Ubiquitination of both adeno-associated virus type 2 and 5 capsid proteins affects the transduction efficiency of recombinant vectors

Abstract

In the presence of complementing adeno-associated virus type 2 (AAV-2) Rep proteins, AAV-2 genomes can be pseudotyped with the AAV-5 capsid to assemble infectious virions. Using this pseudotyping strategy, the involvement of the ubiquitin-proteasome system in AAV-5 and AAV-2 capsid-mediated infections was compared. A recombinant AAV-2 (rAAV-2) proviral luciferase construct was packaged into both AAV-2 and AAV-5 capsid particles, and transduction efficiencies in a number of cell lines were compared. Using luciferase expression as the end point, we demonstrated that coadministration of the viruses with proteasome inhibitors not only increased the transduction efficiency of rAAV-2, as previously reported, but also augmented rAAV-5-mediated gene transfer. Increased transgene expression was independent of viral genome stability, since there was no significant difference in the amounts of internalized viral DNA in the presence or absence of proteasome inhibitors. Western blot assays of immunoprecipitated viral capsid proteins from infected HeLa cell lysates and in vitro reconstitution experiments revealed evidence for ubiquitin conjugation of both AAV-2 and AAV-5 capsids. Interestingly, heat-denatured virus particles were preferential substrates for in vitro ubiquitination, suggesting that endosomal processing of the viral capsid proteins is a prelude to ubiquitination. Furthermore, ubiquitination may be a signal for processing of the capsid at the time of virion disassembly. These studies suggest that the previously reported influences of the ubiquitin-proteasome system on rAAV-2 transduction are also active for rAAV-5 and provide a clearer mechanistic framework for understanding the functional significance of ubiquitination.

FIG. 1.

Production of rAAV-2 and rAAV-2cap5. (A) Principles underlying the pseudotyping of rAAV-2 genomes into AAV-5 particles. In the presence of AAV-2 Rep proteins and helper adenovirus, sequences flanked by the rAAV-2 ITRs are excised from the proviral plasmid (pcisAAV-2) and replicated. Depending on the serotype of capsid proteins provided by a second trans-plasmid, the rAAV-2 genome can be packaged in either rAAV-2 or pseudotyped rAAV-5 (rAAV-2cap5) particles. (B) Various helper plasmids that were tested for packaging of rAAV-2 proviral genomes into AAV-5 particles. AAV-2 Rep proteins are necessary for pseudopackaging of the rAAV-2 genome into AAV-5 particles and were provided by the helper plasmid pAV2-Rep. This plasmid was derived from pAAV-2/Ad, the routine helper plasmid for rAAV-2 production, by deletion of the AAV-2 capsid coding region. pAV5-Trans was generated by replacing the AAV-2 genome with the full-length AAV-5 Rep and Cap coding sequences. It can be used as the helper for generating authentic rAAV-5 vectors or for pseudotyping AAV-2 in an AAV-5 capsid. The AAV-5 capsid expression plasmid p40Av5Cap(1614) contains the native p40 promoter for Cap gene transcription. pCMVAv5Cap(1924) is similar, except that the human cytomegalovirus (hCMV) immediate-early (IE) promoter-enhancer replaces the p40 promoter. pCMVAv5Cap(2196) is derived from pCMVAv5Cap(1924) with the splicing signal deleted so that the CMV promoter is immediately upstream of the VP1 start codon. (C) Effects of the different AAV-5 helper plasmids on virus production. The yields of rAAV-2 and rAAV-2cap5 are the mean and standard error of the mean for three independent preparations.

FIG. 2.

Evaluation of rAAV-2 and pseudotyped rAAV-2cap5 vectors. Both rAAV-2 and pseudotyped AAV-5 (rAAV-2cap5) contained the same luciferase reporter gene derived from the proviral plasmid pcisAV2RSVluc. The titers of both viral stocks were adjusted to equivalent physical particles per milliliter. (A) Titration of these two recombinant viral stocks by slot blot analysis against plasmid DNA standards (STD). The quantity of plasmid and the volume of virus evaluated in each well are marked above and below the blot, respectively. (B) Differences in transduction efficiencies following infection with either rAAV-2 or pseudotyped rAAV-2cap5 in a series of cell types (HeLa cells, primary fetal fibroblasts, IB3 cells, 293 cells, and undifferentiated or differentiated C2C12 muscle cells). Experiments were performed by infecting cells with 5 × 10⁸ total particles/well in 12-well plates. Luciferase activity (relative light units/well) was determined 24 h after infection. Data represent the mean and standard error of the mean for four independent experiments. (C) Time course of transgene expression and viral genome persistence in HeLa cells following infection with rAAV-2 or rAAV-2cap5. A total of 10⁹ particles of rAAV-2 or rAAV-2cap5 were used for infection of six-well plates. (Left panel) Luciferase activity was assayed at 24, 48, and 72 h postinfection. Data represent the mean and standard error of the mean for three independent experiments. (Right panel) Low-molecular-weight Hirt DNA was harvested from infected HeLa cells at 24 h (lanes 1 and 4), 48 h (lanes 2 and 5), and 72 h (lanes 3 and 6) and separated on a 1% agarose gel for Southern blotting and hybridization with a ³²P-labeled luciferase probe. Each lane consists of Hirt DNA from one-half of a 35-mm dish of infected cells (note that microgram normalization of the DNA load did not provide an accurate comparison because cells were actively dividing over the time course of the study and contaminant genomic DNA in the Hirt DNA is dependent on cell number). Lanes 1 to 3, rAAV-2-infected cells; lanes 4 to 6, rAAV-2cap5-infected cells; 7, noninfected negative control.

FIG. 3.

Effects of proteasome inhibitors on rAAV-2 and rAAV-2cap5 transduction. (A) HeLa cells were infected with luciferase-expressing rAAV-2 or rAAV-2cap5 at an MOI of 250 particles/cell in the presence of different concentrations of the proteasome inhibitor LLnL or Z-LLL. (B) HeLa cells were infected with different doses of rAAV-2 or rAAV-2cap5 in the presence of 40 μM LLnL. In all panels, luciferase activity was measured 24 h after infection by using 12-well plates as described in Materials and Methods. Data represent the mean and standard error of the mean relative luciferase activity (per well) for four independent experiments.

FIG. 4.

Ubiquitination of AAV-2 and AAV-5 capsid proteins. (A) Western blot analysis for ubiquitinated AAV-2 and AAV-5 capsid proteins in HeLa cells. HeLa cells were infected with luciferase-expressing rAAV-2 or rAAV-2cap5 with or without 40 μM LLnL. At 4 h after infection, cells were trypsinized, washed twice with phosphate-buffered saline, and lysed in 1 ml of RIPA buffer. As described in Materials and Methods, virus from HeLa cell lysates was immunoprecipitated with antibody B1 and subjected to Western blotting against antiubiquitin monoclonal antibody. Lane 1, rAAV-2 infection without LLnL; lane 2, rAAV-2 infection with LLnL; lane 3, mock-infected cells without LLnL; lane 4, mock-infected cells with LLnL; lane 5, rAAV-2cap5 infection without LLnL; lane 6, rAAV-2cap5 infection with LLnL. Ubiquitinated capsid proteins (Ub-cap; bracket) and cross-reactive immunoglobulin G (IgG) (arrowhead) are marked to the left of the blot. (B) Southern blot analysis of low-molecular-weight Hirt DNA from noninfected HeLa cells (lane 1) and HeLa cells infected with rAAV-2 (lanes 2 and 3) or rAAV-2cap5 (lanes 4 and 5) in the presence (lanes 2 and 4) or absence (lanes 3 and 5) of 40 μM LLnL. Single-stranded viral DNA is marked by an arrowhead to the left of the blot. (C) In vitro ubiquitin conjugation to rAAV-2 or rAAV-2cap5 particles. A total of 3 × 10⁸ particles of rAAV-2 or rAAV-2cap5 were incubated with ubiquitin conjugation enzymes at 37°C for 120 min and then resolved by SDS-10% PAGE. Western blotting with anti-AAV capsid mouse monoclonal antibody B1 and enhanced chemiluminescence detection visualized increased apparent sizes of ubiquitinated AAV capsid proteins. The conditions for each conjugation reaction are marked below the blot. Lanes 1 to 4, virus without conjugation mixture; lanes 5 to 10, conjugation reactions with and without ubiquitin and ATP substrates as a control; lane 11, conjugation mixture without AAV. Both intact (I) and denatured (D) virus samples were evaluated to determine whether native and/or denatured capsid proteins are targets for ubiquitination. In reactions where virus was denatured, the virus was heated at 95°C for 10 min before addition to the conjugation reactions. Reactions in lanes 5 (rAAV-2) and 8 (rAAV-2cap5) were performed as the negative controls; the reaction conditions were the same as those for lane 6 (rAAV-2) and lane 9 (rAAV-2cap5), except that no ubiquitin or ATP was applied. VP1, VP2, and VP3 are marked by arrowheads to the right of the gel, and ubiquitinated capsids (Ub-cap) are bracketed.