{"paper_id":"089e34e1-ca35-4f36-948d-bcdbc1da8e57","body_text":"Despite advances in diagnostic methods, 15% of infertility causes remain unknown\n( Szczepańska  et al .,\n2011 ). The interaction between the endometrium and the embryo, as well as\nthe receptivity of the endometrium is responsible for approximately 60% of\nimplantation failures ( Riyanti  et\nal ., 2020 ). Human embryonic implantation is a complex\nprocess that requires fine synchronization between the embryo and a receptive\nendometrium, as well as an intricate molecular dialog between the two ( Mahajan, 2015 ). Such events occur in a\nreceptive endometrium stimulated by ovarian steroid hormones, namely, estrogen and\nprogesterone, to provide an anatomical site for embryo implantation ( Koot & Macklon, 2013 ).\nAmong the histomorphological criteria for embryonic implantation to occur,\ndecidualization of the stroma and development of pinopods and microvilli of the\nluminal epithelium are required ( Dunn  et\nal ., 2003 ). At the molecular level, alterations in the gene\nexpression of cytokines and growth factors and the transcription of adhesion\nmolecules, such as  HOX  genes, are involved ( Kalma  et al ., 2009 ;  Paria  et al ., 2002 ).\nThroughout the menstrual cycle, the cells of the uterine endometrium constantly\nundergo proliferation and differentiation processes that are analogous to many\nchanges in embryonic development ( Bagot  et\nal ., 2001 ). Sex steroids dictate the pattern of\n HOX  gene expression in the adult endometrium, playing a key\nrole in embryo implantation ( Taylor, 2000 ).\n HOXA10  and  HOXA11  are expressed in the\nendometrial glands and stroma throughout the menstrual cycle. In the endometrial\nglands, these genes have peak expression during the secretory phase and play a\ncrucial role in the success of implantation. In endometrial stroma, gene expression\nis constant and invariable at all stages of the menstrual cycle ( Taylor, 2000 ;  Sarno  et al ., 2005 ). Women with a high implantation\nrate have high expression of  HOXA10  and  HOXA11  in\nthe secretory phase, suggesting that maternal expression of these genes is essential\nfor successful implantation ( Du & Taylor,\n2015 ).\nChronic endometritis is defined as chronic inflammation of the endometrium. It is\nassociated with an increased prevalence of recurrent pregnancy loss and implantation\nfailure after in vitro fertilization ( Zargar\n et al ., 2020 ;  Goldberg  et al ., 2019 ;  Johnston-MacAnanny  et al ., 2010 ;  Cicinelli  et al ., 2005 ). Chronic endometritis\nlesions are often associated with pelvic inflammatory diseases and intermenstrual\nbleeding ( Smith  et al .,\n2010 ). Pathophysiologically, it is characterized by the presence of plasma\ncells in the endometrium, especially in the stroma ( Nucci & Oliva, 2009 ). Diagnosis is usually based on endometrial\nbiopsy based on hysteroscopy, endometrial stromal plasma cell count and\nimmunohistochemical staining for CD138 ( Kitaya &\nYasuo, 2011 ;  Kasius  et\nal ., 2011 ). The possible relationship between chronic\nendometritis, infertility and perinatal complications has recently been reported in\nthe literature ( Kitaya  et al .,\n2016 ). Studies indicate that even after treatment with antibiotics,\npatients with chronic endometritis continue to have low implantation rates,\nsuggesting that the pathology may cause endometrial anomalies other than those\ngenerated by microorganisms that lead to tissue inflammation ( Johnston-MacAnanny  et al ., 2010 ). The molecular\nmechanisms that may promote implantation failure in these patients have not yet been\nfully elucidated ( Kitaya  et al .,\n2016 ). Therefore, in this study we attempted to demonstrate a possible\nimpact of chronic endometritis on the expression of HOXA10 and HOXA11 in infertile\nwomen.\n\nA prospective, translational cohort study was conducted in partnership with the\nAntônio Pedro University Hospital in Niterói, the Ginendo Clinic in\nRio de Janeiro. Cellular Interactions Laboratory of the Institute of Biomedical\nSciences at UFRJ for immunohistochemical processing. CD138,  HOXA10 \nand HOXA11 after approval was obtained from the Research Ethics Committee of\nFluminense Federal University. The patients were selected and invited to participate\nin the study after showing an indication for hysteroscopy due to infertility in the\ncouple. The invitation was made before the examination without any impairment to the\ntreatment or follow-up of the patient.\nThe material for biopsy was removed, and part of it was targeted for research. All\nparticipants signed informed consent forms. The patients who participated in the\nstudy were divided into three groups: women with endometritis and infertility women;\nwomen without endometritis but with infertility; and women without endometritis and\ninfertility. Infertility was defined as the inability to conceive after 12 months\nfor women under the age of 35 years or after 6 months for women over 35 years of\nage. The division of the endometritis groups occurred after hysteroscopy with biopsy\nand confirmation with positive CD138.\nThe inclusion criteria for the 3 groups were 1) aged between 30 and 40 years; 2)\nperforming hysteroscopy during the same period of the menstrual cycle, between the\n8th and 12th day of the cycle, with the dates of the cycle confirmed by the\npatient’s history; 3) fertile women, defined as women with previous pregnancy and\nchildbirth; and 4) indication for hysteroscopy due to infertility if the woman had\nendometritis and infertility. The exclusion criteria for the endometritis and\ninfertility group were as follows: 1- polycystic ovary syndrome/anovulation; 2-\npremature ovarian failure; 3- hyperprolactinemia; 4- hyper/hypothyroidism; 5- tubal\nobstruction; 6- history of endometriosis; 7- previous surgery to remove cervical\ncancer; 8- endometrial hyper/hypoplasia; 9- uterine synechiae; 10- uterine\nmalformation; 11- signs of vaginosis, such as those caused by  Streptococcus\nagalactiae, Candida  spp.,  Gardenerella vaginalis,\nTrichomonas  spp.,  Mycoplasma genitalium, Chlamydia trachomatis,\nNeisseria gonorrhea , or  Mycoplasma hominis ; 12- BMI\n> 40 kg/m 2 ; 13- previous endometrial ablation; 14- anterior\nembolization of the uterine artery; 15- presence of submucosal myoma or myoma that\ndistorts the cavity on hysteroscopy; 16- presence of type 4 myoma with (intramural)\nmean diameter > 4 cm; 17- presence of polyps; 18- presence of hydrosalpinx on\nhysterosalpingography or ultrasound; 19- suspected clinical signs of gonorrhea or\nchlamydia; or 20- chronic use of glucocorticoids (except nasal preparations).\nTwo samples of approximately 1 mm in size were taken from each patient during the\nprocedure in all groups: one sample was sent for histopathological examination, as\nsuggested by the gynecologist and the other sample was used for the\nimmunohistochemistry procedures of the present study.\nThe samples destined for immunohistochemistry for CD138,  HOXA10  and\n HOXA11  staining were subjected to the following procedure. 1-\nAfter fixation in 10% buffered formalin for a maximum period of 24 hours, the\nfragments were dehydrated and clarified. 2- Each cassette was immersed for a total\nof 45 minutes in 70% alcohol, 80% alcohol, 90% alcohol, 100% absolute alcohol,\nfollowed by Xylene I, Xylyl III. 3- The pieces were then subjected to paraffin\nbaths, namely, paraffin I and paraffin II (cleaner), and kept in an oven at 58°C to\n62°C for 45 minutes each. 4- Then, the segments were included in metal molds with\nliquid paraffin for blocking. 5- The tissues were cut into 5 µm-thick\nsections in a microtome (Semiautomatic Sleeve, NiederOlm), followed by immersion in\na water bath at 50°C with distilled water for distension. Then, they were placed in\nsalinized slides and placed in an oven for 20 minutes between 58 and 62°C.\nThe CD138,  HOXA10  and  HOXA11  immunohistochemical\nprotocol was used, and the reagents were applied with the aid of a pipette. The\nslides were placed in a humidified chamber under the surface to allow the flow of\nthe reagents.\nThe slides were deparaffinized following sequential xylene baths, rehydrated in\ndecreasing concentrations of alcohol and washed for 5 minutes in running water.\nThen, antigenic retrieval was performed by incubating the sections in Trilogy buffer\n(Cell Marque, Rocklin) at 98°C for 20 minutes in a water bath. Then, to block\nnonspecific binding (endogenous peroxidase) the sections were incubated with\n“Peroxidase Block” reagent (Leica, São Paulo) for 5 minutes at room\ntemperature, followed by two washes of 5 minutes each with PBS solution. Soon after,\nthe sections were incubated with primary antibodies against  HOXA10,\nHOXA11  and CD138 (Thermo Fisher, Waltham) diluted 1:25, 1:50 and 1:25,\nrespectively, for 1 hour at room temperature. Then, the slides were washed twice for\n5 minutes with PBS solution.\nThe development reaction for the detection of cells labeled with CD138,\n HOXA10  and  HOXA11  was developed using the\n“Novolink Polymer Detection System” kit (Leica, São Paulo) according to the\nmanufacturer’s recommendations. In the case of CD138, after incubation with the\nprimary antibody, the sections were incubated with the “Post Primary” reagent\n(Leica, São Paulo), equivalent to the secondary antibody for amplification of\nthe first signal in antibodies produced in mice, for 30 minutes at room temperature,\nfollowed by two 5 minutes washes each with PBS.\nFor  HOXA11 , the sections were incubated using “Novolink Polymer”\n(Leica, São Paulo), corresponding to the secondary antibody that binds to\nantibodies produced in rabbits, for 30 minutes at room temperature, followed by two\nmore washes with PBS for 5 minutes each. For  HOXA10 , the sections\nwere incubated using “Novolink Polymer” (Leica, São Paulo), corresponding to\nthe secondary antibody that binds to antibodies produced in goats, for 30 minutes at\nroom temperature, followed by two more 5 minutes washes with PBS.\nThe chromogen substrate used in the development reaction of the 3 markers was\ndiaminobenzidine (DAB) associated with hydrogen peroxide (Leica, São Paulo),\nwith a dilution of 50 mL of DAB in 1mL of dilution buffer, which was added to the\ntissue for 2 to 5 minutes. Positive staining in cells was indicated by brown\nstaining. Counterstaining was performed with hematoxylin (Leica, São Paulo)\nfor 5 minutes.\nFor the diagnosis of chronic endometritis, the samples were selected by hysteroscopy\nand stained with antibodies against CD138 (plasma cell transmembrane protein) and\nthe number of plasma cells was quantified. Samples with five or more plasma cells\nlabeled by immunohistochemistry were considered positive using 20 high-power fields\n(most selective criterion) ( Johnston-MacAnanny\n et al ., 2010 ;  Xu\n et al ., 2020 ;  Kimura\n et al ., 2019 ;  Bouet\n et al ., 2016 ;  McQueen  et al ., 2014 ).\nThe analysis of  HOXA10  and  HOXA11  was performed\nusing the HScore technique of immunohistochemistry with polyclonal antibodies.\nAnti-HOXA10 and anti-HOXA11 primary antibodies conjugated with secondary antibodies\nthat provided the characteristic color of the reaction were used. The technique\nfollowed the recommendations of the manufacturer (Thermo Fisher, Waltham).\nThe HScore technique is a semiquantitative method that correlates the number of\nstained cells and the intensity of intracellular staining at 40X magnification under\na light microscope ( Lessey  et al .,\n1988 ;  Budwit-Novotny  et\nal ., 1986 ). The intensity of staining was classified as\nundetectable (0), weak but detectable (1), distinct (2) or very strong (3) ( Figures 1  and  2 ). The score was calculated with the formula H = Σ P x i, where\nP is the percentage of stained cells that were labeled by the antibody and is the\nintensity score (0, 1, 2 or 3). The slides were analyzed individually by two\ndifferent blinded observers, and the percentage of labeled cells was recorded with\nHScore. Statistical analysis of the HScore was performed by ANOVA with Tukey’s post\nhoc test using a significance level of 0.05.\nFigure 1 Average percentage of HOXA10 expression in the proliferative phase of the\nmenstrual cycle (8 th  to 12 th  day) between the\ngroups: fertile without endometritis, infertile without endometritis and\ninfertile with endometritis. This represents a significant difference\ncompared to the control group (fertile without endometritis)\n( p  value<0.05; ANOVA with Tukey’s post hoc\ntest).\nAverage percentage of HOXA10 expression in the proliferative phase of the\nmenstrual cycle (8 th  to 12 th  day) between the\ngroups: fertile without endometritis, infertile without endometritis and\ninfertile with endometritis. This represents a significant difference\ncompared to the control group (fertile without endometritis)\n( p  value<0.05; ANOVA with Tukey’s post hoc\ntest).\nFigure 2 Average percentage of HOXA11 expression in the proliferative phase of the\nmenstrual cycle (8 th  to 12 th  day) between the\ngroups: fertile without endometritis, infertile without endometritis and\ninfertile with endometritis. This represents a significant difference\ncompared to the control group (fertile without endometritis).\n( p  value<0.05; ANOVA test with Tukey’s post-hoc\ntest).\nAverage percentage of HOXA11 expression in the proliferative phase of the\nmenstrual cycle (8 th  to 12 th  day) between the\ngroups: fertile without endometritis, infertile without endometritis and\ninfertile with endometritis. This represents a significant difference\ncompared to the control group (fertile without endometritis).\n( p  value<0.05; ANOVA test with Tukey’s post-hoc\ntest).\n\nThe sociodemographic data regarding age, time to pregnancy and number of children\nwere compared to assess the homogeneity of the sample. For this purpose, normality\nwas assessed using the Kolmogorov-Smirnov test, and if the data followed a normal\ndistribution, the comparison was performed using ANOVA with Tukey’s post hoc test.\nOtherwise, the Kruskal-Wallis test with Dunn’s post hoc test was used. All data were\nevaluated using SPSS statistical software, version 20.0, and the significance level\nused was 0.05.\nThe groups were homogeneous in their initial characteristics in relation to their\nages ( Table 1 ). However, the time to\npregnancy was longer in infertile group than in the control group, and the number of\nchildren was lower in the infertile group than in the control group. It is worth\nnoting that all patients diagnosed with EC had CD138 positivity in more than five\ncells at 20x magnification.\nDemographic data of the fertile women without endometritis, infertile women\nwithout endometritis and infertile women with endometritis.\nThe demographic data were taken from questionnaires answered by the patients before\nthe hysteroscopy. The data was measured in years; the time for pregnancy in the\nfertile group was smaller than 1 year.\nSome of the patients distributed in the infertile groups had already had children\nbefore the present study. This number of children was put on the table as\ndemographic data, at the time of the study they were classified as infertile\n(secondary infertility) since they were attempting to conceive for more than 12\nmonths with no success being younger than 35 years old or 6 months in women being\nolder than 35 years old.\nFigure 1  shows the comparison of de\n HOXA10  averages according to the evaluated groups in the\nproliferative phase of the menstrual cycle (8th to 12th day). There was a\nsignificant difference between the groups, with infertile women with endometritis\nhaving lower  HOXA10  expression values than the women in the control\ngroup (fertile without endometritis). This was also observed in the infertile women\nwithout endometritis compared to the control group ( p  value=0.003;\nfertile without endometritis: 116.75±46.31; infertile without endometritis:\n71.32±41.04; infertile with endometritis: 54.00±27.03). Although there\nwas no significant difference between infertile women without endometritis and\ninfertile women with endometritis, a reduction in  HOXA10  expression\nwas observed in infertile women with endometritis (-24.3%).\nThe comparison of the means of  HOXA11  according to the evaluated\ngroups is shown in  Figure 2 . There was a\nsignificant difference between the groups, with infertile women with endometritis\nhaving lower  HOXA11  expression values than women in the control\ngroup (fertile without endometritis), as was observed for infertile women without\nendometritis compared to the control group ( p  value=0.001; fertile\nwithout endometritis: 212.80±32.99; infertile without endometritis:\n159.32±36.16; infertile with endometritis: 137.25±55.87). Although\nthere was no significant difference between infertile women without endometritis and\ninfertile women with endometritis, a reduction in  HOXA11  expression\nwas observed in infertile women with endometritis (-13.8%).\n\nBased on the HScore analysis by immunohistochemistry, it was possible to observe a\nreduction in the expression of  HOXA10  and  HOXA11 \nin infertile women compared to women in the control group (fertile without\nendometritis). These data have already been investigated by  Taylor  et al ., 1998 , and according to these\nauthors, a few markers are indispensable for implantation, with\n HOXA10  being the most important. In this sense, the low\nexpression of  HOXA10  and  HOXA11  leads to abnormal\nuterine development and impairment of implantation. A difference between that study\nand the present study is that for ethical reasons, endometrial biopsy was performed\nin the proliferative phase of the menstrual cycle, unlike previous studies in which\nthis was performed in the secretory phase where  HOXA10  could\nprobably also be expressed.\nIn contrast to what was found in the present study and that of  Taylor  et al ., 1998 , the study conducted by\n Szczepańska  et al ., 2011 \nin patients with idiopathic infertility, showed that there was no difference between\nthe level of HOXA10 or HOXA11 expression between the infertility and control groups\n( Szczepańska  et al .,\n2011 ).\nThere was no significant difference in the expression of  HOXA10  and\n HOXA11  between infertile patients with endometritis and those\nwithout endometritis ( Marin  et al .,\n2022 ). In the present study, a decrease in the expression of both genes\nwas observed, but this result was not significant. This may have occurred due to the\nnumber of patients evaluated, the use of a single analytical method\n(immunohistochemistry) or the period of the cycle in which the endometrial\ncollection was performed. However, it is worth noting that the power of the test was\n88% and certifies the significance of the study result.\nA recent study with a similar sample size (n=13 with chronic endometritis and n=16\nwithout chronic endometritis) that evaluated the expression of\n HOXA10  and  HOXA11  in patients with and without\nendometritis also found similar data, namely, that there was no significant\ndifference between groups ( Marin  et\nal ., 2022 ). There is a significant difference between the\npresent study and that performed by ( Marin\n et al ., 2022 ), because in the present study, a\ncontrol group of fertile patients without endometritis was created, that is, a\nnegative control that, from the methodological point of view, reduces the chances of\nerror. In the current study, immunohistochemistry and semiquantitative analysis were\nperformed using the HScore to determine the percentage of  HOXA10 \nand  HOXA11  expression.\nOur analysis was at the protein level, and we did not have an evaluation methodology\nat the molecular level, for example quantitative techniques, such as polymerase\nchain reaction (PCR). If this second technique was used, which was not possible for\nstructural reasons, we might also be able to find a difference in relation to our\nstudy. However ( Marin  et al .,\n2022 ), although they used PCR, they also found no significant difference,\nshowing that eventually a new molecular level of analysis would also corroborate our\nresults. Due to issues imposed by the Ethics Committee so that there was no harm to\nthe patients who participated in the study, the period of hysteroscopy performance\nwas in the late follicular phase, which may have impacted the HScore evaluations\nsince the best result would be in the secretory phase.\nIn normal fertile women, the expression of HOXA10 and HOXA11 significantly increases\nduring the luteal phase, with the peak occurring during the implantation window, and\nremains at an elevated level until the end of the cycle ( Du & Taylor, 2015 ). Other studies concluded that this\ndifference in expression during the menstrual cycle occurs only in the region of the\nendometrial glandular epithelium and not in the stromal compartment ( Sarno  et al ., 2005 ). Taking\nthis into account, this phase of the menstrual cycle was used for the study, but it\nis worth noting that the ideal would be in the secretory phase, and this may have\nbeen a factor influencing the result. Likewise, there is a study by  Marin  et al ., 2022 , in which\nthe authors found no significant difference even when performing the biopsy in the\nsecretory phase of the menstrual cycle.\n\nThe expression of  HOXA10  and  HOXA11  during the\nproliferative phase is not significantly different between infertile women with\nendometritis and infertile women without endometritis. Translational studies with a\nlarger number of patients should be performed.","source_license":"CC0","license_restricted":false}