In this report we characterized the relationship between the ATM- and ATR-mediated DDR and lytic BKPyV infection. Our results indicated that both branches of the DDR were activated by BKPyV; moreover, ATM and ATR functioned in parallel and contributed to the activation of individual DDR pathways (). In particular, ATM was mainly responsible for activating pChk2, whereas ATR was more important in activating pChk1 and also might be important for regulating TAg and its modified forms. Our results showed that both ATM and ATR were required to achieve maximal viral DNA replication and infectious progeny production. This is the first such report for any polyomavirus family member to our knowledge. Knocking down each individual kinase resulted in a partial defect in viral replication and the double knockdown had a more dramatic effect. When comparing ATM and ATR double knockdown cells to control cells, the degree of decrease in viral titer was consistently greater than the decrease in viral DNA, suggesting that there may be some post-DNA replication regulation that involves both ATM and ATR function. For example, ATR may be involved in alleviating the DNA replication stress generated during the resolution of the two daughter circular BKPyV chromosomes, and therefore ATR knockdown caused a more marked defect in viral progeny production than in DNA replication. Our results also showed that in the absence of ATM or ATR, severe chromosome damage accumulated upon BKPyV infection. These data point out that another possible function of ATM and ATR during BKPyV infection is to help repair the DNA damage caused by BKPyV ().
Model of the roles of the DDR during BKPyV infection.
The exact role of ATM during polyomavirus early gene expression remains unclear, but it may be different depending on the specific virus and cell type. For example, one report showed that ATM knockdown in African green monkey CV1 cells reduced the level of SV40 TAg-pS120, but it did not have much effect on the total level of TAg 
. In another report using BSC40 monkey kidney cells, the ATM inhibitor KU55933 reduced total SV40 TAg levels 
. Our ATM knockdown data suggest that ATM is not required for TAg expression during BKPyV infection. Consistent with this, BKPyV TAg expression was similar in wild-type and ATM-knockout mouse embryonic fibroblast cell lines (data not shown). Unfortunately we cannot study the entire BKPyV life cycle in mouse cells due to a block to DNA replication 
Our data demonstrate the importance of ATM and ATR during BKPyV DNA replication, DDR activation, and progeny production. It is possible that ATM, ATR and downstream effectors directly participate in replication-related events. For example, ATM has been reported to phosphorylate SV40 TAg in vivo
and therefore is required for optimal SV40 replication 
. In our ATR single knockdown or ATM+ATR double knockdown cells, we observed a decrease in a modified form of TAg (TAg*) at 3 dpi, and the appearance of a third band of TAg at 2 dpi (). Preliminary experiments indicated that phosphatase treatment did not affect the level of TAg or TAg* (data not shown), but this does not definitively rule out the possibility that TAg* may represent a phosphorylated form of TAg. Additionally, proteins involved in homologous recombination or the Fanconi anemia pathway (required for repair of stalled replication forks) have been found to be necessary for SV40 replication 
. It is possible that DDR proteins that are downstream of ATM and ATR directly contribute to BKPyV DNA replication. Alternatively, instead of being directly involved in viral replication, ATM and ATR may participate indirectly in BKPyV infection by affecting cell cycle status. For example, ATM and ATR activation leads to G2 arrest in JCPyV-infected human neuroblastoma cells and oligodendrocytes 
. It is hypothesized that G2 arrest contributes to JCPyV replication by maintaining the cellular replication machinery and preventing mitosis 
. Our cell cycle analyses also showed that BKPyV infection of RPTE cells results in the accumulation of polyploid cells, consistent with the idea that mitosis is inhibited to allow for maximal DNA replication.
A number of DDR proteins including γH2AX and Mre11 were re-localized into small TAg-containing foci during BKPyV infection. Similar co-localization between DDR proteins and TAg foci is seen during SV40 infection, although the MRN proteins are eventually degraded during SV40 infection 
. It has been hypothesized that these foci represent sites of viral DNA replication because they also contain proteins that are required for viral replication including RPA and Pol-prim, but not Pol-prim-associated host replication factor 
. In addition to the small, bright foci of γH2AX, there was also an increase in pan-nuclear staining of γH2AX in BKPyV-infected cells. Although the functional significance of this increase is not clear, it has previously also been reported in other DNA virus infections such as adenovirus and AAV 
. It is thought that this pan-nuclear staining increase may represent modification of histones on cellular chromatin and that the modification can be stimulated by viral replication.
Intriguingly, the induction of γH2AX by BKPyV still occurred in ATM, ATR, and DNA-PKcs triple knockdown cells, suggesting that there might be other cellular kinase(s) responsible for phosphorylating this molecule. Alternatively, it is possible that BKPyV infection alters a cellular phosphatase activity that, together with residual PI3KK activity due to incomplete knockdown, leads to an increase in the steady state level of γH2AX. One phosphatase candidate is PP2A, which has been demonstrated to dephosphorylate γH2AX in an ATM-, ATR-, and DNA-PKcs-independent manner 
. Polyomavirus small T antigen is well known for its interaction with PP2A and its ability to inhibit PP2A enzymatic activity 
. It will be interesting to determine the mechanism of γH2AX induction and its functional significance during BKPyV infection in the future.
What triggers the activation of both the ATM- and ATR-mediated DDR during BKPyV infection requires more careful examination. It is possible that a viral protein(s) alone is able to achieve the induction. For example, expression of SV40 TAg without a viral replication origin in normal human BJ/tert fibroblasts induces both an ATM- and ATR-mediated DDR, and this induction is dependent on the interaction of TAg with Bub1, a mitotic spindle checkpoint kinase 
. Moreover, polyomavirus TAg alone is able to induce cellular DNA damage as judged by comet assays and cytogenetic analyses 
. For JCPyV, the ability of TAg to associate with cellular DNA is important for TAg induction of G2 cell cycle arrest 
. It is also possible that either incoming or replicating viral genomes serve as the trigger for the DDR. For example, both wild-type and UV-inactivated AAV2, but not recombinant AAV2 vectors, are capable of inducing a DDR, suggesting that it is the viral DNA sequence, but not the viral capsid, that is responsible for the activation of DDR 
In conclusion, our results demonstrate the unique activation of various DDR components upon BKPyV infection and the essential roles of both ATM and ATR for viral replication and growth. The study of BKPyV infection and the DDR not only reveals novel knowledge about the cellular pathways with which the virus needs to interact in order to complete the lytic life cycle, but may also have important clinical implications for BKPyV reactivation and its related disease. For example, BKPyV reactivation is a severe problem in bone marrow transplant patients, who might have experienced radiation as part of the preparative regimen. Thus, research focusing on BKPyV and DDR may shed light on the analysis of the causality of BKPyV reactivation in these patients.