Induction and utilization of an ATM signaling pathway by polyomavirus

Abstract

Progression from G(1) to S is essential for polyomavirus DNA replication and depends on the interaction of large T with the retinoblastoma gene product pRb. This virus-induced replication pathway is accompanied by p53 activation resembling a DNA damage response (12). We sought to determine whether this pathway depends in part on activation of the ATM (ataxia telangiectasia mutated) kinase and whether the virus gains advantages from this pathway beyond that of entry into S. We show that polyomavirus infection activates the S- and G(2)-phase checkpoints in primary as well as established mouse cells. Infected cells undergo a prolonged S phase compared to uninfected serum-stimulated cells and show no evidence of a G(2)-->M transition before lytic death ensues. Infection is accompanied by increases in ATM activity in vitro and in the level of ATM-S1981-P in vivo. The incubation of infected cells with caffeine, a known ATM inhibitor, did not block entry into S but reduced the rate of viral compared to cellular DNA synthesis. Importantly, caffeine lowered the yields of viral DNA an average of 3- to 6-fold and those of infectious virus by as much as 10-fold. Virus yields were 10-fold lower in ATM (-/-) p53(-/-) than in ATM(+/+) p53(-/-) mouse embryo fibroblasts, indicating a p53-independent role of ATM in productive infection. Replacement of the normal SMC1 (structural maintenance of chromosomes, or cohesin) protein, a critical ATM substrate in the DNA repair pathway, with its phosphorylation mutant SMC1(S957AS966A) also lowered virus yields by roughly 90%. We suggest that polyomavirus activates and utilizes a component(s) of an ATM pathway of DNA repair to prolong S phase and aid its own replication.

FIG. 1.

Polyomavirus (Py)-infected BMK (A) and A31 (B) cells accumulate in S and G₂ phase before undergoing cell death. Serum-starved monolayers were infected with virus at an MOI of 10 to 20 and supplemented with 10% serum. At the times indicated, samples were taken and prepared for cell cycle analysis by FACS. The positions of the G₁ and G₂ phases are marked. The percentage of cells in S phase is given on each panel. Approximately 90% of cells were infected.

FIG. 2.

Polyomavirus-infected cells do not enter M phase. BMK or A31 cells were serum starved for 2 to 3 days and infected at an MOI of 10 to 20 PFU per cell. The percentage of T antigen-positive cells was 80% or greater. Uninfected and infected cells were fed with low-serum medium. Uninfected cells were fed with medium containing 10% serum. (A) At 44 h postinfection (hpi) (BMK) or 36 hpi (A31), lysates were prepared, and cyclin B1 or cyclin A immunoprecipitates were analyzed for CDC2 phosphorylated on Tyr15 (CDC2-Y15-P) or for p21. (B) At the times indicated, A31 cell extracts were prepared. Levels of Mpm2 proteins were analyzed by immunoblotting. (C) BMK cells were fixed at 48 hpi and analyzed by immunofluorescence staining for T antigen and lamin B.

FIG. 3.

The activity of ATM is elevated in virus-infected cells. (A) BMK cells were harvested at 27 hpi, and lysates were prepared. In vitro ATM activity was assayed as described in Materials and Methods. (B, upper panel) Infected BMK cells were incubated for the indicated times. Caffeine was added 1 h before lysates were prepared and analyzed for γ-H2AX levels by immunoblotting. (B, lower panel) Infected BMK cells were treated with caffeine at the times indicated. At 48 hpi, lysates were prepared and analyzed for p53-S18-P and p21 by immunoblotting. (C) BMK cells were harvested at 27 hpi, and 150 μg of protein was immunoblotted for ATM or ATM-S1981-P. Act D treated, BMK cells were treated for 25 h with 10 mM actinomycin D as a positive control.

FIG. 4.

Caffeine lowers the yield of viral DNA and infectious particles without inhibiting entry into S phase. BMK cells were infected at an MOI of 1 PFU per cell. Following adsorption of virus, monolayers were washed twice, and fresh medium containing caffeine was added. (A) DNAs were isolated at the times indicated, dot blotted to a membrane, and hybridized to ³²P-labeled polyomavirus DNA. (B) Infected cells were harvested at 54 hpi, and the virus yield was determined by a plaque assay. (C) Infected BMK cells treated with or without 3 mM caffeine were harvested at the times indicated and analyzed by FACS for cell cycle distribution. (D) Infected A31 cells treated or not with 1 mM caffeine were measured for the rate of viral and cellular DNA synthesis by pulse labeling with [³H]thymidine at 20 hpi. Data for the rate of DNA synthesis from duplicate samples are presented as percentages of the control activity (0 mM caffeine).

FIG. 5.

Polyomavirus growth is independent of p53 and dependent on ATM and SMC1-S957/966-P. MEFs were infected at an MOI of 0.2 to 0.5 PFU/cell and incubated for the indicated times. The virus was titrated by a plaque assay on UC1B cells. (A) p53^+/+ and p53^−/− MEFs. (B) ATM^+/+ p53^−/− and ATM^−/− p53^−/− MEFs. (C) SMC1^WT and SMC1^S957/966A MEFs. For immunoblot analyses of p53, ATM, and SMC1, cells were either treated with 10 nM actinomycin D or infected with polyomavirus for 24 h before cell lysates were prepared.

FIG. 6.

SMC1 enhances viral DNA yields and virus-induced cell cycle arrest. (A) Cells were infected at an MOI of 1 PFU/cell. At 48 and 72 hpi, DNAs were analyzed by Southern blotting. About 10% of SMC1^WT and mutant cells were infected. (B) Cells were serum starved for 24 h in medium containing 1% fetal calf serum, infected at an MOI of 10 to 20, and supplemented with 10% fetal calf serum. At the times indicated, samples were prepared for FACS analysis. The positions of the G₁ and G₂ phases are marked. The percentage of cells in the S and G₂ phases is given on each panel. More than 90% of the cells were infected.

FIG. 7.

Possible ATM pathways contributing to polyomavirus replication (see Discussion).