Impact of ZEB2 on specification, zonation and regeneration of liver sinusoidal endothelium

Liver sinusoidal endothelial cells (LSECs) are a highly specialized type of endothelial cells (ECs) characterized by a high permeability due to the presence of transcellular pores, called fenestrations, a poorly developed basement membrane, and high expression levels of scavenger receptors. Due to this high permeability and their position at the interface between blood and hepatocytes, LSECs act as a selective filter of the blood, allowing transport of nutrients to the hepatocytes while clearing waste substances and intruding pathogens. At the same time, their strategic position also makes LSECs vulnerable, which is reflected by their rapid dedifferentiation during chronic liver disease towards a continuous capillary EC phenotype, a process known as capillarization, accompanied by LSEC dysfunction. LSEC capillarization is an important catalyzer of chronic liver disease progression. Although restoring LSEC specialization represents a promising strategy for chronic liver diseases of various etiologies, understanding its molecular regulation is only in its infancy.

The most prevalent type of chronic liver disease nowadays is non-alcoholic fatty liver disease (NAFLD). NAFLD has a prevalence of ~30% worldwide. Due to its close relation with the metabolic syndrome and obesity, its prevalence is however still increasing. NAFLD is responsible for 4% of all deaths and is becoming the leading cause of liver transplantation. Unfortunately, no effective pharmacological treatments exist, thereby surrendering patients to supportive therapy and encouragement for weight loss and lifestyle interventions. NAFLD represents a spectrum of chronic liver disease, comprising steatosis in its initial stage, which can evolve to the inflammatory stage non-alcoholic steatohepatitis (NASH), which causes fibrosis. In its end stage, the chronically diseased liver becomes cirrhotic, characterized by angiogenesis to create portal-systemic shunts and regenerative nodules. LSECs are involved in all main pathological processes of chronic liver disease progression, via cell autonomous and non-autonomous mechanisms.

Transcription factor zinc-finger E-box binding homeobox (ZEB) 2, well described in development and cancer due to its role in neuronal differentiation and endothelial-to-mesenchymal transition, respectively, was amongst other transcription factors previously identified by my host lab as a potential regulator of LSEC specialization owing to its enrichment in liver compared to heart and brain ECs. We decided to study if ZEB2 is indeed important for LSEC specialization during maintenance, fibrosis and steatosis. Therefore, we manipulated ZEB2 expression specifically in ECs of EC-specific Zeb2 knock-out (ECZeb2KO) or overexpression (ECOE) mice. To study its role during chronic liver disease, we used a toxin-induced fibrosis model and a western-type diet (WD)-induced model of NAFLD.

Sequencing of liver EC RNA early after deleting ZEB2 revealed an increase in continuous EC markers suggesting that Zeb2-deficient LSECs are more prone to dedifferentiation and capillarization. However, fenestrations were preserved upon ZEB2 deficiency. While functional annotation on differentially expressed genes (DEGs) 2 weeks after tamoxifen induction showed increased Platelet-derived growth factor (PDGF)-B signaling and angiogenesis but not proliferation, increased proliferation was the predominant functional annotation 6 weeks after Zeb2 deletion, likely reflecting different modes of angiogenesis triggered by endothelial Zeb2 deletion. Accordingly, the vascular area was expanded and the presence of pillars inside ECZeb2KO liver vessels indicated that this was likely due to increased intussusceptive angiogenesis. While, shRNA-mediated ZEB2 knockdown in HUVECs did not affect proliferation, it did affect the number of branching points and branch length formed by HUVECs which are parameters rather associated with sprouting angiogenesis. Hence, we identified ZEB2 as a novel multi-modal regulator of LSEC maintenance.

Changes in the endothelial expression of ligands that may be involved in hepatic stellate cell (HSC) quiescence together with significant changes in the expression profile of HSCs showed that Zeb2 regulates LSEC-HSC communication and HSC activation. Accordingly, upon exposure to the hepatotoxin carbon tetrachloride (CCl4), livers of ECKO mice showed increased capillarization, HSC activation and fibrosis compared to livers from wild-type (WT) littermates. The vascular maintenance and anti-fibrotic role of endothelial Zeb2 was confirmed in ECOE mice, as the latter resulted in reduced vascularity and attenuated CCl4-induced liver fibrosis.

To study the role of ZEB2 in LSECs during NAFLD, we fed ECZeb2KO mice a WD or standard diet (SD). In healthy and diseased WT livers, Zeb2 was ubiquitously and similarly expressed across blood vascular EC types. RNA sequencing of liver ECs revealed that Zeb2 deficiency triggers gene expression changes greatly overlapping with those evoked by short-term WD-feeding. Functional annotation on DEGs revealed that these gene expression changes are related to cell division and Rho GTPase signaling. Immunofluorescence staining for KI67 confirmed increased proliferation in LSECs upon ZEB2-loss, WD-feeding or the combination of both. Furthermore, endothelial ZEB2-loss and WD-feeding interacted to boost capillarization, evident from profoundly affected LSEC marker expression. Functional annotation on DEGs emerging from the interaction of ZEB2-loss and WD-feeding specifically revealed gene expression changes related to ‘peroxisome proliferator activated receptor (PPAR) signaling’ and the ‘complement and coagulation cascade’. The latter represents an aspect of EC dysfunction and was shown by reduced F8 expression and impaired blood coagulation. Intriguingly, altered communication among LSECs after combined endothelial ZEB2-loss and WD-exposure revealed similar functional repercussions. Remarkably, Peptidylarginine deiminase 4 (Padi4), the gene encoding an enzyme mainly described in neutrophil extracellular trap formation, emerged as an early-disease gene with citrullination activity induced by short-term WD-feeding and ZEB2-loss in LSECs. Despite causing increased signs of LSEC capillarization, endothelial ZEB2-loss eventually corrected WD-induced hypo-vascularization of the liver parenchyma while decreasing weight gain, hepatic damage and steatosis.

In conclusion, endothelial ZEB2-loss induces LSEC capillarization and angiogenesis during physiological conditions. In the context of toxin-induced liver fibrosis, ZEB2-loss in ECs accelerates liver fibrogenesis, while it decreases steatosis development in a model of NAFLD. During NAFLD, endothelial Zeb2 deficiency decreases steatosis development potentially by boosting LSEC fat metabolism. Due to the context-dependent role of endothelial ZEB2, further research is needed to decipher whether ZEB2 inhibition or stimulation would be most favorable for disease outcome, before exploiting ZEB2 and ZEB2-dependent genes in liver ECs to design novel therapeutic strategies for hepatic fibrosis or steatosis.