Helper-dependent adenoviral vector achieves prolonged, stable expression of interleukin-10 in rabbit carotid arteries but does not limit early atherogenesis

Abstract

Vascular gene therapy could potentially complement or replace current therapies for human atherosclerosis, while avoiding their side effects. However, development of vascular gene therapy is limited by lack of a useful vector. Helper-dependent adenovirus (HDAd) shows promise to overcome this barrier because, unlike first-generation adenovirus, HDAd achieves durable transgene expression in the artery wall with minimal inflammation. To begin to test whether HDAd, delivered to the artery wall, can limit atherosclerosis we constructed HDAd that expresses rabbit interleukin (IL)-10, a potent atheroprotective cytokine, and tested its activity in a rabbit model of early carotid atherogenesis. HDAd expressed immunoreactive, active IL-10 in vitro. In contrast to other HDAd-expressed transgenes, IL-10 expression from HDAd increased significantly between 3 days and 2 weeks after infusion and remained stable for at least 8 weeks. Rising, persistent IL-10 expression was associated with relative persistence of HDAdIL-10 genomes 4 weeks after infusion, compared with HDAdNull genomes. Surprisingly, IL-10 expression had no significant effects on atherosclerotic lesion size, macrophage content, or expression of either adhesion molecules or atherogenic cytokines. These results might be due to inadequate protein expression in vivo or lack of suitability of this rabbit model to reveal IL-10 therapeutic effects. IL-10 remains a promising agent for vascular gene therapy and HDAd remains a promising vector; however, proof of efficacy of HDAdIL-10 is elusive. Future preclinical studies will be aimed at increasing IL-10 expression levels and improving the sensitivity of this animal model to detect atheroprotective effects.

FIG. 1.

Helper-dependent adenoviral (HDAd) vectors express interleukin (IL)-10 protein in vitro. (a) Expression cassette structure. CMV, cytomegalovirus promoter from pCI plasmid; i, β-globin/IgG chimeric intron from pCI; MCS, multiple cloning site; pA, simian virus 40 poly(A) signal; 5U, 5′-untranslated region of the rabbit IL-10 gene; blocks 1, 2, 3, 4, and 5, rabbit IL-10 exons; 3U, 3′-untranslated region of the rabbit IL-10 gene. Drawing is not to scale. (b) HEK 293 cells were transduced with HDAdNull, HDAdcIL-10, or HDAdgIL-10, at the viral doses indicated (particles/ml). Conditioned medium was collected and analyzed by Western blotting for IL-10. Each lane is from cells in a separate well. Size marker is presented as kilodaltons (kDa).

FIG. 2.

Activity of vector-expressed IL-10. Rabbit peripheral blood mononuclear cells were treated with conditioned medium generated by transduction of 293 Cre4 cells with HDAdNull, HDAdgIL-10, or HDAdcIL-10. Aliquots of cells in all three groups were treated with LPS. After 5 hr, RNA was extracted and the expression of (a) TNF-α and (b) IL-1β was measured by qRT-PCR. Columns represent means of n = 3 separate wells of treated cells per group, with normal distribution assumed based on extrapolation from a larger experiment (Supplementary Fig. S2).

FIG. 3.

Identification of an optimal in vivo dose for HDAdgIL-10. (a) HDAdgIL-10 was infused at the indicated concentrations and arteries were harvested 3 days later. Vector genomes (measured by qPCR) are expressed as genomes per vascular wall cell, calculated on the basis of total DNA in the extracts. (b) RNA was extracted from the arteries in (a), from five additional HDAdgIL-10-infused arteries from a separate study (5 × 10¹⁰ particles/ml; Supplementary Fig. S3), and from control arteries infused with HDAdNull. IL-10 mRNA was measured by qRT-PCR and normalized to GAPDH mRNA in the same extracts. Data points represent individual arteries; bars indicate group means.

FIG. 4.

Persistence of IL-10 expression and vector genomes after HDAd infusion. (a) Arteries were infused with HDAdgIL-10 (solid circles) or HDAdNull (open circles) and harvested at the indicated times. IL-10 mRNA was measured in all samples simultaneously by qRT-PCR and normalized to GAPDH mRNA in the same extracts. RNA from a single HDAdNull-infused artery harvested on day 3 served as the “calibrator” for ΔΔC_t calculations (Schmittgen and Livak, 2008). Data points represent individual arteries; bars indicate group means. (b) Arteries were infused with either HDAdNull (open circles) or HDAdgIL-10 (closed circles) and harvested at the indicated times. Vector DNA was measured by qPCR and expressed as a percentage of vector DNA measured in arteries infused with the same vector and harvested on day 3 (set at 100% for both groups). Data represent means ± SEM, with n = 8–14 arteries per group. *p = 0.01; **p = 0.07 versus HDAdNull.

FIG. 5.

Intimal area and macrophage content in arteries infused with HDAdNull or HDAdgIL-10. Carotid arteries of cholesterol-fed rabbits were removed at the indicated time points after vector infusion, sectioned, and stained. (a) Mean intimal areas. (b) Mean percentages of intimal area occupied by macrophages. (c–e) Representative sections of HDAdNull-infused arteries harvested at 14, 28, and 56 days, respectively. Arrows indicate intimal lesions. Total intimal areas were as follows: (c) 0.02 mm²; (d) 0.038 mm²; (e) 0.032 mm². Data points in (a) and (b) are from individual arteries and are means of areas measured on four step sections per artery; bars represent group means. Sections in (c)–(e): Verhoeff–Van Gieson stain; size bars, 200 μm.

FIG. 6.

HDAdgIL-10 does not alter expression of adhesion molecules (ICAM-1 and VCAM-1) in rabbit carotid artery intima. Carotid arteries were transduced with HDAdNull or HDAdgIL-10 and harvested 14, 28, or 56 days later. (a and b) Sections of arteries were stained, and intimal ICAM-1 and VCAM-1 staining intensity was scored, using semiquantitative immunohistochemistry. Data points represent individual arteries; bars indicate group means.

FIG. 7.

No evidence of systemic immunosuppression in rabbits expressing IL-10 in carotid arteries. (a and b) Liver and spleen RNA were harvested 3 days after carotid artery infusion with HDAdNull or HDAdgIL-10. IL-1β and IFN-γ mRNA were measured by qRT-PCR. Expression of IFN-γ in the liver (data not shown) was at or below the limit of detection in all samples. (c) Serum was collected from rabbits before arterial transduction with HDAdNull or HDAdgIL-10 (preoperative) and then again at harvest (14 or 28 days postoperative). Antibodies to human adenovirus type 5 were measured by ELISA, using a substrate that absorbs light at 405 nm (OD₄₀₅). Data points represent individual rabbits; bars indicate group means. *p ≥ 0.2 for comparison of these HDAdgIL-10 groups with the HDAdNull control groups at the same time point.