Widespread phosphorylation of histone H2AX by species C adenovirus infection requires viral DNA replication

Abstract

Adenovirus infection activates cellular DNA damage response and repair pathways. Viral proteins that are synthesized before viral DNA replication prevent recognition of viral genomes as a substrate for DNA repair by targeting members of the sensor complex composed of Mre11/Rad50/NBS1 for degradation and relocalization, as well as targeting the effector protein DNA ligase IV. Despite inactivation of these cellular sensor and effector proteins, infection results in high levels of histone 2AX phosphorylation, or gammaH2AX. Although phosphorylated H2AX is a characteristic marker of double-stranded DNA breaks, this modification was widely distributed throughout the nucleus of infected cells and was coincident with the bulk of cellular DNA. H2AX phosphorylation occurred after the onset of viral DNA replication and after the degradation of Mre11. Experiments with inhibitors of the serine-threonine kinases ataxia telangiectasia mutated (ATM), AT- and Rad3-related (ATR), and DNA protein kinase (DNA-PK), the kinases responsible for H2AX phosphorylation, indicate that H2AX may be phosphorylated by ATR during a wild-type adenovirus infection, with some contribution from ATM and DNA-PK. Viral DNA replication appears to be the stimulus for this phosphorylation event, since infection with a nonreplicating virus did not elicit phosphorylation of H2AX. Infected cells also responded to high levels of input viral DNA by localized phosphorylation of H2AX. These results are consistent with a model in which adenovirus-infected cells sense and respond to both incoming viral DNA and viral DNA replication.

FIG. 1.

Phosphorylated H2AX accumulates at late times after infection with wild-type adenovirus. (A) HeLa cells were mock infected, exposed to 40 Gy of radiation, or infected at an MOI of 5 with the wild-type virus dl309. Cells were processed for immunofluorescence microscopy for the E2A DNA-binding protein (E2A-DBP) and phosphorylated H2AX (γH2AX) immediately after irradiation or 24 h after infection. Representative images predominantly of the nucleus are presented. Multiple foci were observed in the irradiated cells, but only those in the plane of focus are seen in this image. (B) HeLa cells were infected at an MOI of 5 and harvested at the indicate times postinfection. Material representing equivalent numbers of initially infected cells was separated by SDS-polyacrylamide gel electrophoresis and analyzed by immunoblotting with the indicated antibodies. (C) Viral DNA levels were measured by quantitative real-time PCR from HeLa cells infected at an MOI of 5 with dl309 at the indicated times after infection (open symbols). The relative level of γH2AX (closed symbols) was determined as the ratio of γH2AX signal to total H2AX signal for each time point from a blot similar to the representative blot in panel B. The ratio measured for mock-infected cells was set to 1.

FIG. 2.

PI3K-like kinase inhibitors decrease H2AX phosphorylation in infected cells. HeLa cells were mock infected or infected with the wild-type virus at an MOI of 5. After 3 h, either caffeine or wortmannin was added to the infected cells at the indicated concentration, and the cells were harvested for analysis after 24 h by immunoblotting for γH2AX and total H2AX. The relative levels of H2AX phosphorylation were determined as the ratio of γH2AX to total H2AX where the ratio measured for mock-infected cells was set to 1. (A to D) Representative results of four experiments for caffeine treatment are shown in panels A and B and for wortmannin treatment in panels C and D. (E and F) HeLa cells were mock infected (−) or infected with the E4 deletion virus, dl366* (+), at an MOI of 10. After 3 h, either caffeine or wortmannin was added to the infected cells at the indicated concentration, and cells were harvested for analysis at 24 h by immunoblotting for RPA32. An open arrowhead indicates nonphosphorylated RPA32, and a filled arrowhead indicates the phosphorylated form of RPA32.

FIG. 3.

Widespread H2AX phosphorylation during adenovirus infection requires viral DNA synthesis. (A) HeLa cells were mock infected, infected with the wild-type virus dl309, or infected with the DNA replication-deficient adenovirus H5wt300-ΔpTP (ΔpTP) at an MOI of 5 and processed for immunofluorescence after 24 h for the E2A-DBP protein and γH2AX. (B) HeLa cells that stably express the E2B preterminal protein gene (HeLa-pTP) were infected and processed as in panel A. The restoration of centers of viral DNA replication is seen by the appearance of foci of E2A-DBP staining.

FIG. 4.

Viral DNA replication-deficient H5wt300-ΔpTP elicits similar changes to the cellular DNA profile as the wild-type virus. HeLa cells were infected with the wild-type virus dl309 (A) or H5wt300-ΔpTP (B) at an MOI of 5 and harvested at the indicated times after infection. Material representing equivalent numbers of initially infected cells was separated by SDS-polyacrylamide gel electrophoresis and analyzed by immunoblotting with antibodies for viral early proteins or cellular proteins. The approximate position of molecular mass standards is indicated on the right in kilodaltons. Mre11 levels were measured in separate infections and are shown with the corresponding β-actin loading control. (C) HeLa cells infected with indicated viruses were harvested after 24 h and the cellular DNA content analyzed by flow cytometry after staining with propidium iodide. Representative histograms are shown where the gates identify populations containing the indicated DNA content for uninfected cells. (D) HeLa cells were infected at an MOI of 5 and analyzed at the indicated times after infection by flow cytometry for DNA content as in panel C. The fraction of cells within with each gate indicated in panel C is plotted as a percentage.

FIG. 5.

Delivery of increasing amounts of nonreplicating viral genomes elicits limited and focal γH2AX. (A) HeLa cells were infected with the replication-deficient virus H5wt300-ΔpTP at the indicated MOI. Viral DNA levels were measured 24 h after infection by quantitative PCR and are presented as the number of viral genomes per infected cell. The solid line is a best-fit exponential saturation curve. The inset shows the relationship between delivered genomes and MOI at low multiplicity. The dashed line in the inset is a linear regression. (B) HeLa cells were infected with the indicated viruses and at the indicated MOI and processed 24 hpi for direct or indirect immunofluorescence for γH2AX. Levels of γH2AX were determined by flow cytometry and are normalized between experiments to the mean fluorescence intensity measured at an MOI of 20. Error bars represent the SEM from three to eight replicates performed on separate occasions. (C) HeLa cells were infected with H5wt300-ΔpTP at the indicated MOI and processed 24 hpi for immunofluorescence to visualize γH2AX and cellular DNA by DAPI staining. Representative fields are shown using equivalent exposures for each image. The pair of images in panel h is an enlargement of the inset seen in panel f. The scale bars in panels a, e, and h represent 10 μm.

FIG. 6.

Exposure to empty viral capsids fails to elicit γH2AX. (A) HeLa cells were exposed to empty viral capsids corresponding to the number of viral particles required for the indicated MOI of wild-type virus. After 24 h, γH2AX was visualized by immunofluorescence and photographed using the identical exposure settings in Fig. 5. (B) HeLa cells were mock infected, infected with the wild-type virus dl309, or exposed to empty capsids at the indicated equivalent MOI. After 24 h, cellular lysates were prepared and analyzed by immunoblotting for γH2AX, total H2AX, and β-actin.