Research Article High-Resolution Genome-Wide Mapping of Genetic Alterations in Human Glial Brain Tumors Markus Bredel,1,5 Claudia Bredel,1,5 Dejan Juric,1 Griffith R. Harsh,2 Hannes Vogel,3 Lawrence D. Recht,4 and Branimir I. Sikic1 1Division of Oncology, Center for Clinical Sciences Research; Departments of 2Neurosurgery, 3Pathology, and 4Neurology, Stanford University School of Medicine, Stanford, California and 5Department of General Neurosurgery, Neurocenter, University of Freiburg, Freiburg, Germany Abstract profiles in a cohort of 54 gliomas of various histogenesis and tumor High-resolution genome-wide mapping of exact boundaries of grade. The generated high-resolution genome-wide maps allowed chromosomal alterations should facilitate the localization and delineating the precise (gene specific) boundaries of known and identification of genes involved in gliomagenesis and may new chromosomal alterations, which is not feasible by classic characterize genetic subgroups of glial brain tumors. We have chromosomal CGH. We show that gliomas can be clustered into done such mapping using cDNA microarray-based comparative distinct subgroups based on their genetic profiles, which include genomic hybridization technology to profile copy number recurrent patterns of interrelated chromosomal changes. The alterations across 42,000 mapped human cDNA clones, in a alteration of a subset of genes can predict astrocytic and series of 54 gliomas of varying histogenesis and tumor grade. oligodendroglial tumor phenotypes. Finally, we have identified in a subset of gliomas five common deleted regions that involve This gene-by-gene approach permitted the precise sizing of critical amplicons and deletions and the detection of multiple potential candidate tumor suppressor genes. new genetic aberrations. It has also revealed recurrent patterns of occurrence of distinct chromosomal aberrations Materials and Methods as well as their interrelationships and showed that gliomas Tumor specimens. Fifty-four fresh-frozen glioma specimens were can be clustered into distinct genetic subgroups. A subset of collected under Institutional Review Board–approved guidelines and detected alterations was shown predominantly associated subjected to standard WHO classification (5). Specimens included with either astrocytic or oligodendrocytic tumor phenotype. astrocytic [3 juvenile pilocytic astrocytomas, 1 low-grade astrocytic glioma, Finally, five novel minimally deleted regions were identified 3 anaplastic astrocytomas, 31 glioblastomas (of these 3 secondary in a subset of tumors, containing putative candidate tumor glioblastomas and 2 gliosarcomas)], oligodendroglial (5 oligodendrogliomas, suppressor genes (TOPORS, FANCG, RAD51, TP53BP1, and 3 anaplastic oligodendrogliomas), and seven anaplastic oligoastrocytomas tumors. One tumor had been classified as glioneuronal neoplasm. Human BIK) that could have a role in gliomagenesis. (Cancer Res 2005; male and female genomic reference DNA was purchased from Promega 65(10): 4088-96) (Madison, WI). Genomic DNA was isolated using the DNeasy Tissue Kit (Qiagen, Valencia, CA), DPNII (New England Biolabs, Beverly, MA) digested, Introduction and purified using the QIAquick PCR Purification Kit (Qiagen). Adult gliomas encompass a highly lethal group of tumors that DNA labeling and microarray hybridizations. Labeling of digested DNA and microarray hybridizations were done essentially as described (4), includes astrocytomas, oligodendrogliomas, and oligoastrocytomas. with slight modifications. Two micrograms of DNA were labeled using Genomic DNA copy number aberrations are key genetic events in random primers (Bioprime Labeling Kit, Invitrogen, Carlsbad, CA). Tumor gliomagenesis. Recurrent genomic regions of alteration in copy DNA and reference DNA were fluorescently labeled with Cy5 (red) and Cy3 number, including net gains and losses, have been found in these (green) dye (Amersham Biosciences, Piscataway, NJ), respectively. Tumor neoplasms. Whereas some of these regions contain known (or DNA was hybridized together with sex-matching reference DNA to a candidate) oncogenes and tumor suppressor genes, the biologically Stanford human cDNA microarray containing 41,421 cDNA elements, relevant genes within other regions remain to be identified (1). corresponding to 27,290 different UniGene cluster IDs. Comparative genomic hybridization (CGH) has been used to Data analysis. Microarrays were scanned on a GenePix 4000B scanner analyze DNA copy number changes in various human cancers, (Axon Instruments, Union City, CA). Primary data collection was done using including gliomas (2, 3). This karyotype-based method, however, GenePix Pro 5.1 software. Raw data were deposited into the Stanford Microarray Database. Measurements with consistent (regression correla- has limited mapping resolution, and gains or losses must be several tion, >0.6) and sufficient fluorescent intensities (reference wavelength megabases in size to be detected. Microarray-based CGH (array- channel, >2.5 above background) were considered reliable. Raw element CGH) provides a higher-resolution means to map DNA copy intensities were background corrected and normalized using SNOMAD data number alterations (4). cDNA microarrays in particular permit analysis tools (http://pevsnerlab.kennedykrieger.org/snomad.htm). Gene gene-by-gene analysis of aberrations in gene copy number. Here, copy numbers were reported as a moving average (symmetrical 3-/5-/7- we have used 42,000-element array-CGH technology with the aim nearest neighbors). to generate highly precise and comprehensive gene copy number The GoldenPath Human Genome Assembly (http://genome.ucsc.edu, National Center for Biotechnology Information build 34) was used to map log intensity ratios of the arrayed human cDNAs to chromosomal positions. Note: Supplementary data for this article are available at Cancer Research Online The CaryoScope (http://genome-www5.stanford.edu/cgi-bin/caryoscope/ (http://cancerres.aacrjournals.org/). nph-aCGH-dev_update.pl) and TreeView software (6) were used to display Requests for reprints: Markus Bredel, Division of Oncology, Stanford University gene copy number ratios along the human genome. Altered regions were School of Medicine, 269 Campus Drive, CCSR-1120, Stanford, CA 94305-5151. Phone: 650-498-6949; E-mail: [email protected]. also identified and visualized by the CGH-Plotter MATLAB toolbox, by I2005 American Association for Cancer Research. means of mean filtering, k-means clustering, and dynamic programming (7). Cancer Res 2005; 65: (10). May 15, 2005 4088 www.aacrjournals.org Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 2005 American Association for Cancer Research. Array-CGH in Human Glial Brain Tumors Unsupervised hierarchical clustering was done in Cluster (6), and two- 10p and terminal 3p, the latter of which was associated with way complete linkage clustering based on Pearson correlation as distance gains of terminal 12p in two cases. metric was applied. A correlation matrix representing all gene-to-gene Characterization of known critical amplicons and deletions. correlations was constructed in MATLAB using the built-in corrcoef Figure 1B shows selected genomic regions that have been strongly function. Supervised class prediction analysis was done using the nearest implicated in gliomagenesis (see also Supplementary Fig. S8). In shrunken centroids method implemented in the prediction analysis of microarrays package (8). Class predictive genes were identified based on each of the regions, the gene primarily implicated as the ‘‘driving’’ minimal misclassification error in balanced 10-fold cross-validation. target gene has been color-coded in red (amplicon) or green Real-time PCR. Quantitative real-time PCR reactions were done with (deletion). The gene-by-gene nature of our approach permitted the the ABI Prism 7900HT Sequence Detection System using SYBR GREEN dissection of exact amplicon and deletion boundaries within these PCR Master Mix (Applied Biosystems, Foster City, CA). Primers targeting regions and thus the identification of coaltered genes, some of introns of the TOPORS, FANCG, RAD51, TP53BP1, BIK, and ADAR genes which may contribute to tumorigenesis. For example, PDGFRA was were designed with the Primer3 program (http://frodo.wi.mit.edu/cgi-bin/ coamplified with the oncogene KIT in two tumors and with the primer3/primer3_www.cgi) and synthesized at the Stanford PAN Facility vascular endothelial growth factor receptor gene KDR and the (for sequences, see Supplementary Fig. S10). Thermocycling for each IGFBP7 gene (data not shown) in one tumor. Several EGFR reaction was carried out in a final volume of 20 AL containing 10 ng of amplicons included the GBAS gene (9) and two glioblastomas genomic DNA, forward and reverse primers at 300 nmol/L final concentration, and 1Â SYBR GREEN PCR Master Mix. After 10 minutes showed (noncontiguous) coamplification of the IGFBP1 and of initial denaturation at 95jC, the cycling conditions of 40 cycles consisted IGFBP3 genes, distal to EGFR. The CDK4 amplicon partly included of denaturation at 95jC for 15 seconds, annealing at 55jC for 30 seconds, the GEFT (10), OS-9 (11), and AKT-stimulating CENTG1 (12) genes, and elongation at 72jC for 30 seconds. All reactions were done in triplicate. and the candidate oncogene CTDSP2 (13). The two glioblastomas Dissociation curve analysis was done after every run to confirm the primer with MDM2 amplification showed coamplification of the putative specificity. Gene quantities