Conducting the Orchestra: Nonparenchymal Cells Direct Hepatocyte Functions in Steatotic Liver Disease Authors: Julius W Jaeger, Robert F Schwabe Published: 2026 Steatotic liver disease (SLD) encompasses multiple liver diseases with increased intrahepatic fat accumulation, including metabolic dysfunction–associated steatotic liver disease (MASLD), alcoholic liver disease (ALD), and a mixed form termed metabolic– and alcohol-associated liver disease (MetALD). SLD has a high prevalence, with MASLD alone affecting >30% of the global population.1 Furthermore, treatment options for SLD remain limited despite the recent approval of resmetirom for the treatment of metabolic-associated steatohepatitis (MASH) and more effective treatments on the horizon.2 Hence, a better understanding of the molecular mechanisms driving hepatic lipid overload is needed. A central feature of hepatocyte protein, carbohydrate and lipid metabolism in the liver is its spatial organization along oxygen and metabolic gradients from the portal tracts towards the central vein, commonly termed zonation.3,4 Metabolic homeostasis is orchestrated by a complex interplay of multiple liver cells, that includes hepatocytes and mainly 2 types of nonparenchymal cells (NPCs), endothelial cells (ECs) and hepatic stellate cells (HSCs).3,5 ECs and HSCs engage in continuous crosstalk with hepatocytes via angiocrine and stellacrine factors. Within the network that regulates metabolic liver zonation, WNT-β-catenin signaling represents a central pathway, controlled by WNT2 and WNT9b secretion from ECs, and additionally by R-spondin 3 (RSPO3) secretion by HSCs.6,7 In this issue of Cellular and Molecular Gastroenterology and Hepatology, Rao et al expand our understanding of the contribution of the endothelial transmembrane receptor HEG1 as a gatekeeper of hepatocyte lipid homeostasis and liver zonation.8 The authors utilize 2 complementary, endothelial-specific Heg1 knockout models across 3 distinct liver disease models: MASLD through a high-fat diet, MASH via a methionine–choline-deficient diet, and ALD via the National Institute on Alcohol Abuse and Alcoholism protocol. In these models of SLD, they demonstrate exacerbated hepatic steatosis following endothelial Heg1 deletion. Through transcriptomic and lipidomic analysis, the authors characterize hepatic suppression of hepatocyte PPARα signaling and downstream fatty acid oxidation (FAO) enzymes, including Cact, Acadvl, Acadl, Acadm, Ech1, Hadh, Ehhadh, Acaa1, and Acaa2, with subsequent accumulation of acylcarnitine and mitochondrial injury, namely mitochondrial swelling, membrane rupture, and fragmentation. Rao et al identify endothelial Wnt signaling, including Wnt2, Wnt9b, and Rspo3 through RNA sequencing analysis of isolated ECs as the relevant drivers responsible for the metabolic phenotype.8 They confirm these findings via Rspo3 re-expression in ECs, which reversed lipid accumulation and PPARα expression levels and rescued metabolic zonation patterning. Mechanistically, the authors demonstrate that Heg1 facilitates bone morphogenetic protein (BMP) signaling in liver ECs. Accordingly, small interfering RNA–mediated knockdown of Heg1 in cultured ECs reduced BMP signaling activation, which could be alleviated through treatment with BMP9 and BMP agonist SY-LB-35. Furthermore, in vivo administration of SY-LB-35 reversed the altered metabolic zonation in mice with EC-specific knockout of Heg1. Rao et al add to our understanding of the WNT/RSPO3-β-catenin pathway as a central regulator of hepatic lipid homeostasis. Liver zonation is severely disturbed in MASLD,9 and previous studies in mice with hepatocyte-specific knockout of beta-catenin or HSC-specific knockout of RSPO3 not only revealed altered zonation but also pericentral lipid accumulation in homeostasis, worsening of steatosis and liver injury in MASLD and ALD models, and an association of RSPO3 levels with survival in patients.7,10,11 Rao et al now describe direct angiocrine modulation of hepatocyte FAO.9 Furthermore, the authors identify the HEG1-BMP signaling conjunction in liver ECs as an additional upstream mediator, consolidating our perception of intercellular WNT/RSPO3 signaling and complementing previous work establishing HSCs as a major source of BMP9/10 and RSPO3.7,12 Interestingly, these findings extend to both MASLD/MASH and ALD, indicating an integrative NPC-dependent mechanism, which modulates FAO in dietary and toxic injury. The study by Rao et al may open up new possibilities for novel targeted therapeutic interventions. The SY-LB-35 rescue provides a potential selective intervention of the BMP–WNT/RSPO3 arm that could restore FAO and consequently improve the metabolic phenotype without the oncogenic risk of WNT/β-catenin activation. Based on these findings, RSPO3 mimetics and BMP9 analogs provide novel therapeutic avenues for the treatment of MASLD and ALD, although a fine balance between WNT under- and overactivation may be difficult to achieve.5 However, several critical questions require additional studies. Whether endothelial RSPO3 plays a critical role in this process remains to be further investigated in mice with EC-specific knockout of RSPO3, as overexpression of RSPO3 in any cell type is sufficient for rescue. Mice with HSC-specific but not with EC-specific knockout of RSPO3 displayed major alterations in liver size, metabolic zonation, and regeneration, suggesting that the small number of RSPO3-positive pericentral ECs may play distinct roles, such as supporting the clearance of ammonia, whereas RSPO3-positive HSCs are the main regulators of most liver metabolic pathways.7 The observation of substantial zone 2 lipid accumulation in mice with EC-specific Heg1 knockout compared with the previously established pericentral zone 3 deposition is interesting and potentially necessitates a re-examination of our current understanding of zone-specific pathologic vulnerability in MASLD. Furthermore, the precise molecular mechanism by which Heg1 modulates BMP signaling warrants further investigation prior to development of targeted therapeutic strategies. Nevertheless, this work by Rao et al consolidates the role of NPC-dependent signaling as central “conductors” of hepatocyte steatosis and potential therapeutic target. Footnotes Conflicts of interest The authors disclose no conflicts. References