SGK1 Affects the Phosphorylation of FOXO1/FOXO3 Promoting Bovine Fat Deposition via the PI3K/Akt Signaling Pathway

Background: Improving the intramuscular fat content of beef not only could improve our living standard, but decrease the the risk of the disease from the aspect of human nutrition and diet. Serum/glucocorticoid-inducible kinase 1 (SGK1) is a critical kinase involved in regulating multiple metabolic processes. Results: In the present study we reveal the molecular mechanism by which SGK1 affects the phosphorylation of FOXO1 and FOXO3 thus regulating bovine fat deposition via the PI3K/Akt signaling pathway. SGK1 was characterized primarily in the bovine kidney, where it peaked at day 6 of preadipocyte differentiation. Overexpression of SGK1 in bovine preadipocytes promoted mRNA expression of adipose-specic genes, while cell cycle-related genes were repressed. Adipocyte differentiation-related proteins were not signicantly changed, but proliferation-related proteins were markedly inhibited. Moreover, mRNA expression of adipose-specic genes were down-regulated while cell cycle-related genes were up-regulated in response to loss-of-function SGK1. The expression of adipose differentiation-related proteins were signicantly inhibited but proliferation-related proteins were not signicantly changed. Overexpression of SGK1 in bovine adipocytes signicantly changed the gene expression enriched in the PI3K/Akt signaling pathways through RNA-seq. Subsequently, we found that overexpression of SGK1 increased phosphorylation of FOXO1/FOXO3 and reduced protein expression, whereas SGK1 deletion had opposite effects. Overall, we show that overexpression of SGK1 promotes adipogenesis mainly via inhibiting preadipocyte proliferation, whereas the loss-of-function of SGK1 represses adipogenesis mainly through inhibition of adipocyte differentiation. Mechanistically, SGK1 changes the phosphorylation of two key genes located downstream in the PI3K/Akt signaling pathway, FOXO1/FOXO3, thus promoting fat deposition in cattle. Conclusion: These data reveal that SGK1 is a positively regulatory during intramuscular adipogenesis of cattle, which provides the security for human healthy from the aspect of high-quality beef.


Introduction
Intramuscular fat (IMF) content is favored by consumers as since it determines sensory characteristics and palatability [1], and also contains fatty acids and proteins that cannot be synthesized [2]. Adipose tissue helps in energy storage [3], and in maintaining homeostasis [4]. Adipogenesis grows by two mechanisms, hyperplasia and hypertrophy, and is regulated by other tissues via paracrine and endocrine mechanisms. Exploring the mechanism by which adipose tissue develops, which depends on transcription factors, hormones, enzymes and non-coding RNAs [5][6][7][8], may improve genetic improvement and beef quality. Meanwhile, revealing more regulatory mechanism of IMF will provide theoretical basis for improving beef quality and healthy diet of human. SGK1 is one of the most conserved proteins in mammals and a member of the serine/threonine kinase AGC family [9]. SGK1 was rst cloned in rat breast cancer cells stimulated by serum and glucocorticoid.
Functionally, It is similar to the phosphorylated mammalian target of rapamycin (mTOR) [10]. SGK1 inhibits autophagy in red muscle by acting upstream of ULK1, which initiates autophagosome formation [11,12]. SGK1 also controls renal tubular transport by inhibiting the degradation of ENaC [13]. The inhibitory role of SGK1 in angiogenesis is regulated by BMP9 and leads to endothelial cell proliferation [14]. SGK1 also plays a key role in regulating adipocyte differentiation and in maintaining glycolipid homeostasis. Both its mRNA and protein levels are high in white adipose tissue, and this has been linked to obesity and type 2 diabetes [15]. Expression of SGK1 in subcutaneous and omental adipose tissue is regulated by obesity-associated in ammation, and SGK1 expression is stimulated in response to in ammatory signals [16]. SGK1 is a mediator of glucocorticoids and high-fat feeding, and induced insulin resistance in adipocytes by phosphorylation of FOXO1 in db/db mice [17].
Multiple pathways participate in the regulation of adipogenesis [18], among which the PI3K/Akt signaling pathway plays a pivotal role in maintaining lipid metabolism and regulating insulin [19]. The family of forkhead box O (FOXO) transcription factors has emerged as a central player that is located downstream of the PI3K/AKT signaling and is negatively regulated by phosphorylated AKT [20,21]. Genetic mutations in FoxO genes, or abnormal expression of FoxO proteins, are associated with metabolic disease, cancer or altered lifespan in humans and animals [22]. FOXO1 is a downstream target of Akt, regulating the cell cycle of adipocytes and lipogenesis through PPARγ [23]. Moreover, FOXO1 and FOXO3 may be targets for in anti-obesity treatments. Indeed, a FOXO1 antagonist increased adipocyte autophagy and inhibited obesity [24], whereas FOXO3 knockdown signi cantly inhibited autophagy and lipid accumulation, reducing LPS-induced in ammation [25].
After previous RNA-seq data that showed that SGK1 is differentially expressed in bovine adipose tissue, herein, we have explored the role of SGK1 in bovine adipogenesisin. We demonstrate that SGK1 affects the PI3K/Akt signaling pathway and further changes the phosphorylation of downstream genes FOXO1/FOXO3. This contributes to expand the number of candidates in a network regulating bovine adipogenesis and elucidate the role of SGK1 in promoting beef IMF.

Results
Establishment of a model of bovine preadipocytes differentiated into mature adipocytes Primary preadipocytes from bovine subcutaneous adipose tissue were isolated using collagenase Type I digestion. Primary preadipocytes were induced to differentiate into mature adipocytes using insulin when the con uence of seeded preadipocytes reached ~90%-100%. There were many lipid droplets present in adipocytes from day 2 to 6 ( Figure 1A). Oil Red O staining results showed that mature adipocytes contained more lipid droplets than primary preadipocytes ( Figure 1B). The absorbance of lipid droplets (measured at 480 nm) of mature adipocytes was also signi cantly higher than in preadipocytes ( Figure  1C). Consistent with the above morphological observation, the relative expression level of adipocyte-speci c genes, e.g., peroxisome proliferative activated receptor gamma (PPARγ), CCAAT enhancer binding protein alpha (C/EBPα), Lipoprotein Lipase (LPL) and fatty acid-binding protein 4 (FABP4), increased with adipocyte differentiation ( Figure 1D-G). These results indicate a successful induction of primary preadipocytes into mature adipocytes.

Expression pattern detection of bovine SGK1
Compared to heart, SGK1 was more expressed in kidney (P < 0.01), followed by liver, lung, spleen and fat tissues, whereas expression was lowest in muscle (P < 0.05) ( Fig. 2A). The relative expression level of SGK1 peaked at day 6 of differentiation during adipocyte differentiation (Fig. 2B). Thus, we successfully obtained the expression pattern of bovine SGK1 gene in different tissues and during adipocyte differentiation as well as we speculated SGK1 might a hub gene during lipid metabolism.

Overexpression of SGK1 promotes bovine adipogenesis
The 1296 bp-long CDS region of full length bovine SGK1 gene was ampli ed without mutations. The recombinant plasmid was digested with Pac I enzyme, producing two expected fragments ( Supplementary Fig. 1A). The digestion products were transfected into taraka cells. Taraka cells grew well 24 h after transfection (Supplementary Fig. 1B-1) and cells expressed green uorescent protein (GFP) that appeared forming a scattered pattern (Supplementary Fig. 1B-2). GFP expression increased after day 5 ( Supplementary Fig. 1B-3) and appeared the shape of comet-like tail ( Supplementary Fig. 1B-4). GFP covered the whole plate and detached from plate after day 8. Subsequently, collected cells with culture medium and repeatedly infected 4-5 times to obtain high-titer virus solution. The virus titers of OE-SGK1 and OE-NC were 2.51*1010 and 3.16*1010 pfu/mL, respectively. Therefore, the overexpression virus was successfully packaged.
Bovine preadipocytes were infected with OE-SGK1 and OE-NC. Adipocytes grew well, reaching ~90% con uence and GFP covered the visual eld after 48 h (Fig. 3A). In OE-SGK1, expression of mRNA was 664 times higher than in OE-NC (Fig. 3B) and SGK1 protein expression was much higher and formed more lipid droplets than in OE-SGK1 ( Fig. 3C-D). RNA integrity was measured by electrophoresis on agarose gel ( Supplementary Fig. 1C). The results of qPCR revealed that mRNA expression of adipocytespeci c genes, e.g., PPARγ and C/EBPα, increased in OE-SGK1 compared with OE-NC (Fig. 3E) whereas cell cycle-related genes decreased (Fig. 3F). Western blots showed no differences in PPARγ and C/EBPα, whereas CDK2 and CCND2 were down-regulated when SGK1 was overexpressed ( Fig. 3G and 3H). Thus, SGK1 promoted bovine adipogenesis mainly via inhibition of preadipocyte proliferation.

Loss-of-function SGK1 represses bovine adipogenesis
Three 61 bp shRNAs were inserted in pENTR/U6 (Supplementary Fig. 2A) and the coding region of SGK1 was joined with psicheck II (Supplementary Fig. 2B). A dual-luciferase reporter assay identi ed shRNA1 as the best interfering sequence ( Supplementary Fig. 2C). Thus, shRNA1 was used to package lentivirus and the recombinant plasmid c-shRNA1 was constructed ( Supplementary Fig. 2D). The titers of sh-SGK1 and sh-NC were 108 and 3 x 108 TU/mL, respectively.
Bovine preadipocytes infected with sh-SGK1 and sh-NC grew well and con uence reached ~90%, with GFP covering the whole visual eld after 72 h (Fig. 4A). In sh-SGK1, reduction of mRNA (Fig. 4B), SGK1 protein expression (Fig. 4C) and lipid droplets (Fig. 4D) were observed, compared with sh-NC. RNA integrity was measured by electrophoresis on agarose gel ( Supplementary Fig. 2E). In sh-SGK1, results qPCR revealed reduction in mRNA expression of adipocyte-speci c genes (Fig. 4E) and increase in cell cycle related genes (Fig. 4F) compared to sh-NC. This is opposite to the results observed under SGK1 overexpressed. Expression of PPARγ and C/EBPα were down-regulated and no difference was observed in CDK2 and CCND2 ( Fig. 4G and 4H). Hence, loss-of-function SGK1 repressed bovine adipogenesis mainly through repressing adipocytes differentiation. A total of 1704 differentially expressed genes (DEGs) were detected, 1126 of which were up-regulated and 578 were down-regulated (Fig. 5B). k-means cluster analysis was performed on DEGs (Fig. 5C). Ten clusters were obtained and key genes were screened at the vertex or in ection points of each cluster (Supplementary Table 3). By GO annotation, DEGs were assigned to metabolic, development and biological processes, among others (Fig. 5D). Subsequently, the terms related to metabolism and development, positive regulation of MAPK cascade, triglyceride biosynthetic process and cellular response to cAMP were selected to draw a GO chord graph, whereas genes of interest in the selected terms were sorted according to their logFC value (Fig. 5E). KEGG enrichment analysis revealed that DEGs were signi cantly enriched in PI3K/Akt, MAPK and lipic acid metabolism signaling pathways, among others ( Fig. 5F). Genes involved in signaling pathways related to lipid metabolism, and those related to PI3K/Akt, MAPK and related-lipid metabolism signaling pathways were screened to perform visual analysis (Fig. 5G).
Subsequently, qRT-PCR was used to verify the reliability of RNA-seq sequencing results (primers shown in Supplementary Table 4). Six up-regulated genes and four down-regulated genes were randomly selected for qRT-PCR. Expression change of up-regulated (CDKL4, LIF, SGK2, CDK1, KLF5 and IL6) and downregulated (ELOVL6, FAP, PLIN4 and FABP7) genes was consistent with RNA-seq results (Fig. 5H).

SGK 1 indirectly affects the phosphorylation of FOXO1 and FOXO3 via the PI3K/Akt signaling pathway
It has been reported that SGK1 affects the phosphorylation of FOXO1 and FOXO3 [26,27]. Indeed, FOXO1 and FOXO3 are downstream genes of the PI3K/Akt signaling pathway [20]. STRING predicts an interaction between SGK1 and FOXO family proteins (Fig. 6A-B). Indeed, mRNA levels of FOXO1 and FOXO3 were inhibited after SGK1 was overexpressed, and opposite results were obtained for loss-offunction SGK1 (Fig. 6C). Protein expression and phosphorylation of FOXO1 and FOXO3 were detected by Western blot after SGK1 was overexpressed or inhibited. Compared with OE-NC, OE-SGK1 led to reduced expression of FOXO1 and FOXO3, with increased phosphorylation (Fig. 6D-E). Compared with sh-NC, expression of FOXO1 and FOXO3 were up-regulated in response to SGK1 loss-of-function, wheras phosphorylation of FOXO3 was down-regulated ( Fig. 6F-G). The regulatory network by which SGK1 regulates bovine adipogenesis through phosphorylation of FOXO1 and FOXO3 was mapped through the PI3K/Akt signaling pathway (Fig. 6H).
Discussion SGK 1 is a critical factor for bovine adipogenesis.
Understanding how regulators control beef IMF content can suggest candidates that promote adipocyte differentiation and lipid deposition. A model of bovine preadipocytes differentiated into mature adipocytes was established as described [28]. The role of SGK1 during bovine adipogenesis was examined by overexpression adenovirus and short hairpin RNA lentivirus packaging. Viral vectors are easily manipulated and mass reproduced in cells. Also, they are widely used in various cell types as they provide strong infection and e ciency, compared with DNA-liposome mixture transfections [29,30]. Our study reveals that SGK1 is a positive regulatory factor during bovine adipocyte differentiation, regulating the expression of speci c-adipocyte genes and proliferation genes. SGK1 changes the gene expression in PI3K/Akt and other signaling pathways related to lipid metabolism. Moreover, SGK1 indirectly affected the phosphorylation level of FOXO1/FOXO3 via the PI3K/Akt signaling pathway. Thus, our research expands the number of candidates involved in a network of regulating bovine adipogenesis.
A previous research reported that SGK1 is mainly expressed in kidney where it regulates multiple renal ion channels [31]. Our study also shows that expression of SGK1 in cattle is higher in kidney and adipose tissue than in heart. The latter is consistent with results in mice [32]. Moreover, SGK1 was highly expressed at day 6 during adipogenic differentiation. Preadipocytes were induced to differentiate into mature adipocytes at day 7, leading to deposition of lipid droplets [33]. It has been reported that SGK1 is mainly expressed at the last stage of adipocyte differentiation [32]. Therefore, together with previous available evidence, we propose that SGK1 regulates fat deposition and promotes lipid accumulation during the late stages of bovine adipocyte differentiation. SGK 1 affects the expression of genes related to adipocyte differentiation and proliferation in cattle.
We successfully packaged the overexpression adenovirus and short hairpin RNA lentivirus of SGK1 to verify the function of SGK1 in the regulation of bovine adipogenesis. Expression of adipocyte differentiation-related genes, e.g., PPARγ and C/EBPα, increased when SGK1 was overexpressed. As a key regulator of adipogenesis, PPARγ controls the transcription of numerous genes related to adipocyte differentiation and lipid accumulation [34,35], and a high level of the C/EBPα gene maintains adipocytes in a fully differentiated state [36,37]. Loss-of-function SGK1 produced opposite results. Consistently, OE-SGK1 produced more lipid droplets than the control. It has been reported that transgene SGK1 increased the lipid droplets formation in 3T3-L1 cells compared with wild type cells [38], whereas expression of proliferation-related genes decreased in OE-SGK1, e.g., PCNA, CCND2 and MCM6, which are marker genes regulating cell proliferation [39,40].
Overexpression of SGK1 did not change the expression of differentiation-related proteins PPARγ and C/EBPα, but decreased that of proliferation-related proteins CCDN2 and CDK2. This suggests that overexpression of SGK1 positively regulates bovine adipogenesis mainly via inhibition of adipocyte proliferation. Inhibition of SGK1 expression by short hairpin RNA lentivirus led to reduced expression of adipogenic differentiation genes and increased expression of cell proliferation genes. The inhibition of insulin signalling by dexamethasone and oleic acid was reversed by lv-shSGK1 [41], whereas we observed that loss-of-function SGK1 led to down-regulation of proteins related to differentiation (PPARγ and C/EBPα) and no changes in proteins related to proliferation (CCDN2 and CDK2). This suggests that sh-SGK1 repressed bovine adipogenesis mainly by inhibiting adipocytes differentiation. Therefore our study proves that SGK1 promotes bovine adipogenesis through adipogenic differentiation and repressing preadipocyte proliferation. We also show that overexpression of SGK1 promotes bovine adipogenesis mainly through inhibiting adipocyte proliferation, whereas loss-of-function SGK1 repressed adipogenesis mainly via inhibiting adipocyte differentiation.

SGK 1 regulates gene expression related to lipid metabolism in bovine adipocytes
1704 DEGs were found when SGK1 was overexpressed. GO annotation analysis found DEGs related to metabolic process, development process and biological process activation. KEGG enrichment analysis revealed DEGs signi cant enriched in the PI3K/Akt signaling pathway, MAPK signaling pathway and lipic acid metabolism signaling pathway.
The MAPK/PI3K/Akt signaling pathway is important in regulating lipid metabolism, energy homeostasis and cell proliferation [42][43][44]. Overexpression of SGK1 up-regulated FGF23, a key factor in the MAPK and Akt signaling pathway. FGF23 is a key endocrine factor involved in the regulation of systemic homeostasis and lipid metabolism [45]. The FOXO signaling pathway is essential in cellular energy production, oxidative stress resistance, and cell viability and proliferation [46]. SGK1 down-regulated PCK1, located in the FOXO and PI3K/Akt signaling pathway. Tissue-speci c knock out of PCK1 in mice produced a phenotype of obesity, lipodystrophy, fatty liver, and death [47]. Thus, SGK1 might promote bovine adipogenesis through inhibiting PCK1 on the PI3K/Akt signaling pathway. Our results show that DEG colony-stimulating factor 3 (CSF-3) was enriched in the PI3K/Akt signaling pathway. The function of CSF-3 is closely related to adipose tissue and whole-body insulin sensitivity and glucose tolerance in human [48]. Furthermore, SGK1 also affects expression of genes enriched in the lipid metabolism signaling pathway. In conclusion, SGK1 mainly regulates the expression of PCK1 in the PI3K/Akt signaling pathway and up-regulates FGF23, FGFR3, promoting mitosis, cell growth and adipogenesis in cattle. SGK 1 phosphorylation of FOXO1 and FOXO3 via PI3K/Akt signaling pathway. SGK1 affect many multiple physiologic processes via phosphorylation of downstream proteins [49][50][51]. SGK1 promotes adipocyte differentiation and controls cell proliferation by regulating the phosphorylation level of FOXO1 and FOXO3, respectively [52,53]. FOXO1 directly inhibited the expression of PPARγ and interfered with promoter DNA occupancy of the receptor via direct protein-protein interaction [23,54]. Our results show that SGK1 affects the expression of the PI3K/Akt signaling pathway. FOXO family proteins are located downstream of this signaling pathway and are regulated by phosphorylated AKT [21]. After overexpression of SGK1, mRNA and protein expression level of FOXO1 and FOXO3 were down-regulated, whereas inhibition of SGK1 showed opposite results in our research. Overexpression of FOXO1 inhibited bovine adipogenesis and induced apoptosis [55], and FOXO1 interacted with PPARγ and repressed its transcription during adipogenesis [56]. We found that overexpression of SGK1 elevates the phosphorylation of FOXO1 and FOXO3, and loss-of-function SGK1 had opposite effects. Notably, the phosphorylation level of FOXO1 did not change in sh-SGK1 and sh-NC. FOXO1 is located in the nucleus where it may control the cell cycle [57]. Phosphorylation of FOXO1 by SGK1 may lead to the loss of its transcriptional activity by nuclear exclusion [58]. However, shRNA is synthesized in an nucleus of cells and is transported to the cytoplasm [59]. This may explain the poor interference e ciency of sh-SGK1 and may have resulted in unchanged phosphorylation level of FOXO1.
The SGK1/Akt signaling pathway affects adipogenesis in mouse white adipose tissue via phosphorylation of FOXO1 [32,60] and the interaction between SGK1 and FOXO3 has been described in several disease contexts [61, 62]. Our study shows that SGK1 can also affect FOXO3 in bovine adipose tissue. Although we show that SGK1 regulates bovine adipogenesis by phosphorylation FOXO1 and FOXO3 via PI3K/Akt signaling pathway, the interaction is not direct.

Conclusion
Overall, SGK1 is a positive regulatory factor during bovine adipogenesis. In mechanism SGK1 indirectly affects the phosphorylation levels of FOXO1/FOXO3 via the PI3K/Akt signaling pathway. Our study provides clues to the regulatory role of SGK1 during bovine adipogenesis and in improving beef quality.

Ethics statement
Three adult Nanyang cattles were slaughtered from Biyin cattle farm Nanyang city, Henan province. Subcutaneous adipose tissue was obtained from calves for isolation of preadipocytes. The Animal Ethics Committees of Ningxia University approved the experimental design and the animal sample collection for the present study (permit number NXUC20211168). This study is reported in accordance with the ARRIVE guidelines.

Isolation and induced differentiation of bovine preadipocytes
Bovine preadipocytes were isolated using collagenase digestion, as described [63]. Brie y, adipose tissue without blood vessels and connective tissues was minced into cubes with size ~1 mm 3 and digested in 1 mg/mL collagenase type I (Sigma, C0130) in a water bath for 90 min at 37 ℃. Subsequently, preadipocytes were seeded and cultured on a 10 cm 2 plate with a growth medium containing Dulbecco's modi ed eagle medium (DMEM), 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin (P/S) at 37 ℃ with 5% CO 2 for 24 h. Cells were then washed with PBS three times and fresh medium was replaced every two days.

Sequence analysis of bovine SGK1
Primers were designed to amplify the CDS region of SGK1 (Genbank No. NM_001102033.1) (Supplementary Table 1) and interacting proteins were predicted using STRING.

Recombinant adenovirus packaging
Endonucleases Kpn I (Takara, Dalian, China) and Hind III (Takara) were added to forward and reverse primers, respectively. Subsequently, the CDS region was inserted into the shuttle vector pAd-tract-CMV and was transformed into competent cells BJ5183. The recombinant vector was transfected into taraka (HEK293a). Subsequently, cells and culture medium were collected named (OE-SGK1) with the and negative control named (OE-NC) eight days after transfection. pAd-track-CMV and taraka were preserved in our laboratory.
Dual luciferase report analysis Three short hairpin RNAs (shRNA1, shRNA2 and shRNA3) were designed using online software (http://rnaidesigner.thermo sher.com.rnaiexpress/) ( Table 1) and were synthesized. They were connected to the pENTR/U6 plasmid and were named pENTR/U6-shRNA1, pENTR/U6-shRNA2 and pENTR/U6-shRNA3. The recombinant plasmid was constructed with psicheck II and the CDS region of SGK1 (psicheck II-SGK1). Cell transfection was performed using Lipofectamine 3000 (Thermo, Waltham, MA, USA) according to the manufacturer's instructions. pENTR/U6-shRNA or pENTR/U6-NC were dissolved using Opti-MEM (Sigma, St. Louis, Missouri, USA) and incubated with psicheckII-SGK1 to form a DNAliposome mixture. Subsequently, the DNA-liposome mixture was added to the culture medium and was incubated 48 hrs at 37°C and 5% CO 2 . The luciferase activity in 24 wells containing HEK293T cells were measured. Table 1 short hairpin RNA sequences information of SGK1 Names of sequence Note: GATCC and TCGAG were the sequence of endonuclease BamHl and Xhol, italic sequence indicated loop sequence.

Short hairpin RNA lentivirus packaging
The selected shRNA1 for single target lentivirus packaging and the negative control were named sh-SGK1 and sh-NC, respectively. Recombinant vector c-shRNA1 and helper vectors psPAX2 and pMD2G were cotransfected with Lipofectamine 3000 into HEK293T cells according to the manufacturer's instructions. Fluorescence was detected with an inverted microscope after 24 h. The supernatant were collected at 48 h and 72 h to concentrate and purify the virus.

Cell infection
Preadipocytes were seeded onto 6-well plates overnight to reach 70% con uence. The OE-SGK1 and sh-SGK1 virus solutions were added to adipocyte culture plates (n = 3) with an optimal MOI (multiplicity of infection) value determined in a preliminary experiment, using OE-NC and sh-NC, respectively, as controls. After 48 h incubation at 37°C and 5% CO2, uorescence was detected with an inverted microscope after 24 h and the medium was replaced by high glucose medium containing 10% FBS.
Total RNA extraction and RT-qPCR Total RNA was extracted from adipocytes or tissues using Trizol reagent. cDNA was synthesized by using a Primescript TM RT reagent kit (TaKaRa) and SYBR premix Ex Taq II kit (TaKaRa) was used to perform the RT-qPCR reaction on a Bio-Red CFX 96 Touch instrument (Bio-Rad, Hercules, CA, USA). The 2 −ΔΔCt method was used to analyze the data. Primers of adipogenic and cell cycle genes were designed by Primer Premier 5 according to the primer designing criteria (Supplementary Table 1). Relative expression levels were normalized with the internal control GAPDH [64].

RNA-seq analysis
Adipocytes infected by OE-SGK1 or OE-NC and induced differentiation for 6 days. Subsequently, cells were used to perform deep sequencing (n = 4) using Illumina xten completed by BioMarker Co (Qingdao, China).

RNA-seq data analysis
RNA-seq data was analyzed in R (R x64 4.1.2) using the R package (such as pheatmap and GOplot) and visual analysis was performed in Cytoscape (cytoscape_3.9.0).

Statistical analysis
For each group at least three independent experiments were performed and data was expressed as mean ± standard deviation (SD). GraphPad Prism v8.0.2 (GraphPad Software, Inc., La Jolla, CA, USA) was used to analyze the experimental data. Comparisons among multiple groups were performed with two-way analysis of variance (ANOVA). P<0.05 was considered to indicate a statistically signi cant difference, and P<0.01 was considered to indicate that the difference was extremely signi cant.

Declarations
Conceived and designed the research:YM and ZXL; Analyzed the data and conducted the experiment: ZXL; Wrote the paper: ZXL; Modi ed manuscript: YM, DWW, YFM, LT, SZW, CLP. All authors read and approved the nal manuscript.

Consent for publication
Not applicable.

Competing interests
There is no con ict of interest exits in this manuscript, and the manuscript is approved by all authors for publication.