The creation of human liver vascular models towards artificial tissue generation

Abstract

The liver’s intricate architecture is critical for its physiological function, and its disruption is pathologic. Liver cirrhosis, the 11th most common cause of death globally, is marked by significant architectural changes and decreased function in the liver, requiring liver transplant in severe cases. The high demand for liver transplantation and unmet need for liver tissue is inspiring research efforts towards creating artificial liver tissue in vitro. As the liver’s function is closely tied to its structure, the development of functional artificial liver tissue may necessitate recapitulating native liver structure in artificial tissue. Towards this end, we first helped elucidate multilobular human liver architecture by creating the largest, cellularly resolved, 3D reconstructions of non-cirrhotic and cirrhotic human liver tissue known to date. We then compared the functional and structural features of non-cirrhotic tissue with cirrhotic tissue to help decipher which features may be essential for maintaining functional liver tissue. Our comparative multilobular analyses of non-fibrotic and cirrhotic tissues revealed that cirrhosis correlates with sinusoid zonation dysregulation, reduction in glutamine synthetase-expressing pericentral hepatocytes, regression of central vein networks, and fragmentation of biliary networks. Taken together, these observations suggest a pro-portalization/decentralization phenotype in correlation with human liver cirrhosis and emphasize the importance of the liver’s central features for maintaining functional liver tissue. After gaining new appreciation for multilobular human liver architecture, we next pursued computational modeling strategies to generate anatomically accurate human liver vascular designs for future incorporation into artificial liver tissue. Towards this end, we first acquired parameterized morphometrics and performed multiscale analyses on central vein, portal vein, and liver sinusoidal networks. We thereafter collaborated with Nervous System to computationally generate an interpenetrating central vein and portal vein networks design and a 3D sinusoidal mesh network design. Future computational work entails combining modeling methodologies to create interpenetrating central vein and portal vein networks connected by a 3D sinusoidal mesh, thereby creating a scalable multilobular human liver vascular model which we aim to incorporate into artificial liver tissue in future studies. Towards these future tissue engineering efforts, we lastly used DLP additive manufacturing and multi-LSL photodegradation to show feasibility for fabricating the initial interpenetrating networks design and 3D sinusoidal mesh network design respectively within biocompatible hydrogels. All in all, this dissertation work presented the largest cellularly-resolved 3D reconstructions of human liver tissue to date, revealed a pro-portalization/decentralization phenotype in correlation with human liver cirrhosis, generated anatomically accurate human liver vascular designs, and assessed the feasibility of two light-based bioprinting strategies for fabricating human liver vascular designs within biocompatible hydrogels. These collective efforts provide foundational knowledge on human liver architecture in its healthy and diseased states and offer insight into the computational modeling and subsequent fabrication of human liver vascular networks, which together may help advance tissue engineering initiatives towards creating functional artificial liver tissue in the future.