Development Of t-MDS In Patients Undergoing Autologous Transplantation For Lymphoma Is Not Associated With Increased Frequency Of Mitochondrial DNA Mutations

Therapy-related myelodysplasia/acute myeloid leukemia (t-MDS) is a lethal complication of cytotoxic cancer treatment. The incidence of t-MDS is particularly high in patients undergoing autologous hematopoietic cell transplantation (aHCT) for Hodgkin lymphoma (HL) or non-Hodgkin lymphoma (NHL). To better understand pathogenetic mechanisms underlying development of t-MDS we are prospectively following a cohort of patients undergoing aHCT for lymphoma at our center. In previous studies using samples from this cohort we identified altered expression of genes related to mitochondrial function in pre-aHCT samples from lymphoma patients who later developed t-MDS (cases) compared to patients from the cohort who did not develop t-MDS (controls) (Cancer Cell 2011 20:591). Altered mitochondrial gene expression was associated with increased levels of mitochondrial reactive oxygen species (ROS) in samples from patients who subsequently developed t-MDS. These results suggest a possible role for mitochondrial dysfunction and increased ROS in pathogenesis of t-MDS. Somatic mutations of mitochondria DNA (mtDNA) are observed at increased frequency in many malignant conditions. Somatic mtDNA mutations may reflect increased susceptibility to mutagenesis, and could contribute to altered mitochondrial function in patients with t-MDS.

Here we sought to investigate the role of mitochondrial genomic instability in t-MDS development by analyzing the mutation profile of mtDNA in hematopoietic cells from pre-HCT samples obtained from cases that developed t-MDS and controls that did not develop t-MDS after aHCT for lymphoma. We isolated myeloid and lymphoid cell populations from t-MDS cases (n=13) and controls (n=18) using flow cytometry, amplified mtDNA by PCR using specific primers, and sequenced pooled, barcoded amplicons using next generation sequencing on a Illumina Hi-Seq instrument. Sequences were alighted to the revised Cambridge Reference Sequence (rCRS) in the MITOMAP database.

We did not identify significant differences in abundance of variations between controls and cases in protein coding genes (control 13.9, case 14.9), hypervariable region (control 7.7, case 6.6) and rRNA/tRNA genes (control 5.7, case 5.6). In addition we did not observe differences in the abundance of variation between cases and controls in individual genes. We also did not detect significant differences in the abundance of SNPs resulting in non-synonymous amino acid changes in lymphoid compared to myeloid populations from both cases (lymphoid 5.0, myeloid 4.7) and controls (lymphoid 5.3, myeloid 4.3). Several new variations in both coding and noncoding regions were identified that were not previously reported in the rCRS database, which were present in greater abundance in myeloid (control 5.3, case 5.3) compared to lymphoid cells (control 1.7, case 2.2). Amongst new variations, the top 5 protein coding genes with the highest number of mutations were found in myeloid cells and included NADH dehydrogenase subunit 2, 4, and 5, and cytochrome c oxidase subunit 1 and 2. We also analyzed the specific mutations that displayed a variation frequency greater than 25% in myeloid control and case samples. Interestingly, the A10398G mutation, in the NADH dehydrogenase subunit 3 (ND3) coding region, associated with metabolic syndrome and increased risk for breast cancer, was found in ∼30% of t-MDS myeloid case samples.

In summary our data suggests that there is increased abundance of novel variations in mtDNA in myeloid compared to lymphoid populations. Although several new, previously unreported variations were identified, we did not observe any significant differences in type of lesions, abundance of lesions in coding and non-coding regions, or in individual genes between cases and controls. We conclude that changes in mtDNA sequence are not significantly different between patients that later develop t-MDS after aHCT (cases) compared to controls that do not develop t-MDS. These observations suggest that differences in mtDNA sequence cannot by itself explain the alterations in altered mitochondrial function seen early during the course of development of t-MDS.