Loss of Early Growth Response Protein 1 in the Liver Leads to Hepatic Lipid Accumulation Driven by an Imbalance Between Fatty Acid β-Oxidation and Oxidative Phosphorylation

TL;DR

Deleting a protein called EGR1 in mice livers causes fat buildup by disrupting fatty acid breakdown and energy production.

Contribution

Shows EGR1 is critical for balancing fatty acid oxidation and mitochondrial respiration in liver cells.

Findings

EGR1-deficient mice had higher liver triglycerides after a high-fat diet.

EGR1 deficiency reduced maximum oxygen consumption in hepatocytes.

Lipid accumulation in EGR1-deficient mice was linked to disrupted fatty acid oxidation.

Abstract

Metabolic-associated fatty liver disease is a worldwide health problem characterized by increased hepatic lipid accumulation, leading to conditions like steatohepatitis, cirrhosis, and liver cancer. Early growth response protein 1 (EGR1) gene encodes an immediate early transcription factor that regulates a wide variety of cellular processes in response to stress and injury. Whole-body Egr1 knockout studies suggest a role for EGR1 in metabolism, affecting insulin sensitivity, energy homeostasis, cholesterol biosynthesis, and circadian rhythm. However, its direct role in hepatic lipid metabolism remains unclear. This study aimed to investigate the function of EGR1 in the adult liver.

Hepatocyte-specific EGR1-deficient male mice were generated using CRISPR/Cas9. Mice were maintained under chow-fed conditions or challenged with a high-fat diet (HFD). Hepatic lipid levels, zonation, and gene expression were analyzed. Oxygen consumption was measured in mouse and human hepatocytes.

After a HFD challenge, hepatic EGR1-deficient mice showed a significant increase in hepatic triglyceride levels. Transcriptome analysis revealed an upregulation of genes related to fatty acid oxidation and downregulation of mitochondrial respiration genes in livers of both chow and HFD hepatic EGR1-deficient mice. Functional analysis showed reduced maximum oxygen consumption in EGR1-deficient mouse and human hepatocytes. Fasting-induced hepatic lipid accumulation indeed indicated reduced fatty acid oxidation efficiency upon ablation of EGR1.

Hepatic EGR1 deficiency significantly alters lipid metabolism and mitochondrial function, indicating a role of EGR1 in regulating the balance between mitochondrial fatty acid β-oxidation and respiration in the liver.

Figures

C–E). In addition, no changes were observed in liver weight, liver-to-body weight ratio, subcutaneous, and gonadal WAT depots ([Figure 1](#fig1)D, E and Supplementary Figure 1F, G). We observed no changes in plasma triglycerides, nonesterified fatty acid, and APOB levels, and only a small reduction in VLDL-associated triglycerides in plasma of hepatic EGR1-deficient was seen (Supplementary Figure 1H–K). Hepatic cholesterol and triglyceride levels were unchanged in EGR1-deficient livers, and H&E staining showed no major abnormalities at a histological level ([Figure 1](#fig1)F–H). Oil Red O sta

Genes involved in fatty acid oxidation are altered upon hepatic *Egr1* depletion. Using RNA sequencing data from livers, heatmaps were made of genes involved in fatty acid oxidation in hepatocyte-specific *Egr1* knockout and control mice fed a (A) chow or (B) HFD. Heatmaps were constructed based on the z-score calculated from the fragments per kilobase of exon per million mapped reads scores. (C) Schematic representation of the genes altered in various processes involved in fat oxidation in mitochondria and peroxisomes. Created with BioRender.

A–D). In contrast to the observed increase in de novo lipogenesis, no significant effects on mRNA levels of genes involved in this process, such as genes encoding enzymes involved in converting acetyl-CoA into C16:0 (eg, *Fasn* and *Acaca*), desaturation (eg, *Scd1*), and elongation of C16 to C18 (eg, *Elovl6*), were observed in hepatocyte-specific EGR1-deficient mice as compared to controls (Supplementary Figure 4E). These findings indicate an increase in de novo lipogenesis of certain lipids upon EGR1 ablation, potentially explaining the altered zonation of hepatic lipids.

Increased plasma acylcarnitines and hepatic triglycerides accumulation after prolonged fasting upon the loss of hepatic *Egr1*. (A) Total acylcarnitines, (B) acylcarnitine distribution (LC = long chain, MC = mid chain, SC = short chain), (C) acetylcarnitine (C0), and (D) C2 acylcarnitine concentrations in plasma after prolonged fasting (overnight for 16 hours) at room temperature at 15 weeks of age of hepatocyte-specific *Egr1* knockout and control mice fed a chow diet. (E) Plasma C0/(C16 + C18) and (F) C4OH + C3DC acylcarnitines in hepatocyte-specific *Egr1* knockout and control mice. (G) Liv

G). These findings suggest that peroxisomal fatty acid oxidation may partially compensate for mitochondrial dysfunction in EGR1-deficient livers. EGR1 deficiency also resulted in reduced mitochondrial capacity in primary hepatocytes and HepG2 cells but not in isolated mitochondria from livers of hepatic EGR1-deficient mice, suggesting that the deficiency is driven by changes in intracellular processes rather than caused by a specific mitochondrial defect. In HepG2 cells, we found that loss of EGR1 resulted in a shift in metabolic flexibility in conditions of high substrate demand, characterize

Full text

Introduction

The liver plays a central role in lipid and carbohydrate metabolism and contributes to whole-body energy metabolism. An imbalance in energy substrate availability and substrate utilization is the main cause of increased hepatic lipid accumulation (steatosis), which underlies the development of metabolic-associated fatty liver disease (MAFLD).1 MAFLD is characterized by hepatic lipid accumulation of at least 5% of liver weight in combination with overweight/obesity and type 2 diabetes mellitus or signs of metabolic dysregulation.1 The hepatic manifestation consists of a spectrum of conditions ranging from simple steatosis to metabolic-associated steatohepatitis, advanced fibrosis, cirrhosis, and end-stage liver disease. While the disease is relatively benign and reversible in early stages, it can further progress to serious nonreversible stages, including cirrhosis and hepatocellular carcinoma.2 While genome-wide association studies have identified several genes (eg, PNPLA3, TM6SF2, and GCKR) that increase the risk of hepatic lipid accumulation, a large part of the genetic basis of MAFLD remains unknown.3 As treatment options are currently limited, a better understanding of genes involved in hepatic lipid accumulation is essential for developing novel drugs to combat MAFLD.

Early growth response protein 1 (EGR1) is an immediate early transcription factor that belongs to the EGR family of DNA-binding proteins.4 EGR1 (also called NGFI-A, CEF5, TIS8, Krox-24, and Zif268) is rapidly and transiently induced within minutes after exposure of cells to a range of external stimuli, including cytokines, nutrients, growth factors, and environmental stressors such as irradiation, hypoxia, and mechanical or shear stress.5, 6, 7, 8, 9 The EGR1 protein has a short half-life and regulates its own expression via a transcriptional feedback mechanism. In addition, the activity of EGR1 is regulated by post-translational modifications.10, 11, 12 EGR1 binds to conserved guanine/cytosine-rich consensus nucleotide sequences and can activate or repress transcription of its target genes.13 As a transcription factor, EGR1 can regulate numerous cellular transcriptional programs, including those involved in cell growth, proliferation, inflammation, and cell cycle entry.

While EGR1 is required for normal development and cellular responses, it also regulates multiple aspects of metabolism. Transgenic overexpression of Egr1 in adipose tissue is associated with obesity and related metabolic dysfunction in mouse models, while whole-body Egr1-deficient mice are protected from diet-induced obesity, likely because of its role in white adipose tissue (WAT) lipolysis.14, 15, 16, 17 Additional studies using whole-body Egr1 knockout mice have also demonstrated a protective role of Egr1 deletion in hepatic fat accumulation in various contexts, such as ethanol-induced steatosis, acetaminophen-induced liver injury, and in db/db mice, a severe genetic model for diabetes.18, 19, 20, 21, 22, 23 These studies suggest that the observed effects are likely due to reduced lipolysis in adipose tissue. However, a potential physiological function of EGR1 in the liver, particularly its role in diet-induced steatosis, obesity, and associated disorders, remains unclear. In the present study, we aimed to delineate the liver-specific role of EGR1 in hepatic energy metabolism using a mouse model with hepatocyte-specific Egr1 deficiency.

Methods

Animals

Hepatocyte-specific Cas9 expressing mice were generated using Rosa26-LSL-Cas9 knock-in mice (#026175, The Jackson Laboratory) crossed with Albumin-Cre mice (#003574, The Jackson Laboratory) as previously described.18 Gene editing was performed using self-complementary adeno-associated virus particles, which were produced as previously described (see Supplemental Methods).18 Egr1 was targeted using CRISPR/Cas9 gene editing, and mice received 1 × 10^11^ vector genomes of adeno-associated virus by a retro-orbital injection. Male mice, aged 8–12 weeks at the time of gene editing, were used for all experiments and were individually housed under a 12-hour light-dark cycle with ad libitum access to water and a standard chow diet (10% fat, 23% protein, and 67% carbohydrates; V1554-703, Ssniff Spezialdiäten GmbH) or a high-fat diet (HFD; 60% fat, 20% protein, and 20% carbohydrates, D12492i, ResearchDiets). Only male mice were used in this study as they are more susceptible to developing diet-induced hepatic steatosis. Animal experiments were performed with the approval of the National Ethics Committee for Animal Experiments of the Netherlands, in accordance with relevant guidelines and regulations (including laboratory and biosafety regulations).

Animal Studies

Body weight and food intake of individually housed male mice were measured on a weekly basis. Measurements of body composition, including fat mass and lean tissue mass, were assessed at several time points (ie, every 4 weeks on chow or HFD and before indirect calorimetry measurements) in nonanesthetized mice using nuclear magnetic resonance MiniSpec (MiniSpec LF90 BCA-analyzer, Bruker). Blood was collected via retro-orbital bleeding under isofluorane anesthesia in citrate-ethylenediaminetetraacetic acid (EDTA) tubes, and plasma was isolated by centrifugation at 1000 × g for 10 minutes at 4 °C and stored at −80 °C until further analysis. Mice were sacrificed under isofluorane anesthesia after a 4- to 16-hour fasting period at temperatures as indicated after which blood was drawn by cardiac puncture followed by cervical dislocation. Fat pads, liver, and quadriceps were weighed, and tissues for mRNA and protein expression analysis were snap-frozen in liquid nitrogen and stored at −80 °C until further analysis.

Histological Analyses

Liver tissues were fixed in 4% (w/v) phosphate-buffered formalin, embedded in paraffin, and sectioned at 4 μm. Sections of liver tissue were stained with hematoxylin and eosin (H&E), glutamine synthetase (Abcam, #ab176562), and arginase 1 (Cell Signaling, #93668S) according to standard protocols. Frozen liver tissues were cut into 5 μm sections and used for Oil Red O staining (Sigma Aldrich, #00625). Slides were scanned using a Hamamatsu slide scanner, and snapshots were taken using ImageJ.19 All histological sections were evaluated and scored by an animal pathologist (R.E.T.). Oil Red O quantification was performed using QuPath, incorporating a modified Stardist extension to enhance lipid droplet detection.20 The analysis focused on the pericentral and periportal regions and was defined as 2–3 hepatocyte layers surrounding the corresponding vein.

Hepatic Lipid Analysis

Lipid extraction was performed on liver homogenates (15% w/v in phosphate buffered saline) following the method described by Bligh and Dyer.21

Glucose and Insulin Tolerance Tests

Glucose homeostasis was evaluated using glucose and insulin tolerance tests. For both tests, mice were fasted for 6 hours after the dark phase. During the glucose tolerance test, mice received an intraperitoneal injection of 2 g/kg dextrose (Sigma-Aldrich, #G6152). For the insulin tolerance test, mice were injected with insulin (NovoRapid; 0.75 U/kg on chow, 1.0 U/kg on HFD), and glucose levels were measured at the same time points. Blood glucose was measured at indicated time points postinjection.

Plasma Lipid Analysis

Plasma levels of total cholesterol were determined using a colorimetric assay (Roche, #11489232) with a Cholesterol Standard (Diasys Diagnostic Systems GmbH, #113009910030). Plasma triglycerides were measured using a colorimetric assay (Roche, #1187771) using a Precimat glycerol standard (Roche, #16658800). Nonesterified fatty acids were measured by colorimetric assay (Roche) using oleic acid as a standard (included in the kit). All measurements were performed as described in the manufacturer’s protocol.

Fast Performance Liquid Chromatography

Total cholesterol content of the major lipoprotein classes (very-low-density lipoprotein [VLDL], low-density lipoprotein [LDL], and high-density lipoprotein [HDL]) was measured using fast performance liquid chromatography analysis. In short, the system consisted of a PU-4180 RHPLC pump and a UV-4075 UV-Vis detector (Jasco). Plasma samples of each experimental group of mice were pooled, diluted in PBS, and loaded onto a Superose Increase 10/300 GL column (GE HealthCare) for separation of lipoproteins at a flow rate of 0.31 mL/min. A second flow mixed in the column (CO4060) with the coil at 37 °C was used to add cholesterol (Roche, #1489232) enzymatic reagents at a flow rate of 0.10 mL/min.

Lipogenesis, Fractional Cholesterol Synthesis, and Bile Flow Measurements

To assess de novo lipogenesis and cholesterol synthesis by mass isotopomer distribution analysis, mice received drinking water containing 2% [1-^13^C] acetate and bloodspots were collected at 0, 10, 24, 34, 48, 58, 72, 82, 96, 106, and 120 hours. De novo lipogenesis, fatty acyl chain elongation, and cholesterol synthesis measurements were determined as previously described.22^,^23 After the final time point, mice were anesthetized by intraperitoneal injection of hypnorm (1 mL/kg fentanyl-fluanisone) and diazepam (10 mg/kg) followed by gallbladder cannulation as described previously.24 Bile was collected for 30 minutes after which mice were terminated by cardiac puncture under isoflurane anesthesia followed by cervical dislocation. Bile acids in plasma, feces, and bile were analyzed by ultrahigh-performance liquid chromatography tandem mass spectrometry, as described previously.25

RNA Sequencing

For RNA sequencing analysis, total RNA was first analyzed for quality using a NanoDrop. Library preparation, sequencing, and basic analysis were done by Novogene Co Ltd, Europe. Differential expression analysis between 2 groups (n = 6 per group) was performed using the DESeq2 R package. The resulting P values were adjusted using the false discovery rate.26 A false discovery rate of 10% was used to produce the differentially expressed genes list. Heatmaps were constructed based on the z-score calculated from the fragments per kilobase of exon per million mapped reads values. Gene set enrichment analysis was performed to determine pathway enrichment.

Acylcarnitine Profiling

Acylcarnitine profiling was done using 10 μL plasma or liver homogenate with the addition of 100 μL acetonitrile and 100 μL of internal standard (L-carnitine-[D3], acetyl-L-carnitine-[D3], propionyl-L-carnitine-[D3], octanoyl-L-carnitine-[D3], and palmitoyl-L-carnitine-[D3]). Samples were centrifuged at 12,000 × g for 10 minutes, collected into glass vials, and analyzed by liquid chromatography tandem mass spectrometry as previously described.27

Indirect Calorimetry

Indirect calorimetry was performed using fully automated metabolic cages (LabMaster, TSE Systems). To measure core body temperature, G2 E-mitters (Starr Life Sciences Corp) were implanted under the abdominal muscle 8 days before the indirect calorimetry experiment under isofluorane anesthesia. After at least 24 hours of acclimatization, O_2_ consumption (VO_2_), CO_2_ production (VCO_2_), and core body temperature were measured for 3 consecutive days at 22 °C, 3 days at 29 °C, and 3 days at 4 °C. The energy expenditure ([3.941 × VO_2_] + 1.106 × VCO_2_ × 1.44) was calculated from the VO_2_ and VCO_2_. Values are corrected for lean mass of each individual mouse.

Isolation of Hepatic Mitochondria and Oxygen Consumption Rates

Mitochondria were isolated from fresh liver tissues as previously described.28 A brief description of the method can be found in Supplementary Methods.

Quantitative Real-Time PCR Analysis

Frozen liver tissue was homogenized in QIAzol Lysis Reagent (Qiagen), and total RNA was isolated using the RNAeasy mini kit (Qiagen, #74104). Two micrograms of total RNA was used for cDNA synthesis according to the manufacturer protocol (Invitrogen, #28025013). Twenty nanograms of cDNA was used for quantitative real-time polymerase chain reaction (qRT-PCR) analysis using SYBR Green Master Mix (Roche) and qRT-PCR primers. qRT-PCR data were analyzed using QuantStudio Real-Time PCR software, and expression was normalized to housekeeping genes and expressed as fold change compared to vehicle-treated mice.

Immunoblotting

Tissue and cell homogenates were obtained using NP40 buffer (0.1% Nonidet P-40, 0.4 M NaCl, 10 mM Tris-HCl [pH 8.0], 1 mM EDTA) supplemented with protease and phosphatase inhibitors (Roche). Protein concentration was determined using the Bradford assay (Bio-Rad). A total of 15–20 μg of protein per sample was separated using sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred to a polyvinylidene fluoride transfer membrane (Amersham Hybond – P, GE HealthCare; RPN303F). Membranes were blocked in 5% bovine serum albumin in Tris-buffered saline with 0.01% Tween 20 (Millipore Sigma) and incubated with the indicated antibodies. Proteins were visualized using a ChemiDoc XRS + System using Image Lab software (version 5.2.1; Bio-Rad). The antibodies used can be found in Supplemental Methods.

Cell Culture

Primary hepatocytes were isolated from mice after 4–8 weeks after gene editing. Hepatocytes were isolated by liver perfusion, EDTA dissociation, and separation from nonparenchymal cells using Percoll gradient as previously described.29 Isolated primary hepatocytes were seeded in Dulbecco’s Modified Eagle Medium + GlutaMAX supplemented with 10% heat-inactivated fetal calf serum and 1% antibiotics mixture in collegen-coated plates. For in d studies, human hepatocyte cell line HepG2 was used. HepG2 cells were maintained in high glucose Dulbecco’s Modified Eagle Medium + GlutaMAX supplemented with 10% heat-inactivated fetal calf serum, 100 U/mL penicillin, and 100 μg/mL streptomycin. The shRNA-mediated knockdown of human EGR1 is described in Supplemental Methods.

Seahorse XFe96 Metabolic Flux Analysis

Cellular oxygen consumption was analyzed using an XFe96 Seahorse Analyzer. A total of 15.000 HepG2 cells or 8.000 primary hepatocyte cells were plated per well in a 96-well Seahorse plate (V3-PS101085) and maintained in their respective culture medium. Fatty acid oxidation experiments and substrate inhibitor assays were performed according to manufacturer’s protocol (Agilent Seahorse XF Palmitate-BSA FAO Substrate Quickstart Guide and Agilent XF Substrate Oxidation Stress Test kits, respectively). A brief description can be found in Supplemental Methods.

Statistical Analysis

Statistical analysis was performed using GraphPad Prism (version 9.0) software package (GraphPad Software, San Diego, CA). For comparisons between 2 groups, data were tested for normality. Normally distributed data were analyzed using an unpaired t-test, while non-normally distributed data were analyzed using the Mann–Whitney test. All values are given as mean ± standard error of the mean. A P value of less than .05 was considered statistically significant.

Results

Generation of Hepatocyte-specific EGR1-deficient Mice Using CRISPR/Cas9-mediated Somatic Gene Editing

To get insight into the physiological role of EGR1 in hepatic lipid metabolism, we generated a mouse model with a hepatocyte-specific knockout of Egr1 using CRISPR/Cas9 somatic gene editing (Figure 1A). Three single-guide RNAs targeting exon 1 of Egr1 were expressed in hepatocytes using adeno-associated virus, resulting in an effective single-guide RNA targeting efficiency (Figure 1B) and a strong reduction in hepatic EGR1 protein levels in mice on a chow diet (Figure 1C and Supplementary Figure 1A). Hepatocyte-specific deletion of Egr1 did not result in changes in body weight, lean mass, or fat mass compared to empty vector control mice fed a chow diet (Supplementary Figure 1B).Figure 1. Gene editing strategy and liver phenotype upon loss of Egr1. (A) Somatic gene editing strategy to edit the Egr1 gene in mice expressing Cas9 specifically in hepatocytes, using 3 guide RNAs targeting exon 1 of the Egr1 gene at different locations packaged in adeno-associated virus serotype 8 viral particles. (B) sgRNA targeting efficiency of the 3 sgRNAs targeting Egr1 using primers targeting the expected cut sites of the relative sgRNA. (C) Relative EGR1 protein levels after Egr1 gene editing (−91%). (D) Liver weight, (E) liver-to-body weight ratio and (F) liver triglyceride, and (G) cholesterol content of hepatocyte-specific Egr1 knockout mice and controls fed chow or HFD. (H) Hematoxylin and eosin staining (upper panel) and Oil-red-O staining (lower panel) of the liver of hepatocyte-specific Egr1 knockout and control mice fed chow. Bars represent 100 μm. (I) Plasma alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels. Following an acetate incorporation experiment and fatty acid methylation, newly synthesized (J) C16:0, (K) C16:1, (L) C18:0, and (M) C18:1 was determined in livers of hepatocyte-specific Egr1 knockout and control mice. Data shown as mean ± standard error of the mean. sgRNA, single-guide RNA.

Hepatic Deficiency of EGR1 Does Not Induce a Liver Phenotype on a Chow Diet

Previous studies using whole-body knockout mice indicated a role for EGR1 in insulin sensitivity. However, 12 weeks after the knockdown, we did not find any differences in glucose homeostasis or insulin sensitivity in hepatic EGR1-deficient mice compared to controls fed a chow diet (Supplementary Figure 1C–E). In addition, no changes were observed in liver weight, liver-to-body weight ratio, subcutaneous, and gonadal WAT depots (Figure 1D, E and Supplementary Figure 1F, G). We observed no changes in plasma triglycerides, nonesterified fatty acid, and APOB levels, and only a small reduction in VLDL-associated triglycerides in plasma of hepatic EGR1-deficient was seen (Supplementary Figure 1H–K). Hepatic cholesterol and triglyceride levels were unchanged in EGR1-deficient livers, and H&E staining showed no major abnormalities at a histological level (Figure 1F–H). Oil Red O staining and quantitative analysis did not reveal major differences. Additional histological staining for established zonation markers, glutamine synthetase (pericentral), and arginase 1 (periportal), as well as gene expression analysis of additional zonation markers, also revealed no significant abnormalities in the livers of EGR1-deficient mice maintained on a chow diet (Figure 1H and Supplementary Figure 2A–D). Furthermore, plasma levels of alanine aminotransferase (ALT) and aspartate aminotransferase (AST), indicators of hepatocellular injury, were not altered in hepatic EGR1-deficient mice compared to controls (Figure 1I).

Hepatic Deficiency of EGR1 Results in Decreased Plasma Cholesterol Levels

Previous studies reported that EGR1 is a regulator of several genes involved in cholesterol biosynthesis, including Hmgcs and Hmgcr.30 Transcriptome and subsequent GSEA analysis of the liver indeed revealed a negative enrichment of cholesterol biosynthesis in EGR1-deficient livers (negative enrichment score = −1.38, P = .05) (Supplementary Figure 3A). In line with this, we observed a decrease in total plasma cholesterol levels in hepatic EGR1-deficient mice compared to controls (Supplementary Figure 3B and C). However, we did not observe differences in fractional de novo cholesterol synthesis as determined by the incorporation of [1-^13^C] acetate, hepatic cholesterol levels, bile flow and biliary cholesterol, and fecal neutral sterol secretion in EGR1-deficient livers compared to controls (Supplementary Figure 3D–H).

Hepatic Deficiency of EGR1 Causes an Increase in De Novo Lipogenesis

To further investigate potential changes in lipid metabolism, we determined de novo lipogenesis as measured by the incorporation of ^13^C-acetate into fatty acids. Significant increases in newly synthesized palmitic acid (C16:0) and stearic acid (C18:0) were observed in EGR1-deficient livers (Figure 1J–M). Total levels, however, of these fatty acid species remained largely unchanged or only showed a trend toward an increase (Supplementary Figure 4A–D). In contrast to the observed increase in de novo lipogenesis, no significant effects on mRNA levels of genes involved in this process, such as genes encoding enzymes involved in converting acetyl-CoA into C16:0 (eg, Fasn and Acaca), desaturation (eg, Scd1), and elongation of C16 to C18 (eg, Elovl6), were observed in hepatocyte-specific EGR1-deficient mice as compared to controls (Supplementary Figure 4E). These findings indicate an increase in de novo lipogenesis of certain lipids upon EGR1 ablation, potentially explaining the altered zonation of hepatic lipids.

Hepatocyte-specific EGR1-deficient Mice Fed a High-fat Diet Display Increased Hepatic Triglyceride Accumulation Compared to Control Mice

Next, we challenged hepatocyte-specific EGR1-deficient and control mice with a HFD containing 60% fat to observe any phenotypic changes. After 15 weeks of HFD feeding, there were no alterations in body weight or liver weight upon ablation of EGR1 (Figure 2A–C). However, analysis of hepatic lipids revealed that triglyceride levels increased significantly by 43% in response to the HFD in hepatocyte-specific EGR1-deficient mice, while cholesterol levels remained unchanged (Figure 2D and E). Plasma ALT and AST levels also remained unaltered (Figure 2F). Oil-Red-O and H&E staining revealed a mild increase in macrovesicular steatosis as pathologically determined, but no changes in lipid zonation or zonation markers were observed (Figure 2G and Supplementary Figure 5A–D). Furthermore, the decrease in total plasma cholesterol levels observed in chow-fed hepatic EGR1-deficient mice was exacerbated, with a clear decrease in plasma LDL-cholesterol levels, and this was accompanied by a stronger negative enrichment of cholesterol biosynthesis genes in EGR1-deficient livers (negative enrichment score = −1.82, P < .0001) (Figure 2H, I and Supplementary Figure 3I).Figure 2. Liver-specific EGR1-deficient mice fed a HFD results in hepatic triglyceride accumulation. (A) Body weight, (B) liver weight, and (C) liver-to-body weight ratio of hepatocyte-specific Egr1 knockout mice and control mice fed a HFD for 15 weeks (n = 8–9). (D) Liver triglycerides and (E) cholesterol content (note: 2 control samples were lost during Bligh and Dyer lipid extraction processing error). (F) Plasma ALT and AST levels. (G) Hematoxylin and eosin staining (upper panel) and Oil-red-O staining (lower panel) of the liver of hepatocyte-specific Egr1 knockout and control mice fed a HFD. Bars represent 100 μm. (H) Plasma cholesterol levels along with (I) the plasma cholesterol FPLC profile of hepatocyte-specific Egr1 knockout mice and control mice fed a HFD for 15 weeks. Data shown as mean ± standard error of the mean. FPLC, fast performance liquid chromatography.

Hepatic Deficiency of EGR1 Causes an Imbalance in Expression of Genes Involved in Fatty Acid Oxidation and Mitochondrial Respiration

Transcriptome analysis of hepatic EGR1-deficient mice fed a chow or HFD revealed an increase in genes involved in fatty acid oxidation compared to wild-type littermates, while genes involved in mitochondrial respiration were decreased. EGR1 deficiency on both diets resulted, for example, in an induction of Cpt1a, which mediates the first step of fatty acid oxidation by transferring an acyl group from CoA to carnitine. Furthermore, gene expression of Acot1, which is responsible for the hydrolysis of Acyl-CoA into fatty acids and free CoA, and Gpd2, which constitutes the glycerol-3-phosphate dehydrogenase shuttle, was induced in EGR1-deficient livers (Figure 3A–C). Interestingly, genes encoding enzymes involved in peroxisomal fatty acid oxidation (ie, Acot3, Acot4, Decr2, Acnat2, and Acnat1) were also induced in EGR1-deficient livers compared to controls (Figure 3A–C). In contrast, genes encoding mitochondrial components of complex I (ie, Ndufa2, Nadufa4l2, Ndufa11, Ndufb10, Ndufb6, and Ndufb9) and adenosine triphosphate (ATP) synthase (ie, Atp5a1 and Atp5o) were strongly downregulated (Figure 3A–C). Taken together, loss of hepatic EGR1 leads to significant alterations in gene expression related to fatty acid metabolism and mitochondrial function under chow as well as HFD conditions.Figure 3. Genes involved in fatty acid oxidation are altered upon hepatic Egr1 depletion. Using RNA sequencing data from livers, heatmaps were made of genes involved in fatty acid oxidation in hepatocyte-specific Egr1 knockout and control mice fed a (A) chow or (B) HFD. Heatmaps were constructed based on the z-score calculated from the fragments per kilobase of exon per million mapped reads scores. (C) Schematic representation of the genes altered in various processes involved in fat oxidation in mitochondria and peroxisomes. Created with BioRender.

Mitochondrial Oxidative Respiration Is Decreased Upon Ablation of EGR1 in Hepatocytes

To further investigate the role of EGR1 in mitochondrial function, we performed extracellular flux analysis using a Seahorse analyzer in primary hepatocytes isolated from chow-fed hepatic EGR1-deficient mice and controls. Interestingly, we observed a marked decrease in the maximum oxygen consumption rate in EGR1-deficient hepatocytes, indicating a reduced respiratory capacity (Figure 4A–C). In contrast, no effect on mitochondrial respiration and substrate-dependent oxidative phosphorylation was observed in isolated mitochondria from the livers of EGR1-deficient mice using an Oroborus Oxygraph (Supplementary Figure 6A and B). Furthermore, there was no difference in citrate synthase activity in EGR1-deficient liver homogenates or isolated mitochondria compared to controls, and only the ratio of citrate synthase activity in isolated mitochondria versus the liver was decreased in EGR1 deficiency, indicating a reduction in mitochondrial content (Supplementary Figure 6C–E). To further investigate this, we assessed mitochondrial biogenesis markers. While TOM20 protein levels in whole liver lysates remained unchanged, we observed a reduction in Polg expression, a key gene involved in mitochondrial DNA replication and repair, whereas other genes involved in mitochondrial biogenesis remained unaltered upon EGR1 ablation (Supplementary Figure 6F and G).Figure 4. Mitochondrial function is altered upon Egr1 depletion in hepatocytes. (A) Mitochondrial stress test seahorse profile, (B) basal oxygen consumption rate (OCR), and (C) maximum OCR of primary hepatocytes derived from hepatocyte-specific Egr1 knockout and control mice on a chow diet. (D) Mitochondrial stress test seahorse profile in nonstarved and 16 hours starved conditions (representative figure of 5 biological replicates), (E) basal and maximal oxygen consumption rate (OCR) of EGR1 knockdown HepG2 cells and controls in the nonstarved condition. Substrate oxidation stress tests assays using (F) etomoxir, (G) UK5009, or (H) BPTES as inhibitors of long-chain fatty acid, glucose/pyruvate, or glutamine transport in EGR1 knockdown HepG2 cells and controls. (I) Summary of panels F–H showing the percentage of basal and maximum respiration after addition of inhibitors compared to vehicle in EGR1 knockdown HepG2 cells and controls. Data are shown as mean ± standard error of the mean.

We next investigated whether the role of EGR1 in oxygen consumption is conserved in humans. To do this, we generated EGR1-deficient HepG2 cells using shRNA-mediated knockdown. Similar to our findings in primary mouse hepatocytes, basal and maximum oxygen consumption rates were reduced, and this effect was diminished when cells were glucose and serum starved (Figure 4D and E). A more in-depth analysis of these cells revealed that the basal and maximum endogenous fatty acid utilization for respiration were decreased in EGR1-deficient HepG2 cells compared to controls, while the exogenous palmitate utilization was unchanged (Supplementary Figure 7). Next, we delineated the specific substrate pathway, that is, long-chain fatty acids, glucose/pyruvate, and glutamine, responsible for the decreased oxygen consumption rate, using etomoxir, UK5099, and BPTES, respectively, as inhibitors. Control HepG2 cells primarily rely on long-chain fatty acids and glucose/pyruvate for basal respiration and long-chain fatty acids and glutamine under conditions of high substrate demand (Figures 4F–I and Supplementary Figure 8). Basal and maximum respiration in EGR1-deficient HepG2 cells was lower compared to HepG2 control cells in all conditions (Figure 4F–I). Furthermore, only a small response was seen in EGR1-deficient HepG2 cells upon the administration of any of the inhibitors. EGR1-deficient HepG2 cells relied on long-chain fatty acids for basal respiration, but no significant substrate reliance was observed under conditions of increased substrate demand, likely because of the already low levels of maximum oxygen consumption rate (Figure 4F–I). Together, these results indicate that EGR1 is important for mitochondrial function, control of mitochondrial respiration, and potentially substrate utilization in response to demand.

Hepatic Deficiency of EGR1 Causes Fasting-induced but not Cold-induced Hepatic Lipid Accumulation

To investigate the physiological consequences of these changes in fatty acid oxidation and mitochondrial respiration, we determined these processes under conditions of increased lipid utilization, as induced by overnight fasting. In response to overnight fasting, chow-fed hepatic EGR1-deficient mice presented with higher levels of total plasma acylcarnitines, indicating reduced or incomplete in fat oxidation in tissues (Figure 5A). This increase was primarily caused by an increase in acetylcarnitine (C2), while there were no changes in other short-chain acylcarnitines or medium- and long-chain acylcarnitines (Figures 5A–D and Supplementary Figure 9). In addition, there was an increase in ratio of C0/(C16 + C18) acylcarnitines, which reflects reduced CPT1α activity, and increased C4OH + C3DC, which reflects increased ketogenesis (Figure 5E and F). The combined observations of increased plasma C2 levels, increased Cpt1a mRNA levels, and decreased expression of Complex I-related genes suggest an imbalance between fatty acid β-oxidation and the demand by the tricarboxylic acid cycle and respiratory chain. We next assessed potential effects on whole-body energy homeostasis using indirect calorimetry. However, no changes were observed in VO_2_, VCO_2_, respiratory exchange ratio, energy expenditure, body temperature, or activity upon loss of hepatic EGR1 at either 22 °C (Supplementary Figure 10).Figure 5. Increased plasma acylcarnitines and hepatic triglycerides accumulation after prolonged fasting upon the loss of hepatic Egr1. (A) Total acylcarnitines, (B) acylcarnitine distribution (LC = long chain, MC = mid chain, SC = short chain), (C) acetylcarnitine (C0), and (D) C2 acylcarnitine concentrations in plasma after prolonged fasting (overnight for 16 hours) at room temperature at 15 weeks of age of hepatocyte-specific Egr1 knockout and control mice fed a chow diet. (E) Plasma C0/(C16 + C18) and (F) C4OH + C3DC acylcarnitines in hepatocyte-specific Egr1 knockout and control mice. (G) Liver triglyceride and (H) cholesterol content after prolonged fasting (overnight for 16 hours). (I) Liver triglyceride and (J) cholesterol content after prolonged fasting at 29 °C with the last 4 hours at 4 °C (indicated as 4 °C) of hepatocyte-specific Egr1 knockout and control mice fed a chow diet. (K) Total acylcarnitines and (L) acylcarnitine distribution after prolonged fasting at 29 °C with the last 4 hours at 4 °C. Data are shown as mean ± standard error of the mean.

Interestingly, similar to the hepatic EGR1-deficient mice on a HFD, hepatic triglyceride content was increased by 24% in response to prolonged fasting at room temperature in hepatic EGR1-deficient mice compared to controls, while no differences were observed in hepatic cholesterol content (Figure 5G and H). To further maximize the changes in the fatty acid oxidation system, mice were exposed to overnight fasting at thermoneutrality (29 °C) followed by cold exposure (4 °C). However, no difference was observed in hepatic lipid accumulation or plasma acylcarnitines upon cold exposure in liver-specific EGR1-deficient mice compared to controls (Figure 5I–L). Overall, hepatic deficiency of EGR1 causes an increase in hepatic triglyceride accumulation under HFD feeding conditions (43% increase) and in response to prolonged fasting on chow (24% increase) but not in response to overnight fasting at thermoneutrality followed by cold exposure.

Discussion

EGR1 is a transcription factor that regulates various cellular and metabolic processes. Previous studies have primarily investigated its function using whole-body Egr1 knockout models, which exhibit complex metabolic phenotypes that obscure tissue-specific effects. In this study, we specifically assessed the role of hepatic EGR1 in adult mice using CRISPR/Cas9-mediated somatic gene editing. We show that hepatocyte-specific deletion of EGR1 results in hepatic lipid accumulation in mice fed a HFD or after prolonged fasting. These mice were not protected against diet-induced obesity, indicating that the phenotype of reduced lipid accumulation observed in previous studies, using whole-body Egr1 knockout mice, is a peripheral effect and not a liver-driven effect. Mechanistically, we observed an imbalance between fatty acid oxidation and mitochondrial respiration demand which likely contributes to these alterations in hepatic lipid metabolism.

While under standard chow-fed conditions, hepatic EGR1 deficiency did not result in changes in total hepatic levels of triglycerides, increased lipid accumulation around the central vein (zone 3) was observed, which is the first area to accumulate fat during metabolic dysfunction-associated steatotic liver disease development. We also observed a mild decrease in plasma cholesterol levels, consistent with the downregulation of cholesterol biosynthetic genes. However, no differences could be detected in hepatic cholesterol content, fractional de novo cholesterol synthesis, and biliary cholesterol excretion. Apparently, a subtle reduction in cholesterol flux still resulted in a significant decrease in plasma cholesterol levels. Alternatively, the decrease in plasma cholesterol levels could also be due to a modest reduction in VLDL-associated triglycerides (Supplementary Figure 1J–K). Under HFD conditions, the reduction in plasma cholesterol levels was primarily caused by a reduction in LDL cholesterol. Previous studies have linked EGR1 to plasma cholesterol regulation,30, 31, 32 particularly in whole-body knockout models showing reduced HDL cholesterol levels.30 Our study indicates that hepatic EGR1 affects LDL but not HDL cholesterol levels. Since besides the liver, the intestine also contributes to the biogenesis of HDL, we speculate that the effect of EGR1 on HDL is primarily driven by intestinal EGR1.

We next investigated whether EGR1 affects de novo lipogenesis and noted an increase in the synthesis of palmitic acid and stearic acid, suggesting enhanced lipogenic activity. When hepatic EGR1-deficient mice were subjected to a HFD, this resulted in an increase in hepatic triglyceride accumulation, without notable changes in plasma triglyceride levels. The increased de novo lipogenesis under a chow diet may contribute to an elevated baseline level of fatty acid synthesis. Consequently, when challenged with an HFD, the liver’s capacity to further synthesize and store triglycerides is likely augmented due to this pre-existing increase in lipogenesis, making these mice more sensitive to hepatic lipid accumulation.

Mechanistically, we identified an imbalance in fatty acid oxidation and oxidative phosphorylation in hepatic EGR1-deficient mice. As an immediate early gene and transcription factor, EGR1 is expected to regulate lipid metabolism and mitochondrial function primarily through transcriptional control of downstream target genes. Changes in these processes were highlighted by the upregulation of genes such as Cpt1a and Acot1 and downregulation of components of Complex I and ATP synthase. These findings are in line with previous studies that found that EGR1 influences fatty acid oxidation and mitochondrial function. Previous studies have also shown that EGR1 overexpression mitigates acetaminophen-induced hepatotoxicity by promoting mitochondrial respiration and fatty acid oxidation, while hepatocyte-specific ablation of Egr1 had opposite effects.33 Guo et al34 reported increased expression and promotor activity of Acox1, Cpt1, and Acaa2 in db/db mice with shRNA-mediated Egr1 interference. However, their model displayed a severe diabetic phenotype with elevated EGR1 expression, and the shRNA approach may have also affected nonhepatocyte liver cells. In contrast, we did not observe changes in Acaa2 gene expression but we identified a broader impairment in fatty acid oxidation as well as a reduced demand by the tricarboxylic acid cycle and electron transport chain.

Our study also identifies a potential link between EGR1 and peroxisomal lipid metabolism. Transcriptome analysis revealed increased activity of genes involved in peroxisomal lipid metabolism, accompanied by elevated protein levels of ACOX1 and PMP70 (Supplementary Figure 6G). These findings suggest that peroxisomal fatty acid oxidation may partially compensate for mitochondrial dysfunction in EGR1-deficient livers. EGR1 deficiency also resulted in reduced mitochondrial capacity in primary hepatocytes and HepG2 cells but not in isolated mitochondria from livers of hepatic EGR1-deficient mice, suggesting that the deficiency is driven by changes in intracellular processes rather than caused by a specific mitochondrial defect. In HepG2 cells, we found that loss of EGR1 resulted in a shift in metabolic flexibility in conditions of high substrate demand, characterized by reduced reliance on long-chain fatty acids and glutamine. The observed mitochondrial dysfunction might also be partially mediated by crosstalk between peroxisomes and mitochondria, as suggested by previous studies.35^,^36

Recent studies have also linked EGR1 to age-related metabolic dysfunction through its regulation of Cidea expression.37 We observed reduced Cidea expression in hepatic EGR1-deficient mice under HFD conditions but not in chow-fed mice (Supplementary Tables 1 and 2). The reduction of Cidea expression in HFD-fed hepatic EGR1-deficient mice may reflect an adaptive response to altered lipid metabolism, potentially contributing to the observed triglyceride accumulation. Hepatic EGR1-deficient mice displayed increased hepatic triglyceride accumulation upon prolonged fasting but not upon cold exposure. This suggests that the liver may be less responsive to cold-induced metabolic adaptations compared to brown and WAT. Given the divergent metabolic phenotypes observed in whole-body and hepatic EGR1-deficient models, future studies should explore the role of EGR1 in other tissues, such as brown adipose tissue, under thermogenic conditions.

Consistent with our observations, analysis of human liver transcriptomes using the MegaMASLD tool revealed that EGR1 expression is decreased in MASL and metabolic-associated steatohepatitis but returns to baseline levels in cirrhotic livers (Supplementary Figure 11).38 This stage-dependent regulation suggests that suppression of EGR1 may contribute to early metabolic dysfunction.

Conclusion

Our findings identify hepatic EGR1 as a key regulator of liver lipid metabolism, particularly in response to dietary and fasting challenges. Mechanistically, EGR1 regulates de novo lipogenesis and fatty acid oxidation pathways, and a deficiency in EGR1 contributes to MAFLD. Insight in the mechanism by which EGR1 regulates hepatic lipid metabolism could open new avenues for the treatment of MAFLD.