Unravelling the Epigenome of Myelodysplastic Syndrome: Diagnosis, Prognosis, and Response to Therapy

Abstract

Myelodysplastic syndrome (MDS) is a malignancy that disrupts normal blood cell production and commonly affects our ageing population. MDS patients are diagnosed using an invasive bone marrow biopsy and high-risk MDS patients are treated with hypomethylating agents (HMAs) such as decitabine and azacytidine. However, these therapies are only effective in 50% of patients, and many develop resistance to therapy, often resulting in bone marrow failure or leukemic transformation. Therefore, there is a strong need for less invasive, diagnostic tests for MDS, novel markers that can predict response to therapy and/or patient prognosis to aid treatment stratification, as well as new and effective therapeutics to enhance patient quality of life and survival. Epigenetic modifiers such as DNA methylation, long non-coding RNAs (lncRNAs) and micro-RNAs (miRNAs) are perturbed in MDS blasts and the bone marrow micro-environment, influencing disease progression and response to therapy. This review focusses on the potential utility of epigenetic modifiers in aiding diagnosis, prognosis, and predicting treatment response in MDS, and touches on the need for extensive and collaborative research using single-cell technologies and multi-omics to test the clinical utility of epigenetic markers for MDS patients in the future.

Keywords: DNA methylation; diagnosis; long non-coding RNA; micro-RNA; myelodysplastic syndrome; prognosis; treatment.

Figure 1

General schematic of DNA methylation and demethylation in somatic cells. (A) A methyl group is added to the 5′ carbon of a cytosine ring by DNMT enzymes giving rise to 5mC. This methyl group is oxidised by TET enzymes resulting in 5hmC which undergoes further oxidisation and base-excision repair (BER) to convert back to an unmodified cytosine. (B) De novo methylation is predominantly carried out by DNMT3A/B enzymes and during DNA replication methylation marks present on the template strand (“old” DNA) are copied onto the daughter strand (“new” DNA) mainly by DNMT1 enzyme. TET1 and TET2 enzymes instigate DNA demethylation via oxidisation.

Figure 2

Schematic depicting interactions between MDS blasts and their bone marrow micro-environment. (A) In MDS, mesenchymal stromal cells (MSCs; green) display global hypermethylation and a global decrease in miRNA expression compared to healthy controls. They also have hypermethylation of SPINT2, resulting in decreased SPINT2 expression leading to adhesion to MDS blasts (yellow) via integrins and increased blast cell survival. Hypermethylation of FRZB is also observed leading to decreased FRZB expression in MSCs which increases WNT/beta-catenin signalling in MDS blasts. MSCs also release exosomes containing miR-10a and miR-15a, resulting in decreased MDM2 and p53 expression in MDS blasts and enhanced viability and clonogenicity. CD8+ T-cells (blue) in the area generally show hypermethylation at the PD-1 promoter, resulting in decreased PD-1 expression and activated T-cells. (B) Treatment with HMAs such as decitabine (DAC) or AZA may result in restoration of methylation levels back to that of healthy controls and potentially lead to increased miRNA expression. HMA treatment results in demethylation of SPINT2 and FRZB promoters, allowing expression of SPINT2 and FRZB and therefore disrupting some interactions between MSCs and MDS blasts, leading to decreased survival of blasts. The effect of HMAs on exosome release from MSCs and subsequent effects on MDS blast function are unknown. Treatment with HMAs also cause demethylation of PD-1 promoters in CD8+ T-cells resulting in increased PD-1 expression and inhibition of T-cell activity. Adapted using information from [161,162,163,164,165,166,167].