Proof-of-concept study for liver-directed miQURE technology in a dyslipidemic mouse model

Abstract

A gene-silencing platform (miQURE) has been developed and successfully used to deliver therapeutic microRNA (miRNA) to the brain, reducing levels of neurodegenerative disease-causing proteins/RNAs via RNA interference and improving the disease phenotype in animal models. This study evaluates the use of miQURE technology to deliver therapeutic miRNA for liver-specific indications. Angiopoietin-like 3 (ANGPTL3) was selected as the target mRNA because it is produced in the liver and because loss-of-function ANGPTL3 mutations and/or pharmacological inhibition of ANGPTL3 protein lowers lipid levels and reduces cardiovascular risk. Overall, 14 candidate miRNA constructs were tested in vitro, the most potent of which (miAngE) was further evaluated in mice. rAAV5-miAngE led to dose-dependent (≤-77%) decreases in Angptl3 mRNA in WT mice with ≤-90% reductions in plasma ANGPTL3 protein. In dyslipidemic APOE∗3-Leiden.CETP mice, AAV5-miAngE significantly reduced cholesterol and triglyceride levels vs. vehicle and scrambled (miSCR) controls when administrated alone, with greater reductions when co-administered with lipid-lowering therapy (atorvastatin). A significant decrease in total atherosclerotic lesion area (-58% vs. miSCR) was observed in AAV5-miAngE-treated dyslipidemic mice, which corresponded with the maintenance of a non-diseased plaque phenotype and reduced lesion severity. These results support the development of this technology for liver-directed indications.

Keywords: AAV; MT: Non-coding RNAs; RNA interference; angiopoietin-like 3 (ANGPTL3); gene therapy; gene-silencing; lipid-lowering; miRNA.

Graphical abstract

Figure 1

Dose-dependent knockdown of ANGPTL3 gene expression by miAngA to N (A and B) Luciferase activity in Huh7 cells co-transfected with LucAng and miAng or miSCR plasmids, and (C) endogenous ANGPTL3 mRNA in Huh7 cells transfected with miAng or miSCR plasmids. For (A) and (B), the relative mean (±SD) luciferase activity was calculated as the ratio between RL and FL activities (n = 3–5). For (C), ANGPTL3 mRNA levels in miAng-transfected cells are expressed as a percentage mean (±SD) relative to miSCR-transfected Huh7 cells (set to 100%, n = 2). FL, firefly luciferase; Luc, luciferase; miRNA, microRNA; RL, Renilla luciferase; SCR, scrambled control.

Figure 2

Vector DNA, Angptl3 mRNA and ANGPTL3 protein quantification in liver or plasma of WT mice treated with vehicle, rAAV5-miAngE or rAAV5-miSCR (A) Vector DNA copy numbers in liver tissue, (B) correlation between Angptl3 mRNA and mature 23 nucleotide miAngE expression levels, and (C) changes in plasma ANGPTL3 protein levels over time in WT mice injected with vehicle control; low-, moderate-, or high-dose rAAV5-miAngE; or rAAV5-miSCR. Low dose, 1 × 10¹³ gc/kg; moderate dose, 5 × 10¹³ gc/kg; high dose, 2.5 × 10¹⁴ gc/kg; control, treatment-naïve plasma. VCN, vector copy number, are shown as genome copies per μg genomic DNA. GC, genome copies; rAAV5, recombinant adeno-associated virus serotype 5; WT, wild type. For (B), Pearson r was computed for X vs. every Y dataset. Pearson correlation calculations were based on the assumption that both X and Y values were sampled from populations that follow a Gaussian distribution. Two-tailed p value and 95% confidence intervals were used (GraphPad Prism 8.0.0). For (C), significant changes vs. vehicle control were determined using one-way ANOVA, Bonferroni post-test (GraphPad Prism 8.0.0). Mean is (± SD). ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001. LLOQ, lower limit of quantification. In (C), one deviating value in the pre-bleed dataset (which is therefore not related to treatment) has been removed to allow the data to be better visualized on the Y scale.

Figure 3

Vector DNA, Angptl3 mRNA and ANGPTL3 protein quantification in liver and/or plasma of dyslipidemic APOE∗3-Leiden.CETP mice treated with vehicle, rAAV5-miSCR or rAAV5-miAngE (with or without atorvastatin) (A) Vector DNA copy numbers in liver, (B) percent change in liver Angptl3 mRNA levels relative to vehicle control, (C) changes in liver ANGPTL3 protein levels vs. vehicle or miSCR controls, and (D) changes in plasma ANGPTL3 protein levels vs. vehicle or miSCR controls in dyslipidemic APOE∗3-Leiden.CETP mice injected with moderate-high dose rAAV5-miAngE or rAAV5-miSCR, with or without atorvastatin. For (A)–(C), analyzed samples are vehicle group n = 20, miSCR group n = 15, miAngE group n = 15, miSCR + statin n = 15, miAngE + statin n = 15. For (D), week 0 analyzed samples vehicle group n = 8, miSCR group n = 15, miAngE group n = 15, miSCR + statin n = 15, miAngE + statin n = 15; for (D), week 2–16 analyzed samples vehicle group n = 14 (except week 16 n = 9), miSCR group n = 9, miAngE group n = 9, miSCR + statin n = 9, miAngE + statin n = 9. The p values for (C) and (D) were calculated using one-way ANOVA, Bonferroni post-test. Moderate-high dose,1 × 10¹⁴ gc/kg of rAAV5-miAngE or rAAV5-miSCR. Control, treatment-naïve plasma; GC, genome copies; LLOQ, lower limit of quantification; rAAV5, recombinant adeno-associated virus serotype 5. Mean is (± SD). ∗p < 0.05; ∗∗p < 0.001; ∗∗∗p = 0.0007; ∗∗∗∗p < 0.0001.

Figure 4

Plasma total cholesterol and triglycerides levels in dyslipidemic APOE∗3-Leiden.CETP mice treated with vehicle, rAAV5-miSCR or rAAV5-miAngE (with or without atorvastatin) Changes over time in mean (±SD) plasma levels of (A) total cholesterol and (B) triglycerides in dyslipidemic APOE∗3-Leiden.CETP mice injected with vehicle control, moderate-high dose rAAV5-miAngE, or rAAV5-miSCR, with and without atorvastatin. The p values were calculated using one-way ANOVA, Bonferroni post-test. Moderate-high dose (1 × 10¹⁴ gc/kg) of rAAV5-miAngE or rAAV5-miSCR. Atorvastatin dose 45 mg/kg rAAV5, recombinant adeno-associated virus serotype 5. Mean is (±SD). ∗p < 0.05 miAngE vs. vehicle; ∗∗p < 0.01 miAngE vs. vehicle; ∗∗∗p < 0.001 miAngE (up to week 12) and miAngE + atorvastatin vs. vehicle; ^###p < 0.001 miAngE and miAngE + atorvastatin vs. miSCR;^##p < 0.01 miAngE and miAngE + atorvastatin vs. miSCR;^δp < 0.05 miSCR+ atorvastatin vs. vehicle; ^δδδp < 0.001 miSCR+ atorvastatin vs. vehicle; ^γp < 0.05 miSCR+ atorvastatin vs. miSCR;^γγp < 0.001 miSCR+ atorvastatin vs. miSCR.

Figure 5

Lipid profiles for cholesterol and triglycerides in dyslipidemic APOE∗3-Leiden.CETP mice treated with vehicle, rAAV5-miSCR or rAAV5-miAngE (with or without atorvastatin) Lipid profiles for (A) cholesterol and (B) triglycerides in dyslipidemic APOE∗3-Leiden.CETP mice 16 weeks after treatment with vehicle control, moderate-high dose rAAV5-miAngE or rAAV5-miSCR, with and without atorvastatin. For (A) and (B), dots represent the mean value of the group per fraction (n = 9 per treatment group). Moderate-high dose (1 × 10¹⁴ gc/kg) of rAAV5-miAngE or rAAV5-miSCR. Atorvastatin dose 45 mg/kg. Tables in (A) and (B) give the area under the curve for fractions 3–8 VLDL, 9–17 LDL, and 18–24 HDL at t = 16 for indicating their cholesterol and triglycerides content, respectively, for each experimental treatment. HDL, high-density lipoprotein; LDL, low-density lipoprotein; rAAV5, recombinant adeno-associated virus serotype 5; TG, triglyceride; VLDL, very low-density lipoprotein.

Figure 6

Atherosclerotic lesion area and lesion severity in dyslipidemic APOE∗3-Leiden.CETP mice treated with vehicle, rAAV5-miSCR or rAAV5-miAngE (with or without atorvastatin) (A) Total atherosclerotic lesion area, (B) severity of atherosclerotic lesions, and (C) number of total lesions per cross-section in dyslipidemic APOE∗3-Leiden.CETP mice 16 weeks after treatment with vehicle control or moderate-high dose rAAV5-miAngE or rAAV5-miSCR, with or without atorvastatin treatment. Severity of atherosclerotic lesions is expressed as a percentage of all segments showing undiseased tissue or type I–V lesions. Representative pictures of a fatty streak, and mild and severe lesions are shown (original magnification ×20). Scale bar, 200 μm. Pink-purple are smooth muscle cells; yellow is connective tissue/collagen; white/translucent are lipid droplets in macrophages and cholesterol clefts. Statistical analyses were performed using a Kruskal-Wallis test followed by individual Mann-Whitney tests for group comparisons. Moderate-high dose (1 × 10¹⁴ gc/kg) of rAAV5-miAngE or rAAV5-miSCR. Atorvastatin dose 45 mg/kg rAAV5, recombinant adeno-associated virus serotype 5. Mean is (±SD). ∗p ≤ 0.05 vs. Vehicle; ∗∗p ≤ 0.01 vs. Vehicle; ∗∗∗p ≤ 0.001 vs. Vehicle; #p ≤ 0.05 vs. miSCR; ##p ≤ 0.01 vs. miSCR; ###p ≤ 0.001 vs. miSCR; &p ≤ 0.05; &&p ≤ 0.01; &&&p ≤ 0.001; ˆp ≤ 0.05; ˆˆp ≤ 0.01; ˆˆˆp ≤ 0.001.