Defects of ATM Gene Involving Mutation Lead to Complete Elimination of ATM Function in Chronic Lymphocytic Leukemia

Abstract 3902

Abnormalities of ATM gene are frequent in chronic lymphocytic leukemia (CLL) patients and represent important predictive and prognostic factor. ATM defects are commonly assessed through monitoring of 11q deletion (11q-) using I-FISH. However, 11q- does not mean ATM inactivation if the other allele remains intact, and there are patients who harbor ATM mutation(s) without accompanying 11q-. In CLL, studies addressing the presence of ATM mutations and their impact on ATM protein function are rare due to the extreme gene size and lack of well characterized (hot-spot) mutations. Our aims were to (i) identify ATM mutations in a cohort of predominantly high-risk CLL patients, and (ii) assess their functional consequences.

We used the following complementary methodologies: (a) resequencing microarray (Affymetrix platform), which was designed to detect 1-nt substitutions (i.e. missense, nonsense and splicing mutations) (b) western blotting (WB) to disclose patients with null ATM protein level (c) functional test based on induction of CDKN1A (p21) gene after treatment of CLL cells with fludarabine and doxorubicin in parallel; in case of ATM mutation, the former drug leads to p21 expression induction, while the latter does not (d) direct Sanger sequencing to confirm alterations detected by microarray or mutations indicated by functional test and/or WB.

The resequencing on microarray (which is neutral regarding a function of identified mutations) was performed in 107 predominantly high-risk CLL patients (83% harbored either 11q or 17p deletion). We detected 13 different polymorphisms with distinct frequency among patients and 14 ATM mutations (11 missense, 2 nonsense, 1 splicing) in 13 patients (12%). These mutated samples were subjected to the WB analysis and functional test, which showed a loss of ATM protein and/or its dysfunction in all cases. In order to disclose other potential mutations not detected by the array, 64 patients were also analyzed by WB and functional test. One or both of these tests indicated mutation in 7 patients and in 5 cases mutation(s) were identified (5 short deletions, 2 missense mutations). To obtain a more balanced cohort of CLL patients, we applied WB analysis and functional test on additional 33 samples (not analyzed by microarray) with prevalent TP53-wt/ATM-wt status. Among them, both or one of these tests indicated mutation in 4 patients and in all these patients a mutation was identified by Sanger sequencing. In total, we identified 22 patients (16%) with demonstrable ATM defect (non-functional state and/or null protein level backed up by mutation).

High specificity of our functional test for identification of ATM mutations (16 out of 18 patients with indicated ATM mutation had causative DNA alteration) prompted us to perform analysis of overall cell viability in subgroups “mut-ATM” vs. sole “del-ATM.” Cells with ATM mutation showed clearly increased resistance to doxorubicin but not fludarabine in comparison with cells having sole 11q-.

In conclusion, several complementary methodologies should preferably be used to effectively assess overall ATM status. All mutations identified in our study led to the loss of ATM protein or its function and they occurred predominantly in patients harboring 11q-. The work was supported by grants MUNI/A/0784/2011 and CZ.1.07/2.3.00/20.0045.