Z27 - Toward Sex-Specific Microphysiological Systems for Modeling Diabetic Retinopathy

Friday, October 10, 2025

10:00 AM - 11:00 AM PDT

Location: Exhibit Hall F, G & H

Presenting Author(s)

Ashley Martier, PhD

Postdoctoral research scientist
Xavier University of Louisiana
New Orleans, Louisiana, United States

Co-Author(s)

Keefer Boone, B.S.E

PhD Student
Rice University, United States

Last Author(s)

MM

Mark Mondrinos

Assistant Professor
Tulane University
New Orleans, Louisiana, United States

Introduction: : Diabetes mellitus (DM) affects an estimated 537 million adults worldwide with an estimated 240 million more believed to be living undiagnosed. Globally, this burden is projected to increase to an estimated 1.3 billion DM patients by the year 2030. While DM can be often be managed via lifestyle changes and pharmaceuticals, cardiovascular complications secondary to DM remain common. Diabetic retinopathy, or the breakdown of the retinal vasculature and accompanying barriers of the outer retina, occurs in up to 50% of diabetic patients and is the leading cause of preventable blindness in adults. The impact of biological sex on DR pathophysiology is poorly defined. Several large epidemiological studies suggest that male sex is correlated with higher incidence of DR, while other studies show conflicting results or no sexual dimorphism at all. Due to difficulties discerning sex-linked and gender-linked dimorphisms in population-based studies, preclinical studies focusing on sex and DR are an ideal option to discern if a true sexual dimorphism exists. Existing preclinical model systems are largely inadequate to answer this question. Cell-based models and existing culture methods fail to account for cell sex and hormonal stimuli. Further, animal models differ from humans in inflammatory response and in sex hormone biology, thereby decreasing translational relevance. To address this, we engineered microphysiological systems (MPS) to model pathological changes of the outer retina in DR, accounting for both chromosomal and hormonal inputs to assess the influence of sex on the progression and severity of DR.

Materials and

Methods: : Two MPS models of the outer retina were developed and utilized in parallel (Figure 1B). Using our previously developed benchtop fabrication workflow, multilayer devices were fabricated using PDMS with 0.25% PDMS-PEG copolymer to minimize hormone absorption. A single-vessel MPS consisted of a channel loaded with human retinal microvascular endothelial cells (HRMVEC, n=3 donors per sex) separated by a polyester membrane from an underlying stromal compartment loaded with collagen gels containing human ocular fibroblasts (HOF, n=3 donors per sex). A vascular bed MPS consisted of retinal pigmented epithelial (RPE) cells seeded on a Bruch’s membrane mimic and an underlying vascular compartment containing a self-assembled vascular network formed by admixing HOF and HRVMEC in a collagen and fibrin gel. Devices were seeded with sex-matched cells, cultured in normal conditions for 48 hours, switched to a hormone-free media for a hormone depletion period, and then cultured in diabetic conditions (1% O2, 75mM glucose, 1ng/ml TNF and 1ng/ml IL-6) with or without 0.1nM dihydrotestosterone (DHT) or 1nM estradiol (E2) for an additional 4 days. Single-vessel devices were assessed for changes in permeability by quantifying the mass of FITC-dextran transported across the endothelial barrier and stromal layer. Parallel single-vessel devices were fixed and stained for ZO-1 localization to assess tight junction formation. Vascular bed MPS were assessed for changes in vascular network morphology, fibroproliferation, and changes in basement membrane composition of the network. 3 donor pairs per sex were used for 3 devices per condition per donor pair.

Results, Conclusions, and Discussions:: In both XX and XY single-vessel devices, the introduction of diabetic (DM) culture conditions significantly decreased ZO-1 localization at cell boundaries (Figure 1C and 1E). FITC-dextran permeability accordingly increased in both sexes in DM conditions (Figure 1D and 1F). E2 protected against this perturbation in XX models in ZO-1 and FITC-dextran permeability. DHT was not protective in XX models. In XY models, neither E2 nor DHT had protective effects on ZO-1 localization and FITC-dextran permeability in DM conditions. In the vascular bed devices, XX and XY models in DM conditions showed decreased maximum diffusion distances and increased mean segment diameter, vessel area fraction, total vessel length, and vessel length density (Figure 1G-I). In addition, vascular bed devices exposed to DM exhibited increased fibroproliferation and increased basement membrane deposition. When exposed to E2, XX devices exhibited restoration to baselines in vessel area fraction, vessel length density, and mean segment diameter as well as restoration to baseline levels of fibroblast proliferation and basement membrane deposition. DHT did not similarly restore XX devices, and neither E2 nor DHT restored XY vascular beds.

Together, these results suggest that the imposed DM culture conditions recapitulated pathological changes seen in DM patients, including vascular leakage, neovascularization, basement membrane thickening, and fibroproliferation (Figure 1A). Further, female sex alone (i.e. XX chromosomes, no hormones) nor male sex alone (i.e. XY chromosomes, no hormones) were not sufficient to cause significant differences in baseline or DM conditions. However, the combination of female chromosomal sex and female hormones (1nM E2) were sufficient to protect against the changes induced by DM. No such protection was seen in XY models.

This data is line with the hypothesis that female sex, namely the interplay between female hormones and female chromosomal make up, is protective in the setting of inflammatory diseases such as DR. Future investigations should continue to account for more complex, physiologically accurate inputs including the impacts of physiological hormone ratios as well as the effects of additional hormones such as progesterone and other estrogens and androgens.

Acknowledgements and/or References (Optional): :