Altered endometrial expression of endothelial nitric oxide synthase in women with unexplained recurrent miscarriage and infertility.

The demographics of the three groups of women studied are shown in Table 1 . Women in the UI group were significantly younger than the controls. There were no differences in BMI among the three groups. Four subjects in the control group, and two in the RM and UI groups were light smokers (<3 cigarettes/day). None of the subjects had chronic medical illnesses. Laparoscopic evaluation in all three groups showed a normal pelvis. As previously demonstrated eNOS was expressed in all layers of the endometrium (Fg.1). Representative sections demonstrating eNOS protein staining in glandular (panels A–C) and luminal epithelium (panels D–F) (most intense staining) for the three studied groups are shown in this figure. Weak staining of eNOS was detected in the stromal layer (panel H), and vascular endothelium (panel G). eNOS expression was significantly higher in glandular epithelium of patients with RM (panel B) and UI (panel C) compared with control (panel A). A similar pattern was found in the luminal epithelium with greater expression of eNOS in patients with RM (panel E) and UI (panel F) compared with controls (Panel D). Negative and positive controls for eNOS are shown in panels I and J respectively. A summary of image analysis for eNOS expression in different layers of the endometrium are shown in Fig. 2 . In women with recurrent miscarriage glandular (P=0.004) and luminal epithelial (P=0.002) expression of eNOS are greater compared with controls, whereas stromal (P=0.14) and vascular endothelial expression of eNOS was not significantly different in RM and UI women as compared with controls. In women with unexplained infertility the expression of eNOS was only higher in luminal epithelium compared with controls with no significant differences in other layers. The expression of eNOS mRNA in the endometrium of women with RM and UI was significantly (P<0.05) higher compared with controls ( Fig. 3 ) confirming the protein data.

Our data confirms our prior reports on endometrial eNOS immunolocalization ( Khorram et al. 1999 ), and demonstrates a differential site specific alteration in expression of eNOS protein in women with UI and RM compared with controls during the window of implantation. The most prominent changes were found in the luminal epithelium with greater expression of eNOS protein in both the RM and UI groups in this area. The expression of eNOS mRNA in ednometirum of women with UI and RM was significantly higher compared with controls thus confirming the protein data. The luminal eNOS protein changes in both RM and UI groups support the importance of this enzyme for implantation and its dysregulation as a possible cause of implantation failures in these patients. The significance of NO in the implantation process has been demonstrated in animal studies in which pharmacological blockers of NOS impair implantation. Novaro et al. demonstrated that that the expression of NOS and PGE and PGF increase 1 day before implantation and suppression of NOS by L-NAME decreased production of PGE and PGF2 on the day of implantation in rats ( Novaro et al. 1996 ). Biswas et al ( Biswas et al. 1998 ) injected L-NAME into the rat uterine horn and demonstrated an inhibition of implantation, and similar results were obtained by Duran-Reyes et al using a different NOS inhibitor ( Duran-Reyes et al. 1999 ). In contrast to these studies knockout mice for various isoforms of NOS do not show reduced litter size ( Huang et al. 1993 ; Huang et al. 1995 ; MacMicking et al. 1995 ), indicating the importance of multiple pathways and redundancy in the implantation process. Since implantation failure secondary to endometrial factors may be common to UI ( Koot et al. 2011 ) and RM ( Li et al. 2002b ) we sought to determine if aberrant endometrial expression of eNOS could be associated with these two conditions. The common finding of luminal increase in the expression of eNOS protein in both RM and UI patients suggest that luminal eNOS is important in the pathogenesis of these disorders. Excess NO could impair implantation through several mechanisms. NO has been shown to induce endometrial epithelial apoptosis ( Castro et al. 2002 ; Johnson et al. 2004 ; Li et al. 2001 ), and increased eNOS expression and therefore localized excess NO production at the luminal surface could induce epithelial apoptosis and implantation failure. Vatansever et al also reported increased eNOS immunoreactivity in the endometrium of unexplained infertility patients, although this was associated with lower number of apoptotic cells ( Vatansever et al. 2005 ). A second mechanism by which NO might impair implantation is through localized nitrosative stress. Nitric oxide by virtue of its unpaired electron is a highly reactive free radical which in excess can damage protein, carbohydrates, nucleotides and lipids ( Agarwal et al. 2008 ). Based on our data we propose that excess eNOS expression in luminal epithelium of patients with RM and UI can create local oxidative stress which could impair implantation, similar to other inflammatory gynecologic conditions such as endometriosis ( Dong et al. 2001 ; Khorram and Lessey 2002 ; Ota et al. 1998 ; Wu et al. 2003 ), adenomyosis ( Ota et al. 1998 ), and adhesions ( Saed and Diamond 2004 ). Exogenous factors such as cigarette smoking which has been associated with recurrent miscarriages ( Cramer and Wise 2000 ) can also induce nitrosative stress through direct endometrial cell stimulation of eNOS expression, an effect which can be blocked by anti-oxidants such as ascorbic acid ( Khorram et al. 2010 ). Similarly ascorbate has recently been shown to activate eNOS activity by rapid modulation of its phosphorylation status ( Ladurner et al., 2012 ). Several studies have attempted to find a genetic link for RM, and in so doing have examined polymorphmism of eNOS gene in different ethnic groups A recent meta- analysis of these studies showed a significant association in eNOS (Glu298Asp) polymorphism and RM ( Su et al. 2011 ). These genetic abnormalities could lead to reduced NO production and impaired endometrial function. Our data did not show any differences in endothelial microvascular eNOS levels in RM and UI patients, suggesting that localized endometrial blood flow mediated by the NO pathway during the implantation window is not a significant factor in the pathogenesis of these disorders, and other blood flow regulating factors not examined in this study may be of greater importance. Although weak expression of eNOS in vascular endothelium as seen in our endometrial samples could be a factor which can lead to infertility and miscarriage, these expression levels were not significantly different compared to the control group. This suggests that that dysregulated expression of eNOS in non-endothelial sites plays a more a significant role in the pathophysiology of recurrent miscarriages and unexplained infertility. In conclusion our data demonstrates an over expression of eNOS protein and mRNA in the endometrium of RM and UI. This aberrant pattern of eNOS protein and mRNA expression is similar to other inflammatory gynecologic conditions such as endometriosis and adenomyosis. Although eNOS expression in some amounts is essential for implantation, excess eNOS expression and therefore excess generation of NO in the endometrium of patients of with RM and UI is deleterious and could induce nitrosative stress which could lead to implantation failures or failure of early pregnancy maintenance.

Recurrent miscarriage (RM) is described as at least three consecutive spontaneous pregnancy losses in the first trimester and compromises 3% of couples ( Regan and Rai 2000 ; Taylor 2003 ). Genetic and developmental abnormalities are the major causes of RM, however immune, endocrine, and endometrial anatomical or non-anatomical factors as well as thrombophilias can lead to RM. In about 50% of cases, the cause of miscarriage remains unknown ( Li et al. 2002a ; Li et al. 2002b ). Unexplained infertility (UI) affects 15% of infertile couples and is defined as when all the tests of a basic infertility evaluation including semen analysis, hysterosalpingogram, ovarian reserve testing, pelvic ultrasound, and possibly laparoscopic evaluation of the pelvis are within normal limits ( Hatasaka 2011 ; Smith et al. 2003 ). Recent data has shown that UI and RM are distinctly different diagnoses and UI is not due to recurrent pre-clinical pregnancy loss ( Koot et al. 2011 ), although these two conditions might share common endometrial function defects. Nitric oxide (NO) is a vasodilator synthesized from L-arginine through the action of nitric oxide synthase (NOS). The three isoforms of NOS which catalyze the formation of NO are expressed in the human endometrium although eNOS is the predominant form ( Khorram et al. 1999 ). The endometrial expression of eNOS is cyclic with peak expression during the window of implantation in humans ( Khorram et al. 1999 ; Ota et al. 1998 ), and rodents ( Purcell et al. 1999 ). Both estrogen and progesterone ( Han et al. 2005 ; Khorram and Han 2009 ; Zervou et al. 1999 ) regulate the expression of eNOS in the human endometrium ( Khorram et al. 1999 ). Nitric oxide by virtue of its properties as a potent vasodilator ( Palmer et al. 1987 ), a myometrial smooth muscle relaxant ( Buxton 2004 ; Norman et al. 1997 ), and its participation in signal transduction pathways ( Thomas et al. 2008 ) might play a significant role in establishment and maintenance of pregnancy. By virtue of these properties of NO we postulated that aberrant endometrial expression of eNOS as in endometriosis ( Dong et al. 2001 ; Khorram and Lessey 2002 ; Ota et al. 1998 ; Wu et al. 2003 ) and adenomyosis ( Ota et al. 1998 ) could occur in patients with UI and RM. Since oxidative stress plays an important role at least in idiopathic recurrent pregnancy loss ( Gupta et al. 2007 ), and NO in high concentrations can induce nitrosative stress ( Agarwal et al. 2008 ) we hypothesized that increased endometrial eNOS expression and thereby NO generation in patients with UI and RM could impair endometrial function by either inducing cellular apoptosis ( Wang et al. 2010 ), or through nitrosylation of key endometrial proteins ( Gu et al. 2010 ; Weiner et al. 2009 ) impair their physiological function.

The protocol for this study was approved by the Human Subjects Committee at Shahid Beheshti Medical University. Endometrial biopsies were obtained from 3 groups of women using a pipelle curette 7–9 days post ovulation as determined by serial ultrasound scans. The RM group (N=10) consisted of women with a mean age of 32.8 with a mean of 4.7 consecutive pregnancy losses. Women with secondary miscarriages or less than three miscarriages were excluded. Evaluation of RM group including karyotype analysis, antiphospholipid antibody and thrombophilia testing were all within normal range. Endometrial cavity as assessed by hysterosalpingogram was normal in RM patients. Women with unexplained infertility (N=10) consisted of individuals with a mean age of 29.8 years who were unable to conceive for more than 2 years with a normal basic infertility evaluation. This evaluation consisted of endocrine tests (TSH, cycle day 3 FSH and estradiol levels, prolactin, progesterone levels greater than 10ng/ml in mid luteal phase); anatomical tests (hysterosalpingography and pelvic ultrasonography), and semen analysis (WHO criteria). The control group (N=10) consisted of women with a mean age of 36.1 years who presented for tubal sterilization. Women in this group had normal menstrual cycles (26–33 days), had a mean parity of 1.4 and had no prior history of pregnancy losses and no prior use of assisted reproductive techniques for conception. Endometrial biopsy specimens were divided into 3 portions; one piece was placed in 4% paraformaldhyde tissue fixative for 24h and then switched to 70% ethanol for later processing. Another piece was placed in RNA Later preservative and stored at −80°C, and one section was fixed for histological dating of the endometrium using the criteria of Noyes et al. ( Noyes et al. 1975 ). Human placental tissue was used as a positive control ( Bhuiyan et al. 2006 ). Endometrial specimens were cut into 6μm sections using a Cryocut and placed on poly l-lysin coated slides and stored at −70°C for immunostaining. A monoclonal mouse antihuman antibody [6H2] (ABcam Company, UK, Cat.NO:91205) was used to detect the intensity and distribution of eNOS immunostaining using standard immunohistochemical protocol. Briefly after rinsing the slides with buffer (0.1 M of phosphate-buffered saline with a pH of 7.4), endogenous peroxidases were quenched by incubation in 0.3% H 2 O 2 in methanol for 15 minutes. Repeated rinses with 0.05% bovine serum albumin in phosphate-buffered saline were performed followed by antigen retrieval using trypsin. 1.5% normal goat serum (DAKO Co., Denmark) was then added to the slides in humidified chambers for 20 minutes at room temperature to prevent nonspecific binding of antibody. The primary antibody against eNOS (1:100) was added to slides and incubated at 37°C for 1h. Slides were then rinsed three times with wash buffer, followed by incubation of sections with the secondary antibody, a rabbit antimouse IgG H&L (ABcam Company, Cat.NO: ab6728, UK) diluted 1:1000 in PBS. Incubation with the secondary antibody was performed for 1h at 37°C in an incubator. Slides were then exposed to 3, 3-diaminobenzidine in H 2 O 2 (DAKO co, Denmark) was for 15 minutes. Thereafter, the sections were counterstained with hematoxylin and mounted. In case of negative controls similar method was used but phosphate-buffered saline replaced the primary antibody. Human full term placental tissue served as an external positive control. Staining intensity of sections was determined by Image Pro Plus software by a blinded reviewer (TN) using previously described methods ( Khorram et al. 2007 ). 6 different areas of the sections were analyzed at a magnification of 40x and the mean was used for statistical analysis. The results were expressed as % integrated Optical Density (%IOD). RNA was isolated using the High Pure RNA Isolation Kit (Roche applied science). Ribonucleic acid was DNase treated and quantitated by measurement of absorbance in a NanoDrop spectrophotometer. One microgram of total RNA was reverse transcribed into single stranded complementary DNA (cDNA) with use of the Omniscript Reverse Transcription kit (GeneON, Germany) at 37°C for 60 minutes in a total volume of 20 uL. The polymerase chain reaction (PCR) mix consisted of 1 mL of 10-fold diluted cDNA, qPCR MasterMix Plus for SYBR green I reagent (GeneON, Germany) and optimized forward and reverse gene-specific primers (300 nmol/L each). Real-time PCR reactions in triplicate were run in 96-well plates with use of an Mx3000P real-time PCR system (Stratagene, Santa Clara, CA). Reactions were started by activation of DNA polymerase at 95°C for 10 minutes followed by 40 PCR cycles of denaturing at 95°C for 15 seconds and annealing/extension at 60°C for 1 minute. The internal control used was human HPRT ( de Kok et al. 2005 ). Data was analyzed to select a threshold level of fluorescence that was in the linear phase of the PCR product accumulation (the threshold cycle [CT]) for that reaction. The CT value for HPRT was subtracted from the CT value of eNOS gene to obtain a delta CT (DCT) value. The relative fold change for each gene was calculated with use of the ΔΔCT method ( Livak and Schmittgen 2001 ). Results were analyzed by ANOVA comparing %IOD for the imunohistochemical data and fold change in case of Real time RT-PCR using SPSS software. Post hoc analysis was done using the Student- Newman- Keuls test. P<.05 was considered statistically significant.