Hepatic Huwe1 loss protects mice from non-alcoholic fatty liver disease through lipid metabolic rewiring

Summary

Non-alcoholic fatty liver disease (NAFLD) is the most pervasive liver pathology worldwide. Here, we demonstrate that the ubiquitin E3 ligase Huwe1 is vital in NAFLD pathogenesis. Using mass spectrometry and RNA sequencing, we reveal that liver-specific deletion of Huwe1 (Huwe1^LKO) in 1-year-old mice (approximately middle age in humans) elicits extensive lipid metabolic reprogramming that involves downregulation of de novo lipogenesis and fatty acid uptake, upregulation of fatty acid β-oxidation, and increased oxidative phosphorylation. ChEA transcription factor prediction analysis inferred these changes result from attenuated PPARɑ, LXR, and RXR activity in Huwe1^LKO livers. Consequently, Huwe1^LKO mice fed chow diet exhibited significantly reduced hepatic steatosis and superior glucose tolerance compared to wild-type mice. Huwe1^LKO also conferred protection from high-fat diet-induced hepatic steatosis by 6-months of age, with increasingly robust differences observed as mice reached middle age. Together, we present evidence that Huwe1 plays a critical role in the development of age- and diet-induced NAFLD.

Graphical abstract

Highlights

• •Liver-specific deletion of Huwe1 prevents NAFLD induced by aging in mice
• •Liver-specific deletion of Huwe1 also protects mice from diet-induced NAFLD
• •Mass spectrometry and RNA-seq analyses identified metabolic remodeling in Huwe1 KO
• •ChEA analysis revealed downregulation of PPARα, LXR, and RXR activity in Huwe1 KO

Biochemistry; Pathophysiology

Introduction

Non-alcoholic fatty liver disease (NAFLD) is the most prevalent form of liver disease worldwide and is estimated to affect nearly a third of the population.¹ The hallmark characteristic of NAFLD is hepatic steatosis, or exacerbated lipid accumulation. Importantly, NAFLD is not a single disorder but rather encompasses a spectrum of liver diseases. NAFLD begins with the development of hepatic steatosis, but chronic inflammation mediates its progression into non-alcoholic steatohepatitis (NASH).^1,2 Subsequent liver damage, such as cirrhosis and fibrosis, then accrues over time and strongly predisposes individuals to hepatocellular carcinoma (HCC).^1,2,3 Several factors are known to contribute to the onset and progression of NAFLD-related disorders, including aging, obesity, hypertension, dyslipidemia, type 2 diabetes (T2DM), and metabolic syndrome.^1,3,4 Although hepatic steatosis seen in early NAFLD can be reversed with lifestyle interventions such as diet and exercise, liver damage brought on by cirrhosis and fibrosis is often regarded as irreversible. Because there are no Food and Drug Administration (FDA)-approved therapeutics for the treatment of NAFLD/NASH, there remains a critical need to identify strategies to prevent and reverse NAFLD-related pathologies.

Huwe1, also known as ARF-BP1, MULE, and HectH9, is a HECT (homology to E6-APC terminus)-domain E3 ubiquitin ligase originally identified as a binding partner of the tumor suppressor ARF (alternate reading frame; p14^ARF in humans, p19^ARF in mice), as well as a direct negative regulator of the tumor suppressor p53.⁵ Overexpressed in a variety of cancer types, Huwe1 demonstrates an additional aspect of pro-tumorigenicity by activating transcriptional activity of the proto-oncogene c-Myc via K-63 polyubiquitination.⁶ Conversely, Huwe1 has also been shown to promote cancer cell apoptosis following DNA damage by facilitating the proteasomal degradation of MCL-1, an anti-apoptotic BCL-2 family member.⁷ These seemingly contradictory observations illustrate that the role of Huwe1 in vivo is often complicated due to the regulation of variable tissue-specific substrates with distinct context-dependent functions. For example, Huwe1 negatively regulates N-Myc in stem cells found in neural^8,9 and hematopoietic compartments¹⁰ to inhibit proliferation, thereby maintaining stemness. In male germ cells, Huwe1 was found to play a critical role in spermatogenesis by ubiquitinating histones H1, H2A, H2B, H3, and H4 to regulate chromatin condensation.¹¹ Contrary to neural and hematopoietic stem cell populations, where inactivation of Huwe1 promotes cellular proliferation and differentiation, Huwe1 knockout in germ cells leads to a defect in spermatogenesis via hyperactivation of the DNA damage response pathway, and ultimately male infertility.^12,13 Moreover, we have previously demonstrated that oocyte-specific Huwe1 knockout also renders female mice infertile, albeit via a p53-independent mechanism.¹⁴ Therefore, the tissue-specific roles of Huwe1 remain heavily under investigation.

To characterize the in vivo role of Huwe1, we generated a liver-specific Huwe1 (Huwe1^LKO) knockout mouse model. Surprisingly, liver-specific knockout of Huwe1 protected mice from the development of age-induced hepatic steatosis. Mass spectrometry and RNA sequencing (RNA-seq) analyses of wild-type (Huwe1^WT) and Huwe1^LKO mice revealed that Huwe1^LKO resulted in marked downregulation of hepatic lipid metabolism genes and concomitant upregulation of genes related to mitochondrial fatty acid (FA) β-oxidation and oxidative phosphorylation. Accordingly, Huwe1^LKO mice were also protected from high-fat diet (HFD)-induced development of hepatic steatosis and exhibited reduced accumulation of body fat percentage and body weight, in addition to lower incidence of HCC. Huwe1^LKO mice also demonstrated resistance to developing glucose intolerance following prolonged HFD administration. Our results place Huwe1 as a novel regulator of age- and diet-induced NAFLD.

Results

Liver-specific inactivation of Huwe1 attenuates the development of age-associated hepatic steatosis

We first generated a liver-specific Huwe1 knockout mouse model by crossing previously established Huwe1^f//fl mice¹⁵ with mice expressing Cre recombinase under the control of the hepatocyte-specific albumin promoter (Alb-cre)¹⁶ (Figure 1A). Although the global deletion of Huwe1 is known to be embryonically lethal,^14,15 liver-specific inactivation allowed for the development of viable pups. Western blot of liver tissue lysates was performed to confirm knockout efficiency in Huwe1^LKO mice (Figure 1B). To investigate the physiological role of hepatic Huwe1, we measured the body weights of Huwe1^WT and Huwe1^LKO mice maintained on a standard chow diet over the span of 1 year (Figure 1C). Although no differences in body weight were observed earlier in life, a statistically significant 15% reduction in body weight was observed in Huwe1^LKO mice relative to Huwe1^WT as they became 1 year old (approximately equivalent to middle age in humans¹⁷ [Figures 1D and 1E]). Importantly, the modest reduction in body weight was not attributed to differences in diet consumption (Figure 1F).

Figure 1.

Liver-specific deletion of Huwe1 reduces body weight gain in 1-year-old mice

(A) PCR genotyping results depicting the detection of wild-type and floxed Huwe1 alleles in wild-type (Huwe1^+/+), heterozygous (Huwe1^+/fl), and homozygous floxed mice (Huwe1^fl/fl).

(B) Western blot confirming knockout efficiency in Huwe1^LKO livers.

(C) Schematic depicting the timeline of the diet study presented in D–F. Huwe1^WT and Huwe1^LKO body weight and food consumption were measured weekly starting at 4 weeks of age. This cohort of mice was evaluated over the span of 1 year. N = 15 mice per group.

(D) Representative photograph of 1-year-old Huwe1^WT and Huwe1^LKO mice.

(E and F) Body weight (E) and diet consumption (F) of Huwe1^WT and Huwe1^LKO mice maintained on chow diet for 1 year. Data are shown as mean ± SEM. See STAR Methods section for further detail of growth curve analysis.

Interestingly, 1-year-old Huwe1^LKO mice presented significantly smaller livers as compared to their wild-type counterparts although no difference was observed at 24 weeks of age (Figures 2A and 2B). Because aging is a prominent risk factor for the development of hepatic steatosis, we next histologically evaluated the liver morphology of 1-year-old mice. Hematoxylin and eosin (H&E) staining revealed that livers from 1-year-old Huwe1^LKO mice exhibited profoundly less macrovesicular steatosis as compared to Huwe1^WT livers (Figure 2C). Moreover, Oil Red O (ORO) staining confirmed that Huwe1^LKO livers from 1-year-old mice exhibited a marked reduction in lipid droplet accumulation (Figure 2D). In addition, hepatic triglyceride (Figure 2E) and cholesterol (Figure 2F) levels were significantly reduced in 1-year-old Huwe1^LKO livers as compared to Huwe1^WT. Since NAFLD prevalence in humans also peaks in middle age (∼30–50 years old),¹⁸ the kinetics of hepatic steatosis onset in Huwe1^WT mice recapitulates the temporal etiology of the human disease. Interestingly, the protein expression levels of both FA synthase (FASN) and the FA transporter CD36 were markedly downregulated in 1-year-old Huwe1^LKO livers, although the level of carnitine palmitoyl transferase I (CPT1), the rate-limiting enzyme of FA oxidation, appeared to be unaffected by Huwe1^LKO (Figure 2G). These results suggest that Huwe1 plays a critical role in the development of age-associated NAFLD in part through diminished de novo lipogenesis and FA uptake.

Figure 2.

Liver-specific deletion of Huwe1 reduces liver size and lipid accumulation in 1-year-old mice

(A) Representative photographs of livers from 1-year-old Huwe1^WT and Huwe1^LKO mice.

(B) Huwe1^LKO mice exhibit no change in liver size relative to Huwe1^WT at 24 weeks of age but a significant reduction at 1 year of age. Liver weights were normalized to body weight at each time point.

(C and D) H&E (C) and ORO (D) staining of liver sections from middle-aged mice.

(E and F) Biochemical analysis measuring triglyceride (E) and cholesterol (F) in livers from 1-year-old mice. Data shown as mean +SD. Significance assessed by two-sided Student’s t test with significance set at ∗ <0.05, ∗∗ <0.005.

(G) Western blot was performed for the indicated proteins in livers isolated from 1-year-old Huwe1^WT and Huwe1^LKO mice.

Huwe1^LKO results in downregulation of de novo lipogenesis and upregulation of mitochondrial β-oxidation

Huwe1 is an E3 ubiquitin ligase with a wide range of substrates which often varies based on the context, such as tissue type. To elucidate the mechanism underlying the healthy liver phenotype observed in Huwe1^LKO mice, we next performed mass spectrometry analysis on livers harvested from 1-year-old mice. A total of 710 proteins were detected in both Huwe1^WT and Huwe1^LKO livers (Figure 3A). Additionally, 190 proteins were measurable only in Huwe1^WT livers and 341 proteins were found solely in Huwe1^LKO livers. However, many of these proteins were detected with low abundance or inconsistently across replicates. We then sought to identify differentially expressed proteins between Huwe1^WT and Huwe1^LKO livers with a log2 fold change (log2FC) of |log2FC| > 0.5 and a p value of p < 0.05 (Figure 3B). Using these parameters, we identified 82 proteins to be significantly enriched in Huwe1^LKO livers (including 4 Huwe1^LKO-specific proteins) and 45 proteins to be significantly depleted in Huwe1^LKO livers (including 8 Huwe1^WT-specific proteins). Gene set enrichment analysis revealed that “Hallmark_Fatty_Acid_Metabolism” was the most significantly upregulated (Figure S1A [Huwe1^LKO livers from 1-year-old mice exhibit enrichment of metabolic proteins and pathways]) and downregulated pathway (Figure 3C) in Huwe1^LKO livers. Indeed, we observed that several differentially expressed proteins in Huwe1^LKO livers had roles in various aspects of lipid metabolism, including FA synthesis (AACS, ACACA/ACC1, ACACB/ACC2, ACLY, ACSA/ACSS2, FASN, MAOX/ME1), FA β-oxidation (ACADL, ACADV/ACADVL, EXHB/HADHB, THIM/ACAA2), lipid transfer (FABPL/FABP1, STARD10), and bile acid synthesis (CP270/CYP2C70, CP240/CYP2C40) (Figure 3D; Figure S1B [Huwe1^LKO livers from 1-year-old mice exhibit enrichment of metabolic proteins and pathways]; Table S1 [Significantly enriched proteins in Huwe1^LKO livers]; Table S2. [Significantly depleted proteins in Huwe1^LKO livers]).

Figure 3.

Mass spectrometry analysis reveals drastic metabolic remodeling in Huwe1^LKO livers from 1-year-old mice

(A) Venn diagram listing total number of proteins detected in Huwe1^WT and Huwe1^LKO livers by mass spectrometry. |log2FC| > 0.5 and p < 0.05 thresholds were set for filtering of significant differentially expressed proteins.

(B) Volcano plot depicting differentially expressed proteins in Huwe1^WT and Huwe1^LKO livers.

(C) Hallmark gene set enrichment analysis was performed on the 45 significantly depleted proteins from Huwe1^LKO livers using MSigDB.

(D) Bar graph showing the top 25 depleted proteins from Huwe1^LKO livers sorted by -log2FC. Notably, many of these proteins were related to lipid metabolism or associated pathways (highlighted in red boxes).

(E) Signaling diagram depicting upregulated (green) and downregulated (red) proteins and metabolic processes in Huwe1^LKO livers.

(F and G) ChEA analysis on all 127 differentially expressed proteins in Huwe1^LKO livers revealed strong enrichment of RXR, PPARα, and LXR activity (F). Many of the differentially expressed proteins in Huwe1^LKO livers are predicted to be regulated by RXR, PPARα, and/or LXR (G). Upregulated (Green) and downregulated (Red) proteins related to fatty acid metabolism are highlighted.

(H) Western blot was performed for the indicated proteins in livers isolated from 2-month-old and 1-year-old Huwe1^WT and Huwe1^LKO mice. Band densitometric values shown indicate Hsp90-normalized expression values relative to average expression of Huwe1^WT samples.

Upon closer scrutiny, we observed downregulation of nearly the entire de novo lipogenic pathway in the Huwe1^LKO livers (Figures 3D and 3E). Particularly, two of the most highly depleted proteins were ACACA and ACACB (Figures 3B and 3D). These proteins catalyze the same irreversible carboxylation of acetyl-coenzyme A (CoA) to malonyl-CoA but differ in their subcellular localizations and functional consequences of their activities (Figure 3E). Specifically, mitochondrial ACACB produces malonyl-CoA to serve as a potent allosteric inhibitor of CPT1, a process which acts as a molecular switch to favor FA synthesis over catabolism. ACACA is the cytosolic isoform which produces malonyl-CoA in the rate limiting step of de novo FA synthesis. FASN then catalyzes the synthesis of FAs from malonyl-CoA and acetyl-CoA through multiple cycles of elongation reactions (Figure 3E). Consistent with our earlier result (Figure 2G), FASN was also markedly depleted in Huwe1^LKO livers (Figures 3B and 3D). Furthermore, we also observed decreased expression of ACLY and ACSS2 (Figures 3B and 3D), which catalyze the synthesis of the lipogenesis substrate acetyl-CoA from citrate and acetate, respectively (Figure 3E).

In addition to the robust downregulation of the lipogenesis pathway, Huwe1^LKO livers showed concomitant increased abundance of proteins involved in FA β-oxidation (ACAA2, ACADL, ACADVL, HADHB), tricarboxylic acid (TCA) cycle (IDHP/IDH2 and ODO1/OGDH), and glutaminolysis (GLSL/GLS2 and DHE3/GLUD1) (Figure 3E; Figure S1B [Huwe1^LKO livers from 1-year-old mice exhibit enrichment of metabolic proteins and pathways]; Table S1 [Significantly enriched proteins in Huwe1^LKO livers]). We also found the electron transfer flavoprotein ETFB and ATP synthase subunit F1 alpha (ATP5F1A) to be upregulated in Huwe1^LKO livers (Table S1 [Significantly enriched proteins in Huwe1^LKO livers]). ETFB facilitates the transfer of electrons from multiple flavoprotein dehydrogenases involved in FA β-oxidation and delivers them to NADH:ubiquinone oxidoreductase (Complex I) of the electron transport chain (ETC),¹⁹ while ATP synthase (Complex V) catalyzes the final step of the ETC to generate ATP (Figure 3E). Interestingly, in contrast to the upregulation of mitochondrial FA β-oxidation enzymes in Huwe1^LKO livers, we observed concurrent downregulation of peroxisomal β-oxidation enzymes (ABCD3, ECHP/EHHADH, THIKB/ACAA1B) (Figures 3C and 3D; Table S2 [Significantly depleted proteins in Huwe1^LKO livers]). Together, these findings demonstrate that liver-specific deletion of Huwe1 results in profound lipid metabolic reprogramming that involves significant downregulation of de novo FA lipogenesis and peroxisomal FA metabolism, in addition to marked upregulation of mitochondrial FA β-oxidation and oxidative phosphorylation.

Huwe1^LKO leads to lipid metabolic reprogramming at a transcriptional level

Because the loss of Huwe1 should result in stabilization of its substrates, we reasoned that this could not explain the drastic downregulation of the entire FA synthesis pathway in 1-year-old Huwe1^LKO livers. We then asked whether the extensive metabolic rewiring observed by our mass spectrometry analysis on Huwe1^LKO livers was the result of a deregulated transcription factor in the absence of Huwe1. To this end, we interrogated our list of 127 differentially expressed proteins identified by mass spectrometry analysis and performed chromatin immunoprecipitation (ChIP) enrichment analysis (ChEA) using Enrichr.²⁰ ChEA is a powerful technique which leverages previously published chromatin immunoprecipitation (ChIP sequencing, ChIP-ChIP, ChIP-PET; ChIP-paired-end-ditag) and DNA adenine methyltransferase identification (DamID)²¹ genome-wide mapping studies to define associations between transcription factors and their target genes.²² ChEA then uses these annotations to infer enrichment of transcription factor activity for a gene set curated by the user.²² Strikingly, ChEA analysis revealed a strong enrichment of PPARɑ, RXR, and LXR as predicted upstream regulators of differentially regulated genes in Huwe1^LKO livers (Figure 3F; Table S3 [ChEA enrichment analysis on differentially expressed proteins in Huwe1^LKO livers]). These three transcription factors belong to a common family of ligand-activated nuclear receptors that collectively serve as master regulators of hepatic lipid homeostasis. Moreover, it is known that the PPARɑ and LXR receptors form obligate heterodimers with RXR receptors, highlighting their intertwining activities.^23,24

We evaluated the regulatory network predicted by ChEA analysis and found 21 genes to be commonly regulated by LXR, RXR, and PPARɑ, 10 genes regulated by LXR and RXR, and 9 genes regulated by PPARɑ and RXR (Figure 3G). Many of the dysregulated lipid metabolism genes previously identified by our mass spectrometry analysis were recapitulated in these three gene lists and were commonly regulated by all 3 transcription factors (Figure 3G; Table S3 [ChEA enrichment analysis on differentially expressed proteins in Huwe1^LKO livers]). Furthermore, many of these dysregulated lipid metabolism genes were specifically involved in FA synthesis and peroxisomal FA β-oxidation (Figure 3G; Table S3 [ChEA enrichment analysis on differentially expressed proteins in Huwe1^LKO livers]). Although Huwe1^LKO did not affect protein levels of PPARɑ and LXR, we observed a subtle decrease in RXRα expression in mice as young as 2-months-old (Figure 3H). RXRα expression remains modestly diminished in 1-year-old Huwe1^LKO mouse livers (Figure 3H).

To get a better sense of the transcriptional landscape of 1-year-old Huwe1^LKO livers, we next performed bulk RNA-seq on livers collected from 1-year-old mice. We identified 524 upregulated and 618 downregulated genes in Huwe1^LKO livers with a |log2FC| > 0.5 and a Benjamini-Hochberg corrected p value (p.adj) of q < 0.05 (Figures 4A and 4B; Table S4 [Significantly enriched mRNA transcripts in Huwe1^LKO livers]; Table S5 [Significantly depleted mRNA transcripts in Huwe1^LKO livers]). As with our proteomics analysis, we observed several of the top downregulated pathways in Huwe1^LKO livers were again related to lipid metabolism (Figure 4C). Specifically, we observed several of the lipogenesis genes identified to be depleted in Huwe1^LKO livers by mass spectrometry were also downregulated at the mRNA level (Aacs, Acaca, Acacb, Acly, Acss2, Fasn, Me1) (Figure 4D; Table S5 [Significantly depleted mRNA transcripts in Huwe1^LKO livers]). Unlike what we saw with the mass spectrometry analysis of Huwe1^LKO livers, we surprisingly observed no change in mRNA levels of FA β-oxidation genes, lipid transfer genes, or bile synthesis genes (Figures S2A and S2B [Huwe1^LKO livers from 1-year-old mice do not exhibit enrichment of lipid metabolic genes and pathways] and Table S4 [Significantly enriched mRNA transcripts in Huwe1^LKO livers]). Notably, although Huwe1 is an E3 ligase for p53, hepatic depletion of Huwe1 did not result in a significant induction of p53 target genes (Figure S2C [Huwe1^LKO livers from 1-year-old mice do not exhibit enrichment of lipid metabolic genes and pathways]).

Figure 4.

RNA-seq analysis reveals drastic metabolic remodeling in Huwe1^LKO livers is recapitulated at the transcript level

(A) Heatmap displaying differentially expressed genes between 1-year-old Huwe1^WT and Huwe1^LKO livers identified by RNA-seq.

(B) Volcano plot showing significantly upregulated (Red) and downregulated (Blue) genes in Huwe1^LKO livers. Significance cutoffs set at |log2FC| > 0.5, padj <0.05.

(C) Hallmark gene set enrichment analysis performed on 618 significantly downregulated genes in Huwe1^LKO livers. As observed in the mass spectrometry analysis (Figure 3C), “HALLMARK_FATTY_ACID_METABOLISM” was also the top downregulated gene set from Huwe1^LKO livers identified by RNA-seq.

(D) Bar graph showing the top 25 depleted genes from Huwe1^LKO livers sorted by -log2FC. Notably, many of these proteins were related to lipid metabolism or associated pathways (highlighted in red boxes).

(E and F) ChEA analysis on all 1,142 differentially expressed genes in Huwe1^LKO livers identified by RNA-seq revealed strong enrichment of RXR, PPARα, and LXR activity (E). As observed in our mass spectrometry analysis, many of the differentially expressed genes in Huwe1^LKO livers are predicted to be regulated by RXR, PPARα, and/or LXR (F). Upregulated (Green) and downregulated (Red) genes related to fatty acid metabolism are highlighted.

Consistent with our mass spectrometry results, our ChEA analysis on the 1,142 differentially expressed genes identified by RNA-seq showed significant enrichment for LXR, RXR, and PPARɑ activity in Huwe1^LKO livers (Figures 4E and 4F; Table S6 [ChEA enrichment analysis on differentially expressed mRNA transcripts in Huwe1^LKO livers]). We identified 76 genes to be commonly regulated by LXR, RXR, and PPARɑ, 60 genes regulated by LXR and RXR, and 67 genes regulated by PPARɑ-RXR (Figure 4F). In concordance with the ChEA analysis performed on the mass spectrometry samples, we detected a striking downregulation of lipid metabolism genes in these gene sets, again with the highest degree of enrichment seen in genes commonly regulated by LXR, RXR, and PPARɑ (Figure 4F; Table S6 [ChEA enrichment analysis on differentially expressed mRNA transcripts in Huwe1^LKO livers]). These results suggest that metabolic reprogramming may occur in Huwe1^LKO livers due to attenuated LXR/RXR/PPARɑ-dependent transcription, as well as LXR/RXR/PPARɑ-independent post-translational stabilization of Huwe1 substrates.

Liver-specific Huwe1 knockout protects against diet-induced NAFLD but not obesity in 24-week-old mice

We next sought to determine whether the robust differential transcriptional and post-translational lipid metabolic gene signatures observed in Huwe1^LKO livers could also confer protection from diet-induced hepatic steatosis. To this end, we challenged 4-week-old mice with HFD (45% kcal fat) or low-fat diet (LFD, 10% kcal fat) ad libitum for 20 weeks (Figure 5A). Although HFD significantly promoted increased weight gain in both genotypes as compared to respective mice on LFD, no differences in body weight were seen between genotypes within each diet group (Figure 5B). A 5% reduction in body weight was observed in Huwe1^LKO mice on HFD relative to Huwe1^WT mice on HFD at the end of the study. However, the differences did not reach statistical significance (p = 0.19; Figure 5B). As seen with the mice raised on chow diet (Figure 1F), LFD- and HFD-fed mice did not exhibit differences in diet consumption between genotypes in both diet groups (Figure 5C). Notably, while HFD induced an increase in the size of subcutaneous inguinal white adipose tissue (iWAT) in Huwe1^WT mice, this diet effect was attenuated in Huwe1^LKO mice (Figure 5D). However, no difference in visceral epidydimal white adipose tissue (eWAT) was observed (Figure 5E). While HFD induced an increased mass of intrascapular brown adipose tissue (iBAT) in Huwe1^WT mice, Huwe1^LKO mice did not exhibit this effect (Figure 5F).

Figure 5.

Young Huwe1^LKO mice are protected from HFD-induced hepatic steatosis but not body weight gain

(A) 4-week-old Huwe1^WT and Huwe1^LKO were challenged with LFD or HFD for 20 weeks. N = 10–16 mice per group.

(B and C) Body weight gain (B) and diet consumption (C) of the mice over the course of the diet study. Huwe1^WT and Huwe1^LKO exhibited no difference in diet consumed and both gained a nearly equal amount of weight following HFD. Data are shown as mean ± SEM. See STAR Methods section for further detail of growth curve analysis.

(D–F) Weight of iWAT (D), eWAT (E), and iBAT (F) from the mice following 20 weeks of LFD or HFD.

(G) Representative pictures of Huwe1^WT and Huwe1^LKO mice challenged with HFD. Scale bar represents 1 cm.

(H) Liver weight normalized to body weight.

(I) H&E and ORO stains showed reduced lipid droplet accumulation in Huwe1^LKO livers following HFD as compared to Huwe1^WT livers.

(J) Biochemical analysis measuring hepatic triglyceride levels. Data shown as mean +SD. Significance assessed by two-sided Student’s t test with significance set at ∗ <0.05, ∗∗ <0.005.

Although these 24-week-old Huwe1^LKO mice were not protected from diet-induced obesity, liver-specific Huwe1 deletion still conferred resistance against the development of diet-induced hepatic steatosis. HFD-fed Huwe1^LKO livers were visually less mottled and lipid laden (Figure 5G). Indeed, Huwe1^LKO mice were protected from HFD-induced increase in liver size, though no difference was observed in liver size between mice fed with LFD (Figure 5H). H&E and ORO stains confirmed less lipid accumulation in the livers of HFD-fed Huwe1^LKO mice compared to those of HFD-fed Huwe1^WT mice (Figure 5I). Moreover, this result corresponded with significantly reduced hepatic triglyceride content in HFD-fed Huwe1^LKO mice compared to HFD-fed Huwe1^WT mice (Figure 5J). These findings suggest that young Huwe1^LKO mice are protected from diet-induced hepatic steatosis, but not from total weight gain.

Liver-specific Huwe1 knockout protects against diet-induced NAFLD and diabetes in 1-year-old mice

It is well established that aging is a prominent contributing factor to the development of NAFLD and diet-induced hepatic steatosis.^4,25 Therefore, we sought to determine whether the protective phenotype observed in HFD-fed 24-week-old Huwe1^LKO mice (Figure 5A) could be better recapitulated in 1-year-old mice, the setting where NAFLD prevalence peaks in humans.¹⁸ To this end, we challenged 24-week-old mice with LFD or HFD until they reached 1 year of age (Figure 6A). As seen in our diet study in 24-week-old mice, HFD promoted body weight gain in both genotypes relative to LFD-fed mice (Figure 6B). No differences were seen in the diet consumed between genotypes in both diet groups (Figure 6C). Additionally, a modest but statistically significant 6.5% reduction in body weight was observed in HFD-fed Huwe1^LKO mice relative to Huwe1^WT mice (p = 0.01; Figure 6B). Body composition analysis by EchoMRI revealed a significant increase in body fat percentage in Huwe1^WT mice following HFD that was not observed in Huwe1^LKO mice (Figure 6D). Conversely, this observation coincided with a significant decrease in lean mass percentage in HFD-fed Huwe1^WT mice and no change in HFD-fed Huwe1^LKO mice (Figure 6D). Although Huwe1^WT mice exhibited marginally higher body fat percentage than Huwe1^LKO mice before LFD/HFD administration (Figure 6E), no significant difference in body fat or lean mass percentage was observed between both 1-year-old LFD cohorts (Figure 6D). As seen in the 24-week-old mice (Figure 5D), HFD promoted the enlargement of subcutaneous iWAT fat pads in 1-year-old mice of both genotypes (Figure 6F). Notably, HFD-fed Huwe1^LKO mice also exhibited a significant reduction in iWAT size relative to HFD-fed Huwe1^WT mice, indicating protection against diet-induced obesity (Figure 6F). Again, no difference in the eWAT was observed between groups (Figure 6G). As seen in the 24-week-old mice, HFD induced the increased mass of iBAT in Huwe1^WT mice but not Huwe1^LKO mice (Figure 6H). These results demonstrate that the reduction in body weight in HFD-fed Huwe1^LKO mice may be in part attributable to reduced subcutaneous fat accumulation but not increased thermogenesis.

Figure 6.

Aged Huwe1^LKO mice are protected against HFD-induced obesity

(A) 24-week-old Huwe1^WT and Huwe1^LKO raised on chow diet were challenged with LFD or HFD for 28 weeks until mice reached 1 year of age. N = 8–15 mice per diet group.

(B and C) Body weight gain (B) and diet consumption (C) of the mice over the course of the diet study. Despite no differences in diet consumption, HFD-fed Huwe1^LKO mice exhibited a modest reduction in body weight accumulation as compared to HFD-fed Huwe1^WT mice. Data are shown as mean ± SEM. See STAR Methods section for further detail of growth curve analysis.

(D and E) Body composition was measured using EchoMRI on 1-year-old mice after LFD/HFD challenge (D) as well as at 24 weeks before randomization (E).

(F–H) Weight of iWAT (F), eWAT (G), and iBAT (H) from the mice following 28 weeks of LFD or HFD. Unless otherwise specified, the data are shown in this figure as mean +SD with significance assessed by two-sided Student’s t test with significance set at ∗ <0.05, ∗∗ <0.005.

We observed drastic differences in liver morphology in Huwe1^LKO mice relative to Huwe1^WT mice, with Huwe1^LKO mice harboring noticeably healthier livers with a less mottled appearance (Figure 7A). Importantly, Huwe1^LKO livers did not increase in size following HFD challenge and remained significantly smaller than HFD-fed Huwe1^WT livers (Figure 7B). In addition, Huwe1^LKO mice exhibit reduced levels of lipid droplets as compared to respective Huwe1^WT diet groups (Figure 7C). Huwe1^LKO livers also did not exhibit an increase in hepatic triglyceride content following HFD challenge, which remained significantly lower relative to HFD-fed Huwe1^WT mice (Figure 7D). As expected, fecal triglyceride levels increased when the mice were on HFD (Figure 7E). However, there was no significant difference in the fecal triglyceride levels between Huwe1^WT and HFD-fed Huwe1^LKO mice (Figure 7E), indicating that reduced hepatic triglyceride content in Huwe1^LKO mice is unlikely due to malabsorption of dietary fat.

Figure 7.

Aged Huwe1^LKO mice are protected against HFD-induced NAFLD

(A) Representative images of livers collected from 1-year-old Huwe1^WT and Huwe1^LKO mice maintained on LFD/HFD.

(B) Liver weight normalized to body weight.

(C) H&E and ORO stains showed reduced lipid droplet accumulation in Huwe1^LKO livers following HFD as compared to Huwe1^WT livers.

(D and E) Biochemical analysis measuring hepatic (D) and fecal (E) triglyceride levels.

(F–J) RT-qPCR analysis on liver samples reveals attenuated expression of Cd36, Fasn, and pro-inflammatory cytokines in Huwe1^LKO livers following HFD as compared to Huwe1^WT.

(K–L) IHC staining for F4/80 shows reduced macrophage infiltration in Huwe1^LKO mice challenged with HFD as compared to Huwe1^WT mice (K). Box and whisker plot depicting quantification of %F4/80+ area (L). Data shown as mean +SD (B, D, andE) or SEM (F–J). Significance assessed by two-sided Student’s t test with significance set at ∗ <0.05, ∗∗ <0.005, ∗∗∗ <0.0005.

Importantly, compared to LFD, HFD significantly induced Cd36 expression, while downregulating Fasn expression in Huwe1^WT livers (Figures 7F and 7G). However, this effect was not observed in Huwe1^LKO livers (Figures 7F and 7G). In fact, the expression levels of both Cd36 and Fasn remained attenuated in Huwe1^LKO livers, regardless of the diets (Figures 7F and 7G). Advanced NAFLD is often associated with inflammation, cirrhosis, and fibrosis. We found that the expression of three pro-inflammatory immune cell markers (Tnfa, Cxcl10, and Cd68) was robustly increased in livers from HFD-fed Huwe1^WT mice, compared to LFD-fed Huwe1^WT mice (Figures 7H–7J). In contrast, this HFD-mediatedinduction of the immune cell markers was mitigated in Huwe^LKO mice (Figures 7H–7J). Likewise, livers from HFD-fed Huwe1^WT mice showed more F4/80 staining-positive macrophages than livers from HFD-fed Huwe1^LKO mice (Figures 7K–7L). Sirius Red staining also revealed mild fibrosis in LFD-fed Huwe1^WT mice which was largely absent in LFD-fed Huwe^LKO mice (Figure S3 [Huwe1^LKO livers exhibit reduced macrophage infiltration and fibrosis]). Notably, HFD-fed Huwe^LKO mice were also protected against the HFD-induced fibrosis observed in Huwe1^WT mice (Figure S3 [Huwe1^LKO livers exhibit reduced macrophage infiltration and fibrosis]). Taken together, these results indicate that Huwe1^LKO not only safeguards mice from age- and diet-induced NAFLD but also confers a mild degree of protection from inflammation and fibrosis.

NAFLD is often viewed as the hepatic manifestation of metabolic syndrome, a collection of disorders that include obesity, T2DM, dyslipidemia, cardiovascular disease, and insulin resistance.²⁶ As NAFLD almost universally presents with insulin resistance,²⁷ studies have shown that NAFLD often predates the development of T2DM.²⁸ Furthermore, obesity and T2DM have been proposed as major risk factors for progression of NAFLD to HCC.^29,30 Since a central metabolic function of the liver is regulating glucose availability,³¹ we next investigated whether liver-specific Huwe1 inactivation affected systemic glucose handling. At 24 weeks of age (prior to LFD/HFD diet administration) no significant differences were seen in the fasting blood glucose levels between groups (Figure 8A). After 28 weeks on LFD/HFD (Figure 6A), HFD 1-year-old Huwe1^WT mice exhibited a significant increase in fasting blood glucose levels compared to LFD Huwe1^WT mice while no change was observed in Huwe1^LKO mice (Figure 8B). However, fasting glucose levels in 1-year-old HFD-fed Huwe1^LKO mice were only modestly reduced as compared to HFD-fed Huwe1^WT mice, with borderline statistical significance (p = 0.087). We then assessed the glucose clearing capability of these mice by performing glucose tolerance tests. Notably, Huwe1^LKO mice exhibited improved glucose clearance compared to Huwe1^WT mice even prior to LFD/HFD administration at 24-week age (Figure 8C). As diet- and age-induced NAFLD progressed in these mice, HFD-fed Huwe1^WT mice demonstrated worsened glucose clearance kinetics as compared to LFD-fed Huwe1^WT mice at 1 year of age (Figure 8D). Importantly, Huwe1^LKO mice showed marked resistance to developing diet-induced glucose intolerance as HFD-fed Huwe1^LKO mice demonstrated no change in glucose clearing capability (Figure 8D).

Figure 8.

Aged Huwe1^LKO mice are protected against HFD-induced glucose intolerance

(A) Fasting blood glucose levels of 24-week-old Huwe1^WT and Huwe1^LKO mice raised on chow diet before enrollment into LFD/HFD-challenge arms.

(B) Fasting blood glucose levels of 1-year-old Huwe1^WT and Huwe1^LKO mice raised on LFD/HFD. Data shown as mean +SD. Statistical significance measured using one-way ANOVA (A) or two-sided Student’s t test with significance set at ∗ <0.05 (B).

(C and D) Glucose tolerance test run on fasted 24-week-old Huwe1^WT and Huwe1^LKO mice raised on chow diet (C) and 1-year-old LFD/HFD-challenged mice (D).

(E) Aged Huwe1^LKO mice exhibit reduced incidence of spontaneous HCC following HFD challenge as compared to Huwe1^WT mice.

(F) Our proposed model of how liver-specific Huwe1 inactivation protects against hepatic steatosis and glucose intolerance induced by aging and/or HFD. Unless otherwise specific, the data in this figure are shown as mean +SD with statistical significance assessed by two-sided Student’s t test with significance set at ∗ <0.05, ∗∗ <0.005, ∗∗∗ <0.0005.

Lastly, we observed that spontaneous liver tumors developed in 2 out of 12 HFD-fed 1-year-old Huwe1^WT mice but not in Huwe1^LKO mice (Figure 8E). No tumors were observed in either of the 1-year-old LFD groups (Figure 8E). Additionally, no tumors were observed in any of the young LFD/HFD-fed mice (Figure 5A), indicating that aging plays a key role in the etiology of these lesions. These findings collectively demonstrate that 1-year-old Huwe1^LKO mice remain markedly protected against HFD-induced obesity, hepatic steatosis, glucose intolerance, and incidence of HCC (Figure 8F).

Discussion

Hepatic steatosis results from an imbalance of lipid homeostatic processes such as FA acquisition through extracellular uptake or de novo synthesis, elimination through β-oxidation or storage in triglycerides, as well as secretion into the circulation in the form of lipoproteins.³² In the current study, we demonstrated that liver-specific deletion of Huwe1 resulted in significant remodeling of several lipid metabolism pathways, including attenuation of de novo lipogenesis, FA uptake, and peroxisomal FA β-oxidation, as well as increased mitochondrial FA β-oxidation and oxidative phosphorylation (Figures 3C–3E, 4C, and 4D). Notably, downregulation of FA synthesis genes in Huwe1^LKO livers was concordant at both protein and mRNA levels (Figures 3D and 4D), suggesting that this alteration likely originates at the transcriptional level. Indeed, ChEA analysis revealed that the dysregulation of lipid metabolism genes in 1-year-old Huwe1^LKO livers was, at least in part, secondary to the attenuated activity of the 3 major hepatic transcription factors: LXR, RXR, and PPARɑ (Figures 3F and 4E). Interestingly, modestly diminished protein expression of RXRɑ, the obligate dimeric partner of LXR and PPARɑ, was seen in 1-year-old Huwe1^LKO livers (Figure 3H). Further study will be required to directly implicate the relationship between these specific transcription factors and their unique subset of targets following inactivation of Huwe1. In contrast, upregulation of several components of the mitochondrial β-oxidation pathway in Huwe1^LKO livers was observed strictly at protein levels, suggesting that this alteration is likely a result of protein stabilization (Figures S1A and S1B [Huwe1^LKO livers from 1-year-old mice exhibit enrichment of metabolic proteins and pathways]; Table S1 [Significantly enriched proteins in Huwe1^LKO livers]). Since Huwe1 is an E3 ubiquitin ligase, it is tempting to speculate that at least some of the proteins involved in mitochondrial β-oxidation might be direct substrates of Huwe1, although this needs to be experimentally validated. All in all, our results strongly suggest that the downregulation of de novo lipogenesis and FA uptake together with the upregulation of the mitochondrial β-oxidation pathways underlies the ameliorated lipid accumulation phenotype observed in the Huwe1^LKO mice (Figure 8F).

Previously, studies in human NAFLD and NASH patients have demonstrated that hepatic expression of the FA transporter CD36 increases with the development of NASH and is significantly associated with elevated plasma insulin levels, insulin resistance, and histologic grade of steatosis.^33,34 Additionally, the expression of de novo lipogenesis genes is also known to be essential for the development of hepatic steatosis.³⁵ We identified Cd36 and several lipogenic genes (e.g., Srebf1, Acly, Acaca, Fasn, and Scd) to be among the most downregulated in Huwe1^LKO mouse livers by mass spectrometry and RNA-seq (Table S2 [Significantly depleted proteins in Huwe1^LKO livers]; Table S5 [Significantly depleted mRNA transcripts in Huwe1^LKO livers]; Figures 3D and 4D). Importantly, many of these genes, including Cd36, are regulated by LXR, RXR, and/or PPARɑ,^23,36,37 consistent with our ChEA analysis (Figures 3G and 4E). It should be noted that PPARɑ (as well as RXR and LXR) requires ligand binding (FA derivatives for PPARɑ-RXR and cholesterol derivatives for LXR-RXR) for its full activation.²³ Given the low levels of triglycerides in Huwe1^LKO mouse livers, it is possible that the reduced expression of PPARɑ, RXR, and LXR-target genes may also be attributable in part due to the lack of available ligands to activate these transcription factors. Interestingly, a recent study has shown that Huwe1 promotes the degradation of PPARɑ through the progestin and adipoQ receptor 3 (PAQR3)-mediated pathway,³⁸ which somewhat contradicts our results (Figures 3G and 4E). However, other E3 ligases were also found to co-immunoprecipitate with PPARɑ³⁸, albeit with lower abundance than Huwe1. We hypothesize that PAQR3 recruits a different E3 ligase in Huwe1^LKO livers to facilitate PPARɑ degradation and that this is why Huwe1 inactivation does not affect PPARɑ protein levels in our model.

In addition to preventing aberrant hepatic lipid accumulation, Huwe1^LKO also confers protection against the development of HFD-induced glucose intolerance (Figure 8). We find that Huwe1 inactivation in the liver improves glucose clearance in mice by 24 weeks of age (Figure 8C) and protects against glucose intolerance induced by HFD in 1-year-old mice (Figure 8D). Notably, it was previously shown that pancreatic β-cell-specific Huwe1 knockout resulted in a progressive hyperglycemia and diabetic phenotype due to p53 activation and cell death of islet cells.¹⁵ Therefore, this study further emphasizes that the role of Huwe1 in vivo is often context dependent and may vary between tissues. Huwe1 is a well-established negative regulator of the tumor suppressor p53.⁵ Although most widely studied for its anti-tumorigenic properties, p53 is also well known to exert functions in regulating various aspects of metabolism including lipid homeostasis, glycolysis, and insulin resistance.^39,40,41 In response to perturbations in nutrient availability, ribosomal stress and ribosomal protein-mediated degradation of MDM2, the major negative regulator of p53, have been shown to mediate the activation of p53.^39,42 It was recently demonstrated that HFD induces p53 in a c-MYC- and RPL11-dependent fashion in wild-type mice but not in global knockin mice bearing a single amino acid substitution in MDM2 (C305F).³⁹ The defect in p53 activation observed in C305F HFD-challenged mice was associated with a reduction in hepatic triglyceride content and improved glucose tolerance, analogous to what we observe in our HFD-fed Huwe1^LKO mice.³⁹ Notably, the phenotype observed in C305F mice was attributable to RPL11-MDM2-p53 signaling specifically in adipose tissue but not the liver, since no differences were observed in hepatic levels of RPL11, p53, or the p53 target gene p21 in C305F mice.³⁹ These results suggest that the role of RPL11-MDM2-p53 axis on the maintenance of lipid homeostasis may be tissue-specific and dispensable in the context of liver biology. As such, because no differences in p53 or p53 target genes were observed in Huwe1^LKO mice from the current study (Figure S2C [Huwe1^LKO livers from 1-year-old mice do not exhibit enrichment of lipid metabolic genes and pathways]), the protection from NAFLD observed in Huwe1^LKO mice appears to arise independent of p53 signaling.

While the biggest functional effects of Huwe1 deletion are seen in 1-year-old mice challenged with HFD, aspects of this protective phenotype begin to manifest earlier in life. For example, despite exhibiting no difference in liver size at 24 weeks of age (Figure 2B), Huwe1^LKO mice on chow diet already demonstrate reduced body fat accumulation (Figure 6E) and improved glucose tolerance (Figure 8C). Therefore, the protective effects observed in HFD-fed 1-year-old Huwe1^LKO mice may accumulate over the entire lifespan. Importantly, although hepatic steatosis is reversible in early NAFLD, liver damage resulting from inflammation, cirrhosis, and fibrosis observed in advanced stages are often irreversible and predispose individuals to HCC.^4,29 Outside of diet and exercise, treatment strategies for NAFLD remain limited as there are no FDA-approved pharmaceutical options available. Ultimately, our study identifies Huwe1 as a potentially valuable therapeutic target for NAFLD prevention.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Rabbit polyclonal anti-Huwe1	Bethyl Laboratories	Cat # A300-486A; RRID:AB_2264590
Mouse monoclonal anti-beta-Actin	Santa Cruz Biotechnology	Cat # SC-69879; RRID:AB_1119529
Mouse monoclonal anti-FASN	BD Biosciences	Cat # 610962; RRID:AB_398275
Goat polyclonal anti-CD36	R&D Systems	Cat # AF2519; RRID:AB_2228767
Rabbit polyclonal anti-PPARα	Rockland Immunochemicals	Cat # 600-401-421; RRID:AB_2165756
Rabbit polyclonal anti-LXR	Proteintech	Cat # 14351-1-AP; RRID:AB_10640525
Rabbit polyclonal anti-RXRα	Cell Signaling Technology	Cat # 5388S; RRID:AB_10621998
Rabbit monoclonal anti-HSP90	Cell Signaling Technology	Cat # 4877S; RRID:AB_2233307
Rabbit monoclonal anti- F4/80 antibody	Cell Signaling Technology	Cat # 70076S; RRID:AB_2799771
Chemicals, peptides, and recombinant proteins
Oil Red O	Cayman Chemical	Cat # 14419
D-(+)-glucose	Sigma	Cat #G7528
Chow Diet (ProLab® RMH 3000)	LabDiet	Cat # 5P00
High Fat Diet	Research Diets Inc.	Cat #D12451
Low Fat Diet	Research Diets Inc.	Cat #D12450B
Critical commercial assays
Triglycerides Liquid Stable Reagent	Thermo Scientific	Cat # TR22421
Total Cholesterol and Cholesteryl Ester Fluorometric Assay Kit	BioVision	Cat #K603-100
Sirius Red Stain Kit	Abcam	Cat # ab150681
Deposited data
RNA-seq data	This paper	GSE241875
Mass Spectrometry data	This paper	PXD045279
Experimental models: Organisms/strains
Mouse: Huwe1f//fl	Kon et al.15	N/A
Mouse: B6.Cg-Speer6-ps1Tg(Alb-cre)21Mgn/J	Postic et al.16	RRID:IMSR_JAX:003574
Software and algorithms
GraphPad Prism	GraphPad	https://www.graphpad.com/
R: A Language and Environment for Statistical Computing	R Foundation for Statistical Computing	https://www.r-project.org/
Molecular signatures database	Broad Institute	https://www.gsea-msigdb.org/gsea/msigdb
ImageJ	National Institutes of Health	https://imagej.nih.gov/ij/index.html

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Manabu Kurokawa (mkurokaw@kent.edu).

Materials availability

All unique/stable reagents generated in this study are available from the lead contact with a completed Materials Transfer Agreement.

Data and code availability

Raw and processed RNA-seq data was deposited in the Gene Expression Omnibus (GEO) database and are publicly available under the accession number GSE241875. Mass spectrometry data was deposited in the ProteomeXchange Consortium under project accession number PXD045279.

Experimental model and study participant details

Mice

Huwe1^f//fl mice were previously described.¹⁵ Alb-Cre mice (B6.Cg-Speer6-ps1Tg(Alb-cre)21Mgn/J, Stock No: 003574¹⁶) were obtained from The Jackson Laboratory. All mice used in this study were on C57BL/6 background. In all experiments, nearly equal numbers of male littermates (WT and LKO) were included, and the results were pooled and analyzed. On average, 3–4 mice were housed per cage. Mice were maintained on 5P00 ProLab RMH 3000 (LabDiet) chow diet ad libitum until initiation of diet experiments. LFD (D12450B) and HFD (D12451) were purchased from Research Diets Inc. Body weights and diet weights were measured weekly. Body composition was measured using an EchoMRI-700 quantitative magnetic resonance body composition analyzer (Echo Medical Systems). All animal studies were performed by protocols approved by the Institutional Animal Care and Use Committee (IACUC) at Dartmouth College and Kent State University.

Method details

Glucose tolerance test

Glucose tolerance tests were performed as previously described.⁴³ Briefly, mice were fasted overnight for 16 h before intraperitoneal injection with 2 mg/kg D-glucose (Sigma). Blood glucose was measured 0, 15, 30, 60, 120 min post-injection using an Accu-Chek Aviva glucometer (Roche).

Immunohistochemistry (IHC)

Livers were harvested and fixed in 10% neutral buffered formalin overnight before tissue dehydration and paraffin embedment. Tissues were sectioned in 10 μm slices and then dried at 60°C for 20 min. The tissues were then subsequently stained with Harris Hematoxylin (VWR) and Eosin (VWR) after deparaffinization with Xylene (VWR) and hydration. Slides were mounted using VectaMount AQ aqueous mounting media (Vector Laboratories). Slides were imaged by brightfield microscopy within 12 h of staining. Sirius red staining was performed using the Sirius Red Stain Kit (Abcam), according to the manufacturer’s instruction. F4/80+ quantification shown in Figure 7L was conducted using ImageJ version 1.54d.

ORO staining

Livers were harvested and fixed overnight in 10% neutral buffered formalin. Tissues were rinsed in PBS followed by 48 h incubation in 20% sucrose at room temperature. Tissues were then frozen in optimal cutting temperature (OCT) compound (Tissue-Tek) followed by sectioning into 8 μm slices. Tissue sections were then dried at room temperature, followed by staining in ORO (Cayman Chemical) working solution for 5 min. Slides were then rinsed in deionized water followed by a 10-min incubation with Mayer’s Hematoxylin (Sigma-Aldrich). Slides were washed in deionized water and mounted using VectaMount AQ aqueous mounting media (Vector Laboratories). Slides were imaged by brightfield microscopy within 6 h of staining.

Triglyceride and cholesterol quantification

Livers were homogenized in 500 μL PBS followed by the addition of 700 μL chloroform:isopropanol:NP-40 (7:11:0.1 proportion). Mixtures were vortexed and incubated in a sonicating water bath for 1 h. Eppendorf tubes were vortexed, then spun at 13,000 rpm for 10 min. The lower organic phase containing liver lipid extracts was isolated and dried. Hepatic triglyceride levels were measured using Infinity Triglycerides Liquid Stable Reagent (Thermo Scientific) and total cholesterol levels were measured using Total Cholesterol and Cholesteryl Ester Fluorometric Assay Kit (BioVision) according to manufacturer’s instructions.

RNA-seq

RNA was isolated using the SV Total RNA Isolation System (Promega, cat no. Z3105) in accordance with the manufacturer’s recommendations. Library prep was performed by BGI Genomics (San Jose, CA). RNA-seq was performed using the BGISEQ-500 platform. Clean reads were mapped to a reference genome using Bowtie2, and gene expression was calculated using RSEM. DEseq2 was used to identify differentially expressed genes with significance set at ∗p < 0.05 after Benjamini-Hochberg correction for multiple comparisons. Hallmark gene set enrichment analyses were performed using the molecular signatures database (MSigDB) v7.2.

Real-time quantitative PCR (RT-qPCR)

RT-qPCR was performed as previously described.⁴⁴ RNA was isolated using an SV Total RNA Isolation System (Promega) according to the manufacturer’s instructions. cDNA was synthesized using a qScriptTM cDNA synthesis kit (Quanta). RT-qPCR was performed using iTaq Universal SYBR Green Supermix (Bio-Rad) on a CFX96TM Real-Time Detection System. The data was analyzed using the Delta-Delta Ct Method, and individual gene expression was normalized to the expression of Rplp0. The RT-qPCR primers were purchased from Integrated DNA Technologies. The corresponding sequences are listed in Table S7 [List of primers used for RT-qPCR].

Western blot

Cells were lysed in RIPA buffer supplemented with protease and phosphatase inhibitors (Roche). Lysates were cleared by centrifugation at 20,000g for 20 min at 4°C and 40 μg protein was loaded per lane on 8% polyacrylamide gels. Proteins were transferred onto PVDF membranes (Millipore) and were blocked in 3% BSA for 30 min at room temperature. Membranes were incubated with primary antibodies overnight at 4°C with gentle agitation. Membranes were developed using Pierce ECL (Thermo Scientific) and X-ray film. Band densitometry was conducted using ImageJ version 1.54d.

Sample preparation for proteomics analysis

Hepatic proteins (100 μg) from frozen liver tissues were fractionated with sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) for 10 min at 150 V. One wide gel band corresponding to a mixture of hepatic proteins was cut and processed for proteomic analyses. Extracted gel bands were reduced with dithiothreitol (DTT), alkylated with iodoacetamide and digested with trypsin (Promega, Madison, WI) at 37°C overnight. The tryptic peptides were extracted from polyacrylamide gel bands in a sonication bath with 100 μL of 40% acetonitrile with 0.1% trifluoroacetic acid (TFA) and 100 μL of 70% acetonitrile with 0.1% TFA, consecutively. In-gel digested tryptic peptides were analyzed by nanospray LC-MS/MS after reconstitution in 0.1% formic acid.

LC-MS/MS analysis

A solution of tryptic peptides was analyzed by Ultimate 3000 UHPLC (Thermo Scientific, CA) coupled online to Q Exactive Plus Hybrid Quadrupole-Orbitrap Mass Spectrometer (Thermo Scientific, CA). After desalting samples on an Acclaim PepMap100 precolumn (300 μm × 5 mm, C18, 5 μm, 100 Å, Thermo Fisher Scientific), peptides were separated on an Acclaim PepMap RSLC reverse phase nanocolumn (75 μm × 15 cm, C18, 2 μm, 100 Å, Thermo Fisher Scientific) at 300 nL/min with a mobile phase A (0.1% formic acid in water) and B (20% water in acetonitrile with 0.1% formic acid). A stepwise gradient was employed with an initial 2% of mobile phase B. After 4 min of de-salting, mobile phase B was linearly increased to 35% in 90 min. Mobile phase B was then ramped to 90% in 5 min and then held at 90% B for 10 min. Subsequently, mobile phase B was decreased to 2% for 2 min and equilibrated for 13 min with 2% of phase B.

Mass spectrometry analysis was performed in data-dependent acquisition mode with a full profile MS scans at 70000 resolution (200 m/z) between 380 and 1300 m/z. MS/MS spectra were collected in data-dependent acquisition mode for the 14 most abundant precursor ions with an isolation window of 1.4 and offset of 0.3 m/z and 17,500 resolution (at 200 m/z). Higher-energy collisional dissociation (HCD) was performed at a normalized collision energy of 25%. The precursor ion masses were dynamically excluded from MS/MS analyses for a duration of 20 s. MS and MS/MS spectra were acquired for 100 ms with the automatic gain control (AGC) target set at 1.0 x 106 and 2.0 x 104 ions for MS and MS/MS scans, respectively.

For protein identification, all the spectra obtained from the mass spectrometer were transformed into mzML file format using ProteoWizard MSconvert Version 3.0.18116 (http://proteowizard.sourceforge.net/tools), and the MZML files were searched using Mascot software (Matrix Science, London, UK) version 2.3 against the mouse subset of the UniProt protein database released on June 29th, 2016 (containing 149,730 entries) with cysteine carbamidomethylation as fixed and methionine oxidation as variable modifications and trypsin digestion with a maximum of two missed cleavages per peptide.

To assess HUWE1-induced relative changes in hepatic proteins levels, tryptic peptides from both groups were quantified using MaxQuant software (MaxQuant V1.5.2.8)1. Raw files were searched using MaxQuant-integrated Andromeda search engine (version 1.5.2.8) using the parameters for the Orbitrap instrument as default settings. MaxLFQ algorithm of MaxQuant measures the area under the curve of high intensity paired peptides’ signals from multiple samples in each group. The MaxQuant results were manually validated using Xcalibur software (version 3.1, Thermo). To account for variations during sample preparation, peptides isolated from equal total protein amounts (30 μg) were quantified. Data were Log2 transformed and were normalized relative to the mean of control group. Comparison between groups was then estimated as fold change. Thus, the results of these measurements represent relative levels of proteins in whole tissue lysate and reflect the HUWE1-induced change in expression rather than the absolute concentrations of analyzed proteins.

The results were summarized using means and standard deviations and were compared between groups using two-sample t-tests. Mean differences between groups with 95%-confidence intervals were presented.

Statistical analysis

Two-tailed Student’s t test was used to determine statistical significance between groups using GraphPad Prism v8.3.1. Growth curve analysis was conducted using the “statmod” R package with significance set at ∗p < 0.05 after Benjamini-Hochberg correction for multiple comparison. Analyses were run in R version 3.4.2. Further statistical details of experiments can be found in the figure legends.

Published: November 7, 2023

Supplemental information

Table S1. Significantly enriched proteins in Huwe1LKO livers, related to Figure 3

List of proteins identified by mass spectrometry to be increased in Huwe1^LKO livers with a log2FC > 0.5 and a p value of p < 0.05.

Table S2. Significantly depleted proteins in Huwe1LKO livers, related to Figure 3

List of proteins identified by mass spectrometry to be decreased in Huwe1^LKO livers with a log2FC < −0.5 and a p value of p < 0.05.

Table S3. ChEA enrichment analysis on differentially expressed proteins in Huwe1LKO livers, related to Figure 3

List of transcription factors predicted to regulate expression of differentially expressed proteins identified by mass spectrometry.

Table S4. Significantly enriched mRNA transcripts in Huwe1LKO livers, related to Figure 4

List of mRNA transcripts identified by RNA-seq to be increased in Huwe1^LKO livers with a log2FC > 0.5 and a p value of p < 0.05 after Bonferroni-Hochberg correction.

Table S5. Significantly depleted mRNA transcripts in Huwe1LKO livers, related to Figure 4

List of mRNA transcripts identified by RNA-seq to be decreased in Huwe1^LKO livers with a log2FC < −0.5 and a p value of p < 0.05 after Bonferroni-Hochberg correction.

Table S6. ChEA enrichment analysis on differentially expressed mRNA transcripts in Huwe1LKO livers, related to Figure 4

List of transcription factors predicted to regulate expression of differentially expressed mRNA transcripts identified by RNA-seq.