53BP1 mediates the fusion of mammalian telomeres rendered dysfunctional by DNA-PKcs loss or inhibition.

Abstract

Telomere dysfunction promotes genomic instability and carcinogenesis via inappropriate end-to-end chromosomal rearrangements, or telomere fusions. Previous work indicates that the DNA Damage Response (DDR) factor 53BP1 promotes the fusion of telomeres rendered dysfunctional by loss of TRF2, but is dispensable for the fusion of telomeres lacking Pot1 or critically shortened (in telomerase-deficient mice). Here, we examine a role for 53BP1 at telomeres rendered dysfunctional by loss or catalytic inhibition of DNA-PKcs. Using mouse embryonic fibroblasts lacking 53BP1 and/or DNA-PKcs, we show that 53BP1 deficiency suppresses G1-generated telomere fusions that normally accumulate in DNA-PKcs-deficient fibroblasts with passage. Likewise, we find that 53BP1 promotes telomere fusions during the replicative phases of the cell cycle in cells treated with the specific DNA-PKcs inhibitor NU7026. However, telomere fusions are not fully abrogated in DNA-PKcs-inhibited 53BP1-deficient cells, but occur with a frequency approximately 10-fold lower than in control 53BP1-proficient cells. Treatment with PARP inhibitors or PARP1 depletion abrogates residual fusions, while Ligase IV depletion has no measurable effect, suggesting that PARP1-dependent alternative end-joining operates at low efficiency at 53BP1-deficient, DNA-PKcs-inhibited telomeres. Finally, we have also examined the requirement for DDR factors ATM, MDC1 or H2AX in this context. We find that ATM loss or inhibition has no measurable effect on the frequency of NU7026-induced fusions in wild-type MEFs. Moreover, analysis of MEFs lacking both ATM and 53BP1 indicates that ATM is also dispensable for telomere fusions via PARP-dependent end-joining. In contrast, loss of either MDC1 or H2AX abrogates telomere fusions in response to DNA-PKcs inhibition, suggesting that these factors operate upstream of both 53BP1-dependent and -independent telomere rejoining. Together, these experiments define a novel requirement for 53BP1 in the fusions of DNA-PKcs-deficient telomeres throughout the cell cycle and uncover a Ligase IV-independent, PARP1-dependent pathway that fuses telomeres at reduced efficiency in the absence of 53BP1.