DNMT3A Mutations Are Connected with Lower DNA Methylation Levels and Numbers of Concurrently Hypermethylated Genes in AML Patients

Abstract 4629

Recently, mutations in DNA methyltransferase 3A (DNMT3A) have been found in patients with acute myeloid leukemia (AML) and have been confirmed to be connected with adverse clinical outcome. As DNMT3A plays direct role in the process of DNA methylation by adding methyl group to the cytosine residue of CpG dinucleotides, the question is obvious: What is the impact of DNMT3A mutations on DNA methylation levels?

We examined 79 AML patients at diagnosis for the presence of aberrant DNA methylation of 12 tumor suppressor genes (TSG) (CDKN2B, CALCA, CDH1, ESR1, SOCS1, MYOD1, DAPK1, TIMP3, ICAM1, TERT, CTNNA1, EGR1) by methylation specific real-time PCR (MethyLight) and for mutations in the gene DNMT3A by direct sequencing. Next we studied methylation status of 24 HOX genes from all four clusters A-D using methylation-restriction endonucleases followed by RQ-PCR arrays in 10 AML samples compared to 4 healthy donor samples.

Sequencing of cDNA between amino acids 300 and 930 revealed that 32 of 79 AML patients had DNMT3A mutation. The reason for higher DNMT3A mutation incidence in our patients’ cohort consists in preferential selection of AML patients with a higher percentage of probability of DNMT3A mutation (it means normal karyotype and mutations in NPM1, FLT3 and/or IDH1/2 genes). MethyLight assessment of 12 TSG showed subsequent frequencies of hypermethylation: CDKN2B (47%), CALCA (43%), CDH1 (22%), SOCS1 (24%), MYOD1 (18%), ESR1 (14%). The remaining 6 genes were weakly methylated in less than 10 % AML patients at diagnosis and were therefore excluded from further analysis. DNA methylation arrays revealed a set of differentially methylated HOX genes (n=12): 11 HOX genes were hypermethylated (HOXA4, HOXA6, HOXB13, HOXB3, HOXB4, HOXB7, HOXB8, HOXC8, HOXD10, HOXD11, HOXD3) and HOXA5 was hypomethylated compared to healthy donor samples. Comparing overall cumulative DNA methylation levels and numbers of simultaneously hypermethylated genes to mutational status of DNMT3A gene, we did observe lower levels of DNA methylation (P<0.0001) as well as displaying lower numbers of concurrently hypermethylated genes (P<0.0001) in patients with DNMT3a mutations. We observed the same trend also in DNA methylation levels of HOX genes when compared mutant (n=4) versus wild-type (n=6) DNMT3A patients.

These results clearly show that numbers of simultaneously hypermethylated genes and DNA methylation levels of chosen tumor suppressor genes (TSG) as well as HOX genes differs between AML patients with wild type and mutant DNMT3A.

This study is part of the COST Action BM0801 (EuGESMA) and is supported by NS10632-3/2009, OC10042 and IHBT00023736.