Novel AAV serotypes isolated from nonhuman primates and human tissues have shown promise as efficient gene transfer vectors, attributed to their high in vivo gene transfer efficiency and distinct immunology properties. In this study, we show that AAV2/7, 2/8, and 2/9 are more efficient and less immunogenic vectors for both liver and muscle-directed gene therapy for hemophilia B than the traditional serotypes. For a liver-directed approach, evaluation of the new serotypes was performed in mouse and dog hemophilic models. In hemophilia B mice (in C57BL/6 background), long-term superphysiological levels of canine FIX (41–70 μg/ml) and shortened aPTT time have been achieved following intraportal injection of 1 × 1011 GC of AAV2/7, 2/8, and 2/9-LSP-cFIX-WPRE vectors. A 10-fold lower dose (1 × 1010 GC) of AAV2/8 vector generated normal level of cFIX and aPTT time. In UNC hemophilia B dogs, we previously demonstrated that administration of an AAV2/8 vector into a dog previously treated with an AAV2 vector was successful; cFIX expression peaked at 10 μg/ml day 6 after the 2nd intraportal injection (dose = 5 × 1012 GC/kg), then gradually decreased and stabilized around 700 ng/ml (16% of the normal level) throughout the study (1 1/2 years). This level was about 3-fold higher than that from a hemophilia B dog that received a single injection of AAV2-cFIX at the similar dose. Recently, two naive hemophilia B dogs were injected with AAV2/8 vectors intraportally at the dose of 5.25 × 1012 GC/kg. cFIX levels in one dog (male) reached 30% of normal level (1.5 μg/ml) ten weeks after injection and has sustained at 1.3–1.5 μg/ml, while the second dog (female) maintained cFIX expression at about 10% of normal level. WBCT and aPTT were both shortened after the injection. Liver enzymes in both dogs remained in the normal range after surgery. We further assessed these new serotypes for muscle-targeted gene therapy of hemophilia B. AAV-CMV-cFIX-WPRE packaged with 6 different AAV serotypes were compared in hemophilia B mice (in C57BL/6 background) after intramuscular injection at the dose of 1 × 1011 GC. cFIX gene expression and antibody formation were monitored. Highest expression was detected in the plasma of the mice injected with AAV2/8 vectors (1460 ± 392 ng/ml at day 42), followed by AAV2/9 (773 ± 171 ng/ml) and AAV2/7 (500 ± 311 ng/ml). Levels were maintained for 5 months. To our surprise, cFIX expression by AAV2/1 ranged from 0–253 ng/ml (average: 66 ± 82 ng/ml). Anti-cFIX inhibitor (IgG) was detected in some of the AAV2/1-injected mice. Further screening of inhibitor formation was performed on day 28 samples for all serotypes. Hemophilia B mice showed highest inhibitor formation against AAV2/2, followed by AAV2/5, and AAV2/1. Only sporadic and low level inhibitors were detected in animals injected with AAV2/7, AAV2/8 and AAV2/9. Thus, we show the advantages of the new AAV serotypes 2/7, 2/8, and 2/9 for muscle-directed gene therapy for hemophilia B as more efficient and safe vectors without eliciting any significant anti-FIX antibody formation. These vectors could also be useful for gene therapy of other genetic diseases.

Cystic fibrosis (CF) is one of the most common life-threatening autosomal recessive disorders affecting the Caucasian population. The malfunction of the cystic fibrosis transmembrane conductance regulator (CFTR) gene results in both severe chronic lung disease, which currently accounts for more than 95% of the morbidity and mortality of CF patients, and manageable pancreatic disease. Up to date CFTR gene transfer to the lung for the treatment of CF airway disease has been limited by poor vector performance combined with the significant barriers that the airway epithelium poses to effective gene transfer. Recently, our group isolated a novel AAV family (AAV9). In an attempt to evaluate its tropism and transfection efficiency we tested the AAV2 genome packaged in the AAV9 capsid (AAV2/9) and compared it to AAV2/5 in various airway model systems. A 50 ml single dose of 1 × 1011 genome copies (gc) of AAV2/9 expressing either the nuclear targeted b-galactosidase (nLacZ) gene or the green fluorescence protein (GFP) gene under the transcriptional control of the chicken b-actin promoter was instilled intranasally into nude and also C57Bl/6 mice. Twenty-one days later, the lung and nose were processed for gene expression. In control animals transduced with AAV2/9-GFP no LacZ positive cells were seen. AAV2/9-nLacZ successfully transduced mainly airways, whereas AAV2/5-nLacZ transduced mainly alveoli and few airways. Across the nasal airway epithelium both AAV2/5 and AAV2/9 transduced ciliated and non-ciliated epithelial cells. We are currently evaluating epithelial cell specific promoters to improve targeting to the airway cells in vivo. Based on the in vivo findings we next tested the gene transfer efficiency of AAV2/9 to human airway epithelial cells. Airway epithelial cells were isolated from human trachea and bronchi and grown at air-liquid-interface (ALI) on collagen coated membrane supports. Once the cells polarized and differentiated they were transduced with AAV2/9 or AAV2/5 expressing GFP from the apical as well as the basolateral side. Both AAV2/5 and AAV2/9 were successful at transducing epithelial cells from the basolateral surface. However, when applied onto the apical surface AAV2/9 resulted in a 10-fold increase in the number of transduced cells compared to AAV2/5. Currently, we are assessing the gene transfer performance of AAV2/9 in the lungs and nasal airways of nonhuman primates. We have demonstrated that AAV2/9 can efficiently transduce the airways of murine lung and well-differentiated human airway epithelial cells grown at ALI. These encouraging results have prompted the evaluation of AAV2/9 in nonhuman primates as well as the evaluation of AAV2/9-CFTR gene transfer to the nasal airway epithelium of cystic fibrosis knockout mice in an attempt to evaluate physiological correction of the CF defect. This novel family of AAV vectors shows great promise for the treatment of CF airway disease.