PreImplantation Factor (PIF) promoting role in embryo implantation: increases endometrial Integrin-α2β3, amphiregulin and epiregulin while reducing betacellulin expression via MAPK in decidua

Synthesis was previously described [25]. In short, synthetic PIF analog (MVRIKPGSANKPSDD) or MVRIKPGSA and scrambled PIF (GRVDPSNKSMPKDIA) PIFscr – used as negative control, were produced by using solid-phase peptide synthesis (Peptide Synthesizer; Applied Biosystems, Foster City, CA) employing Fmoc (9-fluorenylmethoxycarbonyl) chemistry [21, 24, 26]. Final purification was carried out by reversed-phase high-pressure liquid chromatography, and peptide identity was verified by mass spectrometry [25].

Approval by the Institutional Review Board at the Yale University School of Medicine and St. Mary Hospital, Waterbury, CT, was obtained to collect endometrium and to carry out the endometrial experiments in this study. The method was previously published [25, 27]. Briefly, discarded endometrial tissues from premenopausal women that undergo hysterectomies due to benign indications (primarily myomas) were analyzed. Women provided consent via the standard hospital consent form. There were no anticipated risks associated with participation in this research over the risks associated with the surgical procedure that women underwent for removal of their uterus. After removal during hysterectomy, the uterus was evaluated in the Department of Pathology at Yale New Haven Hospital and in St. Mary Hospital, CT, Dr. M. Albini. The pathologist provided a sample of discarded endometrial tissue, transported to our research laboratory for sPIF experiments. Collected endometrial tissues 400 mg were digested with 0.1% DNAase 0.005%/1h/37C with gentle mixing [28, 29]. Following sedimentation of glandular structures and separation of ESC, the glands were purified of ESC and macrophage contaminants were removed by incubation at 37°C in a Falcon flask. This was followed by a second enzymatic digestion isolating the individual endometrial epithelial cells (EEC) from the glands for flow cytometric measurements. Glands were digested with trypsin (0.25%)–EDTA (0.03%)–DNase (0.1%) for 10 min at 37°C. Subsequently EEC were washed with DMEM–fetal bovine serum (FBS, 2%) and filtered out through a 37 μm mesh sieve. Stromal and epithelial cell dispersions were counted in a haemocytometer and cell viability was assessed using Trypan Blue exclusion method. ESC and EEC percent viability was >80%. Polyclonal antibodies to vimentin (Vm) and cytokeratin (Ck) were used to examine the homogeneity of ESC (Vimentin+) and EEC (Cytokeratin+) in cell smears of cellular dispersions. The maximum cross-contamination between ESC and EEC was <1%.[30]. Subsequently cells were cultured separately in cell bottom plates (8 well) for 5 days in DMEM-MCDB105 medium with 10% FCS (Sigma Chemical Co., St. Louis, MO) and for 24hrs in serum-free media containing 1–125 nM sPIF (15, or 9 aa). Epithelial and stromal were detached with Trypsin 0.01% EDTA 1nM (10min) and washed with PBS and cells were counted. Cells were stained with an anti-α2β3 integrin mAb (clone SSA6, 1:100) and an anti-mouse-FITC conjugate (1:100). α2β3 integrin expression was then determined by flow cytometry. A negative control with nonspecific IgG isotype-matched mAb (Dako, Carpinteria, CA) and an anti-HLA-I mAb (clone w6/32, 1:100) were introduced in each determination (n = 5). α2β3 integrin concentrations Mean ± SEM) were expressed as percentage of basal antigen expression in epithelial cells without sPIF. P < 0.05 was considered statistically significant. The sPIF-induced effect was compared to cells exposed to BSA 2%, 10% FCS, or PIFscr 1-500nM used as controls.

In other experiments endometrial cells were isolated and resuspended in RPMI-1640, grown to confluence, leucocyte free (<1%). After reaching confluence, cells were decidualized using 10^-8 M estradiol and 10^-7 M of the synthetic progestin analogue R5020 (Du Pont/New England Nuclear Products, Boston, MA). Cells were switched to defined medium containing insulin, transferrin, and selenium, (Collaborative Research, Inc., Waltham, MA) with 5 μM trace elements (GIBCO), and 50 μg/ml ascorbic acid (Sigma-Aldrich) and treated overnight with sPIF (100 nM) or vehicle only. The media and tissues were collected and frozen at −80°C. Products secreted into the media were analyzed using the bead-analysis kit (Bio-Rad, CA) and the lysates phospho-proteins by BioPlex kit (Bio-Rad, CA) in triplicate determining (Mean±SEM) activity, following the manufacturer instructions. sPIF and PIFscr effect (10 nM- 10mg/ml) on isolated kinases was tested by using the proprietary Fastkinase method analyzing direct action of peptides generating IC50 data, XlFit by software (MDSPS, Bothell, WA).

Total RNA was extracted from each cell culture as previously reported [25]. Analysis of HESC ± sPIF 100 nM (n = 3/group) was carried out by using Affymetrix (Santa Clara, CA) U133 Plus 2.0 Array (18,400 human genes), followed by fluorescence labeling and hybridization with Fluidics Station 450 and optical scanning with GeneChip Scanner 3000 (Affymetrix) at W. M. Keck Foundation Biotechnology Resource Laboratory, Yale University, New Haven, CT. Raw data were analyzed by GeneSpring software (Agilent, Santa Clara, CA), normalized for interchip and intrachip variation to eliminate false-positive results. Microchip was analysed by National Institutes of Health’s ImageJ image-analysis software (http://​rsb.​info.​nih.​gov/​ij). P < 0.05 and two fold change was considered statistically significant. See Additional file 1: Table S1 describes signal intensity and P value associated with gene detection.

Data was analyzed using Meta-plex software, t-test; significance was set at p < 0.05. The Bioplex data was analyzed by using log transformed data, followed by ANOVA P < 0.05. For gene experiments the data were first analyzed using Student’s t-test with P < 0.05 results followed by fold change testing with a cutoff equal or >2-fold reported.

sPIF effect on endometrial receptivity was investigated modeling the early luteal phase at a stage when circulating estrogen and progesterone levels are low (no direct embryo-endometrial contact). For this end, primary epithelial and stromal endometrial cells cultured for five days reaching confluence were used, exposing them to sPIF for 24hrs. Two different versions of sPIF were tested, both found in the embryo culture media, having the same core 9 first aminoacids. It was found that in epithelial cell cultures both, sPIF(15aa) and sPIF (9aa) at low physiologic concentrations increased α2β3 integrin, an important marker of endometrial receptivity, up to 4-fold in a dose-dependent manner (P < 0.01). (Figure 1a,b) Two different complimentary modalities of analysis were conducted; first qualitative analysis using an anti-α2β3 integrin-antibody carrying out staining of epithelial cells at 5 days of culture. Second, this was corroborated by a semi-quantitative analysis using flow cytometry (FC). The data showed that sPIF is most effective at low concentrations similar to physiologic range found in pregnancy. Significant stimulatory effect was noted with both ligands already at a low 1nM concentration, effect being dose dependent. sPIF specificity was confirmed since PIFscr (scrambled peptide, same amino acids as sPIF but in different order), bovine serum albumin (2% BSA), or 10% FCS effect tested in parallel and used as controls, had no effect. Importantly, the increase seen in α2β3 integrin in epithelial cells was not replicated by exposure to primary culture of non-decidualized stromal cells when tested with two independent methods (not shown). Overall, these experiments indicated that sPIF promotes a critical pro-implantation marker in uterine epithelium independently of sex-steroid exposure.

The endometrial epithelial cells are the site where embryo-derived signaling primes post-fertilization to promote implantation. sPIF increased α2β3 integrin in epithelial cells in primary culture and not in non-decidualized stromal cells. Therefore, we inquired whether sPIF also has similar effect on integrins in HESC, which are estrogen and progestin-induced decidualized stromal cells. sPIF in HESC as expected, does not affect ITGB3 expression the gene that encodes for α2β3 integrin, however, sPIF promotes ITGA2- (1.96 fold), a platelet membrane integrin. An increase in ITGB1(1.92, fold) a fibronectin receptor which is activated by Netrin-4 involved in vascular development was also noted. In contrast, sPIF down-regulated ITGA9 (−2.1 fold), expression, a receptor for VCAM1, cytotactin and osteopontin which binds to VEGF and promotes adhesion and cell migration. Thus, we showed that sPIF effect is dynamic and modulates the endometrium in a different manner, dependent on the evolving stage of embryo-maternal cross-talk at pre-and during implantation period.

In line with the described beneficial pro-inflammatory effects of sPIF on HESC genes expression, we examined whether they are also translated into changes in secreted pro-inflammatory mediators- relevant for uterine environment conditioning. There was a significant up-regulation of major inflammatory interleukins, IL8 (82-fold) IL1β (7-fold) and IL6 (10.7-fold). The increase in intercellular adhesion molecule 1, (ICAM-1, CD54) gene expression was coupled with ligand secretion, a leukocyte adhesion protein (integrin alpha-L/beta-2). The chemokines (growth-related oncogene-α (GRO-α, CXCL1) gene expression chemotactic for neutrophils, was coupled with protein secretion. The monocyte chemotactic protein-1 (MCP-1, CCL7) gene expression was also coupled with protein secretion. (Figure 2a,b) We show that sPIF-induced gene expression leads to pro-inflammatory ligands secretion favorably conditioning the uterine environment for embryo implantation.

We reported that sPIF did not synergize with exogenously added EGF in promoting trophoblast invasion. Since EGF is a member of a major proliferation-promoter family expressed by HESC, we examined sPIF effect on EGF and related GFs expression. Amphiregulin and epiregulin genes expression increased which encode proteins that promote decidual and cultured embryos development. In contrast, pro-proliferative betacellulin expression decreased an EGF receptor ligand as well as IGF1 while TGFβ2 decreased mildly. (Table 1) HBEGF (heparin binding) and EGF gene itself or its receptor expression was not affected. Significantly, increased expression of several fibroblast growth factors (FGF) was observed (Table 2). sPIF exerts a highly selective modulatory effect on HESC, balancing decidual pro-implantation properties while controlling excessive pro-proliferative signals expression.

Since sPIF modulates GFs in HESC, we aimed to determine whether it is exerted by regulation of downstream transcription factors. Kinases added a phosphate molecule (serine, thrionine, or tyrosine) are activated. sPIF markedly down-regulates critical transcription factors involved in proliferative pathways (Figure 3a,b,c). A number of activated-kinases activity involved in MAPK pathway decreased; two proliferation promoters; ERK- activator kinase (p-Thr202, Tyr204) -ERK1/2, (P < 0.01) which was followed downstream by a major decrease in p-MEK1 (p-Ser 218, Ser 222), (P < 0.04). In addition to the decrease in TGFβ2 expression, p-38-MAPK (p-Thr180, Tyr182), (P < 0.04) also decreased, involved in response to stress stimuli. There was also a significant decrease in MAPK8IP2 (−2.1 fold) (MAPK8 interacting protein-2) expression which encodes a scaffold protein that mediates c-Jun amino-terminal kinase signaling pathway, protecting against reactive oxygen species. The mild decrease in pro-inflammatory p-NFkB (P = 0.06) was accentuated with a major decrease in TNFRS11 expression, (−25-fold) and FAS (−3.3 fold) major NFkB activators, while p-IKB, and p-AKT were not affected by sPIF. Upon examining associated phosphatases gene expression that inactivate kinases, it was found that the phosphatases PTPRZ1 (phosphacan, 5.5-fold) which interacts with P53-a major proliferation controller, and PPP2R2C (2.9-fold) increased, both of which inhibit MAPK and proliferative pathways. Thus sPIF has a dual coordinated effect on both kinases and phosphatases. In addition, these pathways are directly involved in down-regulation of both proliferation and inflammation promoting pathways in HESC.

We found that sPIF modulates differentially endogenous GFs expression and this is associated with regulation of MAPK pathways. To examine whether sPIF has a direct effect on MAP kinases enzymes catalytic activity, the Fastkinase method was used to evaluate a given ligand effect on the enzyme activity itself as part of a panel. In contrast to the marked inhibitory effect on phosphoproteins expression seen in HESC, sPIF did not affect recombinant kinases involved in the MAPK pathway ERK1,2, MEK-1, p-38-MAPK, or NFkB activity in vitro (Figure 4). sPIF also did not affect directly the EGF protein tyrosine kinase activity, the down-stream activator of the growth factor. Thus, sPIF down-stream inhibitory effects on proliferative elements in HESC is exerted by down-regulation of pro-proteins involved in the MAPK pathway, which is independent of a direct action on the catalytic site of these enzymes.