A Microphysiological System Approach to Studying Lesion-Vasculature Interactions in Endometriosis

Introduction:: Endometriosis, defined by the growth of endometrial tissue outside of the uterus, is a chronic inflammatory disease that causes debilitating pain and infertility in up to 10% of women. Despite this prevalence, little is known about endometriotic lesion establishment, and our efforts to study and treat the disease are hampered by its heterogeneous clinical presentation, delays in diagnosis, and a lack of relevant preclinical animal models. In vitro models offer a promising avenue for probing the mechanisms of disease, but many of these culture systems are lacking in their physiological relevance. Notably, most existing in vitro models are limited by their use of immortalized cell lines and do not include functional vasculature, which is known to be a major hallmark of lesion development [1].



In this study, we employ a microphysiological system approach to build an in vitro model system that more accurately captures the in vivo complexity of endometriosis. We designed and characterized a microfluidic culture platform that enables co-culture of primary human endometrial cells — grown in a tunable synthetic hydrogel that allows for recapitulation of key endometrial phenotypes [2] — with 3D perfusable microvasculature. We demonstrate that this system supports vascularization of endometrial cell spheroid aggregates and is suitable for both morphological and functional metrics of how local vasculature may change near endometriotic lesions, such as angiogenesis and vascular permeability.


Materials and Methods:: 1) Spheroid co-culture system for modeling endometriotic lesions: endometrial stromal cells (ESCs) and endometrial epithelial organoids (EEOs) were isolated from tissue biopsies collected from consented adult donors. ESCs and EEOs were expanded on tissue-culture plastic and in 50uL Matrigel droplets, respectively, before being harvested and added to ultra-low-attachment microplates to form spheroid aggregates. For initial characterization studies, endometrial spheroids were encapsulated in a peptide-functionalized, synthetic polyethylene glycol (PEG) hydrogel [2]. ESC decidualization response in spheroid co-culture format was confirmed by measuring secreted prolactin via ELISA. For subsequent microfluidic device experiments, ESCs and EEOs were stained with CellTracker dyes before spheroid formation for later visualization.



2) Vascularized microfluidic lesion model: Microfluidic devices, consisting of a square gel compartment flanked by two media channels, were fabricated using polydimethylsiloxane (PDMS). GFP-labeled human umbilical vein endothelial cells (HUVECs) and supporting lung fibroblasts were mixed in fibrin, seeded into the gel compartment of these devices, and allowed to self-assemble into perfusable microvascular networks (MVNs) around an empty central port. After initial MVN formation, endometrial spheroids as described above (or lung fibroblast spheroids, for initial validation studies) were inserted into the central port and encapsulated in PEG (Figure 1A). Over 8 days, confocal imaging was used to evaluate changes in angiogenic sprouting. Culture media was collected daily for measuring vascular endothelial growth factor (VEGF) levels via ELISA. On Day 3, vascular permeability was measured via time-lapse confocal imaging of a fluorescent dextran molecule added to the MVNs, following previously established protocols [3].


Results, Conclusions, and Discussions:: Using lung fibroblast spheroids as a previously established model for inducing angiogenesis, we first validated that this microfluidic culture platform supports growth of angiogenic sprouts from a perfusable vascular bed into our synthetic PEG matrix. Over 7 days of co-culture with MVNs, PEG-encapsulated lung fibroblast spheroids induced angiogenic sprouts with open lumens that extended across the fibrin-PEG interface and into the spheroids (Figure 1B). We then confirmed that this spheroid culture format can be adapted for endometrial cells by observing that ESCs maintain a functional decidualization response in spheroid form, as measured by prolactin secretion after progestin stimulation.



Upon incorporating the endometrial spheroids into our vascularized microfluidic devices, we observed that by the second day of co-culture, the spheroids produce pro-angiogenic factors such as VEGF (Figure 1C) and begin inducing growth of angiogenic sprouts (Figure 1D), neither of which are observed in the cell-free PEG control. On Day 3 of co-culture, after the MVNs were fully formed, we used a fluorescence-based imaging assay to quantify whether vascular permeability changes in the presence of endometrial cells. Spheroids with only ESCs led to higher vascular permeability compared to spheroids with both ESCs and EEOs (Figure 1E), suggesting that the presence of epithelial cells may stabilize stromal cell dysregulation of local vasculature. Since lesions are known to consist primarily of stromal cells, further examination of these permeability trends may implicate lesion cellular composition as one indicator of disease severity.



With this enhanced ability to recapitulate and investigate disease-relevant vascularization, we aim to follow up on our phenotypic observations with more mechanistic studies of lesion-vasculature interactions. In addition, we are in the process of assessing how molecular cues such as reproductive hormones and inflammatory cytokines may further influence how lesions interface with local vasculature. Ultimately, we hope that this microphysiological system offers a more physiologically relevant model of the endometriotic lesion microenvironment, laying the groundwork for improved studies of diverse lesion phenotypes and their potential therapeutic treatments.


Acknowledgements (Optional): : We would like to thank the National Institutes of Health (NIH U01 EB029132 and NIH R01 HD110335), the National Science Foundation Graduate Research Fellowship Program (NSF GRFP), and the John and Karine Begg Fund for generously supporting this work. We would also like to thank our clinical collaborators at the Newton-Wellesley Hospital (NWH) Center for Minimally Invasive Gynecologic Surgery: Dr. Keith Isaacson, Dr. Peter Movilla, and Mollie O’Brien. Finally, we thank the NWH gynecology patients who generously consented to donate their endometrial biopsies in support of this research.

References (Optional): : [1] Laschke & Menger. Basic mechanisms of vascularization in endometriosis and their clinical implications. Human Reproduction Update 2018, 24(2):207-224. https://doi.org/10.1093/humupd/dmy001.

[2] Gnecco, J.S., Brown, A.T., et al. Organoid co-culture model of the cycling human endometrium in a fully-defined synthetic extracellular matrix reveals epithelial-stromal crosstalk. Cell Med 2023 (accepted). https://www.biorxiv.org/content/10.1101/2021.09.30.462577v2.full (preprint).

[3] Hajal, C. et al. Engineered human blood–brain barrier microfluidic model for vascular permeability analyses. Nature Protocols 2022, 17:95-128. https://doi.org/10.1038/s41596-021-00635-w.