Figure S1, Related to Figure 1. Expression Levels of 13Q14 Mirs Is Not Different in Patients

Figure S1, related to Figure 1. Expression levels of 13q14 miRs is not different in patients with deletion of the critical region in 13q14.

(A) MiR-16 levels in purified CLL cells. Displayed data are taken from (Fulci et al.). Patients with 40% or more deleted cells were considered 13q+/-. Two patients with 11% and 19% deleted cells were considered 13q+/+. Patients with no FISH data were not considered for the analysis.

(B) MiR-16 levels in purified CLL cells. Displayed data are taken from (Ouillette et al.). Chromosomal copy number was quantified for a ~0.4 Mb chromosomal region between rs9316484 (centromeric to the miR16/15a locus) and rs3118650 (telomeric to the miR16/15a locus). Patients were grouped as follows: mean copy number>1.50 = 13q+/+, 0.75<mean copy number<1.50 = 13q+/-.

(C) MiR-16 levels were measured by qRT-PCR in PBMCs from our cohort of CLL patients.

(D) MiR-15b levels were measured by qRT-PCR in normal B-cells and PBMCs from CLL patients. *p<0.05.

(E) Inverse correlation between miR-15a and pri-miR-15a levels in CLL patients. Levels of miR-15a and pri-miR-15a were measured by qRT-PCR, normalized to 2 house-keeping genes and to the average of healthy controls.

(F) MiR-155 levels were measured by qRT-PCR in normal B-cells and PBMCs from CLL patients. ***p<0.001.

(G) Pri-miR-155 levels were measured by qRT-PCR in normal B-cells and PBMCs from CLL patients. ***p<0.001.

Figure S2. Schematic of amplicons specific for mature and pre-miRNA transcripts.

(A) Available assays for the quantification of mature miRNAs by qRT-PCR after poly-A tailing are not sufficiently specific and also detect pre-miRNAs e.g. upon knockdown of DICER activity. (B) Specificity can be increased by anchoring the miRNA forward primer in the oligo-A tail. (C) Primers binding to the loop-sequence of pre-miRNAs cannot produce an amplicon from pri-miRNAs and are therefore specific.

Figure S3. Intermediate transcripts of miRNA processing are present at different levels in non-malignant B-cells from healthy donors.

Processing intermdiates from the miR-15/-16 family derived from chromosomal bands 13q14 and 3q25 were measured with qRT-PCR. Ct values were normalized to RNU6B and SNORA73A used as internal housekeeping controls and displayed on an inverted y-axis to reflect levels of transcripts. Pre-miRNAs were the rarest transcript type and mature miRs the most abundant. 13q14 miRNA processing transcripts had higher levels than the homologs localized in chromosomal band 3q25. Whiskers display 95% confidence intervals and points delineate maximum and minimum outliers.

Figure S4. Patients with processing defect do not show different overall survival, IGHV mutational status or different cytogenetic aberrations. (A) Overall survival is depicted in two patient groups segregated by median expression levels of the processing intermediates of the miR-15/-16 family. No significant differences were observed (logrank test, see also supplemental table S2). (B) Patients were stratified according to cytogenetic aberrations as defined in the hierarchical model (Dohner et al. 2000), i.e. patients with del13q and del 17p are stratified as del17p. Shown is the log fold expression compared to the respective housekeeping gene in boxed scatter plots with median (bold line), lower quartile and upper quartile and whiskers displaying the 95% distribution. No significant differences could be observed except for pre-miR15a and b that showed lower levels in del(13q) cases and normal karyotype patients (miR15b) and for pre-miR-16-2 that was higher in del(17p) cases. (C) Patients were stratified according to the IGHV mutational status as described previously (Dohner, Stilgenbauer et al. 2000). Expression levels were not significantly different with the exception of levels of pre-miR15a. For p-values see supplemental Table S1.

Figure S5, related to Figure 3. Quantification of mature miR-16, -15a and -15b correlates significantly between qRT-PCR and microarray.

Levels of mature miR-16 (upper panel), -15a and -15b (lower panel) were quantified with qRT-PCR and microarray profiling. Correlations were highly significant (miR-16 and miR-15b p<0.0001, mir-15a p<0.0002).

Figure S6, related to Figure 3. Levels of mature miR-15a assessed by miRNA-profiling in processing defective CLL patient cells compared to processing competent cells, both with retention of both copies of 13q14.3

Levels of mature miR15a are lower in CLL patients with a defective processing (n=7) compared to patients with normal miR processing (n=5) only when CLL patients with a deletion of 13q that spans miR15a (n=8) are excluded. Difference is not statistically significant (p=0.054, student´s t-test). Displayed are normalized signal intensities from miRNA microarrays for miR-15a.

Figure S7, related to Figure 3. MiRNAs downregulated in CLL cells with a processing defect target components of the T-cell receptor signaling pathway. Using DIANA miRPath (Vlachos et al. 2012) and miRSystem (Lu et al. 2012), target genes of the miRNAs underrepresented in CLL cells with a processing defect were associated with KEGG and INTERACTION functional pathways (See Suppl. Table S3). Intriguingly, signaling components downstream of the T-cell receptor and the CLL B-cell receptor were significantly enriched among these targets, suggesting a functional correlation of the miRNAs with a putative processing defect. Plasma membrane shown as grey lines, proteins / genes in boxes, targets of selected miRs are shown in yellow.

Figure S8. ß-catenin stimulation by LiCl reduces processing of pri-miR-16-1 but not pri-miR-16-2. MiR processing intermediates were measured as described in Figure 6 after stimulation with LiCl. Shown are the levels of miRNA intermediates compared to untreated samples (0; horizontal line), where miR-16-1 shows an enrichment of pr1-miR compared to pre-miR and mature miR-16. However, this was not the case for miR-16-2, suggesting that Wnt-signaling does not fully account for the processing defect described here.

Vlachos, I. S., N. Kostoulas, T. Vergoulis, G. Georgakilas, M. Reczko, M. Maragkakis, M. D. Paraskevopoulou, K. Prionidis, T. Dalamagas and A. G. Hatzigeorgiou (2012). "DIANA miRPath v.2.0: investigating the combinatorial effect of microRNAs in pathways." Nucleic Acids Research40(W1): W498-W504.

Lu, T.-P., C.-Y. Lee, M.-H. Tsai, Y.-C. Chiu, C. K. Hsiao, L.-C. Lai and E. Y. Chuang (2012). "miRSystem: An Integrated System for Characterizing Enriched Functions and Pathways of MicroRNA Targets." PLoS ONE7(8): e42390.

Ouillette, P., H. Erba, L. Kujawski, M. Kaminski, K. Shedden and S. N. Malek (2008). "Integrated genomic profiling of chronic lymphocytic leukemia identifies subtypes of deletion 13q14." Cancer Res68(4): 1012-1021.

Fulci, V., S. Chiaretti, M. Goldoni, G. Azzalin, N. Carucci, S. Tavolaro, L. Castellano, A. Magrelli, F. Citarella, M. Messina, R. Maggio, N. Peragine, S. Santangelo, F. R. Mauro, P. Landgraf, T. Tuschl, D. B. Weir, M. Chien, J. J. Russo, J. Ju, R. Sheridan, C. Sander, M. Zavolan, A. Guarini, R. Foa and G. Macino (2007). "Quantitative technologies establish a novel microRNA profile of chronic lymphocytic leukemia." Blood109(11): 4944-4951.

Dohner, H., S. Stilgenbauer, A. Benner, E. Leupolt, A. Krober, L. Bullinger, K. Dohner, M. Bentz and P. Lichter (2000). "Genomic aberrations and survival in chronic lymphocytic leukemia." N Engl J Med343(26): 1910-1916.