A Vascularized Endometriosis Lesion Model for Parsing Responses to Progestins

Introduction: Endometriosis, defined by ectopic growth of endometrial-like tissue, is a chronic inflammatory disease that causes debilitating pain and infertility in about 10% of women. In the healthy eutopic endometrium, a surge in progesterone after ovulation suppresses estrogen-induced endometrial proliferation and initiates maturation (termed decidualization) in preparation for embryo implantation. However, the endometrium of some women, including many endometriosis patients, exhibits a “progesterone resistance” phenotype characterized by abnormal response to progesterone. An aberrant response to progesterone by endometrial-like cells in ectopic lesions may also underlie why many patients respond poorly to commonly prescribed progestin-based therapies for endometriosis — though the mechanisms of this impaired response remain poorly understood [1].

As the response to progesterone involves a complex interplay between multiple cell types, including endometrial epithelia, fibroblasts (“stroma”), and endothelial cells, we engineered a microphysiological system (MPS) to recapitulate features of the vascularized lesion microenvironment in patients, in order to analyze lesion responses to progesterone perturbations. We co-cultured endometrial epithelial and stromal cells from patients in a tunable synthetic hydrogel and incorporated this model lesion within 3D perfusable microvasculature in a microfluidic device. We then used this vascularized system to assess how hormones influence angiogenesis, local vascular permeability, and other vascular metrics along with properties of the endometrial cells.

Materials and

Methods: (1) Endometrial cell culture and characterization: Endometrial stromal cells (ESCs) and endometrial epithelial organoids (EEOs) were isolated from pipelle biopsies collected from consented adult donors. For progesterone responsiveness studies, ESCs with or without EEOs were encapsulated in a peptide-functionalized, synthetic polyethylene glycol (PEG) extracellular matrix (ECM) hydrogel [2] and stimulated with estradiol, progesterone, and cyclic adenosine monophosphate (cAMP) according to defined temporal protocols [3]. A functional assay to characterize the baseline response of donor cell populations was conducted by assessing the rate of prolactin secretion via ELISA for each donor, for ESCs encapsulated in the 3D synthetic ECM as either spheroid aggregates or dispersed cells. In some experiments, IL-1β was added to simulate an inflamed state with lower progesterone responsiveness.

(2) Vascularized microfluidic lesion model (Figure 1A): Microfluidic devices, comprising 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 were inserted into the central port and encapsulated in PEG. Over 7 days of co-culture, confocal imaging was used to evaluate changes in angiogenic sprouting and measure vascular permeability following previously established protocols [4]. Culture media was collected daily for downstream analysis of secreted cytokines, chemokines, and growth factors. In some experiments, the antiprogestin mifepristone (RU-486) was used to simulate progesterone withdrawal.

Results, Conclusions, and Discussions: We first validated that this microfluidic culture platform supports integration of endometrial cells into a perfusable vascular bed. When ESC spheroids were encapsulated in our synthetic PEG matrix and co-cultured with MVNs, they induced angiogenic sprouts with open lumens that extended across the fibrin-PEG interface and into the spheroids (Figure 1B). We also confirmed that this vascularized system enables functional measurements of vascular permeability, observing that spheroids with solely ESCs induced higher permeability in their surrounding vasculature compared to spheroids composed of both ESCs and EEOs (Figure 1C). This finding suggests that epithelial-stromal crosstalk may play an important role in regulating lesion interactions with local vasculature, and the inclusion of both of these endometrial cell types is critical for improving physiological relevance of lesion models.

With baseline vascular metrics characterized, we next evaluated responses to progesterone perturbations in the vascularized lesion model. We first established two different approaches for studying a “progesterone resistance” phenotype: (1) inducing the phenotype via stimulation with IL-1β, a pro-inflammatory cytokine known to be elevated in the peritoneal lesion microenvironment, and (2) screening for natural donor variation by characterizing ESCs derived from patients with a range of clinical presentations. Primary ESCs treated with IL-1β exhibited a blunted response to stimulation with progesterone and cAMP (Figure 1D). Differences in progesterone responsiveness also emerge across donors both with and without endometriosis (Figure 1E), highlighting the importance of using primary cells for proper recapitulation of disease heterogeneity.

Finally, we extended the microfluidic system to investigate progesterone withdrawal, a canonical cue for initiation of menses in the eutopic tissue. An outstanding debate is whether lesions “bleed” — i.e., whether vessels break down. Bleeding is observed in vivo, but what features precipitate the vessel fragility are unknown. We used the antiprogestin mifepristone to block progesterone signaling and analyzed the response of vascular features for donor cell sources across a spectrum of progesterone responsiveness. This vascularized lesion model provides unprecedented insights into the interplay between hormone perturbations and intrinsic donor material in modeling 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 tissue samples in support of this research.

References
[1] Patel, B.G., et al. Progesterone resistance in endometriosis: origins, consequences and interventions. Acta Obstet Gynecol Scand. 2017, 96(6):623-632. https://doi.org/10.1111/aogs.13156.
[2] Gnecco, J.S., Brown, A.T., et al. Organoid co-culture model of the human endometrium in a fully synthetic extracellular matrix enables the study of epithelial-stromal crosstalk. Cell Med. 2023, 4(8):554-579.E9. https://doi.org/10.1016/j.medj.2023.07.004.
[3] Ma, Q., et al. Extracellular vesicles secreted by human uterine stromal cells regulate decidualization, angiogenesis, and trophoblast differentiation. PNAS. 2022, 119(38):e2200252119. https://doi.org/10.1073%2Fpnas.2200252119.
[4] Hajal, C. et al. Engineered human blood–brain barrier microfluidic model for vascular permeability analyses. Nat. Protoc. 2022, 17:95-128. https://doi.org/10.1038/s41596-021-00635-w.