{"paper_id":"16caba06-93ee-46d0-aa65-aeefd16bfebf","body_text":"Vol.:(0123456789)1 3\nReproductive Sciences (2023) 30:2457–2467 \nhttps://doi.org/10.1007/s43032-023-01181-4\nENDOMETRIOSIS: ORIGINAL ARTICLE\nGenome‑Wide Gene Expression Profiling Reveals the Direct Effect \nof Dienogest on Ovarian Endometriotic Stromal Cells\nHiroshi Honda1  · Norihisa Nishimichi2 · Mayumi Kaneko3 · Michinori Yamashita1,4 · Yumiko Akimoto1,5 · \nHirotoshi Tanimoto1,5 · Mitsue Teramoto1,6 · Hideki Teramoto1,7 · Yasuyuki Yokosaki2\nReceived: 1 March 2022 / Accepted: 24 January 2023 / Published online: 8 February 2023 \n© The Author(s) 2023\nAbstract\nEndometriosis affects up to 10% of women of reproductive age, causing dysmenorrhea, chronic pelvic pain, and infertil-\nity. The current key drug for endometriosis is dienogest, a progestin with high specificity for the progesterone receptor. To \nreveal the direct anti-endometriotic effect of dienogest on ovarian endometriotic cells, we investigated the genome-wide gene \nexpression profiles of ovarian endometriotic stromal cells with (Dienogest group) or without dienogest treatment (Control \ngroup) and compared the groups’ gene expression profiles. We performed a gene ontology (GO) analysis and Ingenuity \npathway analysis using these data. To validate the microarray data, we performed real-time RT-PCRs and immunohisto-\nchemistry for the differentially expressed genes between the two groups. Of 647 genes differentially expressed between the \ntwo groups, 314 genes were upregulated and 333 were downregulated in the Dienogest group versus the Control group. The \nGO analysis showed that the regulation of macrophage chemotaxis, the collagen catabolic process, and the proteoglycan \nbiosynthetic process are the main biological processes closely associated with the differentially expressed genes. We identi-\nfied 20 canonical pathways that were most significantly differentially expressed in the Dienogest group versus the Control \ngroup. We observed that matrix metalloproteinases (MMPs) are the genes in these pathways that are most closely associated \nwith dienogest treatment. Of components involved in the regulation of macrophage chemotaxis, colony-stimulating factor \n1 and macrophage-stimulating 1 are potential upstream regulators of MMPs and were observed herein to be suppressed by \ndienogest. Our results suggest that dienogest may thus exert its anti-endometriotic effect by directly suppressing MMPs.\nKeywords Endometriosis · Dienogest · Microarray · Gene ontology · Ingenuity pathway analysis · MMP\nAbbreviations\nESC  endometriotic stromal cell\nGnRH  gonadotropin-releasing hormone\nIPA  ingenuity pathway analysis\nMMP  matrix metalloproteinase\nTIMP  tissue inhibitor of metalloproteinase\nCSF1  colony-stimulating factor 1\nMST1  macrophage stimulating 1\nIntroduction\nEndometriosis, which is defined by the presence of ectopic \nendometrial tissue outside the uterus, is a chronic disease \naffecting up to 10% of women of reproductive age [1 ]. \nInfertility and chronic pelvic pain are the main symptoms \nof endometriosis. For chronic pelvic pain, mainly hormonal \ntherapies have been provided in the past decades, particu-\nlarly therapies involving gonadotropin-releasing hormone \n * Hiroshi Honda \n h-honda@qg8.so-net.ne.jp\n1 Department of Obstetrics and Gynecology, Hiroshima \nCity North Medical Center Asa Citizens Hospital, 1-2-1 \nKameyamaminami, Asakita-ku, Hiroshima 731-0293, Japan\n2 Integrin-Matrix Biomedical Science, Translational Research \nCenter, Hiroshima University, Hiroshima, Japan\n3 Department of Diagnostic Pathology, Hiroshima City North \nMedical Center Asa Citizens Hospital, Hiroshima, Japan\n4 Department of Surgery and Palliative Medicine, Fujita \nHealth University School of Medicine, Toyoake, Japan\n5 Sumire Women’s Clinic, Hiroshima, Japan\n6 Department of Obstetrics and Gynecology, Sera Central \nHospital, Sera, Hiroshima, Japan\n7 Department of Obstetrics and Gynecology, Shobara Red \nCross Hospital, Shobara, Japan\n\n2458 Reproductive Sciences (2023) 30:2457–2467\n1 3\n(GnRH) analogs. However, GnRH analogs cannot be used \nfor a long period because of their side effects, such as bone \nloss. It has thus been difficult for individuals with chronic \npelvic pain due to endometriosis to obtain relief from pain \nover a long period by using a GnRH analog.\nDienogest is classified as a fourth-generation progestin \nwith high specificity for the progesterone receptor. It has \nan effect on endometriosis-induced pain that is equivalent \nto that of a GnRH analog [2 ]. However, unlike GnRH ana-\nlogs, dienogest can be used for long periods and has brought \nendometriosis patients long-term relief from pain [3]. Dien-\nogest has thus replaced GnRH analogs as the key drug of \nhormonal therapies for endometriosis. The anti-endome-\ntriotic effect of dienogest is thought to be due mainly to \nits ability to suppress ovulation [4 , 5], and several studies \nhave shown that dienogest directly inhibits the inflammatory \nresponses or aromatase expression in endometriotic cells \n[6–8]. Another investigation indicated that dienogest directly \ninhibits the proliferation of ovarian endometriotic stromal \ncells (ESCs) [9]. Although many studies have attempted to \nreveal the biological mechanisms that underlie the effects of \ndienogest on endometriosis, the precise mechanisms remain \nunknown.\nGenome-wide gene expression profiling, using a microar-\nray, and its subsequent pathway analysis have revealed novel \nbiological cell-related findings for several diseases. How -\never, to the best of our knowledge, no reports of genome-\nwide gene expression profiling to investigate the biologi-\ncal actions of dienogest in endometriotic cells have been \npublished.\nWe conducted the present study to determine the genome-\nwide gene expression profile of ovarian ESCs treated with \ndienogest. We also performed a pathway analysis using the \ndata of the gene expression profile. The results demonstrated \nthat, in ovarian ESCs, matrix metalloproteinases (MMPs) \nwere particularly important among the genes that are directly \nmodified by dienogest.\nMaterials and Methods\nPatients and Samples\nTissue specimens of ovarian endometriotic cysts were \nobtained during gynecological surgeries from 15 patients \n(5 for the present microarray analysis and 10 for immuno-\nhistochemistry) with stage III–IV endometriosis evaluated \nin accord with the American Society for Reproductive Medi-\ncine classification of endometriosis. Before the surgeries, the \npatients provided written informed consent for their mate-\nrials to be used. All 15 patients were of reproductive age \n(age range 24–46 years old, mean ± SD: 34.9 ± 8.5 years \nold) with a body mass index in the normal range (mean ± \nSD: 21.3 ± 2.6), had regular menstrual cycles, and were \nclinically and pathologically confirmed to have no gyneco-\nlogical disease other than endometriosis. None of the five \npatients for the microarray analysis had received any hormo-\nnal therapies and, of the 10 patients whose specimens were \nused for immunohistochemistry, five had also not received \nany hormonal therapies; the other five patients had received \ndienogest in oral doses (1 mg twice a day) for 12–18 weeks \n(mean ± SD: 15 ± 2.2 weeks) months until the surgery. All \nspecimens were pathologically confirmed as ovarian endo-\nmetriotic cyst tissues after the surgeries.\nThe protocol of the present study was approved by the \nEthical Committee of Hiroshima City Asa Citizens Hospital, \nand all experiments were carried out in accord with relevant \nguidelines and regulations.\nIsolation and Cell Culture of Ovarian ESCs\nThe isolation of ovarian ESCs was performed as described \nby Honda et al. [10]. Briefly, endometriotic tissue layers \nwere scraped from the inner wall of the cyst, minced into \nsmall pieces, and enzymatically dissociated by incubation \nwith 0.25% collagenase (Sigma-Aldrich, St. Louis, MO, \nUSA) and 0.02% DNase I (Sigma-Aldrich) in phenol-red-\nfree Dulbecco’s modified Eagle’s Medium/Ham’s F-12 (Inv-\nitrogen, Carlsbad, CA) supplemented with 10% charcoal-\nstripped fetal bovine serum (FBS) (Invitrogen) for 1 h at 37 \n°C in an atmosphere of 5%  CO2 under magnetically driven \nagitation. Enrichment of the ESCs was performed by serial \nfiltration using 100 μm and 40 μm nylon sieves (BD Falcon, \nFranklin Lakes, NJ), and filtered cells were collected onto \ntwo 6 cm culture dishes per sample. After incubation of the \nfiltered cells at 37 °C for 30 min to allow the ESCs to attach \nto the dishes, the media were removed, and the dishes were \nwashed for complete removal of the floating endometriotic \nepithelial cells and other cells such as blood cells in the \nsupernatant.\nThe ESCs were cultured in DMEM/Ham’s F-12 medium \nsupplemented with 10% charcoal-stripped FBS and 1% \npenicillin and streptomycin (100 mg/ml) (Invitrogen) under \nthe conditions described above. The medium was changed \nevery other day and, when the cells reached 80% conflu-\nence, the culture was serum-starved in serum-free DMEM/\nHam’s F-12 medium before hormone treatment. After 24 h \nof culture, the medium was replaced with either serum-free \nDMEM/Ham’s F-12 medium with estradiol  (10−8  M; Sigma-\nAldrich) alone or estradiol  (10−8  M) + dienogest  (10−6  M; \nSanta Cruz Biotechnology, Dallas, TX). The ESCs treated \nwithout dienogest were used as a control (Control group), \nand those treated with dienogest were allocated to the Dien-\nogest group.\nSince endometriosis is an estrogen-dependent disease, the \nESCs of the Control group were cultivated with estradiol at \n\n2459Reproductive Sciences (2023) 30:2457–2467 \n1 3\na physiological concentration  (10−8  M). The ESCs of the \nDienogest group were cultivated with estradiol at the same \nconcentration as that used for the Control group together \nwith dienogest at a concentration  (10−6  M) matching that in \nprevious studies [11– 13]. After these hormonal treatments \nfor 48 h, the ESCs were directly lysed on the culture dishes \nwith TRIzol™ reagent (Invitrogen), immediately snap-fro-\nzen, and stored at −80 °C until further processing.\nRNA Isolation\nTotal RNA was purified from the cell lysates of ESCs using \nthe RNeasy Mini Kit (Qiagen, Valencia, CA) in accord with \nthe manufacturer’s instructions. The quantity and quality \nof the purified RNA were measured and assessed using a \nNanodrop ND-1000 spectrophotometer (Thermo Fisher Sci-\nentific, Waltham, MA) and an Agilent Bioanalyzer (Agilent \nTechnologies, Santa Clara, CA) before cRNA amplification \nand labeling.\nThe following processes of cRNA amplification and \nlabeling, the sample hybridization, and the microarray data \nanalysis were performed by a slightly modified version of \nthe protocol described by Yokoi et al. [14].\ncRNA Amplification and Labeling\nTotal RNA was amplified and labeled with cyanine 3 (Cy3) \nusing the Agilent Low Input Quick Amp Labeling Kit, one \ncolor (Agilent Technologies), following the manufacturer’s \ninstructions. Briefly, 100 ng of total RNA was reversed-\ntranscribed to double-stranded cDNA using a poly(dT-T7) \npromoter primer. The primer, template RNA, and quality-\ncontrol transcripts of known concentration and quality were \nfirst denatured at 65 °C for 10 min and incubated for 2 h \nat 40 °C with 5× first strand buffer, 0.1 M DTT, 10 mM \ndNTP mix, and AffinityScript RNase Block Mix (Agilent \nTechnologies). The AffinityScript enzyme was inactivated \nat 70 °C for 15 min.\nThe cDNA products were then used as templates for \nin vitro transcription to generate fluorescent cRNA. The \ncDNA products were mixed with a transcription master mix \nin the presence of T7 RNA polymerase and Cy3-labeled \nCTP and incubated at 40 °C for 2 h. Labeled cRNAs were \npurified using Qiagen’s RNeasy mini-spin columns and \neluted in 30 μl of nuclease-free water. After amplification \nand labeling, the cRNA quantity and the incorporation of \ncyanine were determined using the Nanodrop ND-1000 \nspectrophotometer and the Agilent Bioanalyzer.\nSample Hybridization\nFor each hybridization, 600 ng of Cy3-labeled cRNA was \nfragmented and then hybridized at 65 °C for 17 h with an \nAgilent SurePrint G3 Human GE v2 8×60K Microarray \n(Design ID: 039494). After washing, the microarrays were \nscanned using an Agilent DNA microarray scanner.\nAnalysis of Microarray Data for the GO and Pathway \nAnalyses\nThe intensity values of each scanned feature were quanti-\nfied using Agilent Feature Extraction software ver. 10.7.3.1, \nwhich performs background subtractions. We used only fea-\ntures that were flagged as “no errors” (present flags), and \nwe excluded features that were not positive, not significant, \nnot uniform, not above the background, saturated, or popu-\nlation outliers (marginal and absent flags). Normalization \nwas performed using Agilent GeneSpring GX ver. 11.0.2 \nsoftware (per chip: normalization to the 75th percentile \nshift; per gene: normalization to the median of all samples). \nThere are a total of 50,599 probes on the Agilent SurePrint \nG3 Human GE v2 8×60K Microarray (Design ID: 039494) \nwithout control probes. The RNA samples of the Control \ngroup were used as the total RNA reference.\nGenes that were differentially expressed between the Die-\nnogest and Control groups with a corrected p -value < 0.05 \nand absolute fold change > 1.5 were considered to be signifi-\ncantly differentially expressed. The differentially expressed \ngenes were used for a gene ontology (GO) analysis within \nthe category of biological processes by Panther software \nver. 14 (Panther Labs, San Francisco, CA). The microar -\nray data were also used for an Ingenuity® pathway analysis \n(IPA; Ingenuity Systems, www. ingen uity. com) to analyze \ncanonical pathways. Fisher’s exact test was used to calculate \na p-value to determine the significance of findings in the GO \nand Ingenuity® pathway analyses, and p-values < 0.05 were \naccepted as significant.\nQuantitative RT‑PCR\nTotal RNA purified for the microarray analysis was also \nused for a quantitative real-time reverse-transcription poly -\nmerase chain reaction (RT-PCR) to validate the microarray \ndata. All five RNA samples used for the microarray analysis \nwere used for this validation experiment. One microgram of \ntotal RNA was reverse-transcribed into first-strand cDNA \nby using a first-strand cDNA synthesis kit, ReverTra Ace-α \n(Toyobo, Osaka, Japan), with random primers. The cDNA \nwas stored at −20 °C until use.\nOf the genes that were differentially expressed between \nthe Dienogest and Control groups, we selected colony-stim-\nulating factor 1 (CSF1), macrophage stimulating 1 (MST1), \nMMP-1, MMP-3, MMP-10, and tissue inhibitors of metal-\nloproteinase-4 (TIMP-4) for the PCR reactions, which were \nperformed in an ABI 7300 Real-Time PCR System (Applied \nBiosystems, Carlsbad, CA) using the KAPA SYBR® FAST \n\n2460 Reproductive Sciences (2023) 30:2457–2467\n1 3\nqPCR kit (Nippon Genetics, Tokyo) under thermal cycling \nconditions in accord with the manufacturer’s instructions. \nThe primer sequences of CSF1, MST1, MMPs, and TIMP-4 \nused in the present analysis were the same as those designed \nin previous studies [15– 17]. Real-time RT-PCR for the \nhousekeeping gene GAPDH was also performed for all of \nthe samples to evaluate the quality of the cDNAs used. The \nprimers used for the GAPDH quantification were designed \nby Wang et al. [18].\nThe relative quantification of gene expression for each \ngene was performed with the  2−ΔΔC T method described by \nLivak and Schmittgen [19]. Each C T value was averaged \nfor each duplicate, and then the ΔΔC T value for each gene \n((CT.target − C T.GAPDH) Dienogest−  (CT.target − C T.GAPDH) Control) \nwas calculated. The relative expression of each gene in the \nDienogest group is shown as a fold change to that in the \nControl group by using the equation of  2−ΔΔC T.\nImmunohistochemistry (IHC)\nFive ovarian endometriotic cysts obtained from patients \ntreated with dienogest (Dienogest group) and five ovarian \nendometriotic cysts obtained from the patients not treated \nwith dienogest (Control group) were evaluated by IHC to \nvalidate the data from the GO and Ingenuity® pathway \nanalyses. Of the genes extracted by the GO and pathway \nanalyses, CSF1, MST1, MMP-1, and MMP-3 were selected \nfor the IHC analysis. Deparaffinized and rehydrated tissue \nsections were incubated in the PT Link pre-treatment sys -\ntem (Dako Agilent, Santa Clara, CA) for antigen retrieval \nand then processed on Autostainer Link 48 (Dako Agilent) \nin accord with the manufacturer’s protocol. The following \nprimary antibodies were used: CST1 (1:100 dilution, cat.# \nab52864; Abcam, Boston, MA), MST1 (1:50 dilution, cat.# \nHPA024036; Sigma-Aldrich), MMP-1 (1:100 dilution, \ncat.# 52631; Abcam), and MMP-3 (1:100 dilution, cat.# \n52915; Abcam). The staining intensity was scored as fol-\nlows: 0, none of the cells stained positively; 1, weak stain-\ning; 2, moderate staining; and 3, strong staining. The results \nwere analyzed with Student’s t-test to compare the staining \nintensity for each gene between the Dienogest and Control \ngroups, and p-values < 0.05 were accepted as significant.\nResults\nIdentification of Genes Differentially Expressed \nBetween the Dienogest and Control Groups\nUnder the criteria of the fold change of at least ± 1.5 with a \ncorrected p-value of < 0.05, 824 genes were revealed to be \ndifferentially expressed between the Dienogest and Control \ngroups. Of these 824 genes, 341 were upregulated and 483 \nwere downregulated in the Dienogest group compared to the \nControl group levels. Tables  1 and 2 show the top 20 genes \nupregulated and downregulated in the Dienogest group, \nTable 1  Top 20 genes upregulated in the Dienogest group\nProbe ID Gene symbol Fold change\nA_23_P8801 CYP3A5 5.96\nA_33_P3422499 Not identified 5.48\nA_33_P3327673 COBL 4.84\nA_19_P00318561 LOC100506516 4.7\nA_24_P127691 DNAH14 4.4\nA_21_P0008435 XLOC_011016 4.27\nA_23_P390958 PCDHB18 4.23\nA_33_P3290748 LOC648044 4.1\nA_33_P3212269 MAP2K3 4.09\nA_24_P7965 ESRRG 3.95\nA_21_P0010916 XLOC_l2_001548 3.92\nA_33_P3409518 TUBBP5 3.77\nA_23_P120504 C20orf46 3.76\nA_21_P0014652 LOC100507411 3.65\nA_19_P00330814 HOTAIR 3.51\nA_33_P3272614 PPP1R26 3.41\nA_33_P3268129 Not identified 3.37\nA_33_P3333156 C11orf70 3.33\nA_21_P0014780 LOC100652747 3.2\nA_33_P3424062 KCNF1 3.18\nTable 2  Top 20 genes downregulated in the Dienogest group\nProbe ID Gene symbol Fold change\nA_23_P1452 NPFFR1 −6.53\nA_23_P313550 SLC25A41 −5.84\nA_23_P420281 PRKCB −5.71\nA_23_P329768 GREB1 −5.61\nA_33_P3376273 GK −4.77\nA_33_P3409266 GAFA2 −4.16\nA_32_P32905 Not identified −3.67\nA_21_P0010584 XLOC_l2_000696 −3.51\nA_33_P3410659 CLEC12B −3.49\nA_24_P265088 PDZD4 −3.45\nA_23_P302681 FIGNL1 −3.42\nA_33_P3262376 OTUD7A −3.31\nA_24_P395814 CGB −3.21\nA_33_P3281171 LOC100130987 −3.08\nA_33_P3270197 DIS3L2 −3.06\nA_33_P3260342 NFASC −2.94\nA_21_P0001141 XLOC_000869 −2.91\nA_21_P0010120 XLOC_013781 −2.78\nA_21_P0001404 XLOC_000511 −2.74\nA_21_P0010200 XLOC_014056 −2.72\n\n2461Reproductive Sciences (2023) 30:2457–2467 \n1 3\nrespectively. An overview of these genes was also created \nby hierarchical clustering (Fig.  1).\nGO Analysis\nThe GO enrichment analysis using the Panther software was \nperformed for the genes that were differentially expressed \nbetween the Dienogest and Control groups to reveal the bio-\nlogical functions of these genes. Table 3 lists the top five GO \nbiological processes according to fold enrichment. The main \nbiological processes identified in the GO enrichment analysis \nwere the regulation of macrophage chemotaxis, the collagen \ncatabolic process, and the proteoglycan biosynthetic process.\nIngenuity® Pathway Analysis (IPA)\nThe IPA comparing the Dienogest and Control groups iden-\ntified 20 significant canonical pathways based on 647 dif-\nferentially expressed genes, which were extracted under the \nsame conditions. Figure  2 depicts the top 10 pathways of \nthese 20 canonical pathways. The most frequently associated \ngenes in the 20 canonical pathways were MMPs (MMP-1, \n-3, -10, -12, -25, and -27). Table 4 shows the eight canonical \npathways with which these MMPs were closely associated.\nQuantitative RT‑PCR to Validate the Microarray Data\nOf the genes that were revealed to be differentially expressed \nbetween the Dienogest and Control groups, we selected CSF1, \nMST1, MMP-1, -3, and -10, and TIMP-4 for validation of the \nmicroarray data. The data illustrated in Fig.  3 demonstrate \nthat CSF1, MSP1, MMP-1, -3, and -10 were significantly \ndecreased in the Dienogest group compared to the levels in the \nControl group, whereas TIMP-4 was significantly increased.\nIHC to Validate the Data of the GO and Pathway \nAnalyses\nThe comparison of the expression levels of CSF1, MST1, \nMMP-1, and MMP-3 between the Dienogest and Control \nFig. 1  Hierarchical clustering \nheatmap of the top 20 upregu-\nlated and downregulated genes \nof the Dienogest group com-\npared to the Control group\n\n\n2462 Reproductive Sciences (2023) 30:2457–2467\n1 3\ngroups revealed significant differences (p  < 0.05), with all \nof these proteins being expressed at lower levels in the Dien-\nogest group compared to the Control group (Fig. 4, Table  5).\nDiscussion\nAlthough dienogest has been used as the main hormonal \ntherapy for endometriosis, its precise anti-endometriotic \nmechanism has remained unclear. Various investigations \nhave shown several anti-endometriotic effects of dienogest \nbut, to the best of our knowledge, no studies have analyzed \ndifferences in genome-wide gene expression profiles \nbetween untreated endometriotic cells and endometriotic \ncells treated with dienogest. In the present study, we com-\nprehensively investigated the direct effects of dienogest on \novarian ESCs. Our findings revealed that the MMPs were the \ngenes in the 20 canonical pathways that were most closely \nassociated with dienogest treatment.\nA set of MMPs controls physiological functions of the \nfemale reproductive tract such as ovulation, menstruation, \nand embryo implantation. In particular, endometrial repair, \nregeneration, and breakdown through the menstrual cycle are \nregulated by a delicate balance of MMPs and their inhibitors, \nTable 3  Top 5 GO biological processes among the genes differentially expressed between the Diengest and Control groups\nGO biological process Gene count Fold enrichment Raw p-value FDR Involved genes\nRegulation of macrophage chemotaxis 6 9.9 7.13E-05 2.29E-02 STK4, CSF1, MST1, C5AR1, MMP28, \nMTUS1\nCollagen catabolic process 7 7.52 8.34E-05 2.53E-02 MMP3, MMP12, MMP1, MMP27, MMP25, \nMMP28, MMP10\nProteoglycan biosynthetic process 9 6.93 1.46E-05 8.56E-03 CSGALNACT1, FOXL1, CYTL1, B3GAT1, \nIGF1, HS3ST3B1, PXYLP1\nHS6ST2, GAL3ST4\nRegulation of cation channel activity 14 3.49 9.40E-05 2.75E-02 NEFL, CAMK2B, CACNB4, KCNG1, GPR35, \nGRIN2A, CACNB2, KCNAB1\nJPH3, FGF13, LRRC52, CACNA1F, \nCACNG7, EPHB2\nCalcium ion transmembrane transport 14 3.4 1.22E-04 3.33E-02 CATSPER3, CACNA1B, PKD1, PKD2L1,  \nHTR2A, CACNB4, GRIN2A\nCACNB2, SLC24A1, ITPR1, CACNA1F, \nCACNG7, PTPRC, ATP2A3\nFig. 2  Top 10 canonical \npathways most significantly \ndifferentially expressed in the \nDienogest group compared to \nthe Control group. The trans-\nverse line indicates the thresh-\nold of the distinct pathways in \nthe Dienogest group under the \ncondition of p<0.05\n-log(p-value)\nThreshold\n0\n1\n2\n3\n4\n5\nAltered T-Cell and B-Cell Signaling\nin Rheumatoid Arthritis\nLXR/RXR Activation\nBladder Cancer Signaling\nArteriosclerosis Signaling\nAgranulocyte Adhesion and Diapedesis\nGranulocyte Adhesion and Diapedesis\nSperm Motility\nInhibition of Matrix Metalloproteinases\nPI3K Signaling in B Lymphocyte\nNeuropathic Pain Signaling \nDorsal Neurons\n\n2463Reproductive Sciences (2023) 30:2457–2467 \n1 3\nTIMPs [20]. Aberrated expressions of MMPs and TIMPs in \nendometriotic cells or tissues have been shown as follows: \nSpecific MMPs have been shown to be abnormally expressed \nin endometriotic lesions compared to the levels in eutopic \nendometrium. Several research groups have also observed \nthat endometriotic cells have significantly higher expres-\nsions of MMP-1, -2, -3, -7, -9, or -10 compared to eutopic \nendometrial cells in in vitro or in vivo experiments, or in \npatients’ tissues [21–27]. Of the MMP genes differentially \nexpressed between the present study’s Dienogest and Con-\ntrol groups, the expressions of MMP-1, -3, -10, -25, and -27 \nin ESCs were downregulated by the addition of dienogest. \nEarlier studies noted that MMP-1, -3, and -10 were inhib-\nited by progesterone or progestin in endometrial explants or \nendometriotic cells [28–32]. Bruner et al. [33] reported that \nthe establishment of ectopic human endometrium in nude \nmice was inhibited by the suppression of MMPs caused by \nprogesterone treatment. The finding of downregulations of \nthese MMPs by dienogest is thus likely to be reliable, and we \ntherefore believe that MMPs play central roles in dienogest’s \nanti-endometriotic effects.\nIn general, the expression of MMP genes is induced \nby growth factors, hormones, and inflammatory cytokines \nthrough the MAPK pathway or NF-kB pathway [34]. \nTable 4  Eight pathways with which MMPs are associated among the 20 canonical pathways. ↑: upregulated in the Dienogest group ↓: down-\nregulated in the Dienogest group. The bolded genes are MMPs associated with the 8 pathways\nPathway Number of associated \ngenes\nAssociated genes\nInhibition of matrix metalloproteases 7 MMP-1↓ MMP-3↓ MMP-10↓ MMP-12↑\nMMP-25↓ MMP-27↓ TIMP-4↑\nGranulocyte adhesion and diapedesis 13 C5AR1↓ CCL7↑ CCL11↑ CCL25↑ CLDN3↓\nCXCL14↑ MMP-1↓ MMP-3↓ MMP-10↓\nMMP-12↑ MMP-25↓ MMP-27↓ VCAM1↑\nAtherosclerosis signaling 11 APOA1↑ APOB↓ CCL11↑ CSF1↓ IL6↑ MMP-1↓\nMMP3↓ PLA2G2A↑ PLA2G4E↓ PON1↑ VCAM1↑\nBladder cancer signaling 7 FGF13↑ MMP-1↓ MMP-3↓ MMP-10↓\nMMP-12↑ MMP-25↓ MMP-27↓\nHIFα signaling 6 MMP-1↓ MMP-3↓ MMP-10↓\nMMP-12↑ MMP-25↓ MMP-27↓\nLeukocyte extravasation signaling 9 CLDN3↓ CXCL14↑ MMP-1↓ MMP-3↓ MMP-10 ↓\nMMP-12↑ MMP-25↓ MMP-27↓ VCAM1↑\nRole of macrophages, fibroblast and endothelial cells in \nrheumatoid arthritis\n12 C5AR1↓ CAMK2B↓ CSF1↓ GNAO1↓ IL6↑ MMP-1↓\nMMP-3↓ PLCE1↑ PLCL2↓ SOST↓ TLR3↑ VCAM1↑\nHepatic fibrosis /hepatic stellate cell activation 8 CSF1↓ HGF↓ IGF1↑ IL6↑ IL4R↓ MMP-1↓\nSERPINE1↑ VCAM1↑\nFig. 3  Validation of the micro-\narray data and the pathway \nanalysis by real-time RT-PCR \nby using all five RNA sam-\nples used for the microarray \nanalysis. The expressions of \nCSF1, MSP1, MMP-1, MMP-3, \nand MMP-10 were signifi-\ncantly decreased whereas that \nof TIMP-4 was significantly \nincreased in the Dienogest \ngroup compared to the Control \ngroup levels\n0\n0.5\n1\n1.5\n2\nCSF1 MST1 MMP-1 MMP-3 MMP-10 TIMP-4\n\r\r\r \r \r \r\nfold\n*: pʽ0.05\nControl group\nDNG group\n\n2464 Reproductive Sciences (2023) 30:2457–2467\n1 3\nThe main source of these inflammatory cytokines, such \nas TNFα and IL-1, is thought to be macrophages [35]. \nAlthough the direct effect of dienogest on macrophages to \ncounter endometriosis is still unknown, the results of the \nGO analysis in the present study suggest that dienogest \nexerts a suppressive effect on macrophage chemotaxis in \novarian ESCs.\nCSF1, which is involved in macrophage chemotaxis, \none of the enriched biological processes in the present GO \nanalysis, has been shown to be upregulated in peritoneal \nendometriotic tissues compared to the level in eutopic endo-\nmetrium [36]. It was also revealed to be downregulated in \novarian ESCs by dienogest in the present study. A similar \nfinding was also reported in the mammary gland; specifi -\ncally, progesterone administration together with estradiol \nresulted in a reduced expression of CSF1 compared to that \nupon estradiol administration alone [37]. As CSF1 promotes \nthe release of proinflammatory cytokines from macrophages, \nand treatment with CSF1 siRNA in MCF-7 cells (a human \nbreast cancer cell line) reduced the expression of MMPs in \nthese cells [38], an inhibition of CSF1 release from ovarian \nESCs by treatment with dienogest may suppress the MMP \nexpression in ovarian endometriotic tissues.\nCSF1R (receptor of CSF1) exerts tyrosine kinase activity \nand transduces its signal activated by CSF1 binding to down-\nstream components through the MAPK pathway or PI3K-\nAkt pathway [39–43]. Our present findings demonstrated the \nsuppressive effect of dienogest on the CSF1 expression in \nESCs, suggesting that dienogest may directly inhibit MMP \nexpression in these cells by suppressing CSF1.\nMST1 is another gene involved in macrophage chemo-\ntaxis as identified in the present GO analysis. Matsuzaki \net al. observed that MST1 and its receptor were upregulated \nin deep endometriotic tissues [44], and Xu et al. described \nelevated MST1R (receptor of MST1) expression in endome-\ntriotic tissues [45]. As MST1R also exerts tyrosine kinase \nactivity, as does CSF1R [46], MST1 is likely to regulate \nMMP expression through the MAPK pathway or PI3K-Akt \npathway. Indeed, MSP1 was shown to upregulate MMP-1 \nexpression in skin fibroblasts [47]. Therefore, the suppres-\nsive effect of dienogest on MST1 expression may result in \nthe direct suppression of MMP expression in ESCs.\n1TSM1FSC MMP-1 MMP-3\nDienogest group\nControl group\nFig. 4  Immunohistochemical staining of CSF1, MST1, MMP-1, and MMP-3. The levels of all of these proteins were decreased in the Dienogest \ngroup compared to those in the Control group\nTable 5  Expression levels of \nCSF1, MST1, MMP-1, and \nMMP-3. All of these proteins \nwere expressed at significantly \nlower levels in the Dienogest \ngroup than in the Control group.\n(*p < 0.05)\n\n\n2465Reproductive Sciences (2023) 30:2457–2467 \n1 3\nTIMPs have also been revealed to be abnormally \nexpressed in endometriotic tissues compared to normal \neutopic endometrium. The expressions of TIMP-1, -2, and \n-3 were reported to be significantly lower in endometriotic \ntissues than normal eutopic endometrium [48, 49], and \nan experimental study showed that the addition of TIMP \nprotein to the peritoneal cavity of nude mice prevented the \nestablishment of endometriosis [33]. However, there are few \nstudies describing a change in the expression of TIMP-1, -2, \nor -3 by an addition of progesterone, and our present findings \ndid not provide any information about the changes in TIMP \nexpressions by the addition of dienogest, with the excep-\ntion of TIMP-4. The results of our present analyses showed \nthat dienogest increased the expression of TIMP-4 in ESCs. \nAlthough there has been no report about the relationship \nbetween endometriosis and TIMP-4, the increase in TIMP-4 \nexpression by treatment with dienogest may be related to the \nanti-endometriotic effect of dienogest.\nThis study has some limitations. Pawitan et al. described \nthe relationship between the false discovery rate (FDR) and \nthe sample size for microarray studies [50]. Since a microar-\nray analysis with a small sample size as in the present study \n(n = 5) is susceptible to a high FDR, more falsely differen-\ntially expressed genes might have been extracted compared \nto the result that would be obtained with a large sample size. \nThe present IHC also had a small sample size. A replication \nof the present IHC analyses using a greater number of sam-\nples would be worthwhile.\nEndometriotic tissues consist of various types of cells \nsuch as ectopic endometrial epithelial, stromal, and immune \ncells. Inflammatory cytokines, which are produced mainly \nby immune cells (especially by macrophages in endometri-\notic tissues) are thought to play important roles in the patho-\ngenesis of endometriosis [35]. However, the experimental \nsystem in the present study lacked immune cells, and it was \nthus not possible to assess the anti-endometriotic effect of \ndienogest through the inflammatory cytokines in this study, \nwhich used only ESCs. It is also not yet possible to recreate \nthe exact biological environment of endometriotic tissues \nin an in vitro experiment such as that in the present study.\nHowever, our findings revealed for the first time the direct \neffect of dienogest on ESCs by genome-wide gene expres-\nsion profiling and a pathway analysis. The results of these \nanalyses demonstrated that MMPs were associated with the \ngenes in the 20 canonical pathways that were most closely \nassociated with dienogest treatment. The results of our GO \nanalysis also suggest that the suppressive effects of dien-\nogest on CSF1 and MST1, which are genes involved in mac-\nrophage chemotaxis, may result in the decreased expression \nof MMPs in ESCs.\nIn line with previous studies [11– 13, 51] in which the \nconcentration of dienogest ranged from  10-8 M to  10-5 M, \nwe used dienogest at  10-6 M. However, because the serum \nconcentration of dienogest under the therapeutic dose is \naround  10-7 M [52], our experimental conditions may not \nreflect those in a clinical context.\nOur results strongly suggest that MMPs play important \nroles in the therapeutic action of dienogest against endo-\nmetriosis. It is clear that dienogest is useful and effective \nfor treating endometriosis, but due to its anti-ovulatory \neffect, it is not used for patients who want to conceive. \nPatients with severe endometriosis who want to conceive \nmore than once are likely to experience disease progres-\nsion during the period without dienogest. A substrate that \nsuppresses the expression of MMPs downstream of dien-\nogest and does not suppress ovulation is desired, because \nsuch a substrate could conceivably be used as a drug to \ntreat patients with endometriosis who want to conceive.\nDienogest has biological actions that are similar to \nthose of progesterone, but dienogest and progesterone are \nalso likely to have some differences, like other progestins. \nThese differences remain unknown, but the setup of the \npresent experiment prevents such differences from being \ndetermined. Other studies pursuing these differences may \ndeepen our understanding of the precise biological actions \nof dienogest and the pathogenesis of endometriosis itself.\nConclusions\nTo investigate the biological actions of dienogest in ovar -\nian ESCs, we compared the genome-wide gene expres-\nsion profiles between Dienogest and Control groups and \nextracted 824 genes that are differentially expressed \nbetween these two groups. The data revealed by the GO \nanalysis and Ingenuity® pathway analysis using the dif-\nferentially expressed genes suggested that MMPs play \nimportant roles in the anti-endometriotic effect of dien-\nogest on ovarian ESCs. Our findings also suggest that the \nsuppression of these MMPs by dienogest may be due to \nthe suppression of components involved in macrophage \nchemotaxis, such as CSF1 and MST.\nAcknowledgements We thank the pathology technicians at the Depart-\nment of Diagnostic Pathology, Hiroshima City North Medical Center \nAsa Citizens Hospital for their assistance with the immunohistochemi-\ncal experiments.\nAuthor’s Contributions H.H. designed the study, carried out most \nof the experiments, and wrote the paper. N.N. assisted with some \nexperiments. M.K. assisted with the immunohistochemical experi-\nments. M.Y., Y.A., H. Tanimoto, M.T., and H. Teramoto provided the \npatients’ samples and clinical information. Y.Y. contributed to the data \nanalysis and interpretation. All of the authors reviewed and revised the \nmanuscript.\nFunding This study was funded by the Tsuchiya Foundation (Grant \nID: 2010(26)-21).\n\n2466 Reproductive Sciences (2023) 30:2457–2467\n1 3\nData availability The datasets used and analyzed in this study are avail-\nable from the corresponding author on reasonable request.\nCode Availability Not applicable\nDeclarations \nEthics Approval and Consent to Participate This study was approved by \nthe Ethical Committee of Hiroshima City Asa Citizens Hospital, and \nall experiments were carried out in accordance with relevant guide-\nlines and regulations. All participants provided their written informed \nconsent.\nConsent for Publication Not applicable\nConflict of Interest The authors declare no competing interests.\nOpen Access This article is licensed under a Creative Commons Attri-\nbution 4.0 International License, which permits use, sharing, adapta-\ntion, distribution and reproduction in any medium or format, as long \nas you give appropriate credit to the original author(s) and the source, \nprovide a link to the Creative Commons licence, and indicate if changes \nwere made. The images or other third party material in this article are \nincluded in the article's Creative Commons licence, unless indicated \notherwise in a credit line to the material. If material is not included in \nthe article's Creative Commons licence and your intended use is not \npermitted by statutory regulation or exceeds the permitted use, you will \nneed to obtain permission directly from the copyright holder. To view a \ncopy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.\nReferences\n 1. Shafrir AL, Farland LV, Shah DK, Harris HR, Kvaskoff M, \nZondervan K, et al. Risk for and consequences of endometrio-\nsis: A critical epidemiologic review. Best Pract Res Clin Obstet \nGynaecol. 2018;51:1–15.\n 2. Strowitzki T, Marr J, Gerlinger C, Faustmann T, Seitz C. Dien-\nogest is as effective as leuprolide acetate in treating the painful \nsymptoms of endometriosis: A 24-week, randomized, multicentre, \nopen-label trial. Hum Reprod. 2010;25:633–41.\n 3. Römer T. Long-term treatment of endometriosis with dienogest: \nRetrospective analysis of efficacy and safety in clinical practice. \nArch Gynecol Obstet. 2018;298:747–53.\n 4. Moore C, Carol W, Gräser T, Mellinger U, Walter F. Influence \nof dienogest on ovulation in young fertile women. Clinical Drug \nInvestigation. 1999;18:271–8.\n 5. Klipping C, Duijkers I, Faustmann TA, Klein SF, Schuett B. Phar-\nmacodynamic study of four oral dosages of dienogest. Fertil Steril. \n2010;94:S181.\n 6. Yamanaka K, Xu B, Suganuma I, Kusuki I, Mita S, Shimizu Y, \net al. Dienogest inhibits aromatase and cyclooxygenase-2 expres-\nsion and prostaglandin  E2 production in human endometriotic \nstromal cells in spheroid culture. Fertil Steril. 2012;97:477–82.\n 7. Miyashita M, Koga K, Takamura M, Izumi G, Nagai M, Harada \nM, et al. Dienogest reduces proliferation, aromatase expression \nand angiogenesis, and increases apoptosis in human endometrio-\nsis. Gynecol Endocrinol. 2014;30:644–8.\n 8. Sacco K, Portelli M, Pollacco J, Schembri-Wismayer P, Calleja-\nAgius J. The role of prostaglandin E2 in endometriosis. Gynecol \nEndocrinol. 2012;28:134–8.\n 9. Fu L, Osuga Y, Morimoto C, Hirata T, Hirota Y, Yano T, et al. \nDienogest inhibits BrdU uptake with G0/G1 arrest in cultured \nendometriotic stromal cells. Fertil Steril. 2008;89:1344–7.\n 10. Honda H, Barrueto FF, Gogusev J, Im DD, Morin PJ. Serial anal-\nysis of gene expression reveals differential expression between \nendometriosis and normal endometrium. Possible roles for AXL \nand SHC1 in the pathogenesis of endometriosis. Reprod Biol \nEndocrinol. 2008;6:59.\n 11. Okada H, Nakajima T, Yoshimura T, Yasuda K, Kanzaki H. The \ninhibitory effect of dienogest, a synthetic steroid, on the growth \nof human endometrial stromal cells in vitro. Mol Hum Reprod. \n2001;7:341–7.\n 12. Prechapanich J, Kajihara T, Fujita K, Sato K, Uchino S, Tanaka K, \net al. Effect of a dienogest for an experimental three-dimensional \nendometrial culture model for endometriosis. Med Mol Morphol. \n2014;47:189–95.\n 13. Yamanaka A, Kimura F, Kishi Y, Takahashi K, Suginami H, \nShimizu Y, et al. Progesterone and synthetic progestin, dienogest, \ninduce apoptosis of human primary cultures of adenomyotic stro-\nmal cells. Euro J Obstet Gynecol Reprod Biol. 2014;179:170–4.\n 14. Yokoi A, Yoshioka Y, Yamamoto Y, Ishikawa M, Ikeda SI, Kato \nT, et al. Malignant extracellular vesicles carrying MMP1 mRNA \nfacilitate peritoneal dissemination in ovarian cancer. Nat Com-\nmuned. 2017;8:14470.\n 15. Baird AM, Easy D, Jarzabek M, Shiels L, Silbermann A, Klebe \nS, et al. When RON MET TAM in mesothelioma: All drugga-\nble for one, and one drug for all? Front Endocrinol (Lausanne). \n2019;10:89.\n 16. Dang W, Qin Z, Fan S, Wen Q, Lu Y, Wang J, et al. miR-1207-5p \nsuppresses lung cancer growth and metastasis by targeting CSF1. \nOncotarget. 2016;7(22):32421–32.\n 17. Pennington C. Real-time PCR expression profiling of MMPs and \nTIMPs. Methods Mol Biol. 2010;622:159–73.\n 18. Wang L, Zhang ZG, Zhang RL, Gregg SR, Hoses-Solgot A, \nLeTourneau Y, et  al. Matrix metalloproteinase 2 (MMP2) \nand MMP9 secreted by erythropoietin-activated endothelial \ncells promote neural progenitor cell migration. J Neurosci. \n2006;26:5996–6003.\n 19. Schmittgen TD, Livak KJ. Analysis of relative gene expression \ndata using real-time quantitative PCR and the 2(-Delta Delta \nC(T)) Method. Methods. 2001;25:402–8.\n 20. Curry TE Jr, Osteen KG. The matrix metalloproteinase system: \nChanges, regulation, and impact throughout the ovarian and uter-\nine reproductive cycle. Endocr Rev. 2003;24:428–65.\n 21. Kokorine I, Nisolle M, Donnez J, Eeckhout Y, Courtoy PJ, Mar -\nbaix E. Expression of interstitial collagenase (matrix metallo-\nproteinase-1) is related to the activity of human endometriotic \nlesions. Fertil Steril. 1997;68:246–51.\n 22. Hudelist G, Keckstein J, Czerwenka K, Lass H, Walter I, Auer M, \net al. Estrogen receptor beta and matrix metalloproteinase 1 are \ncoexpressed in uterine endometrium and endometriotic lesions of \npatients with endometriosis. Fertil Steril. 2005;84:1249–56.\n 23. Ueda M, Yamashita Y, Takehara M, Terai Y, Kumagai K, Ueki \nK, et al. Gene expression of adhesion molecules and matrix \nmetalloproteinases in endometriosis. Gynecol Endocrinol. \n2002;16:391–402.\n 24. Cox KE, Piva M, Sharpe-Timms KL. Differential regula-\ntion of matrix metalloproteinase-3 gene expression in endo-\nmetriotic lesions compared with endometrium. Biol Reprod. \n2001;65:1297–303.\n 25. Gilabert-Estellés J, Estellés A, Gilabert J, Castelló R, España F, \nFalcó C, et al. Expression of several components of the plasmi-\nnogen activator and matrix metalloproteinase systems in endome-\ntriosis. Hum Reprod. 2003;18:1516–22.\n 26. Chung HW, Wen Y, Chun SH, Nezhat C, Woo BH, Lake PM. \nMatrix metalloproteinase-9 and tissue inhibitor of metalloprotein-\nase-3 mRNA expression in ectopic and eutopic endometrium in \nwomen with endometriosis: A rationale for endometriotic inva -\nsiveness. Fertil Steril. 2001;75:152–9.\n\n2467Reproductive Sciences (2023) 30:2457–2467 \n1 3\n 27. Pitsos M, Kanakas N. The role of matrix metalloproteinases in the \npathogenesis of endometriosis. Reprod Sci. 2009;16:717–22.\n 28. Vassilev V, Pretto CM, Cornet PB, Delvaux D, Eeckhout Y, \nCourtoy PJ, et al. Response of matrix metalloproteinases and tis-\nsue inhibitors of metalloproteinases messenger ribonucleic acids \nto ovarian steroids in human endometrial explants mimics their \ngene- and phase-specific differential control in vivo. J Clin Endo-\ncrinol Metab. 2005;90:5848–57.\n 29. Bruner-Tran KL, Eisenberg E, Yeaman GR, Anderson TA, \nMcBean J, Osteen KG. Steroid and cytokine regulation of matrix \nmetalloproteinase expression in endometriosis and the establish-\nment of experimental endometriosis in nude mice. J Clin Endo-\ncrinol Metab. 2002;87:4782–91.\n 30. Bruner-Tran KL, Zhang Z, Eisenberg E, Winneker RC, Osteen \nKG. Down-regulation of endometrial matrix metalloproteinase-3 \nand -7 expression in vitro and therapeutic regression of experi-\nmental endometriosis in vivo by a novel nonsteroidal proges-\nterone receptor agonist, tanaproget. J Clin Endocrinol Metab. \n2006;91:1554–60.\n 31. Lockwood CJ, Krikun G, Hausknecht VA, Papp C, Schatz F. \nMatrix metalloproteinase and matrix metalloproteinase inhibitor \nexpression in endometrial stromal cells during progestin-initiated \ndecidualization and menstruation-related progestin withdrawal. \nEndocrinology. 1998;139:4607–13.\n 32. Mönckedieck V, Sannecke C, Husen B, Kumbartski M, Kimmig \nR, Tötsch M, et al. Progestins inhibit expression of MMPs and \nof angiogenic factors in human ectopic endometrial lesions in a \nmouse model. Mol Hum Reprod. 2009;15:633–43.\n 33. Bruner KL, Matrisian LM, Rodgers WH, Gorstein F, Osteen KG. \nSuppression of matrix metalloproteinases inhibits establishment \nof ectopic lesions by human endometrium in nude mice. J Clin \nInvest. 1997;99:2851–7.\n 34. Brinckerhoff CE. Regulation of Matrix Metalloproteinase (MMP) Gene \nExpression. Matrix Metalloproteinase in Health and Disease: Sculpting \nthe Human Body. Singapore: World Scientific; 2017. p. P30–62.\n 35. Iwabe T, Harada T. Inflammation and cytokine in endometriosis. \nEndometriosis. Tokyo: Springer; 2014. p. P87–106.\n 36. Budrys NM, Nair HB, Liu YG, Kirma NB, Binkley PA, Kumar S, \net al. Increased expression of macrophage colony-stimulating fac-\ntor and its receptor in patients with endometriosis. Fertil Steril. \n2012;97:1129–35.\n 37. Dasari P, Sharkey DJ, Noordin E, Glynn DJ, Hodson LJ, Chin PY, \net al. Hormonal regulation of the cytokine microenvironment in \nthe mammary gland. J Reprod Immunol. 2014;106:58–66.\n 38. Aharinejad S, Paulus P, Sioud M, Hofmann M, Zins K, Schäfer \nR, et al. Colony-stimulating factor-1 blockade by antisense oli-\ngonucleotides and small interfering RNAs suppresses growth \nof human mammary tumor xenografts in mice. Cancer Res. \n2004;64:5378–84.\n 39. Kelley TW, Graham MM, Doseff AI, Pomerantz RW, Lau SM, \nOstrowski MC, et al. Macrophage colony-stimulating factor pro-\nmotes cell survival through Akt/protein kinase B. J Biol Chem. \n1999;274:26393–8.\n 40. Baran CP, Zeigler MM, Tridandapani S, Marsh CB. The role of \nROS and RNS in regulating life and death of blood monocytes. \nCurr Pharm Des. 2004;10:855–66.\n 41. Varticovski L, Druker B, Morrison D, Cantley L, Roberts T. The \ncolony stimulating factor-1 receptor associates with and activates \nphosphatidylinositol-3 kinase. Nature. 1989;342:699–702.\n 42. Himes SR, Sester DP, Ravasi T, Cronau SL, Sasmono T, Hume \nDA. The JNK are important for development and survival of mac-\nrophages. J Immunol. 2006;176:2219–28.\n 43. Gobert Gosse S, Bourgin C, Liu WQ, Garbay C, Mouchiroud G. \nM-CSF stimulated differentiation requires persistent MEK activ-\nity and MAPK phosphorylation independent of Grb2-sos asso-\nciation and phosphatidylinositol 3-kinase activity. Cell Signal. \n2005;17:1352–62.\n 44. Matsuzaki S, Canis M, Pouly JL, Dechelotte P, Okamura K, Mage \nG. The macrophage stimulating protein/RON system: A potential \nnovel target for prevention and treatment of endometriosis. Mol \nHum Reprod. 2005;11(5):345–9.\n 45. Xu P, Ding S, Zhu L, Le F, Huang X, Tian Y, Zhang X. Elevated \nRON protein expression in endometriosis and disease-associated \novarian cancers. Arch Gynecol Obstet. 2017;295(3):631–9.\n 46. Danilkovitch A, Leonard EJ. Kinases involved in MSP/RON sign-\naling. J Leukoc Biol. 1999;65:345–8.\n 47. Zhao J, Hu L, Gong N, Tang Q, Du L, Chen L. The effects of \nmacrophage-stimulating protein on the migration, proliferation, \nand collagen synthesis of skin fibroblasts in vitro and in vivo. \nTissue Eng Part A. 2015;21(5-6):982–91.\n 48. Gottschalk C, Malberg K, Arndt M, Schmitt J, Roessner A, \nSchultze D, et al. Matrix metalloproteinases and TACE play a \nrole in the pathogenesis of endometriosis. Adv Exp Med Biol. \n2000;477:483–6.\n 49. Kotronis K, Zafrakas M, Papasozomenou P, Timologou A, Mili-\naras D, Tarlatzis BC, et al. Protein expression pattern of tissue \ninhibitor of metalloproteinase-3 (TIMP3) in endometriosis and \nnormal endometrium. Gynecol Endocrinol. 2019;35:1103–6.\n 50. Pawitan Y, Michiels S, Koscielny S, Gusnanto A, Ploner A. False \ndiscovery rate, sensitivity and sample size for microarray studies. \nBioinformatics. 2005;21:3017–24.\n 51. Choi J, Jo M, Lee E, Lee DY, Choi D. Dienogest enhances \nautophagy induction in endometriotic cells by impairing activa-\ntion of AKT, ERK1/2, and mTOR. Fertil Steril. 2015;104:655–64.\n 52. Foster RH, Wilde MI. Dienogest. Drugs. 1998;56:825–33.\nPublisher’s Note Springer Nature remains neutral with regard to \njurisdictional claims in published maps and institutional affiliations.","source_license":"CC0","license_restricted":false}