The DNA damage response is initiated by Phosphoinositide 3-Kinase-Related Kinases (PIKKs) ATR (ATM- and Rad3-Related), DNA-PK (DNA-Dependent Protein Kinase), and ATM (Ataxia Telangiectasia Mutated), the focus of this review. Most of our understanding of PIKK activation and effects on the DNA damage response are informed by experiments in dividing cells, where these enzymes are activated by DNA double-strand breaks (ATM and DNA-PK) or replication stress. Once activated, PIKKs phosphorylate target proteins directly involved in DNA repair, cell cycle regulation, metabolism, and other processes favorable to cell population survival following DNA damage [1], [2]. PIKKs share a phosphorylation motif S/T-Q, a serine or threonine followed by a glutamine. The simplicity of the motif means that the targets for ATR, DNAPK, and ATM are numerous and in some cases are partially overlapping [3], [4], [5], [6]. The kinases also target each other when activated [7], [8], [9], [10], adding to the complexity of DDR pathway regulation.

ATM is canonically activated during DSB repair by the Mre11/Rad50/Nbs1 (MRN) protein complex [11], [12]. Specifically, MRN binds to double-strand breaks coincident with ATP-dependent local unwinding of the DNA ends [13], [14], [15]. Association of the inactive ATM homodimer with DNA end-bound MRN activates the kinase, inducing monomerization, auto-phosphorylation, and an increase in affinity for ATM substrates [16], [17], [18], [19]. ATM can also be activated through a non-canonical mechanism by reactive oxygen species (ROS) [20]. ATM activation by ROS is independent of MRN and involves a redox-formed disulfide bridge between cysteine residues on each ATM monomer, resulting in activation as an ATM dimer. Recent structural analysis of the oxidized dimer shows that disulfide bond formation in the c-terminal FATC domain induces rotation and displacement of an auto-inhibitory loop, thus promoting access of substrates to the active site [21]. We have previously demonstrated that the MRN-dependent and oxidation-dependent pathways of ATM activation are genetically separable and regulate a diverse set of cellular responses including cell cycle checkpoint arrest, DNA damage survival, autophagy, mitochondrial turnover, and protein homeostasis [22].

ATM was first discovered in the context of Ataxia Telangiectasia (A-T), an autosomal recessive, childhood-onset spinocerebellar ataxia caused by mutations in the ATM gene [23]. Symptoms include cerebellar neurodegeneration, ataxia, oculomotor apraxia, increased cancer susceptibility, and immunodeficiency. Patients typically show symptoms at an early age (1 to 4 years), and usually live until early adulthood. The most striking phenotype is the cerebellar neurodegeneration which is understood to cause the severe ataxia. Cerebellar Purkinje neurons appear to die disproportionately in A-T, and thinning of the granule cell layer is also observed [23], [24]. Why ATM loss disproportionately affects the cerebellum or Purkinje cells is unknown, although many disorders associated with DNA repair are specifically linked to cerebellar dysfunction [25], [26].

A hypomorphic mutant of ATM found in several "variant" forms of A-T alleles (R3047X), exhibits deficiency only in the ROS activation pathway of ATM [20], [27], [28], [29]. Patients expressing this mutant exhibit neurodegeneration similar to that of classical A-T patients despite showing greatly reduced immunodeficiency and radiation sensitivity, thus we infer that ROS activation is a function of ATM that is essential for maintaining cerebellum integrity.

The molecular events leading to cerebellum neurodegeneration in A-T patients are not understood, but recent data in animal models, human neuronal cells in culture, and patient autopsy tissue supports the idea that changes in chromatin accessibility and alterations in transcription patterns accompany loss of cerebellum function in the absence of ATM function. Whether these changes are direct causes of cerebellum neurodegeneration is a topic of ongoing debate. Here we review contributions from several groups using a variety of model systems and approaches and discuss the intersections between DNA damage, innate immunity signaling, transcriptional changes, and other alterations that occur at the chromatin level with ATM loss.