Adeno-associated virus induces apoptosis during coinfection with adenovirus

Abstract

Adeno-associated virus (AAV) is a nonpathogenic parvovirus that efficiently replicates in the presence of adenovirus (Ad). Exogenous expression of the AAV replication proteins induces caspase-dependent apoptosis, but determining if AAV infection causes apoptosis during viral infection is complicated by Ad-mediated programmed cell death. To eliminate Ad-induced cytolysis, we used an E3 adenoviral death protein (ADP) mutant, pm534. AAV and pm534-coinfected cells exhibited increased cell killing compared to pm534 alone. Relative to cells infected with Ad alone, AAV and wild-type Ad-infected cells displayed decreased ADP expression, increased cytolysis until the third day of the infection, and decreased cytolysis thereafter. Biochemical and morphological characteristics of apoptosis were observed during coinfections with AAV and pm534 or Ad, including a moderate degree of caspase activation that was not present during infections with pm534 or Ad alone. AAV coinfection also increased extracellular pH. These studies suggest that AAV induces caspase-dependent and caspase-independent apoptosis.

Figure 1.

Western analysis of Ad and AAV proteins during coinfections. A549 cells were infected with Ad5 (10 moi), pm534 (10 moi), and/or AAV (100 IU) as indicated at the top of the figure. Cultures were harvested 48h pi. Equal amounts of protein were separated by SDS-PAGE and probed with the antibody denoted on the left of the images.

Figure 2

AAV altered lysis of Ad infected cells. A549 cells were infected with Ad5 or the ADP mutant pm534 in combination with (A) wild-type AAV or (B) a recombinant AAV vector lacking AAV gene expression. Cell lysis results in lactate dehydrogenase (LDH) release into the culture medium, and LDH catalyzes the conversion of NADH to NAD. Therefore increased NADH oxidation, which was measured by the change in absorbance at 340 nm, was indicative of increased cell lysis. Cells infected with Ad5 alone are shown as white circles, and cells coinfected with Ad5 and AAV are shown as black circles. Black and white triangles represent pm534 infections with and without AAV, respectively. Uninfected cells are shown as white squares, and AAV-infected cells are shown as black squares. LDH assays were conducted in triplicate, and the averages are reported (+/− standard deviation). The data shown are representative of three independent experiments.

Figure 3

AAV induced apoptosis during coinfection with Ad5 or pm534. Cells were infected, harvested by trypsinization on the indicated day, stained with Annexin V-PE and 7-AAD, and analyzed using the Guava PCA flow cytometer. (A) Time course analysis of Annexin V-PE staining. Flow cytometry results are reported as the percent of Annexin V-PE positive cells (y-axis) on the given day of the infection (x-axis). Ad5-infected cells are depicted as white circles, and Ad5/AAV coinfected cells as black circles. ADP mutant pm534-infected cells are shown as white triangles without AAV and black triangles with AAV. Uninfected cells are depicted by white squares, and AAV-infected cells are represented by black squares. (B) Representative diagrams of Annexin V-PE and 7-AAD staining of uninfected, AAV-infected, and pm534-infected cells. Cells were infected with the virus indicated on the left of each panel and harvested 4 days pi. Fluorescence intensity of Annexin V-PE is represented on the x-axis, and that of 7-AAD is represented on the y-axis. (C) Annexin V-PE and 7-AAD flow cytometry analysis of Ad5-, Ad5/AAV-, and pm534/AAV-infected cells. Cells were harvested between 2-5 days pi, as noted on top of the panels. Fluorescence intensity of Annexin V-PE and 7-AAD are given on the x- and y-axes.

Figure 4

Coinfection with AAV and Ad induces caspase activation. Approximately 1 × 10⁶ cells were infected with AAV and/or pm534 or Ad5. Cells were harvested 4 days pi, and equal amounts of cell lysates were incubated 1 hour with the fluorogenic caspase substrate Ac-DEVD-AMC. Fluorescence was reported in arbitrary relative fluorescence units (RFU) as an indicator of caspase activation. Uninfected and doxorubicin-treated cells were included as negative and positive controls. Each bar indicates the average of 6 samples with the indicated standard deviation. The data shown are representative of three independent experiments. There was no statistical difference in average caspase activity of uninfected cells compared with AAV-infected cells (p = 0.437). AAV in the presence of either Ad5 or pm534 significantly increased caspase activation relative to that of the corresponding adenovirus alone, (p ≤ 0.001).

Figure 5

Helper virus-supported AAV infections alter cell morphology. A549 cells were infected with AAV (100 IU) and/or Ad5 (5 moi) or pm534 (5 moi). Unfixed cells were microscopically examined 3 days pi. Uninfected cells and cells infected with Ad alone are on the top panels, and the corresponding infections with the addition of AAV are below. Cells infected with either Ad5 or pm534 alone were rounded up and enlarged, but cells infected with a helper virus and AAV exhibited cell shrinkage and membrane blebbing. The arrows indicate cells with extensive membrane blebbing. The average area of cells exhibiting cytopathic effect is shown in nm².

Figure 6

Viral infections alter extracellular pH. Approximately 4.5 × 10⁵ cells were infected with 100 IU of AAV and/or 10 moi of Ad5 or pm534. Six days pi, the media was removed from each well and measured using a standard pH meter. The assay was conducted in triplicate, and the average pH values are reported (+/− standard deviation).