Comparative Metabolomics and Lipidomics of Meat from Duroc × Guangdong Small-Eared Spotted Pigs and Commercial Duroc × (Landrace × Yorkshire) Pigs

TL;DR

This study compares the meat quality and lipid profiles of two pig breeds, finding that indigenous crossbred pigs have better tenderness and distinct metabolomic and lipidomic profiles.

Contribution

The study provides novel molecular insights into meat quality differences between indigenous and commercial crossbred pigs using metabolomics and lipidomics.

Findings

DG pigs showed better tenderness but lower meat color and marbling scores compared to DLY pigs.

Metabolomics identified 13 differential metabolites, including L-norleucine and L-phenylalanine.

Lipidomics revealed 77 differential lipids, mainly triglycerides and ceramides, with 76 more abundant in DG pigs.

Abstract

Crossbreeding with indigenous breeds is an important approach for improving pork quality. In this study, untargeted metabolomics and targeted lipidomics were applied to comprehensively characterize meat quality, metabolites, and lipids in Duroc × Guangdong small-ear spotted (DG) and commercial Duroc × (Landrace × Yorkshire) (DLY) pigs. Multivariate statistical analysis was used for differential comparison, compound screening, and breed discrimination. DG pigs presented better tenderness than DLY pigs, although their meat color and marbling scores were lower. Protein, amino acid, and fatty acid contents did not differ significantly between breeds (p > 0.05), but their metabolomic and lipidomic profiles showed marked differences. Metabolomics identified 13 differential metabolites, such as L-norleucine and L-phenylalanine. Lipidomics revealed 77 lipids with differential abundance between the two breeds, predominantly triglycerides and ceramides, with 76 being more abundant in DG pigs. KEGG enrichment analysis showed that amino acid metabolism was the main pathway enriched by the differential metabolites, whereas the differential lipids were primarily involved in glycerolipid metabolism and other related pathways. Correlation analysis indicated that breed influenced relationships among meat quality traits, metabolites, and lipids. These findings offer molecular insights into the meat quality characteristics of indigenous crossbred pigs.

Figures

Full text

1. Introduction

As a key animal-derived food source in the human diet, pork is rich in nutrients, including protein, B-complex vitamins, and essential minerals [1]. China stands as the largest global producer and consumer of pork, with an increasing demand for high-quality products [2,3]. This rising demand has driven targeted research on the quality characteristics of pork, focusing particularly on flavor, nutritional composition, and the metabolic mechanisms that influence these characteristics [4]. The genetic background of pigs plays an important role in determining pork quality [5]. Utilizing high-quality indigenous pig breeds for crossbreeding improvement is an essential breeding strategy to improve pork quality [6,7,8]. The Guangdong small-eared spotted pig, an indigenous breed native to South China, is known for its adaptability to low-input rearing systems, strong stress tolerance, and superior meat flavor [9]. This breed is frequently utilized as the maternal line in crossbreeding systems to improve both meat quality and production performance [10,11]. To date, research on crossbred pigs has primarily focused on meat quality traits and nutritional differences between breeds [12,13]. Nevertheless, comprehensive studies at the molecular level remain limited.

Metabolomics is a robust analytical technology for the comprehensive profiling of metabolites in biological systems [14]. Using high-resolution instruments such as liquid chromatography-mass spectrometry (LC–MS), untargeted metabolomics enables the simultaneous screening of multiple metabolites and provides a comprehensive understanding of the metabolic characteristics of complex biological samples [15]. This approach has been widely applied to compare metabolite profiles between different pig breeds and to identify differential metabolites [16,17]. Lipids are an important class of metabolites in meat and are closely associated with meat flavor and tenderness [18]. As a key branch of metabolomics, targeted lipidomics enables the accurate identification and quantification of pork lipid molecules, offering molecular-level insights into meat quality variations from a lipid metabolism perspective. This approach also facilitates the discovery of lipid markers capable of discriminating among pig breeds [19,20]. The combination of untargeted metabolomics and lipidomics has been increasingly employed to characterize molecular differences among pig breeds. These studies typically compared comprehensive molecular profiles between indigenous and commercial crossbred pigs and identified key differential compounds [21,22,23,24].

In recent years, the integration of multi-omics approaches has emerged as a powerful strategy to elucidate the complex molecular mechanisms underlying meat quality. For instance, integrated untargeted metabolomics, transcriptomics, and flavoromics technologies have been used to investigate the molecular characteristics in different parts of the Guangdong small-ear spotted pig and to map potential metabolic pathways involved in flavor formation [25]. Integrated transcriptomic and metabolomic analysis has also been employed to explore molecular mechanisms of pig muscle fiber type transformation in pigs [26] and to reveal potential genetic-metabolic interaction mechanisms underlying meat quality in cattle [27].

Herein, we present an integrated analysis using untargeted metabolomics and quantitative lipidomics to characterize key differences between a local Chinese crossbreed (Duroc × Guangdong small-ear spotted, DG) and a global commercial breed (Duroc × (Landrace × Yorkshire), DLY). The study encompasses meat quality traits, nutritional composition, small-molecule metabolites, and lipid profiles, aiming to provide a molecular-level explanation for phenotypic differences between pig breeds. These insights are crucial for supporting targeted genetic improvement of pork quality and developing high-value products.

2. Materials and Methods

2.1. Experimental Design and Sampling

All experimental protocols were approved by the Research Ethics Committee of the Institute of Quality Standard and Monitoring Technology for Agro-products, Guangdong Academy of Agricultural Sciences. Ten DG and ten DLY pigs used in this study were from the Foshan Sanshui District Lejiazhuang Breeding Co., Ltd. (Foshan, Guangdong Province, China). All pigs were raised under the same environment, receiving the same commercial feed. The nutrient level of the basal diet is presented in Table S1. The slaughter weights were 115.3 ± 16.4 kg for DG pigs (287–306 days of age) and 106.7 ± 15.1 kg for DLY pigs (154–168 days of age), respectively. Pigs were fasted for 24 h with free access to water before slaughter. The slaughtering was performed by a licensed commercial slaughtering company (Guohong Meat Processing Co., Ltd., Sanshui District, Foshan, China). Longissimus dorsi muscles between the 11th and 13th ribs of each carcass were collected within 2 h post-slaughter for analysis of meat quality and nutritional composition. Each muscle sample was ground to a fine powder in liquid nitrogen and then stored in an ultra-freezer at –80 °C until metabolomics and lipidomics analysis.

2.2. Meat Quality Analysis

A colorimeter (CR-400, Konica Minolta, Osaka, Japan) was used to measure the color of the meat (lightness L*, redness a*, and yellowness b*). Three random positions on the cross-section of each sample were measured, and the average values were calculated. Marbling scores were measured following the China agricultural industry standard (NY/T 821-2019) [28] after a 24 h storage period at 4 °C. Each sample was scored independently by three trained assessors, and the final score was calculated as the average of the three assessments. Water content was measured by weighing approximately 10 g of sample, transferring it to a pre-weighed weighing bottle, drying in a 103 °C oven for 3 h, cooling to room temperature in a desiccator, and reweighing. A tenderness meter (C-LM3B, Northeast Agricultural University, Harbin, China) was used to measure shear force. Before measurement, samples were heated to a core temperature of 70 °C in an 80 °C water bath, cooled to 20 °C, and then cored parallel to the fiber orientation (1 cm in diameter). The tenderness meter was used to shear the muscle in a longitudinal direction perpendicular to the muscle fibers to obtain the shear force value.

2.3. Nutritional Composition Analysis

Protein content was determined according to GB 5009.5-2016 [29]. Briefly, 20 mL of sulfuric acid, 0.4 g of copper sulfate, and 6 g of potassium sulfate were mixed with 2 g of the homogeneous sample. The mixture was digested at 420 °C for 1 h, cooled, diluted with 50 mL of water, and then analyzed by an automatic Kjeldahl nitrogen analyzer (8400, Foss, Hilleroed, Denmark). Amino acid composition was analyzed following GB 5009.124-2016 [30]. 0.5 g of homogeneous sample was mixed with 15 mL of 6 mol/L hydrochloric acid solution. The mixture was hydrolyzed at 110 °C for 22 h, filtered, diluted, evaporated under vacuum, re-dissolved in 2 mL of sodium citrate buffer (pH 2.2), and prepared for detection. Sixteen amino acids were quantified using a fully automated amino acid analyzer (S433D, Sykam GmbH, Munich, Germany). Fat content was determined following GB 5009.6-2016 [31]. Briefly, 5 g of homogeneous sample was weighed, wrapped in filter paper, and extracted with anhydrous ether using a Soxhlet extractor in a water bath for 6 h. This extract was dried in a 100 °C oven for 1 h and weighed to calculate the fat content based on weight change. Fatty acid composition was analyzed according to GB 5009.168-2016 [32]. 5 g of homogeneous sample was hydrolyzed with a mixture containing 2 mL of triundecanoin, 100 mg of pyrogallol, 2 mL of 95% ethanol, 4 mL of water, and 10 mL of hydrochloric acid at 80 °C for 40 min. This hydrolysate was then mixed with 10 mL of 95% ethanol, extracted with 50 mL of diethyl ether/petroleum ether solution (1:1, v/v), and concentrated to dryness. The extract was mixed with 8 mL of 2% sodium hydroxide–methanol solution and 7 mL of boron trifluoride–methanol solution, refluxed at 80 °C for 2 min, and extracted with heptane to prepare the test solution. The test solution was analyzed by gas chromatography (GC-2010 Plus, Shimadzu Corp., Kyoto, Japan).

2.4. Untargeted Metabolomics Analysis

100 mg of homogeneous sample was extracted with 1 mL of methanol/water solution (80:20, v/v) containing 1% formic acid. The extract was vortexed for 1 min and ultrasonicated for 10 min, followed by incubation at −80 °C for 90 min and centrifugation at 11,000 rpm for 15 min at 4 °C. After being collected and filtered via a 0.22 μm membrane, the supernatant was moved to vials for LC-MS analysis. An ultra-high performance liquid chromatography system (U3000, Thermo Fisher Scientific, San Jose, CA, USA) equipped with a BEH amide column (2.1 × 100 mm, 1.7 μm, Waters) was used for chromatographic separation. The mobile phases were acetonitrile/water (95:5, v/v; A) and acetonitrile/water (50:50, v/v; B), both containing 10 mM ammonium formate. Gradient elution was carried out as follows: 0–15 min, 1–99% B; 15–17 min, 99% B; 17–20 min, 1% B. The following chromatographic conditions were applied: injection volume, 2 μL; flow rate, 0.3 mL/min; column temperature, 40 °C. Mass spectrometric analysis was performed on a high-resolution mass spectrometer (Q-Exactive, Thermo Fisher Scientific, SanJose, CA, USA). Data were collected in positive electrospray ionization mode and analyzed using Compound Discoverer 3.3 (Thermo Fisher, CA, USA) with reference to publicly available databases [33].

2.5. Targeted Lipidomics Analysis

40 mg of homogeneous sample was extracted with 2 mL of methyl tert-butyl ether/methanol solution (75:25, v/v) containing internal standard lipids. After vortexing for 15 min, the extract was supplemented with 200 μL of ultrapure water, vortexed for 1 min, and subjected to centrifugation at 12,000 rpm for 10 min at 4 °C. The supernatant was dried under vacuum, dissolved in 200 μL of acetonitrile/isopropanol (50:50, v/v), vortexed for 3 min, and centrifuged again under the same conditions (3 min). Finally, it was filtered via a 0.22 μm membrane and moved to vials for LC–MS analysis. Lipidomics analysis was performed using an ExionLC ultra-high performance liquid chromatography system coupled with a mass spectrometer (QTRAP 6500+, SCIEX, Framingham, MA, USA). An Accucore C30 column (2.1 × 100 mm, 2.6 μm, 150 Å, Thermo Fisher Scientific) was used for lipid separation. The mobile phases were acetonitrile/water (60:40, v/v, A) and acetonitrile/isopropanol (10:90, v/v, B), both with 10 mM ammonium formate and 0.1% formic acid. A multi-step linear gradient was applied: 0–2 min, 20–30% B; 2–4 min, 30–60% B; 4–9 min, 60–85% B; 9–14 min, 85–90% B; 14–15.5 min, 90–95% B; 15.5–17.3 min, 95% B; 17.3–17.5 min, 95–20% B; 17.5–20 min, 20% B. The following chromatographic conditions were applied: injection volume, 2 μL; flow rate, 0.35 mL/min; column temperature, 45 °C. Data were obtained in both positive and negative electrospray ionization modes. Lipids were identified with a custom database (Metware Database) and quantified relatively based on multiple reaction monitoring mode combined with internal standards.

2.6. Data Analysis

Data were presented in the form of means ± standard deviations. Significant differences between groups were assessed using the Mann–Whitney U test at p < 0.05 with SPSS software (version 22.0, SPSS Inc., Chicago, IL, USA). Principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA) were performed using SIMCA software (version 14.1, Umetrics, Umea, Sweden). PCA was applied to visualize the multidimensional data characteristics of metabolomics and lipidomics. OPLS-DA was used to enhance group separation, and the importance of each variable was estimated using variable importance in projection (VIP) scores. Pearson correlation analysis and hierarchical clustering analysis (HCA) were carried out in the R environment (version 4.2.0, R Development Core Team, Vienna, Austria). Pearson correlation analysis was employed to examine relationships among meat quality, metabolites, and lipids. HCA was used to evaluate variations in differential metabolites and lipids among different groups. Annotation of metabolites and lipids was performed using the KEGG compound database, followed by mapping to the KEGG pathway database (http://www.kegg.jp/kegg/compound/, accessed on 15 November 2025).

3. Results

3.1. Meat Quality Traits and Nutritional Composition Analysis

The comparative analysis of meat quality and nutritional composition is presented in Table 1. DG pigs exhibited significantly lower a* and b* values, marbling scores, and shear force compared to DLY pigs. However, no significant differences were observed in L* value or water content between DG and DLY pigs. Nutritional composition analysis indicated that the two breeds had largely similar nutrient profiles. Although the mean values of protein and amino acids were slightly higher in DG pigs, the differences were not statistically significant (p > 0.05). Eight fatty acids were quantified, consisting of three monounsaturated fatty acids (MUFAs), three saturated fatty acids (SFAs), and two polyunsaturated fatty acids (PUFAs). Their abundances followed the order: MUFAs > PUFAs > SFAs. Compared with DLY pigs, the mean values of SFAs, MUFAs, and PUFAs were higher in DG pigs; however, these differences did not reach statistical significance (p > 0.05).

3.2. Comparison of Metabolite Composition of Pork from Different Breeds

A total of 834 metabolites were detected in the longissimus dorsi muscles of DG and DLY pigs using untargeted metabolomics analysis (Table S2). Subsequently, 79 metabolites were identified by matching mass spectra against the mzCloud and mzVault databases in Compound Discoverer 3.3 (Table S3). These 79 identified metabolites spanned multiple classes, mainly including organic acids and derivatives (31.65%), organic heterocyclic compounds (18.99%), and organic nitrogen compounds (18.99%) (Figure 1). The PCA score plot showed that DG pigs and DLY pigs could be clearly distinguished based on metabolites, indicating distinct metabolite compositions between the two breeds (Figure 2A). A clear separation was achieved using OPLS-DA, with distinct group clustering observed in the score plot (Figure 2B). The OPLS-DA model evaluation parameters were R^2^X = 0.612, R^2^Y = 0.967, and Q^2^ = 0.679, indicating good explanatory power and predictive ability. Using the criteria of VIP > 1, p < 0.05, and fold change (FC) > 1.5 or FC < 0.67, 13 significantly differential metabolites were identified between the two breeds (Table S4). These metabolites included amino acids, peptides, and analogues; pyridinecarboxylic acids and derivatives; quaternary ammonium salts; alkaloids and derivatives; N-acylpyrrolidines, nitrobenzenes; and others. Their distribution is shown in the volcano plot (Figure 3A), where red and blue indicate metabolites that were significantly upregulated and downregulated, respectively, in DG pigs relative to DLY pigs. Specifically, DG pigs exhibited higher concentrations of 3 metabolites and lower concentrations of 10 metabolites. HCA of the 13 differential metabolites further confirmed a clear separation between the two breeds. The corresponding heatmap (Figure 3B) visualizes the relative abundance of each metabolite across samples, with a blue-to-red color gradient representing low to high abundance. Furthermore, these 13 differential metabolites were used to discriminate between the two breeds using an OPLS-DA model. The model evaluation parameters were R^2^X = 0.768, R^2^Y = 0.87, and Q^2^ = 0.824. The discrimination accuracy reached 98%, and permutation testing confirmed no overfitting of the model. Based on VIP values, the seven most important metabolites were, in descending order: L-norleucine, L-phenylalanine, N6,N6,N6-trimethyl-L-lysine, nicotinamide, choline, betaine, and guanine.

The differential metabolites were mainly enriched in several KEGG pathways, including phenylalanine, tyrosine and tryptophan biosynthesis; phenylalanine metabolism; lysine degradation; and glycine, serine and threonine metabolism (Figure 3C). The majority of these enriched pathways are linked to amino acid biosynthesis and metabolism. Specifically, in the phenylalanine, tyrosine and tryptophan biosynthesis pathway, L-phenylalanine exhibited a significantly lower relative abundance in DG pigs compared to DLY pigs. In the glycine, serine and threonine metabolism pathway, betaine was upregulated, whereas choline was downregulated in DG pigs relative to DLY pigs.

3.3. Comparison of Lipid Composition of Pork from Different Breeds

A total of 1024 lipid molecules were identified using lipidomics analysis and classified into 27 subclasses (Table S5). The composition of the major lipid subclasses is illustrated in Figure 4. The abundances of these subclasses followed the order: phosphatidylethanolamine (PE, 21.19%), triglycerides (TG, 19.73%), ceramide (CER, 9.77%), phosphatidylcholine (PC, 9.57%), phosphatidylglyceride (PG, 5.66%), and others. Comparative analysis revealed a generally higher overall lipid content in DG pigs compared to DLY pigs. Statistical analysis showed that several lipid subclasses were significantly more abundant in DG pigs, including cholesterol, eicosanoids, N-acyl-1,2-diacylglycerol-3-ethanolamine phosphate (LNAPE), phosphatidic acid (PA), PE, sphingomyelin (SM), and TG.

Multivariate statistical methods were applied to analyze the lipidomics data. PCA indicated a clear separation between DG and DLY pigs in the score plot (Figure 5A). To enhance group separation, an OPLS-DA model was constructed, which also showed distinct clustering of the two breeds (Figure 5B). The model demonstrated strong explanatory and predictive power, with R^2^X = 0.543, R^2^Y = 0.993, and Q^2^ = 0.831.

Using the criteria of VIP > 1, p < 0.05, and FC > 2 or FC < 0.5, 77 lipids were identified as differentially abundant between the two breeds (Table S6). Comparative analysis revealed that 76 lipids were significantly upregulated in DG pigs, while only one was downregulated, as shown in the volcano plot (Figure 6A). Subclass analysis showed that among the upregulated lipids, 63 were TG, 11 were CER, one was an eicosanoid, and one was free fatty acid (FFA). The only downregulated lipid belonged to the HexCer subclass. The HCA heatmap illustrates the relative abundance of these lipids across all samples, with a blue-to-red color gradient indicating increasing abundance (Figure 7). Moreover, an OPLS-DA model based on the 77 differential lipids successfully distinguished the two breeds, with R^2^X = 0.85, R^2^Y = 0.768, and Q^2^ = 0.572. The model achieved 100% discrimination accuracy with no overfitting, as confirmed by permutation testing. Among the lipids with the strongest discriminatory influence (VIP > 1), all belonged to the TG subclass.

An analysis of pathway enrichment was carried out on the differential lipids with the KEGG database, focusing on those mapped to known metabolic pathways. The analysis revealed eight significantly enriched pathways (p < 0.05), including lipid and atherosclerosis, cholesterol metabolism, fat digestion and absorption, regulation of lipolysis in adipocytes, vitamin digestion and absorption, insulin resistance, glycerolipid metabolism, and thermogenesis (Figure 6B). Notably, lipids belonging to the TG and FFA subclasses were involved in several of these pathways. TG and FFA lipids were significantly more abundant in DG pigs than in DLY pigs.

The relationships between meat quality traits and the abundance of significantly differential metabolites and lipid subclasses were assessed by Pearson correlation analysis (Figure 8). In DG pigs, lightness showed positive correlations with CER, eicosanoid, and TG, but was not significantly correlated with any differential metabolites. Similarly, yellowness was positively correlated only with HexCer; marbling score was positively correlated with L-norleucine; and water content was positively correlated with choline. All of these correlation coefficients exceeded 0.9, indicating strong positive relationships. No significant correlations were observed between redness or shear force and any differential metabolite or lipid in this breed. In DLY pigs, a different correlation pattern was observed. A significant negative correlation was found between water content and both FFA and TG. No other significant correlations were found between meat quality traits and differential metabolites or lipids in this breed. These results suggest breed-specific relationships between lipids, metabolites, and meat quality traits.

4. Discussion

Meat quality traits mainly include meat color, marbling, tenderness, and intramuscular fat content [5]. Meat color is a key sensory attribute that directly affects the perception of freshness and influences consumer purchase decisions. In this study, DG pigs exhibited significantly lower a* and b* values compared to DLY pigs. Many studies have reported that Chinese indigenous pig breeds, such as Jianhe White Xiang, Laiwu, and Ningxiang, produce pork with higher a* values than Western commercial breeds [21,23,24]. However, others have observed lower a* values in some indigenous breeds [13,22]. These differences are likely attributable to breed-specific differences in fat deposition and variation in slaughter age. The eating quality of pork—particularly its juiciness, tenderness, and flavor—is closely linked to intramuscular fat content. An intramuscular fat content exceeding 2.5% is generally regarded as beneficial for enhancing the sensory characteristics of cooked pork [34]. In this study, both DG and DLY pigs had intramuscular fat contents above 3%, with no statistically significant difference between them. Although DG pigs received a lower marbling score, they exhibited greater tenderness than DLY pigs. This result is consistent with previous findings that pork from some indigenous pig breeds tends to be more tender [13], suggesting that factors beyond IMF content, such as muscle fiber characteristics [35], may also contribute to tenderness.

Amino acids also play a crucial role in determining pork quality, serving not only as the building blocks of proteins but also as key precursors of flavor compounds. Different amino acids in meat contribute to distinct taste experiences, such as sourness (e.g., phenylalanine, tyrosine, alanine) and umami (e.g., aspartate and glutamate) [36]. The levels of both flavor-related and essential amino acids were similar between DG and DLY pigs in our analysis. This indicates similar nutritional value and flavor potential between the two breeds.

Fatty acids are another essential class of flavor precursors, as their oxidation leads to the release of volatile compounds [4]. The overall fatty acid profiles of the two breeds were largely similar. Palmitic acid (C16:0), stearic acid (C18:0), oleic acid (C18:1n9c), and linoleic acid (C18:2n6c) were the predominant fatty acids detected in this study. Consistent with previous studies [12,37], these saturated fatty acids (SFAs, C16:0 and C18:0) and unsaturated fatty acids (C18:1n9c and C18:2n6c) together accounted for over 90% of the total fatty acids. Previous studies have shown that the dietary intake of oleic acid is associated with health benefits, such as reducing the risk of obesity and metabolic syndrome as well as decreasing inflammation [38]. Its content in meat is also considered a potential modulator of pork flavor [39,40]. Linoleic acid is rich in omega-6 fatty acid and plays an important role in immune regulation and inflammation [41]. Although DG pigs exhibited higher mean values of SFAs, PUFAs, and MUFAs, these differences were not statistically significant. MUFA content is known to be a key factor influencing pork flavor, with higher levels associated with improved meat quality.

Among the 79 metabolites identified in this study, organic acids and derivatives accounted for 31.65%. This class mainly included amino acids, peptides, and analogues, which are essential in regulating meat quality and flavor [42]. The 13 differential metabolites between the two breeds spanned multiple classes, including amino acids, peptides, and analogues; pyridinecarboxylic acids and derivatives; quaternary ammonium salts; and alkaloids and derivatives. Three metabolites were significantly upregulated in DG pigs compared with DLY pigs: betaine, 1-hexadecanoylpyrrolidine, and 5-(hydroxyethyl)-4-methoxy-2,5-dihydrofuran-2-one. Betaine is a secondary metabolite derived from the oxidation of choline. It functions as an important methyl donor in the methionine–homocysteine cycle and serves as an osmolyte to maintain osmotic balance of the body. In pigs, betaine supplementation has been shown to reduce carcass fat, increase muscle content, and improve tenderness [43]. Similarly, in a study of Shaziling pigs, serum betaine level was negatively correlated with body weight, carcass length, backfat thickness, and fat percentage [13]. Another upregulated metabolite, 1-hexadecanoylpyrrolidine, belongs to the N-acylpyrrolidine class and may act as a flavor compound derived from lipid oxidation and the Maillard reaction, potentially contributing to meat aroma [44]. 5-(Hydroxyethyl)-4-methoxy-2,5-dihydrofuran-2-one is a furanone derivative; however, research on its role in meat remains limited. Ten metabolites were downregulated, including three amino acids or derivatives: L-norleucine, L-phenylalanine, and N6,N6,N6-trimethyl-L-lysine. L-norleucine is an isomer of leucine. While leucine specifically affects skeletal muscle protein synthesis, L-norleucine does not share this effect [45]. Correlation analysis revealed a strong positive correlation between L-norleucine and marbling score in DG pigs, suggesting a potential link to fat deposition. This finding is consistent with a study in chickens, in which L-norleucine was identified as a differential metabolite potentially associated with fat deposition [46]; however, its functional mechanism remains unclear. L-phenylalanine is a well-known flavor precursor that can be converted into volatile compounds such as phenylethanal via the Maillard reaction and Strecker degradation [44]. It has also been used as a characteristic metabolite to distinguish beef from pork [47]. N6,N6,N6-trimethyl-L-lysine is a key precursor in the biosynthesis of L-carnitine, which is involved in fatty acid metabolism and linked to fat deposition and meat quality regulation in pigs [22]. Other downregulated metabolites included organonitrogen compounds, pyridines and derivatives. Specifically, the organonitrogen compound nicotinamide has been associated with pork redness [48] and is essential for synthesizing coenzymes NAD+ and NADP+. These coenzymes are crucial for energy metabolism, DNA repair, and cell signaling [49]. Guanine is another downregulated organonitrogen compound. Lower purine content in pork has been shown to correlate significantly with higher ultimate pH, better meat color, and more marbling patterns [50]. For pyridines and derivatives, choline supplementation has been reported to promote intramuscular fat deposition in pigs [51]. The positive correlation between choline and water content in DG pigs may be attributed to choline’s role as a component of phosphatidylcholine, a major structural lipid in cell membranes [52]. Adequate choline availability supports membrane integrity, which may enhance water-holding capacity in muscle. These differential metabolites were enriched in several key metabolic pathways, mostly for amino acid synthesis and metabolism. L-phenylalanine was associated with phenylalanine metabolism and phenylalanine, tyrosine and tryptophan biosynthesis. In these pathways, it serves as a precursor for tyrosine in the former and an end-product in the latter. These pathway-related amino acids are closely tied to pork flavor formation [53]. N6,N6,N6-trimethyl-L-lysine participated in lysine degradation, leading to L-carnitine production [54]. Betaine and choline participated in glycine, serine and threonine metabolism, primarily functioning as methyl donors [55]. In this study, seven metabolites with VIP > 1—L-norleucine, L-phenylalanine, N6,N6,N6-trimethyl-L-lysine, nicotinamide, choline, betaine, and guanine were identified as potential biomarkers to distinguish DG from DLY pigs.

Lipid metabolism and deposition are key processes influencing pork flavor [56]. Lipidomics revealed that the major lipid classes in pork were glycerophospholipids (GP) and glycerolipids (GL), followed by sphingolipids (SP) and fatty acids (FA). GP, which is rich in unsaturated fatty acids, serves as important precursors for pork flavor [57]. GL mainly consists of TG and diacylglycerol (DG), with TG being particularly abundant. Seventy-seven lipids differed significantly between DG and DLY pigs. TG constituted the largest proportion (81.81%), followed by CER (14.29%), with the remainder comprising thromboxane B3 (TxB3) and a free fatty acid (FFA (35:0)). Although intramuscular fat content did not differ between the two breeds, most differential lipids were significantly more abundant in DG pigs. These upregulated lipids mainly included TG, most CER, TxB3, and FFA (35:0). TG, as the main lipid component in pork, is the primary form of energy storage [58] and an important factor influencing intramuscular fat [59]. Previous studies on Chinese indigenous pig breeds have reported upregulated TG expression. For example, in a comparison of Luchuan and Duroc pigs, 61 TG molecules were significantly upregulated in Luchuan pigs [19]. Similarly, in a comparative study of Laiwu pigs and Yorkshire pigs, the contents of TGs were mostly higher in Laiwu pigs [23]. These results align with the upregulation of TG observed in DG pigs in this study. Notably, the Guangdong small-ear spotted pig—the maternal line of DG—along with Luchuan and Laiwu pigs, are all indigenous obese pig breeds in China. CER is a class of sphingolipids formed by linking a sphingosine backbone to a fatty acid side chain via an amide bond. It functions as a key signaling molecule that participates in regulating various cellular processes (such as growth, aging, and death) [60]. In contrast, studies on TxB3 and FFA (35:0) in meat remain limited, and their effects on the characteristics or flavor of meat are still unclear. The sole significantly downregulated lipid was the glycosphingolipid HexCer (t16:0/22:2(2OH)), which has been linked to fat deposition in chickens [46] and may serve as a characteristic molecule for distinguishing meat quality between the two breeds. Correlation analysis revealed that lipid subclasses, including TG, CER, and eicosanoids, showed positive correlations with lightness in DG pigs. Although lipid composition itself may not directly affect meat color, lipid oxidation products may deteriorate color [61,62]. Yellowness was positively correlated with HexCer in DG pigs. While no direct reports on this relationship exist, studies suggest that sphingolipid metabolism may be associated with beef color development [63]. The differential lipids were primarily annotated to pathways within organismal systems (e.g., cholesterol metabolism; fat digestion and absorption; vitamin digestion and absorption; regulation of lipolysis in adipocytes; thermogenesis), followed by metabolism (e.g., glycerolipid metabolism) and human diseases (e.g., lipid and atherosclerosis; insulin resistance). TG plays a key role in these pathways. Similar findings have been observed in lamb research regarding the function of TG in fat digestion, absorption, and cholesterol metabolism [64]. TG is synthesized via glycerolipid metabolism. This metabolic pathway works in concert with fat digestion and absorption to jointly maintain lipid homeostasis and energy supply. Additionally, TG serves as a significant indicator closely linked to cholesterol metabolism. This metabolic pathway is associated with atherosclerosis [65]. The role of the thermogenesis pathway in pork quality remains unclear. In this study, lipidomics combined with multivariate statistical analysis effectively distinguished DG from DLY pigs, with TG as the key characteristic lipid class.

5. Conclusions

In this study, metabolomics and lipidomics were employed to analyze meat quality traits, small-molecule metabolites, and lipids in the longissimus dorsi muscle of DG and DLY pigs. Results demonstrated that tenderness was significantly higher in DG pigs compared to DLY pigs. However, lower a*, b*, and marbling scores were observed in DG pigs. Although DG pigs showed higher mean values of protein, amino acids, and fatty acids, these differences were not statistically significant. Metabolomic analysis identified 13 differential metabolites, including amino acids and their derivatives, pyridine derivatives, and quaternary ammonium compounds. These metabolites effectively discriminated between the two breeds with 98% accuracy. L-norleucine, L-phenylalanine, N6,N6,N6-trimethyl-L-lysine, nicotinamide, choline, betaine, and guanine were identified as potential biomarkers. Lipidomic analysis revealed a largely similar lipid profile between the two breeds; however, 77 lipids showed differential abundance. The majority of these, particularly TG and CER, were present at higher levels in DG pigs. Lipidomics combined with multivariate statistical analysis enabled complete discrimination between the two breeds (100% accuracy), and TG was identified as a key biomarker. Correlation analysis suggested breed-specific relationships between lipids, metabolites, and meat quality traits. Only in DG pigs was the marbling score significantly positively correlated with L-norleucine, and meat color traits were significantly positively correlated with several lipid subclasses, such as TG, CER, eicosanoids and HexCer. In conclusion, this study provides a deeper understanding of the meat quality characteristics in indigenous crossbred pigs and identifies potential biomarkers. This study could offer a scientific reference for targeted breeding strategies aimed at pork quality improvement.