Integrating tissue engineering and microfluidics to model the spatial niches of the human endometrium in vitro with guidance from in vivo multiomics data

The endometrium -- the innermost mucosal lining of the uterus that provides the site of embryo implantation- is crucial for our species preservation and alterations on its function underlay infertility and disease. Developments in single-cell and spatial technologies have illuminated previously unknown subtypes of cells and their spatial arrangements in the endometrium, providing valuable insights into the signalling pathways involved in the different endometrial compartments and hinting at their roles in determining cell specification in the luminal or basal zones. While inferences can be made from these static pictures about physiological functions ranging from blastocyst implantation and invasion to heavy menstrual bleeding, without tractable in vitro models that capture the endometrium's dynamic spatial interactions, mechanistic hypotheses about endometrial function remain challenging to test. Here, we develop in vitro models of the endometrium by combining tissue-level spatial mapping approaches with in vitro tissue engineering and microfluidic approaches. We will refine existing approaches and develop/test new models, prioritizing robustness and reproducibility, and allowing dissemination and implementation by the broader community, as follows Aim 1: Design a robust and scalable microphysiological systems (MPS) model that replicates the spatial niches of the human endometrium in the early-to-mid secretory phase. We we will use a co-culture of epithelia and stroma in a microfluidic device, in which organoids undergo morphogenesis in a special hydrogel. We will inform the media composition and other metrics by querying the in vivo cellular atlases. Aim 2: Quantify the interplay between cell origin (“nature”) and spatial environment (“nurture”) in defining epithelial cell identity in the in vitro models built in Aim 1. We will use single-cell and spatial transcriptomics technologies to profile the devices, and we will use clonal tracing to map the origin of a cell type. By integrating vivo/in vitro datasets, we will conduct a systematic investigation into the various parameters that characterize the cell source (such as luminal / basalis), as well as the magnitudes of gradients in diffusible signaling molecules and nutrients. Aim 3: Investigate the functional consequences of introducing endothelial cells into the in vitro model from Aim 1, using a simple “monolayer” protocol. We will focus on the cell state transitions that are modulated by endothelial-stromal-epithelial crosstalk during decidualization, and use the model to investigate immune cell trafficking. Our ultimate goal is to obtain a systems biology view of the tissue, which will allow us to inform the development of new treatments for endometrial-related disorders. Our tools and knowledge have potential to bring preclinical models of the endometrium more firmly into the realm of humanised research. This work has significant implications for both academic and industrial research, leading towards a personalised medicine approach for gynaecology.