{"paper_id":"5b25db7d-e4e2-43eb-9a38-46bad821c8fb","body_text":"Physiomimetic Models of Adenomyosis\nJuan S. Gnecco, PhD 1,2 Alex T. Brown, PhD 1,2 Ellen L. Kan, BS 1,2 Lauren Baugh, PhD 1,2\nClara Ives, BS 1,2 Megan Loring, MD 1,3 Linda G. Grif fith, PhD 1,2\n1 Center for Gynepathology Research, Massachusetts Institute of\nTechnology, Cambridge, Massachusetts\n2 Department of Biological Engineering, Massachusetts Institute of\nTechnology, Cambridge, Massachusetts\n3 Endometriosis and Adenomyosis Care Collaborative, Center for\nMinimally Invasive Gynecologic Surgery, Newton Wellesley Hospital,\nNewton, Massachusetts\nSemin Reprod Med 2020;38:179 –196\nAddress for correspondence Linda G. Grif fith, PhD, Massachusetts\nInstitute of Technology, Room 16 -429, 77 Massachusetts Avenue,\nCambridge, MA 02139 (e-mail: griff@mit.edu).\nDeconstructing Adenomyosis:\nWhat to Model?\nIn constructing any in vitro disease model, thefirst question is\n“what to model.” This question is often relatively easily answered for\ndiseases with germ line genetic mutations and clearly defined\nclinical phenotypes directly linked to the molecular mutation, such\nas cysticfibrosis. In such cases, high throughput phenotypic screens\nin cell lines, based on correcting a known molecular defect, have\nbeen remarkably effective in identifying blockbuster clinical thera-\npies.\n1 Adenomyosis is arguably at the opposite end of the spectrum.\nFirst, the clinical landscape is convoluted: adenomyosis is\nover 10 times more prevalent than most single-gene defect\ndiseases; symptoms overlap and are comorbid with myriad\nother diseases and conditions; there is no consensus on the\ndefinition of “disease” nor a clearly de fined set of metrics to\nKeywords\n► adenomyosis\n► organoids\n► models of\nadenomyosis\n► tissue engineering\n► microfluidic device\nAbstract Adenomyosis remains an enigmatic disease in the clinical and research communities.\nThe high prevalence, diversity of morphological and symptomatic presentations, array\nof potential etiological explanations, and variable response to existing interventions\nsuggest that different subgroups of patients with distinguishable mechanistic drivers\nof disease may exist. These factors, combined with the weak links to genetic\npredisposition, make the entire spectrum of the human condition challenging to\nmodel in animals. Here, after an overview o f current approaches, a vision for applying\nphysiomimetic modeling to adenomyosis is presented. Physiomimetics combines a\nsystem’s biology analysis of patient populations to generate hypotheses about\nmechanistic bases for strati fication with in vitro patient avatars to test these hypothe-\nses. A substantial foundation for three-dimensional (3D) tissue engineering of adeno-\nmyosis lesions exists in several disparate areas: epithelial organoid technology;\nsynthetic biomaterials matrices for epithelial –stromal coculture; smooth muscle 3D\ntissue engineering; and microvascular tissue engineering. These approaches can\npotentially be combined with micro fluidic platform technologies to model the lesion\nmicroenvironment and can potentially be coupled to other microorgan systems to\nexamine systemic effects. In vitro patient-derived models are constructed to answer\nspecific questions leading to target identi fication and validation in a manner that\ninforms preclinical research and u ltimately clinical trial design.\nIssue Theme Adenomyosis and\nEndometriosis; Guest Editors,\nLinda C. Giudice, MD, PhD, MSc,\nLisa M. Halvorson, MD, Elizabeth\nA. Stewart, MD, and Luk\nR o m b a u t s ,P h D ,F R A N Z C O G ,M D\nDOI https://doi.org/\n10.1055/s-0040-1719084.\nISSN 1526-8004.\nCopyright © 2020 by Thieme Medical\nPublishers, Inc., 333 Seventh Avenue,\nNew York, NY 10001, USA.\nTel: +1(212) 760-0888.\nTHIEME\n179\nArticle published online: 2020-11-09\n\ndiagnose it and assess response to therapy; and diagnostic bias\ntoward older multiparous patients clouds true incidence. 2–5\nPatients with pain, heavy menstrual bleeding, and/or infertili-\nty may have adenomyosis—or something else entirely—as an\nunderlying cause.6 About 40% of patients with symptoms, but\nno indication of adenomyosis on magnetic resonance imaging\n(MRI) or ultrasound, are found to have adenomyosis lesions\nupon histopathological assessment.7 Symptom-targeting ther-\napies for adenomyosis, like those for endometriosis, 8 work\nremarkably well in some patients, but are not at all effective or\ntolerated in others. Moreover, there is neither reliable manner\nto predict who will respond nor alternative therapies (other\nthan developing coping mechanisms around eternal suffering)\nfor those who do not.\n9,10 As such, adenomyosis is not one\ndisease, and it is unlikely there is one model that captures all\nsalient features for all patients.\nSecond, the etiologies —given the prevalence and vast\ndiversity in patient presentations, there is likely more than\none—are obscure: adenomyosis appears to arise from a\ncomplex web of gene –environment interactions that may\nstart in utero; it might be triggered perinatally via seeding of\nthe myometrium by cells that escape the endometrium; it\nmight arise later in life due to uterine mechanical injury; or it\nmight arise due to invasion of the myometrium by endome-\ntrium injured by an in flammatory insult.\n2,4,11–13 This diver-\nsity of possible etiologies, along with the diversity of clinical\npresentations, underscores a shortcoming in how the disease\nresearch has traditionally been approached: adenomyosis is\n“a” disease. Yet, every patient is different. While there are not\nlikely millions of individual diseases, there are likely multiple\ndifferent constellations of molecular networks —immune\nnetworks, invasion networks, and metabolic networks —\nthat go awry to instigate symptomatic and phenotypic\nappearance of ectopic endometrium in the myometrium,\ngiving rise to groups of patients who could potentially be\nstratified molecularly as has been proposed for endometri-\nosis.\n8,9,14–17 Animal models—where strain variation, genetic\nperturbations, and targeted interventions to modulate spe-\ncific pathways can be deployed to probe potential contribu-\ntions to the adenomyotic etiology —offer insights into where\nto look for such mechanistic strati fication in humans.\nThis last point brings us to the third big challenge with\nmodeling adenomyosis—the relative lack of deep clinical and\npathological phenotyping to guide patient strati fication into\ntractable and mechanistically targetable subgroups. 5 Such\nclinical phenotyping and disease staging/classi fication is\nintertwined with therapeutic development through impacts\non reimbursements for treatment 18 and evaluation of effica-\ncy of various forms of treatment for clusters of patients with\nwell-defined characteristics, as is done most thoroughly for\ncancer.\n19,20 In this way, adenomyosis is the sister, though not\nprecisely the twin, of endometriosis, which is inadequately\nstaged based on lesion number, size, and location.\n8 Like\nendometriosis, adenomyosis is currently only de finitively\ndiagnosed via surgical intervention (usually, a hysterectomy)\nand histopathological characterization. However, adeno-\nmyosis, as one of several pathologies that contribute to\nabnormal uterine bleeding (AUB), is included in the PALM-\nCOEIN system for classifying such disorders, as first de-\nscribed in 2011\n6 and updated in 2018. 21 PALM-COEIN\nserves as a crucial tool for patient –clinician–researcher\ncommunication; hence, it is a step toward personalized\nmedicine for the patients suffering from AUB as a symptom\nof adenomyosis.\nThe state of personalized medicine for adenomyosis is still\ndistant from as compared with cancer and other diseases. In\nthis era of molecular strati fication, oncologists routinely\ntailor treatments to a combination of physical features in\naddition to molecular biosignatures linked to disease mech-\nanism and prognosis, sometimes using patient-derived orga-\nnoids or tissue models to test drug sensitivity.\n22 Mechanistic\nmarkers also inform the development of preclinical in vitro\nmodels, providing scientists an essential connection to the\ntranslational medicine. Such approaches are nascent only for\nendometriosis and even more primitive for adenomyosis, in\npart because somatic mutations, which are highly informa-\ntive in cancer, are at best weakly associated with ectopic\nendometrial diseases. Hence, classifications based on protein\nor metabolic network states (or possibly epigenetics), com-\nplemented with clinical phenotypic data incorporating pa-\ntient symptoms and life-long prognosis,\n17,23,24 are therefore\nan appealing route to classi fication; such approaches will\nlikely yield constellations of “molecular signatures ” of dis-\nease state, rather than one single indicator of “diseased, or\nnot.”14 An exemplary model for how other noncancer im-\nmune-mediated disease communities have spurred such\nphenotyping is the Juvenile Diabetes Research Foundation-\nsponsored Network for Pancreatic Organ Donors with Dia-\nbetes (nPOD) program.\n25 Application of programs like these\nwould significantly improve our understanding of adenomy-\notic disease.\nFinally, a fourth challenge is which stage(s) of disease to\nmodel. Modeling the features of disease as they present in\ndiagnosed patients, who often have advanced varieties of\ndisease progression, is arguably essential for identifying\nmolecular targets and developing new therapies to treat\nthese patients, and even more so for judicious prescription\nof existing therapies. Modeling proposed etiologies may\nyield preventive measures, or early treatment options, pre-\nsuming a diagnostic for nascent stage disease, can be defined.\nThe features of such models —and the throughputs required\nfor getting useful information out of them —are somewhat\nbut not entirely overlapping. Patients with advanced disease\nmay also have nascent progressing disease, for example.\nPrecise de finition of the clinical phenotype, along with\nmechanistic hypotheses about the features of disease being\nstudied, as they relate to the clinic , is an essential first step in\ndesigning, and then in implementing, in vitro models of\nadenomyosis.\nA Roadmap of Phenomena to Model\nHow do we de fine what to model? Adenomyosis studies\ngenerally lack the kind of deep clinical phenotyping that links\nmolecular and cellular pathological findings to patient symp-\ntoms, comorbidities, and treatment responses. However, several\nSeminars in Reproductive Medicine Vol. 38 No. 2-3/2020\nPhysiomimetic Models of Adenomyosis Gnecco et al.180\n\n\ncommonly observed pathological phenotypes can be modeled\nnow using in vitro models, with a longer-term goal of more\nspecialized patient avatars focusing on specifics u b g r o u p s .T h e\noverall vision for the role of in vitro models is illustrated\nin ►Fig. 1 .B r i efly, we envision that there are a relatively small\nnumber of subsets of patients who are phenomenologically\nsimilar, but distinct enough from each other that different\ntherapeutic approaches are needed. Tissues and body fluids\nharvested from patients are used on the one hand for “omics”\nanalysis—which can include genomics, proteomics, metabolo-\nmics—of molecular/cellular networks that enable stratification\nof patients into groups with data-informed machine learning\nmethods,\n14 based at least in part on mechanistic hypotheses\nthat may distinguish the groups. Mechanistic strati fication\noffers a path to target identification and experimental hypothe-\nsis testing. On the other hand, tissue banks created from the\npatients are used to create patient avatars that can be used to\ntest hypotheses about mechanistic strati fication and targets,\nwith the goal of informing translational research and employing\npersonalized therapies ( ►Fig. 1 ). In vitro models can thus\npotentially inform directions in clinical phenotyping, in an\niterative fashion.\nConceptualizing Models of Adenomyosis\n“Physiomimetics” is the process of conceptualizing a complex\nphysiological process to define essential features, then building\nin vitro models that capture the most relevant aspects of\nphysiology in ways that can ultimately yield clinically action-\nable outcomes. Physiomimetic models encompass a range of\nexperimental complexities, from relatively standard culture of\nindividual cell types up through complex three-dimensional\n(3D) microfluidic models, driven by biological questions. We\nstart here by conceptualizing established adenomyosis lesion\nphenotypes and the dynamic behaviors implicated in their\npathophysiology, as the etiology of nascent adenomyotic\nlesions is still obscure, and there is great need to treat\nFig. 1 Physiomimetic approach for developing targeted therapies for adenomyosis . Hypotheses regarding different mechanisms of disease\nthat may be operative in patient sub groups are tested with tissue and fluid samples from a large patient population containing the subgroups (1);\nanalysis of tissue and fluid samples to re fine hypotheses about mechanism (2a and 3) are performed in tandem with development of cell banks\nand construction of patient avatars (2b). Mechanistic hypothesis about patient subgroups and interventions can then be tested in patient\navatars representing the subgroups to de fine strati fied clinical trials, or inform more jud icious use of existing therapies.\nSeminars in Reproductive Medicine Vol. 38 No. 2-3/2020\nPhysiomimetic Models of Adenomyosis Gnecco et al. 181\n\n\nestablished adenomyotic lesions. Furthermore, the mecha-\nnisms contributing to disease progression may be operative\nto some extent in disease onset. We focus on diffuse adeno-\nmyosis lesions, as cystic adenomyosis lesions can in some\ninstances be treated surgically.\n26 Conceptually, here we define\nadenomyoticlesion as the ectopic presence of lesion-initiating\ncells—endometrial epithelial glands and stroma —surrounded\nby thelesion microenvironmentcomprising myometrial muscle\ncells, nerves, vasculature, and immune cells with associated\nextracellular matrix (ECM; ►Fig. 2 ).\nDynamic behaviors of lesion-initiating cells that are im-\nplicated in pathologies include (1) invasion into surrounding\nmicroenvironment, potentially seeding new, distant lesions;\n(2) proliferation, which may drive lesion growth, invasion, or\nfeed into a hormone-driven cycle of cell death and prolifera-\ntion; and (3) production of in flammatory cytokines, chemo-\nkines, proteases, and other modulators of signaling and\nremodeling of the local microenvironment\n27,28 (►Fig. 2 ).\nThe dynamic responses of the microenvironment associated\nwith pathologies include (1) smooth muscle hyperplasia of\nlocal myometrial cells (or recruitment of fibroblasts or\nmesenchymal stem cells that acquire that phenotype 29);\n(2) in flux and activation of immune cells; (3) recruitment\nof vasculature and modulation of vascular permeability,\npossibly with local bleeding and clotting; (4) enhancement\nof sensory innervation and regression of sympathetic inner-\nvation; and (5) stiffening of the local microenvironment, via\nmatrix deposition and cellular changes. Cells within lesions\nrespond to local cues, including mechanical stimulation from\nmyometrial contractions, and to systemic cues, notably sex\nsteroids, but are also modulated by nutrition, systemic\ninflammation, stress hormones, and other factors. These\nresponses play out in an orchestrated fashion, as each cell\ntype expresses a unique repertoire of receptors for the cues,\nincluding estrogen and progesterone receptors.\n30–32 Many of\nthe same phenomena are the subject of substantial investi-\ngation in carcinomas, but the spotlight on dynamic changes\nspurred by cyclic variation in sex steroids is omnipresent in\nFig. 2 Conceptualization of an adenomyosis lesion, showing the biological components and pathological processes to consider in building an in\nvitro model.\nSeminars in Reproductive Medicine Vol. 38 No. 2-3/2020\nPhysiomimetic Models of Adenomyosis Gnecco et al.182\n\n\nadenomyosis. After a review of the known features of the\nmyometrium and its changes in adenomyosis, we then\nconsider how each of these dynamic phenomena can be\nmodeled in vitro.\nThe Uterus in Adenomyosis\nThe presence of adenomyosis in the myometrium induces\nchanges throughout the uterus in ways that suggest interest-\ning hypotheses that can be tested using well-designed in\nvitro models. The myometrium is composed of the inner\nmyometrium (or junctional zone, JZ), characterized as having\na relatively greater cell density with relatively less ECM, and\nthe outer myometrium, which is more ECM-rich and has a\nlower cell density.\n33 On a cellular level, the myometrium\ncomprises layers of smooth muscle cells (SMCs) arranged\ninto fibular bundles approximately 0.3 mm in diameter,\ninterwoven with blood vessels and connective, lymphatic,\nand nerve tissue.\n33,34 These layers of SMC bundles\nare oriented in different directions around the uterus to\nenable complex contractile patterns that drive fluid flow in\nthe nonpregnant uterus. 33 In adenomyosis, these contractile\nfunctions can become dysregulated. 35,36 Like the endome-\ntrium, the inner myometrium/JZ exhibits molecular and\ncellular changes in response to cyclic hormones through\nthe menstrual cycle —such as tissue thickening and modula-\ntion of estrogen and progesterone receptor expression —but\nthe outer myometrium appears less sensitive to menstrual\ncycle hormonal changes at both the gross anatomical level\nand the molecular level.\n37–39 How these differences translate\ninto the propensity for adenomyosis lesions to drive symp-\ntoms is unknown, but it is a potential facet that could be\nmodeled in vitro. The healthy human myometrium is typi-\ncally well delineated from the endometrium, but this clear\ndelineation is blurred in adenomyosis, suggesting invasion of\ncells from the endometrium into the myometrium, attribut-\ned by one theory to microtraumas at the endometrial –\nmyometrial interface during uterine peristalsis.\n11,38,40–43\nChanges in ECM architecture and composition, as well as\nsmooth muscle hyperplasia of the region immediately\nsurrounding lesions, characteristically affect multiple cell\ntypes and may involve chemical and mechanical cues.\nPatients with adenomyosis have been shown to have in-\ncreased levels of matrix metalloproteinases (MMP)-2 and\n-9, suggesting abnormal matrix remodeling may be present\nin these patients.\n44 Inflammation may be driven by repeat-\ned local bleeding in the lesion, 45 which may further en-\nhance contractile phenotypes that in turn drive fibrogenic\nprocesses.46\nEndometrial cells in adenomyotic tissue exhibit many aber-\nrant behaviors, including apparent resistance to progesterone,\nepithelial–mesenchymal transitions, decreased apoptosis re-\ngardless of menstrual phase, and other processes.11 Although\nadenomyosis-targeting therapeutic approaches are intended\nto suppress aberrant proliferation of endometrial-derived tis-\nsues by targeting proestrogenic pathways or inducing proges-\nterone action,\n11 these approaches fail in some patients, a\nfinding consistent with dysregulation of speci fich o r m o n e\nreceptors ESR1, ESR2, and PGR isoforms A and B in adenomyo-\nsis.47,48 The downstream consequences of estrogenic signaling\nand acquired progesterone resistance that may contribute to\nthe etiology of disease are still not well understood, and in vitro\nmodels offer an opportunity to parse the individual\nmechanisms.\nAdenomyosis is associated with both pain and heavy men-\nstrual bleeding. Several studies have reported an increased\nsensory nerve fiber density in both ectopic lesions and eutopic\nendometrium in women withendometriosis, along with reduc -\ntion in sympathetic nerves, suggesting a possible diagnostic\ncriterion.\n49–52 Nerve fibers are increased in the fibrotic endo-\nmetriosis nodules in the rectovaginal space and greater nerve\nfiber in filtration of lesions is found in patients with greater\npain.53–55 Increased nervefiber density is also associated with\nan increase in macrophages,56 with some evidence of estrogen\nmodulation of the crosstalk. 57 Similar analysis of nerve fiber\nand macrophage density inadenomyosisis relatively scarce, but\none study examined nerve fiber density in the endometrium\nand myometrium of women with adenomyosis and fibroids,\nand found no difference.58\nThe immune environment in the uterus is dynamic, with\ndramatic increases in the number of macrophages, uterine\nnatural killer (NK) cells, and other immune components\nduring the secretory phase. The extent to which the\nincreased number of immune cells arises from in situ prolif-\neration of tissue-resident cells or from recruitment of circu-\nlating monocytes is still unclear.\n59–61 Furthermore, whether\nCD45þ intraepithelial lymphocytes (IELs) that are character-\nistically observed in the epithelial layer in adenomyosis\nlesions\n62 derive from locally recruited cells (and are there-\nfore part of the microenvironment) or derive from IELs\npresent in the epithelial layer of the eutopic endometrium\n(and are therefore lesion-initiating cells) is unknown. The\ncrosstalk between nerves and macrophages or other immune\ncells may contribute to pain and/or bleeding. An intriguing\nclinical finding, albeit from a small study, examined macro-\nphage density in the uteri of 54 patients undergoing hyster-\nectomy for adenomyosis due to pain as the primary\nsymptom, with a comparison group of 20 women undergo-\ning hysterectomy for adenomyosis with heavy menstrual\nbleeding as the primary symptom,\n63 where adenomyosis\nwas tentatively diagnosed by both transvaginal ultrasound\nand MRI prior to surgery. Greater leukocyte in filtration was\nseen in both the myometrium and endometrium of patients\nwith pain, and groups of macrophages were seen near lesions\nand interspersed in the muscle in the patients with pain; in\nboth groups, there were macrophages around blood vessels,\nas macrophages are known to regulate vessel permeability.\n64\nAlthough the detailed interactions between the vasculature\nand adenomyotic lesions are not fully appreciated in the\nestablishment and progression of adenomyosis, increased\nmicrovessel density has been observed in adenomyosis, 65,66\nand a recent meta-analysis of studies examining vascular\nmorphology and marker expression in the myometrium and\nendometrium suggests increased angiogenesis in the endo-\nmetrium of women with adenomyosis.\n67 Early studies that\nexplored immune responses in adenomyosis showed that\nthe disruption of the JZ at the endometrial –myometrial\nSeminars in Reproductive Medicine Vol. 38 No. 2-3/2020\nPhysiomimetic Models of Adenomyosis Gnecco et al. 183\n\n\ninterface may result from the abnormal aggregation of\nmacrophages in the myometrium, similar to the increased\nnumber of macrophages previously observed in endometri-\nosis patients.\n68 Finally, intriguing data on sex hormone\nresponsiveness of mast cells, which are among the most\ncommon immune cell types in the myometrium and impli-\ncated in the pain associated with interstitial cystitis, may\ncontribute to adenomyosis symptoms.\n69 These observations\nsuggest that recapitulating immune –epithelial–neuronal\ninteractions in vitro using patient-speci fic cells may yield\ninsights between symptomology and molecular and cellular\nbehaviors characteristic of distinct patient populations.\nInterestingly, the eutopic endometrium of patients with\nadenomyosis confer dysregul ated angiogenesis, innerva-\ntion, immune cell presence, and gene and protein expres-\nsion that suggest greater invasive and survival potential at\nectopic sites and contribution to symptoms.\n70–74 Global\nchanges in the myometrium outside the peri-lesion envi-\nronment, including smooth muscle hypertrophy, increased\nstiffness, and immune cell in filtration,38,75 raise questions\nabout whether these global changes precede, or are the\nresult of, lesion establishment —a hypothesis that could\npotentially be tested in vitro.\nDesigning, Building, and Interpreting In\nVitro Models\nThe growing evidence that animal models do not capture\nessential features of human diseases involving chronic inflam-\nmation, sex steroid signaling, or other complex, molecular\nphenomena is driving development of sophisticated in vitro\nmodels that capture complex tissue architecture, and even\ndynamic perfusion of microvascular networks, using human\npatient-derived cells. While these approaches are just now\nemerging, especially driven by applications in cancer, cardio-\nvascular, and liver diseases, they offer promise for modeling\nadenomyosis. In this section, we review basic approaches to\ntissue engineering of the uterine environment, including\ndiscussion of cell sourcing and 3D model design. Then, in the\nnext section, we turn to modeling specific phenotypic facets of\nadenomyosis including invasion, vascular and systemic inter-\nactions, mechanics, and innervation, drawing from impactful\nwork in disparate tissue systems.\nBasic Approaches in Tissue Engineering of the\nEndometrium and Myometrium\nThe functions of the uterus arise from the integrated actions\nof epithelia, stromal fibroblasts, myometrial smooth muscle,\nimmune cells, and endothelial cells, which are each directly\nand indirectly responsive to variations in sex hormones.\nThus, many responses of various cell types to hormone\nchanges are indirect and governed by dynamic communica-\ntions with other cells. The endometrial epithelial response to\nprogesterone, for example, is governed by signaling proteins\nproduced by the progesterone-responsive stromal fibro-\nblasts, which in turn respond to endothelial-derived signals\nto enhance the decidualization process.\n31,76 While parsing\nresponses of individual cell types to hormonal and other\nperturbations in vitro is an essential first step to understand-\ning the integrated responses, building models that capture\nthe dynamic interlinked hormone responsiveness in vitro is\nultimately essential for illuminating complex phenomena\nlike adenomyosis.\nCell Sources: Explants, Cell Lines, Primary Cells\nFor decades, explant cultures of the endometrium and myome-\ntrium were the only external window into the functions of the\nhuman uterus in vitro\n36,77–81 and, until recent advances in\nprimary cell culture, they remained preferred models for anal-\nysis of complex primary tissue functions.\n31,81,82 Endometrial\nexplants in the absence of ECM quickly deteriorate,83 but can\nundergo outgrowth and remodeling when embedded in afibrin\nmatrix.84 For the endometrium —the presumed instigator of\nadenomyosis lesions—definition of protocols for isolating and\nculturing epithelial cells and stromalfibroblasts from endome-\ntrial biopsies85 enabled mechanistic studies on patient-derived\nsamples, allowing diverse healthy and diseased donors to be\ncombined in different ways to parse contributions of each cell\ntype to health and disease. Primary cells remain the gold\nstandard for mechanistic disease studies, but permanent epi-\nthelial\n86,87and stromal88–90 cell lines that exhibit many (but not\nall) tissue-speci fic hormone-responsive features, and some\ndisease states, have been routinely used for protocol develop-\nment, pilot studies, and investigations of general features of\nendometrial cell behavior. The carcinoma-derived Ishikawa\ncell line is most commonly used for “normal” epithelia,\n91–93\nthe h-TERT-immortalized endometriosis-derived“12Z” line is\nused for invasive endometrial epithelia,15,87,94 and the h-TERT-\nimmortalized stromal cell line “tHESC”89,91 is used for endo-\nmetrial stromal cells. (Caution: cell line contamination was\nreported for the endometrial epithelial cell line“HES”\n86;h e n c e ,\nroutine SNP profiling of cell lines is recommended for these, as it\ni sw i t ha l lc e l ll i n e s ,t ov a l i d a t ep r o v e n a n c e . )W h i l ec e l ll i n e s\noffer convenience and relatively good reproducibility, including\nfrom laboratory to laboratory (subject to the typical cell line\ndrift), they have some shortcomings for reproducing primary\ncell behaviors,\n91 including greatly skewed cytokine and growth\nfactor production profiles.95 Pilot studies with cell lines can,\nhowever, offer insights that can be investigated further in more\ndifficult-to-culture primary cells.15\nThe majority of studies on primary endometrial cells focus\non stromal cells, as they can readily be expanded in standard\nculture and frozen to create tissue banks, allowing repeated\nstudies from the same donor within limits of passage num-\nber.\n96,97 For years, however, investigators were limited by the\nlack of robust protocols for expansion and cryopreservation of\nendometrial epithelial cells, with few laboratories reporting\nsuccess.\n98 The landscape changed dramatically in 2017 with\npublication of the first two studies describing robust expan-\nsion and cryopreservation of human endometrial epithelial\ncells as organoids (\n►Fig. 3a ), with retention of hormone\nresponsiveness, cellular architecture, polarity, heterogeneity,\nand other functions, 99,100 using modi fications of protocols\nfirst described by Clevers and coworkers for expansion of\nhuman intestinal epithelia.101 In the Clevers approach, epithe-\nlial cells are either dispersed as individual cells or fragments\nSeminars in Reproductive Medicine Vol. 38 No. 2-3/2020\nPhysiomimetic Models of Adenomyosis Gnecco et al.184\n\n\nand cultured in Matrigel, a basement membrane-rich isolate of\na tumor cell line propagated in mice. 101 While Matrigel had\nlong been used as an ECM for 3D culture of endometrial\nglands99,102 and many other types of epithelial cells, enabling\nremarkable retention of tissue phenotype, Clevers de fined a\ncocktail of growth factors and other medium supplements that\nspur dramatic proliferation of the stem cell compartment,\nallowing the cells in an initial biopsy to expand orders of\nmagnitude in culture and withstand cryopreservation, while\npreserving their ability to differentiate into proper tissue\nphenotypes upon a change in medium composition to remove\nstem cell cues.\n101 This organoid approach has also been\napplied to derive continuously propagated organoid cultures\nfrom ectopic endometrium and endometrial tumors. 103\nAn additional advance essential to building and character-\nizing physiologically relevant 3D models is the identification of\nboth mesenchymal and epithelial stem and progenitor cell\ncompartments within the human endometrium. 96,104–106\nIntriguingly, recent data suggest that stem or at least progeni-\ntor cells are present in the endometrial functional layer and\nluminal regions of the endometrium, as evidenced by in situ\nhybridization for expression of the epithelial stem cell marker\nLGR5.\n107 The tools to identify putative stem and progenitor\ncells are valuable in assessing the phenotypic states of cells that\nare used to initiate models, and to assess their phenotypes\nunder long-term culture conditions, especially to test hypoth-\neses regarding possible contributions of these stem cells to\nectopic adenomyosis lesions.\n108 Related to efforts to identify\nFig. 3 Development of 3D in vitro models using synthetic hydrog els and endometrial epithelial organoids technologies .( a)E n d o m e t r i a l\nepithelial organoids (EEOs) promote the culture, expansion, and propagation of epithelial cells using 3D hydrogel systems. EEOs retain epithelial\nstructure, heterogeneity, and function of the native tissue glands. ( b) Bio-labile, synthetic extracellular matrices (ECMs) can be designed to\nestablish EEOs cultures and cocultures th at include additional cell populations, such as endometrial stromal cells (ESCs). ( c) Polyethylene glycol\n(PEG)-derived hydrogels are fully de fined, modular, and can be tuned by modifying their molecular and biophysical properties to mimic key\nfeatures of the adenomyotic and endometriotic phenotype.\nSeminars in Reproductive Medicine Vol. 38 No. 2-3/2020\nPhysiomimetic Models of Adenomyosis Gnecco et al. 185\n\n\nand characterize functions of stem and progenitor cells in\npostnatal endometrium are efforts to create uterine tissue de\nnovo from induced pluripotent stem cells (iPSCs), which could\npotentially advance patient-specific disease modeling as they\nhave for other diseases. Encouragingly, a protocol for creation\nof human stromal cells with responses to progesterone char-\nacteristic of endometrial stromal cells undergoing deciduali-\nzation in vitro has been developed,\n109 providing a foundation\nfor future efforts to reproduce other uterine tissues.\nTo model the pathogenesis of adenomyosis, it is important\nto consider its ectopic environment in addition to the lesion-\ninitiating cells. The myometrium comprises several cell\ntypes, but the major focus in in vitro models is on myometrial\nSMCs, which have different properties in the inner (JZ) and\nouter myometrium and express different receptor and sig-\nnaling repertoires.\n33,37,110 Protocols for deriving and cultur-\ning primary SMCs developed for many other tissues\n(vascular, intestinal, etc.) facilitated development of robust\nprotocols for isolation and culture of myometrial SMCs,\nwhich are commonly obtained from hysterectomy or caesar-\nean section specimens. Cell origin (geographic location,\npregnant or nonpregnant uterus) is an important consider-\nation in de fining primary cell source for myometrium. Per-\nmanent myometrial cell lines are also widely used for\ncircumscribed studies such as screening drug actions.\n111–113\nFinally, all studies with patient-derived cells must carefully\nconsider the donor pro files in design and interpretation of\nexperiments. Age and parity may influence donor cell pheno-\ntype. Unlike endometriosis lesions, which can be collected\nfrom even very young patients, almost all myometrial tissue,\nalong with adenomyosis lesion tissue, is derived from hyster-\nectomy specimens. Although protocols for resecting and biop-\nsying adenomyotic tissue have been described,\n7,26,114 these\nprocedures are not widely practiced. Thus, significant barriers\nexist in creating tissue banks from young patients who suffer\nfrom adenomyosis.\nGeneral Approaches to 3D Tissue Engineering of\nEndometrium and Myometrium Using Natural and\nSynthetic ECMs\nAn enduring challenge at present is to de fine suitable con-\nditions to sustain long-term cocultures of multiple endome-\ntrial cell types that each has their own ECM and growth factor\nrequirements. Matrigel comprises a mix of proteins and\nproteoglycans typical of the epithelial basement membrane,\nwhich filters proteins from the stromal compartment that\nrequires a type I collagen and fibronectin-rich ECM with a\nvery different composition of proteoglycans and associated\nmatrix-bound growth factors.\n115 The first 3D primary cell\nendometrial coculture model, tailored for studying blasto-\ncyst implantation, addressed this by embedding stromal\nfibroblasts in a collagen gel, mimicking features of the\nstromal matrix, which was coated by a thin layer of Matrigel\nonto which the epithelial cells were seeded, resulting in a\nconfluent well-differentiated epithelia monolayer compris-\ning ciliated and secretory (glandular) epithelia. 116 Versions\nof this model using cell lines, with stromal cells in collagen\ntopped with Matrigel, were adapted to mimic the upregula-\ntion of MMPs and accompanying matrix degradation follow-\ning progesterone withdrawal as a trigger of\nmenstruation\n28,117 as well as some features of the menstrual\nresponse to hormones, and other shorter-term aspects of\nstromal epithelial communication. 91,118 In a simpler version,\nin which stromal cells were embedded in Matrigel with\nepithelial cells on top, epithelial cells underwent 3D mor-\nphogenesis and proliferated, but stromal cells were relatively\nnonproliferative within the gel,\n119 suggesting the need for\ncell-type–tailored ECM. An alternate approach, adapted from\na successful clinical matrix for healing full-thickness (der-\nmal–epidermal) wounds in skin, is to seed stromal cells in a\nporous collagen matrix fabricated by freezing and lyophiliz-\ning a type I collagen solution, and overlaying the matrix with\nepithelial cells, which produce their own basement\nmembrane.\n120\nThese models employing natural ECMs have shortcomings,\nin part because there are no “one-size-fits-all” ECMs for both\nepithelia and stroma (and potentially other cell types), and also\nbecause natural proteins are impure (Matrigel has myriad\ngrowth factors); are substantially variable from lot-to-lot; are\nrelatively rapidly degraded by cells; and are subject to variation\nin structural and mechanical properties depending on how\nquickly or slowly the gels are polymerized.\n115,121 Variation is\nfurther exacerbated as some investigators use atelocollagen,\nwhich yields very different outcomes than whole collagen.122\nThese shortcomings have prompted a sustained effort in the\nbiomaterials community to create synthetic alternatives to\nMatrigel and collagen.123,124 A popular approach exploits the\nrelative inertness of poly(ethylene glycol) (PEG), which is\ncommercially available in multi-arm star/branched configura-\ntions activated with cell-compatible reactive groups, to create\nmodular cell-encapsulating hydrogels crosslinked with short\nmatrix metalloprotease (MMP)-sensitive peptides and incor-\nporating synthetic integrin-binding motifs.\n121,123–126 Other\ntypes of synthetic gel ECMs based entirely on synthetic proteins\nor semisynthetic gels based on modi fied hyaluronic acid are\nalso widely used.124,127 The local and bulk mechanical modu-\nlus, permeability, degradation properties, and biochemical\nrecognition motifs—all of which have been correlated to cell\nresponses115,121,125—can be tuned independently to match the\nneeds of individual cells and tissues.\nRecently, a locally responsive,“one-size-fits-all” PEG-based\nmodular synthetic ECM was developed speci fically for cocul-\nture of human endometrial epithelial and stromal cells, incor-\nporating synthetic integrin-binding peptides for both cell\ntypes, and synthetic peptides that capture and sequester the\ndifferent ECM proteins produced by epithelial and stromal\ncells.\n92,95,121 A version of this synthetic ECM also supports\nexpansion of human endometrial organoids121 and formation\nof networks by human microvascular endothelial cells,125 thus\nproviding a foundation for engineering complex 3D cocultures\n(\n►Fig. 3b). This matrix can be dissolved by a microbial enzyme\nto release cells and local cell –cell signaling molecules for\nanalysis.92,121 A relatively unsolved problem, however, is\nformulation of culture media compositions that support mul-\ntiple different cell types. Solution of this problem may require\nfiner speci fication and validation of “physiological” cell\nSeminars in Reproductive Medicine Vol. 38 No. 2-3/2020\nPhysiomimetic Models of Adenomyosis Gnecco et al.186\n\n\nphenotypes, particularly for stromal cells which are often\ngrown in serum-containing media optimized for maximal\ncell proliferation rates, rather than specific tissue functions.\nTissue engineering of the myometrium has bene fitted ex-\ntensively from the intense activity in using 3D models to\nilluminate arterial smooth muscle pathophysiology and regen-\nerate blood vessels, where the roles of basement membrane, 3D\narchitecture, cell density, and biomechanical stimulation on cell\nand tissue phenotype have been analyzed.\n128 Understanding\nand modulating contractions in the pregnant uterus, to inter-\nvene therapeutically in preterm birth, has been the major\napplication focus of in vitro 3D myometrial explant and tis-\nsue-engineered models.\n36,111–113,129–132 Tissue-engineered\nmodels employing myometrial SMCs embedded in 3D collagen\ngels, adapted from successful models for vascular applications,\noffer advantages over explants by enabling a reproducible cell\nsource from a tissue bank to be employed in multiple experi-\nments,\n131 and providing control over which cell types are\npresent, to parse contributions of various immune cell pop-\nulations, for example. An innovation involving an ECM-free\nmagnetic cell aggregation\n112 has been adapted to the myome-\ntrial contraction assays, but approaches involving synthetic\nECMs are still only nascent for other SMC applications.133\nTo summarize this section, the essential cell-based founda-\ntion for creating complex, 3D models of myometrial–endome-\ntrial interactions has been established through a combination\nof advances in basic methods of cell culture, combined with\ntechnologies for controlling the ECM microenvironment with\nsynthetic biomaterials (\n►Fig. 3c ). These tools can be further\ncombined with micro fluidic devices to model pathologic\nphenotypes in a physiomimetic sense.\nModeling the Dynamic Processes of the\nAdenomyotic Microenvironment\nThe fundamental tools for building 3D tissue models are the\nconstituent cells, scaffolds, matrices, platforms, and soluble\ncues that facilitate reconstruction or morphogenesis of\nphysiologically relevant mimics of tissues. These fundamen-\ntal tools are then deployed in specialized ways to model\nspecific dynamic processes and phenotypes. Here, we\ndescribe the processes illustrated in\n►Fig. 2 ,a n dd r a w\nfrom applications in other disease research areas to illustrate\nhow fundamental tissue engineering tools are often com-\nbined with other devices to further control elements of the\nculture microenvironment, such as mechanical stimulation,\ninnervation, or fluid flow.\nInvasion and Survival\nWhether adenomyosis initially arises from a true invasion of\nthe myometrium from the endometrium or through trans-\nport of endometrial cells into the myometrium by other\nmeans is debated, but once present, the displaced endome-\ntrial cells form lesions that grow and invade the surrounding\ntissue as the lesion enlarges. Cancer, wound healing, immune\nresponse to infection, and myriad other pathophysiological\nbehaviors—including endometriosis and adenomyosis —have\nmotivated a vast array of mechanistic experimental studies\nand biophysical models of cell migration and inva-\nsion.\n15,122,134–136 Although various authors often inter-\nchange the terms migration and invasion, here we will\nrefer to migration as a phenomenon of movement along a\ntwo-dimensional (2D) surface (e.g., endothelial cells moving\nacross a denuded region of a vessel) or a quasi-2D surface\n(highly porous 3D matrix such as a large-pore transwell\nmembrane), where no ECM degradation is required for\nmovement. In contrast, we de fine invasion as a 3D process\nwhere breakdown of matrix or cell–cell junctions is required.\nMigrating and invading cells inherently integrate an array of\nindividual molecular processes to generate biophysical forces\nresulting in cell movement, which may be random or influenced\nby chemotactic (chemical-based), durotactic (matrix stiffness-\nbased), or haptotactic (adhesion-based) gradients.\n134,136 This\nintegration can result in nonintuitive outcomes, necessitating\ncareful attention to quantitative experimental parameters and\nmetrics. For example, experimentally, cell migration speeds\nexhibit maxima for intermediate values of ECM-coating densi-\nties, across many different cell types and ECM coatings, as\npredicted by a biophysical model describing the balance\nbetween cell-matrix adhesion forces and cell-generated con-\ntractile forces.\n134 Growth factors, which modulate both cell-\nmatrix adhesion and cell contractile force generation, can\nappear to increase or decrease cell speed for a given ECM-\ncoating density, depending on which side of the biphasic curve\nis operative for the chosen condition.\n137 3D invasion assays\ninherently integrate more complex processes. The diversity in\ntypes of 3D movement, together with cell-mediated matrix\ndegradation, results in different biophysical phenomena gover-\nnance compared with that of 2D migration assays.\n122,138\nAlthough there are numerous studies focusing on migration\nand invasion of cells derived from eutopic or ectopic tissue in\npatients with endometriosis,15,139 studies on adenomyosis are\nmore limited and focused on stromal cells. 136,140,141 Consid-\neration of how the individual molecular components in the\nextracellular environment (e.g., adhesion molecules, growth\nfactors, porosity and stiffness of the matrix, matrix remodel-\ning) and intracellular environment (e.g., integrin–cytoskeletal\nlinks, actin –myosin contractile forces, signaling pathways\ngoverning protein–protein associations) are integrated bio-\nphysically provides insight into why literature reports com-\nparing adenomyotic and healthy cell migration and invasion\nare con flicting.\n136 When eutopic endometrial stromal cells\nfrom adenomyosis patients and controls were compared in an\nassay format involving migration/invasion of cells across a thin\nporous membrane coated with Matrigel for 24 hours, no\ndifferences between the two groups were observed.\n140 How-\never, when eutopic endometrial stromal cells from patients\nwith adenomyosis were compared with controls in an assay\nthat involved invasion for 10 days into a 1 cm-thick slab of type\nI collagen gel, adenomyotic cells invaded further than controls;\ninvasion of both control and adenomyotic cells was further\nenhanced when myometrial cells from control donors were\nincluded in the gel, and even further enhanced again if the\ndonor myometrial cells were from an adenomyosis patient.\n141\nFurthermore, this same assay format revealed that estradiol\nand tamoxifen drive additional increases in the invasion depth\nSeminars in Reproductive Medicine Vol. 38 No. 2-3/2020\nPhysiomimetic Models of Adenomyosis Gnecco et al. 187\n\n\nof cells in each condition, while progesterone blunts the\neffects.142 The 24-hour Matrigel invasion assay may not have\ncaptured differences between control and diseased popula-\ntions, due to a combination of the relatively short assay time,\nuse of an epithelial basement membrane-type matrix (Matri-\ngel) rather than a stromal-like matrix (collagen I), and the\npotential for the many growth factors in Matrigel to stimulate\nchemokinesis (random motility) in a way that drove maximal\ninvasion and obscured the differences. Still, whether these\nendometrial cells would invade a true smooth muscle struc -\nture, and if so, whether the factors regulating such invasion are\nthe same as those regulating migration in ECM, is an interest-\ning question for future models to address.\nThe enduring challenge, as noted at the outset of this article,\nis whether new targets for therapeutics are revealed, and if so,\nwhether drugs that treat them (as well as existing drugs) can\nbe circumscribed to certain groups of patients who might be\nidentified by accessible clinical tests. Answering these related\nquestions, especially in a manner that captures the multiple\nphenotypes leading to symptomatic pain and bleeding in\nadenomyosis, is still at an early stage of physiomimetic analysis\nfor any type of therapeutic. With respect to identifying new\ntargets, a study employing a high-throughput 3D invasion\nassay for the 12Z endometriotic cell line in collagen gels that\nexamined invasion as a function of growth factor/cytokine\nstimulation, intracellular kinase signaling pathway response,\nand proteolytic shedding of cell surface receptors and growth\nfactors\n15 might be considered relevant for adenomyosis, as the\n12Z line likely shares features with adenomyotic epithelia. As\nnoted earlier, experiments with cell lines offer insights that can\nbe followed up with primary cells. In the study, a Bayesian\nanalysis linking the recursive protease-growth factor-kinase\nsignaling loops revealed potential vulnerabilities in the net-\nwork involving, among others, JUN kinase (JNK), and this\nprediction was experimentally veri fied with inhibitors.\n15 In-\nterestingly, a JNK inhibitor was also identi fied as a driver of\ninflammation in a separate study of peritonealfluid in patients\nwith surgically veri fied endometriosis, where a subset of\npatients exhibited a constellation of signature cytokines. 14\nTwo preclinical studies indicated effectiveness of a JNK inhibi-\ntor against endometriosis,143,144 yet a clinical study in a non-\nstratified patient population (NCT01630252, NCT01631981)\nfailed to meet endpoints. (None of the patients in either study\nwere evaluated for possible adenomyosis, which is often\ncomorbid with endometriosis and may contribute to symp-\ntoms and which may also respond to therapies for endometri-\nosis.) Further exploration of this hypothesis with the types of\nmodels now being developed that include multicellular com-\npartments and robust metrics may provide insights into how\nto design a clinical trial to treat subsets of patients with certain\ncharacteristics.\nAdenomyosis invasion phenocopies features of carcino-\nma invasion into smooth muscle. Although the overwhelm-\ning majority of in vitro invasion studies have investigated\ninvasion into ECM-rich tissues, the conceptual framework\nfor building 3D invasion models in muscle has been mapped\nout over the past decade by observations in clinical speci-\nmens and intravital microscopy of carcinoma invasion in\nmuscle in mouse models.\n135 Epithelial invasion into muscle\nis associated with changes in multiple facets of cell cyto-\nskeletal and mechanosensing behaviors, driving a prefer-\nence for stiff environments.\n135 Cells can invade into muscle\nas individual cells, or as chains, clusters, or collective\nstrands, following the collagenous matrix between bundles\nor migrating along individual SMCs.\n135 Features of collec -\ntive cell invasion have been observed in deep in filtrating\nendometriosis lesions in the rectocervical space, 55 suggest-\ning the possible invasion from adenomyotic lesions ema-\nnating from the cervical region of the uterus. The role of the\nregular smooth muscle contractions in the invasion process\nis still speculative, as there are consequences for mechanical\ndamage, as well as promoting mitosis.\n135 Hence, the devel-\nopment of models that experimentally recapitulate these\nmultiple features is the goal o f adenomyosis physiomimetic\nmodeling.\nVascularization and Immune Interactions\nWith a vision toward building an adenomyosis lesion model\nthat incorporates large-scale features of the lesion microenvi-\nronment, including a vascular bed, creation of a microvascular\nnetwork becomes essential for survival and homeostasis of the\nengineered tissue. First, the microvascular system in the\nuterus, as in other organs, is essential for providing oxygen\nand nutrients. The dimensions of 3D in vitro models are\ninherently limited by diffusion of oxygen and nutrients to\napproximately 0.2 mm for metabolically active tissue like liver\nand muscle, and slightly larger for relatively acellular connec -\ntive tissues such as dermis.\n115 Second, the microvascular\nbarrier also regulates trafficking of immune cells (and possibly\ncirculating stem cells) and plays essential roles in tissue\nhomeostasis through paracrine signaling to tissues, serving\nas the nidus for commencing decidualization of the endome-\ntrium.\n145 Vascular function is modulated by systemic and local\nlevels of sex hormones and inflammatory cues.146,147 In vitro,\ncultures of endometrial stromal and epithelial cells in the\npresence of estrogen regulate angiogenesis,\n148 and signals\nfrom flow-activated endothelial cells enhance hormone re-\nsponse in endometrial stromal cells.145\nEncouragingly, a fusion of tissue engineering and micro-\nfluidic cell culture technologies has spurred development of\ndevices and protocols that could be applied to modeling\nadenomyosis lesions. Micro fluidic devices with planar cul-\ntures that probe endothelial –stromal–immune interactions\nhave already been applied to study the role of environmental\nchemicals on facets of endometriosis phenotypes.\n149 Micro-\nfluidic and mesofluidic devices incorporating 3D tissue struc -\ntures that can be perfused continuously with culture media,\nimaged, and sampled in situ are endemic tools in the emerging\nfield of “microphysiological systems (MPSs)”—in vitro repre-\nsentations of complex physiological phenomena involving\nmultiple cell types and dynamic behaviors in 3D. A crucial\nfirst step was to de fine protocols for the creation of stable,\nperfusable microvascular networks in a collagen or fibrin gel\nsituated between two channels perfused with cell culture\nmedia (\n►Fig. 4a ).150–152 Stable microvessels from endothelial\ncells andfibroblasts seeded in the central gel region (►Fig. 4b),\nSeminars in Reproductive Medicine Vol. 38 No. 2-3/2020\nPhysiomimetic Models of Adenomyosis Gnecco et al.188\n\n\nenabling phenomena involving traf ficking of immune\n(►Fig. 4c ) and tumor cells across the endothelial barrier to\nbe observed in detail ( ►Fig. 4d ).153 These foundational\napproaches have now been widely adapted to support appli-\ncations as diverse as in vitro tumor and islet microvasculari-\nzation and formation of functional blood –brain\nbarriers.\n154–157 These studies establish principles for how a\nmicrovascularized endometrial–stromal lesion could be mod-\neled; indeed, the authors are currently leveraging them to\nbuild multicellular microvascularized models of endometri-\nosis lesions, employing primary patient-derived endometrial\nepithelial, stromal, and immune cells for analysis of how\nmicrovascular permeability and immune cell recruitment\nare regulated as a function of circulating sex steroids. A model\nof vascularized muscle developed for analyzing fibrosis and\nmuscle damage\n158 could potentially be adapted for creating a\nmodel of vascularized adenomyosis in the host tissue. One\nsignificant technical challenge relevant for modeling repro-\nductive tissues is that the material most commonly used for\nfabricating micro fluidic devices, polydimethylsiloxane\n(PDMS), absorbs lipophilic compounds like estrogen and pro-\ngesterone, making it almost impossible to control their con-\ncentrations and thus tissue exposure.\n159 Among fabrication\nmaterials, PDMS is uniquely permeable to oxygen—a limiting\nnutrient in most cultures; hence, significant design changes to\nprovide oxygenation are required in transitioning to more\nsuitable materials, but these shifts are gradually occurring as\ncommercial chips based on thermoplastics are coming into\nuse.\n160\nBiomechanical Stimulation\nIt is now well recognized that the mechanical compliance of the\nECM microenvironment—“stiffness” or “softness”—dramatical-\nly influences the phenotypes of epithelial cells, stem cells, and\nother cell types.161 Matrices that are relatively stiff compared\nwith the normal mechanical properties that a cell experiences\ncan drive transition of fibroblasts to myofibroblasts and epi-\nthelial cells to a tumor-like phenotype.\n161 Similarly, tumor cells\naccustomed to a stiff environment may exhibit poor survival\nand growth when placed into a more normal mechanical\nenvironment—a factor that may contribute to the dif ficulty in\ngrowing cells from endometriosis and adenomyosis lesions.\nThese adaptations occur through a compendium of mechanical\nsignaling pathways that are integrated in part by the intracel-\nlular translocation of transcriptional regulators YAP and TAZ to\nthe nucleus, allowing differences in mechanical signaling be-\ntween two different tissue environments in vivo to be inferred\nfrom immunohistochemistry.\n162 An elegant in vivo –in vitro\nmechanical signaling study involving the myometrium first\ndemonstrated links between the mechanical properties of\nuterine leiomyomas (fibroids) relative to surrounding myome-\ntrium and the corresponding relative amounts of nuclear YAP/\nTAZ (via immunohistochemistry), and then showed that isolat-\ned myometrial SMCs or fibroid cells cultured in 2D on a set of\nsynthetic hydrogel substrates with systematically varied me-\nchanical stiffness recapitulated the trends in nuclear YAP/TAZ\ns e e nf o re a c hc e l lt y p ei nv i v o ,a l o n gw i t he n h a n c e dE C M\ndeposition by fibroid cells as seen in vivo.\n162 This in vivo –in\nvitro correspondence suggests that the in vitro model may be a\nuseful proxy for analyzing potential therapeutic interventions\ninto mechanical signaling pathways. Other behaviors, such as\ninvasion/survival, may require a 3D environment to capture\nthese in vivo behaviors. Toward this goal, we have observed that\nnormal human endometrial epithelial cells, which exhibit a\nprototypical spherical organoid morphology when cultured in\nsoft synthetic hydrogels,\n121 adopt a lesion-like morphology\nwith invasive protrusions and epithelial-mesenchymal transi-\ntion (EMT)-like cellular morphology when cultured in stiff\nFig. 4 Micro fluidic model of immune and tumor cell traf ficking between the microvasculature and tissues .( a)Am i c r ofluidic device\ncomprising a central tissue-containing channel, flanked by two channels for flow of culture medium, is inoculated with a mixture of fibrin\ncontaining endothelial and stromal cells. Over the initial few days, ce lls undergo morphogenesis to form perfusable microvascular networks\nwhich are stable for weeks in culture as shown in ( b) for a day 23 culture (actin stain). ( c) Immune cells (green) can be perfused through the\nmicrovasculature (red) to model peripheral cell recruitment. ( d) The dynamic cell-level phenomena in the devices can be imagined by confocal or\ntwo-photon microscopy to observe phenomena such as neutrophil –tumor cell interactions in the extrav asation of tumor cells through the\nvascular wall into tissue. (Images from Zhang et al, 110 permission is in progress.)\nSeminars in Reproductive Medicine Vol. 38 No. 2-3/2020\nPhysiomimetic Models of Adenomyosis Gnecco et al. 189\n\n\nsynthetic hydrogels (Gnecco et al, unpublished data Aug, 15,\n2020; ►Fig. 3c ). These type of reductionist models may help\nparse the relative contributions of in flammatory cues, hor-\nmones, and other microenvironment factors reported in adeno-\nmyosis lesions. However, the myometrium is not just a static\nmicroenvironment. The nonpregnant myometrium undergoes\nconstant contractions to movefluid, in ways that are disrupted\nin adenomyosis.\n35,36 The propensity of endometrial cells to\nform endometriosis lesions in smooth muscle of the intestinal\ntract and adenomyosis lesions in the myometrium, and to\nsimilarly be associated with hyperplastic smooth muscle (or\nmyofibroblasts) in other ectopic contexts such as the bladder,\nureter, rectocervical space, and relatively absent from the\nmesentery and omentum, suggests an interesting tropism for\nmicroenvironments that may offer mechanical stimulation.\nAlthough the ability of drugs and hormones to alter\ncontractile phenotype of myometrial cells has been used\nas a screen in explant and 3D culture assays,\n81,113,129 and to\ncompare general contractile abilities of normal and\nadenomyotic myometrium, 36 extension to more complex\nquestions of etiology and interaction with endometrium\nunder dynamic conditions is only nascent. Dynamic me-\nchanical stimulation of human uterine cells in 2D has\nshown that endometrial stromal cells can acquire a con-\ntractile phenotype\n163 and that human myometrial cells\nexhibit dramatic shifts in the phosphoproteome under\nstretch.\n164 A study designed to illuminate the possible\neffects of myometrial contractions on endometrial pheno-\ntype used a novel micro fluidic reactor to coculture an\nendometrial cell line on a layer of myometrial cells, in a\nmanner that exposed the coculture to dynamic peristalsis\nthat generated shear flow in the fluid impinging on the\nendometrial epithelial cell layer. 165 This proof-of-principle\nstudy performed with cell lines showed morphological\nchanges in the epithelial barrier in response to the mechan-\nical forces,165 and provides a foundation for next-generation\nexperiments with primary cells and a more complex tissue\narchitecture, building on the well-developed observations\nabout tissue engineering of vascular smooth muscle under\npulsatile mechanical stimulation.\n166\nInnervation\nIn addition to the well-known animal dorsal root ganglia\nsources of neurons, the ef ficient derivation of genetically\ndiverse peripheral sensory neurons from human cells has\nprovided novel avenues for investigating and culturing hu-\nman nerve fibers in vitro\n167,168 and for building models for\nneuromotor action and pain. 167 While models of the endo-\nmetrium have not yet been explored, micro fluidic models of\nenteric nerve –epithelial interactions in the intestine are\nexemplary of the approaches. 169 Similarly, a micro fluidic\nmodel of neuronal activation of vascularized skeletal muscle\ncontraction, which allows quantitative analysis of axonal\ngrowth, muscle maturations, and contraction, provides a\ntemplate for design of a similar device for the myome-\ntrium.\n170 Ultimately, these tools might be combined to build\nmodels of innervated, vascularized adenomyosis lesions,\nusing patient-specific samples to understand the manifesta-\ntion of debilitating menorrhea.\nSystemic and Organ –Organ interactions\nFinally, adenomyosis and other chronic inflammatory diseases\nboth exert and respond to systemic effects, potentially includ-\ning those emanating from the gut microbiome. A growing\nnumber of micro fluidic and meso fluidic models are being\ndeveloped to connect MPSs representing multiple different\norgan systems (e.g., gut, liver, heart, brain) in a continuously\ncommunicating fluidic network for extended (weeks) culture\nperiods, using fabrication materials that avoid the problems of\nPDMS.171–173 A new pumping technology driven by integrat-\ning microfluidic pumps that are safe for immune cells onto the\nplatform has been applied to examine the interplay between\ntissue-resident cells in the gut and liver and circulating im-\nmune cells in response to short-chain fatty acid metabolites\nproduced in the colon, revealing paradoxical responses.\n173\nSuch technologies might be used in the future to examine\nchronic cell trafficking between the vasculature and lesions in\na single-MPS model of adenomyosis as a function of hormone\ncycles, or, more ambitiously, the interplay between the bone\nmarrow–adenomyosis lesion axis with respect to circulating\ncells and factors, given the interplay between bone marrow –\nderived cells and wound healing/lesion phenomena.\n174\nData-Driven Analysis and Predictive\nModeling of Human Responses\nA premise of the physiomimetics approach ( ►Fig. 1 )i st h a t\npatients can first be stratified on the basis of a combination of\nclinical phenotypes and molecular network analysis, to gener-\nate hypotheses for mechanistically distinguishable subgroups.\nIn vitro tissue-engineered models of these subgroups can then\nbe constructed using well-characterized patient-derived\nspecimens and perturbed. Finally, multi-omic measurements\ncombined with phenotypic metrics can be interpreted to drive\nidentification and validation of therapies for patient sub-\ngroups, thus providing a foundation for stratified clinical trials.\nMolecular stratification of endometrium-derived diseases is\nstill in early stages. A meta-analysis of genome-wide associa-\ntion studies for endometriosis showed stronger reproducibili-\nty across eight of nine disease-associated loci for patients with\nStage III/IV disease compared with Stage I,\n175 suggesting there\nmay be clues to processes that amplify lesion characteristics.\nHowever, the loci implicate genes widely expressed through-\nout the body; thus, one could reasonably construct hypotheses\naround gut permeability in fluencing systemic immune func -\ntion, for example, as a possible contributing mechanism. While\nsystems-level physiomimetic modeling of complex systemic\nimmune reactions involving gut permeability are emerging,\n173\nthese are not the most likely place to begin to parse adeno-\nmyosis. Similarly, transcriptomic changes in endometrial bi-\nopsies are more pronounced in severe versus mild\nendometriosis,176 suggesting several possible pathways for\nintervention. However, without other measurements of lesion\ntissue context and associated symptom characteristics, these\nSeminars in Reproductive Medicine Vol. 38 No. 2-3/2020\nPhysiomimetic Models of Adenomyosis Gnecco et al.190\n\n\ndatasets provide incomplete directions on designing physio-\nmimetic models to test mechanisms.\nThese types of analyses might be paired, though, with\nphenotypic assays representing dynamic lesion properties.\nFor example, the fine-grained in vitro analysis of pancreatic\ntumor organoid invasion characteristics, used for parsing\nmechanistic relationships between pancreatic cancer muta-\ntion subtypes and survival,177 might provide phenotypic dis-\ncrimination among patient subgroups, which could in turn\nlead to additional molecular phenotyping based on hypotheses\nregarding mechanisms governing the phenotypes. Even the\nassay used for the pancreatic tumors, however, revealed\nparadoxical relationships with survival\n177—suggesting an as-\nsay incorporating more complex tumor –stroma interactions\ncould be more revealing. A piece of a physiomimetic puzzle for\nendometriosis, involving a bioinformatics prediction that JNK\nwould govern cytokine release from peritoneal macrophages,\nwhich was then confirmed in vitro with patient samples,\n14 may\nalso have implications for adenomyosis, as macrophages are\nalso involved. A physiomimetic model incorporating patient\nmacrophages along with lesions may reveal disease-related\nphenotypic characteristics that could be modulated with a JNK\ninhibitor. At a systems level, the effects of potential new\ntherapeutics can also be assessed using multi-MPS systems\nincluding liver and the immune system, incorporating features\nof liver metabolism and other metabolic and systemic trans-\nformations, using multi-omics to parse complex system\nresponses.\n172,173 These technologies are developing quickly\nand may greatly enhance physiomimetic modeling of gynecol-\nogy diseases in general once resources to clinically phenotype\npatients and build corresponding tissue banks are identified.\nConclusions and Future Directions\nThe enormous spectrum of symptoms, histological/morpho-\nlogical appearances, and associated comorbidities that occur\nin patients with adenomyosis suggest that there may be\ndistinct subsets of patients who could be targeted with\npersonalized therapies—if the rules for defining these patient\npopulations and approaches for developing the targeted\ntherapies are established and implemented in an integrated\nfashion. The complexity of adenomyosis requires a physio-\nmimetic approach: parsing the phenomena that may con-\ntribute using computational modeling approaches and then\nbuilding a physiological model that provides information\nthat translates back into the clinic. The fundamental in vitro\ntissue engineering approaches necessary to create physio-\nmimetic models have largely been established: methods to\nconceptually and computationally characterize the disease\nto generate hypotheses about patient strati fication\napproaches; methods to create tissue banks of relevant cell\ntypes derived from carefully phenotyped patients; 3D syn-\nthetic matrices to engineer microenvironments in a repro-\nducible way; micro fluidic platforms to control vascular and\nnerve interactions with 3D tissues; and analytical outputs\nusing functional and molecular assays. The future of model-\ning adenomyosis lesion complexity arguably depends on\nhow well these tools become democratized, whether large\nenough patient populations can be pooled to de fine the\nbedside-to-bench-to-bedside paradigm of physiomimetic\nmodeling, and whether standardized methods can be de-\nfined and implemented for access to patient myometrial/\nadenomyotic tissue in younger patients who undergo fertili-\nty-sparing procedures. We are optimistic that the roadmap\ndescribed here will spur multidisciplinary teams to hasten\ndevelopment and implementation of better treatments for\npatients.\nFunding\nNational Institutes of Health http://dx.doi.org/10.13039/\n100000002 EB029132 National Science Foundation\nhttp://dx.doi.org/10.13039/100000001\nConflict of Interest\nNone declared.\nAcknowledgments\nWe thank Hilary Critchley, Stacey Missmer, and Doug\nLauffenburger for critical reading of the manuscript.\nThis work was supported by the John and Karine Begg\nFoundation, the Manton Foundation, the National Science\nFoundation, and NIH U01 EB029132.\nReferences\n1 Lopes-Pacheco M. CFTR modulators: the changing face of cystic\nfibrosis in the era of precision medicine. Front Pharmacol 2020;\n10:1662\n2 Benagiano G, Habiba M, Brosens I. The pathophysiology of uterine\nadenomyosis: an update. Fertil Steril 2012;98(03):572–579\n3 Chapron C, Tosti C, Marcellin L, et al. Relationship between\nthe magnetic resonance imaging appearance of adenomyosis\nand endometriosis phenotypes. Hum Reprod 2017;32(07):\n1393–1401\n4 García-Solares J, Donnez J, Donnez O, Dolmans MM. Pathogene-\nsis of uterine adenomyosis: invagination or metaplasia? Fertil\nSteril 2018;109(03):371 –379\n5 Habiba M, Gordts S, Bazot M, Brosens I, Benagiano G. Exploring\nthe challenges for a new classi fication of adenomyosis. Reprod\nBiomed Online 2020;40(04):569 –581\n6 Munro MG, Critchley HO, Broder MS, Fraser ISFIGO Working\nGroup on Menstrual Disorders. FIGO classi fication system\n(PALM-COEIN) for causes of abnormal uterine bleeding in non-\ngravid women of reproductive age. Int J Gynaecol Obstet 2011;\n113(01):3–13\n7 Tellum T, Qvigstad E, Skovholt EK, Lieng M. In vivo adenomyosis\ntissue sampling using a transvaginal ultrasound-guided core\nbiopsy technique for research purposes: safety, feasibility, and\neffectiveness. J Minim Invasive Gynecol 2019;26(07):\n1357–1362\n8 Zondervan KT, Becker CM, Koga K, Missmer SA, Taylor RN, Viganò\nP. Endometriosis. Nat Rev Dis Primers 2018;4(01):9\n9 Koninckx PR, Ussia A, Adamyan L, Wattiez A, Gomel V, Martin DC.\nHeterogeneity of endometriosis lesions requires individualisa-\ntion of diagnosis and treatment and a different approach to\nresearch and evidence based medicine. Facts Views Vis ObGyn\n2019;11(01):57–61\n10 Osuga Y, Fujimoto-Okabe H, Hagino A. Evaluation of the efficacy and\nsafety of dienogest in the treatment of painful symptoms in patients\nwith adenomyosis: a randomized, double-blind, multicenter, pla-\ncebo-controlled study. Fertil Steril 2017;108(04):673–678\nSeminars in Reproductive Medicine Vol. 38 No. 2-3/2020\nPhysiomimetic Models of Adenomyosis Gnecco et al. 191\n\n\n11 Vannuccini S, Tosti C, Carmona F, et al. Pathogenesis of adeno-\nmyosis: an update on molecular mechanisms. Reprod Biomed\nOnline 2017;35(05):592 –601\n12 Inoue S, Hirota Y, Ueno T, et al. Uterine adenomyosis is an\noligoclonal disorder associated with KRAS mutations. Nat Com-\nmun 2019;10(01):5785\n13 Shaked S, Jaffa AJ, Grisaru D, Elad D. Uterine peristalsis-induced\nstresses within the uterine wall may sprout adenomyosis. Bio-\nmech Model Mechanobiol 2015;14(03):437 –444\n14 Beste MT, Pfäf fle-Doyle N, Prentice EA, et al. Molecular network\nanalysis of endometriosis reveals a role for c -Jun-regulated\nmacrophage activation. Sci Transl Med 2014;6(222):222ra16\n15 Miller MA, Meyer AS, Beste MT, et al. ADAM-10 and -17 regulate\nendometriotic cell migration via concerted ligand and receptor\nshedding feedback on kinase signaling. Proc Natl Acad Sci U S A\n2013;110(22):E2074–E2083\n16 Shafrir AL, Missmer SA. Towards subtypes - deep endometriosis\noestrogen receptor- α expression. Nat Rev Endocrinol 2020;16\n(10):541–542\n17 Zondervan KT, Rahmioglu N, Morris AP, et al. Beyond endome-\ntriosis genome-wide association study: from genomics to phe-\nnomics to the patient. Semin Reprod Med 2016;34(04):242 –254\n18 Conklin JE, Lieberman JV, Barnes CA, Louis DZ. Disease staging:\nimplications for hospital reimbursement and management.\nHealth Care Financ Rev 1984;1984:13 –22\n19 Gonnella JS, Hornbrook MC, Louis DZ. Staging of disease. A case-\nmix measurement. JAMA 1984;251(05):637 –644\n20 Gonella J, Louis DZ, Guozum MVE, Callahan CA, Barnes CA.\nDisease staging. In: Gonella J, ed. Disease Staging. MI: Ann\nArbor; 2001:927\n21 Munro MG, Critchley HOD, Fraser ISFIGO Menstrual Disorders\nCommittee. The two FIGO systems for normal and abnormal uterine\nbleeding symptoms and classification of causes of abnormal uterine\nbleeding in the reproductive years: 2018 revisions. Int J Gynaecol\nObstet 2018;143(03):393–408\n22 Driehuis E, van Hoeck A, Moore K, et al. Pancreatic cancer\norganoids recapitulate disease and allow personalized drug\nscreening. Proc Natl Acad Sci U S A 2019;116(52):26580 –26590\n23 Vitonis AF, Vincent K, Rahmioglu NWERF EPHect Working\nGroup, et al; World Endometriosis Research Foundation Endo-\nmetriosis Phenome and Biobanking Harmonization Project: II.\nClinical and covariate phenotype data collection in endometri-\nosis research. Fertil Steril 2014;102(05):1223 –1232\n24 Rahmioglu N, Fassbender A, Vitonis AFWERF EPHect Working\nGroup, et al; World Endometriosis Research Foundation Endo-\nmetriosis Phenome and Biobanking Harmonization Project: III.\nFluid biospecimen collection, processing, and storage in endo-\nmetriosis research. Fertil Steril 2014;102(05):1233 –1243\n25 Pugliese A, Yang M, Kusmarteva I, et al. The Juvenile Diabetes\nResearch Foundation Network for Pancreatic Organ Donors with\nDiabetes (nPOD) program: goals, operational model and emerg-\ning findings. Pediatr Diabetes 2014;15(01):1 –9\n26 Gordts S, Campo R, Brosens I. Hysteroscopic diagnosis and\nexcision of myometrial cystic adenomyosis. Gynecol Surg\n2014;11(04):273–278\n27 Critchley HO, Jones RL, Lea RG, et al. Role of in flammatory\nmediators in human endometrium during progesterone with-\ndrawal and early pregnancy. J Clin Endocrinol Metab 1999;84\n(01):240–248\n28 Critchley HOD, Maybin JA, Armstrong GM, Williams ARW.\nPhysiology of the endometrium and regulation of menstruation.\nPhysiol Rev 2020;100(03):1149 –1179\n29 Vigano P, Candiani M, Monno A, Giacomini E, Vercellini P,\nSomigliana E. Time to rede fine endometriosis including its\npro-fibrotic nature. Hum Reprod 2018;33(03):347 –352\n30 Thiruchelvam U, Drans field I, Saunders PT, Critchley HO. The\nimportance of the macrophage within the human endometrium.\nJ Leukoc Biol 2013;93(02):217 –225\n31 Osteen KG, Rodgers WH, Gaire M, Hargrove JT, Gorstein F,\nMatrisian LM. Stromal-epithelial interaction mediates steroi-\ndal regulation of metalloproteinase expression in human\nendometrium. Proc Natl Acad Sci U S A 1994;91(21):\n10129 –10133\n32 Henderson TA, Saunders PT, Moffett-King A, Groome NP, Critch-\nley HO. Steroid receptor expression in uterine natural killer cells.\nJ Clin Endocrinol Metab 2003;88(01):440 –449\n33 Myers KM, Elad D. Biomechanics of the human uterus. Wiley\nInterdiscip Rev Syst Biol Med 2017;9(05):\n34 Aguilar HN, Mitchell BF. Physiological pathways and molecular\nmechanisms regulating uterine contractility. Hum Reprod Up-\ndate 2010;16(06):725 –744\n35 Mehasseb MK, Bell SC, Pringle JH, Habiba MA. Uterine adeno-\nmyosis is associated with ultrastructural features of altered\ncontractility in the inner myometrium. Fertil Steril 2010;93\n(07):2130–2136\n36 Guo SW, Mao X, Ma Q, Liu X. Dysmenorrhea and its severity are\nassociated with increased uterine contractility and overexpres-\nsion of oxytocin receptor (OTR) in women with symptomatic\nadenomyosis. Fertil Steril 2013;99(01):231 –240\n37 Brosens JJ, Barker FG, de Souza NM. Myometrial zonal differen-\ntiation and uterine junctional zone hyperplasia in the non-\npregnant uterus. Hum Reprod Update 1998;4(05):496 –502\n38 Benagiano G, Brosens I, Habiba M. Structural and molecular\nfeatures of the endomyometrium in endometriosis and adeno-\nmyosis. Hum Reprod Update 2014;20(03):386 –402\n39 Strauss JF, Lessey BA, eds. The structure, function, and evaluation\nof the female reproductive tract. . In: Yen & Jaffe ’s Reproductive\nEndocrinology, Elsevier, NY. 2009:191 –233\n40 Leyendecker G, Wildt L. A new concept of endometriosis and\nadenomyosis: tissue injury and repair (TIAR). Horm Mol Biol Clin\nInvestig 2011;5(02):125 –142\n41 Ibrahim MG, Chiantera V, Frangini S, et al. Ultramicro-trauma in\nthe endometrial-myometrial junctional zone and pale cell\nmigration in adenomyosis. Fertil Steril 2015;104(06):\n1475–83.e1, 3\n42 Zhang Q, Duan J, Liu X, Guo SW. Platelets drive smooth muscle\nmetaplasia and fibrogenesis in endometriosis through epitheli-\nal-mesenchymal transition and fibroblast-to-myofibroblast\ntransdifferentiation. Mol Cell Endocrinol 2016;428:1 –16\n43 Kobayashi H, Kishi Y, Matsubara S. Mechanisms underlying\nadenomyosis-related fibrogenesis. Gynecol Obstet Invest 2020;\n85(01):1–12\n44 Li T, Li YG, Pu DM. Matrix metalloproteinase-2 and -9 expression\ncorrelated with angiogenesis in human adenomyosis. Gynecol\nObstet Invest 2006;62(04):229 –235\n45 Guo SW. Fibrogenesis resulting from cyclic bleeding: the Holy\nGrail of the natural history of ectopic endometrium. Hum Reprod\n2018;33(03):353–356\n46 Li H, Yu Y, Shi Y, et al. HoxA13 stimulates myometrial cells to\nsecrete IL-1β and enhance the expression of contraction-associ-\nated proteins. Endocrinology 2016;157(05):2129 –2139\n47 Mehasseb MK, Panchal R, Taylor AH, Brown L, Bell SC, Habiba M.\nEstrogen and progesterone receptor isoform distribution\nthrough the menstrual cycle in uteri with and without adeno-\nmyosis. Fertil Steril 2011;95(07):2228 –2235, 2235.e1\n48 Jichan Nie, Xishi Liu, Guo SW. Promoter hypermethylation of\nprogesterone receptor isoform B (PR-B) in adenomyosis and its\nrectification by a histone deacetylase inhibitor and a demethyl-\nation agent. Reprod Sci 2010;17(11):995 –1005\n49 Scheerer C, Frangini S, Chiantera V, Mechsner S. Reduced sym-\npathetic innervation in endometriosis is associated to Sema-\nphorin 3C and 3F expression. Mol Neurobiol 2017;54(07):\n5131–5141\n50 Wang G, Tokushige N, Fraser IS. Nerve fibers and menstrual cycle\nin peritoneal endometriosis. Fertil Steril 2011;95(08):\n2772–2774\nSeminars in Reproductive Medicine Vol. 38 No. 2-3/2020\nPhysiomimetic Models of Adenomyosis Gnecco et al.192\n\n\n51 Al-Jefout M, Dezarnaulds G, Cooper M, et al. Diagnosis of\nendometriosis by detection of nerve fibres in an endometrial\nbiopsy: a double blind study. Hum Reprod 2009;24(12):\n3019–3024\n52 Browne AS, Yu J, Huang RP, Francisco AM, Sidell N, Taylor RN.\nProteomic identi fication of neurotrophins in the eutopic endo-\nmetrium of women with endometriosis. Fertil Steril 2012;98\n(03):713–719\n53 Anaf V, Simon P, El Nakadi I, et al. Relationship between endo-\nmetriotic foci and nerves in rectovaginal endometriotic nodules.\nHum Reprod 2000;15(08):1744 –1750\n54 Mechsner S, Kaiser A, Kopf A, Gericke C, Ebert A, Bartley J. A pilot\nstudy to evaluate the clinical relevance of endometriosis-associ-\nated nerve fibers in peritoneal endometriotic lesions. Fertil Steril\n2009;92(06):1856–1861\n55 García-Solares J, Dolmans MM, Squif flet JL, Donnez J, Donnez O.\nInvasion of human deep nodular endometriotic lesions is asso-\nciated with collective cell migration and nerve development.\nFertil Steril 2018;110(07):1318 –1327\n56 Tran LV, Tokushige N, Berbic M, Markham R, Fraser IS. Macro-\nphages and nerve fibres in peritoneal endometriosis. Hum\nReprod 2009;24(04):835 –841\n57 Greaves E, Temp J, Esnal-Zu fiurre A, Mechsner S, Horne AW,\nSaunders PT. Estradiol is a critical mediator of macrophage-\nnerve cross talk in peritoneal endometriosis. Am J Pathol 2015;\n185(08):2286–2297\n58 Zhang X, Lu B, Huang X, Xu H, Zhou C, Lin J. Innervation of\nendometrium and myometrium in women with painful adeno-\nmyosis and uterine fibroids. Fertil Steril 2010;94(02):730 –737\n59 Vacca P, Vitale C, Montaldo E, et al. CD34 þ hematopoietic\nprecursors are present in human decidua and differentiate\ninto natural killer cells upon interaction with stromal cells.\nProc Natl Acad Sci U S A 2011;108(06):2402 –2407\n60 Armstrong GM, Maybin JA, Murray AA, et al. Endometrial\napoptosis and neutrophil in filtration during menstruation\nexhibits spatial and temporal dynamics that are recapitulated\nin a mouse model. Sci Rep 2017;7(01):17416\n61 Russell P, Anderson L, Lieberman D, et al. The distribution of\nimmune cells and macrophages in the endometrium of women\nwith recurrent reproductive failure I: Techniques. J Reprod\nImmunol 2011;91(1-2):90 –102\n62 Bulmer JN, Jones RK, Searle RF. Intraepithelial leukocytes in\nendometriosis and adenomyosis: comparison of eutopic and\nectopic endometrium with normal endometrium. Hum Reprod\n1998;13(1O):2910–2915\n63 Orazov MR, Radzinsky VE, Nosenko EN, Khamoshina MB, Dukhin\nAO, Lebedeva MG. Immune-in flammatory predictors of the\npelvic pain syndrome associated with adenomyosis. Gynecol\nEndocrinol 2017;33(Suppl 1):44 –46\n64 He H, Mack JJ, Güç E, et al. Perivascular macrophages limit\npermeability. Arterioscler Thromb Vasc Biol 2016;36(11):\n2203–2212\n65 Schindl M, Birner P, Obermair A, Kiesel L, Wenzl R. Increased\nmicrovessel density in adenomyosis uteri. Fertil Steril 2001;75\n(01):131–135\n66 Liu X, Shen M, Qi Q, Zhang H, Guo SW. Corroborating evidence for\nplatelet-induced epithelial-mesenchymal transition and fibro-\nblast-to-myofibroblast transdifferentiation in the development\nof adenomyosis. Hum Reprod 2016;31(04):734 –749\n67 Harmsen MJ, Wong CFC, Mijatovic V, et al. Role of angiogenesis in\nadenomyosis-associated abnormal uterine bleeding and subfertil-\nity: a systematic review. Hum Reprod Update 2019;25(05):647–671\n68 Ota H, Igarashi S, Hatazawa J, Tanaka T. Is adenomyosis an\nimmune disease? Hum Reprod Update 1998;4(04):360 –367\n69 De Leo B, Esnal-Zu fiaurre A, Collins F, Critchley HOD, Saunders\nPTK. Immunopro filing of human uterine mast cells identi fies\nthree phenotypes and expression of ER β and glucocorticoid\nreceptor. F1000 Res 2017;6:667\n70 Benagiano G, Brosens I. The endometrium in adenomyosis.\nWomens Health (Lond) 2012;8(03):301 –312\n71 Tremellen KP, Russell P. The distribution of immune cells and\nmacrophages in the endometrium of women with recurrent\nreproductive failure. II: adenomyosis and macrophages. J Reprod\nImmunol 2012;93(01):58 –63\n72 Ota H, Tanaka T. Stromal vascularization in the endometrium\nduring adenomyosis. Microsc Res Tech 2003;60(04):\n445–449\n73 Ota H, Igarashi S, Hatazawa J, Tanaka T. Immunohistochemical\nassessment of superoxide dismutase expression in the endome-\ntrium in endometriosis and adenomyosis. Fertil Steril 1999;72\n(01):129–134\n74 Herndon CN, Aghajanova L, Balayan S, et al. Global transcriptome\nabnormalities of the eutopic endometrium from women with\nadenomyosis. Reprod Sci 2016;23(10):1289 –1303\n75 Acar S, Millar E, Mitkova M, Mitkov V. Value of ultrasound shear\nwave elastography in the diagnosis of adenomyosis. Ultrasound\n2016;24(04):205–213\n76 Al-Sabbagh M, Lam EW, Brosens JJ. Mechanisms of endometrial\nprogesterone resistance. Mol Cell Endocrinol 2012;358(02):\n208–215\n77 Hellweg G, Shaka JA. Endometrial granulocytes; tissue culture\nstudies of endometrium and decidua with special attention to\nthe endometrial granulocytes. Obstet Gynecol 1959;13(05):\n519–529\n78 Kawaguchi K, Fujii S, Konishi I, Okamura H, Mori T. Ultrastruc -\ntural study of cultured smooth muscle cells from uterine leio-\nmyoma and myometrium under the in fluence of sex steroids.\nGynecol Oncol 1985;21(01):32 –41\n79 Gellersen B, Bonhoff A, Hunt N, Bohnet HG. Decidual-type\nprolactin expression by the human myometrium. Endocrinology\n1991;129(01):158–168\n80 Heidari Kani MH, Chan E-C, Young RC, Butler T, Smith R, Paul JW.\n3D cell culturing and possibilities for myometrial tissue engi-\nneering. Ann Biomed Eng 2017;45(07):1746 –1757\n81 Arrowsmith S, Keov P, Muttenthaler M, Gruber CW. Contractility\nmeasurements of human uterine smooth muscle to aid drug\ndevelopment. J Vis Exp 2018;(131):56639\n82 Gaide Chevronnay HP, Lemoine P, Courtoy PJ, Marbaix E, Henriet\nP. Ovarian steroids, mitogen-activated protein kinases, and/or\naspartic proteinases cooperate to control endometrial remodel-\ning by regulating gene expression in the stroma and glands.\nEndocrinology 2010;151(09):4515 –4526\n83 Schäfer WR, Fischer L, Roth K, et al. Critical evaluation of human\nendometrial explants as an ex vivo model system: a molecular\napproach. Mol Hum Reprod 2011;17(04):255 –265\n84 Fasciani A, Bocci G, Xu J, et al. Three-dimensional in vitro culture\nof endometrial explants mimics the early stages of endometri-\nosis. Fertil Steril 2003;80(05):1137 –1143\n85 Osteen KG, Hill GA, Hargrove JT, Gorstein F. Development of a\nmethod to isolate and culture highly puri fied populations of\nstromal and epithelial cells from human endometrial biopsy\nspecimens. Fertil Steril 1989;52(06):965 –972\n86 Korch C, Spillman MA, Jackson TA, et al. DNA pro filing analysis of\nendometrial and ovarian cell lines reveals misidenti fication,\nredundancy and contamination. Gynecol Oncol 2012;127(01):\n241–248\n87 Banu SK, Lee J, Starzinski-Powitz A, Arosh JA. Gene expression\nprofiles and functional characterization of human immortalized\nendometriotic epithelial and stromal cells. Fertil Steril 2008;90\n(04):972–987\n88 Samalecos A, Reimann K, Wittmann S, et al. Characterization of a\nnovel telomerase-immortalized human endometrial stromal cell\nline, St-T1b. Reprod Biol Endocrinol 2009;7:76\n89 Krikun G, Mor G, Alvero A, et al. A novel immortalized human\nendometrial stromal cell line with normal progestational re-\nsponse. Endocrinology 2004;145(05):2291 –2296\nSeminars in Reproductive Medicine Vol. 38 No. 2-3/2020\nPhysiomimetic Models of Adenomyosis Gnecco et al. 193\n\n\n90 Wu J, Taylor RN, Sidell N. Retinoic acid regulates gap junction\nintercellular communication in human endometrial stromal\ncells through modulation of the phosphorylation status of\nconnexin 43. J Cell Physiol 2013;228(04):903 –910\n91 Schutte SC, James CO, Sidell N, Taylor RN. Tissue-engineered\nendometrial model for the study of cell-cell interactions. Reprod\nSci 2015;22(03):308 –315\n92 Valdez J, Cook CD, Ahrens CC, et al. On-demand dissolution of\nmodular, synthetic extracellular matrix reveals local epithelial-\nstromal communication networks. Biomaterials 2017;130:90–103\n93 Arnold JT, Lessey BA, Seppälä M, Kaufman DG. Effect of normal\nendometrial stroma on growth and differentiation in Ishikawa\nendometrial adenocarcinoma cells. Cancer Res 2002;62(01):79–88\n94 Zeitvogel A, Baumann R, Starzinski-Powitz A. Identi fication of an\ninvasive, N-cadherin-expressing epithelial cell type in endome-\ntriosis using a new cell culture model. Am J Pathol 2001;159(05):\n1839–1852\n95 Cook CD, Hill AS, Guo M, et al. Local remodeling of synthetic\nextracellular matrix microenvironments by co-cultured endo-\nmetrial epithelial and stromal cells enables long-term dynamic\nphysiological function. Integr Biol 2017;9(04):271 –289\n96 Barragan F, Irwin JC, Balayan S, et al. Human endometrial fibro-\nblasts derived from mesenchymal progenitors inherit progester-\none resistance and acquire an in flammatory phenotype in the\nendometrial niche in endometriosis. Biol Reprod 2016;94(05):118\n97 von Wolff M, Stieger S, Lumpp K, Bücking J, Strowitzki T, Thaler\nCJ. Endometrial interleukin-6 in vitro is not regulated directly by\nfemale steroid hormones, but by pro-in flammatory cytokines\nand hypoxia. Mol Hum Reprod 2002;8(12):1096 –1102\n98 Chen JC, Hoffman JR, Arora R, et al. Cryopreservation and recovery\nof human endometrial epithelial cells with high viability, purity,\nand functional fidelity. Fertil Steril 2016;105(02):501–10.e1\n99 Turco MY, Gardner L, Hughes J, et al. Long-term, hormone-\nresponsive organoid cultures of human endometrium in a\nchemically defined medium. Nat Cell Biol 2017;19(05):568 –577\n100 Boretto M, Cox B, Noben M, et al. Development of organoids from\nmouse and human endometrium showing endometrial epithe-\nlium physiology and long-term expandability. Development\n2017;144(10):1775–1786\n101 Sato T, Vries RG, Snippert HJ, et al. Single Lgr5 stem cells build\ncrypt-villus structures in vitro without a mesenchymal niche.\nNature 2009;459(7244):262 –265\n102 Valentijn AJ, Saretzki G, Tempest N, Critchley HO, Hapangama DK.\nHuman endometrial epithelial telomerase is important for epithe-\nlial proliferation and glandular formation with potential implica-\ntions in endometriosis. Hum Reprod 2015;30(12):2816–2828\n103 Boretto M, Maenhoudt N, Luo X, et al. Patient-derived organoids\nfrom endometrial disease capture clinical heterogeneity and are\namenable to drug screening. Nat Cell Biol 2019;21(08):1041–1051\n104 Hapangama DK, Drury J, Da Silva L, et al. Abnormally located\nSSEA1þ/SOX9þ endometrial epithelial cells with a basalis-like\nphenotype in the eutopic functionalis layer may play a role in\nthe pathogenesis of endometriosis. Hum Reprod 2019;34(01):\n56–68\n105 Gargett CE, Schwab KE, Zillwood RM, Nguyen HP , Wu D. Isolation\nand culture of epithelial progenitors and mesenchymal stem cells\nfrom human endometrium. Biol Reprod 2009;80(06):1136–1145\n106 Tempest N, Maclean A, Hapangama DK. Endometrial stem cell\nmarkers: current concepts and unresolved questions. Int J Mol\nSci 2018;19(10):E3240\n107 Tempest N, Baker AM, Wright NA, Hapangama DK. Does human\nendometrial LGR5 gene expression suggest the existence of\nanother hormonally regulated epithelial stem cell niche? Hum\nReprod 2018;33(06):1052 –1062\n108 Valentijn AJ, Palial K, Al-Lamee H, et al. SSEA-1 isolates human\nendometrial basal glandular epithelial cells: phenotypic and\nfunctional characterization and implications in the pathogenesis\nof endometriosis. Hum Reprod 2013;28(10):2695 –2708\n109 Miyazaki K, Dyson MT, Coon V JS, et al. Generation of progester-\none-responsive endometrial stromal fibroblasts from human\ninduced pluripotent stem cells: role of the WNT/CTNNB1 path-\nway. Stem Cell Reports 2018;11(05):1136 –1155\n110 Zhang Y, Zhou L, Li TC, Duan H, Yu P, Wang HY. Ultrastructural\nfeatures of endometrial-myometrial interface and its alteration\nin adenomyosis. Int J Clin Exp Pathol 2014;7(04):1469 –1477\n111 Hutchinson JL, Rajagopal SP, Yuan M, Norman JE. Lipopolysac -\ncharide promotes contraction of uterine myocytes via activation\nof Rho/ROCK signaling pathways. FASEB J 2014;28(01):94 –105\n112 Souza GR, Tseng H, Gage JA, et al. Magnetically bioprinted human\nmyometrial 3D cell rings as a model for uterine contractility. Int J\nMol Sci 2017;18(04):E683\n113 Vaes RDW, van den Berk L, Boonen B, van Dijk DPJ, Olde Damink\nSWM, Rensen SS. A novel human cell culture model to study\nvisceral smooth muscle phenotypic modulation in health and\ndisease. Am J Physiol Cell Physiol 2018;315(04):C598 –C607\n114 Osada H. Uterine adenomyosis and adenomyoma: the surgical\napproach. Fertil Steril 2018;109(03):406 –417\n115 Griffith LG, Swartz MA. Capturing complex 3D tissue physiology\nin vitro. Nat Rev Mol Cell Biol 2006;7(03):211 –224\n116 Bentin-Ley U, Pedersen B, Lindenberg S, Larsen JF, Hamberger L,\nHorn T. Isolation and culture of human endometrial cells in a\nthree-dimensional culture system. J Reprod Fertil 1994;101(02):\n327–332\n117 Kelly RW, King AE, Critchley HO. Cytokine control in human\nendometrium. Reproduction 2001;121(01):3 –19\n118 Schutte SC, Taylor RN. A tissue-engineered human endometrial\nstroma that responds to cues for secretory differentiation,\ndecidualization, and menstruation. Fertil Steril 2012;97(04):\n997–1003\n119 Arnold JT, Kaufman DG, Seppälä M, Lessey BA. Endometrial\nstromal cells regulate epithelial cell growth in vitro: a new co-\nculture model. Hum Reprod 2001;16(05):836 –845\n120 Abbas Y, Brunel LG, Hollinshead MS, et al. Generation of a three-\ndimensional collagen scaffold-based model of the human endo-\nmetrium. Interface Focus 2020;10(02):20190079\n121 Hernandez-Gordillo V, Kassis T, Lampejo A, et al. Fully synthetic\nmatrices for in vitro culture of primary human intestinal\nenteroids and endometrial organoids. Biomaterials 2020;\n254:120125\n122 Wolf K, Alexander S, Schacht V, et al. Collagen-based cell migration\nmodels in vitro and in vivo. Semin Cell Dev Biol 2009;20(08):931–941\n123 Lutolf MP, Hubbell JA. Synthesis and physicochemical characteriza-\ntion of end-linked poly(ethylene glycol)-co-peptide hydrogels formed\nby Michael-type addition. Biomacromolecules 2003;4(03):713–722\n124 Caliari SR, Burdick JA. A practical guide to hydrogels for cell\nculture. Nat Methods 2016;13(05):405 –414\n125 Brown A, He H, Trumper E, Valdez J, Hammond P, Grif fith LG.\nEngineering PEG-based hydrogels to foster ef ficient endothelial\nnetwork formation in free-swelling and con fined microenviron-\nments. Biomaterials 2020;243:119921\n126 Kyburz KA, Anseth KS. Synthetic mimics of the extracellular\nmatrix: how simple is complex enough? Ann Biomed Eng 2015;\n43(03):489–500\n127 DiMarco RL, Dewi RE, Bernal G, Kuo C, Heilshorn SC. Protein-\nengineered scaffolds for in vitro 3D culture of primary adult\nintestinal organoids. Biomater Sci 2015;3(10):1376 –1385\n128 Stegemann JP, Hong H, Nerem RM. Mechanical, biochemical, and\nextracellular matrix effects on vascular smooth muscle cell\nphenotype. J Appl Physiol (1985) 2005;98(06):2321 –2327\n129 Dallot E, Pouchelet M, Gouhier N, Cabrol D, Ferré F, Breuiller-\nFouché M. Contraction of cultured human uterine smooth\nmuscle cells after stimulation with endothelin-1. Biol Reprod\n2003;68(03):937–942\n130 Rajagopal SP, Hutchinson JL, Dorward DA, Rossi AG, Norman JE.\nCrosstalk between monocytes and myometrial smooth muscle in\nculture generates synergistic pro-in flammatory cytokine\nSeminars in Reproductive Medicine Vol. 38 No. 2-3/2020\nPhysiomimetic Models of Adenomyosis Gnecco et al.194\n\n\nproduction and enhances myocyte contraction, with effects\nopposed by progesterone. Mol Hum Reprod 2015;21(08):\n672–686\n131 Devost D, Zingg HH. Novel in vitro system for functional assess-\nment of oxytocin action. Am J Physiol Endocrinol Metab 2007;\n292(01):E1–E6\n132 Wendremaire M, Hadi T, Pezze M, et al. Macrophage-induced\nreactive oxygen species promote myometrial contraction and\nlabor-associated mechanisms. Biol Reprod 2020;102(06):\n1326–1339\n133 Peyton SR, Raub CB, Keschrumrus VP, Putnam AJ. The use of poly\n(ethylene glycol) hydrogels to investigate the impact of ECM\nchemistry and mechanics on smooth muscle cells. Biomaterials\n2006;27(28):4881–4893\n134 Lauffenburger DA, Horwitz AF. Cell migration: a physically\nintegrated molecular process. Cell 1996;84(03):359 –369\n135 Beunk L, Brown K, Nagtegaal I, Friedl P, Wolf K. Cancer invasion\ninto musculature: mechanics, molecules and implications.\nSemin Cell Dev Biol 2019;93:36 –45\n136 Weimar CH, Macklon NS, Post Uiterweer ED, Brosens JJ, Gellersen\nB. The motile and invasive capacity of human endometrial\nstromal cells: implications for normal and impaired reproduc -\ntive function. Hum Reprod Update 2013;19(05):542 –557\n137 Maheshwari G, Brown G, Lauffenburger DA, Wells A, Grif fith LG.\nCell adhesion and motility depend on nanoscale RGD clustering. J\nCell Sci 2000;113(Pt 10):1677 –1686\n138 Zaman MH, Trapani LM, Sieminski AL, et al. Migration of tumor\ncells in 3D matrices is governed by matrix stiffness along with\ncell-matrix adhesion and proteolysis. Proc Natl Acad Sci U S A\n2006;103(29):10889–10894\n139 Tan CW, Lee YH, Tan HH, et al. CD26/DPPIV down-regulation in\nendometrial stromal cell migration in endometriosis. Fertil Steril\n2014;102(01):167–177.e9\n140 Yang J-H, Wu MY, Chen MJ, Chen SU, Yang YS, Ho HN. Increased\nmatrix metalloproteinase-2 and tissue inhibitor of metalloprotei-\nnase-1 secretion but unaffected invasiveness of endometrial stro-\nmal cells in adenomyosis. Fertil Steril 2009;91(05):2193–2198\n141 Mehasseb MK, Taylor AH, Pringle JH, Bell SC, Habiba M. Enhanced\ninvasion of stromal cells from adenomyosis in a three-dimen-\nsional coculture model is augmented by the presence of myo-\ncytes from affected uteri. Fertil Steril 2010;94(07):2547 –2551\n142 Taylor AH, Kalathy V, Habiba M. Estradiol and tamoxifen enhance\ninvasion of endometrial stromal cells in a three-dimensional cocul-\nture model of adenomyosis. Fertil Steril 2014;101(01):288–293\n143 Palmer SS, Altan M, Denis D, et al. Bentamapimod (JNK inhibitor\nAS602801) induces regression of endometriotic lesions in ani-\nmal models. Reprod Sci 2016;23(01):11 –23\n144 Hussein M, Chai DC, Kyama CM, et al. c -Jun NH2-terminal kinase\ninhibitor bentamapimod reduces induced endometriosis in\nbaboons: an assessor-blind placebo-controlled randomized\nstudy. Fertil Steril 2016;105(03):815 –824.e5\n145 Gnecco JS, Ding T, Smith C, Lu J, Bruner-Tran KL, Osteen KG.\nHemodynamic forces enhance decidualization via endothelial-\nderived prostaglandin E2 and prostacyclin in a micro fluidic\nmodel of the human endometrium. Hum Reprod 2019;34(04):\n702–714\n146 Goddard LM, Murphy TJ, Org T, et al. Progesterone receptor in the\nvascular endothelium triggers physiological uterine permeabili-\nty preimplantation. Cell 2014;156(03):549 –562\n147 Maggioli E, McArthur S, Mauro C, et al. Estrogen protects the\nblood-brain barrier from in flammation-induced disruption and\nincreased lymphocyte traf ficking. Brain Behav Immun 2016;\n51:212–222\n148 Pence JC, Clancy KBH, Harley BAC. Proangiogenic activity of\nendometrial epithelial and stromal cells in response to estradiol\nin gelatin hydrogels. Adv Biosyst 2017;1(09):1700056\n149 Gnecco JS, Pensabene V, Li DJ, et al. Compartmentalized culture\nof perivascular stroma and endothelial cells in a micro fluidic\nmodel of the human endometrium. Ann Biomed Eng 2017;45\n(07):1758–1769\n150 Zervantonakis IK, Hughes-Alford SK, Charest JL, Condeelis JS,\nGertler FB, Kamm RD. Three-dimensional micro fluidic model for\ntumor cell intravasation and endothelial barrier function. Proc\nNatl Acad Sci U S A 2012;109(34):13515 –13520\n151 Hsu YH, Moya ML, Hughes CC, George SC, Lee AP. A micro fluidic\nplatform for generating large-scale nearly identical human\nmicrophysiological vascularized tissue arrays. Lab Chip 2013;\n13(15):2990–2998\n152 Chen MB, Whisler JA, Fröse J, Yu C, Shin Y, Kamm RD. On-chip\nhuman microvasculature assay for visualization and quanti fica-\ntion of tumor cell extravasation dynamics. Nat Protoc 2017;12(05):\n865–880\n153 Chen MB, Hajal C, Benjamin DC, et al. In flamed neutrophils\nsequestered at entrapped tumor cells via chemotactic con fine-\nment promote tumor cell extravasation. Proc Natl Acad Sci U S A\n2018;115(27):7022–7027\n154 Rambøl MH, Han E, Niklason LE. Microvessel network forma-\ntion and interactions with pancreatic islets in three-dimen-\nsional chip cultures. Tissue Eng Part A 2020;26(9-\n10):556 –568\n155 Aref AR, Campisi M, Ivanova E, et al. 3D micro fluidic ex vivo\nculture of organotypic tumor spheroids to model immune\ncheckpoint blockade. Lab Chip 2018;18(20):3129 –3143\n156 Campisi M, Shin Y, Osaki T, Hajal C, Chiono V, Kamm RD. 3D self-\norganized microvascular model of the human blood-brain bar-\nrier with endothelial cells, pericytes and astrocytes. Biomaterials\n2018;180:117–129\n157 Sobrino A, Phan DT, Datta R, et al. 3D microtumors in vitro\nsupported by perfused vascular networks. Sci Rep 2016;\n6:31589\n158 Bersini S, Gilardi M, Ugolini GS, et al. Engineering an environ-\nment for the study of fibrosis: a 3D human muscle model with\nendothelium specificity and endomysium. Cell Rep 2018;25(13):\n3858–3868.e4\n159 Regehr KJ, Domenech M, Koepsel JT, et al. Biological implications\nof polydimethylsiloxane-based micro fluidic cell culture. Lab\nChip 2009;9(15):2132 –2139\n160 Ivanova E, Kuraguchi M, Xu M, et al. Use ofex vivo patient-derived\ntumor organotypic spheroids to identify combination therapies\nfor HER2 mutant non-small cell lung cancer. Clin Cancer Res\n2020;26(10):2393–2403\n161 Gilbert PM, Weaver VM. Cellular adaptation to biomechanical\nstress across length scales in tissue homeostasis and disease.\nSemin Cell Dev Biol 2017;67:141 –152\n162 Purdy MP, Ducharme M, Haak AJ, et al. YAP/TAZ are activated by\nmechanical and hormonal stimuli in myometrium and exhibit\nincreased baseline activation in uterine fibroids. Reprod Sci\n2020;27(04):1074–1085\n163 Kim J, Ushida T, Montagne K, et al. Acquired contractile ability in\nhuman endometrial stromal cells by passive loading of cyclic\ntensile stretch. Sci Rep 2020;10(01):9014\n164 Copley Salem C, Ulrich C, Quilici D, Schlauch K, Buxton ILO,\nBurkin H. Mechanical strain induced phospho-proteomic sig-\nnaling in uterine smooth muscle cells. J Biomech 2018;\n73:99–107\n165 Elad D, Zaretsky U, Kuperman T, et al. Tissue engineered endo-\nmetrial barrier exposed to peristaltic flow shear stresses. APL\nBioeng 2020;4(02):026107\n166 Huang AH, Niklason LE. Engineering biological-based vascular\ngrafts using a pulsatile bioreactor. J Vis Exp 2011;(52):2646\n167 Wainger BJ, Buttermore ED, Oliveira JT, et al. Modeling pain in\nvitro using nociceptor neurons reprogrammed from fibroblasts.\nNat Neurosci 2015;18(01):17 –24\n168 Lemke KA, Aghayee A, Ashton RS. Deriving, regenerating, and\nengineering CNS tissues using human pluripotent stem cells.\nCurr Opin Biotechnol 2017;47:36 –42\nSeminars in Reproductive Medicine Vol. 38 No. 2-3/2020\nPhysiomimetic Models of Adenomyosis Gnecco et al. 195\n\n\n169 Puzan M, Hosic S, Ghio C, Koppes A. Enteric nervous system\nregulation of intestinal stem cell differentiation and epithelial\nmonolayer function. Sci Rep 2018;8(01):6313\n170 Osaki T, Sivathanu V, Kamm RD. Crosstalk between developing\nvasculature and optogenetically engineered skeletal muscle improves\nmuscle contraction and angiogenesis. Biomaterials 2018;156:65–76\n171 Oleaga C, Bernabini C, Smith AS, et al. Multi-organ toxicity\ndemonstration in a functional human in vitro system composed\nof four organs. Sci Rep 2016;6:20030\n172 Edington CD, Chen WLK, Geishecker E, et al. Interconnected\nmicrophysiological systems for quantitative biology and phar-\nmacology studies. Sci Rep 2018;8(01):4530\n173 Trapecar M, Communal C, Velazquez J, et al. Gut-liver physiomi-\nmetics reveal paradoxical modulation of IBD-related in flamma-\ntion by short-chain fatty acids. Cell Syst 2020;10(03):223–239.e9\n174 Critchley HOD, Babayev E, Bulun SE, et al. Menstruation: science\nand society. Am J Obstet Gynecol 2020; (epub ahead of print).\nDoi: 10.1016/j.ajog.2020.06.004\n175 Rahmioglu N, Nyholt DR, Morris AP, Missmer SA, Montgomery\nGW, Zondervan KT. Genetic variants underlying risk of endome-\ntriosis: insights from meta-analysis of eight genome-wide asso-\nciation and replication datasets. Hum Reprod Update 2014;20\n(05):702–716\n176 Aghajanova L, Giudice LC. Molecular evidence for differences in\nendometrium in severe versus mild endometriosis. Reprod Sci\n2011;18(03):229–251\n177 Huang W, Navarro-Serer B, Jeong YJ, et al. Pattern of invasion in\nhuman pancreatic cancer organoids is associated with loss of\nSMAD4 and clinical outcome. Cancer Res 2020;80(13):\n2804–2817\nSeminars in Reproductive Medicine Vol. 38 No. 2-3/2020\nPhysiomimetic Models of Adenomyosis Gnecco et al.196","source_license":"CC0","license_restricted":false}