Leveraging menstrual effluent (ME) to investigate vitamin D and human decidualization, a potential link with fertility

Abstract

Context:

Low vitamin D levels have been associated with female infertility, but the underlying mechanisms are not understood.

Objectives:

To examine vitamin D treatment, in vitro and in vivo, on human endometrial stromal cell (eSC) decidualization, including mediating pathways. To assess feasibility of vitamin D measurement in menstrual effluent (ME).

Design:

ME, ME-eSCs, and peripheral plasma were collected from participants in the Research Outsmarts Endometriosis (ROSE) study (2017–2024) and the Investigation of Vitamin D and Menstrual Cycles trial (inVitD) (2022–2024; NCT05050916).

Setting:

ROSE study: North America; inVitD study: Detroit, MI and Durham, NC.

Participants:

Healthy females aged 19–40 years. The inVitD trial participants provided ME pre- and post-cholecalciferol supplementation.

Intervention:

Oral cholecalciferol supplementation (50,000 IU/week or 4,200 IU/week × 3 months) was provided to vitamin D deficient inVitD trial participants.

Main outcome measure:

First, we examined the effect of active vitamin D (calcitriol) on ME-eSC decidualization in vitro. Second, we compared ME-eSC decidualization pre- and post-oral cholecalciferol supplementation. Mechanistic studies investigated target signaling proteins and mRNA expression. Third, we compared levels of vitamin D biomarkers in peripheral plasma versus ME.

Results:

Calcitriol increased eSC decidualization measured by IGFBP1 (p<0.0001) and prolactin levels (p<0.0001). Calcitriol and decidualization alone induced VDR mRNA expression. Calcitriol reduced AKT and PRAS40 phosphorylation during decidualization. Decidualization capacity increased following in vivo cholecalciferol supplementation (p=0.07). Peripheral plasma 1,25(OH)₂D levels were correlated with ME levels (r=0.66, p=0.01).

Conclusions:

Vitamin D enhances eSC decidualization, likely through reduced AKT signaling and downstream events, highlighting its potential role in fertility. Vitamin D biomarkers were measurable in ME.

Introduction

Globally, there is a growing prevalence of infertility ¹ and factors that affect fertility are not well-understood. Menstrual cycles are a vital sign ² and menstrual cycle disturbances may predict difficulties in conceiving a pregnancy.³ While known for its role in bone health, the role of vitamin D in female reproductive health is not completely understood, vitamin D deficiency has been associated with infertility, polycystic ovary syndrome (PCOS), and endometriosis.^4,5

The accepted clinical biomarker of vitamin D status is total 25-hydroxyvitamin D (25(OH)D), the sum of 25(OH)D₃ and 25(OH)D₂.⁶ (Supplemental Table 1)⁷ The Institute of Medicine suggests that almost all people maintain healthy bones with a circulating total 25(OH)D level of at least 20 ng/ml.⁶ However, low levels of 25(OH)D are common in the U.S., especially among reproductive-aged non-Hispanic Black women. In the most recent data available (2015–2018), the median 25(OH)D levels for this group were around 15 ng/ml for ages 12–39 and 25% have levels of 10 ng/ml or lower.⁸

Vitamin D₃, or cholecalciferol, is produced in the skin in response to ultraviolet B radiation and can also be obtained from the diet or supplements. Cholecalciferol is hydrolyzed to the circulating form 25(OH)D₃, calcifediol, which can again be hydrolyzed to the active form, 1,25(OH)₂D₃, or calcitriol.^6,9 The active form of vitamin D binds to the vitamin D receptor (VDR) which is transported into the nucleus where it promotes gene transcription.¹⁰ The VDR is expressed in numerous tissues, including the brain, ovary, and placenta,¹¹ as well as the human endometrium.¹² Thus, the vitamin D pathway may be important for reproductive function. Vitamin D has also been associated with uterine receptivity and embryonic implantation,^13–15 which might be relevant mechanisms for the observed associations with conception.^16–18

Specifically, treatment of endometrial biopsy-derived eSCs with 1,25(OH)₂D₃ (calcitriol) increased prolactin, a biomarker of decidualization.^19,20 Decidualization, a progesterone driven differentiation process during the mid-secretory phase, is essential for embryo implantation in humans.¹⁹ Two studies in women using artificial reproductive technologies (ART) suggest that vitamin D deficiency may reduce endometrial thickness²¹ and clinical pregnancy rates,^21,22 albeit inconsistently.²³

In summary, this literature suggests that the vitamin D pathway may impact the ability to conceive a pregnancy potentially through effects on the endometrium. Our objective was to explore the effects of vitamin D on eSC proliferation and decidualization, using endometrial stromal cells derived from menstrual effluent (ME), a non-surgical source of living endometrial cells.

Materials and Methods

ROSE study design.

Since 2013, the IRB-approved ROSE study (13–376A) at Northwell, has been enrolling participants from North America who are either diagnosed with endometriosis (by laparoscopy with verification of endometriosis lesions by pathology), participants who self-report symptoms consistent with endometriosis but have not yet been diagnosed, and unaffected controls (who self-report the absence of endometriosis and chronic symptoms of endometriosis, adenomyosis, and pelvic inflammatory disease). For functional eSC studies, we used ME-eSCs obtained from unaffected controls (N=18) between the ages of 20–40 years, who reported not using hormones/oral contraception, as previously described.^24,25 While limited demographic and medical/gynecologic information was collected, information regarding participant fertility/infertility status and use of vitamin D supplements was not collected. ME-eSCs from the ROSE study were used for all stromal cell experiments except for the paired analyses of ME-eSCs collected pre- and post-vitamin D supplementation in the inVitD trial (see below).

All ROSE participants provided written, informed consent prior to enrollment. ME was collected at home during peak flow (typically days 1–3 of menses) using a menstrual collection cup (provided by DIVA International, Toronto, Canada) and processed as previously described.^24–26

ME-derived endometrial stromal cells (ME-eSCs) were isolated, cultured, phenotyped, and cryopreserved at low passages (0–1), as previously described.^24–26 Briefly, ME-eSCs were grown in DMEM containing 10% mesenchymal stem cell-fetal bovine serum (FBS), 1% penicillin-streptomycin-glutamine (PSQ) (Gibco/Thermo-Fisher), and normocin (1:500) (Invivogen, San Diego, CA, US) (maintenance media) at 37°C with 5% CO₂. When confluent, stromal cells were cryopreserved in 10% dimethyl sulfoxide (DMSO)/90% FBS.

For another subset of ROSE participants (N=14, aged 20–40 years, could be using hormonal contraceptives), peripheral blood (10mL) was collected within 9 days of their ME collection. Information regarding participant use of vitamin D supplements was not collected. The peripheral blood draw was collected into EDTA (ethylenediaminetetraacetic acid) vacutainer tubes and centrifuged for plasma and stored at −80°C. The ME sample was collected and processed as described above, however after a brief centrifugation the cell-free supernatant was also collected and stored as the ME soluble fraction (or “ME”). These paired peripheral plasma and ME samples were sent for vitamin D biomarker quantification at a certified lab with expertise in measuring vitamin D metabolites and proteins by liquid chromatography tandem mass spectrometry (LC-MS/MS) (Department of Laboratory Medicine and Pathology at the University of Washington, Seattle, WA).²⁷ This was a pilot study to determine the feasibility of measuring vitamin D metabolites and vitamin D binding protein (VDBP) in ME and to compare plasma and ME levels. In addition to VDBP, total 25-hydroxyvitamin D (25(OH)D) was measured along with total 1,25-dihydroxyvitamin D (1,25(OH)₂D), and 24,25-dihydroxyvitamin D₃ (24,25(OH)₂D₃) in both ME and peripheral plasma samples.

inVitD trial design.

The inVitD trial was a partially randomized, clinical trial of vitamin D (cholecalciferol) supplementation in women with spontaneous, natural menstrual cycles recruited from Detroit, Michigan and Research Triangle Park, North Carolina.²⁸ Participants with circulating total 25(OH)D levels of less than 20ng/ml were randomized to receive supplementation with either 4,200 IU per week or 50,000 IU per week of cholecalciferol for their next three menstrual cycles. The inVitD Trial was approved by the NIEHS Institutional Review Board and is listed on clinicaltrials.gov (NCT05050916).

A subset of inVitD participants who opted to collect their ME were sent an ‘at home ME collection kit’ containing a menstrual cup and instructions as previously described.^24–26,29 Participants collected ME during their first menses of Phase 2 and were asked to collect prior to taking their first dose of cholecalciferol (Pre-supplementation collection, “PRE”). They collected a second sample of ME during the second menses after starting supplementation, the start of their third and final menstrual cycle of Phase 2 (Post-supplementation collection, “POST”). At the time of POST sample collection, the participants had been on supplementation for an average of 11.9 weeks (range: 10.1 – 14.3 weeks). ME samples were overnight shipped to the Northwell lab and processed to isolate, culture, and cryopreserve ME-eSCs, as described previously for the ROSE study.^24–26 Nine participants provided usable paired ME samples (both PRE and POST). Note: formal sample size calculations to determine ME collection numbers were not performed for this sub-study because it was a pilot aim, the first of its kind to assess the effects of any in vivo supplementation on in vitro ME-eSCs outcomes. ME samples were processed and ME-eSCs were cultured in maintenance media, as described above.

Decidualization assays.

When confluent, ME-eSCs (passages 2–4) from the ROSE and inVitD studies were plated at 2.5×10⁵ cells/ml (100 μl/well in 96-well plates) in maintenance media. After at least 3–4 hrs of adherence, media was aspirated and replaced with DMEM containing 2%FBS and PSQ with normocin (assay media). After an overnight incubation, ME-eSCs were treated with vehicle or calcitriol (MedChem Express, Monmouth Junction, NJ, US) at 10⁻⁷M based on optimization experiments (Supplemental Figure 1)⁷ and as previously described for eSCs ¹⁴, or at the indicated concentrations in assay media. After 4 hrs, eSCs were treated with 8-bromoadenosine 3′,5′-cyclic monophosphate sodium salt (cAMP, 0.5mM) with or without medroxyprogesterone acetate (MPA [1×10⁻⁷M], a stable analogue of progesterone) (N=3–4 per condition) (both purchased from Sigma-Aldrich, St. Louis, MO, US). For some experiments, calcitriol was added either 20 hrs pre-, 4 hrs pre-, at the time of treatment or 20 hrs post-treatment with cAMP (0.5mM) with or without MPA (1×10⁻⁷M) (N=3–4 per condition) (Sigma-Aldrich). Where noted, cells were treated with either cAMP alone or cAMP+MPA+E₂ (E₂ = estradiol, 1×10⁻⁸M) (purchased from Sigma-Aldrich). Also, in a separate set of experiments either cholecalciferol or calcifediol (MedChem Express) was added to ME-eSCs 4 hrs prior to treatment at the indicated concentrations to test their effects on decidualization. All vitamin D metabolites/analogs were added to the cultures once daily. After 48 hrs (post-cAMP±MPA, cAMP alone, or cAMP+MPA+E₂ treatment), cell-free supernatants were collected and frozen at −80°C until assayed for insulin-like growth factor-binding protein 1 (IGFBP1) by ELISA (RRID: AB_2814878) and/or prolactin (PRL) protein by ELISA (RRID: AB_2814879) (both purchased from R&D Systems, Minneapolis, MN, US), as previously described.^24,25 IGFBP1 and PRL are the most well-established markers of decidualization.^19,20 Data are shown as either IGFBP1 concentrations (pg/ml) or PRL concentrations (pg/ml) or as fold-change over vehicle-treated. The 48 hr time point was chosen because it is consistent with our prior studies using ME-derived eSCs where the decidualization response is robust enough to be assessed.^26,29

Proliferation assays.

Confluent ME-eSCs from the ROSE study were plated in 96-well plates at 1.5×10⁴/mL in 100 μL/well in maintenance media. The next day, cells were treated with vehicle or calcitriol (10⁻⁷M) (N=4–6 per condition) and incubated at 37°C with 5% CO₂. ME-eSCs were treated with calcitriol once daily. After 72 hrs incubation, cells were washed once with cold PBS, aspirated, and frozen at −80°C until assayed for cell proliferation using the CyQUANT^™ Cell Proliferation Assay, according to the manufacturer’s directions (Thermo-Fisher, Waltham, MA, US).

Quantitative real-time PCR (qPCR).

To assess the effects of calcitriol on ME-eSCs, confluent cells from the ROSE study were plated at 2.5×10⁵ cells/ml (100 μl/well in 96-well plates) in maintenance media, as described above. After an overnight incubation in assay media cells were then treated with vehicle or calcitriol (at 10⁻⁷M) and then either maintained under basal conditions or treated with deciduogenic agents (cAMP+MPA) 4 hrs later, as described above. At the indicated times, cells were analyzed for the expression of VDR and OSR2 (odd-skipped related 2 protein) mRNA transcripts. Real time qPCR (RT-qPCR) reactions using TaqMan primers (ThermoFisher) (Supplemental Table 2)⁷ were performed in triplicate using the Cells-to-Ct^™ 1 step kit (Fisher Scientific) and the ABI ViiA7 Realtime PCR System (Life Technologies, Carlsbad, CA). Human UBE2D2 (ubiquitin-conjugating enzyme E2 D2) was used as the housekeeping gene for normalizing transcript levels, as it was previously identified as a stable housekeeping gene for the endometrium.³⁰ Relative changes in gene expression were calculated as fold-changes using the comparative $2^(-\mathrm{\Delta }\mathrm{\Delta }\mathrm{C}\mathrm{t})$ method, as previously described.²⁴

Target protein assessment by western blotting.

To determine pathways involved in mediating the effects of vitamin D on eSC decidualization, confluent eSCs from the ROSE study (after an overnight incubation in assay media) were treated with vehicle or calcitriol (10⁻⁷M) for 4 hrs and then treated with cAMP+MPA, as described above (except cells were plated in 2 wells of a 6-well plate, 2mL/well, per condition). After 24 hrs, cell lysates were prepared in RIPA lysis buffer containing protease and phosphatase inhibitors (Halt^™ Protease and Phosphatase Inhibitor Cocktail, Thermo Fisher). Supernatants were collected after centrifugation at 13,000 rpm for 10 min at 4°C. and protein levels were quantified using the Pierce BCA Protein Assay Kit (Thermo Fisher). Gels (4–12% NuPAGE Bis-Tris 1.5 mm) were loaded with up to 65 μg/lane and electrophoresed in NuPAGE MOPS (3-(N-morpholino)propanesulfonic acid) running buffer (Invitrogen), as previously described.²⁴ Proteins were transferred onto Immobilon Immobilon^®-FL-PVDF (fluorescent polyvinylidene fluoride) membranes. After blocking for 1 hr in PBS 0.1% (volume/volume, v/v) Tween-20 (PBS wash buffer) and 5% (weight/volume, w/v) nonfat milk powder, membranes were incubated sequentially overnight at 4°C in primary antibody specific for each analyte: protein kinase B (AKT, RRID: AB_329827), ERK1/2 (extracellular signal-regulated kinase 1 and 2, RRID: AB_390779), phospho-ERK1/2 (RRID: AB_2315112), GAPDH (glyceraldehyde-3-phosphate dehydrogenase, RRID: AB_561053), and PRAS40 (proline-rich AKT substrate of 40 kDa, RRID:AB_2225033), phospho-PRAS40 (RRID:AB_2258110), and p53 (RRID: AB_331476). Antibodies were diluted in PBS wash buffer containing 5% (w/v) bovine serum albumin or 5% (w/v) milk powder, depending on the manufacturer’s recommendation (sources of antibodies and dilutions are listed in Supplemental Table 3)⁷. The next day, membranes were washed and incubated for 1 hr in horseradish peroxidase (HRP)-conjugated secondary antibody (RRID: AB_2099233) diluted in PBS wash buffer containing 5% nonfat milk. Membranes were then washed in PBS wash buffer and imaged after adding Clarity ECL substrate (BioRad) or SuperSignal West Dura Extended Substrate (Thermo Fisher) using a ChemiDoc system (BioRad). SeeBlue Plus2 Pre-stained molecular mass standard (Thermo Fisher) was used to estimate protein weight. Western blots were stripped with ReBlot Plus (Millipore), according to the manufacturer’s suggestions, and then re-blocked and probed with antibodies as described above. Band densities were quantified using NIH ImageJ v1.54r.²⁴

Statistical analyses

Endometrial stromal cell studies (in vitro):

Analyses and graphical presentations were performed using GraphPad Prism 10.4.1 software. The results are presented as the mean for each subject’s cells with the median and interquartile range (IQR) shown for each group, unless indicated. Where shown, fold-change was calculated as the ratio of the measurement in the treated group to the measurement in the vehicle-treated or control group. For paired data with two groups, groups were compared using the Wilcoxon matched-pairs signed-rank test. For data with multiple groups, data were analyzed using the Kruskal-Wallis (analysis of variance [ANOVA] on ranks test), followed by appropriate post hoc comparisons.

Comparison of peripheral plasma and ME vitamin D metabolites and VDBP.

Spearman correlation coefficients (and associated p-values) were calculated to compare plasma and ME levels of vitamin D metabolites and VDBP. We further examined these correlations after excluding ME samples that were highly hemolyzed, which is a known issue particularly for the assessment of VDBP. ³¹

Results

Timing of calcitriol addition to endometrial stromal cells influences decidualization

Using eSCs from the ROSE study, time course experiments were performed to determine the optimal time to add calcitriol to effectively enhance eSC decidualization capacity in vitro. Adding calcitriol (at 10⁻⁷M) 20 hrs prior to, 4 hrs prior to, and at the same time as the addition of cAMP+MPA significantly enhanced decidualization over vehicle-treated eSCs when analyzed by IGFBP1 production(Figure 1A–1C). Optimal enhancement was observed when calcitriol was added prior to cAMP+MPA (Figure 1B), with a slightly more significant effect when added 4 hrs prior to stimulation. By contrast, the addition of calcitriol (10⁻⁷M) 20 hrs after the addition of cAMP+MPA did not significantly affect decidualization capacity over vehicle treatment (Figure 1D). Additional experiments were performed to confirm our initial observations that pre-treatment of eSCs with calcitriol (10⁻⁷M) significantly enhanced decidualization capacity induced by cAMP+MPA: IGFBP1 levels were increased approximately 2.5-fold, and PRL levels (another well-known marker of decidualization) were increased 1.5-fold (Figure 1E–1F and 1G–1H, respectively).

Figure 1.

Calcitriol treatment enhances ME-eSC decidualization. Time course experiment (A-D): ME-eSCs were treated with vehicle (Veh) or calcitriol (10⁻⁷M) for either 20 hrs prior to (A), 4 hrs prior to (B), at the same times as, or (D) 20 hrs after cAMP+MPA stimulation (to induce decidualization). IGFBP1 concentrations in the culture supernatants were measured by ELISA. Validation experiments (E-H): 4 hr pre-treatment of eSCs with calcitriol (10⁻⁷M) enhances cAMP+MPA-stimulated decidualization capacity when assessed by measuring IGFBP1 (E, F) and PRL (G, H). Data are shown as fold-change (median, 25^th, 75^th percentiles) with vehicle as 1 (A-E and G) or as paired data points (vehicle vs. calcitriol-treated) connected by a line showing raw IGFBP1 and PRL concentrations (F and H, respectively). Each symbol represents the results of an individual’s ME-eSCs. P-values are shown.

Calcitriol enhances decidualization regardless of inducing agents

To determine whether the enhancing effect of calcitriol on decidualization potential was dependent on the deciduogenic agents used, we compared adding calcitriol to ROSE study eSCs 4 hrs prior to decidualization induction with cAMP alone, cAMP+MPA, or cAMP+MPA+E2. As shown in Figure 2, the addition of calcitriol prior to decidualization significantly improved decidualization capacity regardless of the deciduogenic cocktail. We mainly continued experiments with cAMP+MPA, as this combination was recently described to best reflect the in vivo process.³²

Figure 2.

Comparison of the effect of vehicle (Veh) vs. calcitriol (10⁻⁷M) on decidualization capacity, as measured by production of IGFBP1 (A-C) or PRL (D-F) after stimulation with deciduogenic agents: cAMP alone (A,D), cAMP+MPA (B, E), or cAMP+MPA + E2 (C, F). Each symbol represents the results of an individual’s ME-eSCs. Data are shown as fold-change (median, IQR), with vehicle as 1. P-values are shown.

Calcitriol does not alter endometrial stromal cell proliferation

Next, we examined the effect of calcitriol treatment on cell proliferation using eSCs from the ROSE study, as numerous agents that enhance decidualization exert anti-proliferative effects. Calcitriol treatment at the dose found to effectively enhance decidualization (10⁻⁷M) had no significant effect on eSC proliferation (Figure 3A–3B).

Figure 3.

The effect of vehicle (Veh) vs. calcitriol (10⁻⁷M) on proliferation of all ME-eSCs (A) and in paired vehicle- vs. calcitriol-treated-ME-eSCs (B). Each symbol represents the results of an individual’s ME-eSCs. Data are shown as median, IQR, with vehicle as 100 (A) or as paired vehicle vs. calcitriol-treated data points (as raw values) connected by a line (B). P-values are shown.

The effects of calcitriol and decidualization on VDR expression by endometrial stromal cells

Since vitamin D’s effects can be mediated through binding to the vitamin D receptor, we examined its effect on VDR mRNA expression using eSCs from the ROSE study. VDR mRNA expression was detectable under basal conditions (i.e. vehicle-treated eSCs) when assessed 48 and 72 hrs after assay start time (Figure 4A and 4B, respectively). The addition of calcitriol (10⁻⁷M) to eSCs under basal conditions significantly induced VDR mRNA when analyzed 48 and 72 hrs later (Figure 4A and 4B). Treatment of eSCs with cAMP+MPA to induce decidualization did not alter VDR mRNA expression when analyzed 48 hrs post-decidualization (Figure 4A). However, when eSCs were pre-treated with calcitriol prior to decidualization VDR mRNA expression was significantly increased when compared to vehicle-treated decidualized cells at 48 hr (Figure 4A), a time point that coincided with enhanced decidualization in prior experiments (Figures 1–2). Interestingly, when eSCs were analyzed 72 hr post-cAMP+MPA, vehicle-treated decidualized eSCs showed significant induction of VDR mRNA expression when compared to vehicle-treated eSCs under basal conditions (Figure 4B). However, this effect was not significantly further increased by pre-treatment with calcitriol (Figure 4B).

Figure 4.

VDR mRNA expression by ME-eSCs under basal conditions (without cAMP+MPA) and during decidualization with cAMP+MPA, with and without calcitriol (10⁻⁷M) pre-treatment. ME-eSCs were assessed by qPCR for VDR mRNA expression 48 hrs (A) and 72 hrs (B) after the onset of decidualization by qPCR. Each symbol represents the results of an individual’s ME-eSCs. Data are shown as fold-change mRNA expression (median, IQR), with vehicle as 1. P-values are shown.

Calcitriol-treated eSCs increase OSR2 mRNA expression

To explore potential mechanisms that might mediate the effects of calcitriol on enhancing decidualization we assessed OSR2 mRNA expression using eSCs from the ROSE study. OSR2 is a known target gene of calcitriol that encodes odd-skipped related 2 protein, previously shown to mediate MYC-associated factor X (MAX)-regulated decidualization.³³ OSR2 mRNA expression was significantly increased (4 to 5-fold) following treatment with vitamin D under both basal and decidualization conditions when assessed at 48 and 72 hr time points (Figure 5).

Figure 5.

OSR2 mRNA expression by ME-eSCs under basal conditions (without cAMP+MPA) and during decidualization with cAMP+MPA, with and without calcitriol (10–7M) pre-treatment. ME-eSCs were assessed 48 hrs (A) and 72 hrs (B) after the onset of decidualization by qPCR for OSR2 mRNA expression. Each symbol represents the results of an individual’s ME-eSCs. Data are shown as fold-change mRNA expression (median, IQR), with vehicle as 1. P-values are shown.

Signaling pathways implicated in decidualization are altered by calcitriol

Next, using eSCs from the ROSE study we analyzed the effects of calcitriol on proteins that mediate signaling pathways implicated in decidualization. Calcitriol pretreatment significantly reduced AKT phosphorylation, an early signal involved in decidualization, without affecting total AKT levels (Figure 6A and 6B). Additionally, calcitriol reduced the phosphorylation of PRAS40, a downstream target of AKT, without affecting total PRAS40 levels (Figure 6A and 6C). By contrast, calcitriol did not alter ERK1/2 phosphorylation or total ERK1/2 expression or p53 expression (Figure 6A, 6D–6F).

Figure 6.

Calcitriol pre-treatment alters levels of signaling mediators implicated in decidualization. (A) Representative western blots for three control participants’ eSCs treated with vehicle (Veh) or calcitriol (10⁻⁷M) for 4 hrs prior to vehicle or cAMP+MPA stimulation to induce decidualization. 24 hrs post-cAMP+MPA stimulation, cell lysates were analyzed by western blotting for various signaling mediators. Quantification of band densities are shown for p-AKT and total AKT (B), p-PRAS40 and PRAS40 and total PRAS40 (C), p-ERK1 and total ERK1 (D), p-ERK2 and total ERK2 (E), p53 (F) and GAPDH (G) in vehicle (Veh) vs. calcitriol-pre-treated eSCs. Band densities were normalized to GAPDH and corrected for total non-phosphorylated proteins, where appropriate, and shown as % control between Veh- vs. calcitriol-treated decidualized eSCs (B-G), with vehicle-treated (non-decidualized) cell values set to 100% (line). Each point represents data from one individual’s eSCs, with median, IQR shown for each group.

The effects of in vivo cholecalciferol supplementation on ME-derived eSC decidualization potential

Paired (n=9) ME-eSCs collected from inVitD participants pre- and post-cholecalciferol supplementation in vivo were assessed for their decidualization capacity following stimulation with cAMP+MPA ex vivo. There was a tendency towards higher IGFBP1 levels in POST-eSCs, compared with PRE-eSCs when decidualization was induced with cAMP, with a p-value of borderline significance, p=0.07 (Figure 7A, as paired data, left panel, and as individual data points with median, IQR, right panel). The p-value was weaker when decidualization was induced with cAMP+MPA (Figure 7B, as paired data, left panel, and as individual data points with median, IQR, right panel). When pre-treated with calcitriol (active vitamin D), eSCs isolated from PRE- and POST-supplementation ME samples showed increased decidualization responses (Figure 8A–D), in a manner similar to that observed with ME-eSCs obtained from the ROSE study (where vitamin D status was unknown, Figures 1–2). The enhancing effect of added calcitriol was statistically significant for both PRE and POST ME-eSCs (Figures 8A–8B and 8C–8D, respectively).

Figure 7.

Effects of in vivo cholecalciferol supplementation on ME-eSC decidualization. ME-eSCs were collected pre- and post-cholecalciferol supplementation (in vivo) and then tested for decidualization capacity following cAMP (A) or cAMP+MPA (B) stimulation. Decidualization capacity was assessed by measuring IGFBP1 production by pre- vs. post-ME-eSCs 48 hrs later. Data are shown as paired data points (from one individual) connected by a line showing raw IGFBP1 concentrations (left panel) and as IGFBP1 concentrations for each subject’s cells with median, IQR for the group (right panel). P-values are shown.

Figure 8.

Effects of exogenously added calcitriol on ME-eSC decidualization pre- and post-cholecalciferol supplementation in vivo. ME-eSCs were collected pre- (A, B) and post-cholecalciferol (C, D) supplementation in vivo and then pre-treated with vehicle (Veh) or calcitriol (10⁻⁷M) 4 hrs prior to decidualization induced by cAMP alone (A, C) or cAMP+MPA (B, D). Decidualization capacity was assessed by measuring IGFBP1 by ELISA. Data are shown as IGFBP1 concentrations (pg/ml), with data points connected by a line representing paired data points from one individual’s eSCs (± calcitriol). P-values are shown.

The effects of cholecalciferol and calcifediol on eSC decidualization

Finally, we used ROSE study ME-eSCs to examine whether the effect of vitamin D on eSC decidualization is limited to the active form of vitamin D, calcitriol, or if decidualization can be enhanced by the higher-level metabolites: vitamin D₃ (cholecalciferol), or 25(OH)D₃ (calcifediol). As shown in Figure 9, cholecalciferol had no effect on decidualization, while calcifediol had a small, significant positive effect on decidualization.

Figure 9.

Effects of cholecalciferol and calcifediol treatment on ME-eSC decidualization. ME-eSCs were pre-treated with vehicle (Veh) vs. cholecalciferol at 10⁻⁷M (A) or Veh vs. calcifediol at 10⁻⁷M (B) 4 hrs prior to decidualization (stimulated with cAMP+MPA). Decidualization capacity was assessed by measuring IGFBP1 by ELISA. Data are shown as fold-change IGFBP1 (median, IQR). Each symbol represents the results of an individual’s ME-eSCs. P-values are shown.

The correlation between peripheral plasma and ME vitamin D metabolite levels

Within peripheral plasma samples, total 25(OH)D was correlated with 24,25(OH)₂D₃ (r = 0.83, p=0.0003). (Table 1) Correlations with other metabolites were weak. Similarly, within ME samples, total 25(OH)D was correlated with 24,25(OH)₂D₃ (r = 0.61, p=0.02) and total 1,25(OH)₂D (r = 0.54, p=0.05). (Table 1) ME 24,25(OH)₂D₃ was also correlated with ME total 1,25(OH)₂D (r = 0.51, p = 0.08). When comparing across matrices, peripheral plasma total 25(OH)D was not correlated with ME total 25(OH)D (r = 0.39, p = 0.16). Peripheral plasma total 25(OH)D was correlated with ME 24,25(OH)₂D₃ (r = 0.79, p=0.0007). Similarly, peripheral plasma 24,25(OH)₂D₃ was correlated with ME 24,25(OH)₂D₃ (r = 0.78, p = 0.001). Plasma total 1,25(OH)₂D was correlated with ME total 1,25(OH)₂D (r = 0.66, p = 0.01) and ME VDBP (r = 0.76, p = 0.002). Plasma VDBP was moderately correlated with ME total 1,25(OH)₂D (r = −0.54, p = 0.05). Most correlations with ME VDBP were stronger when excluding ME samples that were highly hemolyzed. (Table 2) The correlation between ME total 25(OH)D and peripheral plasma total 25(OH)D was also stronger when excluding highly hemolyzed samples. Other correlations with ME metabolites were similar in strength when excluding highly hemolyzed samples.

Table 1.

Spearman correlation coefficients for paired peripheral plasma (PP) and menstrual effluent (ME) vitamin D metabolites and vitamin D binding protein (VDBP) for N=14 paired samples^a (28 total) collected in the ROSE study.

	PP 24,25(OH)2D3	PP total 1,25(OH)2D	PP VDBP	ME total 25(OH)D	ME 24,25(OH)2D3	ME total 1,25(OH)2D	ME VDBP
PP total 25(OH)D	0.83****	−0.25	0.08	0.39	0.79****	0.05	−0.24
PP 24,25(OH)2D3		−0.26	0.05	0.38	0.78***	−0.005	−0.16
PP total 1,25(OH)2D			−0.10	0.31	0.02	0.66**	0.76***
PP VDBP				−0.42	−0.23	−0.54*	0.09
ME total 25(OH)D					0.61*	0.54*	0.31
ME 24,25(OH)2D3						0.51*	0.01
ME total 1,25(OH)2D							0.41

^*0.05<= p <=0.08

^**0.01<= p <0.05

^***0.001<= p<0.01

^****p < 0.001

^aN = 13 pairs for ME total 1,25(OH)₂D

Table 2.

Spearman correlation coefficients for paired peripheral plasma (PP) and menstrual effluent (ME) vitamin D metabolites and vitamin D binding protein (VDBP) for N=8 paired samples^a (16 total) that were not highly hemolyzed, collected in the ROSE study.

	PP 24,25(OH)2D3	PP total 1,25(OH)2D	PP VDBP	ME total 25(OH)D	ME 24,25(OH)2D3	ME total 1,25(OH)2D	ME VDBP
PP total 25(OH)D	0.71**	0.14	−0.33	0.71**	0.76**	−0.11	0.43
PP 24,25(OH)2D3		0.17	−0.02	0.48	0.76**	0.11	0.64*
PP total 1,25(OH)2D			0.38	0.17	0.05	0.57	0.67*
PP VDBP				−0.07	−0.43	−0.36	0.60
ME total 25(OH)D					0.76**	0.21	0.50
ME 24,25(OH)2D3						0.32	0.36
ME total 1,25(OH)2D							0.18

^*0.05<= p <=0.09

^**0.01<= p <0.05

^aN = 7 pairs for ME total 1,25(OH)₂D

Discussion

We found that pre-treatment with calcitriol enhanced eSC decidualization in vitro. Calcitriol enhanced decidualization when eSCs were treated with cAMP alone or cAMP+MPA and when using ME-derived eSCs collected from participants in two separate study populations (ROSE and the inVitD Trial).These results agree with one study of calcitriol and eSCs from human endometrial biopsies¹⁵ and with an in vivo study of rats that reported calcitriol treatment induced decidualization of endometrial cells.³⁴ Calcitriol has been reported to exert progesterone-like activity ³⁵ and may promote apoptosis ³⁶, both of which are important during decidualization. ^19,24 The cross-talk between progesterone and calcitriol should be further investigated.

We observed the strongest effect on eSC decidualization with calcitriol pre-treatment, compared with calcifediol (weak effect), and cholecalciferol (no effect). These results are consistent with at least two possibilities. First, enhanced decidualization occurs through VDR binding, as calcitriol has the highest VDR binding affinity, followed by calcifediol, and cholecalciferol, which has low affinity for the VDR. Or second, calcifediol treatment leads to a higher availability of calcitriol, but at a lower concentration than direct calcitriol treatment.

VDR, the major receptor mediating the effects of vitamin D, is expressed by most tissues in the body. We found that VDR mRNA is expressed in eSCs under basal conditions and was enhanced 72 hrs post-decidualization implicating the importance of vitamin D signaling in sustained decidualization. Pre-treatment with calcitriol increased VDR mRNA expression under both basal conditions and during decidualization. Consistent with our findings, calcitriol has been found to promote VDR protein expression in several cell types such as kidney, intestine, and parathyroid gland, but not in others (of malignant transformation human leukemia and osteosarcoma).³⁷ Here we provide the first evidence in human eSCs of in vitro regulation of VDR expression by its ligand, calcitriol (termed homologous regulation).

Quercetin, a natural flavonoid with anti-inflammatory, antioxidant, and senolytic properties also promotes decidualization.²⁴ This was accompanied by inhibition of phosphorylated AKT and phosphorylated PRAS40, its downstream target (important for protein synthesis and cell metabolism), along with reduced ERK1/2 phosphorylation and elevated p53 expression.^38,39 While calcitriol had no effect on ERK1/2 activation or p53 expression, like quercetin, calcitriol-enhanced eSC decidualization was accompanied by reduced AKT phosphorylation (without affecting total AKT levels) and reduced PRAS40 phosphorylation (without affecting total PRAS40 levels). Previous literature shows that decidualization is accompanied by reduced AKT signaling and AKT inhibitors enhance eSC decidualization, while excess AKT activation suppresses decidualization.^24,40–42 Also, consistent with our findings, previous studies report that calcitriol treatment reduced AKT phosphorylation in a hepatocyte-derived cellular carcinoma cell line⁴³ and in macrophages.⁴⁴ Although we have not examined additional downstream signaling events, we speculate that loss of phosphorylated PRAS40 (a negative regulator of mTORC1 [mechanistic target of rapamycin]) following calcitriol addition promotes mTORC1 activation leading to enhanced decidualization, as previously described.⁴⁵ However, signaling effects are often cell-specific and defining the downstream effectors will require further studies.

We found that calcitriol pre-treatment increased OSR2 mRNA expression under both basal and decidualization conditions. OSR is a known target gene of vitamin D that encodes odd-skipped related 2 protein, previously shown to directly mediate MAX-regulated decidualization.³³ MAX is a transcription factor and MAX was a significantly downregulated transcript in eSCs collected from decidual tissues of women with recurrent spontaneous abortions.³³ MAX has been reported to regulate cAMP+progesterone-induced human eSCs decidualization in vitro by directly targeting OSR2 transcription.³³ Although the direct link between MAX-OSR2 and AKT phosphorylation has not been established, MAX directly binds MYC⁴⁶, a transcriptional activator and repressor that regulates numerous cellular processes, including cell proliferation and differentiation. Furthermore, OSR2 was shown to be indispensable for decidualization.³³ Thus, we propose that calcitriol probably regulates OSR2 expression to enhance decidualization.

To further explore the presence of the vitamin D signaling pathway in ME and the endometrium, we searched for vitamin D related mRNA transcripts in our previously published single cell RNA sequence (scRNAseq) data of ME-tissues deposited in the National Center for Biotechnology Information ^29,47. We found that cells from fresh ME, including eSCs, epithelial cells, myeloid cells and lymphocytes express VDBP mRNA and CYP2R1 (25-hydroxylase) mRNA. Additionally, CYP27B1 (1-α hydroxylase) mRNA is expressed in ME, mainly by eSCs and myeloid cells (Supplemental Figure 2)⁷. Expression of CYP27B1 mRNA in fresh ME samples suggests that calcitriol could be synthesized in these cells from circulating calcifediol⁴⁸ and that increases in calcifediol could be directly converted to increased calcitriol, and its beneficial effects, in these cells. There was almost no expression of FGF23 (which encodes fibroblast growth factor 23) and LPR2 (which encodes low-density lipoprotein receptor-related protein 2, also known as megalin) mRNA in eSCs (Supplemental Figure 2)⁷, suggesting that the canonical VDR signaling pathway may be more important for eSCs than the alternative non-genomic pathways. VDBP is moderately expressed in multiple gastrointestinal tissues, the kidney, testis, prostate, endometrium, and breast.⁴⁹ VDBP is predominantly a hepatic protein, but expression in the uterus has been described, and may modulate immuno-regulatory activities including macrophage activation, relevant for the establishment of endometriosis.⁵⁰

In our analyses of human eSCs collected before and after in vivo supplementation with cholecalciferol, we observed a potential increase in decidualization response post-supplementation. The findings show that all but two participants had higher decidualization responses post-supplementation. To our knowledge this is the first report where ME was leveraged to assess the effects of in vivo supplementation on eSC function measured ex vivo. Future research could examine this finding using a larger sample size, a longer duration of cholecalciferol supplementation, a longer follow-up time, and/or a shorter eSC culture time prior to decidualization assessment.

We found that vitamin D metabolites were measurable in ME samples, providing novel support for the use of ME to represent exposures at the level of the endometrium. While our study is small, it is a first of its kind and provides several avenues for future research. We also found that circulating peripheral plasma levels of some vitamin D metabolites and VDBP were correlated with ME levels of the same proteins. The correlation between peripheral plasma 25(OH)D and ME total 25(OH)D was influenced by the presence of hemolysis in the ME samples. Similarly, correlations between ME VDBP and other metabolites were also affected by the exclusion of hemolyzed samples. Further exploration of the effect of hemolysis and optimizing ME sample preparation for LC-MS/MS would be informative; additional measurements of vitamin D-related proteins in ME versus circulating plasma are warranted.

In conclusion, this research shows for the first time that pre-treatment with calcitriol improves human ME-derived eSC decidualization in vitro. Possible explanations for this include progesterone-like effects of calcitriol, direct modulation of OSR2 expression, and reduced AKT signaling. Vitamin D metabolites and VDBP can be measured in menstrual effluent providing a window into uterine tissue-levels of these proteins. The in vivo cholecalciferol supplementation results are promising and the first of their kind to study the effects of any in vivo supplementation or treatment on ME-eSC function. Moreover, these results highlight the utility of ME for studying other in vivo exposures that may alter endometrial cell function that could then be examined in vitro. Our findings provide a mechanistic link between epidemiologic studies that report associations between higher circulating total 25(OH)D levels and an improved probability of conceiving a pregnancy. It is possible that higher circulating total 25(OH)D levels lead to higher calcitriol levels in the endometrium which improves eSC decidualization and therefore implantation and pregnancy establishment. This may be particularly important for those with infertility or implantation issues. Vitamin D supplementation may be a low-cost intervention for improving fertility by improving eSC decidualization.

Abbreviations used:

total 1,25(OH)₂D

1,25-dihydroxyvitamin D

24,25(OH)₂D₃

24,25-dihydroxyvitamin D₃

25(OH)D

total 25-hydroxyvitamin D

AKT

protein kinase B

ANOVA

analysis of variance

cAMP

8-bromoadenosine 3′,5′-cyclic monophosphate sodium salt

CYP2R1

25-hydroxylase

CYP27B1

1-α hydroxylase

DMSO

dimethyl sulfoxide

E₂

estradiol

EDTA

ethylenediaminetetraacetic acid

ELISA

enzyme-linked immunoabsorbent assay

ERK1/2

extracellular signal-regulated kinase 1 and 2

FBS

mesenchymal stem cell-fetal bovine serum

FGF23

fibroblast growth factor 23

FL-PVDF

Immobilon^®-FL Fluorescent polyvinylidene fluoride

GAPDH

glyceraldehyde-3-phosphate dehydrogenase

HRP

horseradish peroxidase

IGFBP1

insulin-like growth factor-binding protein 1

inVitD

investigation of vitamin D and menstrual cycles trial

IQR

interquartile range

LPR2

low-density lipoprotein receptor-related protein 2, a.k.a. megalin

MAX

MYC-associated factor X

ME

menstrual effluent

ME-eSC

menstrual effluent-derived endometrial stromal cells

MPA

medroxyprogesterone acetate

mTORC1

mechanistic target of rapamycin

OSR2

odd-skipped related 2 protein

PCOS

polycystic ovary syndrome

PRAS40

proline-rich AKT substrate of 40 kDa

PRL

prolactin

PSQ

penicillin-streptomycin-glutamine

ROSE

Research Outsmarts Endometriosis study

RRID

Research Resource Identifier

UBE2D2

ubiquitin-conjugating enzyme E2 D2

VDBP

vitamin D binding protein

VDR

vitamin D receptor

Data availability:

The data underlying this article will be shared upon reasonable request to the corresponding author.