Abstract
Deletions in chromosome 11q are an established prognostic marker in CLL. One copy of the ATM gene is deleted in these tumours, as detected by FISH analysis. However, it remains unclear whether the ATM gene is the main tumour suppressor gene that is accounting for the poor outcome in tumours with 11q deletions.
We have recently reported that patients whose tumours have mutations in the ATM gene have an impaired overall and treatment free survival. In our large cohort of 155 patients, tumours with an ATM mutation only partly correlated with tumours with an 11q deletion. We have therefore investigated the relationship between 11q deletions and mutations in the ATM gene. Using the highly sensitive DHPLC method, we have screened the 60 ATM coding exons for mutations in a cohort of 46 tumours, all with a deletion of chromosome 11q. We have found ATM mutations in 19 tumours, indicating a prevalence of 41%.
The ATM protein is vital in the cell’s response to DNA damage including that induced by chemotherapy. ATM acts upstream from p53 and defects in ATM function, like p53, lead to impaired DNA damage induced apoptosis. Furthermore, we have previously shown that loss of ATM function is associated with both in vitro and in vivo chemo-resistance.
Therefore, we next assessed whether the status of the remaining ATM allele in the 11q deleted tumours affected the response to DNA damage. Firstly we induced DNA damage with irradiation and measured both the phosphorylation of ATM protein targets and the induction of p53 dependent transcription responses in representative samples. We found that 11q deleted tumours with a remaining wild type ATM allele had responses that were similar to those seen in tumours with two wild type ATM alleles. In contrast, 11q deleted tumours with a mutation in the remaining ATM allele had defective DNA damage induced responses. We then analysed the effects of in vitro treatment with Fludarabine. First we demonstrated that fludarabine induces ATM dependent phophosphorylation responses in CLL tumours. Then we analysed its effect in the two 11q deleted CLL subgroups. We showed defective phosphorylation responses to fludarabine in the tumours with a mutation in the second ATM allele, but normal responses in those with a second wild type ATM allele.
In summary, we have shown that approximately 40% of CLL tumours with an 11q deletion have a mutation in their remaining ATM allele. Furthermore, we have demonstrated that the 11q deleted tumours appear to form two functional subgroups based on the presence of a mutation in the remaining ATM allele. In contrast to the subgroup with a wild type ATM allele, CLL tumours with a mutant ATM allele have defective in vitro responses to DNA damage with both irradiation and fludarabine. We expect that the functional differences between the two 11q deleted subsets will translate into differences in clinical outcome.
Sentinel interaction mapping – a generic approach for the functional analysis of human disease gene variants using yeast