Background: Although genomic mutations play an important role in the initiation or progression of myeloid leukemias, the role of mutations at the transcript level is less well understood. Myeloid neoplasms have recurrent somatic mutations in genes for pre-mRNA splicing, cohesin complex, histone modification, and DNA methylation. However, the number of somatic mutations identified in myeloid neoplasms (average 9-13) is usually lower than that of solid tumors. Recent analyses of baseline transcriptome sequences in various cancers have revealed abundant adenosine to inosine (A>I) RNA editing events, mainly at intronic and untranslated regions of genes. The role of condition-specific RNA editing in cancer, especially of the cytosine to uracil (C>U) type, however, remains essentially unknown. We recently discovered antiviral restriction enzyme APOBEC3A (A3A) cytidine deaminase as a novel cellular site-specific C>U RNA editing enzyme in monocytes. RNA editing by A3A is dynamically induced by interferons and/or severe hypoxia in monocytes, resulting in stop codon gains or missense changes in transcripts of scores of genes. The effect of hypoxia in inducing RNA editing by A3A can be mimicked by inhibition of mitochondrial respiration. Whether A3A-mediated RNA editing occurs in monocytic leukemias, including acute monocytic leukemia (AMoL) and chronic myelomonocytic leukemia (CMML), is unknown. We hypothesize that dynamic C>U RNA editing by A3Acontributes to mutational load of leukemic cells under stressful conditions of hypoxia and interferon exposure. In this study, we examined transcript sequences of monocytic leukemias for evidence of inducible RNA editing. We also examined the role of mitochondrial respiratory inhibition, which induces RNA editing by A3A, in growth of leukemic progenitors of CMML.

Methods: Monocytic leukemia samples are obtained from the RPCI Leukemia Bank. Standard diagnostic information included classification (WHO 2008 and FAB), fraction of monocytic component (mature monocytes, promonocytes or monoblasts by flow cytometry and/or aspirate differential count), cytogenetic analyses (conventional karyotype and targeted FISH), and survival data. Leukemic BM mononuclear cells (BM-MNC) prepared by the Ficoll gradient method are thawed and cultured in hypoxia (1% O2) with interferon type 1 (IFN-1). DNA and RNA are isolated. AMoL samples are subjected to nextgen high throughput exome and RNA sequencing. Once variants from DNA and RNA of corresponding samples are obtained, the set difference of variants is extracted to keep those detected on RNA samples. We also examined 3 CMML bone marrow samples for evidence of hypoxia-inducible RNA editing by analyzing selected RNA editing sites. Colony forming unit (CFU) assays are also performed in CMML samples.

Results: We find that the A3A gene is robustly expressed inboth in AMoL and CMML. Hypoxia (with or without IFN-1) induced A3A-mediated RNA editing in AMoL and CMML samples. We previously showed that several RNA edited genes by A3A in normal monocytes function in recurrently altered pathways in myeloid leukemias. Sanger sequencing confirmed RNA editing of such genes in monocytic leukemia samples as well (Table 1), suggesting that epitranscriptomic sequence changes may contribute to molecular defects driving leukomogenesis. RNA seq analysis of AMoL samples (n=3) showed that dozens of genes acquire nonsense or missense mutations at the transcript level under the conditions of severe hypoxia and IFN-1. CFU assays of CMML samples showed that inhibition of mitochondrial respiration, which triggers A3A-mediated RNA editing, inhibits the growth of progenitor cells (Figure 1).

Conclusions: We show that A3A-mediated RNA editing is active in monocytic leukemias and targets genes functioning in pathways recurrently altered in myeloid leukemias. RNA editing by A3A may contribute to growth arrest and survival of leukemic progenitors under certain stress conditions. To our knowledge, these results provide the first experimental evidence that conditional RNA editing by A3A confers substantial mutational load in myeloid leukemias. Such RNA editing events will dynamically increase the total mutational load in monocytic leukemias, and may create neo-antigens for immune therapy by checkpoint inhibition.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution