The extremes of linear chromosomes are dangerous locations. Their structural resemblance to double-stranded DNA breaks can trigger unwanted DNA-damage-response pathways with catastrophic results for the cell. Telomeric DNA is therefore enveloped by a 'cap' of protective proteins that conceal these troublesome ends. Now, researchers have shed new light on the function of one of these telomere-binding proteins — TRF2.

TRF2 is a telomeric double-stranded-DNA-binding protein that has a role in the protection of chromosome ends, but its mechanism of action is unknown. A dominant-negative mutation of TRF2 that displaces the protein from its binding site leads to telomeric association of the ATM kinase — a critical mediator of the DNA-damage-response pathway. These findings indicate that TRF2 might oppose the action or the recruitment of ATM.

To investigate this further, Titia de Lange and co-workers overexpressed TRF2 in human primary fibroblasts, exposed these cells to ionizing radiation (IR) and then analysed the ATM-mediated response to DNA damage. Microscopic analysis revealed that an increased percentage of these cells had entered mitosis — indicating a failure of ATM-mediated cell-cycle arrest. Quantitative immunoblots showed decreased levels of the p53 protein and its downstream targets, and ATM-dependent phosphorylation of NBS1 — a DNA-double-strand-break-repair protein — was also impaired in these cells.

DNA damage that is induced by low levels of IR is usually associated with activation of ATM through autophosphorylation. However, when the authors co-expressed TRF2 and ATM, the level of phosphorylated ATM after IR exposure was reduced compared with controls. Similarly, endogenous ATM phosphorylation was reduced in cells that overexpressed TRF2.

So, does TRF2 directly interfere with ATM activation and function? Anti-ATM antibody co-precipitated endogenous and exogenous TRF2 in primary human fibroblasts, which indicates that these proteins interact in vivo. GST pulldown using fusion proteins that comprised different fragments of ATM showed that TRF2 bound to a specific domain of ATM, close to serine 1981 — the main site of autophosphorylation. Finally, immunofluorescence of IR-treated primary fibroblasts that overexpressed TRF2 detected the protein at telomeric foci, but not at chromosomal sites of DNA damage.

Based on these results, the authors propose a model in which TRF2 binds to ATM and inhibits its activation by preventing phosphorylation at S1981. Importantly, the subnuclear location of TRF2 indicates that this inhibition is restricted to telomeres, which would allow ATM to mediate vital DNA repair elsewhere in the nucleus.