|
|
Blood, Vol. 106, Issue 1, 274-286, July 1, 2005
HOXA genes are included in genetic and biologic networks defining human acute T-cell leukemia (T-ALL)
Blood Soulier et al.
106: 274
Supplemental materials for: Soulier et al, Vol 106, Issue 1, 274-286
Microarray data and statistical analysis for the probe-sets specifically used throughout this work can be found on the EBI database at http://www.ebi.ac.uk/arrayexpress.
Files in this Data Supplement:
- Figure S1. RQ-PCR analysis of oncogene expression in T-ALL sample groups (PDF, 42 KB)
- Figure S2. Stability of unsupervised classifications of T-ALL samples (PDF, 2.25 MB): -
(A) Unsupervised classifications obtained using various filtering parameters for gene selection. (B) Unsupervised classifications of 50 randomly selected T-ALL samples (n=100 experiments). (C) Stability of the sample clusters (BRB tools). (D) Class prediction (BRB tools; see also “Materials and methods”).
- Figure S3. Genomic characterization of TCRB-HOXA rearrangements (PDF, 181 KB) -
(A) Partial restriction map of a 20-kb region encompassing the BCR-HOXA region. The four breakpoints are indicated. Probes used for Southern blot are indicated (see page 9 of the Figure-S2 PDF for materials and methods). B indicates BamHI; H, HindIII; R, EcoRI; X, XhoI; A, AflIII. The restriction map was established according to the genome sequence available in the UCSC; only relevant sites are indicated. (B) Southern blot analysis of the HOXA rearrangements in cases TL43, TL44 and TL45. Southern blot was performed on genomic tumoral DNA and control DNA as indicated. Rearranged bands are shown by arrows. (C) Breakpoint and flanking sequences of both derivative chromosomes from the TCRB-HOXA rearrangement TL46. Recognition sequence signal (RSS) heptamer and putative heptamer-like sequence are underlined according to consensus (Marculescu R, Le T, Simon P, Jaeger U, Nadel B. V(D)J-mediated Translocations in Lymphoid Neoplasms: A Functional Assessment of Genomic Instability by Cryptic Sites. J Exp Med. 2002;195:85-98). Untemplated nucleotides are in lowercase
- Figure S4. Unsupervised classification of the T-ALL samples from the present study (PDF, 513 KB) -
Classification is based on 47 genes derived from Ferrando et al.19 Using a supervised approach based on selected oncogene expression in defining patient groups, and a microarray tool exploring a more limited set of genes (n = 6800), Ferrando et al19 identified a set of 62 genes differentially expressed in 27 T-ALL samples expressing TAL1 (14 cases), TLX1 (8 cases) or LYL1 (5 cases). The Ferrando et al19 study and the present study can be compared with reasonable confidence on TAL_R and TLX1-expressing cases (n=40 and n=11 cases in our study, respectively). Of the 62 genes from the study by Ferrando et al,19 54 were also analyzed in the present study. We observed the previously demonstrated trend (differential gene expression) in samples of the present study in 53 of these genes, with a statistical significance for 47 genes. An unsupervised classification of our samples was performed using the average expression of these 47 genes. As expected, TAL_R and TLX1-expressing cases were properly clustered using these genes. Of importance, three cases here identified as HOXA-translocated co-cluster using this former set of genes, within a branch containing the vast majority of TLX1 and TLX3 cases.
- Supplementary Materials and Methods (PDF, 369 KB)
- Table S1. List of genes used in Fig. 2B (PDF, 21.2 KB) -
(a) List of 255 probe-sets used to construct the unsupervised classification in Figure 2; (b) probe-sets expressed in bone marrow cells and excluded from Figure 2; (c) probe-sets expressed in erythroid cells and used to exclude contaminated samples from TAL1 annotations in Figure 2.
- Table S2. Compendium of the 469 probe-sets used for the supervised classification in Figure 4, and annotations (PDF, 29.7 KB)
- Table S3. Probe-sets selected to construct the unsupervised classifications in Figure 5 (variable expression in thymic sub-populations) n=168 (PDF, 16.6 KB)
|
|
