Large-Scale Study Maps the "Chromatin Landscape" of Acute Myeloid Leukaemia

Acute myeloid leukaemia (AML), an aggressive blood cancer marked by the uncontrolled growth of immature myeloid cells, can be divided into 16 distinct subgroups based on patterns of chromatin accessibility, according to a large study published in Nature.

The research, which analysed samples from 1,563 patients newly diagnosed with AML in Sweden and Japan, used a technique called ATAC-seq (assay for transposase-accessible chromatin with sequencing) to map which regions of DNA are "open" and accessible to the cellular machinery that reads genes. The researchers refer to this large patient collection as the eCHROMA cohort — the Encyclopaedia of Chromatin in AML.

While previous research on AML has focused heavily on gene mutations that drive the disease, the study's authors note that genetic changes alone do not fully explain why AML varies so much between patients, or why outcomes differ. Chromatin — the physical packaging of DNA that determines which genes can be switched on or off — represents a separate, non-genetic layer of regulation that has received comparatively little attention in large cohorts, according to the paper.

Sixteen chromatin subgroups

Using ATAC-seq, the team identified 176,853 recurrent regions of open chromatin shared between the Japanese and Swedish patient groups, most of them located outside gene promoters. These regions overlapped with sites recurrently marked by known chromatin features — including H3K27ac, SMC1, CTCF, RNA polymerase II and H3K27me3 — detected in roughly 200 AML samples, the study reports.

Clustering patients by their chromatin accessibility patterns split the cohort into 16 subgroups. According to the study, each subgroup showed its own distinct combination of driver mutations, differentiation states, gene expression, DNA methylation and "super-enhancer" profiles — clusters of regulatory DNA that strongly boost gene activity. DNA methylation profiling of 424 samples confirmed that each subgroup had its own methylation signature, and that chromatin accessibility and methylation levels were inversely correlated across all subgroups.

Single-cell ATAC sequencing further showed that leukaemic cells within a given subgroup shared a common chromatin accessibility "fingerprint," suggesting each subgroup has a coherent underlying epigenetic identity rather than being a mix of unrelated cell states.

Importantly, the chromatin-based subgroups were linked to clinical outcomes and had a prognostic effect independent of standard genomic classification, the researchers found. The subgroups were also associated with particular sensitivities to drugs, and the findings were validated in separate, independent patient cohorts.

The targeted sequencing conducted alongside the chromatin analysis also reproduced known genomic features of AML reported in earlier studies, while identifying additional, more recently described driver mutations, including in the genes MED12, PHIP, MYB and UBTF.

Taken together, the authors conclude that chromatin profiling combined with multiple layers of molecular data offers insight into AML biology that goes beyond what genetic sequencing alone can reveal, and constitutes what they describe as a valuable resource for future AML research.