{"paper_id":"65b0efe0-6d55-439c-9dd1-b7bcbad37d0d","body_text":"Vol.:(0123456789)1 3\nArchives of Gynecology and Obstetrics (2019) 300:711–716 \nhttps://doi.org/10.1007/s00404-019-05205-w\nGYNECOLOGIC ONCOLOGY\nExpression of annexin A2 in adenomyosis and dysmenorrhea\nFeng Liu1 · Lixue Liu1 · Jian Zheng1 \nReceived: 10 March 2019 / Accepted: 30 May 2019 / Published online: 10 June 2019 \n© The Author(s) 2019\nAbstract\nPurpose To investigate the expression of annexin A2 (ANXA2) in ectopic and eutopic endometrium and serum of women \nwith adenomyosis, and their relationships with adenomyosis-associated dysmenorrhea.\nMethods The expression of ANXA2 was detected by immunohistochemical S-P method in ectopic and eutopic endometrium \ntissues from 30 patients with adenomyosis who underwent hysterectomy. The correlation between ANXA2 expression and \ndysmenorrhea degree was evaluated. The endometrium tissues from 15 patients with uterine fibroids which underwent hys-\nterectomy were used as controls. The preoperative serum level of ANXA2 was measured by enzyme-linked immunosorbent \nassay in 30 patients with adenomyosis and 15 patients with hysteromyoma.\nResult The expression of ANXA2 in eutopic and ectopic endometrium of adenomyosis was higher than in normal endo-\nmetrium (P  < 0.05), but no significant difference of ANXA2 expression was observed between the eutopic endometrium \nand the ectopic endometrium (P  > 0.05). In the ectopic endometrium, but not in the eutopic endometrium, of women with \nadenomyosis, ANXA2 expression was positively correlated with the severity of dysmenorrhea (R  = 0.831, P = 0.000). The \npreoperative serum level of ANXA2 was markedly higher in patients with adenomyosis compared with the patients with \nhysteromyoma (P < 0.05).\nConclusion The increased ANXA2 may contribute to the occurrence and development of adenomyosis, and may play a \nimportant role in the dysmenorrhea. The present study may provide a new idea of diagnosis and treatment to adenomyosis-\nassociated dysmenorrhea.\nKeywords Annexin A2 · Adenomyosis · Dysmenorrhea\nIntroduction\nAdenomyosis is a benign gynecological disease character -\nized by the presence of aberrant growth and invasion of \nendometrial tissue embedded within the myometrium, lead-\ning to dysfunctional myometrial hyperperistalsis, increased \nintra-uterine pressure and impairment of proper uterine \nfunction [1, 2]. Detection of serum CA125 is good for diag-\nnosis, but definitive diagnosis still requires surgery and \npathology. Due to the lack of reliable diagnostic indicators, \nit is inevitable to cause misdiagnosis or missed diagnosis. \nTherefore, it is of profound significance to find a diagnos-\ntic marker. In addition, moderate and severe dysmenorrhea \nseriously affects the physical and mental health and qual-\nity of life of women of childbearing age, and the relevant \nmechanism is not yet very clear, the root cause of urgent \nneed to investigate. Recent studies have found that epithe-\nlial–mesenchymal transition (EMT) is the initiating factor of \nAM [3, 4]. The in vitro AM model confirmed that estrogen \nsignificantly up-regulated ANXA2 and induced EMT [5 ]. \nAdditionally, evidence-based data unraveled an active role \nfor ANXA2 in the pathogenesis of adenomyosis through \nconferring the ability of endometrial carcinomas to metasta-\nsize and proangiogenic capacity [6]. Annexin A2 (ANXA2), \na calcium-binding cytoskeletal protein found on various cell \ntypes, has a diverse range of cellular functions including \nangiogenesis, proliferation, apoptosis, calcium signaling and \ncell growth regulation [7–9]. The up-regulated expression of \nANXA2 has been reported in breast cancer, pancreatic can-\ncer and laryngeal cancer tissue [10]. However, little is known \nabout the relationship between ANXA2 and adenomyosis-\nassociated dysmenorrhea. In this study, the expression of \n * Jian Zheng \n zheng_jian625@aliyun.com\n1 Department of Gynecology and Obstetrics, The Affiliated \nHospital of Inner Mongolia Medical University, Huhhot, \nPeople’s Republic of China\n\n712 Archives of Gynecology and Obstetrics (2019) 300:711–716\n1 3\nANXA2 was detected by immunohistochemical S-P method, \nfollowed by the Pearson correlations for the correlation \nanalysis of ANXA2 with adenomyosis-associated dysmenor-\nrhea. Meanwhile, the levels of preoperative serum ANXA2 \nof patients with adenomyosis (n  = 30) and uterine myoma \n(n = 15) were also measured by enzyme-linked immunosorb-\nent assay (ELISA). These findings are to provide a reliable \ntheoretical basis for the clinical diagnosis and treatment of \nadenomyosis-associated dysmenorrhea and the development \nand application of related markers.\nMaterials and methods\nTissue collection\nFreshly resected matched ectopic and eutopic endometrial \ntissues of adenomyosis and adenomyoma patients undergo-\ning hysterectomy were collected from the department of \ngynecology and obstetrics, the affiliated hospital of Inner \nMongolia medical university from January 2018 to Decem-\nber 2018. None of pre-menopausal had intra-uterine device \n(IUD), typical endometrial hyperplasia; none unincorpo-\nrated endocrine, immune and metabolic disease, malignant \ntumors, and none received any hormone and immune agents \nprior to surgery for 3 months. Tissue and serum samples \nwere immediately frozen in liquid nitrogen and stored at \n− 80 °C.\nImmunohistochemistry\nExperimental group: ectopic and eutopic endometrial tissues \nof 30 donors with an average of 42 years old (31–48 years \nold) were documented by pathology department. There were \n15 cases of proliferative phase and secretory phase in 15 \ncases with a history of uterine cavity operation no more than \n2 times (diagnostic curettage or abortion, etc.), respectively. \nOf these, there were ten cases of mild dysmenorrhea, ten \ncases of moderate dysmenorrhea, and ten cases of severe \ndysmenorrhea.\nControl group : 15 normal endometrium tissues were \nobtained from adenomyoma patients with an average of \n40 years old (33–50 years old).\n4-μm tissue sections were set in oven at 60 °C for 20 min, \nand then xylene was used to dewaxing twice, 10 min each \ntime. Tissue sections were incubated with 100% alcohol for \n10 min twice, 95, 90 and 80% alcohol for 5 min, respec-\ntively, to block endogenous peroxidase activity. After wash-\ning, sections were incubated with a repairing solution for \n9 min. 50 μL ANXA2 antibody (Bio-synthesis, Lewisville, \nUSA) 1:100 was diluted in blocking solution and incu-\nbated in a humidified chamber at 4 °C for one night. After \nwashing, 50 μL secondary antibody (biotin-labeled goat \nanti-mouse IgG) was applied for 10 min at room temperature \nfollowed by 3 washing steps of 3 min each, and then 50 μL \nstreptavidin peroxidase solution was used. Sections were \nthen washed, and color was developed using DAB substrate \nchromogen system (Zymed Inc., South San Francisco, CA, \nUSA). Sections were counterstained with Mayer hematoxy-\nlin for 3 min. Finally, sections were analyzed with a micro-\nscope (BH-2 OLYMPUS, Tokyo, Japan).\nELISA\nExperimental group: Preoperative serum samples were \nobtained from 30 adenomyosis patients, with an average of \n42 years (ranging from 31 to 48 years old).\nControl group: preoperative serum samples of 15 patients \nwith uterine myoma ranged from 33 to 50 years old, with an \naverage of 40 years.\nThe level of serum ANXA2 was detected using a human \nANXA2 ELISA kit (Uscn Life Science Inc, Wuhan, China) \naccording to the manufacturer’s instructions. 50 μL of serum \nsample or standard separately was added to each well, and \nthen 50 μL of detection reagent was applied and incubated \nfor 30 min at 37 °C. Subsequently, 50 μL color development \nreagent A and 50 μL reagent B were added and incubated for \n15 min at 37 °C. Finally, 50 μL of stop solution was added to \neach well, and absorbance was read at 450 nm. During the \nprocedure, washing the plate was according to the ELISA \nroutine method.\nEvaluation criteria\nStaining and scoring: ANXA2 exhibited positive expres-\nsion with yellow or brown particles in membrane and \ncytoplasm. To evaluate the immunostaining, a score cor -\nresponding to the product of both (a) staining (0 = negative; \n1 = canary; 2 = yellow; 3 = brown) and (b) percentage of \npositive cells (0 = < 5% positive cells; 1 = 5–25% positive \ncells; 2 = 26–50% positive cells; 3 = 51–75% positive cells; \n4 ≥ 75%) was established. The product of (a) × (b) was con-\nsidered as comprehensive evaluation scores (0–2 = nega-\ntive; 3–4 = weakly positive; 5–8 = positive; 9–12 = strongly \npositive). A score greater than 2 was the value of a positive \nimmunohistochemical assay [11, 12]. The results of stain-\ning were evaluated by two independent pathologists without \nknowledge of the clinico-pathological features, and any dif-\nference in interpretation was resolved by consensus.\nDysmenorrhea degree: The examiners evaluated the \ndegree of dysmenorrhea using a visual analog scale [13]. \n(VAS, ranged 0–10 cm, a score of “10” entailed the best \noutcome, while a score of “0” entailed the worst). Pain \nassessment criteria: no dysmenorrhea (−), mild pain (+, \nl–4 cm), moderate pain (++ , 5–7 cm) and severe pain (+++, \n8–l0 cm).\n\n713Archives of Gynecology and Obstetrics (2019) 300:711–716 \n1 3\nStatistical analysis\nIBM SPSS statistical software, version 13 for Windows \n(IBM SPSS Inc., Chicago, IL, USA) was used for the sta-\ntistical analysis. The experimental data were shown as \nmean ± SD. One-way ANOVA was used with Tukey’s mul-\ntiple comparison tests for multiple groups. Bivariate Pearson \ncorrelation was used for the correlation analysis. The level \nof statistical significance was set at 0.05.\nResults\nANXA2 antibody staining\nAccording to the immunohistochemistry (Fig. 1a–c) and the \ncomprehensive scoring criteria, ANXA2 was strongly posi-\ntive expressed in eutopic endometrium, and was positively \nexpressed in ectopic endometrium. The result of semi-quan-\ntitatively for ANXA2 showed that there was a significant \ndifference between control group and adenomyosis groups \n(Fig.  1d). ANXA2 was significantly increased in adenomyo-\nsis. However, the expression of ANXA2 had no significant \ndifference between eutopic endometrium group and ectopic \nendometrium group.\nANXA2 expression and dysmenorrhea degree \nin adenomyosis groups\nAs shown in Fig.  1, ANXA2 was significantly increased in \neutopic endometrium and ectopic endometrium. Further, \nto investigate the correlation between ANXA2 expression \nand dysmenorrhea degree in eutopic endometrium and \nectopic endometrium, Pearson correlations analysis was \nperformed. The result revealed that there was a significant \npositive correlation between ANXA2 expression in ectopic \nFig. 1  The expression of ANXA2 in controls, eutopic endome-\ntrium and ectopic endometrium. a ANXA2 was weakly positively \nexpressed in normal endometrium tissues. b ANXA2 was strongly \npositively expressed in eutopic endometrium. c ANXA2 was posi-\ntively expressed in ectopic endometrium. d Semi-quantitatively for \nANXA2 of three groups (*P < 0.05, **P < 0.01) compared with con-\ntrol group  × 400\n\n714 Archives of Gynecology and Obstetrics (2019) 300:711–716\n1 3\nendometrium and the degree of dysmenorrhea (R  = 0.831, \nP = 0.000, Fig. 2), however, ANXA2 expression in eutopic \nendometrium had no obviously linear dependence on dys-\nmenorrheal degree (R = 0.187, P = 0.121).\nLevel of ANXA2 in serum of patients \nwith adenomyosis\nThe ANXA2 level in preoperative serum was further \ndetected by ELISA (Fig.  3). Serum level of ANXA2 in \nadenomyosis group was significant higher than that in con-\ntrol group (P  < 0.01).\nDiscussion\nMultiple pregnancies, childbirth, induced abortion, chronic \nendometritis and other factors make the endometrial basal \nlayer thinning and loss of protective function. The endome-\ntrium grows through direct contact with the myometrium \nthrough the unprotected areas and invades the myometrium, \nand further invades the surrounding tissue. In the observa-\ntion of pathological sections of AM patients undergoing \nhysterectomy, endometrial glands and stromal tissues were \nfound in 10–47% of the myometrium, which were closely \nconnected with the endometrium. Therefore, although AM \nis a benign pathological manifestation, it has biological \ncharacteristics similar to malignant tumors, such as implan-\ntation, growth, infiltration, recurrence and so on. In addi-\ntion to the basement membrane invagination theory, the \npathogenesis of AM may be related to estrogen metabolism \ndisorder, EMT, eutopic endometrial lesion, immune factors \nand genetic factors. From the perspective of EMT, exploring \nthe pathogenesis of AM will become an important research \nFig. 2  ANXA2 expression \nwas positively correlated with \ndysmenorrhea degree in ectopic \nendometrium\nFig. 3  The ANXA2 level in preoperative serum of control and adeno-\nmyosis groups. (**P < 0.01, compared with control group)\n\n715Archives of Gynecology and Obstetrics (2019) 300:711–716 \n1 3\ndirection in the future. Research finds promotion for the \ngrowth, distant metastasis and angiogenesis of AM endo-\nmetrial tissue is implemented through the mechanism on \nANXA2-inducing EMT [5]. ANXA2 was first discovered in \n1979 by Rade and Martin. Its basic structure contains 339 \namino acids, consisting of the N-terminal of 3 kD and the \nC-terminal domain of 33 kD. As a calcium-binding cytoskel-\neton protein, it has many functions including angiogenesis, \nproliferation, apoptosis, calcium signal transduction and cell \ngrowth regulation [7–9]. The mechanism on induction EMT \nin AM may be through binding fibrinogen then hydrolyz-\ning to fibrinolysis enzyme, which can degrade extracellular \nmatrix and peripheral vascular basement membrane [14]. It \nis possible to change cell-to-cell and cell-to-matrix adhesion \nby binding to cell surface adhesion molecules to enhance \nthe anti-apoptosis and motility of cells, and then induce \nepithelial-to-mesenchymal cell transformation (EMT) [ 3, \n4]. Research finds that ANXA2 abnormal expression in \ncervical cancer, ovarian cancer, choriocarcinoma and other \ngynecological malignancies [15] and AM also has biological \nbehavior similar to a malignant tumor. According to ANXA2 \nfunction and the above in vitro study AM, the mechanism is \nspeculated. In the human body of AM, that ANXA2 abnor-\nmal expression probably promoted the occurrence and devel-\nopment of AM have not been reported. In this study, the \nexpression of ANXA2 in AM was detected by immunohis-\ntochemistry S-P method. There was no significant difference \nbetween the two groups (P  > 0.05), but the expression was \nhigher in eutopic endometrium and ectopic endometrium \nthan in normal endometrium (P  < 0.05). In vitro studies \nfound that the increased expression about ANXA2 in AM \nectopic lesions was consistent [5]. This not only confirms the \nabnormal expression of ANXA2 in human AM tissues, but \nalso indicates that the effect of ANXA2 in AM epitope and \nectopic endometrium may be the same, and may be different \nfrom that in uterine leiomyoma. It indicates that ANXA2 \nmay play an important role in the development of AM.\nProgressive dysmenorrhea is the main clinical mani-\nfestation of AM, seriously affecting the quality of life of \npatients, some patients just because of dysmenorrhea can \nnot bear the reluctance to remove the uterus. It is reported \nthat the dysmenorrhea rate of AM patients is as high as \n64.8–77.8% [16]. It is noted that dysmenorrhea is closely \nrelated not only to estrogen but also to prostaglandins and \ncyclooxygenases 2. A study confirmed that estrogen sig-\nnificantly up-regulates ANXA2 [5 ]. Another study found \nthat prostaglandin E2 increased significantly in uterine tis-\nsues of AM patients, especially in patients with severe dys-\nmenorrhea. Cyclooxygenase 2 is the rate-limiting enzyme \nfor the conversion of arachidonic acid to prostaglandins. \nIt was observed that ANXA2 significantly increased the \nexpression of Cyclooxygenase 2 in peritoneal macrophages \nof patients with endometriosis by PCR and Western-blot \nmethods [17]. It can be seen that AM-elevated estrogen \nlevels in patients may increase cyclooxygenase 2 by up-\nregulating ANXA2. There by increasing prostaglandin E2 \nproduction and promoting dysmenorrhea. EMT is strongly \nrelated to high estrogen environment. Estrogen-induced \nEMT is one of the important mechanisms of AM devel-\nopment. Among them, ANXA2 may play a key role. The \nrelationship between ANXA2 expression and dysmenor -\nrhea in AM tissues was tested in this study. The expression \nlevel of ANXA2 in AM ectopic endometrium was posi-\ntively correlated with dysmenorrhea degree (R  = 0.831, \nP = 0.000). The expression level of ANXA2 in AM ectopic \nendometrium increased gradually with the degree of dys-\nmenorrhea. These results suggest that the exacerbation of \ndysmenorrhea is closely related to the up-regulation of \nANXA2 expression in AM lesions, and there is a certain \ncorrelation between them. It is suggested that ANXA2 is \ninvolved in the occurrence of AM and dysmenorrhea, and \npromotes the development of AM and the aggravation of \ndysmenorrhea.\nPreoperative diagnostic coincidence rate of AM by \nultrasonography was only 52.9–60.5% and misdiagnosis \nrate was high. Preoperative diagnostic coincidence rate \nof MRI was 88.2%, but the high price limited the clinical \napplication. Although the detection of serum CA125 has \nsome reference value in the diagnosis of AM, the posi-\ntive rate and the value of serum CA125 have not been \nreported in the literature. There is still a lack of sensitivity \nand specificity, so it is necessary to develop more reliable \nmarkers. In this study, the serum concentration of ANXA2 \nin AM patients was determined by ELISA for the first \ntime, which was significantly higher than that in hystero-\nmyoma patients (P  < 0.05), suggesting that ANXA2 might \nbe a new marker for the diagnosis of AM. Gene-targeting \ntherapy for ANXA2, a biological target, provides a new \ntherapeutic approach to alleviate dysmenorrhea in AM \npatients and to meet the conservative and fertility require-\nments of some patients after the open second-child policy.\nConclusion\nThe increased ANXA2 may contribute to the occurrence \nand development of adenomyosis, and may play a impor -\ntant role in the dysmenorrhea. The present study may pro-\nvide a new idea of diagnosis and treatment to adenomyo-\nsis-associated dysmenorrhea.\nAcknowledgements The authors thank the support of Department of \nPathology, The Affiliated Hospital of Inner Mongolia Medical Univer-\nsity, Huhhot, People’s Republic of China.\n\n716 Archives of Gynecology and Obstetrics (2019) 300:711–716\n1 3\nAuthor contributions FL: performed the experiments, collected and \nanalyzed the data and wrote the manuscript. LL: performed the experi-\nments and collected the data. JZ: designed the experiments, analyzed \nthe data and edited the manuscript.\nFunding The study was funded by the Natural Science Foundation of \nInner Mongolia Autonomous Region of China (2018LH08037).\nCompliance with ethical standards \nConflict of interest The authors declare that they have no conflict of \ninterest.\nEthical approval  Pessary is a noninvasive test and treatment, which \nis voluntarily chosen and informed about to the patients before using, \nso Institutional Review Board approval was exempted for the study.\nOpen Access This article is distributed under the terms of the Crea-\ntive Commons Attribution 4.0 International License (http://creat iveco \nmmons .org/licen ses/by/4.0/), which permits unrestricted use, distribu-\ntion, and reproduction in any medium, provided you give appropriate \ncredit to the original author(s) and the source, provide a link to the \nCreative Commons license, and indicate if changes were made.\nReferences\n 1. Tamai K, Togashi K, Ito T, Morisawa N, Fujiwara T, Koyama T \n(2005) MR imaging findings of adenomyosis: correlation with \nhistopathologic features and diagnostic pitfalls. Radiographics \n25:21–40\n 2. Wang PH, Su WH, Sheu BC, Liu WM (2009) Adenomyosis and \nits variance: adenomyoma and female fertility. Taiwan J Obstet \nGynecol 48:232–238\n 3. Mathias RA, Simpsion RJ (2009) Towards understanding epithe-\nlial-mesenchymal transition: a proteomics perspective. Biochim \nBiophys Acta 1794:1325–1331\n 4. Voulgari A, Pintzas A (2009) Epithelial-mesenchymal transi-\ntion in cancer metastasis: mechanisms, markers and strategies to \novercome drug resistance in the clinic. Biochim Biophys Acta \n1796:75–90\n 5. Zhou S, Yi T, Liu R, Bian C, Qi X, He X, Wang K, Li J, Zhao \nX, Huang C, Wei Y (2012) Proteomics identification of annexin \nA2 as a key mediator in the metastasis and proangiogenesis of \nendometrial cells in human adenomyosis. Mol Cell Proteomics \n11:M112017988\n 6. Alonso-Alconada L, Santacana M, Garcia-Sanz P, Muinelo-\nRomay L, Colas E, Mirantes C, Monge M, Cueva J, Oliva E, \nSoslow RA, Lopez MA, Palacios J, Prat J, Valls J, Krakstad C, \nSalvesen H, Gil-Moreno A, Lopez-Lopez R, Dolcet X, Moreno-\nBueno G, Reventos J, Matias-Guiu X, Abal M (2015) Annexin-A2 \nas predictor biomarker of recurrent disease in endometrial cancer. \nInt J Cancer 136:1863–1873\n 7. Pianta A, Drouin EE, Crowley JT, Arvikar S, Strle K, Costello CE, \nSteere AC (2015) Annexin A2 is a target of autoimmune T and B \ncell responses associated with synovial fibroblast proliferation in \npatients with antibiotic-refractory Lyme arthritis. Clin Immunol \n160:336–341\n 8. Zhou X, Deng S, Liu H, Liu Y, Yang Z, Xing T, Jing B, Zhang X \n(2015) Knockdown of ubiquitin protein ligase E3A affects prolif-\neration and invasion, and induces apoptosis of breast cancer cells \nthrough regulation of annexin A2. Mol Med Rep 12:1107–1113\n 9. Onishi M, Ichikawa T, Kurozumi K, Inoue S, Maruo T, Otani Y, \nFujii K, Ishida J, Shimazu Y, Yoshida K, Michiue H, Antonio \nChiocca E, Date I (2015) Annexin A2 regulates angiogenesis and \ninvasion phenotypes of malignant glioma. Brain Tumor Pathol \n32:184–194\n 10. Deng S, Wang J, Hou L, Li J, Chen G, Jing B, Zhang X, Yang Z \n(2013) Annexin A1, A2, A4 and A5 play important roles in breast \ncancer, pancreatic cancer and laryngeal carcinoma, alone and/or \nsynergistically. Oncol Lett 5:107–112\n 11. Mattern J, Koomägi R, Volm M (1996) Association of vascular \nendothelial growth factor expression with intratumoral microves-\nsel density and tumour cell proliferation in human epidermoid \nlung carcinoma. Br J Cancer 73:931–934\n 12. Zhao ZS, Zhou JL, Yao GY, Ru GQ, Ma J, Ruan J (2005) Cor -\nrelative studies on bFGF mRNA and MMP-9 mRNA expressions \nwith microvascular density, progression, and prognosis of gastric \ncarcinomas. World J Gastroenterol 11:3227–3233\n 13. Hawker GA, Mian S, Kendzerska T, French M (2011) Measures of \nadult pain: visual analog scale for pain (vas pain), numeric rating \nscale for pain (nrs pain), mcgill pain questionnaire (mpq), short-\nform mcgill pain questionnaire (sf-mpq), chronic pain grade scale \n(cpgs), short form-36 bodily pain scale (sf-36 bps), and measure \nof intermittent and constant osteoarthritis pain (icoap). Arthritis \nCare Res 63:S240–S252\n 14. Sharma MC, Sharma M (2007) The role of annexin II in angio-\ngenesis and tumor progression: a potential therapeutic target. Curr \nPharm Des 13:3568–3575\n 15. Zhang X, Liu S, Guo C, Zong J, Sun MZ (2012) The association \nof annexin A2 and cancers. Clin Transl Oncol 14:634–640\n 16. Levgur M, Abadi MA, Tucker A (2000) Adenomyosis: symp-\ntoms, histology and pregnancy terminations. Obstet Gynecol \n95:688–691\n 17. Wu MH, Chuang PC, Lin YJ, Tsai SJ (2013) Suppression of \nannexin A2 by prostaglandin  E\n2 impairs phagocytic ability of peri-\ntoneal macrophages in women with endometriosis. Hum Reprod \n28:1045–1053\nPublisher’s Note Springer Nature remains neutral with regard to \njurisdictional claims in published maps and institutional affiliations.","source_license":"CC0","license_restricted":false}