Multi-omics integration uncovers the molecular mechanism underlying variation in meat quality across different muscles of Pingliang Red cattle Authors: Jiawei Lu, Junwei Hu, Jin Wang, Yanxiang Zhang, Yifan Ma, Rui Duan, Xuemei Gao, Tianliang Yang, Yongfeng Wang, Minghao Ma, Shuzhen Li, Lili Chen, Yi Ma, Lupei Zhang Published: 2026 ABSTRACT AbstractThis study integrated metabolomics and proteomics to uncover core metabolic pathways underlying variation in meat quality among three cuts of Pingliang Red cattle: topside, striploin, and clod. The pH value and b* value of the topside were significantly higher than those of the two muscles (P < 0.05). Metabolomics identified 29, 15, and 13 notably abundant metabolites in topside, striploin, and clod, respectively, including 5-diphosphate, PS(18:0/20:4), and L-threonine (P < 0.05). Proteomics illustrated that there were 35, 64, and 19 differentially expressed proteins in topside, striploin, and clod, respectively. Integrated correlation and protein-protein interaction analyses verified that energy metabolism, amino acid turnover, and lipid biosynthesis served as central regulatory modules dominated by hub molecules such as NME1 and MYH7. These findings elucidate the multi-omics basis of meat quality divergence in Pingliang Red cattle and provide potential molecular markers for targeted breeding and the segmented utilization of this local breed's meat.Keywords: Pingliang Red cattle, Meat quality, Multi-omics, Meat cuts Introduction Meat quality determines the purchasing intention of consumers and the economic benefit of the livestock industry as a vital indicator (Wang et al., 2026; Wang et al., 2026; Wang, Mu, Zhang, & Wang, 2026), which is determined by genetic structure, breeding, decomposition part, and storage conditions (Dashdorj, Amna, & Hwang, 2015). Beef has abundant essential amino-acid, fatty acids and B vitamins as an important source of nutrition for the consumers in the daily diet (Wu, 2020). In traditional markets, beef is typically sold in cut. Beef carcass meat can be cut into topside, striploin, clod, tenderloin, rib, neck, shank, and brisket (McAfee et al., 2010). The meat quality of these beef cuts differs from their natural physiological function in the body (Zhang et al., 2022), resulting from the differences of the muscle fiber types, fatty acid composition and metabolites (Oliveira et al., 2015; Scozzafava, Corsi, Casini, Contini, & Loose, 2016; Yu et al., 2019). Previous studies have mainly focused on how breed, nutrition, and rumen microbiota affect the beef quality traits (Liu et al., 2026; Polizel et al., 2026; Wang, Miao, et al., 2026; Wang, Mu, et al., 2026; Wang, Song, et al., 2026), but mechanistic studies remain unknown about the meat quality differences among the beef cuts. The meat quality and flavor arise from a complicated biochemical transformation process involving nucleotides, organic acids, and bioactive peptides (Ramalingam, Song, & Hwang, 2019). Protein acts as the dominant structural constituent of beef muscle and serves as the primary reservoir of flavor precursors. It releases cysteine, glutamic acid and other essential amino acids that act as core substrates driving the Maillard reaction during thermal processing (Zhang et al., 2026). Small peptides further enrich meat aroma by modulating the formation of Maillard-derived flavor compounds (Ogasawara, Katsumata, & Egi, 2006). In addition, hydrolysates generated from endogenous protein degradation, such as hydrogen sulfide and ammonia, can react with lipid oxidation-derived aldehydes to form alkyl thiazoles and other heterocyclic volatile substances, which collectively determine the characteristic flavor of cooked beef (Zhang et al., 2025; Zhang et al., 2025). Despite the well-established roles of proteins, peptides and lipid derivatives in shaping beef flavor, the integrated molecular regulatory networks governing tissue-specific meat quality disparities remain poorly characterized when only relying on single analytical methods. Multi-omics strategies, combining metabolomics and proteomics, offer a systematic solution to decode the full spectrum of biomolecular alterations linked to phenotypic traits (Wang, Miao, et al., 2026; Wang, Mu, et al., 2026; Wang, Song, et al., 2026). Metabolomics captures dynamic small-molecule metabolic signatures that directly mirror post-slaughter muscle physiological status, whereas proteomics quantitatively profiles functional proteins and enzymes driving upstream biochemical cascades (Goswami et al., 2026). Specifically, data-independent acquisition (DIA)-based proteomics enables precise high-throughput quantification of muscle proteins to pinpoint key regulatory proteins and associated biological pathways linked to meat quality (Wang, Miao, et al., 2026; Wang, Mu, et al., 2026; Wang, Song, et al., 2026). Compared with independent genomic or transcriptomic detection, the complementary strengths of metabolomic and proteomic profiling allow comprehensive dissection of cascaded metabolic events from protein expression to terminal metabolite accumulation (Johnson, Ivanisevic, & Siuzdak, 2016; Li et al., 2026; Li et al., 2026). For this reason, integrated multi-omics analysis represents a powerful framework to unravel the molecular basis of divergent meat quality phenotypes. To date, however, no multi-omics integrated research has systematically compared proteomic and metabolomic variations across three commercially valuable Pingliang Red cattle cuts: topside, striploin, and clod. Pingliang Red cattle are a well-known breed in China, famous for their tender muscles, abundant intramuscular fat, and high myoglobin content. The marble score of Pingliang Red cattle exceeds 4.5, and the meat quality of Pingliang Red cattle is comparable to that of Japanese Wagyu (Liu, Pan, Ye, & Cao, 2013). Pingliang Red cattle have abundant flavor precursor amino acids, oleic acid, and linoleic acid (Liu et al., 2025). Despite its commercial value, the scientific basis of flavor difference among topside, striploin, and clod remains unexplored. In this study, the differentially expressed metabolites and proteins were distinguished by integrative metabolomics and proteomics. This study provides a scientific basis for molecular breeding, cut-specific fine processing, and product grading of Pingliang Red cattle. Materials and methods Animals and sample collection All animal experiments were approved by the Animal Care and Use Committee of the Tianjin Institute of Animal Husbandry and Veterinary Medicine (2023003) and were performed in compliance with the committee's guidelines. Muscles from three-year-old male Pingliang Red cattle (n = 6) with an average body weight of 530.42 ± 43.07 kg were collected from a local slaughterhouse in Pingliang, China. From each carcass, three target muscle cuts (topside, striploin, and clod) were collected simultaneously. All three meat cuts originated from the same set of six cattle. Three technical replicates were conducted for each muscle sample to guarantee experimental reliability. Fig. 1 illustrates the anatomical locations and corresponding meat samples of three representative cuts from the cattle carcass: clod, striploin, and topside. The meat samples were collected in frozen pipes and stored in liquid nitrogen for metabolomic and proteomic analyses. Schematic anatomical location diagram of three representative beef cuts (topside, striploin, clod) from Pingliang Red cattle, with corresponding muscle sample photographs attached. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Meat quality measurement pH value The pH value of the muscle samples was measured by a pH meter (206-pH 2, testo, China). The measurements were repeated three times to obtain the average value (Liang et al., 2020). Cooking loss Meat samples (about 100 g) were vacuum-packaged, and cooked in a water bath (HH-3SY, NY, China) at 70 °C until the internal temperature reached 70 °C. After cooking, the samples were cooled to room temperature, surface-dried, and reweighed (Liang et al., 2020). Shear force After measurement of cooking loss, cylindrical cores (1.27 cm diameter) were prepared parallel to muscle fibers and sheared perpendicular to fiber direction using a Warner-Bratzler device on a texture analyzer (FTM-50, METROTEC, Spain). Shear force (N) was recorded and averaged for each sample (AMSA, 1995). Water holding capacity (WHC) A meat sample (about 0.5 g) was placed on a pre-weighed filter paper, then covered with another filter paper and subjected to a constant pressure of 35 kg for 5 min using a press device. After pressing, the filter paper was reweighed. The amount of water released was calculated, and WHC was expressed as the percentage of retained water relative to the initial moisture content of the sample (Liang et al., 2020). Meat color The color parameters of the muscle samples, including lightness (L*), redness (a*), and yellowness (b*), were determined using a Konica Minolta CR-10 colorimeter (Tokyo, Japan). Measurements were performed under the D65 illuminant with an 8 mm aperture, a 10° standard observer angle, and with the instrument oriented transversely to the muscle fiber direction. Before measurement, 10 mm-thick cross-sectional steaks were prepared perpendicular to the muscle fibers and allowed to bloom for 15 min under ambient light conditions. Moreover, chroma (c*) = [(a*)2 + (b*)2]0.5, total color difference (E*) = [(a*)2 + (b*)2 + (c*)2]0.5 and hue angle (H*) = [arctan(b*/a*)] were calculated (Liu et al., 2025). Metabolome analysis Metabolite extraction Approximately 50 mg of tissue sample was transferred into a 2 mL centrifuge tube, together with a 6 mm grinding bead for tissue homogenization. The extraction solvent consisting of methanol and water (4:1, v/v) containing four internal standards was then added to the tube. Samples were homogenized using a Wonbio-96c low-temperature tissue grinder (Wonbio-96c, Wonbio, China) at 50 Hz and − 10 °C for 6 min, followed by ultrasonic extraction at 40 kHz and 5 °C for 30 min. After incubation at −20 °C for 30 min, the mixture was centrifuged at 13,000g and 4 °C for 15 min using a high-speed refrigerated centrifuge (5430R, Eppendorf, Germany). The resulting supernatant was collected and transferred to sample vials for subsequent ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) detection (Zhang, Tan, et al., 2025; Zhang, Zhang, et al., 2025). UPLC-LC-MS/MS analysis Metabolite profiling was performed on a UHPLC-Orbitrap Exploris 480 mass spectrometer coupled with an ACQUITY HSS T3 column (100 mm × 2.1 mm, 1.8 μm, Waters, USA) at Majorbio Bio-Pharm Technology Co., Ltd. The mobile phase system was composed of solvent A (0.1% formic acid in ultrapure water) and solvent B (0.1% formic acid in acetonitrile). The flow rate was set at 0.40 mL/min and column temperature maintained at 40 °C, with an injection volume of 5 μL. An electrospray ionization (ESI) source was adopted and operated in both positive and negative ion modes. The ion source temperature was set to 400 °C; sheath gas and auxiliary gas flow rates were 40 arbs and 10 arbs, respectively. The ion spray voltage was set at 3500 V in positive mode and − 2800 V in negative mode. A stepped normalized collision energy of 20–40-60 V was applied for MS/MS fragmentation. Proteome analysis Total protein extraction Tissue samples were lysed with protein extraction buffer supplemented with protease inhibitors to prevent protein degradation. The mixtures were processed by high-throughput tissue disruption for three cycles (180 s per cycle), followed by non-contact low-temperature ultrasonication. After incubation, samples were centrifuged at 16,000g and 8 °C for 30 min. Total protein concentration in the supernatant was quantified using the bicinchoninic acid (BCA) assay kit (Thermo Scientific), and sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was conducted for preliminary quality verification. Protein digestion A total of 100 μg of quantified protein was dissolved in 100 mM triethylammonium bicarbonate buffer (TEAB) solution. Protein reduction was achieved by adding 10 mM tris(2-carboxyethyl) phosphine (TCEP) and incubating the mixture at 37 °C for 60 min. Subsequently, 40 mM iodoacetamide (IAM) was added for alkylation in dark conditions at room temperature for 40 min. The solution was centrifuged at 10,000g and 4 °C for 20 min, and the precipitate was re-suspended in 100 mM TEAB buffer. Sequencing-grade trypsin was added at a protein-to-enzyme mass ratio of 50:1, and the digestion reaction was performed overnight at 37 °C (Qiao, Zhu, Du, & Morigen, 2025). DIA mass detection Peptide samples were analyzed using a Vanquish Neo liquid chromatography system coupled to an Orbit Astral mass spectrometer (Thermo, USA) at Majorbio Bio-Pharm. A uPAC high-throughput column (75 μm × 5.5 cm) was used for chromatographic separation. The mobile phase consisted of 2% acetonitrile +0.1% formic acid (solvent A) and 80% acetonitrile +0.1% formic acid (solvent B), with a total run time of 8 min. DIA mode was applied for mass spectral collection, and the mass range was set at 100–1700 m/z (Cao et al., 2026). Protein identification Raw DIA data were analyzed using Spectronaut (Version 19). The search criteria were set as follows: trypsin/P was designated as the digestion enzyme, allowing a maximum of two missed cleavage sites. Cysteine carbamidomethylation was defined as a fixed modification; methionine oxidation and protein N-terminal acetylation were set as variable modifications. The false discovery rate (FDR) for both proteins and peptides were kept below 0.01, and peptide confidence was required to exceed 99%. The XIC extraction window was limited to ≤75 ppm, and the MaxLFQ algorithm was adopted for protein relative quantification (Ye et al., 2026). Statistical analyses All proteomic bioinformatics analyses were performed on the Majorbio Cloud Platform. Student's t-test in R was used to calculate P-values and fold changes (FC) between sample groups. Proteins with FC > 1.2 or FC < 0.83 and P < 0.05 were regarded as differentially expressed proteins (DEPs). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were performed to interpret the biological functions of DEPs. Protein-protein interaction (PPI) networks were constructed based on the STRING v11.5 database for further functional exploration (Wang, Song, Wang, & Quan, 2024). Results Meat quality of striploin, topside, and clod As shown in Table 1, the pH values of topside were significantly higher than those of striploin (P < 0.05). Numerically, topside had higher b* values than clod (P < 0.05). However, there were no significant differences in the cooking loss, shear force, and water holding capacity in topside, striploin, and clod (P > 0.05). Table 1Meat quality traits of striploin, topside, and clod in Pingliang Red cattle (n = 6).ItemsStriploinTopsideClodP-valuepH5.62 ± 0.03b5.80 ± 0.03a5.70 ± 0.06ab0.045Cooking loss, %24.29 ± 2.93a28.92 ± 1.66a23.29 ± 1.39a0.127Shear force, N36.46 ± 8.51a38.92 ± 3.00a31.71 ± 3.30a0.589Water holding capacity, %21.85 ± 2.27a23.68 ± 0.63a25.44 ± 1.89a0.357L*38.15 ± 1.46a42.92 ± 2.25a37.52 ± 0.88a0.067a*13.68 ± 1.77a13.24 ± 1.47a10.81 ± 0.71a0.293b*13.75 ± 1.30ab15.64 ± 0.82a11.99 ± 0.78b0.045c*19.42 ± 2.14a20.77 ± 0.71a16.17 ± 0.98a0.061E*27.46 ± 3.03a29.38 ± 1.01a22.86 ± 1.39a0.061H*45.60 ± 1.46a50.10 ± 4.17a48.01 ± 1.41a0.552Notes: The values were shown as the mean ± SEM. Different lowercase superscripts within the same row indicate significant differences at P < 0.05; identical lowercase letters mean no significant difference among groups (P < 0.05). Meat quality traits of striploin, topside, and clod in Pingliang Red cattle (n = 6). Notes: The values were shown as the mean ± SEM. Different lowercase superscripts within the same row indicate significant differences at P < 0.05; identical lowercase letters mean no significant difference among groups (P < 0.05). Differential metabolites and metabolome data analysis To explore metabolic differences among the three groups (topside, striploin, and clod), partial least squares discriminant analysis (PLS-DA) was performed. The PLS-DA score plot showed clear separation of the three groups along the two main components, indicating distinct metabolic profiles among the treatments (Fig. 2A). A Venn diagram illustrated the significant metabolites among topside, striploin, and clod. There were 29, 15, and 13 significant metabolites in topside, striploin, and clod, respectively, with only 1 common metabolite altered across all three comparisons (P < 0.05, Fig. 2B; Table S1). To identify metabolites with the highest contribution to the group, variable importance projection (VIP) values from the PLS-DA model were calculated for each comparison. The results showed that topside had higher contents of 2-octylpent-2-enedioic acid, violacein, 5,7-dihydroxy-3-methoxyflavone, 2-deoxy-28-homocastasterone, and arachidoyl ethanolamide (P < 0.05, Fig. 2C). The contents of uridine, uridine diphosphate-N-acetylgalactosamine, violacein, sulfonamidoethanol, and other metabolites were significantly upregulated in topside than those in striploin (P < 0.05, Fig. 2D). Moreover, striploin showed higher contents of N-eicosapentaenoyl tyrosine, phytosphingosine, and remimazolam than the clod (P < 0.05, Fig. 2E). To visualize the expression patterns of significantly altered metabolites across all samples, hierarchical clustering analysis with heatmaps was performed for each comparison. The clustering dendrograms revealed distinct metabolite expression patterns that clearly separated the samples according to different groups (Fig. 2F, G, H). Partial least squares discriminant analysis (PLS-DA) and hierarchical clustering heatmap analysis of differential metabolites from topside, striploin and clod of Pingliang Red cattle. (A) PLS-DA score plot; (B) Venn diagram showing the overlap of significantly altered metabolites among three muscle groups; Variable importance in projection (VIP) plots for metabolite comparison between clod vs topside (C), striploin vs topside (D), striploin vs clod (E); Hierarchical clustering heatmaps of differential metabolites for clod vs topside (F), striploin vs topside (G), striploin vs clod (H). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) KEGG enrichment analysis of metabolome data To explore the biological functions associated with the differential metabolites identified between muscle parts, KEGG pathway enrichment analysis was performed. Metabolites differing between the clod and topside muscles were significantly enriched in pathways related to purine metabolism, nucleotide metabolism, steroid hormone biosynthesis, and other terms (P < 0.05, Fig. 3A; Table S2). KEGG enrichment analysis indicated that amino acid and fatty acid metabolism pathways were altered between striploin and topside, including nucleotide metabolism, arachidonic acid metabolism, glycine, serine, threonine metabolism, and other pathways (P < 0.05, Fig. 3B; Table S3). In the comparison between striploin and clod, differential metabolites were mainly enriched in glycerophospholipid metabolism, taurine and hypotaurine metabolism, sulfur metabolism, and other terms (P < 0.05, Fig. 3C; Table S4). Kyoto Encyclopedia of Genes and Genomes (KEGG) differential abundance (DA) score plots of differential metabolites between three muscle cuts: clod versus topside (A), striploin versus topside (B), striploin versus clod (C). Correlation analysis between metabolites and enriched KEGG pathways To explore the metabolic basis underlying quality differences in different parts of Pingliang Red cattle, correlation network analysis between differential metabolites and their enriched KEGG pathways was performed (Fig. 4). Multiple core metabolic and signaling pathways were significantly associated with key metabolites, including oxidative phosphorylation, purine metabolism, biosynthesis of cofactors, FoxO signaling pathway, and steroid hormone biosynthesis. Notably, adenosine 5′-diphosphate (ADP) was identified as a central metabolite, connecting to energy-related pathways such as oxidative phosphorylation and thermogenesis, as well as the FoxO signaling pathway between clod and topside (P < 0.05, Fig. 4A). In the correlation analysis between striploin and topside, the network revealed two major functional modules associated with lipid and nucleotide/energy metabolism. The first module was centered on glycerophospholipid metabolism and arachidonic acid metabolism, linked to lipid-related metabolites including PS(18:0/20:4), 6-ketoprostaglandin F1α, and 20-hydroxy leukotriene B4. These associations highlight differences in membrane lipid composition, phospholipid turnover, and eicosanoid signaling between the two cuts. The second module focused on nucleotide and energy metabolism, anchored by purine metabolism, nucleotide metabolism, and biosynthesis of cofactors. Key metabolites in this module included ADP, adenosine 3′,5′-diphosphate, uridine, and d-ribose 5-phosphate, reflecting variations in cellular energy status, nucleotide recycling, and redox homeostasis. Additional pathways, such as ABC transporters and sulfur metabolism, were linked to metabolites like taurine and L-threonine, further supporting differences in amino acid metabolism and cellular transport processes (P < 0.05, Fig. 4B). Notably, lipid-related metabolites such as Pc(16:0/20:5), Pe(36:4), and Gpcho(20:5/20:4) were strongly linked to lipid metabolism pathways, indicating variations in membrane lipid composition and lipid signaling between the striploin and clod muscles (P < 0.05, Fig. 4C). Correlation network integrating differential metabolites and enriched Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways for pairwise muscle comparisons: clod vs topside (A), striploin vs topside (B), striploin vs clod (C). Proteomics analysis PLS-DA graph indicated that the samples were clearly separated in different muscles, highlighting that specific proteins were vital contributors (Fig. 5A). Venn diagram showed that there were 35, 64, and 19 proteins notably different in topside, striploin and clod, respectively (Fig. 5B). As shown in Fig. 5C–F, the number of DEPs was 214 between topside and striploin (109 upregulated proteins and 105 downregulated proteins). 239 DEPs were identified between clod and striploin (116 upregulated proteins and 123 downregulated proteins). 140 DEPs were detected between topside and clod (76 upregulated proteins and 64 downregulated proteins). Statistical analysis of differentially expressed proteins (DEPs) in three Pingliang Red cattle muscle cuts. (A) Partial least squares discriminant analysis (PLS-DA) score plot reflecting sample grouping similarity; (B) Venn diagram illustrating the number of overlapping DEPs across three pairwise comparisons; (C) Bar chart showing upregulated and downregulated DEP counts; Volcano plots for DEPs of topside vs clod (D), topside vs striploin (E), clod vs striploin (F). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) GO and KEGG enrichment analysis of proteome data GO pathway enrichment analysis illustrated that the DEPs between topside and striploin were involved in cyclic nucleotide transport, acetyl-CoA metabolic process, lysine catabolic process and other terms (P < 0.05, Fig. 6A; Table S5). Similarly, KEGG enrichment analysis indicated that DEPs between topside and striploin were involved in metabolic pathways, including lysine degradation, retinol metabolism, Hippo signaling pathway, purine metabolism and other pathways (P < 0.05, Fig. 6B; Table S6). GO pathway enrichment analysis illustrated that the DEPs between topside and clod were involved in negative regulation of triglyceride metabolic process, GTP metabolic process, protein processing and other terms (P < 0.05, Fig. 6C; Table S7). Moreover, KEGG enrichment analysis indicated that DEPs between topside and clod were predominantly enriched in cAMP signaling pathway, cGMP-PKG signaling pathway, pyrimidine metabolism, purine metabolism and other pathways (P < 0.05, Fig. 6D; Table S8). GO pathway enrichment analysis illustrated that the DEPs between clod and striploin were enriched in triglyceride biosynthetic and metabolic process, neutral lipid biosynthetic process, lactate metabolic process and other terms (P < 0.05, Fig. 6E; Table S9). KEGG enrichment analysis indicated that DEPs between clod and striploin were enriched in p53 signaling pathway, cholesterol metabolism, autophagy and other pathways (P < 0.05, Fig. 6F; Table S10). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis of differentially expressed proteins (DEPs). GO enrichment (A) and KEGG enrichment (B) for topside versus striploin; GO enrichment (C) and KEGG enrichment (D) for topside versus clod; GO enrichment (E) and KEGG enrichment (F) for clod versus striploin. Protein-protein interaction network analysis To explore the functional interactions among DEPs between topside, striploin and clod muscles of Pingliang Red cattle, PPI network was constructed (Fig. 7). The network comprised multiple interconnected modules reflecting distinct biological processes underlying muscle functional differences. As shown in Fig. 7A, a central hub was formed by nucleoside diphosphate kinase (NME1), which interacted with key proteins including AK4, RRM1, GFM1, and MTIF2. This core module is closely linked to nucleotide metabolism, mitochondrial translation, and energy homeostasis, indicating differences in cellular energy status and mitochondrial function between the two muscle types. The association of NME1 with ribosomal proteins (RPL13, RPL23A-2) and mitochondrial ribosomal components (MRPL17, MRPS36) further suggests variations in protein synthesis machinery and mitochondrial biogenesis. A second major module centered on energy metabolism and redox balance, including mitochondrial respiratory chain proteins such as COX15, COX7C, and UQCR10, as well as tricarboxylic acid cycle enzymes like DLST. Protein-protein interaction (PPI) networks of differentially expressed proteins (DEPs) constructed based on the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) database. PPI network for DEPs of topside vs striploin (A), topside vs clod (B), clod vs striploin (C). Each node represents a single protein, and line thickness corresponds to the strength of protein interaction correlation. As shown in Fig. 7B, a central hub was formed by NME1, interacting with multiple proteins, including RRM1, AK4, MRPL17, MRPS15, and GFM1. This module links nucleotide metabolism, ribosomal biogenesis, and mitochondrial translation, indicating differences in cellular energy status, nucleotide turnover, and mitochondrial function between the two muscle types. A second prominent module centered on ANXA5, connected to structural and signaling proteins such as SMAD3, FBN1, FLNA, and STAT6. This cluster points to variations in extracellular matrix organization, cytoskeletal dynamics, and TGF-β/STAT signaling, which are closely related to muscle fiber structure, connective tissue content, and meat tenderness. Additional modules involved proteins associated with RNA processing (FUS, NEFH), autophagy and protein degradation (ATG7, DLST), and lipid/energy metabolism (APOC3, CST6). These interconnected networks collectively demonstrate that differences in nucleotide/energy metabolism, structural protein organization, and cellular signaling underpin the divergent meat quality traits of topside and clod muscles in Pingliang Red cattle. As shown in Fig. 7C, a prominent hub centered on NOP58 and MRPS15, with connections to ribosomal proteins (RPL9, RPL13, RPL21), translation initiation factors (EIF4E), and mitochondrial ribosomal components. A second key module was anchored by myosin heavy chain 7 (MYH7) and TPM2, interacting with structural and contractile proteins such as ATP2A1, CASQ2, and ILK. Additional modules involved proteins associated with lipid and energy metabolism (COX15, HSD11B1), cellular stress response (GSS, THNSL1), and signal transduction (RPTOR, EIF4E). Discussion Topside, striploin, and clod are commercially valuable cuts of beef meat. The meat quality is evaluated by pH value, tenderness, cooking loss, and meat quality (Hou et al., 2023). The results showed that topside had a higher pH value and b* value than the striploin and clod muscles, indicating that topside had a lower extent of post-mortem acidification, less susceptibility to PSE (pale, soft, exudative) meat, and superior water-holding capacity and flavor stability. The difference of pH value among the muscle cuts may be resulting from the differences of the fiber type, oxygen consumption rate and iron content (Holman, Kerr, Morris, & Hopkins, 2019). Topside has a higher b* value than striploin from dark cutting and control cutting in Australian beef (Holman & Hopkins, 2019). The changes in L* and b* are caused by pH-mediated differences in physical muscle fibers and their darkening effect on light scattering characteristics (Hughes, Clarke, Purchas, & Warner, 2017). Differences among muscle cuts of Pingliang Red cattle were only observed in pH value and b* value, the profound variations in flavor and meat quality are primarily governed by molecular metabolic networks. Metabolomics distinguishes the small molecule metabolites with a molecular weight < 1500 Da in tissue, fluid, and cell samples by qualitative and quantitative analysis (Muthubharathi, Gowripriya, & Balamurugan, 2021). Topside, striploin, and clod were separated by PLS-DA analysis. There were 32 differential metabolites between topside and clod, 54 differential metabolites between striploin and clod, and 69 differential metabolites between topside and striploin. These differential metabolites were mainly nucleotide, amino acids, and their metabolites. Similarly, KEGG analysis indicated that differential metabolites were enriched in nucleotide metabolism, glycerophospholipid metabolism, glycine, serine, and threonine metabolism. Amino acid metabolism is associated with the flavor, tenderness, and nutritional quality (Wang et al., 2024; Wang, Song, et al., 2024), and nucleotide metabolism is related with the taste characteristics and tenderization (Lv et al., 2024; Mu et al., 2022). The metabolic variation and metabolic characteristics explain the molecular proposition of the meat quality among topside, striploin, and clod in Pingliang Red cattle. Glycerophospholipid metabolism is a core pathway affecting beef lipid characteristics and flavor formation. Glycerophospholipids are important components of cell membranes and major precursors of volatile flavor compounds (Li, Lu, et al., 2026; Li, Zhang, et al., 2026). After slaughter or during thermal processing, phospholipids are hydrolyzed and oxidized to produce aldehydes, ketones, and other volatile substances, which constitute the characteristic aroma of beef (Jung et al., 2026). Multiple differential glycerophospholipids such as PS(18:0/20:4) were screened in this study, and their variable contents among the three cuts indicated that the composition and turnover of membrane lipids were tissue-specific, which directly led to differences in lipid-derived flavor. Purine metabolism dominates the formation of meat umami substances: ATP and its degradation products (ADP, inosine monophosphate, inosine) are critical taste-active components (Xu et al., 2025). Pingliang Red cattle exhibited higher inosine monophosphate (IMP) content in muscle compared with Qinchuan cattle and Simmental cattle (Liu et al., 2025). The significant enrichment of purine metabolism in differential metabolites suggested that the energy metabolism and nucleotide turnover rates of the three muscles were distinct, further causing differences in umami intensity after cooking. Purine and glutathione metabolism pathways are linked to the improvement of meat tenderness and comprehensive meat quality traits (Liu et al., 2025). Correlation network analysis confirmed that ADP, PS(18:0/20:4), and L-threonine acted as core differential metabolites, connecting energy metabolism, lipid metabolism, and amino acid pathways, and jointly shaping the metabolic characteristics of different muscle parts. Quantitative proteomics using DIA can systematically identify proteins and pathways associated with meat quality traits (Song et al., 2022). In this study, there were 104 DEPs between topside and striploin, 145 DEPs between clod and striploin, and 75 DEPs between topside and clod. These proteins were notably enriched in purine metabolism, muscle system process, protein processing, negative regulation of lipid metabolic process, protein maturation, and lysine metabolic process. These proteins and related pathways may contribute to the meat differences among the topside, striploin, and clod muscles in Pingliang Red cattle. The vital proteins exert a functional impact on the meat quality features of beef cuts (Jerez-Timaure et al., 2019). PPI network analysis identified NME1 and MYH7 as hub proteins regulating meat quality differences. NME1 is a key enzyme in nucleotide metabolism, which catalyzes the mutual conversion of nucleoside triphosphates and maintains intracellular energy homeostasis (Abu-Taha et al., 2017). It interacts with AK4, RRM1, and other proteins to jointly regulate purine metabolism and mitochondrial function, which matched the metabolomic results of differential nucleotide metabolites. MYH7 is a classic skeletal muscle structural protein, closely related to muscle fiber type, muscle contraction, and beef tenderness (Yan et al., 2025). The differential expression of MYH7 and its interacting proteins (TPM2, ATP2A1) altered the structure and mechanical properties of muscle fibers, resulting in subtle differences in texture among the three cuts. In addition, a large number of DEPs were enriched in lipid metabolism, autophagy, and signal transduction pathways, including p53 and FoxO signaling pathways, indicating that complex signaling networks were involved in regulating muscle metabolism and meat quality after slaughter. The combined analysis of metabolomics and proteomics revealed that the divergence of meat quality among topside, striploin, and clod of Pingliang Red cattle was jointly driven by energy metabolism, amino acid turnover, and lipid biosynthesis. The tissue-specific expression of metabolic enzymes, structural proteins, and differential metabolites reshaped the post-mortem biochemical reactions of different muscles. The higher pH and b* values of the topside were attributed to a weakened glycolytic pathway, reduced lactic acid production, and active lipid metabolism; striploin and clod showed distinct characteristics in amino acid and nucleotide metabolism, resulting in differentiated flavor profiles. The molecules NME1 and MYH7 linked upstream protein expression and downstream metabolite changes, constructing a complete molecular regulatory chain for meat quality traits. Conclusions In conclusion, this study systematically expounded the multi-omics mechanism of meat quality variation in different muscle cuts of Pingliang Red cattle. The differential metabolites and proteins were distinguished among these muscle cuts. The meat quality analysis indicated that the topside had higher pH value and b* value than striploin and clod muscles, suggesting a phenotypic basis for the unique flavor characteristics. There were 29, 15, and 13 notable metabolites in topside, striploin, and clod, respectively, including 5-diphosphate, PS(18:0/20:4), and L-threonine. Proteomics detected that there were 35, 64, and 19 DEPs in topside, striploin, and clod, which were involved in glycine-serine-threonine metabolism, glycerophospholipid metabolism, and lysine degradation. The screened biomolecules are only preliminary candidate molecular markers that require systematic verification before practical application. Follow-up research will first expand the sample population and employ parallel reaction monitoring (PRM) proteomics and real-time quantitative polymerase chain reaction (RT-qPCR) to quantify core molecules and validate their consistent differential expression among muscles. Furthermore, hub genes NME1 and MYH7 will be used to screen trait-associated single nucleotide polymorphisms (SNPs), and population association analyses will be conducted to facilitate molecular-assisted breeding. CRediT authorship contribution statement Jiawei Lu: Writing – original draft, Investigation, Formal analysis. Junwei Hu: Writing – original draft, Investigation, Formal analysis. Jin Wang: Investigation, Formal analysis. Yanxiang Zhang: Investigation, Formal analysis. Yifan Ma: Investigation, Formal analysis. Rui Duan: Formal analysis. Xuemei Gao: Investigation. Tianliang Yang: Validation. Yongfeng Wang: Validation. Minghao Ma: Investigation. Shuzhen Li: Investigation. Lili Chen: Investigation. Yi Ma: Project administration, Funding acquisition. Lupei Zhang: Writing – review & editing, Project administration, Funding acquisition. Ethics declaration Animal subject This study was conducted in accordance with the following guidelines for animal welfare and/or reporting: All animal experiments were approved by the Animal Care and Use Committee of the Tianjin Institute of Animal Husbandry and Veterinary Medicine (2023003) and were performed in compliance with the committee's guidelines. This study was approved by the Animal Care and Use Committee of the Tianjin Institute of Animal Husbandry and Veterinary Medicine. (Approval No. 2023003) Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgments This work was supported by the Kongtong District East-West Science and Technology Cooperation Program (Tianjin-Gansu Collaboration) (PL-KT-2024-2-011), Gansu Provincial Special Program for East-West Science and Technology Cooperation (Tianjin-Gansu Collaboration) (24CXNL006), Integrated Innovation of Key Technologies for Precision and Rapid Propagation of Pingliang Red Cattle & Talent Training-Construction of an Integrated Embryo Engineering Platform, Special Fund for Central Government Guiding Local Science and Technology Development (24ZYCGSN00930), the Major Science and Technology Project for Innovation Consortium of Tianjin (25ZXCKQY00060). Footnotes Supplementary data to this article can be found online at https://doi.org/10.1016/j.fochx.2026.104408. Contributor Information Yi Ma, Email: tjnnyz@126.com. Lupei Zhang, Email: zhanglupei@caas.cn. Supplementary data Supplementary material Data availability Data will be made available on request. References Associated Data Supplementary Materials Supplementary material Data Availability Statement Data will be made available on request.