Statins improve cardiac endothelial function to prevent heart failure with preserved ejection fraction through upregulating circRNA-RBCK1

Abstract

Heart failure with preserved ejection fraction (HFpEF) is associated with endothelial dysfunction. We have previously reported that statins prevent endothelial dysfunction through inhibition of microRNA-133a (miR-133a). This study is to investigate the effects and the underlying mechanisms of statins on HFpEF. Here, we show that statins upregulate the expression of a circular RNA (circRNA-RBCK1) which is co-transcripted with the ring-B-box-coiled-coil protein interacting with protein kinase C-1 (RBCK1) gene. Simultaneously, statins increase activator protein 2 alpha (AP-2α) transcriptional activity and the interaction between circRNA-RBCK1 and miR-133a. Furthermore, AP-2α directly interacts with RBCK1 gene promoter in endothelial cells. In vivo, lovastatin improves diastolic function in male mice under HFpEF, which is abolished by loss function of endothelial AP-2α or circRNA-RBCK1. This study suggests that statins upregulate the AP-2α/circRNA-RBCK1 signaling to suppress miR-133a in cardiac endothelial cells and prevent diastolic dysfunction in HFpEF.

Fig. 1. Statins increase circRNA-RBCK1 gene expression in endothelial cells.

A Cultured human umbilical vein endothelial cells (HUVECs) were treated lovastatin (10 μM) for 24 h. Total RNAs were extracted and subjected to perform RNA sequencing analysis. 13 circRNAs were increased in red and 37 circRNAs were decreased in green. B Cultured HUVECs were treated with lovastatin (10 μM), pravastatin (20 μM), and atorvastatin (10 μM) for 24 h. C Primary human endocardium endothelial cells (EEC), human coronary arterial endothelial cells (CAEC), human myocardial capillary endothelial cells (MCEC), and human aortic endothelial cells (AEC) were incubated with lovastatin (10 μM) for 24 h. The levels of circRNA-RBCK1 were measured by quantitative PCR in (B) and (C). All experiments were repeated from five different donors per cell type. A one-way ANOVA followed by the Dunnett test was used to determine P value in (B). A two-sided unpaired Student’s t test was used to determine P value in (C). Data are presented as mean ± SD. Source data are provided as a Source Data file.

Fig. 2. CircRNA-RBCK1 binds to miR-133a in human endocardium endothelial cells (EEC).

A and B Human EECs were transfected with a plasmid expressing circRNA-RBCK1 for 24 h under ox-LDL (100 μg/ml). Total cell lysates were subjected to perform biotinylated-circRNA-RBCK1 pull-down assay followed by quantitative PCR analyses of circRNA-RBCK1 in (A) and top five candidate miRNAs predicted by CircNet database in (B). C The biotinylated wildtype (WT) or mutant (MT) miR-133a was, respectively, transfected into HEK293 cells with circRNA-RBCK1 overexpression. The levels of circRNA-RBCK1 were tested by quantitative PCR after streptavidin capture. D Plasmid of luciferase reporter construction containing circRNA-RBCK1 sequences with WT or mutated miR-133a binding sites (MT1, MT2, MT3, MT1/2/3) was co-transfected with miRNA negative control (miR-NC) or miR-133a in HEK293 cells. The luciferase activities in total cell lysates were assayed. E Human EECs were pretreated lovastatin (10 μM) for 2 h followed by ox-LDL (100 μg/ml) for 24h incubation. FISH was conducted to determine the co-location between circRNA-RBCK1 and miR-133a in human EECs. Scale bar, 5 µM. Red, circRNA-RBCK1; Green, miR-133a; Blue, nucleus. N = 5 per group in (A–D). Representative microscopy image was obtained from five independent experiments in (E). A one-way ANOVA followed by Tukey post-hoc tests was used to determine P value in A and D. A two-sided unpaired Student’s t test was used to determine P value in (B) and (C). Data are presented as mean ± SD. Source data are provided as a Source Data file.

Fig. 3. Lovastatin via circRNA-RBCK1 suppresses the function of miR-133a in human endocardium endothelial cells (EEC).

Human EECs infected with lentivirus harboring scramble shRNA or circRNA-RBCK1 shRNA for 48 h were incubated with ox-LDL (100 μg/ml) plus lovastatin (10 μM) for 24 h. Cells were harvested to assay (A) GTPCH1 mRNA by quantitative PCR, (B) BH4 contents by HPLC, and (C) ROS productions by DHE/HPLC, and (D) NO levels by DAF/HPLC. N = 5 per group. A one-way ANOVA followed by Tukey post-hoc tests was used to determine P value in (A–D). Data are presented as mean ± SD. Source data are provided as a Source Data file.

Fig. 4. AP-2α is a transcriptional factor of circRNA-RBCK1 in human endocardium endothelial cells (EEC).

A–C Human EECs were treated with lovastatin (10 μM), pravastatin (20 μM), and atorvastatin (10 μM) for 24 h. The protein levels of phosphorylated AP-2α at serine 219 (pAP-2α) and total AP-2α were measured in (A). The transcriptional activity of AP-2α was assayed by EMSA in (B) and (C). D Human EECs were treated with lovastatin (10 μM, 24 h). Cells were used for the detection of the affinity of AP-2α to RBCK1 gene promoter by ChIP assay. The representative image was obtained from five independent experiments. E and F HEK293 cells transfected with plasmid of luciferase reporter construction containing wildtype (WT) or mutant (MT) RBCK1 promoter were treated with lovastatin (10 μM, 24h) in (E) or co-transfected with AP-2α cDNA in (F). N = 5 per group. A one-way ANOVA followed by Dunnett test was used to determine P value in (A) and (C). A one-way ANOVA followed by Tukey post-hoc tests was used to determine P value in (E) and (F). Data are presented as mean ± SD. Source data are provided) as a Source Data file.

Fig. 5. Lovastatin upregulates circRNA-RBCK1 to recouple eNOS in human endocardium endothelial cells (EEC), which is AP-2α dependent.

Human EECs were infected with lentivirus expressing scramble shRNA or AP-2α shRNA for 48 h followed by incubation with lovastatin (10 μM) for 24 h in presence or absence of ox-LDL (100 μg/ml). Cells were harvested to assay the levels of circRNA-RBCK1 in (A) and GTPCH1 mRNA in (B) by quantitative PCR, BH4 contents by HPLC in (C), ROS productions by DHE/HPLC, and NO levels by DAF/HPLC in (D). N = 5 per group. A one-way ANOVA followed by Tukey post-hoc tests was used to determine the P value in this figure. Data are presented as mean ± SD. Source data are provided as a Source Data file.

Fig. 6. Lovastatin improves diastolic function in HFpEF mice.

The protocols and experimental designs were described in Supplementary Fig. 5A. A–C Representative pulsed-wave Doppler (top) and tissue Doppler (bottom) tracings were shown in (A). The values of E/E′ in (B) and E/A in (C) were presented. D Hearts isolated from mice were subjected to perform FISH analysis of circRNA-RBCK1 gene expression in the endocardium. Red, circRNA-RBCK1; Green, vWF; Blue, nucleus. Representative microscopy image was obtained from ten mice in each group. E Left ventricle wall was isolated to determine pAP-2α level by western blot. In (B) and (C), N = 10 (control), N = 12 (control + lovastatin), N = 13 (HFpEF and HFpEF + lovastatin), a one-way ANOVA followed by Scheffe tests were used to determine the P value. In (E), N = 10 per group, a one-way ANOVA followed by Tukey post-hoc tests was used to determine the P value. Data are presented as mean ± SD. Source data are provided as a Source Data file.

Fig. 7. Tamoxifen-induced endothelial cell-specific AP-2α knockout eliminates the effects of lovastatin in HFpEF mice.

The protocols were described in Supplementary Fig. 8C. A–C Representative pictures of pulsed-wave Doppler, tissue Doppler, M-mode echocardiographic tracings, HE staining, and Masson staining. Representative microscopy images of HE staining and Masson staining in (A) were obtained from ten mice in each group. The values of E/E′ in (B) and E/A in (C) were presented. D–F Ratio between wet and dry lung weight (LW) in (D), the ratio of heart weight to tibia length (HW/TL) in (E), running distance during exercise exhaustion test in (F) were calculated. N = 10 in per group. A one-way ANOVA followed by Tukey post-hoc tests was used to determine the P value in (B–F). Data are presented as mean ± SD. Source data are provided as a Source Data file.

Fig. 8. AAV9-mediated circRNA-RBCK1 gene knockdown abolishes the effects of lovastatin in HFpEF mice.

The protocols and experimental designs were described in Supplementary Fig. 10A. A–C Representative pulsed-wave Doppler (top) and tissue Doppler (bottom) tracings were shown in (A). The values of E/E′ in (B) and E/A in (C) were presented. D–F Ratio between wet and dry lung weight (LW) in (D), running distance during exercise exhaustion test in (E), and the ratio of heart weight to tibia length (HW/TL) in (F) were calculated. G and H Left ventricle wall was isolated to determine GTPCH1 mRNA by quantitative PCR and BH4 contents by HPLC. N = 11 (AAV9-NC alone), N = 12 (AAV9-NC + lovastatin), N = 13 (circRNA shRNA alone), N = 12 (circRNA shRNA + lovastatin. A one-way ANOVA followed by Scheffe tests was used to determine the P value in (B–H). Data are presented as mean ± SD. Source data are provided as a Source Data file.