BIOLOGY OF REPRODUCTION (2012) 87(4):79, 1–10Published online before print 18 July 2012.DOI 10.1095/biolreprod.111.096735

The Hedgehog System in Ovarian Follicles of Cattle Selected for Twin Ovulations andBirths: Evidence of a Link Between the IGF and Hedgehog Systems1

Pauline Y. Aad,3,4 Sherrill E. Echternkamp,5 David D. Sypherd,5 Nicole B. Schreiber,4 and Leon J. Spicer2,4

4Department of Animal Science, Oklahoma State University, Stillwater, Oklahoma5U.S. Meat Animal Research Center, Agricultural Research Service, U.S. Department of Agriculture, Clay Center,Nebraska

ABSTRACT

Hedgehog signaling is involved in regulation of ovarianfunction in Drosophila, but its role in regulating mammalianovarian folliculogenesis is less clear. Therefore, gene expressionof Indian hedgehog (IHH) and its type 1 receptor, patched 1(PTCH1), were quantified in bovine granulosa (GC) or theca(TC) cells of small (1–5 mm) antral follicles by in situhybridization and of larger (5–17 mm) antral follicles by real-time RT-PCR from ovaries of cyclic cows genetically selected(Twinner) or not selected (control) for twin ovulations.Expression of IHH mRNA was localized to GC and cumuluscells, whereas PTCH1 mRNA was greater in TC than in GC.Estrogen-active (E-A; follicular fluid concentration of estradiol .progesterone) versus estrogen-inactive follicles had a greaterabundance of mRNA for IHH in GC and PTCH1 in TC.Abundance of IHH mRNA in GC was not affected by cowgenotype, whereas TC PTCH1 mRNA was less in large E-Afollicles of Twinners than in controls. In vitro, estradiol andwingless-type (WNT) 3A increased IHH mRNA in IGF1-treatedGC. IGF1 and BMP4 treatments decreased PTCH1 mRNA insmall TC. Estradiol and LH increased PTCH1 mRNA in IGF1-treated TC from large and small follicles, respectively. Insummary, functional status of ovarian follicles was associatedwith differences in hedgehog signaling in GC and TC. Wehypothesize that as follicles grow and develop, increased freeIGF1 may suppress expression of IHH mRNA by GC and PTCH1mRNA by TC, and these effects are regulated in a paracrine wayby estradiol and other intra- and extragonadal factors.

follicular development, follicular maturation, granulosa cells,ovary, theca cells

INTRODUCTION

In growing follicles, communication among the oocyte,granulosa cell (GC), and theca cell (TC) compartmentsmodulates proliferation and differentiation of these cells [1,

2]. The hedgehog (Hh) gene family was identified in 1990 as afamily of developmentally regulated morphogens controllingbasic embryonic developmental processes [3]. Three secretedlipid-modified protein ligands—Sonic (SHH), Indian (IHH),and Desert (DHH) Hh—have been identified in mammaliancells [3–6], including localization primarily within mouseovarian GC and TC [1, 7]. These ligands work through twomembrane-bound receptors: Patched (PTCH) 1 and 2 [5]. Bothreceptors bind smoothened (SMO), the seven-transmembraneG-protein-coupled coreceptor, in the absence of Hh ligands andsuppress Hh-induced intracellular transcriptional effectorsGlioma-associated oncogene homolog (GLI) 1, 2, and 3 [1,5, 8–10]. All three Hh ligands bind to PTCH1 with equalaffinity [11] and have been used interchangeably to invokecellular responses [11–14]. In Drosophila, Hh signaling andstimulation of PTCH promotes proliferation of both ovariansomatic stem cells [15] and germ cells [16]. In vertebrates, Hhsignaling is involved in regulating cell proliferation andsurvival and in determining cell fate, differentiation, andpolarity of embryonic cells [17, 18]. Recently, SHH was foundto stimulate mitogenesis in mouse GC [7] and bovine TC [19].In addition, GC IHH mRNA abundance is less, while TCPTCH1 mRNA is greater in small (2–6 mm) than in large (8–22 mm) ovarian follicles of cattle [19], and bovine GC IHHmRNA abundance was decreased by IGF1 but unaffected byFSH [19]. Although it is known that expression of Ihh, Dhh,and Ptch1 mRNA is rapidly lost in atretic follicles of mice [1]and PTCH1 mRNA abundance is less in large than small antralfollicles of cattle [19], it has not been determined if the Hhsystem is physiologically regulated during ovarian follicularselection and dominance in mammals. In the uterus, Ihhexpression and signaling patterns were shown to be modulatedby ovarian progesterone (P4) in mouse [20–22] and hamster[23], indicating endocrine regulation of the Hh system inreproductive tissues.

Cattle selected for twin ovulations and births (Twinner)have a twofold greater density of secondary preantral follicles[24], 50% more small (�5 mm) and medium (6–12 mm) antralfollicles [25], and .70% frequency of twin or multipleovulations [25–27] compared with females not selected fortwins (control). Consequently, Twinner cattle constitute anexcellent model to study the effect of Hh on folliculardynamics. We hypothesized that an increase in ovarianfollicular development in Twinner females may be facilitatedby increased activation of the Hh signaling pathway. Also, withgreater levels of serum and follicular fluid IGF1 in Twinnercattle [25, 28] and the importance of gonadotropins and IGF1as key regulators of ovarian follicular development andsteroidogenesis [29–32], cell culture experiments served toinvestigate the effects of gonadotropins, steroids, IGF1, andother hormones on components of the Hh system. Because thepathway for Hh signaling works together with that of bone

1Supported in part by the Oklahoma Agricultural Experiment Stationunder project H-2510 and approved for publication by the Director.Mention of trade names or commercial products in this publication issolely for the purpose of providing specific information and does notimply recommendation or endorsement by the U.S. Department ofAgriculture.2Correspondence: Leon J. Spicer, 114 Animal Science Bldg., Depart-ment of Animal Science, Oklahoma State University, Stillwater, OK74078. E-mail: leon.spicer@okstate.edu3Current address: Department of Natural and Applied Sciences, NotreDame University, Louaizeh, Lebanon.

Received: 30 September 2011.First decision: 30 October 2011.Accepted: 17 July 2012.� 2012 by the Society for the Study of Reproduction, Inc.eISSN: 1529-7268 http://www.biolreprod.orgISSN: 0006-3363

1 Article 79

Dow

nloaded from w

ww

.biolreprod.org.

morphogenetic protein (BMP)/transforming growth factor-bsignaling during embryogenesis [33, 34], we evaluated theeffect of BMP4, GDF9, activin, and gonadotropin hormones onovarian cell IHH and/or PTCH1 mRNA abundance. Objectivesof this study were to 1) identify whether the ovarian Hh systemwas associated with bovine ovarian follicular growth and/or agreater incidence of double ovulations in cattle and 2) study thehormonal regulation of the ovarian cell Hh system, especiallythe effects of gonadotropins, steroids, transforming growthfactor-b superfamily proteins, and IGF1 on components of theHh system.

MATERIALS AND METHODS

Experimental Design

The experimental design and procedures employed in experiments 1 and 2of this study were approved by the U.S. Department of Agriculture,Agricultural Research Service, U.S. Meat Animal Research Center (USMARC)Animal Care and Use Committee. Experimental procedures were conducted inaccordance with the Guide for the Care and Use of Agricultural Animals inAgricultural Research and Teaching. The USMARC Twinner population is acomposite population of cattle selected for the production of twin ovulationsand fraternal twin births (Twinner) since 1981 [27, 35]. Control females werefrom a contemporary crossbred population at USMARC that has not beenselected for twinning.

Experiment 1: Genetic Effect and Cellular Localization ofthe Hh System in the Bovine Ovary

Expression of mRNAs for IHH, PTCH1, and aromatase (CYP19A1) wasevaluated in small (1–5 mm) antral follicles of ovaries obtained from mature,cyclic Twinner and control cows. Estrous cycles were synchronized among 15Twinner and 15 control cyclic cows by the administration of a single injectionof prostaglandin F

2a (PGF; 30 mg i.m.; Lutalyse; Pfizer Animal Health) duringthe luteal phase of the estrous cycle. Presence of a functional corpus luteum(CL) was determined transrectally by real-time ultrasonography using a 7.5-MHz linear-array probe, Aloka 500 instrument (Corometrics Medical Systems).Cows were subsequently monitored for estrous behavior twice daily, andovulation was confirmed by ultrasonography by the absence of a dominantfollicle and the presence of a corpus hemorrhagicum 2 days after estrus.Ovaries were collected at slaughter on Day 3, 6, or 9 after estrus, storedimmediately on ice, and transported to the laboratory for processing. Days 3, 6,and 9 were selected because this is the period of the 21-day estrous cycle duringwhich the dominant follicle is selected and initiates its dominance oversubordinate follicles [25, 36]. Pieces of ovarian cortical tissue containingmultiple small antral follicles were dissected from the ovarian surface, fixed inneutral buffered formalin, dehydrated in ethanol and then xylene, andembedded in paraffin [37].

Experiment 2: Changes in the Hh System DuringFolliculogenesis in Twinner and Control Cows

Ovaries of 15 Twinner [25, 35] and 15 MARC I, II, and III (control) cowswere evaluated transrectally to determine presence of functional CL and torecord the follicular population. Cows with CL were injected with PGF2a (30mg; Lutalyse; Pfizer Animal Health) and ultrasounded 2 days prior toslaughter to confirm ovulation and record follicular populations. Cows wereslaughtered (no more than six cows per day) at follicle recruitment (Days 3–4;D3) or deviation (Days 5–6; D5) of the estrous cycle to evaluate if the Hhsystem changes during the time the dominant follicle is being selected andinitiates its dominance over subordinate follicles [25, 36]. Ovaries wereimmediately recovered and transported on ice to the laboratory. Up to 10 ofthe largest antral follicles (.4 mm in diameter) per pair of ovaries wereexcised from the ovaries and individually snap frozen and stored in liquidnitrogen.

Frozen follicles (range 5–17 mm) were processed as described previously[38] with modifications. Briefly, frozen follicles were bisected, and the outerfollicle wall was slightly thawed in order for the TC to be peeled out of thefollicle shell, as described previously [38]. Follicular fluid was then thawed byincubation at 378C for 5 min, then centrifuged at 1000 3 g at 48C for 5 min. TheGC pellet was lysed in 0.5 ml of TRIzol reagent (Invitrogen) vortexed,incubated for 5 min at 378C, and stored at �808C until RNA extraction. Thefollicular fluid supernatant was transferred into clean Eppendorf tubes and

stored at �208C until hormone assays. TC were suspended in 0.5 ml ofRNAlater (Ambion) at 48C overnight, then homogenized and treated withTRIzol Reagent (Invitrogen) and stored at �808C until RNA extraction, asdescribed below.

Estradiol (E2) and P4 levels in follicular fluid were determined byradioimmunoassay (RIA) as described previously [39, 40]. The intraassaycoefficient of variation was 10% for the P4 RIA and 12% for the E2 RIA.Follicles with E2:P4 ratio � 1 were considered healthy estrogen-active (E-A)follicles, whereas follicles with E2:P4 ratio , 1 were considered atreticestrogen-inactive (E-I) follicles [41]. Two control cows had no E-A follicles,and one control cow had no functional CL, and their data were removed fromall subsequent analyses.

Experiments 3–9: In Vitro Evaluation of the HormonalControl of the Hh Components

To evaluate the hormonal regulation of IHH and PTCH1 mRNA in GC andTC, respectively, ovaries of cattle obtained at slaughter from a nearby abattoirwere brought to the laboratory on ice and processed as described previously forobtaining GC and TC from small (GC, 1–5 mm; TC, 3–6 mm) and large (.8mm) follicles [42, 43]. For GC isolation, follicular fluid was aspirated using 20-gauge needles and syringes and centrifuged at 200 3 g for 5 min. For TCisolation, follicles were bisected after aspiration and washed with serum-freemedium to remove GC, and the theca layer was peeled from the follicular shelland digested for 1 h in 0.01 mg/ml DNase, 1.0 mg/ml hyaluronidase, 1.0 mg/mlcollagenase, and 1.0 mg/ml protease (Sigma) in serum-free medium mixture, asdescribed previously [43, 44]. These follicle-size categories were selectedbecause of their E2 production potential and their responsiveness to FSH andLH in the presence of IGF1, as described previously [43–45]. Prior to plating,GC and TC were resuspended in medium containing 1.25 mg/ml of collagenaseand 0.5 mg/ml of DNase (Sigma). Using the trypan blue exclusion method,small-follicle GC, large-follicle GC, small-follicle TC, and large-follicle TCviability averaged 74 6 7%, 66 6 6%, 94 6 1%, and 93 6 2%, respectively,at time of plating. Approximately 2 3 105 viable cells per well (24-well Falconplates; BD Biosciences) were seeded in 1 ml of medium; medium was a 1:1mixture of Dulbecco modified Eagle medium and Ham F12 containing 0.12mM gentamycin, 2.0 mM glutamine, and 38.5 mM sodium bicarbonate(Sigma). Cultures were kept at 38.58C in a 95% air, 5% CO

2atmosphere, and

for all experiments medium was changed every 24 h, as described previously[42, 43].

Unless stated otherwise, small- and large-follicle GC and TC were culturedfor 48 h in medium containing 10% FCS, washed twice with 0.5 ml of serum-free medium, and then cultured for an additional 24 h with hormonal treatmentsdescribed below. To evaluate the effects of three major inducive hormones ofthe ovary—E2, FSH, and IGF1—on abundance of IHH mRNA, GC wereuntreated or treated with E2 (300 ng/ml; Sigma) in the presence of ovine FSH(30 ng NIDDK-oFSH-20/ml; activity: 175 3 NIH-FSH-S1 U/mg; experiment3) or with E2 (300 ng/ml), recombinant human IGF1 (100 ng/ml; R&DSystems), or both (experiment 4). To compare hormonal regulation of IHHversus DHH mRNA (experiment 5), GC from small and large follicles weretreated for 24 h with either cortisol (300 ng/ml; Sigma), PGE2 (300 ng/ml;Sigma), recombinant human SHH (500 ng/ml; R&D Systems), recombinanthuman wingless-type mouse mammary tumor virus integration site (WNT) 3A(300 ng/ml; R&D Systems), or recombinant human angiogenin (ANG; 300 ng/ml; R&D Systems); all treatments included 30 ng/ml IGF1. Cortisol, PGE2,SHH, WNT3A, and ANG were tested because of their implication in cysticfollicle development [46] and studies showing that these doses significantlyalter GC function [19, 44–48].

Hormonal regulation of PTCH1 mRNA was studied in small- and large-follicle TC treated with IGF1 (0 or 100 ng/ml) and/or ovine LH (0 or 30 ng/ml;NIDDK-oLH-26; activity: 1.0 3 NIH-LH-S1 U/mg) for 24 h (experiment 6), inlarge-follicle TC treated with either IGF1 (0 or 30 ng/ml) or E2 (0 or 300 ng/ml) for 24 h (experiment 7), or in TC treated with recombinant human bonemorphogenetic protein-4 (BMP4; 0 or 30 ng/ml; R&D Systems), recombinantrat growth differentiation factor-9 (GDF9; 0 or 500 ng/ml; provided by Dr.Aaron J. W. Hsueh, Stanford University School of Medicine, Stanford, CA)[43], or recombinant human activin (0 or 25 ng/ml; R&D Systems) for 24 h(experiment 8). The effect of culture on PTCH1 expression (experiment 9) wascompared among large-follicle TC uncultured (freshly isolated), cultured for 48h in 10% FCS, or cultured for an additional 48 h in serum-free medium after theinitial 48 h in 10% FCS.

At the end of the treatment period, medium was aspirated, cells were lysedin 0.5 ml of TRI reagent, and RNA was extracted as described below.Abundance of IHH, DHH, and PTCH1 mRNA were quantified using real-timeRT-PCR and was normalized to constitutively expressed 18S rRNA asdescribed below.

AAD ET AL.

2 Article 79

Dow

nloaded from w

ww

.biolreprod.org.

In Situ Hybridization

Relative abundance of IHH, PTCH1, and CYP19A1 mRNA expressionwithin individual small antral follicles was determined by in situ hybridizationanalysis described previously [37]. Antisense and sense 35S-rUTP-labeledcRNA probes were transcribed from linearized cDNA using an in vitrotranscription kit (Ambion), according to the manufacture’s recommendation,and 35S-rUTP (PerkinElmer Health Sciences). The 223-base-pair (bp) ampliconof bovine IHH and 240-bp amplicon of bovine PTCH1 were synthesized byPCR, cloned into pCR4TOPO plasmid vector (Invitrogen) and transformed intoTOP10 chemically competent E. coli cells (Invitrogen). Forward and reverseprimers for IHH synthesis were AGAGTGGCAGCTGTCTCCA andGGTAGAGCAGCTGGGGGTA (NCBI accession no. NM_001076870),respectively. Identity and orientation of the 223-bp amplicon of bovine IHHwere confirmed by sequencing; identity was 100% with mRNA sequence(bases 755–977) of NCBI accession NM_001076870. Forward and reverseprimers for PTCH1 were TGTCAGGCATCAGTGAGGAG and AGCATTCTCTGGGGGTTTCT, respectively, and sequencing confirmed thatthe 240-bp amplicon for PTCH1 had 100% identity with mRNA sequence(bases 3836–4075) of NCBI accession no. NM_001205879. Transfected cellscontaining cDNA plasmids for CYP19A1 [37, 49] were provided by Dr. AllenGarverick, University of Missouri, Columbia, and CYP19A1 antisense andsense probes were transcribed as described previously [37, 49].

Multiple sections (8 lm) of the embedded tissue were cut, mounted on glassmicroscope slides, deparaffinized, and subsequently incubated with proteinaseK for 10 min. Tissue sections were prehybridized at 428C and hybridized withthe respective probes overnight at 558C and 100 ll hybridization solution/slideat a probe concentration of 5000 dpm/ll. After hybridization, slides werewashed, treated with ribonuclease (RNase) for 30 min, washed, lightlycounterstained with hematoxylin, and dehydrated. Slides were dipped in KodakNTB-2 emulsion (Eastman Kodak), followed by exposure for 4 wk at 48C in adesiccated dark box. Slides were developed in Dektol (Eastman Kodak), fixedwith Kodak Fix (Eastman Kodak), dehydrated and cleared, and mounted forexamination by bright-field and dark-field microscopy. For each animal, twosections were hybridized to the antisense probe, and one section was hybridizedto the sense probe for each gene. Sections of cortical tissue from Twinner andcontrol cows from each day of the estrous cycle were included within ahybridization run to minimize biases among runs.

Hybridization intensity was quantified using the Bioquant Nova Prime imageanalysis system (BI-OQUANT Image Analysis Corporation). Within a markedarea of interest within the GC or TC layer of an individual antral follicle, thesystem determined the total number of pixels and the number of graphic pixelsoccupied by the silver grains. Hybridization intensity was defined as theproportion of total pixels occupied by graphic pixels. Specific hybridizationintensity within an individual follicle was defined as the average of four fieldswithin a follicle cell type for the antisense probe minus the sense probe andexpressed as the proportion of the area occupied by specific grains. Measurementswere collected on two to five follicles per cow, and similar numbers of small antralfollicles were evaluated for Twinner (n¼56) and control (n¼50) cows.

RNA Extraction and RT-PCR for mRNA Quantification

RNA from GC samples of experiment 2 was extracted in 13 batches with anaverage of 13 samples per extraction batch, and RNA from TC was extracted ineight batches with an average of 20 samples per extraction batch. Each batchconsisted of an equal number of samples from each treatment group and asimilar number of different size follicles. GC, stored in 0.5 ml of TRIzol at�808C, had RNA extracted using the TRIzol protocol as described previously[43, 45]. TC, stored in 0.5 ml of RNAlater (Ambion), were transferred into0.75 ml of TRIzol Reagent and homogenized for 2–3 min on ice using theOmni TH tissue homogenizer (Omni International Inc.) with Omni Tipdisposable generator probes to prevent contamination between treatments. Afterextraction, GC and TC RNA was quantified via ultrasensitive fluorescentnucleic acid staining using the RiboGreen RNA Quantitation Reagent and Kit(Molecular Probes) with modifications as described previously [50] and using afluorescent plate reader (Wallac 1420, PerkinElmer). The intraassay coefficientof variation was ,10%. GC and TC from cell culture experiments (experiments3–7) were lysed in 0.5 ml of TRIzol Reagent; RNA was extracted andquantified at 260 nm using NanoDrop spectrophotometer (ND-1000; Nano-Drop Technologies) as described previously [19, 45].

The differential expression of target gene mRNA in TC and GC wasquantified using the multiplex one-step real-time RT-PCR reaction forTaqman Gold RT-PCR Kit (Applied Biosystems Inc.) as describedpreviously for IHH, PTCH1, and the housekeeping gene 18S rRNA [19,51]. The DHH (848 bp; accession no. XM002687306.1) forward primersequence was constructed between 46 and 66 bp with a Tm of 598C andsequence of CAACCCCGACATCATCTTCAA, its reverse primer sequence

was CACATGTTCATCACGGCTATGG and was constructed between 156and 135 bp with a Tm of 608C, and its TaqMan probe wasCGCCTGATGACCGAGCGTTGTAAGG and was constructed between 92and 116 bp with a Tm of 708C. The DHH primers spanned introns (forwardprimer was located in exon 1, bases 243–263 of its gene [accession no. AC000162] and reverse primer was located in exon 2, bases 3038–3059 of itsgene; 15 bases of the probe are located in exon 1, bases 289–303, and theremaining 10 bases of the probe are located in exon 2, bases 3010–3019 ofgene; exons 1 and 2 are separated by a 2706-base intron 1). The IHHprimers also spanned introns (forward primer was located in exon 2, bases3061–3082 [accession no. AC_000159], and reverse primer was located inexon 3, bases 4865–4882 of its gene; the TaqMan Probe was located inexon 2 bases 3088–3109). The PTCH1 primers did not span introns. For allRT-PCR runs, a no-template control and a no-reverse-transcriptase controlwere included to ensure the lack of contaminants in the master mix and theabsence of any genomic DNA contamination, respectively. In addition, theRT-PCR products were run on agarose gels to verify the length and size ofthe expected target genes, and the same RT-PCR cDNA samples were usedto verify the amplified sequences.

Because the total number of samples in experiment 2 was greater than the 96-well-plate capacity, cows were sorted by genotype and cycle. Cows withingenotype and cycle were assigned randomly to each plate, with all follicles fromone individual cow on the same plate; all individual samples were run in duplicate.The 18S rRNA values were used as internal controls to normalize samples for anyvariation in amounts of RNA loaded as described previously [19, 43, 52]. Relativequantity of target gene mRNAs was expressed as 2�DDCt using the relativecomparative threshold cycle method as described previously [50, 53].

Statistical Analyses

In experiment 1, genoytpe and day of estrous cycle effects on mRNAexpression for the three genes within the small antral follicles were tested foreach gene by PROC MIXED procedure (Windows version 8.02; SAS InstituteInc.) for repeated measurements. Independent fixed effects and two-wayinteractions in the initial statistical model were genotype and day of the estrouscycle. Individual follicles were subsequently classified as E-I or E-A wheneither �10% or .10%, respectively, of the analyzed area was occupied byspecific grains for CYP19A1, and estrogen status and its interaction were addedto the repeated measures analysis as independent fixed effects. Relationshipsamong measured traits were assessed by PROC CORR of SAS.

Data from experiment 2 were analyzed as a completely randomized designwith a 2 3 2 3 2 factorial treatment structure with main effects of genotype(control vs. Twinner), day of estrous cycle (recruitment D3, or deviation D5),and follicle estrogenic profile (healthy E-A, E2:P4 ratio � 1, or atretic E-I,E2:P4 ratio , 1) [41]. Treatment effects on the dependent variables (i.e.,abundance of IHH and PTCH1 mRNA) were determined using the MIXEDprocedure of SAS, where cow was included as a random effect. Outliers weredetected according to the procedure described by Grubbs [54] provided byGraphPad Software (http://www.graphpad.com/quickcalcs/Grubbs1.cfm).Mean differences were determined by Fisher’s protected least significantdifferences test [55] if significant treatment effects in analysis of variance(ANOVA) were detected. The slice option in the least squares means(LSMeans) statement of SAS was used to separate the main effects meandifferences when interactions were significant (P , 0.05). Results are presentedas LSMeans 6 SEM.

Data from cell culture experiments were analyzed as a completelyrandomized design with a one-way ANOVA (experiments 3, 5, 7, 8, and 9)or a 2 3 2 factorial treatment structure (experiments 4 and 6) using the GLMprocedure of SAS. Because only 49% of samples for large GC of experiment 5expressed IHH mRNA, the Ct for IHH mRNA for undetectable samples was setat 45 and then analyzed. Experimental data are presented as means 6 SEM ofmeasurement from replicated experiments. Each experiment from cell culturewas replicated three or more times, and within each experiment, treatmentswere applied in four culture wells, two of which were combined to generateduplicate samples for each treatment within each experiment. Each experimentwas conducted on a separate pool of GC and TC obtained from 3 to 15 cows orheifers. To correct for heterogeneity of variance, abundance of IHH, DHH, andPTCH1 mRNA were analyzed after natural log (x þ 1) transformation.

RESULTS

Genetic Effect and Localization of the Hh Components inthe Bovine Ovary by In Situ Hybridization (Experiment 1)

Expression of IHH mRNA, measured by in situ hybridiza-tion, within individual small antral follicles was localized

HEDGEHOG SYSTEM IN TWINNER CATTLE

3 Article 79

Dow

nloaded from w

ww

.biolreprod.org.

primarily within the GC (Fig. 1C), including cumulus cells(Fig. 2A), whereas PTCH1 mRNA was expressed in TC (Figs.1D and 2B) and, to a lesser degree in GC, of small ovarianfollicles from both Twinner and control cows. Abundance ofIHH mRNA in GC and of PTCH1 mRNA in TC was notaffected (P . 0.10) by genotype, but their abundance wasgreater (P , 0.01) in small E-A vs. E-I antral follicles (Table1); E-A vs. E-I defined as follicles with . 10% vs. , 10% ofthe analyzed area occupied by specific grains for CYP19A1,respectively. Thus, abundance of IHH mRNA was correlatedpositively with CYP19A1 mRNA in GC (r ¼ 0.85; P , 0.01)and with PTCH1 mRNA in TC (r ¼ 0.70; P , 0.01) but wascorrelated negatively with PTCH1 mRNA in GC (r¼�0.28; P, 0.05). Expression for GC IHH mRNA was not affected (P .0.10) by day of cycle, whereas abundance of PTCH1 mRNA inE-I follicles but not in E-A follicles decreased from D3 to D9(estrogen status by day; P , 0.05; Fig. 3).

Gene Expression in GC and TC of Control Cows and CowsSelected for Double Ovulations (Experiment 2)

Ratio of E2:P4 concentration in the follicular fluid was usedto classify the health (estrogen) status of individual antralfollicles [41, 56]. The E2:P4 ratio differed (P , 0.001)between healthy (E-A; 6.62 6 0.41) and atretic (E-I; 0.12 60.33) follicles, but the ratio was not affected (P . 0.10) bygenotype, day of cycle, or any of the two- or three-wayinteractions (data not shown). Health status (P , 0.01) andstatus by day of cycle interaction influenced (P , 0.05)follicular fluid E2 such that E2 concentrations were greater inE-A follicles on D5 (179 6 21 ng/ml) vs. D3 (166 6 34 ng/ml). Concentrations of E2 averaged 5.2 6 1.5 ng/ml and 2.8 60.7 ng/ml in E-I follicles on D3 and D5, respectively. Onlyhealth status affected (P , 0.01) P4 concentrations with E-Ifollicles (92.7 6 17 ng/ml) having greater P4 concentrationsthan E-A follicles (29.6 6 2.6 ng/ml).

Relative abundance of IHH mRNA in GC of antral folliclesincreased from D3 to D5 in healthy (E-A) follicles but did notchange in atretic (E-I) follicles (follicle status by day of cycle;P ¼ 0.02; Fig. 4), abundance of IHH mRNA being 3.3-foldgreater (P , 0.05) in healthy (E-A) follicles than in atretic (E-I)follicles on D5. The abundance of IHH mRNA did not differ (P. 0.10) in antral follicles between Twinner and control cattleregardless of follicle status (E-A or E-I) or day of the estrouscycle.

Relative abundance of PTCH1 mRNA in TC of antralfollicles increased (P , 0.05) between D3 and D5 (Table 2 andFig. 5A). In addition, there was a genotype by follicle statuseffect (P , 0.05) on PTCH1 gene expression in TC of antralfollicles (Fig. 5B). Healthy (E-A) follicles of control cows had3.5-fold greater (P , 0.05) TC PTCH1 mRNA abundance thanatretic (E-I) follicles of control cows and 3.2-fold greater (P ,0.05) TC PTCH1 mRNA abundance than healthy (E-A)follicles of Twinner cows (Fig. 5B).

Regulation of IHH and DHH mRNA in Bovine GranulosaCells (Experiments 3–5)

Hormonal and signaling-pathway regulation of IHH andDHH mRNA was evaluated in cultured small-follicle GCtreated with FSH without or with E2 (experiment 3) or IGF1(experiment 4) and in small- and large-follicle GC treated with

FIG. 1. Adjacent bright-field and dark-field photomicrographs of in situ hybridization images for CYP19A1, IHH, and PTCH1 mRNA with 35S-rUTP-labeled antisense probes (experiment 1). A) Bright-field view of the wall of a small antral follicle identifying the location of the granulosa (g) and thecainterna (t) layers. B) Localization of CYP19A1 mRNA transcription. C) Localization of IHH mRNA transcription. D) Localization of PTCH1 mRNAtranscription. Original magnification 320.

FIG. 2. Dark-field photomicrographs of in situ hybridization images forIHH and PTCH1 mRNA with 35S-rUTP-labeled antisense probes(experiment 1). Localization of IHH mRNA in granulosa and cumuluscells (A) and localization of PTCH1 mRNA in theca interna cells (B) ofadjacent cross sections of small antral follicles. Original magnification30.75.

AAD ET AL.

4 Article 79

Dow

nloaded from w

ww

.biolreprod.org.

cortisol, PGE2, SHH, WNT3A, and ANG (experiment 5).FSH, E2, and IGF1 were tested because they are three majorinducive hormones of the ovary, and cortisol, PGE2, SHH,WNT3A, and ANG were tested because of their implication incystic follicle development [46] and studies showing that thesedoses significantly alter GC function [19, 44–48]. In culturedGC, a 2-day treatment with 300 ng/ml E2 increased (P , 0.05)IHH mRNA abundance in the presence of FSH (Fig. 6A) aswell as in the presence of IGF1 (P , 0.05; Fig. 6B), the latterof which suppressed (P , 0.05) IHH mRNA abundance.WNT3A increased (P , 0.05) IHH mRNA abundance in GCfrom small and large follicles by 5.5- and 8.6-fold, respectively(Fig. 7). Similarly, cortisol increased (P , 0.05) IHH mRNAabundance in GC from small and large follicles by 1.9- and2.85-fold, respectively (Fig. 7). In contrast, PGE2, SHH, andANG had no effect (P . 0.10) on IHH mRNA abundance inGC of small and large follicles (Fig. 7). Abundance of DHHmRNA in GC of small and large follicles did not differ (P .0.10) among treatments (Fig. 7); only 49% of samples for largeGC of experiment 5 had detectable IHH mRNA.

Regulation of PTCH1 mRNA in Bovine Theca Cells(Experiments 6–9)

Experiments 6–9 were conducted to evaluate the effects ofmajor inducive hormones (i.e., IGF1, LH, and E2) andsignaling pathways (i.e., BMP4 and GDF9) of TC on PTCH1mRNA. In cultured TC from small follicles (experiment 6),both IGF1 and LH decreased (P , 0.05) PTCH1 mRNAexpression independently (Fig. 8A), whereas the combinedtreatment of IGF1 and LH was less inhibitory on PTCH1mRNA expression than either treatment alone (Fig. 8A).Likewise, IGF1 decreased (P , 0.05) PTCH1 mRNAexpression in control and LH-treated TC from large follicles(Fig. 8B), but the effect of LH alone on abundance of PTCH1mRNA was not significant (P . 0.10). In experiment 7, E2increased (P , 0.05) PTCH1 mRNA expression by 36% inIGF1-treated TC from large follicles (18.1 vs. 13.3 6 0.9relative abundance), but E2 did not affect (P . 0.10)abundance of PTCH1 mRNA in the absence of IGF1 (datanot shown). In experiment 8, treatment of small-follicle TC

TABLE 1. Comparison of CYP19A1, IHH, and PTCH1 mRNA expression among small antral follicles from Twinner and control cattle assessed by in situhybridization (experiment 1).a

Itemb Nc

Twinner Control Combinedd

E-I, % E-A, % E-I, % E-A, % E-I, % E-A, %

CYP19A1, GC 15 5.8 6 0.1 36.4 6 2.2 4.1 6 3.1 34.4 6 2.7 5.0 6 2.8 35.0 6 2.0**IHH, GC 15 1.3 6 4.2 31.7 6 3.7 0.7 6 0.6 34.0 6 4.1 1.0 6 2.9 32.9 6 2.3**PTCH1, GC 15 14.0 6 1.9 10.9 6 1.6 12.2 6 1.9 11.7 6 2.1 13.1 6 1.6 11.3 6 1.4PTCH1, TC 15 24.8 6 3.4 57.9 6 2.8 21.5 6 1.9 55.1 6 3.2 23.1 6 2.6 56.9 6 2.3**

a Percentage of area occupied by specific grains; values are least squares means 6 pooled SEM.b Granulosa (GC) and theca (TC) cells from small antral follicles; diameter¼ 1–5 mm.c Number of cows/treatment; number of follicles/cow ranged from two to five.d Averages of combined values from control and Twinner cows.** Means differ between estrogen-inactive (E-I) and estrogen-active (E-A) follicles; P , 0.01.

FIG. 3. Effects of estrogen status and day of estrous cycle on relativeabundance of PTCH1 mRNA determined by in situ hybridization ingranulosa and theca cells of small antral follicles (estrogen status by dayfor theca cells; P¼ 0.02; experiment 1). Because the genotypic effect wasnot significant, data were combined for Twinner and control cows; valuesare means 6 pooled SEM of in situ localized gene expression expressed asthe proportion of the area occupied by specific grains. a–cP , 0.01; d–fP ,0.05.

FIG. 4. Gene expression for IHH quantified by real-time RT-PCR inbovine granulosa cells from cows genetically not selected (control, n¼12)or selected (Twinner, n ¼ 15) for multiple ovulations (experiment 2).Because the genotypic effect was not significant, data were combined forTwinner and control cows. Follicles were collected between D3 and D4(D3) or D5 and D6 (D5) of an estrous cycle and classified as healthy (E-A;E2:P4 . 1) or atretic (E-I, E2:P4 , 1). Means expressed as relativeabundance to 18S rRNA. a,bMeans (6 pooled SEM) without a commonsuperscript differ (P , 0.05).

HEDGEHOG SYSTEM IN TWINNER CATTLE

5 Article 79

Dow

nloaded from w

ww

.biolreprod.org.

with BMP4 reduced (P , 0.05) PTCH1 abundance by 49%(from 2.41 to 1.24 6 0.1 relative abundance), whereastreatment with GDF9 (2.42 6 0.1 relative abundance) oractivin (2.19 6 0.9 relative abundance) was without effect (P. 0.10). In experiment 9, abundance of PTCH1 mRNA wascompared in large-follicle TC that were uncultured (freshlyisolated), cultured for 48 h in 10% FCS or cultured for anadditional 48 h in serum-free medium. Relative abundance ofPTCH1 mRNA decreased (P , 0.05) from 13.9 6 1.7 in freshuncultured TC to 4.2 6 0.4 after 48 h culture in 10% FCS andthen decreased further (P , 0.05) to 2.6 6 0.4 after anadditional 48 h culture in serum-free medium.

DISCUSSION

Expressions of IHH mRNA and PTCH1 mRNA werelocalized primarily within GC and TC, respectively, of bovineovarian follicles and were correlated positively with GCCYP19A1 mRNA expression, and PTCH1 mRNA was reducedin TC of antral follicles of Twinner vs. control cows.Moreover, expression of IHH mRNA increased in E-A (i.e.,

TABLE 2. Hedgehog system gene expression in granulosa (GC) and thecacells (TC) from control and Twinner cows (experiment 2).

Main effect Group No. GC IHH mRNA*� TC PTCH1 mRNA*

Genotype Control 46 991 59.4Twinner 85 1381 43.9

Day of cycle 3 66 1347 36.7a

5 65 1025 66.6b

Follicle status E-I 83 1011 38.7E-A 48 1361 64.6

SEM 320 13.4

* GC IHH and TC PTCH1 mRNA were measured using real-time RT-PCR.Abundance of mRNA is expressed as relative abundance per 18S rRNA;values are least squares means 6 pooled SEM. Follicle size rangedbetween 5 and 17 mm.� A day of cycle by follicle status existed (P , 0.05).a,b Within main effect, means without a common superscript differ (P ,0.05).

FIG. 5. Gene expression for PTCH1 in theca cells from cows geneticallynot selected (control, n ¼ 12) or selected (Twinner, n ¼ 15) cows formultiple ovulations (experiment 2). Follicles were collected between D3and D4 (D3) or D5 and D6 (D5) of an estrous cycle and classified ashealthy (E-A, E2:P4 . 1) or atretic (E-I, E2:P4 , 1), and RNA wasextracted. PTCH1 mRNA in theca cells was quantified using real-time RT-PCR and means expressed as relative abundance to 18S rRNA. A) Datawere averaged across genotypic groups and follicle classification andpresented as day of cycle effect on PTCH1 gene expression in theca cells.B) Data were averaged across day of cycle and presented as genotype byfollicle status effect on PTCH1 gene expression in theca cells. a,bWithineach panel, means without a common superscript differ (P , 0.05).

FIG. 6. Effects of E2 and IGF1 on IHH mRNA in small-follicle granulosacells treated in vitro. Granulosa cells from small (1–5 mm) follicles werecultured for 2 days in 10% FCS, washed in serum-free medium, and thentreated for 24 h with E2 (0 or 300 ng/ml) in the presence of FSH (30 ng/ml;experiment 3; A) or with IGF1 (0 or 100 ng/ml; experiment 4; B).a,b,cWithin a panel, means without a common superscript differ (P ,0.05).

AAD ET AL.

6 Article 79

Dow

nloaded from w

ww

.biolreprod.org.

healthy) follicles during follicle deviation between D3 and D5(experiment 2) and was greater in E-A vs. E-I (i.e., atretic)small follicles (experiment 1), and E2 treatment in vitroincreased IHH mRNA abundance in FSH-treated small-follicleGC (experiment 3). Together, these findings provide evidencethat link Hh signaling components to follicular status in cattle.

The positive association between Hh signaling and ovarianfollicular development agrees with previous studies in whichabundance of GC IHH mRNA was greater in small vs. largeantral follicles of cattle [19], concentrations of SHH infollicular fluid were greater in small vs. medium or largefollicles of pigs [57], and SHH treatment in vitro increasedmurine follicle diameter above controls [7] and induced bovineTC proliferation [19]. In addition, E-A follicles on D5 of theestrous cycle had levels of IHH mRNA and E2 that weregreater than in E-I follicles on D5 and in E-A follicles on D3(experiment 2), whereas expression of Ihh and Dhh was lost

rapidly in atretic follicles of mice [1], further supporting thenotion that Hh affects cell differentiation and survival in bothmammals [1, 8, 12, 19] and Drosophila [15, 58].

Because SHH increases cell proliferation and androstenedi-one production in cultured bovine TC [19], possible stimula-tory effects of the Hh system on E2 production in the bovineovary are likely mediated via increased proliferation of GCand/or TC within a follicle, increased stimulation of andro-stenedione production by TC (and its subsequent stimulation ofaromatase), or both. Based on results of the present study,increased E2 production may provide positive feedbackregulation to further increase IHH mRNA. Because differencesin ovarian follicle numbers between Twinner and control cattleare already present at early stages of follicular development[27, 35], further research is needed to identify the possible roleof Hh signaling in the development of the bovine ovary duringembryogenesis and to determine genotypic effects on germ cellproliferation.

In antral follicles (experiment 2), TC PTCH1 mRNA levelsincreased at deviation (D5) but were less in E-A follicles of

FIG. 7. Hormonal effects on IHH and DHH mRNA quantified by real-time RT-PCR in small (A) and large (B) follicle granulosa cells treated invitro (experiment 5). Granulosa cells isolated from small (1–5 mm) andlarge (>8 mm) follicles, cultured for 2 days in 10% FCS, washed in serum-free medium, and then treated for 24 h with IGF1 (30 ng/ml) and either noadditional hormones (control; Con), cortisol (Cort; 300 ng/ml), PGE2 (300ng/ml), SHH (500 ng/ml), WNT3A (300 ng/ml; WNT3), or ANG (300 ng/ml). a,b,cFor IHH mRNA, means without a common superscript differ (P ,0.05).

FIG. 8. Effects of IGF1 and LH on PTCH1 mRNA expression in thecacells from small (3–6 mm; A) or large (.8 mm; B) follicles (experiment 6).Cells were cultured for 2 days in 10% FCS, washed in serum-free medium,and treated for 24 h with IGF1 (0 or 100 ng/ml) and/or LH (0 or 30 ng/ml).a,b,cWithin a panel, means without a common superscript differ (P ,0.05).

HEDGEHOG SYSTEM IN TWINNER CATTLE

7 Article 79

Dow

nloaded from w

ww

.biolreprod.org.

Twinner vs. control cows. Like IHH mRNA, the expression ofPTCH1 mRNA was greater in TC of E-A vs. E-I small antralfollicles (experiment 1), and TC PTCH1 mRNA was positivelycorrelated with GC CYP19A1 mRNA. Consistent with thisPTCH1-CYP19A1 correlation, we observed that E2 treatmentincreased PTCH1 mRNA abundance in IGF1-treated large-follicle TC. Ptch1 gene expression was detected previously inTC of antral follicles evaluated at different sizes and stages ofthe mouse estrous cycle [59], and, like Ihh, expression of Ptch1mRNA was lost rapidly in morphologically atretic follicles ofmice [1]. In addition, PTCH1 mRNA abundance was greater inlarge vs. small bovine follicles [19]. Although SHH treatmentin vitro increased PTCH1 mRNA in bovine TC [19] and GlimRNA in mouse GC [7], differences in GC IHH mRNAcannot fully explain the difference in TC PTCH1 mRNAbetween E-A follicles of Twinner and control cows since GCIHH mRNA abundance was similar between control andTwinner cows. The differential expression of PTCH1 in controlvs. Twinner cows of the present study does indicate a possiblegenetic difference in the regulation of PTCH1 and thus apossible involvement of PTCH1 and the Hh system in theselection of two or more rather than one dominant follicle inTwinner cows. Because IGF1 treatment decreased PTCH1mRNA in cultured TC from small and large follicles (presentstudy) and Twinner cows have greater IGF1 concentrationsthan control cows [25, 28], perhaps IGF1 is suppressing thePTCH1 gene expression during growth of the dominantfollicle. Thus, exposure of follicles in Twinner cows to greaterconcentrations of IGF1 may have both directly and indirectly(via reduced IHH) reduced TC PTCH1 mRNA. In support ofthe idea that IHH secreted by GC may be needed for continuedPTCH1 mRNA expression in TC, we found in experiment 9that PTCH1 mRNA decreased with time in TC culturedwithout GC (i.e., presumed source of IHH).

Information regarding hormonal regulation of the Hh systemin the mammalian ovary is very limited. Wijegerde et al. [1]observed that expression of IHH and DHH mRNA in GC and,subsequently, that of PTCH1 mRNA in TC was lost quickly inpreovulatory follicles after mice were treated with hCG. Inaddition, Ren et al. [59] showed that Ihh mRNA in GC andPtch1 mRNA in ovarian tissue enriched for theca interstitialcells decreased within 4 h following hCG injection in mice,suggesting possible hormonal control of Ihh and Ptch1 duringfollicular differentiation and luteinization. Consistent withhCG-induced decreases in ovarian Ptch1 mRNA in rodents[1, 59], we observed that LH treatment in vitro decreasedPTCH1 mRNA abundance by 41% in TC from small bovinefollicles in the absence of IGF1, whereas in the presence ofIGF1, LH increased PTCH1 mRNA in TC from small follicles.However, LH had no effect on PTCH1 mRNA in TC of largefollicles, suggesting that as bovine follicles develop, the Hhsystem response to LH changes. We also found that BMP4 (butnot GDF9 or activin) decreased PTCH1 mRNA levels in small-follicle TC treated with IGF1 and LH, indicating that specificTGFb family proteins may regulate the Hh system in the ovaryas reported for other tissues [33, 34]. Further research will berequired to elucidate the regulation of the Hh system by IGF1,E2, LH, and the TGFb family proteins.

Consistent with a previous finding [19], we found that incultured small-follicle GC, IGF1 decreased IHH mRNA butthat the decrease was diminished in the presence of E2.Moreover, E2 increased IHH mRNA abundance in thepresence of FSH, but E2 alone, similar to FSH alone [19],did not affect abundance of IHH mRNA, indicating that IHHgene expression is controlled by IGF1, E2, and FSH.Interestingly, in GC from both small and large follicles, we

found that WNT3A stimulated IHH mRNA abundance several-fold, suggesting the WNT signaling system may be another keyregulator of the Hh system in the ovary. During fetaldevelopment in mice, the canonical WNT pathway directlyregulates Ihh transcription in chondrocytes [60]. In addition,we observed that 2 days of treatment with cortisol increasedIHH mRNA by 1.9-fold in GC. Another study supporting ourfinding that glucocorticoids may regulate IHH mRNAexpression found that mouse metanephroi cultured for 6 daysin the presence of 14 lM cortisol exhibited a 4.68-fold increasein Ihh mRNA [61]. Further research will be required toelucidate the timing, hormonal control, and role of IHH andPTCH1 in regulating various stages of follicular growth.

Assessment of DHH mRNA expression indicated that DHHmRNA in GC from small and large follicles was not regulatedby any of the hormones tested in experiment 5. Differences inthe regulation of expression of IHH and DHH mRNA are notsurprising since IHH is located on chromosome 2 and DHH onchromosome 5 in cattle. Differential expression of DHH andIHH mRNA is not without precedence. In previous studies, GCDhh mRNA was expressed earlier in primary mouse follicledevelopment than was GC Ihh mRNA [1], and rat pancreas cellexpression of Ihh was decreased by 50% after alloxantreatment in vitro with no concomitant change in Dhhexpression [62].

In conclusion, findings of the present study suggest that anassociation exists between modulation of the Hh pathway andselection of the dominant follicle(s). The observed patterns ofIHH and PTCH1 mRNA in GC and TC further indicate apotential paracrine role for the Hh signaling pathway in bovineovarian folliculogenesis, which may result in enhancingaromatase activity in early developing follicles. In addition,differences in TC PTCH1 mRNA abundance in Twinner vs.control cows indicate a genetic control of the Hh system inmammals, but further research is needed to verify thissuggestion. We hypothesize that an intrafollicular increase ofIGF1 in Twinner cows leads to the regulation of the Hh systemvia directly inhibiting the production of both GC IHH mRNAand TC PTCH1 mRNA. Whether the interaction between theHh system and the IGF system is a key contributor to theselection of two codominant follicles in Twinner cows willrequire further investigation.

ACKNOWLEDGMENT

The authors thank Jennifer Dentis, Morgan Totty, and John Evans fortheir laboratory assistance; the OSU Microarray Core Facility and OSURecombinant DNA/Protein Resource Facility for use of equipment; Dr.A.F. Parlow, National Hormone & Pituitary Program (Torrance, CA) forpurified LH and FSH; Dr. A.J. Hsueh for the recombinant rat GDF9; andCreekstone Farms (Arkansas City, KS) for their generous donations ofbovine ovaries.

REFERENCES

1. Wijgerde M, Ooms M, Hoogerbrugge JW, Grootegoed JA. Hedgehogsignaling in mouse ovary: Indian hedgehog and desert hedgehog fromgranulosa cells induce target gene expression in developing theca cells.Endocrinology 2005; 146:3558–3566.

2. Nilsson E, Skinner MK. Cellular interactions that control primordialfollicle development and folliculogenesis. J Soc Gynecol Invest 2001; 8:S17–S20.

3. Ingham PW, McMahon AP. Hedgehog signaling in animal development:paradigms and principles. Genes Dev 2001; 15:3059–3087.

4. Tasouri E, Tucker KL. Primary cilia and organogenesis: is Hedgehog theonly sculptor? Cell Tissue Res 2011; 345:21–40.

5. Pangas SA. Growth factors in ovarian development. Semin Reprod Med2007; 25:225–234.

6. Ingham PW. Localized hedgehog activity controls spatial limits of

AAD ET AL.

8 Article 79

Dow

nloaded from w

ww

.biolreprod.org.

wingless transcription in the Drosophila embryo. Nature 1993; 366:560–562.

7. Russell MC, Cowan RG, Harman RM, Walker AL, Quirk SM. Thehedgehog signaling pathway in the mouse ovary. Biol Reprod 2007; 77:226–236.

8. Cohen MM Jr. Hedgehog signaling update. Am J Med Genet A 2010;152A:1875–1914.

9. Theunissen JW, de Sauvage FJ. Paracrine Hedgehog signaling in cancer.Cancer Res 2009; 69:6007–6010.

10. Huang CC, Yao HH. Diverse functions of Hedgehog signaling information and physiology of steroidogenic organs. Mol Reprod Dev 2010;77:489–496.

11. Krishnan V, Ma Y, Moseley J, Geiser A, Friant S, Frolik C. Bone anaboliceffects of sonic/indian hedgehog are mediated by BMP-2/4-dependentpathways in the neonatal rat metatarsal model. Endocrinology 2001; 142:940–947.

12. Mimeault M, Batra SK. Frequent deregulations in the hedgehog signalingnetwork and cross-talks with the epidermal growth factor receptor pathwayinvolved in cancer progression and targeted therapies. Pharmacol Rev2010; 62:497–524.

13. Vortkamp A, Lee K, Lanske B, Segre GV, Kronenberg HM, Tabin CJ.Regulation of rate of cartilage differentiation by Indian hedgehog andPTH-related protein. Science 1996; 273:613–622.

14. Deckelbaum RA, Chan G, Miao D, Goltzman D, Karaplis AC. Ihhenhances differentiation of CFK-2 chondrocytic cells and antagonizesPTHrP-mediated activation of PKA. J Cell Sci 2002; 115:3015–3025.

15. Zhang Y, Kalderon D. Regulation of cell proliferation and patterning inDrosophila oogenesis by Hedgehog signaling. Development 2000; 127:2165–2176.

16. King FJ, Szakmary A, Cox DN, Lin H. Yb modulates the divisions of bothgermline and somatic stem cells through piwi- and hh-mediatedmechanisms in the Drosophila ovary. Mol Cell 2001; 7:497–508.

17. Ingham PW, Placzek M. Orchestrating ontogenesis: variations on a themeby sonic hedgehog. Nat Rev Genet 2006; 7:841–850.

18. Ingham PW, Kim HR. Hedgehog signalling and the specification ofmuscle cell identity in the zebrafish embryo. Exp Cell Res 2005; 306:336–342.

19. Spicer LJ, Sudo S, Aad PY, Wang LS, Chun SY, Ben-Shlomo I, Klein C,Hsueh AJ. The hedgehog-patched signaling pathway and function in themammalian ovary: a novel role for hedgehog proteins in stimulatingproliferation and steroidogenesis of theca cells. Reproduction 2009; 138:329–339.

20. Lee K, Jeong J, Kwak I, Yu CT, Lanske B, Soegiarto DW, Toftgard R,Tsai MJ, Tsai S, Lydon JP, DeMayo FJ. Indian hedgehog is a majormediator of progesterone signaling in the mouse uterus. Nat Genet 2006;38:1204–1209.

21. Takamoto N, Zhao B, Tsai SY, DeMayo FJ. Identification of Indianhedgehog as a progesterone-responsive gene in the murine uterus. MolEndocrinol 2002; 16:2338–2348.

22. Matsumoto H, Zhao X, Das SK, Hogan BL, Dey SK. Indian hedgehog as aprogesterone-responsive factor mediating epithelial-mesenchymal interac-tions in the mouse uterus. Dev Biol 2002; 245:280–290.

23. Khatua A, Wang X, Ding T, Zhang Q, Reese J, DeMayo FJ, Paria BC.Indian hedgehog, but not histidine decarboxylase or amphiregulin, is aprogesterone-regulated uterine gene in hamsters. Endocrinology 2006;147:4079–4092.

24. Cushman RA, Hedgpeth VS, Echternkamp SE, Britt JH. Evaluation ofnumbers of microscopic and macroscopic follicles in cattle selected fortwinning. J Anim Sci 2000; 78:1564–1567.

25. Echternkamp SE, Roberts AJ, Lunstra DD, Wise T, Spicer LJ. Ovarianfollicular development in cattle selected for twin ovulations and births. JAnim Sci 2004; 82:459–471.

26. Echternkamp SE, Cushman RA, Allan MF, Thallman RM, Gregory KE.Effects of ovulation rate and fetal number on fertility in twin-producingcattle. J Anim Sci 2007; 85:3228–3238.

27. Echternkamp SE, Gregory KE, Dickerson GE, Cundiff LV, Koch RM,Van Vleck LD. Twinning in cattle: II. Genetic and environmental effectson ovulation rate in puberal heifers and postpartum cows and the effects ofovulation rate on embryonic survival. J Anim Sci 1990; 68:1877–1888.

28. Echternkamp SE, Spicer LJ, Gregory KE, Canning SF, Hammond JM.Concentrations of insulin-like growth factor-I in blood and ovarianfollicular fluid of cattle selected for twins. Biol Reprod 1990; 43:8–14.

29. Spicer LJ, Alpizar E, Echternkamp SE. Effects of insulin, insulin-likegrowth factor I, and gonadotropins on bovine granulosa cell proliferation,progesterone production, estradiol production, and(or) insulin-like growthfactor I production in vitro. J Anim Sci 1993; 71:1232–1241.

30. Spicer LJ, Echternkamp SE. Ovarian follicular growth, function andturnover in cattle: a review. J Anim Sci 1986; 62:428–451.

31. Spicer LJ, Echternkamp SE. The ovarian insulin and insulin-like growthfactor system with an emphasis on domestic animals. Domest AnimEndocrinol 1995; 12:223–245.

32. Spicer LJ, Santiago CA, Davidson TR, Bridges TS, Chamberlain CS.Follicular fluid concentrations of free insulin-like growth factor (IGF)-Iduring follicular development in mares. Domest Anim Endocrinol 2005;29:573–581.

33. Katoh Y, Katoh M. Hedgehog signaling pathway and gastrointestinal stemcell signaling network (review). Int J Mol Med 2006; 18:1019–1023.

34. Katoh M, Katoh M. Transcriptional regulation of WNT2B based on thebalance of Hedgehog, Notch, BMP and WNT signals. Int J Oncol 2009;34:1411–1415.

35. Gregory KE, Echternkamp SE, Dickerson GE, Cundiff LV, Koch RM,Van Vleck LD. Twinning in cattle: I. Foundation animals and genetic andenvironmental effects on twinning rate. J Anim Sci 1990; 68:1867–1876.

36. Adams GP, Jaiswal R, Singh J, Malhi P. Progress in understanding ovarianfollicular dynamics in cattle. Theriogenology 2008; 69:72–80.

37. Echternkamp SE, Aad PY, Eborn DR, Spicer LJ. Increased abundance ofaromatase and follicle stimulating hormone receptor mRNA and decreasedinsulin-like growth factor-2 receptor mRNA in small ovarian follicles ofcattle selected for twin births. J Anim Sci 2012; 90:2193–2200.

38. Murdoch WJ, Dailey RA, Inskeep EK. Preovulatory changes prostaglan-dins E2 and F2 alpha in ovine follicles. J Anim Sci 1981; 53:192–205.

39. Spicer LJ, Enright WJ. Concentrations of insulin-like growth factor I andsteroids in follicular fluid of preovulatory bovine ovarian follicles: effectof daily injections of a growth hormone-releasing factor analog and(or)thyrotropin-releasing hormone. J Anim Sci 1991; 69:1133–1139.

40. Stewart RE, Spicer LJ, Hamilton TD, Keefer BE, Dawson LJ, Morgan GL,Echternkamp SE. Levels of insulin-like growth factor (IGF) bindingproteins, luteinizing hormone and IGF-I receptors, and steroids indominant follicles during the first follicular wave in cattle exhibitingregular estrous cycles. Endocrinology 1996; 137:2842–2850.

41. Ireland JJ, Roche JF. Development of nonovulatory antral follicles inheifers: changes in steroids in follicular fluid and receptors forgonadotropins. Endocrinology 1983; 112:150–156.

42. Langhout DJ, Spicer LJ, Geisert RD. Development of a culture system forbovine granulosa cells: effects of growth hormone, estradiol, andgonadotropins on cell proliferation, steroidogenesis, and protein synthesis.J Anim Sci 1991; 69:3321–3334.

43. Spicer LJ, Aad PY, Allen DT, Mazerbourg S, Payne AH, Hsueh AJ.Growth differentiation factor 9 (GDF9) stimulates proliferation andinhibits steroidogenesis by bovine theca cells: influence of follicle sizeon responses to GDF9. Biol Reprod 2008; 78:243–253.

44. Spicer LJ, Chamberlain CS. Influence of cortisol on insulin- and insulin-like growth factor 1 (IGF-1)-induced steroid production and on IGF-1receptors in cultured bovine granulosa cells and thecal cells. Endocrine1998; 9:153–161.

45. Spicer LJ, Aad PY. Insulin-like growth factor (IGF) 2 stimulatessteroidogenesis and mitosis of bovine granulosa cells through the IGF1receptor: role of follicle-stimulating hormone and IGF2 receptor. BiolReprod 2007; 77:18–27.

46. Grado-Ahuir JA, Aad PY, Spicer LJ. New insights into the pathogenesis ofcystic follicles in cattle: microarray analysis of gene expression ingranulosa cells. J Anim Sci 2011; 89:1769–1986.

47. Lee HS, Lee I-S, Kang T-C, Jeong GB, Chang S-I. Angiogenin is involvedin morphological changes and angiogenesis in the ovary. BiochemBiophys Res Comm 1999; 257:182–186.

48. Lapointe E, Boerboom D. WNT signaling and the regulation of ovariansteroidogenesis. Front Biosci (Schol Ed) 2011; 3:276–285.

49. Xu Z, Garverick HA, Smith GW, Smith MF, Hamilton SA, YoungquistRS. Expression of messenger ribonucleic acid encoding cytochrome P450side-chain cleavage, cytochrome p450 17 alpha-hydroxylase, andcytochrome P450 aromatase in bovine follicles during the first follicularwave. Endocrinology 1995; 136:981–989.

50. Santiago CA, Voge JL, Aad PY, Allen DT, Stein DR, Malayer JR, SpicerLJ. Pregnancy-associated plasma protein-A and insulin-like growth factorbinding protein mRNAs in granulosa cells of dominant and subordinatefollicles of preovulatory cattle. Domest Anim Endocrinol 2005; 28:46–63.

51. Voge JL, Santiago CA, Aad PY, Goad DW, Malayer JR, Spicer LJ.Quantification of insulin-like growth factor binding protein mRNA usingreal-time PCR in bovine granulosa and theca cells: effect of estradiol,insulin, and gonadotropins. Domest Anim Endocrinol 2004; 26:241–258.

52. Spicer LJ, Aad PY, Allen D, Mazerbourg S, Hsueh AJ. Growthdifferentiation factor-9 has divergent effects on proliferation and

HEDGEHOG SYSTEM IN TWINNER CATTLE

9 Article 79

Dow

nloaded from w

ww

.biolreprod.org.

steroidogenesis of bovine granulosa cells. J Endocrinology 2006; 189:329–339.

53. Livak KJ, Schmittgen TD. Analysis of relative gene expression data usingreal-time quantitative PCR and the 2(-Delta Delta C(T)) method. Methods2001; 25:402–408.

54. Grubbs FE. Sample criteria for testing outlying observations. Ann MathStat 1950; 21:27–58.

55. Ott L. An introduction to statistical methods and data analysis. NorthScituate, MA: Duxbury Press; 1977.

56. Ireland JJ, Roche JF. Growth and differentiation of large antral folliclesafter spontaneous luteolysis in heifers: changes in concentration ofhormones in follicular fluid and specific binding of gonadotropins tofollicles. J Anim Sci 1983; 57:157–167.

57. Nguyen NT, Lin DP, Yen SY, Tseng JK, Chuang JF, Chen BY, Lin TA,Chang HH, Ju JC. Sonic hedgehog promotes porcine oocyte maturationand early embryo development. Reprod Fertil Dev 2009; 21:805–815.

58. Forbes AJ, Lin H, Ingham PW, Spradling AC. Hedgehog is required for

the proliferation and specification of ovarian somatic cells prior to egg

chamber formation in Drosophila. Development 1996; 122:1125–1135.

59. Ren Y, Cowan RG, Harman RM, Quirk SM. Dominant activation of the

hedgehog signaling pathway in the ovary alters theca development and

prevents ovulation. Mol Endocrinol 2009; 23:711–723.

60. Spater D, Hill TP, O’Sullivan RJ, Gruber M, Conner DA, Hartmann C.

Wnt9a signaling is required for joint integrity and regulation of Ihh during

chondrogenesis. Development 2006; 133:3039–3049.

61. Chan SK, Riley PR, Price KL, McElduff F, Winyard PJ, Welham SJ,

Woolf AS, Long DA. Corticosteroid-induced kidney dysmorphogenesis is

associated with deregulated expression of known cystogenic molecules, as

well as Indian hedgehog. Am J Physiol Renal Physiol 2010; 298:

F346–F356.

62. Mfopou JK, Baeyens L, Bouwens L. Hedgehog signals inhibit postnatal

beta cell neogenesis from adult rat exocrine pancreas in vitro. Diabetologia

2012; 55:1024–1034.

AAD ET AL.

10 Article 79

Dow

nloaded from w

ww

.biolreprod.org.