recombination. Last year we reported that no-end double stranded
(ds) linear monomer genomes accumulated and predominated in
mouse tissues in the absence of DNA-PKcs or Artemis, but not in wild type mice, when Artemis/DNA-PKcs-independent DNArepair pathways were saturated by an excess viral genome load. This indicates that exposing DNA-PKcs- or Artemis-proficient cells to such a high viral load should allow us to investigate how DNA-PKcs and Artemis open viral hairpin termini. In the present study, we extensively characterizedstructures ofrAAV hairpin termini in transduced tissues at sequence levels. For this, we established a PCR-based AAV-ITR sequencing methodology that employs bisulfitechemical modificationofviral DNA. First, we tested our method to determine AAV-ITR sequence of single-stranded viral genomes extracted from rAAVvirions and a double D-ITR (DDITR) plasmid template. Both 145-base ITR and 165-bp DDITR were successfully sequenced following PCR amplification. Second, we purified no-end ds linear monomer genomes that emerged in DNA-PKcs (-) or Artemis (-) mouse livers by Exo III digestion followed by size fractionation, and viral terminal sequences were determined. As expected, viral hairpin termini of no-end ds linear genomes exhibited a 165-base DDITR. This observation confirmed our previous presumption that viral hairpintermini in the no-end ds linear rAAVgenornes remained closed and had never been modified in the absence of DNA-PKcs or Artemis. This also provided a rationale for our approach. Finally, we determined rAAV-rAAV junction sequences in wild type mouse livers transduced with a high vector genome load that supposedly saturated out ArtemislDNA-PKcs-independent DNA repair pathways. Mapping of breakpoints and microhornology regions found at junctions strongly indicated that it is the AAV-ITR hairpin tips that ArtemislDNA-PKcs-associated endonuclease activity cleaves. Thus, our approach wiII not only further our understanding of the in vivo rAAVbiology, but also provide significant insights into fundamental DNA repair systems in quiescent somatic cells in various tissues of living animals.
388. Adeno-Associated Virus 2-Mediated Gene Transfer: Tyrosine-Phosphorylation of Capsid Proteins and Its Consequences on Transgene Expression Li Zhong, I Baozheng Li,' Sergei Zolotukhin,'> Lakshmanan Govindasamy,' Mavis Agbandje-Mckenna,' Arun Srivastava.t-' Pediatrics, University 0/Florida College ofMedicine, Gainesville, FL; /Moleculor Genetics and Microbiology, University 0/ Florida College 0/Medicine, Gainesville, FL; 'Btochemistry and Molecular Biology, University ofFlorida College ofMedicine, Gainesville, FL. J
The transduction efficiency of recombinant adeno-associated virus 2 (AAV) vectors varies greatly in different cells and tissues in vitro and in vivo. Others and we have carried out systematic studies to elucidate the fundamental steps in the life cycle ofAAV. For example, we have documented that a cellular protein, FKBP52, phosphorylated at tyrosine residues by epidermal growth factor receptor protein tyrosine kinase (EGFR-PTK), inhibits AAV second-strand DNA synthesis and consequently, transgene expression in vitro (J. Virol., 75: 8968,2001 ; J. Virol., 77: 2741, 2003) as well as in vivo (Mol. Ther., 10: 950, 2004; Gene Ther., I I: I 165, 2004; Hum. Gene Ther., 15: 1207,2004). We have also demonstrated that EGFR-PTKsignaling modulates the ubiquitin/proteasomepathwaymediated intracellular trafficking as well as FKBP52-mediated second-strand DNA synthesis of AAV vectors. In those studies, inhibition ofEGFR-PTK signaling led to decreased ubiquitination of AAVcapsid proteins, which in tum, facilitated nuclear transport by limiting proteasome-mcdiatod degradation ofAAV vectors, implicating EGFR-PTK-mcdiatcdphosphorylationof tyrosine residues on AAV capsids (Mol. Ther., 13: S3, 2006). In the present studies, Molecular Therapy Volume 15.Supplement I. ~br Copyright © The American Soc iety of Gene Therapy
we observed that AAV capsids could indeed be phosphorylated at tyrosine residues by EGFR-PTK in in vitro phosphorylation assays and that phosphorylated AAVcapsids retained their structural integrity. However, although phosphorylated AAV vectors could enter cells as efficiently as their unphosphorylated counterparts, their transduction efficiency was significantly reduced. This reduction was not due to impaired viral second-strand DNA synthesis since transduction efficiency of both single-stranded AAV(ssAAV) and self-complementary AAV (scAAV) vectors was decreased by -68% and -74%, respectively. We also observed that intracellular traffickingof tyrosine-phosphorylatedAAVvectors from cytoplasm to nucleus was significantly decreased, which, consistent with our previous studies (Mol. Ther., in revision, 2007), most likely led to ubiquitination of AAV capsids followed by proteasome-mcdiatcd degradation. Our currently ongoing studies on site-directed mutational analyses of surface-exposed tyrosine residues on AAVcapsid proteins should yield new and useful information on the role of AAV capsid phosphorylation in various steps in the life cycleAAV,which may have implications in the optimal use of recombinantAAV vectors in human gene therapy.
389. Subcellular Trafficking of Recombinant AAV: Nucleolar Accumulation of Capsids Is Unique to Infectious Virions Jarrod S. Johnson,' R. Jude Samulski.' 'Pharmacology; University ofNorth Carolina at Chapel/Jill, Chapel/Jill, NC.
Several adeno-associatedvirus(AAV)serotypes are beingadapted into sophisticated gene delivery vehicles; however, fundamental questions still remain regarding how virion components modulate their subcellular trafficking. Successful infection depends on the capsid actively directing and enduring a cascade of subcellular events that include internalization into an endocytic compartment, endosomal processing, endosomal escape, nuclear targeting, and uncoating. Each ofthese events has been reported to limit incoming virions from achieving nuclear delivery of the transgene. To more clearlyunderstandthe traffickingprofileof AAV2and to explore how capsid components modulate subcellular trafficking, we examined the trafficking profiles of recombinant virions by immunofluorcscence. The subcellular trafficking ofrecombinant virions (rAAV2) were compared to a series of non-infectious virions containing mutations that span regions ofthe capsid such as the phospholipase domain, the putative nuclear localization signals, and the 5-fold pore. Using antibodies that recognize intact capsids, exposed VPI/2 N-tennini, or disassembled subunits, we are able to track virions to a perinuclear region within 2hrs after infection. Surprisingly, after 4hrs rAAV2intact eapsids are found to accumulate in nucleoli. This unique nucleolar targeting of rAAV2 capsids was confirmed by confocal z-stack analysis and found to be significantly enhanced by proteasome inhibitors.These data, taken with the findingthat capsids localized in nucleoli are not associated with the viral genome during assembly (Wistuba et al., 1997), implicate the nucleoli as a site for uncoating or capsid degradation. In contrast to full rAAV virions, empty rAAVcapsids and non-infectious mutants do not accumulate in nucleoli, irrespectiveof their ability to expose VP1/2N-termini. In summary, using immunofluorescencedetection of virions in concert with well established biochemical methods of virion analysis, we can more clearly discern how capsid components direct subcellular steps of infection as we progress toward designing ideal virions for gene therapy applications.