Background and Aims: Acetaminophen (APAP) overdose is the leading cause of acute liver failure, with one available treatment, N-acetyl cysteine (NAC). Yet, NAC effectiveness diminishes about ten hours after APAP overdose, urging for therapeutic alternatives. This study addresses this need by deciphering a mechanism of sexual dimorphism in APAP-induced liver injury, and leveraging it to accelerate liver recovery via growth hormone (GH) treatment. GH secretory patterns, pulsatile in males and near-continuous in females, determine the sex bias in many liver metabolic functions. Here, we aim to establish GH as a novel therapy to treat APAP hepatotoxicity. Approach and Results: Our results demonstrate sex-dependent APAP toxicity, with females showing reduced liver cell death and faster recovery than males. Single-cell RNA sequencing analyses reveal that female hepatocytes have significantly greater levels of GH receptor expression and GH pathway activation compared to males. In harnessing this female-specific advantage, we demonstrate that a single injection of recombinant human GH protein accelerates liver recovery, promotes survival in males following sub-lethal dose of APAP, and is superior to standard-of-care NAC. Alternatively, slow-release delivery of human GH via the safe non-integrative lipid nanoparticle-encapsulated nucleoside-modified mRNA (mRNA-LNP), a technology validated by widely used COVID-19 vaccines, rescues males from APAP-induced death that otherwise occurred in control mRNA-LNP-treated mice. Conclusions: Our study demonstrates a sexually dimorphic liver repair advantage in females following APAP overdose, leveraged by establishing GH as an alternative treatment, delivered either as recombinant protein or mRNA-LNP, to potentially prevent liver failure and liver transplant in APAP-overdosed patients.

A reduction in GH, as well as IGF1, is associated with non-alcoholic fatty liver disease (NAFLD). However, the relative contribution of changes in circulating GH and IGF1, to hepatic triglyceride accumulation (steatosis), remains to be clearly defined. To study the direct actions of GH on hepatocyte metabolism, we have utilized a mouse model of adult-onset, hepatocyte-specific, GHR knockdown (aHepGHRkd; 10–12 week-old, GHRfl/fl male mice, treated with AAV8-TBGp-Cre). In this and previous reports, we have observed that aHepGHRkd male mice rapidly develop steatosis (after 7 days) associated with enhanced de novo lipogenesis (DNL; measured by deuterated H2O labeling, 10h after 0800h food removal), and low ketone levels, suggestive of reduced hepatic β-oxidation. Of note, aHepGHRkd also reduces plasma IGF1 levels to >80% of GHR-intact controls (GHRfl/fl mice treated with AAV8-TBGp-Null), leading to a rise in GH, due to loss of IGF1 negative feedback to the pituitary/hypothalamus. This reciprocal shift in IGF1/GH is associated with an increase in insulin levels. Therefore, it is possible that the steatosis that develops in aHepGHRkd mice is the consequence of systemic insulin resistance supplying excess substrates (glucose and NEFA) for hepatic lipogenesis. However, inconsistent with this theory is the fact that glucose and NEFA levels are not altered after aHepGHRkd. To tease out the indirect (perhaps driven by high insulin levels) vs. direct effects of GH on hepatocyte lipid accumulation, male aHepGHRkd mice were injected with a vector expressing rat IGF1 (AAV8-TBGp-rIGF1). Reconstitution of hepatocyte IGF1 in aHepGHRkd mice, raised plasma IGF1 and normalized GH, insulin and ketone levels, but hepatic steatosis and DNL remained greater than that of GHR-intact controls, indicating GH directly suppresses hepatic fat accumulation. RNAseq analysis of livers from aHepGHRkd mice showed expression of genes related to carbohydrate metabolism (Gck, Khk) and fatty acid synthesis (Fasn, Srebf1, Usf1), processing (Scd1) and uptake (Cd36) were increased, while genes related to gluconeogenesis (Pck1, Fbp1, G6pc) were reduced. Remarkably, IGF1 reconstitution had no major impact on the hepatic transcriptome of aHepGHRkd mice, with the exception of reducing the expression of Srebf1, consistent with the reduction in circulating insulin levels. Interestingly, carbohydrate-responsive element-binding protein (CHREBP) levels, but not mRNA levels, were greater in aHepGHRkd mice with or without IGF1 reconstitution, consistent with upregulation of CHREBP target genes (Khk and Fasn among others). Taken together, these results suggest GH directly regulates steatosis, at least in part, by suppressing carbohydrate-driven DNL, where additional studies are underway to test this hypothesis.