The epidemiologic observation that HDL cholesterol is highly inversely associated with coronary risk drove these studies. The Framingham Heart Study has also shown that even when the LDL levels are at acceptable levels, a very significant association between risk and low HDL remained. Therefore, increasing HDL levels might be a powerful modality to reduce cardiovascular (CV) risk.

HDL metabolism is quite complex involving a variety of gene products, and each is a potential therapeutic target for atherosclerosis (Fig.1). The ABC1 gene product facilitates removal of cholesterol and phospholipids from cells. HDL and its major protein, Apo A-1 transports cholesterol through a series of steps ultimately resulting in the deposition of the cholesterol ester in the liver by a cell surface receptor, SR-B1, with excretion into the bile. Conceptually, promotion of this pathway through a variety of potential interventions could promote removal of cholesterol from the vessel wall and be a therapeutic intervention for atherosclerosis.

Somatic gene transfer to the liver as a way of overexpressing proteins at a time when atherosclerotic lesions had been induced in the animals was used (Fig. 2). Their studies were designed to study the impact of gene overexpression in terms of regression of atherosclerosis or changes in lesion morphology when atherosclerosis was already present. This is in contrast to transgenic mice models used to determine which overexpressed gene prevent and reduce the development of atherosclerosis.

A second-generation replication defective adenovirus with an E1 deletion and a mutation of the E2^a gene, making the vector somewhat less responsive to the E1-like cellular factors found in many cells, particularly the liver, was used. A somewhat reduced late gene transcription and therefore reduced immunoresponse to the vector results. Expression is prolonged to about 12 weeks in mice, allowing enough time to look at atherosclerosis.

ApoA-1 studies

To test whether expression of HDL-associated proteins might induce regression or changes in lesion morphology, LDL receptor- (LDLR) deficient mice (n=38) were fed a high fat, Western-type diet for 5 weeks which induces fatty-streak lesions, not advanced lesions. Of these 38 mice, 15 were assessed at baseline for the extent of atherosclerosis, 12 were injected with a second-generation adenovirus expressing human

ApoA-1, and 11 were injected with an adenovirus not encoded with transgene. After 4 weeks quantitation of the aortic atherosclerosis and cellular composition of the atherosclerotic lesions were analyzed.

Significant regression of pre-existing atherosclerotic lesions was found at 4 weeks in LDLR knockout mice to less than a 1% aortic lesion from nearly 3% at baseline, while the lesion continued to progress in the adenovirus null group to nearly 4%. Further, morphological changes were seen, with the lesions in the ApoA-1 injected mice being smaller, flatter, and less rich with foam cells. This is proof-of-principle that even short-term expression of this major HDL protein was able to induce fairly substantial regression.

ApoE studies

ApoE is associated with HDL and other lipoproteins, and is anti-atherogenic (Fig. 3). Plasma-derived ApoE, mostly derived from the liver, is known to have major effects on lipoprotein metabolism. Vessel wall-derived ApoE is derived from macrophages, both of which have been shown to have has direct effects on atherosclerosis, including progression.

Could the liver be targeted with an adenovirus to cause ApoE expression into the blood, then the plasma ApoE access the vessel wall and contribute to the vessel wall pool of ApoE, and thereby induce regression and morphologic changes?

ApoE expressed in ApoE knockout mice results in a very profound reduction in plasma cholesterol levels. In 12-week old ApoE knockout mice with early fatty streak-type lesions at baseline, nearly complete regression of lesions was found, whereas the lesions progressed in the control mice. In 26-week old ApoE knockout mice with more advanced lesions substantial regression was seen, while lesions progressed in the control mice. In the ApoE knockout mice, the regressed lesions were flatter and more fibrous,and the foam cells appeared to be very disrupted. Fibrous-type caps also appeared to increase significantly, consistent with the concept that lesions altered in this manner results in more stable lesions.

Expression of ApoE in the liver only for 8 weeks resulted in a large amount of ApoE within the lesions. Therefore, the effect was not due to the adenovirus directly targeting the lesion of the vessel wall. They believe the ApoE came from the liver, directly targeted the lesions and was retained by the lesions. So, this approach might be useful to target atherosclerosis even when the cholesterol levels are not being reduced.

ApoE in LDLR mice

Data from pilot studies suggested that cholesterol levels were not lowered with overexpression of ApoE in high-fat diet, LDLR-deficient mice. In an experiment directed at advanced atherosclerotic lesions, LDLR-deficient mice (6 months of age) were fed a high fat, Western-type diet for 12 weeks. A baseline group was assessed for extent of atherosclerosis. The remaining mice were injected with either human ApoE3 expressing a second-generation adenoviral vector or the control vector.

Little change in the plasma ApoE levels during the 6-week study was seen in the control mice. In the ApoE injected mice, a significant increase in plasma ApoE to about 40 mg/dL at 5 days, compared to about 10 mg/dL in the control, was seen. This peak dropped off over time due to the immunoresponse to the vector. But even at 6 weeks, the plasma ApoE levels in the ApoE injected mice was twice that in control mice (about 20 mg/dL vs 10 mg/dL, respectively).

Changes in cholesterol can not explain the changes in atherosclerosis, as no difference in plasma cholesterol between the two groups was found. The ApoE resulted in substantial regression of the lesion in the thoracic and abdominal aorta and some in the aortic arch. In the baseline mice, the lesion area was about 15%, in the control mice about 18%, and in the ApoE mice about 5% (p=0.0001).

ApoE has been reported to be an anti-inflammatory and anti-oxidant protein in vitro. To test this in vivo, isoprostanes were measured in the mice. Isoprostanes 1) are produced by free radical catalyzed peroxidation of fatty acids, 2) formed in vivo, certainly in plasma and excreted in the urine, 3) levels specifically reflect in vivo oxidative stress.

The control mice showed no change in isoprostanes over the 6-week period. The ApoE-injected mice had a dramatic reduction in isoprostanes, from 7.0 ng/mg LDL at day zero, to 2.5 ng/mg LDL at day 7, to just above zero at day 14. This reduction was maintained for the 6 weeks. This is consistent with the possibility that ApoE expression has an anti-inflammatory or anti-oxidant effect that may have contributed to its anti-atherogenic effects.

A pilot study using the AAV vector was performed, as adenovirus vectors will not be used in human studies. A recombinant AAV encoding the human ApoE3 gene yielded a fairly long-term expression of ApoE in a dose-dependent manner. However, the plasma levels of ApoE are substantially lower than those achieved with adenovirus. Much work remains to refine a vector that is less inflammatory and will express for a long enough period of time.

SR-B1 is a cell surface protein that mediates ester uptake into the liver and targets it for excretion into the bile. A substantial reduction in atherosclerosis was seen in SR-B1 injected LDLR-deficient mice fed a Western diet compared to control, consistent with the concept that overexpression of SR-B1, despite its reducing HDL, actually reduced atherosclerosis perhaps by stimulating this pathway and reversing cholesterol transport. LDLR-deficient mice with established atherosclerotic lesions injected with an adenoviral vector expressing SR-B1, which targets the liver and remains there, had a dramatic 55% reduction in the HDL peak with little effect on VLDL and LDL at day 7. These results indicate that HDL metabolism is affected by transient SR-B1 overexpression even when VLDL and LDL are substantially elevated. Further, it was found that atherosclerosis was significantly reduced at two different time points by the transient hepatic overexpression of SR-B1 in these LDL-R-deficient mice. More studies are needed to determine how hepatic SR-B1 overexpression is protective, but three potential mechanisms may explain this effect. One, the expected increased movement of HDL cholesterol through the reverse cholesterol transport pathway may decrease the rate of plaque deposition and may even have effects on plaque regression. Two, HDL itself may be altered and have increased antiatherogenic properties. Three, non-HDL proteins may be changed in amount or structure diminishing their atherogenecity. Based on the analysis in this study of the lipoprotein component levels and lesion size, it seems likely that the changes in HDL cholesterol level had a major impact on atherosclerosis. Their overall findings suggest that interventions aimed at increasing hepatic SR-B1 overexpression may be a novel approach to the prevention and treatment of atherosclerosis, since tissue distribution and regulation of SR-B1 expression are similar in mice, cultured human cells, and humans.