Seminal plasma promotes decidualization of endometrial stromal fibroblasts in vitro from women with and without inflammatory disorders in a manner dependent on interleukin-11 signaling

Abstract

STUDY QUESTION

Do seminal plasma (SP) and its constituents affect the decidualization capacity and transcriptome of human primary endometrial stromal fibroblasts (eSFs)?

SUMMARY ANSWER

SP promotes decidualization of eSFs from women with and without inflammatory disorders (polycystic ovary syndrome (PCOS), endometriosis) in a manner that is not mediated through semen amyloids and that is associated with a potent transcriptional response, including the induction of interleukin (IL)-11, a cytokine important for SP-induced decidualization.

WHAT IS KNOWN ALREADY

Clinical studies have suggested that SP can promote implantation, and studies in vitro have demonstrated that SP can promote decidualization, a steroid hormone-driven program of eSF differentiation that is essential for embryo implantation and that is compromised in women with the inflammatory disorders PCOS and endometriosis.

STUDY DESIGN, SIZE, DURATION

This is a cross-sectional study involving samples treated with vehicle alone versus treatment with SP or SP constituents. SP was tested for the ability to promote decidualization in vitro in eSFs from women with or without PCOS or endometriosis (n = 9). The role of semen amyloids and fractionated SP in mediating this effect and in eliciting transcriptional changes in eSFs was then studied. Finally, the role of IL-11, a cytokine with a key role in implantation and decidualization, was assessed as a mediator of the SP-facilitated decidualization.

PARTICIPANTS/MATERIALS, SETTING, METHODS

eSFs and endometrial epithelial cells (eECs) were isolated from endometrial biopsies from women of reproductive age undergoing benign gynecologic procedures and maintained in vitro. Assays were conducted to assess whether the treatment of eSFs with SP or SP constituents affects the rate and extent of decidualization in women with and without inflammatory disorders. To characterize the response of the endometrium to SP and SP constituents, RNA was isolated from treated eSFs or eECs and analyzed by RNA sequencing (RNAseq). Secreted factors in conditioned media from treated cells were analyzed by Luminex and ELISA. The role of IL-11 in SP-induced decidualization was assessed through Clustered regularly interspaced short palindromic repeats (CRISPR)/Cas-9-mediated knockout experiments in primary eSFs.

MAIN RESULTS AND THE ROLE OF CHANCE

SP promoted decidualization both in the absence and presence of steroid hormones (P < 0.05 versus vehicle) in a manner that required seminal proteins. Semen amyloids did not promote decidualization and induced weak transcriptomic and secretomic responses in eSFs. In contrast, fractionated SP enriched for seminal microvesicles (MVs) promoted decidualization. IL-11 was one of the most potently SP-induced genes in eSFs and was important for SP-facilitated decidualization.

LARGE SCALE DATA

RNAseq data were deposited in the Gene Expression Omnibus repository under series accession number GSE135640.

LIMITATIONS, REASONS FOR CAUTION

This study is limited to in vitro analyses.

WIDER IMPLICATIONS OF THE FINDINGS

Our results support the notion that SP promotes decidualization, including within eSFs from women with inflammatory disorders. Despite the general ability of amyloids to induce cytokines known to be important for implantation, semen amyloids poorly signaled to eSFs and did not promote their decidualization. In contrast, fractionated SP enriched for MVs promoted decidualization and induced a transcriptional response in eSFs that overlapped with that of SP. Our results suggest that SP constituents, possibly those associated with MVs, can promote decidualization of eSFs in an IL-11-dependent manner in preparation for implantation.

STUDY FUNDING/COMPETING INTEREST(S)

This project was supported by NIH (R21AI116252, R21AI122821 and R01AI127219) to N.R.R. and (P50HD055764) to L.C.G. The authors declare no conflict of interest.

Introduction

Seminal plasma (SP) is the liquid fraction of semen produced by the male accessory sex organs. In addition to serving as a vehicle to transport sperm, SP can also elicit responses in the female reproductive tract (FRT) that promote conception, in particular implantation, as shown in mice (Bromfield et al., 2014) and humans (Bellinge et al., 1986; Coulam et al., 1995; Tremellen et al., 2000) and as confirmed in a recent meta-analysis (Crawford et al., 2015). Rodent models indicate that SP components may promote implantation, in part, through the induction of regulatory T cells that mediate maternal immune tolerance of newly formed semi-allogenic embryos (Robertson et al., 1996; Johansson et al., 2004; Moldenhauer et al., 2009; Robertson et al., 2009; Balandya et al., 2012; Meuleman et al., 2015) and through enhancing secretion of embryotrophic cytokines and chemokines by FRT epithelial cells (Robertson et al., 1992; Sharkey et al., 2012a; 2012b; 2018).

Implantation occurs in the endometrium, and successful pregnancy requires activity of the endometrial stromal fibroblasts (eSFs). During the secretory phase of the menstrual cycle, eSFs undergo decidualization, a progesterone (P₄)-driven differentiation process that is essential for embryo implantation and subsequent maintenance of pregnancy (Cha et al., 2012). In vitro, eSFs can be induced to decidualize by treatment with estradiol (E₂) together with P₄ (Irwin et al., 1989). Endometrial SFs from women with inflammatory disorders associated with suboptimal implantation (e.g. endometriosis or polycystic ovary syndrome (PCOS)) have reduced decidualization capacity (Klemmt et al., 2006; Piltonen et al., 2015), suggesting that aberrations in eSF function may be an underlying cause of female infertility.

Interestingly, SP can accelerate and increase P₄-induced decidualization (Doyle et al., 2012), providing another potential explanation for its ability to improve implantation rates. However, the components of SP and the mechanisms responsible for this effect are unknown. Furthermore, to what extent SP can promote decidualization of eSFs from women with inflammatory disorders and comprised decidualization has not been explored. A recent microarray study conducted by our group revealed that eSFs respond dramatically to SP exposure and that one of the most highly SP-induced genes is interleukin (IL)-11 (~15-fold induction relative to vehicle treatment) (Chen et al., 2014). IL-11 belongs to the gp130 family of cytokines, which includes IL-6 and leukemia inhibitory factor (LIF); both of which are also induced by eSFs upon SP exposure (Gutsche et al., 2003; Chen et al., 2014). IL-11, IL-6 and LIF play important roles in implantation—LIF and IL-11 as essential factors (Stewart et al., 1992; Robb et al., 1998; Menkhorst et al., 2009) and IL-6 in a supporting role (Jasper et al., 2007). Furthermore, IL-11 promotes P₄-induced decidualization of eSFs in vitro and eSFs from patients with infertility have diminished IL-11 production and decidualization capacity (Dimitriadis et al., 2002; Karpovich et al., 2005).

We sought to better understand how SP can improve implantation rates by investigating the effects of SP and SP components on decidualization and on global gene expression in eSFs. SP contains a variety of bioactive agents that are good candidates for the mediators of decidualization. Because the SP factor(s) promoting decidualization is >3500 Da (Doyle et al., 2012), SP proteins (which are typically >3500 Da) are attractive candidates. Proteins in SP may exist in a soluble form or may be associated with extracellular vesicles (EVs) that are highly abundant in semen (Machtinger et al., 2016). Among SP proteins, semen amyloids seemed particularly promising candidates. Initially identified for their ability to enhance HIV infection (Munch et al., 2007; Roan et al., 2011; Usmani et al., 2014) and more recently implicated in reproduction via sperm selection (Roan et al., 2017), semen amyloids are structurally similar to amyloids associated with neurodegenerative diseases. In these diseases, amyloids are generally pro-inflammatory and induce IL-6 (Griffin et al., 2016) and LIF (Rensink et al., 2002). Furthermore, IL-11 levels in cerebrospinal fluid are associated with the levels of Aβ42 amyloids (Galimberti et al., 2008). We therefore postulated that semen amyloids may also induce these IL-6 family members in the FRT, thereby promoting a receptive state. We first assessed the ability of SP to promote decidualization of eSFs from healthy women and from women with inflammatory disorders with comprised decidualization. We then carried out a detailed global transcriptomic analysis of eSFs exposed to semen amyloids and other constituents of SP. Our findings revealed that unlike other SP constituents, SP amyloids are not potent signaling agents in the FRT. Furthermore, we demonstrate that the ability of SP to promote decidualization is dependent on IL-11.

Materials and Methods

Ethical approval

Research procedures were performed in accordance with the Declaration of Helsinki. All donors were confirmed not to be pregnant and provided written, informed consent.

Tissue processing and cell culture

Human endometrial tissues were obtained from the Women’s Health Clinic of Naval Medical Center Portsmouth (NMCP) in Virginia (CIP no. NMCP.2016.0068) as well as the University of California, San Francisco (UCSF) Endometrial Tissue Bank (IRB no. 14-15 361), under established operating procedures (Fassbender et al., 2014). Details regarding the age, cycle phase and clinical diagnosis (if available) of each donor is included in Supplementary Table SI.

Endometrial epithelial cells (eECs) and eSFs were isolated as previously reported (Chen et al., 2014). Briefly, endometrial tissue was digested with 6.4 mg/ml collagenase type I (Worthington Biochemical Corporation, Lakewood, NJ, USA) and 100 U/ml hyaluronidase (Sigma-Aldrich, St. Louis, MO, USA) in Hanks’ buffered salt solution with Ca++ and Mg++ (Life Technologies, Carlsbad, CA, USA). To isolate single cells from glandular fragments and luminal epithelial sheets, digested cells were passed through a Falcon 40-μm cell strainer (Fisher Scientific, Waltham, MA, USA). The eSFs were then further purified using selective attachment, as previously described (Chen et al., 2014, 2015.

Following isolation, eSFs were cultured in serum-containing fibroblast growth medium (SCM; 75% phenol red-free Dulbecco’s modified Eagle’s medium (DMEM; Life Technologies) and 25% MCDB-105 (Sigma-Aldrich) supplemented with 10% charcoal-stripped fetal bovine serum (FBS; BenchMark from Gemini), 1-mM sodium pyruvate (Sigma-Aldrich) and 5-μg/ml insulin) until confluent. eSFs were passaged serially as previously described (Irwin et al., 1989). The eECs were then plated in 24-well plates coated with Matrigel (VWR Scientific, Radnor, PA, USA) with defined keratinocyte serum-free medium (KSFM; Life Technologies). The eEC cultures were monitored for epithelial morphology and the absence of eSF contamination. When an epithelial monolayer with dome-like folding structures (Chen and Roan, 2015) was visible, eEC cultures were considered ready for treatments. Images were captured by bright field microscopy (×100 magnification) using an Olympus CKX41 microscope (Olympus, Tokyo, Japan) with a QIClick CCD Camera and QCapture Suite Plus software (QImaging, Surrey, British Columbia).

SP preparation

Individual semen samples from 20 de-identified donors obtained from the UCSF Center for Reproductive Health (IRB no. 11-06115) were liquefied at room temperature for 2 h and frozen. The samples were then all simultaneously thawed, pooled and centrifuged at 828g for 10 min at room temperature to remove spermatozoa and other cells. Supernatant containing SP was collected, aliquoted and frozen at −80°C and served as the SP stock. Endotoxin levels in SP were quantified using the ToxinSensor Chromogenic LAL Endotoxin Assay Kit (GenScript, Piscataway, NJ, USA).

SP protein degradation

For assessing the role of SP proteins in SP-induced changes in eSFs, SP proteins were denatured by heating at 100°C for 1 h or degraded by treatment with Proteinase K (200 μg/ml) (Sigma-Aldrich) for 18 h at 37°C followed by the addition of phenylmethylsulphonyl fluoride at a final concentration of 5 mM (Roche, Basel, Switzerland) to inactivate the Proteinase K. Protein degradation by Proteinase K was confirmed by running samples at 100 V for 1 h on a Bis-Tris 4–12% gradient protein gel (ThermoFisher) followed by staining with Bio-Safe™ Coomassie (Bio-Rad, Hercules, CA, USA) following the manufacturer’s protocol. To degrade the HIV infection-enhancing amyloids (Roan et al., 2011, 2014), SP was incubated for 37°C for 20 h as previously described (Roan et al., 2014).

Preparation of reconstituted SP and seminal microvesicles

SP was first centrifuged on 0.65-μm centrifugal filters (Millipore, Burlington, MA, USA) at 860g for 10 min at room temperature or until the majority of supernatant had passed through filter. The filtrate was then collected and transferred to 0.22-μm centrifugal filter units (Millipore) and further centrifuged at 860g for 10 min. The retenate on the 0.22-μm centrifugal filter was resuspended with PBS and this solution was used as the microvesicle (MV)-enriched fraction. In order to generate reconstituted SP, the MV-enriched fraction was resuspended in the filtrate from both the 0.65- and 0.22-μm centrifugal filters.

Generation of pure seminal amyloids

As it is not possible to purify endogenous semen amyloids (Usmani et al., 2014), human SEM1(86-107) (sequence: DLNALHKTTKSQRHLGGSQQLL) was custom-synthesized to generate semen amyloids (Celtek Bioscience, Franklin, TN, USA). The peptide was resuspended in PBS to a concentration of 2.5 mg/ml, and amyloid fibrils were formed by agitating overnight at 37°C at 1400 rpm using an Eppendorf Thermomixer (Eppendorf, Hamburg, Germany). Fibrillation was confirmed by quantifying the binding of thioflavin T (ThT) (5 μM) to the agitated peptides (125 μg/ml) using a PerkinElmer Enspire (excitation 440 nm, emission 482 nm; PerkinElmer, Waltham, MA, USA), as described (Roan et al., 2014). Monomeric peptide, corresponding to freshly resuspended SEM1(86-107), was confirmed not to be fibrillar by ThT binding.

TZM-bl infectivity assay

To assess for functional activity of the amyloid fibrils, a TZM-bl infectivity assay was conducted as previously described (Roan et al., 2011, 2014). The TZM-bl cell line was generated from HeLa cells engineered to express CD4, CCR5 and CXCR4, as well as firefly luciferase and Escherichia coli β-galactosidase reporter genes under the control of an HIV-1 long terminal repeat (Wei et al., 2002). A total of 5 × 10³ TZM-bl cells/well were seeded in 100-μL DMEM (Life Technologies) supplemented with 10% FBS (BenchMark cat 100–106 from Gemini Bioproducts, West Sacramento, CA, USA) and 1% Pen-Strep (Gemini Bioproducts). The day after seeding, 180 μl of additional media was added to each well. Peptides, fibrils and SP were serially diluted and then pre-treated with CCR5-tropic 81A HIV-1 (100 ng/ml p24^Gag) obtained from 293T-transfected cells using methods previously described (Roan et al., 2014). A total of 20 μl of the pre-treated 81A was added to each well and infection was allowed to proceed for 2 h. Next, media were replaced with fresh DMEM (Life Technologies) supplemented with 10% FBS, 1% Pen-Strep and 50-μg/ml gentamicin (Gibco, Waltham, MA, USA) and cells were cultured for additional 3 days. Infection was assessed by the way of beta-galactosidase activity (Tropix, Bedford, MA, USA) by quantifying luminescence on a PerkinElmer Enspire (PerkinElmer). Viability was quantified in parallel using the Cell Titer Glo assay (Promega, Madison, WI, USA).

Decidualization assay

Endometrial SFs were treated with ovarian steroids as described (Aghajanova et al., 2009). This treatment protocol allows eSFs to differentiate in response to E₂ and P₄ (Aghajanova et al., 2009; Piltonen et al., 2015), including the induction of decidualization. eSFs were seeded into 24-well plates and cultured in SCM until confluent, and then cultured in SFM media, which consists of SCM media made with 2% instead of 10% FBS. To test the ability of hormones, SP or SP components to promote decidualization, eSFs were treated for 6 h at 37°C with various combinations of the following: 0.1% ethanol (EtOH) as a vehicle control, 10 nM E₂ (17β-estradiol; Sigma-Aldrich) + 1-μM P₄ (E₂P₄; Sigma-Aldrich), 10% SP (a concentration previously used by Sharkey et al., 2012a, 2015, 2018), MV-enriched fraction (equivalent to that present in 10% SP), MV-enriched fraction + E₂P₄, 100-μg/ml amyloids (equivalent to the average physiological concentration in semen; Roan et al., 2014) or 10% SP + E₂P₄. After 6 h, media were replaced with fresh SFM containing hormones as appropriate for the treatment condition and cultured for another 42 h, at which time another cycle of treatment was initiated. Hormone and SP treatments were repeated every 2–3 days for 14 days and supernatants were collected immediately prior to each subsequent round of treatment. Decidualization was assessed by quantifying levels of human insulin-like growth factor binding protein 1 (IGFBP1) and prolactin in the cell culture supernatant by ELISA (Abcam), according to the manufacturer’s protocol. Samples were assayed in triplicates and a standard curve was run for each assay. Images of eSFs (×100 magnification) were captured by bright-field microscopy using an Olympus CKX41 microscope with a QIClick CCD Camera and QCapture Suite Plus software (QImaging). Viability of cells was assessed by the Cell Titer Glo assay (Promega).

Treatment of eSFs and eECs with SP and SP components for RNAseq

The responses of eSFs and eECs to exposure to SP and SP components were examined by RNA sequencing (RNAseq). Of note, similar to our prior microarray study (Chen et al., 2014), we tested the transcriptional response of eECs and eSFs in the absence of hormones. This was because in order to have a system where the response of eSFs and eECs to SP can be assessed in the presence of ovarian hormones, the cells must be co-cultured since eECs require paracrine support from eSFs to differentiate in response to ovarian hormones (Cunha, 1976; Cunha et al., 1983). Because this system is technically challenging as cells are less viable in co-culture than in monoculture conditions (Chen et al., 2013), we generated our RNAseq datasets in the absence of hormones. eSFs were seeded into 24-well plates and cultured in SCM until confluent, while eEC glands were seeded into 24-well, thin layer Corning Matrigel Matrix Plates (VWR Scientific) and cultured in KSFM until domes were apparent. Cells were treated for 6 h at 37°C as follows: with media alone, with 1% SP (a concentration previously used in microarray studies; Chen et al., 2014), with 1% SP pretreated for 20 h at 37°C, with 100-μg/ml peptides, with 100-μg/ml amyloids or with 0.1-Eu/ml lipopolysaccharides (LPS) from E. coli O111:B4 (Sigma-Aldrich). Note the amount of LPS used was equivalent to the levels of LPS present in 1% SP, as determined by limulus amebocyte lysate (LAL) endotoxin analysis. In a separate set of experiments, cells were treated with as follows: with media alone, with 1% reconstituted SP or with 1% of the MV-enriched fraction. Cultured cells were harvested and RNA was isolated, as detailed below.

RNA extraction and library preparation

Total RNA was extracted from eSF or eEC cultures (1 μg of total RNA from each sample) using the High Pure RNA Isolation kit (Roche) according to the manufacturer’s protocol. Quality of isolated RNA was confirmed using Nanodrop for RNA purity (OD 260/280) and Bioanalyzer (Agilent, Santa Clara, CA, USA) for RNA integrity. RNA samples with an RNA integrity number > 6 were used for library preparation for mRNA sequencing.

All RNAseq procedures were performed by Novogene. Library preparation was performed using 1 μg of total RNA from each sample with the NEBNext Ultra RNA Library Prep Kit (New England BioLabs, Inc., Ipswich, MA, USA). Briefly, polyadenylated (polyA) RNAs were selected using NEBNext Magnetic Oligo d(T)25 Beads (New England BioLabs, Inc.). Following polyA selection, RNA was fragmented and primed for first-strand cDNA synthesis, followed by second-strand synthesis. Double-stranded cDNA was purified using Agencourt AMPure XP Beads (Beckman Coulter, Brea, CA, USA), the fragments were end-repaired, 3′ adenylated and NEBNext Adaptor with hairpin loop structure (Illumina, San Diego, CA) was ligated to prepare for hybridization. cDNA fragments ~350 bp in length were selected using the Agencourt AMPure XP system (Beckman Coulter) and PCR amplification was performed to create the final cDNA library.

RNAseq and data analysis

Paired-end RNAseq at >20 million reads per sample was sequenced on an Illumina HiSeq platform (Illumina), as per the manufacturer’s instructions. Sequencing data were deposited in the Gene Expression Omnibus repository under series accession number GSE135640. Trimmed reads were aligned to the Genome Reference Consortium Human Build 38 reference genome (GenBank assembly accession: GCA_000001405.15) using TopHat2 v2.1.0 with the following parameters: global alignment, no mismatch in the 20 base-pair seed, up to two mismatches in the read, library-type fr-unstranded and mate inner distance 50. Aligned reads were filtered by removing reads with low mapping quality (below 20) and keeping singletons. Filtered reads were annotated to the Genome Reference Consortium Human Build 38 reference transcriptome (RefSeq Assembly Accession: GCF_000001405.26) using the Partek E/M annotation model (Partek Flow, build version 5.0.16.1128; Partek, St Louis, MO, USA) with the following parameters: junction reads required to match introns (set to true), strict paired-end compatibility not required (set to false) and strand specificity not required (set to false). Filtered reads were normalized using transcripts per kilobase million. Low-expressing transcripts were removed, with the lowest maximum coverage set to 1.0. Differential gene expression was calculated by ANOVA in a paired test (paired on donor ID), by using treatment as the fixed factor and ID as the random factor. Differentially expressed transcripts were identified based on fold change (≥ ±1.5), with significance level set at P < 0.05. Principal component analysis (PCA) and hierarchical clustering was performed using Partek Flow software. Ingenuity pathway analysis (IPA) software (Qiagen, Hilden, Germany) was used for pathway analysis through the core analysis function to determine biological context of differentially expressed genes.

Luminex multiplex assay

Culture supernatants from samples analyzed by RNAseq were analyzed by the Luminex multiplex assay (ThermoFisher) to assess the effects of SP and SP components on the secretion of cytokines/soluble factors by eSFs and eECs. Supernatants were analyzed by Luminex using a 30-plex human cytokine panel (ThermoFisher) consisting of epidermal growth factor (EGF), eotaxin, basic fibroblast growth factor (FGFβ), granulocyte-colony stimulating factor (G-CSF), granulocyte-macrophage colony-stimulating factor (GM-CSF), hepatocyte growth factor (HGF), interferon alpha (IFN)-α, IFN-γ, IL-10, IL-12, IL-13, IL-15, IL-17, IL-1β, IL-1 receptor antagonist (IL-1RA), IL-2, IL-2 receptor (IL-2R), IL-4, IL-5, IL-6, IL-7, IL-8, IFN gamma-induced protein 10 (IP-10), monocyte chemoattractant protein-1 (MCP-1), monokine induced by gamma (MIG), macrophage inflammatory protein (MIP)-1α, MIP-1β, regulated upon activation normal T cell expressed and secreted (RANTES), tumor necrosis factor (TNF)-α and vascular endothelial growth factor (VEGF). The assay was performed by the Endocrine Technologies Support Core (ETSC) at the Oregon National Primate Research Center (ONPRC) according to the manufacturer’s instructions (ThermoFisher). Briefly, 50 μl of each serum sample was diluted in assay diluent and incubated overnight with antibody-coated, fluorescent-dye capture microspheres specific for each analyte, followed by detection antibodies and streptavidin-phycoerythrin. Washed microspheres with bound analytes were resuspended in reading buffer and analyzed on a Milliplex LX-200 Analyzer (EMD Millipore, Burlington, MA, USA) bead sorter with XPonent Software version 3.1 (Luminex, Austin, TX, USA). Data were calculated using Milliplex Analyst software version 5.1 (EMD Millipore). An in-house generated rhesus macaque serum pool was run in quadruplicate as a quality control (QC). Intra-assay coefficients of variation (CVs) were <11% for all targets.

Enzyme-linked immunosorbent assay

Supernatants from samples analyzed by RNAseq were additionally analyzed for LIF by the ETSC at the ONPRC by ELISA following the manufacturer’s instructions (ThermoFisher). Two vendor-supplied QCs were run in each assay (Ctrl 1: 1179 ± 295 pg/ml; Ctrl 2: 2420 ± 606 pg/ml). The intra- and inter-assay CVs were 1.7% and 4.6%, respectively, for Ctrl 1; and 0.5% and 0.5%, respectively, for Ctrl 2. An in-house generated rhesus macaque serum pool spiked with LIF was used as a QC. Intra-assay CV was 6.9%. The assay dynamic was 280–7100 pg/ml LIF. Where indicated, IL-8 and IL-11 levels were quantified by ELISA (R&D Systems, Minneapolis, MN, USA) using similar methods, following the manufacturer’s directions.

CRISPR/Cas-9-mediated editing of IL-11RA in primary eSFs

Synthetic guide RNAs (sgRNAs) targeting the human IL-11 receptor (IL-11RA) were designed using the Synthego CRISPR Gene Knockout (KO) design tool (Synthego, Menlo Park, CA, USA). IL-11RA sgRNA (sequence: 5′-CCAUCCCGAAACCAGGACAC-3′) and negative control scrambled sgRNA (sequence: 5′-GCACUACCAGAGCUAACUCA-3′) were chemically synthesized by Synthego. For each reaction, the sgRNA (30 pmol/μl; Synthego) and Cas9 protein (20 pmol/μl; Synthego) were added at a 6:1 ratio in 23 μl P2 buffer (Amaxa P2 Primary Cell 96-well 4D-Nucleofector Kit, Lonza, Basel, Switzerland) and incubated for 20 min at 37°C. eSFs (250 000 cells/reaction, three reactions per condition) were washed with PBS, resuspended with Cas9/sgRNA RNP complex and transferred to a 16-well reaction cuvette. Electroporation was performed with the 4D-Nucleofector (Lonza; program DT-130). Following electroporation, cells were incubated in the reaction cuvette for 15 min at 37°C. Nucleofected cells were then gently collected and cultured in T-75 flasks at 37°C for 6 days and passaged for further expansion. On Day 10 post-nucleofection, eSFs were harvested for decidualization assays (see Materials and Methods above) or for DNA extraction and evaluation of editing efficiency.

PCR amplification of IL-11RA and ICE analysis

To assess KO efficiency, genomic DNA was extracted from eSFs using the QuickExtract DNA Extraction Solution as per manufacturer’s protocol (Epicentre, Madison, WI, USA). Genomic primers flanking the proposed editing site (forward: 5′-GGGGAGAGAGATAGGAGCTG-3′; reverse: 5′-CCCTATGTGTCATCAGTGCC-3′) were designed using the Primer3 online tool (http://bioinfo.ut.ee/primer3/), chemically synthesized (ELIM Biopharmaceuticals, Inc., Hayward, CA, USA) and suspended at 100 μM in water. The PCR amplification of the IL-11RA gene using Phusion High Fidelity Hot Start Polymerase x2 Mastermix (ThermoFisher) was conducted according to the manufacturer’s instructions. The thermocycler setting included one step at 98°C for 30 sec, followed by 35 cycles at 98°C for 10 sec, 68°C for 30 sec, 72°C for 30 sec, with one final step at 72°C for 5 min. PCR cleanup and capillary sequencing was performed by ELIM Biopharmaceuticals, Inc. Synthego’s ICE CRISPR Analysis Tool was utilized to determine the KO efficiency and percentage of indels from sequencing traces.

Statistical analysis

Two-tailed Student’s t-tests were conducted with Microsoft Excel 2016 (Microsoft Corporation, Redmond, WA, USA), with mean and SD reported. Differential gene expression analysis of RNAseq data was calculated by ANOVA in a paired test, using treatment as the fixed factor and ID as the random factor (Partek Flow, build version 5.0.16.1128). Statistical significance was established at P < 0.05. For pathway analysis of RNAseq data, IPA was used to identify pathways affected by differentially expressed genes (>1.5-fold change) and Z-scores >1.5 fold were included in the analyses similar to methods previously implemented (Chen et al., 2014).

Results

SP enhances decidualization of eSF from women with and without inflammatory disorders

To characterize the effects of SP on decidualization, we treated eSFs from endometrial biopsies (Supplementary Table SI) with 10% SP, in the presence or absence of E₂P₄, for up to 12 days and monitored decidualization by quantifying the production of prolactin and IGFBP1 in the conditioned media, as previously described (Chen et al., 2014). Endometrial SFs from a woman without an inflammatory disorder (‘Donor A’) were capable of decidualizing upon treatment with E₂P₄, but addition of SP further increased the levels of secreted prolactin (Fig. 1A, left) and decreased the time at which IGFBP1 levels peaked (Fig. 1A, right). By contrast, eSFs from a second donor without an inflammatory disorder (‘Donor B’) could not decidualize with E₂P₄ alone but their decidualization was rescued upon the addition of SP (Fig. 1B, left). Morphological assessment of Donor B cultures on Day 12 revealed that the combination of SP and E₂P₄ induced a change from spindle-like to polygonal/cobblestone cell shape, further confirming their decidualization (Zhu et al., 2014) (Fig. 1B, right).

Figure 1.

Seminal plasma enhances decidualization of endometrial stromal fibroblasts from women with and without inflammatory diseases. (A) Kinetic assessment of prolactin and insulin-like growth factor binding protein 1 (IGFBP1) secretion by endometrial stromal fibroblasts (eSFs) from a woman without inflammatory disease following treatment with media alone, estradiol with progesterone (E₂P₄), seminal plasma (SP) or E₂P₄ + SP. In this donor, SP increased the kinetics and extent of E₂P₄-induced decidualization. Each value corresponds to the mean ± SD of cultures set up in triplicates. (B) Kinetic and morphologic analysis of eSFs from a woman without inflammatory disease. Left: kinetic assessment of IGFBP1 secretion. Each value corresponds to the mean ± SD of triplicate cultures. Right: representative images (×100 magnification) of eSFs from Donor B after 10 days of treatment with media alone, E₂P₄, SP or E₂P₄ + SP. In this donor, decidualization only occurred in the presence of SP. (C) eSFs from women without inflammatory disease (Donors A–C), women with polycystic ovary syndrome (PCOS) (Donors D–F) or women with endometriosis (Donors G–I) were treated with media alone, E₂P₄, SP or E₂P₄ + SP. After 4 or 12 days of treatment, culture supernatants were assessed for decidualization by measuring IGFBP1 levels by ELISA. Each value corresponds to the mean ± SD of cultures set up in triplicates. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001 and ^****P < 0.0001 as determined using a two-tailed Student’s t-test comparing each treatment condition (E₂P₄, SP and E₂P₄ + SP) to the mock-treated samples. ns indicates non-significant.

We next analyzed the ability of SP to promote decidualization in eSFs from multiple donors, including healthy women, women with PCOS and women with endometriosis (Supplementary Table SI). eSFs were treated with vehicle alone, E₂P₄ alone, 10% SP alone or E₂P₄ together with 10% SP for 4 or 12 days. After 4 days of treatment, the combination of SP with E₂P₄ led to higher levels of IGFBP1 than E₂P₄ treatment alone in all eSF samples, including those from normal, PCOS and endometriosis donors (Fig. 1C, left). Furthermore, data from Day 12 of treatment demonstrated that eSFs from women that failed to decidualize in response to E₂P₄ alone (Donors B, G, H and I) could be induced to decidualize when the E₂P₄ was supplemented with SP (Fig. 1C, right). Collectively, these data suggest that SP promotes decidualization of eSFs from healthy women as well as those with inflammatory disorders, including those where E₂P₄ alone could not initiate a decidualization response.

Degradation or denaturation of seminal proteins eliminates SP-enhanced decidualization

Since the SP factor(s) promoting decidualization is >3500 Da (Doyle et al., 2012), we hypothesized it was a protein. To test this, we treated SP for 18 h with Proteinase K, which degraded the majority of seminal proteins (Fig. 2A) and then tested the ability of the Proteinase K-treated SP to promote decidualization in the presence of E₂P₄. Proteinase K treatment completely abrogated SP-enhanced decidualization in eSFs (Fig. 2B). The lack of activity was not due to cellular toxicity (Fig. 2C). To further demonstrate that the decidualization-promoting factor is a protein, we denatured SP proteins by boiling. Consistent with the Proteinase K treatment data, boiled SP combined with E₂P₄ poorly enhanced decidualization of eSFs (Fig. 2D). Boiled SP was also minimally cytotoxic to eSFs (Fig. 2E).

Figure 2.

Degradation or denaturation of seminal proteins eliminates SP-enhanced decidualization. (A) SP was mock-treated or incubated with Proteinase K (200 μg/ml) for 18 h at 37°C, followed by the addition of phenylmethylsulphonyl fluoride to inactivate the Proteinase K. Protein content was then analyzed by Coomassie staining. (B) Proteinase K treatment abrogates SP-enhanced decidualization. IGFBP1 levels were measured in supernatants of eSF cultures (Donors D and J) after 14 days of treatment with media alone, E₂P₄, SP, E₂P₄ + SP or E₂P₄ + Proteinase K-treated SP. ^*P < 0.05, ^**P < 0.01 and ^***P < 0.001, as determined using a two-tailed Student’s t-test comparing each treatment condition (E₂P₄, SP, E₂P₄ + SP and E₂P₄ + Proteinase K-treated SP) to the corresponding mock-treated samples. (C) Proteinase K-treated SP is not cytotoxic to eSFs. eSF cultures set up as described in panel B were assessed for cellular viability using the CellTiter-Glo luminescent assay. (D) Denaturation of seminal proteins by boiling diminishes SP-enhanced decidualization. IGFBP1 levels were measured in supernatants of eSF cultures (Donors C and F) after 14 days of treatment with media alone, E₂P₄, SP, E₂P₄ + SP or E₂P₄ + boiled SP. ^**P < 0.01, ^***P < 0.001 and ^****P < 0.0001 as determined using a two-tailed Student’s t-test comparing each treatment condition (E₂P₄, SP, E₂P₄ + SP and E₂P₄ + boiled SP) to the corresponding mock-treated samples. (E) Boiled SP is not cytotoxic to eSFs. eSF cultures set up as described in panel D were assessed for cellular viability using the CellTiter-Glo luminescent assay.

Semen amyloids do not promote eSF decidualization and do not induce a potent transcriptional response in eSFs or eECs

Semen amyloids are protein-derived fibrils that have recently been suggested to play a role in sperm selection during fertilization (Roan et al., 2017) and are known to enhance HIV infection (Munch et al., 2007; Roan et al., 2011; Usmani et al., 2014). To test whether semen amyloids might be responsible for SP-mediated decidualization, we chemically synthesized the highly abundant semenogelin amyloids (Roan et al., 2011, 2014). After confirming the functional activity of the synthetic amyloids by demonstrating their ability to enhance HIV infection (Supplementary Fig. S1), we tested their ability to promote decidualization of eSFs. We found that they did not (Fig. 3A) and that this was not a result of SP toxicity (Fig. 3B).

Figure 3.

Semen amyloids do not promote eSF decidualization and induce a limited transcriptional response in eSFs and endometrial epithelial cells. (A) Purified SP amyloids do not enhance decidualization. IGFBP1 levels were measured in supernatants of eSFs treated for 14 days with media alone, E₂P₄, SP or purified amyloids, for which HIV infection-promoting activity was confirmed. ^**P < 0.01 as determined by a two-tailed Student’s t-test. (B) Purified SP amyloids are not cytotoxic. eSF cultures were prepared as described in panel A and were assessed for cellular viability using the CellTiter-Glo luminescent assay. (C) Semen amyloids, unlike SP, do not induce a potent transcriptional response in eSFs or endometrial epithelial cells (eECs). eSFs or eECs were treated for 6 h with SP, SP depleted of HIV-enhancing amyloids, purified SP amyloids or the monomeric control peptide. Because SP contains a low amount of contaminating lipopolysaccharides (LPS), a negative control of LPS matching the levels measured in SP stock was also included. After treatment, samples were harvested for RNAseq analysis. Left: Principal component analysis (PCA) of eSFs treated as indicated. Right: PCA of eECs treated as indicated. (D) Heatmap illustrating hierarchical clustering of eSF and eEC samples treated as indicated. Top: Heatmap showing the most differentially expressed genes following treatment of eSFs with the indicated factors, based on RNAseq analysis. Bottom: Heatmap showing the most differentially expressed genes following treatment of eECs with the indicated factors. The hierarchical clustering was based on differentially expressed genes across treatment (2467 and 57 genes for eSFs and eECs, respectively), not taking donor ID into account. (E) Venn diagram showing the number of SP-induced genes in eSFs and eECs.

To confirm the inability of the synthetic amyloids to promote decidualization, we assessed whether they elicited any transcriptional changes in eSFs associated with decidualization. To this end, we performed global RNAseq analysis of eSFs treated for 6 h with purified SP amyloids, control monomeric peptide or SP. As another control, we also tested SP depleted of HIV-enhancing amyloids (Supplementary Fig. S1). Furthermore, because low amounts of bacteria are sometimes present in SP, we quantified the amount of LPS in our SP stock by the LAL endotoxin assay and tested the effect of an equivalent amount of pure LPS as an additional negative control. We found no evidence that the amyloids induced transcription of any genes associated with decidualization. In fact, PCA of treated eSFs revealed that while SP and SP depleted of HIV-enhancing amyloids induce a potent transcriptional response, purified amyloids, like the negative control peptide and the LPS control, did not (Fig. 3C, left).

Because eECs are also a major cell type in the endometrium and one of the main cells exposed to the SP that enters this tissue, we also assessed the effects of SP and the SP amyloids on eECs. The transcriptional response of eECs to SP and SP depleted of HIV-enhancing amyloids was attenuated relative to that observed in eSFs, with the majority of the variance due to donor (Fig. 3C, right). In both eECs and eSFs, hierarchical clustering analyses of gene expression profiles identified two major subdivisions (Fig. 3D): SP and SP depleted of HIV-enhancing amyloids and amyloids and controls (mock, monomeric peptide, LPS equivalent). Subsequent branches separated samples by donor. Treated eECs showed reduced differential gene expression to SP in comparison with eSFs (69 versus 914 genes; Fig. 3E; Supplementary Table SII), suggesting that SP induces a more potent response in eSFs than in eECs and confirming our previous microarray study (Chen et al., 2014).

To understand how decidualization may be enhanced by SP, we conducted IPA on genes that were differentially expressed between mock and SP-treated eSFs (Table I). SP treatment of eSFs induced activation of pathways related to cell survival, cell morphology, cellular development, cellular movement and gene expression. Concomitantly, pathways related to cell death were inhibited. In eECs, SP activated pathways related to cellular movement and inhibited pathways related to cell death (Table II). Selections of eSF and eEC genes that are differentially expressed in response to SP are presented in Table III, along with their associated pathways.

Table I.

Selected pathways significantly affected by SP in eSFs.

Category	Functions annotation	Activation State	Z-score
Cell Death and Survival	Cell survival	Increased	2.89
Cell Death and Survival	Cell viability	Increased	2.696
Cell Death and Survival	Necrosis	Decreased	3.215
Cell Death and Survival	Cell death	Decreased	3.072
Cell Death and Survival	Apoptosis	Decreased	2.118
Cell Death and Survival	Cell death of connective tissue cells	Decreased	1.928
Cell Morphology	Tubulation of endothelial cells	Increased	1.935
Cell Morphology	Morphogenesis of endothelial cells	Increased	1.664
Cell Morphology	Tubulation of cells	Increased	1.656
Cellular Development	Proliferation of tumor cells	Increased	2.311
Cellular Development	Differentiation of stromal cells	Increased	2.19
Cellular Development	Proliferation of cancer cells	Increased	1.852
Cellular Development	Endothelial cell development	Increased	1.775
Cellular Development	Differentiation of endothelial cells	Increased	1.709
Cellular Development	Proliferation of endothelial cells	Increased	1.656
Cellular Movement	Migration of vascular endothelial cells	Increased	2.564
Cellular Movement	Movement of vascular endothelial cells	Increased	2.547
Cellular Movement	Invasion of endothelial cells	Increased	2.537
Cellular Movement	Cell movement of endothelial cells	Increased	2.443
Cellular Movement	Migration of endothelial cells	Increased	2.293
Cellular Movement	Cell movement	Increased	2.273
Cellular Movement	Migration of cells	Increased	2.055
Cellular Movement	Chemotaxis of endothelial cells	Increased	2.012
		Increased
Gene Expression	Activation of DNA endogenous promoter	Increased	2.107
Gene Expression	Binding of DNA	Increased	1.709

P < 0.05; Z-score > |1.5|.

SP indicates seminal plasma; eSFs, endometrial stromal fibroblast.

Table II.

Selected pathways significantly affected by SP in endometrial epithelial.

Category	Functions annotation	Activation State	Z-score
Cell Death and Survival	Cell Death	Decreased	1.698
Cellular Movement	Migration of cells	Increased	2.978
Cellular Movement	Chemotaxis	Increased	2.122
Cellular Movement	Migration of endothelial cells	Increased	1.751
Cellular Movement	Movement of vascular endothelial cells	Increased	1.526

P < 0.05; Z-score > |1.5|.

Table III.

Selected genes regulated in eSFs and eECs.

eSF SP versus eSF Mock			eEC SP versus eEC Mock
Category	Gene	Fold Change	Category	Gene	Fold Change
Cell Death and Survival	IL33	74.78	Cell Death and Survival	MT1X	4.29
	CXCR4	16.81		NR4A1	3.23
	CXCL8	11.00		IL11	1.93
	PTGS2	9.65		PTGS2	1.77
	LIF	8.18		CXCL8	1.62
	STC1	7.53		IGFBP3	1.55
	VEGFA	3.96		IL1RN	1.51
	DUSP1	3.67		FOSL1	1.51
	FOXO1	3.61		ZNF43	−2.43
	HIF1A	2.7
	FGF2	2.65
	WNT5A	2.61
	BMP2	2.55
	EGFR	2.33
	IGF1R	2.21
	TP53	−2.07
	IRF1	−2.16
	JAG1	−2.18
	JUN	−2.48
	IL24	−2.74
	HEY1	−2.97
	KLF11	−3.27
	TLR4	−3.36
	PDGFB	−5.18
	IFIT3	−5.18
	IFIT2	−7.72
Category	Gene	Fold Change	Category	Gene	Fold Change
Cellular Movement	IL33	74.78	Cellular Movement	NR4A1	3.23
	IL11	58.38		NR4A2	2.68
	CXCR4	16.81		CXCL3	2.37
	CXCL8	11.00		IL11	1.93
	PTGS2	9.65		CXCL2	1.81
	LIF	8.18		PTGS2	1.77
	STC1	7.53		CCL2	1.81
	TP53	−2.07		CXCL8	1.62
	HEY1	−2.97		IGFBP3	1.55
	TLR4	−3.36
	PDGFB	−5.18
	IFIT2	−7.72

P < 0.05.

SP: seminal plasma, eSF: endometrial stromal fibroblasts, eEC: endometrial epithelial cells.

In summary, our decidualization and RNAseq analyses reveal that semen amyloids do not induce decidualization, semen amyloids signal poorly to both eSFs and eECs and SP induces a more potent transcriptional response in eSFs than in eECs.

Secretome analyses of SP-exposed eSFs and eECs

Given the importance of cytokines in decidualization and implantation (Stewart et al., 1992; Robb et al., 1998; Jasper et al., 2007), we also characterized the effects of SP and SP components on the secretion of 32 different growth factors and cytokines by eSFs. To distinguish between cytokines secreted by eSFs in response to SP from endogenous cytokines present in SP, empty wells (without cells) were treated in parallel with SP (‘SP alone’) as a negative control. Secreted growth factors and cytokines were quantified through a combination of Luminex and ELISA. We found that treatment of eSFs with SP and SP depleted of HIV-enhancing amyloids increased (P < 0.05) the secretion of EGF, FGFβ, HGF, IFN-γ, IL-8, MCP-1, MIP-1β and VEGF (Table IV). In addition, we also found SP upregulated secretion of LIF, IL-6 and IL-11 (Table IV; Fig. 4A, B), which have known roles in implantation (Stewart et al., 1992; Robb et al., 1998; Jasper et al., 2007).

Table IV.

The effect of SP components on the production of cytokines in eSF.

Analytes	Vehicle	SP	SP(depleted) a	Peptide b	Amyloid c	LPS equivalent	SP, no eSF
EGF	2.97 ± 0.61	42.52 ± 3.52 *	43.67 ± 4.11 *	3.56 ± 0.41	3.32 ± 0.71	ND	ND
Eotaxin	ND	ND	ND	ND	ND	ND	160.23
FGFβ	6.15 ± 1.21	21.00 ± 0.71 *	21.68 ± 3.47 *	ND	ND	6.97 ± 1.86	ND
G-CSF	ND	ND	ND	ND	ND	ND	15.16
GM-CSF	ND	ND	ND	ND	ND	ND	ND
HGF	9.64 ± 2.04	61.80 ± 21.01 *	46.52 ± 11.38 *	8.46 ± 2.044	8.46 ± 2.04	7.85 ± 0.99	ND
IFN-⍺	ND	11.39 ± 2.76	10.19 ± 2.08	ND	ND	ND	23.79
IFN-γ	3.49	18.94 ± 3.30 *	20.71 ± 0.89 *	3.49	3.49	ND	ND
IL-10	ND	4.92 ± 1.13	5.23 ± 1.17	ND	ND	ND	16.63
IL-11	206.70 ± 111.57	4520.96 ± 821.96 *	4480.81 ± 591.74 *	ND	ND	ND	86
IL-12	ND	7.43 ± 0.82	7.19 ± 0.71	ND	ND	ND	3.3
IL-13	ND	19.16 ± 4.28	20.13 ± 2.34	ND	ND	ND	6.48
IL-15	ND	48.24 ± 6.27	50.77 ± 7.63	ND	ND	ND	16.69
IL-17	ND	ND	ND	ND	ND	ND	52.36
IL-1β	ND	13.12 ± 1.72	11.99 ± 1.56	ND	ND	ND	ND
IL-1RA	ND	38.37 ± 1.89	34.04 ± 1.86	ND	ND	ND	7.57
IL-2	ND	ND	ND	ND	ND	ND	42.76
IL-2R	ND	28.05	24.92 ± 3.13	ND	ND	ND	ND
IL-4	ND	ND	ND	ND	ND	ND	ND
IL-5	ND	ND	ND	ND	ND	ND	ND
IL-6	ND	87.50 ± 13.06	106.73 ± 45.92	ND	ND	ND	ND
IL-7	ND	143.13 ± 48.24	72.03 ± 7.98	ND	7.61	8.63 ± 3.33	19.78
IL-8	516.51 ± 457.14	6497.67 ± 2236.15 *	4654.00 ± 1298.72 *	78.56 ± 238.6	349.19 ± 292.53	411.83 ± 314.94	421.71
IP-10	ND	ND	ND	ND	ND	ND	49.34
LIF	ND	1310.66 ± 344.29	1160.45 ± 567.06	ND	ND	ND	ND
MCP-1	198.17 ± 248.15	572.59 ± 575.39 *	399.31 ± 17.53 *	145.45 ± 180	139.06 ± 169.34	177.75 ± 224.34	113.08
MIG	ND	74.01 ± 15.74	49.21 ± 3.38	ND	ND	ND	81.52
MIP-1⍺	ND	22.88 ± 1.74	23.27	ND	ND	ND	18.62
MIP-1β	2.60 ± 0.38	20.06 ± 4.29 *	17.28 *	2.40 ± 0.722	2.82	2.82	11.32
RANTES	ND	ND	ND	ND	ND	ND	ND
TNF-⍺	ND	ND	ND	ND	ND	ND	ND
VEGF	6.03 ± 3.70	213.57 ± 36.40 *	239.80 ± 39.50 *	5.49 ± 3.356	6.09 ± 4.01	6.26 ± 3.42	196.82

^{a a}Seminal plasma devoid of amyloids.

^{b b}Monomeric peptide.

^{c c}Synthesized human SEM1(86–107) amyloid fibrils.

LPS: lipopolysaccharide^*P < 0.05eSF (pg/ml)

Figure 4.

Secretome analysis of eSFs following SP treatment reveals a potent upregulation of interleukin-11, a cytokine crucial for implantation. (A) eSFs from three donors (Donors K–M) were treated for 6 h with media alone, SP or SP depleted of HIV-enhancing amyloids, after which cell supernatants were collected for Luminex analysis. Both SP and SP depleted of HIV-enhancing amyloids induced secretion of multiple cytokines in eSFs (see Table IV for full list). To control for endogenous levels of the cytokines, the equivalent concentration of SP (‘SP alone’) was also assayed in parallel. ^*P < 0.05 as determined using a two-tailed Student’s t-test comparing each treatment condition (SP and SP(depleted)) to the corresponding mock-treated samples. nd indicates not detectable. (B) eSF culture supernatants set up as described in panel A were assayed for interleukin (IL)-11 levels by ELISA. ^**P < 0.01 as determined using a two-tailed Student’s t-test comparing each treatment condition (SP and SP(depleted)) to the mock-treated samples. (C) eSFs were treated with media alone, estradiol with progesterone (E₂P₄), SP or E2 + SP. After 4 or 12 days of treatment, culture supernatants were assessed for decidualization by measuring IGFBP1 levels by ELISA. ^**P < 0.01 and ^***P < 0.001 as determined using a two-tailed Student’s t-test comparing the E₂P₄ and E₂P₄ + SP treatment conditions.

Given the particularly important role of IL-11 in both decidualization and implantation (Robb et al., 1998; Dimitriadis et al., 2002; Karpovich et al., 2005; Menkhorst et al., 2009), we focused on it further by investigating whether SP-induced secretion of IL-11 occurs during E₂P₄-mediated decidualization. As shown in Fig. 4C, E₂P₄ alone induced only low levels of IL-11 production consistent with a prior report (Dimitriadis et al., 2002). However, SP in the presence of E₂P₄ potently stimulated eSF secretion of IL-11, both at 4 (P < 0.001 versus E₂P₄) and 12 days (P < 0.01 versus E₂P₄; Fig. 4C) post-treatment.

Induction of IL-11 by SP is mediated by a proteinaceous component associated with MVs

We postulated that SP-induced IL-11 secretion by eSF played a critical role in SP’s ability to promote decidualization. If this were true, then SP that had been devoid of proteins, which lack the ability to promote decidualization (Fig. 2), should not induce IL-11. To test this, we measured IL-11 levels in culture supernatants of Proteinase K-treated SP. SP devoid of proteins was unable to elicit IL-11 secretion (Fig. 5A), consistent with IL-11 signaling as an important pathway in SP-enhanced decidualization.

Figure 5.

SP-mediated induction of IL-11 is mediated by proteins associated with microvesicles. (A) eSFs from three donors (Donors D, J and Q) were treated for 6 h with vehicle, SP or SP treated with Proteinase K to determine whether seminal proteins are necessary for IL-11 induction. IL-11 levels in culture supernatants were quantified by ELISA. ^*P < 0.05 as determined using a two-tailed Student’s t-test comparing the SP and the SP + Proteinase K conditions. (B) eSFs from five donors (Donors B and Q–T) were treated for 6 h with vehicle, the microvesicle (MV)-enriched fraction, or reconstituted SP, and cell supernatants were assessed for IL-11 levels by ELISA. ^**P < 0.01 as determined using a two-tailed Student’s t-test comparing the MV-enriched treatment to the mock-treated samples. (C) eSFs from three donors (Donors B, Q and R) were treated for 6 h with vehicle, the MV-enriched fraction, or reconstituted SP, and cell supernatants were assessed for IL-8 levels by Luminex. (D) eSFs from two donors (Donors C and T) were treated with media alone, E₂P₄, the MV-enriched fraction or E₂P₄ + the MV-enriched fraction. After 4 or 12 days of treatment, culture supernatants were assayed for IL-11 levels by ELISA. ^*P < 0.05, ^**P < 0.01 and ^****P < 0.0001 as determined using a two-tailed Student’s t-test comparing treatment condition E₂P₄ to E₂P₄ + the MV-enriched fraction.

Proteins in SP can exist in either a soluble form or associated with EVs, including large MVs (typically 50–1000 nm in diameter) (Machtinger et al., 2016). To determine whether the SP proteins with decidualization capacity may be MV-associated, we assessed whether SP enriched for large MVs (220–650 nm) induce IL-11 secretion by eSF. Indeed, fractionated SP enriched for MVs induced IL-11 secretion by eSF at a level of ~57% of that induced by SP (Fig. 5B). This effect was specific for IL-11 since other cytokines, such as IL-8, were highly induced by SP but not by the MV-enriched fraction (compare Figs 4A and 5C). We further confirmed that the ability of the MV-enriched fraction to upregulate IL-11 secretion by eSF also occurred in the presence of E₂P₄ (Fig. 5D).

The MV-enriched fraction of SP induces decidualization and an intermediate transcriptional response in eSFs

In addition to upregulating IL-11 in the presence of E₂P₄, the MV-enriched fraction also promoted E₂P₄-mediated eSF decidualization (Fig. 6A). We then assessed the global effects of the MV-enriched fraction on eSFs. eSFs were treated for 6 h with vehicle, the MV-enriched fraction, or reconstituted SP (comprised a solution where the MV-enriched and non-MV fractions were added back together to reconstitute the SP; this was used instead of SP to account for any effect fractionation may have had on the signaling effects of SP). RNAseq analysis was then conducted on the specimens. PCA revealed that the MV-enriched fraction elicited a transcriptional response intermediate between that of vehicle and reconstituted SP (Fig. 6B). Hierarchical clustering analyses indicated that cells treated with the MV-enriched fraction clustered more closely with cells exposed to reconstituted SP than to vehicle-treated eSFs (Fig. 6C). Furthermore, the majority of transcripts induced by the MV-enriched fraction overlapped with those induced by reconstituted SP (347 genes in common; Fig. 6D; Supplementary Table SIII).

Figure 6.

MV-enriched fraction from SP induces a potent transcriptional response and decidualization in eSFs. (A) eSFs were treated with media alone, E₂P₄, the MV-enriched fraction or E₂P₄ + the MV-enriched fraction. After 4 or 12 days of treatment, culture supernatants were assessed for decidualization by measuring IGFBP1 levels by ELISA. Each value corresponds to the mean ± SD of cultures set up in triplicates. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001 and ^****P < 0.0001 as determined using a two-tailed Student’s t-test comparing each treatment condition (E₂P₄, the MV-enriched fraction or E₂P₄ + the MV-enriched fraction) to the mock-treated samples. (B) PCA of RNAseq data from eSFs treated for 6 h with vehicle, the MV-enriched fraction or reconstituted SP. (C) Heatmap illustrating hierarchical clustering of eSFs treated with vehicle, the MV-enriched fraction, or reconstituted SP, based on the RNAseq datasets. Hierarchical clustering was based on differentially expressed genes across treatment (281 genes), not taking donor ID into account. (D) Venn diagram showing the number of transcripts induced by the MV-enriched fraction and reconstituted SP.

IPA analysis of differentially expressed genes revealed similar regulated biological pathways between cells treated with the MV-enriched fraction versus reconstituted SP (P < 0.05, Z-score > 1.5; Table V). Following both treatments, pathways related to cell survival, cell movement and cellular development increased while pathways related to cell death decreased (Table V). Selected differentially expressed genes and their associated pathways affected by the MV-enriched fraction and reconstituted SP in eSFs are presented in Table VI.

Table V.

Selected pathways significantly affected by microvesicles in eSF.

Reconsituted SP versus Mock				MV versus Mock
Category	Functions annotation	Activation State	Z-score	Category	Functions annotation	Activation State	Z-score
Cell Death and Survival	Cell survival	Increased	2.485	Cell Death and Survival	Cell survival	Increased	2.703
Cell Death and Survival	Cell viability	Increased	2.125	Cell Death and Survival	Cell viability	Increased	2.186
Cell Death and Survival	Cell viability of endothelial cells	Increased	2.101	Cell Death and Survival	Cell viability of endothelial cells	Increased	1.869
Cell Death and Survival	Cell death	Decreased	3.45	Cell Death and Survival	Cell death	Decreased	2.143
Cell Death and Survival	Necrosis	Decreased	3.141	Cell Death and Survival	Necrosis	Decreased	2.292
Cell Death and Survival	Apoptosis	Decreased	2.564	Cell Death and Survival	Apoptosis	Decreased	1.893
Cell Death and Survival	Cell death of tumor cell lines	Decreased	2.232
Cellular Movement	Cell movement	Increased	2.964	Cellular Movement	Cell movement	Increased	2.212
Cellular Movement	Migration of cells	Increased	2.886	Cellular Movement	Migration of cells	Increased	2.033
Cellular Movement	Chemotaxis of endothelial cells	Increased	1.952	Cellular Movement	Chemotaxis of endothelial cells	Increased	1.952
Cellular Movement	Migration of tumor cell lines	Increased	1.859	Cellular Movement	Chemotaxis	Increased	1.658
Cellular Movement	Invasion of endothelial cells	Increased	1.756
Cellular Movement	Movement of vascular endothelial cells	Increased	1.573
Cellular Movement	Chemotaxis	Increased	1.509
Cellular Development	Proliferation of tumor cells	Increased	2.499	Cellular Development	Proliferation of tumor cells	Increased	1.634
Cellular Development	Differentiation of stromal cells	Increased	1.96	Cellular Development	Tubulation of endothelial cells	Increased	1.585
Cellular Development	Tubulation of endothelial cells	Increased	1.79	Cellular Development	Differemtiation of endothelial cells	Increased	1.709
Cellular Development	Differentiation of endothelial cells	Increased	1.709
Cellular Development	Proliferation of cancer cells	Increased	1.649
Cellular Development	Differentiation of epithelial tissues	Increased	1.51
Cellular Growth and Proliferation	Colony formation of cells	Increased	3.527
Gene Expression	Binding of DNA	Increased	2.239

P < 0.05; Z-score > |1.5|.

MV indicates microvesicles.

Table VI.

Selected genes regulated in eSF by MV.

Reconstituted SP versus Mock			MV versus Mock
Category	Gene	Fold Change	Category	Gene	Fold Change
Cell Death and Survival	SST	379.6	Cell Death and Survival	SST	385.22
	IL11	24.65		IL11	12.98
	IL33	21.19		CCL7	11.99
	CCL7	14.06		IL33	7.49
	NR4A2	5.14		CCL2	4.17
	CCL2	5.05		PTGS2	3.00
	PTGS2	4.21		FGF7	2.83
	FGF7	3.42		FOXO1	2.13
	HIF1A	2.58		HIF1A	2.01
	STC1	2.51		WNT5A	1.98
	VEGFA	2.45		IFIT3	−2.05
	DUSP1	2.37		ICAM1	−2.07
	FOXO1	2.37		TLR4	−2.35
	EGFR	2.35		JUN	−2.46
	WNT5A	2.28		IFIT2	−3.1
	FGF2	2.07		IL24	−3.21
	IRF1	−1.95		EGR3	−4.79
	PDGFB	−2.16
	ICAM1	−2.21
	KLF11	−2.25
	IFIT3	−2.84
	JUN	−3.07
	IL24	−3.31
	BMF	−3.34
	EGR3	−5.58
	IFIT2	−5.16
Cellular Movement	IL11	24.65	Cellular Movement	IL11	12.98
	IL33	21.19		CCL7	11.99
	CCL7	14.06		IGFBP1	8.53
	IGFBP1	7.07		IL33	7.49
	CCL2	5.05		CCL2	4.17
	FGF7	3.42		PTGS2	3.00
	VEGFA	2.45		FGF7	2.83
	FOXO1	2.37		ICAM1	−2.07
	EGFR	2.35		KLF2	−2.14
	WNT5A	2.27		TLR4	−2.35
	PDGFB	−2.16		IFIT2	−3.1
	IFIT2	−5.16		RGS4	−6.84

P < 0.05.

IL-11 signaling in eSFs is required for SP-promoted decidualization

The data presented thus far suggest that proteinaceous factors associated with seminal MVs promote decidualization of eSFs, possibly through the induction of IL-11 secretion. To directly assess a role for IL-11, we used CRISPR gene-editing technology to knock out the IL-11RA in eSFs and disrupt the cells’ ability to respond to IL-11 signaling. Using specific guide RNAs for IL-11RA, we knocked out IL-11RA in eSFs from two donors with PCOS (Donors W and X). Efficient gene editing (average 95.5% indel and 92% knockout efficiency) was demonstrated by inference of CRISPR edit (ICE) analysis (Fig. 7A). Control eSFs (treated with non-targeting guides) and IL-11RA-knockout eSFs were treated with SP, in the presence or absence of E₂P₄, and decidualization was monitored by the production of IGFBP1 and morphological characterization. As expected, vehicle-treated eSFs as well as eSFs treated with scrambled guides decidualized in response to 10% SP and 10% SP + E₂P₄ (Fig. 7B). In comparison, in the presence of either 10% SP or 10% SP + E₂P₄, IL-11RA eSF knockout lines secreted lower levels of IGFBP1 on both Days 4 and 12 post-treatment (Fig. 7B). Similar results were observed when treatment was performed with 1% SP (Supplementary Fig. S2). These results suggest that IL-11 signaling in eSFs is necessary for SP-enhanced decidualization.

Figure 7.

CRISPR/Cas-9 mediated knockout of IL-11 receptor reveals IL-11 signaling is important for SP-enhanced decidualization. (A) Average percentage of indels and knockout (KO) efficiency at the interleukin-11 receptor (IL-11RA) locus in eSFs treated with the IL-11RA guide or the scrambled control, as determined by ICE analyses following Clustered regularly interspaced short palindromic repeats (CRISPR)/Cas-9-mediated gene editing. (B) IL-11RA KO eSFs, scrambled negative control eSFs and unedited (untreated) eSFs were treated with media alone, E₂P₄, 10% SP or E₂P₄ + 10% SP. After 4 or 12 days of treatment, culture supernatants were assessed for decidualization by measuring IGFBP1 levels by ELISA. Each value corresponds to the mean ± SD of cultures set up in triplicates. ^*P < 0.05, ^**P < 0.01 and ^****P < 0.0001, as determined by two-tailed Student’s t-test, indicate a significant decrease in IGFBP1 levels in the IL-11RA KO as compared with the scrambled negative control.

Discussion

While beneficial effects of SP on implantation are recognized (Robertson et al., 2016), the underlying molecular mechanisms are not well understood. In this study, we have used in vitro assays to characterize how SP signals to eSFs. Our major findings are as follows: SP promotes decidualization of eSFs from healthy women as well as those with PCOS and endometriosis; MV-associated SP proteins are, in part, responsible for this decidualization-promoting activity; the SP proteins promoting decidualization induce transcriptional responses in eSFs associated with cellular differentiation, including potent induction of IL-11; and the ability of SP to support decidualization is dependent on IL-11 signaling. These data support the notion that SP may promote implantation by either initiating or increasing the efficiency of decidualization.

SP promotes ex vivo eSF decidualization

Doyle et al. (2012) previously reported that dialyzed SP accelerates and increases the extent of P₄-induced decidualization in eSFs, providing a possible explanation for how SP may increase implantation rates. In this study, we used both E₂ and P₄, as this more closely mimics the in vivo development of a receptive endometrium during the luteal phase, which requires the concerted action of both E₂ and P₄ (Cha et al., 2012). Our findings confirm that SP promotes decidualization and further demonstrate that this effect also holds true in eSFs from women with PCOS and endometriosis whose eSFs are more commonly resistant to E₂P₄-mediated decidualization. Furthermore, we found multiple patterns of SP effects on E₂P₄-mediated decidualization, including enhanced kinetics and instances where eSFs that normally do not decidualize with E₂P₄ can be rendered capable to do so in the presence of SP. These results demonstrate that SP can both enhance and, in some cases, potentiate E₂P₄-mediated eSF decidualization.

What in SP is promoting decidualization?

Given that the factor(s) in SP promoting decidualization is >3500 Da (Doyle et al., 2012) and SP proteins are typically this size, it was perhaps not surprising to find that protein degradation or denaturation eliminated SP-enhanced decidualization in eSFs. On the other hand, the fact that amyloids, a very abundant class of proteins in SP, did not mediate this effect was somewhat unexpected since other amyloids have been reported to induce IL-6, whose family members have been associated with the decidualization response (Rensink et al., 2002; Galimberti et al., 2008; Griffin et al., 2016). Unlike other amyloids that signal potently to cells (Meda et al., 1995; Hu et al., 1998), semen amyloids are distinct in that they are produced in a non-disease state (Usmani et al., 2014) and, as demonstrated in this study, are relatively inert with regards to signaling a transcriptional response in both eSFs and eECs. It is possible that the role of semen amyloids in the FRT is more structural in nature and limited to promoting phagocytosis of excess and/or defective sperm, as recently suggested (Roan et al., 2017).

The proteinaceous moiety responsible for SP-induced decidualization seems, at least in part, associated with MVs. SP proteins can exist in a soluble form but many are also packaged in EVs, which consist of larger MVs (typically 50–1000 nm in diameter) (Machtinger et al., 2016) and smaller exosomes (typically 40–100 nm in diameter) (Vojtech et al., 2014). SP EVs originate from several cell types in the male reproductive tract (Renneberg et al., 1997; Vojtech et al., 2014) and have been documented to fuse with the sperm membrane to assist with survival in the acidic vaginal environment (Arienti et al., 1997). In this study, we focused on larger MVs and found that an SP fraction enriched for MVs promotes eSF decidualization. Whether MVs fuse with eSFs to deliver a decidualization signal is not known but would be of interest to examine in future studies.

Molecular pathways underlying SP-enhanced decidualization

To better understand the mechanisms by which SP promotes decidualization, we assessed how eSFs respond to treatment with SP and SP components with or without decidualization-promoting activity. To characterize the signaling pathways in eSFs that are elicited upon treatment with SP or SP constituents, we conducted global gene expression analysis and Luminex assays. Consistent with our prior microarray study (Chen et al., 2014), SP treatment resulted in a potent transcriptional response in eSFs, eliciting signaling pathways related to cell death and survival and cellular development and movement. We also assessed the effects of the MV-enriched fraction on eSFs and observed an intermediary transcriptional response as compared with SP. Since the MV-enriched fraction was sufficient to promote decidualization, the signaling pathways associated with transcripts induced by the MV-enriched fraction may be the ones responsible for enhancing the decidualization response. These pathways include those related to cell survival and cellular movement.

The RNAseq datasets generated in this study of how eSFs respond to SP and specific SP constituents provide a useful resource for the field. In comparison with the previous microarray datasets examining how eSFs respond to SP exposure (Chen et al., 2014), the RNAseq datasets generated here provide a more unbiased assessment of transcripts, since RNAseq technology is independent of probe availability and intensity of transcript expression (Wang et al., 2009). Furthermore, RNAseq has a larger dynamic range and can detect low abundance transcripts, transcript variants and novel transcripts, which is not possible with probe hybridization in microarrays. Compared with our prior microarray study (Chen et al., 2014), we identified additional transcripts affected by unfractionated SP and MV-enriched SP in eSFs, including those with essential roles in uterine receptivity and implantation. For example, RNAseq demonstrated that both unfractionated and MV-enriched SP stimulated the expression of FOXO1 in eSFs, which has recently been defined as critical for embryo implantation through downregulation of progesterone receptor (PGR) in the uterine epithelium (Vasquez et al., 2018). The FOXO1-PGR interaction has also been documented as important for in vitro eSF decidualization (Labied et al., 2006; Takano et al., 2007; Vasquez et al., 2015). Another transcript induced by SP and uniquely identified by RNAseq was Stanniocalcin-1 (STC1), which is upregulated in endometrial fluid during the mid-secretory phase in fertile women and increases during ex vivo eSF decidualization but is downregulated in endometriosis (Aghajanova et al., 2016).

We also identified differentially expressed transcripts that were shared between the microarray and RNAseq datasets, thus confirming our prior analyses. One transcript upregulated by SP in both the microarray and RNAseq datasets was IL-33, which has previously been associated with the decidual response (Salker et al., 2012). Furthermore, altered IL-33 signaling is observed in decidualizing eSFs from women with recurrent pregnancy loss and pretreatment of mice uteri with conditioned media from those cultures results in pregnancy failure (Salker et al., 2012), supporting the notion that SP promotes eSFs decidualization and can potentially facilitate implantation.

Of most interest in our microarray and RNAseq datasets, however, was the potent induction of IL-11 transcription (3.9-fold as assessed by microarray, >74-fold as assessed by RNAseq) by SP in eSFs. Consistent with the transcriptomic datasets, IL-11 protein was secreted by eSF at significantly higher levels following treatment with SP in both the presence (>25-fold) and absence (>32-fold) of E₂P₄ at 12 days post-treatment. Moreover, it was also induced by the MV-enriched fraction of SP that is sufficient to promote decidualization. In contrast, semen amyloids, which did not promote decidualization, did not induce IL-11 secretion in eSFs (data not shown). To directly examine the role of IL-11 signaling in SP-enhanced decidualization, we abrogated the ability of eSFs to respond to IL-11 signaling by mutating IL-11RA via CRISPR/Cas9 technology. To our knowledge, this is the first reported instance of successful application of CRISPR/Cas9 technology to disrupt genes in primary eSFs. We found that knockout of IL-11RA in eSFs diminishes SP-induced decidualization, suggesting that SP does not overcome the requirement for IL-11 signaling in eSF decidualization. Future studies should establish if SP exposure after coitus induces similar effects in vivo and the identity of the seminal protein(s) responsible for such effects.

Inflammation and SP-induced decidualization

The mechanism by which SP restores endometrial decidualizing ability in women with inflammatory disorders is unclear, and at face value may seem at odds with the notion that SP harbors pro-inflammatory cytokines and generally elicits an inflammatory response in cells of the FRT. We speculate that the reason SP can elicit a decidual response is because pro-inflammatory factors including cytokines are pleiotrophic in their actions and can be pro-inflammatory in one setting yet anti-inflammatory in another. Furthermore, decidualization is biphasic and involves an acute-phase inflammatory response followed by an anti-inflammatory response required for embryo implantation (Salker et al., 2012). For example, both IL-11 and IL-33 are important for initiating decidualization (Dimitriadis et al., 2002, 2006; Menkhorst et al., 2009; Salker, et al., 2012), but if these cytokines are elevated later in pregnancy, complications can occur (Salker et al., 2012; Winship, et al., 2015). Furthermore, while PCOS eSFs that do not decidualize in response to steroid hormones secrete increased levels of IL-6 and IL-8 relative to eSFs from control women, exogenously added IL-6 and/or IL-8 do not inhibit decidualization in the control eSFs (Piltonen et al., 2015). These results suggest that high levels of these inflammatory cytokines in non-decidualizing PCOS samples are likely not the cause for impaired decidualization. We postulate that SP does not amplify the inflammation already present in eSFs from women with PCOS or endometriosis but rather elicits a different type of response that promotes decidualization.

Under what in vivo conditions might SP be able to access the endometrium and promote decidualization?

For SP to induce decidualization of the endometrium in vivo, semen and SP components must reach the upper FRT, which may occur through peristalsis (Egli et al., 1961; Leyendecker et al., 1996; Kunz et al., 1997; Wildt et al., 1998; Barnhart et al., 2001). Once semen components reach the uterine lumen, they would need to access the underlying eSFs to promote decidualization. This could occur during P₄-dominant hormonal conditions of the secretory phase when permeability of the endometrial epithelium increases and luminal contents can gain increased access to the stroma (Someya et al., 2013; Goddard et al., 2014; Neidleman et al., 2017). Another possibility is that SP proteins deposited in the vagina access eSFs via countercurrent exchange from vaginal venous blood into uterine arterial blood, as has been shown with steroid hormones (Cicinelli et al., 1999, 2000, 2001, 2003).

The timing when eSFs encounter SP may also be important. Morphological changes associated with decidualization are evident ~9 days post-ovulation in humans (Gellersen et al., 2014). Because SP components are largely cleared from the lower FRT (Pandya et al., 1985) and few sperm are present in the uterus (Rubenstein et al., 1951; Moyer et al., 1970) by 24 h post-coitus, it is unlikely that the SP carrying the fertilizing sperm is also the source of SP that promotes in vivo decidualization; rather, the latter SP would likely arise from a subsequent coital event. Thus, we speculate that repeated SP exposure during the secretory phase will increase the efficiency of decidualization, but the extent to which this occurs in vivo remains to be determined. Alternatively, it is also possible that SP proteins could be sequestered within the endometrial epithelial glycocalyx to act days after a coital event. However, it remains to be determined whether coital events in the proliferative phase results in premature decidualization of the eSFs.

Limitations

One limitation of this study is that our in vitro culture system did not incorporate paracrine interactions between eECs and eSFs, which eECs require to differentiate in response to ovarian hormones that vary throughout the menstrual cycle (Cunha, 1976; Cunha et al., 1983). This could explain why in our experimental conditions, SP treatment in the absence of E₂P₄ did not induce as many transcriptional changes in eECs as in eSFs. Nonetheless, SP in the absence of steroid hormones did induce transcriptional changes in eECs (69 genes), with MT1G (metallothionein 1G) as one of the most highly upregulated genes. MT1G is a P₄-regulated gene that is highly increased in the secretory-phase endometrium of women without disease (Talbi et al., 2006) and is decreased in the early secretory-phase endometrium of women with endometriosis (Burney et al., 2007). Thus, eECs do respond to SP in a manner that is likely to support decidualization and endometrial receptivity, and future studies to better characterize the effects of SP on eECs are warranted.

Conclusion

In conclusion, our results suggest that MV-associated proteins actively signal to cells of the endometrium and can promote decidualization of eSFs from women with and without inflammatory disorders through induction of IL-11.

Future directions

Future studies should focus on determining the identity of the protein(s) in SP responsible for initiating and potentiating decidualization in eSFs. Because some proteins/peptides with previously described direct or indirect roles in decidualization—including transforming growth factor-β (Cork et al., 2001), relaxin (Huang et al., 1987), activin A (Jones et al., 2002), EGF (Irwin et al., 1994), FSH (Tang et al., 1993) and laminin (Brar et al., 1995)—are present in semen; one or more of these proteins may be responsible for SP-induced decidualization.

It will also be important to more precisely define the cellular mechanisms of SP-induced decidualization, particularly in the context of endometrial disorders and compromised decidualization. One candidate pathway downstream of IL-11/IL-11RA signaling is the Janus kinase (JAK)-signal transducer and activator of transcription (STAT) signaling cascade (Negahdaripour et al., 2016), since the activation of JAK results in STAT3 phosphorylation and activation, which is required for decidualization and implantation (Catalano et al., 2005; Dimitriadis et al., 2006; Wang et al., 2012; Pawar et al., 2013; Sun et al., 2013).

While SP may not be required to achieve fertilization or implantation, our results suggest it may enhance reproductive success, particularly in women with decidualization deficiencies. Indeed, women with infertility have been found to have compromised IL-11 production and eSF decidualization in vitro (Karpovich et al., 2005). Additionally, even if women with decidualization deficiencies manage pregnancy with ART, abnormalities in decidualization can lead to miscarriage, aberrant placentation and preterm birth (Cha et al., 2012; Schatz et al., 2016). Should follow-up studies demonstrate similar effects of SP on decidualization in vivo, it would support the notion that exposure of the endometrium of women to SP or SP components prior to standard IVF procedures may rescue compromised endogenous eSF decidualization disorders and enhance pregnancy establishment and success.

Authors’ roles

A.F.G.: study design, sample preparation and experiments, data analysis, figure preparation and manuscript preparation; K.S.J.: study design, sample preparation and experiments, data analysis and figure preparation; M.N.: data analysis and manuscript revision; J.N.: study design and manuscript revision; T.S.: sample collection; G.X.: sample preparation, technical troubleshooting and manuscript revision; J.C.: study design, data interpretation and manuscript revision; E.H.: experiments; A.L.: technical troubleshooting and manuscript revision; E.G.M.: sample preparation and manuscript revision; S.H.: sample selection and collection; C.D.P.: sample selection and collection; P.J.N.: study design; M.R.J.: study design; W.C.G.: study design, technical troubleshooting and manuscript revision; L.C.G.: study design, technical troubleshooting, clinical interpretations and manuscript revision; N.R.R.: corresponding author, study design, technical troubleshooting, data analysis and interpretation and manuscript preparation.

Funding

National Institutes of Health (R21AI116252, R21AI122821 and R01AI127219) to N.R.R. and (P50HD055764) to L.C.G.

Conflict of interest

None declared.

The views expressed in this article reflect the results of research conducted by the authors and do not necessarily reflect the official policy or position of the Department of the Navy, Department of Defense, or the United States Government. Coauthor T.S. is a military service member, and her contribution to this work was prepared as part of her official duties. Title 17 U.S.C. 105 provides that “Copyright protection under this title is not available for any work of the US Government.” Title 17 U.S.C. 101 defines a US Government work as a work prepared by a military service member or employee of the US Government as part of that person’s official duties.