Lipid analysis of meat from Bactrian camel (Camelus bacterianus), beef, and tails of fat-tailed sheep using UPLC-Q-TOF/MS based lipidomics

Introduction As a source of low-cost and high-quality meat for human beings, the consumption of camel meat was increasing, and beef has similar texture and nutritional characteristics with camel meat. Camel hump and fatty-tails are important parts of fat storage for camels and fat-tailed lambs, respectively, which were to adapt and endure harsh environments. Considering their similar physiological functions, their fat composition might be similar. Lipidomics is a system-level analysis of lipids method, which play an important role in the determination and quantification of individual lipid molecular specie, food adulteration and labeling. Methods A GC/MS was used to analyze fatty acids composition of Xinjiang Bactrian camel meat, hump, beef, and fatty-tails. UPLC-Q-TOF/MS based on lipidomics approach was used to analyze lipid composition, characterize and examine the lipid differences in Xinjiang Bactrian camel meat, hump, beef, and fatty-tails. Results and discussion The major fatty acids of the four samples were C16:0, C18:0, and C18:1cis, and camel meat had a significant low SFA content and high MUFA content. A total of 342 lipid species were detected, 192, 64, and 79 distinguishing lipids were found in the groups camel hump compared to camel meat, camel meat compared to beef, and camel hump compared to fatty-tails, respectively. Lipid metabolisms of ether lipid, glycerophospholipid, glycerolipid, and sphingolipid were the most influential pathways revealed by KEGG analysis. The results contributed to enrich the lipid information of Bactrian camel meat, and indicated that UPLC-Q-TOF/MS based on lipidomics was an alternative method to distinguish meat samples.


. Introduction
Dietary lipids comprise glycerol and long-chain fatty acids, which can provide essential nutrients for maintaining human health and contribute to various intra and extracellular signaling pathways (1). Dietary lipids are also involved in many pathological processes, such as microbial imbalance, metabolic complications and neuropathological processes (2). Meat is rich in lipids, and the different content and type in meat have been linked to dietary health concerns, plus contribute as important flavor components (3). Demand for meat continues markers in pork cuts, and black pig varieties were analyzed using UPLC-ESI-MS/MS, respectively (16). Many lipids, such as glycerophospholipids, were found to be the marker candidates between chicken meat and abdominal fat from different sources with an LC-MS-based lipidomics approach (14). At present, lipidomics has been widely used in various research fields, such as dairy product adulteration, disease diagnosis, mechanism of food lipid oxidation mechanism (17)(18)(19)(20).
In this study, the meat lipids from Xinjiang Bactrian camels, beef, and fatty-tails were analyzed using UPLC-Q-TOF/MS based on lipidomics to establish a characterization of the lipid profiles. The lipid datasets of groups, camel meat compared with camel hump, camel meat compared with beef, and camel hump compared with fatty-tails, were analyzed to identify the differential lipids and related lipid metabolisms. The study will provide lipid information from Bactrian camel meat and camel hump, give lipid species clue to discriminate camel meat and beef, camel hump and fatty-tails.

. . Preparation of meat samples
The longissimus lumborum (LL) and hump of Bactrian camels were supplied by Xinjiang Wangyuan Biological Technology Group Co., Ltd. (Xinjiang, China); the male camel was about 4 years old. The LL of beef (Bos Taurus) and tails of fat-tailed sheep (Ovis Aries) were obtained from a same distributor. The LL were taken from the 13th/14th rib immediately after slaughter, quick-frozen, stored at −80 • C for further analysis.

. . Histology
Meat samples (2 × 1 × 1 cm) were trimmed and were fixed using 10% formalin. The samples were dehydrated with gradient ethanol, serially hyalinized in xylene: ethanol (v/v, 1:1) and xylene, and the tissue mass was imbedded in paraffin wax. After, the samples were sliced to 5-µm thickness by a microtome, and fixed between a slide and cover slip. Sections were dewaxed and rehydrated using xylene, gradient ethanol, and distilled water. After each section was stained with hematoxylin and eosin, the samples mounted using neutral resins after being dehydrated and hyalinized. Finally, images were taken using an ETHOS1 microscope (LabTechq, Beijing, China) at a 100× magnification.

. . Fatty acids analysis
The Folch method was used to extract the fat (21). The fat transesterification/methylation procedures followed: 40 mg of fat were weighted in tubes, 240 µL of 10% BF 3 -methanol were added and heated at 100 • C water for 90 min. After cooling to laboratory temperature, the samples added 480 µL of nhexane and 1.5 mL of deionized water. Shaken the samples and centrifuged at 2,000 rpm for 5 min, collected the organic phase (upper layer). The upper solution was repeatedly washed (deionized water) and centrifuging (2,000 rpm, 5 min) three time. After that, 0.2 g of anhydrous sodium sulfate was added to the samples and held at 4 • C for 30 min. Finally, the samples were centrifuged (2,000 rpm, 5 min) and collected upper solution.
Fatty acid analysis was performed by a GC/MS (70090B/5977A, Agilent Technologies, USA) with a FID detector, using a DB -5MS capillary column (30 m long, film 0.25 µm thickness and 0.25 mm internal diameter). The GC parameters: helium flow rate, 1.0 mL/min; vaporization chamber temperature, 250 • C; injection volume, 1 µL; split ratio, 30:1. The oven temperature started at 80 • C for 5 min, then which was increased to 260 • C at 5 • C/min and held for 15 min. The MS parameters: temperature of the inlet, 290 • C; ionization mode, EI; ion source temperature, 200 • C; electron energies, 70 ev; scan of collecting in m/z range 3-550 amu.
. . Analysis of lipidomics . . . Lipid extraction procedure The lipid fraction was extracted with MTBE. The muscle samples (camel meat and beef) were minced, and 10 mg was precisely weighed, and was spiked with 1,600 µL of MS-grade water in a glass bottle with a screw cap. After, 200 µL of the homogenization sample was diluted with 200 µL of water. Likewise, 3 mg of the adipose samples (hump and fatty-tails) were weighed and spiked with 400 µL of MS-grade water. 960 µL of MTBE: methanol (v/v, 5:1), which comprised 9 µL of 10 ppm d7-PE, 9 µL of 10 ppm d7-LPC, and 9 µL of 100 ppm d7-TG, was added to the bottles. After the samples vortex-mixed for 60 s and ultra-sonicated for 10 min, the mixture was frozen centrifuged at 3,000 rpm for 15 min. Collected the upper solution (organic phase), and the lower layer was re-extracted with 500 µL of MTBE for extraction twice more. The collected organic phases were combined and evaporated under nitrogen. Lipid extractions were re-dissolved by 100 µL of DCM: methanol (v/v, 1:1) and 60 µL of the solution was removed to analysis. Five replications of each sample were performed.

. . . UPLC-Q-TOF/MS conditions
Lipids analysis was performed with 1290 UPLC (Agilent, USA) equipped with Triple TOF 6600 (Q-TOF, AB Sciex, USA), using a Phenomen Kinetex C18 100A Column (100 × 2.1 mm, 1.7 µm). The mobile phase A was comprised 10 mM Ammonium acetate, 40% H 2 O and 60% ACN, and the mobile phase B was comprised 10 mM Ammonium acetate, 10% CAN and 90% IPA. An elution gradient started with 40% B for 12 min; subsequently changed to 100% B within 1.5 min; at 13.7 min, changed to 40% B again. Run duration was 18 min with 300 µL/min. The injection volumes were 1 µL (positive mode) and 3 µL (negative mode). The Analyst TF 1.7 (AB Sciex, USA) constantly assessed the comprehensive scan survey MS data as it collected and touched off the acquisition of MS/MS spectra depending on preselected criteria. The program of ESI source: ion source gas 1 and 2 were 60, curtain gas was 30; the source temperature was held at 550 • C, ion spray voltage floating was 5,500 and −4,500 V.

. . Statistics analysis
The original data format was converted to mzXML using the ProteoWizard version 3.0.6150 (California, USA). The mzXML .

. . Tissue sections of meat
The histological sections of LL of meat from camel and beef as well as adipose tissue from the hump and fatty-tails were shown in Figure 1, displaying the distribution variation of fat. Visually, the muscle fiber structure was unbroken with shape of cable in muscle tissue from camel meat and beef, and comparing with beef, the camel muscle had a bright red color. The sizes of the camel muscle fibers were much smaller than the beef 's, further, the extracellular spaces were smaller in the camel muscle. Moreover, unilocular adipocytes comprised most of the adipose tissue mass in hump and fatty-tails, and the cell size of the hump was larger than that of fatty-tails.

. . Comparison of lipid profiles between the camel meat and camel hump
In the study, OPLS-DA scores were applied to elucidate intragroup separation and variables responsible for classification, A distinction was observed in the score scatter plot of OPLS-DA model for the group of camel hump compared with camel meat ( Figure 3A). In addition, a volcano plot and hierarchical clustering were constructed, which showed the lipid composition of the camel hump and camel meat were significantly different ( Figures 3B, C). A total of 236 lipids with differential abundance (VIP > 1 and p < 0.05). Among them, 192 species were found to have an FC < 0.5 or FC > 2, with 99 lipids were high in camel hump and 93 lipids were high in camel meat, which were potential markers for distinguishing between camel hump and camel meat groups (Supplementary material S1). To determine the possible pathways contributing to differentially expressed lipids, KEGG analysis was applied to identify metabolic pathways and their networks. A total of 8 lipid metabolisms were enriched (Supplementary material S2). The result of metabolic pathway analysis was shown in bubble plots ( Figures 3D, E). The metabolic pathways of glycerophospholipid metabolism, sphingolipid metabolism, ether lipid metabolism, and glycerolipid metabolism have the most impact values (0.4649, 0.2807, 0.2143, and 0.1070, respectively) and might play a key role in developing lipids.

. . Comparison of lipid profiles between camel meat and beef
There was good visual separation between camel meat and beef group in OPLS-DA score plot ( Figure 4A). Furthermore, the volcano plot presented a differential compound between camel meat and beef ( Figure 4B). To identify the potential biomarkers from the differential metabolites, we clustered the differential metabolites using heat maps and hierarchical clustering ( Figure 4C). The results showed that PLs, DGs, and TGs differed in camel meat and beef. Based on the OPLS-DA model and multivariate statistics (VIP > 1 and p < 0.05, FC < 0.5 or FC > 2), 64 individual lipid species were screened as potential molecules for distinction to the camel meat and beef (Supplementary material S1). 41 lipids were more abundant in camel meat than those in beef. Different expressed lipids of camel meat and beef were mainly related to seven lipid pathways (Supplementary material S2). Among the identified metabolic pathways, glycerolipid metabolism, sphingolipid metabolism, and glycerophospholipid metabolism were observed to be the most important, their impact values were 0.1070, 0.2807 and 0.2682, respectively ( Figures 4D, E).

. . Comparison of lipid profiles between camel hump and fatty-tails
The camel hump and fatty-tails were also investigated to identify potential lipid markers. The result of OPLS-DA demonstrated that camel hump and fatty-tails samples were separated from each other, and a permutation test indicated that the model had robustness and reliability ( Figure 5A). The volcano plot and heat map (Figures 5B, C) on the different metabolites presented the clustering of lipids with differential abundance  ). (B) Volcano plot using log FC as the x-axis, -log (p < . ) as the y-axis, FC values were the ratios of lipid content in camel hump to that in camel meat. (C) Heat map analysis and hierarchical clustering of di erential abundance lipids between camel hump and camel meat. To build the heat map, normalized abundances were used; the red color represents high abundance, whereas the blue color represents low abundance. Bubble plots of KEGG pathway analysis in positive (D) and negative (E) ionization mode, and the supplementary data showed in Supplementary material S . between the camel hump and fatty-tails. A total of 79 lipid molecules (VIP > 1 and p < 0.05; FC < 0.5, FC > 2) were selected as chemical descriptors, 54 lipids were more abundant in camel hump compared with fatty-tails (Supplementary material S1). The result of pathway analysis showed that there were constructed eight metabolic pathways in the group (Supplementary material S2). Among the identified metabolic pathways, lipid metabolisms of ether lipid, glycerolipid, sphingolipid, and glycerophospholipid had the greatest impact on differential lipids, their values of impact were 0.1071, 0.1070, 0.2807, and 0.4649, respectively ( Figures 5D, E).

. Discussion
There was a difference in color, structure, and IMF content between the LL of camel and beef. Factors such as breed, diet, production systems, and genetics may influence display color traits variability and IMF deposition of muscle (22). Likewise, fat deposition of the adipose tissue in domestic animals was affected by the same factors, and was also influenced by resistin and leptin (23).
In this study, fatty acid contents and species were different in camel meat, beef, hump and fatty-tails. These differences are in agreement with the literatures (10). C16:0, C18:0, and C18:1cis were the major fatty acids in all samples. Compared with dromedary camel muscle, Xinjiang two-hump camels have higher content of C18:1 cis (41.02 vs. 25.9%), lower content of C16:0 (15.55 vs. 22.40%), and C18:0 (12.8 vs. 16.9%) (10). Furthermore, palmitic acid, stearic acid, and oleic acid were shown as the predominant mixtures of fatty acids in camel humps (24). Kadim et al. reported the proportions of C18:1 cis (33.5%) and total MUFAs (41.4%) in dromedary camel longissimus thoracis, which were lower than LL of camel meat from Xinjiang, the difference may be attributed to the camel species and cuts (25). Effectiveness of a MUFA-enriched diet to improve insulin sensitivity and associated cardiometabolic risk, as well as improve blood lipids and systemic inflammatory responses and endothelial dysfunction (26). The ratio of n-6: n-3 PUFAs was closely related to human health, including multiple biological processes, metabolic homeostasis, and diseases (27). Among the samples, the ratio of n-6: n-3 PUFAs in beef was within the recommended values of British Department of Health, which should not exceed 4.0 (28), in this respect, beef had better nutritional value than camel meat. The P:S is approximately 0.31 in camel meat, it was similar to the value (around 0.3) in dromedary camel indicated by Abdelhadi et al. (8). From a consumer-health viewpoint, nutritional advice for this ratio is 0.4 or higher (29). In the study, the ratio of P:S in beef (0.28) was slightly lower than camel meat, the value of beef was similar with Limousin bulls (0.29) (30), but camel meat was inconsistent with Maqsood et al. (4).  , Q = .
). (B) Volcano plot using log FC as the x-axis, -log (p < . ) as the y-axis, FC values were the ratios of lipid content in camel meat to that in beef. (C) Heat map and hierarchical clustering of di erential abundance lipids between camel meat and beef. To build the heat map, normalized abundances were used; the red color represents high abundance, whereas the blue color represents low abundance. Bubble plots of KEGG pathway analysis in positive (D) and negative (E) ionization mode, and the supplementary data showed in Supplementary material S . n-3 PUFAs-PI was high, and PI (18:0/20:4) was the major species. It had been reported that PI is the primary source of C20:4 n-6, and required for biosynthesis of eicosanoids (31). Beef has the highest content of Cer, which was associated with regulating cellular proliferation and apoptosis, and can be an important mediator of cancer-promoting effects of chronic alcohol consumption (32). PC was the main lipid differential species in camel meat and beef, accounting for about 25% of the total lipid differential species. PC is an important source of choline sphingomyelin and plasmalogen, and the supplement of choline in the form of phosphatidylcholine could reverse fatty livers (33).
To further explore the generation of differential lipid molecules, we used the KEGG pathway analysis to identify eight lipid metabolic pathways. Lipid metabolisms of ether lipid, glycerolipid, sphingolipid, and glycerophospholipid were the most influential pathways for three groups. LysoPC (O-18:0) was mainly enriched in ether lipid metabolism. Ether lipids were an abundant subclass of glycerophospholipid present in membranes, and were a fundamental building block of glycosylphosphatidylinositol (GPI) anchors (34). TG was a primary participant of glyceride metabolism, and serving as the primary form of energy storage in mammalian cells, synthesis and lipolysis of TG is associated with disease, for example, obesity, hypertension, and lipodystrophy (35). In the study, TGs in camel meat were significantly lower than in camel hump. TG was the major lipid class in adipose tissue (>90%), whereas there was a considerable proportion of PL in the muscle tissue (36). Here, the TGs with long-chain and medium-chain FAs were high in camel hump and fatty-tails, respectively. This was probably related to TGs containing long-chain and/or mediumchain fatty acids, and different metabolic processes to synthesize and store TGs (37). Besides, cold stimulates could also cause a highly selective increase in the abundance of long-chain/ultra-longchain/odd-chain acyls TG (38).
We observed that Cer (d18:1/18:0) was more abundant in camel meat compared with camel hump, however, Cer (d18:1/18:0) and Cer (d18:1/24:1) were lower in camel meat than in beef. Cer intermediates in sphingolipid metabolism, as lipid second messengers, it played an important role in pathophysiological mechanisms, and the plasma concentrations of Cer increased after diet (39). However, recent research has indicated that the higher Cer (d18:1/18:0) and Cer (d18:1/24:1) concentrations in plasma were related with the development of adverse cardiovascular events (40). Differential metabolites of PLs, such as PC, PE, and PS, were mainly regulated by glycerophospholipid metabolism, and the majority were PC and PE. Dietary PL could improve antioxidant capacity, promote lipid utilization and reduce liver lipid accumulation (41). In the study, PUFAs had side chains coupled with mostly PL, such as PC   ). (B) Volcano plot using log FC as the x-axis, -log (p < . ) as the y-axis; FC values were the ratios of lipid content in camel hump to that in fatty-tails. (C) Heat map and hierarchical clustering of di erential abundance lipids between camel hump and fatty-tails. To build the heat map, normalized abundances were used; the red color represents high abundance, whereas the blue color represents low abundance. Bubble plots of KEGG pathway analysis in positive (D) and negative (E) ionization mode, and the supplementary data showed in Supplementary material S .

. Conclusion
In this study, comparative lipidomics analysis of camel meat, camel hump, beef, and sheep fatty-tails were performed. There was a difference in fatty acid composition and camel meat had a significant low SFA content and high MUFA content. The potential lipid markers for distinguishing meat from different species were identified. PS (18:0/20:3), PI (16:0/20:4), PE (12:0/17:0) and DG (16:1/16:1), PC (P-16:0/17:2), TG (17:2/17:2/17:2) and TG (18:0/18:0/22:5), Hex2Cer (d14:1/16:0), TG (17:0/18:1/18:2) were the lipid components with the largest fold differences between camel meat and camel hump, camel meat and beef, camel hump to that in fatty-tails, respectively. UPLC-Q-TOF-MS based on lipidomics could be a feasible strategy to distinguish the meat samples. The results of pathway analysis suggested that ether lipid metabolism, glycerophospholipid metabolism, glycerolipid metabolism, and sphingolipid metabolism might be the most impactful to develop lipid profile in camel meat, hump, beef, and sheep fatty-tails. This study contributed to enrich lipids information of Bactrian camels, and provided a groundwork to discriminate camel meat, camel hump, beef, and fatty-tails. In the future, further investigation is required with a larger number of camel meat samples collected from different areas and breeds for an extended characterization of lipid profiles, meanwhile, comparison with other livestock meat is needed for exploring differences in lipids.

Data availability statement
The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding authors.  . /fnut. .