Interference with DGAT Gene Inhibited TAG Accumulation and Lipid Droplet Synthesis in Bovine Preadipocytes

Simple Summary Triacylglycerol (TAG) is the primary component of intramuscular fat, an important factor in determining meat quality. The synthesis of TAG is regulated by diacylglycerol o-acyltransferase (DGAT). This study investigated the regulatory mechanisms of two subtypes of DGAT, namely DGAT1 and DGAT2, in the differentiation of Yanbian bovine preadipocytes and their roles in lipid metabolism-related signaling pathways. sh-DGAT1 and sh-DGAT2 were prepared using short interfering RNA (siRNA) interference technique targeting DGAT1 and DGAT2 genes and infected with bovine preadipocytes separately or simultaneously. Differentially expressed genes (DEGs) in bovine preadipocytes were analyzed using RNA sequencing and genome databases. During bovine preadipocytes differentiation, interference with DGAT1 and DGAT2 inhibited the formation of lipid droplets, content, and expression of lipid-forming genes at the mRNA level. DGAT2 showed a stronger inhibitory effect. Transcriptome analysis revealed 2070 and 2242 DEGs between pre-adipocytes and normal cells that inhibited DGAT1 and DGAT2 expression, respectively, and 2446 DEGs when both were simultaneously inhibited. Our results indicate that compared with DGAT1, DGAT2 plays a more important role in regulating bovine fat metabolism, which provides a theoretical basis for producing high-quality marbled beef. Abstract Triacylglycerol (TGA) is the primary component of intramuscular fat. Expression of diacylglyceryl transferase (DGAT) determines the polyester differentiation ability of precursor adipocytes. The two DGAT isoforms (DGAT1 and DGAT2) play different roles in TAG metabolism. This study investigates the roles of DGAT1 and DGAT2 in signaling pathways related to differentiation and lipid metabolism in Yanbian bovine preadipocytes. sh-DGAT1 (sh-1), sh-DGAT2 (sh-2), and sh-DGAT1 + sh-DGAT2 (sh-1 + 2) were prepared using short interfering RNA (siRNA) interference technique targeting DGAT1 and DGAT2 genes and infected bovine preadipocytes. Molecular and transcriptomic techniques, including differentially expressed genes (DEGs) and Kyoto Encyclopaedia of Genes and Genomes (KEGG) pathway analysis, were used to investigate the effects on the differentiation of Yanbian bovine preadipocytes. After interference with DGAT1 and DGAT2 genes, the contents of TAG and adiponectin were decreased. The TAG content in the sh-2 and sh-1 + 2 groups was significantly lower than that in the sh-NC group. RNA sequencing (RNA-seq) results showed 2070, 2242, and 2446 DEGs in the sh-1, sh-2, and sh-1 + 2 groups, respectively. The DEGs of the sh-2 group were mainly concentrated in the PPAR, AMPK, and Wnt signaling pathways associated with adipocyte proliferation and differentiation. These results demonstrated that at the mRNA level, DGAT2 plays a more important role in lipid metabolism than DGAT1.


Introduction
The beef cattle industry plays an important role in China's livestock industry and economy. Producing premium beef with rich intramuscular fat deposits, such as "Snowflake beef", can enhance the profitability and prospects of the dairy-fattening business. The aim of any fattening operation is to produce meat with excellent quality characteristics, including tenderness, shear strength, and marbling [1]. These quality traits are associated with the intramuscular fat content of beef [2], indicating the importance of intramuscular fat content in determining meat quality. Intramuscular fat is mainly located in skeletal muscle fibers and stored as lipid droplets (LD) primarily composed of triglycerides (TAGs) and cholesterol esters [3].
TAG is the main form of energy storage in animals, and fat deposition in beef cattle muscle is mainly due to TAG accumulation. TAG is synthesized in vivo by two pathways: The glycerophosphate pathway (Kennedy pathway) in most cells and the glycerophosphoryl pathway in specialized cells [4]. These two pathways catalyze the synthesis of diacylglycerol and fatty acid acyl via the action of the microsomal enzyme diacylglycerol o-acyltransferase (DGAT) [5]. Studies have shown that DGAT is not only the last reactive enzyme controlling TAG synthesis in adipocytes but also the only key and rate-limiting enzyme in TAG synthesis [6]. DGAT comprises two subtypes, and the genes encoding these two enzymes are DGAT1 and DGAT2. DGAT1 belongs to the acyl-coA cholesterol acyl transferase (ACAT) gene family, while DGAT2 to the acyl-coa monoacylglyceryl transferase (MGAT1) gene family, both of which have significantly different membrane topologies [7] and are widely expressed in various mammalian tissues where they play different roles [8].
Harris et al. [9] studied the effects of DGAT1 and DGAT2 genes on TAG synthesis and LD formation in mouse adipose cells and found that knocking out one gene in DGAT1 and DGAT2 does not interfere with TAG synthesis or LD formation in adipose cells. Conversely, when the two genes were simultaneously knocked out, no TAG was synthesized in the fat cells, and no LD were formed inside the cells. This study further demonstrates that DGAT1 and DGAT2 are closely associated with TAG synthesis, lipid deposition, and LD formation.
Given that enzymes in the DGAT1 and DGAT2 families have several different lipid acyltransferase activities [10], and since the functional activities of several of these enzymes are unknown, whether TAG is produced by enzymes other than DGAT1 and DGAT2 and the molecular mechanism by which DGAT regulates the differentiation and lipid metabolism of bovine preadipocytes remain unclear. We hypothesize that DGAT2 plays an important role in TAG synthesis and storage and in regulating the differentiation of bovine preadipocytes.
In this study, sh-DGAT1 (sh-1), sh-DGAT2 (sh-2) and sh-DGAT1+ sh-DGAT2 (sh-1 + 2) were generated via small interfering RNA (siRNA)interference targeting the DGAT1 and DGAT2 genes in infected bovine preadipose cells. Molecular and transcriptomic techniques were also used to investigate the role of DGAT in lipid metabolism and preadipocyte differentiation in cattle and to provide a theoretical basis for producing high-quality marbled beef.
When the cell density reached more than 80%, lipid-induced differentiation was performed with a low serum concentration, and cell growth was observed and recorded on day 8 of culture.

Detection of DGAT-siRNA Interference Effect
To determine a better interference effect, three DGAT1-siRNA and DGAT2-siRNA from the two genes were screened and evaluated. When the cell growth density reached >80%, cell infection was performed, and three replicates were used for each group. Cells were collected 48 h later, RNA was extracted, and infection efficiency was detected using fluorescence quantitative PCR to screen for the siRNA with the best interference effect for subsequent tests.

RNA Extraction and Quantitative Real-Time PCR (qRT-PCR) Detection
Total RNA was extracted from bovine preadipocytes using TRIzol™ reagent (Thermo Fisher Scientific, Waltham, MA, USA), and RNA integrity was resolved using NanonodropnD-100 spectrophotometer (2000C, Thermo Fisher Scientific, Waltham, MA, USA) and 1% agarose gel electrophoresis. A complementary DNA (cDNA) template was synthesized using the FastKing gDNA Dispelling RT SuperMix Kit (Tiangen Biotech, Beijing, China). The qRT-PCR was performed on an Agilent Mx3000/5p Real-Time PCR Detection System (Agilent Technologies, Santa Clara, CA, USA), using 20 µL SYBR Green SuperReal PreMix Plus (Tiangen Biotech, Beijing, China) containing 10 µL SYBR, 0.6 µL up-and downstream primers (10 µmol/L), 1 µL cDNA template, 7.5 µL Rnase-free ddH 2 O, and 0.3 µL Rox. PCR was performed as follows: Pre-denaturation at 95 • C for 15 min. There were 35 cycles of denaturation at 95 • C for 10 s, annealing at 55 • C for 30 s, and extension at 72 • C for 32 s. The dissolution stage was 95 • C for 15 s and 65 • C for 5 s. The relative expression of the target gene was normalized to that of glyceraldehyde-3-phosphate dehydrogenase (GAPDH; internal control) and calculated using the 2 − Ct method [12]. Primers for lipid metabolism-and phosphoglycerol pathway-related genes were designed online through primer premier 6.25 and synthesized by Shenggong Bioengineering Co., Ltd. (Shanghai, China), and their sequences are listed in Table 2. The triglyceride content was determined using an unkude triglyceride determination kit (Applygen Technologies, Beijing, China). The cells were washed three times with PBS and lysed in lysis buffer (R0010; Solarbio, Beijing, China). TAG content was determined via enzymatic colorimetry at 570 nm using a microplate reader (iMark; Bio-Rad, Hercules, CA, USA).

Determination of Aiponectin (ADP) Concentration
After 96 h of induction and differentiation, the cell culture medium was collected and analyzed using a bovine adiponectin ELISA Kit (Mlbio, Shanghai, China). A microplate reader (iMark; Bio-Rad, Hercules, CA, USA) was used to measure the optical density at 450 nm and generate a standard curve to determine adiponectin concentration.

RNA Sequencing (RNA-Seq)
Bovine preadipocytes were infected with sh-NC, sh-DGAT1, sh-DGAT2, and sh-DGAT1 + sh-DGAT2, with three replicates per group. Cell samples were collected 96 h after differentiation, and RNA was extracted using 1 mL of TRIzol reagent. Shanghai Parsonol Biotechnology Co. Ltd. (Shanghai, China) was used for RNA purification, cDNA library construction, and sequencing. Standard library preparation was sampled with an RIN value > 7.
The samples were sequenced by Illumina ® NovaSeq 6000 (2 × 150 bp ultra-long read length sequence has a better splicing effect), and the image file was obtained, which was converted by the software of the sequencing platform to generate the original FASTQ data, and each sample was analyzed separately. The sample name was Q30 with fuzzy base percentage, Q20 (%) and Q30 (%). There were some low-quality reads containing joints in the sequencing data, which might have interfered with the subsequent information analysis. Therefore, FastQC v0.11.8 was used to check the quality of the disembarkation data.
Reference Genome and Gene Model Annotation files (GTF files) were downloaded directly from the Genome website. Using HISAT2 v2.0.5 (http://ccb.jhu.edu/software/ hisat2/index.shtml (accessed on 18 December 2021)) to construct the reference index in the genome, the paired-end clean reads were compared with the reference genome using HISAT2 v2.0.5. We used gene coverage parity and saturation analyses to assess the sequencing quality for the volume of the sequencing data. Under ideal conditions, the read distribution of all expressed genes should be uniform. We used RSeQC analysis of expression saturation to assess whether the amount of data measured was sufficient to correctly calculate gene expression levels. HTSeq (0.9.1) was used to calculate and compare the Read Count value of each gene with the original gene expression level. Non-stromalization of expression was performed using FPKM, and genes with FPKM >1 were generally considered to be expressed in the reference transcriptome. We used the DESeq software (version 1.20.0) to perform differential expression analysis between the two comparison combinations. For the gene expression via DESeq variance analysis, screening of differentially expressed gene conditions was as follows: Multiple expression differences |log2FoldChange| > 1, significant p-value < 0.05. The R language ggplots2 software package was used to map the volcano plots of the differentially expressed genes.

Gene Ontology (GO) and Kyoto Encyclopaedia of Genes and Genomes (KEGG) Enrichment Analysis
TopGO was used for GO enrichment analysis, and the hypergeometric distribution method was used to calculate the p-value (the standard of significant enrichment was p-value < 0.05) to determine the GO term of significant enrichment of differential genes so as to determine the main biological functions of differential genes. ClusterProfiler (3.4.4) software was used for the enrichment analysis of KEGG pathways, focusing on significantly enriched pathways with a p-value < 0.05.

Statistical Analyses
Data analysis and graph generation were performed using GraphPad Prism 6.07 (GraphPad Software, La Jolla, CA, USA) and SPSS Statistical software v19.0 (IBM, Armonk, NY, USA). The unpaired t-test was used for calculating p values. The results are presented as the mean ± standard error of the mean (SEM) from experiments performed in triplicate. Differences were considered statistically significant at p < 0.05.

Interfering DGAT Gene Inhibited the Differentiation of Bovine Adipose Cells Induced by Oleic Acid
Prior to the experiment, we first cultured and identified the extracted preconditioned adipocytes in vitro. The results showed that the extracted precursor adipocytes had general cell growth characteristics (Supplementary Figure S1a,b), and the expression patterns of precursor adipocyte marker genes (Supplementary Figure S1d) and adipogenic marker genes (Supplementary Figure S1e) were consistent with the growth characteristics of precursor adipocytes, which could be used for subsequent experiments.
We performed sequential expression of DGAT1 and DGAT2 and found similar expression patterns between DGAT1 and PPARγ and C/EBPα, that is, the expression levels of DGAT1 gradually increased with the extension of differentiation time. However, the expression pattern of DGAT2 was the opposite of that of DGAT1 (Supplementary Figure S1e).
In the absence of oleic acid, only a small number of lipid droplets were formed in bovine precursor adipose cells on day 8 of differentiation (Supplementary Figure S1c). When 100 uL oleic acid was added, the accumulation of triglycerides and adiponectin significantly increased for 96 h (Supplementary Figure S2b), and the expression of adipose marker genes and related genes in the triglyceride synthesis pathway significantly increased (Supplementary Figure S2c,d). Therefore, the amount of oleic acid added was 100 uL.
The siRNA-mediated interference effects on DGAT1 and DGAT2 are shown in Figure 1A. The interference efficiency of siRNA-DGAT1-1157 with DGAT1 was 77.5% compared with that of the negative control. Therefore, siRNA-DGAT1-1157 was selected for subsequent infection testing and was named sh-1. siRNA-DGAT2-108 of DGAT2 had the highest interference efficiency. Therefore, siRNA-DGAT2-108 was selected for subsequent infection tests and named sh-2. The experimental group coinfected with these two genes was named sh-1 + 2.
software was used for the enrichment analysis of KEGG pathways, focusing on significantly enriched pathways with a p-value < 0.05.

Statistical Analyses
Data analysis and graph generation were performed using GraphPad Prism 6.07 (GraphPad Software, La Jolla, CA, USA) and SPSS Statistical software v19.0 (IBM, Armonk, NY, USA). The unpaired t-test was used for calculating p values. The results are presented as the mean ± standard error of the mean (SEM) from experiments performed in triplicate. Differences were considered statistically significant at p < 0.05.

Interfering DGAT Gene Inhibited the Differentiation of Bovine Adipose Cells Induced by Oleic Acid
Prior to the experiment, we first cultured and identified the extracted preconditioned adipocytes in vitro. The results showed that the extracted precursor adipocytes had general cell growth characteristics (Supplementary Figure S1a,b), and the expression patterns of precursor adipocyte marker genes (Supplementary Figure S1d) and adipogenic marker genes (Supplementary Figure S1e) were consistent with the growth characteristics of precursor adipocytes, which could be used for subsequent experiments.
We performed sequential expression of DGAT1 and DGAT2 and found similar expression patterns between DGAT1 and PPARγ and C/EBPα, that is, the expression levels of DGAT1 gradually increased with the extension of differentiation time. However, the expression pattern of DGAT2 was the opposite of that of DGAT1 (Supplementary Figure  S1e).
In the absence of oleic acid, only a small number of lipid droplets were formed in bovine precursor adipose cells on day 8 of differentiation (Supplementary Figure S1c). When 100 uL oleic acid was added, the accumulation of triglycerides and adiponectin significantly increased for 96 h (Supplementary Figure S2b), and the expression of adipose marker genes and related genes in the triglyceride synthesis pathway significantly increased (Supplementary Figure S2c,d). Therefore, the amount of oleic acid added was 100 uL.
The siRNA-mediated interference effects on DGAT1 and DGAT2 are shown in Figure  1A. The interference efficiency of siRNA-DGAT1-1157 with DGAT1 was 77.5% compared with that of the negative control. Therefore, siRNA-DGAT1-1157 was selected for subsequent infection testing and was named sh-1. siRNA-DGAT2-108 of DGAT2 had the highest interference efficiency. Therefore, siRNA-DGAT2-108 was selected for subsequent infection tests and named sh-2. The experimental group coinfected with these two genes was named sh-1 + 2. As shown in Figure 1B, different amounts and sizes of lipid droplets were formed in each experimental group due to the addition of oleic acid. The number and size of the lipid droplets in the sh-1 + 2 group were significantly lower than those in the other experimental groups. Meanwhile, compared with the control group, the contents of TAG and ADP in the sh-2 and sh-1 + 2 groups were significantly lower (p < 0.05) ( Figure 1C).
The effects of DGAT1 and DGAT2 knockdown on the expression of genes related to the phosphoglycerol pathway synthesized by TAG are shown in Figure 1D. The expression of DGAT1 in the sh-1 group was significantly lower than that in the other experimental groups (p < 0.05). However, there was no significant difference between the sh-2 As shown in Figure 1B, different amounts and sizes of lipid droplets were formed in each experimental group due to the addition of oleic acid. The number and size of the lipid droplets in the sh-1 + 2 group were significantly lower than those in the other experimental groups. Meanwhile, compared with the control group, the contents of TAG and ADP in the sh-2 and sh-1 + 2 groups were significantly lower (p < 0.05) ( Figure 1C).
The effects of DGAT1 and DGAT2 knockdown on the expression of genes related to the phosphoglycerol pathway synthesized by TAG are shown in Figure 1D. The expression of DGAT1 in the sh-1 group was significantly lower than that in the other experimental groups (p < 0.05). However, there was no significant difference between the sh-2 and sh-1 + 2 groups (p > 0.05). The expression level of DGAT2 in the sh-1 test group was the lowest. There was no significant difference between the sh-1 group and sh-2 group (p > 0.05), but there was between the sh-1 test group and the other test groups (p < 0.05). AGPAT4 gene expression in the sh-NC and sh-2 groups was significantly lower than that in the other experimental groups (p < 0.05), and that in the sh-1 group was significantly higher than in the sh-2 and sh-1 + 2 groups (p < 0.05). LIPIN1 gene expression in the sh-1 group was significantly lower than that in the other experimental groups (p < 0.05), but there was no significant difference between the sh-2 group and sh-NC groups (p > 0.05), while that in the sh-1 + 2 group was significantly higher than in the other experimental groups (p < 0.05). The expression of MGAT1 and GPAT4 genes in the sh-1 + 2 group was higher than that in the sh-NC group. However, the difference was not statistically significant (p > 0.05). The expression levels of MGAT1 and GPAT4 in the sh-1 and sh-2 experimental groups were the lowest. However, there were no significant differences between the two groups (p > 0.05).
The effects of DGAT1 and DGAT2 on the expression of lipid metabolism-related genes are shown in Figure 1E. The PPARγ gene expression in the sh-1 group was lower than that in the sh-2 and sh-1 + 2 groups, while the C/EBP-α gene expression was significantly higher than that in the sh-2 and sh-1 + 2 groups (p < 0.05). The SCD gene, sh-2 and sh-1 + 2 groups were significantly lower than the other experimental groups (p < 0.05). The expression level of FABP4 in the sh-1 + 2 group was significantly higher than that in the other groups (p < 0.05), and the expression level in the sh-2 group was the lowest and significantly lower than that in the other groups (p < 0.05). The expression of PLIN2 in the sh-1 group was significantly lower than that in the other experimental groups (p < 0.05), while the expression of PLIN2 in the sh-2 and sh-1 + 2 groups was not significantly different (p > 0.05) but was significantly lower than that in the negative control group (p < 0.05). The expression of CD36 in the sh-2 and sh-1 + 2 groups was significantly higher than that in the other groups (p < 0.05).

Difference Analysis of Bovine Preadipocytes Infected with sh-DGAT1/sh-DGAT2
Results from total RNA integrity tests showed that the quality of the extracted RNA was consistent with the requirements of the sequencing experiment for library construction (Figure 2A). Quality evaluation of the sequencing data showed that the proportion of high-quality clean reads was >91.00% in all the groups (Table S1). The libraries were aligned against the Bos taurus (https://ftp.ensembl.org/pub/release-86/gtf/bos_taurus/ Bos_taurus.UMD3.1.86.gtf.gz (accessed on 18 December 2021)) genome (Table S2). The uniform-level gene coverage results showed no obvious bias toward the front ( Figure 2B), and log10(FPKM + 1) showed a normal distribution ( Figure 2C).
A total of 2070 DEGs were screened in the sh-1 group, including 1214 upregulated and 856 downregulated genes ( Figure 3A). A total of 2242 DEGs were screened in the sh-2 group, of which 1255 were upregulated and 987 were downregulated ( Figure 3B). In the sh-1 + 2 group, the total number of DEGs was 2446, of which 1317 were upregulated and 1129 were downregulated ( Figure 3C).
was consistent with the requirements of the sequencing experiment for library construction (Figure 2A). Quality evaluation of the sequencing data showed that the proportion of high-quality clean reads was >91.00% in all the groups (Table S1). The libraries were aligned against the Bos taurus (https://ftp.ensembl.org/pub/release-86/gtf/bos_taurus/Bos_taurus.UMD3.1.86.gtf.gz (accessed on 18 December 2021)) genome (Table S2). The uniform-level gene coverage results showed no obvious bias toward the front ( Figure  2B), and log10(FPKM + 1) showed a normal distribution ( Figure 2C). A total of 2070 DEGs were screened in the sh-1 group, including 1214 upregulated and 856 downregulated genes ( Figure 3A). A total of 2242 DEGs were screened in the sh-2 group, of which 1255 were upregulated and 987 were downregulated ( Figure 3B). In the sh-1 + 2 group, the total number of DEGs was 2446, of which 1317 were upregulated and 1129 were downregulated ( Figure 3C).

Functional Analysis of Differentially Expressed Genes GO
The biological functions of DEGs were elucidated using GO enrichment analysis, and the 20 most significantly enriched GO terms were ranked according to the significance level. As shown in Figure 4, the biological processes (BP) of each treatment group accounted for most single gene annotations, followed by cell components (CC), and none of them were enriched in molecular function (MF). BP mainly includes multicellular organismal development and regulation of multicellular organismal processes, cellular developmental processes, cell differentiation, and cell surface receptor signaling pathways. CC enrichment mainly included the extracellular matrix, extracellular regions, and extracellular regions.

Functional Analysis of Differentially Expressed Genes GO
The biological functions of DEGs were elucidated using GO enrichment analysis, and the 20 most significantly enriched GO terms were ranked according to the significance level. As shown in Figure 4, the biological processes (BP) of each treatment group accounted for most single gene annotations, followed by cell components (CC), and none of them were enriched in molecular function (MF). BP mainly includes multicellular organismal development and regulation of multicellular organismal processes, cellular developmental processes, cell differentiation, and cell surface receptor signaling pathways. CC enrichment mainly included the extracellular matrix, extracellular regions, and extracellular regions.

Functional Analysis of Differentially Expressed Genes GO
The biological functions of DEGs were elucidated using GO enrichment analysis, and the 20 most significantly enriched GO terms were ranked according to the significance level. As shown in Figure 4, the biological processes (BP) of each treatment group accounted for most single gene annotations, followed by cell components (CC), and none of them were enriched in molecular function (MF). BP mainly includes multicellular organismal development and regulation of multicellular organismal processes, cellular developmental processes, cell differentiation, and cell surface receptor signaling pathways. CC enrichment mainly included the extracellular matrix, extracellular regions, and extracellular regions.

KEGG Enrichment Analysis of DEGs
KEGG pathway enrichment analysis showed that the main enrichment pathways of DEGs in the experimental groups included ECM-receptor interaction and the PI3K-Akt, MAPK, TGF-beta, and Hippo signaling pathways ( Figure 5A-C).

KEGG Enrichment Analysis of DEGs
KEGG pathway enrichment analysis showed that the main enrichment pathways of DEGs in the experimental groups included ECM-receptor interaction and the PI3K-Akt, MAPK, TGF-beta, and Hippo signaling pathways ( Figure 5A-C).

Analysis of Signal Pathways Related to Differentially Expressed Genes and Lipid Metabolism
Transcriptome sequencing results showed that after DGAT2 knockdown, several differentially enriched genes were associated with fat deposition in the selected pathways. After interference with DGAT2, PPARγ-regulated target genes also changed in the PPAR signaling pathway, and matrix metalloproteinase (MMP1), long-chain lipoyl CoA synthetase (ACSL1), peroxidase acyl-CoA oxidase 2 antibody (ACOX2) were downregulated. FABP3, CD36 and oxidized low-density lipoprotein receptor 1 (OLR1) were upregulated ( Figure 6A). There were 22 differential genes in the Wnt signaling pathway, among which 16 genes were upregulated, including PLCB4, SFRP2, PRKCB, WNT2B, MYC, TCF7, DKK2, CCN4, and PLCB1, and six were downregulated genes, namely SFRP4, SFRP1, SERPINF1, WNT2, PRKCG, and PORCN ( Figure 6B). There were 12 differentially expressed genes in the AMPK signaling pathway, including seven upregulated genes and five downregulated genes. The upregulated genes were PIK3CD, CREB3L3, FASN, CD36, HMGCR, STRADB, and PPP2R2C, while those downregulated were FBP1, INSR, PFKFB1, CFTR, and PFKFB3 ( Figure 6C). Twelve DEGs were screened from the above signaling pathways for real-time PCR validation. The primer sequences are listed in Table S3, while the results of real-time quantitative fluorescence PCR are shown in Figure 6A-C. The q-PCR results of different genes were consistent with the expression trends of the RNA-seq results. Figure 5. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of differentially expressed genes (DEGs) in the sh-NC vs. sh-1 groups (A), sh-NC vs. sh-2 groups (B) and sh-NC vs. sh-1 + 2 groups (C).

Discussion
Short interfering RNA (siRNA) can bind to the target gene or messenger RNA of the target gene to inhibit expression with high specificity and efficiency [13]. TAGs are the most important form of energy storage in eukaryotic cells, most of which are stored in the LD of fat cells. As a key rate-limiting enzyme in TAG synthesis, DGAT plays an important role in lipid accumulation [14], and its expression level directly determines the polyester differentiation ability of fat precursor cells. DGAT includes two subtypes, DGAT1 and DGAT2, which, although they catalyze the same biochemical reaction and have similar broad lipid acyl-CoA substrate specificity, play completely different roles in mammalian TAG metabolism [15]. In this study, small interfering RNA technology targeting DGAT1 and DGAT2 genes was used to infect bovine preadipocytes separately or jointly with sh-DGAT1 and sh-DGAT2, respectively, to explore their effects on TAG and lipid droplet synthesis during lipid differentiation. The interference effect of DGAT1 was higher than 75%, and the interference effect of DGAT2 was as high as 70%, which met the requirements for subsequent tests.
Unsaturated fatty acids and C-carotenoids from cattle feed can increase the expression of PPARγ and are natural activators of PPARγ [16]. As an agonist of PPARγ, the exogenous addition of oleic acid affects the expression of DGAT. However, the effects of oleic acid on the DGAT1 and DGAT2 genes of large mammal cattle remain nebulous. Different concentrations of oleic acid could promote the accumulation of lipid droplets in adipocytes of foie gras, as well as the expressions of lipid metabolism-related genes DGAT1, DGAT2, PPARγ, and PLIN, but with higher concentrations, the effect of oleic acid was gradually decreased [17]. In this study, the expressions of DGAT1, DGAT2, PPARγ, CEBP/α, FABP4, and PLIN2 genes showed a significantly increasing trend with the increase of oleic acid concentration, and the higher the concentration, the greater the increase, which was contrary to previous studies [17], possibly due to the differences in species and cell sources.
The increased expression of DGAT2 is associated with the formation of large lipid droplets, whereas the overexpression of DGAT1 only produces small lipid droplets [18].
DGAT2 has been shown to regulate the accumulation of TAGs in the tissues of DGAT1-deficient mice [19]. Additionally, DGAT2 was highly expressed in various lipid-

Discussion
Short interfering RNA (siRNA) can bind to the target gene or messenger RNA of the target gene to inhibit expression with high specificity and efficiency [13]. TAGs are the most important form of energy storage in eukaryotic cells, most of which are stored in the LD of fat cells. As a key rate-limiting enzyme in TAG synthesis, DGAT plays an important role in lipid accumulation [14], and its expression level directly determines the polyester differentiation ability of fat precursor cells. DGAT includes two subtypes, DGAT1 and DGAT2, which, although they catalyze the same biochemical reaction and have similar broad lipid acyl-CoA substrate specificity, play completely different roles in mammalian TAG metabolism [15]. In this study, small interfering RNA technology targeting DGAT1 and DGAT2 genes was used to infect bovine preadipocytes separately or jointly with sh-DGAT1 and sh-DGAT2, respectively, to explore their effects on TAG and lipid droplet synthesis during lipid differentiation. The interference effect of DGAT1 was higher than 75%, and the interference effect of DGAT2 was as high as 70%, which met the requirements for subsequent tests.
Unsaturated fatty acids and C-carotenoids from cattle feed can increase the expression of PPARγ and are natural activators of PPARγ [16]. As an agonist of PPARγ, the exogenous addition of oleic acid affects the expression of DGAT. However, the effects of oleic acid on the DGAT1 and DGAT2 genes of large mammal cattle remain nebulous. Different concentrations of oleic acid could promote the accumulation of lipid droplets in adipocytes of foie gras, as well as the expressions of lipid metabolism-related genes DGAT1, DGAT2, PPARγ, and PLIN, but with higher concentrations, the effect of oleic acid was gradually decreased [17]. In this study, the expressions of DGAT1, DGAT2, PPARγ, CEBP/α, FABP4, and PLIN2 genes showed a significantly increasing trend with the increase of oleic acid concentration, and the higher the concentration, the greater the increase, which was contrary to previous studies [17], possibly due to the differences in species and cell sources.
The increased expression of DGAT2 is associated with the formation of large lipid droplets, whereas the overexpression of DGAT1 only produces small lipid droplets [18].
DGAT2 has been shown to regulate the accumulation of TAGs in the tissues of DGAT1-deficient mice [19]. Additionally, DGAT2 was highly expressed in various lipidmetabolizing tissues [20]. Moreover, DGAT2 can compensate for LD formation in DGAT1-deficient intestinal stem cells [21]. In the present study, DGAT2 overexpression increased the formation and accumulation of LD in Yanbian cattle preadipocytes, whereas DGAT2 knockdown inhibited LD formation in the cells, which was consistent with previous findings [22]. Several studies have shown that low expression of DGAT2 can cause a decrease in the TAG content of adipocytes [18].
DGAT2 overexpression increased the expression of lipid-forming genes and the accumulation of TAGs in skeletal muscle cells (BSCs) [14]. In this study, inhibition of DGAT2 expression was positively correlated with TAG and ADP content in preadipocytes. The importance of DGAT2 in TAG synthesis was further explained [18].
Transcriptome sequencing is an important method for exploring gene function [23]. The PPAR signaling pathway plays an important role in regulating various lipid activities such as lipogenesis, fatty acid transport, and adipocyte differentiation. In yaks, there is a significant negative correlation between FABP3 mRNA expression and MUFA level [24]. The transcriptome sequencing results of this study showed that the expression of FABP3 in the pathway regulating lipid metabolism in the PPAR signaling pathway was significantly upregulated following interference with DGAT2. If the mRNA expression level of FABP3 in Yanbian cattle is negatively correlated with MUFA levels, then the expression level of FABP3 will be downregulated after the addition of oleic acid. However, the expression of this gene was upregulated after interference with DGAT2, which suggests a negative regulatory effect between DGAT2 and FABP3. The regulation of the adipocyte differentiation pathway by MMP-1 was significantly downregulated. We further speculate that the DGAT2 gene plays an important role in lipid metabolism and differentiation regulation. The ACSL enzyme family is crucial to fatty acid metabolism in mammals and includes five members: ACSL1, ACSL3, ACSL4, ACSL5, and ACSL6 [25]. ACSL1, involved in the activation of TAG fatty acid synthesis [26], is found in the liver, heart, and fat cells and has a wide range of fatty acid specificities [27]. In mice, overexpression of cardiac ACSL1 increased the accumulation of TAG in cardiomyocytes by 12-fold [28]. Our sequencing results showed that following interference with DGAT2, ACSL1 was significantly downregulated in the PPAR signaling pathway, consistent with the results of previous studies, indicating that DGAT2 and ACSL1 are positively regulated.
AMPK, an important cellular energy sensor [29], is a key factor in controlling cellular energy homeostasis and metabolism [30] and can reduce the expression of SREBP-1, PPARγ, and C/EBp-α, thereby inhibiting the accumulation of fat during fat formation [31]. FASN is a key enzyme involved in the process of fatty acid regeneration and plays a crucial role in energy homeostasis by converting excess carbon intake into fatty acid storage [32]. CFTR is a Cl-channel in the apical membrane of epithelial cells regulated by cAMP and protein kinase a (PKA). AMPK acts as a "biorheostat" of CFTR, that is, activation of AMPK can inhibit CFPT [33]. In addition, the loss of functional expression of CFTR is thought to upregulate AMPK activity in cystic fibrosis (CF)-deficient epithelial cells [33]. Transcriptional sequencing results showed that following interference with DGAT2, FASN expression in the AMPK signaling pathway was upregulated, while CFPT expression was downregulated. This is possibly because, after DGAT2 expression is inhibited, intracellular energy is reduced, and AMPK activity is activated, which further promotes fatty acid synthesis, thus upregulating FASN expression and downregulating CFPT expression. This finding is consistent with those of the previous studies.
Wnt is involved in various important biological processes, including tissue regeneration, animal development, cell proliferation, and differentiation, and is a member of a conserved glycoprotein family. The typical Wnt/b-catenin signaling pathway is a highly conserved pathway critical to cell fate and patterns during development [34]. The catenin signaling pathway is regulated by a series of secreted molecules, including Wnt inhibitory signaling factor-1 (WIF1), Cerebrus, Sclerostin, Dickkopf-1 (DKK1) and secreted curl-associated proteins (SFRP). Sclerostin and DKK1 antagonize typical signals by binding to LRP5/6, whereas WIF1, cerebrus and SFRP2 interact directly with Wnt proteins [35]. In mammals, SFRPs comprise five protein families: Frzb (SFRP3), SFRP1, SFRP2, SFRP4, and SFRP5 [36]. Since SFRPs are homologous to the WNt-binding domain sequence of the Fz receptor, they are considered typical Wnt signaling antagonists that can bind to Wnt proteins and block signal transduction [37]. However, it has recently been reported that SFRPs synergistically mimic Wnt activity by interacting directly with Fz receptors [38], antagonizing each other's actions [39], enhancing the extracellular transport of Wnt proteins [40], or exerting other roles besides direct control of Wnt signaling pathways [41]. The sequencing results showed that SFRP2 and Wnt2 expression was upregulated following interference with DGAT2. Furthermore, the Wnt signaling pathway is closely related to the upstream gene process of lipogenic differentiation [42], which promotes the differentiation of mesenchymal stem cells into myoblasts or osteoblasts while inhibiting the differentiation of precursor adipocytes.
The identification of candidate genes is an important step in promoting marker breeding in Yanbian cattle. The results of this study confirm that DGAT1 and DGAT2 play important regulatory roles in adipocyte differentiation and lipid metabolism, providing insights into improving the bovine genome annotation and molecular breeding of beef cattle.

Conclusions
In this study, owing to the addition of OA, LD generation was observed after the DGAT1 and DGAT2 genes were disrupted, the contents of TAG and ADP were significantly reduced, and the expression of genes related to fat metabolism was inhibited at the mRNA level. RNA-seq was used to analyze the differentially expressed genes interfering with DGAT1 and DGAT2 in bovine preadipocytes, and 2070, 2242, and 2446 DEGs were detected in the sh-DGAT1 and sh-DGAT2 infected groups alone and co-infected groups, respectively. In the sh-DGAT2 treatment group, DEGs were enriched in the AMPK, TGF-β, and PPAR signaling pathways associated with adipocyte proliferation and differentiation, thus regulating the production of lipids by regulating the transcription of related genes in adipocytes.

Supplementary Materials:
The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/ani13132223/s1, Figure S1: Isolation and culture of bovine adipose cells and time sequence expression of DGAT gene; Figure S2: Differentiation of bovine preadipocytes induced by different concentrations of oleic acid; Table S1: Information of clean data; Table S2: Comparison of results and statistics; Table S3: Sequence of PCR primers used in this study.