Hematopoetic Prostaglandin D Synthase: An ESR1-Dependent Oviductal Epithelial Cell Synthase

Abstract

Oviductal disease is a primary cause of infertility, a problem that largely stems from excessive inflammation of this key reproductive organ. Our poor understanding of the mechanisms regulating oviductal inflammation restricts our ability to diagnose, treat, and/or prevent oviductal disease. Using mice, our objective was to determine the spatial localization, regulatory mechanism, and functional attributes of a hypothesized regulator of oviductal inflammation, the hematopoietic form of prostaglandin D synthase (HPGDS). Immunohistochemistry revealed specific localization of HPGDS to the oviduct's epithelium. In the isthmus, expression of HPGDS was consistent. In the ampulla, expression of HPGDS appeared dependent upon stage of the estrous cycle. HPGDS was expressed in the epithelium of immature and cycling mice but not in the oviducts of estrogen receptor α knockouts. Two receptor subtypes bind PGD₂: PGD₂ receptor and G protein-coupled receptor 44. Expression of mRNA for Ptgdr was higher in the epithelial cells (EPI) than in the stroma (P < 0.05), whereas mRNA for Gpr44 was higher in the stroma than epithelium (P < 0.05). Treatment of human oviductal EPI with HQL-79, an inhibitor of HPGDS, decreased cell viability (P < 0.05). Treatment of mice with HQL-79 increased mRNA for chemokine (C-C motif) ligands 3, 4, and 19; chemokine (C-X-C motif) ligands 11 and 12; IL-13 and IL-17B; and TNF receptor superfamily, member 1b (P < 0.02 for each mRNA). Overall, these results suggest that HPGDS may play a role in the regulation of inflammation and EPI health within the oviduct.

Tubal or oviductal disease is a major reproductive concern in women (1, 2). Tubal dysfunction is comparable in etiology with ovulatory defects or endometriosis as an indication for the treatment of female infertility (3–5), and tubal ectopic pregnancies are the primary cause of mortality for a woman in her first trimester (6, 7). Excessive or chronic inflammation of the oviduct is the primary precursor to tubal disease (8–13), yet our understanding of inflammatory mechanisms in this organ, and hence our ability to diagnose, treat, and/or manage for tubal disease, remains poor.

In this article, we provide evidence suggesting that the hematopoietic form of prostaglandin (PG) D synthase (HPGDS) plays a pivotal role in the regulation of inflammation in the oviduct. This synthase acts downstream of PG-endoperoxide synthase 2 (cyclooxygenase-2) to catalyze the conversion of PGH₂ to PGD₂ (14–16). PGs are known to mediate inflammation (17), maintaining homeostasis of the epithelium in the lung (18, 19), digestive tract (20, 21), urinary (22, 23) and other reproductive (24, 25) systems. In addition, the involvement of PGs in epithelial-derived cancers (26–28) is well documented. Although PGs of the E and F series and their respective receptors are well established as regulators of many aspects of reproductive biology, HPGDS-derived PGD₂ has received little attention as a regulator of reproductive function and represents a potential target to improve our ability to specifically diagnose and/or treat oviductal inflammation and disease.

When originally cloned, levels of mRNA for Hpgds were found to be highly expressed in the oviduct of the rat (29). However, with the interests of those authors lying outside of reproductive biology, no comprehensive analysis of this synthase, or its end product, was reported for the oviduct. Interestingly, the spatial pattern of expression for HPGDS in the various tissues and organs of the body is unique. In their original study, Kanaoka et al. (29) conducted a broad distribution analysis by Northern blotting and observed relatively specific expression of Hpgds in samples of the spleen, followed by the oviduct. In that study, HPGDS was not detectable in the ovary and uterus, or in many of the nonreproductive tissues they evaluated.

With inflammation-induced oviductal epithelial cell (EPI) death known to precede tubal occlusion and infertility (8–13), strong and relatively tissue-specific localization of HPGDS to the oviduct and the knowledge that PG are known regulators of inflammation, we hypothesized that HPGDS-catalyzed PGD₂ may be acting as a key regulator of inflammation in this organ and hence performing a vital role in the maintenance of cellular homeostasis, patency, and function. Using mice, we report that HPGDS is localized specifically to the EPI of the oviduct, expressed before puberty, temporally regulated over the course of the estrous cycle, and dependent upon functional expression of the transcription factor estrogen receptor-α (ESR1). In addition, inhibition of HPGDS was found to decrease the viability of human oviductal EPI (hOEC) in vitro and increase the level of expression of multiple genes affecting inflammation within the oviducts of mice in vivo.

Together, our results suggest that HPGDS-catalyzed PGD₂ may be a regulator of oviductal inflammation and that the role of this synthase and PGD₂ in oviductal biology needs to be examined in greater detail. This synthase was identified within the oviduct over 10 yr ago (29), in such quantities that the oviduct was then used as a positive control in other studies (30, 31), yet HPGDS and PGD₂ had not been investigated with respect to the actual function of the oviduct itself. Identification of HPGDS as a key regulator of oviductal inflammation would have dramatic implications to the management of female reproductive health, an especially important finding when considering the lack of noninvasive treatment options for oviductal disease, as well as the financial and emotional burden posed on both women and their families by oviductal-based infertility (3–5, 32–34).

Materials and Methods

Animals

Animal procedures involved in these studies were approved by the University of Kentucky Animal Care and Use Committee and/or the Institutional Animal Care and Use Committee of Hanyang University. With the exception of experiments using transgenic mice, female CD1 mice were used throughout. For the transgenic analysis, mice with a global deletion of ESR1 [ESR1 knockout (ESR1KO)] were generated on a C57BL6 background using the cre/loxP approach, and female littermates that expressed ESR1 served as wild-type (WT) controls (35). Genotypes were confirmed by the analysis of genomic DNA extracted from ear punches using the Easy DNA kit (Invitrogen, Carlsbad CA) according to the manufacturer's directions, as described previously (35, 36). For the analysis of oviducts collected on defined days of the estrous cycle, females were staged by analysis of vaginal cytology, as described before (37). Briefly, vaginal smears were collected daily, at the same time each day, with 0.9% sodium chloride using a bent, blunted borosilicate glass pipette. Vaginal cytology was analyzed under a Motic AE21 inverted microscope (Motic Instruments, Richmond, British Columbia, Canada) and classified according to well-established morphological criteria (38, 39). Digital images were recorded for later reference.

Immunohistochemistry

The spatial localization of HPGDS in the oviduct was determined in regularly cycling CD1 females, as well as immature females treated with 5 IU of pregnant mare's serum gonadotropin (PMSG) (Sigma-Aldrich, St. Louis, MO) to induce follicular development and the production of estradiol. In addition, the dependence of HPGDS on ESR1 was evaluated in 10-wk-old ESR1KO and WT females, killed on diestrus. Diestrus was chosen for consistency among genotypes, because the ESR1KO mice do not display regular estrous cycles (40). Oviducts were retrieved from all mice and fixed for 4 h in Bouin's fixative (Sigma-Aldrich), then embedded in paraffin. Sections were cut to 5 μm, mounted on poly-l-lysine-coated glass slides, deparaffinized, and rehydrated. Antigen retrieval was performed by autoclaving for 10 min in 10 mm citrate buffer (pH 6.0). The slides were then incubated for 15 min in 3% H₂O₂ in PBS to remove endogenous peroxidase activity, blocked for 60 min in 3% BSA in PBS, and incubated overnight at 4 C with a primary (rat monoclonal) antibody against HPGDS (no. MBS601438; MyBioSource, San Diego, CA) at a dilution of 1:1000. After washing in PBS, the slides were then incubated for 60 min at room temperature in a 1:500 dilution of rabbit antirat IgG (Abcam, Cambridge, MA) in PBS with 1% BSA. After further rinses in PBS, the coloring reaction was conducted with ImmPACT diaminobenzidine-based peroxidase substrate for 10 min at room temperature. Sections were rinsed again in PBS, then dehydrated in a graded ethanol series, dipped in xylenes, and permanently mounted. Images were obtained using a digital imaging system (DFC320; Leica Microsystems, Wetzlar, Germany), as described previously (41, 42). Immunohistochemistry was performed on multiple sections in at least two independent experiments for any genotype and/or time of collection.

Western blotting

Western blotting was used to determine the temporal expression of HPGDS in whole oviducts collected from 8- to 10-wk-old mice killed on each day of the estrous cycle. Oviducts were collected from mice killed between 1300 and 1400 h on each day of the estrous cycle. Single oviducts collected from each of four mice were pooled to generate a sample; the analysis was replicated in three experiments. Oviducts were lysed by homogenization in radioimmunoprecipitation assay buffer (Cell Signaling Technology, Danvers, MA). The concentration of protein in each sample was determined using the bicinchoninic acid protein assay (Pierce, Rockford, IL) and protein (50 μg per lane) separated by 10% SDS-PAGE before being transferred to nitrocellulose membranes (Whatman, Kent, UK) using a semidry transfer. Membranes were blocked with 5% nonfat milk at room temperature for 2 h to reduce nonspecific binding and then incubated at 4 C overnight with the same antibody used for the immunohistochemical localization of HPGDS at a dilution of 1:1000. The immunoreaction was determined by incubation for 1 h at room temperature with a horseradish peroxidase-conjugated antirat IgG (Abcam, Cambridge, UK). Subsequent incubation with antiglyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Lab Frontier, Seoul, Korea) and anti-β-tubulin (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) polyclonal rabbit antibodies, each diluted at 1:5000, was used to verify protein loading, with all images acquired for subsequent quantification using an enhanced chemiluminescence kit (Amersham Pharmacia, Freiburg, Germany), as described before (37).

Gene expression I

Real-time PCR was performed to determine the level of expression of mRNA encoding the two receptors that bind PGD₂, PGD receptor (Ptgdr), and G protein-coupled receptor 44 (Gpr44). Gene expression was determined in isolated oviductal EPI and residual oviductal stroma collected from immature CD1 females killed at 0 and 48 h after treatment with 5 IU of PMSG. Briefly, oviductal EPI were separated from the residual stroma (RES) by gently squeezing small sections of the oviduct with fine forceps under a dissecting microscope, as described before (41). The EPI or RES from three or four mice was pooled to generate a single sample, and a total of three samples were generated at each time point. Total RNA was then extracted using TRIzol (Invitrogen) and purified with RNeasy (QIAGEN, Valencia, CA). Oligonucleotide primers (Table 1) were generated for Ptgdr and Gpr44 as well as L19 as an endogenous reference gene. The specificity of each primer set was confirmed by running the PCR products on a 1.5% agarose gel as well as analyzing the melting (dissociation) curve after each PCR reaction. Real-time PCR was performed with a total volume of 25 μl per reaction, with each reaction containing 5 μl of cDNA, 1 μl of a 10 μm stock of each primer (forward and reverse), 12.5 μl of 2× SYBR Green PCR Master Mix, and 5.5 μl of diethylpyrocarbonate-treated water. Real-time PCR were performed on an Eppendorf ep realplex² Thermal Cycler (Eppendorf, Hamburg, Germany), as described previously (37, 41). The relative amount of each transcript was calculated using the 2^−ΔΔCT method (43).

Table 1.

Primer sequences (forward and reverse) used for independent real-time PCR analyses

Name	Primer sequence
Ccl3	5′-CCTCTGTCACCTGCTCAACA-3′
	5′-GATGAATTGGCGTGGAATCT-3′
Ccl4	5′-TCCCACTTCCTGCTGTTTCT-3′
	5′-GAGGAGGCCTCTCCTGAAGT-3′
Ccl11 #1	5′-TTCCATCTGTCTCCCTCCAC-3′
	5′-CTATGGCTTTCAGGGTGCAT-3′
Ccl11 #2	5′-CTCCACAGCGCTTCTATTCC-3′
	5′-CTTCTTCTTGGGGTCAGCAC-3′
Ccl11 #3	5′-CTCCACAGCGCTTCTATTCC-3′
	5′-CTATGGCTTTCAGGGTGCAT-3′
Ccl11 #4	5′-CTCCACAGCGCTTCTATTCC-3′
	5′-CTTCTTCTTGGGGTCAGCAC-3′
Ccl19	5′-CAGGAAACCAAGGACCAGAA-3′
	5′-CGGCTTTATTGGAAGCTCTG-3′
Cxcl11	5′-CAGCTGCGACAAAGTTGAAG-3′
	5′-GCATGTTCCAAGACAGCAGA-3′
Cxcl12	5′-CTCCCTCTCTTCCCTTTGCT-3′
	5′-TCAGAGCCCATAGAGCCACT-3′
Gapdh	5′-CCCCCAATGTGTCCGTCGTGG-3′
	5′-TGAGAGCAATGCCAGCCCCG-3′
Gpr44	5′-AGAGACACCATCCCGCGGCT-3′
	5′-ACAGAATGGGCACGCGCCTC-3′
Hpgds	5′-GCGCCAAACCCAGAAGGCCT-3′
	5′-GCTTGGGCCTGGGCCACATT-3′
Il13	5′-GGTTCTCTCACTGGCTCTGG-3′
	5′-CCACACTCCATACCATGCTG-3′
Il17b	5′-CTGACTTGGTGGGATGGACT-3′
	5′-CTCTTCCATTCGAGCGTAGG-3′
L19	5′-TGGTTGGATCCCAATGAGAC-3′
	5′-GTCTGCCTTCAGCTTGTGGAT-3′
Ptgdr	5′-CTTCGGTGCGGGGCTTCTGG-3′
	5′-CCAGCAAAGGGGAGCGCACA-3′
Tnfrsf1b	5′-AAGGACTGTTCCTGGGTGTG-3′
	5′-GTAAAGGTGGGATGGGAGGT-3′

Gene expression II

To determine whether inhibition of HPGDS in vivo affected the expression of genes regulating inflammation in the oviduct, a real-time PCR-based superarray approach was used. Beginning on the day of diestrus, 8-wk-old CD1 mice were treated orally, twice daily for 4 d, with 30 mg/kg body weight of an inhibitor of HPGDS (HQL-79; Cayman Chemical, Ann Arbor, MI) in 2% methyl cellulose, or 2% methyl cellulose as the vehicle control. Treatment with HQL-79 will specifically inhibit HPGDS-catalyzed production of PGD₂; treatment will have minimal effect on the production of any other PG (44). This dose was chosen based upon reports by others (45–48). At 1300 h on the fifth day, mice were killed and the oviducts collected. Oviducts were collected from four mice for each treatment, and the paired oviducts from each mouse were handled as individual samples for this analysis. Total RNA was isolated from each sample and purified, as described above. The mouse Inflammatory Cytokines and Receptors RT² Profiler PCR arrays and RT² Real-Timer SYBR Green reagent were purchased from SuperArray Bioscience Corp. (Frederick, MD) and used, as described before (37). Briefly, each superarray (96-well plate) contains the primers to identify 84 key genes involved in the inflammatory response as well as housekeeping genes. A genomic DNA contamination control, RT controls and positive PCR controls are also included. The RT² First Strand kit contained all the reagents required to reverse transcribe the total RNA to cDNA and eliminate potential genomic DNA. The level of expression of those genes identified by the superarray analysis as having a probability of being affected by treatment (P < 0.10) was then examined by real-time PCR using independently designed primers. Oligonucleotide primers were generated and tested (Table 1) and real-time PCR performed, as described above. For this analysis, ROX dye, a stabilized conjugate of caboxy-X-rhodamine in water (at a 1:500 dilution) was also included in each PCR, and consistent with analysis of the superarray dataset, the relative level of expression of each mRNA was standardized against GAPDH as a housekeeping gene (43).

Gene expression III

To determine whether the expression of mRNA encoding Hpgds in the oviduct was dependent upon ESR1, oviducts were collected from transgenic mice bearing a global deletion of ESR1 for analysis by real-time PCR. Immature female mice (ESR1KO and WT) were killed at 23 d of age or treated with 5 IU of PMSG and killed 48 h later. Whole oviducts were collected for subsequent extraction of RNA, and the oviducts from three or four mice of each genotype were pooled to generate a single sample. A total of three samples were generated for each group. Oligonucleotide primers (Table 1) for Hpgds were generated, tested, and real-time PCR performed, as described above.

EPI viability analysis

Inflammation-induced oviductal EPI death is a common precursor to oviductal disease (8–13). We therefore sought to determine whether inhibition of HPGDS in vitro affected oviductal EPI viability using a well-characterized, immortalized line (OE-E6/E7) of hOEC (49) and the CellTiter 96 AQueous One Solution (MTS, [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt]) Cell Proliferation Assay system (Promega, Madison, WI). The cell line was a generous gift from William S.B. Yeung at the University of Hong Kong and was used according to reported guidelines (49–52). Cells were maintained in T-25 (25 cm²) or T-75 (75 cm²) flasks in DMEM/F12 medium with 10% fetal bovine serum plus antibiotics at 5% CO₂, 37 C (all from Invitrogen). For the cell viability analysis, cells were subcultured to 96-well plates at 2 × 10⁴ cells per well. Six hours before experimental treatment, culture medium was changed to serum-free DMEM/F12. Cells were then treated with vehicle, 50 μm PGD₂, 300 μm HQL-79, 300 μm HQL-79 plus 5 μm PGD₂, or 300 μm HQL-79 plus 50 μm PGD₂ (Cayman Chemical). After overnight incubation, cell viability was determined using the MTS assay according to the manufacturer's protocol. Twenty microliters of the MTS reagent were added to each well, the cells incubated for an additional 3 h, and then absorbance measured at 492 nm using an Infinite F200 plate reader (Tecan Group, Männedorf, Switzerland), as described before (53). As described by the manufacturer, cell viability is directly proportional to the quantity of formazan product as measured by absorbance.

Statistical analysis

Datasets were first tested for normality and homogeneity of variance. When appropriate, data were transformed before statistical analysis. Nontransformed data are depicted in all the figures. One-way ANOVA using SigmaStat 3.5 (Systat Software, Inc., Point Richmond, CA) was used to determine differences in EPI viability, levels of mRNA for the analyses of Hpgds in WT and ESR1KO mice, and for the quantification of HPGDS in the oviducts of cycling mice. If differences were detected, Tukey's test was used to determine which means differed. The Student's t test was used to determine the effect of treatment with HQL-79 on gene expression in the superarray and independent real-time PCR analyses, to determine whether levels of mRNA for Ptgdr and Gpr44 differed in EPI vs. RES and to determine whether HPGDS differed in the oviducts of immature mice vs. those treated with PMSG.

Results

Expression of HPGDS in the oviduct

The oviduct is known as a dominant site for the expression of mRNA encoding Hpgds (29). However, no analysis of the spatial or temporal patterns of expression had been performed, an analysis that could provide clues as to the functional significance of what appears to be an active modulator of oviductal function. Hence, immunohistochemistry was used to determine the cellular and spatial localization of HPGDS, as well as provide an indication of whether the expression of this synthase changes in response to treatment with exogenous gonadotropins, or over the course of the estrous cycle. In the immature, prepubertal mouse, HPGDS was consistently expressed in the EPI lining the isthmus of the oviduct (Fig. 1). Forty-eight hours after follicular development was stimulated in these immature mice by treatment with PMSG, staining for HPGDS was also readily observed in the oviduct's ampullary EPI layer. Our analysis of HPGDS in the oviducts of mature, naturally cycling mice proved consistent with these results. Epithelial localization of HPGDS was observed in the isthmus of mice killed on all days of the estrous cycle, with an induction of HPGDS in the EPI of the ampulla observed in mice killed at proestrus. As a complement to the analysis of HPGDS by immunohistochemistry, Western blotting was performed on lysates of whole oviducts collected from mice at the same stages of development. The hematopoetic form of PGDS was detectable in whole oviducts of immature (0 h) mice and increased after treatment with PMSG (P < 0.01) (Fig. 2). The level of expression of HPGDS also differed over the course of the estrous cycle; HPGDS was increased in whole oviducts collected on the day of estrus vs. the day of diestrus (P < 0.05) (Fig. 2). Because differences in the spatial localization of HPGDS were observed by immunohistochemistry, especially an induction of HPGDS in the EPI of the ampulla in mice killed at PMSG + 48 h and at proestrus, the quantification of HPGDS in whole oviducts should be interpreted accordingly.

Fig. 1.

Immunohistochemical localization of HPGDS in the ampulla and isthmus of the mouse oviduct. Representative images are shown. Upper panels, Immature mice were killed at 23 d of age (0 h) or treated with 5 IU of PMSG to induce follicular development and killed 48 h later. Lower panel, Eight- to ten-week-old regularly cycling mice were killed on each day of the estrous cycle. Tissues were fixed in Bouin's fixative and embedded in paraffin. Scale bars, 50 μm.

Fig. 2.

Detection of HPGDS in whole oviducts of mice. Upper panels, Images of representative Western blottings are shown. Upper left, Immature mice were killed at 23 d of age (0 h) or treated with 5 IU of PMSG to induce follicular development and killed 48 h later. Upper right, Eight- to ten-week-old regularly cycling mice were killed on each day of the estrous cycle. Di, Diestrus; Pro, proestrus; Est, estrus; Met, metestrus. GAPDH and β-tubulin were used as loading controls. Lower panels, Quantification of Western blottings (HPGDS/GAPDH). Lower left, An asterisk indicates differences in gene expression (P < 0.01). Lower right, Values with different superscript letters differ (P < 0.05).

Expression of Ptgdr and Gpr44 in the oviduct

Two receptor subtypes are reported to bind PGD₂, PTGDR and GPR44, which is also known as chemoattractant receptor-homologous molecule expressed on TH2 cells (CRTH2). To begin to understand the receptor-mediated signaling pathways involved in EPI-catalyzed PGD₂, a real-time PCR analysis was performed to determine levels of mRNA encoding Ptgdr and Gpr44 in the oviducts of immature mice, before and 48 h after treatment with exogenous PMSG. On a per microgram of total RNA basis, the expression of mRNA for Ptgdr was higher in oviductal EPI than RES (P < 0.05) (Fig. 3A). In contrast, mRNA for Gpr44 was higher in the RES than isolated EPI (P < 0.05) (Fig. 3B). This was observed in mice killed at 0 h and at PMSG + 48 h. Overall, these results suggest diversity to the effects of PGD₂ on oviductal function, with biological action(s) apparently not confined to the dominant, epithelial site of HPGDS expression.

Fig. 3.

Expression of mRNA encoding Ptgdr and Gpr44 in EPI and RES collected from whole oviducts of mice. Immature mice were treated with 5 IU of PMSG to induce follicular development and killed 48 h later. Data are the means ± sem for real-time PCR analysis of three samples for each cell type, with EPI and RES cells pooled from three to four mice for each sample. Levels of mRNA were obtained by real-time PCR and are expressed as fold changes. For each time point within a panel, an asterisk indicates differences in gene expression (P < 0.05).

Dependence of HPGDS on ESR1

Immunohistochemistry indicated an induction of HPGDS in the EPI of the ampulla from immature mice treated with PMSG and cycling mice killed at proestrus. With these results suggesting possible regulation through estradiol and ESR1, the expression of mRNA encoding Hpgds and its protein was determined in oviducts collected from mice with or without a global deletion of ESR1. Real-time PCR revealed readily detectable levels of mRNA for Hpgds in whole oviducts collected from immature WT mice, regardless of treatment with PMSG (Fig. 4A), with the lack of any induction of Hpgds by PMSG likely reflecting the dominance of EPI in the isthmus vs. the ampulla of the oviduct (where the differential staining for HPGDS was observed). The expression of mRNA encoding Hpgds was decreased to approximately 20% of WT controls in ESR1KO females, again regardless of treatment with PMSG (P < 0.05). This apparent dependence on ESR1 was confirmed by immunohistochemistry, with HPGDS readily detectable in the EPI of the isthmus in 10-wk-old WT mice but not in 10-wk-old ESR1KOs (Fig. 4B).

Fig. 4.

A, Expression of mRNA encoding Hpgds in whole oviducts collected from WT and ESR1KO mice. Immature mice were killed at 23 d of age (0 h) or treated with 5 IU of PMSG to induce follicular development and killed 48 h later. Data are the means ± sem for real-time PCR analysis of three samples for each treatment and genotype, with oviducts pooled from three to four mice for each sample. Levels of mRNA were obtained by real-time PCR and are expressed as fold changes. Values with different superscript letters differ (P < 0.05). B, Localization of HPGDS in the ampulla and isthmus of 10-wk-old WT, and ESR1KO mice killed on diestrus. Representative images are shown. Tissues were fixed in Bouin's fixative and embedded in paraffin. Scale bars, 50 μm.

Regulation of inflammatory gene expression in vivo

To test our hypothesis that HPGDS is regulating inflammatory processes in the oviduct, we treated mice for 4 d with an orally active inhibitor of HPGDS (HQL-79) before collecting their oviducts for a superarray, real-time PCR-based analysis of inflammatory gene expression. The expression of nine key genes identified in the superarray analysis (P < 0.10) was then reevaluated by independent real-time PCR. With the exception of chemokine (C-C motif) ligand 11 (Ccl11), the results of the two real-time PCR analyses were consistent (Table 2). Expression of mRNA's encoding chemokine (C-C motif) ligands 3, 4 and 19 (Ccl3, Ccl4, and Ccl19) were each increased by 2- to 3-fold after treatment of mice with HQL-79 (P < 0.02 for each mRNA by independent real-time PCR). Expression of mRNA encoding chemokine (C-X-C motif) ligands 11 and 12 (Cxcl11 and Cxcl12, respectively) was also increased by inhibition of HPGDS in vivo; however, our secondary analysis of mRNA for Cxcl11 indicated a reduced magnitude of change for the expression of that gene (3.03- vs. 1.69-fold, P = 0.082 by independent real-time PCR). Similarly, expression of mRNA encoding IL-13 and IL-17B (Il13 and Il17b), and TNF receptor superfamily, member 1b (Tnfrsf1b), was increased by approximately 2-fold by treatment of mice with HQL-79 (P < 0.02 for each mRNA by independent PCR). Given the −87-fold (P < 0.001) effect of treatment on the expression of mRNA encoding Ccl11 in the superarray analysis, we generated a total of four sets of primers to validate those results. We did not confirm the superarray data for Ccl11, and our secondary, independent analysis was consistent using all four primer pairs that were generated. Although levels of mRNA for Ccl11 in the oviducts of HQL-79-treated mice appeared to average approximately 5-fold increased levels of expression, this was largely due to a dramatic induction of mRNA for this gene in the oviducts of one animal alone. Overall, the results of this experiment suggest that HPGDS does regulate inflammation in the oviduct of the mouse. Inhibition of HPGDS by treatment of mice with HQL-79 increased the expression of mRNA for eight of the inflammatory genes examined.

Table 2.

Effect of treatment with HQL-79 on inflammatory gene expression in vivo

Gene name	Symbol	Fold change HQL-79/Vehicle		P value (t test)
		Array	PCR	Array	PCR
Chemokine (C-C motif) ligand 3	Ccl3	2.04	2.31	0.036	0.008
Chemokine (C-C motif) ligand 4	Ccl4	1.97	2.78	0.083	0.004
Chemokine (C-C motif) ligand 11	Ccl11	−87.4	4.97	0.001	.120
Chemokine (C-C motif) ligand 19	Ccl19	2.14	2.42	0.026	0.015
Chemokine (C-X-C motif) ligand 11	Cxcl11	3.03	1.69	0.013	0.082
Chemokine (C-X-C motif) ligand 12	Cxcl12	2.03	2.93	0.063	0.018
IL-13	II13	2.52	2.27	0.097	0.012
IL-17B	II17b	2.31	2.24	0.033	0.005
TNF receptor superfamily, member 1b	Tnfrsf1b	1.95	1.83	0.057	0.006

Fold change in gene expression and P values are indicated after analysis by superarray (array) and by independent real-time PCR (PCR).

Regulation of oviductal EPI viability in vitro

Inflammation-induced oviductal EPI death is known to precede tubal blockage and infertility (8–13). To address the functional significance of identifying HPGDS as a regulator of inflammation in the oviduct, we sought to determine whether treatment of an immortalized line of hOEC with this same inhibitor affected EPI viability in vitro. Withdrawal of serum decreased the viability of hOEC (P < 0.05) (Fig. 5). Treatment of serum-free hOEC with HQL-79 decreased the viability of EPI to approximately 60% of serum-free controls (P < 0.05). Treatment of hOEC with PGD₂ alone did not affect EPI viability (P > 0.05); however, concurrent treatment of hOEC with HQL-79 plus either dose of PGD₂ was partially able to reverse the effect of HQL-79, suggesting that the effect of treatment with HQL-79 was specific to the HPGDS-PGD₂ pathway and not a deleterious effect of treatment with this inhibitor. Real-time PCR confirmed that mRNA for Ptgdr and Gpr44 was expressed in cultured hOEC (data not shown).

Fig. 5.

Effect of HQL-79 and PGD₂ on the viability of hOEC in vitro. Viability is expressed as absorbance, using the Cell Titer MTS assay system. Data are the means ± sem from six replicates per treatment. Values with different superscript letters differ (P < 0.05).

Discussion

Our results suggest that HPGDS-catalyzed PGD₂ may act as a regulator of inflammation in the oviduct. Inhibition of HPGDS in vivo consistently increased the expression of mRNA for inflammation-regulating chemokines and ILs. Inhibition of HPGDS in vitro decreased the viability of hOEC. We found that HPGDS was abundant in the EPI of the isthmus, even in the immature mouse, and appeared largely dependent upon functional expression of ESR1. In addition, some temporal regulation of HPGDS was observed over the course of the estrous cycle, especially an induction in the EPI of the ampulla under an estrogen-dominant hormonal environment (at PMSG + 48 h and at proestrus). Considering that freshly ovulated cumulus-oocyte complexes, associated follicular debris, spermatozoa, seminal fluids, and possibly an array of foreign pathogens must traverse the oviduct after each ovulation or after mating, it can be postulated that high levels of HPGDS are constitutively expressed in the oviduct as part of a mechanism that ensures this organ the ability to rapidly respond to an inflammatory insult and, therefore, maintain homeostasis and continued function, i.e. we hypothesize that upon an insult to the oviduct, the expression of HPGDS is decreased, allowing the inflammatory response to proceed rapidly.

When first cloned, mRNA for Hpgds was detected in the oviduct but not in the ovary or uterus (29). This appears to be due to the sensitivity of the methods employed rather than total specificity of this synthase to the oviduct as a female reproductive organ. In cultures of luteinizing granulosa cells collected from rats, the expression of mRNA for Hpgds was reported to increase after treatment with human chorionic gonadotropin (53). mRNA for Gpr44 has also been localized to the granulosa cells of the ovarian follicle (54). In addition, Zelinski-Wooten et al. (55, 56) have performed analyses of all the major PG, including PGD₂, on various aspects of luteal function. Most pertinent to the results described here, treatment of granulosa cells collected from rats with PGD₂ was reported to increase cell viability (53). Although we did not observe an increase in oviductal EPI viability after treatment with PGD₂ alone, the effect of PGD₂ on HQL-79-treated oviductal EPI viability can be considered consistent with this. The ovary must therefore be considered a potential site of action for HPGDS-catalyzed PGD₂; however, from a functional viewpoint, the abundance of HPGDS to the oviduct needs to be recognized. We evaluated the relative level of expression of mRNA for Hpgds in the reproductive tract of the mouse and consistently found that the level of mRNA for Hpgds was approximately 15-fold higher in the oviduct than the associated ovary (Bridges, P.J. and M. Jeoung, unpublished data).

When considering the inflammatory process and the potential regulation within the oviduct by HPGDS, cytokines can be broadly described as signaling molecules that regulate immune function. Chemokines and ILs are both members of the family of cytokines; chemokines are recognized for their chemotactic activity toward leukocytes (57); however, the breadth in the immune response to ILs makes them harder to classically define. Interestingly, chemotactic recruitment of leukocytes into the oviduct appears consistent with the endocrine-regulated immune response to the physiological inflammatory process that is ovulation. Leukocytes are well-established infiltrators of the ovary around ovulation (58, 59), and in this study, we observed an increase in expression of mRNA for chemokine (C-C motif) ligands 3, 4, and 19 and chemokine (C-X-C motif) ligands 11 and 12 in the oviducts of mice treated with HQL-79. Overall, the extent of chemokine regulation that we observed suggests a role for HPGDS in the regulation of leukocyte infiltration in preparation for EPI damage after ovulation and/or mating. Additional research specifically investigating inflammatory infiltration after inhibition of HPGDS is warranted to delineate the precise mechanisms involved. Given that the spleen appears to acts as a reservoir for the leukocytes released around ovulation (58, 59), the specificity of dominant sites of expression for Hpgds, i.e. to the spleen plus oviduct (29), is intriguing.

An increase in mRNA for Il13 and Il17b was also observed after inhibition of HPGDS. IL-17 is reported to contribute to the influx of neutrophils after genital infection of mice with Chlamydia muridarum (60), and an increase in IL-2, IL-6, and IL-10 was observed after infection of pig-tailed macaques with Chlamydia trachomatis (61). Infection with C. trachomatis is also reported to induce chemokine (C-X-C motif) ligand 13 within the human oviduct (62). In addition to HPGDS-regulated inflammation as a mechanism to ensure oviductal homeostasis after ovulation and mating, potential regulation of cytokines induced in response to pathological infection appears likely. Given that the incidence of infection with C. trachomati is currently at record high levels in the United States (63) and that treatment of Neisseria gonorrhoeae, the second most reported of these notifiable sexually transmitted diseases, is becoming increasingly plagued by resistance to antibiotic therapy (64), further examination of HPGDS as a potential regulator of the immune response to pathological infection is warranted.

If HPGDS does indeed act as a primary regulator of inflammation within the oviduct, the apparent dependence of this synthase on ESR1 is interesting. This is especially true when considering that the subtle temporal regulation of HPGDS over the course of the estrous cycle, and apparent induction of this synthase in the EPI of the ampulla in response to PMSG and at proestrus suggests only minor regulation by ovarian-produced estradiol. Considering that HPGDS was also found to be highly expressed in the oviducts of immature, prepubertal mice, it appears that this requirement for ESR1 is a developmental phenomenon rather than an active postpubertal steroid/receptor signaling mechanism. Within the oviduct, ESR1 is distributed in a homogenous manner. Immunostaining reveals distinct nuclear localization to both ciliated and secretory EPI, as well as to the nuclei of the oviduct's smooth muscle (65). This nuclear receptor is also abundant in the oviducts of prepubertal mice (65). It is important to note, however, that in contrast to the well-recognized phenotype of the ovaries collected from ESR1KO mice (66, 67), sections of both the ampulla and isthmus appear to be morphologically normal in these ESR1KOs (Fig. 4).

Further examination of gene expression in the oviducts of ESR1KO mice also revealed what appears to be a partial mechanism to compensate for decreased HPGDS activity. In the oviducts of mice devoid of ESR1 and therefore HPGDS, we observed an increase in the expression of mRNA for a genetically distinct, alternate synthase that can also catalyze the production of PGD₂ (68, 69), lipocalin-type PGDS (LPGDS). This synthase is recognized for its role in the central nervous system (70, 71), especially in the regulation of sleep-wake cycles (72). Although the overall level of expression of mRNA for Lpgds in the oviduct is low when compared with Hpgds, the expression of mRNA for Lpgds in the oviducts of ESR1KO mice was approximately 1.5-fold greater than for their WT littermates (Bridges, P.J. and M. Jeoung, unpublished data). LPGDS is established as an estrogen-regulated gene (73, 74), yet consistent with our analysis of mRNA for Hpgds, we observed no affect of treatment with gonadotropins on the expression of mRNA for Lpgds in WT or ESR1KOs.

Overall, our results support the hypothesis that HPGDS is a regulator of inflammation in the oviduct, an effect likely required to maintain homeostasis and function of this key reproductive organ. Given the prevalence of oviductal dysfunction causing infertility in both women (3–5) and domestic species (75), further investigation of not only HPGDS-regulated inflammation but how the oviduct responds to pathological challenges appears to be needed. Our understanding of oviductal biology remains limited as a whole, and delineating mechanisms such as HPGDS-regulated inflammation can only increase our ability to manage for optimal reproductive health.