Transcriptomics and metabolomics insights into the seasonal dynamics of meat quality in yak on the Qinghai-Tibetan Plateau Authors: Tongqing Guo, Xungang Wang, Qian Zhang, Yuna Jia, Hongjin Liu, Linyong Hu, Na Zhao, Shixiao Xu, Yingzhu Duan, Ke Jia Published: 2024 ABSTRACT BackgroundMeat quality in yak is influenced by the fluctuation of nutritional composition in different grazing seasons on the Qinghai-Tibetan Plateau. However, the molecular mechanism underlying in yak meat remains unknown. Therefore, this study aimed to investigate the seasonal dynamics of meat quality in yak by transcriptomics and metabolomics techniques. Twelve healthy female yaks with a similar weight were divided into two groups, including the warm season group (WS) and cold season group (CS). After slaughter, samples of longissimus lumborum were collected and subjected to transcriptomics and metabolomics to explore the effects of different seasons on meat quality.ResultsYak in the WS group had higher contents of n-3 Polyunsaturated fatty acid (PUFA), n-6 PUFA, threonine, and valine compared to the CS group, but the pH45min and b* values were lower. A total of 75 differentially expressed metabolites in the longissimus lumborum muscle were identified, with 23 metabolites upregulated and 52 metabolites downregulated in the WS group. These metabolites were mainly enriched in the pathway of glycine, serine and threonine metabolism, tryptophan metabolism, and carbohydrate digestion and absorption. In comparison, the WS group exhibited 262 upregulated genes in the longissimus lumborum muscle and 81 downregulated genes relatives to the CS group, which were enriched in the fat deposition of TGF-beta, ECM-receptor interaction, MAPK, and PPAR signaling pathway.ConclusionsAmong these, downregulated genes NPNT, GADL1, SESN3, and CPXM1 were associated with lipid metabolism and fat deposition in grazing yaks. It was found that DDC, DHTKD1, CCBL1, GCDH, and AOC1 involved in the tryptophan metabolism played an important role in the regulation of energy metabolism in yak.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12864-024-11093-5. Background On the Qinghai-Tibetan Plateau, yaks (Bos grunniens), esteemed as “plateau treasures”, constitute a unique bovine species primarily inhabiting the Himalayan region [1]. In recent years, yak meat has gained popularity due to its high nutritional value, serving as a crucial component in the diet and health regimen of pastoralists. Yak meat is rich in protein, essential amino acids, and polyunsaturated fatty acids (PUFA), while low in fat content compared with cattle meat [2]. Recent studies on the pasture-grazed yak meat have highlighted the influence of grazing environments on meat composition, flavor, and nutritional properties [2]. Studies have shown that the diverse botanical composition of grazing pastures contributes to the unique flavor profile and nutritional richness of yak meat compared to grain-fed yak. However, traditional grazing remains the primary feeding approach for yaks. There is a series of issues that arise when yaks do not get enough grass to graze in the winter, leading them to lose weight. The growing season for pasture is short, approximately 120 days, and forage nutrients are insufficient to sustain the growing yak. Consequently, it is postulated that the fluctuation in nutritional composition significantly influences meat quality in grazing yaks. Meat quality and fatty acids are involved in many biological processes [3]. Previous research suggests that muscle metabolites and genes serve as primary phenotypic components, determining the physiological characteristics of muscles and pork meat quality [4]. Transcriptomic RNA-Seq sequencing has gained increasing popularity for assessing gene expression levels in muscle development, fat metabolism, and flavor formation, thereby providing a comprehensive understanding of the molecular processes shaping pork meat quality [5]. Furthermore, bioinformatics plays a crucial role in elucidating the molecular mechanisms contributing to developing specific characteristics, providing a theoretical basis for future animal production [6]. Xiong et al. identified that grazing yaks exhibited higher contents of n-3 PUFA and n-6 PUFA compared to stall-feeding yaks through the integration of lipidomic and metabolomics approaches [7]. Zheng et al. revealed the inhibition of hepatic adipogenesis and fat catabolism in yak for adaptation to forage shortage during cold season by the transcriptomic and metabolomic [8]. Wang et al. found the important biological pathways associated with longissimus dorsi muscle of Ashdan yaks under different feeding regimes using transcriptome and metabolome [9]. Hence, the integration of multi-omics analyses has the potential to provide critical insights that influence meat quality. This study hypothesized that different grazing seasons could affect the meat quality of longissimus lumborum by modulating the expression of genes and metabolites associated with muscle development. These findings would be provided fundamental insights for enhancing the meat quality of grazing yak across varying seasons. Additionally, these findings contribute to refining feeding regimes to produce high-quality yak meat. Results Muscle quality As illustrated in Fig. 1, the diameters of muscle fiber tended to be lower in the cold season (CS) group than in the warm season (WS) group (P = 0.061). However, there were no differences in the perimeters, areas, and densities of muscle fibers between the two groups (P > 0.05). It was observed that the pH45min and b* values were lower in the WS group compared to the CS group (P < 0.05, Fig. 2), while the L* value was higher (P < 0.001). Furthermore, the WS group exhibited higher drip loss compared to the CS group (P = 0.042). The profiles of fatty acids in the longissimus lumborum are presented in Table S3. The saturated fatty acid (SFA), such as C12:0, C14:0, and C17:0, were higher in the WS group than in the CS group (P < 0.05). In addition, the WS group had a higher concentration of SFA, monounsaturated fatty acid (MUFA), n-6 PUFA, n-3 PUFA, and total fatty acids (TFA) compared to the CS group, while the ratio of n-6 to n-3 was lower (P < 0.05). The profiles of amino acids in the longissimus lumborum are described in Table S4. The contents of threonine and valine were higher in the WS group compared to the CS group (P < 0.05). No differences were found in other individual amino acids among the experiment treatments (P > 0.05). Fig. 1Comparison of muscle fiber index between WS and CS groups Comparison of muscle fiber index between WS and CS groups Fig. 2Comparison of meat color and meat quality index between WS and CS groups Comparison of meat color and meat quality index between WS and CS groups Metabolomic analysis Latent structures discriminant analysis (OPLS-DA) in the WS and CS groups indicates that the samples had a distinct separation and were no overlapping areas. The R2Y and Q2 values, obtained from response permutation testing, were 0.998 and 0.79, respectively (Fig. 3A). The upward trend in regression line demonstrated the stable and effectiveness of the model. In total, 669 metabolites were observed in muscle samples from both the WS and CS groups (Fig. 3B). Among them, 75 differential metabolites (DMs) were observed based on the screening criteria. Compared to the CS group, 23 DMs were upregulated and 52 DMs were downregulated in the WS group. With respect to muscle, a total of 18 DMs were significantly associated with 28 metabolic pathways. Figure 3C illustrates the enrichment of the top 20 pathways. Among these, the most significantly enriched pathways included glycine, serine and threonine metabolism, tryptophan metabolism, and carbohydrate digestion and absorption. Fig. 3Metabolic spectrum of longissimus in the WS and CS groups analyzed by LC-MS. (A) PCA diagram of the WS and CS group. (B) Volcanic diagram of DMs; red represents the significantly upregulated DMs, green represents the significantly downregulated DMs, and gray represents the non-significant DMs. (C) Bubble diagram of the top 20 positions of KEGG pathway enrichment in the comparison between WS and CS group. The ordinate is the name of metabolic pathway, and the abscissa is the enrichment factor ratio Metabolic spectrum of longissimus in the WS and CS groups analyzed by LC-MS. (A) PCA diagram of the WS and CS group. (B) Volcanic diagram of DMs; red represents the significantly upregulated DMs, green represents the significantly downregulated DMs, and gray represents the non-significant DMs. (C) Bubble diagram of the top 20 positions of KEGG pathway enrichment in the comparison between WS and CS group. The ordinate is the name of metabolic pathway, and the abscissa is the enrichment factor ratio Transcriptome profiling of muscle In RNA-seq sequencing, a total of 73.73 G of clean data was obtained (Table S5). The effective data size of the samples ranged from 5.87 to 6.6 G, with a Q30 distribution ranging from 91.71 to 93.25%. Additionally, a great number of genes were detected in the ranges of 0–1, 3–10, and 3–15 FPKM (Fig. S1); relatively fewer genes were observed in the ranges of 15–60 and above 60 FPKM. The PCA diagram demonstrated that samples clustered within a narrow range (Fig. 4A). The differential expression analysis showed a total of 343 differentially expressed genes (DEGs) in longissimus lumborum (Fig. 4B). More specifically, 262 genes exhibited higher expression levels, while 81 genes showed lower expression levels in the WS group relative to the CS group. Fig. 4Functional enrichment analysis of differentially expressed genes (DEGs). (A) PCA diagram of the comparison of WS and CS groups; (B) Volcano diagram of DEGs in the comparison between WS and CS group; (C) GO term enrichment for DEGs in the WS vs. CS comparison Functional enrichment analysis of differentially expressed genes (DEGs). (A) PCA diagram of the comparison of WS and CS groups; (B) Volcano diagram of DEGs in the comparison between WS and CS group; (C) GO term enrichment for DEGs in the WS vs. CS comparison Genes were classified into three major categories: biological process (BP), cellular component (CC), and molecular function (MF). As depicted in Fig. 4C, the DEGs were classified into 30 subcategories, including 10 BP terms, 10 CC terms, and 10 MF terms. In the BP category, the genes were involved in the chromosome organization and protein modification by small protein conjugation or removal category (GO:0051276, GO:0070647). In the CC category, the genes were annotated in the nucleus category (GO:0005634). In the MF category, the genes were located in the cytoskeletal protein binding category (GO:0008092). The identified DEGs were categorized into different pathways using the Kyoto Encyclopedia of Genes and Genomes (KEGG) bioinformatic database (P < 0.05). As shown in Fig. 5, the enrichment analysis indicated that the top 20 enriched terms were primarily associated with the metabolism of the TGF-beta, ECM-receptor interaction, glucagon, MAPK, Glycolysis/Gluconeogensis, PPAR, and HIF-1 signaling pathway. Fig. 5Bubble chart of the top 20 pathway in KEGG analysis of DEGs Bubble chart of the top 20 pathway in KEGG analysis of DEGs Integration analysis of metabolome and transcriptome In the correlation analysis (Fig. 6), there was a strong correlation between DMs and DEGs. Metabolites such as 1-Naphthol, VPH, L-Tryptophan, 23-Nordeoxycholic acid, 3-(propan-2-yl)-octahydropyrrolo[1,2-a]pyrazine-1,4-dione, L-Argininosuccinate, 4-Hydroxyisoleucine, Loxoprofen, o-Cresol, and Asp-glu were positively correlated with PATL2, GPA33 and ABCA1, and negatively correlated with the genes of NPNT, LOC102274962, LOC102271093, IGFS3, CCDC69, CCDC88C, GADL1, SESN3, CPXM1, IGDCC4, LOC102280840, CIB2, ST8SIA2, AMOTL1, LOC102270897, GPI, and SLC9A5. The metabolites of Ala-lle, 4-ethoxy-N-[(4-methoxyphenyl)methylene]aniline, 3-(methylsulfanyl)-5 H- [1, 2, 4]triazino[5,6-b]indole, D-Ala-D-Ala, and N1-isopropyl-2-(1 H-2-pyrrolylcarbonyl)1-hydrazinecarboxamide were positively associated with the genes of NPNT, LOC102274962, LOC102271093, IGFS3, CCDC69, CCDC88C, GADL1, SESN3, CPXM1, IGDCC4, LOC102280840, CIB2, ST8SIA2, AMOTL1, LOC102270897, GPI, and SLC9A5, while negatively correlated with the genes of PATL2, GPA33, and ABCA1. Fig. 6The correlation heat maps of top 20 DMs and DEGs in the WS and CS groups. Red indicates positive correlation and blue indicates negative correlation The correlation heat maps of top 20 DMs and DEGs in the WS and CS groups. Red indicates positive correlation and blue indicates negative correlation KEGG enrichment results showed that DEGs and DMs were mainly concentrated in the tryptophan metabolism pathway (Fig. 7A). In the context of tryptophan metabolism (Fig. 7B), it was found that the metabolite of N-formylkynurenine was positively correlated with DHTKD1, while it was negatively correlated with the CCBL1, GCDH, and CAT. The metabolite of L-tryptophan was positively associated with the CAT and negatively associated with the DDC, DHTKD1, and AOC1. L-kynurenine was positively correlated with the DDC, DHTKD1, and AOC1 and negatively correlated with the CAT and GCDH. The pathway map of highly correlated DEGs and DMs in tryptophan metabolism are shown in Fig. 7C. Fig. 7KEGG enrichment for DEGs and DMs in the WS vs. CS comparison. Bubble chart of the top 20 pathway (A); The correlation analysis of DEGs and DMs in the tryptophan metabolism, red lines indicate positively correlation, green lines indicate negatively correlation (B). Pathway map of highly correlated DEGs and DMs in tryptophan metabolism (C) KEGG enrichment for DEGs and DMs in the WS vs. CS comparison. Bubble chart of the top 20 pathway (A); The correlation analysis of DEGs and DMs in the tryptophan metabolism, red lines indicate positively correlation, green lines indicate negatively correlation (B). Pathway map of highly correlated DEGs and DMs in tryptophan metabolism (C) Detection of DEGs expression by quantitative real time polymerase chain reaction (qRT-PCR) In this study, 6 DEGs of longissimus lumborum were selected and their expression levels were performed by qRT-PCR verification (Fig. 8). The results demonstrated that the relative gene expression levels obtained from qRT-PCR were consistent with those from transcriptome sequencing, confirming the reliability and effectiveness of the sequencing. Fig. 8The expression pattern of DEGs was detected by qRT-PCR. (A) NPNT; (B) CPXM1; (C) DDC; (D) GADL1; (E) CCBL1; (F) GCDH The expression pattern of DEGs was detected by qRT-PCR. (A) NPNT; (B) CPXM1; (C) DDC; (D) GADL1; (E) CCBL1; (F) GCDH Discussion As an important source of protein in herdsman diets, the quality and taste of yak meat affect the consumers attention. Muscle fiber with a smaller size had a lower shear force and higher tenderness. Muscle fiber size is a crucial determinant of meat quality, which decides the size of muscle bundles [10]. Additionally, the development of muscle fibers is primarily influenced by the genetic and environmental factors, such as gender, muscle type, age, and hormones. No differences were found in the indexes of diameter, perimeter, and area in this study. Lengerken et al. found that pigs with smaller fiber size tended to exhibit higher muscle pH45min and lower drip loss [11]. Drip loss and cooking loss are crucial indicators of water-holding capacity, which determines the meat yield and eating quality to a certain extent [12]. In this study, the grazing yaks in the CS group exhibited higher pH45min and lower drip loss in the longissimus lumborum muscle, indicating better water holding capacity compared to the WS group. Welzenbach et al. identified that the glycolysis/gluconeogenesis pathway significantly influence on drip loss [13]. Hence, the lower drip loss observed in the CS group could be attributed to the differential genes enriched in the glycolysis/gluconeogenesis pathway. Color is one of the decisive factors in the selection of meat and is directly related to the content of myoglobin. The higher a* as well as lower L* and b* values were associated with better muscle color within a defined range. Muscle from pastured steers was darker than grain-finished steers due to the dietary effects [14]. Similarly, meat from fattened yak had higher L* values than grazing yaks [2]. Maybe due to the reason the fat content from fattened yak meat was higher than grazing yak, resulting in the higher L* value of fattened yak meat. In the current study, meat in the WS group exhibited higher L* values and lower b* values in our study, which attributed to the relationship between lipid and pigment oxidation, particularly the instability of PUFAs [15]. In the current study, the WS group had higher contents of SFA, MUFA, n-6 PUFA, n-3 PUFA, and TFA compared to the CS group, although the ratio of n-6 to n-3 was lower. The fatty acids composition of yak meat is influenced by the nutrient composition of grass in different season. The findings suggest that the lower fat deposition in yak meat may result from the cold plateaus where yaks reside, necessitating a substantial fat content for body temperature regulation. Liu et al. found that the longissimus lumborum of grazing yaks had higher levels of MUFA, PUFA, and the ratio of PUFA to SFA than that of the fattening yaks [16]. In fact, ruminants feeding pasture has higher PUFA, particularly n-3 PUFA compared with their grain-fed counterparts [17–19]. In addition, secondary metabolites in pasture may act as a protective mechanism against microbial biohydrogenation of PUFA, facilitating their increased absorption and deposition [20]. In recent years, consumers have increasingly concerned about the health fatty acid ratio in meat [21], which is an important indicator for evaluating the nutritional quality of food [22]. It is noteworthy that PUFA are crucial in preservation the integrity of cell membranes, controlling the expression of genes, balancing cytokines and lipoproteins, protecting against cardiovascular diseases, and supporting growth and development [23]. The recommended ratio of PUFA/SFA in a healthy diet is above 0.4 [2]. In the present study, the ratio of PUFA to SFA was 1.18 and 0.68 in the WS and CS groups, respectively, as they reached the recommended values. The ratio of n-6/n-3 is associated with the risk factor in cancers and coronary heart disease, with recommendations suggesting it should be below 4.0 [24]. It was found that the ratio of n-6/n-3 was 1.48 and 2.03 in the WS and CS groups, respectively, which also reached the recommended standard. The current data indicate that the grazing yak in the WS group had higher PUFA content and a favorable ratio of PUFA to SFA, which contributed to the consumers health. It was found that the contents of threonine, valine, and isoleucine were higher in the WS group than those in the CS group. Branched-chain amino acids (BCAAs: valine, isoleucine, and leucine) are structurally related to branched-chain fatty acids and serve as both building blocks for nitrogenous compounds and signaling molecules that control the metabolism of glucose, lipids, as well as protein synthesis. BCAA and their catabolized products (succinyl-CoA, acetyl-CoA, acetoacetate), some of which might enter the tricarboxylic acid cycle, are used to partially meet the increased energy and nutrient demands [25]. These findings indicated that the BCAAs play a role in modulation of energy metabolism in yaks grazing during different seasons. Through untargeted metabolomics, we found that the enriched pathways of glycine, serine and threonine metabolism, tryptophan metabolism, and carbohydrate digestion and absorption metabolism were different between WS and CS groups. Within these pathways, the metabolites of L-tryptophan, L-serine, L-kynurenine, N-formylkynurenine, hippuric acid, phenylacetylglutamine, L-tyrosine, and xanthosine in the WS group were upregulated, whereas the metabolites of betaine, sucrose, maltotriose, and theophylline were downregulated. The results for L-tyrosine and L-serine were inconsistent with the targeted analysis of amino acid concentrations. However, the upregulated sucrose and maltotriose in the CS group were enriched in the pathway of carbohydrate digestion and absorption. As such, it is possible that the upregulation of carbohydrate digestion and absorption was a potential indicator of fat deposition impediment [26]. This finding was inconsistent with the results of fatty acids. It was speculated that the upregulation of carbohydrate digestion and absorption was not used the fat deposition but for energy requirement to live on a cold plateau. The glycine, serine, and threonine metabolism are involved in regulating energy metabolism and organic acids, providing essential precursors for the citric acid cycle [27]. Threonine is transformed into glycine and serine, while betaine is a methylated form of glycine [28]. In addition, betaine activates AMPK, which enhances genes encoding proteins involved in fatty acid transport and fatty acid oxidation, while decreasing fatty acid synthesis [29]. In this study, the upregulated betaine in the CS group may inhibit the fatty acid synthesis of yak in the cold season. Domestic animal muscle undergoes a sophisticated multi-step growth and development process influenced by various factors, with a particular focus on genes associated with fat deposition [30]. Compared to the CS group, the WS group exhibited 530 upregulated genes and 228 downregulated genes. The candidate DEGs of NPNT (Nephronectin), GADL1 (glutamate decarboxylase like 1), SESN3 (Sestrin 3), and CPXM1 (Carboxypeptidase X, M14 Family Member 1) were found to be downregulated in the CS group and associated with fat deposition. The NPNT gene, encoding an extracellular matrix protein that plays a role in cell adhesion, has been found to enhance myoblast fusion during cellular differentiation [31]. Furthermore, NPNT contributes to the fatty acid composition in skeletal muscle [32]. Clark et al. observed a decrease in the bovine gene NPNT in highly marbled animals, indicating its involvement in intramuscular fat deposition within skeletal muscle [33]. Additionally, there has been a report indicating a positive correlation between the NPNT gene and total n-3 PUFA content [34]. GADL1 is an enzyme known for its versatile biological function and plays a crucial role in the conversion aspartic acid to β-alanine in carnosine biosynthesis [35]. Nevertheless, in humans, overexpression of GADL1 inhibited cell migration and led to morphological alterations [36]. Additionally, it has been found that this gene was enriched in a specific pathway related to the gamma-aminobutyric acid synthesis associated with meat tenderness [37, 38]. In steers, the GADL1 gene was associated with lipid metabolism and carcass. SESN3 is a part of the SESNs family by modulating the AMPK-mTOR axis, which maintains metabolic balance and prevents the metabolic syndrome linked to aging and obesity [39]. Increased expression of SESN3 can enhance the triglycerides accumulation. On the other hand, a significant decrease in triglyceride accumulation was observed with the downregulation of SESN3 in cells [40]. In pigs, a previous study showed that SENS3 prevents preadipocyte development [37]. CPXM1 belongs to the carboxypeptidase family, which is a glycoprotein secreted in N-glycosylation and can bind to collagen [41]. Moreover, CPXM1 serves as a positive regulatory factor of fat formation, promoting adipose tissue proliferation through its impact on extracellular matrix remodeling [41, 42]. CPXM1 has been identified as a potential gene for fat deposition in sheep tail, which is also reported as a positive regulator of adipogenesis in mice and humans [42]. Ramos et al. reported found that CPXM1 was involved in fat storage and metabolism and correlated with body condition score [43]. Therefore, the downregulated NPNT, GADL1, SESN3, and CPXM1 may play a crucial role in the fat deposition, along with the decrease in PUFAs, in grazing yaks transitioning from the warm season to the cold season. According to that analysis, most of genes focused on the hexose metabolic process, and monosaccharide metabolic process, system development, and DNA replication, which was consistent with the muscle as an important tissue in the biological process of lipid metabolism and cell growth. Similarly, we observed that some signaling pathways related to fat deposition were significantly enriched, such as TGF-beta, ECM-receptor interaction, MAPK, and PPAR. These results were in agreement with previous findings, Fei et al. identified that the relative metabolism pathway was associated with fat deposition between different tail sheep breeds, including TGF-beta, ECM-receptor interaction, MAPK, and PPAR [44]. The TGF-beta signaling pathway plays a role in the regulation of cell proliferation, differentiation, tissue development, and apoptosis [45]. The PPAR signaling pathway consisted of three different PPAR isotypes (PPARα, PPARβ/δ, and PPARγ) and highly expression in tissues, regulating fatty acid catabolism in the skeletal muscles and cellular energy metabolism [46, 47]. MAPK signaling pathway, including extracellular signal-regulated kinase (Erk), p38, and cJUN NH2-terminal kinase (JNK), is crucial in adipocyte proliferation and differentiation. It can impact lipid metabolism by regulating genes related to lipogenesis, lipolysis, and fatty acid oxidation [48]. In addition, JNK signaling stimulates cell proliferation by activating the AP1 transcription factor [49]. The primary components of Extracellular matrix (ECM)-receptor interaction signaling pathway in adipose tissue consist of collagen, fibronectin, laminin, hyaluronan, and proteoglycan. During development, the ECM performs functions that provide structural support for cells and direct cell differentiation [50]. ECM components are synthesized in the cytoplasm and subsequently secreted into the extracellular region for further modifications into their final structures [51]. These studies suggested that the ECM-receptor interaction signaling pathway is vital for tissue architecture and plays a crucial role in adipogenesis. Therefore, according to the enrichment information of these signaling pathways, it could potentially serve as an important mechanism regulating differences in fat deposition and lipid metabolism in grazing yak. Protein-protein interaction (PPI) network offers a conceptual structure for better comprehension of the roles of DEGs and DMs. In this study, the PPI network analysis revealed a close relationship between 6 DEGs and 3 DMs involved in tryptophan metabolism. Tryptophan, the first amino acid identified, stimulates weight gain for normal growth in mice [52]. Tryptophan is metabolized along the kynurenine pathways, producing an array of kynurenine metabolites such as kynurenic acids and N-formylkynurenine [53]. In this study, L-kynurenine was positively correlated with the DDC, DHTKD1, and AOC1. L-tryptophan was converted into 5-hydroxytryptophan (5-HTP), where DDC (L-DOPA decarboxylase) catalyzes its conversion to 5-hydroxytryptamine [54]. Approximately 95% of tryptophan is metabolized via the kynurenine pathway. The upregulation of DDC may limit 5-hydroxytryptamine to generate the serotonin and enhance the kynurenine pathway. The kynurenine pathway has two major branches, terminating in the production of the neuroactive metabolite kynurenic acid or nicotinamide adenine dinucleotide (NAD+), respectively. NAD+ is an important co-factor in cellular reactions linked to energy metabolism and enzymatic redox [55]. In mice, restricted energy intake decreased tryptophan-kynurenine pathway activity, whereas a high-fat diet increased it [56]. These findings suggest that tryptophan metabolism significantly regulates energy metabolism in grazing yaks. In addition, CCBL1 (Kynurenine aminotransferase 1) encodes for kynurenine aminotransferase, converting L-kynurenine and L-3-hydroxykynurenine into kynurenate and xanthurenate, respectively [57], which was involved in the regulation of immune response. DHTKD1 (Transketolase domain containing 1), is associated with lysine and tryptophan catabolism [58]. Lysine and tryptophan are degraded via a shared pathway in which the dehydrogenase DHTKD1 catalyzes the synthesis of the intermediate glutaryl CoA [59]. Subsequently, GCDH (Glutaryl-CoA Dehydrogenase) converts glutaryl CoA to crotonyl CoA, metabolized to acetyl CoA to enter the tricarboxylic acid cycle [60]. In cancer cells, DHTKD1 affects the switch of metabolism towards glycolysis [61]. Wu et al. reported that GCDH was related to meat quality, especially beef meat color [62]. AOC1 (Amine oxidase copper-containing 1) is a secreted glycoprotein catalyzing the degradation of putrescine and histamine [63]. AOC1 plays an important role in physiologic and pathological processes by regulating polyamine homeostasis [64]. Given the above finding, the DDC, DHTKD1, CCBL1, GCDH, and AOC1 were involved in tryptophan metabolism and associated with the regulation of energy metabolism and immune response in grazing yaks. Conclusions In this study, the meat quality and muscle fiber of longissimus lumborum were measured in grazing yak during different seasons, and the transcriptomes and metabolomics of longissimus lumborum were comprehensively analyzed. Yak in the warm season had higher contents of PUFAs . Down-regulated genes NPNT, GADL1, SESN3, and CPXM1 were associated with lipid metabolism and fat deposition in grazing yak. Through comprehensive analysis of transcriptomics and metabolomics, it was found that DDC, DHTKD1, CCBL1, GCDH, and AOC1 involved in the tryptophan metabolism played an important role in the regulation of energy metabolism in grazing yak. Thus, our findings contribute to a better understanding of the molecular mechanisms underlying meat quality in grazing yaks across different seasons. Materials and methods Animal and sample preparation The experimental animals were selected from Haibei Demonstration Zone of Plateau Modern Ecological Animal Husbandry Science and Technology, Haiyan County, Qinghai Province, China. It has a plateau continental climate with an average altitude of 3100 m. The average annual temperature is 1.5℃, and the average precipitation is 400 mm. The annual sunshine duration is 2980 h, and the growth period of pasture is 100–150 days. On the Qinghai-Tibetan Plateau, May to October is considered the warm season, and November to April is as the cold season. At the beginning of the experiment, a total of twelve healthy female yaks with similar birth dates and weights (110.42 ± 13.13 kg) were selected from a herd of 200 yaks. Due to the special geographical conditions of the Qinghai-Tibetan Plateau, yaks were divided into two groups: warm season group (WS, n = 6) and cold season group (CS, n = 6). Throughout the year, yaks grazed freely in the same natural grassland located in Haiyan County, Qinghai Province (E100°95′, N36°91′, 3100 m altitude). All yaks fed ad libitum to grass and had free access to fresh water. The primarily consisted of alpine meadow, which includes Kobresia pygmaea, Kobresia humilis, Elymus dahuricus, and Poa annua. Nutrients and chemical composition in the grass are presented in Table S1. At the end of the warm season and cold season, six yaks with similar weights were selected from each group following 12 h fasting and water deprivation, and transported to a nearby commercial slaughterhouse (Xiahua Halal Meat Food Co., Ltd., Qinghai, China). The slaughter processes consisted of stunning with a pneumatic hammer and subsequent exsanguination (5 min). Then, within 30 min of slaughter, samples of longissimus lumborum were collected, ensuring the removal of visible external fat and connective tissue. Concurrently, the meat quality of longissimus lumborum was assessed, and samples were promptly preserved at -80 °C until subsequent analysis. Meat quality and muscle fiber The meat quality in longissimus lumborum were evaluated at 45 min. The pH of the muscle samples was measured using a pH meter (Matthaus pH Star, Germany). Meat color was assessed using the Germany Matthaus meat color tester Opto-Star, and the assessment of meat color was conducted utilizing three parameters by the Minolta system (L*, a*, and b*). Drip loss was determined at 45 min post-slaughter following the procedure of Li et al. [65]. Cooking loss was performed in accordance with the method described by Bertram et al. [66]. The characteristics of myofibers in the longissimus lumborum were assessed using the traditional method of hematoxylin and eosin (H&E) according to the procedure of Guo et al. [67]. Total RNA extraction and transcriptome sequencing analysis Total RNA was isolated from muscle tissue samples using Trizol reagent kit (Invitrogen, USA), according to the manufacturer’s instructions. RNA concentration was measured using the Qubit® RNA Assay Kit with a Qubit2.0 Fluorometer (Life Technologies, CA, USA). RNA integrity was assessed using an RNA Nano 6000 Assay Kit on the Bioanalyzer 2100 system (Agilent Technologies, CA, USA). Sequencing was performed on the platform of Illumina NovaSeq 6000 (Illumine, USA), generating 150 bp paired-end reads. Low-quality reads were filtered out to obtain the clean reads for subsequent analysis. Clean reads were mapped to the genome reference of wild yak (https://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/000/298/355/GCF_000298355.1_BosGru_v2.0/) using HISAT2 v 2.0.5. Read numbers of each gene was obtained by featureCounts v1.5.0-p3. FPKM of each gene was calculated, and reads counts were obtained by HTSeq-count. Differential expression analysis was calculated using the edge R package (3.22.5). P values were adjusted using the Benjamini & Hochberg method. A significant expression gene was considered with a P-value of 0.05 and absolute foldchange of 2. Gene Ontology (GO) enrichment analysis and KEGG pathways for DEGs were conducted using the clusterProfiler R package. Metabolites extraction and ultra-high performance liquid chromatography coupled to tandem mass spectrometry (UHPLC-MS/MS) analyses Metabolites extraction of samples and UHPLC-MS/MS analysis were performed following the procedure outlined by Want et al. [68]. A Vanquish UHPLC system (ThermoFisher, Germany) coupled with an Orbitrap Q ExactiveTM HF mass spectrometer (Thermo Fisher, Germany) was used to perform UHPLC-MS/MS analysis (Novogene Co., Ltd Beijing, China). Samples were injected into a Hypersil Gold column (100 × 2.1 mm, 1.9 μm) with a 17-min linear gradient at a flow rate of 0.2 mL/min. The water with 0.1% formic acid was as eluent A and methanol was as eluent B in positive polarity mode. In the negative polarity mode, 5 mM of ammonium acetate (pH 9.0) was as eluent A and methanol was as eluent B. Solvent gradient conditions were set as follows: 2% B, 1.5 min; 2–85% B, 3 min; 85–100% B, 10 min; 100-2% B, 10.1 min; 2% B, 12 min. The Q ExactiveTM HF mass spectrometer was operated in positive/negative polarity mode with a spray voltage of 3.5 kV, capillary temperature of 320 °C, sheath gas flow rate of 35 psi, aux gas flow rate of 10 L/min, and aux gas heater temperature of 350 °C. Data processing and metabolite identification Raw data files generated were processed using Compound Discoverer 3.1 (CD3.1, Thermo Fisher) to perform peak alignment, peak picking, and quantitation for each metabolite. The normalized data was used to predict the molecular formula based on additive ions, molecular ion peaks, and fragment ions. Subsequently, peaks were matched with the mzCloud, mzVault, and MassList databases to obtain accurate qualitative and relative quantitative results. Generated data were imported into SIMCA-P software (version 14.1, Umetrics AB, Umea, Sweden) for OPLS-DA analysis. Metabolites with VIP > 1 and P-value of t-test < 0.05 and fold change ≥ 2, or FC ≤ 0.5 were considered as significantly differential metabolites. Volcano plots were used to filter metabolites based on log2(Fold Change) and -log10(p-value) via ggplot2 in R language. The functions of these metabolites and metabolic pathways were investigated using the KEGG database. Fatty acids extraction and data processing Samples were suspended in liquid nitrogen and then diluted in water by thorough vertexing. Then, 100 µL of each diluted samples were homogenized with 300 µL of isopropanol/acetonitrile (1:1) containing mixed internal standards, followed by centrifugation at 12,000 rpm for 10 min. Finally, the supernatant was injected into the LC-MS/MS system for analysis. Fatty acids were quantified using a UHPLC-MS/MS system (ExionLC™ AD UHPLC-QTRAP 6500+, AB SCIEX Corp., Boston, MA, USA) by Novogene Co., Ltd. Beijing, China. Separation was conducted on a Waters ACQUITY UPLC BEH C18 column (2.1 × 100 mm, 1.7 μm), maintained at 50 °C. The mobile phase consisted of 0.05% formic acid in water as solvent A and isopropanol/acetonitrile as solvent B. The gradient was programmed as follows: initial 30% B, 1 min; 30–65% B, 2 min; 65–100% B, 11 min; 100% B, 13.5 min; 100 − 30% B, 14 min; 30% B, 15 min. The mass spectrometer was operated in negative multiple reaction mode (MRM). Instrument parameters for the UHPLC-MS/MS system were configured following the methodology outlined by Guo et al. [69]. Amino acid extraction and data processing Meat sample was suspended in liquid nitrogen and then diluted in water through well vertexing. Then, 50 µL of it was taken and homogenized with 200 µL of acetonitrile/methanol (1:1) containg mixed internal standards through thorough vortexing. After that, the mixture was centrifuged at 12,000 rpm for 10 min. Finally, the supernatant was injected into the LC-MS/MS system for analysis. A UHPLC-MS/MS system (ExionLC™ AD UHPLC-QTRAP 6500+, AB SCIEX Corp., Boston, MA, USA) was used to quantity amino acids (Novogene Co., Ltd. Beijing, China), which was performed on an ACQUITY UPLC BEH Amide column (2.1 × 100 mm, 1.7 μm). The mobile phase consisted of 0.1% formic acid in 5 mM ammonium acetate (solvent A) and 0.1% formic acid in acetonitrile (solvent B), and delivered at a flow rate of 0.30 mL/min. Solvent gradient and mass spectrometer was operated following the procedure outlined by Guo et al. [69]. RT-qPCR analysis Six DEGs were validated via qRT-PCR, and GAPDH was used as the internal reference gene. Gene primers are presented in Table S2. qRT-PCR was conducted using the SLAN 96 S real-time PCR instrument (Hongshi, Shanghai, China), with three replicates per gene. The reaction procedure began at 95℃ 2 min for activation, followed by 40 cycles at 95℃ for 15 s, 58℃ for 30 s, and 60℃ for 10 s for amplification. Relative gene expression was calculated using 2−ΔΔCt method. Statistics analysis The top 20 of DMs and DEGs were selected in transcriptome and metabolome, respectively. Pearson analysis was used to determine the relationship between the DEGs expression level and DMs response intensity data. Shared DEGs/DMs between transcriptome and metabonomic data using the KEGG database to identify common metabolic pathways. Then, in this metabolic pathway, the DEGs and DMs were used to perform the Pearson correlation analysis. A correlation network diagram was constructed using Cystoscope v3.9.1 based on statistically significant results with P-values < 0.05 [70]. The paraments of muscle fiber, meat color, and meat quality were analyzed using SPSS 24.0 (IBM, Armonk, NY, United States). Independent-samples t-tests was conducted to analysis the significant differences between WS and CS groups. A P ≤ 0.05 was considered statistically significant, while tendency for a treatment effect was noted when 0.05 < P ≤ 0.10. All figures in this study were generated using GraphPad Prism version 8.0.2 (GraphPad Software Inc., CA, USA). Electronic supplementary material Below is the link to the electronic supplementary material. Supplementary Material 1: Supplementary Table S1. Nutrients and chemical composition in the grass (%, DM basis). Supplementary Table S2. Primers for qRT-PCR of Longissimus lumborum. Supplementary Table S3. Effects of the different pastoral season on the content of fatty acids in the longissimus lumborum in yaks. Supplementary Table S4. Effects of the different pastoral season on the content of amino acids in the longissimus lumborum in yaks. Supplementary Table S5. Sequencing data evaluation statistics. Supplementary Material 1: Supplementary Table S1. Nutrients and chemical composition in the grass (%, DM basis). Supplementary Table S2. Primers for qRT-PCR of Longissimus lumborum. Supplementary Table S3. Effects of the different pastoral season on the content of fatty acids in the longissimus lumborum in yaks. Supplementary Table S4. Effects of the different pastoral season on the content of amino acids in the longissimus lumborum in yaks. Supplementary Table S5. Sequencing data evaluation statistics. Supplementary Material 2: Supplementary Figure S1. Distribution of FPKM expression in the WS and CS groups. Supplementary Material 2: Supplementary Figure S1. Distribution of FPKM expression in the WS and CS groups.