Adiponectin limits differentiation and trophoblast invasion in human endometrial cells

Successful human embryo implantation requires a proper differentiation of endometrial stromal cells (ESCs) into decidual cells, during a process called decidualization. ESCs express specific molecules, such as prolactin, insulin-like growth factor-binding protein-1 (IGFBP-1) and connexin-43. Decidual cells are also involved in the control of trophoblast invasion, by secreting various factors, such as matrix metalloproteinases (MMPs) and tissue inhibitors of metalloproteinases (TIMPs). Adiponectin is an adipokine with insulin-sensitizing, anti-inflammatory and anti-proliferative effects. At the embryo– maternal interface, adiponectin promotes differentiation and invasion of human trophoblastic cells. We hypothesize that the effects of adiponectin on endometrium could counteract its pro-invasive effects previously described in the human trophoblast. In this context, we have firstly demonstrated that adiponectin downregulates IGFBP-1 and connexin-43 mRNA expressions, as well as prolactin secretion in ESCs, suggesting an anti-differentiative effect of adiponectin. Secondly, we found that invasive capacities of trophoblastic cell line HTR-8/SVneo are reduced in the presence of conditioned media from ESC cultured in the presence of adiponectin. Adiponectin’s anti-invasive action is associated with a decreased activity of MMP-2 and MMP-9, and an increased TIMP-3 mRNA expression in ESCs. Finally, adiponectin receptors (ADIPOR1 and ADIPOR2) knockdown abolishes the anti-differentiative and anti-invasive effects of adiponectin in human ESCs. Altogether, our results suggest that adiponectin reduces the decidualization process and inversely induces the production of endometrial factors that limit trophoblast invasion. Thus, through a dual control in trophoblast and endometrial cells, adiponectin appears as a pivotal actor of the embryo implantation process.


Introduction
Human embryo implantation is a multi-step process that begins with apposition of trophoblastic cells from a competent blastocyst to the maternal endometrium.This apposition occurs during the 'window of implantation', a short period of uterine receptivity corresponding to the mid-secretory phase of the menstrual cycle (Diedrich et al. 2007).After the blastocyst has attached to the endometrial epithelium, its extravillous trophoblasts (EVTs) penetrate into the endometrial stroma and acquire an invasive phenotype resulting in the placenta anchoring in the endometrium (Lunghi et al. 2007).
During trophoblast invasion, EVTs secrete large amounts of matrix metalloproteinases (MMPs) (Burrows et al. 1996).The actions of MMP-2 and MMP-9 are crucial for trophoblast invasion (Staun-Ram et al. 2004).Indeed, these gelatinases are able to degrade the major component of the endometrial extracellular matrix (ECM) i.e. collagen IV (Burrows et al. 1996).MMP activity is regulated by specific tissue inhibitors of metalloproteinases (TIMPs), which bind and inactivate MMPs with a 1:1 stoichiometry (Huppertz et al. 1998).Thus, correct regulation of the MMP/TIMP balance in EVTs is essential for an appropriate embryo implantation (Nissi et al. 2013).Trophoblast invasion also requires adequate endometrial receptivity (Singh et al. 2011).This is characterized by morphological, biochemical and vascular modifications, including (i) the appearance of epithelial pinopodes, (ii) the modulated expression of cytokines, growth factors and adhesion molecules (particularly the upregulation of epithelial αv-and β3-integrins), (iii) the disappearance of an adhesive molecular surface (loss of mucin-1 and mucin-12) and (iv) the decidualization of endometrial stromal cells (ESCs).Decidualization occurs in response to the ovarian hormones, 17β-estradiol (E2) and progesterone (P4).It is accompanied by the secretion of prolactin and insulinlike growth factor-binding protein-1 (IGFBP-1), and the expression of the gap junction connexin-43 (CX-43) in human ESCs (Brar et al. 1995, Yu et al. 2011).Decidual ECM remodeling events have also been observed.These processes are associated with MMP-2,-9 secretion and the modulation of TIMP-1,-2,-3 expression in human ESCs (Zhang & Salamonsen 1997, Estella et al. 2012, Tapia-Pizarro et al. 2013).This type of ECM remodeling enhances ESC motility and thus facilitates trophoblastic infiltration into the endometrial stroma (Grewal et al. 2008).Decidual ECM remodeling is a complex process, tightly controlled by factors intrinsically produced by both placental and decidual cells (Sharma et al. 2016).For example, human chorionic gonadotropin (hCG) and tumor necrosis factor alpha (TNF-α) facilitate EVT migration by enhancing endometrial MMP-2,-9 activities and by inhibiting endometrial TIMP-1 secretion (Haider & Knöfler 2009, Tapia-Pizarro et al. 2013).Conversely, the antifungal antibiotic trichostatin A limits trophoblast invasion by increasing the production of TIMP-1,-3 by ESCs and reducing the activity of endometrial MMP-2,-9 (Estella et al. 2012).Hence, the local MMP/TIMP balance, at the invasion site, requires an appropriate interaction between trophoblasts and ESCs.
Adiponectin is a 30 kDa cytokine, predominantly produced by adipose tissue (Kern et al. 2003).Adiponectin exerts its action via two specific receptors (ADIPOR1 and ADIPOR2) containing seven transmembrane domains but which are structurally and functionally distinct from G-protein-coupled receptors (Yamauchi et al. 2003a).By binding to ADIPORs, adiponectin activates a variety of targets, such as protein kinase A, phosphoinositide-3 kinase, mitogen-activated protein kinase and most notably, the AMP-activated protein kinase through which adiponectin exerts its insulin-sensitizing effects (Yamauchi et al. 2003a, Campos et al. 2008).This adipokine has also been shown to exert anti-inflammatory, anti-angiogenic, antiatherosclerotic, anti-proliferative and pro-apoptotic roles in various cell types (Yamauchi et al. 2003b, Dieudonne et al. 2006, Duval et al. 2016).
During pregnancy, adiponectin seems to act as an endo/paracrine signal at the embryo-maternal interface.
It is now established that adiponectin is not produced by human trophoblastic cells but seems to be expressed in human endometrium (Takemura et al. 2006, Benaitreau et al. 2010a, Dos Santos et al. 2012).ADIPOR1 and ADIPOR2 are expressed in human placenta and endometrium (Takemura et al. 2006, Benaitreau et al. 2009).A growing body of evidence has highlighted adiponectin role in the establishment of a functional placenta (Benaitreau et al. 2010b).Our laboratory has shown that adiponectin promotes human trophoblast invasion by enhancing MMP-2,-9 activities and downregulating TIMP-2 mRNA expression (Benaitreau et al. 2010a).
However, little is known about adiponectin effects in human endometrium.A study showed that expression of ADIPOR1 and ADIPOR2 peaks in the mid-secretory phase of menstrual cycle (Takemura et al. 2006).We have also observed significantly low endometrial expression of ADIPOR1 and ADIPOR2 in women with unexplained recurrent implantation failure relative to fertile women (Dos Santos et al. 2012).In vitro, adiponectin exerts an anti-proliferative action by promoting cell death (Bohlouli et al. 2013).Taken as a whole, these data strongly suggest that adiponectin plays a crucial role during embryo implantation.With a view to better understand embryomaternal dialog in the early stages of pregnancy, we studied the in vitro effects of adiponectin on (i) the human ESC decidualization program and (ii) the endometrial control of trophoblast invasion.

Study population and tissue collection
A total of 23 normally cycling women aged 27-40 undergoing biopsy for fertility evaluation, were recruited into the present study.Endometrial tissues were obtained during the window of implantation.Patients with low endometrial volume (<2 cm) or presence of submucous fibroma or polyps were excluded.A good hormonal reserve and normal responses to ovarian stimulation were also included as additional criteria.This study was approved by the local investigational review board (Comité Consultatif de Protection des Personnes dans la Recherche Médicale, approval reference protocol 01-78).All participants provided their informed consent before tissue sampling.

Human ESC culture
Human ESCs were isolated and cultured as described by González and coworkers (Gonzalez et al. 1999).Briefly, tissue samples were minced into small pieces and digested in a two-step process.Tissues were incubated for 1 h at 37°C in a phenol-red-free DMEM/F12 medium containing collagenase (0.1%), DNase type I (0.0005%), penicillin (10 µg/mL) and streptomycin (100 U/mL).
The supernatant was filtered through a 100 µm nylon screen and then centrifuged at 200 g for 10 min.A second enzymatic digestion was performed on undigested tissue for 10 min at 37°C in DMEM/F12 medium-containing trypsin (0.25%), DNase type I (0.1%), EDTA (0.03%), penicillin (10 µg/mL) and streptomycin (100 U/mL).The digested tissue was filtered through a 40 µm nylon screen and then centrifuged at 200 g for 10 min.Cell pellets from both digestions were pooled and centrifuged at 200 g for 10 min.

Invasion assay for HTR-8/SVneo cells co-cultured with ESC culture medium
The invasiveness of trophoblastic cells in response to ESCsecreted molecules was assessed using the HTR-8/SVneo immortalized EVT cell line, kindly provided by Dr Nadia Alfaidy (CEA Grenoble, France) in agreement with Dr Charles Graham.HTR-8/SVneo cells were cultured in RPMI medium supplemented with HEPES 1M (2%), penicillin (100 U/mL), streptomycin (100 µg/mL) and FCS (10%) until they reached confluence.Invasion assays were performed in 24-well plates containing Matrigel-coated polycarbonate membrane (pore size: 8 µm) invasion chamber inserts (Greiner Bio-One SAS, Courtaboeuf, France), according to the modified protocol from Tapia-Pizarro and coworkers (Tapia-Pizarro et al. 2013).HTR-8/SVneo cells were suspended (5 × 10 4 cells per well) in 250 µL of conditioned medium (CM) from 15-day-decidualized ESCs treated or not with adiponectin (25 ng/mL and 250 ng/mL).Media containing only E2 and P4 in the absence or presence of adiponectin (25 ng/mL and 250 ng/mL), were used as controls.RPMI medium supplemented with FCS (10%) was added to the lower well as a chemoattractant.After 48 h of incubation at 37°C, medium containing noninvading cells was removed from the upper well.Invasive HTR-8/SVneo cells at the lower surface of the insert were washed and fixed with paraformaldehyde (4%) for 30 min.The nuclei were counterstained with 1 µg/mL Hoechst reagent and visualized with an inverted laser scanning confocal microscope (Leica white light laser TCS SP8-X, Leica Microsystems, Wetzlar, Germany).On each insert, invasive cells were counted on five randomly selected fields by the post-imaging procedure in ImageJ software.Invasive cells were defined as those whose nucleus exceeded 8 µm (i.e. the equivalent of a pore size).

Zymography
After 1, 8 or 15 days of culture (D1, D8 and D15) with differentiation medium in the absence or presence of adiponectin (25 ng/mL and 250 ng/mL), total gelatinase activities from ESCs were analyzed by zymography.Aliquots of CM containing 60 µg of protein were resolved under non-reducing conditions in 10% polyacrylamide gels containing 1 mg/mL gelatin (Difco).Gels were washed in Triton X-100 (2.5%) for 30 min to remove SDS and incubated overnight at 37°C in a renaturing buffer (50 mM Tris-HCl, pH 7.5, 5 mM CaCl 2 , 150 mM NaCl, and 0.02% sodium azide).Gels were stained with Coomassie Brilliant Blue, and destained in methanol/acetic acid (20%/5% v/v).Proteolytic activity was identified as a clear band on a blue background.The images were scanned and quantitative enzyme analysis was carried out, using ImageJ software.

RNA interference for ADIPOR1 and ADIPOR2
Two pairs of small-interfering RNAs (siRNAs) corresponding to different regions of each receptor gene were chemically synthesized by Qiagen.The sequences of the sense siRNAs are given in Table 2.A fluorescently labeled, non-silencing control siRNA (siNS) was used to optimize the transfection conditions and as a control for nonspecific silencing effects.For the knockdown experiments, ESCs were plated in 12-well dishes at 3.5 × 10 5 cells per well.After 13 days, cells were transfected with siNS (10 nM) or siADIPOR1 (25 nM) or siADIPOR2 (25 nM) using a Lipofectamine RNAiMAX transfection reagent (Invitrogen) according to the manufacturer's instructions.Adiponectin (250 ng/mL) was added 24 h after transfection.mRNA expression was analyzed by RT-qPCR and protein expression was assessed using Western blot, as described previously.

Prolactin secretion
Prolactin secretion into the culture medium was measured using an automated immuno-chemiluminescence system (KRYPTOR, BˑRˑAˑHˑMˑS, Saint-Ouen, France).In order to compare the levels of prolactin secretion, results were normalized to 1 µg of total protein.The protein concentration was measured according to Bradford's method, with BSA as the standard.

Statistical analyses
Statistical analyses were performed on the raw data from 5 to 10 separate experiments.Comparisons between controls (absence of the adiponectin) at D1, D3, D8 and D15, were made using analysis of variance (ANOVA) with Bonferroni P values.A non-parametric, paired Wilcoxon test was used to compare the effect of adiponectin concentration  (25 ng/mL or 250 ng/mL) with the control situation (the absence of adiponectin) for a given day of exposure.

Characterization of human ESCs
To assess the purity of our cell preparation, immunocytochemistry experiments were performed, 48 h after cell isolation.We have studied the expression of vimentin and pan-cytokeratin, which are specific markers of endometrial stroma and epithelium, respectively.Our microscopy analyses revealed that 95% of isolated endometrial cells stained positive for vimentin and only 5% expressed pan-cytokeratin (Fig. 1).This clearly demonstrated that the great majority of isolated cells were ESCs.These latter results were confirmed by measuring prolactin secretion into cell culture supernatant, which is a specific marker of ESC decidualization.We showed that this hormonal production increased progressively during cell differentiation (16 ± 3, 66 ± 16, 571 ± 157, and 2374 ± 607 IU/µg of protein after 1, 3, 8, and 15 days of cell culture, respectively).

Expression of adiponectin and its specific receptors in human ESCs
Human ESCs did not express adiponectin at D3, D8 and D15 of cell decidualization.However, this adipokine was expressed in human adipose tissue, used as a positive control (Fig. 2A).The RT-qPCR results demonstrated that ESCs expressed ADIPOR1 and ADIPOR2, whatever the decidualization status.Furthermore, the respective expression levels were 2.1-and 2.3-fold higher at D15 of cell culture, over D3 (Fig. 2B).Western blot analyses confirmed the presence of the two ADIPOR subtypes in human ESCs, at D3, D8 and D15 of cell decidualization.Human firsttrimester placenta was used as a positive control (Fig. 2C).However, ESC treatment with adiponectin (25 ng/mL and 250 ng/mL) did not seem to affect ADIPOR expression in these cells (data not shown).

Effect of adiponectin on human ESC decidualization
Prolactin, IGFBP-1 and CX-43 mRNA expressions and prolactin secretion were used as biochemical markers in order to study the differentiation of ESC into decidual cells.As shown in Fig. 3A, B and C, we observed all along the ESC decidualization process, a significant increasing prolactin, IGFBP-1 and CX-43 mRNA expressions, with a maximal expression at D15 (8.8-fold, 715.2-fold and 3.3-fold increase over D3, respectively).Likewise, we observed a rise in prolactin secretion between D1 and D15 of cell decidualization (118.4-fold) (Fig. 3D).These results validated our in vitro differentiation protocol and showed that ESCs are fully decidualized at D15.

Involvement of adiponectin in the endometrial control of trophoblast invasion
In order to determine how adiponectin might be involved in the endometrial control of trophoblast invasion, we performed Matrigel Transwell invasion assays.We cultured the HTR-8/SVneo immortalized EVT cell line in the presence of CM collected from differentiating ESCs treated or not with adiponectin (25 ng/mL and 250 ng/mL) for 15 days (Fig. 4A).

Effect of adiponectin on endometrial TIMP mRNA expression
We next sought to specify the molecular mechanisms involved in adiponectin anti-invasive action.We focused on the mRNA expression of three invasion inhibitors, such as TIMP-1, TIMP-2, and TIMP-3, strongly expressed in the human endometrium.Using RT-qPCR, we found that TIMP-1 and TIMP-2 mRNA expression did not significantly change throughout ESC decidualization regardless of the presence or absence adiponectin (data not shown).However, as shown in Fig. 5A, TIMP-3 mRNA expression increased significantly throughout the ESC decidualization process with a peak effect at D15 of cell culture (1.8-fold increase over D3).Furthermore, this increase was amplified by the presence of 25 ng/mL adiponectin (+117%, +81%, +69%, on D3, D8 and D15, respectively) and 250 ng/mL adiponectin (+122%, +100% and +125% at D3, D8 and D15, respectively).

Effect of adiponectin on endometrial MMP activity
Zymography assays revealed two bands corresponding to the gelatinase activities of MMP-2 (72 kDa) and MMP-9 (92 kDa) in CM collected on D1, D8 and D15 of cell culture.Both gelatinases showed a higher activity at D8 and D15.An additional band of gelatinase activity (45 kDa) appeared in CM collected at D1 but had disappeared by D8 and D15 of cell culture (Fig. 5B). Figure 5C and D showed that adiponectin significantly decreased MMP-2 and MMP-9 activities at D8 (-19% for both MMPs) but only at the higher dose tested.At D15, however, we found that adiponectin downregulated MMP-2 activity (-11% at 25 ng/mL and -15% at 250 ng/mL) and MMP-9 activity (-28% at 25 ng/mL and -31% at 250 ng/mL).

Effects of ADIPOR1 and ADIPOR2 knockdown in human ESCs
Considering the late effect of adiponectin on decidualization (significant at D15 of differentiation), we decided to knockdown ADIPOR1 and ADIPOR2 mRNA expressions with siRNA for the last 2 days of decidualization (from D13 to D15).We observed a significant decrease in ADIPOR1 and ADIPOR2 mRNA expression levels (-56 and -58%, respectively) after 48 h of knockdown.This ADIPOR invalidation was confirmed using Western Blot analysis (Fig. 6).The partial suppression of ADIPOR1 and ADIPOR2 abolished the decrease of CX-43 and IGFBP-1 mRNA expressions observed in the presence of 250 ng/mL adiponectin (Fig. 7).In parallel, silencing the ADIPOR1 or ADIPOR2 genes suppressed adiponectin (250 ng/mL) effect on the endometrial control of trophoblast invasion (Fig. 8).These results were also observed when ADIPOR1 and ADIPOR2 were concomitantly invalidated (data not shown).

Discussion
Effective embryo-maternal communication is the key for a successful pregnancy.Adiponectin is a cytokine to induce the formation of a functional placenta (Benaitreau et al. 2010a,b).In this study, we sought to examine the role of adiponectin in maternal compartment by studying its effects on decidualization process and endometrial control of trophoblast invasion.Firstly, we attempted to determine whether human ESCs transcribe the ADIPOR1 and ADIPOR2 genes.Our results reveal that isolated human ESCs do indeed express these two receptors and strongly suggest that the endometrium is a target tissue for this adipokine.We also observed that ADIPOR1 and ADIPOR2 mRNA levels increased moderately throughout the in vitro endometrial decidualization (with a 2-fold increase between D0 and D15 of decidualization).Our results differ quantitatively from those observed by Gamundi-Segura and coworkers who observed a 10-fold increase in ADIPOR mRNA expression in ESCs during decidualization (Gamundi-Segura et al. 2015).As the endometrium undergoes major remodeling over the ovarian cycle, this discrepancy might be due to the time point at which the biopsy was performed (window of implantation in our study and oocyte pickup in the study by Gamundi-Segura and coworkers).Furthermore, and in accordance with a previous study in human ESCs, we did not observe any regulation of ADIPOR1 or ADIPOR2 by adiponectin in these cells (Bohlouli et al. 2013).We also found that adiponectin did not seem to be produced by isolated human ESCs under our experimental conditions.This result contrasts with previous publications, including ours (Takemura et al. 2006, Dos Santos et al. 2012).However, these earlier studies were performed with endometrial tissue that contains various cell types (stromal cells, epithelial cells, blood vessels, etc.).Detection of adiponectin in endometrial tissue might well have been due to the presence of peripheral blood.Experiments to establish whether isolated endometrial epithelial cells express adiponectin are currently underway in our laboratory.
Secondly, we assessed ESC decidualization by measuring the expression of decidualization markers, as prolactin, IGFBP-1 and connexin-43.We observed a time-dependent increased expression of these three markers (between D3 and D15), in the presence of the differentiation medium, indicating that ESCs are fully differentiated at D15.This result is commonly observed for in vitro ESC decidualization in the presence of E2 and P4 (Kasahara et al. 2001, Godbole et al. 2011).We also studied adiponectin effect on ESC decidualization.To this end, we used subphysiological doses of recombinant human adiponectin (25 ng/mL and 250 ng/mL), which are much lower than those observed in the human circulation.This recombinant human adiponectin effectively mainly  comprises the high-molecular-weight form (Bub et al. 2006), considered to be the active isoform (Kobayashi et al. 2004).Moreover, adiponectin is commonly used at these concentrations for in vitro experiments with various tissue and cell types (Bråkenhielm et al. 2004, Dos Santos et al. 2008, Duval et al. 2016).Our results clearly show that adiponectin negatively controls biochemical ESC decidualization in a dose-independent manner.We effectively observed a significant decrease of prolactin secretion in ESCs treated with adiponectin.There was no effect on prolactin mRNA expression.This lack of correspondence between RNA and protein levels is not surprising and has been already described (Salari et al. 2012, Beileke et al. 2015).One study specifically evidenced the mismatch between the prolactin mRNA expression and hormone release (Castaño et al. 1997).
Thirdly, we demonstrated that adiponectin downregulates IGFBP-1 and connexin-43 mRNA expression levels.These effects were mainly observed when ESCs were fully differentiated (after 15 days of decidualization).As levels of ADIPOR mRNA doubled between D0 and D15 of decidualization, this observation might explain the delayed, negative effect of adiponectin on decidualization.Taken as a whole, these data suggest that like hCG, leptin and interleukin-1, adiponectin could be a new regulator of human ESC decidualization (Kariya et al. 1991, Kasahara et al. 2001, Tanaka et al. 2003).
As one of the major roles of decidual cells is the control of trophoblast invasion (Godbole et al. 2011, Sharma et al. 2016), we next looked at whether adiponectin could modify the ability of decidualized ESCs to regulate this process.Therefore, we used Transwell assays to estimate the trophoblast HTR-8/SVneo cell line invasive capacities when seeded with CM from ESCs treated or not with adiponectin.Our results clearly demonstrate that CM from ESC cultured in the presence of adiponectin significantly decreased trophoblastic cell ability to invade Matrigel.In order to gain more information at molecular level, we next investigated adiponectin effect on the MMP/ TIMP balance in human ESCs.Our data demonstrate that adiponectin significantly and specifically upregulated TIMP-3 mRNA expression and, conversely, reduced MMP-2,-9 activities in ESCs.Thus, it seems that in human endometrium, adiponectin can modulate, at least in part, ECM remodeling via regulation of the MMP-2,-9/TIMP-3 balance.Adiponectin effects on human ESCs were once again dose independent, but is not surprising, since the majority of in vitro studies performed in various cell types such as placental cells, did not show a dose-response effect for adiponectin (Benaitreau et al. 2009, Duval et al. 2016).Taken as a whole, these data suggest that adiponectin induces the production of endometrial factors, which could limit trophoblast invasion.At present, it is difficult to precise whether the secretion of endometrial factors is directly controlled by adiponectin, or indirectly via the induced production of hormones or cytokines.A recent study effectively demonstrated that TNF-α facilitates EVT invasion in vitro by regulating the MMP/TIMP balance in human ESCs (Haider & Knöfler 2009).Thus, endometrial TNF-α signal could be modified by adiponectin, as described in macrophages and in adipocytes (Tsatsanis et al. 2005, Greenberg & Obin 2006).Additional experiments will be needed to clarify this hypothesis.
In order to confirm the specific involvement of ADIPOR1 and ADIPOR2 in adiponectin regulation of ESC decidualization and the control of trophoblast invasion, we silenced ADIPORs in ESCs.We found that ADIPOR1 and ADIPOR2 knockdown (separately or concomitantly) suppressed adiponectin anti-differentiative and antiinvasive effects in human ESCs.These results clearly demonstrated the critical involvement of the ADIPORs in adiponectin control of ESC functions.
Lastly, we also attempted to determine whether or not the human ESC decidualization status is associated with the cell ability to control ECM remodeling.We clearly demonstrated that TIMP expression and MMP activities are closely related to ESC decidualization status.Indeed, we confirmed that TIMP-3 mRNA expression significantly rose throughout decidualization, as described by Zhang and Salamonsen (Zhang & Salamonsen 1997).We also found that non-decidualized ESCs secrete an active MMP, which might be MMP-3.Indeed, MMP-3 is a 45 kDa protein able to digest gelatin.However, the putative MMP-3 activity falls during decidualization.With regard to MMP-9, and more especially MMP-2, which appears to be the most abundantly produced gelatinase in ESCs, we observed an upregulation of their activities during decidualization.These results confirmed that ECM remodeling occurs during ESC differentiation in order to maintain endometrial integrity, which is essential for embryo implantation.
Interestingly, we noted that although adiponectin enhances the MMP/TIMP ratio in human trophoblastic cells, and thus, promotes invasion (Benaitreau et al. 2010a); this adipokine decreases the MMP/TIMP ratio in human ESCs during cell decidualization.This would limit cell migration and consequently avoid excessive invasion.Taken as a whole, these data strongly suggest that the adiponectin system has a critical role in the establishment of embryo implantation by maintaining the balance of pro-invasive and anti-invasive signals in placenta and endometrium, respectively.In conclusion, it should be noted that dysregulation of the placental invasive potential is associated with several pregnancy pathologies, as spontaneous miscarriages, preeclampsia, intrauterine growth restriction (reduced trophoblast invasion), placenta accreta and choriocarcinoma (excessive trophoblast invasion) (Lala et al. 2002, Anin et al. 2004, Ball et al. 2006, Kadyrov et al. 2006).Hence, our results suggest that adiponectin, through dual control in trophoblast and decidual cells, could have a causative role in the appearance of these pathologies by modulating the invasive capacities.

Figure 2
Figure 2 Adiponectin receptors are expressed in human ESCs.Human ESCs were cultured in DMEM/F12 medium supplemented with E2 and P4.(A) Total RNA was extracted after 3 days (D3), 8 days (D8) and 15 days (D15) of cell differentiation.Adiponectin, ADIPOR1 and ADIPOR2 mRNA expression levels were analyzed by RT-PCR with the primers listed in Table 1.PCR products were analyzed by agarose gel electrophoresis.The figure represents one representative of nine separate experiments.(B) ADIPOR1 and ADIPOR2 mRNA levels were quantified by RT-qPCR as described in Materials and methods section.The data are quoted as the mean ± s.e.m. of nine separate experiments.(C) Western blot analysis of cell lysates (20 µg) using anti-ADIPOR1 and anti-ADIPOR2 antibodies, as described in the Materials and methods section.The figure shows one representative of five separate experiments.*P<0.05;**P<0.01;NS: non-significant.ANOVA test.

Figure 4
Figure 4Adiponectin improves the endometrial control of trophoblastic migration.(A) Transwell migration assays of HTR-8/SVneo cells were performed as described in the Materials and methods section.(B) HTR-8/ SVneo cells were suspended in the presence of conditioned medium (CM) from 15-day decidualized ESCs treated or not with adiponectin (25 ng/mL and 250 ng/mL) or in the presence of control medium, supplemented or not with adiponectin (25 ng/mL and 250 ng/mL).The data are quoted as the mean ± s.e.m. of seven separate experiments.*P<0.05;**P<0.01;NS: non-significant.Wilcoxon test.

Figure 5
Figure 5 Adiponectin modifies endometrial TIMP expression and MMP activities.Human ESCs were cultured in DMEM/F12 medium supplemented with E2 and P4, and exposed or not to adiponectin (25 ng/mL and 250 ng/mL) for 15 days.(A) Total RNA was extracted after 3 days (D3), 8 days (D8) and 15 days (D15) of cell differentiation.The data are quoted as the mean ± s.e.m. of ten separate experiments.(B) Activities of gelatinases in conditioned media (CM) from decidualized ESCs after 1 day (D1), 8 days (D8) and 15 days (D15) of cell differentiation were measured as described in the Materials and methods section.This figure shows one representative of seven separate experiments.(C) Quantification of the gelatin zymography results for the MMP-2 band.The data are quoted as the mean ± s.e.m. of seven separate experiments.(D) Quantification of the gelatin zymography results for the MMP-9 band.The data are quoted as the mean ± s.e.m. of seven separate experiments.*P<0.05;**P<0.01;NS: non-significant.(A) vs control situation (without adiponectin).Wilcoxon test.(B) vs D3; (C) vs D8.ANOVA test.

Table 1
Primers used for RT-PCR.
SVneo cells were suspended in the presence of conditioned medium (CM) from ESCs transfected with siADIPOR1, siADIPOR2 or siNS, supplemented or not with adiponectin (250 ng/mL) to perform Transwell invasion assays.The data are quoted as the mean ± s.e.m. of six separate experiments.*P<0.05;NS: non-significant.Wilcoxon test.