Gene Therapy Targeting LDL Cholesterol but not HDL Cholesterol Induces Regression of Advanced Atherosclerosis in a Mouse Model of Familial Hypercholesterolemia

Abstract

A reduction in low density lipoprotein (LDL) cholesterol or an increase in high density lipoprotein (HDL) cholesterol can reduce the risk of development of atherosclerosis through overlapping or independent mechanisms. However, the clinical outcome of combined therapy remains in debate. In this study, we first characterized effects of various constructs of helper-dependent adenoviral vector (HDAd) expressing apolipoprotein E3 or LDL receptor (LDLR) in vivo on plasma cholesterol levels. Using this information, we designed experiments and compared the effects of long-term (28 weeks) LDL cholesterol lowering or raising HDL cholesterol, or a combination of both on advanced atherosclerosis in Ldlr(-/-) mice, a mouse model of familial hypercholesterolemia. Our major findings are: (i) various factors influence in vivo functional activity, which appear to be context dependent; (ii) apolipoprotein AI (APOAI) gene transfer, which raises HDL cholesterol, retards progression of atherosclerosis but does not induce regression; (iii) LDLR or LDLR and APOAI combination gene therapy induces lesion regression; however, LDLR gene transfer accounts for the majority of the effects of combined gene therapy; (iv) LDLR gene therapy reduces interleukin-7, which is a master regulator of T-cell homeostasis, but APOAI gene therapy does not. These results indicate that LDL cholesterol lowering is effective and sufficient in protection against atherosclerosis and induction of regression of pre-existing atherosclerosis.

Figure 1

Structure of helper-dependent adenoviral vectors expressing APOE3. A. Schematic presentation of helper-dependent adenoviral vectors (HDAd) containing phosphoenolpyruvate carboxykinase (PEPCK) promoter on C4HSU backbone. L-ITR and R-ITR indicate left and right adenovirus (Ad) inverted terminal repeat sequence, respectively; Ψ, Ad packaging signal; PEPCKpr, PEPCK promoter; AI intron, human apolipoprotein AI intron, hGH polyA, human growth hormone polyadenylation signal; WPRE, Woodchuck hepatitis virus post-transcriptional regulatory element; LCR, human APOE gene liver control region. B. HDAd containing human APOE gene promoter and liver specific enhancer, LCR. In HDAd-E-E3, the fragment containing exon 2–4 was removed and replaced with APOE3 cDNA. HDAd-gE3 contains APOE3 5′ flanking region as well as all exons. C. HDAd containing human APOAI promoter and 3′ flanking region. In HDAd-AI-E3, the APOAI gene corresponding to exon 2 to exon 4 was removed and replaced with APOE3 cDNA. D. HDAd containing human APOE3 gene and LCR on pΔ21 backbone. HPRT, intron region of human genomic HPRT; C346, cosmid C346 human genomic stuffer sequence.

Figure 2

Structure of HDAd expressing low-density lipoprotein receptor (LDLR). A. HDAd vector containing LDLR cDNA on pΔ28 or pΔ21 backbone. HDAd-P-LDLR has the pΔ28 backbone and HDAd-AI-LDLR and HDAd-AI-LDLR-E4 have the pΔ21 backbone. AdE4, Ad E4 promoter. Other abbreviations are as described in Figure 1 legend. B. HDAd containing human LDLR gene. The intron 1 of LDLR gene was removed and replaced with cDNA.

Figure 3

Plasma cholesterol levels. Six to eight week old Apoe^−/− mice received i.v. injection of 5 × 10¹² v.p./kg of HDAd expressing apoE3 and plasma cholesterol was measured over the following 32 weeks. A. PEPCK expression cassette. *p<0.05 vs. PBS, **p<0.001 vs. PBS, †p<0.01 vs. HDAd-P-E3, ‡p<0.05 vs. HDAd-P-E3 and HDAd-PW-E3. All time points in the HDAd-PW-E3 and HDAd-PWL-E3 groups were significantly different from the PBS group after vector treatment (p<0.001), but not indicated. B. APOAI expression cassette. p<0.001 vs. PBS, group, except in the HDAd-AI-E3 group 1 week after treatment. There was no statistical significance among treatment groups. C. APOE expression cassette. Plasma cholesterol levels in all treatment groups were significantly lower than those in the PBS group (p<0.001) except 1 week after treatment in the HDAd-E-E3 and HDAd-ghE3 groups. *p<0.05 vs. HDAd-EW-E3. **p<0.01 vs. HDAd-EW-E3 and HDAd-gE3. D. Plasma cholesterol levels in mice treated with HDAd-gE3 on pΔ21 backbone. *p<0.05 vs. HDAd-gE3, **p<0.01 vs. HDAd-gE3. n=5/group.

Figure 4

Human ApoE3 levels in Apoe^−/− mice after treatment with HDAd. A. PEPCK expression cassette. *p<0.05 vs. HDAd-P-E3 (n=5/group). B. APOAI cassette. *p<0.05 vs. HDAd-AIW-E3 at 2 weeks, **p<0.01 vs. HDAd-AIW-E3 and HDAd-AIWL at 1 week. C. APOE expression cassette. *p<0.05 vs. HDAd-EW-E3 and HDAdgE3, **p<0.01 vs. HDAd-EW-E3 and HDAd-gE3, and †p<0.05 vs. HDAd-EW-E3. D. Plasma apoE3 levels in mice treated with HDAd-gE3-D21. *p<0.05 vs. HDAd-gE3, **p<0.01 vs. HDAd-gE3.

Figure 5

En face quantification of aortic atherosclerotic lesions. 1. PBS (n=5); 2. HDAd-0 (n=4); 3. HDAd-P-E3 (n=3); 4. HDAd-PW-E3 (n=5); 5. HDAd-PWL-E3 (n=4); 6. HDAd-AI-E3 (n=2); 7. HDAd-AIW-E3 (n=5); 8. HDAd-AIWL-E3 (n=4); 9. HDAd-E-E3 (n=3); 10. HDAd-EW-E3 (n=4); 11. HDAd-gE3 (n=5), 12. HDAd-gE3 (n=5). HDAd-0 was on pΔ28 backbone and HDAd-gE3 was on pΔ21 backbone. All other HDAd vectors were on C4HSU backbone. *p<0.05 vs. PBS, **p<0.01 vs. HDAd-gE3. Mann-Whitney Rank Sum Test was used. Median, 25^th percentile and 75^th percentile values are indicated.

Figure 6

Plasma cholesterol levels in Ldlr^−/− mice after hepatic LDLR gene transfer. 6–8 week old Ldlr^−/− mice on high cholesterol diet were treated with a single injection of HDAd expressing LDLR and plasma cholesterol was measured at indicated time. All data are presented as mean ± SD. A. Effects of expression cassette and Ad E4 promoter. HDAd vectors are on pΔ21 or pΔ28 backbone. *p<0.01, **p<0.001 vs. PBS. n=5/group. At 12 weeks, plasma cholesterol levels in the HDAd-P-LDLR group were no longer different from those of the PBS group. B. Incorporation of LDLR gene structure into HDAd. n=3.

Figure 7

Comparison of single gene therapy and combined gene therapy on plasma cholesterol. A. Experimental design. 6–8 week old female Ldlr^−/− mice were fed a diet containing 0.2% (w/w) cholesterol and 10% (v/w) coconut oil for 36 weeks to induce atherosclerosis. Mice were separated in 5 groups. One group (baseline) was sacrificed and other 4 groups received single i.v. injection of HDAd vector. A high cholesterol diet was maintained throughout experiment. n=13 (baseline), 13 (HDAd-0), 14 (HDAd-LDLR), 12 (HDAd-AI) and 15 (HDAd-LDLR + HDAd-AI). B. Plasma cholesterol levels. *p<0.01, **p<0.001 vs. HDAd-0. C. FPLC analysis. Plasma collected 28 weeks after vector treatment. 0.2 ml of pooled plasma was separated with FPLC and cholesterol was determined. VLDL: very low density lipoprotein; IDL/LDL: intermediate-density-lipoprotein/low density lipoprotein; HDL: high density lipoprotein.

Figure 8

LDLR gene therapy induces regression of advanced atherosclerosis. A. En face quantification of aortic lesions. Atherosclerotic lesions were stained with Oil Red-O and quantified by morphometric analysis. Data were analyzed by Mann-Whitney Rank Sum Test. Median, 25^th percentile and 75^th percentile values are indicated. *p<0.05 (vs. HDAd-0), **p<0.001 (vs. baseline), ***p<0.001 (vs. baseline, HDAd-0, HDAd-AI). n=13 (baseline), 13 (HDAd-0), 14 (HDAd-LDLR), 12 (HDAd-AI), and 15 (HDAd-LDLR + HDAd-AI). Representative staining of aortas is shown in right panel.

Figure 9

Immunological features in atherosclerotic lesions. A. Mac3 (macrophage) staining. *p<0.01 vs. baseline, HDAd-0 and HDAd-AI. Data are presented as mean ± SD. n=5/group. B. Representative Mac3 immunoreactivity in cross-section of aorta 28 weeks after treatment. Bar, 100 μm. C. Relative α-actin immuno positive areas. *p<0.01 vs. baseline and HDAd-0, and p<0.05 vs. HDAd-AI. **p<0.001 vs. baseline, HDAd-0 and HDAd-AI. n=4 (baseline), 3 (HDAd-0), 4 (HDAd-LDLR), 4 (HDAd-AI) and 4 (HDAd-LDLR+HDAd-AI). D. Representative α-actin immuno staining. Bar, 100 μm.

Figure 10

IL-7 immunoreactive areas are reduced in mice treated with HDAd-LDLR. A. Relative IL-7 immunoreactive area in aorta. *p<0.05 vs. baseline and HDAd-AI, and p<0.01 vs. HDAd-0. n=6 (baseline), 6 (HDAd-0), 4 (HDAd- LDLR), 3 (HDAd-AI), and 3 (HDAd-LDLR+HDAd-AI). Bar represents mean value. B. Representative staining of aortas. Bar, 100 μm.