The E2F1^–/– mice were obtained from the Jaxson laboratory¹ and were bred, maintained, and operated in the Animal Experimental Center of Tongji Hospital, Huazhong University of Science and Technology. In this study, the E2F1^–/– mice (male, 6–8 w) were compared with WT littermates of the same age and gender. All mice were allowed free access to food and water under controlled conditions (12/12 h light/dark cycle with humidity of 60 ± 5%, and a temperature of 22 ± 3°C). All surgical procedures were performed under anesthesia. The study was performed following guidelines on animal experimentation of the Ethical Committee of the Tongji Hospital.

The subcutaneous adipose tissues were extracted from the inguinal region of mice, and the superficial fascia and blood vessels were removed. Then, the adipose tissues were washed with phosphate-buffered saline (PBS) and minced into small pieces, followed by digestion using 0.15% type I Collagenase (Sigma, United States) under 37°C for 35 min. The Dulbecco’s modified Eagle’s medium (DMEM) (Gibco, United States) high glucose medium containing 20% fetal bovine serum (FBS) (Gibco, United States) was added to terminate the digestion, followed by filtration using a 75 μm screen mesh and centrifugation at 400 × g for 5 min. Lastly, the cells were added to the primary medium with 20% FBS and inoculated into a culture dish, and placed in an incubator with 5% CO₂ at 37°C. After 24 h, the non-adherent cells were removed. These obtained ADSCs were passaged (1:3) using the culture medium containing 10% FBS every 2–3 days when they reached approximately 80–90% confluence.

In a 6-well plate, 1 × 10⁵ ADSCs/cm² were plated in complete media for 24 h. Differentiation was initiated by replacing complete media with adipogenic differentiation media, which contained 1 μM dexamethasone, 10 μg/μL insulin, 0.5 mM isobutylmethylxanthine, and 200 μM indomethacin. Cells were incubated for 21 days, and the media changed every 3 days. Murine 3T3-L1 preadipocyte cells were cultured and differentiated as previously described (Hwang and Lee, 2020). To assess the effect of autophagy on adipocyte differentiation, the autophagic inhibitor (3-MA), an inhibitor of phosphoinositide 3 kinase that specifically inhibits autophagosome formation at 5 mM (MCE, China), were added to the media before adipogenic differentiation.

The Oil Red O staining was used to evaluate adipogenic differentiation. The cells were rinsed with PBS and fixed with 4% formaldehyde, then were rinsed with distilled water, dehydrated with 60% isopropanol, stained with Oil Red O, and finally were observed under a light microscope (SDPTOP, China).

The fourth-passage ADSCs were tested by flow cytometry analysis. Briefly, the adherent cells were harvested by trypsinization, centrifuged and washed with sterile PBS, and finally resuspended in PBS. For biomarker identification, cell suspensions were incubated with anti-CD29-APC, anti-CD90-FITC, anti-CD105-APC, anti-CD34-Alexa Fluor 647, and anti-CD44-PE⋅Cy^TM 7 antibodies (Becton Dickinson, United States). Finally, ADSCs were analyzed using a FACS Calibur cytometer (Becton Dickinson, United States) and Flow Jo VX software.

The adipose tissues were fixed in 4% formaldehyde and embedded in paraffin. Tissues were then cut into sections of 3 μm thickness with a microtome, deparaffinized, and stained with H&E. H&E images were taken with the microscope (SDPTOP, China) at 200 × and 400 × magnification. The adipocytes area quantification, represented by adipocyte diameter was performed in 6 fields of each preparation.

For IHC staining of UCP-1 expression level in WAT, formalin-fixed and paraffin-embedded adipose tissues were deparaffinized to prevent non-specific protein binding. Samples were then incubated with diluted primary anti-UCP-1 antibody overnight at 4°C. After the wash steps, the sections were incubated with secondary antibody for 2 h and were visualized by incubating with DAB substrate. The UCP-1 expressions were evaluated by high-power light microscopy examination. Images were taken with the microscope (SDPTOP, China) at 200 × and 400 × magnifications. The expression level of UCP-1 in adipose tissue was indicated with a stained positive index (integrated density/area).

A small portion of WAT sections (1 mm³) were fixed with fresh TEM fixative at 4°C overnight and then washed the tissues using 0.1 M PB. Tissues were fixed with 1% OsO4 in 0.1 M PB (pH 7.4) for 2 h at room temperature. The tissues were dehydrated by a series of graded concentrations of alcohol and acetone and embedded in araldite. The resin blocks were cut to 60–80 nm thin and then stained with 2% uranium acetate and lead citrate, and observed under the TEM (Hitachi, Japan). The number of mitochondria in a region containing the complete view of an adipocyte was counted.

The E2F1 siRNAs were purchased from Ribobio (Guangzhou, China). The sequences for the AGER siRNA used for the experiments were as follows: E2F1 siRNA1: 5′-GCAGAAACG GCGCATCTAT-3′; E2F1 siRNA2: 5′-GGGTGAGGGCATTAGA GAT-3′. ADSCs were plated in six-well plates and transient transfection using Lipofectamine 3000 Reagent (Invitrogen, United States) according to the manufacturer’s instructions.

The total RNA was isolated from the inguinal WAT of WT and E2F1^–/– mice using TRIzol (Takara, Japan), and 1 μg of RNA in a final reaction volume of 20 μL was then reversed-transcribed into complementary DNA (cDNA) using the 1st Strand cDNA Synthesis SuperMix (Yeason, China) according to the manufacturer’s instructions. qRT-PCR was performed using SYBR Green Master Mix (Yeason, China) and was detected with an ABI Q1 PCR System (Thermo Fisher Scientific, United States). Threshold cycles of primer probes were normalized to the housekeeping gene β-actin and translated to relative values. The following primer set sequences were used, LC3-II (forward 5′- TTATAGAGCGATACAAGGGGG AG-3′, reverse 5′-CGCCGTCTGATTATCTTGATGAG-3′), p62 (forward 5′-GAGGCACCCCGAAACATGG-3′, reverse 5′-ACTTATAGCGAGTTCCCACCA-3′), ATG5 (forward 5′-TG TGCTTCGAGATGTGTGGTT-3′, reverse 5′-ACCAACGTCA AATAGCTGACTC-3′), ATG7 (forward 5′-TCTGGGAAGCC ATAAAGTCAGG-3′, reverse 5′-GCGAAGGTCAGGAGCAGA A-3′), UCP-1 (forward 5′-GTGAACCCGACAACTTCCGA A-3′, reverse 5′-TGCCAGGCAAGCTGAAACTC-3′), E2F1 (forward 5′-GAGAAGTCACGCTATGAAACCTC-3′, reverse 5′-CCCAGTTCAGGTCAACGACAC-3′), and β-actin (forward 5′-GGCTGTATTCCCCTCCATCG-3′, reverse 5′-CCAGTTGG TAACAATGCCATGT-3′).

Briefly, the protein concentration was quantified by the BCA method. The proteins were separated using 12% SDS-PAGE and transferred to a PVDF membrane. The membrane was blocked with 5% skim milk for 1.5 h and with 1:1000 dilutions of anti-LC3 antibody (Proteintech, United States), anti-p62 antibody (Proteintech, United States), anti-UCP-1 antibody (Proteintech, United States), anti-E2F1 antibody (ABclonal, China), and anti-β-actin antibody (ABclonal, China) overnight at 4°C. The membrane was then washed with TBST and incubated with a 1: 4000 diluted HRP-conjugated secondary antibody (ABclonal, China) for 1 h at 37°C. At last, the immunoreactive bands were detected by using an enhanced chemiluminescent (ECL) reagent kit (Yeason, China), and were scanned by ChemiDoc XRS⁺ imaging system (Bio-Rad, United States) and analyzed by Image Lab software. The analysis represented the ratio of the target protein expression relative to β-actin expression.

All experiments were performed with at least three replicates per group. Data are shown as the means ± standard deviation (means ± SD). Two treatment groups were compared by Student’s t-test. Statistical analysis was conducted using GraphPad Prism 9.0 software, and statistical significance was declared as ^∗p < 0.05, ^∗∗p < 0.01 and ^∗∗∗p < 0.001, respectively.

Firstly, we intended to explore the impact of E2F1 knockout on the general morphology of mice and WAT. There was no obvious difference in the size and weight of subcutaneous inguinal WAT between the WT and E2F1^–/– groups (Figure 1A). Mice weight growth curve was measured weekly from the birthday, and there was no significant difference in weight between the two groups (Figure 1B). Interestingly, the microscopic observation revealed that the E2F1^–/– WAT exhibited substantially denser and smaller adipocytes (Figure 1C). The adipocyte diameter in the E2F1^–/– group was significantly shorter than that in the WT group (P < 0.001) (Figure 1D). These results certified that E2F1 knockout did not change the body and WAT weight, but reshaped the morphology of adipose. Thus, E2F1^–/– WAT might tend to be beige adipose morphology with multiple small lipid droplets.

Emerging evidence has confirmed that the increased number of mitochondria and high UCP-1 expression in WAT are strong indicators for browning. Consequently, we hypothesized that E2F1 knockout might induce WAT browning activity, which is related to mitochondria and UCP-1 in WAT. Intriguingly, E2F1^–/– WAT showed small intracellular lipid droplets and dramatically increased mitochondria, compared with the WT WAT (P < 0.01) (Figures 2A,F). In the WT group, the presence of multiple degraded vesicles was consistent with the structure of autophagosomes and autophagolysosomes, which suggested that autophagy activity might occur. The UCP-1 of IHC staining demonstrated that there were more UCP-1 expression and beige cells formation in the E2F1^–/– group (P < 0.01) (Figures 2B,G). Moreover, the Western Blot analysis showed that the expression of UCP-1 protein was significantly increased in the E2F1^–/– WAT (P < 0.05) (Figures 2C, 3C). Similarly, the qRT-PCR results consistently confirmed that E2F1^–/– WAT possessed higher mRNA expression of UCP-1 than WT WAT (P < 0.05) (Figure 2D). In addition, to validate the adipogenic gene expression during WAT browning, the PPARγ mRNA expression in the E2F1^–/– WAT group was detected, which exhibited a downward trend compared with the WT group (Figure 2E). These data indicated that E2F1 knockout induced WAT into beige adipose tissue by increasing mitochondrial number and expressing highly UCP-1 both on mRNA and protein levels.

Since the autophagy-related autophagosomes expression was decreased in E2F1^–/– WAT, we further explored the underlying autophagy mechanism of E2F1 in regulating WAT browning. The E2F1^–/– group demonstrated the protein expressions of E2F1 and LC3-II/I were significantly decreased, and p62 was increased inversely (P < 0.05, P < 0.01) (Figures 3A,C). Compared with the WT group, the reduced mRNA expressions of E2F1, LC3-II, ATG5, and ATG7 in E2F1^–/– WAT were further confirmed (P < 0.05, P < 0.01) (Figure 3B). The results suggested that autophagy participated in the intracellular beige adipogenesis, while E2F1^–/– WAT exhibited lower autophagy gene expression than WT WAT.

The cultured WT and E2F1^–/– ADSCs both presented a typical fibroblast-like morphology, which was similar to previous studies (Figure 4A). Flow cytometry analysis also verified the surface markers of obtained ADSCs, by staining positively for CD29, CD90, CD105, and CD44, and negatively for CD31 and CD34 (Figure 4C). The Oil Red O staining images indicated the presence of oil droplet formation, suggesting the acquisition of mature adipocytes in vitro (Figure 4B). To demonstrate the role of E2F1 in beige adipocyte modulation in vitro, the UCP-1 protein, and mRNA expression were analyzed. The Western Blot analysis testified that E2F1^–/– adipocytes possessed significantly high expression of UCP-1 (P < 0.05) (Figures 5A,E). The qRT-PCR results further proved the higher mRNA expression of UCP-1 in the E2F1^–/– group than WT group in vitro (P < 0.01) (Figure 5B). Thus, E2F1 knockout could increase the UCP-1 expression in adipocytes, which was necessary to the white-to-beige process.

The autophagy inhibition was proved to occur in the E2F1^–/– WAT in vivo, thus participating in the WAT browning remodeling. To confirm that the role of E2F1 in beige adipocytes regulation depends on autophagy in vitro, the autophagy-related genes were detected. At the protein level, the LC3-II/I expression was decreased, and the p62 expression was increased in the E2F1^–/– adipocytes group, compared with the WT adipocyte group (P < 0.05, P < 0.01) (Figures 5C,E). Also, the reduced mRNA expression of LC3-II and the induced mRNA expression of p62 were further verified in E2F1^–/– adipocytes (P < 0.05, P < 0.01) (Figure 5D). The results proved that E2F1 could be involved in the upregulation of autophagy-related genes in adipocytes in vitro, and E2F1 knockout inhibited autophagy activity to promote WAT browning.

Firstly, we used the 3T3-L1 cell line as an in vitro model to generate adipocyte-like cells. To figure out whether the reduced autophagy could affect the UCP-1 and E2F1 expression, the 3T3-L1 adipocytes were treated with autophagy inhibitor 3-MA at the early differentiation stage. After adipogenic induction, the inhibition of autophagy using 3MA resulted in lower autophagy activity. The LC3-II and ATG5 mRNA expressions were relatively reduced compared to the normal cultured 3T3-L1 cells (Figures 6A,B,I). The p62 mRNA expression was increased in the 3T3-L1 + 3MA group (Figure 6B). Furthermore, the 3MA-treated 3T3-L1 adipocytes also expressed higher UCP-1 on mRNA and protein levels, and lower PPARγ mRNA expression than the 3T3-L1 adipocytes, which demonstrated that the autophagy inhibition leads to reduced lipogenesis and enhanced WAT browning transition (Figures 6A,C,I). Importantly, there was no significant difference in E2F1 expression between the 3T3-L1 + 3MA group and 3T3-L1 group (Figure 6D). It was indicated that the autophagy inhibition in the WAT could be induced by E2F1 knockout, leading to the increased mitochondria and UCP1 expression in the browning WAT.

Next, three groups were designed to examine the effect of E2F1 on autophagy inhibition and WAT browning, including ADSCs + 3MA group, ADSCs + E2F1 siRNA group, and ADSCs + siNC group. After ADSC adipogenic differentiation, the expression levels of E2F1 mRNA were reduced by si-E2F1 in ADSCs (Figure 6E). The 3MA addition did not lead to a decrease of E2F1 mRNA expression in the WT ADSCs + 3MA group. The Western Blot analysis proved that the LC3-II/I protein expression indeed was reduced in both ADSCs + 3MA group and ADSCs + E2F1 siRNA group (Figures 6F,J). The ADSCs treated with E2F1 siRNA mimicked the effect of autophagy inhibitor to increase the UCP-1 expression at protein and mRNA level (Figures 6F,G,J). In addition, the p62 mRNA expression was relatively induced in ADSCs + 3MA group and ADSCs + E2F1 siRNA group, compared to the ADSCs + siNC group (Figure 6H). These results further confirmed that E2F1 deficiency could be involved in UCP-1 expression via autophagy inhibition in WAT and adipocytes.