Background
Long non-coding RNAs are RNAs longer than 200 bps that do not encode any proteins and are able to alter gene expression by acting on different steps of regulation, including DNA methylation and chromatin structure. They represent a class of biomarkers of crescent interest in the hematologic and oncologic fields. Recent studies showed that the expression levels of specific lncRNAs correlate with the prognosis of paediatric patients with Acute Lymphoblastic Leukaemia.
Methods
We used NGS approaches to analyse the transcriptome of 9 childhood B-ALL patients and 6 childhood T-ALL patients, in comparison with B and T healthy lymphocytes from cord blood. We validate our findings both ex vivo, in a different cohort of 10 B-ALL and 10 T-ALL patients, and in silico using public datasets.
Results
We characterised the lncRNA landscape for B-ALL, T-ALL, healthy B, and T cell progenitors. From the characterised signature, we selected candidate lncRNAs able to discriminate not only B-ALL and T-ALL from healthy subjects but also between the two types of leukaemia, and subsequently validated their potential as a diagnostic tool in an additional cohort of paediatric patients. We confirmed our finding with open access transcriptomic data, comparing ALL lncRNAs with AML lncRNA landscape as well. Finally, expression correlation analyses of T-ALL selected lncRNA biomarkers suggested a possible role in lymphocyte activation and the β-catenin signalling pathway for AC247036.1 and involvement in hedgehog signalling for HHIP-AS1.
Conclusions
Our work identified a lncRNA signature discriminating paediatric B-ALL and T-ALL from healthy subjects, between them and from AML. This study provides the keystone to future clinical studies determining the theragnostic value of the characterised long non coding transcriptome panorama in a clinical setting for childhood patient management.