, the proportion of LH cells (in rela-tion to all other cells) was higher in BS than NBS animals (P < 0.05); this resulted from an increasednumber of non-gonadotroph cells during the NBS (P < 0.05). In geldings, no significant seasonal effectswere detected for any of the variables investigated (P > 0.05). These results reveal robust seasonal effectson common a-subunit and FSHb gene expression in the pituitary of the mare, in the absence of detectablechanges in the content of LHb or GnRH-R mRNA.

� 2008 Elsevier Inc. All rights reserved.

1. Introduction

Reproduction in mammalian species is dependent on GnRH act-ing through its specific receptors on the pituitary gonadotrophcells to cause the synthesis and release of the gonadotrophins, LHand FSH. Although over the last two decades studies have investi-gated the regulation of gonadotrophin subunit gene and proteinsynthesis, most of this work has focused on animal models suchas the rat and the sheep. Very little research has been carried outin this area in the horse, particularly at the molecular level. Thehorse is a long-day seasonal breeder in which the sexually activephase of the annual reproductive cycle is normally restricted tothe long days of spring and summer. As for other photoperiodicspecies, this annual pattern of reproduction is dependent on sea-sonal variations in gonadotrophin secretion in both male and fe-males. In gonadal-intact females (mares), serum concentrationsof LH (Garcia and Ginther, 1976) and FSH (Silvia et al., 1987) arelow during the non-breeding season (NBS), and pituitary contentof LH is also significantly decreased when compared to breedingseason (BS) values (Hart et al., 1984). As daylength increases, basal

ll rights reserved.

tonese).

gonadotrophin concentrations and pituitary content of LH risegradually (Silvia et al., 1986). In the gonadal-intact male (stallion),serum LH concentrations vary with season, being 3–5 times higherin the BS compared to NBS (Harris et al., 1983; Irvine and Alexan-der, 1982; Thompson et al., 1977). These effects appear to be gona-dal-dependent in the male, as smaller (Hoffman et al., 1987) or no(Irvine and Alexander, 1982) changes are observed in long-termcastrates (geldings). Interestingly, however, gonadal-independenteffects of season on gonadotroph function are apparent in the fe-male, since seasonal variations in gonadotrophin secretion havebeen reported in ovariectomised mares (Fitzgerald et al., 1983;Freedman et al., 1979a). Pituitary GnRH receptor (GnRH-R) num-bers have also been determined using ligand-binding techniquesin intact mares at different stages of the annual reproductive cycle,but no significant effect of season was observed (Hart et al., 1984).While cDNA sequences for the equine gonadotrophin subunit(Chopineau et al., 1995; Saneyoshi et al., 2001; Sherman et al.,1992; Stewart et al., 1987) and GnRH-R genes (Porter et al.,1998) have been cloned and sequenced, as yet no studies have re-ported the effects of season on their expression.

Morphological changes in the gonadotroph population may alsohave important implications for reproductive activity in photoperi-odic species such as the horse. Recent studies by our group have re-

J. Townsend et al. / General and Comparative Endocrinology 160 (2009) 236–242 237

vealed important seasonal and gonadal effects on the cytologicalconfiguration of the equine pituitary. Proportions of immunoposi-tive LH–gonadotrophs in the pars distalis (PD) are significantly in-creased in stallions when compared to castrated males in thebreeding season (Tortonese et al., 2001; Townsend et al., 2004),and during the NBS gonadotroph numbers are increased in the parstuberalis (PT), but not the PD, of mares who continue to exhibitreproductive activity under these short day conditions when com-pared to non-cyclic (anoestrous) mares (Eagle and Tortonese,2000; Tortonese et al., 2001). This latter finding may represent along day breeder-specific characteristic, as no seasonal or gonadaleffects were reported for the gonadotroph population of the ovinePT (Skinner and Robinson, 1996). Comparisons of gonadotrophcells between cyclic mares in the BS and anoestrus mares in theNBS have, as yet, not been investigated.

In order to further delineate the pituitary mechanisms thatunderlie seasonal changes in fertility in the horse, this study wasconducted to characterise the expression of mRNA for the gonado-trophin subunits and GnRH-R in intact female and castrate malehorses during the breeding and non-breeding seasons. In addition,the prevalence of LH–gonadotroph cells in the PD in the sameexperimental groups was assessed to allow comparisons of LHbmRNA expression with that of its translated protein.

2. Materials and methods

2.1. Experiment 1

Thoroughbred mares (all gonadal-intact, without a clinical his-tory of reproductive dysfunction, and of an age range of 5–15years) were killed at an abattoir (Potter Abattoir, Taunton, UK)for commercial reasons and pituitary glands dissected out. Tissuewas collected during July/August in the breeding season (BS) andduring January/February in the non-breeding season (NBS). A de-tailed visual examination of the ovaries was conducted to confirmwhether NBS mares were non-cyclic and BS mares were cyclic;only such mares took part in the study. Mares were consideredto be non-cyclic if only a corpus albicans, small follicles and no cor-pora lutea were present in the ovaries, whereas mares were con-sidered to be cyclic if a recent corpus luteum together with afollicle (P2 cm) were present. Therefore, the two groups of mareswere at opposite stages of their annual reproductive cycles. It isimportant to note that variability within group resulting from le-vel/depth of anoestrus in NBS animals and from day of the oestrouscycle in BS animals is taken into account by the experimental de-sign for the test of hypothesis of this study. For RNA extraction,the PD region was carefully dissected from the pituitaries (BS,n = 5; NBS, n = 5), snap frozen in liquid nitrogen, and stored at�70 �C. For immunohistochemistry, dissected pituitaries (BS,n = 5; NBS, n = 5) were fixed in Bouins and embedded in paraffinwax as described previously (Eagle and Tortonese, 2000; Townsendet al., 2004). The pituitaries used here had not been utilised in anyprevious immunohistochemical analysis.

2.1.1. RNA extraction and Northern analysisTotal RNA was extracted from frozen PD tissue (250 mg) using a

RNA extraction kit (RNeasy, Qiagen, Crawley, UK), and 15–25 lgsamples were electrophoresed through denaturing 1.5% agarose/0.66 M formaldehyde gels in MOPS buffer. RNA was then trans-ferred to Hybond XL membrane (Amersham Biosciences Ltd, Chal-font St. Giles, UK) using 20� SSC buffer. Membranes were pre-hybridised for 6 h at 65 �C with 0.5 M sodium phosphate (pH7.2), 1 mM EDTA (pH 8.0), 7% SDS, 1% BSA, and 15% deionised form-amide. Hybridisation was carried out at 65 �C overnight with the32P labelled cDNA probes (see below) in fresh hybridisation buffer.

Membranes were then washed 3 times in 2� SSC, 0.1% SDS bufferat 65 �C, and exposed to X-ray film (Biomax MS, Kodak, UK) at�70 �C for 1–2 days. Each membrane was then stripped in boiling0.1% SDS solution and rinsed in 2� SSC and re-hybridised with a32P-labelled 18S ribosomal RNA oligonucleotide probe (kindly sup-plied by Dr. P. Saunders, MRC Human Reproductive Sciences Unit,Edinburgh, UK) to correct results for efficiency of loading. Eachautoradiograph was scanned and the intensity of the bands quan-tified using Quantity One analysis software (Biorad, Hemel Hemp-stead, UK). Changes in specific mRNA levels were expressedrelative to 18S rRNA concentrations.

2.1.2. Preparation of radiolabelled cDNA probescDNA probes for ovine GnRH receptor (Brooks et al., 1993), bo-

vine FSHb (Brooks et al., 1992) and equine common a-subunit (agift from Dr. P. Brown, MRC Human Reproductive Sciences Unit,Edinburgh, UK) were radiolabelled with [a-32P]dCTP (AmershamBiosciences) using a DNA labeling kit (Rediprime, Amersham Bio-sciences). A cDNA probe specific for equine LHb was radiolabelledas above after being generated by PCR as follows. Oligonucleotideprimers were designed (forward primer, 50-GCATCTGTGCCGGC-TACTGCC-30; reverse primer, 50-GGATGAGAGGAACGTCTG TGG-30)from the published equine LHb cDNA sequence (Sherman et al.,1992), and were predicted to generate a 350 bp product. Equinepituitary RNA (1 lg) was treated with Deoxyribonuclease I (Invit-rogen, Paisley, UK) and cDNA synthesised using a first strand cDNAsynthesis kit (Expand Reverse Transcriptase, Roche Diagnostics Ltd,Lewes, UK). Equine LHb cDNA was amplified using 5 ll cDNA as atemplate, with 1.25 U Taq DNA polymerase (AB Gene, Epsom, UK)in PCR buffer with 0.2 mM of each dNTP, 1.5 mM MgCl2 and2.5 pmol of the forward and reverse oligonucleotide primers in atotal volume of 50 ll. The PCR amplification conditions were: aninitial denaturation step at 95 �C for 2 min, followed by 35 cyclesof denaturation at 95 �C for 20 sec, annealing at 55 �C for 45 sec,and extension at 72 �C for 1 min; a final extension period at72 �C for 10 min completed the amplification. After amplification,15 ll of each PCR reaction was electrophoresed through a 1.5% aga-rose gel containing 0.1 lg/ml ethidium bromide, and the PCR prod-ucts were visualised under UV light. The PCR product was thensubcloned into a TA cloning vector (pCR II-TOPO, Invitrogen), andthe resulting clones sequenced to confirm they contained equineLHb cDNA sequence.

2.1.3. ImmunohistochemistySagittal sections (6 lm) were mounted on vectorbond-coated

slides (Vector Laboratories, Peterborough, UK) and dried overnightat 37 �C. After being dewaxed, rehydrated and washed in 0.05 MTris-buffered saline (TBS, pH 7.4), sections were blocked for 20 minfor endogenous peroxidase activity (one part 3% hydrogen peroxi-dase: four parts methanol). Sections were rinsed briefly with TBSand incubated overnight at 4 �C with normal rabbit serum (DAKOLaboratories Ltd. Ely, UK; 1:5 dilution in TBS). Each section was thenincubated with 100 ll of mouse monoclonal antibody raised againstbovine LHb subunit (518 B7, a gift from Dr. J. Roser, University of Cal-ifornia, Davis, CA, USA; 1:5000 dilution) overnight at 4 �C. This anti-body has been previously demonstrated to immunoreact in theequine pituitary and showed no cross reactivity with other pituitaryhormones (Eagle and Tortonese, 2000; Townsend et al., 2004). Neg-ative control sections received mouse IgG1 (DAKO) in place of pri-mary antibody. The sections were washed (4 � 5 min) in 0.01 MPBS containing 0.1% BSA (PBS/BSA, pH 7.4), and 100 ll of horseradishperoxidase-conjugated rabbit anti-mouse IgG (DAKO; 1:200 dilu-tion) applied at room temperature for 2 h. Finally, sections werewashed (3 � 10 min) in PBS/BSA, developed in 3,30-diaminobenzi-dine (DAKO) for 2 min, counterstained with haematoxylin andmounted using DPX (Merck Ltd. Poole, UK).

238 J. Townsend et al. / General and Comparative Endocrinology 160 (2009) 236–242

2.1.4. Quantitative analysis of immunohistochemical dataThree sections (separated by at least 30 lm) per pituitary were

immunostained to identify and count gonadotrophs and allremaining cell types using a light microscope (Reichert Biovar,Vienna, Austria). LH-immunopositive cells (brown staining) werecounted from five fields per section within the PD, totaling 15fields counted per animal. A field was defined by a 10 mm2 grati-cule divided into 100 squares, 20 of which were randomly selectedto be counted from. To account for the regional variation in theprevalence of cells throughout the PD, and to standardise thecounting procedure, the 5 different regions of the PD from whichthe cells were counted remained constant for each section. Onefield was taken from the rostral border of the PD, three from thecentre, and one from the border adjacent to the pars intermedia.The average cell density per field for each animal was calculated,as was the proportion of gonadotrophs, where the average densityof gonadotrophs was expressed as a percentage of the average totalcell density.

2.1.5. Statistical analysisThe effects of season on (i) the pituitary content of the gonado-

trophin subunits and GnRH-R mRNA and (ii) the number of LH–gonadotrophs and their proportion in relation to the entire pitui-tary cell population were examined by ANOVA using statisticalanalysis software (Statview 5.0, SAS Institute Inc., Cary, NC). Statis-tically significant differences were set at P < 0.05.

2.2. Experiment 2

2.2.1. Animals, tissue preparation and experimental proceduresPituitary glands were obtained from castrated male horses (all

Thoroughbreds in good body condition, age range 5–15 years)killed at an abattoir (Potters) for commercial reasons in eitherJuly/August (subjective BS) or January/February (subjective NBS);as orchidectomy in horses in the United Kingdon is carried outwithin the first year of age these were long-term castrates. Pitui-tary specimens were subjected to the same experimental proce-dures and analyses as for the mares (see above). Pituitaries werefrozen for RNA extraction (BS, n = 6; NBS, n = 6), or fixed for immu-nohistochemistry (BS, n = 5; NBS, n = 5) as before.

2.2.2. Statistical analysisThe effects of season on (i) the pituitary content of the gonado-

trophin subunits and GnRH-R mRNA and (ii) on the number of LH–gonadotrophs and their proportion in relation to the entire pitui-tary cell population in the castrated males were examined by AN-OVA as in Experiment 1.

3. Results

3.1. Experiment 1

3.1.1. Seasonal effects on gonadotrophin subunits and GnRH-R geneexpression in the PD of the mare’s pituitary gland

Messenger RNA for GnRH-R and all the gonadotrophin subunitswas expressed in the PD of the pituitary gland of intact femalehorses both in the BS and the NBS (Fig. 1). Pituitary contents ofGnRH-R and LHb mRNA were not significantly different (P > 0.05)between the BS and NBS. In contrast, FSHb mRNA contents weresignificantly lower (P < 0.01) in the BS (0.60 ± 0.22 arbitrary units(a.u.) when compared to the NBS (1.55 ± 0.17 a.u.). Conversely,common a-subunit mRNA contents were significantly higher(P < 0.05) in the BS (1.12 ± 0.20 a.u.) when compared to the NBS(0.44 ± 0.12 a.u.).

3.1.2. Seasonal effects on the number and proportion of LH–gonadotrophs and non-gonadotroph cells in the PD of the mare’spituitary gland

Immunohistochemical studies revealed abundant populationsof LH–gonadotrophs in the PD of the mare in both the BS andNBS (Fig. 2A and B). No staining for LHb was detected in the controlsections (Fig. 2A insert). While no significant effect of season wasapparent on the number of LH–gonadotrophs (Fig. 2C), significantdifferences (P < 0.05) were detected between the BS and NBS forthe number of non-gonadotroph cells (Fig 2D; 278.8 ± 13.6 vs.354.5 ± 10.0 cells/field, for BS and NBS groups, respectively). Whencombined, the above observations revealed a significant effect ofseason (P < 0.05) on the proportion (%) of LH–gonadotrophs in rela-tion to the entire pituitary cell population (Fig 2E; 22.9 ± 1.4 vs.16.5 ± 0.8%, for BS and NBS groups, respectively).

3.2. Experiment 2

3.2.1. Seasonal effects on gonadotrophin subunits and GnRH-R geneexpression and on the number and proportion of LH–gonadotrophsand non-gonadotroph cells in the PD of the guelding’s pituitary

While mRNA for the gonadotrophin subunits and GnRH-R wereexpressed in the gelding PD, no significant effects of season on theabundance of either the gonadotrophin subunits or GnRH-R wereapparent (Fig. 3). Similarly, no significant effect of season on thenumber of LH–gonadotrophs, on the number of non-gonadotrophsor in the proportion (%) of LH–gonadotrophs in relation to the en-tire pituitary cell population was found in the castrated males(Fig. 4).

4. Discussion

The results of this study provide clear evidence for significantseasonal effects on steady-state levels of FSHb- and common a-subunit mRNA in the pituitary PD of the mare. During the BS, thepituitary content of mRNA for FSHb was reduced, while that ofcommon a-subunit was increased in comparison to the NBS. Whilesimilar trends were observed in castrated male horses, these didnot reach statistical significance. In contrast to these observations,no apparent effect of season was found on the content of mRNA forGnRH-R or LHb-subunit in either mares or geldings. Interestingly,although seasonal effects on numbers of LH–gonadotrophs werenot apparent, a distinctive increase in the prevalence of LH–gonadotrophs was observed in mares during the BS which resultedfrom a statistically significant decrease in the number of non-gonadotroph cells at this time. Overall, these results reveal dra-matic changes in gonadotrophin subunit gene expression andalterations in the cytological configuration of the equine pituitarygland throughout the annual reproductive cycle.

The increase in pituitary content of FSHb-subunit mRNA ob-served in mares during the NBS was unequivocal. Several mecha-nisms can be proposed to account for these observations. Firstly,GnRH pulse frequency has been shown to cause differential effectson gonadotrophin subunit synthesis, with lower pulse frequencies(as those characteristic of the NBS) preferentially stimulating FSHbmRNA expression (Kaiser et al., 1997; Kirk et al., 1994). Previousstudies have also reported that both hypothalamic content ofGnRH in mares (Hart et al., 1984), and mean secretory rates ofGnRH in pony mares (Sharp and Grubaugh, 1987) were reducedduring seasonal anoestrus. Secondly, during this anovulatory peri-od, low circulating concentrations of oestradiol occur in the mare(Donadeu and Ginther, 2002; Watson et al., 2002), and this hor-mone is known to reduce FSHb gene expression in other species.Although there does not appear to be an oestrogen response ele-ment (ERE) in the promoter region of the FSHb gene (Hirai et al.,

Fig. 1. (A) Autoradiograph showing hybridisation of 32P-labelled cDNA probes for GnRH-R, LHb, FSHb and common a-subunit (CAS) with total RNA extracted from the parsdistalis of representative mares in the breeding (BS) and non-breeding (NBS) season. (B) Changes in amounts of mRNA for GnRH-R, LHb, FSHb and CAS in the pars distalis ofmares in the BS and NBS. Results were corrected for efficiency of loading by re-hybridising with a 32P-labelled 18S oligonucleotide probe. Data are expressed as means ± SEM(n = 5 per group); asterisks denote significant (*P < 0.05; **P < 0.01) differences between groups.

Fig. 2. (A and B) Immunohistochemical staining for LHb-subunit showing distribution of LH–gonadotroph cells in representative sagittal sections of pars distalis tissue frommares in the breeding (BS) and non-breeding (NBS) season. No staining was observed in control sections (insert). Scale bar (B) represents 50 lM. (C–E) Number of LH–gonadotroph cells/field, number of non-LH–gonadotroph cells/field, and proportion (%) of LH–gonadotrophs in relation to all pituitary cell types in the pars distalis of mares inthe breeding (BS) and non-breeding (NBS) season. Data are expressed as means ± SEM (n = 5 per group); asterisks denote significant (P < 0.05) differences between groups; NS,non-significant.

J. Townsend et al. / General and Comparative Endocrinology 160 (2009) 236–242 239

Fig. 3. (A) Autoradiograph showing hybridisation of 32P-labelled cDNA probes for GnRH-R, LHb, FSHb and common a-subunit (CAS) with total RNA extracted from the parsdistalis of representative geldings in the breeding (BS) and non-breeding (NBS) season. (B) Changes in amounts of mRNA for GnRH-R, LHb, FSHb and CAS in the pars distalis ofgeldings in the BS and NBS. Results were corrected for efficiency of loading by re-hybridising with a 32P-labelled 18S oligonucleotide probe. Data are expressed asmeans ± SEM (n = 6 per group).

Fig. 4. (A and B) Immunohistochemical staining for LHb-subunit showing distribution of LH–gonadotroph cells in representative sagittal sections of pars distalis tissue fromgeldings in the breeding (BS) and non-breeding (NBS) season. Scale bar (B) represents 50 lM. (C–E) Number of LH–gonadotroph cells/field, number of non-LH–gonadotrophcells/field, and proportion (%) of LH–gonadotrophs in relation to all pituitary cell types in the pars distalis of geldings in the breeding (BS) and non-breeding (NBS) season.Data are expressed as means ± SEM (n = 5 per group); asterisks denote significant (P < 0.05) differences between groups; NS, non-significant.

240 J. Townsend et al. / General and Comparative Endocrinology 160 (2009) 236–242

1990; Miller and Miller, 1996), evidence suggests that oestradiolmay indirectly affect FSHb mRNA expression via its action on the

activin bD gene (Baratta et al., 2001). Lastly, inhibin is known to de-crease both FSHb mRNA transcription (Clarke et al., 1993) and sta-

J. Townsend et al. / General and Comparative Endocrinology 160 (2009) 236–242 241

bility (Attardi and Winters, 1993) in other species, and in maresconcentrations of inhibin are low during anestrus (Donadeu andGinther, 2002; Watson et al., 2002). These factors may all accountfor the differences in patterns of expression of FSHb mRNA ob-served in the mares in the current study. The trend in the samedirection for increased FSHb mRNA expression during the NBS inthe gelding (although statistically non-significant) suggests thatthe seasonal/photoperiodic regulation of gonadotrophin subunitmRNA may be fully expressed only if the gonads are present.

The finding that steady-state levels of FSHb-subunit mRNAwere significantly higher in the NBS is in contrast with a previousreport showing that plasma FSH concentrations are lower at thistime (Silvia et al., 1987). Previous studies in sheep have indicatedthat FSH is secreted via a constitutive pathway, with very littlestorage within the pituitary; ovine FSH synthesis and secretionare thus tightly regulated (Brooks et al., 1992; Crawford et al.,2004; Winters, 1996). However, the results of the present worksuggest that in the horse this close coupling between synthesisand secretion is unlikely to occur, at least in the NBS, as at this timethe pituitary content of FSHb mRNA was unambiguously increased.Although one possible explanation is that FSH may be stored in thepituitary in the NBS, a previous study reported no significant effectof season on the pituitary content of FSH in mares (Hart et al.,1984), although careful examination of the data reveals a tendencyfor contents to be increased at this time. An alternative explanationis that during the NBS the translation of FSHb-subunit mRNA intoprotein is inhibited. Within this context, it is possible that a factorthat binds FSHb mRNA is produced to block its translation. Indeed,the existence of this mechanism of regulation has been demon-strated in the ovarian follicle, where LH inhibits the synthesis ofthe gap junction protein, connexin 43 (Kalma et al., 2004).

The present study has revealed that steady state levels of mRNAfor the common a-subunit in the mare are significantly decreasedduring the NBS. Again, similar but non-significant trends were ob-served at the same time in the geldings. This inverse relationshipbetween common a-subunit and FSHb-subunit mRNA levels isworthy of note, as while it is well known that activin increasestranscription of the FSHb gene (Bailey et al., 2004; Baratta et al.,2001; Carroll et al., 1989; Weiss et al., 1992), it also decreases tran-scription of the common a-subunit gene (Attardi et al., 1995).Although it has been suggested that the synthesis of the commona-subunit protein is not the rate-limiting step in the productionof gonadotrophin hormones (Brown and McNeilly, 1999; Shupnik,1996), it remains possible that the decreased content of its mRNAin the NBS may account, at least in part, for the reported decreasedsecretion of LH and FSH observed in horses at this time. Notwith-standing, the common a-subunit is also required for the synthesisof the structurally-related glycoprotein hormone, thyroid-stimu-lating hormone (TSH). While TSH concentrations also increase dur-ing the BS in mares (Buff et al., 2007), the effect of season onproportions of thyrotrophs or TSHb mRNA content in the horsepituitary is as yet unknown.

The lack of effect of season on steady-state levels of GnRH-RmRNA in either the mares or the geldings confirms the results ofan earlier study where concentrations of GnRH-R in mares, asdetermined by binding of GnRH analogues in pituitary tissue, wereunaltered throughout the annual reproductive cycle (Hart et al.,1984). In addition, in ovine pituitary cell culture experiments,GnRH-R mRNA contents were reported to be similar in cyclic andanoestrous animals (Ghosh et al., 1996). Therefore, the findingsof the present study are indicative that GnRH-R mRNA expressionis resistant to both the increased content of hypothalamic GnRH(Hart et al., 1984), and the higher ovarian oestradiol output (Freed-man et al., 1979b) previously reported in cyclic mares.

Steady-state levels of LHb-subunit mRNA were unaffected byseason in either mares or geldings. In contrast, a previous report

suggested that equine LHb mRNA levels increase during the longdays of summer (Sherman et al., 1992); however, as values fromonly three pituitaries, collected at a different time of the year, werepresented in that study, those results should be interpreted withcaution. Here, while no significant effect of season on the numberof immunopositive LH–gonadotroph cells/field was detected ineither mares or geldings, a significantly reduced proportion ofimmunopositive LH–gonadotroph cells was apparent in anoestrousmares. This finding resulted from an increase in the number ofnon-gonadotroph cells in the PD at this time. Interestingly,although no significant effects were noted for any of these vari-ables in the geldings, confirming the results of a previous studyby our group (Townsend et al., 2004), the number of non-gonado-troph cells/field showed a tendency to be higher in the NBS(P = 0.06). Despite our findings that steady-state levels of LHb-sub-unit mRNA and the numbers of immunopositive LH–gonado-trophs/field are unaffected by season in mares, previous workhas shown that both blood (Garcia and Ginther, 1976) and pitui-tary (Hart et al., 1984) LH concentrations are significantly lowerin anoestrous animals. Together, these results are indicative ofuncoupling between the rates of LH synthesis and release in thisspecies. Similarly, in sheep, clear changes in LH pulse frequencyand amplitude occur in the absence of detectable changes in stea-dy-state levels of LHb-subunit mRNA, which has been attributed tothe regulatory mode of secretion of this gonadotrophin (Brookset al., 1993; Crawford and McNeilly, 2002; Curry et al., 1989).

The uncoupling between LH synthesis and release could partlyresult from paracrine/juxtacrine regulation of the gonadotroph.In this regard, folliculostellate (FS) cells have been shown to affectgonadotroph function in rodents (Baes et al., 1987) and primates(Kawakami et al., 2002). FS-cells are non-endocrine glial-like cellsthat, amongst other factors, express annexin-1, a protein capableof activating receptors in neighbouring cells (Raynal and Pollard,1994). Interestingly, annexin-1 was shown to suppress forskolin-induced LH release in rat pituitary explants (John et al., 2003). Inthe horse, we have recently reported the widespread distributionof FS-cells throughout the PD (Henderson et al., 2008), and alsoprovided evidence that annexin-1 expression in the equine pitui-tary is seasonally regulated (Biedrzycki et al., 2006). Specifically,a significant increase in annexin-1-positive cells was detected dur-ing the NBS. Thus, a functional relationship is apparent betweenthe increased incidence of non-gonadotroph cells observed in thecurrent study at this time of year and the reported decrease inperipheral LH concentrations which occur despite unaltered LHb-subunit mRNA expression. Moreover, while the increase in annex-in-1-containing FS-cells is likely to contribute to the uncouplingbetween LH synthesis and release in the NBS, their decrease inthe BS may facilitate the enhanced LH secretory response to GnRHcharacteristic of this stage of the annual reproductive cycle.

In conclusion, this study has provided evidence for clear oppo-site effects of season on common a- and FSHb-subunit geneexpression in the female equine pituitary gland. Whereas steady-state levels of common a-subunit mRNA were found to be signifi-cantly increased in the BS, those of FSHb-subunit mRNA weregreatly reduced. In contrast, no seasonal effects on LHb-subunitor GnRH-R mRNA were apparent. In addition, while no changeswere observed in numbers of immunopositive LH–gonadotrophs,an increase in the number of non-gonadotroph cells during theNBS resulted in a reduced proportion of LH–gonadotrophs at thisstage of the annual reproductive cycle.

Acknowledgments

We thank Dr. J.F. Roser and Quidel Corporation (San Diego, CA,USA) for the gift of LHb monoclonal antibody, Dr. P. Brown (Edin-burgh, UK) for the equine a-subunit cDNA, and Dr P. Saunders

242 J. Townsend et al. / General and Comparative Endocrinology 160 (2009) 236–242

(Edinburgh, UK) for the 18S ribosomal RNA oligonucleotide probe.We also thank Potters Abattoir (Taunton, UK) for providing equinespecimens. These studies were financially supported by theDepartment of Anatomy, University of Bristol, and the BBSRC, UK.

References

Attardi, B., Klatt, B., Little, G., 1995. Repression of glycoprotein a-subunit geneexpression and secretion by activin in a-T3-1 cells. Mol. Endocrinol. 9, 1737–1749.

Attardi, B., Winters, S.J., 1993. Decay of follicle-stimulating hormone-b messengerRNA in the presence of transcriptional inhibitors and/or inhibin, activin, orfollistatin. Mol. Endocrinol. 7, 668–680.

Baes, M., Allaerts, W., Denef, C., 1987. Evidence for functional communicationbetween folliculo-stellate cells and hormone secreting cells in perfused anteriorpituitary cell aggregates. Endocrinology 120, 685–691.

Bailey, J.S., Rave-Harel, N., McGillivray, S.M., Coss, D., Mellon, P.L., 2004. Activinregulation of the follicle-stimulating hormone b-subunit gene involves Smadsand the TALE homeodomain proteins Pbx1 and Prep1. Mol. Endocrinol. 18,1158–1170.

Baratta, M., West, L.A., Turzillo, A.M., Nett, T.M., 2001. Activin modulates differentialeffects of estradiol on synthesis and secretion of follicle-stimulating hormone inovine pituitary cells. Biol. Reprod. 64, 714–719.

Biedrzycki, A., Maw, C., Tortonese, D.J., Townsend, J., 2006. Seasonal and gonadaleffects on annexin 1 expression in the male equine pituitary gland. Hormones 5(Suppl 1), P123.

Brooks, J., Crow, W.J., McNeilly, J.R., McNeilly, A.S., 1992. Relationship betweengonadotrophin subunit gene expression, gonadotrophin-releasing hormonereceptor content and pituitary and plasma gonadotrophin concentrationsduring the rebound release of FSH after the treatment of ewes with bovinefollicular fluid during the luteal phase of the cycle. J. Mol. Endocrinol. 8, 109–118.

Brooks, J., Taylor, P.L., Saunders, P.T.K., Eidne, K.A., Struthers, W.J., McNeilly, A.S.,1993. Cloning and sequencing of the sheep pituitary gonadotropin-releasinghormone receptor and changes in expression of its mRNA during the estrouscycle. Mol. Cell. Endocrinol. 94, R23–R27.

Brown, P., McNeilly, A.S., 1999. Transcriptional regulation of pituitarygonadotrophin subunit genes. Rev. Reprod. 4, 117–124.

Buff, P.R., Messer IV, N.T., Cogswell, A.M., Johnson, P.J., Keisler, D.H., Ganjam, V.K.,2007. Seasonal and pulsatile dynamics of thyrotropin and leptin in maresmaintained under a constant energy balance. Dom. Anim. Endocrinol. 33, 430–436.

Carroll, R.S., Corrigan, A.Z., Gharib, S.D., Vale, W., Chin, W.W., 1989. Inhibin, activin,and follistatin: regulation of follicle-stimulating hormone messengerribonucleic acid levels. Mol. Endocrinol. 3, 1969–1976.

Chopineau, M., Stewart, F., Allen, W.R., 1995. Cloning and analysis of the cDNAencoding the horse and donkey luteinizing hormone b-subunits. Gene 160,253–256.

Clarke, I.J., Rao, A., Fallest, P.C., Shupnik, M.A., 1993. Transcription rate of the FSHbgene is reduced by inhibin in sheep but this does not explain the decrease inmRNA. Mol. Cell. Endocrinol. 91, 211–216.

Crawford, J.L., McNeilly, A.S., 2002. Co-localisation of gonadotrophins and granins ingonadotropes at different stages of the oestrous cycle in sheep. J. Endocrinol.174, 179–194.

Crawford, J.L., McNeilly, J.R., McNeilly, A.S., 2004. No evidence for pituitary primingto gonadotropin-releasing hormone in relation to luteinizing hormone (LH)secretion prior to the preovulatory LH surge in ewes. Biol. Reprod. 71, 224–235.

Curry, T.E.J., Dean, D.D., Sanders, S.L., Pedigo, N.G., Jones, P.B., 1989. The role ofovarian proteases and their inhibitors in ovulation. Steroids 54, 501–521.

Donadeu, F.X., Ginther, O.J., 2002. Follicular waves and circulating concentrations ofgonadotrophins, inhibin and oestradiol during the anovulatory season in mares.Reproduction 124, 875–885.

Eagle, R.C., Tortonese, D.J., 2000. Characterization and distribution of gonadotrophsin the pars distalis and pars tuberalis of the equine pituitary gland during theestrous cycle and seasonal anoestrus. Biol. Reprod. 63, 826–832.

Fitzgerald, B.P., I’Anson, H., Loy, R.G., Legan, S.J., 1983. Evidence that changes in LHpulse frequency may regulate the seasonal modulation of LH secretion inovariectomized mares. J. Reprod. Fertil. 69, 685–692.

Freedman, L.J., Garcia, M.C., Ginther, O.J., 1979a. Influence of ovaries and photoperiodon reproductive function in the mare. J. Reprod. Fertil. (Suppl.), 79–86.

Freedman, L.J., Garcia, M.C., Ginther, O.J., 1979b. Influence of photoperiod andovaries on seasonal reproductive activity in mares. Biol. Reprod. 20, 567–574.

Garcia, M.C., Ginther, O.J., 1976. Effects of ovariectomy and season on plasmaluteinizing hormone in mares. Endocrinology 98, 958–962.

Ghosh, B., Wu, J., Miller, W.L., 1996. Gonadotropin-releasing hormone-stimulatedcalcium mobilization is altered in pituitary cultures from anoestrous ewes. Biol.Reprod. 54, 753–760.

Harris, J.M., Irvine, C.H.G., Evans, M.J., 1983. Seasonal changes in serum levels ofFSH, LH and testosterone and in semen parameters in stallions. Theriogenology19, 311–317.

Hart, P.J., Squires, E.L., Imel, K.J., Nett, T.M., 1984. Seasonal variation in content ofGnRH, pituitary receptors for GnRH, and pituitary content of LH and FSH in themare. Biol. Reprod. 30, 1055–1062.

Henderson, H.L., Hodson, D.J., Gregory, S.J., Townsend, J., Tortonese, D.J., 2008.Gonadotropin-releasing hormone stimulates prolactin release from lactotrophsin photoperiodic species through a gonadotrophin-independent mechanism.Biol. Reprod. 78, 370–377.

Hirai, T., Takikawa, H., Kato, Y., 1990. The gene for the b-subunit of porcine FSH:absence of consensus oestrogen-responsive element and presence ofretroposons. J. Mol. Endocrinol. 5, 147–158.

Hoffman, L.S., Adams, T.E., Evans, J.W., 1987. Circadian, circhoral and seasonalvariation in patterns of gonadotrophin secretion in geldings. J. Reprod. Fertil. 35(Suppl.), 51–58.

Irvine, C.H.G., Alexander, S.L., 1982. Importance of testicular hormones inmaintaining the annual pattern of LH secretion in the male horse. J. Reprod.Fertil. 32 (Suppl.), 97–102.

John, C.D., Christian, H.C., Morris, J.F., Flower, R.J., Solito, E., Buckingham, J.C., 2003.Kinase-dependent regulation of the secretion of thyrotrophin and luteinizinghormone by gloucocorticoids and annexin 1 peptides. J. Neuroendocrinol. 15,946–957.

Kaiser, U.B., Jakubowiak, A., Steinberger, A., Chin, W.W., 1997. Differential effects ofgonadotropin-releasing hormone (GnRH) pulse frequency on gonadotropinsubunit and GnRH receptor messenger ribonucleic acid levels in vitro.Endocrinology 138, 1224–1231.

Kalma, Y., Granot, I., Galiani, D., Barash, A., Dekel, N., 2004. Luteinizing hormone-induced connexin 43 down-regulation: inhibition of translation. Endocrinology145, 1617–1624.

Kawakami, S., Fujii, Y., Okada, Y., Winters, S.J., 2002. Paracrine regulation of FSH byfollistation in folliculostellate cell-enriched primate pituitary cell cultures.Endocrinology 143, 2250–2258.

Kirk, S.E., Dalkin, A.C., Yasin, M., Haisenleder, D.J., Marshall, J.C., 1994.Gonadotropin-releasing hormone pulse frequency regulates expression ofpituitary follistatin messenger ribonucleic acid: a mechanism for differentialgonadotrope function. Endocrinology 135, 876–880.

Miller, C.D., Miller, W.L., 1996. Transcriptional repression of the ovine folliclestimulating hormone-b gene by 17-b-estradiol. Endocrinology 137, 3437–3446.

Porter, M.B., Green, M.L., Simmen, R.C.M., Sharp, D.C., 1998. Cloning and sequencingof the equine GnRH receptor: amino acid homology with other species andexpression of the equine GnRH receptor. Biol. Reprod. 58 (Suppl.), 230.

Raynal, P., Pollard, H.B., 1994. Annexins—the problem of assessing the biologicalrole for a gene family of multifunctional calcium- and phospholipid-bindingproteins. Biochim. Biophys. Acta 1197, 63–93.

Saneyoshi, T., Min, K.S., Jing Ma, X., Nambo, Y., Hiyama, T., Tanaka, S., Shiota, K.,2001. Equine follicle-stimulating hormone: molecular cloning of b-subunit andbiological role of the asparagine-linked oligosaccharide at asparagine (56) of a-subunit. Biol. Reprod. 65, 1686–1690.

Sharp, D.C., Grubaugh, W.R., 1987. Use of push-pull perfusion techniques in studiesof gonadotrophin-releasing hormone secretion in mares. J. Reprod. Fertil. 35(Suppl.), 289–296.

Sherman, G., Wolfe, M., Farmerie, T., Clay, C., Threadgill, D., Sharp, D., Nilson, J.,1992. A single gene encodes the b-subunits of equine luteinizing hormone andchorionic gonadotropin. Mol. Endocrinol. 6, 951–959.

Shupnik, M.A., 1996. Gonadotropin gene modulation by steroids and gonadotropin-releasing hormone. Biol. Reprod. 54, 279–286.

Silvia, P.J., Squires, E.L., Nett, T.M., 1986. Changes in the hypothalamic–hypophysealaxis of mares associated with seasonal reproductive recrudescence. Biol.Reprod. 35, 897–905.

Silvia, P.J., Squires, E.L., Nett, T.M., 1987. Pituitary responsiveness of mareschallenged with GnRH at various stages of the transition into the breedingseason. J. Anim. Sci. 64, 790–796.

Skinner, D.C., Robinson, J.E., 1996. The pars tuberalis of the ewe: no effect of seasonor ovariectomy on the distribution, density or presence of immunoreactivecells. Cell Tiss. Res. 284, 117–123.

Stewart, F., Thompson, J.A., Leigh, S.E., Warwick, J.M., 1987. Nucleotide (cDNA)sequence encoding the horse a-subunit. J. Endocrinol. 115, 341–346.

Thompson, D.L., Pickett, B.W., Berndtson, W.E., Voss, J.L., Nett, T.M., 1977.Reproductive physiology of the stallion. VIII. Artificial photoperiod, collectioninterval and seminal characteristics, sexual behavior and concentrations of LHand testosterone in serum. J. Anim. Sci. 44, 656–664.

Tortonese, D.J., Gregory, S.J., Eagle, R.C., Sneddon, C.L., Young, C.L., Townsend, J.,2001. The equine hypophysis: a gland for all seasons. Reprod. Fertil. Dev. 13,591–597.

Townsend, J., Sneddon, C.L., Tortonese, D.J., 2004. Gonadotroph heterogeneity,density and distribution, and gonadotroph-lactotroph associations in thepars distalis of the male equine pituitary gland. J. Neuroendocrinol. 16,432–440.

Watson, E.D., Heald, M., Tsigos, A., Leask, R., Steele, M., Groome, N.P., Riley, S.C.,2002. Plasma FSH, inhibin A and inhibin isoforms containing pro- and -aCduring the winter anoestrus, spring transition and the breeding season inmares. Reproduction 123, 535–542.

Weiss, J., Harris, P.E., Halvorson, L.M., Crowley, W.F., Jameson, J.L., 1992. Dynamicregulation of follicle-stimulating hormone-b messenger ribonucleic acid levelsby activin, gonadotropin-releasing hormone in perifused rat pituitary cells.Endocrinology 131, 1403–1408.

Winters, S.J., 1996. Relationship between gonadotropin subunit messengerribonucleic acid levels and plasma gonadotropin concentrations in intact andorchidectomized adult rats. Biol. Reprod. 55, 828–832.