Abstract
Our lab has previously reported the successful systemic delivery of AAV6 to skeletal muscle achieving nearly a 100% transduction of muscle fibers (Gregorevic et al., Nat Med 2004). We used the same technique to study the efficiency of recombinant adeno-associated virus pseudotype 6 vectors (rAAV6) to transduce vasculature. Because endothelial cells (EC) and smooth muscle cells of the vasculature can proliferate in vivo and AAV does not integrate in the genome efficiently; we used the Cre-loxP technology to study in vivo transduction of vasculature. For this, AAV6 constructs expressing Cre recombinase under the control of the CMV promoter were systemically injected into loxP-ROSA26 mice. In these transgenic mice Cre expression results in lacZ expression in the cells/tissues transduced with AAV. In order to determine the optimal dose for transduction of vasculature in vivo, 3 viral titers were used, 10^11, 10^12, 10^13 vector genomes (n=3 per viral dose). 4 weeks after transplant, muscle mononuclear cells were harvested from tibialis anterior, quadriceps and gastrocnemius muscles for culture in endothelial medium (abstract by Reyes). In this medium endothelial cells can be expanded for more than 10 cell doublings. LacZ and bgal staining of muscle EC cells (at multiple times during a 10 week culture) revealed 35–40% positive cells. In addition, tissue sections of the heart, liver, GI, spleen, esophagus and multiple skeletal muscles were analyzed for lacZ staining. We used multiple EC (vWF, CD31, CD34, Sca-1) and smooth muscle (smooth muscle actin and smooth muscle myosin) markers to determine AAV transduction efficiency of these two types of vascular cells in these tissues. Highest transduction was observed in the vessels of the heart and skeletal muscles obtaining approximately 50% at the lowest dose but nearly 100% at the highest dose (10^13). Vessel of the liver, GI and spleen obtained variable transduction at the lowest and intermediate dose, but more than 50% at the highest dose.
In a second set of experiments, the same CMV-Cre AAV6 construct was injected into another type of floxed mice, ROSA26loxP-GFP. These transgenic mice constitutively express lacZ but when infected with a Cre expressing AAV, lacZ expression is replaced with enhanced GFP expression in tissues infected with AAV. The advantage of these mice is that GFP positive cells (AAV transduced cells) can be analyzed and sorted by flow cytometry. For these studies three mice were injected with the highest dose, 10^13. We analyzed the transduction efficiency of cells from the bone marrow and skeletal muscles. Whereas very poor transduction was observed in bone marrow cells (<1% CD45+ cells, <0.5% of Sca-1+/CD45+ cells were GFP+) higher transduction was observed in muscle cells (10% of all muscle mononuclear cells were GFP+, of those 30% expressed endothelial markers such as Sca-1+/34+/31+). These cells were sorted and cultured in endothelial medium to demonstrate that they were bona fide ECs. In addition, freshly sorted GFP+ECs were IM transplanted into wt mice. 2 weeks after, GFP positive cells were seen in multiple vessels and perivascular tissues of the muscle. This is first demonstration that AAV6 can efficiently transduce the vasculature when administered systemically, thus this technique is potentially applicable to target diseases of the vasculature such as ischemia and atherosclerosis. Studies to determine the in vivo persistence of AAV transgene expression in the vasculature are ongoing using vectors constitutively expressing reporter genes.
Scleraxis (Scx) directs lacZ expression in tendon of transgenic mice