Introduction

A transient window of uterine receptivity ensures that embryos implant in an optimal endometrial environment [1]. Failure to establish [2] or premature closure [1, 3, 4] of the implantation window is thought to be a major cause of infertility, which affects ~ 15% of couples. Conversely, prolonged receptivity may lead to out-of-phase implantation and miscarriage [3, 4] or recurrent pregnancy loss [1, 4, 5]. The stromal signals responsible for terminating the window of implantation are not well characterized. A putative candidate is Left-Right Determination Factor 2 (LEFTY2), a cytokine highly expressed by decidualizing stromal cells during the late secretory phase of the menstrual cycle [6], i.e., following closure of the window of implantation and prior to menstruation [7]. LEFTY2, initially designated endometrial bleeding associated factor (EBAF), is a secreted ligand of the transforming growth factor-beta (β) superfamily of proteins [7, 8]. Following proteolytic processing of the secreted precursor, LEFTY2 acts as an antagonist of the Nodal signaling pathway by interfering both with the binding of NODAL to the activin receptor and the formation of a receptor complex [9]. During the late luteal phase of the menstrual cycle, strong LEFTY2 immunoreactivity was found in the stroma and to a lesser extent in the endometrial glands [10]. Furthermore, LEFTY2 has also been detected in the endometrial fluid of fertile women, indicating that LEFTY2 is secreted into the lumen of the uterus. Interestingly, induction of the proprotein convertase (PC) 5/6, which processes LEFTY2, is triggered in the mouse uterus in response to artificial decidualization with oil or upon mechanical injury [11]. Endometrial LEFTY2 is elevated during the receptive phase in some patients with “unexplained infertility” suggesting that dysregulation of LEFTY2 contributes to infertility [12, 13]. Further, in vivo gene transfer of Lefty2 in the mouse uterus leads to implantation failure [14]. However, the mechanisms underlying the negative impact of LEFTY2 on endometrial receptivity remain largely unknown.

Several implantation events, including blastocyst-endometrium adhesion [15], regulation of growth factor signaling [16], transcription factor activity [17], epithelial tight junctions [18], protease activity [19], cyclooxygenase 2 (COX2)-dependent prostaglandin production [20, 21], and epithelial transport [22, 23] are regulated by Ca2+ signaling. Embryo-derived tryptic serine proteases have emerged as important implantation signals, capable of inducing COX2-dependent prostaglandin E2 (PGE2) production in response to cytosolic Ca2+([Ca2+]i) oscillations [21]. [Ca2+]i is tightly regulated by several mechanisms, including Ca2+ release from intracellular stores and subsequent activation of store-operated Ca2+ entry (SOCE) [24]. SOCE is initiated by the Ca2+ sensor proteins, stromal interaction molecule 1 (STIM1) and STIM2, that are located within the endoplasmic reticulum (ER) [25]. Following store depletion, STIMs cluster and trap the plasma membrane (PM) proteins ORAI1–3 into ER-PM junctions [24]. These regions become sites of highly selective Ca2+ entry, predominantly through ORAI assembled channels [26, 27]. Prolonged and disordered Ca2+ oscillations in decidualizing stromal cells in response to serine proteases secreted by low-quality human embryos are thought to trigger ER- stress, which in turn facilitates early maternal rejection of a non-viable conceptus [28]. Recently, we reported that Orai1 is expressed in the secretory phase endometrium as well as in endometrial carcinoma cells [29]. Both Orai1 expression and function are upregulated by TGFß1, an effect presumably participating in the regulation of endometrial regeneration [29]. The role of Orai1 in implantation remains to be determined.

The mechanisms regulating [Ca2+]i levels in endometrial cells remain incompletely understood but presumably involve Ca2+ release from intracellular stores with subsequent activation of SOCE [27, 30,32,33,34,34]. Here, we show that LEFTY2 inhibits the expression of Ca2+ responsive receptivity genes by downregulating Orai1 expression and limiting SOCE. Our findings provide new insights into the molecular mechanisms that define a limited window of implantation in the human endometrium.

Results

Mining of publicly available microarray data (Gene Expression Omnibus accession number: 24460960) demonstrated a marked increase (30-fold) in LEFTY2 mRNA levels in whole endometrial biopsies upon transition from mid-secretory (MS; receptive) to late secretory (LS; refractory) phase of the menstrual cycle (Fig. 1a). In culture, decidual transformation of HESCs increased LEFTY2 expression. After 8 days of differentiation, LEFTY2 mRNA levels increased approximately 14,000-fold (Fig. 1b).

Fig. 1
figure 1

Expression of LEFTY2 in human endometrium. a Arithmetic means ± SEM of LEFTY2 transcripts expressed in arbitrary units (a.u.) in whole endometrial biopsies throughout the menstrual cycle. Expression levels were derived from in silico analysis of publicly available microarray data (GEO profile LEFTY2: ID 24460960). The phases of the cycle are indicated as follows: P proliferative, ES early-secretory, MS mid secretory, and LS late secretory. b Induction of LEFTY2 in decidualizing HESCs. Arithmetic means ± SEM (n = 3) of transcript levels of LEFTY2 measured in undifferentiated HESCs and cells decidualized with 8-br-cAMP and MPA for 2, 4 or 8 days. The data are expressed relative to that in undifferentiated cells (D0). Different letters above the error bars indicate that those groups are significantly different from each other at *P < 0.05 using the Kruskal-Wallis test

Full size image

In order to explore the functional significance of increased LEFTY2 expression, we examined the effect of LEFTY2 on embryo implantation in the mouse uterus. Embryo implantation in mice is a stepwise process that starts with attachment of the blastocysts to the endometrial luminal epithelial layer between 3.5 to 4.5 days postcoitus (dpc), triggering decidualization of the underlying stromal cells [35] and closure of the uterine lumen [36].To confirm that LEFTY2 renders the endometrium refractory to implantation [14], murine uterine flushing experiments were performed in C57BL/6 female mice mated with vasectomized males. The uterine lumen of pseudopregnant female mice was injected with recombinant LEFTY2 before transfer of 10 cultured blastocysts equivalent of 3.5 dpc into a single uterine horn. Phosphate-buffered saline (PBS; vehicle) was flushed through the uterus prior to embryo transfer in control animals. The number of implantation sites was determined 72 h later. Out of 50 embryos transferred in each group, 43 (86%) embryos failed to implant in uteri treated with recombinant LEFTY2. By contrast, only 5 (10%) embryos failed to implant following flushing of the uterine lumen with PBS (Fig. 2a). Viable embryos were observed in the implantation sites of control (Veh) animals (Fig. 2b). The few established implantation sites following LEFTY2 exposure appeared smaller and the presence of collapsed blastocysts suggested imminent demise (Fig. 2b). To exclude a possible effect of LEFTY2 on the blastocysts, we performed further flushing experiments. C57BL/6 female mice were mated with fertile males to induce a normal pregnancy. At day 3.0 dpc, the uterus was flushed once either with PBS or LEFTY2. In this natural mating model, the blastocysts are still within the oviduct at day 3.0 dpc and are therefore not exposed to recombinant LEFTY2. Our results show again a significant decrease of implantation sites in LEFTY2 flushed mice (Supplementary Fig. 1), further suggesting that LEFTY2 blocks implantation by acting directly on the endometrium.

Fig. 2
figure 2

Recombinant LEFTY2 blocks embryo implantation. Vas mated C57BL/6 female mice were subjected to laparotomy and both uterine horns gently flushed with either vehicle (PBS) or recombinant LEFTY2 (500 ng/ml). Ten embryos (equivalent to stage 3.5 dpc) were then transferred to the right horn. Each treatment group consisted of five mice. a The gross morphological appearance of the uteri and implantation sites at 9.5 dpc (scale bar, 1 cm); the right panel shows arithmetic means ± SEM (n = 5) of the number of implantation sites following uterine flushing with PBS (Veh) or recombinant LEFTY2. b Corresponding histological appearance of implantation sites. The upper panel (Veh; control) shows a normal blastocyst. The lower panel shows fetal demise following uterine LEFTY2 exposure prior to embryo transfer (scale bar, 100 μm). c Arithmetic means ± SEM (n = 5) of the relative expression of murine Orai1, Cox2 Wnt4 and Bmp2, transcript levels, normalized to Cyclo (housekeeping) mRNA and expressed in arbitrary units (a.u.), following uterine flushing with LEFTY2 or vehicle (PBS, dotted line). **P < 0.01 and ***P < 0.001 indicate statistically significant difference to absence of LEFTY2 (Student’s t test)

Full size image

We speculated that failed implantation following LEFTY2 exposure could reflect inhibition of key murine implantation genes. In agreement, murine levels of Orai1, Bmp2, Cox2 and Wnt4 were significantly downregulated in LEFTY2-treated uteri when compared to control mice (Fig. 2c). Other cardinal murine implantation genes, including Lif, Ihh, Hoxa10 and Hbegf were unaffected by LEFTY2 (Supplementary Fig. 2).

To determine the relevance of these observations to the human endometrium, we decidualized HESCs with 8-Br-cAMP and MPA (medroxyprogesterone acetate, a progestin) for 6 days in the presence or absence of LEFTY2. As was the case in mice, recombinant LEFTY2 also attenuated the expression of WNT4, BMP2, and COX2 at protein and mRNA level in decidualizing HESCs (Fig. 3 and Supplementary Fig. 3). By contrast, endometrial WNT4, BMP2 and COX2 expression was markedly upregulated in response to the Ca2+ ionophore, ionomycin (Fig. 3), indicating that the expression of these implantation factors is responsive to increase in [Ca2+]i. Interestingly, LEFTY2 abrogated the induction of COX2 and WNT4 in response to Ionomycin, although it failed to reverse the induction of BMP2. Ionomycin treatment had no impact on LIF, IHH, HOXA10, or HBEGF mRNA levels in decidualizing HESCs (Supplementary Fig. 4).

Fig. 3
figure 3

Differential regulation of Ca2+-dependent receptivity factors in response to LEFTY2 and Ionomycin. Primary HESC cultures were treated with 8-Br-cAMP and MPA for 6 days with or without LEFTY2 (25 ng/ml) or Ionomycin (1 μM) or in combination. a Original Western blots showing COX2, WNT4, BMP2 and GAPDH. b – d The abundance of COX2, WNT4 and BMP2 was quantified by ImageJ analysis of six Western blots and normalized to GAPDH. The data shows arithmetic means ± SEM. *P < 0.05, **P < 0.01 indicate significant difference to absence of LEFTY2 (Student’s t test)

Full size image

In order to determine if the inhibition of Ca2+-responsive receptivity genes following LEFTY2 treatment was a consequence of SOCE inactivation, primary HESCs were decidualized with 8-Br-cAMP and MPA for 6 days in the absence or presence of Orai1 inhibitors 2-aminoethoxydiphenyl borate (2-APB) [29], YM-58483 [37] or MRS-1845 [38]. Induction of IGFBP1 and PRL transcript levels confirmed that the cells mounted a decidual response (data not shown). As shown in Fig. 4, exposure of decidualizing HESC Orai1 inhibitors 2-APB, YM-58483 and MRS-1845 significantly decreased COX2, BMP2 and WNT4 transcript levels.

Fig. 4
figure 4

Orai1 inhibitors 2-APB, MRS-1845, and YM-58483 inhibit Ca2+-responsive implantation genes. Arithmetic means ± SEM (n = 6) of BMP2, WNT4, and COX2 transcript levels relative to that of L19 (housekeeping) in primary HESC cultures that were treated with 8-Br-cAMP and MPA for 6 days in the absence (dotted line) or in presence of the Orai-1 inhibitors 2-APB (50 μM), MRS-1845 (10 μM), or YM-58483 (10 μM); **P < 0.01 indicates significant difference from the control using Student’s t test

Full size image

We hypothesized that LEFTY2 may similarly impair SOCE in luminal endometrial epithelial cells at implantation. To test this hypothesis, we first determined the transcript levels of the ORAI1–3 and STIM1–2 isoforms in endometrial Ishikawa cells treated with or without LEFTY2. Ishikawa cells are a commonly used cell line to study human implantation events [28, 39, 40]. As illustrated in Fig. 5a, LEFTY2 selectively decreased the transcript levels of ORAI1, the dominant ORAI isoform in these cells. The transcript levels for ORAI2-3 were low in Ishikawa cells, irrespective of LEFTY2 treatment. Similarly, STIM1-2 expression did not change in response to LEFTY2 treatment. The decrease of ORAI1 transcript levels following LEFTY2 treatment was paralleled by a decrease of ORAI1 at the protein level. Flow cytometry showed a significant reduction in ORAI1 in response to LEFTY2 treatment (Fig. 5b–c).

Fig. 5
figure 5

LEFTY2 downregulates endometrial Orai1 expression. a Endometrial Ishikawa cells were treated with or without LEFTY2 (25 ng/ml) for 6 days, and ORAI1, ORAI2, ORAI3, STIM1 and STIM2 mRNA levels were measured. The results show arithmetic means ± SEM (n = 6). Expression levels of the ORAI2 and ORAI3 isoforms are relative to the ORAI1 transcript level in the absence of recombinant LEFTY2 treatment. b and c Original FACS plot and bar graph of ORAI1 expression in parallel cultures treated with and without LEFTY2. **P < 0.01, ***P < 0.001 indicates statistically significant difference to absence of LEFTY2. ### P < 0.001 indicates statistically significant difference from Control-Orai1 transcripts (Student’s t test)

Full size image

In order to explore whether LEFTY2 impacts on Ca2+ signaling, Ishikawa cells were treated with thapsigargin, a potent inhibitor of the sarco/endoplasmic reticulum Ca2+ ATPase (SERCA), in the presence or absence of recombinant LEFTY2. Treatment with thapsigargin in the absence of extracellular Ca2+elicited a transient increase in [Ca2+]i(Δ[Ca2+]i) (Fig. 6a). Both peak and slope of Δ[Ca2+]i were significantly blunted in response to LEFTY2 treatment. Subsequent addition of extracellular Ca2+sharply increased [Ca2+]i, reflecting SOCE. Peak and slope of Ca2+entry were significantly decreased by LEFTY2 (Fig. 6b–c). Basal Ca2+ levels were not significantly modified by LEFTY2 (Supplementary Fig. 5). Taken together, the data show that LEFTY2 blunts intracellular Ca2+release and SOCE.

Fig. 6
figure 6

LEFTY2 reduces SOCE. a Original tracing of Fura-2 fluorescence-ratio in fluorescence spectrometry during and after Ca2+ depletion (1 μM thapsigargin) in LEFTY2 treated (black circles) and vehicle treated (PBS; white circles) Ishikawa cells. b – c Arithmetic means ± SEM (n = 6, each experiment 15–30 cells) of the peaks (left panels) and slopes (right panels) of b Ca2+ release and c Ca2+ entry, respectively **P < 0.01, ***P < 0.001 indicates significant difference from absence of LEFTY2 (Student’s t test)

Full size image

Discussion

LEFTY2, a member of the TGF-β superfamily, is implicated in embryo development and stem cell differentiation through its antagonistic action on the TGF-β/Smad signaling pathway [41,43,43]. LEFTY2 is also expressed in the human endometrium, most prominently during the late secretory phase of the menstrual cycle [10, 44, 45]. Previous studies reported that overexpression of LEFTY2 is associated with unexplained infertility [46] and irregular menstrual bleeding [8], and in vivo gene transfer studies in mice have also shown inhibition of implantation [14]. Furthermore, auto/paracrine LEFTY signaling attenuates decidual transformation of HESCs [47, 48], although the underlying mechanisms are incompletely understood.

In this study, we demonstrate that LEFTY2 renders the endometrium refractory to implantation, an effect paralleled by downregulation of ORAI1 expression and of SOCE in endometrial epithelial cells. By attenuating [Ca2+]i signaling, LEFTY2 decreases expression of Ca2+-responsive implantation genes such as COX2, WNT4 and BMP2. Other key implantation genes, including LIF, IHH, HOXA10 and HBEGF transcripts [12, 49], were not affected by LEFTY2 or by Ionomycin treatment.

The pathways that couples LEFTY2 to transcriptional repression of ORAI1 and the mechanisms linking ORAI1 function and increases of [Ca2+]i to expression of receptivity genes requires further elucidation. The present observations uncover a novel function of Orai1 and store-operated Ca2+ entry. SOCE is known to trigger Ca2+ oscillations [27, 50], which stimulate several cellular functions [51,53,54,55,55] including entering into the S and the M phase of the cell cycle [56, 57] as well as cell survival [58, 59]. In contrast, sustained increase of cytosolic Ca2+ activity leads to apoptosis [53, 55, 60,62,63,64,65,66,67,68,68]. By mediating SOCE and oscillations of cytosolic Ca2+ activity, Orai1-3 and their regulators STIM 1–2 are decisive for cell survival [69,71,72,72]. It is noteworthy that ORAI1/STIM1 is upregulated by the serum and glucocorticoid inducible kinase (SGK1) and TGF-β in endometrial cancer cells [29], but downregulated by the AMP activated kinase (AMPK) [73]. Although as yet untested, downregulation of Orai1/STIM1 in response to AMPK activation potentially plays a role in preventing implantation of an embryo into an energy-deficient endometrium.

The present observations point to a new role of ORAI1-dependent SOCE in the regulation of endometrial receptivity genes, and hence the likelihood of successful embryo implantation. However, our findings do not rule out the involvement of additional mechanisms of Ca2+ release and/or entry during implantation. For example, evidence has emerged to implicate voltage-gated Ca2+ channels in Ca2+ entry in endometrial cells exposed to embryonic proteases [21]. Moreover, progesterone-dependent induction of phospholipase C (PLC)-related catalytically inactive protein 1 (PRIP-1) in decidual cells blocks Ca2+ release from the ER by inhibiting inositol 1,4,5-trisphosphate (IP3) signaling [74]. Most importantly, knockout of Orai1 on the specific Institute of Cancer Research (ICR) genetic background mice did not abrogate female fertility [75]. Unlike mice lacking Orai1 on a C57/DBA/129 background, the Orai1-deficient ICR mice did not suffer from high incidence of perinatal lethality [75] and may have partially replaced the function of Orai1 by other mechanisms. Further experiments are required to identify those mechanisms and explore their LEFTY2 sensitivity.

Our findings provide a novel mechanistic explanation for the clinical observation that elevated LEFTY2 levels are associated with implantation failure and infertility [13]. Taken together, our observations provide further evidence that coordinated temporal and spatial regulation of Ca2+ signaling in the endometrium across the window of receptivity is critical for reproductive success.

Materials and methods

Patient selection and sample collection

The study was approved by the NHS National Research Ethics–Hammersmith and Queen Charlotte’s & Chelsea Research Ethics Committee (1997/5065). The biopsies were timed between 6 and 10 days after the pre-ovulatory luteinizing hormone (LH) surge. None of the subjects were on hormonal treatments for at least 3 months prior to the procedure. Written informed consent was obtained from all participants in accordance with the guidelines in The Declaration of Helsinki 2000.

Cell culture

Human endometrial stromal cells (HESCs) were isolated from endometrial tissues as described previously [76]. Purified HESCs were expanded in maintenance medium of DMEM/F-12 (Invitrogen, Schwerte, Germany) containing 10% dextran-coated charcoal-treated fetal bovine serum (DCC-FBS; Invitrogen, UK) and 1% antibiotic-antimycotic solution (Invitrogen). Confluent monolayers were decidualized in DMEM/F-12 containing 2% DCC-FBS with 0.5 mM 8-bromo-cAMP (8-Br-cAMP; Sigma, Munich, Germany) with or without 10−6 M medroxyprogesterone acetate (MPA; Sigma) to induce a differentiated phenotype. Where indicated, the cells were treated with recombinant LEFTY2 (25 ng/ml; R&D Systems, Germany) as described previously [77]. Ionomycin was used at 1 μM (Sigma) and the Orai inhibitors: 2-APB, YM-58483, and MRS-1845(TOCRIS, Germany). Ishikawa cells, an endometrial epithelial-like cell line (ECACC 99040201) [28, 29], were maintained in DMEM/F12 (Invitrogen) supplemented with 10% fetal bovine serum (Invitrogen), 2 mM L-glutamine, and 100 U/ml penicillin/streptomycin (Invitrogen). All cells were incubated at 37 °C in a humid atmosphere maintained at 5% (vol/vol) CO2, and routinely tested for mycoplasma infection.

Animal experiments

C57BL/6 mice were purchased from Charles River Ltd. (Margate, UK). All experiments were carried out in accordance with the UK Home Office regulations (PPL70/6867). Mice had free access to food and water ad libitum, and were kept under constant humidity (55 ± 10%), temperature (22 ± 2 °C), and 12 h light-dark cycle conditions. To assess implantation, C57BL/6 female mice were mated by vasectomized males and the day of the appearance of the vaginal plug designated as day 1.0 dpc. Laparotomy was performed at 3.0 dpc. Both uterine horns were injected with 100 μL PBS or LEFTY2 dissolved in PBS (500 ng/ml). After 10 min, 10 cultured blastocysts (equivalent of 3.5 dpc) were transferred to a single treated uterine horn. The uteri were harvested 72 h following surgery, implantation sites counted, and tissues fixed in formalin or snap frozen for further analysis.

Natural mating model

We conducted timed matings by placing wild-type C57BL/6 female mice with fertile wild-type males to induce pregnancy. The day when a vaginal plug was apparent was designated as 1.0 dpc, and mice were anesthetized at 3.0 dpc and subjected to laparotomy to expose the uterus [12]. We used two groups of mice: a control group flushed with PBS (Veh) and a study group flushed with LEFTY2 (as described above). The incision was then closed to allow the mice to recover. The uteri were then harvested at 9.5 dpc and the implantation sites counted.

Western blot analysis

Whole cell protein extracts were prepared by lysing cells in RIPA buffer. Protein yield was quantified using the Bio-Rad DC protein assay kit (Bio-Rad). Equal amounts of protein were separated by 10% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) before wet-transfer onto a PVDF membrane (Amersham Biosciences, UK). Nonspecific binding sites were blocked by overnight incubation with 5% nonfat dry milk in Tris-buffered saline with 1% Tween (TBS-T; 130 mmol/L NaCl, 20 mmol/L Tris, pH 7.6, and 1% Tween) as previously described [76]. The following primary antibodies were used: anti-BMP2 (♯sc6895, Santa Cruz Biotechnology Inc., Texas, USA), anti-WNT4 (♯ sc376279, Santa Cruz Biotechnology Inc), anti-Cox2 (♯15191; Abcam, Cambridge, UK), and anti-GAPDH (♯21185; Cell Signaling, Leiden, The Netherlands). All primary antibodies were diluted 1:1000. Protein complexes were visualized with a chemiluminescent detection kit (WesternBright™ ECL, Advansta, CA, USA).

Real-time quantitative (qRT)-PCR

Total RNA was extracted from cell cultures or from snap frozen whole uteri using Trizol (Invitrogen) based on a phenol-chloroform extraction protocol [78]. Equal amounts of total RNA (2 μg) were reverse transcribed using the Superscript III First-Strand synthesis system for RT-PCR (Invitrogen) with oligo dT priming. The resulting first-strand cDNA was diluted and used as a template in qRT-PCR analysis. L19 and Cyclophilin A (Cyclo), representing non-regulated human and murine housekeeping genes, respectively, were used to normalize for variances in input cDNA. Detection of gene expression was performed with KappaFast-SYBR Green (Peqlab, Erlangen, Germany), and qRT-PCR was performed on a BioRad iCycler iQ™ Real-Time PCR Detection System (Bio-Rad Laboratories, Munich, Germany). The non-template control (NTC) reactions (cDNA was substituted with DNase/RNase free water) and reverse transcriptase (RT) controls were included in each PCR reaction. The PCR products were not detected in NTC or RT control reactions (data not shown). Transcript levels were determined by the ΔΔCt method [79] and expressed in arbitrary units. All measurements were performed in triplicate. Melting curve analysis and agarose gel electrophoresis confirmed amplification specificity. Primer sequences are provided on request.

Flow cytometry

Orai1 expression was analyzed by flow cytometry. Cultured cells were detached, washed three times with phosphate-buffered saline (PBS), and fixed with 4% paraformaldehyde for 15 min on ice. The cells were then incubated for 60 min (37 °C) with anti-Orai1 monoclonal primary antibody (1:200, Abcam) washed once in PBS, and stained in 1:250 diluted CF™ 488A-labeled anti-rabbit secondary antibody (Sigma) for 30 min (37 °C). Samples were immediately analyzed on a FACSCalibur flow cytometer (BD Biosciences, Heidelberg, Germany). Data were analyzed using the FlowJo software (FlowJo LLC, Ashland, Oregon, USA).

Calcium measurements

Fura-2 fluorescence was utilized to determine [Ca2+]i [80]. Cells were loaded with Fura-2/AM (2 μM; Invitrogen) for 20 min at 37 °C. Cells were excited alternatively at 340 and 380 nm through an objective (Fluor ×40/1.30 oil) built in an inverted phase-contrast microscope (Axiovert 100, Zeiss, Oberkochen, Germany). Emitted fluorescence intensity was recorded at 505 nm. Data were acquired using specialized computer software (Metafluor, Universal Imaging, Downingtown, USA). Cytosolic Ca2+ activity was estimated from the 340 nm/380 nm ratio. SOCE was determined by extracellular Ca2+ removal in the presence of sarcoendoplasmatic Ca2+ ATPase inhibitor thapsigargin (1 μM, Invitrogen) and subsequent Ca2+ re-addition [81]. For quantification of Ca2+ entry, the slope (delta ratio) and peak (delta ratio) were calculated following re-addition of Ca2+ [80, 82]. Experiments were performed with Ringer solution containing (in mM) 125 NaCl, 5 KCl, 1.2 MgSO4, 2 CaCl2, 2 Na2HPO4, 32 HEPES, 5 glucose, and pH 7.4. To reach nominally Ca2+-free conditions, experiments were performed using Ca2+-free Ringer solution containing (in mM) 125 NaCl, 5 KCl, 1.2 MgSO4, 2 Na2HPO4, 32 HEPES, 0.5 EGTA, 5 glucose, and pH 7.4. For calibration purposes, ionomycin (10 μM, Sigma) was applied at the end of each experiment.

Histology

For histological assessment, uterine horns were formalin-fixed and embedded in paraffin, cut in 5-μm sections, and stained with H&E as previously described [12].

Statistical analysis

Data were analyzed with the statistical package GraphPad Prism (GraphPad software Inc., CA, USA). Student’s t test or Kruskal-Wallis test was used when appropriate. Statistical significance was assumed when P < 0.05.