NEW YORK (GenomeWeb) – Two studies by independent research teams have identified epigenetic subgroups in children with a rare pediatric leukemia called juvenile myelomonocytic leukemia (JMML) that may offer prognostic clues.
For one of the studies, appearing in Nature Communications today, researchers at the German Cancer Research Center, Otto-von-Guericke University in Magdeburg, and elsewhere performed array-based methylation and copy number analyses on samples from 167 children with JMML or Noonan syndrome, a self-limiting myeloproliferative disorder with blood features that closely resemble those found in JMML. From these and other data, they uncovered methylation subgroups that tended to coincide with recurrently mutated genes.
"[O]ur data provide strong evidence for the existence of three methylation subgroups in JMML, which are characterized by distinct clinical and biological features and provide the rationale for future work to dissect the molecular mechanisms underlying the methylation subgroups in JMML," senior author Christoph Plass, a cancer epigenetics researcher at the German Cancer Research Center, and colleagues wrote.
His team began by profiling DNA methylomes in 19 JMML cases and one case of Noonan syndrome with Illumina HumanMethylation450 Bead Chip arrays, identifying two broad methylation clusters that coincided with hematopoietic cell-of-origin types.
When they took a closer look at methylation and mutation patterns in these samples, and samples from another 147 children with JMML or Noonan syndrome, the researchers narrowed in on low-, intermediate-, and high-methylation subgroups that corresponded to mutations in different RAS pathway genes. Ramped up RAS pathway signaling fueled by mutations in genes such as KRAS, NRAS, or PTPN11 has been linked to JMML development previously.
In the JMML subgroup with muted methylation, for example, the team saw an overrepresentation of mutations affecting genes such as NRAS, along with mutations to a development-related gene called CBL.
That subgroup appeared to have more favorable outcomes, the researchers reported, while a subgroup with frequent mutations to PTPN11 and high methylation had poorer prognoses. An intermediate methylation subgroup typically had higher-than-usual levels of KRAS mutations or monosomy involving chromosome 7.
By combining methylome and exome sequence data for a subset of cases, they further fleshed out the genes that are frequently altered, particularly the high- and intermediate-methylation groups. And through additional analyses, the researchers unraveled apparent ties between certain RAS signaling changes, enhanced expression of the DNMT1 and DNMT3B methyltransferase enzyme-coding genes, and hyper-methylation in JMML.
"Our present data on JMML suggest that RAS-activating mutations in different genes might have distinct effects on epigenome remodeling correlating with disease aggressiveness," they wrote.
For their own Nature Communications study, members of team lead by the University of California, San Francisco profiled genome-wide methylation patterns in 39 JMML cases. They, too, found evidence for three JMML methylation clusters, ranging from a low-methylation subgroup with enhanced four-year event-free survival patterns and some spontaneous remissions to high-methylation cases with poorer outcome.
"We found that JMML patients cluster into three subgroups that were independently predictive of outcome based on their methylomes," senior author Mignon Loh, a pediatrics and cancer researcher at UCSF, and her colleagues wrote.
Starting with a pilot analysis using reduced representation bisulfite sequencing on samples from three children with JMML, the team used Illumina arrays to characterize methylation patterns in 39 JMML cases, focusing on more than 1,500 of the most variably methylated CpG sites. The methylation clusters and their apparent ties to JMML outcomes were subsequently validated using samples from 40 individuals with JMML and 22 unaffected controls.
The set of 79 cases considered in that study included 27 patients with methylation clusters that were closer to those found in samples from unaffected controls than in other individuals with JMML, the researchers reported, and more than half of those near-normal methylation cases went on to experience spontaneous disease resolution.
The team noted that the clusters were distinct from those identified using expression data for JMML, though Ras signaling and other pathways appeared to have frequent expression and methylation alterations in JMML cases with relatively poor outcomes.
Even so, the authors cautioned that "after analyzing a small but representative subset of our JMML patients at the gene expression level, we believe that altered DNA methylation in JMML is functioning in a more complex manner than simply altering expression of associated genes."
"[O]ur results show the potential of DNA methylation as a biomarker that can both identify patients who are predicted to fail [hematopoietic stem cell transplantation (HSCT)] as well as those who are most likely to experience spontaneous resolution and could be observed to avoid the acute and late side effects of HSCT," they wrote.