Comparative gene expression pro ling of mouse ovaries upon stimulation with native equine chorionic gonadotropin (eCG) and tethered recombinant-eCG


 Background: Equine chorionic gonadotropin (eCG) induces super-ovulation in laboratory animals. Notwithstanding its extensive usage, limited information is available regarding the differences between the in vivo effects of native eCG and recombinant eCG (rec-eCG). This study aimed to investigate the gene expression profiles of mouse ovaries upon stimulation with native eCG and rec-eCG produced from CHO-suspension (CHO-S) cells. eCG and rec-eCG were cloned and transfected into CHO-S cells and quantified. Thereafter, we determined the metabolic clearance rate (MCR) of native eCG and rec-eCG up to 24 h after intravenous administration through the tail vein and identified differentially expressed genes in both ovarian tissues, via quantitative real-time PCR (qRT-PCR) and immunohistochemistry (IHC).Results: Rec-eCG was markedly up-regulated initially after transfection and maintained until recovery on day 9. Oligosaccharide chains were substantially modified in rec-eCG protein produced from CHO-S cells and eliminated through PNGase F treatment. The MCR was slightly lower for rec-eCG than for eCG, and no significant difference was observed after 60 min. Notwithstanding their low concentration, rec-eCG and native eCG were detected in the blood at 24h post-injection. Microarray analysis of ovarian tissue revealed that 20 of 12,816 genes assessed therein were significantly up-regulated and 43 genes were down-regulated by >2-fold in the group receiving rec-eCG (63 [0.49%] differentially regulated genes in total). The microarray results were concurrent with and hence validated by those of RT-PCR, qRT-PCR, and IHC analyses.Conclusions: The present results indicate that rec-eCG can be adequately produced through a cell-based expression system through post-translational modification of eCG and can induce ovulation in vivo. These results provide novel insights into the molecular mechanisms underlying the up- or down-regulation of specific ovarian genes and the production of rec-eCG with enhanced biological activity in vivo.

. Moreover, eCG administration to cows is reportedly associated with an increase in their ovulation rate [16], particularly in early postpartum calves [17,18].
The glycosylation sites at amino acid residue 52 in the α-subunit of human FSH (hFSH) [19] and hCG [5] and residue 56 in eCG [7] are important for signal transduction when the cAMP response is impaired, and the binding activity of these hormones increased by 20-to 3-fold [20], concurrent with our previous ndings [2,6]. Thus, post-translational glycosylation of glycoprotein hormones plays a pivotal role in receptor-mediated signal transduction. N-and O-linked oligosaccharides at 56 of α-subunit of eCG and a C-terminal extension (residues 114-149) in the β-subunit were included in vectors expressing eCG to produce rec-eCG and to investigate the role of these regions in the biological activity of eCG.
High-throughput RNA sequencing and microarray analysis are useful during transcriptome pro ling and gene expression analysis [21,22]. A microarray contains thousands of millions of complementary DNA fragments or oligonucleotides that hybridize with speci c RNA molecules in a sample [22]. A recent study revealed differentially expressed genes (DEGs) upon RNA-seq using ovarian tissue of dairy goats upon repeated eCG treatment [23], indicating that three-time eCG treatment dysregulated several ovarian genes including glucagon, follistatin-related protein 3 (FSTL3), and aquaporin-3 (AQP3), thereby reducing reproductive function.
Thus far, we have attempted to assess the different roles of rec-eCG with respect to their attached oligosaccharides [2,24], glycosylation sites for LH-and FSH-like activity [2], tethered rec-eCGs [6], internalization of rat FSH and LH receptors by rec-eCG [25], and signal transduction through eel FSH and LH receptors by rec-eCG and native eCG [26]. Furthermore, we have analyzed the ovulation rates between native eCG and rec-eCGs in mice [27]. These results suggest that rec-eCGs are induced at markedly lower in nonfunctional oocytes upon deglycosylation. Approximately 20% of non-functional oocytes with native eCG and only 2% with the rec-eCGs have been determined thus far. Numerous studies have reported the effects of a combination of eCG and hCG on reproductive performance and estrous synchronization [13][14][15][16]. However, no studies have investigated the effects of native eCG and rec-eCG on gene regulation through RNA-based microarray analysis.
In the present study, we hypothesized that treatment of ovarian tissues with native eCG and rec-eCG results in different DEG pro les. We produced rec-eCGβ/α proteins in CHO-S cells, characterized their physiological function in vivo, and analyzed the difference in gene expression pro les through microarray analysis.

Results
Production of rec-eCG and western blot analysis eCG contains two N-linked glycosylation sites at amino acid positions 56 and 82 in the α-subunit of eCG. The β-subunit of eCG contains one N-linked glycosylation site at position 13 and approximately 11 Olinked glycosylation sites at the C-terminal region (Fig. 1). Thus, we constructed an expression vector encoding the tethered eCG mutant, which was linked with the α-subunit without the signal sequence at the C-terminal region of the β-subunit. Rec-eCG expression levels were markedly increased to 210±10.3 mIU/mL on day 1 after transfection. These levels were consistently maintained until day 9, being 212±12.7, 227±16.1, 230±15.6, and 202±7.8 mIU/mL at 3, 5, 7, and 9 days, respectively (Fig. 2a). Rec-eCG was e ciently secreted into the cell culture medium. eCG levels markedly increased initially upon transfection and were maintained until recovery on day 9.
Further, we analyzed the molecular weight of tethered rec-eCG. On western blot analysis, the approximate molecular weight of rec-eCG was 40-46 kDa (Fig. 2b). After deglycosylation with PNGase F, the molecular weights signi cantly decreased to approximately 30-36 kDa (Fig. 2b). The oligosaccharide chains were substantially modi ed post-translation in tethered rec-eCG, con rming the loss of the oligosaccharide chains upon PNGase F treatment.
Metabolic clearance rates (MCRs) of natural eCG and tethered rec-eCG in vivo To analyze the MCR, eCG was detected in both groups (~550 mIU/mL) in the serum at 1 h after injection, as shown in Fig. 3. Although the MCR was slightly lower in the rec-eCG-treated groups, no signi cant difference was observed between native-eCG and rec-eCG treatment after 60 min. Their concentration was low (~100 mIU/mL) until 24 h. These results indicate that rec-eCG produced herein had a normal MCR and induced ovulation, as previously described [27].
Comparison of ovarian gene expression pro les between groups treated with native eCG and rec-eCG Global gene expression pro les were analyzed in mouse ovarian tissue treated with native eCG and rec-eCG via microarray analysis. The ovarian tissues were harvested at 13 h upon combinational treatment (10 IU of eCG followed by 10 IU of hCG after 48 h). Gene expression levels were analyzed via microarray analysis with 12,816 gene probes. Genes showing a >2-fold difference in expression levels were identi ed in eight ovaries (native eCG: four, rec-eCG: four). Figure 4 shows the differences in gene expression pro les between the two samples.
Further, we analyzed the data to gain insight into the biological processes and functions of the DEGs. The distribution of the 63 DEGs (at least 2-fold) between ovaries treated with native eCG and rec-eCG and their distribution in different Gene Ontology (GO) categories were analyzed (Supplementary Material Fig. 1).
Gene expression analysis through quantitative reverse-transcription PCR (qRT-PCR) analysis To validate the results of microarray analysis, we performed RT-PCR and qRT-PCR analyses using speci c primers (Supplementary Material Table 1) for the 14 genes identi ed herein (Fig. 5a, b). Among the upregulated genes identi ed through microarray analysis of rec-eCG-treated ovaries, six genes, i.e., Tex19.2, Sectm1b, Ctsk, Gpnmb, Sectm1a, and Hsd17b1, were con rmed to be up-regulated through qRT-PCR (Fig.  5a). Among the down-regulated genes, eight genes, i.e., OVGP1, BC048546, Tmem68, Dcpp1, Prkg2, Edn2, Adamts1, and Akr1b7, were con rmed to be down-regulated by >2-fold in the rec-eCG-treated mouse ovarian tissue, of which seven, i.e., OVGP1, BC048546, Tmem68, Dcpp1, Edn2, Adamts1, and Akr1b7, were con rmed to be down-regulated via qRT-PCR analysis (Fig. 5b). Nonetheless, one gene, Prkg2, displayed no signi cant change in expression levels upon qRT-PCR analysis. The fold-change in the expression levels of these genes was consistent with the results of microarray analysis, con rming that the results of qRT-PCR analysis correlated with those of the microarray analysis.

Immunohistochemical analysis of ovarian tissue
To determine the cell types expressing four proteins (HSD17b1, ADAMTS1, Edn2, and OVGP1), immunohistochemical analysis was performed for the same ovarian tissues used for microarray analysis ( Fig. 6). Among the up-regulated genes in the rec-eCG-treated ovarian tissue, HSD17b1 was localized in the granulosa cells and theca folliculi. Among the down-regulated genes in rec-eCG-treated ovarian tissue, ADAMTS1, which is required for normal ovulation and is localized in the cumulus oocyte complex during the preovulatory stage, was also localized in granulosa cells. Edn2, which is transiently expressed in granulosa cells immediately prior to ovulatory follicle rupture, was also strongly expressed in the ovarian stroma of a native eCG-treated ovarian tissue. OVGP1, which improves the e ciency of in vitro fertilization and increases the number of fertilized eggs, was weakly expressed in the ovaries after ovulation. These results indicate that the expression of these four proteins was directly correlated with the time of ovulation in mice.

Discussion
This study examined the biological activity of tethered rec-eCG, containing N-and the O-linked oligosaccharide chains and their MCR in vivo. Furthermore, this study evaluated differential gene expression pro les in mouse ovaries stimulated with native eCG and rec-eCG in combination with hCG. The present study shows differences in the up-and down-regulated genes (>2-fold) in ovarian tissues treated with native eCG and rec-eCG.
Thus, far, we have expressed rec-eCG in only CHO-K1 cells and stable CHO-K1 cells under G418 selection [6,[25][26][27]. Hence, levels of secreted rec-eCG at 24 h post-transfection have remained unknown. However, supernatants of the culture media of CHO-S cells were recovered until 9 days after transfection. In the present study, single-chain rec-eCG was markedly up-regulated on day 1 after transfection in CHO-S cells. However, rec-eCG with a C-terminal deletion in the β-subunit was detected at a low concentration on day 1 and 3 post-transfection (data now shown). The present results indicate that the CTP region including approximately 13 O-linked oligosaccharides plays a pivotal role in the early secretion of eCG from cells into the supernatant medium after transfection.
Various studies have reported that rec-eCG proteins lead to the production and secretion of stable heterodimeric eCG in COS-7 cells [28] and infected Sf9 cells [29], with thermal stability similar to that of native pituitary LH [30]. Secreted single-chain eCG in COS-7 cells is detectable as a doublet of 46 and 44 kDa [12]. The present results show that the molecular weight of rec-eCG greatly decreases upon elimination of the N-linked oligosaccharide chains via PNGase F treatment, decreasing the molecular weight to approximately 30-36 kDa. Our results are consistent with those of other studies, suggesting that rec-eCG contained highly modi ed N-linked glycosylation sites in COS-7 cells and CHO-K1 cells posttranslation [12,27].
Furthermore, rec-eCG mutants deglycosylated through site-directed mutagenesis was markedly low in number (< 2.4%) in nonfunctional oocytes in comparison with native eCG (21.2%) [27]. These results suggest a speci c model for ovulation without displaying a long-half-life and to only induce functional oocytes in experimental animals despite using native eCG. Furthermore, the MCR of rec-eCG was somewhat lower than that of native eCG at 10-60 min after injection and was similarly maintained at 2 and 24 h. The present results suggest that the potency of rec-eCG can be assessed at only 10 IU, as previously described [27]. These MCR results suggest that rec-eCG derivatives can be attentively utilized for animal experiments.
Furthermore, we previously reported that rec-eCG exerts dual LH-and FSH-like activity in in vitro bioassays involving rat Leydig and granulosa cells, respectively [2,6]. Moreover, we previously reported that rec-eCG has both LH-and FSH-like activity in cells expressing rLH/CGR and rFSHR [25]. Nevertheless, no studies have examined differential gene expression in ovaries stimulated with native eCG and rec-eCG. We performed gene expression pro ling for ovarian tissue through microarray analysis after administration of native eCG or rec-eCG. We identi ed genes up-and down-regulated by >2-fold. The present results show that 63 genes were up-and down-regulated (0.49% of 12,816 genes in rec-eCGinjected ovaries). These changes in gene expression pro les directly render oocytes nonfunctional upon comparing native eCG-treated and rec-eCG-treated ovaries, suggesting that rec-eCGb/a derivatives used herein can cause slightly aberrant gene expression in the ovaries and produce functional oocytes without nonfunctional oocytes, in comparison with native eCG-treated ovaries. In the "biological process" category, the largest number of deregulated genes included signal transduction (16 proteins; Supplementary Material Figs. 1 and 2). In contrast, the largest number of genes (6 genes) among the 18 "molecular function" categories were present in "proteases." Seven genes were categorized as "molecular function unclassi ed." We assessed differences in the expression of ovary-speci c genes between groups treated with native eCG and rec-eCG through qRT-PCR analysis. The differences in gene expression were con rmed for six genes. These genes, Tex19.2, Sectm1b, Ctsk, Gpnmb, Sectm1a, and Hsd17b1, were speci cally over-expressed in rec-eCG-treated ovaries. Among the genes found to be down-regulated on microarray analysis, seven genes were con rmed to be down-regulated through qRT-PCR analysis. These differences should be further assessed through a systematic study.
Immunohistochemical analysis was conducted to determine the cell type responsible for protein expression in the ovaries. We rst con rmed that 17b-hydroxysteroid dehydrogenase type 1 (17b-HSD1), which catalyzes the conversion of estrone to estradiol, is primarily localized in ovarian granulosa cells after rec-eCGb/a injection. Our results are consistent with those of another study, showing that 17b-HSD1 is expressed in the placenta and ovarian granulosa cells [31]. Some studies have reported that ADAMTS-1 is induced in granulosa cells in preovulatory follicles after LH administration [32] and is important for follicular development and the maintenance of normal granulosa cell layers in follicles [33]. Endothelin-2 (Edn2), a potent vasoconstrictive peptide, is abundantly produced by preovulatory follicles during ovulation at the onset of CL formation [34]. Edn2 directly induces vascular endothelial growth factor in granulosa cells of the bovine ovary [35] and ovulation and CL formation are signi cantly impaired in Edn2-knockout mice [34].
Oviduct-speci c glycoprotein (OVGP1), also known as oviductin, is the major non-serum glycoprotein in the oviduct uid during fertilization and increases the number of fertilized eggs and promotes early embryonic development [36]. Furthermore, Edn2 and OVGP1 are primarily localized in the ovaries after native eCG administration. These results suggest that 17b-HSD1, ADAMTS-1, Edn2, and OVGP1 perform pivotal functions as ovulatory factors during ovulation in mice.

Conclusions
This study shows that rec-eCG produced from CHO-S cells has high biological activity in vivo. Although the MCR of rec-eCG is low soon after its administration, rec-eCG displays a wide range of biological activity including the induction of ovulation and oogenesis. The present results show that 63 ovarian genes were differentially expressed between native eCG-treated and rec-eCG-treated ovaries. Differential expression patterns of these genes were further con rmed through RT-PCR, qRT-PCR, and immunostaining analyses. The present results suggest that these differences may have resulted from the nature of the hormone, including oligosaccharides and folding. Further systematic analyses are required to investigate the role of these DEGs in ovulation. Therefore, rec-eCG derivatives can potentially be produced at high levels with high biological activity to induce oocytes in vivo.

Materials
The oligonucleotides used hereiny were synthesized by Genotech (Daejon, Korea). The restriction enzymes and the DNA ligation kit were purchased from Takara (Tokyo, Japan). The QIAprep-Spin plasmid kit was acquired from QIAGEN, Inc. (Hilden, Germany). The Lumi-Light western blot kit was purchased from Roche (Basel, Switzerland), and the pcDNA3 mammalian expression vector, FreeStyle CHO-S suspension cells, PNGase F, FreeStyle MAX transfection reagent, and TRIzol reagent were obtained from Invitrogen (Carlsbad, CA, USA). The PMSG ELISA kit was purchased from DRG International, Inc. Construction of tethered eCG cDNA encoding the tethered rec-eCGβ/α was inserted into the mammalian expression vector pcDNA3, as previously reported [6]. The same method was used to insert a myc tag (Glu-Gln-Lys-Leu-Ile-Ser-Glu-Glu-Asp-Leu) between the rst and second amino acid residues of the β-subunit of the mature eCG protein [27]. Plasmid DNA was then puri ed and sequenced in both directions through automated DNA sequencing to ensure correct inserts. The cloned expression vector of tethered eCGβ/α was designated as pcDNA3-eCGb/a, as previously reported [6]. A schematic representation for tethered eCG β/α is shown in Fig. 1.

Cell culture and generation of tethered rec-eCG
In CHO-S cells, the tethered eCG expression vector was transfected into CHO-S cells using the FreeStyle MAX reagent (Invitrogen; Carlsbad, CA, USA) transfection method, in accordance with manufacturer's instructions. Flasks were placed on an orbital shaking platform, rotating at 120-135 rpm at 37°C in a humidi ed atmosphere of 8% CO 2 in air. On transfection, the cell density was approximately 1.2-1.5 × 10 6 cells/mL. The plasmid DNA (260 ug) and a FreeStyle TM MAX Reagent complexes were gradually added to 200 mL of medium containing cells. Finally, culture media were sampled on day 9 after transfection and centrifuged to eliminate cell debris. The supernatant was sampled and stored at -20°C until the assay. The samples were concentrated using a Centricon lter or by freeze-drying and mixed with PBS.
Quanti cation of rec-eCG proteins rec-eCG protein was quanti ed with the PMSG ELISA kit (DRG Diagnostics; Mountain side, NJ, USA). Brie y, the PMSG standard and rec-eCG samples (100 µL) were dispensed into the wells of a plate coated with the antibody and incubated for 60 min at ambient temperature. After rinsing thrice, 100 µL of anti-PMSG antibody conjugated with horseradish peroxidase was added into each well and incubated for 60 min. The plate wells were rinsed ve times, and substrate solution (100 µL) was added and incubated for 30 min at ambient temperature. Finally, 50 µL of a stop solution was added and the absorbance was measured at 450 nm, using a microtiter plate reader Cytation TM 3 (BioTeK, Winooski, VT, USA). The average absorbance of each standard was plotted against its corresponding concentration in a linearlog graph. We determined the average absorbance of each sample to determine the corresponding PMSG value via simple interpolation through a standard curve. Finally, 1 IU was considered 100 ng in accordance with the conversion factor of the suggested assay protocol.

Detection of rec-eCGs via western blotting and enzymatic digestion of N-linked oligosaccharides
Concentrated sample media were subjected to SDS-PAGE (12.5% resolving gel) via the Laemmli method [37]. After SDS-PAGE, the proteins were electro-transferred to a nitrocellulose membrane for 2 h in a Mini Trans-Blot Electrophoretic Transfer cell. To eliminate all N-linked oligosaccharides, the rec-eCG sample was incubated for 24 h at 37°C with PNGase F [2 mL of the enzyme (2.5 U/mL) per 30 mL of sample+8 mL of 5´ reaction buffer]. The reaction was terminated by boiling for 10 min, and the samples were subjected to SDS-PAGE and the proteins were electro-transferred on to a membrane. After blocking the membrane with a 1% blocking reagent for 1 h, followed by probing with monoclonal anti-myc antibody (1: 5,000) for 2 h, the membrane was washed and probed with a secondary antibody (peroxidase-conjugated anti-mouse IgG antibody 37.5 μL/15 mL of the blocking solution) for 30 min. The membrane was then incubated for 5 min with 2 mL of the Lumi-Light substrate solution and X-ray lm was exposed to the membrane for 1-10 min.

Assessment of the MCR of eCGs
Each animal was intravenously administered 5 IU of native eCG or rec-eCG through the tail vein to determine the 50% dose for the induction of superovulation. Blood was sampled from the transorbital vein in heparinized microhematocrit tubes. Blood samples were obtained at 10 and 30 min and at 1, 2, and 24 h and centrifuged for 15 min at 5,000 rpm at 4°C, and plasma eCG concentrations were estimated using the PMSG ELISA kit (DRG Diagnostics).

Animals
The MCRs of native eCG and rec-eCG were determined in 8-week-old male B6D2F1 (C57BL6 ´ DBA/2) 12 mice. The female 16 mice (8-week-old B6D2F1; Oriental Bio, Gyeonggi, Korea) were superovulated by injection of 10 IU of native eCG and rec-eCG and then 10 IU hCG after 48 h. The ovarian tissues were sampled at 13 h after hCG administration. All mice were euthanized with carbon dioxide inhalation, and the ovarian tissues were collected at the end of study. All the mice were raised in an environment with the temperature of 23 ± 1 °C with regular 12 h light/dark cycle and allowed free access to feed and water. The animals were processed by the Animal Care and Use Committee procedure. The protocol was approved by the Committee on Ethics of Animal Experiments at the Hankyong National University (Approval ID: 2015-8).

Microarray analysis
Total RNA was extracted from ovaries, using TRIzol reagent, and puri ed using RNeasy columns in accordance with the manufacturers' protocols, as previously described [15].

1) Labeling and puri cation
Total RNA was ampli ed and puri ed using an Ambion Illumina RNA ampli cation kit (Ambion, Austin, TX, USA) in accordance with the manufacturer's instructions to obtain biotinylated cRNA. Brie y, 550 ng of total RNA was reverse-transcribed into cDNA with a T7 oligo(dT) primer. Second-strand cDNA was synthesized, transcribed in vitro, and labeled with biotin-NTP.

3) Raw data preparation and statistical analysis
Raw data were extracted using the software provided by the manufacturer (Illumina Genome Studio v.2009.2) and ltered using a detection p-value of <0.05 (a signal value higher than that of the background was necessary to set the detection p-value of <0.05). The selected gene signal value was logarithmically transformed and normalized to XYZ. Comparative analysis between two groups was conducted on the basis of the p-value evaluation, via the local-pooled-error test (adjusted Benjamini-Hochberg false discovery rate had to be<5%) and the fold-change. Biological ontology-based analysis was performed for the Panther database (http://www.pantherdb.org). Furthermore, genes whose expression levels differed by >2-fold were considered differentially expressed between the two groups.

Immunohistochemistry
Immunohistochemical staining of ovarian samples was performed using the Vectastain ABC kit (Vector Laboratories, Burlingame, CA, USA) in accordance with the manufacturer's instructions. The samples were xed in 10% neutral-buffered formalin at ambient temperature for 24 h and washed with PBS. Thereafter, the xed samples were rehydrated in graded ethanol (EtOH) solutions (3 min each in 100% 2´; 95% 1´; 70% 1´; and 50% 1´) and embedded in para n. Para n-embedded tissues were sectioned into 8µm-thick sections, which were then mounted onto Poly-l-lysine-coated slides. The slides were boiled in 10 mM sodium citrate for 10 min and chilled on ice for 20 min. Thereafter, they were washed with 3% hydrogen peroxide for 10 min and blocked for 1 h at ambient temperature. The slides were incubated with the primary antibody and then with an anti-rabbit IgG antibody (secondary antibody). Finally, the slides were immunostained using the ABC detection kit in accordance with the manufacturer's instructions and stained with DAB. The slides were examined under a Nikon Eclipse TE-2000-E confocal microscope (Tokyo, Japan).

Data and statistical analysis
Data are presented as mean ± SEM values. One-way ANOVA with Tukey's multiple-comparison test was conducted to compare the results between samples. In gures, the superscripts indicate signi cant differences from groups (p< 0.05).  Bold genes were adjusted to RT-PCR and qRT-PCR    The metabolic clearance rate (MCR) of native eCG and of rec-eCG / . Both eCGs were intravenously administrated at 5 IU through the tail vein. Blood samples were collected after 10 and 30 min and 1, 2, and 24 h. The samples were centrifuged at 5,000 rpm for 15 min at 4°C, and eCG concentrations in the serum were estimated using a PMSG ELISA kit. The levels of eCG were analyzed via sandwich ELISA in triplicate.

Figure 4
Hierarchical clustering of gene expression pro les in native eCG-treated and rec-eCG-treated ovarian tissues. The ovaries were excised from 8-week-old ICR female mice. The mice were induced to superovulate with 10 IU of native eCG or rec-eCG and then 10 IU of hCG after 48 h, and the ovulated oocytes were collected in an oviduct ampulla after 13 h. Thereafter, the ovaries were harvested after 13h and RNA was analyzed via microarray analysis. Gene expression levels were evaluated through microarray analysis with 12,816 gene probes. Genes showing >2-fold differences in expression were identi ed. The expression of 63 (0.49%) of 12,816 genes differed by at least 2-fold between native eCGtreated and rec-eCG-treated ovaries. Twenty of 63 genes were up-regulated in rec-eCG-treated ovaries and 43 genes were down-regulated  Localization of HSD17 1, ADAMTS1, EDN2, and OVGP1. The ovaries were induced to superovulate with 10 IU of either natural-eCG or rec-eCG / , followed by 10 IU of hCG after 48 h. Representative immunohistochemical analyses for HSD17 1, ADAMTS1, EDN2, and OVGP1 were conducted with antisera, and a goat anti-rabbit IgG antibody (secondary antibody). According to the microarray and qRT-PCR results, HSD17 1 was up-regulated in the rec-eCG-treated ovaries, while the other three proteins (ADAMTS1, EDN2, and OVGP1) were up-regulated in the native eCG-treated ovaries.
Immunohistochemistry was performed with a Vectastain ABC kit. Scale bar = 200 μm.