Unique Long Non-Coding RNA Expression Signature in ETV6/RUNX1-Driven B-Cell Precursor Acute Lymphoblastic Leukemia

Emerging evidence suggests that long non-coding RNAs (lncRNAs) are critically involved in a variety of human tumor entities and the identification of cancer-associated lncRNAs might reveal new prognostic biomarkers or novel therapeutic targets for the treatment of human cancer. In this study, we characterized the lncRNA expression signature associated with ETV6/RUNX1-positive pediatric B-cell precursor acute lymphoblastic leukemia (BCP-ALL), one of the most prevalent genetic subtypes of childhood leukemia.

First, we used primary leukemia patient samples to identify an ETV6/RUNX1 specific expression signature consisting of 596 lncRNA transcripts. Besides using primary BCP-ALL patient samples, we also integrated these lncRNA expression data with RNA sequencing results from a panel of human BCP-ALL leukemia cell lines and identified a unique lncRNA expression profile of 16 lncRNAs exclusively associated with the presence of the ETV6/RUNX1 fusion protein. Notably, lnc-SARDH-1 (also known as DBH-AS1) was the only lncRNA from this list for which an oncogenic role has previously been postulated.

Given that previous studies revealed putative overlap between lncRNA expression and the presence of regulatory enhancer elements, we also evaluated the distribution of H3K27ac, a histone modification associated with enhancer activity, at the genomic loci of the 16 ETV6/RUNX1 specific lncRNAs mentioned above. Notably, broad H3K27ac binding was identified for lnc-NKX2-3-1, lnc-RTN4R-1, lnc-GIP-1, lnc-LRP8-3, lnc-TCF12-2, lnc- C8orf4-1, lnc-C8orf4-2, lnc-TINAGL1-1 and lnc-LSM11-4 in ETV6/RUNX1-positive REH cells, whereas more discrete H3K27ac peaks were identified at putative promoter regions for lnc-TIMM21-5 and lnc-CPT2-7. Interestingly, coding genes adjacent to some of these lncRNAs showed unique overexpression in ETV6/RUNX1-positive BCP-ALLs, suggesting a possible cis regulatory relationship between these lncRNAs and their nearby protein coding genes. Finally, we applied shRNA-mediated silencing of endogenous ETV6/RUNX1, integrated these expression profiles with the patient and cell line data, to show that lnc-NKX2-3-1, lnc-TIMM21-5, lnc-ASTN1-1 and lnc-RTN4R-1 were the only lncRNAs that were truly regulated by the oncogenic fusion protein.

Interestingly, subsequent lncRNA modulation experiments using LNA GapmeR technology revealed that lnc-TIMM21-5 and lnc-ASTN1-1 modulation has no effect on overall transcription, suggesting that these lncRNAs might act at the translational level or/and at various steps of mRNA processing. In contrast, lnc-NKX2-3-1 and lnc-RTN4R-1 perturbations resulted in severe changes in gene expression, suggesting an alternative mechanism of action for these lncRNAs that is more transcriptionally oriented. Most notably, loss of lnc-RTN4R-1expression significantly affected part of the ETV6/RUNX1-specific mRNA expression signature as exemplified by reduced levels of AK7, PTPRKand GBA3.

Altogether, our study identified a panel of ETV6/RUNX1 specific lncRNAs that might be implicated in the biology of human BCP-ALL and could serve as novel biomarkers or novel therapeutic targets for the treatment of this prevalent subtype of human leukemia.

TLM and PVV are shared last authors