The cloning of infectious AAV genomes in bacterial plasmids facilitated a molecular genetic analysis of AAV (70,71). These studies showed that the rep and cap genes are required in trans to provide functions for replication and encapsidation of viral genomes, respectively, and that the ITR is required in cis (4,6,63). Mutations in the ITR have an Ori phenotype and cannot be complemented in trans.
Mutations that affect the Rep78 and Rep68 proteins have a Rep phenotype, and are deficient for both the bulk replication and amplification of duplex RF molecules and for accumulation of ss, progeny genomes. A mutation that affected only the Rep52 and Rep40 proteins showed an Ssd phenotype in which duplex RF replication occurred normally but no ss progeny DNA accumulated.
The cap gene encodes the proteins VP1, VP2, and VP3 that share a common overlapping sequence, but VP1 and VP2
contain additional amino terminal sequence. All 3 proteins are required for capsid production. Mutations that affect VP2 or VP3 have a Cap phenotype, and block capsid assembly and prevent any accumulation of ss DNA. This indicates that VP3 and VP2 are primarily responsible for forming the capsid and that ss DNA does not accumulate unless it can be packaged into capsids. Mutations that affect only the amino terminus of the VP1 protein do not prevent accumulation of capsids or ss DNA, but no infectious AAV particles accumulate. This phenotype has been described as either Inf or Lip (low infec-tivity particles).
These genetic studies, together with additional biochemical studies, show that Rep68 and Rep78 are required for replication, that VP2 and VP3 are required to form the capsid, and that Rep52 and Rep40 appear to act in concert with VP1 to encapsidate the DNA and stabilize the particles (4,5,66). An additional role of VP1 appears to be to provide the phospholi-pase activity that is required for infectivity (58).
The rep proteins exhibit several pleiotropic regulatory activities, including positive and negative regulation of AAV genes and expression from some other viral or cellular promoters, as well as inhibitory effects on the host cell. Because of the inhibitory effects of expression of rep gene products on cell growth, expression of rep proteins in stable cell lines was difficult to achieve, and this delayed development of AAV packaging cell lines (72). For this reason, various approaches to AAV vector production have employed transient transfec-tion of cells with AAV vector plasmids and complementing rep-cap plasmids. However, even in these transfection systems, the closely coordinated regulation of rep and cap gene
Uncoating n__n lr "U
Transcription j Translation
Figure 3 Metabolic pathway of AAV genomes in cells. Binding of AAV to cells is independent of helper virus functions. Trafficking of AAV to the nucleus may be enhanced by agents that interact with the ubiquitin pathway and proteosome processing. Conversion of the infecting single-strand genome to a duplex structure (or parental RF) through the process of metabolic activation (second-strand synthesis) can occur independently of helper virus. This process may be enhanced by infection with helper adenovirus genes such as E4orf6 or by other metabolic insults, including genotoxic stress or heat shock. Treatments that enhance metabolic activation may enhance gene expression from the vector template. The single strand and duplex strands are drawn to show the ITR in the base-paired hairpin conformation that allows self-priming of replication to form a duplex template using cellular DNA poly-merases. For further details see text. (Reprinted with permission from Ref. 165.)
expression and the interactions of the 3 AAV promoters (69) are important considerations in optimizing vector production.
A U.S. government screening program to assess human cell lines for vaccine production led to the observation that infection of primary cultures of human embryonic kidney cells with adenovirus resulted in rescue of infectious AAV. The hypothesis that some cultures may have carried a latent form of AAV was tested directly (34) by infecting a human cell line, Detroit 6, with AAV at a high multiplicity of infection and passaging the cell cultures until no infectious AAV genomes were present, which required at least 10 cell passages. Following this, superinfection of the cultures with adenovirus resulted in rescue of infectious AAV. This provided an important demonstration of a way in which AAV may survive in a cell if conditions are not permissive for replication.
Analysis of a human cell line, Detroit 6, carrying latent AAV showed that the cells contained a relatively low number of AAV genomes that were integrated into the host cell chromosome, mostly as tandem repeats. Early studies of cell lines stably transduced with AAV vectors expressing selectable markers also showed that most stable copies in the cell existed as tandem repeats with a head-to-tail conformation (35). Analysis of chromosomal flanking sequences showed that, for wild-type AAV, 50% to 70% of these integration events occurred in a defined region (36,37). When wild-type AAV infects human cell lines in culture, up to 50% to 70% of these integration events occur at a region known as the AAVS1 site on chromosome 19 at 19q13ter. Both the specificity and efficiency of AAV integration are mediated by the AAV rep protein that binds to the ITR and to a site in the AAVS1 site on human chromosome 19 (38). These studies were performed on cells in culture, and it is noteworthy that naturally occurring, latent, integrated AAV genomes have not been well characterized in humans or any other animal species. In a single study, wt AAV infection of rhesus macaques was examined in the presence and absence of adenovirus infection (73). In this study, a wtAAV-AAVS1 host cell DNA junction could be detected by polymerase chain reaction (PCR) amplification in only 1 of 9 animals, suggesting that integration of AAV even under favorable conditions is very rare.
Some early studies with AAV vectors expressing selectable markers suggested that these vectors also integrated at AAVS1 (36,74). However, the vectors used in these studies were also were contaminated by wild-type AAV particles and may represent rare integration events mediated by rep and enriched by the selectable marker. In contrast, AAV vectors that contain no AAV rep coding sequences and no selectable marker have reduced efficiency and specificity for integration at the chromosome 19 AAVS1 site (40,41), and more usually persist in an episomal state. Initial evidence for this episomal persistence came from fluorescent in situ hybridization (FISH) analysis of cell lines transduced with AAV vectors, which carried a low number of copies of an AAV vector as determined by Southern blot hybridization. FISH analysis of interphase nuclei compared with FISH analysis of metaphase chromosomes showed a reduced proportion of the cells carrying all copies at a metaphase chromosomal site (41).
Studies performed in vivo in a variety of animal models now indicate that, in the absence of selective pressure, AAV
Second strand synthesis
Transcription j Translation vectors generally persist as episomal genomes. A number of studies have now shown that AAV can persist for extended periods of time when administered in vivo (51-56) and that the predominant form of the persisting vector genomes appears to be multimeric structures, which are head-to-tail con-catemers (52,75-78) that are circular (79-81). How these head-to-tail multimers are formed is unknown, but it cannot involve the normal AAV replication process because that requires rep protein and gives only head-to-head or tail-to-tail concatemers. Whether the circular concatemers are integration intermediates, as has been suggested for AAV integration (39,82), is also unknown. However, available evidence indicates that the majority of these head-to-tail concatemers are episomal and that integrated copies of vector in organs such as liver or muscle are very rare (83,84).
Was this article helpful?