1 Introduction
Adipose tissue is a critical organ in mammals, with essential functions such as energy storage, cold resistance and regulation of body homeostasis. Adipose tissue development is a dynamic process involving sequential stages of preadipocytes proliferation, differentiation and subsequent adipocyte hypertrophy. In the neonatal period, the preadipocytes undergo committed differentiation into mature adipocytes. The triglycerides accumulation within mature adipocytes leads to lipid droplet enlargement, ultimately results in increased adipose tissue mass and deposition
[1]. Adipose tissue can be divided into white adipose tissue (WAT) and brown adipose tissue (BAT). WAT primarily stores excess energy in the form of triglycerides and is distributed across subcutaneous, visceral and epididymal fat depots
[2]. In contrast, BAT is mainly found in rodents and infants
[3], and is located in regions such as the scapular area, clavicles and axillae
[4]. BAT converts chemical energy into heat via non-shivering thermogenesis to resistance cold
[5]. There are three types of adipocytes: white, brown and beige adipocytes
[6]. White adipocytes contain a unilocular lipid droplet, have few mitochondrial and have low expression of
UCP1, with their primary function being energy storage
[7]. Brown adipocytes have multilocular lipid droplets, are rich in mitochondria, express the classical thermogenic gene
UCP1 at high levels and primarily function to convert chemical energy into heat
[8]. Beige adipocytes, distinct cells found within white adipose tissue, share characteristics with brown adipocytes, including multilocular lipid droplets, high mitochondrial content and
UCP1 expression, and are also involved in thermogenesis
[9,
10].
Both white and beige adipocytes originate from mesenchymal stem cells (MSC), which proliferate and differentiate into preadipocytes. Key adipogenic regulators such as
PPARG and
CEBPA drive the differentiation of these preadipocytes into mature white adipocytes, enabling triglyceride storage
[4,
11]. However, in response to environmental stimuli, such as cold exposure and hormonal signals, MSC differentiate into beige preadipocytes. Under the regulation of the
PRDM16 transcription factor, these preadipocytes differentiate into beige adipocytes, expressing
UCP1 and contribute to thermogenesis
[12]. Brown adipocytes and skeletal muscle cells share a common origin from
EN1+/MYF5+ progenitor cells
[12].
PRDM16 recruits the
CEBPB transcription factor to promote the differentiation of progenitor cells into brown adipocytes
[4]. The SVF cells derived from adipose tissue comprises a variety of cell types and have significant cellular heterogeneity, including MSC, preadipocytes, vascular endothelial cells (VEC), fibroblasts and immune cells. Of these, MSC and preadipocytes are crucial for promoting adipocyte proliferation and regulating angiogenesis, thereby contributing to adipose tissue expansion and fat deposition
[13]. Recent studies have demonstrated that SVF cells isolated from neonatal mice can form adipose spheroids and secrete
LEPTINin vitro, it can also respond to adrenergic lipolytic stimulation, further supporting their essential role in adipogenesis
[14]. In human adipose tissue, three distinct progenitor cell types within the SVF cells have been identified, which can differentiate into preadipocytes and subsequently mature adipocytes, thus promoting lipid accumulation
[15]. Also, scRNA-seq of adipose-derived stem cells (ADSC) from women with pear-shaped and apple-shaped obesity revealed depot-specific adipogenic potential; abdominal ADSC preferentially differentiate into adipocytes promoting fat accumulation whereas gluteal ADSC are more associated with lipid and cholesterol metabolism
[16]. Collectively, SVF cells includes multiple cell types and early lineage markers involved in adipocyte differentiation and adipose tissue expansion, supporting its critical role in regulating fat deposition.
Due to the absence of brown adipose tissue and a functional
UCP1 gene in pigs, white adipose tissue becomes the primary fat organ, having a key role in regulating metabolic processes
[17]. Based on body fat percentage and lean meat content, pigs are generally classified as either fatty or lean types. Tibetan pigs, a typical fatty breed, inhabit high-elevation regions at about 2800 and 4000 m above sea level. To adapt to the harsh conditions of cold and hypoxia at these elevations, adult Tibetan pigs have significantly higher fat content and backfat thickness compared to Duroc pigs, with an average fat percentage of 41% and backfat thickness of 2.59 cm
[18]. In our study, we focused on the Diqing Tibetan pig, a high-elevation breed in China known for its exceptional fat deposition, thick backfat, and superior meat quality traits
[19]. In contrast, Duroc pigs, a globally recognized lean-type breed, are extensively used as sires in crossbreeding programs to improve the production performance of other pig breeds
[20]. Adult Duroc boars have a lean meat percentage of 65% and a backfat thickness of around 1.5 cm
[21]. As a representative Chinese indigenous pig breed, Tibetan pigs have significantly higher intramuscular fat content compared to the commercial Duroc pigs
[22]. Studies on porcine adipose tissue have demonstrated that adipose tissue growth predominantly relies on adipocyte hyperplasia during early developmental stages, characterized by the proliferation and differentiation of MSC into mature adipocytes. Meishan pigs, a representative fatty breed, have significantly larger adipocytes and thicker backfat than lean-type breeds such as Landrace. Notably, differences in backfat thickness between Meishan and Landrace pigs are evident as early as 1 week of age, with Meishan pigs having both a greater number and larger size of adipocytes
[23]. Meishan pigs consistently have thicker backfat and more advanced adipose development than Landrace pigs with same bodyweight
[24], it is suggested that early adipocyte morphology is closely associated with subsequent fat deposition. Similarly, in intramuscular adipose tissue, Wujin pig is a fat-type breed and it has larger intramuscular adipocytes than Landrace pigs. Local pig breeds generally have larger adipocytes and earlier adipogenic maturation than commercial breeds
[25], indicating that the degree of early adipocyte development could a crucial determinant for the extent of fat accumulation during later stages of growth. To explore the differences in adipocyte development between Tibetan and Duroc pigs, we conducted investigated the heterogeneity of their adipose tissues.
scRNA-seq technology is widely used to reveal cellular heterogeneity within tissues. Unlike flow cytometry, which is limited by the availability of antibodies and its sorting capacity, scRNA-seq allows for the detection of all cell types within a tissue sample. For example, scRNA-seq has been applied to mouse adipose tissue, where it identified a novel population of
DPP4+ MSC capable of differentiating into preadipocytes
[26]. Additionally,
CD81+ preadipocytes were observed to differentiate into thermogenic beige adipocytes under cold stimulation in mouse inguinal white adipose tissue, as revealed by scRNA-seq analysis
[27]. In human adipose tissue, scRNA-seq identified cellular heterogeneity and had an increase in immune cell populations with enhanced metabolic activity in obese individuals
[28]. As scRNA-seq technology continues to evolve, it has provided deeper insights into cellular heterogeneity in various tissues, including mouse brain neural cells, cardiac tissue and human spermatogenesis
[29–
31].
However, the cellular heterogeneity of adipose tissue in pigs, a significant agricultural animal, remains poorly understood. In this study, we investigated the cellular heterogeneity in the neck white adipose tissue of pigs and explored the differences in adipogenesis between Tibetan and Duroc pigs. Given the significant disparity in fat content between Tibetan and Duroc pigs
[18,
21], we hypothesized that there may be differences in the proportion of adipocytes within their adipose tissues, as well as in the proliferation and differentiation capacities of SVF cells. These differences could potentially contribute to the increased fat deposition observed in Tibetan pigs.
2 Materials and methods
2.1 Animal sources and adipose tissue collection
Three 4-day-old male Diqing Tibetan pigs were purchased from Yunnan Agricultural University in China and three 4-day-old male Duroc pigs were obtained from the Pig Breeding Farm (Beijing Zhongyu, China). After 12 h of fasting with free access to water, anesthesia was administered via ketamine injection, followed by exsanguination through the anterior vena cava. The carcasses were then cleaned with water and 75% ethanol. Subcutaneous white adipose tissue from the neck and inguinal regions was immediately collected and placed in 1.5 mL centrifuge tubes, flash-frozen in liquid nitrogen, and stored at −80 °C. Some adipose tissue samples were preserved in 4% paraformaldehyde solution and stored at 4 °C. The remaining subcutaneous white adipose tissue from the neck and inguinal regions was placed in high-glucose DMEM for SVF cell isolation.
2.2 Cell isolation and culturing
The adipose tissue from each pig was removed from the DMEM solution, briefly rinsed in 75% ethanol for 2 s, and then placed in phosphate buffered saline (PBS) within a sterile laminar flow hood. The tissue was minced into a homogenous, meat-like consistency. All minced tissues were added to HEPES buffer (Gibco, Thermo Fisher Scientific, Waltham, MA, USA) containing collagenase II (Gibco) at a final concentration of 0.2% (w/v) and incubated at 37 °C with shaking for 45 min. After digestion, an equal volume of cell culture medium consisting of high-glucose DMEM (Gibco), 200 U·mL−1 penicillin/streptomycin (Gibco), and 10% FBS (Gibco), then it was added to terminate collagenase activity. The mixture was then sequentially filtered and the collected solution was centrifuged. The resulting pellet contained SVF cells. Red blood cells were removed by adding red blood cell lysis buffer (Beyotime, Beyotime Biotech Inc, Haimen, Jiansu, China), followed by centrifugation. The isolated cells from each pig were resuspended in cell culture medium in 6-well plates. After incubating for 2 h at 37 °C to allow for cell adhesion, the medium was replaced and the cells further cultured for 3 days. Subsequently, a portion of the cells was maintained in culture for further experiments, while the remaining cells were cryopreserved in liquid nitrogen for future use.
2.3 Preparation of single-cell suspensions and cDNA library construction
After culturing the SVF cells for 1 day, a portion of the cells was collected using 0.25% trypsin digestion (Gibco). The SVF cells from three Tibetan pigs and three Duroc pigs were then separately mixed. After centrifugation at 1000 r·min−1, the pellet was resuspended in 0.04% bovine serum albumin and filtered through a 40 μm cell strainer to remove large cell fragments and debris. A 10 μL sample of the cell suspension was mixed with 10 μL of Trypan blue (Invitrogen, Thermo Fisher Scientific) and counted using the Countness II Automated Cell Counter. The cell concentration was adjusted to 700–1200 cells μL−1, ensuring a viability of over 80%. Single-cell gel beads in emulsion (GEMs) were prepared using the 10x Genomics Chromium Controller Instrument (Genomics Ltd., Oxford, UK), following the 10x Genomics protocol.
To obtain 10,000 cells for library construction, 10.2 μL of single-cell suspension from Tibetan pigs and 21.8 μL from Duroc pigs were combined and diluted with nuclease-free water to a final volume of 33.9 μL, following the 10x Genomics protocol (Genomics). The resulting single-cell suspension was then mixed with a master mix and loaded into the 10x Genomics Chromium Controller Instrument to produce GEMs. Subsequently, the GEM-RT program and cDNA amplification were performed to construct the cDNA library. The cDNA library was purified and size-selected according to the manufacturer’s protocol (Genomics), and quality was assessed using the Agilent 2100 Bioanalyzer System (Agilent, Santa Clara, CA, USA). The qualified library was then sequenced by BGI Genomics.
2.4 Detection of cell proliferation
According to the protocol of a BeyoClick EdU Cell Proliferation Kit with Alexa Fluor 488 (Beyotime), SVF cells from Tibetan and Duroc pigs, with equal volumes of cell suspension, were seeded separately into three replicate wells each of a 12-well plate and a 6-well plate, and cultured for 24, 48 and 72 h. After each time point (24, 48 and 72 h), cells in the 6-well plate were used for Hoechst nuclear labeling, while those in the 12-well plate were used for cell proliferation detection.
To perform nuclear labeling, prewarmed EdU working solution was added to the cells in three replicate wells of the 6-well plate at a final concentration of 1×. The cells were incubated at 37 °C for 2 h. After incubation, the cell culture medium was removed and 1 mL of 4% paraformaldehyde added to fix the cells. Following fixation, the cells were washed with buffer and permeabilized with Triton X-100 for 15 min. The Click Additive Solution, prepared using the reagents provided in the kit, was then added to cover the cells. After incubating in the dark for 30 min, Hoechst 33342 (provided in the kit) was used for nuclear staining, and the cells were imaged under a microscope.
For the CCK-8 assay, the CCK-8 solution from a Cell Counting Kit-8 (Beyotime,) was added to the cells in three replicate wells of the 12-well plate, containing cells from both Tibetan and Duroc pigs, at 24, 48, and 72 h. After incubating at 37 °C for 4 h, the absorbance values were measured at 450 nm using the Tecan Infinite F50 machine (Tecan, Männedorf, Switzerland).
2.5 Adipogenesis of SVF cells
Adipose SVF cells from Tibetan and Duroc pigs were cultured in adipogenesis medium in a 12-well plate, with three replicate wells for each group. The adipogenesis medium consisted of 4 nmol·L
−1 insulin, 10 mmol·L
−1 HEPES and 4 mmol·L
−1 glutamine
[32]. The medium was refreshed every other day. On day 8, cells were collected using TRIzol for RNA extraction.
2.6 Oil red O staining and lipid content measurement
The cells from Tibetan and Duroc pigs with three replicate wells were differentiated on day 8, the cells were washed for three times with PBS and subsequently fixed with 4% paraformaldehyde at room temperature for 15 min. Concurrently, an Oil Red O working solution was prepared by dissolving oil red O powder (Sigma, Washington, DC, UAS) in isopropanol, followed by mixing with deionized water in a 3:2 ratio and double filtration before use. After removal of paraformaldehyde, the cells were incubated with the Oil Red O working solution at room temperature for 20 min, followed by PBS washing until the red color disappeared. Images were captured under a microscope.
For lipid content quantification, PBS was removed, and 100% isopropanol was added to both the samples and blanks, followed by incubation at room temperature for 15 min. Subsequently, absorbance values were measured at 510 nm
[33].
2.7 RT-qPCR
A 500-ng sample of total RNA was reverse-transcribed into cDNA using a PrimeScript RT reagent kit with gDNA Eraser (Takara, Kyoto, Japan). The cDNA was diluted with water and used for qPCR analysis. The RT-qPCR reaction mixture was prepared using TB Green Premix Ex Taq (Takara). Briefly, 0.4 μL of primers as shown in Table S1 were dissolved in the reagent of TB Green Premix Ex Taq to achieve a final concentration of 0.2 μmol·L−1. Subsequently, 2 μL of cDNA was added to a total reaction volume of 20 μL.
The mixture was loaded into a 96-well plate and subjected to qPCR using the 7300 Fast Real-Time PCR System (Applied Biosystems, Foster City, CA, UAS). The thermal cycling program included an initial denaturation step at 95 °C for 30 s, followed by 40 cycles of denaturation at 95 °C for 5 s and annealing/extension at 60 °C for 30 s. Expression values were normalized to 18S mRNA levels. Relative mRNA expression was determined using the ΔΔCt method.
2.8 Cell oxygen consumption detection
On day 3, cells were transferred to Agilent Seahorse XF24 Cell Culture Microplates (Agilent) at a density of 10,000 cells per well and continued to differentiate in adipogenesis medium for 4 days. On day 7, XFe24 FluxPak plates (Agilent) were pre-incubated in a non-CO2 incubator for calibration, preparing them for the experiment.
On day 8, oligomycin, FCCP (carbonyl cyanide-p-trifluoromethoxyphenylhydrazone) and a mixture of antimycin A and rotenone were added to the XFe24 FluxPak plates (Agilent) at final concentrations of 1, 2 and 2 μmol·L−1, respectively. The cell medium was then replaced with Seahorse XF Assay medium (Agilent) including 4.5 mg·mL−1 glucose (Gibco), 1mM Sodium pyruvate (Gibco) and 1x GlutaMax (Gibco). The plates were then placed in a non-CO2 incubator. After calibration and equilibration of the XFe24 FluxPak plates in the Seahorse XFe24 Analyzer, the Agilent Seahorse XF24 Cell Culture Microplates were used to measure oxygen consumption capacity. At the end of the experiment, oxygen consumption data from each group (with five replications) were statistically analyzed and graphed.
2.9 Histological analysis
Adipose tissue was embedded in paraffin and sectioned into 5 μm slices. The slices were then placed in a heater at 68 °C for 1 h to melt the paraffin. Following this, xylene and ethanol with different concentrations were used to remove the paraffin. For staining, the slices were sequentially stained with hematoxylin and eosin to visualize nuclear and cytoplasmic features. The slices were processed by gradient dehydration with increasing concentrations of ethanol. Finally, xylene and neutral resin were used to mount coverslips on the slices. Three replicate slices from each pig were examined and images were taken at least three times per slice using a microscope.
2.10 Analysis of scRNA-seq data
Single-cell RNA sequencing data were processed using the Cell Ranger Single Cell Software Suite (v3.0.2; Genomics) and Seurat v3.1.2
[26]. This pipeline included sample demultiplexing, read alignment, filtering, unique molecular identifier (UMI) counting, inter-sample normalization and unsupervised clustering. Cell cluster visualization was performed using Loupe Cell Browser (v3.0; Genomics). Dimensionality reduction and clustering were based on principal component analysis (PCA) with the top 20 variable genes, followed by visualization using t-distributed stochastic neighbor embedding (t-SNE) with resolution of 0.5 for clustering.
Pseudo-time trajectory analysis was conducted with Monocle 3 using default parameters. Clusters associated with adipocyte development were identified based on the t-SNE plot and corresponding barcodes were extracted. Dimensionality reduction was performed using the DDRTree method via the reduceDimension function with max_components set to 2, and cell ordering was achieved accordingly. Significant genes driving cell ordering were identified using the differentialGeneTest function, and cells were subsequently arranged along the inferred developmental trajectory.
Differentially expressed genes (DEGs) were identified using a threshold of P≤ 0.05 and fold change > 2. Cell types were annotated based on known marker genes and DEGs. Functional enrichment of DEGs was conducted using the Metascape platform (metascape.org) for gene ontology (GO) analysis.
To assess differences in cell-type composition between Tibetan and Duroc pigs, the total number of cells from each pig breed was compared. The proportional change in each cell cluster was calculated by subtracting the number of cells in a given cluster from Duroc pigs from that of Tibetan pigs, followed by normalization to the total cell number in Duroc pigs to obtain the relative change in cell-type proportion.
3 Results
3.1 Tibetan pigs had greater adipogenic capacities in white adipose tissue than Duroc pigs
To elucidate the biological characteristics of WAT in Tibetan and Duroc pigs, subcutaneous WAT was collected from the necks of 4-day-old Tibetan and Duroc pigs and performed histological analysis using hematoxylin and eosin staining. We observed that both Tibetan and Duroc adipocytes contained unilocular large lipid droplets in WAT, with some multilocular lipid droplet cells present among the white adipocytes Fig.1.
We analyzed genes involved in adipogenesis and found that the expression of the early adipogenic regulator genes
CEBPA and
CEBPB was significantly higher in Tibetan pigs than in Duroc pigs Fig.1. Additionally,
Leptin expression was significantly elevated in Tibetan pigs Fig.1.
PPARG contributes to thermogenesis and the induction of browning of WAT
[34]. We also assessed the expression of beige adipocyte marker genes, including
PPARG,
DIO2 and
UCP3, and found that
UCP3 expression was higher in Tibetan pigs than in Duroc pigs Fig.1. However, glucose transporter and mitochondrial gene expressions were significantly higher in Duroc pigs than in Tibetan pigs Fig.1.
These findings indicate that Tibetan pigs had significantly higher adipogenic capacity but lower metabolic activity than Duroc pigs.
3.2 Single-cell sequencing revealed the SVF cells heterogeneity in subcutaneous adipose tissue from Tibetan and Duroc pigs
To understand adipocyte development in Tibetan and Duroc pigs, we performed scRNA-seq to uncover the SVF cells heterogeneity in subcutaneous WAT. SVF cells were isolated from neck WAT and cultured in a 6-well plate for 1 day. The cell viability of Tibetan pig was 95% and of Duroc pig was 86% as shown in Table S2. Subsequently, single-cell transcriptome library construction was performed, followed by sequencing. DNA libraries were evaluated and optimal fragment sizes ranging from 350 to 1200 bp were selected for sequencing (Fig. S1(a)).
Sequencing yielded around 480 million reads for both Tibetan and Duroc pigs. Tibetan pig generated 9874 cells with a mean of 50,018 reads per cell, a median of 3236 genes per cell, a median of 17,179 UMI counts per cell and a total of 16,352 detected genes as shown in Table S2. Duroc pig obtained 9051 cells with a mean of 53,335 reads per cell, a median of 3504 genes per cell, a median of 19,532 UMI counts per cell and a total of 16,090 detected genes as shown in Table S2. Subsequently, cell clustering was performed after removing low-quality cells based on criteria including a mitochondrial proportion exceeding 25%, nFeature-RNA less than 200 and greater than 4500, Fig. S1(b,c).
We identified 12 clusters in Tibetan pigs and 11 clusters in Duroc pigs through t-SNE clustering Fig.2 and named them based on different expressed genes and marker genes as shown in Table S3 and Table S4.
We defined clusters expressing
CD34 in Tibetan and Duroc pigs as mesenchymal stem cells (MSC), Fig.2. Previous studies have found the expression of
CD34 in progenitor cells isolated from porcine subcutaneous adipose tissue
[35].
CD34 has also been used as a marker for MSC in human adipose tissue
[36]. We identified the cluster with high
CD44 expression as adipose stem cells (ASC), Fig.2 because
CD44 is commonly used as a marker gene to identify and isolate mouse ASC
[37,
38]. Here, we name clusters expressing
CD44 and
CD142 as preadipocytes Fig.2.
Cd142 is a newly discovered marker gene for preadipocytes, used in single-cell sequencing of mice to identify preadipocytes
[26]. In human adipose tissue,
CD26+ adipocyte stem cells express
CD142[39]. In addition, we also identified clusters specifically expressing
ADIPOQ as mature adipocytes and clusters with specific proliferation marker gene
MKI67 as proliferating cells, Fig.2).
For clusters in Tibetan and Duroc pigs without differentially expressed genes, we could not define cell types and thus named them as unknown clusters. In Tibetan pigs, we extracted upregulated genes from the unknown cluster and performed GO analysis to understand their functions. We found that the regulation of protein catabolic processes, organelle biogenesis and maintenance and organelle assembly pathways were upregulated, Fig. S2(a). Similarly, for the unknown cluster in Duroc pigs, we performed GO analysis on upregulated genes and found that pathways related to the negative regulation of cell proliferation, extracellular matrix organization, and inflammatory response were enhanced, Fig. S2(b).
3.3 Mature adipocytes numbers were greater in Tibetan than in Duroc pigs
We compared the SVF cell heterogeneity between Tibetan and Duroc pigs and observed distinct differences. In Tibetan pigs, we identified separate clusters for proliferation and MSC, proliferation, VEC, myoblasts. In contrast, Duroc pigs had a single cluster combining myoblasts and VEC, while MSC formed a distinct cluster. Notably, Tibetan pigs lacked the MAC cluster, Fig.3.
Tibetan pigs had an increased proportion of mature adipocytes, oxidative metabolism cells, proliferation and preadipocytes, endothelial cells clusters. Conversely, the proportion of proliferation and ASC, preadipocytes, VSMC and MSC clusters decreased in Tibetan pigs more than in Duroc pigs, Fig.3.
To study the adipocyte development in Tibetan and Duroc pigs, we extracted cell population by marker genes at different stages of adipose development. CD34, CD44, CD142 and ADIPOQ were used to identify MSC, ASC, preadipocytes, mature adipocytes, respectively. Therefore, we found that ADIPOQ+ cells numbers were greater whereas the CD34+ cells, CD44+ cells and CD142+ cells were fewer in Tibetan pigs than in Duroc pigs, Fig.3.
3.4 A greater proportion of stem cells differentiated into mature adipocytes in Tibetan than in Duroc pigs
To investigate the adipocyte differentiation process in Tibetan and Duroc pigs, we conducted a pseudo-time analysis Fig.4. In Tibetan pigs, the proliferation and ASC cluster served as the starting point for adipose differentiation. This cluster bifurcated into two branches, reaching Branch Point 1, where cells went into proliferation cells and preadipocytes. Subsequently, at Branch Point 3, some cells continued to remain in the proliferation and preadipocytes states, while others progressed through Branch Points 2 and 4 to go to mature adipocytes. Additionally, some cells persisted in the preadipocytes state Fig.4.
In Duroc pigs, the proliferation and ASC cluster also as the starting point for adipose differentiation, differentiated into three directions at Branch Point 2. Some cells went to proliferation and preadipocytes, while others become preadipocytes, with a minority eventually differentiated into mature adipocytes. In addition, some cells went into preadipocytes at Branch Point 3, and some cells maintained at ASC state Fig.4.
3.5 Fatty acid transporters and thermogenesis pathways were upregulated during adipocyte differentiation in Tibetan pigs
In addition, we selected upregulated genes in Tibetan and Duroc pigs from all clusters by aggregated scRNA-seq analysis (Fig. S3) and performed GO analysis, Fig.4). In Tibetan pigs, compared to Duroc pigs, pathways related to actin cytoskeleton organization, VEGFA-VEGFR2 signaling, fatty acid transporters and thermogenesis were upregulated, Fig.4. In Duroc pigs, compared to Tibetan pigs, pathways related to the innate immune response, cell activation and regulation of defense response were upregulated, Fig.4.
Subsequently, we used CD34, CD44, CD142 and ADIPOQ to indicate different stages of adipogenic differentiation to study the process of fat deposition in Tibetan pigs. We selected CD34+ cells from the integrated data and considered them as representing the MSC state. We extracted the upregulated genes in Tibetan pigs for comparison with Duroc pigs and conducted GO analysis. In Tibetan pigs, we found upregulated pathways, including the negative regulation of growth, fatty acid transporters and thermogenesis, Fig.5.
Next, we selected CD44+ cells to represent the state of ASC. We extracted the upregulated genes in Tibetan pigs for comparison with Duroc pigs and performed GO analysis. In Tibetan pigs, we observed upregulated pathways including fatty acid transporters, regulation of actin filament-based processes, thermogenesis and negative regulation of cell population proliferation, Fig.5. We extracted the upregulated genes in Tibetan pigs from CD142+ cells as the state of preadipocytes and performed GO analysis. Compared to Duroc pigs, we observed that pathways related to fat development, such as fatty acid transporters, thermogenesis and positive regulation of lipid localization, were upregulated in Tibetan pigs, Fig.5.
Finally, we selected upregulated genes in Tibetan pigs from ADIPOQ+ cells as the state of mature adipocytes and performed GO analysis. In Tibetan pigs, compared to Duroc pigs, we observed enhanced pathways including fatty acid transport, regulation of plasma membrane-bound cell projection organization and VEGFA-VEGFR2 signaling, Fig.5.
These findings indicate that the adipose SVF cells of Tibetan pigs possessed greater adipogenic potential and had enhanced thermogenic capacity compared to those of Duroc pigs.
3.6 Tibetan pig SVF cells had higher adipogenic potential and fatty acid transport ability compared to Duroc pig SVF cells
To further validate the predictions mentioned above, we first assessed the proliferative capacity of SVF cells derived from Tibetan and Duroc pigs. We cultured SVF cells and detected the proliferation at 24, 48 and 72 h. This revealed that Duroc SVF cells had significantly higher proliferation capacity at 24, 48 and 72 h compared to Tibetan pig SVF cells, Fig.6. However,
MKI67 and
CENPF were not significantly different in expression at these time points between Tibetan pig and Duroc SVF cells Fig. S4(a,b).
MKI67 and
CENPF were have been to mark proliferation cells in mice
[38]. Then, we selected
TUBB6, a significantly expressed gene in proliferation cell clusters in scRNA-seq, and found a significant higher in
TUBB6 expression in Duroc compared to Tibetan pig SVF cells at 24 and 72 h of proliferation, Fig.6.
In addition, we assessed the adipogenic potential and metabolic capacity of ASC from Tibetan and Duroc pigs. The gene CEBPA, an early regulator of adipocyte differentiation, had higher expression in Tibetan pig SVF cells compared to Duroc pig SVF cells, Fig.6. Additionally, genes related to fatty acid transport, such as FABP4 and CD36, were significantly more expressed in Tibetan pig SVF cells than in Duroc pig SVF cells, Fig.6.
Consistent with earlier findings that mitochondrial gene expression was higher in Duroc neck fat tissue, Fig.1, we also observed higher COX5B expression in Duroc SVF cells compared to Tibetan pig SVF cells, Fig.6.
The results suggested that SVF cells from Tibetan pigs had significantly higher adipogenic potential and fatty acid transport capability, while exhibiting lower proliferation and metabolic capacity compared to Duroc SVF cells.
3.7 Tibetan pigs have higher adipogenic capacity and lower metabolic activity than Duroc pigs at the mature adipocytes state
To further investigate the adipogenic process in Tibetan and Duroc pigs, we differentiated SVF cells into mature adipocytes by adipogenesis medium, and allowed them to differentiation until day 8. On day 8, oil red O staining revealed the presence of differentiated adipocytes in both Tibetan and Duroc cells, characterized by numerous lipid droplets. We quantified the lipid droplet content and found that mature adipocytes in Tibetan pigs had significantly higher lipid droplet content than in Duroc pigs, Fig.7. Subsequently, we detected the expression of adipogenesis-related genes and observed that CEBPA, CEBPB and ADIPOQ were significantly higher in Tibetan mature adipocytes and CD142 expression was lower than in Duroc pigs, Fig.7. Similarly, we identified fatty acid transport genes and found that, compared to Duroc mature adipocytes, the expression of FABP4 and CD36 were significantly higher in mature adipocytes of Tibetan pigs, Fig.7. It is suggested that Tibetan mature adipocytes provide stronger adipogenic and fatty acid transport capacity than Duroc pigs.
In addition, we evaluated cellular metabolic capability. It was observed that mature adipocytes of Tibetan pigs had lower expression of cytochrome complex genes CYCS and COX5B than in Duroc pigs (Fig.7). We further assessed energy metabolism in mature adipocytes by Seahorse experiments. While mature adipocytes of Tibetan and Duroc pigs did not differ in basal oxygen consumption rates and maximal mitochondrial oxygen consumption capacity Fig. S3(d), the mature adipocytes spare respiratory capacity of Tibetan pigs was lower than in Duroc pigs, Fig.7.
4 Discussion
In this study, despite no morphological differences evident in white adipocytes between Tibetan and Duroc pigs, we found a stronger fat deposition capacity, with notably higher expression of adipogenesis genes. including
CEBPA,
CEBPB and
LEPTIN, in Tibetan pigs. The evolutionary trajectories of Tibetan and Duroc pigs have led to distinct physiologic characteristics and adaptations. Tibetan pigs have gradually evolved into a breed emphasizing fat deposition, while Duroc pigs have been breed for lean meat production for commercial purposes. Previous research indicated that Tibetan pigs have thicker backfat and higher intramuscular fat content compared to Duroc pigs
[19,
22]. Additionally, Tibetan pigs have significantly higher expression of fat deposition-related genes such as
ADIPOQ and
LEPTIN[40].
In addition, it was found that Tibetan pigshad higher expression of
UCP3, although their metabolism gene expression in adipose tissue was significantly lower compared to Duroc pigs. The absence of brown adipose tissue and the
UCP1 thermogenic gene in pigs leads to lower thermogenic capability
[41]. Previous studies found cold-resistant pig breeds can activate non-shivering thermogenesis pathways through
UCP3 expression in cold environments, promoting adipocyte browning and aiding adaptation to harsh conditions
[42]. In addition, pigs also depend on shivering thermogenesis and physical measures, including clustering and optimizing husbandry conditions, to maintain body temperature in cold stress scenarios
[43]. Previous study have shown that high-altitude animals and humans exhibit reduced oxygen sensitivity and decreased oxygen demand as adaptive responses to hypoxic environments
[44]. These results further confirm Tibetan pigs possessed robust high-elevation adaptation.
However, the adipocyte development and metabolism regulation in fat- and lean-type pigs remain unclear. Therefore, we performed scRNA-seq analysis on the SVF cells of WAT from 4-day-old Tibetan and Duroc pigs to investigate heterogeneity of SVF cells and the adipocyte differentiation process in these pigs. We found in addition to clusters associated with adipocyte development, including mesenchymal stem cells (MSC), adipose stem cells (ASC), preadipocytes, proliferation cells and mature adipocytes, there are about 10% less endothelial cells and fewer myoblasts, VEC, vascular smooth muscle cells, and macrophages in adipose tissue. As previously described, adipose tissue is composed of many types of cells, but adipose cells predominant. Consistent with the results of mouse white adipose SVF cells, there are different types of cells in the SVF cells of porcine white adipose tissue, but the types of immune cells were fewer than in mice
[38].
In addition, compared to Duroc pigs, mature adipocyte proportions were found to be significantly higher in Tibetan pigs. The pseudo-time analysis indicated that Tibetan pig adipocyte stem cells differentiated into mature adipocytes during the differentiation process, with fewer cells remaining in the preadipocytes state. In contrast, in Duroc pigs, the majority of cells remained in the preadipocytes state without differentiation into mature adipocyte. Adipocytes undergo three main processes during differentiation. First, the MSC that have the potential to differentiate into various cell types and perhaps committed differentiation into preadipocytes. Subsequent proliferation and differentiation of preadipocytes, regulated by transcription factors such as
PPARG,
CEBPA and
CEBPB, led to the formation of mature adipocytes. Lastly, some mature adipocytes re-entered the cell cycle, replenishing new cells to continue growing in adipose tissue
[40,
45,
46]. It is suggested that Tibetan pig adipocyte stem cells once differentiated into preadipocytes more easily and more rapidly differentiated into mature adipocytes than in Duroc pigs. Additionally, the higher expression of adipogenic genes in Tibetan pig adipose tissue further indicates the higher adipogenic capacity of Tibetan pig preadipocytes to contribute to the lipid accumulation.
Subsequent GO analysis revealed that, compared to pathways upregulated in Duroc pigs, which mainly involved immune response processes such as innate immune response and oxidative stress response, Tibetan pigs enhanced pathways related to fatty acid transporters and fatty acid formation. In MSC, ASC and preadipocytes states, Tibetan pigs also had an upregulated thermogenic pathway, consistent with the high-elevation adaptive phenotype and cold resistance properties of these pigs. In Tibetan pigs, fatty acid oxidation consumes fuel for glucose oxidation and glycolysis processes, thereby reducing the demand for oxygen. Simultaneously, they accumulated more unsaturated fatty acids, which provides greater cold tolerance and is thus an adaptation for the harsh environment of high-elevation coldness
[26,
44]. In addition, the enhanced positive regulation of the lipid localization pathway at the preadipocytes stage further confirmed that Tibetan pig preadipocytes more readily progress toward maturity.
ASC in SVF cells from pig adipose tissue can differentiate into preadipocytes to ultimately become mature adipocytes
in vitro[47]. In this study, we cultured white adipose SVF cells for 3 days and observed that the adipogenic ability and fatty acid transport capacity were significantly higher in Tibetan pigs than in Duroc pigs. This revealed the activation of the fatty acid transport pathway during the differentiation process of Tibetan pig ASC. However, SVF cells in Duroc pigs had higher proliferative capacity at 24, 48 and 72 h of culture. In the analysis of proliferative gene expression,
TUBB6 was expressed substantively more in the proliferation cluster of SVF cells in Duroc pigs than in Tibetan pigs. In contrast, the expression of proliferative markers
MKI67 and
CENPF did not differ between Duroc and Tibetan pigs
[38,
48]. These results indicate that
TUBB6 can serve as an indicator of proliferation in pig SVF cells. Similarly, the metabolic capacity of ASC in Duroc pigs remained higher than in Tibetan pigs, indicating that Tibetan pigs have lower respiratory metabolic capacity at the adipose stem cell stage.
At the mature adipocyte stage (on day 8), Tibetan pigs also had higher adipogenic capabilities than Duroc pigs. Additionally, in comparison to the SVF stage, Tibetan pigs had greater abilities in fat formation and fatty acid transport than Duroc pigs. Consistent with earlier pseudo-time analysis results, Tibetan pig preadipocytes more readily progressed toward mature adipocytes. Additionally, the expression of
CD142, a marker gene for preadipocytes
[38], was significantly higher in Duroc pigs than in Tibetan pigs, and was consistent with the proportion of preadipocytes in Duroc pigs being higher than in Tibetan pigs as revealed by scRNA-seq analysis. At the mature adipocyte stage, the metabolic capacity and oxygen consumption rate of Tibetan pig cells remained lower than those of Duroc pigs. This indicates that Tibetan pig adipocytes have lower oxygen demand during the initial stages of differentiation, while maintaining consistently higher adipogenic and fatty acid transport capability than Duroc pigs. This highlights the enduring high-elevation adaptability of Tibetan pig during the differentiation process of ASC differentiation to mature adipocytes.
This study primarily focused on the developmental characteristics and cellular heterogeneity of adipose tissue in piglets. However, a comprehensive understanding of the epigenetic regulatory mechanisms underlying adipocyte development is still lacking. Also, there is a lack of direct evidence linking SVF cells differentiation into adipocyte formation and fat deposition in adult pigs. In future studies, it will be important to undertake single-nucleus ATAC-seq (snATAC-seq) analysis on adipose tissues from Tibetan and Duroc pigs to elucidate the epigenetic regulatory landscape of adipogenesis. In addition, there would be merit in examining the in vivo relationship between SVF cells fate determination and fat deposition in adult pigs.
5 Conclusions
In the present study, we observed that Tibetan pigs had greater adipogenic and fatty acid transport capabilities, but with a lower metabolic rate than Duroc pigs in vitro and invivo. Our scRNA-seq results revealed the heterogeneity of subcutaneous adipose tissue SVF cells in the high-elevation fat deposition pig breed compared to the lean meat pig breed. Our study holds significant implications for guiding efforts to enhance the fat content and backfat thickness in agricultural animals.
The Author(s) 2025. Published by Higher Education Press. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0)
PDF
(3544KB)
