SUMMARY:We have used a new method of genomic microarray to investigate amplification of oncogenes throughout the genome of glioblastoma multiforme (GBM). Array-based comparative genomic hybridization (array CGH) allows for simultaneous examination of 58 oncogenes/amplicons that are commonly amplified in various human cancers. Amplification of multiple oncogenes in human cancers can be rapidly determined in a single experiment. Tumor DNA and normal control DNA were labeled by nick translation with green-and red-tagged nucleotides, respectively. Instead of hybridizing to normal metaphase chromosomes in conventional comparative genomic hybridization (CGH), the probes of the mixed fluorescent labeled DNA were applied to genomic array templates comprised of P1, PAC, and BAC clones of 58 target oncogenes. The baseline for measuring deviations was established by performing a series of independent array CGH using test and reference DNA made from normal individuals. In the present study, we examined fourteen GBMs (seven cell lines and seven tumours) with CGH and array CGH to reveal the particular oncogenes associated with this cancer. High-level amplifications were identified on the oncogenes/ amplicons CDK4, GLI, MYCN, MYC, MDM2, and PDGFRA. The highest frequencies of gains were detected on PIK3CA (64.3%),

This is the first genome-wide survey of multiple oncogene amplifications involved in the development of medulloblastoma. Gains of several candidate oncogenes such as D17S1670, ERBB2, PIK3CA, PGY1, MET, and CSE1L were frequently detected. These genes may be used as molecular markers and therapeutic targets of medulloblastomas.

No abstract

To identify critical tumor suppressor loci that are associated with the development of medulloblastoma, we performed a comprehensive genome-wide allelotype analysis in a series of 12 medulloblastomas. Non-random allelic imbalances were identi®ed on chromosomes 7q (58.3%), 8p (66.7%), 16q (58.3%), 17p (58.3%) and 17q (66.7%). Comparative genomic hybridization analysis con®rmed that allelic imbalances on 8p, 16q and 17p were due to loss of genetic materials. Finer deletion mapping in an expanded series of 23 medulloblastomas localized the common deletion region on 8p to an interval of 18.14 cM on 8p22 ± 23.2. We then searched within the region of loss on 8p for loci that might contain homozygous deletion using comparative duplex PCR. An overlapping homozygous deletion region was identi®ed in a 1.8 cM interval on 8p22 ± 23.1, between markers D8S520 and D8S1130, in two medulloblastomas. This region of homozygous deletion also encompasses the 1.4 cM minimal deletion region detected on 8p in ductal carcinoma in situ of breast. In conclusion, we reported for the ®rst time a detailed deletion mapping on 8p in medulloblastoma and have identi®ed a region of homozygous deletion on 8p22 ± 23.1 that is likely to contain a critical tumor suppressor gene involved in the development of medulloblastoma.