| Background and Purpose:Endometriosis is defined by the presence of endometrial tissue outside the uterus. It is one of the most common benign gynecologic disorders associated with pelvic pain and infertility. Endometriosis, most commonly involving the ovary, affects approximately10%of women in their reproductive age and up to50%of women suffering infertility and abdominal pain. The patho genesis of endometriosis remains unclear and elusive since it was first described by Von Rokitansky over100years ago. In1920s, Sampson’s retrograde menstruation theory has been widely accepted, but remains controversial as retrograde menstruation occurs in up to90%of women in reproductive age but only6%to10%of those women have endometriosis. Retrograde menstruation may explain occurrence of endometriosis within the pelvis or abdominal cavity but fails to explain the presence of endometriosis in remote sites outside the peritoneal cavity. These controversial points led Iwanoff and Meyer to propose the coelomic metaplasia theory, which explained that endometriosis may derive from mesothelial cells through metaplasia. The presence of endometriosis in remote areas and the rare endometriosis in males support the metaplasia theory. Although this metaplastic theory may explain the majority of endometriosis both in and outside of the ovary and within the peritoneal cavity, it lacks a cellular or molecular basis. Lymphatic and vascular transportation of the endometrium has also been proposed as a complimentary theory to explain rare cases of endometriosis occurring in unusual locations far from the pelvis, but this is unlikely to be the primary mechanism of disease spread. Another compelling proposal suggests that bone marrow derived stem cells may differentiate into endometriotic tissue within and outside of endometrial cavity. However, the stem cell theory is premitive and not completely understood. Overall, no single theory perfectly accounts for the pathogenesis of all cases of endometriosis.The fallopian tube was previously recognized only as a carrier for the menstrual endometrium to pass into the peritoneal cavity or the ovarian surface. The fallopian tube has never been examined as a possible tissue or cellular source of endometriosis. Based on our clinical observations of tubal specimens, we think that the fallopian tube is likely in the certain degree to contribute the formation of ovarian endometriosis. Tubal mucosa is known to be able to form endometrial-like tissue. For instance, endometrialization is commonly seen within the tubal lumen after tubal ligation. It is also known that tubal epithelia shed viable cells onto the ovarian surface forming endosalpingiosis or ovarian epithelial inclusions, a common finding seen within the ovary in approximately30%of the cases. In a recent study onthe cell origin of low-grade serous carcinoma of the ovary, we demonstrated that the majority of the OEIs are derived from tubal epithelia and the OEIs could be transformed into ovarian endometriosis through a probable metaplastic process. We hypothesize that tubal epithelium contributes to the formation of ovarian endometriosis. To examine this provocative hypothesis, we searched for biomarkers which are specific to help differentiate tubal from endometrial origin of endometriosis within the ovary. In this study, we identified a set of novel genes which are either highly expressed in the fallopian tube or in the endometrium through a gene differential array study. We validated these unique genes and their corresponding protein expression in ovarian endometriosis by comparing their expression levels to paired specimens of fallopian tube and endometrium within individual patients.Material and Methods:Tissue samples including human tubal fimbria, paired endometrium and ovarian endometriosis were obtained from surgical pathology specimens within30minutes of resection at the Department of Gynecology, Qilu Hospital of Shandong University from2010to2013. A total of147specimens derived from56patients were studied. Among them, each of35patients with ovarian endometriosis generated a set of samples including fallopian tube, endometrium, and ovarian endometriosis. The remaining non endometriosis patients generated21paired fallopian tube and endometrial samples. All these patients underwent total hysterectomy and bilateral salpingo-oophorectomy for either ovarian endometriosis or benign gynecologic disease without endometriosis. The patients’ age ranged from35to51years with a mean age of41.5. Representative portions of the same tissue specimen were either snap frozen and stored in liquid nitrogen until use or fixed in10%neutral formaldehyde overnight and embedded in paraffin for routine histological examination. Tissue sections containing areas of endometriosis, tubal mucosa and endometrium with both glandular epithelia and stroma were confirmed under microscope and hand microdissected for either real-time PCR or Western analysis. In order to identify tissue specific bio markers, we compared the gene expression between the fallopian tube and the endometrium from patients without evidence of endometriosis by whole-genome microarray analysis. Three pairs of fresh human endometrium and corresponding tubal fimbria specimens were selected from the pool of the21paired samples mentioned above, labeled, and sent to perform whole-genome expression microarray analysis. The criteria for selection of differentially expressed genes were (1) The cut-off value differentially expressed level between the tubal fimbria and the endometrium is more than2-fold;(2) p＜0.05in fold change filtering. Genes that fit these criteria were considered significant for discrimination. The target genes for this study were selected based on the level of expression after comparing with tubal expression:FMO3and DMBT1. To verify the gene expression data obtained from the microarray, real-time PCR was performed on2selected genes, using total RNA from21paired tubal fimbria and corresponding endometrial samples including those three paired specimens for the gene array analysis. Thirty-five cases with ovarian endometriosis and paired tubal and endometrial samples were analyzed by real-time PCR. All samples mentioned above were subsequently evaluated for protein expression of FMO3and DMBT1by Western blot. Immunohistochemistry (IHC) with antibodies to FMO3and DMBT1was performed on the sections of fallopian tubal mucosa, corresponding endometrium and ovarian endometriosis respectively. Results:1. Multiple unique genes identified in the fallopian tube over the endometrium A total of4114and3451genes were identified in the fallopian tube and endometrium, respectively. There were1796genes identified with more than2-fold differential expression between human fallopian tube and endometrial tissues. All these differentially expressed genes were further scrutinized and the highly differentially expressed genes were summarized. Compared with the endometrium, the fallopian tube showed a total911up-regulated genes. These included50-fold or more (n=8),20-fold or more (n=28), and2-fold or more (n=875). Compared with the fallopian tube, the endometrium showed a total of885up-regulated genes including20-fold or more (n=7) and2-fold or more (n=878). There were no genes with more than50-fold up-regulation found in the endometrial tissue.2. FMO3was highly expressed in the fallopian tube, while DMBT1was in the endometrium After screening, we identified2genes, one was highly expressed in the fallopian tube (FM03) and the other in the endometrium (DMBT1), which matched our specified conditions. We validated these two genes with real-time PCR, Western blot and immunohistochemistry. Both the FMO3and DMBT1genes follow the trend differences of the microarray results. In the21pairs of the tubal and corresponding endometrial samples in real-time PCR analysis, FM03was highly expressed in the fallopian tube compared with the paired endometrium, with average fold change of44.38(p＜0.001). In contrast, DMBT1was highly expressed in endometrium compared with the fallopian tube with average fold change of22.11(p＜0.001). None of the21pairs of samples showed discordance with the trend towards co-expression of FMO3and DMBT1in fallopian tube versus endometrial tissue. Among the21pairs of tubal and endometrial samples,14pairs were adequate for Western blot analysis. FMO3protein expression was significantly higher in the fallopian tube samples than that in the endometrium, with an average fold of increment11.05(p=0.006). In contrast, the DMBT1protein level was significantly lower in the fallopian tube (average decreasing fold=32.08) compared with the expression in the endometrium (p＝0.001). These results were compatible with the findings from real-time PCR validation, indicating FMO3and DMBT1do not change significantly at the transcriptional and post-transcriptional levels. All21pairs of the tissue samples were studied for the cellular location of both FM03and DMBT1by immunohistochemistry. Both bio markers were mainly cytoplasmic. No nuclear stainings were identified for these2genes. Of the21paired samples,19(91%) showed moderately to strongly positive staining of FM03in the majority epithelial cells of the fallopian tube, while the remaining2tubal samples were weakly stained. In contrast, FM03was only weakly and focally expressed in the3(14%) endometrial samples, mainly within the endometrial glands. There were some stromal and endothelial cell stainings identified with no specific pattern. DMBT1was strongly and diffusely positive in the majority of glandular cells in all21endometrial samples, but not in the tubal sections we studied.3. FMO3and DMBT1expression in ovarian endometriosis. There were a total of35patients with the requisite three samples for the study. These gene expression levels were compared in patients with ovarian endometriosis and those without ovarian endometriosis from the data presented above. Both FM03and DMBT1expression levels in the fallopian tubal and endometrial samples showed no statistical difference between the patients with ovarian endometriosis and those without ovarian endometriosis. Among the35cases with ovarian endometriosis,32paired samples were adequate for real-time PCR analysis. FMO3was highly expressed in18(56%) of the32samples studied, with the average fold increment of7.21(p=0.016) compared with the endometrium. However, FMO3expression in the remaining14ovarian endometriosis samples was significantly lower than that in fallopian tube (p<0.01), but similar to the level of expression in the corresponding endometrium (p=0.184). In contrast, the18patients with high level of FM03expression, showed a significantly low expression of DMBT1in the ovarian endometriosis lesions compared with the paired endometrial samples (average decreasing fold=6.94, p=0.022). Meanwhile, DMBT1expression showed no statistical differences between the ovarian endometriosis and fallopian tube samples (p=0.144). The18ovarian endometriosis samples with high FMO3and low DMBT1expression also showed concordant level of the protein expression by Western blot. The FM03expression increased with an average fold increment of7.2(p=0.007) compared with the expression in the endometrium. Among the remaining17ovarian endometriosis samples,8showed a similar level of FMO3expression to the endometrium but lower than the tube, while9showed no statistical difference compared to either the endometrium or the fallopian tube. Among the35paired patients,19cases showed that DMBT1protein expression in ovarian endometriosis samples was similarly low in the fallopian tube, while significantly high in the endometrium. DMBT1in the remaining16cases showed high expression in areas of ovarian endometriosis and the endometrium (n=12) and no difference (n=4) between ovarian endometriosis and paired endometrium. Cellular localization of these gene products was examined by immunohistochemistry. Among the35paired samples,19(54%) showed that FMO3was positively staining in the cytoplasm of ovarian endometriosis epithelium, while the remaining16ovarian endometriosis samples showed either low expression in focal areas (n=10) or inadequate expression for analysis due to loss of glandular epithelia within the sections (n=6). In terms of DMBT1,17(49%) ovarian endometriosis samples were negative, while the remaining18ovarian endometriosis samples showed positive staining (n=10) or were inadequate (n=8).Conclusion:Based on our preliminary study, the findings suggest that approximately60%of the ovarian endometriosis we studied may be derived from the fallopian tube, while about40%of the cases may be of endometrial origin. The fallopian tube epithelia may represent one of the tissue sources contributing to ovarian endometriosis. Such novel findings, which require confirmation, may have a significant clinical impact in searching for alternative ways of prevention and treatment of endometriosis. |
PDF Full Text Request