ith so much modern molecular biology utterly dependent on mutations of various kinds it is easy to lose sight of how faithful DNA duplication is normally. Millions upon millions of base pairs are copied accurately for every mistake, and damage to the DNA is generally repaired before it can do real damage to the cell. At #3 in the current Top Ten is a paper elucidating a crucial component in the system that repairs potentially the most destructive damage of all, double-stranded breaks.

Existing evidence pointed to the idea that activation of ATM is itself caused by phosphorylation, and that this is an auto-phophorylation, that ATM activates itself. Bakkenist and Kastan digested active ATM with trypsin and looked for phosphorylated peptides. They found a single newly phosphorylated peptide, in which serine was the target amino acid. But which serine? After deploying a barrage of standard techniques, the researchers were no nearer the answer. Eventually, by a careful process of elimination they identified a tryptic sequence that contained a phosphorylated serine at position 1981 in ATM. This single amino acid is conserved across a wide range of species, from toad to mouse to human. Having identified the target, Bakkenist and Kasten used antibodies to show that within 30 minutes of ionizing radiation almost all the ATM in the cell has become activated. Ultra-violet radiation causes a similarly rapid response.

ATM is a large protein; it was a tour de force to locate the single target amino acid. But there’s more: serine 1981 is part of the ATM domain called FAT. In the normal cell, without DNA damage, two ATM molecules line up head to tail as a dimer. The kinase domain of one molecule is bound to, and thus inactivated by, the FAT domain of the other, and vice versa. This is a novel method for inactivating a kinase in the cell. Something about double-stranded breaks in the DNA triggers autophosphorylation of serine 1981 by the kinase opposite. The two molecules of ATM separate and each is now free to phosphorylate its targets.

The mechanism is incredibly sensitive. Bakkenist and Kastan dosed cells with radiation that would cause just one or two double-stranded breaks per cell, and that was enough to trigger ATM to autophosphorylate. The response was so swift that the researchers reasoned that ATM must be sensitive to a stimulus distant from the actual break. That stimulus seems to be a change in the shape of the chromatin, the complex structure formed by the DNA as it binds with proteins called histones. Drugs that loosen chromatin but do not damage DNA cause ATM to activate and to activate p53, a crucial switch in the cell’s choice between delayed division or programmed cell death. But other ATM targets, which are directly involved in repairing broken DNA, are not phosphorylated. Other molecules must direct DNA repair and, perhaps, convey the message from the distorted chromatin to the ATM dimers.