Reproductive, Physiological, and Molecular Outcomes in Female Mice Deficient in Dhh and Ihh

Abstract

Ovarian development requires coordinate communications among oocytes, granulosa cells, and theca cells. Two Hedgehog (Hh) pathway ligands, Desert hedgehog (Dhh) and Indian hedgehog (Ihh), are produced by the granulosa cells and work together to regulate theca cell specification and development. Mice lacking both Dhh and Ihh had loss of normal ovarian function, which raised the question of which biological actions are specifically controlled by each ligand during folliculogenesis. By comparing the reproductive fitness, hormonal profiles, and ovarian transcriptomes among control, Dhh single-knockout (KO), Ihh KO, and Dhh/Ihh double-knockout (DKO) mice, we examined the specific roles of Dhh and Ihh in these processes. Dhh/Ihh DKO female mice were infertile because of a lack of theca cells and their steroid product androgen. Although Dhh and Ihh KO mice were fertile with normal folliculogenesis, they had decreased androgen production and alterations in their ovarian transcriptomes. Absence of Ihh led to aberrant steroidogenesis and elevated inflammation responses, which were not found in Dhh KO mouse ovaries, implicating that IHH has a greater impact than DHH on the activation of the Hh signaling pathway in the ovary. Our findings provide insight into not only how the Hh pathway influences folliculogenesis but also the distinct and overlapping roles of Dhh and Ihh in supporting ovarian development.

Loss of Dhh and/or Ihh in the ovary results in distinct physiological and molecular aberrations in female mouse reproduction.

The Hedgehog (Hh) signaling pathway is fundamental to development, controlling growth, survival, and fate determination in multiple cell lineages, as well as patterning of embryonic structures. Loss of Hh functions results in defects in organ morphogenesis, whereas aberrant activation of Hh signaling is linked to cancer development in adult organs (1, 2). In mammals, three Hh ligands have been identified: Sonic hedgehog (Shh), Desert hedgehog (Dhh), and Indian hedgehog (Ihh) (2, 3). In the absence of the ligands, the Hh receptor Patched homolog 1 or Patched homolog 2 inhibits the activity of another membrane-bound receptor Smoothened (SMO). Binding of Hh to Patched homolog suppresses its activity, releasing the inhibition on SMO. SMO then transduces the Hh signals to the cytoplasm, activates zinc-finger transcription factors [glioma-associated oncogene homolog (GLI)1, GLI2, and GLI3 in mammals], and ultimately controls the expression of Hh target genes (2, 4). In addition, SMO may function as a G protein–coupled receptor to stimulate biological processes in vivo, independent of the activation of GLI1 or GLI2 (5). Hh ligands have distinct tissue-specific expression patterns but may also be coexpressed in the same tissue with partially overlapping functions (2). For example, Shh, Dhh, and Ihh are predominantly expressed in the adrenal, testis, and endochondral bones, respectively (2, 6), whereas Shh and Ihh are simultaneously expressed in organs such as embryonic heart (7), stomach (8), gut (8, 9), bladder (10) and prostate (11, 12).

The Hh pathway is also important for gonadal development. DHH derived from Sertoli cells in the fetal testis regulates the specification and appearance of androgen-producing fetal Leydig cells (13). Downstream transcription factors implicated in this process include Gli1 and Gli2 (13, 14). In the mouse ovary, Dhh and Ihh are expressed in granulosa cells of the developing follicles (15, 16), whereas their downstream targets Patched homolog 1 (Ptch1) and Gli1, are localized specifically in the mesenchymal stromal cells surrounding the follicles (6, 16). Loss of both Dhh and Ihh in the ovary caused defects in theca-cell differentiation, leading to follicular arrest at preantral stages accompanied by the absence of corpora lutea (CL) (6). On the contrary, aberrant activation of Hh signaling in the ovary led to ectopic appearance of fetal Leydig cells during fetal life (17) or defects in smooth-muscle cell development and ovulation failure in adulthood (18). It is apparent that balanced Hh signaling is critical for interstitial/theca cell differentiation in the fetal ovary.

The fact that two Hh ligands are expressed in the granulosa cells of growing follicles raises the question of whether the two Hh ligands have distinct functions in the ovary. Female mice deficient for Dhh are fertile (19), which implies Ihh may have a redundant role during folliculogenesis. In this study, we investigated the individual roles of Dhh and Ihh in the mouse ovary by examining fertility, histological, hormonal, and transcriptional changes in the ovaries of control, Dhh knockout (KO), Ihh KO, and Dhh/Ihh double-KO (DKO) mice.

Materials and Methods

Animals

Dhh^+/− (catalog no. 002784) and Ihh^+/− (catalog no. 004290) mice were purchased from the Jackson Laboratory (Bar Harbor, ME). Sf1-Cre (20) and Ihh^{floxed/floxed} mice (21) were provided by the late Dr. Keith Parker at UT Southwestern Medical Center and Dr. Francesco DeMayo at Baylor College of Medicine, respectively. The genotypes of the ovaries used in this study were as follows: control, Sf1-Cre; Ihh^f/+ (where f indicates floxed); Dhh^+/−; Dhh KO, Sf1-Cre; Ihh^f/+; Dhh^−/−; Ihh KO, Sf1-Cre; Ihh^f/-; Dhh^+/−; DKO, Sf1-Cre; Ihh^f/-; Dhh^−/−. All animal procedures were approved by the National Institute of Environmental Health Sciences Animal Care and Use Committee and are in compliance with National Institute of Environmental Health Sciences–approved animal study proposals. All experiments were performed on at least three animals for each genotype.

Fertility study

Ten-week-old female mice were paired with proven fertile CD-1 male mice for 3 months. The breeding pairs were checked for the birth date and size of litters. The pups were immediately euthanized after birth when the litter size was documented.

Superovulation study

Adult female mice of various genotypes (6 to 15 weeks old) were injected IP with pregnant mare serum gonadotropin (PMSG, 5 IU; Calbiochem/MilliporeSigma, Burlington, MA) followed by an IP injection of human chorionic gonadotropin (hCG) 48 hours later (5 IU; Calbiochem). The animals were euthanized 14 hours after the hCG injection. Body weight and ovarian weight were recorded. Ovulated oocytes were collected from the ampulla of the oviduct and counted.

Histological analysis and immunohistochemistry

Ovaries from the female mice were collected at age 2 months and fixed in 4% paraformaldehyde in PBS at 4°C overnight. Vaginal smears were performed at every tissue collection and DKO mice never exhibited an estrous phase. These samples were dehydrated through an ethanol gradient, embedded in paraffin wax, and serially sectioned at 6-μm thickness. Sections at 60-μm intervals (every 10th section) were stained for systematic histology analyses.

For immunofluorescence, the sections were dewaxed and rehydrated in a series of alcohol to PBS. The slides were pretreated in citric acid–based Antigen Unmasking Solution (catalog no. H-3300; Vector Laboratories, Burlingame, CA) for 20 minutes in the microwave (power level, 10), followed by incubation with 5% normal donkey serum in PBS for 1 hour. The sections were incubated with primary antibodies against FOXL2 (1:500; catalog no. NB100-1277; RRID: AB_2106188; Novus Biologicals, Littleton, CO), HSD3β (1:500; catalog no. K0607; RRID: AB_2722746; Cosmo Bio, Carlsbad, CA), and αSMA (1:500; catalog no. ab5694; RRID: AB_91982; Abcam, Cambridge, MA) (6) in PBS-Triton X-100 solution with 5% normal donkey serum at 4°C overnight. The sections were then washed and incubated in the appropriate secondary antibodies (1:250; catalog nos. A11055 and A10042; RRIDs: AB_2534102 and AB_2534017; Invitrogen, Carlsbad, CA) before mounting in Vector Mount with 4′,6-diamidino-2-phenylindole (Vector Laboratories). Slides were viewed under a Leica DMI4000 confocal microscope (Leica Microsystems, Buffalo Grove, IL). For histological analysis, sections were stained with hematoxylin and eosin and scanned using an Aperio ScanScope XT Scanner (Aperio Technologies/Leica Microsystems, Buffalo Grove, IL) for digital image analysis.

Hormone assays

Serum samples were collected, aliquoted, and stored at −80°C. For dehydroepiandrosterone and testosterone, serum samples were assayed using the Steroid Hormone Panel kit (catalog no. N45CB-1; Meso Scale Discovery, Gaithersburg, MD) according to manufacturer’s protocols (6). Testosterone standard is not supplied with the Steroid Hormone Panel kit; it was purchased separately from Steraloids (catalog no. A6950; Newport, RI) and used in the range of 16 ng/mL to 0.1 ng/mL. For progesterone, FSH, and LH measurement, the samples were analyzed by the Center for Research in Reproduction Ligand Assay and Analysis Core at University of Virginia using Progesterone ELISA (ALPCO, Salem, NH) and Millipore Pituitary Panel Multiplex kits at 25 uL per singlet determination (FSH and LH), respectively.

Gene expression analysis

Total RNA was isolated from ovaries using the PicoPure RNA isolation kit (Arcturus, Mountain View, CA) according to the manufacturer’s protocol. The cDNA preparation was synthesized from 500 ng of RNA using random hexamers and the Superscript II cDNA synthesis system (Invitrogen), following manufacturer’s instruction. Gene expression was analyzed by RT-PCR using the Bio-Rad CFX96 Real-Time PCR Detection system (6). Taqman assays and SYBR Green–based detection were used to examine the fold changes of the transcripts. The following Taqman probes were used: Dhh (Mm01310203_m1), Ihh (Mm00439613_m1), Gli1 (Mm00494654_m1), Ptch1 (Mm00436026_m1), hedgehog interacting protein (Hhip; Mm0469580_m1), chicken ovalbumin upstream promoter–transcription factor II (Nr2f2; Mm00772789_m1), cholesterol side-chain cleavage enzyme (Cyp11a1; Mm00490735_m1), prostaglandin F receptor (Ptgfr; Mm00436055_m1), hydroxysteroid (17-β) dehydrogenase 7 (Hsd17b7; Mm00501703_m1), luteinizing hormone/choriogonadotropin receptor (Lhcgr; Mm00442931_m1), steroid 5-α-reductase 3 (Srd5a3; Mm04243702_m1), cytochrome P450 17A1 (Cyp17a1; Mm00484040_m1), steroidogenic factor 1 (Nr5a1; Mm00446826_m1), thrombospondin type I motifs-1 (Adamts1; Mm01344169_m1), tachykinin 1 (Tac1; Mm01166996_m1), chemokine (C-C motif) ligand 5 (Ccl5; Mm01302427_m1), interleukin 1 receptorlike 1 (Il1rl1; Mm00516117_m1), and 18S rRNA (Mm03928990_g1). Fold changes in gene expression were determined by quantitation of cDNA from target (mutant) samples relative to a calibrator sample (control). All real-time PCR analyses were performed in duplicate, and the results were analyzed from a minimum of five biological replicates for each experiment. The relative fold change of transcript was calculated using the mathematical model of Pfaffl (22) and was normalized to 18S rRNA (Taqman probes) as an endogenous reference.

Microarray and gene ontology analysis

Transcriptome analysis was conducted using Affymetrix Mouse Genome 430 2.0 GeneChip arrays (Affymetrix, Santa Clara, CA). Total RNA (50 ng) was amplified with the WT-Ovation Pico RNA Amplification System and labeled with biotin, following the Encore Biotin Module (Nugen, San Carlos, CA) protocols. Amplified biotin–antisense RNAs (5 μg) were fragmented and hybridized to each array for 18 hours at 45°C in a rotating hybridization oven. Array slides were stained with streptavidin/phycoerythrin using a double-antibody staining procedure and then washed for antibody amplification according to the GeneChip Hybridization, Wash, and Stain Kit and user manual, following protocol FS450-0004. Arrays were scanned in an Affymetrix Scanner 3000 and data were obtained using the GeneChip Command Console Software, version 3.2 (Affymetrix). The microarray raw data were analyzed using the Partek Genomics Suite software (St. Louis, MO).

The arrays for single KO ovaries (Dhh KO and Ihh KO) and DKO ovaries were carried out at different times with the same control samples. Batch effects between the two arrays were removed using Partek Genomic Suite software. Two-way ANOVA was performed to determine the statistical significance (P < 0.05) between the means of the groups. The heat map was created using a one-way ANOVA (P value with a false discovery rate < 0.05 and fold change >2) comparing the batch corrected robust multi-array average normalized log2 average intensities for Dhh KO, Ihh KO, and DKO vs control. The heat map was divided into four clusters according to the polygenetic relevance of the probes. The clusters of gene lists were then subjected to gene ontology (GO) analysis using DAVID Bioinformatics Resources 6.7 (https://david-d.ncifcrf.gov/). Microarray data have been deposited in Gene Expression Omnibus under accession code GSE109824.

Statistical analysis

For ovarian weight and oocyte counts after superovulation, an unpaired t test with equal SD was used. Fertility test, hormonal assay, quantitative PCR and raw log2 intensity data were analyzed using Prism, version 6 (GraphPad Software, La Jolla, CA) by ANOVA with the Tukey multiple comparison test. Values are expressed as mean ± SEM. A minimum of three biological replicates were used for each experiment. The specific number of biological replicates for each experiment is listed in the figure legends.

Results

Ovaries deficient in either Dhh or Ihh alone exhibited normal folliculogenesis, whereas inactivation of both genes resulted in severe follicular defects

To investigate the contributions of Dhh and/or Ihh to ovarian development, we generated Dhh and Ihh single KOs and DKOs, in which Dhh and/or Ihh were ablated from Sf1-positive gonadal somatic cells (6). Dhh KO and Ihh KO ovaries exhibited normal folliculogenesis, evident by different stages of follicle development as well as the formation of corpus lutea or CL (Fig. 1A–1C). On the other hand, DKO ovaries lacked CL, and follicle development was arrested at the preantral stage (Fig. 1D). Immunostaining for α-SMA, a marker for smooth-muscle cells in the theca layer (6, 18), revealed normal formation of the theca layer in Dhh KO and Ihh KO ovaries, but a failure in theca layer formation in DKO ovaries (Fig. 1E–1H). Although Dhh KO and Ihh KO ovaries appeared morphologically similar, they exhibited some differences in steroidogenesis. The steroidogenic cell marker 3βHSD was normally found in both stroma areas (arrowheads in Fig. 1I–1J) and theca layer (arrows in Fig. 1M and 1N) of control and Dhh KO ovaries. In the Ihh KO ovaries, 3βHSD was decreased in the stroma area (arrowhead in Fig. 1K) and was absent from the theca cell layer (Fig. 1O). In the DKO ovaries, 3βHSD expression was abolished from the stromal area (arrowhead in Fig. 1L) and the theca layer (Fig. 1P), a phenotype more severe than that of the Ihh KO (Fig. 1K and 1O). Unexpectedly, 3βHSD expression was elevated in the granulosa cells of growing follicles in the Dhh KO, Ihh KO, and DKO ovaries, compared with the minimal 3βHSD staining observed in the control ovaries (Fig. 1I and 1P). Although Sf1-cre is known to be active in the pituitary and adrenal, and our Dhh model is a global one, we do not expect that knocking out Ihh or Dhh will affect pituitary or adrenal functions, based on the fact that Dhh and Ihh were either undetectable or expressed at basal level in these tissues (Supplemental Fig. 1) (23, 24).

Figure 1.

Consequences of inactivation of Dhh and/or Ihh in the mouse ovary. (A–D) Histological analysis of ovarian sections from control (n = 10), Dhh KO (n = 8), Ihh KO (n = 6), and DKO (n = 7) ovaries. *Indicate the presence of CL. (E–H) Immunofluorescence in control, Dhh KO, Ihh KO, and DKO ovaries: red, smooth-muscle cell marker α-SMA; green, granulosa cell marker FOXL2. (I–L) Immunofluorescence in control, Dhh KO, Ihh KO, and DKO ovaries: red, steroidogenic marker 3βHSD. (I–L) Arrowhead indicates the ovarian stroma areas. (M, N) Arrows indicate the theca layer. Scale bar, 200 μm. (M–P) Higher magnification of the outlined areas in (I–L). Scale bars, 50 μm. DAPI, 4′,6-diamidino-2-phenylindole; GC, granulosa cell compartment; H&E, hematoxylin and eosin; IT, interstitial-theca layer.

DKO female mice were sterile, whereas Dhh KO and Ihh KO female mice were fertile

To assess the fertility of single- and double-KO female mice, we performed a 12-week fertility study beginning when the mice were 10 weeks old. The total number of pups and the average litter size from Dhh KO and Ihh KO female mice were not significantly different from those of control female mice (Fig. 2A and 2B). In contrast, DKO female mice did not produce any pups during this time (Fig. 2A and 2B). In addition, DKO mice had smaller ovaries. The ovary-to-body weight ratio was significantly deceased in DKO mice compared with control female mice (Fig. 2C). This outcome was likely due to the loss of the theca cell layer and the absence of CL, which occurred as a result of defective follicular development (6). The observation that both Dhh KO and Ihh KO female mice were fertile indicates the DKO phenotype requires the loss of both Dhh and Ihh. To determine if the fertility defect of the DKO female mice was due to a lack of ovulatory response, superovulation was performed in adult females. After pregnant mare serum gonadotropin and hCG treatments, 36.9 ± 3.1 ovulated eggs were found in the ampulla of control female mice. Conversely, the ovulatory response was significantly reduced in the DKO mice (3.5 ± 1.4 ovulated eggs; (Fig. 2D).

Figure 2.

Reproductive parameters of female mice. (A) Total pup number per breeding pair, (B) average litter size per breeding pair, (C) ovary–to–body weight ratio, and (D) oocyte number after superovulation procedure in control (white), Dhh KO (light gray), Ihh KO (dark gray), and DKO (black) female mice. The data were analyzed with ANOVA with Tukey multiple comparison test (P < 0.05; a < b) for (A) and (B) (sample sizes: control, n = 4; Dhh KO, n = 3; Ihh KO, n = 3; DKO, n = 3), and Student t test for (C) and (D) (P < 0.001; samples sizes: control, n = 11; DKO, n = 14). Error bars represent SEM.

Female mice deficient in Dhh and/or Ihh displayed differences in serum hormone levels

To assess the roles of Dhh and Ihh in ovarian hormone production, we measured serum hormone levels in control, Dhh KO, Ihh KO, and DKO female mice at 2 months of age. Serum levels of dehydroepiandrosterone, testosterone, and progesterone were progressively lower in the order of Dhh KO, Ihh KO, and DKO female mice compared with that of control female mice (Fig. 3A–3C). FSH level was not significantly different among the four groups (Fig. 3D). On the other hand, LH concentration was elevated in the DKO female mice but did not differ in Dhh KO or Ihh KO female mice compared with control mice (Fig. 3E).

Figure 3.

Hormone measurements for serum levels of (A) dehydroepiandrosterone (DHEA), (B) testosterone, (C) progesterone, (D) FSH, and (E) LH in control (white; n = 10), Dhh KO (light gray; n = 8), Ihh KO (dark gray; n = 6), and DKO (black; n = 7) female mice. The data were analyzed with ANOVA with Tukey multiple comparison test (P < 0.05; a < b). (F) Log2 intensity values from microarray analyses for genes Ar, Esr1, Esr2, Pgr, and Cyp19a1 in control (white; n = 8), Dhh KO (light gray; n = 5), Ihh KO (dark gray; n = 6), and DKO (black; n = 12) ovaries are presented. Error bars represent SEM of the log2 values. The data were analyzed with ANOVA with Tukey multiple comparison test, but no statistical significance was detected.

We also examined expression of the genes associated with the steroid hormones. The levels of Ar, Esr1, Esr2, Pgr, and Cyp19a1 were not significantly different among the KOs compared with that of the control mice (Fig. 3F).

Distinct transcriptomic changes in Dhh KO, Ihh KO, and DKO ovaries

We performed microarray analyses on control, Dhh KO, Ihh KO, and DKO ovaries from 2-month-old mice with the goal to pinpoint possible biological processes or signaling pathways controlled by each Hh ligand. A heat map of 1663 differentially expressed probes is shown in Fig. 4A. GO analysis revealed that Dhh and Ihh were implicated, redundantly or differentially, in at least four physiological functions and/or signaling pathways, including Hh signaling pathway (n = 232 probes), steroidogenesis (n = 414 probes), ovulation regulation (n = 931 probes), and inflammatory response (n = 86 probes; Fig. 4). Differentially expressed probes were subdivided into two categories (Fig. 4A): The first category (n = 1577 genes) included probes that were downregulated in the absence of Dhh and Ihh (or stimulated by DHH and IHH under normal circumstance), and the second category (n = 86 genes) contained probes that were upregulated in the absence of Dhh and Ihh (or inhibited by DHH and IHH under normal circumstances). The selection of the four physiological functions and/or signaling pathways was based on differentially regulated biological processes (GO analysis) shown in Fig. 4B in the context of ovarian development and their physiological relevance to our study. For example, GO-generated biological processes, including cell proliferation, smoothened signaling, sex differentiation, cell-cell signaling, vasculature development and extracellular region, all included genes that are components of the Hh signaling pathway.

Figure 4.

Distinct transcriptomic changes in Dhh KO, Ihh KO, and DKO ovaries. (A) Heat map of 1663 differentially expressed probes (P value with a false discovery rate <0.05; fold change >2) in Dhh KO, Ihh KO, and DKO ovaries compared with those of the control mice. The enriched GO terms for the four most influenced pathways/physiological processes are shown. Red indicates high expression; green indicates low expression. (B) Detailed GO analyses in relation to the signaling pathway/ovarian function categories (i.e., Hh signaling, steroidogenesis, ovulation regulation, and inflammatory response) shown in (A). The percentage after each biological process denotes the percentage of the microarray-detected genes in the GO lists. P < 0.05 [presented by –log(P value)] was used to identify differentiated biological processes. C1, category 1; C2, category 2.

Genes downregulated in the absence of Dhh and/or Ihh

The genes most highly downregulated in the absence of Dhh and Ihh belonged to three physiological processes or pathways: (1) Hh signaling, (2) steroidogenesis, and (3) ovulation regulation (Fig. 4). Using real-time quantitative PCR, we first investigated genes in the Hh signaling pathway, including Dhh, Ihh, the Hh target Gli1 (25), the Hh receptor Ptch1 (25), Nr2f2 (26), and Hhip (27, 28). Compared with the control ovary, Dhh expression was decreased in the Ihh KO ovaries, whereas expression of Ihh was not affected in the Dhh KO ovary. Neither Dhh nor Ihh showed a compensatory increase in mRNA level in the reciprocal KOs (Fig. 5). Expression of Gli1 was not changed in the Dhh KO whereas its expression was significantly decrease in DKO ovaries and moderately reduced in Ihh KO ovaries (Fig. 5). Expression of Ptch1, Nr2f2, and Hhip shared a similar pattern: no changes in Dhh KO ovaries and a significant decrease in Ihh KO and DKO ovaries (Fig. 5). These observations were consistent with the heat map analysis, in which a consensus of gene downregulation was observed in Ihh KO and DKO ovaries for these genes, whereas their expression profiles were largely unaffected in Dhh KO ovaries compared with those of control mice (Fig. 4A). These results also imply that IHH has a greater impact than DHH in the activation of the Hh pathway in the ovary.

Figure 5.

Differentially expressed genes that are involved in the Hh signaling pathway. Quantitative PCR analyses for representative genes (Dhh, Ihh, Gli1, Ptch1 Nr2f2, and Hhip) implicated in Hh signaling in control (white; n = 8), Dhh KO (light gray; n = 5), Ihh KO (dark gray; n = 6), and DKO (black; n = 12) ovaries. Error bars represent SEM. The data were analyzed with ANOVA with Tukey multiple comparison test (P < 0.05; a < b < c).

In addition to the Hh signaling pathway, several genes associated with steroidogenesis were differentially regulated in Dhh KO, Ihh KO, and DKO ovaries (Fig. 4A and Fig. 6A). Examples of these genes include Cyp11a1, Ptgfr, Hsd17b7, Lhcgr, Srd5a3, and Cyp17a1 (Fig. 6A). Cyp11a1 catalyzes the conversion of cholesterol to pregnenolone, the first reaction in the synthesis of all steroid hormones (29–31). Ptgfr mediates the process of luteolysis and regulates progesterone synthesis during the formation of CL (32–34). Hsd17b7 is critical in converting less-active forms of estrogen and androgen to more active forms (35–37). Lhcgr is expressed in granulosa cells and theca cells, and is important for the expression of steroidogenic acute regulatory protein (38). Srd5a3 (39, 40) and Cyp17a1 (41) are critical genes implicated in the production of androgens in the theca cells. Expression of Cyp11a1 was unaffected in Dhh KO and Ihh KO ovaries, whereas its expression was significantly downregulated in DKO ovaries, as previously reported (6) (Fig. 6A). Ptgfr and Hsd17b7 exhibited similar expression patterns: unchanged in Dhh KO and DKO ovaries, but increased in Ihh KO ovaries compared with those of control mice. Lhcgr expression followed the pattern of Ptgfr and Hsd17b7 despite the statistical insignificance. In a similar manner, Srd5a3 shared the expression trend with Cyp17a1: decreased in Ihh KO and DKO ovaries but unchanged in Dhh KO ovaries.

Figure 6.

Quantitative PCR analyses of differentially expressed genes that are involved in (A) steroidogenesis, (B) ovulation regulation, and (C) inflammatory response in control (white; n = 8), Dhh KO (light gray; n = 5), Ihh KO (dark gray; n = 6), and DKO (black; n = 12) ovaries. Error bars represent SEM. The data were analyzed with ANOVA with Tukey multiple comparison test (P < 0.05; a < b < c).

In addition to genes implicated in ovarian steroidogenesis, multiple genes important for ovulation were significantly downregulated in DKO ovaries (Fig. 4A). Here we focused on genes with an apparent link to ovulation regulation (Fig. 6B). Research has shown that ovaries deficient in either Nr5a1 (42) or Adamts1 (43) exhibit defects in ovulation and CL formation. Expression of Nr5a1 and Adamts1 expression patterns were not significantly changed among control, Ihh KO, and DKO ovaries, but were remarkably increased in the Dhh KO ovaries. These observations suggest that DHH might have a regulatory role for Nr5a1 and Adamts1, despite that Dhh KO female mice exhibited equivalent ovulatory capacity to that of the control animals (Figs. 2, 3, and 6A).

Genes upregulated in the absence of Dhh and/or Ihh

We found 86 genes involved in the innate immune response that were upregulated in Ihh KO and DKO ovaries compared with control and Dhh KO samples (Fig. 4A). Examples of genes include Tac1 (44, 45), Ccl5, and Il1rl1 (46, 47) (Fig. 6C). Expression of Tac1 went from mild to a significant increase in Dhh KO, Ihh KO, and DKO ovaries compared with those of the control mice. Ccl5 and Il1rl1 exhibited a similar pattern: unchanged in Dhh KO and Ihh KO ovaries, but increased significantly in the DKO ovaries. The upregulation of inflammatory genes in the DKO ovaries was consistent with the hemorrhagic follicular defects previously reported (6).

Discussion

This study reveals the distinct roles that Dhh and Ihh play in the mouse ovary. When both Dhh and Ihh are inactivated in the ovary, folliculogenesis is compromised, resulting in complete infertility. Loss of either Dhh or Ihh alone does not affect follicular development or fertility, but they lead to distinct, altered gene expression profiles. The histology, hormone production, and ovarian transcriptome analyses included in this report strongly imply that Dhh and Ihh are differentially involved in many aspects of ovarian development, such as the activation of Hh signaling pathway, steroidogenesis, and inflammation responses.

Redundant roles of Dhh and Ihh in ovarian development

Although Dhh KO and Ihh KO female mice were fertile with normal folliculogenesis, they showed decreased androgen production and alterations in ovarian transcriptomes. These observations suggest that Dhh and Ihh perform partially redundant roles in the ovary, and both ligands are required for full ovarian function. Hh ligands can be expressed individually in tissues such as testis (Dhh), bone (Ihh), and adrenal (Shh) (2, 6), yet corporately present in other organs such as embryonic heart (7), stomach (8), gut (8, 9), bladder (10), and prostate (11, 12). For example, both Shh and Ihh are expressed in, and required for the development of, the stomach, gut, and heart (7, 8, 48). To our knowledge, the ovary is the first organ known to produce both Dhh and Ihh and to require both Hh ligands for its function.

IHH has a greater impact than DHH on the activation of the Hh pathway in the ovary

Although Dhh and Ihh are produced in the ovary, IHH appears to be the major Hh ligand responsible for activation of the Hh pathway. Our findings are consistent with those of previous studies in which Hh responsive cell line–based assays and mouse tissue explant assays were used to assess the activation of the Hh pathway in response to full-length Hh ligands. Together, these findings demonstrate that IHH is a strong activator of the Hh signaling, whereas DHH has lower activity (49, 50). It is known that Hh signaling in the ovary is critical for theca progenitor cells to gain their androgen-producing capacity. If this is the case, then one would assume that IHH, being the predominant Hh ligand, is more influential to theca cell differentiation. Indeed, Ihh KO female mice, compared with their Dhh KO counterparts, had more pronounced changes in ovarian histology such as a loss of 3βHSD in the theca layer, and a mild reduction in androgen production. The finding of Ihh being a more important ovarian Hh is also supported by a study in bovines, where Ihh, rather than Dhh, is much more responsive to hormonal effects in the cow ovary (51). Compared with Ihh, Dhh appears to contribute minimally to activation of the Hh signaling pathway. The minor role of Dhh in ovarian development is in stark contrast to its profound role in the testis, where absence of Dhh is associated with Leydig (13) and peritubular myoid cell defects (52), a loss of mature sperm, and complete infertility in adult testes (19). One explanation for the surprising differences between the testis and ovarian phenotypes of Dhh KO mice is that DHH is the sole Hh ligand produced in the testis and, therefore, bears greater responsibility for Hh signaling-related biological processes. In the ovary, however, two Hh ligands, DHH and IHH, are produced, and IHH supersedes DHH to mediate long-distance Hh signaling in this tissue (49).

Ihh is implicated in the regulation of steroidogenesis in the ovary

By comparing the ovarian transcriptomes among different KO models, we found that expression of Srd5a3 and Cyp17a1 was decreased in the Ihh KO and DKO ovaries. Srd5a3 and Cyp17a1 were both involved in the production of androgens in the theca cells (39–41). Interestingly, other steroidogenesis-related genes were upregulated in Ihh KO ovaries, meaning that under normal circumstances, IHH suppresses expression of these steroidogenic genes. This observation seems contradictory to the existing knowledge that Ihh promotes the differentiation of steroidogenic theca cells (6, 53). However, unlike Cyp17a1 that is specifically expressed in the theca layer, many of the steroidogenesis-related genes, such as Cyp11a1 (54), Ptgfr (55, 56), Hsd17b7 (57), and Lhcgr (58) are also present in the granulosa cells of large antral follicles. It is possible that the upregulation of steroidogenesis-related genes in Ihh KO ovaries is the result of elevated steroidogenesis in granulosa cells. In fact, granulosa cells of small follicles in the Ihh KO ovaries appear to have aberrantly high 3βHSD expression. This phenomenon of aberrant 3βHSD expression in granulosa cells becomes even more evident in DKO ovaries. In line with these observations, previous studies have shown that growth differentiation factor 9 (GDF9), an upstream regulator of Ihh in the ovary (6), is capable of suppressing the steroidogenesis of granulosa cells in culture (59). Furthermore, in the Gdf9 KO ovaries, granulosa cells of small follicles become highly steroidogenic (60). These results suggest a role of GDF9 in preventing premature differentiation of granulosa cells in small follicles. Although there is an apparent link among GDF9, IHH, and granulosa cell steroidogenesis, the mechanisms by which GDF9/IHH signaling regulates granulosa cell development are not clear. Because the expression of Hh downstream targets such as Gli1, Gli2, and Ptch1 are restricted within the theca layer, one would presume that GDF9/IHH signaling regulates the differentiation program of granulosa cells indirectly through their actions on theca cells. Indeed, several theca cell–derived factors, such as KGF, HGF (61), and TGF-β (62), have been shown to regulate granulosa cell development using similar mechanisms (63).

It is not yet clear how DHH and IHH exert different physiological functions during ovarian development. Studies examining the posttranslational processing of Hh ligands have revealed that under cell culture conditions, DHH undergoes little autoprocessing, and mostly induces cell contact–mediated juxtacrine signaling. By contrast, IHH is subject to high levels of autoprocessing and secretion, thereby inducing not only localized but also long-distance signaling (49). Consistent with this finding, addition of cholesterol moieties during autoprocessing is required for maximal activity of the Hh ligands (64). These results suggest that posttranslational processing of Hh ligands (likely through the addition of cholesterol moieties) alters Hh functions, which may explain the differential roles of DHH and IHH in ovarian development. Although DHH and IHH act through the membrane-bound regulator SMO, previous findings indicate that SMO is capable of translating different levels of Hh into distinct intracellular responses (65). One recent finding shows that SMO is able to signal through G-proteins to stimulate cell proliferation in mammary glands in vivo (5). In addition to SMO, transduction of Hh signals is also mediated through the activities of the GLI proteins (i.e., GLI1, GLI2, and GLI3), which harbor repressor and activator functions (25). Cellular responses to the activities of GLI proteins are further complicated by differential promoter affinities as well as the involvement of region-specific cofactors (66). Unfortunately, published data on the direct links between GLI proteins and the target genes shown in this manuscript are very limited. The potential interactions between GLI proteins and the target genes are further complicated by the fact that Hh signaling intersects with WNT, Notch, BMP, and other regulatory networks (67). Thus, it is unclear whether the changes of target genes were due to direct interactions with the GLI proteins or resulted from other signaling networks. Although Gli3 is barely detectable, Gli1 and Gli2 are enriched in the theca cell layer of the adult ovary (6, 16). It is not clear whether Gli1 and Gli2 are redundantly expressed in the same cell types within the theca layer, or present in distinct cell populations. Nonetheless, the spatial and temporal distribution of GLI factors during ovarian development may act to refine redundant Hh signals into distinct cellular decisions.

In summary, we have identified a substantial number of genes that are differentially expressed during ovarian development in the absence of Dhh and/or Ihh. The possible redundant and distinct roles of Dhh and Ihh are involved in at least four unique signaling pathways and biological processes: Hh signaling, granulosa cell-steroidogenesis, ovulation regulation, and inflammation response. This information is important to dissect the differential roles of Hh signaling during ovarian development and will aid future studies concerning the developmental regulation of follicular development as well as the pathogenesis of ovarian diseases.

Glossary

Abbreviations:

Adamts1

thrombospondin type I motifs-1

CL

corpora lutea

Cyp11a1

cholesterol side-chain cleavage enzyme

Cyp17a1

cytochrome P450 17A1

Dhh

Desert hedgehog

DKO

double knockout

f

floxed

GDF9

growth differentiation factor 9

GLI

glioma-associated oncogene homolog

GO

gene ontology

hCG

human chorionic gonadotropin

Hh

hedgehog

Hhip

hedgehog interacting protein

Hsd17b7

hydroxysteroid (17-β) dehydrogenase 7

Ihh

Indian hedgehog

Il1rl1

interleukin 1 receptorlike 1

KO

knockout

Lhcgr

luteinizing hormone/choriogonadotropin receptor

Nr2f2

chicken ovalbumin upstream promoter-transcription factor II

Nr5a1

steroidogenic factor 1

Ptch1

Patched homolog 1

Ptgfr

prostaglandin F receptor

Shh

Sonic hedgehog

SMO

Smoothened

Srd5a3

steroid 5-α-reductase 3

Tac1

tachykinin 1