Low frequency of DNMT3A mutations in pediatric AML, and the identification of the OCI-AML3 cell line as an in vitro model

Article metrics

Recently, next generation sequencing technology has been applied to discover tumor-specific mutations in acute myeloid leukemia (AML) genomes.1 After the finding of recurrent mutations in the enzyme isocitrate dehydroxygenase 1 (IDH1) in AML,2 two studies recently reported the identification of somatic mutations in the DNA (cytosine-5-)-methyltransferase 3 alpha (DNMT3A) gene in adult AML cases.3, 4 Ley et al.3 reported the presence of DNMT3A mutations in 22% of de novo adult AML cases, after the discovery of this mutation by sequencing a whole AML genome. They found that DNMT3A mutations were highly associated with cytogenetically normal (CN)-AML (37% (44 of 120) CN-AML cases). Yan et al.4 detected DNMT3A mutations by sequencing the complete coding region of the genome (exome sequencing) of nine AML-M5 samples. They found that these mutations were restricted to the myelomonocytic (French–American–British (FAB)-M4) and monocytic (FAB-M5) AML subtypes, presenting in 13.6 and 20.5% of these cases, respectively. DNMT3A mutations were localized in the methyltransferase and the plant homeo domain fingers, and impaired DNMT3A methyltransferase activity, or altered histone H3 affinity in vitro. Both studies reported a poor clinical outcome for patients with DNMT3A-mutated AML.3, 4

We identified the presence of a somatic heterozygous R882C mutation in the DNMT3A gene by performing exome sequencing of a pediatric AML case and confirmed this mutation by Sanger sequencing (Figure 1). This index case concerned an 8-year-old boy with a CN-AML of the FAB-M1 subtype, who is currently in continuous complete remission 3.6 years after diagnosis, following treatment according to the AML MRC15-protocol. Furthermore, we identified a Wilms tumor 1 (WT1) R394W mutation in this patient, resulting from a missense mutation in exon 9, which was known to be present from previous screening for molecular mutations. This case was also characterized by a large insertion/deletion in WT1 exon 7 and an internal tandem duplication in the FLT3 gene (Table 1).

Figure 1

Sequence chromatograms of the DNMT3A mutations detected in the pediatric AML index case and the OCI-AML3 cell line. A heterozygous mutation (C>T), changing arginine into cysteine at codon 882, is present in the pediatric AML index case and the OCI-AML3 cell line. The germline sample of the index case did not harbor this mutation. Mutated nucleotides are indicated by arrows.

Table 1 Clinical and genetic characteristics of the three DNMT3A-mutated pediatric AML cases

As data on DNMT3A mutations in pediatric AML were not yet available, we subsequently screened cDNA of a large representative pediatric AML series (n=140; including 34 FAB-M4 and 27 FAB-M5 cases, and including 46 CN-AML cases) from the Dutch Childhood Oncology Group and AML-Berlin Frankfurt Münster Study Group studies for DNMT3A mutations in the region, including amino acids 460–912, in which all but one of the previously reported mutations were found. Remarkably, only two additional cases with DNMT3A mutations were detected; one case harbored a R484W mutation, and the other case harbored a V716F mutation. The characteristics of these cases are presented in Table 1. Of interest, two of the three pediatric cases with DNMT3A mutations concerned an AML of the FAB-M1 subtype, which is in contrast with the (myelo)monocytic morphology predominantly found in adult DNMT3A-mutated AML cases.3, 4 Unlike the poor clinical outcome of adults with DNMT3-mutated AML, all three pediatric cases were in continuous complete remission at last follow-up. In conclusion, the estimated frequency of DNMT3A mutations in our pediatric AML series is only 2.1% (95% confidence interval (CI) 0–4.5%), and 6.4% (95% CI 0–13.4%) in the CN-AML subset. In line with our data, Ho et al.5 recently published that no DNMT3A mutations were observed in their pediatric AML series (n=180). Consistent with this low frequency in children, both adult AML studies reported a high mean age (54.9 and 53.1 years, respectively) for DNMT3A mutants, indicating that these mutations are associated with a disease onset at advanced age.3, 4 Of note, our frequency might be slightly underestimated, as we performed mutational screening on cDNA, and Ley et al.3 showed that 2 out of 21 (10%) mutated alleles were not expressed. However, in the study of Yan et al.,4 all 23 DNMT3A-mutated alleles were expressed.

Furthermore, we screened 12 AML cell lines derived from the German Resource Centre for Biological Material (DSMZ, Germany), and found that the OCI-AML3 cell line harbored a DNMT3A R882C mutation (Figure 1). The OCI-AML3 cell line was derived from a 57-year-old male with FAB-M4 AML, and carries an NPM1 mutation.6 Hence, this cell line can be used as an in vitro model to further study the leukemogenic and drug-resistance aspects of DNMT3A mutations in AML.

Our findings further illustrate the large differences in the frequency of genetic aberrations found in pediatric and adult AML. NPM1 mutations are highly associated with DNMT3A mutations in adult AML,3, 4 which is consistent with our finding of a DNMT3A mutation in one pediatric NPM1-mutated case and in the NPM1-mutated OCI-AML3 cell line. In pediatric AML, there is a four to five-fold lower frequency of NPM1 mutations, compared with adult AML,7 which may partially explain why the frequency of DNMT3A mutations is also lower. Moreover, pediatric AML is characterized by a different base pair insertion in NPM1 as compared with the type of NPM1 mutations found in adult AML,7, 8 which points towards a different ontogeny of pediatric AML. Furthermore, IDH1/2 mutations also occurred more frequently in adult DNMT3A-mutated cases,3 but these mutations are, similar as NPM1 mutations, less frequent found in pediatric AML.9 None of our pediatric DNMT3A mutants carried an IDH1/2 mutation. Because the recently discovered mutations in AML are less frequent or even absent in pediatric AML, further genome-wide sequencing studies in pediatric AML are warranted separately from adult AML to map genetic aberrations underlying pediatric AML specifically. The discovery of these aberrations is needed to further improve the current stagnated survival rates of pediatric AML, because they provide insight in leukemogenesis and might serve as druggable targets. As genome-wide sequencing studies have become more affordable in recent years, they promise to be a valuable tool to further dissect pediatric AML.


  1. 1

    Ley TJ, Mardis ER, Ding L, Fulton B, McLellan MD, Chen K et al. DNA sequencing of a cytogenetically normal acute myeloid leukaemia genome. Nature 2008; 456: 66–72.

  2. 2

    Mardis ER, Ding L, Dooling DJ, Larson DE, McLellan MD, Chen K et al. Recurring mutations found by sequencing an acute myeloid leukemia genome. N Engl J Med 2009; 361: 1058–1066.

  3. 3

    Ley TJ, Ding L, Walter MJ, McLellan MD, Lamprecht T, Larson DE et al. DNMT3A mutations in acute myeloid leukemia. N Engl J Med 2010; 363: 2424–2433.

  4. 4

    Yan XJ, Xu J, Gu ZH, Pan CM, Lu G, Shen Y et al. Exome sequencing identifies somatic mutations of DNA methyltransferase gene DNMT3A in acute monocytic leukemia. Nat Genet 2011; 43: 309–315.

  5. 5

    Ho PA, Kutny MA, Alonzo TA, Gerbing RB, Joaquin J, Raimondi SC et al. Leukemic mutations in the methylation-associated genes DNMT3A and IDH2 are rare events in pediatric AML: a report from the Children's Oncology Group. Pediatr Blood Cancer 2011; 57: 204–209.

  6. 6

    Quentmeier H, Martelli MP, Dirks WG, Bolli N, Liso A, Macleod RA et al. Cell line OCI/AML3 bears exon-12 NPM gene mutation-A and cytoplasmic expression of nucleophosmin. Leukemia 2005; 19: 1760–1767.

  7. 7

    Hollink IH, Zwaan CM, Zimmermann M, Arentsen-Peters TC, Pieters R, Cloos J et al. Favorable prognostic impact of NPM1 gene mutations in childhood acute myeloid leukemia, with emphasis on cytogenetically normal AML. Leukemia 2009; 23: 262–270.

  8. 8

    Thiede C, Creutzig E, Reinhardt D, Ehninger G, Creutzig U . Different types of NPM1 mutations in children and adults: evidence for an effect of patient age on the prevalence of the TCTG-tandem duplication in NPM1-exon 12. Leukemia 2007; 21: 366–367.

  9. 9

    Damm F, Thol F, Hollink I, Zimmermann M, Reinhardt K, van den Heuvel-Eibrink MM et al. Prevalence and prognostic value of IDH1 and IDH2 mutations in childhood AML: a study of the AML-BFM and DCOG study groups. Leukemia 2011; e-pub ahead of print 7 June 2011.

Download references


We acknowledge funding from the Childhood Oncology Foundation Rotterdam, The Netherlands.

Author information

Correspondence to C M Zwaan.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Rights and permissions

Reprints and Permissions

About this article

Further reading