In diffuse large B-cell lymphoma (DLBCL), a number of studies have shown that dual disruption of the p16/Rb and the ARF/p53 pathway, e.g. deletion of the INK4A/ARF locus, are negative prognostic factors for survival. Molecular studies have shown that the p53-miR34 axis may be another link between the ARF/p53 pathway and the p16/Rb pathway. In a complex circuit, p53 promotes transcription of MIR34A and MIR34B/C, and the miR34s' in turn act as mediators of p53 signaling. In addition, miR34s' have been shown to inhibit MYC and several proto-oncogens that counteract the p16/Rb tumor suppressor pathway. These observations place the miR34s' at the center of cell cycle and apoptosis regulation, and loss of miR34 expression has been associated with adverse response to therapy in several tumor types. In CLL it has been suggested that miR34A expression is a surrogate marker for TP53 disruption and the MDM2 SNP309 GG genotype, both associated with a poor prognosis.
In CD19+ B-cells isolated from healthy donors (PBL-B) and reactive lymph nodes, the expression of the –5p and –3p species of miR34A, miR34B, and miR34C was analysed by qRT-PCR using LNA-based primers from Exiqon that discriminate the individual miR34 members. Histone H3K4 and H3K27 methylation status at the MIR34A and MIR34B/C promoters was analyzed by ChIP. In 120 newly diagnosed DLBCLs we studied the combined status of TP53 mutations, MIR34A and MIR34B/C promoter DNA methylation, and the MDM2SNP309 GG genotype using a panel of PCR based methods.
We initially determined the expression of the individual miR34 family members in PBL-B and in reactive lymph nodes. While miR34A-5p was expressed at low levels in normal PBL-B, no expression of miR34A-3p, miR34B-5p, miR34B-3p, miR34C-5p, and miR34C-3p was observed. In reactive lymph nodes the expression of miR34A-5p was induced in average 70 fold (P<0.001), while no significant induction was observed for the rest of the investigated miRs, including the homologous miR34B-5p and miR34C-5p. No promoter DNA methylation was observed in PBL-B, but ChIP experiments showed a relative enrichment for the bivalent H3K4me3/H3K27me3 silencing mark at the MIR34B/C promoter.
In 100 of the 120 (83%) DLBCLs one or more of the investigated genes carried (epi)-genomic alterations including TP53 mutations: 20(17%), MIR34A methylation: 34(28%), MIR34B/C methylation: 93(78%) and the MDM2 SNP309 GG-genotype 11(9%). All but one of the MIR34A methylated cases also had methylation of MIR34B/C. Seven cases carried combined TP53 mutation and methylation of both MIR34A and MIR34B/C. A borderline inferior overall survival for patients carrying TP53 mutations was found (p=0.058), while none of the other parameters influenced survival. However, the 7 cases with concomitant mutation of TP53 and methylation of both MIR34A and MIR34 B/C had a median survival of 14 months (p=0.02). In a DLBCL-cell line with concomitant TP53 mutation and methylation of both MIR34A and MIR34B/C, miR34s' were upregulated by 5-aza-2'deoxycytidine (5-aza-CdR).
In normal B-cells miR34A seems to play a dominant role over miR34B/C, however in DLBCL the concomitant DNA methylation of MIR34B/C in the MIR34A methylated cases suggest they may compensate for each other. Our results suggest that in DLBCL the p53-miR34 axis is not linear, and that the TP53, MIR34A and MIR34B/C “triple-hit” DLBCLs may represent more aggressive lymphomas that may benefit from treatment with 5-aza-CdR. This is, to the best of our knowledge, the first study to address these issues in a large sample of diagnostic DLBCL cases.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.