FoxOs integrate pleiotropic actions of insulin in vascular endothelium to protect mice from atherosclerosis

Abstract

Atherosclerotic cardiovascular disease is the leading cause of death in insulin-resistant (type 2) diabetes. Vascular endothelial dysfunction paves the way for atherosclerosis through impaired nitric oxide availability, inflammation, and generation of superoxide. Surprisingly, we show that ablation of the three genes encoding isoforms of transcription factor FoxO in endothelial cells prevents atherosclerosis in low-density lipoprotein receptor knockout mice by reversing these subphenotypes. Paradoxically, the atheroprotective effect of FoxO deletion is associated with a marked decrease of insulin-dependent Akt phosphorylation in endothelial cells, owing to reduced FoxO-dependent expression of the insulin receptor adaptor proteins Irs1 and Irs2. These findings support a model in which FoxO is the shared effector of multiple atherogenic pathways in endothelial cells. FoxO ablation lowers the threshold of Akt activity required for protection from atherosclerosis. The data demonstrate that FoxO inhibition in endothelial cells has the potential to mediate wide-ranging therapeutic benefits for diabetes-associated cardiovascular disease.

Figure 1. Insulin signaling in aortae of WTD-fed Ldlr^−/− mice

(A) Glucose tolerance tests. (B-C) Representative Oil Red-O staining of en face aorta preparations (B) and quantification of lesion area (C) (n=3-4 for each genotype). (D-E) Representative immunoblots (D) and quantification (E) of insulin signaling, eNOS, and Erk1/2 in aortae isolated 5 min after intravenous insulin injection from mice fed standard (SD) or Western diet (WTD) for 6 weeks. We determined band intensity from three independent experiments. (F-J) We carried out similar experiments in mice fed SD or WTD for 20 weeks. Data indicate means ± SEM. * P< 0.05, ** P< 0.01, *** P< 0.001 vs. chow.

Figure 2. Glucose metabolism, vascular function, and atherosclerosis in VECKO; Ldlr^−/− mice

(A-B) Foxo1, 3a, and 4 mRNA (A) or FoxO1 and 3a protein levels (B) in cultured aortic EC from VECKO and control mice. (C-D) Glucose (C) and insulin tolerance tests (D) in VECKO; Ldlr^−/− and Ldlr^−/− mice after 18-19 weeks on WTD (n=7-10). (E-F) Acetylcholine– (E) and Na nitroprusside–induced (F) vasorelaxation in femoral arteries from VECKO; Ldlr^−/− and Ldlr^−/− mice after 18 weeks on WTD (n=5). Data indicate means ± SEM. * P< 0.05, ** P< 0.01, *** P< 0.001 vs. WT or Ldlr^−/−.

Figure 3. Aortic and coronary immunohistochemistry

(A-B) Representative en face Oil Red-O staining (A) and quantification (B) of lesion area in total, aortic arch, and descending thoracic aortae in VECKO; Ldlr^−/− and Ldlr^−/− mice after 20 weeks WTD (n=11-13). (C) Histological analysis of the brachiocephalic and thoracic aorta with hematoxylin and eosin (HE) and Masson-Trichrome staining. Values indicate quantification of atherosclerotic lesions and positive area as % of total area, respectively (n=4). (D) Immunostaining of thoracic aorta with α-smooth muscle actin and Mac-3 antibodies. (E) Immunostaining of thoracic aorta with Vcam-1 antibody. Values indicate quantification of immunoreactive areas as % of total area (n=4). (F) Histological analysis of coronaries by HE staining. Values indicate plaque-positive coronaries as % of total coronaries scored in 4 independent sections of each mouse (n=8). Data indicate means ± SEM.* P< 0.05, ** P< 0.01, and *** P< 0.001 vs. Ldlr^−/−. Scale bars: 200μm

Figure 4. eNOS expression and NO production in aortic EC and aorta

(A) eNos levels in cultured aortic EC from VECKO and control mice (n=4). (B-C) Representative immunoblots (B) and quantification (C) of p-eNOS-S1176 and total eNOS in cultured aortic EC from VECKO and control mice following insulin treatment (n = 3). The loading control is shown in Figure 6A. (D-E) Representative immunoblots (D) and quantification (E) of eNOS dimer and monomer in cultured aortic EC from VECKO and control mice analyzed by low-temperature SDS-PAGE immunoblotting (n = 4). (F-G) Representative immunoblots (F) and quantification (G) of p-eNOS-S1176 and total eNOS in cultured aortic EC from VECKO and control mice (n = 3). The loading control of Akt is shown in Figure 6A. Scale bar: 200μm. (H-I) Insulin-stimulated NO production in cultured aortic EC from VECKO and control mice pretreated with L-NAME (0.5mM) or vehicle. NO production visualized by DAF2-DA fluorescence (H) and quantified in (I). Data indicate means ± SEM. AU= arbitrary mRNA units. * P < 0.05, ** P < 0.01, and *** P < 0.001 vs. controls.

Figure 5. Pro-inflammatory cytokine expression, NF-κB signaling, and oxidative stress in aortic or lung EC

(A) Time courses of Icam1, Vcam1, Mcp1, iNos, IL-1β, and IL-6 induction in aortic EC from VECKO and control mice. Basal (without LPS stimulation) gene expression in control cells is set to 1. (B) Immunoblot analysis of lung EC from VECKO and control mice following stimulation with LPS for the indicated times. (C) NF-κB luciferase activity in lung EC from VECKO and control mice transduced with adenovirus encoding Nf-κB–responsive luciferase reporter construct followed by LPS stimulation (n=4). (D-E) Free cholesterol (FC)- and H₂O₂-stimulated reactive oxygen species (ROS) production in aortic EC from VECKO and control mice visualized by DAF2-DA fluorescence (D) and quantified from 4 independent wells in (E). (F-G) H₂O₂-stimulated superoxide production in aortic EC from VECKO and control mice visualized by DHE fluorescence (F) and quantified in (G). (H) Expression of NADPH oxidase subunits in aortic EC from VECKO and control mice (n=4). Data indicate means ± SEM. AU= arbitrary mRNA units. * P < 0.05, ** P < 0.01, and *** P < 0.001 vs. controls. Scale bar: 250μm.

Figure 6. Insulin signaling in aortic EC of VECKO mice

(A-B) Representative immunoblots (A) and quantification (B) of insulin signaling in cultured aortic EC from VECKO following insulin treatment. (C) Irs1, Irs2 and Igf1r expression in cultured aortic EC from VECKO and control mice. AU= arbitrary mRNA units. (D) Illustration of conserved putative FoxO-binding sequences (red) in the Irs1 promoter and primers used for amplification of region A and B during ChIP. TSS: transcription start site. (E) ChIP assay of cultured aortic EC using immunoprecipitation (IP) with rabbit IgG or antibody against FoxO1 and RNA polymerase (pol) II. Size of predicted PCR products sizes is 88 and 208 bp, respectively. The asterisk on the gel indicates a non-specific band. Data indicate means ± SEM. * P < 0.05, ** P < 0.01, and *** P < 0.001 vs. controls.