|
|
Blood, Vol. 113, Issue 19, 4586-4594, May 7, 2009
Patterns of microRNA expression characterize stages of human B-cell differentiation
Blood Zhang et al.
113: 4586
Supplemental materials for: Zhang et al
Files in this Data Supplement:
- Document 1. Luciferase indicator assays (PDF, 80.1 KB)
- Table S1. MicroRNAs detected in at least one B-cell subset (PDF, 145 KB) -
Normalized expression values for B-cell subsets are shown, along with significance analysis of microarrays q values. Empty microRNA expression cells indicate below-threshold values, defined as RT-QPCR Ct values greater than 36 or undetected. SAM q values greater than 5 were considered non-significant and not displayed.
- Table S2. miRNAs measured using the multiplex RT-PCR assay, but not consistently detected in any B-cell population (PDF, 43.7 KB) -
miRNAs marked with * were not used in analyses because they have been reclassified as non-miRNAs.
- Table S3. Average expression of the genes depicted in Fig. S2 (PDF, 992 KB) -
Empty cells indicate values not appreciably measured.
- Table S4. Differentially expressed transcription factors in mature B cells that are predicted targets of differentially expressed miRNAs (PDF, 188 KB) -
Normalized expression values are shown. q values are only shown for genes differentially expressed between B-cell subsets.
- Table S5. Predictor microRNAs that distinguish germinal center B-cell (GCB) DLBCL from other lymphoma types in each pair-wise comparison (PDF, 26.1 KB) -
BL=Burkitt lymphoma, ABC=activated B-cell DLBCL, CLL=chronic lymphocytic lymphoma.
- Table S6. Predictor microRNAs that distinguish activated B-cell (ABC) DLBCL from other lymphoma types in each pair-wise comparison (PDF, 24.6 KB) -
BL=Burkitt lymphoma, CLL=chronic lymphocytic lymphoma.
- Table S7. Predictor microRNAs that distinguish Burkitt lymphoma from other lymphoma types in each pair-wise comparison (PDF, 22.5 KB) -
CLL=chronic lymphocytic lymphoma.
- Figure S1. Distribution of miRNAs present in B-cell subsets (JPG, 42.1 KB)
-
- Figure S2. microRNA targets are expressed at lower levels (JPG, 41.5 KB)
-
(A) Density plots of the expression frequency of predicted mRNA targets of miRNAs expressed highly in the germinal center B cells compared to naive cells. mRNAs from Fig. 1E that were also predicted targets of the miRNAs (N=830) were plotted for both naive cells and germinal center B cells. The depicted p-value was calculated using a 1-sided Kolmogorov-Smirnov test. (B) Density plots of the expression frequency of predicted targets of miRNAs expressed highly in the germinal center B cells compared to plasma cells. mRNAs from Fig. 1G that were also predicted targets of the miRNAs (N=1098) were plotted for both plasma cells and germinal center B cells. The depicted p-value was calculated using a 1-sided Kolmogorov-Smirnov test. (C) Density plots of the expression frequency of predicted targets of miRNAs expressed highly in the germinal center B cells compared to memory B cells. mRNAs from Fig. 1I that were also predicted targets of the miRNAs (N=269) were plotted for both naive cells and germinal center B cells. The depicted p-value was calculated using a 1-sided Kolmogorov-Smirnov test. Analysis: In the Naïve to Germinal Center transition (A), genes with at least a two-fold change in expression were analyzed by SAM using a false discovery rate of less than 1% (Larsson et. al, BMC Bioinformatics, May 2005), depicted in figure 1e. From that list of significantly differentially expressed genes, we determined those that were predicted targets of miRNAs (defined as genes with 3′UTR sequence complementarity to microRNA nucleotides 2-7) expressed more highly in Germinal Center cells. The gene expression distribution of these targets was graphed for Naïve B Cells (blue curve) and Germinal Center B cells (orange curve). Thus, in Fig. S2A, we show genes differentially expressed between Naïve and Germinal Center B cells that are also predicted targets of miRNA expressed more highly in Germinal Center B cells. We observe that the expression of these microRNA target genes is lower in germinal center B cells compared to naïve B cells. The corresponding depictions are shown for the Germinal Center to Plasma Cell (Fig. S2B) and Germinal Center to Memory Cell (Fig. S2C) transitions. In all three cases, the distribution of germinal center miRNA target gene expression is statistically significantly lower (leftward shift of orange curve), which suggests a possible gene regulatory role for the GC high miRNA. The observed lower expression of the GC high miRNA targets could also be caused by other factors such as downstream effects of particular transcription factors. The complete set of genes plotted above is listed in Table S3.
- Figure S3. Differentially expressed transcription factors are frequently microRNA targets (JPG, 43.0 KB)
-
(A) The left panel indicates the proportion of transcription factors that are differentially expressed in the naive to germinal center B-cell transition that are also predicted targets of differentially expressed miRNAs. The right panel indicates the proportion of transcription factors that are not differentially expressed and also are predicted targets of differentially expressed miRNA in that stage-transition. The p-value indicates the results of a chi-squared test for the enrichment of predicted miRNA targets among the differentially expressed transcription factors. (B and C) demonstrate a similar analysis for the GC to plasma cell and GC to memory cell transition.
- Figure S4. Mybl1 transcript levels decrease with miR-223 overexpression (JPG, 44.8 KB)
-
(A) Base-pairing of the 3′UTR of the MYBL1 gene with nucleotides 2-8 of miR-223. This 7-mer is highly conserved across a number of species and serves as a potential binding site for miR-223. (B) Effects of over-expression of miR-223 in germinal center lymphoma-derived BJAB cells in 3 separate experiments. The blue bars depict expression of MYBL1 mRNA 24 hours after transfection with a scrambled control that does not possess complementarity to the human genome. The orange bars depict the expression of Mybl1 24 hours after transfection with a precursor for miR-223. The expression of LMO2 was consistently lower in the cells treated with the miR-223 precursor (P<0.05 in all cases).
- Figure S5. Specificity of real-time PCR probes for members of miR-30 family (JPG, 31.4 KB)
-
Three separate transfection experiments using high concentrations of miR-30b precursors are shown in the left panel with measurement of miR-30b and miR-30d. Similar experiments were performed with precursors for miR-30d (right panel).
- Figure S6. Results of leave one out cross validation applied to the predictors for Burkitt lymphoma, chronic lymphocytic leukemia, activated B-cell diffuse large B-cell lymphoma, and germinal center B-cell DLBCL (JPG, 27.6 KB)
-
For a sample prediction to be correct, it had to be classified correctly in each pair-wise comparison with all remaining entities.
|
|
