Promoter Hypermethylation in MLL-r Leukemia: Biology and Therapeutic Targeting.

Tumor suppressor silencing via promoter CpG island hypermethylation is important in oncogenesis. DNA methyltransferase (DNMT) maintains promoter hypermethylation, and DNMT inhibitors (DNMTis) are of proven benefit for MDS and are in trials for AML and CML. Infant ALL is a rare, aggressive and distinctive subset of childhood ALL with 80% of cases harboring MLL gene rearrangements (MLL-r). Cooperating events leading to leukemia in these cases are largely unknown and recent studies suggest that oncogenesis is related to neither copy number alterations nor mutations in the tyrosine kinome. MLL has multiple domains with epigenetic activity, and silencing of several single tumor suppressor genes in MLL-r infant ALL cases has been described. We explored the role of genome-wide CpG island hypermethylation in MLL-r infant ALL, with a focus on its potential role as a novel molecular target for therapy.

HELP (HpaII-tiny fragment Enrichment by Ligation-mediated PCR), which accurately measures the abundance of DNA methylation at ∼30,000 CpG sites covering ∼13,000 promoter regions, was used to examine a cohorts of MLL-r infant ALL primary samples (n=5), other childhood ALL primary samples (TEL-AML1+, n=2 and hyperdiploid, n=3) and normal controls (CD34-selected- (n=2) and CD19-selected- (n=3) cord blood). To correlate HELP results with gene expression, we performed qRT-PCR on cDNA from the same samples for a set of 10 “genes of interest”, which were selected due to differential methylation in the HELP assay (n=6; DAPK1, DAXX, CASP9, LIFR, CCR6, HRK), or due to known biological significance (n=4; FLT3, HOXA9, MEIS1, FHIT). GAPDH and ABL were used as housekeeping genes. To explore the therapeutic potential of DNMTi in infant ALL, we treated MLL-r cell lines (n=3; SEMK2, KOPN8, HB1119) and an MLL-wt cell line (n=1; NALM6) with various concentrations (0, 0.5, 1, 2, 4 mM) of decitabine over multiple durations (8, 24, 48 and 72 hours), then performed MTT assays and qRT-PCR on the genes of interest for each concentration/time point combination. To validate the HELP assay findings and to determine whether the effects of decitabine on re-expression of silenced genes was due to DNMT activity, we also performed methylation specific PCR (MSP) for 3 of the genes of interest (DAPK1, CCR6, HRK) on the primary samples and the cell lines.

Unsupervised analysis of the HELP assay showed tight clustering of samples into their known biological groups, indicating large differences in global methylation patterns, and the ability of these patterns to accurately sub-classify ALL samples. Global hypermethylation was seen in the MLL-r cohort compared to both the normals and the other ALLs, with ratios of significantly (p<0.01) hyper- to hypomethylated CpG's of 2.1 and 2.2, respectively.

Of the 10 genes assayed by qRT-PCR, 4 (FLT3, MEIS1, FHIT, and DAPK1) showed marked and significant (p<0.05) differences in expression by cohort. As expected, FLT3 and MEIS1 were over-expressed in the MLL-r cohort, while FHIT and DAPK1 were under-expressed. LIFR was silenced in all samples. CCR6 and HRK were relatively under-expressed in the MLL-r group

In the cell line treatment assays with decitabine, MLL-r cell lines showed dose and time dependent cytotoxicity and re-expression of 3 of the 4 silenced or under-expressed genes[DAPK1, CCR6, HRK]. These phenomena were not seen in the MLL-wt cell line. For the 3 re-expressed genes, MSP confirmed preferential promoter methylation in MLL-r samples at baseline, and a relative decrease in methylation after treatment with decitabine.

MLL-r infant ALL samples demonstrate global hypermethylation. The HELP assay can reliably identify individual hypermethylated genes that can be confirmed to be hypermethylated by MSP and can be shown to be silenced or under-expressed at the transcript level. Treatment with DNMTi demonstrates selective cytotoxicity for MLL-r cell lines in a dose and time dependent manner that correlates with baseline hypermethylation/silencing of key genes, and correlates temporally with demethylation and re-expression of these same hypermethylated genes. DNMTi may represent a novel molecularly targeted therapy for infant ALL.

No relevant conflicts of interest to declare.