Unraveling the genetic underpinnings of fat metabolism and SDR16C5 mediated molecular mechanisms in adipogenesis of domestic pigs

Abstract

Background

Fat metabolism plays a pivotal role in determining meat quality and economic value in domestic pigs, yet its underlying genetic mechanisms remain poorly understood. To clarify the molecular basis of adipose deposition, this study integrated whole-genome resequencing data from 226 individuals across 61 Eurasian pig breeds and applied population genetic and selective sweep analyses to identify and validate candidate genes involved in fat metabolism.

Results

Significant genetic divergence was observed between Asian and European domestic pigs, with higher linkage disequilibrium (LD) levels in domestic breeds likely driven by artificial selection during domestication. High-fat groups (backfat thickness: 39 mm; fat percentage: 35%) exhibited substantially greater fat deposition than low-fat groups (backfat thickness: 15 mm; fat percentage: 15%). Combined analyses of Pi-ratio, XP-EHH, and F_ST revealed 13 key candidate genes, including SDR16C5 and CPE, functionally enriched for lipid metabolism pathways such as fatty acid metabolism and insulin signaling. Cellular assays further confirmed that SDR16C5 promotes adipogenesis by enhancing preadipocyte proliferation and lipid accumulation.

Conclusion

This study identifies SDR16C5 as a candidate regulator of adipogenesis and fat metabolism in pigs. These findings establish a genetic framework that may facilitate targeted breeding strategies to improve fat composition and enhance premium pork production.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12864-026-13159-y.

Background

The domestication of pigs (Sus scrofa) began approximately 10,000 years ago and has resulted in extensive phenotypic and genetic diversification among pig populations [1–3]. Among the traits shaped during domestication and breeding, fat deposition represents an economically important characteristic that strongly influences pork quality. Appropriate fat content and distribution contribute to meat tenderness, juiciness, and flavor, thereby enhancing overall product quality [4, 5]. In addition, the fatty acid composition of pork fat is closely associated with human nutritional value and health concerns such as cardiovascular disease [6, 7]. Substantial variation in fat deposition exists among pig breeds, reflecting different breeding objectives and selection pressures. Chinese indigenous breeds, such as Meishan pigs (fat content: 27.15%) and Huai pigs (fat content: 26.7%), typically exhibit pronounced fat deposition and superior meat quality [8–10]. In contrast, Western commercial pig breeds (fat content: 14.13%) have been intensively selected for lean meat production [11, 12]. These marked phenotypic differences between Chinese and Western breeds suggest that divergent artificial selection has acted on genomic regions associated with lipid metabolism and fat deposition.

Selection signal analysis has become an effective genomic approach for identifying genetic loci underlying economically important traits in livestock. Because fat deposition has been a major target of artificial selection during pig domestication and breeding, genomic regions associated with fat-related traits are expected to exhibit detectable signatures of selection. Therefore, selection signal analysis is particularly suitable for identifying genes associated with fat deposition, as divergent breeding objectives among pig populations have left detectable genomic signatures. Selection signals refer to the phenomenon where the frequency of genetic variation in certain regions of the genome significantly deviates from the neutral expectation under selection pressures [13–15]. Population divergence among pig breeds with contrasting fat deposition phenotypes provides a natural framework for identifying genomic regions under selection associated with lipid metabolism. By comparing the genomes of breeds with extreme phenotypic differences (such as the high-fat Chinese breeds vs. the low-fat Western breeds), we can detect genomic regions that have been targeted by selection. This approach has been widely used to identify genomic regions associated with important economic traits like body size and disease resistance [3, 16], and is particularly instrumental in discovering key genes that regulate fat metabolism and deposition [17, 18]. Fat traits are influenced by multiple factors, including genetics, nutrition, and the environment, with genetic factors playing a dominant role [19]. The genetic regulation of these traits is a complex molecular network involving interactions between transcription factors, gene expression, and signaling pathways [17]. For instance, PPARγ and C/EBPα play crucial roles in adipocyte differentiation, while the mTOR signaling pathway plays a central role in both lipogenesis and lipolysis [20–23]. Research has shown that genes such as FABP4 and LPL significantly influence pig fat traits, regulating the formation and function of adipose tissue [24–26]. Despite these findings, the biological functions of many candidate genes identified through population genetic analyses remain poorly characterized, highlighting the need for further functional validation.

Population divergence among pig breeds with contrasting fat deposition phenotypes provides a natural framework for identifying genomic regions under selection associated with lipid metabolism. In particular, the pronounced differences between high-fat Chinese indigenous breeds and lean Western commercial breeds provide a useful framework for identifying genomic regions potentially associated with fat-deposition-related divergence. In this study, we applied selection signal analysis to whole-genome resequencing data from Chinese and Western pig breeds and identified SDR16C5 and CPE as two key candidate genes potentially involved in fat deposition. Previous studies have hinted at their association with lipid metabolism: for example, a study on 129 Iberian × Duroc backcross pigs found that SDR16C5 is associated with fatty acid composition [27]. Xiong et al. identified SDR16C5 as an important candidate gene for growth traits in pigs through a genome-wide analysis [28]. Analysis of liver sequencing data from cattle suggests that SDR16C5 is linked to differences in production traits [29]. Additionally, Niamh et al. demonstrated the impact of CPE on fat accumulation and its important role in insulin secretion regulation [30, 31]. However, no experimental evidence has yet conclusively verified the specific functions of SDR16C5 and CPE in lipid metabolism. Therefore, a deeper investigation into the functions of these two genes is the central objective of this study, aiming to fill the knowledge gap and provide crucial insights into the molecular regulatory mechanisms of lipid metabolism.

This study utilizes whole-genome resequencing data from 226 individuals representing 61 pig breeds to identify key genes associated with fat deposition in two pig populations with extreme phenotypic differences in backfat thickness and overall fat percentage, two widely used indicators of fat deposition. Based on these contrasting phenotypic characteristics, populations representing extreme differences in fat deposition were selected in this study for selection signal analysis. Comparing populations with divergent phenotypes is an effective strategy for identifying genomic regions associated with fat-related traits. The identified key genes were annotated for function using genomic annotation databases, and functional validation of the key genes SDR16C5 and CPE was conducted at the cellular level. The aim of this study is to apply population genetics methods combined with selection signal analysis to identify key genes influencing fat deposition traits in pigs and to experimentally validate their genetic mechanisms. This research will provide a scientific foundation for the production of high-quality pork and genetic improvement in pig breeding.

Results

Regional characteristics and genetic structure analysis of pig populations

To investigate the genetic structure and evolutionary relationships within Sus scrofa, we analyzed whole-genome sequencing data from 226 individuals representing 61 pig breeds. Based on geographic origin and taxonomic classification, individuals were assigned to five groups: Asian domestic pigs (ASD), European domestic pigs (EUD), Asian wild boars (ASW), European wild boars (EUW), and outgroup species (OGP), including Phacochoerus africanus, Sus barbatus, Sus cebifrons, and Sus celebensis (Fig. 1A). Principal component analysis (PCA) based on genome-wide SNP data revealed clear genetic differentiation among these populations. The first principal component (PC1), explaining 39% of the total variance, separated Asian and European lineages, while the second principal component (PC2), accounting for 12.72% of the variance, distinguished domestic pigs from wild boars within each geographic region (Fig. 1B; Fig. S1A–B). Asian domestic pigs clustered closely with Asian wild boars, and European domestic pigs clustered with European wild boars, indicating regional genetic continuity between domestic and wild populations. In contrast, the outgroup species were clearly separated from all Sus scrofa populations in the PCA space, indicating substantial genetic divergence.

Fig. 1.

Regional Characteristics and Genetic Structure Analysis of Pig Populations. (A) Schematic diagram of pig population groups based on global geographic distribution, illustrating the worldwide distribution of diverse pig breeds and reflecting the global diversity and geographic characteristics of pig populations. (B) Principal Component Analysis (PCA) plot showing two distinct and significantly separated clusters for Asian domestic pigs (ASD) and European domestic pigs (EUD). Asian wild boars (ASW) and European wild boars (EUW) cluster closely with their respective domestic counterparts, while the outgroup species (OGP) is clearly separated, indicating genetic divergence. (C) Phylogenetic tree confirming the close genetic relationships between domestic pigs and their regional wild counterparts. Asian and European domestic pigs cluster with Asian and European wild boars, respectively, while the outgroup species occupies a distant position. (D) Linkage disequilibrium (LD) decay analysis demonstrates higher LD levels in domestic pigs compared to wild boars, reflecting selection pressure and genetic drift during domestication, as well as the higher genetic diversity in wild boars. (E) Population structure analysis using Admixture software, displaying genetic component partitioning under different K-values (K = 2 to K = 6). The model with K = 5 shows the best fit to the genetic data (lowest cross-validation error), with population structure aligning with geographic distribution. This indicates that the genetic architecture of domestic and wild pigs is primarily driven by regional backgrounds in Asia and Europe

A maximum-likelihood phylogenetic tree constructed using genome-wide SNPs showed patterns consistent with the PCA results (Fig. 1C). Asian domestic pigs grouped with Asian wild boars, while European domestic pigs clustered with European wild boars, forming region-specific clades. The outgroup species formed distinct branches outside the Sus scrofa clade, consistent with their evolutionary divergence from domestic pigs and wild boars. Linkage disequilibrium (LD) decay analysis further showed differences between domestic pigs and wild populations. Domestic pig populations exhibited higher LD levels and slower LD decay compared with wild boars in both Asia and Europe (Fig. 1D), indicating stronger genetic drift and reduced effective population size in domesticated populations.

Population structure was further investigated using Admixture analysis with ancestral components ranging from K = 2 to K = 8. Cross-validation error analysis indicated that K = 5 provided the best fit to the data (Fig. S2). Under this model, individuals were largely partitioned into five genetic clusters corresponding to ASD, ASW, EUD, EUW, and OGP. At K = 2, the primary genetic division corresponded to the Asian–European split. With increasing K values, finer population substructure became apparent, particularly within European populations, where European domestic pigs and European wild boars gradually formed distinct ancestry components. Collectively, these analyses confirm the expected geographic genetic structure among pig populations and demonstrate clear differentiation between Asian and European lineages as well as between domestic pigs and their wild counterparts, consistent with previous studies. This population divergence reflects differences in breeding history and phenotypic characteristics, including fat deposition traits, and provides a suitable framework for comparative genomic analyses aimed at identifying genetic loci associated with fat deposition.

Selection signal analysis for identifying candidate genes associated with fat deposition

To detect genomic regions associated with fat deposition in pigs, breeds were classified into high-fat and low-fat groups using phenotypic information from the Pig Chronicles. Breed-level phenotypic summaries showed clear contrasts between the two groups. Breeds assigned to the high-fat group exhibited higher average backfat thickness (39 mm vs. 15 mm) and body fat percentage (35% vs. 15%) than breeds assigned to the low-fat group (Fig. 2A–B). These descriptive comparisons are consistent with the classification criteria used to define the two groups. Each point shown in Fig. 2A–B represents a breed rather than an individual animal. These results confirm clear phenotypic divergence between the breed groups. Together with the observed genetic differentiation among pig populations, these contrasts provide a framework for identifying genomic regions potentially associated with divergent evolutionary histories and fat-deposition phenotypes.

Fig. 2.

Identifying Candidate Genes Associated with Fat Deposition. (A) Breed-level average backfat thickness of pig breeds assigned to the high-fat and low-fat groups. (B) Breed-level average body fat percentage of pig breeds assigned to the high-fat and low-fat groups. Each point represents one breed included in the high-fat or low-fat group used for selective sweep analyses. (C) Genomic analysis results based on different methods (LogR-in-Ratio, XP-EHH, and F_ST), showing the signal distribution characteristics across genomic regions. (D-E) Overlapping genes detected by XP-EHH, Pi-Ratio, and F_ST analyses. Numbers in the diagram indicate the quantity of genes overlapping between different methods

Genome-wide selection scans were performed using three complementary statistics—π-ratio, F_ST, and XP-EHH—calculated across sliding windows of 50 kb. Candidate regions were defined as windows within the top 5% of the empirical distribution for each metric. This analysis identified 2,513 genes using F_ST, 1,259 genes using XP-EHH, and 1,945 genes using the π-ratio. An integrative analysis revealed 1,139 consensus genes detected by at least two methods (Fig. 2D). Among these, 13 genomic regions ranked within the top 1% across all three statistics and contained 13 annotated genes, including SDR16C5, CPE, PLAG1, AQP9, and VAV1 (Fig. 2C). These genes were considered candidate loci located within regions exhibiting strong signatures of divergent selection and therefore warrant further functional investigation. To investigate the biological functions of genes under selection, we performed functional enrichment analysis on the top 5% of genes ranked by Pi-ratio statistic. This approach was adopted because the intersecting gene set across methods was too small for reliable enrichment analysis. KEGG pathway analysis highlighted the Rap1 signaling pathway (P = 4.64 × 10⁻⁸), MAPK signaling pathway (P = 8.27 × 10⁻⁸), and Ras signaling pathway (P = 6.79 × 10⁻⁷) as the most significantly enriched pathways (Fig. 3B). GO analysis further identified enrichment in biological processes such as angiogenesis, apoptotic process, and regulation of gene expression, as well as molecular functions including kinase activity, GTP binding, and protein phosphorylation (Fig. 3A). These pathways are known to participate in multiple cellular processes, including cell proliferation, differentiation, and metabolic regulation. Previous studies have suggested that MAPK and Ras signaling pathways participate in adipocyte differentiation and lipid metabolism. Therefore, genes located within these selective regions may represent biologically plausible candidates for further investigation of fat-deposition-related processes.

Fig. 3.

Functional enrichment analysis of the top 5% of candidate genes subjected to selection. (A) Gene Ontology (GO) enrichment analysis: The bar plot shows the enrichment of GO terms categorized into three main domains: biological process (BP), cellular component (CC), and molecular function (MF). Key biological processes, cellular components, and molecular functions significantly enriched in the selected genes are shown, including “Positive regulation of angiogenesis,” “Early endosome,” and “Myosin binding.” (B) KEGG pathway enrichment analysis: The dot plot shows the enrichment of KEGG pathways in the top 5% of candidate genes subjected to selection. Major enriched pathways include Rap1 signaling, MAPK signaling, and Ras signaling pathways. Representative genes contributing to these pathways include PLAG1, VAV1, and AQP9, which have previously been implicated in cellular proliferation, differentiation, and metabolic regulation

Functional validation of candidate genes in cell proliferation, cycle regulation, and lipid deposition

To examine the potential involvement of candidate genes in adipocyte differentiation and lipid metabolism, the expression patterns of 5 genes identified from the selection signal analysis were first assessed during different stages of porcine adipocyte differentiation using RT–qPCR. Among the tested genes, SDR16C5 and CPE exhibited dynamic expression changes during differentiation (Fig. S3A). Both genes showed relatively low expression during the early stages of differentiation, followed by a marked increase in expression at later stages. To further characterize their expression profiles across tissues, transcriptomic data from the PigGTEx database were analyzed. SDR16C5 showed relatively higher expression in metabolically active tissues, including adipose tissue, placenta, and small intestine, while lower expression was observed in liver and kidney (Fig. S3B). In contrast, CPE was expressed across multiple tissues, with relatively higher expression levels in arteries, the hypothalamus, and the frontal cortex (Fig. S3C).

To investigate the functional roles of SDR16C5, gain- and loss-of-function experiments were performed in the mouse pre-adipocyte cell line 3T3-L1. Overexpression (pcDNA3.1-SDR16C5) and knockdown (si-SDR16C5) efficiency were confirmed by RT–qPCR (Fig. 4A–B). Flow cytometry analysis indicated that overexpression of SDR16C5 did not significantly alter the distribution of cells across cell cycle phases (Fig. 4C). However, CCK-8 assays showed that overexpression of SDR16C5 significantly increased cell proliferation (Fig. 4E). Conversely, knockdown of SDR16C5 significantly increased the proportion of cells in the G0/G1 phase and reduced the proportion of cells in the S phase (Fig. 4D), and was accompanied by a decrease in cell proliferation (Fig. 4F). During adipogenic differentiation, 3T3-L1 cells were induced for 8 days while maintaining gene overexpression or knockdown. Oil Red O staining and quantitative analysis demonstrated that SDR16C5 overexpression significantly increased lipid accumulation, whereas knockdown resulted in reduced lipid deposition (Fig. 5C–F). Microscopic observations of Oil Red O–stained cells further confirmed these differences in lipid droplet formation (Fig. 5E).

Fig. 4.

Modulation of 3T3-L1 Cell Proliferation and Cell Cycle Progression by SDR16C5 Overexpression and Knockdown. (A) Detection of overexpression efficiency of SDR16C5. (B) Detection of knockdown efficiency of SDR16C5. (C) Effects of overexpression of SDR16C5 on the cell cycle of 3T3-L1 cells. (D) The effect of knockdown with SDR16C5 on the cell cycle of 3T3-L1 cells. (E) CCK-8 assay by overexpressing SDR16C5 in 3T3-L1 cells. (F) CCK-8 assay by knockdown SDR16C5 in 3T3-L1 cells. All experiments were performed in triplicate, and the data are expressed as the mean ± S.E.M. (* p < 0.05; ** p < 0.01)

Fig. 5.

Modulation of 3T3-L1 Cell Differentiation by SDR16C5 Overexpression and Knockdown. (A) The overexpression efficiency of SDR16C5 was assayed after induction of differentiation. (B) The knockdown efficiency of SDR16C5 was assayed after induction of differentiation. (C) Oil red O staining results after overexpression of SDR16C5. (D) Oleoresin O staining results after knockdown of SDR16C5. (E-F) Analysis of oil red O content (OD value read at 500 nm). All experiments were performed in triplicate, and the data are expressed as the mean ± S.E.M. (* p < 0.05; ** p < 0.01)

Functional experiments were also performed for CPE using similar experimental approaches. Overexpression (pcDNA3.1-CPE) and knockdown (si-CPE) efficiencies were validated by RT–qPCR (Fig. S4A–B). Flow cytometry analysis indicated that overexpression of CPE did not significantly affect cell cycle distribution (Fig. S4C), whereas knockdown significantly increased the proportion of cells in the G0/G1 phase and decreased the S-phase fraction (P < 0.01) (Fig. S4D). CCK-8 assays showed that CPE overexpression significantly reduced cell proliferation (Fig. S4E), while knockdown significantly increased proliferation, particularly at 36 and 48 h of culture (P < 0.01) (Fig. S4F). After 8 days of adipogenic induction, RT–qPCR confirmed sustained overexpression and knockdown of CPE (Fig. S5A–B). Quantitative Oil Red O staining showed no significant difference in lipid accumulation following CPE overexpression, whereas knockdown resulted in a reduction in lipid content (Fig. S5C–D). Microscopic staining images further confirmed these differences in lipid deposition (Fig. S5E–F). Collectively, these results support functional roles for SDR16C5 and CPE in adipocyte proliferation and lipid deposition.

Discussion

Fat deposition is a key trait shaped during pig domestication and breeding. Previous studies have shown clear phenotypic differences between domestic pigs and wild boars as well as between Asian and European pig lineages [32]. Consistent with previous studies, our analyses confirmed clear genetic differentiation between Asian and European pig populations. Building upon these findings, our study further identified candidate genes potentially associated with fat deposition and provided functional evidence supporting their roles in adipocyte biology. The significant differences in fat content between domestic pigs and wild boars highlight the considerable impact of artificial selection on fat deposition [33, 34]. In contrast, many European commercial pig breeds exhibit lower fat content and have been intensively selected for lean meat production in modern breeding programs [11, 35]. These differences illustrate the phenotypic divergence between regional pig populations during domestication and subsequent breeding. Our selection signal analysis identified SDR16C5 and CPE within genomic regions showing strong signatures of divergent selection between pig populations with contrasting fat-deposition phenotypes. However, because the analyses were performed among domesticated breeds, these signals may reflect a combination of demographic history, breed-specific selection, and selection related to fat deposition. Despite the high genetic similarity between wild boars and domestic pigs, wild boars exhibit significantly lower fat deposition, indicating the strong selective pressure on fat storage during domestication [36, 37]. The introduction of wild pig species, such as the African warthog (Phacochoerus africanus) and bearded pig (Sus barbatus), further emphasizes the critical role of fat metabolism in pig domestication. Wild boars prioritize mobility and survival for environmental adaptation over energy storage, which helps explain their lower fat deposition capacity [38, 39]. The functional validation of SDR16C5 and CPE provides biological evidence supporting their potential involvement in adipocyte-related processes, although additional studies are required to determine whether these genes were direct targets of selection during pig domestication and breed improvement. Genetic analysis reveals significant genetic differentiation between Asian and European domestic pigs, with their corresponding wild boar populations also showing distinct genetic relationships with domestic pigs.

Our genetic analyses revealed significant genetic differentiation between Asian and European domestic pigs, with each regional wild boar population clustering closely alongside its corresponding domestic lineage. PCA and phylogenetic tree reconstruction supported independent domestication events in these populations. Compared to wild boars, elevated LD levels in domestic pigs indicate profound impacts of intensive artificial selection and genetic drift during domestication, consistent with previous studies [40]. Admixture analysis further unveiled the complex genetic architecture of porcine populations. At lower K-values (K = 2, K = 3), distinct divergence between Asian and European genetic backgrounds was observed, whereas increasing K-values (K = 4, K = 5) revealed finer-scale admixture patterns. These analyses primarily reflect the demographic history and population structure of pig populations rather than direct selection on fat-related traits. However, understanding this population structure is essential for interpreting subsequent selection signal analyses [36, 41]. Such historical gene flow likely contributed to the genetic diversity of modern pig breeds, providing an adaptive foundation for diverse environments and production systems.

The identification of candidate genes related to fat metabolism provides important clues to understanding the molecular mechanisms of fat deposition [42, 43]. Functional enrichment analyses further supported the biological relevance of the identified candidate regions. In particular, MAPK, Ras, and Rap1 signaling pathways were significantly enriched and contained several candidate genes involved in cellular growth, differentiation, and metabolic regulation. Previous studies have demonstrated important roles for MAPK and Ras signaling in adipocyte differentiation and lipid metabolism, suggesting that these pathways may contribute to fat-deposition-related phenotypes through coordinated regulation of adipogenesis and energy homeostasis. The phenotypic differences between the high-fat and low-fat breed groups are consistent with historical differences in breeding objectives and selection intensity among pig populations [44, 45]. This finding supports previous studies highlighting fat deposition as one of the important traits selected during pig domestication. Using various selection signal detection methods (Pi-ratio, XP-EHH, and F_ST), we identified several candidate genes located within genomic regions showing signatures of divergent selection, including SDR16C5, CPE, AQP9, VAV1, and PLAG1. These genes are involved in lipid metabolism, energy storage, and cellular processes, indicating their biological significance. Notably, the significant expression changes of SDR16C5 and CPE during adipocyte differentiation suggest their potential involvement in adipogenesis and lipid accumulation, as indicated by our previous experimental results. SDR16C5, a member of the short-chain dehydrogenase/reductase family, has been linked to lipid metabolism and oxidative stress response, which aligns with its high expression in metabolic tissues like fat, placenta, and small intestine in the PigGTEx database, supporting its role in regulating energy homeostasis [27, 46, 47]. Functional experiments in 3T3-L1 pre-adipocytes confirmed that SDR16C5 promotes cell proliferation and lipid accumulation, consistent with its association with obesity and metabolic disorders in human and mouse studies [48, 49]. CPE has garnered attention due to its roles in neuropeptide processing and metabolic modulation [50, 51]. Its elevated expression in the hypothalamus and prefrontal cortex suggests potential involvement in central energy homeostasis [52]. Functional validation demonstrated that CPE exhibits significant regulatory effects on cell cycle progression, yet exerts a weaker influence on lipid accumulation compared to SDR16C5. These findings imply that CPE’s role in lipid metabolism may exhibit complexity and context dependency, warranting further investigation. Additionally, genes such as AQP9 and PLAG1 play critical roles in lipid transport and phospholipid metabolism. AQP9, functioning as a glycerol channel, facilitates glycerol transport into adipocytes, providing essential substrates for triglyceride synthesis [53, 54]. PLAG1 mediates phospholipid hydrolysis and is intricately linked to inflammatory responses and metabolic pathways [55, 56]. The functional relevance of these genes in porcine lipid metabolism further underscores the importance of integrating genomic and functional approaches to elucidate molecular mechanisms underlying economically significant traits.

In summary, this study integrates population genomic analyses with functional validation to investigate the genetic basis of fat deposition in pigs. Through comparative population genomic analyses, we identified genomic regions exhibiting divergent selection between pig populations with contrasting fat-deposition characteristics and highlighted several biologically plausible candidate genes for further investigation. Notably, SDR16C5 and CPE were identified as candidate genes showing strong selection signals and dynamic expression during adipocyte differentiation. Functional experiments further demonstrated that SDR16C5 promotes adipocyte proliferation and lipid accumulation, while CPE influences cell cycle progression and may play a regulatory role in adipogenesis. These findings provide new insights into the genetic mechanisms underlying fat deposition during pig domestication and contribute to a better understanding of economically important traits in pig breeding.

Limitation

Despite these findings, several limitations should be acknowledged. First, the functional validation in this study was conducted using the murine pre-adipocyte cell line 3T3-L1 rather than porcine adipocytes. Although 3T3-L1 cells are widely used to investigate adipogenesis, species-specific differences may exist between mouse and pig adipocyte biology, and therefore the observed effects may not fully reflect gene functions in pigs. Second, the candidate genes identified in this study were inferred from genomic selection signal analyses. Such signals do not necessarily indicate direct causal relationships with fat deposition traits, as they may also arise from linkage with nearby loci or selection acting on correlated traits such as growth, reproduction, or immune function. Therefore, further validation using larger pig populations, porcine adipocyte models, and in vivo functional studies will be necessary to confirm the precise roles of these genes in porcine fat metabolism. Future studies integrating multi-omics approaches and functional validation in pig-specific systems will help to further clarify the molecular mechanisms underlying fat deposition in pigs. Third, the selective sweep analyses were conducted primarily among domesticated pig breeds with contrasting fat-deposition phenotypes. Consequently, some of the detected signals may reflect demographic history, population structure, or breed-specific selection rather than direct selection on fat deposition itself. Future studies incorporating wild boar reference populations, individual-level phenotypic measurements, and association analyses will help disentangle demographic effects from trait-specific selection. In addition, because the high-fat and low-fat groups were composed predominantly of Chinese indigenous breeds and European commercial breeds, respectively, phenotypic divergence and population ancestry could not be completely separated in the current study design.

Conclusion

In conclusion, this study provides new insights into the genetic mechanisms underlying fat deposition in pigs by integrating population genomics and functional validation approaches. The identification and characterization of key genes associated with adipogenesis contribute to a deeper understanding of the genetic architecture of economically important traits in pigs. These findings provide a valuable foundation for future studies on lipid metabolism and offer potential genetic targets for improving meat quality and production efficiency in pig breeding programs.

Materials and methods

Sample collection and grouping

This study utilized whole-genome resequencing data from 226 individuals representing 61 pig breeds worldwide. All samples were obtained from the National Germplasm Resource Center for Domestic Animals, encompassing domestic pigs, wild boars, and related wild pig species. Based on geographic distribution and genetic background, these breeds were categorized into 5 groups: Asian domestic pig, Asian wild pig, European domestic pig, European wild pig, and outgroup (Phacochoerus africanus, Sus barbatus, Sus cebifrons, and Sus celebensis). Detailed information on these pig breeds is provided in the supplementary table S1.

Based on breed-level average phenotypic data obtained from the Pig Chronicles database (https://zypc.nahs.org.cn/pzml/classify.html), which compiles backfat thickness and fat percentage measurements reported in previous studies, samples were classified into high-fat and low-fat groups according to their fat phenotype. These breed-level averages represent the typical fat deposition characteristics of each breed rather than individual-level measurements. Based on the reported fat deposition characteristics of different breeds, Chinese indigenous breeds with relatively high fat deposition (e.g., Meishan and Huai pigs) were classified as the high-fat group, while Western commercial breeds characterized by lean meat production (e.g., Large White, Landrace, and Duroc) were classified as the low-fat group. Only breeds representing the contrasting ends of the fat-deposition spectrum were included in the high-fat and low-fat groups used for selective sweep analyses, whereas breeds with intermediate fat-deposition phenotypes were not included. Wild pig populations and outgroup species were not included in the fat phenotype grouping due to the lack of comparable phenotypic records and were only used in population structure analyses. Because comparable fat-deposition phenotypes were unavailable for wild pig populations and outgroup species, selective sweep analyses were performed only between domesticated breed groups with contrasting fat-deposition characteristics. The backfat thickness values correspond to measurements of subcutaneous adipose tissue taken at the last rib, the standard anatomical site for pig backfat assessment. Body fat percentage was calculated as the ratio of fat weight to carcass weight, with adipose tissue comprising subcutaneous, abdominal, and visceral fat. For detailed data, see Supplementary Table S1 and Table S3.

Sequencing and variant calling

Raw sequencing reads were generated using the Illumina sequencing platform. Quality control of raw FASTQ files was performed using Fastp (v0.20.1) to remove adapter sequences and low-quality bases [57]. High-quality reads were aligned to the pig reference genome (Sscrofa11.1) using BWA-MEM (v0.7.17) [58]. The resulting alignment files were coordinate-sorted using SAMtools (v1.10), and PCR duplicates were removed using Picard. Variant discovery was performed using the GATK (v4.0.11.0) “HaplotypeCaller” pipeline. Individual GVCF files were generated for each sample and subsequently subjected to joint genotyping via “GenomicsDBImport” and “GenotypeGVCFs” to produce a multi-sample VCF file [59].

Variant filtering and quality control

SNP filtering was conducted using GATK “VariantFiltration” with the following hard-filtering criteria: QD < 2.0, MQ < 40.0, FS > 60.0, SOR > 3.0, MQRankSum < − 12.5, ReadPosRankSum < − 8.0, and QUAL < 30.0 [60]. Further quality control was applied using PLINK (v1.90), with SNPs retained based on MAF > 0.05, a SNP missing rate < 0.1, and an individual missing rate < 0.1. The final dataset showed an average sequencing depth of approximately 10× and a mean genome coverage of approximately 95%. In addition, only autosomal SNPs were retained for downstream analyses.

Population genetic analysis

We performed linkage disequilibrium (LD) pruning using PLINK (Version 1.90) with the command “--indep-pair-wise 50 10 0.2” to remove SNPs in strong linkage disequilibrium. These thresholds are commonly used in population genomic analyses to reduce marker redundancy while retaining informative genetic variation. The decay of LD relative to physical distance among SNPs was computed and graphically represented using PopLDdecay (Version 3.26). Principal component analysis (PCA) was executed with PLINK software (Version 1.90). Population structure was analyzed using ADMIXTURE software (version 1.3.0), with the number of ancestral clusters (K) ranging from 2 to 8. The optimal K value was determined based on the lowest cross-validation (CV) error estimated by the ADMIXTURE cross-validation procedure, and the results were visualized using the “RColorBrewer” R package. Genetic distance matrices were calculated using identity-by-state (IBS) distances implemented in PLINK software (--distance-matrix), and the tree was visualized in MEGA11 and iTOL (Version 7.1, https://itol.embl.de/).

Selective sweep identification

To identify genomic regions under selection associated with fat deposition, selective sweep analysis was performed between the high-fat and low-fat pig populations defined in this study. Three complementary statistics were used to detect selection signals, including nucleotide diversity ratio (π-ratio), fixation index (F_ST), and cross-population extended haplotype homozygosity (XP-EHH). The π-ratio and F_ST were calculated using vcftools (Version 0.1.16) with a sliding window size of 50 kb and a step size of 10 kb (--window-size 50,000 --window-step 10,000). The π-ratio was calculated as the ratio of nucleotide diversity between the low-fat and high-fat populations (πlow-fat / πhigh-fat) across sliding windows. The selected window size has been widely used in pig population genomic studies and provides a balance between detection resolution and statistical robustness for identifying candidate selective regions. For XP-EHH analysis, genotype data were first phased using BEAGLE software (Version 4.1) to infer haplotypes. XP-EHH statistics were then calculated using selscan (Version 2.0.0) to detect recent positive selection between the two populations. The normalized XP-EHH scores were summarized for each 50-kb genomic window. For each statistic, the top 5% of genomic windows were considered candidate selective regions. Genes located within overlapping regions identified by at least two methods were defined as candidate genes potentially associated with fat deposition traits.

Moreover, the KOBAS database (http://bioinfo.org/kobas/) was used to understand the function and complex pathways of candidate genes, including the Kyoto Encyclopaedia of Genes and Genomes (KEGG) and Gene Ontology (GO) in the study (corrected p-value < 0.05).

Gene expression analysis

The expression levels of candidate genes (SDR16C5 and CPE) in pig adipose tissue and other tissues were determined by querying the Pig Genotype-Tissue Expression (PigGTEx) database (http://piggtex.farmgtex.org/) [46]. The beige adipocyte differentiation model in pigs was provided by Professor Yanfang Wang’s research group. During the cell differentiation process, RNA was extracted at various time points, and RT-qPCR was used to assess the expression changes of the target genes, with β-actin serving as the internal control.

Cell culture

The murine preadipocyte cell line 3T3-L1 was purchased from the Cell Resource Center, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences. 3T3-L1 cells were cultured in high-glucose DMEM supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin, under conditions of 37 °C and 5% CO₂.

RNA extraction, cDNA synthesis, and quantitative PCR (qPCR)

Total RNA was extracted from cells using the FreeZol Reagent Total RNA Extraction Kit (Vazyme, China). RNA concentration and purity were assessed using a NanoDrop 2000 spectrophotometer (Thermo, Waltham, MA, USA). For cDNA synthesis, 1 µg of total RNA was reverse transcribed using the HiScript II Q RT SuperMix for qPCR (+ gDNA wiper) kit (Vazyme, China) according to the manufacturer’s instructions. Quantitative real-time PCR (qPCR) was performed using the ChamQ Universal SYBR qPCR Master Mix (Vazyme, China) on a LightCycler 96 qPCR System (Roche, Basel, Switzerland). Each 10 µL reaction contained 5 µL of 2× SYBR Master Mix, 0.4 µL each of forward and reverse primers, 1 µL of cDNA template, and 3.2 µL of nuclease-free water. The thermal cycling conditions consisted of an initial denaturation at 95 °C for 30 s, followed by 40 cycles of 95 °C for 5 s and 60 °C for 30 s. Melt curve analysis was performed after each run to confirm primer specificity. β-actin was used as the internal reference gene. All experiments were conducted with at least three biological replicates, each including three technical replicates. Relative gene expression levels were calculated using the 2^−ΔΔCt method. The primer sequences used in this study are provided in Supplementary Table S2.

Vector construction and siRNA oligonucleotide synthesis

The CDS sequences of the genes SDR16C5 and CPE were downloaded from the NCBI database (https://www.ncbi.nlm.nih.gov/). Using the pcDNA3.1 expression vector retained in the laboratory, corresponding overexpression vectors, pcDNA3.1-SDR16C5 and pcDNA3.1-CPE, were constructed. The interfering siRNA fragments, si-SDR16C5, si-CPE, and the corresponding negative control si-NC, were designed and synthesized by Guangzhou RiboBio. Primer information is provided in Supplementary Table S2.

Cell transfection

Cells were transfected using Lipofectamine 3000 (Invitrogen, USA) according to the manufacturer’s protocol. 3T3-L1 cells were seeded in 6-well plates and cultured to 70–80% confluence. Plasmids were transfected into the cells as required for the experiments. Transfection complexes were prepared in Opti-MEM (Gibco, USA) and added to the cells according to the Lipofectamine 3000 protocol. After 6 h of transfection, the medium was replaced with fresh complete culture medium. Following the manufacturer’s instructions, Lipofectamine 3000 reagent (Invitrogen, USA) was used to transfect pcDNA3.1-SDR16C5, pcDNA3.1-CPE, si-SDR16C5, si-CPE, and their respective controls into cells. In each well of the 6-well plate, 2000 ng of plasmid or 25 nM of oligonucleotide or inhibitor was added, and cell samples were collected at the corresponding time points as needed for the experiments.

CCK-8 assay and flow cytometry

CCK-8 Assay: Cell viability was assessed using the CCK-8 kit (Vazyme, China) according to the manufacturer’s instructions. Two hours prior to detection, 10 µL of CCK-8 solution was added to each well of a 96-well plate. The plate was then incubated in a 37 °C, 5% CO₂ incubator. Cell viability of preadipocytes cultured in the 96-well plate was measured every 12 h at a wavelength of 450 nm using a fluorescence microplate reader (BD BioTek, Winooski, USA).

Flow Cytometry: For cells seeded in 6-well plates, transfection was performed at concentrations of 70% (overexpression) and 40% (interference). When cells reached 90%-100% confluence, they were processed by washing with PBS, digesting with trypsin, and terminating with complete culture medium to obtain cell pellets. The cells were then resuspended in 70% ethanol and fixed at 4 °C. Cell cycle analysis was performed using a BD Accuri C6 flow cytometer (BD Biosciences, California, USA).

Induction of differentiation

The 3T3-L1 cell line is the most well-characterized and widely accepted in vitro model for studying adipocyte differentiation and lipid metabolism. Its differentiation protocol, molecular markers, and signaling pathways have been extensively documented over decades of research. This allows for direct comparison with a vast body of existing literature and provides a robust platform for initial functional screening [61, 62]. 3T3-L1 cells were first cultured in complete medium at 37 °C with 5% CO₂ to maintain the cells in the logarithmic growth phase. The cells were seeded in 6-well plates and further cultured until they reached full confluence, then maintained for an additional 2 days to achieve contact inhibition. Following this, the culture medium was replaced with differentiation medium containing inducers (1µmol/L dexamethasone, 5mmol/L 3-isobutyl-1-methylxanthine (IBMX), 2µM Rosiglitazone, and 5 µg/mL insulin) [63]. The differentiation induction process typically lasted for 2 days and activated a series of signaling pathways and gene expressions associated with adipocyte differentiation. Afterward, the medium containing insulin was replaced every 2 days to maintain and further promote differentiation into adipocytes until day 8, when cell samples were collected.

Oil Red O staining

Lipid accumulation was assessed via Oil Red O staining on day 8 of cell differentiation. Cells were washed three times with phosphate-buffered saline (PBS) and fixed in 4% paraformaldehyde solution for 30 min. After three PBS washes, cells were stained with Oil Red O (Sigma, St. Louis, MO, USA) dissolved in 100% isopropanol at room temperature for 30 min. Excess dye was removed by washing three times with distilled water. Cells were then observed and photographed under an inverted microscope. For quantitative analysis, Oil Red O dye was dissolved by washing with 100% isopropanol for 5 min. Lipid droplet content was quantified by measuring absorbance at 500 nm wavelength using a spectrophotometer.

Statistical analysis

Statistical analysis of qRT-PCR expression data was performed using GraphPad Prism 8.0 (GraphPad Software, USA). All qRT-PCR experiments were conducted with at least three independent biological replicates, each measured with three technical replicates. Data are presented as the mean of these independent replicates. Prior to analysis, normality and homogeneity of variance were assessed. Statistical significance between groups was evaluated using a two-tailed Student’s t-test when assumptions were met; otherwise, appropriate non-parametric tests were applied. Significance levels are indicated (*p < 0.05, **p < 0.01).

Abbreviations

AQP9

Aquaporin-9

CCK-8

Cell Counting Kit-8

CDS

Coding Sequence

CPE

Carboxypeptidase E

CV

Cross-validation

DMEM

Dulbecco’s Modified Eagle Medium

FBS

Fetal Bovine Serum

GO

Gene Ontology

GTEx

Genotype-Tissue Expression

GTP

Guanosine Triphosphate

IBMX

3-Isobutyl-1-methylxanthine

KEGG

Kyoto Encyclopedia of Genes and Genomes

LD

Linkage disequilibrium

MAPK

Mitogen-Activated Protein Kinase

mTOR

Mammalian Target of Rapamycin

NCBI

National Center for Biotechnology Information

PBS

Phosphate-buffered saline

PCA

Principal component analysis

PLAG1

PLAG1 zinc finger

RNA

Ribonucleic Acid

RT-qPCR

Reverse Transcription Quantitative Polymerase Chain Reaction

SDR16C5

Short chain dehydrogenase/reductase family 16 C member 5

SNPs

Single Nucleotide Polymorphisms

VAV1

Vav guanine nucleotide exchange factor 1

VCF

Variant Call Format

XP-EHH

Cross Population Extended Haplotype Homozygosity

Authors’ contributions

PY and PS: Validation, Visualization, Writing - original draft. PS: Data curation and Software. RM, PY and YM: Validation and Data curation. YM and LJ: Formal analysis, Investigation and Methodology. LJ and DW: Conceptualization, Formal analysis, Supervision and Resources, Funding acquisition, Project administration, Writing - review & editing. All authors approved the final manuscript.

Funding

This work was supported by the National Key R&D Program of China (No.2020YFA0509500), and the National Natural Science Foundation of China (Nos. 32222079 and 32202621). The funding bodies had no role in study design or in any aspect of the collection, analysis and interpretation of data or in writing the manuscript.

Data availability

The variation data generated during this study are publicly available in the Genome Variation Map (GVM) repository under accession number GVM001472 and can be accessed at https://ngdc.cncb.ac.cn/gvm/getProjectDetail?project=GVM001472. All other data supporting the findings of this study are included in this published article and its Supplementary Information files.

Declarations

Ethics approval and consent to participate

The experiments and animal care were performed according to the Regulations for the Administration of Affairs Concerning Experimental Animals (Ministry of Science and Technology, China, 2004). All experiments were approved by the Animal Care Committee of the Beijing Academy of Agricultural Sciences (Approval number: IAS2019-60).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.