Distinct Methylation Patterns among Infants with MLL Rearranged Acute Lymphoblastic Leukemia.

Acute Lymphoblastic Leukemia (ALL) in infants (i.e. children <1 year of age) is characterized by a high incidence of rearrangements of the MLL gene (∼80%) which is associated with a poor prognosis. The most frequent MLL rearrangements in infant ALL are translocations t(4;11), t(11;19) and t(9;11). Recently, gene expression profiling has established MLL rearranged leukemia as a unique type of leukemia (denoted MLL), that is clearly distinguishable from other ALL subtypes. Currently, these gene expression profiles are slowly revealing important genetic properties underlying this aggressive type of leukemia, however, any epigenetic data on MLL are still lacking. Therefore, the present study was designed to unravel the MLL-specific methylation patterns underlying infant MLL by applying differential methylation hybridization (DMH) using CpG island microarrays containing ∼9000 CpG island probes, in duplicate. Primary infant ALL samples carrying t(4;11) (n=21), t(11;19) (n=17) and t(9;11) (n=6) were compared to infant ALL (n=13) and non-infant pediatric ALL (n=15) samples without MLL rearrangements. In addition, healthy pediatric bone marrow samples (n=9) were included as a reference. Compared to healthy controls, 656 CpG island probes were identified as significantly hypermethylated in t(4;11) positive samples, and 131 CpG island probes in t(11;19) positive samples (p<0.01, false discovery rate <5%). Interestingly, t(11;19) positive ALL patients shared 95% of their methylated probes with t(4;11) patients, suggesting a common methylation pattern which is completely absent in both infant and non-infant ALL patients lacking MLL rearrangements. Remarkably, displaying only a single probe significantly methylated as compared to healthy bone marrow, this common methylation pattern is also absent in t(9;11) positive ALL patients, indicating that based on genome-wide methylation, these patients represent a distinct entity clearly distinguishable from other MLL subgroups. Moreover, the fact that t(4;11) patients exhibit 532 methylated CpG island probes that were not found to be methylated in t(11;19) patients, demonstrates that these patients also exhibit a t(4;11) specific set of methylated genes. Identification of the genes represented by these CpG island probes and subsequent validation of the results obtained in this study is currently being performed (using pyrosequencing and methylation specific PCR analyses). In conclusion, these data reveal that different types of MLL rearranged infant ALL show distinct genome-wide methylation patterns. Specifically, infant ALL patients carrying t(4;11) and t(11;19) are characterized by severe CpG island hypermethylation, as compared to both t(9;11) positive infant ALL patients, as well as pediatric ALL patients lacking MLL rearrangements. Therefore, t(4;11) and t(11;19) patients in particular may well be suitable candidates for DNA methylation inhibiting therapeutic intervention. Finally, these promising results for the first time provide epigenetic insights into the complex biology of infant MLL, and clearly warrant further investigation currently being performed at our laboratory.