Abstract This study employed UPLC-MS/MS and GC-IMS techniques to compare and analyze the lipid metabolites and volatile flavor compounds in raw abdominal muscle (CK), sour video abdominal muscle (SV), steamed abdominal muscle (ST), and oven-cooked abdominal muscle (OC). A total of 42 subclasses and 1230 lipids were identified. Among these, lysophosphatidylethanolamine (LPE) 18:2/0:0, lysophosphatidylcholine (LPC) 18:2/0:0, and triacylglycerol (TG) 16:0_18:1_18:1 enhanced the aroma retention of steamed abdominal muscle, whereas phosphatidylcholine (PC) 16:0_18:2 and phosphatidylethanolamine (PE) P-18:18:18:2 influenced the aroma retention of roasted abdominal muscle. Additionally, 250 differentially abundant metabolites were identified as potential markers for differentiating various cooking methods. Seven compounds were recognized as potential indicators for distinguishing cooking methods: propanal-D, n-pentanal-M, n-pentanal D, butanal-D, 3-methylbutanal, 1-hexanal-M, and 1-hexanal D. Correlation analysis results indicated a significant positive correlation between aldehydes and phospholipid molecules, including PC, PE, LPC, and LPE. Keywords: Abdominal muscle, Cooking methods, UPLC-MS/MS, GC-IMS Graphical abstract [37]Unlabelled Image [38]Open in a new tab Highlights * • A total of 42 subclasses and 1230 lipids were identified from abdominal muscle. * • A total of 49 volatile flavor compounds were identified. * • Cooking exerted a significant impact on both lipid composition and flavor of pork. * • A significant positive correlation between aldehydes and phospholipid molecules. * • LPC, LPE, PC and PE were critical in flavor development in pork dishes. 1. Introduction Meat and its products are popular among consumers because of their unique flavor and significant nutritional benefits. The taste of meat strongly affects acceptance among buyers ([39]Han, Zhang, Fauconnier, & Mi, 2020). Uncooked meat primarily contributes a flavor reminiscent of blood and metal ([40]Fu, Cao, Yang, & Li, 2022). Approximately 90 % of the aroma compounds found in cooked meat originate from the breakdown of lipids, whereas the other 10 % are produced through various other chemical reactions ([41]Amjad et al., 2022). Lipids are vital for the development of flavor in meat products, and this process is influenced by numerous factors, such as the type of meat, genetic background, feeding practices, and methods of food processing. A variety of cooking methods exist for abdominal muscle, a widely favored type of meat, which can influence its nutritional value by causing the loss of lipid components while enhancing flavor and color ([42]Ángel-Rendón et al., 2020). The cooking process enhances the creation of new characteristic flavors. Compared with other cooking methods, baking is characterized by longer durations and higher temperatures, leading to increased oxidation and volatile compounds. Consequently, the Maillard reaction and lipid oxidation contribute to flavor enhancement. Vacuum steaming, in particular, minimizes the production of harmful byproducts, thereby promoting a healthier cooking outcome ([43]Jiang et al., 2022). Lipids, which constitute 38 % of the dry weight of meat, serve as the primary substrates for flavor production in roasted meat products ([44]Liu et al., 2020). Lipids have the capacity to generate a wide variety of aromatic compounds via autooxidation or thermal oxidation processes. The generation of these aromatic compounds during thermal treatment is intricately linked to the particular lipid types and fatty acids present (Li [45]Zhou, Zhao, Bindler, & Marchioni, 2014). [46]Liu et al. (2024) reported that phosphatidylcholine (PC), phosphatidylserine (PS) and phosphatidylethanolamine (PE) play crucial roles in the production of key aroma substances in roasted pork. [47]Liu et al. (2021) reported that PC and PE may play major roles in the formation of aromatic compounds. Lipid molecules, such as LPE (22:6), LPE (20:4), LPE (20:5), LPE (18:3), LPE (16:1), and LPE (18:2), are likely key lipids responsible for generating distinctive flavor compounds during the hot processing of crayfish ([48]Zhou et al., 2023). Unsaturated fatty acids (UFAs) are recognized as the primary flavor components in meat (Longzhu [49]Zhou et al., 2024). In recent years, advancements in metabolomics technology have led to high reliability and accuracy in measuring volatile flavor compounds and lipid components. Gas chromatography-ion mobility spectrometry (GC-IMS) is a technique that offers high sensitivity, resolution, and rapid analysis, making it advantageous for various applications. This approach is especially useful for detecting functional groups, including amino, thiol, aldehyde, ketone, and ether groups, present in food flavorings ([50]Wang, Chen, & Sun, 2020). Currently, some scholars have investigated the effects of various cooking methods on the flavor components of beef ([51]Watanabe et al., 2015), mutton ([52]Roldán, Ruiz, Del Pulgar, Pérez-Palacios, & Antequera, 2015), and fish ([53]Nieva-Echevarría, Manzanos, Goicoechea, & Guillén, 2017). However, there is a scarcity of research reports focusing on pork, and a comprehensive comparison of the influence of common Chinese cooking methods on pork remains lacking. Lipidomics, which uses mass spectrometry, has become a crucial method for studying the lipid makeup of food products ([54]Song et al., 2022). This approach offers detailed insights into the complete lipid profiles of samples without the need for prior screening. Currently, lipidomics is extensively applied to detect lipid components and potential biomarkers in both raw and processed meats ([55]Li, Al-Dalali, Wang, Xu, & Zhou, 2022). Research has indicated that 97 distinct lipids exist among donkey, beef, and lamb, with 3 phospholipids proposed as potential markers for differentiating between these meats and 13 characteristic VOCs exhibiting negative correlations with 21 phospholipid markers ([56]Man et al., 2023). [57]Jia, Shi, and Shi (2021) employed fragmentation mechanisms along with UHPLC-Q-Orbitrap MS/MS to investigate the transformations of lipids and their molecular mechanisms, revealing alterations in the lipid profile and quality of Tannan sheep after refrigeration periods of 12 and 24 days. To date, few studies have combined lipidomics with volatile flavor compounds to examine differences in the cooking of abdominal muscle. The Nanyang black pig, a highly esteemed breed from China, is celebrated by consumers for its flavorful meat, high fatty acid content, and diverse nutrient profile. Unsaturated fatty acids are prevalent in black pork and are prone to oxidation and decomposition, resulting in the formation of aldehydes, ketones, alcohols, and other small volatile flavor compounds during processing. Consequently, this study aims to employ UPLC-MS/MS and GC-IMS techniques to (i) delineate the lipidomic and volatile profiles of raw abdominal muscle, sous vide abdominal muscle, steamed abdominal muscle, and oven-cooked abdominal muscle; (ii) identify key lipids and volatile compounds responsible for these differences; and (iii) explore the relationships between significant lipid molecules and volatile compounds. The findings of this study will enhance our understanding of lipid profile alterations during the processing of abdominal muscle products and serve as a reference for advancing related production technologies. 2. Materials and methods 2.1. Chemicals and reagent HPLC-grade acetonitrile (ACN), methanol (MeOH), isopropanol (IPA), dichloromethane (CH2Cl2), and tert-butyl methyl ether (MTBE) were procured from Merck (Darmstadt, Germany). HPLC-grade formic acid (FA) and ammonium formate (AmFA) were obtained from Sigma-Aldrich (St. Louis, MO, USA). Ultrapure water was obtained via a Milli-Q system (Millipore, Billerica, MA). Lipid standards were acquired from either Sigma-Aldrich or Avanti Polar Lipids (Alabaster, AL). 2.2. Cooking processing of abdominal muscle The abdominal muscle of black pigs (n = 3, 100–110 kg live weight, about 6 months old) was sourced from a farm in Nanzhao County, Nanyang City, Henan Province, China. The fresh abdominal muscle (1 kg) was refrigerated and transported to the laboratory within one hour of the pigs were slaughtered and dissected. Divide one kilogram of abdominal muscle into four equal portions, each weighing 250 g. Prior to cooking, the skin of the abdominal muscle was removed, and each sample was cut to dimensions of 0.5 × 1 × 2 cm. Four cooking methods were employed in this study: raw abdominal muscle, steam cooking, sous vide, and oven cooking. These widely used techniques were selected because of their significant variations in terms of food heating mechanisms. 2.2.1. Raw abdominal muscle (CK) Unprocessed raw abdominal muscle. 2.2.2. Sous vide (SV) The samples were pretreated and then vacuum-sealed using a Lavezzini Univac vacuum sealer (Lavezzini Univac, Fiorenzuola d'Arda, PC, Italy) in OPA/PP 15/65 bags (Orved, Musile di Piave, Italy). These sealed samples were subsequently boiled at 75 °C in a water bath for 45 min. 2.2.3. Steaming (ST) (100 % steam) The samples were subjected to steaming at an atmospheric pressure of 100 °C for 180 min in a Combo-Steam SL oven (V-Zug, Zurich, Switzerland). 2.2.4. Oven cooking (OC) An electric oven (T7-L3840, Guangdong Midea Kitchen Appliance Manufacturing Co., Ltd.) was preheated to 180 °C. After 15 min, the samples were removed and allowed to cool. During the baking process, the samples were turned every 5 min. Each of the four groups of experimental samples was repeated five times in for a total of 20 samples. Once cooked, the samples were placed at room temperature, minced using a meat grinder, vacuum-packed, and stored at −80 °C until further analysis. 2.3. Lipid oxidation indices 2.3.1. Determination of the peroxide value (POV) The POVs of the abdominal muscle samples were determined via the titration method according to GB/T5009.227–2016. 2.3.2. Determination of thiobarbituric acid reactive substances (TBARS) Based on the research method reported by ([58]Domínguez, Gómez, Fonseca and Lorenzo, 2014a), the TBARS of various samples were determined. Six grams of chopped and mixed abdominal muscle were combined with 20 mL of a 10 % trichloroacetic acid solution containing 0.1 % ethylene diamine tetraacetic acid (EDTA) and filtered through Whatman filter paper. Subsequently, 5 mL of the filtrate was combined with 5 mL of a 0.02 mol/L TBA solution, swirled for 10 s, and incubated at 90 °C for 40 min. After cooling to room temperature, the absorbance was measured at 532 nm. 2.4. Lipid extraction procedure About 20 mg of sample was allowed to thaw and then transferred into a correctly labeled 2 mL centrifuge tube, and the weights were recorded. To this end, 1 mL of extraction solvent (methyl tert-butyl ether/methanol: 3/1, V/V) was introduced. The resulting mixture was vortexed for 15 min. Following this step, 200 μL of water was added, and the mixture was vortexed for 15 min. Afterward, the mixture was centrifuged at 4 °C (12,000 rpm) for 10 min via a 5424 R centrifuge from Eppendorf GmbH. Subsequently, 200 μL of the supernatant was subsequently transferred to a 1.5 mL centrifuge tube for concentration. Once concentrated, the lipid extract was reconstituted in 200 μL of isopropanol (1:1, v/v), vortexed for 3 min, and then centrifuged again at 12,000 rpm for 3 min. The supernatant was then collected for analysis via LC-MS/MS. To create a quality control sample, twenty microliters of supernatant from each sample were pooled together. 2.5. HPLC conditions The sample extracts were analyzed using an LC-ESI-MS/MS system (UPLC, ExionLC AD, [59]https://sciex.com.cn/; MS, QTRAP® 6500+ System, [60]https://sciex.com/). The column temperature was 45 °C. Flow rate: 0.35 mL/min. Mobile phase A was composed of acetonitrile (Merck, Darmstadt, Germany) and water at a ratio of 60:40, whereas mobile phase B was composed of acetonitrile/isopropanol (Merck, Darmstadt, Germany) (95/5). The program for the chromatographic gradient elution was designed as follows: at 0 min: A/B (80:20, V/V), at 2.0 min: 70:30 (V/V), at 4 min: 40:60 (V/V), at 9 min: 15:85 (V/V), at 14 min: 10:90 (V/V), at 15.5 min: 5:95 (V/V), at 17.3 min: 5:95 (V/V), then returning to 80:20 (V/V) at 17.3 min, and finally at 20 min: 80:20 (V/V). 2.6. ESI-MS/MS conditions The experimental parameters for ESI-MS/MS were established with ionization voltages of (+) 5.5 kV and (−) 4.5 kV. The temperatures for the ion source (TEM), as well as the settings for nebulizer gas (GC1), auxiliary gas (GC2), and curtain gas (CUR), were controlled at 500 °C, 45 psi, 55 psi, and 35 psi, respectively. Instrument tuning and mass calibration were performed with 10 and 100 μmol/L polypropylene glycol solutions in QQQ and LIT modes, respectively. 2.7. Volatile compound analysis Volatile compound assessment was performed utilizing a GC-IMS device (FlavourSpec®, G.A.S., Dortmund, Germany). A 2 g sample of abdominal muscle was introduced into a 20 mL headspace vial and incubated at 60 °C for 15 min. Next, 500 μL of the sample was injected, with the injector temperature set to 85 °C and the incubation speed adjusted to 500 rpm. The MXT5 (15 m × 0.53 mm ID) capillary column was used, and the temperature of the column was held constant at 60 °C throughout the entire procedure. Nitrogen gas, with a purity of ≥99.999 %, was utilized as the carrier gas. The flow rate commenced at 2.0 mL/min for the initial 2 min, subsequently ramped linearly to 10.0 mL/min within an 8-min timeframe, and further increased linearly to 100.0 mL/min over the course of 10 min, which was then maintained at 100.0 mL/min for an additional 10 min. Ionization of analytes occurred within the IMS chamber before being directed into the drift tube (150 mL/min). 2.8. Relative odor activity value (ROAV) The ROAV method was employed to assess how particular aroma components influence the overall flavor profile. The component that had the greatest impact on the flavor of the sample was designated ROAV[x] = 100. The ROAV values for the other components (x) were derived via the formula outlined below: [MATH: ROAVX100×C%X C%stan×TstanTX< /mfrac> :MATH] In the formula, C%X signifies the percentage content relative to the volatile component, whereas TX denotes the sensory threshold associated with it. Additionally, C% reflects the percentage of the component that has the most significant impact on the sample's overall flavor, and T is the sensory threshold of that particular flavor-contributing component. 2.9. Statistical analysis Data on lipids were analyzed via Analyst 1.6.3 software, whereas R software (available at [61]www.r-project.org) was used for multivariate statistical analyses. Flavor data analysis was carried out utilizing the instrumental data processing software VOCal. Analysis of variance (ANOVA) and Duncan's multipole difference test (P < 0.05) were used to evaluate the differences between samples. Both PCA (utilizing UV scaling) and OPLS-DA (employing Par scaling) analyses were executed with Simca 14.1 to differentiate the smell and taste characteristics of abdominal muscle samples. 3. Results and discussion 3.1. Lipid oxidation The effects of various cooking methods on abdominal muscle POV and TBARS are illustrated in [62]Fig. 1A and [63]Fig. 1B, respectively. Compared with those in the control group (CK), the POV and TBARS levels in the SV, ST, and OC groups tended to increase. Specifically, the CK group presented POV and TBARS values of 0.28 meq/kg and 0.35 mg MDA/kg, respectively, whereas the ST group presented values of 1.91 meq/kg and 3.85 mg MDA/kg, indicating increases of approximately 7 and 11 times, respectively, compared with those of the CK group. Although the POV and TBARS values in both the SV and OC groups were lower than those in the ST group were, they remained elevated compared with those in fresh abdominal muscle. This finding is consistent with Broncano et al.([64]Broncano, Petrón, Parra, & Timón, 2009), who demonstrated that cooking significantly increases physicochemical indicators of pork, such as the TBARS value and cholesterol oxide content, thereby reflecting the extent of fat oxidation. These cooking methods can considerably increase the degree of oxidation of pork fat, potentially contributing to the development of flavor compounds. Fig. 1. [65]Fig. 1 [66]Open in a new tab Changes in lipid oxidation in abdominal muscle with different cooking methods: (A) changes in the content of peroxide value (POV); (B) changes in the content of thiobarbituric acid reactants (TBARS). Lipid composition of abdominal muscle treated with different cooking methods: number and percentage of lipid subclasses (C). Changes in lipid content of abdominal muscle under different cooking methods (D). Changes in lipid subclass content in abdominal muscle under different cooking methods (E-H). The asterisk above the bar chart indicates the results of a t-test comparing the two groups of data, with the following criteria for P-value determination: * P < 0.05; ** P < 0.01; and *** P < 0.001. CK: raw abdominal muscle, SV: sous vide abdominal muscle, ST: steamed abdominal muscle, OC: oven-cooked abdominal muscle. 3.2. Lipid changes in abdominal muscle under different cooking methods Abdominal muscle samples treated with four cooking methods were analyzed by UPLC-MS/MS to assess changes in lipid components resulting from these techniques. A mixed solution was utilized as the QC sample. As depicted in Fig. S1, the curves representing the total ion flow for lipid detection showed considerable overlap, with consistent retention times and peak intensities. These findings indicate the robust stability of signals when identical samples are analyzed across various time points, which suggests high reproducibility and dependability of the data. The lipid composition of the abdominal muscle processed by the four cooking methods was shown in [67]Fig. 1C, and a total of 1230 lipids were identified. These included 553 glycerophospholipids (GP), 248 sphingolipids (SP), 345 glycerolipids (GL), 74 fatty acids (FAs), and 10 other lipid compounds. Among these lipid subclasses, GP were the most prevalent, followed by GL and SP. This observation aligns with earlier studies on pork lipids, which also indicated that the GL, GP, and SP subclasses constituted more than 90 % of the total lipid content (H. [68]Liu et al., 2024). These lipid compounds can be further categorized into 42 subclasses, predominantly consisting of triglycerides (TG), phosphatidylethanolamines (PE), phosphatidylcholines (PC), and ceramides (Cer). Triglycerides are the primary form of lipids found in both animals and plants, and their degradation into free fatty acids is crucial for flavor development. This finding aligns with previous research on pork lipids ([69]Wu, He, Yang, & Li, 2024). The alterations in the lipid content of the abdominal muscle following various heat treatments are illustrated in [70]Fig. 1D. GLs were the most abundant lipids both before and after heat treatment, followed by GP, FA, and SP. The contents of various lipid subclasses were further analyzed, among which the lipid contents of DG and TG were the highest in ST, significantly higher than those of the other three samples ([71]Fig. 1E). Under heat treatment conditions, DGs and TGs are susceptible to thermal degradation and hydrolysis, yielding free fatty acids (FFA). These FFAs can subsequently undergo β-homolysis, ketone-enol tautomerization, or isomerization at carbon‑carbon double-bond reaction sites, forming various types of hydroperoxides ([72]Yin, Xu, & Porter, 2011). These compounds are then decomposed into smaller molecular species, including hydrocarbons, ketones, aldehydes, and alcohols ([73]Zhang, Qin, Lin, Shen, & Saleh, 2015). These findings indicate that the FA content in the abdominal muscle decreased following different heat treatments, likely due to the oxidation of FAs into smaller molecular weight compounds ([74]Fig. 1H). Compared to CK, the levels of LPC, LPE, and LPS were significantly elevated in both ST SV and OC, with the highest concentration observed in ST ([75]Fig. 1F). This increase may be attributed to the hydrolysis of PC and PE into LPC and LPE under aerobic heating conditions. Cooking temperature is a crucial factor influencing phospholipase activity, which in turn regulates the conversion of PC and LPC in abdominal muscles ([76]Jia, Li, Wu, Liu, & Shi, 2021). Notably, CK and SV do not contain LPA. The contents of PE, PG, PI, PS, PC, and PA were highest in OC, followed by SV ([77]Fig. 1F). This finding may be associated with the alteration and degradation of lipid molecules during oxidation ([78]Tu et al., 2022). Additionally, the content of SM increased significantly following heat treatment, particularly during baking ([79]Fig. 1G). The contents of TG, LPC, and LPE in the ST group were significantly greater than those in the other treatment groups. The highest levels of LPE, LPC, and TG were observed for LPE (18:2/0:0), LPC (18:2/0:0), and TG (16:0_18:1_18:1), respectively. Moreover, the concentrations of PE and PC in the OC group significantly increased relative to those in the other groups. The lipid molecules found in the highest abundance within PC and PE were identified as PC (16:0_18:2) and PE (P-18:0_18:2), respectively. It has been reported that lipid content has a greater influence on aroma than does lipid type (M. [80]Zhou et al., 2023). Additionally, the findings revealed that the aroma retention of steamed abdominal muscle was influenced by LPE (18:2/0:0), LPC (18:2/0:0), and TG (16:0_18:1_18:1), whereas PC (16:0_18:2) and PE (P-18:0_18:2) could significantly contribute to the aroma retention of roasted abdominal muscle. With respect to double bond equivalents (DBE), triglycerides (TG), phosphatidylethanolamine (PE), and phosphatidylcholine (PC), with DBE values of 2, dominate the abdominal muscle subjected to various cooking methods. The most frequently observed DBE value for free fatty acids (FFA) is 1 ([81]Fig 2ABC). An analysis of lipid compounds with different chain lengths revealed that TG (chain lengths: 46, 48, 50, 52, and 54) presented the highest content in the abdominal muscle. Additionally, PE (chain lengths: 36, 38), PC (chain lengths: 34, 36, 38), and FFA (chain length: 18) were among the most abundant (Fig. S2). The concentration of TG (chain lengths: 46, 48, 50, 52, and 54) in the ST sample was significantly greater than that in the other three samples, whereas the concentration of PE (chain lengths: 36 and 38) was significantly lower than that in the other samples ([82]Fig 2EF). Compared with that in the control (CK), the concentration of PC (chain lengths: 34, 36, 38) increased significantly following heat treatment, whereas the concentration of FFA (chain length: 18) decreased significantly ([83]Fig 2GH). Fig. 2. [84]Fig. 2 [85]Open in a new tab Content of different double bond equivalents (DBEs) in abdominal muscle (A) TG, (B) PE, (C) PC, (D) FFA. Content of lipid compounds with different chain lengths in abdominal muscle (E) TG, (F) PE, (G) PC, (H) FFA. * P < 0.05, ** P < 0.01, *** P < 0.001. CK: raw abdominal muscle, SV: sous vide abdominal muscle, ST: steamed abdominal muscle, OC: oven-cooked abdominal muscle. 3.3. Differential lipidomics analysis in cooking abdominal muscle Principal component analysis (PCA) allows for the visual interpretation of complex datasets, revealing groupings, trends, and outliers within the data ([86]Guo et al., 2022). Orthogonal partial least squares discriminant analysis (OPLS-DA) is a supervised pattern recognition multivariate statistical analysis method that effectively removes unrelated influences to identify differentially abundant metabolites. Utilizing a dataset comprising 1230 lipids, PCA and OPLS-DA analyses were conducted on the metabolite data of cooked abdominal muscle to discern differences among samples subjected to various cooking treatments ([87]Fig 3AB). The PCA results showed that PC 1 and PC 2 account for 36.19 % and 20.1 % of the total model, respectively, highlighting differences among the four sample groups. Notably, the difference between SV and OC was minimal, whereas the differences between CK and ST, and OC were substantial. The classification metrics R2Y and Q2 were 0.995 and 0.785, respectively, demonstrating that the OPLS-DA model is reliable and has good fitting and predictive capabilities ([88]Bi et al., 2023). These findings indicate that OPLS-DA is effective for differentiation applications. Furthermore, the results demonstrate that the metabolic profile of abdominal muscle undergoes significant changes after cooking, aligning with previous studies. Fig. 3. [89]Fig. 3 [90]Open in a new tab PCA score plot (A), orthogonal partial least squares discriminant analysis (OPLS-DA) score plots (B). Venn diagram of the number of differential lipid molecules in abdominal muscle between different cooking methods (C). Volcano plot of differential lipid molecules in ST vs CK (D), SV vs CK (E), OC vs CK (F). Each point in the volcano plot represents a metabolite, with the x-axis reflecting the fold change in lipid differences between groups (log[2]FoldChange) and the y-axis indicating the level of significance (-log[10]FDR). Significantly upregulated lipids are represented as red points, significantly downregulated lipids as green points, and gray points indicate metabolites detected but not significantly different. The size of the points corresponds to the VIP value. Effect of heat treatment on the amount of differentially changed lipids in abdominal muscle: ST vs CK (G), SV vs CK (H), OC vs CK (I). (For interpretation of the references