Every so often, an innovative new technology comes along that is powerful enough to reveal entire new worlds. Hans Lippershey’s 1608 refracting telescope was quickly improved by Galileo and Kepler and used to detect extraterrestrial stars, planets, and satellites invisible to the naked eye. Antonie van Leeuwenhoek’s handcrafted microscopes brought previously unknown life forms to the attention of biologists after he published his first observations on “animalcules” in 1673. The introduction of the Iberian caravel in the mid-15th century dramatically improved ocean-going travel and led to rediscovery of Oceania and the “New World” by Europeans. And while DNA sequencing is far from new, it is only now, at the beginning of the 21st century, that unbiased highthroughput nucleic acid resequencing at a practical cost is revealing the molecular engines of human diseases in a way not previously possible, generating novel insights that will take many years to functionally validate and fully understand.
In the field of cancer biology, acute myeloid leukemia (AML) was the first malignancy scrutinized with whole-exome and whole-genome sequencing. Since 2008, AML genome projects coordinated by Timothy Ley, Elaine Mardis, and Richard Wilson at Washington University’s Cancer Genome Initiative (WUCGI) in St. Louis led to the discovery of recurrent disease-associated mutations in genes encoding isocitrate dehydrogenase 1 and 2 (IDH1/2) among other proteins.1,2 Most recently this team reported that DNMT3A, which encodes DNA (cytosine-5)-methyltransferase 3A, is recurrently mutated in AML. This is an enzyme responsible for de novo cytosine methylation essential for processes such as DNA imprinting, modulating gene expression, and X-chromosome inactivation.3
WUCGI investigators analyzed 281 samples from patients with AML finding that 22.1 percent had DNMT3A mutations, including 18 different missense mutations, a smaller number of frameshift, nonsense, and splice-site mutations, and a 1.5-Mbp deletion. DNMT3A mutations were absent in patients with favorable karyotypes but were frequent (33.7%) in patients with intermediate-risk cytogenetic profiles; their presence predicted a markedly shorter survival (12.3 months with mutations vs. 41.1 months without). Surprisingly, there was no clear difference in global DNA methylation levels between AML cells with DNMT3A mutations and those without.
Tim Graubert is leading WUCGI’s efforts in myelodysplastic syndromes (MDS), and his team now report heterozygous DNMT3A mutations with translational consequences in 8 percent of 150 patients with MDS. DNMT3A mutations in MDS, as with AML, conferred a poorer prognosis with a median survival of 14.2 months versus 31.7 months for those without mutations (p=0.02), and a higher risk of progression to AML. Patients with DNMT3A mutations were a median of nine months older than patients with wild-type DNMT3A but otherwise clinically indistinguishable.
In Brief
DNMT3A mutations appear to predict poorer outcomes in both MDS and AML. In MDS, DNMT3A joins a growing list of recurrent disease-associated somatic mutations implicated in altering gene expression by disturbing epigenetic patterning and chromatin conformation, including TET2, EZH2, IDH1/2, ASXL1, UTX, and ATRX. Aberrant splicing patterns of DNMT3A and DNMT3B transcripts have also been reported in related myeloid neoplasms. Although the finding of recurrent DNMT3A mutations suggests a potential link to the clinical responses observed with DNA methyltransferase inhibitors such as azacitidine and decitabine, it is not yet clear whether there is a direct connection between mutations that alter DNA methylation, MDS pathogenesis, and treatment response. Ongoing high-throughput sequencing projects in MDS promise even more discoveries that may better illuminate MDS pathobiology.
References
Competing Interests
Dr. Steensma indicated no relevant conflicts of interest.
