Inhibition of epidermal growth factor receptor restores decidualization markers in stromal fibroblasts from women with endometriosis

Endometrial stromal fibroblasts (eSF) from women with (eSF endo) and without (eSF non-endo) endometriosis were treated with inhibitors to EGFR tyrosine kinase (gefitinib), mTOR (rapamycin) and MAPK kinase 1/2 (MEK1/2) (UO126) during 8-bromoadenosine 3’,5’-cyclic monophosphate (8-br-cAMP)–stimulated decidualization. Decidualization was assessed by evaluating expression of insulin growth factor binding protein 1 (IGFBP1), prolactin (PRL) and forkhead box protein O1A (FOXO1A) by quantitative real-time PCR.

Gefitinib restored expression of decidualization markers in eSF endo to levels consistent with those in eSFnon-endo. Elevated levels of phosphorylated mTOR in eSF endo were reduced to levels found in eSF non-endo, by gefitinib during treatment with 8-br-cAMP. Additional gene expression analyses suggested dysregulation of EGFR negative feedback regulators in eSFendo.

Results implicate EGFR signaling as an underlying cause for aberrant cAMP-induced decidualization in women with endometriosis, and provide a potential target for management of infertil- ity associated with the disease. The reduction of p-mTOR levels in eSF endo during 8-br-cAMP treatment suggests cooperation between EGR and protein kinase A signaling in the regulation of mTOR in eSF.

Decidualization, Endometriosis, Epidermal growth factor receptor, Fibroblast, Mammalian target of rapamycin, Mitogen-activated protein kinase Accepted: October 28, 2014

Endometriosis is a chronic estrogen-dependent disorder affecting 6%-10% of reproductive aged women (1, 2). It is characterized by extrauterine endometrial tissue that elic- its an inflammatory response mainly in the pelvis (3), con- tributing to chronic pelvic pain and associated infertility due to ovulatory dysfunction, poor egg quality, abnormal © 2014 Wichtig Publishing - ISSN 2035-9969 2 EGFR inhibition restores decidual markers women with endometriosis (eSF endo) do not express ap- preciable amounts of the decidual marker insulin growth factor binding protein 1 (IGFBP1) due to inherent insen- sitivity to PKA pathway stimulation (14, 15). t reatment of ectopic eSF from ovarian endometriomas with small molecule inhibitors of AKt or PI3K during in vitro decidu- alization increased expression of IGFBP1 and its upstream transcriptional regulator forkhead box 1 (FOXO1) (16). We previously demonstrated that treatment of stromal fibroblasts from eutopic endometrium of women without endometriosis (eSFnon-endo) with 8-br-cAMP resulted in up- regulation of IGFBP1; whereas eSF endo were refractory to 8-br-cAMP (14). In addition, the inhibition of MAPK kinase 1/2 (MEK1/2) with UO126 did not restore 8-br- cAMP up-regulation of IGFBP1 in these cells from women with disease (2). thus, the limited use of small molecule inhibitors, reported in a few studies, has met with fair success in restoring decidualization marker expression in eSF endo. Herein, we hypothesized that a combination of overactive PI3K/AKt/mtOR, MAPK and EGFR signaling pathways is causative of aberrant 8-br-cAMP-induced decidualization marker expression in eSF from women with endometriosis. We treated eSF with the selective chemical inhibitors ge- fitinib (EGFR tyrosine kinase inhibitor), rapamycin (mtOR inhibitor) and UO126 concomitantly with 8-br-cAMP , and performed multiplex and quantitative real-time polymerase chain reaction (qPCR) analyses to investigate their effects on decidualization marker expression. Our results indicate that inhibition of EGFR tyrosine kinase activity restored expression of specific decidual markers in eSF endo to levels in eSFnon-endo.

Human endometrial samples Human eutopic endometrial tissue samples were obtained in accordance with the guidelines of the Declaration of Helsin- ki, from 24- to 42-year-old women undergoing endometrial biopsy or hysterectomy for diagnosis or treatment of pelvic pain, fibroids, prolapse and/or endometriosis (tab. I). no ec- topic lesions or tissues were collected from patients or used in the present study. Presence or absence of endometriosis was confirmed by laparoscopic visualization and histologi- cal analysis of peritoneal lesions. Staging of endometriosis uterine endometrium and/or compromised embryo im- plantation (1, 4). Ectopic endometrial lesions (outside the uterus) and eutopic endometrium (within the uterus) of women with this disease are stimulated by estradiol, but are resistant to the action of progestins and progesterone (P 4) (5, 6). the latter regulates nociceptive pain thresh- olds in innervated endometriosis lesions (7) and, within the uterus, is essential for endometrial epithelial secretory transformation and stromal fibroblast decidualization, crit- ical for the establishment and maintenance of pregnancy. thus, resistance to P 4 and progestins can have a major impact on pain relief, disease control and fertility success of affected women. Several signaling pathways have been implicated in these aspects of the pathophysiology of progesterone resistance and compromised stromal fibroblast decidualization, in- cluding protein kinase A (PKA), mitogen-activated protein kinase (MAPK), and phosphatidylinositide-3 kinase/AKt/ mammalian target of rapamycin (PI3K/AKt/mtOR) (8). Un- derstanding involvement and cross-talk among these path- ways, and whether pathway-specific inhibitors can restore endometrial cell function or biomarkers, is of great interest as they hold promise to mitigate endometrial cellular dys- function in the setting of disease. Constitutive activation of PI3K/AKt and MAPK pathways has been reported in both eutopic and ectopic endometrial tissues in endometriosis (9). We previously found higher levels of phosphorylated (p)-ERK1/2 in stromal fibroblasts from women with severe endometriosis versus women without disease in vivo (2). Higher levels of phosphorylated (p)-mtOR have been found in ovarian endometrioma tis- sues (10), and active AKt1 is elevated in eutopic endo- metrium of women with versus without endometriosis (11). In addition, mRnA expression of epidermal growth factor receptor (EGFR), which is upstream of both PI3K/AKt and MAPK pathways, is increased in eutopic endometrium in the early secretory phase in severe versus mild disease (12). these data suggest a role for EGFR in increased acti- vation of these pathways within the stromal compartment of the uterus in women with endometriosis. Consistent with this notion is the finding that members of the EGFR family HER1, HER2 and HER3 exhibit greater mRnA ex- pression in whole eutopic endometrium from women with versus without disease (13). When treated with decidualizing stimuli (P 4, progestin and/ or 8-bromoadenosine 3’,5’-cyclic monophosphate [8-br- cAMP]), eutopic endometrial stromal fibroblasts (eSF) from © 2014 Wichtig Publishing - ISSN 2035-9969 3 Erikson et al was defined according to the Revised American Society for Reproductive Medicine classification system (17). no patients used hormone treatments within 3 months of sur- gery. Samples were obtained through the UCSF nIH Human Endometrial tissue and DnA Bank with appropriate institu- tional review, approvals and written informed consent from all participating subjects, as approved by the University of California San Francisco Committee on Human Research. Endometrial samples were processed for cell culture experi- ments as described below. eSF isolation and culture eSF from non-endometriosis (eSF non-endo) (n = 5) and endo- metriosis (eSFendo) (n = 5) subjects were isolated by digest- ing endometrial tissue samples with collagenase followed by filtration as previously described (18, 19). Isolated eSF were cultured in growth medium (phenol red-free medium of 3:1 high-glucose phenol red-free DMEM/MCDB-105, 0.676 mM sodium pyruvate, 10% charcoal-stripped fe- tal bovine serum [FBS], 1% penicillin-streptomycin mix, 50 µg/ml gentamycin, 5 µg/ml insulin), with medium re- placed every 2-3 days for up to 6 passages. All cells used in the experiments described herein were from the fourth passage. For qPCR and multiplex experiments, confluent cells were incubated for 24 hours in low-serum medium (3:1 high-glucose phenol red-free DMEM/MCDB-105, 0.75 mM sodium pyruvate, 50 µg/ml gentamycin, 2% FBS) before treatment. For multiplex experiments, cells were incubated in either EGF (20 ng/ml) for 30 minutes or 8-br-cAMP (0.5 mM) for 60 minutes with inhibitors to mtOR (rapamycin; 10 nM), MEK1/2 (UO126; 10 µM) or EGFR t yr1173 t yr992 (gefitinib; 10 µM) alone or in combi- nation. these time points were determined by treatment of eSF with either EGF or 8-br-cAMP and then observ- ing the time point at which optimal p-ERK1/2 activation TABLE I - HUMAn PARt ICIPAnt CHARACt ERISt ICS Subject ID Phase Age Procedure Ethnicity Diagnosis Analysis S01 PE 28 l aparoscopic supracervical hysterectomy White Dysmenorrhea, menometrorrhagia, ovarian cyst qPCR S02 PE 42 Supracervical hysterectomy White Symptomatic fibroids, adenomyosis qPCR, multiplex S03 SE 24 EBx Asian n atural cycle donor qPCR, multiplex S04 PE 37 t ubal ligation, EBx Asian Undesired fertility qPCR S05 PE 39 t ubal ligation, EBx Asian Undesired fertility qPCR, multiplex S06 SE 37 Operative laparoscopy, EBx White Endometriosis, stage IV, pelvic adhesion, adenomyosis qPCR, multiplex S07 PE 30 Operative laparoscopy, EBx Unknown Endometriosis, stage IV, pelvic pain, uterine bleeding qPCR S08 PE 32 Operative laparoscopy, EBx White Endometriosis, stage IV, pelvic pain, uterine bleeding, dysmenorrhea qPCR, multiplex S09 PE 42 Operative laparoscopy, EBx White Endometriosis, stage IV, pelvic pain, pelvic adhesions qPCR S10 SE 40 Operative laparoscopy, EBx Asian Endometriosis, stage IV, pelvic pain, uterine bleeding qPCR, multiplex EBx = endometrial biopsy; PE = proliferative endometrium; qPCR = quantitative real-time reverse transcription polymerase chain reaction; SE = secretory endometrium. © 2014 Wichtig Publishing - ISSN 2035-9969 4 EGFR inhibition restores decidual markers (EGF) or inhibition (8-br-cAMP) occurred (2, 20). All inhibi- tors were delivered in dimethyl sulfoxide (DMSO), and all treatments were adjusted to contain 0.1% DMSO. Cells were then harvested for protein analysis as described be- low. For decidualization, cells were treated with 0.5 mM 8-br-cAMP for 96 hours in low-serum medium, condi- tions previously determined optimal for decidualization (2, 14), with or without the inhibitors listed above. Media were refreshed after 48 hours. Optimal inhibitor concen- trations were determined using previously published data on effective doses (2, 21, 22). Samples without treatment were collected at baseline (t = 0) to serve as controls. After 96 hours of treatment, cells were harvested for RnA or protein isolation as described below. All sample treat- ments were performed in duplicate in independent ex- periments to ensure reproducibility of results. Quantitative RT-PCR t otal RnA was isolated from cultured, treated eSF using the nucleoSpin RnA kit (Macherey-nagel, Bethlehem, PA, USA) following the manufacturer’s instructions. t otal RnA was quantified by UV spectrophotometry (nanodrop, Wilmington, DE, USA), and reverse transcription was per - formed using random primers and the iScript R t reagents following the manufacturer’s protocol (Bio-Rad laborato- ries, Hercules, CA, USA). mRnA levels were determined by qPCR using the Mx 3005 Pro (Stratagene, la Jolla, CA, USA). Primer sets used are listed in Supplemental t able I. cDnA (10 ng) was amplified by SYBR Supermix (thermo Fisher Scientific, Waltham, MA, USA) using the following protocol: (i) DnA polymerase activation at 95°C for 15 min- utes; (ii) 40 cycles of denaturation at 95°C for 15 seconds, annealing at 60°C for 45 seconds and extension at 72°C for 60 seconds. the absence of primer dimers was deter - mined through dissociation curve analysis by performing a cycle of denaturation at 95°C for 1 minute followed by 35 cycles of increasing temperature for 33 seconds from 65°C to 97°C at the end of each analysis. Relative expression was calculated using the relative standard curve method (23), where a standard curve is created for each primer set using pooled cDnAs from the entire experimental sample set, and efficiency of the cDnA standards is between 90% and 110%. Samples were assayed in duplicate, and rela- tive gene expression was normalized with beta actin (ACt- nB) and t = 0 controls within each group to control for time in culture. Multiplex assays Following treatment, cells were washed in phosphate- buffered saline (PBS) and lysed with RIPA buffer contain- ing a protease inhibitor cocktail (2 mM AEBSF , 1 mM EDtA, 130 µM Bestatin, E-64 14 µM, 1 µM leupeptin, 0.3 µM Aprotinin) (Millipore, Bedford, MA, USA). After centrifuga- tion to clear cellular debris (10,000 g, 10 minutes, 4°C), protein quantity was evaluated by Bradford assay (Bio- Rad) using the manufacturer’s instructions. Milliplex MAP multiplex protein assays were performed according to the manufacturer’s protocol (Millipore). Briefly, cell lysates (3 µg protein/well) were diluted 1:1 with Milliplex MAP As- say Buffer 2. the filter plate was pre-wetted with Assay Buffer 2, and the buffer removed by vacuum filtration. Mil- liplex MAPmates were combined and 25 µl was added per well (Suppl tab. II). After adding 25 µl of diluted cell lysate to the appropriate wells, the plate was incubated overnight at 4°C with shaking. the lysate was removed by vacuum filtration, washed 3 times with Assay Buffer 2, and 25 µl of detection antibody was added for 1 hour at room tem- perature. Diluted streptavidin-phycoerythrin (SAPE) was then added after filtration, the plate was incubated for 15 minutes at room temperature, and amplification buffer was added before a final incubation of 15 minutes. the SAPE/ amplification buffer mix was removed by filtration, 150 µl of Assay Buffer 2 was added, and the plate was analyzed on a BioPlex 200 (Bio-Rad). the ratio of phospho protein to glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was calculated for each target. p-EGFR multiplex assays Analysis of p-EGFR (t yr1173) was performed using a mul- tiplex approach similar to that for other phospho protein targets described above. After treatment and total protein extraction from cells, lysates were diluted in cell lysis buf- fer to a concentration of 10 µg/50 µl and incubated with BioPlex Pro magnetic cell signaling beads (Bio-Rad) for p-EGFR (tyr1173) and GAPDH overnight at 4°C with shak- ing. Beads were washed and incubated for 30 minutes with antibodies for p-EGFR and GAPDH at room temperature, washed and further incubated with SAPE for 10 minutes at room temperature. Beads were then washed, resuspended in Bead Resuspension Buffer and analyzed on a BioPlex 200 (Bio-Rad). the ratio of phospho protein to GAPDH was calculated for p-EGFR. © 2014 Wichtig Publishing - ISSN 2035-9969 5 Erikson et al Data analysis All quantitative data were subjected to analyses of variance (AnOVA) using general linear model procedures available with SAS (SAS 2012-2013; SAS Institute, Cary, nC, USA). For transcript comparisons, data were generated from fold change of each transcript to ACtnB transcripts using individually determined standard curves. For protein com- parisons fold change to GAPDH protein was performed. Statistical models considered variation due to the main effect of inhibitor treatment, including no treatment ve- hicle control (DMSO). t reatment effects were identified by performing a set of preplanned contrasts to include com- parisons of specific groups for each target transcript or protein (24). In all cases, significance was set at a p value of <0.05.

t o prove/disprove our central hypothesis that overactive EGFR signaling in eSF endo is causative of the compro- mised decidualization response observed in these cells, we pursued the following: (i) analyzed EGFR pathway ac- tivation, including downstream ERK1/2 and mtOR, in eSF to determine activation status of these pathways in dis- ease; (ii) demonstrated compromised eSF decidualization response on a background of increased EGFR signaling (12, 13), and recreated this effect in eSF non-endo by activat- ing EGFR; and (iii) used small molecule inhibitors to EGFR, mtOR and MEK1/2 in an attempt to restore the decidual- ization response in eSFendo. eSF from women with endometriosis do not decidualize in response to cAMP Previous data have demonstrated that eSF endo exhibit an altered response to PKA stimulation (14), which manifests in reduced decidualization capacity (2, 15). Our objective in this experiment was to demonstrate compromised de- cidualization response by eSF on a presumed background of increased EGFR signaling (12, 13). After treatment with cAMP for 96 hours, we measured expression of 3 key de- cidualization genes by PCR, IGFBP1, PRL and FOXO1A, and calculated fold changes against expression levels at t = 0. IGFBP1 expression increased over 100,000-fold in eSF non-endo, while there was a minimal increase in eSF endo (Fig. 1A). PRL expression increased over 35,000-fold in eSFnon-endo, with a small increase in eSF endo (Fig. 1B). We also observed a significant increase in FOXO1A expression in eSFnon-endo, but little increase in eSF endo (Fig. 1C). these data demonstrate a limited capacity of eSF endo to respond to cAMP , and suggest that overactive EGFR found in these subjects could be responsible for aberrant decidualization in eSF endo. Components of the EGFR signaling pathway are activated in endometriosis to better understand the role of signaling in aberrant decidu- alization, we analyzed mtOR, EGFR and ERK phosphory- lation by 8-br-cAMP at 60 minutes, a time point found to decrease ERK1/2 phosphorylation in eSFnon-endo (2). Contrary to our hypothesis, there were no significant differences in p- EGFR at baseline (t = 0), and p-EGFR increased in eSFnon-endo when treated with 8-br-cAMP for 60 minutes, with no con- comitant increase in eSF endo (Fig. 2A). p-mtOR was elevated in eSFendo at t = 0 and increased in both eSFnon-endo and eS- Fendo when treated with cAMP , with levels in diseased cells remaining significantly higher than in nondiseased cells after 60 minutes of treatment (Fig. 2B). Finally, p-ERK1/2 was also elevated in eSF endo at t = 0, and decreased in eSFnon-endo while remaining elevated in eSFendo with 60 minutes of cAMP treat- ment (Fig. 2C). these data indicate differences in response of EGFR and mtOR to cAMP and subsequent phosphor - ylation states that may affect the onset of decidualization in eSF . EGF inhibition of decidualization in eSF from women without endometriosis is restored by gefitinib Previous experiments have described the inhibition of cAMP-induced decidualization by EGF in eSF (25). the

of this experiment was to inhibit decidualization in eSFnon-endo by stimulating EGFR with EGF . Figure 3 shows IGFBP1 mRnA expression after 96-hour treatment with EGF + 8-br-cAMP and compared with 8-br-cAMP alone.

showed that expression of IGFBP1 in eSF non-endo, when treated with EGF + 8-br-cAMP was equal to that in eSFendo and similar to levels found in eSF endo treated with 8-br-cAMP alone. When the EGFR tyrosine kinase in- hibitor gefitinib was added to EGF + 8-br-cAMP , IGFBP1 expression in eSFnon-endo was fully restored and allowed par- © 2014 Wichtig Publishing - ISSN 2035-9969 6 EGFR inhibition restores decidual markers Fig. 1 - Endometrial stromal fibroblasts from women with endome - triosis do not decidualize in response to 8-br-cAMP. Endometrial stromal fibroblasts from women without endometriosis (Non-Endo) or with endometriosis (Endo) were cultured for 96 hours in low- serum medium (DMEM-MCDB105 + 2% FBS) with 0.5 mM 8-br- cAMP to induce decidualization. Separate cells were collected prior to 8-br-cAMP treatment and served as baseline (t = 0 + control). After 96 hours, cells were harvested for total RNA extraction and subjected to quantitative real-time PCR analysis for A) insulin- like growth factor binding protein 1 (IGFBP1); B) prolactin (PRL); and C) forkhead box protein 01A (FOXO1A). Gene expression data were calculated using the standard curve method and normalized to ACTNB levels in each sample. All data were then normalized to t = 0 to control for time in culture. Data are represented as average fold change to t = 0 ± s.e.m. Asterisks indicate significant differ - ences by ANOVA (p<0.05). Fig. 2 - 8-br-cAMP regulation of cell signaling activation in en - dometrial stromal fibroblasts. Endometrial stromal fibroblasts from women without endometriosis (Non-Endo) and with endo - metriosis (Endo) were cultured for 60 minutes in low-serum me - dium (DMEM-MCDB105 + 2% FBS) containing 0.5 mM 8-br-cAMP. Cells (n = 3 each Non-Endo and Endo) were harvested for total protein extraction and subjected to multiplex protein analysis for A) phosphorylated-epidermal growth factor receptor (p-EGFR); B) phosphorylated-mammalian target of rapamycin (p-mTOR); and C) phosphorylated-extracellular-signal-regulated kinase 1/2 (p-ERK1/2). Phosphorylated-protein values were normalized to GAPDH levels for each sample. Data are represented as average ratios ± s.e.m. Asterisks indicate significant differences by ANOVA (p<0.05) where Non-Endo and Endo values were compared within treatment (t = 0 or cAMP). © 2014 Wichtig Publishing - ISSN 2035-9969 7 Erikson et al tial restoration of IGFBP1 expression in eSFendo. these re- sults indicate that overactivation of EGFR in disease could be a cause for decreased decidualization in response to 8-br-cAMP , and that blockade of EGFR tyrosine kinase with gefitinib restores decidualization marker expression in eSF endo. EGF increases activation of mTOR in endome- triosis We treated eSF with EGF for 30 minutes alone or with in- hibitors to specific signaling pathways, to evaluate EGFR, ERK and mtOR activation (in a pilot study, 30 minutes was found to be the optimal time for EGF activation of ERK1/2 in eSF – data not shown). p-EGFR increased similarly in eSF non-endo and eSF endo with all treatments, except for gefi- tinib + EGF which decreased p-EGFR in eSF independent of disease status (Fig. 4A). p-ERK1/2 was also detected in similar levels in eSF non-endo and eSF endo, except for t = 0 where ERK1/2 activation was increased in eSF endo versus eSFnon-endo. t reatment with both gefitinib and UO126 inhib- ited ERK activation by EGF (Fig. 4B). mtOR phosphoryla- tion was increased in eSFendo versus eSFnon-endo by EGF in all treatments, including treatment with EGF in combination with gefitinib (Fig. 4C). these data demonstrate that EGF is able to activate EGFR, ERK and mtOR pathways in eSF , independent of disease state. they also demonstrate that while gefitinib is able to decrease activation of both EGFR and ERK by EGF , the same was not evident with p-mtOR, suggesting an alternate route of activation of mtOR by EGF that is EGFR-independent. Gefitinib restores decidualization marker expres- sion in endometriosis We treated eSF with inhibitors to EGFR, mtOR and MEK1/2 during cAMP treatment to determine their ef- fect on restoring decidualization marker expression, and investigated the roles of these signaling pathways on the decidualization process in eSF . IGFBP1 expression remained decreased in eSFendo versus eSFnon-endo with 8-br- cAMP and rapamycin + 8-br-cAMP , and UO126 + 8-br- cAMP . In contrast, when treated with gefitinib, levels of IGFBP1 increased in eSF endo to levels equal to that found in eSF non-endo (Fig. 5A). Similar results were also found with 2 other decidualization markers, PRL (Fig. 5B) and FOXO1A (Fig. 5C). these results indicate that inhibition of EGFR, but not downstream mtOR or ERK pathways, is able to restore cAMP-induced decidualization marker expression in eSF endo. Fig. 3 - EGF inhibition of decidualiza- tion in endometrial stromal fibroblasts is restored by gefitinib. Endometrial stromal fibroblasts from women with- out endometriosis (Non-Endo) or with endometriosis (Endo) were cultured for 96 hours in low-serum medium (DMEM-MCDB105 + 2% FBS) with 0.5 mM 8-br-cAMP, 0.5 mM 8-br- cAMP + 20 ng/mL EGF or 8-br-cAMP + EGF + 10 µM gefitinib. Inhibitors were delivered in DMSO, and all treat- ments including cAMP were adjusted to contain 0.1% DMSO. Cells were harvested for total RNA extraction and subjected to quantitative real- time PCR for insulin-like growth factor binding protein 1 (IGFBP1). Relative gene expression was calculated using the standard curve method and nor - malized to ACTNB levels in each sam- ple. Data are represented as average fold change to t = 0 ± s.e.m. Different letters indicate significant differences by 2-way ANOVA (p<0.05). © 2014 Wichtig Publishing - ISSN 2035-9969 8 EGFR inhibition restores decidual markers Gefitinib affects mTOR activation in endometriosis Because gefitinib restored decidualization marker expres- sion in eSF endo, we investigated signaling pathway activa- tion when EGFR tyrosine kinase was inhibited with gefitinib. p-EGFR was increased in eSF non-endo versus eSF endo with 8-br-cAMP treatment, but increased in eSF endo versus eSFnon-endo with rapamycin + 8-br-cAMP (Fig. 6A). there were no differences in p-ERK1/2 between eSF non-endo and eSFendo with any treatment, aside from the t = 0 time point and 8-br-cAMP treatment, as observed earlier (Fig. 6B). UO126 did not decrease p-ERK1/2 levels in the presence of 8-br-cAMP , as it did when cells were treated with EGF , suggesting the possibility of direct activation of ERK1/2 by PKA stimulation in eSF (Fig. 6B). Activation of mtOR was higher at t = 0 in eSF endo as described earlier, and remained elevated in eSF endo versus eSF non-endo with 8-br-cAMP , ra- pamycin + 8-br-cAMP , and UO126 + 8-br-cAMP treatment. However, when cells were treated with gefitinib + 8-br- cAMP , the treatment that restored decidualization marker levels in eSF endo, p-mtOR levels were equal in eSF non-endo and eSFendo, although levels in eSF endo with this treatment were lower than those found in eSF endo treated with 8-br- cAMP alone (Fig. 6C). thus, these data demonstrate that inhibition of the tyrosine kinase domain of EGFR during cAMP treatment results in decreased levels of p-mtOR in Fig. 4 - EGF regulation of cell signaling activation in endometrial stromal fibroblasts. Endometrial stromal fibroblasts from women without endometriosis (Non-Endo) and with endometriosis (Endo) were cultured for 30 minutes in low-serum medium (DMEM-MCDB105 + 2% FBS) containing 20 ng/mL EGF, EGF + 10 µM gefitinib, EGF + 10 nM rapamycin, or EGF + 10 µM UO126. Inhibitors were delivered in DMSO, and all treatments including cAMP were adjusted to contain 0.1% DMSO. Cells were harvested for total protein extraction and subjected to multiplex protein analysis including A) phosphorylated-epidermal growth factor receptor (p-EGFR); B) phosphorylated-extracellular-regulated-kinase 1/2 (p-ERK1/2); and C) phosphorylated mammalian target of rapamycin (p-mTOR). Phosphorylated-protein values were normalized to GAPDH levels for each sample. Data are represented as average ratios ± s.e.m. Asterisks indicate significant differences between Non-Endo and Endo within treatment by ANOVA (p<0.05). © 2014 Wichtig Publishing - ISSN 2035-9969 9 Erikson et al eSFendo to levels equal to eSFnon-endo, suggesting that EGFR- mtOR signaling is critical during cAMP-induced decidual- ization in eSF . When combined with the data above showing EGF stimulating mtOR through EGFR-independent means, the data suggest interplay of an EGFR tyrosine kinase and PKA signaling in the regulation of mtOR activation in eSF . EGF receptor and negative feedback regulator expression is dysregulated in eSFendo Because differences in EGFR activation between eSF non- endo and eSF endo were minimal, we evaluated expression of several EGF-ligand binding receptors and negative feedback regulators to determine possible alternate regulation of EGFR signaling. Figure 7A shows expression of receptors EGFR and ERBB2, and negative regulators of EGFR, ERRFI1 and PTPRK at t = 0. levels of EGFR, ERBB2 and ERRFI1 mRnA were all higher in eSFendo versus eSFnon-endo; levels of PTPRK did not differ between eSFnon-endo and eSFendo. However, after 96 hours of 8-br-cAMP treatment, levels of EGFR, ERBB2, ERRFI1 and PTPRK mRnA all significantly increased in eS- F non-endo versus eSFendo (Fig. 7B). these data suggest some level of involvement of EGFR in decidualization in eSFnon-endo; however, when stimulated with cAMP , eSFendo do not prop- erly regulate activation of EGFR, as suggested by the lower levels of ERRFI1 and PTPRK expression. thus, increased Fig. 5 - Gefitinib restores decidualization marker expression in endometrial stromal fibroblasts from women with endometriosis. Endometrial stromal fibroblasts from women without endometriosis (Non-Endo) and with endometriosis (Endo) were cultured for 96 hours in low-serum medium (DMEM-MCDB105 + 2% FBS) containing 0.5 mM 8-br-cAMP, cAMP + 10 µM gefitinib, cAMP + 10 nM rapamycin, or cAMP + 10 µM UO126. Inhibitors were delivered in DMSO and all treatments including cAMP were adjusted to contain 0.1% DMSO. Cells were harvested for total RNA extraction and subjected to quantitative real-time PCR analysis for A) insulin-like growth factor binding protein 1 (IGFBP1); B) prolactin (PRL); and C) forkhead box protein 01A (FOXO1A). Gene expression data were calculated using the standard curve method and normalized to ACTNB levels in each sample, and normalized to t = 0 to control for time in culture. Data are represented as average fold change to t = 0 ± s.e.m. Asterisks indicate significant differences by ANOVA (p<0.05). © 2014 Wichtig Publishing - ISSN 2035-9969 10 EGFR inhibition restores decidual markers activation of EGFR may not be causative of aberrant de- cidualization in eSF endo, as levels of p-EGFR are actually higher in eSFnon-endo following 8-br-cAMP treatment. Rather, decreased decidualization in response to cAMP may be due to sustained unopposed activation of EGFR in eSFendo. Expression of select EGFR ligands are increased in endometriosis In addition to evaluating EGF family receptors and nega- tive regulators, we evaluated mRnA expression of several EGFR ligands. Expression of AREG and HBEGF was higher in eSFnon-endo than eSFendo at t = 0, while TGFA expression oc- curred at similar levels in eSFnon-endo and eSFendo at t = 0 (Fig. 8A). However, with 8-br-cAMP treatment, AREG increased in eSFendo versus eSFnon-endo and expression was inhibited in eSF independent of disease with UO126 + 8-br-cAMP (Fig. 8B). TGFA increased dramatically in eSF endo versus eSFnon- endo with 8-br-cAMP , gefitinib + 8-br-cAMP , and UO126 + 8-br-cAMP; rapamycin + 8-br-cAMP decreased expression of TGFA in eSF endo (Fig. 8C). Finally, HBEGF expression was not different between eSFnon-endo and eSFendo except with the combination of rapamycin + 8-br-cAMP , where expression was elevated in eSF endo versus eSF non-endo (Fig. 8D). these data demonstrate the possibility that autocrine/paracrine activation of EGFR by tGFA or AREG with cAMP treat- Fig. 6 - 8-br-cAMP regulation of cell signaling pathways in the presence of chemical inhibitors. Endometrial stromal fibroblasts from women without endometriosis (Non-Endo) and with endometriosis (Endo) were cultured for 60 minutes in low-serum medium (DMEM-MCDB105 + 2% FBS) containing 0.5 mM 8-br-cAMP, cAMP + 10 µM gefitinib, cAMP + 10 nM rapamycin, or cAMP + 10 µM UO126. Inhibitors were deliv- ered in DMSO, and all treatments including cAMP were adjusted to contain 0.1% DMSO. Cells were harvested for total protein extraction and subjected to multiplex protein analysis for A) phosphorylated-epidermal growth factor receptor (p-EGFR); B) phosphorylated-extracellular- regulated-kinase 1/2 (p-ERK1/2); and C) phosphorylated-mammalian target of rapamycin (p-mTOR). Phospho-protein values were normalized to GAPDH levels for each sample. Data are represented as average ratios ± s.e.m. Asterisks indicate significant differences between Non- Endo and Endo within treatment by ANOVA (p<0.05). © 2014 Wichtig Publishing - ISSN 2035-9969 11 Erikson et al Fig. 7 - EGF receptor and negative regulator expression in endometrial stromal fibroblasts. Endometrial stromal fibroblasts from women without endometriosis (Non-Endo) or with endometriosis (Endo) were cultured for 96 hours in low-serum medium (DMEM-MCDB105 + 2% FBS) with 0.5 mM 8-br-cAMP. Cells were harvested for total RNA extraction and subjected to quantitative real-time PCR analysis for epidermal growth factor receptor (EGFR), v-erb-b2 avian erythroblastic leukemia viral oncogene homolog 2 (ERBB2), ERBB receptor feedback inhibitor 1 (ERRFI1), and receptor-type tyrosine-protein phosphatase kappa (PTPRK) at t = 0 (A) and after 96-hour 8-br-cAMP treatment (B). Gene expression data were calculated using the standard curve method and normalized to ACTNB levels in each sample, and normalized to t = 0 to control for time in culture. Data are represented as average fold change to t = 0 ± s.e.m. Asterisks indicate significant differences by ANOVA (p<0.05). ment could result in EGFR activation in eSF endo. the dif- ferent expression patterns of these ligands depending on pathway inhibition also provide insight into the complexity of their regulation in eSF . the similar levels of AREG in eSFnon- endo and eSF endo with gefitinib, as well as the higher levels of TGFA in eSFendo versus eSFnon-endo with gefitinib suggest that dysregulation of EGFR signaling occurs downstream of ligand receptor binding. no significant differences in p-AKt , p-PtEn, p-P70S6K, cleaved PARP or active caspase 3 between eSF non-endo and eSFendo were observed regardless of treatment time or inhibi- tors present (Suppl Figs. 1 and 2). Simultaneous treatment with any 2-inhibitor combination did not have an additive effect on decidual marker gene expression (Suppl Fig. 3).

Herein, we used selective chemical inhibitors to block the EGFR, mtOR and MAPK signaling pathways in an effort to induce decidualization marker expression in eSF from women with endometriosis. Using gefitinib, an inhibitor of the EGFR tyrosine kinase domain (t yr1173 and t yr992), we were able to restore gene expression of 3 key mark- ers of decidualization: IGFBP1, PRL and FOXO1A. t o our knowledge, this is also the first study providing evidence of the dysregulation of protein tyrosine phosphatase recep- tor type kappa (PtPRK), a negative feedback regulator of EGFR, during 8-br-cAMP-induced decidualization in eSF . these observations are crucial to understanding the de- cidualization process in eSF and its aberrant regulation in endometriosis. Decidualization Decidualization is a complex process that involves a number of autocrine/paracrine loops involving cytokines and growth factors and other secreted molecules includ- ing interleukin-11 (Il-11), PRl, IGFBP1 and HBEGF (26). Previously, studies have demonstrated a blunted expres- sion of specific decidual markers in response to activa- tion of the cAMP/PKA pathway in eSF from women with endometriosis versus those without disease (2, 14, 15). We also previously determined that there is no defect in cAMP hydrolysis in eSF endo, indicating that the diminished response of these cells to cAMP/PKA stimulation may be due to insensitivity of certain genes to PKA stimulation or a difference in the regulation of cell signaling path- ways in eSF endo (14). In the present study, IGFBP1, PRL and FOXO1A expression increased in eSF endo to levels consistent with those found in eSF non-endo, by treatment with the EGFR tyrosine kinase domain inhibitor gefitinib (Fig. 5), suggesting dysregulation of EGFR signaling in eSFendo during cAMP-induced decidualization. © 2014 Wichtig Publishing - ISSN 2035-9969 12 EGFR inhibition restores decidual markers We detected elevated levels of p-ERK1/2 and p-mtOR in eSFendo when compared with normal counterparts prior to treatment with 8-br-cAMP , both of which are downstream of EGFR in their respective pathways. We previously found an increase in EGFR gene expression and gene expression of selected EGFR ligands in eSF from women with severe endometriosis versus women with mild disease (12), while another study showed decreased expression of EGFR in eSF from women with disease (27). A decreased response to EGF by eSF endo (28) and an inhibitory effect of EGF on 8-br-cAMP-induced decidualization (25) have also been reported. these data collectively suggest it is possible that several mechanisms are operating with regard to EGFR and other ERBB receptors in eSF endo, which may be affect- ed by increased copy number or higher levels of tyrosine kinase activity. It is also possible that negative regulators of EGFR, including ERRFI1 and PtPRK, do not respond to stimuli in eSF endo as they do in eSF non-endo, thus affect- ing down-regulation of EGFR tyrosine kinase. Evidence for dysregulation at different levels within the EGFR system is apparent in the current study from the impaired regulation of EGFR, ERBB2, ERRFI1 and PTPRK expression by 8-br- cAMP in eSF endo, which remained essentially nonrespon- sive to 8-br-cAMP (Fig. 7). Collectively, these data coupled Fig. 8 - Expression of select EGFR ligands in endometrial stromal fibroblasts. Endometrial stromal fibroblasts from women without endome - triosis (Non-Endo) and with endometriosis (Endo) were cultured for 96 hours in low-serum medium (DMEM-MCDB105 + 2% FBS) containing 0.5 mM 8-br-cAMP, cAMP + 10 µM gefitinib, cAMP + 10 nM rapamycin, or cAMP + 10 µM UO126. Inhibitors were delivered in DMSO, and all treatments including cAMP were adjusted to contain 0.1% DMSO. Cells were harvested for total RNA extraction and subjected to quantita - tive real-time PCR analysis for amphiregulin (AREG), heparin-binding epidermal growth factor-like growth factor (HBEGF), and transforming growth factor alpha (TGFA) at t = 0 (A) or after 96-hour treatment with 8-br-cAMP (B-D) . Gene expression data were calculated using the standard curve method and normalized to ACTNB levels in each sample, and normalized to t = 0 to control for time in culture. Summary of 8-br-cAMP-induced pathways in decidualization of human endometrial stromal fibroblasts in women without endometriosis (Normal) and with disease (Endometriosis). A) Relative expression data calculated at t = 0. B-D) Average fold change to t = 0 ± s.e.m. Asterisks indicate significant differences by ANOVA (p<0.05). © 2014 Wichtig Publishing - ISSN 2035-9969 13 Erikson et al sion of a number of cancers, presumably by stimulating tumor cell proliferation (34). this raises the possibility that endogenous tGFA acts as an autocrine/paracrine ligand stimulating EGFR signaling that inhibits decidualization. However, changes in TGFA or AREG expression did not correlate with decidualization marker changes consistently across treatments. Addition of rapamycin in combination with 8-br-cAMP reduced TGFA expression, but this was not correlated with increased expression of decidualization markers in eSF endo. Conversely, 8-br-cAMP in combination with gefitinib had no effect on TGFA expression, but in- creased expression of decidualization markers in eSF endo. therefore, increased availability of EGFR ligand alone does not seem to account for the inhibition of decidualization marker expression in eSF endo, suggesting a potential dys- regulation of the EGFR signaling pathway residing down- stream from ligand-receptor binding. mTOR and decidualization markers Another observation made in this study is the relationship between mtOR activation and expression of decidualization marker genes. levels of p-mtOR were constitutively elevat- ed in eSFendo versus eSFnon-endo and in all cases when treated with 8-br-cAMP (Fig. 6). the one exception occurred when eSF were treated with gefitinib in addition to 8-br-cAMP , which decreased p-mtOR in eSF endo to levels equal to those in eSFnon-endo. Addition of gefitinib to 8-br-cAMP treatment was also when we observed equivalent levels of expression of IGFBP1, PRL and FOXO1A in eSFendo and eSFnon-endo (Fig. 5). this would suggest that inhibition of EGF signaling in eS- Fendo causes a sufficient decrease in the elevated levels of p-mtOR for decidualization to occur. cAMP or its analogues are not known to directly activate the mtOR signaling path- way. A study previously performed in thyroid cells demon- strated that cAMP , in combination with tSH, could activate mtORC1, which then led to downstream activation (35). this activation of mtOR by the cAMP/tSH combination occurred independent of AKt phosphorylation, suggesting direct ac- tivation of mtOR by cAMP/tSH. this observation, coupled with our own data, suggests that cAMP may stimulate mtOR signaling through nonclassic yet still unclear mechanisms. thus, it is possible that the dysregulation of EGFR signal- ing we propose is occurring downstream of ligand-receptor binding results in hyperactivation of mtOR in eSF endo that is normalized to that of eSFnon-endo with inhibition of the EGFR tyrosine kinase domain. with restoration of decidual marker expression in eSF endo following treatment with gefitinib, demonstrate that the EGFR signaling system plays a key role in the aberrant ex- pression of decidualization markers seen in endometriosis. EGFR inhibitors Our study showed lower levels of ERRFI1 expression in 8-br-cAMP-treated eSF endo compared with their normal counterparts, consistent with its known altered expres- sion in endometrium from women with endometriosis (26, 29-31). ERRFI1 has also previously been shown to be in- volved in regulation of uterine homeostasis in response to E2 in mice (31), and knockdown of ERRFI1 in human bronchiolar epithelial leads to an increase in both p-EGFR and p-AKt (32), resulting in increased cell proliferation. the current study also revealed the dysregulation of another negative receptor of EGF signaling in eSF endo, receptor- type protein tyrosine phosphatase kappa (PtPRK), to our knowledge not previously reported. PtPRK acts as a regu- lator of EGFR activity by dephosphorylating the receptor at tyrosines 1068 and 1173 (33), and its silencing increases basal and EGF-stimulated EGFR tyrosine kinase phos- phorylation and downstream ERK activation. In the cur - rent study, 8-br-cAMP-stimulated levels of both ERRFI1 and PTPRK were markedly increased by gefitinib, and showed less prominent changes with the other inhibitors, but invariably with a persistent blunted response in eSF endo compared with controls. Alternatively, it is possible that the effect of gefitinib on dephosphorylation of p-EGFR tyr1173 resulted in decreased expression of PTPRK in eSF endo. the same effect was not seen in eSF non-endo, however, indicat- ing differences in how eSF respond to these stimuli in the context of disease. Combined, these results indicate that, regardless of EGFR expression or activation, eSF endo may demonstrate a compromised capacity to regulate EGFR activity during 8-br-cAMP decidualization. EGFR ligands the expression of several EGFR ligands was evaluated to investigate the potential for autocrine or paracrine activa- tion of EGFR signaling in eSF endo. In the present study, 8-br- cAMP increased TGFA and AREG expression selectively in eSF endo. While similar observations in endometriosis have not been reported, the interaction of tGFA with EGFR in an autocrine manner has been implicated in the progres- © 2014 Wichtig Publishing - ISSN 2035-9969 14 EGFR inhibition restores decidual markers and aromatase P450 inhibitors to block ovarian estradiol synthesis and/or action on endometrial tissue, prevent retrograde menstruation and provide a less optimal envi- ronment for lesion survival and growth (36). Most of these current medical therapies affect ovulation and endometrial cyclicity, thus having no positive effect of fertility in women with disease, and in fact delay fertility potential. Further - more, no drug that has been used to date has proven to be effective at restoring infertility in subfertile endometriosis patients versus placebo (36). the data presented herein demonstrating that the use of gefitinib restores 8-br-cAMP- induced decidualization marker expression in women with endometriosis indicate that chemical inhibitors of EGFR tyrosine kinase activation warrant further exploration as potential treatments for endometriosis-related infertility. the use of an EGFR inhibitor could be one alternative, as it may not have negative effects on reproductive functions, although this remains undetermined. Summary and conclusions Herein, we have demonstrated the restoration of the de- cidualization markers IGFBP1, PRl and FOXO1A in eSF from women with endometriosis by inhibiting EGFR tyrosine kinase activity with gefitinib. these experiments have also provided insight into the roles of EGFR, mtOR and MAPK signaling in 8-br-cAMP-induced decidualization. Future studies are warranted to investigate the involvement of these pathways in the complex process of decidualization, and evaluate the potential for, and safety of, EGFR inhibitors in the clinical management of endometriosis-related infertility. Financial Support: this project was supported by the Eunice Kennedy Shriver national Institute of Child Health and Human Development/national Institutes of Health (nICHD/nIH) through cooperative agreement U54HD 055764-06 as part of the Special- ized Cooperative Centers Program in Reproduction and Infertility Research (l.C.G.) and F32HD074423 (J.C.C.). Conflict of Interest: none of the authors has any financial interest related to this study to disclose. Address for correspondence: linda C. Giudice, MD, PhD, MSc Distinguished Professor and Chair Department of Obstetrics, Gynecology and Reproductive Sciences University of California – San Francisco 505 Parnassus Avenue M1496 San Francisco, CA 94143-0132, USA [email protected] Summary of observations EGF is known to signal via multiple pathways including ERK1/2 and mtOR via AKt . It is evident that in eSF non-endo 8-br-cAMP activates mtOR, but not ERK1/2 in our model. In eSF endo 8-br-cAMP has no effect on constitutively high levels of p-ERK1/2, but activates mtOR such that it re- mains higher than in eSF non-endo. this differential activation of mtOR correlates with the differential expression of decidu- alization markers in 8-br-cAMP-treated eSF non-endo/eSFendo, and further, parallels changes in eSF non-endo in response to inhibitor treatments. the involvement of EGFR is inferred from the effects of gefitinib, which normalizes decidualiza- tion marker expression in eSFendo and at the same time re- duces mtOR activation to normal eSF non-endo levels, while nonselectively activating p-ERK1/2 in both eSF non-endo and eSFendo. this is consistent with the known involvement of mtOR signaling in protein synthesis versus MAPK path- way in proliferation. Despite these parallel observations, there are inconsistencies in the model. Activation of mtOR by 8-br-cAMP is associated with decidualization marker expression in eSF non-endo, but rapamycin has no effect on decidualization marker expression. In addition, overactive mtOR correlates with no expression of decidualization markers. this suggests that cAMP signaling in eSF non-endo does not depend on mtOR to induce decidualization mark- er expression. While it is not entirely clear from these data how constitutively active EGFR may prevent decidualiza- tion marker expression in eSF endo, inhibition of EGFR tyro- sine kinase with gefitinib restores expression of 3 classic decidualization markers to similar levels to those found in eSFnon-endo. this posits overactive EGFR as key in the ab- errant decidualization response of eSF endo to 8-br-cAMP through dysregulation of downstream pathways. Clinical translation Endometriosis-related infertility is presumably due, in part, to endometrium that is nonreceptive to embryo implanta- tion, as well as poor egg quality and reduced ovarian re- serve due to intrapelvic inflammation. Current therapies have focused on a variety of fertility medications, surgery or assisted reproductive technologies to achieve preg- nancy, which are costly and can result in treatment-related side effects (1). t reatment of endometriosis-related pain has largely relied on the use of drugs such as combined contraceptive steroids, progestogens, GnRH agonists © 2014 Wichtig Publishing - ISSN 2035-9969 15 Erikson et al 15. Klemmt PA, Carver JG, Kennedy SH, Koninckx PR, Mardon HJ. Stromal cells from endometriotic lesions and endometrium from women with endometriosis have re- duced decidualization capacity. Fertil Steril. 2006;85(3): 564-572. 16. Yin X, Pavone ME, lu Z, Wei J, Kim JJ. Increased activation of the PI3K/AKt pathway compromises decidualization of stromal cells from endometriosis. J Clin Endocrinol Metab. 2012;97(1):E35-E43. 17. American Society for Reproductive. Revised American Soci- ety for Reproductive Medicine classification of endometriosis: 1996. Fertil Steril. 1997;67(5):817-821. 18. t ulac S, Overgaard Mt , Hamilton AE, Jumbe nl, Suchanek E, Giudice lC. Dickkopf-1, an inhibitor of Wnt signaling, is regulated by progesterone in human endometrial stromal cells. J Clin Endocrinol Metab. 2006;91(4):1453-1461. 19. Aghajanova l, Velarde MC, Giudice lC. the progesterone re- ceptor coactivator Hic-5 is involved in the pathophysiology of endometriosis. Endocrinology. 2009;150(8):3863- 3870. 20. Yamamoto t , Cui XM, Shuler CF . Role of ERK1/2 signaling during EGF-induced inhibition of palatal fusion. Dev Biol. 2003;260(2):512-521. 21. Gong HC, Wang S, Mayer G, et al. Signatures of drug sensitivity in nonsmall cell lung cancer. Int J Proteomics. 2011;2011:215496. 22. Eum KH, lee M. t argeting the autophagy pathway using ec- topic expression of Beclin 1 in combination with rapamycin in drug-resistant v-Ha-ras-transformed nIH 3t3 cells. Mol Cells. 2011;31(3):231-238. 23. larionov A, Krause A, Miller W. A standard curve based

for relative real time PCR data processing. BMC Bioinformatics. 2005;6(1):62. 24. Chen JC, Frankshun Al, Wiley AA, et al. Milk-borne lacto- crine-acting factors affect gene expression patterns in the developing neonatal porcine uterus. Reproduction. 2011; 141(5):675-683. 25. Sakamoto t , t anaka t , Umesaki n, Ogita S. Epidermal growth factor inhibits 8-Br-cAMP-induced decidualization of human endometrial stromal cells. Horm Res. 2000;53(6): 294-299. 26. Aghajanova l, t atsumi K, Horcajadas JA, et al. Unique tran- scriptome, pathways, and networks in the human endome- trial fibroblast response to progesterone in endometriosis. Biol Reprod. 2011;84(4):801-815. 27. Huang JC, Yeh J. Quantitative analysis of epidermal growth factor receptor gene expression in endometriosis. J Clin En- docrinol Metab. 1994;79(4):1097-1101. 28. Mellor SJ, thomas EJ. the actions of estradiol and epider - mal growth factor in endometrial and endometriotic stroma in vitro. Fertil Steril. 1994;62(3):507-513. 29. Burney RO, t albi S, Hamilton AE, et al. Gene expression analysis of endometrium reveals progesterone resistance and candidate susceptibility genes in women with endome- triosis. Endocrinology. 2007;148(8):3814-3826.

1. Giudice lC. Clinical practice. Endometriosis. n Engl J Med. 2010;362(25):2389-2398. 2. Velarde MC, Aghajanova l, nezhat CR, Giudice lC. In- creased mitogen-activated protein kinase kinase/extra- cellularly regulated kinase activity in human endometrial stromal fibroblasts of women with endometriosis reduces 3′,5′-cyclic adenosine 5′ -monophosphate inhibition of cy- clin D1. Endocrinology. 2009;150(10):4701-4712. 3. Gerbie AB, Merrill JA. Pathology of endometriosis. Clin Ob- stet Gynecol. 1988;31(4):779-786. 4. Papaleo E, Ottolina J, Viganò P , et al Deep pelvic endome- triosis negatively affects ovarian reserve and the number of oocytes retrieved for in vitro fertilization. Acta Obstet Gyne- col Scand. 2011;90(8):878-884. 5. Kitawaki J, Kado n, Ishihara H, Koshiba H, Kitaoka Y , Honjo H. Endometriosis: the pathophysiology as an estrogen- dependent disease. J Steroid Biochem Mol Biol. 2002;83 (1-5): 149-155. 6. Aghajanova l, Velarde MC, Giudice lC. Altered gene expres- sion profiling in endometrium: evidence for progesterone re- sistance. Semin Reprod Med. 2010;28(1):51-58. 7. Alvarez P , Chen X, Hendrich J, et al. Ectopic uterine tissue as a chronic pain generator. neuroscience. 2012;225:269-282. 8. Kim JJ, Yin X. Signaling pathways in endometriosis (eutopic/ ectopic). In: Endometriosis: science and practice. Giudice lC, Evers JlH, Healy Dl, eds. Blackwell, Oxford, UK. 2012: 164-172. 9. Honda H, Barrueto FF , Gogusev J, Im DD, Morin PJ. Serial analysis of gene expression reveals differential expression between endometriosis and normal endometrium. Possible roles for AXl and SHC1 in the pathogenesis of endometrio- sis. Reprod Biol Endocrinol. 2008;6(1):59. 10. Yagyu t, tsuji Y , Haruta S, et al. Activation of mammalian tar- get of rapamycin in postmenopausal ovarian endometriosis. Int J Gynecol Cancer. 2006;16(4):1545-1551. 11. laudanski P , Szamatowicz J, Kowalczuk O, Kuz ´micki M, Grabowicz M, Chyczewski l. Expression of selected tumor suppressor and oncogenes in endometrium of women with endometriosis. Hum Reprod. 2009;24(8):1880-1890. 12. Aghajanova l, Giudice lC. Molecular evidence for differ - ences in endometrium in severe versus mild endometriosis. Reprod Sci. 2011;18(3):229-251. 13. Ejskjaer K, Sorensen BS, Poulsen SS, Mogensen O, Forman A, nexo E. Expression of the epidermal growth factor sys- tem in eutopic endometrium from women with endometrio- sis differs from that in endometrium from healthy women. Gynecol Obstet Invest. 2009;67(2):118-126. 14. Aghajanova l, Horcajadas JA, Weeks Jl, et al. the protein kinase A pathway-regulated transcriptome of endometrial stromal fibroblasts reveals compromised differentiation and persistent proliferative potential in endometriosis. Endocri- nology. 2010;151(3):1341-1355. © 2014 Wichtig Publishing - ISSN 2035-9969 16 EGFR inhibition restores decidual markers lates epidermal growth factor receptor function. J Biol Chem. 2005; 280(52):42694-42700. 34. Ciardiello F , tortora G. Interactions between the epidermal growth factor receptor and type I protein kinase A: biological significance and therapeutic implications. Clin Cancer Res. 1998;4(4):821-828. 35. Blancquaert S, Wang l, Paternot S, et al. cAMP-dependent activation of mammalian target of rapamycin (mtOR) in thy- roid cells. Implication in mitogenesis and activation of CDK4. Mol Endocrinol. 2010;24(7):1453-1468. 36. Evers JlH. Medical therapy of endometriosis: subfertility. In: Endometriosis: science and practice. Giudice lC,Evers JlH, Healy Dl, eds. Blackwell, Oxford, UK. 2013:449-454. 30. t albi S, Hamilton AE, Vo KC, et al. Molecular phenotyp- ing of human endometrium distinguishes menstrual cycle phases and underlying biological processes in normo- ovulatory women. Endocrinology. 2006;147(3):1097-1121. 31. Jeong JW, lee HS, lee KY , et al. Mig-6 modulates uterine steroid hormone responsiveness and exhibits altered ex- pression in endometrial disease. Proc natl Acad Sci USA. 2009;106(21):8677-8682. 32. Jin n, Cho Sn, Raso MG, et al. Mig-6 is required for appro- priate lung development and to ensure normal adult lung homeostasis. Development. 2009;136(19):3347-3356. 33. Xu Y , t an lJ, Grachtchouk V, Voorhees JJ, Fisher GJ. Receptor-type protein-tyrosine phosphatase-kappa regu- View publication statsView publication stats