{"paper_id":"23daa98a-c9af-4bf1-98ff-b87f4979df34","body_text":"1 \n \nReduced frequency of perforin-positive CD8+ T cells in menstrual \neffluent of endometriosis patients compared to healthy controls \nTimo Schmitz1,2, Verena Hoffmann1, Elisabeth Olliges1,3, Alina Bobinger1,3, Roxana \nPopovici4,5, Elfriede Nößner6*, Karin Meissner1,3* \n1 Institute of Medical Psychology, Medical Faculty, LMU Munich  \n2 Chair of Epidemiology, Medical Faculty, LMU Munich at UNIKA-T Augsburg \n3 Division of Health Promotion, Coburg University of applied Sciences, Coburg, Germany \n4 kïz), Munich, Germany \n5 Department of Gynecologic Endocrinology and Fertility Disorders, Heidelberg University \nWomen’s Hospital, Heidelberg, Germany \n6Immunoanalytics Research Group Tissue control of immunocytes, Helmholtz Zentrum \nMünchen, Munich, Germany \n \n* authors contributed equally \n \nCorresponding authors: \nTimo Schmitz, MD, email: t.schmitz3@gmx.de\n  \nProf. Dr. Karin Meissner, MD. M.Sc., email: karin.meissner@hs-coburg.de  \n \n  \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \nNOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice.\n\n2 \n \nAbstract \nBackground: Endometriosis is widespread among women in reproductive age and quite \ncommonly reduces life quality of those affected by symptoms like dysmenorrhea, \ndyspareunia or infertility. The scientific literature indicates many immunological changes like \nreduced cytotoxicity of natural killer cells or altered concentrations of cytokines and cell \nadhesion molecules. Frequently examined tissues are peripheral blood, endometrial tissue \nand peritoneal fluid. Yet, knowledge on immunological differences in menstrual effluent (ME) \nis scarce.  \nMethods: 12 women with endometriosis and 11 healthy controls were included in this study. \nME was collected using menstrual cups and venous blood samples (PB) were taken. \nMononuclear cells were obtained from ME (MMC) and PB (PBMC) and analyzed using flow \ncytometry. Furthermore, concentrations of cell adhesion molecules (ICAM-I and VCAM-I) and \ncytokines (IL-6, IL-8 and TNF-α ) were measured in ME and PB. \nResults: CD8+ T cells obtained from ME were significantly less often perforin-positive in \nwomen with endometriosis compared to healthy controls. Additionally, plasma ICAM-I \nconcentrations were significantly lower in the endometriosis group. A comparison between \nMMC and PBMC revealed that MMC contained significantly less T cells and more B cells. \nThe CD4/CD8 ratio was significantly higher in MMC, and Tregs were significantly less \nfrequently in MMC. In ME, T cells and NK cells expressed significantly more CD69. NK cells \nobtained from ME were predominantly CD56bright/CD16dim and had a lower frequency of \nperforin+ cells compared to PBMC NK cells. NKp46 was significantly more expressed on NK \ncells from PBMC.  \nConclusion: CD8+ T cells obtained from the ME were significantly less perforin-positive in \nendometriosis patients indicating a reduced cytotoxic potential. MMC are distinctively \ndifferent from PBMC and, thus, seem to be of endometrial origin.  \n \nKeywords: Endometriosis, Menstrual effluent, T cells, Perforin \n \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n3 \n \nIntroduction \nEndometriosis is a common disease that affects up to 10% of the female population in the \nreproductive age [1]. It is characterized by the appearance of ectopic endometrial tissue \npredominantly in the lower abdomen. Frequently affected structures are the uterus, \nperitoneum, ovary, tube, rectum and bladder. Conduction symptoms are dysmenorrhea, \ndeep dyspareunia, dyschezia and chronic abdominopelvic pain; subfertility is widespread \namong women with endometriosis [2]. The diagnosis of the disease can be challenging. If \nendometriosis is suspected, in many cases an abdominal laparoscopy is performed, which is \nan invasive method that allows assessment of peritoneal infestation with the possibility to \ntake biopsies to histologically confirm the diagnosis [2]. Nevertheless, late diagnosis is very \ncommon and 10 years or more may pass from the onset of symptoms to confirmed diagnosis \n[3]. Treatment includes amongst others hormonal therapy, pain medication and surgery. \nToday’s most accepted theory on the pathogenesis of endometriosis is that of retrograde \nmenstruation [4]. It postulates that vital endometrial cells get into the tubes, the peritoneal \ncavity and other structures in the lower abdomen via retrograde menstruation. Some of those \nendometrial cells attach and grow to form herds [5]. Since retrograde menstruation affects a \nmajority of women but only few develop endometriosis [6], the question arises as to which \nmechanisms are responsible for attachment and growth of endometrial cells outside of the \nendometrium. Hormonal and immunological abnormalities in women suffering from \nendometriosis have been intensely investigated. One hypothesis suggests that immune cells \nof women with endometriosis are incapable of clearing those vital endometrial cells \ndislocated by retrograde menstruation due to reduced cytotoxicity [7]. Prior studies examined \nimmune cells isolated from endometriotic lesions, endometrial tissue, peritoneal fluid, \nperipheral blood and other tissues and found alteration in the number and function of \nimmune cells. Differences were found in the number of naive natural killer (NK) cells, \ncytotoxic T cells, regulatory T cells (Tregs) and other lymphocyte subsets in women with \nendometriosis compared to healthy individuals [7–9]. Additionally, cytokines like interleukin-6 \n(IL-6), interleukin-8 (IL-8) or tumor necrosis factor alpha (TNF-\nα ) and cell adhesion \nmolecules, like ICAM-1 and VCAM-1, are suspected to play a potential role in the \npathogenesis of endometriosis and altered concentrations have been observed in women \nwith endometriosis [10–13]. Nevertheless, only limited data is available on specific \ndifferences in mononuclear cells obtained from menstrual effluent (ME) or levels of cytokines \nand cell adhesion molecules in ME. Samples of ME can be collected with the help of so \ncalled menstrual cups. The aim of this study was to investigate menstrual effluent \nmononuclear cells (MMC) and peripheral blood mononuclear cells (PBMC) for potential \ndifferences in women with endometriosis and healthy controls to gain a deeper \nunderstanding of the underlying pathophysiologic mechanisms of endometriosis. A further \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n4 \n \naim was to compare MMC and PBMC, since it is not yet clear whether MMC originate \npredominately from PB or mainly from endometrial tissue. A third question was whether ME \ncould be useful for non-invasive diagnostic for endometriosis to reduce the delay in \ndiagnosing the disease. \nMaterials and Methods \nStudy population \nBetween January 2018 and August 2019, 12 women with histologically confirmed \nendometriosis and 11 age matched healthy volunteers without endometriosis or menstrual \npain were recruited for this study. Inclusion criteria were the following: age between 18 and \n45 years, regular menstrual cycle and sufficient knowledge of German language. Further \ninclusion criteria for the patients with endometriosis was biopsy-confirmed endometriosis. An \nadditional criterion was maximum pelvic pain during the last three menstrual cycles \n(dysmenorrhea) assed by using a numeric rating scale (0 = no pain, 10 = worst pain \nimaginable). Inclusion criterion was 5 or more for patients with endometriosis, and 3 or less \nfor healthy controls. All participants underwent gynecological examination in order to confirm \n(endometriosis group) or exclude (healthy controls) the presence of endometriosis. Exclusion \ncriteria for both groups were a manifest mental illness, a malignant disease, the acute need \nof treatment of a gynecological disease, and medication with hormonal drugs. All study \nparticipants gave written informed consent. The study protocol was approved by the ethical \ncommittee of the Medical Faculty at LMU Munich (no. 17-695) and the study was performed \nin accordance with the Declaration of Helsinki.  \nProcedure \nStudy participants collected menstrual blood using a menstrual cup (Mooncup®, Mooncup \nLtd, Brighton, UK). The collection started 12 hours after the beginning of the menstruation \nand lasted for 24 hours. The collection was divided into two 12-hour cycles, after each cycle \nthe samples were decanted from the menstrual cup into a tube. The tube contained 10 ml of \nthe following medium: RPMI 1640 medium, pyruvate (1 mM), glutamax (2 mM), penicillin \n(100 U/ml), streptomycin (100 \nμ g/ml), 10% human pooled serum (HPS), and 0.3% sodium \ncitrate. The samples were stored by the participants at room temperature. After the \ncollection, study participants came to the institution to hand over the samples and peripheral \nblood (PB) was drawn. Menstrual blood was filtered using cell strainers (70 and 40 \nmicrometer by OMNILAB-LABORZENTRUM GmbH & Co.Kg, Munich, Germany) in order to \nremove bigger accumulation of endometrial tissue.  \n  \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n5 \n \nIsolation of MMC and PBMC and Flow cytometry  \nThe mononuclear cells of ME and PB were isolated using density gradient centrifugation. \nAfter isolation, the vital cells were frozen using dimethylsulphoxide (DSMO) and stored in the \nvapor face of liquid nitrogen. Less vital cells were obtained from the ME samples than from \nPB. Nevertheless, the number and quality of the mononuclear cells was sufficient for flow \ncytometry measurement. After sample collection was completed, MMC and PBMC were \ndefrosted and subjected to antibody staining and flow cytometry. Each experimental run \nincluded samples of healthy women and endometriosis patients, and MMC and PBMC of the \nsame person were stained in parallel in the same experiment to facilitate comparison. Next to \nLIVE/DEAD™ Fixable Blue Dead Cell Stain (Thermofischer, Waltham, Massachusetts, USA, \nCatalog-Nr. L23105), the flow cytometry was performed using two panels with the following \nantibodies (BioLegend, San Diego, California, United States): Panel 1: CD19-A700 (Catalog-\nNr. \n302226), CD20-A700 (Catalog-Nr. 302322), CD3-PerCP (Catalog-Nr. 300326), CD14-\nPacificBlue (Catalog-Nr. 301828), CD56-PE-Cy7 (Catalog-Nr. 362510), CD16-Alexa Fluor® \n488 (Catalog-Nr. 302019), CD335-APC (NKp46) (Catalog-Nr. 331918) and CD69-PE \n(Catalog-Nr. 310906). Panel 2: CD56-PE-Cy7 (Catalog-Nr. 362510), CD3-PerCP (Catalog-\nNr. 300326), CD14-APC-Cy7 (Catalog-Nr. 301820), CD4-APC (Catalog-Nr. 344614), CD25-\nPE (Catalog-Nr. 356104), FoxP3-Alexa Fluor® 488 (Catalog-Nr. 320012), Perforin-\nPacificBlue (Catalog-Nr. 308118) and CD8-BUV496 (Becton Dickinson, Franklin Lakes, New \nJersey, United States, Catalog-Nr. 564804). Isotype controls were used: Panel 1: IgG1-Alexa \nFluor® 488, IgG1-APC, IgG1-PE, IgG1-PE-Cy7, IgG2a-PerCP; Panel 2: IgG1-APC, IgG1-\nPE, IgG1-PE-Cy7, IgG2a-PerCP. Staining of panel 1 was surface only, panel 2 staining \ninvolved fixation and permeabilization using FOXP3 buffer (BioLegend Transcription Factor \nBuffer Set True Nuclear TM Fix). Optimal antibody concentrations were identified by serial \ndilutions. After staining, data were acquired using the CytoFlex cytometer with Beckman \nCoulter Cytexpert 2.3 software and analyzed using FlowJo version 10.6.2 (Treestar). Isotype \ncontrols were used for gate setting. \nGating strategy \nFigure 1 displays the gating strategy to identify the main cell types. In the first step, cell \ndebris was excluded and the mononuclear cells were chosen using FSC (forward-scatter) \nand SSC (side-scatter). LIVE/DEAD™ Fixable Blue Dead Cell staining identified the live \ncells, followed by doublet exclusion using FSC-H/FSC-A-dot plots. B cells were identified by \nCD19/CD20 antibody staining. All other mononuclear non-B cells were further discriminated \nbased on CD3 versus CD14 staining, identifying CD3+ T cells and CD3-negative non-T cells. \nThe T cells were then divided into CD8+ T cell and CD4+ T cells. Among the non-T cells, NK \ncells (CD56+) and monocytes/macrophages (CD14+) were identified. These main cell types \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\nwere further analyzed regarding the expression of CD16 (for NK cells and \nmonocytes/macrophages), CD69 and CD25 (T cells, NK cells), NKp46 (NK cells), perforin \n(CD8+ T cells, NK cells) and FOXP3 (CD4 T cells).  \n \n \nFigure 1: Gating strategy to identify the main mononuclear cell types, exemplified mainly by a PB \nsample. For details see text.  \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n7 \n \nELISA \nSamples of ME and PB were centrifuged for 15 min at 4° Celsius with 1,000 (cytokines) and \n3,000 (ICAM-1 and VCAM-1) rpm, respectively. The pipetted aliquots were stored at -20° \nCelsius and -80° Celsius, respectively. Levels of IL-6, IL-8 and TNF-α , ICAM-1 and VCMA-1 \nwere assessed using ELISA-Kits (Thermofischer, Waltham, Massachusetts, USA): Invitrogen \nICAM-1 (Soluble) Human ELISA Kit (Catalog-Nr. BMS201), Invitrogen VCAM-1 Human \nELISA Kit (Catalog-Nr. KHT0601), Invitrogen IL-6 Human Uncoated ELISA Kit with Plates \n(Catalog-Nr. 88-7066-22), Invitrogen IL-8 Human Uncoated ELISA Kit with Plates (Catalog-\nNr. 88-8086-22), Invitrogen TNF-\nα  Human Uncoated ELISA Kit with Plates (Catalog-Nr. 88-\n7346-22). The ELISA were carried out according to the manufacturer’s protocol. \nConcentrations of cytokines in ME were higher than in plasma of PB and, therefore, requiring \nhigh dilution (IL-6: 1:120, IL-8: 1:400, TNF-\nα : 1:40). Peripheral blood plasma levels of \ncytokines were mainly under the lowest level of detection (detection limits: IL-6: 2-200 pg/ml, \nIL-8: 2-250 pg/mL, TNF-alpha: 4-500 pg/mL, sVAM-1: 0.59-75 ng/mL, sICAM-1: 6.25-100 \nng/mL).  \nStatistical analysis \nMann-Whitney U-test was used for comparison of independent samples (endometriosis \npatients vs. healthy controls) and Wilcoxon-Signed-Rank test was applied for paired samples \n(MMC vs PBMC). The tests were performed two tailored and a p-value of < 0.05 was \nconsidered significant. The statistical analysis was performed with R Version 3.6.1.  \nResults \nParticipants \n12 women with endometriosis and 11 healthy controls fulfilled the inclusion criteria and were \nincluded in the study. There was no significant difference in mean age of women with \nendometriosis (average age: 33.7 years, standard deviation (SD): 6.3) and healthy controls \n(average age: 32.5 years, SD: 7.2). Women with endometriosis had significantly greater \nmaximum pelvic pain (dysmenorrhea) symptoms during the last menstruation compared to \nthe healthy control. On a numeric rating scale from 0 = no pain to 100 = maximum pain, \nwomen with endometriosis reported a median of 78 (range: 65 - 100) compared to a median \nof 22 (range: 0 - 68) reported by healthy controls (p-value: < 0.0001). According to the \nrASRM classification [14], 4 women of the endometriosis groups had stage I endometriosis, 2 \nwomen had stage II endometriosis, 3 women had stage III endometriosis and 3 women had \nstage IV endometriosis. On average, the mean time between the first surgery due to \nendometriosis and the participation in this study was 19.5 months (minimum: 2 months, \nmaximum: 58 months). No significant differences between the two groups were found for \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n8 \n \nplasma levels of hemoglobin, C reactive protein (CRP), estrogen, progesterone, follicle \nstimulating hormone (FSH), luteinizing hormone (LH), Ca-125 or differential blood count.  \nAll women described the use of the menstrual cup as tolerable. The average volume of the \ncollected ME samples was 30.4 ml (min: 10 ml, max: 95 ml) for patients with endometriosis \nand 21.4 ml (min: 10 ml, max: 60 ml) for healthy women (measured before filtering with cell \nstrainer), p-value: 0.1917.  \nFlow cytometry of ME and PB samples comparing endometriosis patients with healthy \ncontrols \nTable 1 summarizes the results of the flow cytometry examinations including all p-values. \nRegarding the main cell types, no significant differences were observed in ME or PB \nbetween women with endometriosis and healthy controls. Both groups had similar \npercentages of T cells, NK cells, monocytes/macrophages and B cells in ME or PB (Fig 2, A \nand B).  \n \n \nFigure 2: Frequencies of the main mononuclear cell types in menstrual effluent (ME) (A) and \nperipheral blood (PB) (B) compared between endometriosis patients and healthy controls. Each \nsymbol corresponds to one sample of either ME or PB of an endometriosis patients or healthy control. \nThe box plot depicts the median of each group with the 75 and 25 percentiles. \n \nReduced frequency of perforin+ CD8 T cells in menstrual effluent of endometriosis patients \nCD8+ T cells were more frequent in MMC of endometriosis patients compared to MMC of \nhealthy controls, yet not reaching significance (Fig 3 A). Among PBMC, the opposite, non-\nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n9 \n \nsignificant, trend was observed. Consistently with the enrichment of CD8+ T cells, the ratio of \nCD4/CD8 T cells was slightly reduced in MMC and enhanced in PBMC of patients (Fig 3 B). \nThe frequency of Treg was enhanced in MMC and PBMC of endometriosis patients, yet not \nreaching significance (Fig 3 C).  \nThe fraction of CD8+ T cells, which contained perforin, identifying them as cytotoxic, was \nsignificantly reduced in ME of endometriosis patients compared to controls (median \nendometriosis 4.0%, median healthy controls: 11.2%, p-value: 0.0295) (Fig 3 D). In PBMC, \nthere was no difference in CD8+ perforin+ T cells between patients and healthy controls. \nRepresentative density plots are shown in Figure 3 E.  \n \n \nFigure 3: Frequencies of T cell subsets in women with endometriosis and healthy controls. A: \nPercentage of CD8+ T cells in ME and PB. B: CD4/CD8 ratio of T cells in ME and PB. C: Percentage \nof Tregs (FOXP3+CD25+) among CD4+ T cells in ME and PB. D: Proportion of perforin-positive CD8+ \nT cells among all CD8+ T cells. Each symbol in the box plots corresponds to one sample of either ME \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n10 \n \nor PB of an endometriosis patient or healthy control. The box plot depicts the median of each group \nwith the 75 and 25 percentiles. E: Representative density plots of CD8+ T cells of MMC and their \nexpression of perforin, 4 examples of individual endometriosis patients and healthy controls.  \n \nAltered NK cell subset frequencies in MMC compared to PMBC with reduced frequencies of \nCD56dim und perforin+ NK cells and enriched CD69 and NKp46 expression in MMC  \nNK cells in PB are distinguished by their expression levels of CD56 and CD16 [15, 16]. In \nhealthy controls, the predominant NK cell population is CD56dim and CD16bright \nrepresenting 90-95% of the peripheral NK cells. The remaining NK cells are CD56bright with \nCD16low. This representation was observed in PBMC of healthy controls as well as \nendometriosis patients (Fig 4 A, B). These two NK cell populations were also found in the \nMMC samples, however with strikingly different frequencies compared to PB. The \nCD56bright/CD16dim NK cells were the dominant NK cell type and frequencies of the \nCD56dim/CD16bright NK cells were low. However, there was no difference in the subset \ndistribution between endometriosis and healthy control MMCs (Fig 4 C, D).  \nNK cells have natural killing capacity due to constitutive expression of perforin, which is \nrestricted mainly to the CD56dim/CD16bright NK cell subset [16]. Similarly, in MMC-NK cells \nthe perforin expression was found in the CD56dim NK cells. The frequency of perforin+ NK \ncells in PBMC was between 90 and 95% in endometriosis patients and healthy controls (Fig \n4 F). Similar to PBMC, the perforin-positive NK cells also were of the CD56dim subset in \nMMC, but in contrast to PBMC, a larger percentage of NK cells of MMC were perforin-\nnegative, in endometriosis as well as controls (Fig 4 E). \n \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n11 \n \n \nFigure 4: NK cell subsets in endometriosis and healthy controls. A: PBMC NK cell subsets \ndistinguished according to their expression of CD56 and CD16 B: representative dot plots for PBMC. \nC: MMC NK cell subsets according to their expression of CD56 and CD16. D: representative dot plots \nfor MMC. E: CD56dim/Perforin-high NK cells of MMC, determined as percentage of all NK cells with a \nrepresentative dot plot. F: CD56dim/Perforin-high NK cells of PBMC as percentage of all NK cells with \nrepresentative dot plot.  \n \nMyeloid cells in MMC do not follow the CD14/CD16 subset distinction of monocytes in PBMC \nMyeloid cells are represented in PBMC by monocytes. They are represented by 3 \npopulations distinguished based on their CD14 and CD16 expression levels [17, 18]. The \nmain population are the classical monocytes, which express high levels of CD14 and no/low \nCD16. Intermediate and non-classical monocytes are represented with much lower \nfrequency than classical monocytes in healthy blood. They are characterized by high \nexpression of CD14 together with varied CD16 (intermediate monocytes) and very low levels \nof CD14 with high CD16 (non-classical monocytes), respectively. All three subsets were \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n12 \n \nidentified in PBMC of healthy and endometriosis patients, without noticeable differences \nbetween endometriosis and healthy controls (Fig 5 A and B). In MMC, the typical \nCD14/CD16 monocyte subsets were no longer discernable. Only two populations could be \ndiscriminated with CD14high/CD16low or CD14low/CD16high expression, respectively (Fig 5 \nC and D). The CD14high/CD16low cells represented the more frequent myeloid cell \npopulation with a stronger enrichment in endometriosis.  \n \n \nFigure 5: Myeloid cells in PBMC and MMC. A: monocyte subsets in PBMC discerned according to \ntheir expression of CD16 and CD14. Box plots summarize all samples with each symbol depicting the \ncell frequency of one patient or healthy control. The box delineates the 75 and 25 percentile of the \ngroup and the median. B: representative dot plot of PBMC monocytes. C: myeloid cell subsets MMC \ndistinguished according to their expression of CD14 and CD16. D: The dot plots depict examples of \nsubset distribution in pre-gated non-B, non-T, non-CD56 cells. On the left: endometriosis, on the right: \nhealthy control. \n \nMMC cell subset composition differs from PBMC identifying ME as a compartment distinct \nfrom PB \nIn order to elucidate differences between MMC and PBMC we compared cell frequencies in \nMMC and PBMC separately in endometriosis patients and healthy controls, and, additionally, \nwithout distinguishing healthy controls and patients. Results are displayed in Figure 6 and \nTable 1. The comparison revealed that ME had significantly less T cells (MMC median: 44.1, \nPB median: 66.0, p-value: 0.00137), while more B cells were present in ME. The reduction in \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n13 \n \nT cells and increase in B cells were less pronounced in samples of patients with \nendometriosis. The frequencies of CD56+ NK cells (gated as non-CD19/20, non-CD3, non-\nCD14) and myeloid cells (gated as non-CD19/20, non-CD3, non-CD56) as percentage of \ntotal mononuclear cells did not differ significantly between MMC and PBMC. However, the \nfrequency of NK cells was mostly lower in MMC compared to PBMC of endometriosis \npatients, but not of healthy controls.  \nThe CD4/CD8-ratio among the CD3+ T cells was significantly higher in MMC (MMC median: \n2.6, PBMC median: 2.2, p-value: 0.0399). CD8+ T cells were also significantly less frequent \namong total CD3+ T cells in MMC than in PBMC (MMC median: 26.0, PBMC median: 29.6, \np-value: 0.0239; not significant for the endometriosis group). Furthermore, significantly less \nCD8+ T cells were perforin-positive in the MMC compared to the PBMC (MMC median: 5.3, \nPBMC median: 9.0, p-value: 0.024). The proportion of Tregs (FOXP3+CD25+) among CD3+ \nT cells was significantly lower in MMC than in PBMC (MMC median: 0.52, PBMC median: \n4.99, p-value: < 0.0001). T cells expressing the activation marker CD69 were significantly \nmore frequent among CD3+ T cells from MMC compared to PBMC (median of CD69+ T cells \nin MMC: 48.2%, median of CD69+ T cells in PBMC: 11.1%, p-value < 0.0001).  \nRegarding the NK cell population, MMC contained significantly more CD56bright/CD16dim \nNK cells as percentage of CD56+ NK cells (MMC median: 67.8%, PBMC median: 3.9%, p-\nvalue: < 0.0001) (Fig 6 D), and significantly more NK cells expressed the activation marker \nCD69 in MMC compared PBMC (MMC median: 42.9%, PBMC median: 11.2%, p-value: \n<0.0001). Moreover, the NK cells from MMC had significantly lower NKp46+ NK cells (MMC \nmedian: 58.4%, PBMC median: 91.5%, p-value: < 0.0001), as well as significantly fewer \nperforin-positive NK cells than PBMC (MMC median: 40.4%, PBMC median: 92.6%, p-value: \n< 0.0001). The cell surface marker CD69+ was significantly more expressed on NK cells in \nME than in PB.  \nThe distribution of myeloid subsets (gated as non-CD19/20, non-CD3 non-CD56, CD14+) \nwas discriminated based on their expression of CD14 and CD16. In PBMC, the expected \ndistinction in classical, intermediate and non-classical monocytes was observed [19], with \ncell frequencies similar in MMC and PBMC (Table 1, Fig 5). As described above, MMC \nappeared to contain only two main groups of myeloid cells, CD14high/CD16low and \nCD14low/CD16high. The CD14high/CD16low cells represented the more frequent myeloid \ncell population in ME; more pronounced in endometriosis patients than in heathy controls.  \n \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n \nFigure 6: A: Frequencies of the main cell types among the mononuclear cells of MMC and PMBC of \neither endometriosis patients or healthy controls. Values were determined by flow cytometry and gated \nas described in Fig 1. Samples of MMC and PBMC corresponding to the same individual are \nconnected by line. B: Stacked bar plots compare the cell type composition of ME and PB of either \nendometriosis patients or healthy controls (presentation of mean values) C: CD4/CD8-ratio, \nfrequencies of Treg cells (gated as FOXP3+CD25+CD4+ among the CD4 T cells), CD69-positive cells \n(among gated CD3+ T cells), CD8+ T cells (of CD3+ T cells) and perforin+ cells gated among the \nCD8+ T cells. D: NK cell subsets as percentage of non-CD19/20, non-CD3, non-CD14 positive cells \n(see figure 1). Frequencies of CD56brigh/CD16dim NK cells, CD69+, NKp46+, and perforin+ cells \namong gated NK cells.  * p-value < 0.05, ** p-value < 0.001, ** p-value < 0.0001 \n \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n15 \n \nEndometriosis patients have lower plasma levels of soluble ICAM-1  \nMedian PB plasma levels of ICAM-1 were significantly lower for women with endometriosis \n(median: 355.2 ng/ml) compared to healthy controls (median: 459.2 ng/ml; p-value: 0.0268; \nFigure 7 A). There were no differences in ICAM-1 concentrations in ME (endometriosis \nmedian: 601.9 ng/ml; healthy controls median: 673.4ng/ml). \nVCAM-1 concentrations were comparable for both groups for menstrual effluent \n(endometriosis median: 551.8 ng/ml; healthy controls median: 543.0 ng/ml) and plasma of \nPB (endometriosis median: 1089.0 ng/ml; healthy controls median: 853.0 ng/ml). Overall, \nICAM-1 levels were higher in ME compared to PB, while VCAM-1 levels were higher in \nplasma of PB compared to ME.  \nConcentrations of IL-6, IL-8 and TMF-alpha did not differ significantly between women with \nendometriosis and healthy controls (IL-6: endometriosis median: 7.7 ng/ml, healthy controls \nmedian: 8.5 ng/ml; IL-8: endometriosis median: 57.0 ng/ml, healthy controls median: 71.8 \nng/ml; TNF-α : endometriosis median: 7.9 ng/ml, healthy controls median: 10.0 ng/ml). \nCytokine concentrations in the PB were below the detection limit of the ELISA-kits used in \nthis study.  \nFigure 7: A: Concentrations of cell adhesion molecules ICAM-1 and VCAM-1 in ME and PB. * p-value \n< 0.05. B: Concentrations of cytokines IL-6, IL-8 and TNF- α  in ME.  \n \nDiscussion \nAlthough immunological alterations have been studied extensively in the diseases of \nendometriosis, mechanisms underlying the pathophysiology are still unclear. The aim of this \nstudy was to contribute to the current knowledge on immunological changes in women with \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n16 \n \nendometriosis by analyzing MMC of ME collected with the help of menstrual cups. ME and \nPB samples were examined by flow cytometry and ELISA. \nThe comparison of the main mononuclear cell groups (T cells, B cell, NK cells and myeloid \ncells) revealed no significant differences between endometriosis and healthy controls for both \nME and PB. A not statistically different deviation, however, was seen in the frequency of NK \ncells, which was less in ME from women with endometriosis than from the control group. This \nconfirms the observation by Warren et al. [20], which suggested that NK cells and reduced \ncytotoxic capacity might be involved in the pathogenesis of the disease [21, 22]. Thus, we \nexpected differences within the NK cell subsets, in particular the frequency of \nCD56\nbright/CD16dim NK cells, which are known to produce high amounts of cytokines; but \ncontain less perforin, granzymes and cytolytic granules and are considered less cytotoxic \n[19]. Indeed, prior studies found increased percentages of immature CD56bright NK cells in \nendometrial tissue and peripheral blood of women with endometriosis and infertility [22–24]. \nIn our study, we did not find significant differences in NK cell subsets of PBMC or MMC \nbetween endometriosis patients and healthy controls. Furthermore, the frequency of NK cells \nexpressing perforin, NKp46 or CD69 was similar for both groups in ME as well as in PB. In \nsummary, our study did not reveal evidence of an altered activation or cytotoxic status of NK \ncells in the ME of women with endometriosis compared to healthy individuals. There are \nseveral possible explanations, why we did not detect differences in NK cells, including the \nrelatively small sample sizes and differences in study design and methodology, like different \ntime-points in menstrual cycle when collecting the samples, varying stages of the disease \nand differences in sample processing (cell isolations and freezing methods, gating strategy \nand used antibodies). \nRegarding T cells, in ME as well as in PB, no significant differences were found for the \nCD4/CD8 ratio and the frequency of CD8\n+T cells as percentage of CD3+ T cells in patients \ncompared to controls. Prior studies indicated an altered CD4/CD8 ratio in patients with \nendometriosis with an increased ratio in the peripheral blood [25] and a decreased ratio in \nperitoneal fluid of women with endometriosis [7, 9]. Furthermore, in our study group, there \nwere no significant differences in the expression of CD69 on T cells between endometriosis \npatients and healthy controls, neither in ME nor in PB. Guo et al. found a higher frequency of \nCD69+ T cells in peritoneal fluid of women with endometriosis [26] suggesting that \nupregulated CD69 expression might be a central characteristic of endometriosis T cells. The \nimmune alterations they found were more prominent in minimal/mild endometriosis, \nindicating an influence of the stage of the disease. The diverging results compared to what \nwe found regarding CD69 expression on T cells in endometriosis might primarily be \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n17 \n \nexplained by the different compartments (ME versus peritoneal fluid) of the sample \ncollection.  \nTregs are immunomodulatory cells that can inhibit cytotoxicity of CD8+ T cells and NK cells \n[27, 28]. Raised numbers of Tregs or a higher activation state of Treg might lead to a less \ncytotoxic environment allowing endometrial cells to grow outside of the endometrium [29]. \nWhile other studies found increased percentages of Treg cells (as percentage of CD4+ T \ncells) in endometriotic lesions and peritoneal fluid, the evidence for altered Treg numbers in \neutopic endometrium and peripheral blood is weak [29, 30]. In the present study, the \npercentage of FOXP3+CD25+ Treg cells within CD4+ T cells was non-significantly higher in \nwomen with endometriosis compared to healthy control in ME as well as in PB.  \nNotably, we observed that CD8+ T cells in the ME of women with endometriosis were \ncharacterized by a reduced percentage of cells that contain perforin, indicating a reduced \ncytotoxicity of the CD8+ T cell subset. The paucity of perforin+ CD8 T cells in ME of \nendometriosis patients was not recapitulated in the CD8+ T cells of PB, which contained \nperforin+ CD8+ T cells comparable to healthy controls. The reduced number of perforin-\npositive CD8+T cells in ME of women with endometriosis could be due to incomplete or \nsuppressed CD8 T cell differentiation caused by cytokines, such as IL-10, VEGF or TGF-ß \n[31–33]. An alternative explanation could be that recent cytotoxic activation led to granule \nexocytosis with consecutive loss of perforin granula [34]. Since no further data is available \n(such as measurements of other time-points in the menstrual cycle or information on T-cell \nactivity modulating substances like IL-10, VEGF or TGF-ß), we cannot distinguish between \nthese two possibilities. However, an involvement of cytotoxic T cells in the pathogenesis of \nthe disease has been suggested by previous studies which reported a defective T-cell \nresponse and reduced cytotoxicity to autologous endometrial cells in endometriosis [7]. \nKonno et al. suspected that perforin and granzyme B of cytotoxic T cell and NK cells cause \napoptosis in human endometrium in order to induce endometrial menstruation [35]. A similar \nmechanism might be responsible for clearing ectopic endometrial cells to prevent \nendometriotic lesions from growing. Deficiency in perforin-mediated cell death pathway plays \nan important role in the susceptibility to cancer [36–39]. Cytotoxic lymphocytes are required \nto detect and destroy transformed and dislocated cells [40]. There are pathophysiological \nsimilarities between cancer growth and benign tumors such as endometriosis. Therefore, \ninefficient cytotoxic activity might enable persistence of ectopic endometrial tissue. \nIn conclusion, our study revealed that ME of endometriosis patients is characterized by \nsignificantly less perforin-positive CD8+ T cells. Together with less NK cells and higher \nnumbers of Tregs in ME, this might create an environment of reduced cytotoxicity in the \nmenstrual effluent in women with endometriosis. This fits well with the hypothesis that \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n18 \n \nreduced cellular cytotoxicity prevents cellular clearing of shed endometrial tissue [7]. Of \ntherapeutic relevance, the cytokine IL-2, which induces perforin expression, has been shown \nto restore in vitro the cytolytic activity especially of T cells in women with endometriosis [41]. \nFuture clinical research is needed to answer the question whether IL-2 or other substances \nmight offer new possibilities in the treatment of endometriosis.  \nIn addition to the identification of distinct differences between patients and healthy controls, \nwhich might be of therapeutic relevance, we also found that the mononuclear cells from ME \nare distinctly different from those of PBMC, independent of disease status. Observed \ndifferences include: lower frequency of T cells; lower frequency of Treg cells; more T cells \nand NK cells expressing CD69, and less T cells and NK cells expressing perforin. In addition, \nthe subset distribution of NK cells was inverted in ME compared to PB with CD56\nbright being \nthe dominant subset in ME.The NK cells in ME showed, moreover, reduced NKp46 \nexpression.  \nThe lower frequency of CD3+ T cells is in line with prior results [42–44]. The CD4/CD8 ratio \nwas significantly higher and proportion of CD8+ T cells was significantly lower in ME \ncompared to PB in the healthy study group, but not in endometriosis patients. A prior study \nreported similar proportions of CD8+ T cells and CD4+ T cells in ME compared to PB [43], \nwhile van der Molen et al. found higher percentages of CD8+ T cells in ME [42]. Our finding \nof significantly less Treg cells (FOXP3+CD25+) among CD4+ T cells in ME compared to PB \nis in contrast to that of Feyaerts et. al who reported similar percentages for ME and PB [43]. \nIn accordance with prior findings is our observation of highly enhanced CD69 expression on \nT cells from ME compared to PB [42]. \nUnlike other studies, [43, 44], we did not observe an enrichment of total NK cells in ME, yet \nthe CD56bright subset was present at significantly higher frequency in ME compared to PB. \nThe CD56bright/CD16dim NK cells represent a less cytotoxic subgroup and are known to be \nabundant cytokine producers [45]. Our result confirms prior studies, which also found \nsignificantly more CD56bright/CD16dim NK cells in ME compared to PB [42, 44]. The \ncomposition of the NK cell subsets is comparable to that of endometrial NK cells (obtained by \nendometrial biopsy), which are also predominantly CD56bright NK cells [46]. In contrast to PB, \nNK cells of ME were largely perforin-negative and the expression of the NK receptor NKp46 \nwas significantly lower in NK cells. These aspects indicate a reduced cytotoxic potential of \nNK cells in ME. Similar to the T cells, the NK cells from ME expressed significantly more \nCD69. Guo et al. analyzed the immune cells of peritoneal fluid (PF) and observed higher \nCD69 expression of PF immune cells compared to their counter parts in PB, and also \nreduced cytolytic activity [26], suggesting that the immune environment in PF might be \nsimilar to that of ME. \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n19 \n \nCD19/CD20+ B cells in our study group showed a statistical trend towards higher \nfrequencies in ME compared to PB. Prior studies reported conflicting results regarding the \nnumber of B cell in ME. One study found lower numbers of B cells in ME compared to PB (n \n= 5 on several consecutive menstrual cycles) [42], another did not find significant differences \n(n = 17) [43], while a third study reported higher percentages of B cells (n = 12) [44]. \nExplanations for the diverging results reported might be the relatively small sample sizes and \ndifferences in the study population, sample collection and sample processing (e.g., ME \ncollection just before pregnancy, time-point of ME collection in menstrual cycle). \nCells of the myeloid lineage were observed in PB as the typical monocytic subsets with \npredominance of classical monocytes (CD14\nhigh/CD16neg), and low frequencies of \nintermediate (CD14high/CD16intermediate) monocytes and non-classical monocytes \n(CD14low/CD16high). In ME, the typical CD14/CD16 monocyte subsets were not discernable \nrather only two groups were seen, one expressing high CD14 together with low levels of \nCD16 (CD14\nhigh/CD16low) and the other group expressing low CD14 together with high CD16 \n(CD14low/CD16high). This classification was appropriate for the majority of the ME samples, \nhowever, 3 of 23 samples did not fit properly into this dichotomization. The differences in \nexpression of myeloid markers in PB and ME suggest that the myeloid cell types of the two \ntissue compartments are distinct in their polarization. It further suggests that ME is not a PB \ncompartment but rather a tissue compartment where myeloid cells seem to differentiate into \nmacrophages [47]. To our knowledge, subtypes of ME myeloid cells have not yet been \ninvestigated. Further research is necessary to gain a deeper understanding of the myeloid \ncells in ME.  \nWhile differences were found between myeloid cell subsets from ME and PB, no differences \nin myeloid cells were observed between endometriosis and healthy controls. Prior studies \ninvestigated peritoneal macrophages and indeed found differences in activation, function and \nexpression of cell markers between women with endometriosis and healthy women [26, 48–\n50]. For instance, Guo et. al reported significantly higher expression of CD16 in peritoneal \nmacrophages in minimal/mild endometriosis compared to healthy controls, but not so in more \nsevere stages [26]. Nevertheless, prior studies on macrophages in endometriosis obtained \nthe cells from peritoneal fluid. In the present study macrophages were obtained from ME and \ntherefore comparability to previous studies on macrophages in endometriosis is limited.  \nIn conclusion, we found significant differences in mononuclear cells between ME and PB not \nonly for the main cell groups, but also in subtypes of T cells, NK cells and myeloid cells. The \ndifferences in mononuclear cells from ME and PBMC indicate the mucosal/endometrial origin \nof the MMC, supporting the assumption that mononuclear cells derived from the collection of \nME closely resemble the uterine immunological environment [42].  \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n20 \n \nCytokines and cell adhesion molecules are immunologic messengers. Differences in \nconcentrations of these messenger substances might be a possible explanation for altered \nimmune cells in patients with endometriosis. [51]. Previous studies found associations \nbetween endometriosis and levels of certain cytokines (e.g. IL-6, IL-8, TNF-α ) and cell \nadhesion molecule in PB or peritoneal fluid [10, 11, 52, 53]. However, to the best of our \nknowledge, no study has analyzed ME in this regard. We measured concentrations of the \ncell adhesion molecules sICAM-1 and sVCAM-I, and cytokines IL-6, IL-8 and TNF-α  in ME \nand PB. All substances are suspected to be involved in the pathophysiology of \nendometriosis. In prior studies it was hypothesized, that soluble ICAM-1 affects the immune \nsurveillance of shed endometrial cells by immune cells, especially NK cells [54, 55]. Cell \nbound ICAM-1 on endometrial cells might initiates cell-cell interaction and, therefore, be \nessential for proper immune surveillance [56]. ICAM-1 also plays a key role in \nleukodiapedesis by mediating cell contacts between leukocytes and endothelial cells [57].  \nWe found significantly lower concentrations of ICAM-1 in PB of women with endometriosis, \nwhich confirms prior findings [12, 53]. Nevertheless, other studies reported no differences or \neven higher concentrations in endometriosis [58, 59]. Kuessel et al. reported significantly \nlower serum ICAM-1 levels in women with endometriosis compared to healthy controls, but \nthe serum levels increased after laparoscopy and were significantly higher compared to the \ncontrol group 6-10 weeks after laparoscopy [12]. Placido et al. found significantly lower levels \nof sICAM-1 in women with stage I-II endometriosis compared to women with stage III-IV \nendometriosis, suggesting an important influence of the stage of the disease on sICAM-\nlevels [60]. Furthermore, sICAM-1 levels might be influenced by menstrual cycle, which \nwould be a further explanation for diverging results reported in scientific literature on sICAM-\n1 levels in endometriosis.  \nIn ME, concentrations of ICAM-1 were not different between women with endometriosis and \nhealthy controls; and VCAM-1 concentrations showed no differences in ME or PB. The levels \nof IL-6, IL-8 and TNF-\nα  in plasma were under detection limit of the ELISA kits that we used in \nthis study. In ME, however, the concentrations of these cytokines were very high, which \nmade high dilutions necessary. It is uncertain whether these high values indeed represent \nhigh concentrations of extracellular cytokines in ME. Cellular production of cytokines and cell \nlysis of leucocytes, but also red blood cells, can lead to artificially high amounts of cytokines \nin centrifugation supernatant [61, 62]. In contrast to the collection of PB, for ME, there was a \ndelay of up to 4 days between the time of collection by the study participants and handing \nover the samples for analysis. Consequently, substantial cell lysis might have occurred \nduring the storage of the samples. Furthermore, the leucocytes might have produced \ncytokines in response to oxidative stress caused by the storage of the samples. Both \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n21 \n \npossibilities might explain the very high amounts of cytokines in ME of women with and \nwithout endometriosis.  \nLuckow Invitti et al. cultivated endometrial cell obtained from biopsy of women with \nendometriosis and healthy controls. The endometrial cells from women with endometriosis \nproduced significantly more IL-6 and IL-8 than endometrial cells from healthy women on day \n7 and 10, but no significant differences were seen in the first days of the cell culture (day 1 \nand 3) [63]. Consistent with these results, we did not detect any significant differences in \nconcentrations of IL-6, IL-8 or TNF-\nα  in ME between women with endometriosis and healthy \ncontrols, which were not cultivated and processed within 1 to 4 days after collection.  \nOur results suggest that collecting and analyzing ME using menstrual cups might represent a \nnew and valuable non-invasive opportunity for further research and diagnostic of \nendometriosis. At the moment, diagnosing the disease remains challenging and several \nyears may pass until a diagnosis is reached, which means a long time of uncertainty for the \npatients. Quite frequently, invasive procedures like laparoscopy are necessary for reliable \ndiagnosis. Until today, ME is not considered in the diagnosis of endometriosis. Previous \nstudies evaluated and approved the use of menstrual cups in order to obtain viable \nendometrial tissue [42, 64]. In this study, the use of menstrual cups was tolerated well by the \nstudy participants and proved to be a suitable method to collect ME. It was possible to isolate \na sufficient number of vital mononuclear cells for flow cytometry measurement. The origin of \nthe endometriotic cells is very likely eutopic endometrium and we are supporting evidence \nthat ME closely resembles the endometrial tissue. Therefore, by collecting and analyzing ME, \nvaluable information from a site close to the origin of the disease can be obtained. The \ncollection is a non-invasive method and might complement the existing non-invasive \ndiagnostic repertoire like ultrasound. In particular, cytotoxic T cells derived from ME in \nregards to their perforin equipment might represent a new approach in the diagnostics of the \ndisease. Further research and larger studies are needed to verify this assumption.  \nOur study has some strengths and weaknesses. First, all participants of the endometriosis \ngroup had biopsy confirmed endometriosis. All participants underwent a gynecological \nexamination to confirm or exclude the presence of endometriosis. This is one of the very first \nstudies to investigate immunological differences in ME between women with endometriosis \nand healthy controls. We examined MMC by flow cytometry using a wide variety of \nantibodies, which allowed a detailed analysis. Limitations are, firstly, the small number of \nparticipants (n = 23), which means the influence of random fluctuations might be substantial \nand smaller differences between the two study groups cannot be detected. Nevertheless, to \nour knowledge, this is the largest study so far examining MMC from women with \nendometriosis. Another limitation is the process of collecting the ME samples by the study \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n22 \n \nparticipants. We must consider differences in handling and storing of the samples among the \nstudy participants. This might influence, amongst others, the viability of the MMC and thereby \nthe results obtained by flow cytometry and ELISA. However, to minimize variations across \nthe samples, the study participants received precise instructions on how to collect and store \nthe sample until handing it over. Moreover, the cryopreservation of the mononuclear cells \nbefore flow cytometry could have affected the viability of the immune cells and could have \ninduced alterations in marker expression. Nevertheless, studies have shown that \ncryopreservation can be an adequate method not only for analyzing PBMC but also for \nmucosal leucocytes [65, 66]. All mononuclear cells were frozen by the same person, who \nalso performed subsequent thawing, staining and flow cytometry to maximize uniformity in \nthe process and to reduce interassay variability. PB and ME of the same individual were \nalways handled together, in parallel with PB and ME of a healthy control. Flow cytometry \nstaining was performed in batches, including each time PB and ME samples of 5 patients \ntogether with 5 healthy donors.  \nConclusions \nMMC were found to be distinctively different from PBMC and exhibited characteristics of \nendometrial origin. Perforin+ T cell were significantly reduced among CD8+ T cells from ME \nof endometriosis patients compared to ME of healthy controls This suggests a reduced \ncytotoxic potential, which might result in a reduced capacity to remove endometrial cells from \nectopic locations. Plasma ICAM-1 levels were significantly lower in endometriosis compared \nto healthy controls.  \n \n \nAuthors´ contributions \nKM, VH, EO, RP, AS and TS conceived and carried out the study. TS performed the \nstatistical analysis and drafted the manuscript. KM and EN supervised data analysis and \nmanuscript drafting. All authors reviewed the manuscript. \nEthics approval and consent to participate \nAll study participants have given written informed consent. The concept and implementation \nof this study was approved by ethical committee of the Medical Faculty at LMU Munich (no. \n17-695) and the study was performed in accordance with the Declaration of Helsinki. \nFunding \nThe study was supported by the Schweizer-Arau-Foundation, Germany. \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n23 \n \nConflict of interests \nThe authors declare that they have no competing interests. \nAcknowledgements \nWe acknowledge B. Mosetter for excellent technical assistance in flow cytometry, L. Ziegler-\nHeitbrock and T. Hofer for insight into myeloid cell characterization.  \n  \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n24 \n \nReferences \n1. Giudice, L.C., Kao, L.C.: Endometriosis. The Lancet 364(9447), 1789–1799 (2004). doi: \n10.1016/S0140-6736(04)17403-5 \n2. Falcone, T., Flyckt, R.: Clinical Management of Endometriosis. Obstetrics & Gynecology \n131(3), 557–571 (2018). doi: 10.1097/AOG.0000000000002469 \n3. Hudelist, G., Fritzer, N., Thomas, A., Niehues, C., Oppelt, P., Haas, D., Tammaa, A., \nSalzer, H.: Diagnostic delay for endometriosis in Austria and Germany. Causes and \npossible consequences. Human Reproduction 27(12), 3412–3416 (2012). doi: \n10.1093/humrep/des316 \n4. Ahn, S.H., Monsanto, S.P., Miller, C., Singh, S.S., Thomas, R., Tayade, C.: \nPathophysiology and Immune Dysfunction in Endometriosis. BioMed Research \nInternational 2015 (2015). doi: 10.1155/2015/795976 \n5. Sourial, S., Tempest, N., Hapangama, D.K.: Theories on the Pathogenesis of \nEndometriosis. International Journal of Reproductive Medicine 2014(2), 1–9 (2014). doi: \n10.1155/2014/179515 \n6. Sasson, I.E., Taylor, H.S.: Stem Cells and the Pathogenesis of Endometriosis. Annals of \nthe New York Academy of Sciences 1127(1), 106–115 (2008). doi: \n10.1196/annals.1434.014 \n7. Osuga, Y., Koga, K., Hirota, Y., Hirata, T., Yoshino, O., Taketani, Y.: Lymphocytes in \nEndometriosis. American Journal of Reproductive Immunology 65(1), 1–10 (2011). doi: \n10.1111/j.1600-0897.2010.00887.x \n8. Riccio, L.d.G.C., Santulli, P., Marcellin, L., Abrão, M.S., Batteux, F., Chapron, C.: \nImmunology of endometriosis. Best practice & research. Clinical obstetrics & \ngynaecology 50, 39–49 (2018). doi: 10.1016/j.bpobgyn.2018.01.010 \n9. Izumi, G., Koga, K., Takamura, M., Makabe, T., Satake, E., Takeuchi, A., Taguchi, A., \nUrata, Y., Fujii, T., Osuga, Y.: Involvement of immune cells in the pathogenesis of \nendometriosis. J. Obstet. Gynaecol. Res. 44(2), 191–198 (2018). doi: 10.1111/jog.13559 \n10. Wu, M.-Y., Ho, H.-N.: The Role of Cytokines in Endometriosis*. American Journal of \nReproductive Immunology 49(5), 285–296 (2003). doi: 10.1034/j.1600-\n0897.2003.01207.x \n11. Králí\nč ková, M., Vetvicka, V.: Immunological aspects of endometriosis. A review. Annals \nof translational medicine 3(11), 153 (2015). doi: 10.3978/j.issn.2305-5839.2015.06.08 \n12. Kuessel, L., Wenzl, R., Proestling, K., Balendran, S., Pateisky, P., Yotova, Yerlikaya, G., \nStreubel, B., Husslein, H.: Soluble VCAM-1/soluble ICAM-1 ratio is a promising \nbiomarker for diagnosing endometriosis. Human reproduction (Oxford, England) 32(4), \n770–779 (2017). doi: 10.1093/humrep/dex028 \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n25 \n \n13. Li, R., Qiu, Y.: Diagnostic value of se rum ICAM-1 for endometriosis. A meta-analysis. \nMedicine 97(31), e11760 (2018). doi: 10.1097/MD.0000000000011760 \n14. Haas, D., Shebl, O., Shamiyeh, A., Oppelt, P.: The rASRM score and the Enzian \nclassification for endometriosis: their strengths and weaknesses. Acta obstetricia et \ngynecologica Scandinavica 92(1), 3–7 (2013). doi: 10.1111/aogs.12026 \n15. Cooper, M.A., Fehniger, T.A., Caligiuri, M.A.: The biology of human natural killer-cell \nsubsets. Trends in Immunology 22(11), 633–640 (2001). doi: 10.1016/S1471-\n4906(01)02060-9 \n16. Freud, A.G., Mundy-Bosse, B.L., Yu, J., Caligiuri, M.A.: The Broad Spectrum of Human \nNatural Killer Cell Diversity. Immunity 47(5), 820–833 (2017). doi: \n10.1016/j.immuni.2017.10.008 \n17. Passlick, B., Flieger, D., Ziegler-Heitbrock , H.W.: Identification and characterization of a \nnovel monocyte subpopulation in human peripheral blood. Blood 74(7), 2527–2534 \n(1989) \n18. Ziegler-Heitbrock, L., Hofer, T.P.J.: Towa rd a refined definition of monocyte subsets. \nFrontiers in immunology 4, 23 (2013). doi: 10.3389/fimmu.2013.00023 \n19. Ziegler-Heitbrock, L.: Blood Monocytes and Their Subsets: Established Features and \nOpen Questions. Frontiers in immunology, 423 (2015). doi: 10.3389/fimmu.2015.00423 \n20. Warren, L.A., Shih, A., Renteira, S.M., Seckin, T., Blau, B., Simpfendorfer, K., Lee, A., \nMetz, C.N., Gregersen, P.K.: Analysis of menstrual effluent. Diagnostic potential for \nendometriosis. Mol Med 24(1), 63 (2018). doi: 10.1186/s10020-018-0009-6 \n21. Jeung, I., Cheon, K., Kim, M.-R.: Decreas ed Cytotoxicity of Peripheral and Peritoneal \nNatural Killer Cell in Endometriosis. BioMed Research International 2016(2), 1–6 (2016). \ndoi: 10.1155/2016/2916070 \n22. Thiruchelvam, U., Wingfield, M., O'Farrelly, C.: Natural Killer Cells. Key Players in \nEndometriosis. American Journal of Reproductive Immunology 74(4), 291–301 (2015). \ndoi: 10.1111/aji.12408 \n23. Lynch, L., Golden-Mason, L., Eogan, M., O'Herlihy, C., O'Farrelly, C.: Cells with \nhaematopoietic stem cell phenotype in adult human endometrium. Relevance to \ninfertility? Human Reproduction 22(4), 919–926 (2007). doi: 10.1093/humrep/del456 \n24. Kikuchi, Y., Ishikawa, N., Hirata, U., Imaizumi, E., Sasa, H., Nagata, I.: Changes of \nperipheral blood lymphocyte subsets before and after operation of patients with \nendometriosis. Acta Obstet Gynecol Scand 72(3), 157–161 (1993). doi: \n10.3109/00016349309013364 \n25. Szyllo, K., Tchorzewski, H., Banasik, M., Glowacka, E., Lewkowicz, P., Kamer-\nbartosinska, A.: The involvement of T lymphocytes in the pathogenesis of endometriotic \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n26 \n \ntissues overgrowth in women with endometriosis. Mediators of Inflammation 12(3), 131–\n138 (2003). doi: 10.1080/0962935031000134842 \n26. Guo, M., Bafligil, C., Tapmeier, T., Hubbard, C., Manek, S., Shang, C., Martinez, F.O., \nSchmidt, N., Obendorf, M., Hess-Stumpp, H., Zollner, T.M., Kennedy, S., Becker, C.M., \nZondervan, K.T., Cribbs, A.P., Oppermann, U.: Mass cytometry analysis reveals a \ndistinct immune environment in peritoneal fluid in endometriosis. A characterisation \nstudy. BMC Med 18(1), 441 (2020). doi: 10.1186/s12916-019-1470-y \n27. Suvas, S., Kumaraguru, U., Pack, C.D., Lee, S., Rouse, B.T.: CD4+CD25+ T cells \nregulate virus-specific primary and memory CD8+ T cell responses. J Exp Med 198(6), \n889–901 (2003). doi: 10.1084/jem.20030171 \n28. Pedroza-Pacheco, I., Madrigal, A., Saudemont, A.: Interaction between natural killer \ncells and regulatory T cells: perspectives for immunotherapy. Cellular & molecular \nimmunology 10(3), 222–229 (2013). doi: 10.1038/cmi.2013.2 \n29. Hou, X.-X., Wang, X.-Q., Li, D.-J.: Roles of regulatory T Cells in pathogenesis of \nendometriosis. Reprod Dev Med 3(2), 117 (2019). doi: 10.4103/2096-2924.262392 \n30. Barros, I.B.L. de, Malvezzi, H., Gueuvoghlanian-Silva, B.Y., Piccinato, C.A., Rizzo, L.V., \nPodgaec, S.: “What do we know about regulatory T cells and endometriosis? A \nsystematic review”. Journal of Reproductive Immunology 120, 48–55 (2017). doi: \n10.1016/j.jri.2017.04.003 \n31. Chen, W.F., Zlotnik, A.: IL-10: a novel cytotoxic T cell differentiation factor. J Immunol \n147(2), 528–534 (1991) \n32. Ziogas, A.C., Gavalas, N.G., Tsiatas, M., Tsit silonis, O., Politi, E., Terpos, E., Rodolakis, \nA., Vlahos, G., Thomakos, N., Haidopoulos, D., Antsaklis, A., Dimopoulos, M.A., \nBamias, A.: VEGF directly suppresses activation of T cells from ovarian cancer patients \nand healthy individuals via VEGF receptor Type 2. International journal of cancer 130(4), \n857–864 (2012). doi: 10.1002/ijc.26094 \n33. Thomas, D.A., Massagué, J.: TGF-beta direct ly targets cytotoxic T cell functions during \ntumor evasion of immune surveillance. Cancer cell 8(5), 369–380 (2005). doi: \n10.1016/j.ccr.2005.10.012 \n34. Isaaz, S., Baetz, K., Olsen, K., Podack, E., Griffiths, G.M.: Serial killing by cytotoxic T \nlymphocytes: T cell receptor triggers degranulation, re-filling of the lytic granules and \nsecretion of lytic proteins via a non-granule pathway. Eur. J. Immunol. 25(4), 1071–1079 \n(1995). doi: 10.1002/eji.1830250432 \n35. Konno, R., Igarashi, T., Okamoto, S., Sato, S., Moriya, T., Sasano, H., Yajima, A.: \nApoptosis of Human Endometrium Mediated by Perforin and Granzyme B of NK Cells \nand Cytotoxic T Lymphocytes. Tohoku J. Exp. Med. 187(2), 149–155 (1999). doi: \n10.1620/tjem.187.149 \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n27 \n \n36. Brennan, A.J., Chia, J., Trapani, J.A., Voskoboinik, I.: Perforin deficiency and \nsusceptibility to cancer. Cell Death Differ 17(4), 607–615 (2010). doi: \n10.1038/cdd.2009.212 \n37. Trapani, J.A., Thia, K.Y.T., Andrews, M., Davis, I.D., Gedye, C., Parente, P., \nSvobodova, S., Chia, J., Browne, K., Campbell, I.G., Phillips, W.A., Voskoboinik, I., \nCebon, J.S.: Human perforin mutations and susceptibility to multiple primary cancers. \nOncoimmunology 2(4), e24185 (2013). doi: 10.4161/onci.24185 \n38. Kontani, K., Sawai, S., Hanaoka, J., Tezuka, N., Inoue, S., Fujino, S.: Involvement of \ngranzyme B and perforin in suppressing nodal metastasis of cancer cells in breast and \nlung cancers. European journal of surgical oncology : the journal of the European \nSociety of Surgical Oncology and the British Association of Surgical Oncology 27(2), \n180–186 (2001). doi: 10.1053/ejso.2000.1060 \n39. Cullen, S.P., Brunet, M., Martin, S.J.: Granzymes in cancer and immunity. Cell Death \nDiffer 17(4), 616–623 (2010). doi: 10.1038/cdd.2009.206 \n40. Osi\nń ska, I., Popko, K., Demkow, U.: Perforin. An important player in immune response. \ncejoi 1, 109–115 (2014). doi: 10.5114/ceji.2014.42135 \n41. Melioli, G., Semino, C., Semino, A., Venturini, P.L., Ragni, N.: Recombinant interleukin-2 \ncorrects in vitro the immunological defect of endometriosis. American Journal of \nReproductive Immunology 30(4), 218–227 (1993). doi: 10.1111/j.1600-\n0897.1993.tb00623.x \n42. van der Molen, R.G., Schutten, J.H.F., van Cranenbroek, B., ter Meer, M., Donckers, J., \nScholten, R.R., van der Heijden, O.W.H., Spaanderman, M.E.A., Joosten, I.: Menstrual \nblood closely resembles the uterine immune micro-environment and is clearly distinct \nfrom peripheral blood. Human Reproduction 29(2), 303–314 (2014). doi: \n10.1093/humrep/det398 \n43. Feyaerts, D., Benner, M., van Cranenbroek, B., van der Heijden, O.W.H., Joosten, I., \nvan der Molen, R.G.: Human uterine lymphocytes acquire a more experienced and \ntolerogenic phenotype during pregnancy. Sci Rep 7(1), 584 (2017). doi: \n10.1038/s41598-017-03191-0 \n44. Sabbaj, S., Hel, Z., Richter, H.E., Mestecky, J., Goepfert, P.A.: Menstrual blood as a \npotential source of endometrial derived CD3+ T cells. PloS one 6(12), e28894 (2011). \ndoi: 10.1371/journal.pone.0028894 \n45. Poli, A., Michel, T., Thérésine, M., Andrès, E., Hentges, F., Zimmer, J.: CD56 bright \nnatural killer (NK) cells. An important NK cell subset. Immunology 126(4), 458–465 \n(2009). doi: 10.1111/j.1365-2567.2008.03027.x \n46. Lukassen, H.G.M., Joosten, I., van Cr anenbroek, B., van Lierop, M.J.C., Bulten, J., \nBraat, D.D.M., van der Meer, A.: Hormonal stimulation for IVF treatment positively \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n28 \n \naffects the CD56bright/CD56dim NK cell ratio of the endometrium during the window of \nimplantation. Molecular Human Reproduction 10(7), 513–520 (2004). doi: \n10.1093/molehr/gah067 \n47. Orekhov, A.N., Orekhova, V.A., Nikiforo v, N.G., Myasoedova, V.A., Grechko, A.V., \nRomanenko, E.B., Zhang, D., Chistiakov, D.A.: Monocyte differentiation and \nmacrophage polarization. VP 2019 (2019). doi: 10.20517/2574-1209.2019.04 \n48. Zeller, J.M., Henig, I., Radwanska, E., Dmowski, W.P.: Enhancement of human \nmonocyte and peritoneal macrophage chemiluminescence activities in women with \nendometriosis. American journal of reproductive immunology and microbiology : AJRIM \n13(3), 78–82 (1987). doi: 10.1111/j.1600-0897.1987.tb00097.x \n49. Rana, N., Braun, D.P., House, R., Gebel, H., Rotman, C., Dmowski, W.P.: Basal and \nstimulated secretion of cytokines by peritoneal macrophages in women with \nendometriosis. Fertility and Sterility 65(5), 925–930 (1996) \n50. Gazvani, R., Templeton, A.: Peritoneal environment, cytokines and angiogenesis in the \npathophysiology of endometriosis. Reproduction (Cambridge, England) 123(2), 217–226 \n(2002). doi: 10.1530/rep.0.1230217 \n51. Condotta, S.A., Richer, M.J.: The immune battlefield: The impact of inflammatory \ncytokines on CD8+ T-cell immunity. PLOS Pathogens 13(10), e1006618 (2017). doi: \n10.1371/journal.ppat.1006618 \n52. Witz, C.A.: Cell Adhesion Molecules and Endometriosis. Semin Reprod Med 21(2), 173–\n182 (2003). doi: 10.1055/s-2003-41324 \n53. Barrier, B.F., Sharpe-Timms, K.L.: Expression of Soluble Adhesion Molecules in Sera of \nWomen With Stage III and IV Endometriosis. Journal of the Society for Gynecologic \nInvestigation 9(2) (2002). doi: 10.1016/s1071-5576(01)00158-7 \n54. Somigliana, E., Viganò, P., Gaffuri, B., Guarneri, D., Busacca, M., Vignali, M.: Human \nendometrial stromal cells as a source of soluble intercellular adhesion molecule (ICAM)-\n1 molecules. Human reproduction (Oxford, England) 11(6), 1190–1194 (1996) \n55. Fukaya, T., Sugawara, J., Yoshida, H., Mu rakami, T., Yajima, A.: Intercellular adhesion \nmolecule-1 and hepatocyte growth factor in human endometriosis. Original investigation \nand a review of literature. Gynecologic and obstetric investigation 47 Suppl 1, 11-6; \ndiscussion 16-7 (1999). doi: 10.1159/000052854 \n56. Barber, D.F., Long, E.O.: Coexpression of CD58 or CD48 with Intercellular Adhesion \nMolecule 1 on Target Cells Enhances Adhesion of Resting NK Cells. The Journal of \nImmunology 170(1), 294–299 (2003). doi: 10.4049/jimmunol.170.1.294 \n57. Yang, L., Froio, R.M., Sciuto, T.E., Dvorak, A.M., Alon, R., Luscinskas, F.W.: ICAM-1 \nregulates neutrophil adhesion and transcellular migration of TNF-alpha-activated \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n29 \n \nvascular endothelium under flow. Blood 106(2), 584–592 (2005). doi: 10.1182/blood-\n2004-12-4942 \n58. Wu, M.-H., Yang, B.-C., Hsu, C.-C., Lee, Y.-C., Huang, K.-E.: The expression of soluble \nintercellular adhesion molecule-1 in endometriosis. Fertility and Sterility 70(6), 1139–\n1142 (1998). doi: 10.1016/s0015-0282(98)00384-7 \n59. Matalliotakis, I.M., Vassiliadis, S., Goumenou, A.G., Athanassakis, I., Koumantakis, \nG.E., Neonaki, M.A., Koumantakis, E.E.: Soluble ICAM-1 levels in the serum of \nendometriotic patients appear to be independent of medical treatment. Journal of \nReproductive Immunology 51(1), 9–19 (2001). doi: 10.1016/s0165-0378(01)00064-x \n60. Placido, G. de, Alviggi, C., Di Palma, G., Carravetta, C., Matarese, G., Landino, G., \nRacioppi, L.: Serum concentrations of soluble human leukocyte class I antigens and of \nthe soluble intercellular adhesion molecule-1 in endometriosis: relationship with stage \nand non-pigmented peritoneal lesions. Human Reproduction 13(11), 3206–3210 (1998). \ndoi: 10.1093/humrep/13.11.3206 \n61. Zhou, X., Fragala, M.S., McElhaney, J.E., Kuchel, G.A.: Conceptual and methodological \nissues relevant to cytokine and inflammatory marker measurements in clinical research. \nCurrent opinion in clinical nutrition and metabolic care 13(5), 541–547 (2010). doi: \n10.1097/MCO.0b013e32833cf3bc \n62. Karsten, E., Breen, E., Herbert, B.R.: Red blood cells are dynamic reservoirs of \ncytokines. Sci Rep 8(1), 3101 (2018). doi: 10.1038/s41598-018-21387-w \n63. Luckow Invitti, A., Schor, E., Martins Parreira, R., Kopelman, A., Kamergorodsky, G., \nGonçalves, G.A., Batista Castello Girão, M.J.: Inflammatory cytokine profile of \nco\n/i1cultivated primary cells from the endometrium of women with and without \nendometriosis. Molecular medicine reports 18(2), 1287–1296 (2018). doi: \n10.3892/mmr.2018.9137 \n64. Koks, C.A., Dunselman, G.A., Goeij, A.F. de, Arends, J.W., Evers, J.L.: Evaluation of a \nmenstrual cup to collect shed endometrium for in vitro studies. Fertility and Sterility \n68(3), 560–564 (1997). doi: 10.1016/S0015-0282(97)00250-1 \n65. Weinberg, A.: Cryopreservation of Peripheral Blood Mononuclear Cells. In: Detrick, B., \nSchmitz, J.L., Hamilton, R.G. (eds.) Manual of molecular and clinical laboratory \nimmunology, pp. 263–268. ASM Press, Washington, DC (2016) \n66. Hughes, S.M., Shu, Z., Levy, C.N., Ferre, A.L., Hartig, H., Fang, C., Lentz, G., Fialkow, \nM., Kirby, A.C., Adams, W.K.M., Veazey, R.S., Germann, A., von, B.H., McElrath, M.J., \nDezzutti, C.S., Sinclair, E., Baker, C.A., Shacklett, B.L., Gao, D., Hladik, F.: \nCryopreservation of Human Mucosal Leukocytes. PloS one 11(5) (2016). doi: \n10.1371/journal.pone.0156293 \n \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n30 \n \n \n  \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n31 \n \nTable \nTable 1: Summary of cell frequencies in menstrual effluent (ME) and peripheral blood (PB) of \nendometriosis patients and healthy controls, determined by flow cytometry.  \n Endometriosis Healthy controls p-\nvalue \n(Endo \nvs. \ncontrol\ns)* \nAll participants \n Median % \n(IQR) \np-\nvalue \n(ME \nvs. \nPB)** \nMedian \n(IQR) \np-\nvalue \n(ME \nvs. \nPB)** \nMedian % \n(IQR) \np-value \n(ME vs. \nPB)** \nMain cell types (as % of mononuclear cells) \nB cells in ME 5.6 (7.1) \n0.151 \n7.2 (7.8) \n0.164 \n0.722 5.9 (7.9) \n0.032 \nB cells in PB 4.7 (1.5) 4.7 (1.9) 0.974 4.7 (1.9) \nT cells in ME 44.2 \n(26.2) \n0.042 \n39.6 \n(26.2) \n0.012 \n0.497 44.1 (28.1) \n0.001 \nT cells in PB 68.2 (8.9) 64.6 \n(12.3) \n0.821 66.0 (12.0) \nNK cells in ME 10.0 (6.0) \n0.301 \n15.2 \n(21.6) 0.496 \n0.228 11.5 (16.7) \n0.891 \nNK cells in B 11.0 (3.2) 13.4 (5.4) 0.821 11.3 (4.3) \nMonocytes ME 14.9 \n(24.5) 0.424 \n14.9 \n(17.4) 0.734 \n0.872 14.9 (22.6) \n0.337 \nMonocytes PB 12.8 (6.5) 13.4 (5.4) 0.872 13.0 (6.4) \nT cells \nCD4/CD8 ratio \nME \n2.5 (0.9) \n0.266 \n3.4 (6.5) \n0.074 \n0.314 2.6 (1.8) \n0.040 \nCD4/CD8 ratio \nPB \n2.4 (0.5) 2.0 (0.5) 0.085 2.2 (0.8) \nCD8 + T cells (% \nof CD3 + T) in ME \n26.7 (7.4) \n0.170 \n18.9 \n(20.0) \n0.039 \n0.197 26.0 (12.6) \n0.024 CD8 + T cells (% \nof CD3 + T cells \nPB \n27.5 (3.9) 31.5 (2.8) 0.084 29.6 (5.5) \nCD69+ T cells ME 53.6 \n(21.5) 0.001 \n43.0 \n(28.7) 0.003 \n0.6744 48.2 (26.4) \n< 0.001 \nCD69+ T cells PB 8.5 (8.2) 12.1 (7.0) 0.1802 11.1 (8.9) \nPerforinhigh  \n(% of CD8+ T) in \nME \n4.0 (4.8) \n<0.001 \n11.2 \n(10.4) \n0.039 \n0.029 5.3 (8.0) \n0.024 \nPerforinhigh (% of \nCD8+ T) PB \n9.0 (6.6) 12.9 \n(17.3) \n0.751 9.0 (14.8) \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint \n\n32 \n \nTregs (% of CD4+ \nT) in ME \n0.9 (1.3) \n0.034 \n0.3 (0.4) \n0.004 \n0.342 0.5 (1.3) \n< 0.001 \nTregs (% of CD4+ \nT) in PB \n5.4 (1.1) 4.5 (1.3) 0.197 5.0 (1.5) \nNK cells \nCD56bright/CD16dim \nNK cells ME \n68.0 \n(46.1) \n<0.001 \n65.7 \n(50.5) \n0.004 \n0.974 67.8 (50.3) \n< 0.001 \nCD56bright/CD16dim \nNK cells PB \n3.9 (1.7) 3.9 (2.9) 3.9 (2.9) \nCD56low/Perforinhi\ngh NK cells ME \n52.7 \n(49.8) \n<0.001 \n36.2 \n(58.2) \n0.004 \n1 \n0.251 \n40.4 (51.2) \n< 0.001 \nCD56low/Perforinhi\ngh NK cells PB \n92.2 (4.1) 94.9 (3.2) 92.6 (4.9) \nNKp46+ NK cells \nME \n61.1 \n(19.9) \n<0.001 \n53.2 \n(20.6) \n0.004 \n0.346 58.4 (22.5) \n< 0.001 \nNKp46+ NK cells \nPB \n91.8 (7.0) 85.1 \n(16.6) \n0.418 91.5 (13.9) \nCD69+ NK cells \nME \n79.5 \n(15.4) \n<0.001 \n84.5 \n(13.7) \n0.004 \n0.314 81.0 (16.3) \n< 0.001 \nCD69+ NK cells \nPB \n12.3 \n(10.6) \n18.5 \n(14.5) \n0.107 15.4 (15.3) \nMyeloid cells (no statistical comparison between ME and PB) \nCD14high/CD16low \nME \n63.3 \n(57.5) \n 46.0 \n(37.6) \n 0.418 57.9 (54.5) \nCD14low/CD16high \nME \n26.6 \n(45.0) \n38.7 \n(26.1) \n0.539 36.2 (39.1) \nClassical \nmonocytes PB \n79.9 \n(11.5) \n82.4 (7.1) 0.372 82 (8.1) \nIntermediate \nmonocytes PB \n4.2 (2.9) 4.4 (1.2) 1 4.3 (2.1) \nNon-classical \nmonocytes PB \n14.7 (8.4) 11.9 (6.0) 0.268 13.4 (8.2) \n* Mann-Whitney-U test \n** Wilcoxon rank-sum test \nAll rights reserved. No reuse allowed without permission. \n(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. \nThe copyright holder for this preprintthis version posted January 28, 2021. ; https://doi.org/10.1101/2020.12.03.20243436doi: medRxiv preprint","source_license":"CC0","license_restricted":false}