The functional expression of transient receptor potential channels in the mouse endometrium

Abstract

Introduction

Embryo implantation, the rate-limiting step in reproduction, is a complex process that requires a competent blastocyst, a receptive endometrium and the intricate embryo-uterine crosstalk mediated by a myriad of factors such as hormones and prostaglandins. The implantation process is initiated by embryo attachment to the endometrial epithelium, and is followed by invasion in the stromal compartment, sustaining and perpetuating decidualization (Lessey, 2000; Wang and Dey, 2006; Zhang et al., 2013). However, human reproduction has proven to be rather inefficient since the monthly fecundity rate is merely 30% (Zinaman et al., 1996; Evers, 2002) and the worldwide infertility prevalence is circa 9% (Boivin et al., 2007). Many underlying causes of infertility have been overcome by tremendous developments in assisted reproductive techniques (ART); pregnancy rates, however, remain relatively low mainly due to implantation failure (Wilcox et al., 1988; Norwitz et al., 2001; Margalioth et al., 2006). While embryo quality may be considered as a minor factor contributing to implantation failure in ART, the ability to achieve a receptive endometrium can be a major cause (Cakmak and Taylor, 2011). Endometrial receptivity is a consequence of spatiotemporal actions of estrogen and progesterone, culminating during the ‘window of implantation’ (Psychoyos, 1973; Yoshinaga, 1988; Achache and Revel, 2006). However, high rates of infertility reflect the nescience of the implantation process, and a comprehensive understanding of the reciprocal interaction between embryo and uterus is still missing.

Calcium is a key player in a variety of important events including fertilization, decidualization and implantation (Sakoff and Murdoch, 1994; Santella et al., 2004; Whitaker, 2006; Banerjee et al., 2009; Ruan et al., 2012). Notably, we recently described the functional expression of calcium-conducting transient receptor potential (TRP) channels TRPV2, TRPV4, TRPC6 and TRPM7 in human endometrial stromal cells (De Clercq et al., 2015). In mammals, the TRP superfamily of the cation conducting ion channels counts 28 members and is involved in a plethora of physiological functions. Hence, considering TRP channels as important cellular sensors, their previously described functional expression in human stromal cells renders them good candidates for intercellular signaling during processes as decidualization and embryo implantation (Voets et al., 2005). However, ethical and practical consideration limit further in-depth mechanistic research in humans and have prompted the use of animal models. Although murine reproduction differs from human in several aspects, their shared features such as implantation during a restricted period of time, the fact that implantation results in stromal decidualization and haemochorial placentation has incited the use of mice as a model to study reproduction. In addition, the mouse model has a high reproducibility and allows for the use of transgenic animals. Nevertheless, the expression pattern of TRP channels in the murine endometrium remains unknown. Moreover, differences in the menstrual cycle and the estrous cycle make it difficult to compare the effect of the spatiotemporal regulation by ovarian hormones estrogen and progesterone. Here, we will apply a menstruating mouse model (MMM) (Greaves et al., 2014) in order to investigate whether manipulating hormonal levels in mice can mimic the human uterine expression profile that was previously published (De Clercq et al., 2015) and to what extent this differs from the expression in their natural estrous cycle. Furthermore, endometrial stromal cells and luminal epithelial cells both have their own distinct function in the implantation process, implying the possibility for a distinct expression profile. Endometrial epithelial cells are the first to make contact with the implanting embryo and have an important role in endometrial receptivity. The underlying stroma will undergo decidualization and promote the maternal-fetal interface. Therefore, primary cultures of murine endometrial epithelial cells (MEEC) and stromal cells (MESC) were investigated separately, both at the mRNA and at the functional level using qRT-PCR and calcium fluorimetry, respectively.

Materials and Methods

Animals

Female, 8–12 week-old C57BL/6 J mice (Janvier, France) were housed in filter-top cages with a maximum of five animals per cage under conventional conditions. They were fed ad libitum with food pellets and tap water, and kept under controlled conditions (23 ± 1.5°C, relative humidity 40–60%, 12:12 light/dark cycle).

Ethical approval

The ethical review committee for animal experiments at the Katholieke Universiteit Leuven (Belgium) approved the use of mice for this study (Project P174/2013).

Analgesia and anesthesia

All surgeries were performed using sterile instruments under general anesthesia induced using a mixture of 1–2% isoflurane (Iso-vet, Dechra veterinary products, Bladel, The Netherlands) in 100% oxygen. Prior to the surgeries, animals were shaved at the appropriate location, and the area was disinfected with an iodine-based solution (iodine solution isopropanolitica 1%, BDH Prolabo-VWR international, Amsterdam, The Netherlands) and 70% ethanol. In order to maintain body temperature, animals were kept on a heating pad during the surgeries. A s.c. injection of 0.05 mg/kg body weight buprenorphine (Vetergesic multidose, Ecuphar, Breda, The Netherlands) diluted in saline (0.9% NaCl Viaflo, Baxter, Utrecht, The Netherlands), was used as analgesia after ovariectomy and pellet implantation. Twenty-four hours after surgery, the animals were again injected with the above-mentioned analgesic.

The estrous cycle

Female C57BL/6 J mice (n = 4 animals/cycle phase) were sacrificed in different phases of the estrous cycle. Of each mouse, the left uterine horn was preserved in RNAlater RNA Stabilization Reagent (Qiagen, The Netherlands) for expression of TRP channels analysis by qRT-PCR. The right uterine horn was weighed. Proof for cycle phase was evaluated by vaginal smear examinations.

Vaginal smear examination

Proof for the cycle phase was provided by the evaluation of the vaginal cytology obtained by vaginal smear examination, which allows for tracking changes in morphology of desquamated vaginal epithelium in response to changes of blood concentrations of ovarian hormones. Phosphate buffered saline (PBS) (50 μl) was used to flush the vagina gently 3–5 times. The final flush containing the vaginal fluid was assembled on a glass slide. Thereafter, the smear was examined under a bright field microscope using a 10× objective. The estrous cycle phases were determined according to the proportion among nucleated epithelial cells, cornified epithelial cells and leukocytes observed in the vaginal smear. As essentially exemplified by Caligioni et al. (Caligioni, 2009), the estrous cycle is divided into four stages which were identified as follows (Supplementary data, Fig. S1): Proestrus: predominance of nucleated epithelial cells clustered or individually. Some cornified cells may have appeared. Estrus: this stage was characterized by clustered irregular cornified cells without nucleus. Metestrus: a mix of different cell types was observed with a plentitude of leukocytes. Diestrus: this stage was indicated when leukocytes prevailed.

The MMM

Female C57BL/6 J mice (n = 28) were ovariectomized and allowed to recover for 1 week. Animals were then sequentially administered with steroid hormones. Days 1, 2 and 3, all mice were s.c. injected daily (09–10 a.m.) with 100 ng of 17β-Estradiol (E2) (Sigma-Aldrich, Bornem, Belgium) in arachis oil. On Day 7 (2 p.m.), a progesterone (P4)-releasing silastic tube was implanted s.c. into the back of the mouse and a s.c. injection of 5 ng E2 was given on the same day and the subsequent 2 days (09–10 a.m.). On the final day of E2 injection, decidualization was induced (2 p.m.) by a vaginal intra-uterine injection of 100 μl/horn sesame oil using a 22 G blunted catheter (Insyte, Vialon, Becton Dickinson, Madrid, Spain). Four days later, the P4 pellet was removed to initiate P4 withdrawal. Uteri were harvested at different time points (Table I) and fat-tissue was removed. Of each mouse, the left uterine horn was preserved in RNAlater RNA Stabilization Reagent (Qiagen, The Netherlands) for expression of TRP channels analysis by qRT-PCR. The right uterus horn was weighed and fixed in 4% paraformaldehyde for at least 12 h, followed by a 1 h wash step in PBS (Gibco, Invitrogen, Belgium) and was eventually stored at room temperature in 70% ethanol until it was embedded in paraffin. The sequence of events is graphically shown in Fig. 1.

Figure 1

Experimental set up of the MMM. Sequence of experimental steps in the menstruation mouse model: endometrial proliferation and differentiation was stimulated by subjecting ovariectomized mice to a series of E2 injections and a P4 releasing implant. Four days after decidualization was induced by the injection of sesame oil into the uterus, P4 was withdrawn by removing the implants. Boxed day numbers indicate time points uteri were harvested. Ovex, ovariectomy; E2, 17β-estradiol; P4, progesterone; MMM, menstruating mouse model.

Open in new tabDownload slide

Table I

Open in new tab

Experimental groups.

Group .	Mimicked cycle phase .	Harvest .	Event .	# .	Referred to as .
1	Baseline	Day 0	7 Days after ovariectomy	6	Ovex
2	Proliferative phase	Day 4	1 Day after estrogen injections	6	E2
3	Early secretory phase	Day 9	2 Days after P4 pellet	6	E2 + P4
4	Late secretory phase	Day 11	2 Days after decidualization	4	Deci
5	Menstruation	Day 13	6 h After P4 pellet removal	5	Menses

Group .	Mimicked cycle phase .	Harvest .	Event .	# .	Referred to as .
1	Baseline	Day 0	7 Days after ovariectomy	6	Ovex
2	Proliferative phase	Day 4	1 Day after estrogen injections	6	E2
3	Early secretory phase	Day 9	2 Days after P4 pellet	6	E2 + P4
4	Late secretory phase	Day 11	2 Days after decidualization	4	Deci
5	Menstruation	Day 13	6 h After P4 pellet removal	5	Menses

P4; progesterone, E2; estradiol.

Table I

Open in new tab

Experimental groups.

Group .	Mimicked cycle phase .	Harvest .	Event .	# .	Referred to as .
1	Baseline	Day 0	7 Days after ovariectomy	6	Ovex
2	Proliferative phase	Day 4	1 Day after estrogen injections	6	E2
3	Early secretory phase	Day 9	2 Days after P4 pellet	6	E2 + P4
4	Late secretory phase	Day 11	2 Days after decidualization	4	Deci
5	Menstruation	Day 13	6 h After P4 pellet removal	5	Menses

Group .	Mimicked cycle phase .	Harvest .	Event .	# .	Referred to as .
1	Baseline	Day 0	7 Days after ovariectomy	6	Ovex
2	Proliferative phase	Day 4	1 Day after estrogen injections	6	E2
3	Early secretory phase	Day 9	2 Days after P4 pellet	6	E2 + P4
4	Late secretory phase	Day 11	2 Days after decidualization	4	Deci
5	Menstruation	Day 13	6 h After P4 pellet removal	5	Menses

P4; progesterone, E2; estradiol.

Ovariectomy

A small dorsal midline skin incision was made caudal to the posterior border of the ribs. Another small lateral incision was made in the peritoneum, opening the abdominal cavity. The ovary was gently exteriorized and removed. Next, a similar incision was made in the contralateral peritoneum and the procedure was repeated. The wound was closed and the skin was sutured using 6-0 non-absorbable monofilament polyamid suture (Ethicon, Norderstedt, Germany).

E2 injections

17β-Estradiol (Sigma-Aldrich, Belgium) was diluted in 96% ethanol in order to make a stock solution of 1 mg/ml. This stock solution was diluted 1:1000 in arachis oil to be injected daily as 100 ng/100 μl on the first 3 days of the MMM protocol. Prior to daily injection of 5 ng/100 μl on Days 7, 8 and 9, the stock solution was diluted 1:20 in ethanol followed by another dilution of 1:1000 in arachis oil.

Construction of progesterone pellet

P4 implants were adapted from Miligan and Cohen (188). Briefly, silastic tubes (1.57 mm inner diameter; Dow Corning, Seneffe, Belgium) were filled with crystalline progesterone (Sigma-Aldrich, Belgium), and sealed at the ends with multipurpose sealant (Dow Corning, Belgium), such that the functional length of each implant was 1 cm. Prior to its use, the implants were incubated overnight, for at least 12 h, at 37°C in PBS (Gibco, Belgium) complemented with 5% fetal bovine serum (FBS, Gibco, Belgium).

Pellet implantation and removal

Concisely, when under anesthesia, a small incision was made between the shoulder blades, and bluntly dissected creating a s.c. pouch. The progesterone pellet was positioned s.c., and the incision was subsequently closed by a 6-0 polyamid suture (Ethicon, Germany). On the day of the pellet removal, the suture was reopened and the pellet was removed.

Immunohistochemistry

Standard hematoxylin & eosin (H&E) staining was used for the assessment of the histological and morphological features of the uteri. Briefly, 4 μm sections were subjected to a series of deparaffinization and rehydration, respectively toluene and 100% ethanol. Nuclei were stained with Gill's haematoxylin (Prosan, Merelbeke, Belgium) during 4 min, followed by a few dips in acid alcohol (1% HCl in ethanol), and subsequently lithium carbonate (saturated in Aqua distillate (AD)). Cytoplasmic staining was achieved with eosin (Prosan, Belgium) during 3 min. Note that sections were rinsed in tap water and subsequently AD between each of the previous steps. Sections were then dehydrated in graded alcohols, cleared in xylene and coverslips were mounted with a histological mounting media (Depex mounting medium, BDH Prolabo, The Netherlands).

Isolation and culture of mouse endometrial epithelial and stromal cells

During 3 days prior to the start of the isolation protocol, mice (n = 5/culture) were injected s.c. with 100 ng E2 solution to synchronize the cycle of all the animals and to standardize the protocol. Before sacrificing the animals, a vaginal smear was performed to ascertain whether the standardization was successful. The isolation of MEEC and MESC was performed as described previously (De Clercq et al., JoVe, JoVE55168, in press). Briefly, the uteri were removed and cut open longitudinally. Then, uterine tissue was incubated in 10 ml Hanks Balanced Salt Solution 1X (HBBS, Gibco, Belgium) complemented with 100 U/ml penicillin and 100 µg/ml streptomycin (further referred to as HBSS+) containing 0.25% trypsin (Sigma-Aldrich) and 2.5% pancreatin (Sigma-Aldrich, Belgium) for 60 min at 4°C followed by 45 min at RT, and 15 min at 37°C. Next, uterine tissue was transferred to Dulbecco Modified Eagle Medium (DMEM)/Hams F-12 nutrient mixed 1:1 (Gibco, Belgium) enriched with 10% FBS (Gibco, Belgium), 0.5 mg/ml fungizone (Gibco, Belgium) and 100 mg/ml gentamicin (Gibco, Belgium) (further referred to as MESC medium) to stop the enzymatic activity. After 5 min, the uteri were transferred into 3 ml of ice-cold HBSS+ and vortexed for 10 s in order to release the epithelial sheets. This step was repeated three times. Next, the obtained epithelial cell suspensions were passed through a 100-micron cell strainer (Falcon, Fisher Scientific, Belgium) and centrifuged at 500 × g for 5 min. After discarding the supernatant, the MEEC pellet was resuspended in 12 ml HBSS+ and the upper 2 ml was removed after 5 min to separate remaining MESC based on differential size using gravity sedimentation. Afterwards, the suspension was centrifuged again and the pellet was resuspended in MEEC medium containing DMEM supplemented with 10% FBS, 25% MCDB-105 (Sigma, Belgium), 0.5 mg/ml fungizone, 100 mg/ml gentamicin and 5 mg/ml insulin (Sigma, Belgium). Subsequently, cells were seeded on collagen-coated coverslips (collagen I from rat tail, 50 μg/mL in 0.02 M acetic acid, BD biosciences, Belgium) and incubated at 37°C and 5% CO₂.

The residual uterine parts were incubated in 3 ml of 0.05% trypsin-EDTA (Gibco, Belgium) supplemented with 1 mg/ml collagenase (Sigma-Aldrich) for 30 min at 37°C, while constant shaking, in order to loosen the stromal cells. After incubation, the uterine parts were shaken gently for 10 s, rinsed in HBSS+ and transferred to 3 ml MESC medium. This step was then repeated another three times. Next, the uterine tissue was dispersed by gently rubbing in medium on a 40-micron cell strainer (Falcon, Belgium) and the obtained cell suspension was centrifuged at 500 × g for 7 min. The supernatant was discarded and the MESC pellet was resuspended in MESC medium and kept under the same conditions.

QRT-PCR experiments

Quantitative RT–PCR (qRT-PCR) experiments were performed on RNA isolated from whole uterine tissues and from primary cultures. RNA quantity was checked with the Thermo Scientific NanoDrop 1000 Spectrophotometer; the RNA quality was determined with the Experion RNA Analysis kit (Bio-rad, Belgium) and only samples with an RNA quality indicator  >7 were used for cDNA synthesis.

Uterine tissues

The tissue, kept in RNAlater, was homogenized by use of a power homogenizer (Polytron). Total RNA was extracted with TriPure Isolation Reagent (Roche, Germany) and subsequently used for cDNA synthesis using the High-Capacity cDNA Reverse Transcription Kit (Life Technologies Europe B.V., Ghent, Belgium).

Primary cultures

For the extraction of RNA out of the MEEC and MESC samples, the RNeasy Mini Kit (Qiagen, The Netherlands) was used according to manufacturer's guidelines. cDNA was synthesized with Ready-To-Go You-Prime First-Strand Beads (GE Healthcare Life Sciences, Belgium).

Triplicate cDNA samples (2.5× diluted) from each independent preparation were used in the StepOne PCR system (Applied Biosystems, Life Technology) using specific TaqMan gene expression assays for all TRP channels (Supplementary data, Table S1). The MS Excel application GeNorm 3.5 indicated β-actin (ACTB) and TATA Box Binding Protein (TBP) as the most stable endogenous controls for further analysis of the TRP expression in tissue samples, while TBP and phosphoglycerate kinase 1 (PGK-1) were most stable for analysis of MESC and MEEC. Data were shown as 2^−ΔCt (mean ± SEM) in which ΔCt = Ct _{TRP channel} – Ct _{geometric mean of endogenous controls}. The protocol consisted of 40 replication cycles. TRP channels with Ct values above 34 cycles were considered as non-detectable (n.d.).

Functional measurements

TRP pharmacology

Since selective pharmacological agents for TRP channels are limited, not all TRP channels have been tested for their functional expression by calcium (Ca²⁺) microfluorimetry. This has hampered the analysis of functional TRPV6, TRPM4, TRPM6 and TRPM7.

TRPV2 activity was assessed by the application of 50 µM Δ⁹-tetrahydrocannabinol (THC) (De Petrocellis et al., 2011). Responses were challenged with 2 µM of the nonspecific inhibitor ruthenium red (RR) (Hu et al., 2004; Leffler et al., 2007).

The functionality of TRPV4 was evaluated by stimulation with 10 nM GSK016790A (GSK), a potent and selective TRPV4 activator (Dunn et al., 2013). Responses were challenged with 100 nM HC-067047, a TRPV4 specific inhibitor (Everaerts et al., 2010a,b).

TRPC1/4 activity was assessed by the application of 250 nM (−)–Englerin A (Akbulut et al., 2015).

TRPC6 functionality was evaluated by the application of 100 µM 1-oleoyl-2-acetyl-glycerol (OAG), the membrane-permeable analog of diacylglycerol. To date, no specific TRPC6-inhibitors are commercially available.

The sensory TRP channels TRPV1, TRPA1, TRPM3 and TRPM8 were evaluated by stimulation with capsaicin (1 µM), allyl isothiocyanate (mustard oil (MO), 100 µM), pregnenolone sulfate (PS, 40 µM) or menthol (100 µM) (Vriens et al., 2014), respectively. Responses to MO were challenged with 10 µM of the specific TRPA1 antagonist HC-030031.

Measurement of intracellular Ca²⁺ concentration

For intracellular Ca²⁺ measurements, cells were incubated with 2 µM Fura-2 acetoxymethyl ester for 30 min at 37°C. Fluorescent signals were evoked during alternating illumination at 340 and 380 nm using a Lambda XL illuminator (Sutter instruments, Novato, USA), and recorded using an Orca Flash 4.0 camera (Hamamatsu Photonics Belgium, Mont-Saint-Guibert, Belgium) on a Nikon Eclipse Ti fluorescence microscope (Nikon Benelux, Brussels, Belgium). The imaging data were recorded and analyzed using NIS-elements software (Nikon). Absolute calcium concentrations were calculated from the ratio of the fluorescence signals at both wavelengths (F340/F380) after correction for the individual background fluorescence signals, using the Grynkiewicz equation (Grynkiewicz et al., 1985). The standard solution contained 150 mM NaCl, 2 mM CaCl₂, 1 mM MgCl_2, 10 mM d-glucose and 10 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) (pH 7.4 with NaOH). The solution for experiments without extracellular calcium was: 150 mM NaCl, 1.5 mM MgCl₂, 10 mM d-Glucose, 10 mM HEPES and 5 mM EGTA (pH 7.4 with HCl). Ionomycin (2 µM, Sigma) was applied at the end of every experiment as a positive control.

For all measurements, cells were considered responders if the calcium influx during agonist application exceeded 100 nM and when the highest value of the derivative of the calcium trace during the application of an activator exceeded at least 3 times the standard deviation of the derivative during basal conditions. Calcium amplitudes were calculated as the difference between the maximum calcium and basal calcium of responding cells during the application of an activator as described elsewhere (De Clercq et al., 2015). Only cells that responded to ionomycin at the end of the experiment were taken into account.

Reagents

Capsaicin, RR, pregnenolone sulfate, GSK1016790A (GSK), HC-067047, HC-030031 and allyl isothiocyanate were purchased from Sigma-Aldrich. Menthol was acquired from Merck Milipore. THC was kindly provided by G. Appendino. OAG was obtained from Calbiochem (Bio-Connect, Huissen, The Netherlands). (−)–Engelrin A was acquired from Phytolabs (Vestenbergsgreuth, Germany). Stock solutions were dissolved in dimethylsulfoxide, EtOH or MetOH.

Data analysis

For data display, the Origin 8.6 software package was used (OriginLab, Northampton, USA). Statistical analysis was performed using Graphpad Prism 5.01 (Graphpad software incorporated, La Jolla, CA, USA). Normality was tested using the D'Agostino-Pearson omnibus test. Results were considered to be statistically significant when P < 0.05. Statistical tests are given in figure legends. For correlations figures from supplementary figures, data was first analyzed using TRPM7 as a housekeeping gene in order to compare data sets that were analyzed with different housekeeping genes. Correlations were tested using the Spearman correlation test for nonparametric data. TRPM7 was excluded to calculate the Spearman correlation coefficient based on previously mentioned reason. On each graph, a theoretical perfect fit was drawn (intercept = 0, slope = 1) to compare expression levels.

For expression data, TRP channels were considered as non-detectable (n.d.) when the overall average Ct value ≥34, it was considered around the detection level when 30 ≤ Ct ≤ 34 and were considered to be expressed when Ct < 30. Statistical analysis were done only on those channels that were expressed (Ct ≤ 30) and was performed on the average ΔCt from three technical replicates for each sample. The analysis of the expression levels throughout the MMM or in the estrous cycle was done using one-way ANOVA with Bonferroni's multiple comparison Post-hoc test when data were normally distributed or with Kruskal–Wallis with Dunn's multiple comparison Post-hoc test for nonparametric data. The comparison of expression levels between MEEC and MESC was done using a two-way ANOVA with bonferroni multiple comparison Post-hoc test.

Results

The MMM to induce menstrual-like changes in murine endometrium

Important differences between the murine estrous cycle and the human menstrual cycle make it difficult to draw conclusions about similarities in uterine expression of TRP channels and their regulation. To investigate whether the expression profile of TRP channels in the human endometrium and its regulation by ovarian hormones is comparable in mice, the human menstrual cycle was mimicked by the use of the MMM (Fig. 2A). The succession of events was verified using vaginal smears, uterine weight and uterine morphology (Fig. 2B–D). Estrogen-induced proliferation of epithelial and stromal cells and engendered the thickening of the endometrium, hence, an increase in uterine weight (31.7 ± 9.8 mg, average weight gain = 23.1 mg). Two days after decidualization was induced, the stromal cells underwent differentiation from fibroblast-like cells into round decidual cells resulting in a massive expansion of the endometrium and a significant increase in uterine weight (153.8 ± 79.7 mg, P < 0.001, average weight gain = 145.5 mg) (Fig. 2C). Interestingly, uterine morphology changed dramatically after the induction of decidualization indicated by the differentiation into round decidual cells starting at the mesometrial side of the uterus, the thinning of the myometrium and the absence of endometrial glands in the decidualized tissue. Menstruation was induced by the removal of the progesterone-releasing pellet 4 days after decidualization, resulting in vasoconstriction and atrophy triggering ischemic and hemorrhagic phenomena, and subsequent shedding of the endometrium accompanied with overt menstrual like-bleeding (Fig. 2D, insert). Overall, these results suggest that manipulation of the hormonal levels can result in menstrual-like modifications in the murine endometrium.

Figure 2

Induced menstrual-like cycle in mouse resembles human menstrual cycle. (A) MMM and time line showing the different time points of harvest, c.f. Fig. 1. (B) Transition of estrus cycle phases assessed by vaginal smear in response to hormonal changes. Insert illustrates endometrial breakdown. M = metestrus, D = diestrus, E = estrus, P = proestrus (C) Wet weight of one uterine horn at the time of sacrifice, mean + SEM (n = 4–6). **P < 0.01, ***P < 0.001, Kruskal–wallis test. (D) Morphological changes shown by H&E staining. Colored frame corresponds to different time points of harvest in the menstrual-like cycle. E2, estrogen; P4, progesterone; P, perimetrium; E, endometrium; M, myometrium; G, glands; L, lumen; DE, decidualized endometrium; ovex, ovariectomy; deci, decidualization; scale bar = 100 µm.

Open in new tabDownload slide

TRP channels expression profile in whole uterus sample throughout the induced menstrual cycle

Using qRT-PCR analysis, the expression profile of TRP channels in uteri harvested at different times during the MMM was investigated (Fig. 3). The mRNA expression of TRPV2, TRPV4, TRPV6, TRPC1, TRPC3, TRPC4, TRPC6, TRPM4, TRPM6 and TRPM7 was above the detection limit whereas TRPA1, TRPC2, TRPC5, TRPM1, TRPM2, TRPM3 and TRPM5 expression was around the detection level (30 ≤ Ct ≤ 34). In contrast, the mRNA levels of TRPV1, TRPV3, TRPV5, TRPC7 and TRPM8 were below the detection level (Ct > 34). The expression level of TRPC6 and TRPM7 showed little variation throughout the different phases of the induced menstrual-like cycle whereas other channels had a variable expression level, suggesting their possible regulation by ovarian hormones.

Figure 3

Quantitative RT–PCR analysis on uteri harvested at different time points in the MMM. mRNA levels (mean + SEM) of the members of the TRPV (A), TRPM (B) and TRPC (C), relatively quantified to the geometric mean of housekeeping genes ACTB and TBP, after ovariectomy (Day 0, n = 6), after 3 days of E2 (Day 4, n = 6), during E2 and P4 (Day 9, n = 6), 2 days after decidualization (Day 11, n = 4) and at menstruation (Day 13, 6 h after pellet removal, n = 5). TRPV1, TRPV3, TRPV5, TRPC7 and TRPM8 were below the detection limit. When significantly expressed during the cycle phases, the P-value for each TRP channel is indicated. More detail can be found in Supplementary data, Table S2. Data are presented as mean ± SEM. n.d., non-detectable (average threshold cycle >34); E2, estrogen; P4, progesterone; ovex, ovariectomy; deci, decidualization; ACTB, β-actin; TBP, TATA Box Binding Protein.

Open in new tabDownload slide

TRP channels expression profile in whole uterus sample throughout estrous cycle

To assess the uterine expression profile of TRP channels during the natural estrous cycle, uteri were isolated at different phases of the cycle, which were assessed by vaginal smear examination (Supplementary data, Fig. S1). Using qRT-PCR, the expression levels of TRP channels during these different phases were assessed (Fig. 4). The mRNA expression of TRPV2, TRPV4, TRPV6, TRPC1, TRPC2, TRPC3, TRPC4, TRPC6, TRPM4, TRPM6, TRPM7 was above the detection limit (Ct < 30) whereas TRPA1, TRPC5, TRPM2, TRPM3 and TRPM5 expression was around the detection level (30 ≤ Ct ≤ 34). In contrast, the mRNA levels of TRPV1, TRPV3, TRPV5, TRPC7, TRPM1 and TRPM8 were below the detection level (Ct > 34). Expression level of TRPM7 was high throughout the different phases of the estrous cycle, whereas other channels had a variable expression. Interestingly, the expression level of TRPV6 was the highest during the proestrus phase suggesting a possible upregulation or modulation by estrogen.

Figure 4

Relative mRNA expression of TRP channels throughout natural murine estrous cycle. mRNA levels (mean ± SEM) of the members of the TRPV (A), TRPM (B) and TRPC (C), relatively quantified to the geometric mean of housekeeping genes ACTB and TBP. TRPV1, TRPV3, TRPV5, TRPC7, TRPM1 and TRPM8 were below the detection limit. When significantly expressed during the cycle phases, the P-value for each TRP channel is indicated. *P < 0.05; **P < 0.01. Data are presented as mean ± SEM. n.d., non-detectable (average threshold cycle >34).

Open in new tabDownload slide

Expression of TRP channels in the murine endometrial cells

The endometrium consists of a monolayer of epithelial cells lining the uterine lumen and a layer of stromal cells, which varies in thickness according to the cycle phase. Given the diverse function of stromal and epithelial cells in reproduction, it is possible that both cell types have a different expression pattern of TRP channels. Therefore, the expression profile of TRP channels was independently investigated and primary cultures of MEEC and MESC were started from murine uterus. Next, mRNA was isolated (n = 3–4) and the mRNA expression pattern of TRP channels was further investigated in MEEC and MESC (Fig. 5). Interestingly, TRPV2 mRNA was highly expressed in MESC, with mRNA levels that were ~20-fold higher compared to MEEC. Furthermore, the expression level of TRPC4 was significantly higher in MESC compared to MEEC (eleven times higher expression). In contrast, the mRNA expression of TRPV4 (seven times more expressed), TRPV6 and TRPM6 was more pronounced in MEEC compared to MESC. Notably, the mRNA expression of TRPV6 and TRPM6 was exclusively detected in MEEC. No significant difference was detected in mRNA levels of TRPC1, TRPC3, TRPC6, TRPM4 and TRPM7 between MESC and MEEC. The expression of sensory TRP channels TRPA1, TRPV1, TRPM3 and TRPM8 was below detection level in both cell populations. Taken together, these results illustrate a distinct expression pattern between epithelial and stromal cells.

Figure 5

TRP channel expression pattern in MEEC and MESC. Quantitative RT–PCR (mean + SEM) on freshly isolated primary MEEC and MESC obtained from mice with E2 injection for 3 days prior to isolation (n = 3–4). Data are presented as mean ± SEM. mRNA levels are quantified relative to the geometric mean of housekeeping genes TBP and PGK. ***P < 0.001 with two-way ANOVA with Bonferroni multiple comparison Post-hoc test. MEEC, murine endometrial epithelial cells; MESC, murine endometrial stromal cells. PGK, phosphoglycerate kinase.

Open in new tabDownload slide

Functional expression of TRP channels in mouse endometrial epithelial cells

The functional expression of TRPV2 in MEEC was evaluated in Fura2-based microfluorimetry using the TRPV2-agonist THC (50 µM). Application of THC resulted in an increase in intracellular calcium concentration ([Ca²⁺]_I) (Δ[Ca²⁺]_i = 318 ± 125 nM) but only in a very small fraction of cells in the epithelial culture (2.1 ± 1.0%; 61 out of 3036 cells) (Fig. 6A and F and Fig. 8). In addition, no increase in intracellular calcium was observed when extracellular calcium was omitted (Supplementary data, Fig. 2A), suggesting that the THC-induced calcium flux originates from extracellular calcium.

Figure 6

Functional expression of TRP channels in MEEC assessed with Ca²⁺ imaging. (A) Example traces of THC (50 µM)-induced intracellular (IC) Ca²⁺ changes ([Ca²⁺]_ic). (B) Example traces of (−)–Englerin A (EA, 100 nM)-induced intracellular Ca²⁺ changes in MEEC. (C) Example traces of GSK1016790A (GSK, 10 nM)-induced intracellular Ca²⁺ changes ([Ca²⁺]_ic), which could be blocked by the TRPV4 inhibitor HC-06704 (D). (E) Example traces of OAG (100 µM)-induced intracellular Ca²⁺changes ([Ca²⁺]_ic) MEEC. (F) Percentage of responding MEEC cells to shown activators. Data are presented as mean ± SD. Example traces are nonproportionally chosen and show a mix of responding and non-responding cells, independent on the exact percentage of responders. Ionomycin (Iono, 2 µM) was added at the end of every experiment as a positive control. OAG, 1-oleoyl-2-acetyl-glycerol. THC, Δ⁹-tetrahydrocannabinol.

Open in new tabDownload slide

Functional activity of TRPC4 or TRPC1/4 heteromers was assessed by the use of Englerin A (EA) and could be detected only in 1.7 ± 1.2% of MEEC (22 out of 1382 cells), inducing a Ca²⁺ influx of (ΔCa²⁺ = 300 ± 138 nM) (Fig. 6B and F and Fig. 8). No increase in intracellular calcium was observed when extracellular calcium was omitted (Supplementary data, Fig. 2B)

Stimulation of MEEC by the specific agonist of TRPV4 GSK016790A, induced a robust and reversible Ca²⁺ influx (ΔCa²⁺ = 385 ± 88 nM) in 17.7 ± 10% of MEECs (Fig. 6C and F and Fig. 8), while the GSK-induced calcium influx was absent when extracellular calcium was omitted (Supplementary data, Fig. S2C). In addition, the GSK-induced responses in MEEC could be blocked in the presence of HC-067047 (Fig. 6D). The HC067047-induced block of [Ca²⁺]_i increase, was reversible and GSK-induced calcium responses recovered afterwards in 28 ± 13% of MEEC.

Finally, application of OAG, a TRPC6 activator, resulted in an increase in [Ca²⁺]_i (ΔCa²⁺ = 211 ± 35 nM) in a minority of the epithelial cell culture (0.6 ± 0.3%, 8 out of 1240) (Fig. 6E and F and Fig. 8). In the absence of extracellular calcium stimulation no Ca²⁺ influx was observed in MESC after stimulation by OAG (Supplementary data, Fig. S2D).

Stimulation of MEEC with capsaicin, allyl isothiocyanate, menthol or pregnenolone sulfate, specific activators of respectively TRPV1, TRPA1, TRPM8 and TRPM3 did not induce a significant rise in [Ca²⁺]_i, while a normal Ca²⁺ response was observed after stimulation by ionomycin in MEEC (Supplementary data, Fig. S3). These observations are in line with the low RNA expression profile of these sensory TRP channels observed in the qRT-PCR experiments.

Functional expression in mouse endometrial stromal cells

Stimulation of MESC with THC resulted in a robust and reversible Ca²⁺ influx (ΔCa²⁺ = 579 ± 370 nM) in 14.1 ± 6.5% of the cells (845 out of 6028 cells) (Fig. 7A and H and Fig. 8). Responses to THC stimulation were absent when Ca²⁺ was omitted from the extracellular solution (Supplementary data, Fig. S4A). In addition, THC-induced Ca²⁺ influx could be blocked by pre-application of the nonspecific TRPV inhibitor RR (Fig. 7B). The RR-induced block was reversible and Ca²⁺ responses to THC recovered afterwards in 26.9 ± 4.4% in MESC when the application of the blocker was stopped. On the contrary, THC-induced calcium influx in MESC could not be prevented by the co-application of the cannabinoid receptor blockers CB1 and CB2 using AM630 and AM251, respectively (Fig. 7C).

Figure 7

Functional expression of TRP channels in MESC assessed with Ca²⁺ imaging. (A) Example traces of THC (50 µM)-induced intracellular (IC) Ca²⁺ changes ([Ca²⁺]_ic), which could be blocked by the nonspecific TRPV inhibitor  RR (2 µM) (B) but not by CB1 and CB2 blockers AM630 and AM251, respectivly (C). (D) Example traces of (−)–Englerin A (EA)-induced intracellular Ca²⁺ changes ([Ca²⁺]_ic). (E) Example traces of GSK1016790A (GSK)-induced intracellular Ca²⁺ changes ([Ca²⁺]_ic), which could be blocked by the TRPV4 inhibitor HC-06704 (F). (G) Example traces of OAG (100 µM)-induced intracellular Ca²⁺changes ([Ca²⁺]_ic). (H) Percentage of responding MESC cells to shown activators. Data are presented as mean + SD. Example traces are nonproportionally chosen and show a mix of responding and non-responding cells, independent on the exact percentage of responders. Ionomycin (Iono, 2 µM) was added at the end of every experiment as a positive control. Ruthenium Red, RR.

Open in new tabDownload slide

Figure 8

Percentage of responders in MEEC and MESC. (A) Percentage of responders to the shown activators in MEEC and MESC. *, **, ***: P < 0.05; P < 0.01, P < 0.001 two-way ANOVA with Bonferroni multiple comparison Post-hoc test. Data are presented as mean + SD.

Open in new tabDownload slide

Functional activity of TRPC4 or TRPC1/4 heteromers was assessed by the use of EA and could be found in 32.2 ± 10.2% of MESC (619 out of 1844 cells) (Fig. 7D and H and Fig. 8). In the absence of extracellular calcium no Ca²⁺ influx was observed in MESC after stimulation by EA (Supplementary data, Fig. S4B).

Stimulation of MESC by GSK016790A-induced a robust Ca²⁺ influx (ΔCa²⁺ = 344 ± 205 nM) in 4.5 ± 2% of the MESC (283 out of 3129 cells) (Fig. 7E and H and Fig. 8). The GSK-induced calcium responses were not observed in the absence of extracellular Ca²⁺ (Supplementary data, Fig. S4C) and could be blocked by HC-067047 (Fig. 7F). The HC-induced block was reversible and responses to GSK recovered in 3.8 ± 1% of MESC.

Finally, application of OAG resulted in a robust Ca²⁺ influx (ΔCa²⁺ = 488 ± 253 nM) in 19.6 ± 10.2% of all MESC (162 out of 1265) (Fig. 7G and H and Fig. 8). In the absence of extracellular calcium no Ca²⁺ influx was observed in MESC after stimulation by OAG (Supplementary data, Fig. S4D).

Stimulation of MESC with capsaicin, menthol or pregnenolone sulfate, specific activators of respectively TRPV1, TRPM8 and TRPM3 did not induce a significant rise in [Ca²⁺]_i, while a normal Ca²⁺ response was observed after stimulation by ionomycin in MESC (Supplementary data, Fig. S5). Allyl isothiocyanate, the activator for TRPA1, caused Ca²⁺ influx in 4.5 ± 2.3% of MESC (Supplementary data, Fig. S5B). However, these responses could not be blocked by HC030031 (10 µM), a specific blocker for TRPA1 and were found in 4.1 ± 1.1% of MESC during the co-application of MO and HC030031, suggesting that allyl isothiocyanate-induced calcium responses are not mediated by TRPA1. These observations are in line with the low RNA expression profile of sensory TRP channels observed in the qRT-PCR experiments.

Discussion

Calcium is an important cation, essential for living organisms and plays an important role as secondary messenger in many cellular processes, including reproduction (Sakoff and Murdoch, 1994; Santella et al., 2004; Banerjee et al., 2009). However, the knowledge about possible calcium-conducting ion channels in the reproductive system is limited. Considering TRP channels as important cellular sensors (Voets et al., 2005), their previously described functional expression in human endometrial stromal cells renders them good candidates for mediating intercellular signaling during processes such as decidualization and embryo implantation. However, ethical and practical considerations limit further mechanistic research concerning implantation in humans. In this study, we assessed the expression of TRP channels during a natural estrous cycle and during an induced-menstrual cycle to evaluate whether the expression pattern in mouse uterus resembles the pattern found in human endometrial biopsies during the menstrual cycle. In addition, we showed for the first time the molecular and functional expression of TRP channels in endometrial epithelial cells, demonstrating a very distinct expression pattern in the epithelial and stromal cellular compartment.

Comparison of the RNA expression pattern in murine and human endometrium is not straightforward, since the estrus cycle in mice is substantially different from the menstrual cycle in humans. Whereas the variance in hormonal levels of the reproductive cycle is very similar, both species differ in cycle length, induction of decidualization and appearance of menses. Therefore, we decided to use a MMM to induce a menstrual-like cycle in mice in order to compare the expression profile during events as decidualization and menstruation. Evidence for an effective succession of events was provided by the assessment of the estrus cycle, the uterine wet weight and the uterine morphology. The proliferative phase was then assigned to Day 4, the early luteal phase as Day 9, which coincides with the ‘window of implantation’ as at this time point the decidualization stimulus is given, the late luteal phase as Day 11, and the menstrual phase was assigned to Day 13, 6 h after pellet removal. Here, we found the RNA-expression of TRPV2, TRPV4, TRPV6, TRPC1, TRPC3, TRPC4, TRPC6, TRPM4, TRPM6 and TRPM7 in the total murine uterus, and this expression level was mostly higher after estrogen activation. In 1984, Finn and Pope were the first to induce menstruation in a mouse model reflecting endometrial breakdown as a consequence of progesterone withdrawal from artificially decidualized endometrium (Finn and Pope, 1984). A first step in this protocol is excluding endogenous ovarian hormones by ovariectomy. In this study, we assessed the expression of TRP channels in the uteri of ovariectomized mice. Hereby, it is shown that several TRP channels are also expressed under ‘basal’ conditions what could indicate that their expression is not exclusively regulated by reproductive hormones. Finn and Pope additionally postulated three fundamental requirements for the occurrence of menstrual-like changes in non-menstruating species. First off all, the endometrium should be prepared by the sequential administration of estrogen and progesterone in ovariectomized animals. Priming the endometrial cells with estrogen is generally mediated by the injection of 100 ng on 3 consecutive days in order to mimic the estrogen-dependent proliferative phase of the menstrual cycle. Unlike other TRP channels, TRPV6 expression tremendously increased during the induced-proliferative phase. The estrogen-induced upregulation of TRPV6 was also described in previous studies in the mouse uterus (Lee and Jeung, 2007; Yang et al., 2011). In addition, RNA expression of TRPV2, TRPC4 and TRPM4 in the whole uterus was increased after estrogen injections. As progesterone counteracts the mitogenic effect of estrogen, the uterine weight decreased and further proliferation of the endometrium was limited. Interestingly, the expression of TRPV2, TRPV6, TRPC4 and TRPM4, whose expression increased after E2 dosing, showed a decreased expression 2 days after implanting of the progesterone pellet. Remarkably, the expression of TRPV4 in the endometrium was unaltered after application of progesterone. This result is different from what is earlier described in human airways and mammary gland epithelial cells where TRPV4 expression was regulated by progesterone (Jung et al., 2009). A second requirement is that the process of decidualization should be triggered in the stroma of the endometrium. Decidualization is a response of endometrial stromal cells to high levels of progesterone, in preparation of accepting a competent embryo. As opposed to humans, decidualization in non-menstruating species will be initiated only in the presence of an implantation stimulus. Throughout the years, the most commonly used stimulus to artificially induce decidualization is the intra-uterine injection of oil in hormonally sensitized ovariectomized animals. Given that decidualization is reflected as the massive expansion of the endometrium, the uterine wet weight increased significantly and endometrial stromal cells differentiated into round decidual cells. Furthermore, endometrial glands are absent and the myometrium is thinner. It has been shown that TRPC channels are markedly expressed in the myometrium and play an important role during pregnancy and labor (Yang et al., 2002; Babich et al., 2004). The fact that the myometrium becomes thinner throughout the MMM could explain the decreased expression levels of some TRPC channels at Days 11 and 13 of the MMM. The last requirement to introduce menstrual-like changes is the withdrawal of progesterone support of the artificially decidualized endometrium, achieved by the removal of the progesterone-releasing pellet. Indeed, vaginal smear examination also showed red blood cells, indicating the shedding of the endometrium and overt menstruation. Although differences in the MMM protocols are being observed, different studies all concluded that menstruation in mice resembles the human situation with regard to degeneration of decidual cells, shedding of the endometrium, leukocyte infiltration, gene expression and involvement of matrix metalloproteinases (Brasted et al., 2003; Xu et al., 2007; Kaitu'u-Lino et al., 2012; Menning et al., 2012; Rudolph et al., 2012; Wang et al., 2013).

The expression pattern of TRP channels found in whole murine uteri resembles the pattern found in human endometrial biopsies (De Clercq et al., 2015). To compare the expression in both species, the human data were reanalyzed using TRPM7 as a same housekeeping gene and the human and murine TRP channel expression was plotted for each cycle phase (Supplementary data, Fig. S6). Important to note is that identical TRP channels are present or absent both in human and murine tissue. Moreover, for all cycle phases, a significant positive correlation was found (Spearman correlation 0.67 for follicular phase compared to E2 phase, 0.73 for early luteal phase compared to E2 + P4 phase, 0.71 for late luteal phase compared to decidualization phase and 0.72 for the menstruation phases; P < 0.01). Every graph contains a theoretical perfect fit which would mean that a TRP channel is equally expressed in both tissues when this point lies on the line. Since most points are not on the line, the relative expression in human tissue is different from murine tissue. One possible explanation is that the murine expression was evaluated in the whole uterus, including the myometrium, while the human expression was assessed in endometrial biopsies. In addition, it is possible that the manipulations used in the MMM resulting in menstruation are events that naturally do not occur in normal murine uterus, which implies that other mechanisms may also be involved. Nevertheless, these results suggest that, in spite of interspecies differences, the same TRP channels were found to be expressed in human endometrium and mouse uterus, albeit at a different level, suggesting that the mouse could be a valuable model to investigate the role of TRP channels in reproduction.

Furthermore, the expression of TRP channels was assessed during the natural estrous cycle. Interestingly, a similar RNA expression profile was identified in murine endometrium artificially exposed to sexual hormones compared to endometrium derived from the natural estrous cycle. To exemplify this, the expression of TRP channels during the proestrus phase was plotted against the E2 phase of the MMM and a positive significant correlation could be observed (spearman correlation = 0.95, P < 0.0001) (Supplementary data, Fig. S7). These results could suggest that the MMM is a representative model to investigate the expression profile in reproductive cycle of mouse and human. In addition, the hormonal regulation in both models is very comparable, e.g. fluctuations in TRPV6 expression correspond to the fluctuations of estrogen during the estrous cycle.

In the second part of this study, the expression of TRP channels was assessed separately for epithelial and stromal cells. The first step aims to isolate the first layer of the uterus, hence the epithelium, the next step will isolate the layer underneath, namely the stroma. Interestingly, a very distinct expression pattern could be observed between the two cell layers. A first difference was the high and exclusive expression of TRPV6 in MEEC. However, it must be noted that this high expression is in part induced by the daily injection of estrogen 3 days prior the cell isolation. The upregulation of TRPV6 by estrogen was previously shown (Lee and Jeung, 2007; Yang et al., 2011) and could be verified in this study where TRPV6 expression was highly expressed compared to TRPM7 in the murine uterus at Day 4 of the MMM. Interestingly, TRPM6 expression was observed only in MEEC. These results are in line with earlier reports showing that TRPM6 is mainly expressed in epithelial cells, and that its expression is regulated by estrogen (Groenestege et al., 2006). Possibly, upregulation of TRPM6 and the consequent increase of magnesium influx could result in increased cell proliferation (Ikari et al., 2008). The expression of TRPV6 and TRPM6 could provide the required calcium and magnesium exchange in the uterus, respectively. Nonetheless, the lack of specific pharmacology (activators, inhibitors) of TRPV6 and TRPM6 limits further functional assessment with calcium imaging experiments. Secondly, TRPV2 and TRPV4 showed an inverse expression pattern in MEEC and MESC, making them reasonable candidate proteins to act as specific markers of the two distinct cell types. Using THC as an activator for TRPV2, its expression was found in merely 2% of MEEC whereas this was 14% in MESC. Moreover, as MEEC and MESC cultures are not 100% pure, it is likely that the low number of THC responders in MEEC might be caused by stromal cells that reside in-between the MEEC. Interestingly, we could observe TRPV2 functionality in 22% of MESC after the co-application of RR and THC compared to the applying THC alone, suggesting that THC might increase TRPV2 surface expression, a process that was previously shown for Cannabidiol, another TRPV2 agonist (Reichhart et al., 2015). In addition, THC responses were not influenced by the presence of the CB receptor antagonists, indicating that the THC-induced calcium influx was not mediated by activation of the cannabinoid receptors.

Of all members of the TRPC family, mRNA of TRPC5 and TRPC7 was not detectable in the whole murine uterus. Concordant with other studies, TRPC1 and TRPC4 are the most abundant at the mRNA level, followed by TRPC6 and TRPC3 (Dalrymple et al., 2002; Yang et al., 2002; Babich et al., 2004; Ku et al., 2006). These channels were also differentially expressed in MEEC compared to MESC. The functionality of TRPC4 or TRPC1/4 heteromers was evaluated using EA, a specific activator for TRPC4 or TRPC4/TRPC5 or TRPC4/TRPC1 (Akbulut et al., 2015; Carson et al., 2015). However, as no mRNA for TRPC5 was observed, we attribute the EA-induced calcium influx to activation of TRPC4 homomers or TRPC4/1 heteromers. Functional TRPC6 expression was only identified in MESC. Although no significant difference was observed in the expression levels of TRPC6 in MEEC and MESC, an important discrepancy was found in the number of OAG-responding cells. Amongst TRPC channels, TRPC3, TRPC6 and TRPC7 are all activated by DAG or the membrane-permeable analog OAG (Hofmann et al., 1999). It is possible that TRPC6 forms homo- and heterotetramers within the confines of the TRPC3/6/7 subfamily (Hofmann et al., 2002). Since no mRNA was found for TRPC7, we only consider the possibility of TRPC3/TRPC6 heterotetramers. Such heteromers may give a possible explanation for the lack of OAG-induced calcium transients in MEEC, since no TRPC3 mRNA was found in these cells. Indeed, OAG-induced calcium influx was found in a minority (<1%) of MEEC whereas the number of OAG-responders was higher (circa 14%) in MESC.

The opposite was found for TRPV4 functionality, which was found in almost 18% in MEEC and only 4.5% in MESC. Concordantly, TRPV4 functionality was described in many other epithelial cells such as urothelium (Everaerts et al., 2010a,b) and bronchial epithelial cells (Alenmyr et al., 2014). Furthermore, it has been described in vascular endothelial cells where it is involved in shear stress-induced vasodilation (Vriens et al., 2005; Hartmannsgruber et al., 2007). The calcium influx via TRPV4 is contributing to mechanosensation, and could therefore play a role in the early stages of embryo implantation since a physical interaction was shown between the blastocyst and the epithelium (Enders and Schlafke, 1978).

These results reveal distinct expression patterns for MEEC and MESC, which may contribute to their specific and unique function in the reproduction process. It should be noted that a larger proportion of TRPV2, TRPV4 and TRPC6 expressing cells was reported for human endometrial stromal cells (ESC) (De Clercq et al., 2015). However, this difference can be due by the fact that a different, more strict, approach was used in this study to differentiate responders from non-responders. Nevertheless, the expression patterns of mouse ESC and human ESC were plotted and a positive correlation was found (spearman correlation = 0.70, P < 0.05) meaning that mouse and human stromal cells have comparable expression patterns (Supplementary data, Fig. S8).

Altogether, this study reports for the first time a fingerprint of the expression pattern of TRP channels in the murine endometrium, including epithelium and stroma (Fig. 9). We have identified mRNA expression of TRPV6 and TRPM6 and functional expression of TRPV4 as characteristic members of endometrial epithelial cells in mouse. In addition, TRPV2 and members of the TRPC family (TRPC1, TRPC3, TRPC4 and TRPC6) showed functional expression exclusively in endometrial stromal cells. Other TRP channels, including TRPM4 and TRPM7, showed a similar expression pattern in epithelial and stromal cells. Most other members of the TRP superfamily, including the sensory TRPs, did not show any detectable expression in the endometrium. By the use of a MMM and an isolation technique for primary endometrial epithelial and stromal cells, the physiological role of TRP channels in MEEC and MESC in the complex process of decidualization and implantation can be further studied in future work.

Figure 9

Cartoon of TRP channel expression in mouse endometrium. Summary of TRP channels detected in epithelial and stromal cells of the mouse uterus. TRP channels shown in green are proven to be expressed on mRNA and functional level. TRP channels shown in black were found on mRNA level but were not further investigated. When mRNA was found but no functionality could be observed, the channel was no longer shown.

Open in new tabDownload slide

In conclusion, this work showed that the expression pattern of TRP channels in human uterine samples is similar to the murine endometrium during the induced menstrual cycle in a murine menstruation model or the natural estrous cycle. Furthermore, an adapted protocol was described for primary murine endometrial cell cultures of epithelial and stromal cells. These primary cell cultures have led to the characterization of TRP channels in murine endometrium and to the generation of a fingerprint of the TRP channel expression in epithelial and stromal cells. The distinct expression pattern is in line with the different role of MEEC or MESC in reproduction and makes these channels interesting candidates for further investigations.

Supplementary data

Supplementary data are available at Human Reproduction online.

Acknowledgements

We thank all members of the Laboratory of Experimental Gynaecology and Obstetrics and the members of the Laboratory of Ion Channel Research for helpful discussions. We are very grateful to G. Appendino for providing us with chemical compounds.

Authors’ roles

K.D.C.—substantial contribution to conception, design, acquisition and interpretations of data, drafting the article and final approval of the version to be published. C.V.d.E.—substantial contribution to the acquisition of data, revising the article critically for important intellectual content and final approval of the version to be published. A.H.—substantial contribution to the acquisition of data, revising the article critically for important intellectual content and final approval of the version to be published. R.V.B.—substantial contribution to the acquisition of data and final approval of the version to be published. T.V.—substantial contribution to the acquisition and interpretations of data, revising the article critically for important intellectual content and final approval of the version to be published. J.V.—substantial contribution to conception and design and acquisition and interpretations of data, revising the article critically for important intellectual content and final approval of the version to be published.

Funding

Research Foundation-Flanders (G.0856.13 N to J.V.); Research Council of the Katholieke Universiteit Leuven (OT/13/113 to J.V. and T.D.); Planckaert-De Waele fund (to J.V.); Fonds Wetenschappelijk Onderzoek Belgium (to K.D.C. and A.H.).

Conflict of interest

None of the authors have any conflict of interests.

References