Poster CC3 - Engineering a Synthetic Hydrogel to Support a Microphysiological Model of Vascularized Endometriosis Lesions

Introduction: Ectopic growth of the endometrium in the myometrium (adenomyosis) or outside the uterus (endometriosis) is a debilitating gynecologic disorder affecting ~10% of reproductive age women with no curative treatment. Endometriosis lesions establish complex microenvironments comprised of epithelial glands with surrounding stroma, vasculature, and immune cells. Animals cannot capture the complex immune-hormone interplay and human genetic diversity, necessitating the development of in vitro models with patient-derived cells which harbor disease phenotypes.
Microphysiological systems (MPSs) are powerful tools for elucidating disease mechanisms and drug testing as they can incorporate physiologically relevant components such as vasculature - a major hallmark of endometriosis lesions that is not accounted for in most in vitro models. However, fibrin, the standard extracellular matrix (ECM) used in vascularized MPS, is rapidly degraded, contains extraneous biological signals, and cannot be tailored to adequately support the range of cell types in the endometrium. Fibrin has been the mainstay for creating perfusable vessels in vitro using endothelial and supporting cell types from a variety of sources as no publication yet reports a reliable synthetic ECM for such vessel formation in a morphogenesis format. To overcome these limitations, we have developed a modular and fully synthetic ECM that has the potential to substantially advance the field of vascularized MPS development and can support the diverse cell types necessary to model endometriosis lesions.

Materials and

Methods: Briefly, the synthetic hydrogels comprise 8-arm polyethylene glycol (PEG) vinyl sulfone molecules crosslinked with a cell-degradable peptide and functionalized with integrin- and matrix- binding peptides. Primary endometrial epithelial organoids (EEOs) and endometrial stromal cells (ESCs) are isolated from patient biopsies. GFP-labeled human umbilical vein endothelial cells (HUVECS) and normal human lung fibroblasts (NHLFs) were expanded according to manufacturer protocols. CD14+ human monocytes were isolated from healthy donors and differentiated into macrophages over 6 days using 50 ng/mL macrophage-colony stimulating factor. Custom microfluidic devices were fabricated with a central gel channel with flanking media channels which can be used in static culture or connected to pumps for continuous flow with a controlled pressure drop.
All cells were mixed with the hydrogel precursor solution at day 0 and injected into the central gel channel (Fig. 1A). Media were changed daily and stored frozen for analysis. We semi-empirically determined the hydrogel formulation, cell concentrations, and HUVEC to stromal (NHLF and/or ESC) ratio to optimize endothelial network formation in custom microfluidic devices prior to co-culture with EEOs and labeled ESCs. Network formation and organoid growth was monitored via daily imaging. Network morphology was quantified by analyzing confocal images at Day 5 using MATLAB. Perfusability was assessed by the flow of fluorescent beads across the network.

Results, Conclusions, and Discussions: We demonstrate the ability to create self-assembled perfusable microvascular networks in our synthetic PEG-based hydrogel in the presence of flow using HUVEC-NHLF coculture, which is the canonical cell combination used in vascularized MPSs, as evidenced by the flow of fluorescent beads through the open vessel structures (Fig. 1B). In parallel, we established this synthetic ECM supports the emergence of EEOs from single cells, serving as a viable alternative for Matrigel, which is typically used for organoid culture but is limited by its tumor-derived origin (Fig. 1C). Next, we were able create microvascular networks using ESCs as a more relevant stromal cell population with two donors (Fig. 1D). Additionally, we developed a media for co-culture of the vascular and endometrial lesion cells which supports viability and growth of all cell types. Altogether, our microfluidic device design and hydrogel formulation supports the co-culture of patient-derived EEOs and ESCs, which can be added to the initial endothelial-fibroblast gel encapsulation mixture at the time of injection to form endometriosis lesion-like structures in the microvascular bed (Fig. 1E).
Current directions are focused on the addition of macrophages to the model, as immune cells are also an important hallmark of endometriosis lesions and a target of novel drug compounds. We have established conditions to achieve microvascular network formation with macrophages at a ratio comparable to that found in endometriosis lesions in vivo and demonstrated macrophage functionality as evidenced by IL-6 secretion detected by ELISA.
In conclusion, we demonstrate our synthetic hydrogel provides a “one size-fits-all” ECM to support the co-culture of self-assembled perfusable microvasculature with epithelial, stromal, and immune cells for modeling endometriosis lesions. This microfluidic culture model has potential for impacts on a better scientific understanding of endometriosis and can be used for testing novel drug compounds.

Acknowledgements (Optional): We would like to acknowledge our funding sources (NIH U01 EB029132, NIH R01 HD110335, Begg Fund) and the patients from Newton-Wellesley Hospital for making this work possible.