Generating microliver tissues to recapitulate hepatic function is of increasing importance in tissue engineering and drug screening. But the limited availability of primary hepatocytes and the marked loss of phenotype hinders their application. Human induced hepatocytes (hiHeps) generated by direct reprogramming can address the shortage of primary hepatocytes to make personalized drug prediction possible. Here, we simplify preparation of reprogramming reagents by expressing six transcriptional factors (HNF4A, FOXA2, FOXA3, ATF5, PROX1, and HNF1) from two lentiviral vectors, each expressing three factors. Transducing human fetal and adult fibroblasts with low vector dosage generated human induced hepatocyte-like... More
Generating microliver tissues to recapitulate hepatic function is of increasing importance in tissue engineering and drug screening. But the limited availability of primary hepatocytes and the marked loss of phenotype hinders their application. Human induced hepatocytes (hiHeps) generated by direct reprogramming can address the shortage of primary hepatocytes to make personalized drug prediction possible. Here, we simplify preparation of reprogramming reagents by expressing six transcriptional factors (HNF4A, FOXA2, FOXA3, ATF5, PROX1, and HNF1) from two lentiviral vectors, each expressing three factors. Transducing human fetal and adult fibroblasts with low vector dosage generated human induced hepatocyte-like cells (hiHeps) displaying characteristics of mature hepatocytes and capable of drug metabolism. To mimic the physiologic liver microenvironment and improve hepatocyte function, we prepared 3D scaffold-free microliver spheroids using hiHeps and human liver nonparenchymal cells through self-assembly without exogenous scaffolds. We then introduced the microliver spheroids into a two-organ microfluidic system to examine interactions between hepatocytes and tumor cells. The hiHeps-derived spheroids metabolized the prodrug capecitabine into the active metabolite 5-fluorouracil and induced toxicity in downstream tumor spheroids. Our results demonstrate that hiHeps can be used to make microliver spheroids and combined with a microfluidic system for drug evaluation. Our work could make it possible to use patient-specific hepatocyte-like cells to predict drug efficacy and side effects in various organs from the same patient.