Optimized Protein Patterning Methods for Human Liver Cultures FerrariErika 2018 Organizing cells on extracellular matrix (ECM) proteins is useful for optimizing cell-cell interactions and therefore cell functions in vitro. In the case of the liver, micropatterned co-cultures (MPCCs) containing primary human hepatocytes (PHHs) and 3T3-J2 murine embryonic fibroblasts adhered to hard surfaces (e.g. glass and plastic) have been shown to exhibit high levels of hepatic functions for several weeks in vitro; the use of PHHs in this model mitigates the differences observed in drug metabolism enzymatic pathways between animal (e.g. rodents) and human livers. Furthermore, MPCCs have been previously augmented with liver sinusoidal endothelial cells (LSECs) or human umbilical vein endothelial cell (HUVECs) non-liver controls to model the interactions between PHHs and endothelial cells as in vivo; however, this study was limited to gene expression analysis of the endothelial phenotype, which may not correlate with functional markers. This thesis seeks to build upon the MPCC platform to a) develop and optimize an alternative technique than currently used for patterning proteins onto softer surfaces that more accurately mimic the stiffness of the liver, and b) determine protein marker expression of the endothelial cells in monocultures and co-cultures with PHHs and/or fibroblasts to complement the published gene expression data set. Towards the first aim, we optimized parameters for a soft lithographic microcontact protein printing method using polydimethylsiloxane (PDMS) stamps and subsequently showed its utility for the creation of highly functional MPCCs on both collagen and fibronectin stamped domains on glass. Additional studies demonstrated the utility of this method for patterning proteins and cells on softer surfaces in the kPa range of stiffness as opposed to the GPa range of stiffness for glass and plastic. Towards the second aim of this thesis, we tested three protein markers (CD31, SE-1, and Factor VIII) for their ability to distinguish LSECs and HUVECs in pure cultures, co-cultures with either PHHs or fibroblasts, and tri-cultures containing endothelial cells, PHHs, and fibroblasts. Our results showed that while CD31 (immunostaining) and Factor VIII secretion (as assessed via enzyme linked immunosorbent assay) were detected in both endothelial cell types, SE-1 could distinguish LSECs from HUVECs, but only in pure monocultures reliably since this marker was downregulated in co-/tri-cultures containing LSECs, cross-reacted with mouse fibroblast protein(s), and showed upregulation in HUVECs in co-culture with PHHs. Overall, our studies set the stage for the creation and phenotypic characterization of multicellular human liver models on surfaces that mimic liver-like stiffness, which will have robust utility for drug screening and ultimately, regenerative medicine.