Plasma apolipoprotein E (apoE) has multiple atheroprotective actions. However, although liver-directed adenoviral gene transfer of apoE reverses hypercholesterolemia and inhibits atherogenesis in apoE-deficient (apoE(-/-)) mice, safety considerations have revived interest in nonviral DNA (plasmid) and nonpathogenic adeno-associated viral (AAV) vectors. Here, we assess the effectiveness of these two delivery vehicles by minimally invasive intramuscular injection. First, we constructed AAV2-based expression plasmids harboring human apoE3 cDNA, driven by two muscle-specific promoters (CK6 and C5-12) and one ubiquitous promoter (CAG); each efficiently expressed apoE3 in transfected cultured C2C12 mouse myoblasts, although muscle-specific promoters were active only in differentiated multinucleate myotubes. Second, a pilot study verified that electrotransfer of the CAG-driven plasmid (p.CAG.apoE3) into tibialis anterior muscles, pretreated with hyaluronidase, of apoE(-/-) mice significantly enhanced (p < 0.001) local intramuscular expression of apoE3. However, in a 7-day experiment, the CK6- and C5-12-driven plasmids produced less apoE3 in muscle than did p.CAG.apoE3 (0.61 +/- 0.38 and 0.45 +/- 0.38 vs. 13.38 +/- 7.46 microg of apoE3 per muscle, respectively), but plasma apoE3 levels were below our detection limit (<15 ng/ml) in all mice and did not reverse the hyperlipidemia. Finally, we showed that intramuscular injection of a cross-packaged AAV serotype 7 viral vector, expressing human apoE3 from the CAG promoter, resulted in increasing levels of apoE3 in plasma over 4 weeks, although the concentration reached (1.40 +/- 0.35 microg/ml) was just below the threshold level needed to reduce the hypercholesterolemia. We conclude that skeletal muscle can serve as an effective secretory platform to express the apoE3 transgene, but that improved gene transfer vectors are needed to achieve full therapeutic levels of plasma apoE3 protein.