{"paper_id":"4132febc-7249-45d9-bb90-79ca4e0c7658","body_text":"Endometriosis is defined as the presence of\ntissues which somewhat resembles endometrial\nglands and stroma outside the uterine cavity,\nmost commonly implanted over visceral and\nperitoneal surfaces within the female pelvis.\nEndometriosis exhibits disturbances of cellular\nproliferation, cellular invasion and neoangiogenesis\n( 1 ). Although the exact prevalence of\nendometriosis in the general population is not\nclear, the prevalence in women of reproductive\nage is estimated to range between 10 and 15% ( 2 ). Endometriosis is a chronic, benign, oestrogen-\ndependent multifactorial and gynaecological\ndisease, with pain being the most common\nand specific symptom. To date, the cardinal\ntreatments for endometriosis are medical and\nsurgical therapies. Pain symptoms may persist\ndespite seeming adequate medical or/and surgical\ntreatment of the disease ( 3 ).\nStem cell therapy as a promising and unprecedented\nstrategy has the potential to be more\neffective than single-agent drug therapies ( 4 ).\nMesenchymal stem cells (MSCs) are especially\nwell suited for cell therapy owing to their ability\nto differentiate into different lineages and\nsecrete a number of cytokines ( 5 ). Human umbilical\ncord-MSCs (hUC-MSCs) have become\nstrong candidates for a cell-based therapy because\nof their key characteristics of long-term\nself-renewal and capacity to differentiate into\ndiverse tissues. In addition, they can be easily\nobtained and cultured without raising ethical issues\n( 6 ), as well as being an excellent alternative\nto bone marrow as a source of MSCs for\ncell therapies ( 6 ,  7 ). Furthermore, hUC-MSCs\nare a subset of primitive stem cells. HUC-MSCs\nneither induce teratomas nor result in acute rejection\nafter being transplanted into non-immune-\nsuppressed animals ( 8 ). In various animal\ndisease models, transplantation of hUC-MSCs\nwas reported to improve neurobehavioral functions\nfollowing ischemic stroke ( 9 ), ameliorate\nmouse hepatic injury ( 10 ), and show effectiveness\nin apomorphine-induced rotations in\na rodent model of Parkinson’s disease ( 6 ,  11 ).\nNevertheless, currently little is known about the\napplication of hUC-MSCs to endometriosis.\nSome researchers have identified nerve fibers\nin endometriotic lesions in women with endometriosis\n( 12 - 14 ). Berkley et al. ( 15 ) and\nOosterlynck et al. ( 16 ) have reported that endometriotic\nimplants developed a sensory and\nsympathetic nerve supply both in rats and in\nwomen, similar to that of the healthy rat uterus.\nThe present study demonstrated the existence of\na much greater density of nerve fibers in deep\ninfiltrating endometriosis than in peritoneal endometriotic\nlesions ( 17 ). These nerve fibers in\nendometriotic lesions could possibly exert their\nfunctions on the pathogenesis or symptoms of\nendometriosis.\nAs a consequence, we established surgically\ninduced endometriosis in a rat model to investigate\nthe effects of the hUC-MSCs transplantation\non nerve fibers and the pathogenesis of the\ndisease.\n\nThe study protocol was approved by the Research\nEthics Committee of Qilu Hospital of\nShandong University (Shandong, P. R. China).\nHUCs (n=10, clinically normal pregnancies)\nwere excised and washed in a 0.1 mol/L phosphate\nbuffer saline (PBS, pH=7.4, Gibco-BRL,\nGrand Island, NY, USA) to remove excess blood\n( 6 ). The cords were dissected and the blood\nvessels were removed. The remaining tissues\nwere cut into small pieces (1-2 mm 3 ) and placed\nin plates with low-glucose Dulbecco-modified\nEagle medium (L-DMEM, Gibco-BRL, Grand\nIsland, NY, USA), supplemented with 10% fetal\nbovine serum (FBS, Gibco-BRL, Grand Island,\nNY, USA), 2 ng/mL vascular endothelial\ngrowth factor (VEGF, R&D Systems, Minneapolis,\nMN), 2 ng/mL epidermal growth factor\n(EGF, R&D Systems, Minneapolis, USA), 2 ng/\nmL fibroblast growth factor (FGF, R&D Systems,\nMinneapolis, USA), 100 U/mL penicillin,\nand 100 μg/mL streptomycin (Gibco-BRL,\nGrand Island, USA). Cultures were maintained\nat 37˚C in a humidified atmosphere with 5%\nCO 2 . The media were changed every 3-4 days.\nAdherent cells proliferated from individual explanted\ntissues 7-12 days after initiating incubation.\nAt this time, the small tissue pieces were\nremoved from the culture and the adherent fibroblast-\nlike cells were cultured to confluence,\nwhich subsequently took 2-3 weeks in culture.\nThe cells were then trypsinized using 0.25%\ntrypsin (Gibco-BRL, Grand Island, USA) and\npassaged at 1×10 4  cells/cm 2  in the medium described\nabove. The cells were used after five or\nmore passages.\nFifth- to seventh-passage cells were collected\nand treated with 0.25% trypsin. The cells were stained with either fluorescein isothiocyanate-\nconjugated or phycoerythrin-conjugated\nmonoclonal antibodies in 100 μL PBS for 15\nminutes at room temperature, as suggested by\nthe manufacturer. The antibodies used were\nagainst human antigens cluster of differentiation\n34 (CD34), CD29, CD44, CD45, CD105,\nand CD106 (SeroTec, Raleigh, NC, USA). Cells\nwere analyzed using flow cytometry (Cytometer\n1.0, CytomicsTM FC500, Beckman Coulter\nInc., USA). Positive cells were counted and\ncompared to the signal of corresponding immunoglobulin\nisotypes.\nTo investigate the differentiation potential of the\nfibroblast-like cells, the fourth passage cells were\ncultured under conditions appropriate for inducing\nthe differentiation of each lineage.\nCells were seeded at a density of 2×10 4  cells/\ncm 2  and the differentiation media were changed\nevery 3-4 days. The osteogenic differentiation\nmedium consisted of L-DMEM supplemented\nwith 10% FBS, 0.1 μM dexamethasone, 50\nmM β-glycerol phosphate, and 0.2 mM ascorbic\nacid (Sigma-Aldrich, St. Louis, MO, USA).\nThe adipogenic differentiation medium consisted\nof high-glucose DMEM supplemented with\n0.25 mM 3-isobutyl-1-methylxanthine, 0.1 μM\ndexamethasone, 0.1 mM indomethacin (Sigma-\nAldrich, USA), 6.25 μg/mL insulin (PeproTech,\nUK), and 10% FBS. Cells were kept in the normal\ngrowth medium served as the control.\nAll animal procedures were conducted in accordance\nwith the institutional guidelines of\nQilu Hospital of Shandong University (Shandong,\nP. R. China). Adult female Wistar rats,\nweighing 180-210 g, were housed in cages in\nan air-conditioned room at 25 ±1˚C with a 12\nhours dark/light cycle. The oestrous stage was\nmonitored daily by vaginal smear every morning,\nbeginning at least 2 weeks before surgery\nand continued until the day of death. Only rats\nwith a regular 4-day cycle both before and after\nsurgery were used. Surgically induced endometriosis\nin a rat model was done as previously\ndescribed ( 18 ) and surgery was done under\naseptic precautions. Rats in estrus were anesthetized\nwith 3% pelltobarbitalum natricum\n(Solarbio, Beijing Solarbio Science & Technology\nCo., Ltd. China) at a dose of 0.2 mL /200 g\nby means of intraperitoneal injection. A midline\nabdominal incision exposed the uterus, and a\n1-cm segment of the middle of the left uterine\nhorn was removed and placed in warm sterile\nsaline. Four pieces of uterine horn (≈2×2 mm)\nwere cut from this segment and sewn with 4.0\nnylon sutures around the alternate cascade mesenteric\narteries that supply the caudal small intestine,\nstarting from the caecum. The incision\nwas closed in layers, and the rats were allowed\nto recover from anesthesia under close observation.\nHereafter the endometriosis model rats\nwere randomly divided into two groups (12 rats\neach), namely the treatment group and the control\ngroup. Two weeks later, the treatment group\nreceived hUC-MSCs by injection of 1×10 6  cell/\nmL normal saline into the tail vein every 5 days\nfor 15 days. Meanwhile, the control group only\nreceived the same volume of normal saline.\nFour weeks later, ectopic implants were collected\nand fixed in 10% neutral buffered formalin\nfor 18~24 hours.\nWe examined the presence of different types\nof nerve fibers in endometriotic implants in a rat\nmodel by immunohistochemistry using highly\nspecific markers. We used neurofilament (NF),\nnerve growth factor (NGF), NGF receptor p75\n(NGFRp75), tyrosine kinase receptor-A (Trk-\nA), calcitonin gene-related peptide (CGRP) and\nsubstance P (SP) to differentiate types of nerve\nfibers.\nThese implants were fixed with formalin, processed\nand embedded in paraffin according to a\nstandard protocol. Each section was cut at 4 um\nand mounted onto slides. These sections were\nroutinely stained with haematoxylin and eosin\n(H&E, Gibco-BRL, Grand Island, NY, USA)\nstaining. For immunohistochemistry, the slides\nwere submitted to antigen retrieval by boiling\nin citrate buffer (0.01 mol/L, pH=6.0) for 15 minutes using a micro-wave oven.\nEndogenous peroxidase activity was prevented\nby incubating in 0.3% hydrogen peroxide for\n15 minutes. Nonspecific binding was blocked\nby 10% goat serum (Zhongshan Golden Bridge\nBiotecnology Co., Ltd., China) for 20 minutes\nat room temperature. The sections were immunostained\novernight at 4˚C using antibodies for\nmonoclonal mouse anti-NF (dilution 1:150; Abcam,\nUK), a highly specific marker for myelinated\nnerve fibers, as follows: polyclonal rabbit\nanti-NGF (dilution 1:200; Abcam, UK), monoclonal\nmouse anti-NGFRp75 (dilution 1:200;\nAbcam, UK), polyclonal rabbit anti-TrkA (dilution\n1:500; Abcam, UK), polyclonal mouse\nanti-SP (dilution 1:250), and polyclonal rabbit\nanti-CGRP (dilution 1:300, Abcam, UK), which\nare sensory fiber markers, and they can be present\nin both Ad and C nerve fibers. The slides\nwere washed and incubated with horseradish\nperoxidase-conjugated secondary antibody at\nroom temperature for 30 minutes.\nPeroxidase activity was visualized by exposure\nto diaminobenzidine tetrahydrochloride\nsolution (DAB kit, Zhongshan Golden Bridge\nBiotecnology Co., Ltd., China) for 3-5 minutes.\nThe sections were then washed, counterstained\nwith hematoxylin for 1 minute, dehydrated, and\nmounted with coverslips. We used normal rat\nskin as a positive control as it reliably contains\nmyelinated and unmyelinated nerve fibers expressing\nNF, NGF, NGFRp75, Trk-A, SP, and\nCGRP.\nThe images were captured using an Olympus\nDP72 camera (Tokyo, Japan). The assessment\nof the mean density of nerve fibers was\nperformed by Image Pro Plus software (Media\nCybernetics, MD, USA). The integrated optical\ndensity (IOD) and area of the images were\ncalculated using Image Pro Plus software. The\narea was divided by integrated optical density\nto obtain the mean density of nerve fibers. All\nlighting conditions and magnifications were\nheld constant. Moreover, the investigator was\nunaware of the experimental groups from which\nthe slices were obtained.\nThe results were expressed as the mean±SD.\nAll analyses were performed using the SPSS\n(SPSS Inc., Chicago, IL, USA) version 17.0. The\ncomparison between two groups was performed\nusing non-parametric 2-tailed t test (Mann-Whitney\ntest). Statistical significance was defined as a\np value of less than 0.5.\n\nAfter several passages, adherent cells from\nUC could form a monolayer of typical fibroblastic\ncells. Flow cytometry results showed\nthat UC-derived cells shared most of their immunophenotype\nwith MSCs, including positive\nexpression for stromal markers (CD29, CD44,\nCD105, and CD106), but negative expression\nfor hematopoietic markers (CD34 and CD45)\n( Fig.1A, B ).\nMSC differentiation was assessed using the\nfourth passage cells. When being induced to\ndifferentiate under osteogenic conditions, MSC\ncongregation increased with increasing induction\ntime and formed a mineralized matrix, as\nconfirmed by alizarin red staining ( Fig.1C ).\nMost of the MSC-like cells became alkalinephosphatase-\npositive by the end of 14 days\n( Fig.1D ). No mineralized matrix was observed\nin the control cells kept in the normal growth\nmedium. The spindle shape of the MSCs flattened\nand broadened after 1 week of adipogenic\ninduction. Small oil droplets gradually appeared\nin the cytoplasm. By the end of the second\nweek, almost all cells contained numerous\noil-red-O-positive lipid droplets ( Fig.1E ). The\ncontrol cells maintained in the regular growth\nmedium did not stain positive for oil red O.\nThe mean density values of nerve fibers are\ngiven in  table 1 . Nerve fibers stained with kinds\nof special markers in ectopic endometriotic lesions\nwere shown in  figure 2 . In summary, there\nwere significant differences (p<0.05) in the\nmean density of nerve fibers in endometriotic\nimplants stained with most of the specific markers\nwhich we used between the treatment group\nand the control group.\nHUC-derived MSC-like cells in passaged cultures. Immunophenotype (A) and H&E staining of UC-derived MSC-like cells (B). Osteogenic\ndifferentiation as indicated by the formation of mineralized matrix shown by alizarin red staining (C) and alkaline phosphatase expression\n(D). Adipocytic differentiation was noted by the presence of broadened morphology and formation of lipid vacuoles (E) (positive\noil-red O staining). Scale bars=80 μm.\nhUC; Human umbilical cord, MSCs; Mesenchymal stem cells, H&E; Haematoxylin and eosin and CD; Cluster of differentiation.\nQuantitative assessment of the endometrial mean nerve fiber density stained against\ndifferent neural markers in model rat with endometriosis\nData are represented by mean density±SD.\nNF; Neurofilament, NGF; Nerve growth factor, Trk-A; Tyrosine kinase receptor-A, NGFRp75; NGF receptor\np75, CGRP; Calcitonin gene-related peptide, SP; Substance P, *; P<0.01 and **; P<0.001.\nNerve fibers in ectopic endometriotic lesions. A. Nerve fibers stained with NF from the control group without the transplantation\nof hUC-MSCs. B. Nerve fibers stained with NF from the treatment group with the transplantation of hUC-MSCs. C. Nerve fibers stained\nwith NGF from the control group. D. Nerve fibers stained with NGF from the treatment group. E. Nerve fibers stained with Trk-A from the\ncontrol group. F. Nerve fibers stained with Trk-A from the treatment group. G. Nerve fibers stained with NGFRp75 from the control group.\nH. Nerve fibers stained with NGFRp75 from the treatment group. I. Nerve fibers stained with CGRP from the control group. J. Nerve fibers\nstained with CGRP from the treatment group. K. Nerve fibers stained with SP from the control group. L. Nerve fibers stained with SP\nfrom the treatment group. Scale bars represent 50 μm in A-R (magnification ×200). Black arrows represent nerve fibers and yellow arrows\nrepresent endometrial glands.\n\nNF as a highly specific marker for myelinated\nnerve fibers stains Aα, Aβ, Aγ, Aδ and B fibers.\nBoth SP and CGRP are sensory nerve fiber markers\nthat can be present in both Aδ and C nerve fibers.\nIn the present study, statistically significant difference\nwas observed in the mean density of the NFimmunoactive\nnerve fibers between the treatment\nand control groups. Lower number of nerve fibers\nstained with NGF, TrkA, and NGFRp75 existed\nin the treatment group than in the control group.\nThe mean densities of the CGRP- and SP- immunoreactive\nnerve fibers were lower in the treatment\ngroup, which indicates that the sensory nerve fibers\nwere reduced. To sum up, our results showed\nthat there were less nerve fibers stained with most\nof the specific markers used in this study in the\ntreatment group compared with the control group.\nIt is believed that rich innervation in endometriosis\nmay be involved in pain generation\n( 17 ,  19 ). Patients with the highest pain scores\ndisplayed significantly more neural invasion into\nendometriosis than those with lower pain scores\n( 20 ). Therefore, less innervation may ameliorate\nthe symptoms of disease. Tokushige et al. ( 21 )\nreported that the nerve fiber density in peritoneal\nendometriotic lesions from women with endometriosis\nwho were on hormone treatment with progestogens\nand combined oral contraceptives was\nstatistically significantly lower than in peritoneal\nendometriotic lesions from untreated women\nwith endometriosis. In the present study, our results\nshowed that there was lower number of nerve\nfibers in the treatment group, which is consistent\nwith the findings of previous studies.\nThe pathogenesis of endometriosis and pathophysiological\nbasis for endometriosis–associated\npain are still unclear. Endometriosis is believed to\nbe a chronic in.ammatory state, with disturbances\nof both cell-mediated and humoral immunity ( 16 ).\nIn women with endometriosis, the peritoneal .uid\nhas high concentrations of cytokines, growth factors,\nand angiogenic factors ( 16 ,  22 - 24 ), derived\nfrom the lesions themselves; secretory products of\nmacrophages and other immune cells; and follicular\n.uid after follicle rupture in ovulating\nwomen. Once endometriotic lesions are formed,\nthey secrete several pro-in.ammatory molecules\n( 23 ,  24 ).\nThese nerve fibers in endometriotic lesions\nprobably play an important role in the pathogenesis\nof pain and hyperalgesia. The nerve\nendings of nerve fibers can potentially be\nstimulated by many inflammatory substances,\nincluding histamine, serotonin, bradykinin,\nprostaglandins, leukotrienes, interleukins (ILs),\nacetylcholine, VEGF, tumor necrosis factor-α\n(TNF-α), epidermal growth factors, transforming\ngrowth factor-β (TGF-β), platelet-derived\ngrowth factor and NGF. Many of the above substances\ncan be secreted by macrophages as well\nas from endometriotic lesions ( 25 ,  26 ), and are\nfound in high concentration in the peritoneal\nfluid of endometriosis patients. Moreover, macrophages\nand their products may play important\nroles in the growth and repair of nerve fibers.\nThe growth of nerve fibers is regulated by many\nsubstances, including NGF, brain-derived neurotropic\nfactor (BDNF) and VEGF, and the synthesis\nof these substances is also affected by\nmacrophage activities.\nHMSCs, first described by Fridenstein et al. ( 27 )\nin 1974, have extensive proliferative potential and\nthe capacity to differentiate into various cell types.\nThe bone marrow has been considered as the\nmajor source of MSCs. Transplantation of bone\nmarrow-MSCs (BM-MSCs), however, may not be\nacceptable because of the variations in cell numbers\nand the proliferative potential of these cells\nfrom different donors ( 28 ). Other sources of MSCs\nhave been considered and currently the presence\nof MSCs has been confirmed in the placenta, amniotic\nfluid, peripheral blood, lungs and teeth ( 29 ).\nBecause there are large numbers of MSCs in neonates\n( 30 ), human umbilical cords may be an ideal\nsource for these cells. Supporting their potential as\na source of cells, MSCs have been isolated from\nhuman umbilical cord ( 9 ,  27 ,  31 ). MSCs are poor\nantigen-presenting cells and do not express major\nhistocompatibility complex class II or costimulatory\nmolecules. HMSCs suppress T-lymphocyte\nproliferation induced by cellular or non-specific\nmitogenic stimuli and inhibit the response of\nnaive and memory antigen-specific T cells to their\ncognate peptide ( 32 ). HMSCs altered the cytokine\nsecretion profile of dendritic cells (DCs), naive\nand effector T cells [T helper 1 (T(h)1) and T(h)2],\nand natural killer (NK) cells to induce a more antiinflammatory\nor tolerant phenotype ( 33 ). MSCs\nhave potent anti-inflammatory effects in multiple disease states ( 34 ). Some researchers have reported\nthat MSCs administered by intravenous injections\npotently inhibit systemic levels of inflammatory\ncytokines and chemokines in the serum of treated\nanimals ( 35 ). In addition, MSCs were able to\nmodulate the immune system through the release\nof anti-inflammatory cytokines, prostaglandin E2\nin many models ( 36 ). Aggarwal and Pittenger ( 33 )\nreported that through the interactions of hMSCs\nwith the various immune cells, hMSCs could inhibit\nor limit inflammatory responses and promote\nthe mitigating and anti-inflammatory pathways.\nThey demonstrated that when hMSCs are present\nin an inflammatory environment (such as that\nartificially created by activating DCs, macrophages,\nNK cells, or T cells using various stimuli), they\nmay alter the outcome of the on-going immune\nresponse by altering the cytokine secretion profile\nof DC subsets (DC1 and DC2) and T-cell subsets\n(TH1, TH2, or TRegs), thereby resulting in a shift\nfrom a proinflammatory environment toward an\nanti-inflammatory or tolerant cell environment.\nThere was significantly lower number of nerve\nfibers stained with specific markers we used in\nthe treatment group than in the control group.\nEndometriosis is a benign oestrogen-dependent\ninflammatory disease and hUC-MSCs could attune\ninflammatory effects of inflammatory factors\nsuch as cytokines, growth factors, and angiogenic\nfactors. Other underlying mechanisms such as\nthe differentiation of hUC-MSCs and/or the paracrine\nmediator secreted by hUC-MSCs may be\nalso involved. A recent study also suggested that\nhUC -MSCs may serve as a promising treatment\napproach to ameliorate endometrial damage ( 37 ).\nOur study was the preliminary exploration of hUC\n–MSC treatment with endometriosis. The exact\nmechanism and outcome of hUC-MSCs remain to\nbe elucidated in future studies.\n\nWe demonstrated that hUC-MSCs could reduce\nnerve fibers density in the treatment group and\nmay provide a new potential therapeutic modality\nto endometriosis.","source_license":"CC0","license_restricted":false}