Maternal Hepatic Adaptations to Pregnancy

Abstract

During gestation, the maternal liver undergoes various adaptive changes to cope with the in-creasing physiological and metabolic demands from both maternal and fetal compartments. Among these changes are robust growth and changes in transcriptome profile. However, how these events happen, and other aspects of this physiological phenomenon remains unexplored. Therefore, we aimed at further understanding how maternal liver responds to pregnancy. We used BrdU labeling combined with a virus-based tracing approach to quantify the percentage of maternal hepatocytes undergoing DNA synthesis and division over the course of gestation in mice. We found that ~50% maternal hepatocytes entered S-phase but, unexpectedly, did not undergo cytokinesis. This strongly suggests that maternal hepatocytes in fact undergo endoreplication instead of hyperplasia, as believed previously. Pericentral Axin2+ hepatocytes were reported to behave as liver stem cells responsible for liver homeostasis and turnover. We generated an in vivo fate-tracing mouse model to monitor the behavior of these cells in the maternal liver. Our results showed that they did not proliferate during pregnancy, homeostasis, and following par-tial hepatectomy. Curiously, we uncovered that, hepatocytes exhibit developmental phenotypes at mRNA level pre-pregnancy and at both mRNA and protein level during pregnancy. In the non-pregnant state, hepatocytes reserved mRNA expression of liver progenitor marker genes Cd133 and Afp, which are localized in the nuclei, without protein translation. During gestation, maternal hepatocytes displayed cytoplasmic translocation of Cd133 and Afp transcripts, con-comitant with corresponding protein expression. Overall, all maternal hepatocytes became CD133+, and a subset of them express AFP. Addi-tionally, in non-pregnant livers, mRNA of Epcam, another liver progenitor marker, was ex-pressed within majority of hepatocytes, whereas its protein was solely translated in the pericen-tral region. In contrast, by end-gestation, EPCAM protein expression switched to the periportal region. These observations indicate that maternal hepatocytes exhibit heterogeneous develop-mental phenotypes, partially resembling fetal hepatocytes. It is intriguing why mature hepato-cytes dedifferentiate into a progenitor state in response to pregnancy. AFP is considered to be produced primarily from fetal liver and thus is used to evaluate fetal development health. A potential clinical relevance of our data is that we identified maternal liver as a new source of AFP. The hippo signaling pathway has been shown to potently control liver growth and hepato-cyte heterogenicity. Surprisingly, we found that pregnancy neither altered the expression nor activities of the components of this pathway and its effector YAP1/TAZ. This finding indicates that pregnancy-induced maternal liver growth is not driven by hippo-YAP1 pathway. However, we demonstrate that the presence of YAP1 is essential for CD133 protein expression in mater-nal hepatocytes. Collectively, we revealed that, as pregnancy advances, maternal hepatocytes likely undergo endoreplication and display developmental phenotypes. Mechanistically, YAP1 dictates the expression of CD133, contributing to the pregnancy-dependent phenotypic changes of maternal hepatocytes.