The longevity-associated BPIFB4 gene supports cardiac function and vascularization in ageing cardiomyopathy

Figures

• 

  Figure 1. Immunohistochemical characterization of human hearts. (A–B) Expression of BPIFB4 in cardiomyocytes (A) and endothelial cells (B) from controls and IHF hearts. (C–E) Microvascular alterations in IHF hearts. Capillary density is decreased in IHF compared with control hearts (C), whereas the reduction in arteriole density did not reach a statistical significance (D). PC density and coverage are lower in hearts explanted from elderly patients with IHF (E). PCs stained with PDGFRβ (red), endothelial with vWF or CAV1 (green) or CD34 (light blue) and cardiomyocytes with α-sarcomeric actin (αSA, green or white). Nuclei are identified by DAPI (blue) and BPIFB4 expression labelled in red. n = 8–9 C hearts and 23-22 IHF hearts. Data were analyzed using the Mann–Whitney U test (panels B, D, and E, pericyte coverage IHF vs. C) or unpaired Student’s t-test (all the other panels).

• 

  Figure 2. Effect of forced BPIFB4 titration on human cardiac pericytes. (A–H) Effects of exposing aged IHF-PCs to recombinant LAV-BPIFB4, WT-BPIFB4, or vehicle (V). The LAV-BPIFB4 protein reduced the proportion of Ki67neg and γH2AXpos senescent IHF-PCs (Ki67 stained green and γH2AX red) (B, n = 7 per group) and the levels of oxidized lipofuscin (green) (C, n = 4 per group). In contrast, the WT-BPIFB4 protein was ineffective in improving these endpoints. Both isoforms were similarly effective in reducing the abundance of mitochondrial O2− radicals (D, n = 6/group) and increasing the levels of nuclear VDR [(E), n = 8/group]. Both markers are stained red in G and H, respectively. In all the panels, nuclei are stained blue by DAPI. (F–H) Transfection of C-PCs with BPIFB4 siRNA or Scramble control. Effective reduction in BPIFB4 expression by siRNA (G, n = 6 per group) was associated with a significantly increased rate of Ki67neg and γH2AXpos senescent cells (H, n = 7 per group). Ki67 stained green and γH2AX red. Data were analyzed using ANOVA followed by Tukey's multiple comparisons test (B, C, D and H) or Kruskal–Wallis followed by Dunn's multiple comparisons test (all the other panels).

• 

  Figure 3. LAV-BPIFB4 enhances the ability of endothelial cells to form vascular networks in vitro. (A) Schematic of the BPIFB4 protein supplementation to P3 and P10 HUVEC. (B) Bar graph showing passaging causes proliferative senescence in HUVECs, as assessed by β-galactosidase activity (n = 3/group, Unpaired t-test). (C) Phase-contrast microscopy image and a bar graph showing LAV-BPIFB4 induces network formation by early and late passage HUVECs (n = 4/group). Data were analyzed using ANOVA followed by Tukey's multiple comparisons test). (D) Schematic of the experiment where the network formation assay was performed with HUVECs and aged IHF-PCs. The latter were conditioned in advance with BPIFB4 recombinant proteins. (E) Representative phase-contrast images and a bar graph show that aged IHF-PCs conditioned with LAV-BPIFB4 promoted networking by early and late passage HUVECs. n = 4 biological replicates/group. Data were analyzed using ANOVA followed by Tukey's multiple comparisons test). (F) Bar graph showing the levels of released angiogenic factors in the media of aged IHF-PC treated with LAV-BPIFB4 protein or vehicle. Data are expressed as fold change vs. vehicle. n = 4 biological…

• 

  Figure 4. BPIFB4 promotes ribosome biogenesis. (A–D) Effect of BPIFB4 protein supplementation on 47S levels in PCs. (A) Basal levels of 47S in C-PCs and IHF-PCs (n = 7/group, Unpaired t-test). (B) Schematic of the conditioning experiment. (C–D) Supplementation of the LAV-BPIFB4 protein did not change the 47S levels in C-PCs [(C), n = 5/group] while rescuing the 47S deficit in aged IHF-PC [(D), n = 5/group]. ANOVA followed by Tukey's multiple comparisons test. (E–F) Effect of BPIFB4 silencing by siRNA on 47S in C-PC. (F) Depletion of BPIFB4 reduced the level of 47S in C-PC compared with scrambles compared with scramble (n = 5/group, Unpaired t-test). (G–H) aged IHF-PC had smaller nucleoli than C-PC (n = 4/group, Unpaired t-test). (I–J) This defect was rescued by LAV-BPIFB4 protein (n = 4/group, ANOVA followed by Tukey's multiple comparisons test). Nucleoli were stained for fibrillarin (red) and nuclei with DAPI (blue).

• 

  Figure 5. LAV-BPIFB4 physically interacts with NCL improving angiogenesis. The binding between BPIFB4 and NCL was determined using co-immunoprecipitation assays. (A) Cartoon showing the BPIFB4 transfection groups. (B) Lysates from Hek-293 transfectants expressing the BPIFB4 isoforms were immunoprecipitated with anti-BPIFB4, resolved by SDS-PAGE (10%), and probed with anti-NCL antibody (left panel). Lysate aliquots were loaded to verify transfection and immunoprecipitation efficiencies (right panel). (C) The subcellular localization of exogenous LAV-BPIFB4 and endogenous NCL in transfected Hek-293 was determined using double staining immunofluorescence using anti-BPIFB4 polyclonal (red) and anti-NCL monoclonal (green) antibodies. White squares point to the punctate area of colocalization between LAV-BPIFB4 and NCL. (D) Schematic representation of BPIFB4 constructs used for transfection. (E) Lysates from Hek-293 transfectants expressing deleted forms of LAV-BPIFB4 were immunoprecipitated with anti-BPIFB4 antibody, resolved by SDS-PAGE (10%), and probed with indicated antibodies (upper panel). (F) Schematic of the experiment in which IHF-PCs were silenced with siRNA against NCL or scramble…

• 

  Figure 6. Early LAV-BPIFB4 gene therapy improves cardiac function in elderly mice. (A) Cardiac function was assessed in female mice at baseline (14 months old) and 4 months post-treatment (18 months old). n = 10 female and 3 male mice/group. (B) Body weight at baseline and the end of follow-up. (C–I) Functional parameters are all expressed as fold change from baseline except for Ejection Fraction which is illustrated as absolute change from baseline. Heart rate (HR) (C), left ventricular mass (D), stroke volume (SV) (E), ejection fraction (EF) (F), cardiac output (CO) (G), cardiac index (CI) (H), and E/A ratio (I). Bar graphs show combined data for male and female mice, including the mean, standard deviation, and individual values. (J–L) Histological analysis of the vascular density in hearts harvested at the end of the follow-up. Representative images of isolectin B4 (green) positive endothelial cells, α-smooth muscle actin (red) positive smooth muscle cells (J), and PDGFRβ (red) positive PCs (K–L) with nuclei identified by DAPI (blue) in heart sections from mice attributed to the 4 groups. Scale bars: 50 μm. Graphs showing the density of capillaries and arteries (J), and pericyte…

• 

  Figure 7. Late LAV-BPIFB4 gene therapy improves cardiac function in elderly mice. (A) Cardiac function was assessed in female mice at baseline (18 months old) and four weeks post-treatment (19 months old). n = 10 female mice per group. (B) Body weight. (C–I) Fold changes in functional parameters from basal measurements except for Ejection Fraction, which is expressed as absolute unit change. Heart rate (HR) (C), left ventricular mass (D), stroke volume (SV) (E), ejection fraction (EF) (F), cardiac output (CO) (G), cardiac index (CI) (H), and E/A ratio (I). (J–K) Graphs show capillaries (J) and arteries (K) density. n = 6 mice per group. (L–M) Graphs showing data of the density of β-galactosidase (L) and p16ink4A (M) positive cells. n = 5 to 10 mice per group. (N) Cardiac fibrosis was assessed after staining with Azan Mallory in female mice at 1-month post-treatment (19 months old). n = 6–7 mice per group. (O) Representative images of PET imaging were performed in subgroups of the early and late intervention studies. Representative images. The bar graph shows the data from basal and Dobutamine stress tests. n = 5 mice per group. Values are presented as mean, standard deviation, and…