Cardiac gene therapy treats diabetic cardiomyopathy and lowers blood glucose

Figures

• 

  Figure 1. cBIN1 is downregulated in diabetic mouse hearts and can be rescued by AAV9-cBIN1 to improve exercise capacity and systemic blood glucose control. (A) Schematic illustration of experimental protocol used. (B) Western blots of cBIN1 and GAPDH with quantification (cBIN1/GAPDH) in heart lysates from posttreatment mice (n = 15 per group). (C–F) Bar graphs of plasma cBIN1 score (CS) values (n = 11–16, C), maximal running distance (n = 16–28, D), nonfasting plasma glucose (n = 9–17, E), and insulin (n = 9–16, F) in 9-week-old pretreatment (left) and 17-week-old posttreatment (right) db/db and littermate control db/m mice. (G) Percent of peak blood glucose during iGTT in AAV9-treated mice at 17 weeks of age (n = 4–6 animals per group). (H) Percent of baseline blood glucose during iITT in AAV9-treated mice at 17 weeks of age (n = 4–6 animals per group). (I) qPCR analysis of V5-tagged exogenous cBin1 (V5 normalized to the housekeeping gene Hprt1 then compared with noninjected control organs) in indicated organs from db/db mice injected with AAV9-cBIN1-V5. SKM, skeletal muscle; Pan, pancreas; K, kidney; Sp, spleen; Adi, adipose tissue; n = 4 animals per group. Data are presented as mean ±…

• 

  Figure 2. Exogenous cBIN1 normalizes membrane microdomains at t-tubules. (A) Representative transmission electron microscopy images of posttreatment hearts (scale bar: 1 μm) from each group (left). Quantification of the degree of contour of t-tubules (n = 100–101 t-tubules from 16–40 images of 2–3 myocardial sections and 3 hearts from each group). Data are presented as percentage of t-tubules. χ2 test was used to compare t-tubule contour between groups. *** indicates P < 0.001 for db/m + GFP versus db/db + GFP; ††† indicates P < 0.001 for db/db + GFP versus db/db + cBIN1. (B) Western blots of cBIN1, SERCA2a, CaV1.2, RyR2, GLUT4, and IRAP in total cardiac microsome and sucrose-gradient isolated TT/jSR fraction (F4) from each group (n = 5 hearts per group). (C) Representative spinning disc confocal images of posttreatment mouse myocardium with power spectrum analysis of boxed areas, and quantification of SERCA2a peak power density at t-tubules (n = 35–44 cells from 3 hearts per group). Nonparametric Kruskal-Wallis test followed by Dunn’s test was used for comparison between selected pairs. *, **, *** indicates P < 0.05, 0.01, and 0.001, respectively, for comparison versus db/m + GFP; †, ††,…

• 

  Figure 3. Exogenous cBIN1 normalizes GLUT4 expression at t-tubules. (A and B) Representative GLUT4 and IRAP confocal images in posttreatment isolated cardiomyocytes with corresponding fluorescence intensity profiles (A), and quantification of GLUT4 and IRAP peak power density (B) at t-tubules (n = 42–58 cells from 3 hearts). Scale bar: 10 μm. Data are presented as mean ± SEM. Kruskal-Wallis test followed by Dunn’s test was used for comparison between selected pairs. *, *** indicates P < 0.05, 0.001, respectively, for comparison versus db/m + GFP; ††† indicates P < 0.001 for comparison between db/db + GFP and db/db + cBIN1. (C and D) Intracellular distribution of GLUT4 and IRAP (peak power density analysis, n = 34–61 cells from 3 hearts) (C) and glucose uptake (D) following insulin (0 and 10 nM) stimulation in cardiomyocytes isolated from each group (n = 10–12 repeats from 4 animals per group). Metformin (Met; 20 μM) was included as a positive control treatment. Two-way ANOVA followed by Tukey’s (3 or more treatment groups) or Bonferroni’s test (2 insulin doses) was used for multiple comparison among groups. **, *** indicates P < 0.01, 0.001, respectively, for comparison versus db/m + GFP…

• 

  Figure 4. Bin1 deletion in cardiomyocytes impairs GLUT4 translocation and glucose utilization, weakening systemic glucose response to insulin in mice. (A) Representative GLUT4 confocal images in isolated cardiomyocytes from WT and Bin1 HT mice with or without prior AAV9-cBIN1 rescue. Cardiomyocytes were treated with 0 or 10 nM insulin for 20 minutes before collection. The images of the boxed areas with the corresponding fluorescence intensity profiles are included in the bottom panels. Scale bar: 10 μm. (B) Quantification of GLUT4 peak power density at t-tubules (n = 30–47 cells from 3 hearts per group). (C) 2DG uptake following insulin (0 and 10 nM) stimulation in cardiomyocytes isolated from each group (n = 10–12 repeats from 3 animals per group). (D) 13C enrichment of extracellular lactate and intracellular glycolysis and TCA cycle intermediates from cardiomyocytes isolated from each group (n = 3–5). Schematic for metabolic pathway of 13C6-glucose is included. Pyr., pyruvate; Lac., lactate; Glc, glucose; F6P, fructose-6-phophate; F1,6BP, fructose-1,6-biphosphate; DHAP, dihydroxyacetone phosphate; 3PG, 3-phosphoglyceric acid; PEP, phosphoenolpyruvate; Cit., citrate; α-KG,…

• 

  Figure 5. AAV9-cBIN1 normalizes cardiac proteomics in diabetic mice. (A) Bar graphs of LFQ LC-MS/MS data (fold changes over control db/m mice) of SERCA2a, RyR2, GLUT4, and IRAP from the db/m mice treated with AAV9-GFP (n = 4 hearts) and db/db mice treated with AAV9-GFP (n = 4 hearts) or cBIN1 (n = 3 hearts). (B) PCA plot of all 3 groups (n = 2 repeats/heart × 3–4 hearts/group) generated based on LFQ LC-MS/MS proteomics. All data are presented as mean ± SEM. One-way ANOVA followed by Bonferroni’s test or Kruskal-Wallis test followed by Dunn’s test was used for comparison between the selected pairs. * indicates P < 0.05 for comparison versus db/m + GFP; † indicates P < 0.05 for comparison between db/db + GFP and db/db + cBIN1.

• 

  Figure 6. Exogenous cBIN1 rescues diastolic heart failure in diabetic mice. (A) Representative images of longitudinal axis view of left ventricles at end diastolic (top) and systolic (bottom) phases in posttreatment mice (17 weeks of age). Quantification of end-diastolic volume (EDV), cardiac output (CO), and ejection fraction (EF) from each group are included in the bar graphs below (n = 12–17). (B) Representative images of mitral valve inflow pulsed wave Doppler (top) and septal mitral valve annulus tissue Doppler (bottom) in posttreatment mice. Quantification of E/A, E/e’, and isovolumic relaxation time (IVRT) from each group are included in the bar graphs below (n = 11–17). (C and D) The ratio of heart weight (C) and lung weight (D) over tibial length after treatment (n = 13–17). All data are presented as mean ± SEM. One-way ANOVA followed by Bonferroni’s test or Kruskal-Wallis test followed by Dunn’s test was used for comparison between selected pairs. *, **, *** indicates P < 0.05, 0.01, 0.001, respectively, for comparison versus db/m + GFP; †, ††, ††† indicates P < 0.05, 0.01, 0.001, respectively, for comparison between db/db + GFP and db/db + cBIN1.

• 

  Figure 7. AAV9-cTnT-cBIN1 improves systemic glucose tolerance in diabetic mice. Percent of peak blood glucose during iGTT in AAV9-cTnT-GFP/cBIN1–treated mice following an i.p. injection of glucose (1 g/kg) after 12-hour fasting (n = 10 mice per group). Data are presented as mean ± SEM. Two-way ANOVA followed by Tukey’s for multiple comparisons among groups is used. *, **, *** indicates P < 0.05, 0.01, 0.001 for comparison versus db/m + GFP; †† indicates P < 0.01 for comparison between db/db + GFP and db/db + cBIN1.

• 

  Figure 8. I.v. AAV9-cBIN1 (6 × 1011 vg/kg) transduces exogenous cBIN1 in minipig hearts. (A) Cardiac expression of exogenous cBin1-V5 (ΔΔCt of V5/HPRT1 when compared with PBS controls) in minipig hearts 6 months after AAV9-cBIN1 injection. (B) Exogenous cBin1-V5 as percentage of endogenous porcine cBIN1 (derived from ΔCt of V5/cBIN1) in minipig hearts 6 months after AAV9-cBIN1 injection (n = 10 tissue samples across left ventricle obtained from 2 minipigs per group). Data are presented as mean ± SEM. Unpaired 2-tailed Student’s t test or Mann-Whitney U test was used. *** indicates P < 0.001 for comparison versus PBS controls.