Integrative microRNA and mRNA deep-sequencing expression profiling in endemic Burkitt lymphoma

Abstract

Background: Burkitt lymphoma (BL) is characterized by overexpression of the c-myc oncogene, which in the vast majority of cases is a consequence of an IGH/MYC translocation. While myc is the seminal event, BL is a complex amalgam of genetic and epigenetic changes causing dysregulation of both coding and non-coding transcripts. Emerging evidence suggest that abnormal modulation of mRNA transcription via miRNAs might be a significant factor in lymphomagenesis. However, the alterations in these miRNAs and their correlations to their putative mRNA targets have not been extensively studied relative to normal germinal center (GC) B cells.

Methods: Using more sensitive and specific transcriptome deep sequencing, we compared previously published small miRNA and long mRNA of a set of GC B cells and eBL tumors. MiRWalk2.0 was used to identify the validated target genes for the deregulated miRNAs, which would be important for understanding the regulatory networks associated with eBL development.

Results: We found 211 differentially expressed (DE) genes (79 upregulated and 132 downregulated) and 49 DE miRNAs (22 up-regulated and 27 down-regulated). Gene Set enrichment analysis identified the enrichment of a set of MYC regulated genes. Network propagation-based method and correlated miRNA-mRNA expression analysis identified dysregulated miRNAs, including miR-17~95 cluster members and their target genes, which have diverse oncogenic properties to be critical to eBL lymphomagenesis. Central to all these findings, we observed the downregulation of ATM and NLK genes, which represent important regulators in response to DNA damage in eBL tumor cells. These tumor suppressors were targeted by multiple upregulated miRNAs (miR-19b-3p, miR-26a-5p, miR-30b-5p, miR-92a-5p and miR-27b-3p) which could account for their aberrant expression in eBL.

Conclusion: Combined loss of p53 induction and function due to miRNA-mediated regulation of ATM and NLK, together with the upregulation of TFAP4, may be a central role for human miRNAs in eBL oncogenesis. This facilitates survival of eBL tumor cells with the IGH/MYC chromosomal translocation and promotes MYC-induced cell cycle progression, initiating eBL lymphomagenesis. This characterization of miRNA-mRNA interactions in eBL relative to GC B cells provides new insights on miRNA-mediated transcript regulation in eBL, which are potentially useful for new improved therapeutic strategies.

Keywords: Endemic Burkitt lymphoma; Lymphomagenesis; RNA sequencing; mRNA; miRNA.

Fig. 1

Sample to sample hierarchal clustering of eBL tumor cells and GC B cells based on a mRNA expression profiles, b miRNA expression profiles with highest correlation of variation (CV) values (calculated using regularized log transformed mRNA and miRNA expression values)

Fig. 2

Differentially expressed mRNAs in eBL compared to GC B cells. a Heatmap of differentially expressed (DE) miRNAs between eBL tumor cells and GC (Germinal center) B cells. The heatmap shows the hierarchical clustering based on the expression profiles of the 211 DE genes with at least 2-fold difference in expression compared to their normal counterpart. b Volcano plot representing the significance genes (−log of the adjusted p-value) vs the fold change difference in eBL compared to GC B-cells. The red and blue colored circles represent genes which are DE with p < 0.01 and FDR < 0.01. The 132 down-regulated genes in eBL are colored blue (have a negative fold-change value) while the 79 up-regulated genes in eBL are colored red (positive fold-change value)

Fig. 3

Gene set enrichment plot and expression heatmap of corresponding genes in the enriched gene set. Left panels include the running enrichment score throughout the gene set and projection of genes in the geneset to the complete list of genes rank ordered based on signal to noise ratio. On the expression heatmap (columns are eBL tumors and GC B cells, rows are genes in the gene set), dark red represent higher expression while dark blue lower expression. Genes in this enrichment are a set of genes regulated by MYC in eBLs tumor cells relative to GC B cells (ES = 0.45, Nominal P = 0.046, FDR q = 0.118)

Fig. 4

Differentially expressed miRNAs in eBL compared to GC B cells. a. Heatmap of differentially expressed (DE) miRNAs between eBL tumor cells and GC (Germinal center) B cells. The heatmap shows the hierarchical clustering based on the expression profiles of the 49 DE miRNAs with at least 2-fold difference in expression compared to their normal counterpart. b. Volcano plot representing the significance miRNAs (−log of the adjusted p-value) vs the fold change difference in eBL compared to GC B-cells. The red and blue colored circles represent miRNAs which are DE with p < 0.01 and FDR < 0.01. The 27 down-regulated miRNAs in eBL are colored blue (have a negative fold-change value) while the 22 up-regulated miRNAs in eBL are colored red (positive fold-change value)

Fig. 5

Aberrant transcriptome expression pivotal to eBL lymphomagenesis. a Schematic illustration of the aberrant gene expression and miRNA mediated regulatory changes that would initiate lymphomagenesis as a result of DNA damage. Combined loss of p53 function due to small interfering RNA-mediated regulation of ATM and NLK together with upregulation of TFAP4, would facilitate survival of cells with the c-myc-Igh chromosomal translocation and MYC induced cell cycle progression initiating eBL tumor development. ATM checkpoint kinase, transduces genomic stress signals to halt cell cycle progression in response to DNA damage. It is critical in the regulation of apoptosis and lymphomagenesis in c-myc induced lymphomas. ATM is downregulated in eBL and it is targeted by 4 miRs that are Upregulated in eBL. NLK is required for the upregulation of P53 expression in response to DNA damage. It interacts with P53 to enhance its stability and activity by abrogating MDM2 mediated degradation. NLK is downregulated in eBL tumor cells and also targeted by 2 miRs that are upregulated in eBL tumor cells. TFAP4/AP4 is a central mediator of cell cycle progression in response to c-MYC activation. b RNA seq. Expression counts of MYC, TFAP4, ATM and NLK in eBL tumor cells and GC B cells. c Hierarchical clustering of eBL and GC B cells based on the expression profiles of MYC, TFAP4, ATM and NLK also revealed a clear separation of the two groups. d. miRNA seq. Expression counts of hsa-miR-26a-5p, hsa-miR-27b-3p, hsa-miR-30b-5p, miR-17~92-cluster members (hsa-miR-19b-3p, and hsa-miR-92a-3p), and let-7-family miRs (hsa-let-7a-5p, hsa-let-7b-5p, hsa-let-7d-5p, hsa-let-7e-5p, and hsa-let-7 g-5p) in eBL tumor cells and GC B cells

Fig. 6

a The significantly enriched signaling pathways of the validated target genes of the DE miRNAs that showed an inverse expression change. b The significantly enriched gene ontologies (GO’s) of the validated target genes of the DE miRNAs that showed an inverse expression change