UCSF

Native cardiac tissue is comprised of many different cell types that work cooperatively for proper tissue function. Combining self-assembled tissue engineering strategies that provide a fully-defined platform to study pairwise interactions between different cardiac cell types, with human pluripotent stem cell (hPSC) technologies, such as robust differentiation strategies and genome editing capabilities, has enabled our comprehensive studies of heterotypic interactions between cardiomyocytes and various non-myocyte sources in order to determine the specific contributions of non-myocytes to cardiac microtissue fate and function. We approached these studies in a systematic manner with narrowing focus.

We first broadly tested the most commonly-used stromal cells in cardiac tissue engineering studies and found that the different sources of stromal cells (primary human-derived vs. stem cell-derived; from different types of primary tissues) were distinct in terms of their surface marker expression, morphometry, and gene expression. These differences carried over into their ability to support engineered cardiac tissue formation and function, where only primary human cardiac fibroblasts and primary human dermal fibroblasts paired with hPSC-cardiomyocytes resulted in microtissues with the most robust tissue self-assembly and advanced calcium handling function.

Since the tissue-specific cardiac fibroblasts were able to positively support cardiac microtissue culture, we further characterized the specific contributions of different types of non-myocytes (endothelial cells, fetal human cardiac fibroblasts, adult human cardiac fibroblasts) in the context of heterotypic cardiac microtissue phenotype and function, at the single-cell and tissue-level, respectively. We found that 1 week after tissue formation, the cardiac microtissues containing the cardiac fibroblasts displayed more mature calcium handling properties relative to the tissues that contained endothelial cells and the tissues made from only cardiomyocytes, and that the cardiomyocytes paired with the cardiac fibroblasts were transcriptionally distinct from cardiomyocytes from the other tissues. However, after extended culture duration (1 month), the distinction between cardiac microtissues with cardiac fibroblasts versus without was lost, with the cardiomyocytes exhibiting similar transcriptomic profiles and the tissues displaying similar calcium transients. Furthermore, at both time points, there were no discernable differences between the different age cardiac fibroblasts, potentially because the source (isolated from primary tissue) was a bigger mismatch with stem cell-derived cardiomyocytes than the ontogenic difference.

Inspired by the pairing of different technologies to assess single-cell-level phenotype in the context of microtissue-level function, we sought to further characterize individual cell properties within intact 3D microtissues in order to better link the single cell building blocks to tissue-level properties. We used light sheet fluorescence microscopy to quantify 3D heterotypic multicellular organization as well as identify individual cardiomyocyte functional heterogeneity within heterotypic cardiac microtissues. Overall, this study demonstrated that advanced imaging techniques can be a powerful tool to dissect complex heterotypic interactions without removing the cells from their 3D environment.

Lastly, to dig deeper into the mechanisms governing the heterotypic interactions between cardiomyocyte and non-myocytes in our 3D engineered microtissues, we first had to generate cardiac tissues made from entirely stem cell-derived cellular constituents in order to take advantage of the robust genome engineering strategies developed for hPSCs. We were able to generate entirely-isogenic tissues when two differentiation protocols for the derivation of cardiac fibroblasts were published in 2019. We evaluated the different hPSC-cardiac fibroblast subtypes generated by these protocols in our heterotypic cardiac microtissue platform and found that they behaved similarly to one another and to microtissues made with primary human fetal cardiac fibroblasts in their ability to quickly self-assemble into tissues and their calcium handing function. We then knocked down one of the most-cited gap junctions that connects cardiomyocytes and cardiac fibroblasts in the heart, connexin 43, in the hPSC-derived cardiac fibroblasts using an inducible CRISPR interference method. Heterotypic cardiac microtissues generated with the knockdown fibroblasts displayed diminished calcium handling function, indicating a potential role for cardiac fibroblast support of cardiomyocyte function. Further mechanistic understanding of the interactions between cardiac cell populations can be determined in a similar manner, within the context of our fully-defined, tailored microtissue platform.

Taken together, this body of work provides a basis for the study of multicellular heterotypic interactions in 3D engineered models of myocardial tissue from human pluripotent stem cells. Future studies can build on this work by generating more complex microtissue constructs (i.e. incorporating more than two cell types, or modulating the cell types) as well as modeling cardiac diseases in tissue format. Combining genome editing, advanced imaging, and next-generation sequencing technologies enables customizable generation and comprehensive characterization of the multicellular interactions within engineered heterotypic tissue constructs.