|
|
Blood, Vol. 113, Issue 26, 6681-6690, June 25, 2009
Copy number abnormalities, MYC activity, and the genetic fingerprint of normal B cells mechanistically define the microRNA profile of diffuse large B-cell lymphoma
Blood Li et al.
113: 6681
Supplemental materials for: Li et al
Files in this Data Supplement:
- Figure S7. Unsupervised hierarchical clustering analysis of primary DLBCL based of genome-wide miRNA copy number data (JPG, 563 KB)
-
Tumors with a higher percentage of infiltrating T cells clustered together (see right side of figure) and had attenuated copy number changes. See Table S1 for details on T-cell infiltrate quantification.
- Figure S8. Distribution of clinical and pathological features of our DLBCL series stratified by MG clusters (JPG, 64.8 KB)
-
The newly defined microRNA-driven subsets of DLBCL are unrelated to tumor site (upper panel), GCB and non-GCB classification (middle panel), or percentage of infiltrating cells (lower panel, p=.17 MG-A vs. MG-B vs. MG-C, and p=.07 MG-A vs. MG-B/C).
- Figure S9. Unsupervised hierarchical clustering analysis of primary DLBCL based of a miRNA signature obtained from DLBCL cell lines (JPG, 89.7 KB)
-
Nine miRNAs reported to be expressed at high and low levels in ABC-type and GCB-type cell lines, respectively, did not segregate our primary tumors in these categories or any other biologically relevant clusters.
- Figure S10. Overall survival in DLBCL and microRNA expression profiling (JPG, 47 KB)
-
Kaplan-Meier curves for 30 DLBCL patients treated with anthracycline-based regimens and assigned to MG-A, MG-B, and MG-C is shown. On the left panel, all three clusters were considered individually, and the differences were only borderline significant (p-value = 0.065, log rank test). On the right side, a direct comparison between MG-A and a combination of MG-B/C highlighted the association between MG-A and worse outcome in this small series of DLBCLs (p=0.02, log rank test). Considering the relatively small sample size, we also used a permutation test (see Supplemental Materials and Methods for details) to determine the validity of these results and found similar p-values (0.07 and 0.02, left and right panels, respectively).
- Figure S11. Integration of microRNA copy number and expression data in DLBCL (JPG, 156 KB)
-
The correlation between miRNA gene expression values and copy number was computed for seventeen primary DLBCL samples for which both global microRNA expression profiling (Fig. 2A–C) and array-CGH analyses (Fig. 1) was performed. Twelve microRNA genes showed a significant correlation (see Supplemental Materials and Methods for details on the analyses) between these two metrics. Box-plot display comparing the expression distribution of the three copy number bins (loss, diploid, and gain) for each relevant miRNA is shown. Five of these miRNAs belong to the miR17-92 cluster; miR-100, miR-125b-1, and miR-130a, all mapped to chromosome 11 and belong to MG-B, whereas miR-223, miR-320, miR-324, and miR-331 have note been previously associated with lymphoid tumors and deserve further investigation. Notably, loss of miR-15a/16-1 did not associate with lower expression, as investigated in details in Fig. S12.
- Figure S12. Box plot display of miR-16 and miR-15a expression in DLBCL (JPG, 19.2 KB)
-
Stem-loop real-time RT-PCR was used to define the expression of mature miR-16-1, and miR-15a in DLBCL with loss or normal copy number at this locus (n=5 and n=25, respectively) as defined by our arrayCGH analysis. As shown, chromosomal material loss did not significantly influence the expression of these miRNAs in DLBCL (p>.1, Mann-Whitney test).
|
|
