Endothelial SIRT3 deficiency predisposes brown adipose tissue to whitening in diet-induced obesity

Abstract

Endothelial dysfunction and vascular rarefaction are supposed to be secondary to metabolic diseases, while recent evidence has revealed the primary roles of endothelium in initiating and accelerating metabolic disorders. Here, the effects and underlying mechanisms of endothelial SIRT3 in modulating the whitening of BAT during obesity progression were explored. Therefore, mice with global or BAT regional endothelium-specific Sirt3 knockout were constructed and fed with high-fat diet (HFD). The results showed that both global and BAT regional endothelium-specific Sirt3 knockout accelerated diet-induced weight gain, accompanied by glucose intolerance, insulin resistance, and BAT whitening. In vitro results revealed that the inhibition or knockdown of endothelial Sirt3 impeded palmitic acid-induced angiogenesis deficiency, while the overexpression of Sirt3 exhibited the opposite effects. Furtherly, endothelial Sirt3 overexpression ameliorated palmitic acid-induced adipocyte dysfunction and proinflammatory macrophages polarization in a paracrine way. Mechanistically, endothelial SIRT3 deficiency increased the acetylation of fatty acid synthase (FASN), which disturbed the fatty acid metabolism and thus, leading to angiogenesis insufficiency. Moreover, loss of SIRT3 promoted adipocytes dysfunction and proinflammatory macrophage polarization via CASP1-mediated pyroptosis. Endothelial SIRT3 loss contributed to diet-induced BAT whitening and obesity progression and thus, could be a therapeutic target in treating obesity and associated metabolic diseases.

Keywords: BAT whitening; SIRT3; angiocrine factor; angiogenesis; endothelium.

Figure 1

Loss of endothelial SIRT3 exacerbated HFD-induced BAT whitening. (A) Representative images of IF staining and corresponding quantitative results of CD31 (Red) in BAT, sWAT, and eWAT of C57 mice fed with ND or HFD (n=3). Scales bars, 50μm. (B) Representative images of IF staining and quantitative results of CD31 (red) in BAT, sWAT, and eWAT in db/m and db/db mice (n=3). Scales bars, 50μm. (C) Representative images of IF staining and quantitative results of CD31 (Red) and SIRT3 (Green) in BAT, sWAT, and eWAT of C57 mice fed with ND or HFD (n=3). Scales bars, 50μm. (D-F) The changes of body weight (D, n=4 or n=5), ITT (E, n=3 or n=4), and IPGTT (F, n=4). (G) Volume O₂, Volume CO₂, Heat, and RER of Sirt3^flox/flox-Tek-Cre mice (hereafter referred to as Sirt3-EKO mice) and Sirt3^flox/flox mice (hereafter referred to as WT mice) fed with HFD, detected by CLAMS (n=4). (H-I) Representative images of ¹⁸F-FDG uptake in BAT and quantitative results WT and Sirt3-EKO mice, detected by PET-CT scanning (n=3). (J) Representative images of H&E staining and quantitative results of the number and size of lipid droplets in BAT, sWAT, and eWAT of WT and Sirt3-EKO mice (n=4). Scales bars, 200μm. (K) Representative images of Western blots and quantitative results of UCP-1 in BAT of WT and Sirt3-EKO mice (n=3). β-actin served as a loading control. (L) Representative images of IHC staining and quantitative results of UCP-1 in BAT of WT and Sirt3-EKO mice (n=3). Scales bars, 200μm. (M) Representative images of TEM of BAT in WT and Sirt3-EKO mice (n=3). Scale bars, 1μm. (N) Representative images of IF staining and quantitative results of CD31 (Red) staining in BAT, sWAT and eWAT of WT an Sirt3-EKO mice (n=3). Scales bars, 50μm. ^*P< 0.05, ^**P< 0.01, ^***P< 0.001, ^****P< 0.0001 compared with ND (A, C), db/m (B), and WT (D-L, N). IF: immunofluorescence; BAT: brown adipose tissue; sWAT: subcutaneous white adipose tissue; eWAT: epididymal white adipose tissue; ND: normal diet; HFD: high fat diet; ITT: insulin tolerance test; IPGTT: Intraperitoneal glucose tolerance test; Volume O₂: Oxygen consumption; Volume CO₂: carbon dioxide production (Volume CO₂); RER: respiratory exchange ratio; H&E: Hematoxylin and eosin; IIHC: immunohistochemical; TEM: Transmission Electron Microscope; UCP-1: uncoupling protein 1.

Figure 2

Endothelial Sirt3 knockout in BAT accelerated HFD-induced BAT whitening. (A) Representative images of IF staining of AAV (Red) in BAT from regional-specific endothelial Sirt3 knockout mice (hereafter referred to as Sirt3-BAT-EKO mice) and their control mice (hereafter referred to as WT mice) (n=4). Scales bars, 50μm. (B) Representative images of IF staining of AAV (Red) and SIRT3 (Green) in BAT of WT and Sirt3-BAT-EKO mice (n=4). Scales bars, 50μm. (C) Representative images of the whole appearance, BAT, sWAT, and eWAT of WT and Sirt3-BAT-EKO mice at the end of the HFD intervention. (D-F) The changes of body weight (D, n=6 or n=8), blood glucose levels during the IPGTT (E, n=6 or n=8) and ITT (F, n=3 or n=4) of WT and Sirt3-BAT-EKO mice fed with HFD. (G) the changes of Volume O₂, Volume CO₂, RER, and Heat of WT and Sirt3-BAT-EKO mice fed with HFD, detected by CLAMS (n=6). (H) Representative images of ¹⁸F-FDG uptake in BAT in WT and Sirt3-BAT-EKO mice, detected by PET-CT scanning (n=3 or n=4). (I) Representative images of H&E staining, IHC staining of UCP-1, TEM of mitochondria, and quantitative results of lipid droplets and IHC staining, of BAT in WT and Sirt3-BAT-EKO mice (n=3 or n=4). Scales bars, 200μm or 200nm (TEM). (J) Representative images of H&E staining and quantitative results of the number or size of lipid droplets in BAT from WT and Sirt3-BAT-EKO mice (n=4). Scales bars, 200μm. (K) Representative images of Oil Red O staining of the liver, muscle, and kidney from WT and Sirt3-BAT-EKO mice (n=3). Scales bars, 200μm. (L) The levels of serum Triglyceride (TG) and Total Cholesterol (TC) in WT and Sirt3-BAT-EKO mice (n=3). (M) Representative images of IF staining (left) and quantitative results (right) of Ki67 (Green) and CD31 (Red) of BAT in WT and Sirt3-BAT-EKO mice (n=3). Scales bars, 50μm. (N-Q) Representative images of IF staining and quantitative results of IB4 (green, N, O), CD11c (green, P), CCR2 (green, P), CD206 (green, P), and GSDMD (green, Q) of BAT in WT and Sirt3-BAT-EKO mice (n=3). Scales bars, 50μm. ^*P< 0.05, ^**P< 0.01, ^***P< 0.001, ^****P< 0.0001 compared with WT mice (D, E, F, G, H, I, J, L, M, O, P, Q). AAV: Adeno-associated Virus; TG: Triglyceride; TC: Total Cholesterol; ITT: insulin tolerance test; IPGTT: Intraperitoneal glucose tolerance test; IB4: Isolectin B4; IF: immunofluorescence; BAT: brown adipose tissue; sWAT: subcutaneous white adipose tissue; eWAT: epididymal white adipose tissue; ND: normal diet; HFD: high fat diet; Volume O₂: Oxygen consumption; Volume CO₂: carbon dioxide production (Volume CO₂); RER: respiratory exchange ratio; H&E: Hematoxylin and eosin; IIHC: immunohistochemical; TEM: Transmission Electron Microscope; UCP-1: uncoupling protein 1.

Figure 3

Endothelial SIRT3 deficiency impeded angiogenesis. (A-B) Representative images of Reactome analysis of down-regulated (A) and up-regulated (B) pathways in HUVECs (hereafter referred to as ECs) treated with BSA (150μM) or PA (150μM), detected by RNA sequencing (n=3). (C-F) Representative images and quantitative results of EdU staining (C), F-actin (D), transwell (E), and SA-β-galactosidase (F) staining in ECs treated with BSA, PA, or 3-TYP (50μM) +PA (n=3). Scale bars, 50μm. (G) The mRNA levels of Il-1β, Vcam, and Il-6 in ECs treated with BSA, PA, or 3-TYP+PA (n=3). β-actin served as a loading control. (H) Representative images of Western blots and quantitative results of the indicated cytokines in ECs treated with BSA, PA, or 3-TYP+PA (n=3). β-actin served as a loading control. (I-L) Representative images and quantitative results of EdU staining (I), F-actin (J), SA-β-galactosidase (K), and transwell (F) assays in ECs treated with LV-con+BSA, LV-con+PA, or LV-Sirt3+PA (n=3). Scales bars, 50μm. (M) The mRNA levels of Il-1β, Vcam, and Il-6 in ECs treated with LV-con+BSA, LV-con+PA, or LV-Sirt3+PA (n=3). β-actin served as a loading control. (N) Representative images of Western blots and quantitative results of indicated cytokines in ECs treated with LV-con+BSA, LV-con+PA, or LV-Sirt3+PA (n=3). β-actin served as a loading control. ^*P< 0.05, ^**P< 0.01, ^***P< 0.001, ^****P< 0.0001 compared with BSA (C, E, G, H), and LV-con+BSA (I, L, M, N). ^#P< 0.05, ^##P< 0.01, ^###P< 0.001, ^####P< 0.0001 compared with PA (C, E, G, H) and LV-con+PA (I, L, M, N). HUVECs: Human umbilical vein endothelial cells; BSA: bovine serum albumin; PA: palmitic acid; IL-6: Interleukin 6; IL-1β: Interleukin 1β; VCAM: Vascular cell adhesion molecule.

Figure 4

Loss of endothelial SIRT3 exacerbated mitochondrial dysfunction and disturbance of fatty acid metabolism. (A-C) The changes of mitochondrial respiration indices, including CⅠ leak, CⅠ OXPHOS, CⅠ+Ⅱ OXPHOS, CⅠ+Ⅱ ETS, CⅡ ETS, and ROC, detected by O2k (A, n=3), cytosolic and mitochondrial ROS measured by DHE and MitoSOX fluorescent staining and individual quantitative results (B, n=3), and changes of mitochondrial Ca²⁺ that labeled with Rhod-2 AM (C, n=3) in ECs treated with LV-con+BSA, LV-con+PA, or LV-Sirt3+PA. Scales bars, 200μm. (D-E) The changes of mitochondrial metabolites profile detected via targeted metabolomics (D, n=6), especially Citric acid and Isocitric acid (E), in ECs treated with LV-con+BSA, LV-con+PA, or LV-Sirt3+PA (n=6). (F) The mRNA levels of Fasn, Acly, and Accs in ECs treated with LV-con+BSA, LV-con+PA, or LV-Sirt3+PA (n=3). β-actin served as a loading control. (G) Representative images of Western blots and quantitative results of indicated cytokines in ECs treated with LV-con+BSA, LV-con+PA, or LV-Sirt3+PA (n=3). β-actin served as a loading control. (H-J) Representative images and the corresponding quantitative results of transwell (H), EdU staining (I), F-actin (J), and mRNA levels of the indicated cytokines (K) in ECs treated with LV-con+BSA, LV-con+PA, or LV-Sirt3+PA (n=3). Scales bars, 50μm. (L) Representative blots of FASN and immunoblotted by Acetyllysine in RAW264.7 cells treated with LV-con+BSA, LV-con+PA, or LV-Sirt3+PA (n=3). (M-O) Representative images and the corresponding quantitative results of EdU staining (M), F-actin (N), and transwell (O) assays of ECs treated with LV-con+PA, LV-Sirt3+PA, or LV-Sirt3+Mal-CoA (20 μM)+PA (n=3). Scales bars, 50μm. (P-S) Representative images and quantitative results of EdU staining (P), transwell assays (Q), F-actin (R), and mRNA levels of the indicated cytokines (S) in ECs treated with LV-con+PA, LV-Sirt3+PA, or LV-Sirt3+si-Cpt-1a+PA (n=3). Scales bars, 50μm. ^*P< 0.05, ^**P< 0.01, ^***P< 0.001, ^****P< 0.0001 compared with LV-con+BSA (A, B, C, E, F, G), and LV-con+PA (H, I, K, M, O, P, Q, S). ^#P< 0.05, ^##P< 0.01, ^###P< 0.001, ^####P< 0.0001 compared with LV-con+PA (A, B, C, E, F, G), and LV-con+Sirt3 (H, I, K, M, O, P, Q, S). BSA: bovine serum albumin; PA: palmitic acid; CⅠ leak: Complex Ⅰ leak; CⅠ OXPHOS: Complex Ⅰ oxidative phosphorylation; CⅠ+Ⅱ OXPHOS: Complex Ⅰ and Complex Ⅱ oxidative phosphorylation; CⅠ+Ⅱ ETS: Complex Ⅰ and Complex Ⅱ electron transport system; CⅡ ETS: Complex Ⅱ electron transport system; ROC: Residual oxygen consumption; O2k: Oxygraph-2k; FASN: fatty acid synthase; ACLY: ATP citrate lyase; ACCS: acetyl-CoA Carboxylase; IL-6: Interleukin 6;; IL-1β: Interleukin 1β; VCAM: Vascular cell adhesion molecule.

Figure 5

Loss of endothelial SIRT3 exacerbated adipocyte dysfunction and pro-inflammatory macrophages polarization in a paracrine way. (A-B) Representative images of lipid droplets formation (A) and mRNA levels of Ucp-1, Pgc-1α, and Cidea (B) in terminally differential brown 3T3-L1 adupocytes treated with the CM of LV-con+BSA, LV-con+PA, or LV-Sirt3+PA-pretreated ECs (n=3). Scales bars, 200μm. β-actin served as a loading control. (C) Representative images of IF staining (left) and quantitative results (right) of CD11c (Green), CD206 (Green), CCR2 (Green), and F4/80 (Red) in BMDMs treated with the CM of LV-con+BSA, LV-con+PA, or LV-Sirt3+PA-pretreated ECs (n=3). Scales bars, 50μm. (D) The mRNA levels of Arg, Il-10, Tnf-α, inos, and Il-1β in BMDMs treated with the CM of LV-con+BSA, LV-con+PA, or LV-Sirt3+PA-pretreated ECs (n=3). β-actin served as a loading control. (E-F) Representative images of cluster analysis of BP (E) and KEGG analysis (F) in CM of ECs treated with LV-con+BSA, LV-con+PA, or LV-Sirt3+PA, detected via proteomics (n=3). The data were normalized to LV-con+BSA. (G) Representative image of Venn diagram of CM from ECs treated with LV-con+BSA, LV-con+PA, or LV-Sirt3+PA, detected via proteomics (n=3). (H-I) Representative images of the Venn diagram (H) and Heat map (I) of secreted proteins (Angiocrine factors) of CM from ECs treated with LV-con+BSA, LV-con+PA, or LV-Sirt3+PA, detected by proteomics (n=3). ^**P< 0.01, ^***P< 0.001, ^****P< 0.0001 compared with the LV-con+BSA (B, C, D). ^##P< 0.01, ^###P< 0.001, ^####P< 0.0001 compared with the LV-con+PA (B, C, D). BSA: bovine serum albumin; PA: palmitic acid; UCP-1: uncoupling protein 1; PGC-1α: Peroxisome proliferator-activated receptor-gamma coactivator 1 α; Cidea: Cell death inducing DFFA like effector A; CM: conditional medium; BMDMs: bone marrow derived macrophages; Arg: Arginase; IL-10: Interleukin 10; iNOS: inducible Nitric oxide synthase.

Figure 6

Endothelial SIRT3 deficiency induced pyroptosis in surrounding cells. (A-B) Representative images of lipid droplets formation and the mRNA levels of Ucp-1, Pgc-1α, and Cidea in terminally differential brown 3T3-L1 adipocytes treated with the CM of BSA, PA, or ZYF (10μM) + PA-retreated ECs (n=3). Scales bars, 200μm. β-actin served as a loading control. (C) The mRNA levels of Arg, Il-10, Tnf-α, inos, and Il-1β in BMDMs treated with the CM of BSA-, PA-, or ZYF+PA-pretreated ECs (n=3). β-actin served as a loading control. (D) Representative images of IF staining (left) and quantitative results (right) of CD11c (Green), CD206 (Green), CCR2 (Green), and F4/80 (Red) in BMDMs treated with the CM of BSA-, PA-, or ZYF+PA-pretreated ECs (n=3). Scales bars, 50μm. (E) The mRNA levels of Gsdmd in terminally differential brown 3T3-L1 adipocytes or BMDMs treated with the CM of LV-con+BSA, LV-con+PA, or LV-Sirt3+PA-pretreated ECs (n=3). β-actin served as a loading control. (F-G) The mRNA levels of Il-1β and Il-18 in BMDMs (F) and terminally differential 3T3-L1 adipocytes (G) treated with the CM of LV-con+BSA, LV-con+PA, or LV-Sirt3+PA-pretreated ECs (n=3). β-actin served as a loading control. (H-I) Representative images of lipid droplets formation (H) and the mRNA levels of Ucp-1, Pgc-1α, and Cidea (I) in si-con or si-Gsdmd-pretreated terminal differential brown 3T3-L1 adipocytes that treated with the CM of PA-pretreated ECs (n=3). Scales bars, 200μm. β-actin served as a loading control. (J) The mRNA levels of Arg, Il-10, Tnf-α, inos, and Il-1β in si-con- or si-Gsdmd-pretreated RAW264.7 cells treated with the CM of PA-pretreated ECs (n=3). β-actin served as a loading control. ^*P< 0.05, ^**P< 0.01, ^***P< 0.001, ^****P< 0.0001 compared with BSA (B, C, D, I, J), LV-con+BSA (I, J, K, M, N). ^#P< 0.05, ^##P< 0.01, ^###P< 0.001, ^####P< 0.0001 compared with PA (B, C, D, I, J), LV-con+PA (E, F, G, I, J). ZYF: Z-YVAD-FMK; GSDMD: Gasdermin D; IL-18: Interleukin 18; BSA: bovine serum albumin; PA: palmitic acid; UCP-1: uncoupling protein 1; PGC-1α: Peroxisome proliferator-activated receptor-gamma coactivator 1 α; Cidea: Cell death inducing DFFA like effector A; IL-1β: Interleukin 1β; IL-18: Interleukin 18.

Figure 7

Working model. Loss of endothelial SIRT3 resulted in vascular rarefaction and impeded angiogenesis via disturbing mitochondrial function and fatty acid metabolism, which then, leading to adipocytes hypertrophy. Moreover, SIRT3 deficiency predisposed ECs to a SASP phenotype that promoted adipocyte dysfunction and pro-inflammatory macrophages polarization in a paracrine way.