Lysophosphatidic acid receptor 1 antagonist (EPGN2154) causes regression of NASH in preclinical NASH models

Abstract

Background:

NASH causes a tremendous health care burden in the United States. A glucagon-like peptide-1 agonist, semaglutide (Sema), treatment resulted in hepatic steatosis reduction in clinical trials of NASH. Lysophosphatidic acid receptor 1 antagonists are known to have antifibrotic effects in several organs. We tested Sema and a novel lysophosphatidic acid receptor 1 antagonist, EPGN2154, individually and in combination to evaluate their efficacy for NASH remission in preclinical models.

Methods:

In the present study, we used (1) C57Bl6/J wild-type mice fed on a high-fat, high-carbohydrate (HFHC) diet for 16 weeks and (2) leptin-deficient mice (ob/ob) fed on an Amylin liver NASH diet for 16 weeks. After 16 weeks, the mice were randomly distributed in equal numbers in (1) no-drug, (2) EPGN2154, (3) Sema, and (4) EPGN2154+Sema treatment groups for 8 additional weeks at a dosage of 10 mg/kg body weight for EPGN2154 (oral gavage, 5 days a week) and 6.17 μg/kg body weight of Sema (subcutaneous injection every alternate day, 3 days a week).

Results:

In the wild-type-high-fat, high-carbohydrate model, we observed the most body weight loss in the EPGN2154+Sema combination group compared to the other treatment groups. All groups led to a significant reduction in alanine transaminase levels when compared to high-fat, high-carbohydrate–fed wild type. However, no significant difference in alanine transaminase levels was observed among the treatment groups. In the ob/ob mice study, Sema did not cause body weight loss. Moreover, the EPGN2154 and the combination groups had a lower NAFLD Activity Score and incidence of advanced-stage hepatic fibrosis than the Sema group.

Conclusions:

EPGN2154 demonstrated a hepato-protective effect independent of body weight loss in preclinical NASH models.

INTRODUCTION

NASH has imposed a sizable health care burden globally.¹ Studies have found many causal factors for NASH, including genetic polymorphisms, type 2 diabetes, dyslipidemia, environment, and lifestyle of patients.² As a result of its multifaceted nature, researchers have tried NASH remission approaches targeting various receptors of molecular signaling pathways, including bile acid receptors,³ peroxisome proliferator–activated receptors,⁴ liver-specific thyroid hormone receptors,⁵ chemokine receptors,⁶ and hepatic lipid metabolism.⁷ Despite promising preclinical and clinical data, the Food and Drug Administration (FDA) of the United States has yet to approve any medication specifically for NASH.⁸

A glucagon-like peptide-1 receptor agonist, semaglutide (Sema), has shown the ability to induce weight loss in humans.⁹ In a placebo-controlled phase 2 clinical trial in patients with NASH, Sema treatment resulted in a significantly higher percentage of patients with NASH resolution than placebo. However, the trial did not show a significant between-group difference in the percentage of patients with improved fibrosis.¹⁰ Small molecule lysophosphatidic acid receptor 1 (LPAR1) antagonists have shown therapeutic efficacy in preclinical models of lung and kidney fibrosis as well as clinical efficacy in idiopathic pulmonary fibrosis.^11–17 Further, pharmacological inhibition of a pro-fibrogenic protein like monocyte chemoattractant protein‐1 (MCP-1) has shown the reduction of macrophage infiltration in the liver of the preclinical NASH models.¹⁸ Interestingly, LPAR1 deficiency causes a reduction of infiltrating macrophages and expression of MCP-1 in response to lipopolysaccharide in a preclinical model of chronic lung diseases.¹⁹ During NASH progression, the infiltrating macrophages secrete proinflammatory cytokines in the liver, activating HSCs, thereby initiating the process of fibrosis.²⁰ HSCs constitute nearly 90% of the collagen-producing cells in a fibrotic liver.²¹ Studies have shown that lysophosphatidic acid (LPA) induces the proliferation of HSC.²² Additional studies have shown that patients with type 2 diabetes have a higher prevalence of NASH.^23–25 Combination drug therapies for NASH have shown better outcomes in clinical trials.²⁶ The present study investigated the efficacy of a novel small molecule LPAR1 antagonist, namely EPGN2154, and the glucagon-like peptide-1 receptor agonist Sema, either individually or in combination, in preclinical NASH models.

Our previous studies reported that wild-type (WT), C57Bl6/J, mice fed a high-fat and high-carbohydrate diet (HFHC) develop hepatic fibrosis.^27–29 Leptin-deficient mice (ob/ob) develop a type 2 diabetes phenotype with mild hepatic fibrosis.³⁰ Ob/ob mice, when fed the amylin liver NASH (AMLN) diet, develop a severe NASH phenotype, including hepatic fibrosis.³¹ Both WT and ob/ob mice NASH models allow for the evaluation of the efficacy of EPGN2154 and Sema, to determine (1) the effect of LPAR1 antagonist, EPGN2154, on body weight and regression of hepatic fibrosis; (2) the effect of Sema in hepatic fibrosis regression in the absence/presence of body weight loss; and (3) the efficacy of EPGN2154 compared to Sema in hepatic fibrosis regression.

METHODS

Determination of mouse pharmacokinetics and oral bioavailability of EPGN2154

All mice studies were approved by the Institutional Animal Care and Use Committee at Siri Pharma Labs, San Jose, CA, approval number A4734-01, and Children's Hospital Los Angeles, Los Angeles, CA, approval number 392. EPGN2154 solution was prepared fresh in a vehicle formulation of 20% Solutol and 40% PEG400 in 40% water at a 5 mg/mL concentration. A total of 12 female CD1 mice were acclimatized to living conditions and then divided into 2 groups of 6 animals each. The study groups are displayed in Supplemental Table S1, http://links.lww.com/HC9/A663. EPGN2154 was administered by oral gavage (dose 20 mg/kg). EPGN2154 was administered intravenously (dose 1 mg/kg) at a dose volume of 5 mL/kg based on mouse weight. Plasma samples were diluted with blank mouse plasma as needed. An aliquot of 20 µL of plasma sample was extracted with 100 µL of acetonitrile containing internal standard (terfenadine). The mixture was vortexed on a shaker for 15 minutes and subsequently centrifuged at 4000 rpm for 15 minutes. An aliquot of 70 µL of the supernatant was mixed with 70 µL of water for the injection into the LC/MS/MS. Calibration standards and quality control samples were prepared by spiking the test compound into blank mouse plasma and then processed with the unknown samples. The instrument setup is provided in the Supplemental Text File, http://links.lww.com/HC9/A663.

Mice experiments

Six- to 8-week-old male C57Bl6/J WT and B6.Cg-Lep^ob/J (ob/ob) mice (Jackson Laboratory, Bar Harbor, ME) were housed in a 12-hour light-dark cycle maintained in a (22±2°C) temperature-controlled room. WT mice were randomized to chow (n = 4) (Research Diets Inc., New Brunswick, NJ) or high-fat, high-carbohydrate (n = 40) (HFHC) diet (D12331i, Research Diets Inc., 58 kcal % fat; Research Diets, New Brunswick, NJ) and drinking water with high fructose (55% fructose by weight; Acros Organics, Morris Plains, NJ) and sucrose (45% sucrose by weight; Acros Organics, Morris Plains, NJ) mixture at a concentration of 42 g/L.⁵ Animals were provided ad-lib access to diets for 24 weeks. Ob/ob mice were randomized to chow (n = 5) (Research Diets Inc., New Brunswick, NJ) or AMLN diet (n = 50) (D09100310i, 40% fat, 22% fructose, and 2% cholesterol, Research Diets, New Brunswick, NJ). After 16 weeks on the AMLN diet, a portion of the ob/ob mice cohort (n = 10) was euthanized to serve as a 16-week sentinel group (AMLN 16 wk). WT and ob/ob mice were provided ad-lib diet access for 24 weeks.

In the experiments with WT mice, after 16 weeks on the HFHC diet, mice were randomly distributed in each experimental group (n = 10): (1) HFHC (no-drug treatment), (2) HFHC+2154 (EPGN2154 treatment at 10 mg/kg body weight), (3) HFHC+Sema ( Sema treatment at 6.17 μg/kg body weight), and (4) HFHC+2154+Sema (combination therapy of EPGN2154 at 10 mg/kg body weight and Sema treatment at 6.17 μg/kg body weight). EPGN2154 was administered orally once daily for 5 days a week. Sema (BOC Sciences, Shirley, NY) was administered by subcutaneous injection every alternate day, that is, 3 days a week. After 8 weeks of treatment, the mice were euthanized, and tissues were harvested for assays and histological analysis.

In the experiments with ob/ob mice, after 16 weeks on the AMLN diet, the mice were randomly distributed in each experimental group (n = 10): (1) AMLN +Vehicle (Vehicle treatment group), (2) AMLN+2154 (EPGN2154 treatment at 10 mg/kg body weight), (3) AMLN+Sema (Sema treatment at 6.17 μg/kg body weight), and (4) AMLN+2154+Sema (combination therapy of EPGN2154 at 10 mg/kg body weight and Sema treatment at 6.17 μg/kg body weight). EPGN2154 was administered orally once daily for 5 days a week. Sema was administered by subcutaneous injection every alternate day, that is, 3 days a week. After 8 weeks of treatment, the mice were euthanized, and tissues were harvested for assays and histological analysis. Body weights and food intake were recorded every week. All animal studies were approved by the Institutional Animal Care and Use Committee at the Children’s Hospital Los Angeles.

Body composition

Body composition analysis (fat mass and lean mass) was done using EchoMRI-100H Body Composition Analyzer (EchoMRI, TX).

Plasma alanine transaminase assay

Plasma isolated from the whole blood was used to estimate alanine transaminase (ALT) concentration using ALT Activity Assay kit (Sigma, St.Louis, MO) as per the manufacturer’s instruction.

Oil Red O staining

Oil Red O staining was performed on the 5 µm frozen liver tissue section using the Oil Red O Stain Kit (Lipid Stain) (ab150678) from Abcam (Boston, MA) according to the manufacturer’s instructions. The Oil Red O staining area percentage in the liver sections was quantified by a single independent pathologist blinded to the experimental design and treatment groups.

Hydroxyproline assay

Frozen liver tissue (100 mg) was used to quantify the hydroxyproline content using Hydroxyproline Assay Kit (Colorimetric) (ab222941) from Abcam (Boston, MA) according to the manufacturer’s instructions.

LPAR1 expression

Total RNA was isolated from 10 mg of liver tissue by TRI Reagent Solution (Invitrogen, Waltham, MA). 5 µg of total RNA was used for cDNA synthesis using SuperScript III Reverse Transcriptase (Invitrogen, Waltham, MA). cDNA was used to quantify the relative expression of LPAR1 using a Taqman probe for LPAR1 (Mm01346925_m1) (Invitrogen, Waltham, MA). The relative expression of LPAR1 was calculated by the ΔΔC_t method using a Taqman probe for Rpl18 (Mm01197265_g1) (Invitrogen, Waltham, MA) as a housekeeping gene.

Liver histology analysis and immunohistochemistry

Liver tissue was harvested from mice, fixed in 10% formalin, and sectioned in a microtome to generate 5-μm sections for histologic analyses. The sections were stained with eosin and hematoxylin and analyzed to determine the NAFLD Activity Score (NAS) by a single independent pathologist blinded to the experimental design and treatment groups. In NAS, liver histology is graded on steatosis (score 0–3), lobular inflammation (score 0–3), and ballooning (score 0–2).³² Sirius Red staining evaluated hepatic fibrosis on a scale (0–4). Immunohistochemistry was performed on the liver section of the ob/ob mice for alpha-smooth muscle actin, collagen I, laminin and galectin-3 using anti-alpha-smooth muscle actin antibody (Cat# ab5694, Abcam, Cambridge, UK), anti-Collagen I antibody (Cat# ab270993, Abcam, Cambridge, UK), anti-Laminin antibody (Cat# ab11575, Abcam, Cambridge, UK), and anti-galectin-3 antibody (Cat# ab76245, Abcam, Cambridge, UK) respectively. The biotinylated Goat Anti-Rabbit IgG (H+L) (Cat# ab64256, Abcam, Cambridge, UK) and streptavidin HRP (Cat# ab64269, Abcam, Cambridge, UK) were used for detection.

The effect on the migration of MCP-1 stimulated RAW264.7 in response to LPAR1 antagonists EPGN696 and EPGN2154

EPGN2154 (Lot EPGN2154.005) and EPGN696 (Lot EPGN696.009) were prepared as a stock solution in DMSO at 10 mM concentration. The stock solution was diluted 50-fold with PBS to the final concentration to be used in the migration assay. For the migration assay, briefly, each test solution (1 μL) was diluted 50-fold in RAW264.7 cell suspension (49 μL). Final concentrations in a dose range of 10 fM to 10 μM were tested. MCP-1 solution (10⁻⁸M, 26 μL) was added to the bottom chamber of the Boyden apparatus. A 5-micron polycarbonate membrane was applied, followed by the gasket and the donor chamber apparatus screwed into place. Once assembled, positive controls and test wells (compound plus cell incubations), all ~50 μL, were pipetted into the upper chamber. A set of background wells was established with a bottom chamber containing MEM only and the upper wells RAW264.7 cells (50 μL). Once complete, the setup is incubated at 35°C for 1 hour. After 1 hour, the apparatus was disassembled and inverted to allow for the removal of the upper chamber and gasket. The filter was clamped, and the underside was rinsed and scraped in PBS before fixing it in methanol and drying on a glass slide. The dried filter was hematoxylin and eosin stained, and cells were visualized by microscope. Cell count was determined for each well as the average of 5 fields under the microscope. Cell counts for the positive control (no antagonist) and test wells were corrected for background by subtraction of any cells counted on the membrane in the negative control (no MCP-1 in the receiving chamber).

Percent inhibition of migration was determined by

$$\left(\mathrm{Cellcountintestwells}\right)\times 100/\left(\mathrm{Cellcountinpositivecontrol}\right)$$

The effect of lysophosphatidic acid stimulated the proliferation of human stellate cells in response to the LPAR1 antagonist, EPGN696

HSCs were isolated as previously described.^33,34 DMEM, fetal calf serum, antimycotic/antibiotic, Cyquant cell proliferation kit, and Ca²⁺/Mg²⁺-free PBS were from Thermofisher (Waltham, MA). Oleoyl-LPA-sodium salt was from Cayman Chemicals (Ann Arbor, MI). Thirty percent fatty acid-free bovine serum albumin was purchased from Sigma (St.Louis, MO). LPA was dissolved in Ca²⁺/Mg²⁺-free PBS containing 0.1% fatty acid-free bovine serum albumin to give a 1 mM stock solution. The solution was sonicated for 15 minutes (min) before use. EPGN696 was diluted from a 10 mM solution in DMSO. HSCs were thawed and cultured in high glucose DMEM with 10% fetal calf serum and antibiotic/antimycotic at 37°C, 5% CO₂. Cells were used between passages 4 and 7. Cells were plated at 1000 or 2000 cells/well of a 96-well plate and incubated overnight. The following day media was replaced with that containing 0.5% fetal bovine serum (FBS) for 24 h. The following day media was replaced with 100 µl media containing 0.5% FBS and 0.1% DMSO or EPGN696 at 1 µM, 0.1 µM or 0.01 µM for 15 min, followed by the addition of a further 100 µl media containing 0.5% FBS, vehicle, 20 µM LPA (2x final concentration) and/or 1 µM, 0.1 µM, 0.01 µM EPGN696. Cells were incubated for 48 hours before removing the media and freezing the plate at −80°C overnight. According to the manufacturer’s instructions, the relative cell number per well was assessed using the Cyquant cell proliferation kit. The Cyquant assay determines cell density via a dye that fluoresces when bound to nucleic acids. Fluorescence was determined using a Clariostar plate reader.

Statistical analysis

Statistical comparison between the experimental groups was performed using one-way or two-way ANOVA and post hoc Bonferroni test. Student t test was used to determine a statistical comparison between the 2 experimental groups. The p-value of <0.05 was considered statistically significant. Results were presented as mean ± SEM.

RESULTS

The rationale for selecting EPGN2154

During our drug discovery program, we have synthesized multiple LPAR1 antagonists. Using a progressive screening cascade, compounds EPGN696 and EPGN2154 have been identified as leads meeting or exceeding program criteria and possessing good drug-like properties. EPGN2154 has demonstrated superior pharmacokinetics and distribution profiles that afford efficacy with much lower doses, leading to an improved risk-benefit profile compared to EPGN696 (Table 1). Furthermore, in preclinical models of diabetic kidney disease, EPGN2154 has demonstrated efficacy in the remission of kidney fibrosis and diabetic nephropathy. Data from ob/ob mice show that EPGN2154 reaches a steady state, and the half-life is appropriate for the once-daily dosing regimen used in this study (Supplemental Table S2, http://links.lww.com/HC9/A663). Oral gavage (PO) administration of EPGN2154 has shown absolute oral bioavailability of 40% and half-life of ~3 hours (Figure 1A and Supplemental Table S3, http://links.lww.com/HC9/A663).

TABLE 1.

Epigen’s LPAR1 antagonists—in vitro pharmacology and ADME-PK properties

	EPGN696	EPGN2154
LPAR1 IC50 Ca2+ influx (nM)	8	8
Inhibition of MCP-1-induced RAW264.7 migration IC50 (nM)	0.95	1.76
Mouse PK (20 mg/kg p.o.) AUC (h.ng/mL)	4069	25,700

Abbreviations: LPAR1, Lysophosphatidic acid receptor 1; MCP-1, monocyte chemoattractant protein‐1; PK, pharmaco kinetics.

FIGURE 1.

EPGN2154 and Sema combination therapy cause the maximum reduction in body weight. (A) EPGN2154 pharmacokinetics mean plasma concentration of EPGN2154 oral gavage (PO: 20 mg/kg) versus i.v. (1 mg/kg) over 24 hours. (B) Body weight of WT mice at week 16 on HFHC or chow diet. (C) Body weight change of WT mice from week 1 to week 16 when fed on HFHC or chow diet. (D) Body weight of WT mice from week 16 to week 24 during drug dosing. (E) Body weight change of WT mice from week 16 to week 24 during drug dosing. Mean± SEM. ****p<0.0001, ***p<0.001, **p<0.01, *p<0.05. Abbreviations: HFHC, high-fat, high-carbohydrate; PO, per-os; Sema, semaglutide.

EPGN2154 and Sema cause a reduction in body weight

WT mice fed the HFHC diet for 16 weeks had body weight higher than that of chow-fed mice (HFHC vs. Chow: 47.383±0.843 g vs. 33.650±0.833 g; p<0.0001) (Figure 1B). The HFHC-fed mice showed a body weight change of 24.021±0.734 g, whereas the chow-fed mice gained 10.15±1.569 g in 16 weeks (Figure 1C). After 8 weeks of drug administration, the HFHC+2154+Sem group had the lowest body weight compared to the HFHC (no drug), HFHC+2154, and HFHC+Sem groups (Supplemental Table S4, http://links.lww.com/HC9/A663) (Figure 1D). The HFHC+2154+Sem and the chow-fed group had no significant difference in body weight. The HFHC+2154+Sem group had shown maximum weight loss among the drug treatment groups (Supplemental Table S5, http://links.lww.com/HC9/A663) (Figure 1E). At 24 weeks of the study, the HFHC+2154+Sem group has shown the lowest body weight among the drug treatment groups (Supplemental Table S4, http://links.lww.com/HC9/A663) (Figure 2A). The body mass composition revealed that the HFHC+Sem+2154 group has the lowest fat mass percentage among the drug treatment groups (Supplemental Table S6, http://links.lww.com/HC9/A663). The HFHC+2154+Sem group has no significant difference in fat mass percentage from the chow group (Figure 2B). The HFHC+Sem+2154 group has the highest lean mass percentage among the drug treatment groups (Supplemental Table S7, http://links.lww.com/HC9/A663). The HFHC+2154+Sem group has no significant difference in lean mass percentage from the chow group (Figure 2C).

FIGURE 2.

EPGN2154 and Sema provide hepato-protection in the HFHC-fed WT mice. (A) Body weight of mice at week 24. (B) Fat mass percentage of mice at week 24. (C) Lean mass percentage of mice at week 24. (D) Liver weight of mice at week 24. (E) Liver-to-body weight ratio of mice at week 24. (F) Plasma ALT concentration of mice at week 24. Mean± SEM. ****p<0.0001, ***p<0.001. Abbreviations: ALT, alanine transaminase; AU, arbitrary unit; HFHC, high-fat, high-carbohydrate; Sema, semaglutide.

EPGN2154 and Sema cause a reduction in the liver weight and liver-to-body weight ratio

Both EPGN2154 and Sema cause a reduction in the liver weight compared to the no-drug treatment group. However, the HFHC+Sem group has the lowest liver weight compared to the other treatment groups (HFHC+Sem vs. HFHC (no-drug) vs. HFHC+2154 vs. HFHC+2154+Sem: 1.623±0.093 g vs. 3.775±0.190 g, p<0.0001, 2.849±0.252 g, p<0.0001, 1.943±0.077 g, p>0.05) (Figure 2D). Further, the HFHC+Sem group has a lower liver-to-body weight ratio of mice compared to the other treatment groups (HFHC+Sem vs. HFHC (no-drug) vs. HFHC+2154 vs. HFHC+2154+Sem: 0.0372±0.001 vs. 0.0694±0.003, p<0.0001, 0.060±0.003, p<0.0001, 0.053±0.001, p<0.0001). There was no observed significant difference between HFHC+Sem and the chow group (Figure 2E).

EPGN2154 and Sema improve hepatic injury and liver physiology

Both EPGN2154 and Sema reduced the plasma ALT concentration compared to that of the HFHC group (HFHC+2154 vs. HFHC (no-drug) vs. HFHC+Sem vs. HFHC+2154+Sem: 62.109±10.746 vs. 116.205±11.950, p<0.001, vs. 45.043±4.033, p>0.05, vs. 39.517±8.924, p>0.05). We observed no significant difference in plasma ALT levels between the drug treatment and chow groups (Figure 2F). The analysis of liver histology (Figure 3A) by NAS revealed that HFHC+2154 and HFHC+Sem groups have no significant difference in the steatosis score (2.100±0.314 vs. 1.400±0.221, p>0.05) and ballooning score (0.800±0.200 vs. 0.300±0.153, p>0.05) (Figure 3B). The liver sections of HFHC+2154 and HFHC+Sem groups have shown no difference in the Oil Red O stain area (Figure 3C). However, both HFHC+2154 and HFHC+Sem groups have lower Oil Red O stain area than HFHC (Supplemental Figure S1, http://links.lww.com/HC9/A663).

FIGURE 3.

EPGN2154 and semaglutide improve hepatic injury and liver physiology. (A) Representative image of hematoxylin and eosin–stained liver cross-section of experimental groups, “#” sign identifies the central hepatic veins. (B) NAS of the liver cross-section. (C) Oil Red O stain area percentage on the liver section of experimental groups at week 24. Mean± SEM. ****p<0.0001, ***p<0.001, **p<0.01, *p<0.05. Abbreviations: HFHC, high-fat, high-carbohydrate; NAS, NAFLD Activity Score; Sema, semaglutide.

EPGN2154 imparts protection from the progression of hepatic fibrosis

Histological analysis of Sirius Red–stained liver cross-section (Figure 4A) revealed that 50% of mice in the HFHC group have advanced-stage hepatic fibrosis (Fibrosis Grade >2). However, only 10% of mice in the HFHC+2154 and HFHC+Sem groups have advanced-stage hepatic fibrosis (HFHC vs. HFHC+2154 vs. HFHC+Sem: 50% vs. 10%, χ² p<0.0001; 10%, χ² p<0.0001, Figure 4B). No advanced-stage fibrosis was observed in the liver sections of mice from the HFHC+2154+Sem groups. HFHC+2154 mice group has lower hydroxyproline concentration in the liver compared to the HFHC group (HFHC+2154 vs. HFHC: 0.22±0.038 ng/µL per mg liver vs. 0.51±0.082 ng/µL per mg liver, p<0.001) (Figure 4C). The drug treatment mice groups did not show any significant differences in the hydroxyproline concentration in the liver. HFHC+2154 mice group has lower lysophosphatidic acid receptor 1 (LPAR1) expression in the liver compared to the HFHC group (HFHC+2154 vs. HFHC: 0.66±0.062 A.U. vs. 1.30±0.277 A.U.) (Figure 4D). The drug treatment and chow-fed groups have no significant difference in the hepatic LPAR1 expression.

FIGURE 4.

EPGN2154 imparts protection from the progression of hepatic fibrosis. (A) Representative image of Sirius Red–stained liver cross-section of experimental groups, “#” sign identifies the central hepatic veins. (B) Percentage incidence of advanced-stage fibrosis (Fibrosis Score >2) in the liver cross-section of the experimental groups at week 24. (C) Hepatic hydroxyproline concentration. and (D) Hepatic LPAR1 expression of the experimental groups at week 24. Chi-square has a significant p-value for all experimental groups compared to the HFHC group. Mean±SEM, ***p<0.001, *p<0.05. Abbreviations: HFHC, high-fat, high-carbohydrate; LPAR1, lysophosphatidic acid receptor 1; Sema, semaglutide.

B6.Cg-Lep^ob/J (ob/ob) mice gain similar body weight when fed either chow or AMLN diet

Genetically modified obese mice (ob/ob), mice either fed a chow diet or an amylin diet (AMLN), have similar body weight (chow: 61.802±0.936 vs. AMLN: 60.680±0.380, Figure 5A). Likewise, we did not observe any significant difference in the weight gain between the chow- and AMLN diet-fed group (chow: 25.3842±0.401 vs. AMLN: 28.076±1.588, Figure 5B).

FIGURE 5.

EPGN2154 and semaglutide reduce body weight in B6.Cg-Lep ^ob /J (ob/ob) (A) Body weight of 6- to 8-week-old male ob/ob mice fed AMLN and chow diet for 16 weeks. (B) Body weight change of ob/ob mice fed AMLN and chow from week 0 to week 16. (C) Body weight of ob/ob mice fed on AMLN diet from week 16 to week 24 with or without drug treatment. (D) Body weight change of ob/ob mice fed on AMLN diet from week 16 to week 24 with or without drug treatment. Mean±SEM. Abbreviations: AMLN, Amylin liver NASH; HFHC, high-fat, high-carbohydrate; Sema, semaglutide.

EPGN2154 reduces the body weight in B6.Cg-Lep^ob/J (ob/ob) mice

After 8 weeks of drug treatment, mice in AMLN+2154, AMLN+2154+Sem, and AMLN+Sem groups have shown lower body weight than the chow+Veh group (AMLN+2154: 64.188±0.594 g, p<0.001; AMLN+2154+Sem: 60.255±0.850 g; AMLN+Sem: 63.493±0.841 g vs. chow+Veh: 70.422±1.034 g) (Figure 5C). Mice in the AMLN+2154 group gained the least body weight compared to the AMLN+2154+Sem and AMLN+Sem groups (AMLN+2154: 4.789±1.040 g, p<0.05; AMLN+2154+Sem: 6.080±0.810 g; AMLN+Sem: 6.833±0.926 g vs. chow+Veh: 7.822±0.956 g) (Figure 5D).

EPGN2154 improves NAS in AMLN-fed ob/ob mice

After 8 weeks of treatment, the liver weight of AMLN+2154, AMLN+2154+Sem, and AMLN+Sem mice groups decreased compared to the AMLN+Veh group (AMLN+2154: 5.861±0.224 g; AMLN+2154+Sem: 4.812±0.203 g; AMLN+Sem: 5.562±0.240 g vs. AMLN+Veh: 6.419±0.206 g) (Figure 6A). The AMLN+2154, AMLN+2154+Sem, and AMLN+Sem mice groups have lower liver-to-body weight ratio (arbitrary unit, AU) compared to the 16-week sentinel AMLN-fed mice group (AMLN 16 week) (AMLN+2154: 0.091±0.003 AU, p<0.05; AMLN+2154+Sem: 0.080±0.003 AU; AMLN+Sem: 0.088±0.004 AU vs. AMLN 16 week: 0.104±0.002) (Figure 6B). Further, the histological analysis of the liver section (Figure 6C) revealed that AMLN+2154 mice have a lower inflammation and ballooning score than AMLN+Sem mice [Inflammation score: 1.571±0.202 (AMLN+2154) versus 2.3±0.153 (AMLN+Sem); p<0.05; Ballooning score: 0.286±0.184 (AMLN+2154) vs. 0.7±0.213 (AMLN+Sem)]. AMLN+2154 mice have a lower NAS than AMLN+Veh and AMLN+Sem groups (AMLN+2154: 4.571±0.430 vs. AMLN+Veh: 6.250±0.250, p<0.0001; vs. AMLN+Sem: 6.000±0.298, p<0.0001). There was no observed significant difference between the AMLN+2154 and AMLM+2154+Sem groups (4.125±0.295) (Figure 6D). AMLN+2154 group has a lower Oil Red O stain area than AMLN+Sem group (Figure 6E). AMLN+2154 has no significant difference with AMLN+2154+Sem in the Oil Red O stain area (Supplemental Figure S2, http://links.lww.com/HC9/A663).

FIGURE 6.

EPGN2154 improves hepatic fibrosis in AMLN-fed ob/ob mice. (A) Liver weight of ob/ob mice at week 24. (B) Liver-to-body weight ratio of ob/ob mice at week 24. (C) Representative image of hematoxylin and eosin–stained liver cross-section of experimental groups. (D) NAS of liver cross-section of experimental groups at week 24 NAS of the liver cross-section. (E) Oil Red O stain area percentage on the liver section of experimental groups at week 24. Mean ± SEM. ****p<0.0001, ***p<0.001, **p<0.01, *p<0.05. Abbreviations: AMLN, Amylin liver NASH; A.U., arbitrary unit; NAS, NAFLD Activity Score; Sema, semaglutide; Veh, vehicle.

EPGN2154 lowers the incidence of advanced-stage hepatic fibrosis in AMLN-fed ob/ob mice

Mice in AMLN+2154 group (14%) have shown a lower incidence of advanced-stage hepatic fibrosis (Fibrosis Score <2) than AMLN+Veh (100%, χ² p<0.0001), AMLN+Sem (100%, χ² p<0.0001), and AMLN 16-week sentinel (33%, χ² p<0.01). In addition, we observed a similar incidence of advanced-stage hepatic fibrosis in AMLN+2154+Sem compared to the AMLN+2154 group (Figure 7A, B). Immunophenotyping of the liver cross-section revealed that the AMLN+2154 group has a lower abundance of hepatic fibrosis markers like α-smooth muscle actin, galectin-3, collagen1a1, and laminin than that of the AMLN+Sem and AMLN+Veh groups (Figure 7C). The differences between the AMLN+2154 and AMLN+2154+Sem groups in α-smooth muscle actin, galectin-3, collagen1a1, and laminin abundance were insignificant. The AMLN+2154 mice have a lower concentration of hepatic hydroxyproline compared to AMLN+Veh, AMLN+2154+Sem, and AMLN+Sem groups (AMLN+2154: 0.72±0.018 ng/µL per mg liver vs. AMLN+Veh: 0.81±0.035 ng/µL per mg liver, p<0.05; vs. AMLN+2154+Sem: 0.97±0.075 ng/µL per mg liver, p<0.01; vs. AMLN+2154+Sem: 0.89±0.038 ng/µL per mg liver, p<0.01) (Figure 7D). AMLN+2154 mice did not significantly differ in the hepatic hydroxyproline concentration compared to the AMLN-fed 16-week sentinel group (0.81 ± 0.092, p>0.05).

FIGURE 7.

EPGN2154 improves hepatic fibrosis in AMLN-fed ob/ob mice by inhibiting macrophage migration and HSC proliferation. (A) Representative image of Sirius Red staining of liver cross-sections. (B) Hepatic fibrosis score of the liver cross-section of experimental groups; chi-square has a significant p-value for all experimental groups compared to the AMLN+Veh group. (C) Immunohistochemistry of the liver cross-section for αSMA, Gal-3, Col1, and Lam. (D) Hepatic hydroxyproline concentration of the experimental groups. (E) EPGN696 (IC₅₀ = 0.95 nM) inhibits migration of MCP-1-treated RAW264.7 cells (mouse macrophages) in a dose-dependent manner. (F) EPGN2154 (IC₅₀ = 1.76 nM) inhibits migration of MCP-1-treated RAW264.7 cells in a dose-dependent manner. (G) EPGN696 inhibits the LPA-stimulated proliferation of primary human HSCs in a dose-dependent manner, Mean±SEM, ****p<0.0001 ***p<0.001, **p<0.01, *p<0.05. Abbreviations: αSMA, α-smooth muscle actin; AMLN, Amylin liver NASH; Col1a1, collagen1a1; Gal-3, galectin-3; Lam, laminin; LPA, lysophosphatidic acid; MCP-1, monocyte chemoattractant protein‐1; mMCP-1, murine monocyte chemoattractant protein-1; Veh, vehicle.

LPAR1 antagonists, EPGN2154 and EPGN696, inhibit the migration of macrophages and the proliferation of HSCs

EPGN2154 and EPGN696 inhibited MCP-1–mediated RAW264.7 cell migration across a concentration range of 0.0001–10 μM (Figure 7E, F). The calculated IC₅₀ for this inhibition was 1.76 nM for EPGN2154 and 0.95 nM for EPGN696. The data obtained successfully demonstrated the inhibition of hepatic fibrosis by the LPAR1 antagonist, EPGN2154. Treatment of human HSCs with a prototypical LPAR1 antagonist, EPGN696, at 1 µM significantly reduced the proliferation of HSCs stimulated by 10μM LPA after 48 hours. No effect was observed with 0.010 µM EPGN696 ex vivo. These data demonstrate a dose-dependent response of EPGN696 on HSC proliferation (Figure 7G).

DISCUSSION

The FDA has approved Sema, a GLP-1 analog, for the treatment of obesity.^35,36 However, there continues to be a lack of FDA-approved therapeutics for obesity-related liver disease (NASH/NAFLD). Lysophosphatidic acid receptor (LPAR1) antagonists have shown antifibrotic effects in patients with idiopathic pulmonary fibrosis.^14,17 Our study demonstrated that a novel LPAR1 antagonist, EPGN2154, reverses NASH-related liver fibrosis. Both diet-induced WT and ob/ob mice, when treated with EPGN2154, have reduced liver inflammation like ALT, lower NAS, and lower incidence of advanced-stage fibrosis compared to the nontreatment group and, more interestingly, both in ob/ob mice and HFHC-fed WT mice, the EPGN2154 treatment groups had lower NAS and a lower incidence of advanced-stage fibrosis.

Although our studies used established preclinical models of NASH and obesity,^29,31 they had clear limitations, including the contrast between the response to Sema in the HFHC-fed WT mice and AMLN-fed ob/ob mice. The HFHC WT had a lower incidence of advanced-stage hepatic fibrosis with Sema treatment, while Sema-treated AMLN-fed ob/ob mice had neither weight loss nor prevention of advanced-stage hepatic fibrosis. Further, though the prototypical LPAR1 antagonist, EPGN696, has been shown to reduce HSC proliferation,^22,37,38 we could only demonstrate a potential mechanism of inhibition of human HSC proliferation with LPAR1 antagonist EPGN2154. Future studies will be required to further outline the detailed mechanism of action through genetically modified knockdown, overexpression, knockout models, and clinical studies.