Advertisement

MicroRNA-155 controls vincristine sensitivity and predicts superior clinical outcome in diffuse large B-cell lymphoma

Hanne Due, Anna Amanda Schönherz, Laura Ryø, Maria Nascimento Primo, Ditte Starberg Jespersen, Emil Aagaard Thomsen, Anne Stidholt Roug, Min Xiao, Xiaohong Tan, Yuyang Pang, Ken H. Young, Martin Bøgsted, Jacob Giehm Mikkelsen and Karen Dybkær

Data supplements

Article Figures & Data

Figures

  • Figure 1.

    Changes in miR-155 expression in GCB-DLBCL cell lines affect vincristine response. (A) Schematic overview of lentiviral vector plasmids used as control (i), for expression of miR-155 (ii) and TuD-155 (iii) inhibitor in fusion with eGFP. (B-E) Up- and downregulation of miR-155 was detected by RT-qPCR in (Bi,Ci) SU-DHL-5 and (Di,Ei) OCI-Ly7 cells and followed by vincristine dose-response analysis. (Bii,Cii,Dii,Eii) Drug response is illustrated as percentage of living cells related to the no-drug treated condition. CMV, cytomegalovirus promoter; LTR, long terminal repeat; NQ, normalized quantity; PGK, phosphoglycerate kinase promoter; WPRE, woodchuck hepatitis virus posttranscriptional regulatory element.

  • Figure 2.

    miR-155 knockout experiments. Two independent miR-155 knockout clones were generated in OCI-Ly7 cells by the CRISPR-Cas9 technology. Wild-type accounts for untransduced OCI-Ly7 cells, whereas control is an OCI-Ly7 population subjected to a sgRNA not targeting the human genome. (A) Cell proliferation was determined by the trypan blue exclusion method after 48 hours of growth in wild-type, control cells, and miR-155 knockout clones. (B) miR-155 expression detected by RT-qPCR. (C) Vincristine dose-response analysis using 0.0005 µg/mL, 0.001 µg/mL, and 0.0015 µg/mL. Drug response is illustrated as percentage of living cells related to the no-drug condition. (D) Detection of Wee1 protein by western blotting. β-actin was used as loading control and Ship-1, a validated target of miR-155,53 as positive control. (E) Quantification of western blot bands. Ship-1 and Wee1 levels are depicted as relative to β-actin levels. (F) Chemical inhibition of Wee1 by MK-1775. Wild-type OCI-Ly7 cells were exposed to saline, 400 nM MK-1775, 0.0015 µg/mL vincristine or 400 nM MK-1175 and 0.0015 µg/mL vincristine for 48 hours, and the number of metabolically cells were determined by 3-(4,5 dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl-2-(4-sulfophenyl)-2H-tetrazolium assay. Drug response is presented as the number of cells relative to the no-drug condition. MK, MK-1775; Vin, vincristine.

  • Figure 3.

    Analysis of prognostic effect of miR-155 expression. Kaplan-Meier plots depicting OS of R-CHOP-treated patients with DLBCL in the in-house cohort. The analysis was conducted for all patients with DLBCL (A), ABC-classified patients (B), and GCB-classified patients (C). For each cohort, patients were dichotomized by median split of miR-155 expression.

  • Figure 4.

    Analysis of association between OS and PFS and miR-155 expression. Kaplan-Meier plot depicting OS (A-C) and PFS (D-F) of R-CHOP-treated patients with DLBCL in the meta-cohort, consisting of R-CHOP–restricted Lymphoma/Leukemia Molecular Profiling Project and International DLBCL Rituximab-CHOP Consortium MD Anderson Project data. Before analysis, data were ComBat normalized to compensate for study-wise batch effect. Analyses were conducted for all patients with DLBCL (A,D), ABC-classified patients (B,E), and GCB-classified patients (C,F). For each cohort, patients were dichotomized by median split of MIR155HG expression.

Tables

  • Table 1.

    MIR155HG expression is an IPI-independent prognostic marker for R-CHOP-treated patients with GCB-DLBCL

    nno.SimpleMultiple
    HR95% CIPHR95% CIP
    All DLBCL
     miR-155
      Low293901
      High2931071.170.88-1.54.28
     IPI
      0-171611
      2-3314843.761.64-8.61.00173.421.49-7.84.0038
      4-52011069.614.22-21.887.19e-088.293.62-18.985.73e-07
     Subclass
      ABC24210511
      GCB248580.450.33-0.621.09e-060.550.40-0.76.00032
      UC96340.730.50-1.08.110.710.48-1.05.089
    ABC-DLBCL
     miR-155
      Low121571
      High121480.780.53-1.14.20
     IPI
      0-11521
      2-3125383.570.86-14.89.080
      4-51026510.372.51-42.87.0012
    GCB-DLBCL
     miR-155
      Low1244111
      High124170.400.23-0.71.00170.460.26-0.81.0071
     IPI
      0-148211
      2-3140326.221.49-25.97.0125.911.42-24.69.015
      4-5602412.532.96-53.06.000611.102.61-47.15.0011
    • Array-based miR-155 expression and outcome were analyzed by simple and multiple Cox proportional hazards regression analyses for OS. IPI (international prognostic index) score information was not available for all patients, thus cohort sizes are reduced in this setting (115 samples were removed). MIR155HG expression was dichotomized by median split in each of the individual cohorts: all DLBCL patients, ABC-classified patients, and GCB-classified patients.

    • —, value not available because variables were only included in multiple Cox proportional hazards regression analysis if significant results were obtained in simple Cox proportional hazards regression analysis; CI, 95% lower and upper confidence intervals; HR, hazard ratio; n, number of samples; no., number of events.