Influence of CPM-dependent sorting on the multi-omics profile of hepatocyte-like cells matured in microscale biochips

Highlights

• Hepatocytes-like cells were differentiated in Petri dishes, sorted, matured, and then inoculated in biochip.

• Cells issued from sorted populations reached higher levels of maturation when compared to unsorted ones.

• Transcriptomics, proteomics, and metabolomics profiles were characterized.

• Characteristic pathways differentially expressed between cells issued from sorted and unsorted populations were analyzed.

Abstract

Liver modeling in disease via advanced in vitro physiological tissues remains a challenging issue, yet crucial for the future development of relevant tools for drug screening. In that regard, advanced technics such as human induced Pluripotent Stem Cells (hiPSCs) and organ-on-chip are promising technologies for the relevant reproduction of the in vivo micro physiology in in vitro culture conditions. In the present work, the maturation of carboxypeptidase M positive (CPM+) Hepatocyte-Like-Cells (HLCs) in microfluidic culture conditions was investigated. hiPSCs were differentiated in culture dishes until sorting and further amplified until seeding in biochips. After 2 weeks of maturation in biochips, evaluation of the cell metabolism was performed, and samples were collected for characterization via RNA sequencing, proteomics and metabolomics. The omics profile of biochips loaded with CPM+ HLCs was then compared with the one of previously cultured biochips loaded with unsorted HLCs. CPM+ HLCs presented advanced liver characteristics in terms of a higher activity of important Transcriptions Factors (TFs) such as HNF1, HNF4A, and CEBP/A, of the upregulation of steroids/corticoids-related nuclear receptors (FXR, NR3C2, NR4A1 genes), of phase I, and phase II metabolism genes (CYP1A1, CYP1B1, CYP2C18, CYP27A1, several UGT, SULT and GST genes). Further differences in the cellular reorganization, in the lipids and steroids metabolisms, in the OXPHOS respiration and in the expression of TGFβ signaling were also observed. The present study introduces a novel protocol using advances hiPSCs differentiation and sorting methods as well as the organ-on-chip technology to highlight important in vitro liver regeneration and hepatic maturation processes.

Introduction

The liver plays a central role in the metabolism of xenobiotics and in their detoxification while also being involved in regulation of lipids, proteins, and amino acids, in glucose homeostasis, and in bile synthesis [23]. In that regard, efficient in vitro models are highly sought for by industrials for drug testing purposes and by clinicians for the understanding of new mechanisms. Especially, microfluidics devices and organ on chip technologies are expected to be able to enhance the precision of prediction of drug safety and efficacy during preclinical tests by reproducing accurately the physiological environment of the targeted organ [32]. By doing so, the development of therapeutical solutions would not only be more accurate but also more cost-efficient which is a recurring issue for the industry [14]. However, numerous challenges in the field remain as compatibility with manufacturing processes and standardization is often low for these models [34] and will need to be solved once satisfying models have been developed. Focusing of the reproduction of the physiology of the liver, organ on chip must go through the reproduction of its microenvironment including the canals of Hering and the liver acinus. Of this, hepatic zonation-like patterns [1], [24], maintenance of primary hepatic cells for drug metabolism applications [7], and liver cocultures [38] have been proposed and have shown significant improvement of functionality of conventional culture methods. However, the combination of these models with new cutting-edge technologies in cell biology such as stem cells is still under investigation.

In that regard, hiPSCs-derived cells have been the topic of intensive research for the past few years as they are potentially allowed to differentiate into any cell type while being free from the constraint of limited cell number seen in primary cells harvested from donors. hiPSCs have also been combined with the organ on chip technology to form liver tissues, composed of Hepatocyte-Like Cells (HLCs), and which could be used for the improvement of the knowledge of liver homeostasis and liver regeneration processes as well as for the formation of mature in vitro liver tissue [11], [12], [18], [42], [45]. While the differentiation and maturation of hiPSCs-derived HLCs is expected to be enhanced by the culture in microfluidic devices either due to the presence of a physiological flow [48] or to the culture in confined environment [36], those cells still suffer from dominant patterns of primitiveness and are yet to lead to practical applications. A possible culprit for the unsatisfying maturation state of hiPSCs-derived cultures could be the low yield resulting from the differentiation process. A low yield in the differentiation would lead to the formation of a heterogeneous tissue which could have an uncontrolled effect on the behavior of HLCs. While positive effects have been observed [11], with the formation of an endothelial-like tissue, due to the effect of the shear stress from the microenvironment, control over the culture is needed to improve the reliability of those models for drug screening applications.

A possible strategy to increase control over the culture is cell sorting according to specific markers. Carboxypeptidase M (CPM) has been reported as a bipotential markers which led to the formation of highly functional, proliferating hepatocytes [25]. CPM+ cells are selected at the hepatic progenitor differentiation stage and can be amplified several times in culture dishes. The cells were then detached and inoculated in microfluidic biochips for maturation. Comparison of the function and of the transcriptome, proteome and metabolome profiles of those cells with previously published protocols (Unsorted HLCs matured in biochips, data published in [12]) was then performed and the influence of sorting for CPM positive populations was investigated. The analysis has led to a clear separation between the two conditions and improvements of the maturation in the CPM+ HLCs model could be confirmed by both the functional analysis and the multiomics analysis. The dataset presented in this study is of value toward the improvement of the culture of HLCs within microfluidic environment and should lead toward efficient liver on chip models for pharmaceutical applications.