PMC

Array probes were grouped together according to their position in distinct cytobands. Cytobands were then scored in each tumor as gained (+1, red), lost (−1, green), or unchanged (0, white), based on the array CGH data for all probes in a particular cytoband region. Subsequently, these cytoband data for gain, loss or no change were clustered using an unsupervised one-way hierarchical clustering. This was done for each molecular subtype separately.

For each medulloblastoma subtype A, B, C, D, and E, the expression data (vertical axes) are shown in each tumor (indicated with colored circles) for 2 markers that are specifically expressed or upregulated in that subtype, and 2 markers that are not expressed or only at very low levels in that subtype. A. Type A markers; B. Type B markers; C. In 2C markers are shown that are expressed either in subtype A and B together or in subtype C, D, and E together; D. Type C and/or CD markers; E. Type DE markers; F. Type E markers.

The clustering identified 4 distinct clusters indicated as A, B, C, and DE. Clusters D and E could not be separated in this series. Clinical annotations are at the bottom. Thompson subtypes as indicated in their paper: yellow = Thompson subtype A, light pink = Thompson subtype B, blue = Thompson subtype C, green = Thompson subtype D, dark pink = Thompson subtype E; Histology: grey = desmoplastic, orange = large cell/anaplastic, white = classic; sex: pink = female, white = male; M stage: yellow = M1, orange = ≥M2, white = M0; Age: purple = age ≤3 yrs, white = age >3 yrs; β-catenin mutations: brown = mutations, white = wild type; PTCH1/SUFU mutations: light blue = PTCH1 mutations, dark blue = SUFU mutation, white = wild type; 17p deletions: dark red = yes, white = no. A cross means not analyzed.

A. Unsupervised two-way hierarchical cluster analysis of 62 medulloblastoma samples and expression data of 1300 most differentially expressed genes identified 5 distinct clusters indicated as A, B, C, D, and E. Clinical annotations are at the bottom. Histology: grey = desmoplastic, orange = large cell/anaplastic, white = classic; sex: pink = female, white = male; Metastatic disease at diagnosis indicated with M stage: yellow = M1, orange = ≥M2, white = M0; Age: purple = age ≤3 yrs, white = age >3 yrs; β-catenin mutations: brown = mutations, white = wild type; PTCH1 mutations: blue = mutations, white = wild type. A cross means not analyzed. B. Schematic pentagram showing the correlations between the 5 molecular subtypes of medulloblastoma. Numbers at the outside near each subtype indicate number of genes that are significantly differently expressed between that subtype and all other subtypes (P<0.001). Numbers at connecting lines indicate number of genes that are significantly differently expressed between medulloblastoma subtypes.

A. One copy of chromosome X is lost in most female cases of clusters C, D, and E. The plot shows array CGH data for the X and Y chromosome in males and females of the different subtypes. The plotted data represent the average logratio for probes on chromosome X vs chromosome Y after normalization on the median of autosomes (see ). The error-bars represent the standard error of the mean. Open circles represent male patients, closed circles are for female patients. Colors correspond to the colors for the different subtypes as indicated in . B. XIST expression (vertical axis) is shown for 21 female cases (indicated with colored circles) in the MB62 series. 17 of them were analyzed by CGH arrays (). XIST expression is lost or strongly reduced in most type C, D, or E tumors, but not in type A or B tumors.

A. Expression data suggest loss of chromosome 6 in all 9 tumors of cluster A. A one-way unsupervised hierarchical clustering of 62 medulloblastoma samples is shown for all expressed genes on chromosome 6. Colors on top correspond with the colors of the different subtypes as shown in . Genes on the vertical axis are not clustered and are placed next to each other according to their position on the chromosome. B. Example of how DNA alterations affect gene expression. Chromosome 9 array CGH data (green) and expression data (Z-scores, red), both plotted on the Y-axis, are shown for two samples. MB0313 has no DNA alterations for chromosome 9, and MB0270 has lost the telomeric part of 9q. C. Overall analysis shows how genetic aberrations affect gene expression. Cytoband regions for all tumors analyzed by array CGH were marked as ‘−1’ for loss, ‘0’ for no change, and ‘+1’ for gain. Then the average expression (average Z-scores) of all expressed genes in these regions was calculated for each tumor and plotted on the Y-axis for each category. Results show that genes in regions with loss are on average expressed at lower levels (green box) and genes in regions with gain at higher levels (red box), compared to regions with no changes (blue box). Numbers at the X-axis between brackets represent the number of probe sets that were taken into account per category.