Dual GIP/GLP-1 receptor agonist tirzepatide ameliorates hepatic steatosis and inflammatory responses in a MASLD mouse model associated with the CCL2/CCR2 axis Authors: Yongning Pan, Fanfan Song, Xuehua Liu, Yan Zhang, Juanjuan Zhang, Guangyao Song, Luping Ren Published: 2026 ABSTRACT AbstractBackground/objectivesMetabolic dysfunction-associated steatotic liver disease (MASLD) is a common multisystem chronic progressive liver disease with a rising burden. Tirzepatide (TZP), a dual agonist of glucose-dependent insulinotropic polypeptide (GIP) and glucagon-like peptide-1 (GLP-1) receptors, has shown beneficial hepatic effects, but the related molecular mechanisms remain unclear. The aim of this study was to investigate the hepatic molecular signatures associated with TZP in MASLD.MethodsA MASLD mouse model was established using a high-fat, high-fructose (HFHFr) diet. Male C57BL/8J mice (n = 32) were randomized to four groups: control (CON), HFHFr, HFHFr + semaglutide (Sema), and HFHFr + TZP. Hepatic transcriptomic and proteomic profiles were generated by RNA sequencing and liquid chromatography-mass spectrometry (LC-MS), respectively. Key molecular targets were validated by quantitative real-time PCR and immunoblotting of liver tissue.ResultsHFHFr feeding induced hyperglycaemia, increased HOMA-IR, elevated ALT/AST, and marked hepatic steatosis and inflammatory injury. Both Sema and TZP ameliorated these abnormalities. TZP was associated with lower hepatosomatic index, improved fasting glucose and HOMA-IR, and reduced hepatic MCP-1, IL-1β, TNF-α and GSDMD, with partial restoration of IL-10. Integrated liver transcriptomic-proteomic profiling highlighted chemokine signaling and PI3K-AKT pathway signatures to be associated with TZP. Further targeted validation showed that TZP treatment was associated with reduced hepatic CCL2/CCR2 axis components, and decreased PI3K abundance and lower AKT phosphorylation.ConclusionTZP ameliorated hepatic steatosis and attenuated inflammatory responses in HFHFr-induced MASLD and was associated with downregulation of the CCL2/CCR2-linked chemokine signaling and PI3K-AKT-related inflammatory responses.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12876-026-04986-0.Keywords: MASLD, Tirzepatide, Semaglutide, Mice, Steatosis, Inflammation Introduction Metabolic dysfunction-associated steatotic liver disease (MASLD) is the most common cause of chronic liver disease worldwide, affecting > 30% of adults and closely linked to the global rise in obesity and type 2 diabetes mellitus [1–3]. MASLD encompasses a disease spectrum ranging from hepatic steatosis to metabolic dysfunction-associated steatohepatitis (MASH), which can progress to fibrosis, cirrhosis and ultimately, hepatocellular carcinoma (HCC) [4]. It is a condition characterized by abnormal lipid accumulation in the liver. Diet is a key modifiable driver: fructose promotes hepatic de novo lipogenesis, lipid accumulation and insulin resistance [5, 6], whereas high-fat intake increases intrahepatic triglyceride content, perturbs lipid homeostasis and contributes to hepatocellular injury [7, 8]. Beyond liver-related outcomes, MASLD is increasingly recognized as a multisystem disorder associated with heightened risks of cardiovascular disease, chronic kidney disease, and extrahepatic malignancies [9, 10]. Although current management remains largely relies on lifestyle modification [11], this is frequently insufficient to achieve complete histological resolution, underscoring the need for mechanism-informed pharmacological strategies. The hepatic inflammation and immune dysregulation are pivotal in the progression from steatosis to MASH and fibrosis [12]. MASLD pathogenesis involves interlocking inflammatory and immune pathways, including hepatocyte stress signals, innate immune activation and altered cytokine/chemokine networks [13]. Among chemokines implicated in hepatic immune trafficking, C–C chemokine ligand 2 (CCL2; also known as monocyte chemoattractant protein-1 [MCP-1]) is a key mediator of monocyte recruitment through its receptor C–C chemokine receptor 2 (CCR2) [14–19]. In an experimental MASH model, inhibition of monocyte-derived macrophage recruitment using small-molecule CCR2/CCR5 antagonists attenuates steatosis and fibrosis [20]. Clinically, MASLD has been associated with elevated serum CCL2 levels that correlate with hepatic lipid accumulation [21]. These findings suggest that targeting the CCL2-CCR2 axis and related inflammatory circuits may offer a rational approach to decelerate the progression of MASLD. Concurrently, incretin-based therapies have expanded the therapeutic landscape for metabolic disease, including dual receptor agonists targeting glucagon-like peptide-1 (GLP-1) and glucose-dependent insulinotropic peptide (GIP) [22]. Studies have shown that these peptides are efficacious in the attenuation of hepatic steatosis in experimental animal models [23–25]. Semaglutide (Sema), a GLP-1RA, has shown promising clinical efficacy in the treatment of diabetic MASLD [26]. The findings reported by Newsome et al. suggest that Sema is effective in ameliorating the early stages of MASLD (steatosis and inflammation), whereas its impact on advanced fibrosis and cirrhosis remain to be established [26, 27]. Notably, a study using pair-fed, middle-aged low-density lipoprotein receptor knockout (Ldlr−/−) mice showed that the GLP-1 receptor agonist dulaglutide reduced hepatic steatosis and inflammatory signaling independent of body weight loss, implicating direct hepatic and adipose tissue mechanisms beyond caloric restriction [28]. Tirzepatide (TZP), a dual GIP/GLP-1 receptor agonist, was found to have a dose-dependent superiority over placebo, mono-receptor agonists and basal insulin in terms of glucose reduction and weight loss in a recent meta-analysis [29]. It has also been reported to attenuate liver injury and lipid dysregulation in a high-fat diet (HFD)-induced MASLD model [30]. Beyond preclinical evidence, recent clinical data support TZP’s efficacy in patients with MASH. In the phase II SYNERGY-NASH trial involving adults with biopsy-confirmed MASH and stage 2–3 fibrosis, once-weekly TZP (5, 10 or 15 mg) administered over 52 weeks led to statistically significant histological resolution of MASH without worsening of fibrosis across all evaluated doses compared with placebo, accompanied by a greater proportion of patients achieving fibrosis improvement [31].Consistent histological benefits were also observed across a range of clinically relevant subgroups in post hoc analyses [32]. Collectively, these data position TZP as a compelling therapeutic strategy for MASH. Although our previous study showed that TZP improved systemic metabolic dysfunction and altered metabolism- and inflammation-related gene expression in brown adipose tissue, its liver-specific mechanisms of action in MASLD remain unclear [33]. Therefore, the aim of our study was to investigate the molecular mechanisms of TZP-mediated improvement in MASLD using a high-fat, high-fructose (HFHFr) diet-induced mouse model, with a focus on pathways associated with hepatic lipid accumulation and inflammatory responses. Materials & methods Animal model The present study was performed under the same animal ethics approval and within the same overall in vivo experimental framework as our previous study focused on brown adipose tissue [33]. The animal sets used in the two studies were therefore not fully independent and included substantially overlapping animals, with minor differences in individual inclusion across specific analyses. Several systemic phenotypic endpoints, including body weight, intraperitoneal glucose tolerance test (IPGTT)-derived measures and its associated AUC, fasting insulin and hepatic triglyceride content, arose from this shared experimental framework. To ensure transparency, these overlapping endpoints are clearly disclosed in the Supplementary Materials and are provided for contextual interpretation only. The principal analyses and conclusions of the present study are based on liver-specific pathological, molecular and omics data. The male C57BL/6J mice (7 weeks old, ~ 20–25 g) were purchased from Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). All mice were housed in an animal laboratory facility under controlled environmental conditions (relative humidity 50%; temperature 24 ± 3 °C; 12-hour light-dark cycle). After one week of acclimatization, mice were randomly divided into two groups: a normal diet group (CON group, n = 8; D12450B, 10% kcal fat, 70% kcal carbohydrates and 20% kcal protein; Research Diets, Inc., New Brunswick, USA) and a HFHFr diet group (HFHFr group, n = 24; M18011801, 40% kcal fat, 34.5% kcal fructose, and 2% cholesterol purchase from Biopike Biotechnology Co., Ltd., Sichuan, China), with four to five mice assigned to each cage. After a 7-week feeding period, the mice in the HFHFr group were divided into three subgroups (n = 8 per group) using a random-number table: (i) HFHFr model group; (ii) HFHFr + Sema group (Sema 30 nmol/kg/3d, subcutaneous injection); Sema (Novo Nordisk, Copenhagen, Denmark) administered for 7 weeks; and (iii) HFHFr + TZP group (TZP 10 nmol/kg/3d, subcutaneous injection; TZP (Abmole Biotechnology Co., Ltd.) administered for 7 weeks [34]. Mice in both the CON and HFHFr groups received an equal volume of 0.9% NaCl solution via subcutaneous injection. At week 14, mice were fasted for 12 h and anaesthetized by an intraperitoneal injection of sodium pentobarbital (40 mg/kg). Blood samples were collected from the retro-orbital venous plexus. Mice were then euthanized by cervical dislocation and liver tissue were removed and stored in a refrigerator at − 80 °C for subsequent analysis. Food intake was measured every three days throughout the intervention. For each mouse, the amount of food provided and the remaining food, including visible food fragments, were weighed at each measurement point. Food intake was calculated as the difference between food provided and food remaining, and values were expressed as grams of food consumed per mouse over each 3-day interval. Intraperitoneal glucose tolerance test (IPGTT) An intraperitoneal glucose tolerance test (IPGTT) was performed at week 14. Mice were fasted for 12 h before receiving an intraperitoneal injection of a glucose solution at a dosage of 2 g/kg. Baseline blood glucose (0 min) was measured from a tail vein using a German Roche Rokan total glucometer. Blood glucose concentrations were subsequently determined at 15, 30, 60, and 120 min post-injection. Biochemical analysis Serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were measured using a fully automatic biochemical analyzer (TBA-FX8; Tokyo Canon, Japan). Hepatic triglyceride (TG) concentration in liver tissue was determined by a TG assay kit (Single Reagent, GPO-PAP Method, Nanjing Jiancheng Biotechnology Research Institute, China). Briefly, ~ 20 mg of snap-frozen liver tissue (collected from the same lobe across animals) was weighed, homogenized in [buffer/solvent, 9 times volumes] and processed for TG determination. TG was measured colorimetrically against a kit-provided standard curve and normalized to the wet mass of the liver tissue used for the assay, and are presented as mmol per g wet tissue. Serum tumor necrosis factor (TNF-α), monocyte chemotactic protein-1 (MCP-1) and interleukin-1β (IL-1β) were measured using a multifactor detection kit (flow cytometry; Wuhan Yunklon Technology Co., Ltd., China) and a Luminex analyzer. Gasdermin D (GSDMD) was detected by a chemiluminescence assay (Beijing Meide Taikang Biotechnology Co., Ltd., China). Fasting insulin (FINS) levels were measured using an ELISA kit (Wuhan Ilerite Biotechnology Co., Ltd., China). All assays were performed according to the manufacturer’s instructions. Insulin resistance (IR) was estimated by calculating the HOMA-IR (Homeostatic Model Assessment of Insulin Resistance) using the following formula: HOMA-IR = fasting blood glucose (FBG, mmol/L) × FINS (mIU/L) / 22.5. Histopathological staining Fresh liver tissues were fixed in 4% paraformaldehyde and then embedded in paraffin. The paraffin-embedded tissue blocks were sectioned into 5 μm-thick slices with a microtome. Slices were dewaxed, rehydrated, and stained with hematoxylin and eosin (H&E). Images were captured using an ECLIPSE Ci-L microscope (Nikon, Tokyo, Japan) under identical illumination and exposure settings. Frozen liver tissues were embedded with O.C.T compound, cut into sections of 10 μm thickness using a cryostat, and subjected to Oil Red O staining (Servicebio Technology Co., Ltd., Catalog numbers: G1015) to evaluate lipid accumulation. For histological quantification, one frozen liver section was analyzed per animal. Three non-overlapping regions of interest were imaged from each section under identical acquisition settings. Lipid accumulation was quantified using Image Pro Plus 6.0 software (Media Cybernetics, Rockville, MD, USA) by calculating the lipid-positive area as a percentage of total tissue area, using a uniform threshold applied across all samples. For each animal, the values from the three regions were averaged to generate a single animal-level value, which was used for subsequent statistical analysis. H&E-stained liver sections were further evaluated using the NAFLD activity score (NAS) system to provide a standardized semi-quantitative assessment of histological disease activity. For each mouse, one liver section was evaluated by a member experienced in liver histopathology who was blinded to treatment allocation. The entire tissue section was reviewed before scoring to minimize field-selection bias. The NAS was calculated as the unweighted sum of the steatosis, lobular inflammation and hepatocellular ballooning scores. Steatosis was scored from 0 to 3 according to the percentage of hepatocytes containing with macrovesicular steatosis: 0, < 5%; 1, 5%–33%; 2, > 33%–66%; and 3, > 66%. Lobular inflammation was scored from 0 to 3 according to the number of inflammatory foci per 200× field: 0, none; 1, fewer than 2 foci; 2, two to four foci; and 3, more than 4 foci. Hepatocellular ballooning was scored from 0 to 2: 0, none; 1, a few ballooned cells; and 2, many ballooned cells or prominent ballooning. The total NAS ranged from 0 to 8. Each mouse contributed one NAS value at the animal level for statistical analysis. Fibrosis was not included in the NAS. RNA-sequencing analysis Total RNA was extracted from liver tissue using TRIzol reagent (Life Tech nology). RNA concentration and purity were assessed by NanoDrop 2000 spectrophotometer (Thermo Scientific, USA), and RNA integrity was evaluated using Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA); only samples with RIN ≥ 8 were used for sequencing. Liver samples from n = 3 biologically independent mice per group were analyzed. Poly(A) + RNA was enriched from total RNA and libraries were prepared using a standard Illumina-compatible protocol. The library was sequenced on an Illumina Novaseq 6000 platform (OE Biotech, Shanghai, China) to generate 150 bp paired-end reads. Raw reads were processed to remove adapters and low-quality reads before alignment. The effective clean data yield ranged from 6.89 to 7.06 Gb per sample. Clean reads were aligned to the reference genome (GRCm38) using HISAT2 software, with total genome-mapping rates ranging from 98.88% to 99.47% across samples. Per-sample sequencing depth and alignment statistics are provided in Supplementary Table S1. Gene-level read counts were generated using HTSeq-count based on the corresponding GRCm38 gene annotation. Differential expression analysis was performed using raw gene-level count data. No additional pre-analysis filtering based on an arbitrary low-count threshold was applied, and all genes with available count information were retained for statistical analysis. FPKM values were calculated only for descriptive expression visualization and were not used for differential expression testing. Sample relations were assessed by principal component analysis (PCA) based on all expressed genes in R (version 3.2.0). All RNA samples were processed, library-prepared and sequenced within the same experimental batch. Consistent with this design, PCA did not reveal an apparent technical batch-associated separation among samples. Therefore, no batch covariate was included in the differential expression model, and no additional batch-correction procedure was applied. Differentially expressed genes (DEGs) were identified using DESeq software, defined as those with a q-value ≤ 0.05 and |log₂ fold-change| ≥ 1. Functional enrichment analysis on the identified DEGs was performed for Gene Ontology (GO) terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways using a hypergeometric distribution algorithm. Statistical power was estimated using RNASeqPower (target power = 0.8). All visualizations were produced in R. Gene set enrichment analysis (GSEA) was performed using the JavaGSEA platform, and predefined gene sets were used for analysis. According to the GSEA user guide, the downregulated pathways in group A (control group) were regarded as the upregulated pathways in group B (intervention/model group), and vice versa. In accordance with GSEA recommendations, statistical significance was set at nominal p < 0.05 and FDR q < 25%. Proteomic analysis Protein extraction and digestion Liver tissues were lysed in lysis buffer containing protease inhibitors [0.1 M Tris-HCl, pH 8.0; 0.1 M dithiothreitol (Sigma, 43815), 1 mM phenylmethylsulfonyl fluoride (Amresco, M145)] and sonicated on ice for 1 min. The supernatant was collected and then solubilized with 4% sodium dodecyl sulfate (SDS). The mixture was centrifuged at 12,000× g for 5 min, and the supernatant was precipitated with acetone. Protein pellets were washed three times with pre-chilled acetone and aspirated using an AQ-type vacuum concentrator (Eppendorf, Germany). Proteins were then suspended in 200 µL of 8 M urea (pH 8.0) and quantified by Bradford assay, followed by separation using SDS-PAGE gel electrophoresis at a constant voltage of 120 V for 60 min. Afterward, the gel pieces were excised, destained with 30% acetonitrile and 100 mmol/L ammonium bicarbonate until clear, and dried in a vacuum centrifuge. For each sample, 200 µg of protein was reduced with 10 mM dithiothreitol for 30 min at 56 °C and subsequently alkylated with 10 mM iodoacetamide for 30 min at room temperature in the dark. The sample containing 200 µg protein was combined with 8 M urea bicarbonate (200 µL), loaded onto the filter unit, vortexed and spun at 14,000 rcf for 20 min. The protein suspension was digested overnight at 37 °C with trypsin (Promega; 4 µg) in 50 mM NH4HCO3 buffer (100 µL). The resulting peptides were collected as the filtrate and vacuum dried (Concentrator Plus, Eppendorf). LC-MS analysis Peptides from proteome analysis samples were analyzed using an Orbitrap mass spectrometer (Thermo Fisher Technology Co., Ltd.) coupled to a high-performance liquid chromatography system (EASYnLC 1200, Thermo Fisher Technology Co., Ltd.). Samples were first loaded onto a trap column (inner diameter 100 μm, Dr. Maisch GmbH) and subsequently separated on an analytical column packed with 1.9 μm ReproSilPurC18-AQ beads (Dr. Maisch GmbH) at a flow rate of 600 nL/min over a 120-min gradient. Mobile phases consisted of 0.1% aqueous formic acid (Solvent A) and 0.1% acetonitrile (ACN) with formic acid (Solvent B). Eluted peptides were ionized at 2.4 kV and then analyzed on the Orbitrap. Full MS (MS1) spectra were acquired over mass-to-charge ratios (m/z) between 300 and 1400 at a high resolution of 120,000, with the automatic gain control (AGC) target of 3 million and the maximum injection time of 80 milliseconds. MS/MS (MS2) acquisition was performed in fast mode with a 1-s cycle time. FAIMS compensation voltages were set to − 45 V and − 65 V. Data were processed for qualitative and quantitative analyses using the iProteome one-stop cloud-based platform. Quantitative real-time PCR Candidate genes for qRT-PCR validation were selected to evaluate a pathway-level mechanism suggested by the transcriptomic analysis, rather than on the basis of DEG ranking alone. The representative nodes of the CCL2/CCR2–PI3K–Akt inflammatory signaling axis were prioritized, including upstream chemokine-related components, downstream PI3K–Akt signaling mediators and inflammatory effector genes. This selection was based on pathway-level enrichment findings, biological relevance to hepatic inflammation in MASLD and previous evidence implicating the CCL2/CCR2–PI3K–Akt pathway in experimental NASH-associated liver inflammatory responses [35]. Total RNA was isolated using the Total RNA Reagents 2.0 kit following the manufacture’s instructions, with sterile enzyme-free reagent consumables used throughout. The quantification of RNA was performed using a NanoDrop 2000 spectrophotometer (Thermo Scientific, USA), and integrity was assessed through agarose gel electrophoresis with ethidium bromide staining. The RNA was then reverse transcribed into complementary DNA (cDNA) using the TransScript All-in-One First-Strand cDNA Synthesis SuperMIX for qPCR. Real-time PCR was performed in accordance with the instructions provided with the PerfectStart™ Green qPCR SuperMix Kit (TransGen Biotech, Beijing, China) utilizing the LightCycler® 480 II Fluorescent Quantitative PCR System (Roche, Switzerland). GAPDH served as the internal reference. Primer sequences were designed based on NCBI mRNA sequences and synthesized by TsingKe Biotech; sequences are outlined in Table 1. Table 1The specific primers for the target genesNumGeneForward primer (5′–3′)Reverse primer (5′–3′)1 GAPDH GCAAGGACACTGAGCAAGAGGATGGAAATTGTGAGGGAG2 Ccl2 TTTGTCACCAAGCTCAAGAGATCTCATTTGGTTCCGATCC3 Ccr2 ACTAACATAGACAGCTCAGGACCTTAGAGTCTAAATCACCATGC4 Pik3r5 TTTGCTAGGACCTGGTACATTGAAGTGCCTTAACTTCCTCT5 Akt1 GAGAACCTCATGCTGGACAAAGAATGTCTTCATAGTGGCAC6 Tnf AATGGGTGTTCATCCATTCTGGCTCTGAGGAGTAGACAATAA7 Il1b TGTCCTGTGTAATGAAAGACGGCTCTTGACTTCTATCTTGTTG The specific primers for the target genes The initial qRT-PCR validation included n = 3/4 biological samples per group, with each sample measured in technical triplicate. To increase the biological coverage of the qRT-PCR validation, an additional qRT-PCR assay batch was performed using the remaining available biological samples from the same animal cohort. In the additional qRT-PCR batch, each remaining biological sample was measured once because of limited remaining material. As the initial and additional qRT-PCR assays were performed in separate batches, the two datasets were analyzed and presented separately rather than pooled for a single statistical analysis. Relative expression was calculated using the comparative Ct method, and fold changes were presented as 2−ΔΔCt values. Statistical comparisons were performed within each qRT-PCR batch. Protein isolation and western blot analysis Total liver protein lysates were prepared from approximately 20 mg of frozen liver tissue. Samples were homogenized on ice in RIPA lysis buffer supplemented with protease and phosphatase inhibitor cocktails (Beijing Solarbio Science & Technology Co.,Ltd.) using a mechanical tissue homogenizer. Lysates were incubated on ice for 30 min and then centrifuged at 12,000× g for 15 min at 4 °C to remove insoluble debris. The supernatants were collected as total protein extracts, and protein concentrations were determined using a Bradford protein assay kit (Wuhan Boster Biological Technology Co., Ltd.Cat. No. AR0145). Equal amounts of proteins were separated by 10% SDS-PAGE gel electrophoresis and transferred to polyvinylidene fluoride (PVDF) membranes. Membranes were blocked with a 5% skim milk solution for 2 h and incubated with primary antibodies overnight at 4 °C. Primary antibodies and working dilutions were as follows: CCL2/MCP-1 (1:500–1:2000, 29548, Signalway Antibody, USA); CCR2 (1:500–1:2000, 36784 Signalway Antibody, USA); PI3K (1:500–1:2000, 39102 Signalway Antibody, USA); AKT (1:500–1:1000, 21054, Signalway Antibody, USA); phospho-Ser 473 AKT (1:500–1:1000, 11054, Signalway Antibody, USA), and GAPDH antibody (1:3000–1:30,000, AF7021, Affinity Biosciences, USA). After three washes with diluted TBS-Tween, membranes were incubated with the corresponding secondary antibody (BA1054, Boster Biological Technology, Pleasanton, USA) for 1 h at room temperature. Signals were captured with a fluorescence imaging system (FluorChem FC2, Alpha, Shanghai). Band intensities were quantified using Image J software (Version 1.8.0) and target protein levels were normalized to GAPDH and expressed as target/GAPDH ratios. The initial Western blot validation batch included [n = 3] biological samples per group. To increase the biological coverage of the protein-level validation, additional Western blot assays were performed using the remaining available biological samples from the same animal cohort. Because the additional Western blot assays were performed in a separate experimental batch, the initial and additional Western blot datasets were quantified, analyzed, and presented separately rather than pooled for a single statistical analysis. Statistical comparisons were performed within each Western blot batch. Statistical analysis Data analysis was conducted using SPSS (version 26.0). Before inferential testing, within-group distributions were assessed for normality using the Shapiro–Wilk test, and homogeneity of variances was assessed using Levene’s test. Normality was assessed using the Shapiro–Wilk test and inspection of Q–Q plots. No statistical outlier-exclusion procedure was applied; all valid measurements were retained unless attributable to a predefined technical or recording error. Data are presented as mean ± standard deviation (Mean ± SD) unless otherwise stated. For normally distributed variables, between-group differences were assessed using one-way ANOVA. Post hoc pairwise comparisons were performed using Fisher’s LSD test when variances were homogeneous and Dunnett’s T3 test when variances were unequal. For variables showing mild departures from normality, Kruskal–Wallis tests were additionally performed as sensitivity analyses. TNF-α, IL-10 and hepatic triglyceride levels showed mild departures from normality; the Kruskal–Wallis sensitivity analyses di not alter the statistical interpretation of the primary ANOVA results. A two-sided p < 0.05 indicated statistical significance. Results Tirzepatide ameliorates HFHFr diet-induced weight gain and blood glucose Representative gross images of mice and excised livers from each group are presented in Fig. 1A. Food intake, assessed every 3 days during the intervention, did not differ significantly among groups, although it varied over the course of treatment (Fig. 1B). HFHFr feeding increased fasting glucose and HOMA-IR versus CON, whereas treatment with Sema or TZP reduced both parameters, with TZP showing a greater improvement in HOMA-IR (Fig. 1C and D). Body weight, IPGTT-related measures and fasting insulin are shown in Supplementary Figure S1A-E for transparency and contextual interpretation, as these endpoints overlap with those associated with our previous BAT-focused study within the same overall experimental framework [33]. The subsequent analyses therefore centered on liver-specific pathological and molecular features relevant to MASLD. Sema and TZP improve liver-relevant phenotypes in HFHFr-fed mice (n = 8). A Representative images of body size and liver tissue of mice in each group; (B) Food intake, recorded for each mouse every 3 days during the 7-week intervention period; (C, D) Serum fasting blood glucose and HOMA-IR values. Body weight, IPGTT-related measures, and fasting insulin, which overlap with those associated with our previous BAT-focused study [32] generated within the same overall experimental framework, are presented in Supplementary Figure S1 for transparency and phenotypic context. CON, normal diet; HFHFr, high-fat high-fructose diet; Sema, Semaglutide; TZP, Tirzepatide; FBG, fasting blood glucose; HOMA-IR, Homeostatic Model Assessment of Insulin Resistance. Data were analyzed using one-way ANOVA and expressed as mean ± SD. *P < 0.05, **P < 0.01,***P < 0.001,****P < 0.0001 Effects of tirzepatide on steatosis and liver injury in HFHFr-induced MASLD mice H&E staining showed relatively preserved hepatic architecture in CON mice, whereas HFHFr feeding was associated with more prominent hepatocellular lipid accumulation, balloon-like degeneration, hepatocellular punctate necrosis and lobular inflammatory foci. Both Sema and TZP attenuated these histopathological lesions and reduced hepatic lipid accumulation, as confirmed by Oil Red O staining (Fig. 2A). NAS-based semi-quantitative scoring, incorporating steatosis, hepatocellular ballooning and lobular inflammation, further supported these histological findings. NAS values were higher in HFHFr-fed mice and were significantly reduced by both Sema and TZP treatment (Fig. 2C). In parallel, the livers of mice were weighted at the end of the experiment and the hepatosomatic index was calculated. HFHFr feeding significantly increased the hepatosomatic index, which was reduced by both Sema and TZP (Fig. 2B). Serum ALT and AST levels were also increased in HFHFr-fed mice, indicating hepatocellular injury, and were significantly lowered by Sema and TZP (Fig. 2D and E). Hepatic triglyceride content is presented in Supplementary Figure S1F for transparency and phenotypic context as this endpoint overlaps with our previous work [33]. TZP and Sema attenuates HFHFr-induced hepatic steatosis and injury. A Representative images of HE-stained and oil red O-stained liver sections and quantitative analysis results of oil red O-staining (n = 3). B Hepatosomatic index, (C) NAFLD activity score (NAS; n = 3 mice per group), (D) serum ALT and AST (E), GSDMD (F), MCP-1 (G), IL-1β (H), TNFα (I), and IL-10 (J) (n = 8). Data were analyzed using one-way ANOVA and expressed as mean ± SD. *P < 0.05, **P < 0.01,***P < 0.001,****P < 0.0001. CON, normal diet; HFHFr, high-fat, high-fructose diet; Sema, Semaglutide, TZP, Tirzepatide;; HE, ematoxylin- Eosin; Oil red O, Oil red O staining; LDL, low-density lipoprotein; ALT, alanine aminotransferase; AST, aspartate aminotransferase; GSDMD, Gasdermin D, key protein for cellular pyroptosis; MCP-1, monocyte chemotactic protein-1; IL-1β, interleukin 1β; TNFα, tumor necrosis factor; IL-10 interleukin 10 Tirzepatide attenuates HFHFr-induced inflammation in MASLD mice HFHFr feeding increased the levels of GSDMD and pro-inflammatory mediators (MCP-1, IL-1β and TNFα) relative to CON, and these markers were all decreased by Sema and TZP, with TZP generally achieving a greater reduction than Sema (Fig. 2F-I). Conversely, IL-10 was markedly reduced by HFHFr and was partially restored by both treatments, with TZP increasing IL-10 toward control levels (Fig. 2J). Effects of tirzepatide on the transcriptional profiling of the liver in MASLD mice To define hepatic transcriptional programs associated with HFHFr-induced MASLD and their modulation by incretin-based therapy, we performed bulk RNA sequencing of liver tissue from CON, HFHFr, Sema and TZP groups. We then used a reversal-oriented analytical framework, prioritizing HFHFr-induced genes that were reduced after treatment, to identify transcriptional features most closely aligned with phenotypic improvement. The PCA of DEGs showed clear separation among the CON, HFHFr, Sema, and TZP groups, indicating distinct transcriptomic profiles across groups (Fig. 3B). Compared to CON, HFHFr feeding group exhibited a total of 1054 DEGs, including 633 upregulated and 421 downregulated genes (Fig. 3C). Relative to HFHFr, Sema treatment was associated with 742 DEGs (376 upregulated and 366 downregulated), whereas TZP treatment was associated with 681 DEGs (272 upregulated and 409 downregulated) (Fig. 3C-E). Overlap analysis further clarified the shared and treatment-specific components of the transcriptional response (Fig. 3A). Effects of Sema and TZP on hepatic gene expression in HFHFr-induced MASLD mice(n = 3–4 per group). A Venn diagrams showing overlap of differentially expressed genes (DEGs) across the indicated pairwise comparisons; (B) Principal component analysis (PCA) scatter plots of differential genes in each group; Volcano plots of the (C) HFHFr versus CON, (D) Sema versus HFHFr, and (E) TZP versus the HFHFr. Gene Ontology (GO) enrichment analysis performed using directionally defined DEG sets: (F) GO terms enriched among genes upregulated in HFHFr vs. CON; (G) GO terms enriched among genes downregulated in Sema vs. HFHFr; and (H) GO terms enriched among genes downregulated in TZP vs. HFHFr. KEGG pathway enrichment analysis using the same directionally defined DEG sets: (I) pathways enriched among genes upregulated in HFHFr vs. CON; and (J) pathways enriched among genes downregulated in Sema vs. HFHFr; and (K) pathways enriched among genes downregulated in TZP vs. HFHFr. L Heatmap of selected pathway-associated DEGs; colours represent per-gene z-scored expression across samples (− 3 to 3), with negative and positive values indicating lower and higher expression relative to the gene-wise mean, respectively. Gene set enrichment analysis (GSEA) on ranked gene lists for (M) HFHFr versus CON, (N) Sema versus HFHFr and (O) TZP versus HFHFr Enrichment analyses were then organized according to the direction of change. GO and KEGG analyses were performed separately for upregulated and downregulated DEG sets for each comparison. Genes upregulated in HFHFr versus CON were enriched for immune- and inflammation-related GO terms and KEGG pathways, including immune response, positive regulation of IL-6 production, cytokine binding, IL-1 receptor binding, chemokine signaling, and PI3K-Akt signaling (Fig. 3F and I). In treatment comparison, the TZP-downregulated DEG set relative to HFHFr was enriched for immune-related GO terms, lysosomal and endoplasmic-reticulum-associated components, oxidoreductase activity, MHC class II protein complex binding, chemokine signaling, and PI3K-Akt signaling (Fig. 3H K). Together, these results indicate that TZP was associated with reduced expression of a treatment-downregulated DEG set enriched for inflammatory and chemokine-related pathway features. Sema also modified the HFHFr-associated transcriptomic profile; however, its downregulated gene set was more prominently enriched for metabolic terms, including tryptophan catabolism and fatty acid metabolism, and the chemokine and PI3K-Akt signal was less apparent than that observed in the TZP comparison (Fig. 3G and J). Enrichment analyses of treatment-upregulated gene sets are provided in Supplementary Figure S2. A heatmap of selected pathway-associated DEGs, scaled by per-gene z-score, further illustrated treatment-associated attenuation of several HFHFr-related expression patterns, with a cleaere pattern in the TZP group (Fig. 3L). GSEA was performed on ranked gene lists to asses pathway-level trends without restricting to DEGs alone. Compared to the CON group, the HFHFr group showed enrichment of several inflammation-related pathways (Fig. 3M), including chemokine signaling, neutrophil extracellular trap formation, and natural killer cell-mediated cytotoxicity. These pathways were upregulated in the HFHFr group, but showed downregulation in the TZP-treated and Sema-treated groups (Fig. 3N and O). The FDR value of the chemokine signaling pathway in the TZP group was 0.26, which is slightly above the commonly used threshold of 0.25, but its direction was concordant with the DEG-based KEGG analysis. In contrast, this pathway showed no significant change in the Sema group. Overall, the GSEA results were directionally consistent with the KEGG enrichment analyses. Tirzepatide treatment altered the proteomic characteristics of the liver in MASLD mice Next, the label-free quantitative proteomics were performed to characterize the alterations in protein levels in MASLD. A total of 6536 proteins was identified and quantified using LC-MS/MS. A Venn diagram summarized the overlap of all quantified proteins across the CON, HFHFr, Sema and TZP groups (Fig. 4A). PCA showed distinct proteomic profiles between CON and HFHFr, and between HFHFr and those in the Sema and TZP treatment groups, with majority of the variance explained by the first two principal components (Fig. 4B–D). Differential expression analysis (screening conditions: FC > 1.5, p < 0.05) identified 137 increased and 386 decreased proteins in HFHFr versus CON (Fig. 4E), 281 increased and 160 decreased proteins in Sema versus HFHFr (Fig. 4F), and 92 increased and 58 decreased proteins in TZP versus HFHFr (Fig. 4G). Effect of Sema and TZP on hepatic protein expression in HFHFr-induced MASLD mice (n = 4). A Venn diagram of all quantified proteins across CON, HFHFr, Sema and TZP groups. PCA scatter plots of differential proteins between (B) CON and HFHFr groups, (C) Sema and HFHFr groups, and (D) TZP and HFHFr groups. Volcano plots showing differentially expressed proteins in (E) HFHFr vs. CON, (F) Sema vs. HFHFr, and (G) TZP vs. HFHFr. H Trend clustering analysis of proteins from the indicated three-group comparisons across CON, HFHFr and treatment groups. The cluster includes proteins with higher abundance in HFHFr relative to CON and lower abundance after treatment relative to HFHFr. I, J GO enrichment analyses of the protein cluster shown in (H) for the CON-HFHFr-TZP and CON-HFHFr-Sema comparisons, respectively. K, L KEGG pathway enrichment analysis of the same cluster shown in (H) for the CON-HFHFr-TZP and CON-HFHFr-Sema comparisons, respectively. M Integrated KEGG pathway analysis of transcriptomic and proteomic changes in the CON-HFHFr-TZP comparison based on the reversal-pattern cluster identified in (H), defined as molecules increased in HFHFr versus CON and decreased after TZP treatment versus HFHFr. The Y-axis represents the number of differential genes or proteins For the three-group comparisons, proteins were clustered according to their expression patterns across CON, HFHFr and treatment groups, and GO and KEGG analyses were focused on the HFHFr-induced, treatment-attenuated cluster, defined as proteins showing a higher level in HFHFr versus CON and a lower level after treatment versus HFHFr (Fig. 4H). In the CON-HFHFr-TZP comparison, this cluster was enriched for GO terms involved in response to endoplasmic reticulum stress, regulation of proteolysis, positive regulation of lymphocyte chemotaxis, ribonucleoprotein granule, rough endoplasmic reticulum, oxidoreductase activity and platelet-derived growth factor receptor binding (Fig. 4I). KEGG analysis of the same cluster identified enrichment for thiamine metabolism, valine, leucine and isoleucine degradation, mTOR signaling, natural killer cell-mediated cytotoxicity, JAK-STAT signaling and chemokine signaling (Fig. 4K). In the CON-HFHFr-Sema comparison, enriched GO terms for the corresponding reversal-like pattern cluster were for peptide biosynthesis process, translation, amide biosynthetic process, ribosomal subunit, ribosomes and cytoplasmic regions, structural constituent of ribosome, mRNA binding and translation initiation factor activity (Fig. 4J). Meanwhile, enriched KEGG terms for the same cluster were related to protein processing in endoplasmic reticulum, ribosome, pyruvate metabolism, thiamine metabolism, valine, leucine and isoleucine degradation and citrate cycle were among the enriched pathways in the CON-HFHFr-Sema comparison (Fig. 4L). GO and KEGG enrichment analyses of proteins clusters showing treatment-increased abundance patterns from CON-HFHFr-Sema and CON-HFHFr-TZP comparisons are shown in Supplementary Figure S3. Together, these proteomic data indicate that TZP treatment was associated with reduced abundance of a subset of HFHFr-induced proteins involved in inflammatory, chemokine-related, and stress-response pathways. An integrated analysis revealed the key pathways by which tirzepatide regulates MASLD Integrated transcriptomic and proteomic analyses were performed to identify candidate pathways associated with the hepatic response to TZP. The HFHFr-induced, TZP-attenuated protein cluster defined above were used as the proteomic basis for pathway-level integration with the transcriptomic results (Fig. 4H). KEGG enrichment analysis of this protein cluster highlighted several pathways, including chemokine signaling, PI3K-Akt signaling and mTOR signaling (Fig. 4K). We then integrated transcriptomic and proteomic analysis using KEGG annotations. The integrated analysis showed that these pathways were represented in both omics layers (Fig. 4M). For example, within the chemokine signaling pathway, annotated proteins were detected at both the protein (Rps6, Sect. 13, Pik3r1, Ptk2b and Aox3) and transcript (Elmo1, Fgr, Gnai1, Pik3r5, Ntrk1, Syk, H2-Ab1 and H2-Aa) levels. These findings provide a descriptive overview of pathway-level overlap between datasets. Effects of tirzepatide on CCL2/CCR2 axis-related gene and protein expression in HFHFr-induced MASLD mice To follow up the transcriptomic and proteomic signals linked to the CCL2/CCR2 axis, the key associated transcripts were quantified using qRT-PCR. In the initial qRT-PCR validation batch, HFHFr-fed mice showed higher hepatic mRNA expression of CCL2 and its receptor CCR2 compared with CON mice (Fig. 5A and B). In parallel, transcripts encoding downstream signaling-related components, including PI3KR5 and Akt1, were also increased in HFHFr versus CON (Fig. 5C and D). Consistently, mRNA expression of the pro-inflammatory cytokines IL-1β and TNF-α was likewise elevated in HFHFr-fed mice (Fig. 5E and F). In contrast, TZP treatment was associated with lower expression of these genes compared with HFHFr mice (Fig. 5A-F). The additional qRT-PCR data from the remaining available biological samples are shown separately in Supplementary Figure S4 and showed directionally consistent expression patterns. These findings provide supplementary support for the proposed involvement of the CCL2/CCR2 inflammatory signaling axis. Initial qRT-PCR validation of selected CCL2/CCR2 inflammatory signaling genes in HFHFr-fed mice (n = 3). Hepatic mRNA expression of (A) Ccl2, (B) Ccr2, (C) Pik3r5, (D) Akt1, (E) Il1β and (F) Tnfα in CON, HFHFr and TZP mice. Data were analyzed using one-way ANOVA and expressed as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 Consistent with the transcriptomic and qRT-PCR findings, the initial western blot analyses showed higher protein levels of CCL2, CCR2 and PI3KCG, together with increased AKT phosphorylation (p-AKT), in HFHFr-fed mice compared with CON mice. TZP treatment was associated with lower levels of these proteins and reduced p-AKT relative to HFHFr mice (Fig. 6A-E). Additional Western blot data from the remaining available biological samples are presented separately in Supplementary Figure S5. The selected protein targets showed trends that were directionally consistent with the initial Western blot results. These data provide supplementary protein-level support for the proposed involvement of the CCL2/CCR2 signaling axis. Initial Western blot validation of CCL2/CCR2 signaling-related proteins in HFHFr-fed mice (n = 3). A Western blotting analysis of CCL2, CCR2, PI3K, AKT and phosphorylated AKT (p-AKT) in CON, HFHFr and TZP-treated mice; GAPDH, loading control. Quantification of (B) CCL2, (C) CCR2, (D) PI3KCG, and (E) p-Akt, normalized to GAPDH and expressed relative to CON. Data were analyzed using one-way ANOVA and expressed as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 Discussion To evaluate the impact of TZP on hepatic lipid accumulation in MASLD, we established a MASLD mouse model using a HFHFr diet and evaluated liver-centered pathological, biochemical, and multi-omic responses to treatment. After 14 weeks of feeding, the HFHFr-fed mice showed markedly increase in body weight, fasting blood glucose, insulin and HOMA-IR compared with controls, accompanied by elevated serum transaminases and hepatic lipid levels (Fig. 1 and Figure S1). Histopathological assessment revealed canonical MASLD features, including steatosis, balloon-like degeneration and lobular inflammation. In line with these morphological observations, NAS scoring showed higher histological activity in HFHFr-fed mice, with significant reductions after Sema or TZP treatment, supporting attenuation of MASLD-related liver pathology. Together with improved metabolic and biochemical indices, these data support mitigation of HFHFr-induced MASLD features by both treatments (Fig. 2). These findings are consistent with prior reports using similar diet composition and duration, supporting the successful establishment of the HFHFr-induced MASLD model [36]. Although the present study was conducted within the same overall in vivo experimental framework as our previously published BAT-focused work [33], the central analyses and conclusions of the current study are based on liver-specific pathological, molecular, transcriptomic and proteomic data, which constitute the principal novel contribution of this study. A notable feature of the metabolic phenotyping is that glucose tolerance was assessed using IPGTT rather than OGTT. IPGTT was used to evaluate systemic glucose disposal in the context of Sema and TZP administration, thereby reducing variability related to intestinal glucose absorption and nutrient-stimulated endogenous incretin release. However, IPGTT does not replicate the postprandial physiology through which incretin-based therapies may act, including effects on gastric emptying, oral glucose handling, endogenous incretin responses, enterohepatic signaling and gut–liver metabolic integration. Therefore, the IPGTT findings are better interpreted as measures of systemic glucose disposal rather than as a complete assessment of postprandial metabolic regulation. An important interpretative issue is whether the hepatic benefits of TZP are explained primarily by reduced caloric intake and body weight loss, or whether additional weight-loss-independent mechanisms also contribute. Incretin-based therapies are known to reduce appetite and food intake, and these effects, together with subsequent weight loss, likely account for part of the improvement in MASLD-related phenotypes. In the present study, TZP treatment was associated with a reduction of hepatosomatic index (Fig. 2B) and an improvement of glucose homeostasis (Fig. 1C and D), supporting an overall improvement in metabolic status. At the same time, the parallel improvement in liver histology, inflammatory mediators, biochemical and molecular readouts raise the possibility that mechanisms beyond caloric restriction alone are also involved, including indirect metabolic effects and possible liver-intrinsic actions. However, in the absence of a pair-fed control group, the present study cannot distinguish these possibilities definitively. The relative contribution of reduced energy intake, weight loss and direct hepatic effects remains to be established. Beyond these systemic effects, hepatic inflammation appeared as a prominent TZP-responsive feature in our model. The pro-inflammatory cytokines (TNF-α and IL-1β) released during the pathophysiological process of MASLD primarily originate from macrophages [37]. TNF-α acts as a key amplifier of inflammatory cascades [38], whereas IL-1β promotes hepatic inflammation and injury, and contributes to steatosis and fibrogenic remodeling [39, 40]. Mechanistically, IL-1β can induce hepatocytes to secrete TNF-α and MCP-1 and increase their sensitivity to TNF-α-induced cytotoxicity [41, 42]. Additionally, pyroptosis may further propagate this inflammatory circuit, as GSDMD, a key executor of pyroptosis, facilitates IL-1β release and induces pyroptosis [43, 44]. Compared with CON, our current model showed that HFHFr feeding increased the levels of circulating TNF-α, IL-1β, MCP-1 and GSDMD. This was supported by histological evidence of hepatic steatosis and lobular inflammation (Fig. 2). Both TZP and Sema interventions shifted these markers toward control levels, with TZP associated with greater reduction. These findings support an association between treatment and decreased inflammation of the liver, and suppression of hepatic inflammation is an important component of the therapeutic response. TZP may exert a quantitatively stronger anti-inflammatory effect than Sema under the present experimental conditions. We next sought to explore the molecular mechanisms of TZP in treating hepatic steatosis and inflammation in MASLD mice through comprehensive transcriptomic and proteomic analyses. In the present HFHFr-induced MASLD model, we observed a marked and coordinated hepatic transcriptome and proteome reprogramming, with PCA separating control, HFHFr and treatment groups, consistent with a robust shift in liver state rather than isolated pathway perturbations (Figs. 3B and 4B-D). Within this remodeling, TZP partially reverse HFHFr-associated alterations without fully restoring the control state, suggesting that pharmacological improvement is accompanied by treatment-specific molecular signatures (Fig. 3L). A salient feature of the HFHFr state was enrichment of immune-response and cytokine-associated ontologies, together with enrichment of chemokine, PI3K–AKT and mTOR signaling—pathways that connect inflammatory recruitment, cellular activation and nutrient-sensing programs (Figs. 3F–K and 4I-L). Notably, relative to HFHFr, TZP was associated with downregulation of pathway annotations linked to chemokine and PI3K–AKT signaling. Sema also shifted the hepatic molecular profile, but the pattern of enrichment was not identical. In particular, chemokine- and PI3K-AKT-related enrichment appeared more evident in the TZP comparisons than in the Sema comparisons in the datasets presented (Figs. 3 and 4). Together, these findings indicate differential pathway association rather than proof of distinct liver-specific mechanisms, and may reflect differences in systemic metabolic improvement, hepatic inflammatory state or the bulk-tissue composition captured by these analyses. The integrated KEGG annotation across transcriptomics and proteomics highlights a clear multi-omic convergence on inflammatory signaling modules (Fig. 4M). The chemokine and the PI3K–AKT signaling pathways were among the most consistently represented across omic layers. Specifically, chemokine- and immune-associated transcripts detected in the integrated analysis included Elmo1, Fgr, Gnai1, Pik3r5, Ntrk1, Syk, H2-Ab1 and H2-Aa, whereas proteins mapped by KEGG to these pathways included Rps6, Sect. 13, Pik3r1, Ptk2b and Aox3 (Fig. 4M). This cross-omic concordance strengthens inference beyond single-omic analyses and supports coordinated regulation of chemokine signaling and PI3K–AKT-associated modules in response to TZP. Chemokines are a family of small, highly conserved cytokines that classically grouped into four subfamilies—C, CC, CXC and CX3C. By directing leukocyte trafficking and positioning, chemokine networks can contribute to hepatic immune-cell recruitment and activation in MASLD and shape downstream inflammatory mediator production and tissue immune organization [45]. Specifically, chemokine signaling contributes to multiple MASLD-related processes, including extracellular matrix remodeling, hepatocyte proliferative responses and hepatic stellate cell activation [46]. The CCL2 signaling has been implicated in metabolic dysfunction and hepatic lipid accumulation in MASLD [46], and CCL2/CCR2 can engage PI3K-AKT signaling to regulate cellular metabolism, growth, survival, and apoptosis [47]. PI3K-AKT is also downstream of cytokine and chemokine receptors that coordinate immune cell recruitment and activation, particularly of macrophages. This process plays a critical role in regulating inflammation and exerting anti-inflammatory, antioxidant stress, and anti-apoptotic effects in liver tissue. Consistent with this framework, activation of the CCL2/CCR2–PI3K-AKT pathway has been linked to high-fat diet-induced liver inflammation in NASH mice [35]. The results of this study are consistent with previous findings. Compared to CON, HFHFr feeding was associated with increased hepatic expression of the chemokine axis CCL2–CCR2 and linked PI3K-AKT signaling (including elevated p-AKT), together with higher transcripts of pro-inflammatory mediators (IL-1β and TNF-α) (Figs. 5A–F and 6A–E). In contrast, TZP treatment reduced these molecular readouts relative to HFHFr (Figs. 5A-F and 6A-E), which is consistent with an anti-inflammatory molecular shift at the level of CCL2/CCR2-associated PI3K-AKT signaling pathway and cytokine expression. Taken together, these findings support a model in which TZP is associated with suppression of a CCL2/CCR2-linked inflammatory signaling axis involving PI3K-AKT-related downstream responses. However, these mechanistic findings should be interpreted with caution. The qRT-PCR and Western blot assays were performed in separate batches and were not designed to serve as an independent replication cohort. Although the additional samples increased biological coverage and showed concordant trends for selected targets, residual batch effects cannot be excluded. Thus, these data support the involvement of the CCL2/CCR2–PI3K-AKT inflammatory signaling axis in the hepatic response to TZP, but do not establish pathway causality. This study has several limitations. First, a pair-fed control group was not included. Therefore, we were unable to rigorously determine the extent to which the hepatic benefits of TZP were attributable to reduced caloric intake and weight loss versus weight-loss-independent mechanisms. Future studies incorporating pair-feeding and liver-specific mechanistic analyses will be needed to address this question. Second, all experiments were performed in male C57BL/6J mice, limiting generalizability given the documented sex differences in metabolic liver disease. Validation in female mice and additional preclinical models will be required. Third, as transcriptomic and proteomic profiling was conducted on bulk liver tissue, pathway-level changes cannot be assigned to specific hepatic cell populations. Single-cell and spatial profiling combined with quantitative immunophenotyping will be needed to determine the cellular sources of the TZP-responsive chemokine and PI3K/AKT-associated signatures, including hepatocytes as well as liver-resident and infiltrating immune populations, such as Kupffer cells and monocyte-derived macrophages. Fourth, although multi-omic integration and targeted validation identified the CCL2/CCR2–PI3K/AKT axis as a TZP-responsive inflammatory signaling node, the absence of causal perturbation (e.g., CCR2 genetic deletion, pharmacologic blockade or cell-type–restricted manipulations) limits mechanistic attribution and precludes inference of necessity or sufficiency. Conclusions In the current HFHFr diet-induced MASLD mouse model, TZP treatment ameliorated hepatic steatosis, liver injury and inflammatory abnormalities. Multi-omics analyses and targeted validations revealed an association between TZP and the downregulation of hepatic chemokine-related inflammatory signaling, particularly the CCL2/CCR2 axis and PI3K-AKT signaling. Together, these data revealed a liver-related inflammatory pathway signature associated with TZP, suggesting that the CCL2/CCR2-associated signaling network as a candidate mechanism underlying its hepatoprotective effects in diet-induced MASLD. Supplementary Information Supplementary Material 1. Supplementary Material 2. Supplementary Material 3. Supplementary Material 4. Supplementary Material 5. Supplementary Material 6. Supplementary Material 7. Acknowledgements The authors are grateful to all participants of this study. Abbreviations One-way analysis of variance Alanine aminotransferase Aspartate aminotransferase Area under the curve C–C chemokine ligand 2 C–C chemokine receptor 2 Control Differentially expressed gene Fasting insulin Glucose-dependent insulinotropic polypeptide Glucagon-like peptide-1 Gene Ontology Gasdermin D Gene set enrichment analysis High-fat diet High-fat, high-fructose Homeostatic Model Assessment of Insulin Resistance Intraperitoneal glucose tolerance test Insulin resistance Kyoto Encyclopedia of Genes and Genomes Liquid chromatography-mass spectrometry Metabolic dysfunction-associated steatohepatitis Metabolic dysfunction-associated steatotic liver disease Monocyte chemoattractant protein-1 Molecular functions Non-alcoholic fatty liver disease NAFLD activity score Principal component analysis Sodium dodecyl sulfate Semaglutide Triglyceride Tirzepatide Authors’ contributions Conceptualization: Yongning Pan and Luping Ren; methodology: Yongning Pan, Fanfan Song, Xuehua Liu; validation: Juanjuan Zhang, Yan Zhang and Yongning Pan; formal analysis: Luping Ren; investigation: Yongning Pan, Fanfan Song; resources: Luping Ren; data curation: Yongning Pan; writing—original draft: Yongning Pan; writing—review and editing: Luping Ren; visualization: Yongning Pan, Fanfan Song; supervision Guangyao Song; project administration: Luping Ren; funding acquisition, Luping Ren. All authors have read and agreed to the published version of the manuscript. Funding This research was funded by Natural Science Foundation of Hebei Provincial funds Project (grant no. H2025307007). Data availability The datasets analyzed for this study can be found in the Figshare (10.6084/m9.figshare.30618578). The transcriptomic data used in the current study are publicly available at the National Library of Medicine (NCBI). They can be accessed via the following link: https://dataview.ncbi.nlm.nih.gov/, under the BioProject numbers: PRJNA1364351. The mass spectrometry proteomics data have been deposited in the ProteomeXchange Consortium (https://proteomecentral.proteomexchange.org) via the iProX partner repository with the dataset identifier PXD071304. Declarations Ethics approval and consent to participate The animal study protocol was approved by the Ethics Committee of Hebei General Hospital (protocol code 2025DW002) on January 9, 2025. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1. Supplementary Material 2. Supplementary Material 3. Supplementary Material 4. Supplementary Material 5. Supplementary Material 6. Supplementary Material 7. Data Availability Statement The datasets analyzed for this study can be found in the Figshare (10.6084/m9.figshare.30618578). The transcriptomic data used in the current study are publicly available at the National Library of Medicine (NCBI). They can be accessed via the following link: https://dataview.ncbi.nlm.nih.gov/, under the BioProject numbers: PRJNA1364351. The mass spectrometry proteomics data have been deposited in the ProteomeXchange Consortium (https://proteomecentral.proteomexchange.org) via the iProX partner repository with the dataset identifier PXD071304.