MOLECULAR CARCINOGENESIS 36:6-14 (2003)
ARTICLES
Chromosomal Abnormalities in Human Glioblastomas: Gain in Chromosome 7p Correlating With Loss in Chromosome 10q
Maria del Mar Inda,1 Xing Fan,1 Jorge Muňoz,1 Christine Perot,2 Didier Fauvet,2 Giselle Danglot,2 Ana Palacio,3 Pilar Madero,3 Idoya Zazpe,4 Eduardo Portillo,4 Teresa Tufión,5
José Maria Martínez-Peňuela,5 Jorge Alfaro,6 José Eiras,7 Alain Bernheim,2 and Javier S. Castresana1*
1Molecular N euro-Oncology Laboratory, University of Navarre Medical School, Pamplona, Spain
2Laboratoire de Génomique Cellulaire des Cancers, UMR 1599 CNRS, Inštitút Gustave Roussy, Villejuif, France
3Centro de Análisis Genéticos-Citogen, S.L., Zaragoza, Spain
4Servicio de Neurocirugia, Hospital de Navarra, Pamplona, Spain
sServicio de Anatómia Patológica, Hospital de Navarra, Pamplona, Spain
6Servicio de Anatómia Patológica, Hospital Miguel Servet, Zaragoza, Spain
7Servicio de Neurocirugia, Hospital Miguel Servet, Zaragoza, Spain
Various genomic alterations have been detected in glioblastoma. Chromosome 7p, with the epidermal growth factor receptor locus, together with chromosome 10q, with the phosphatase and tensin homologue deleted in chromosome 10 and deleted in malignant brain tumors-1 loci, and chromosome 9p, with the cyclin-dependent kinase inhibitor 2A locus, are among the most frequently damaged chromosomal regions in glioblastoma. In this study, we evaluated the genetic status of 32 glioblastomas by comparative genomic hybridization; the sensitivity of comparative genomic hybridization versus differential polymerase chain reaction to detect deletions at the phosphatase and tensin homologue deleted in chromosome 10, deleted in malignant brain tumors-1, and cyclin-dependent kinase inhibitor 2A loci and amplifications at the cyclin-dependent kinase 4 locus; the frequency of genetic lesions (gain or loss) at 16 different selected loci (including oncogenes, tumor-suppressor genes, and proliferation markers) mapping on 13 different chromosomes; and the possible existence of a statistical association between any pair of molecular markers studied, to subdivide the glioblastoma entity molecularly. Comparative genomic hybridization showed that the most frequent region of gain was chromosome 7p, whereas the most frequent losses occurred on chromosomes 10q and 13q. The only statistically significant association was found for 7p gain and 10q loss.  © 2002 Wiley-Liss, Inc.
Key words: phosphatase and tensin homologue deleted in chromosome 10; deleted in malignant brain tumors-1; cyclin-dependent kinase inhibitor 2A; epidermal growth factor receptor; comparative genomic hybridization
INTRODUCTION
Glioblastoma (GBM) is the most frequent and aggressive kind of brain tumor. GBMs have been studied extensively at the molecular level. Usually, only a few loci are explored per study [1 -9]. Genome-wide scans also have been performed to detect the genetic lesions linked to GBM development, by the use of comparative genomic hybridization (CGH), a technique that shows the gained and lost chromosomal regions in tumor DNA in a single experiment [10-13]. In our study we first evaluated the genetic changes in 32 GBMs by a CGH approach. We specifically explored 16 loci, most of them documented so far to be involved in GBM development: 11 correspond to oncogenes and loci related to cell proliferation (epidermal growth factor receptor [EGFR], platelet-derived growth factor receptor [PDGFR],
myc, cyclin-dependent kinase 4 [CDK4], PCNA, and the [E2F] transcription factor family of genes, E2F1-E2F6) and five tumor-suppressor gene-like loci (phosphatase and tensin homologue deleted in
* Correspondence to: Departmento de Genetica, Universidad de Navarra, Irunlarrea, s/n, E-31008 Pamplona, Spain.
Received 16 January 2002: Revised 16 August 2002: Accepted 2 October 2002
Abbreviations: GBM, glioblastoma: CGH, comparative genomic hybridization: EGFR, epidermal growth factor receptor: PDGFR, platelet-derived growth factor receptor: CDK4, cyclin-dependent kinase 4; PTEN, phosphatase and tensin homologue deleted in chromosome 10: DMBT1, deleted in malignant brain tumors-1: CDKN2A, cyclin-dependent kinase inhibitor 2A: PCR, polymerase chain reaction: LOH, loss of heterozygosity.
DOI 10.1002/mc. 10085
© 2002 WILEY-LISS, INC
7p GAIN/1 Oq LOSS CORRELATION IN GLIOBLASTOMA
chromosome 10 [PTEN], deleted in malignant brain tumors-1 [DMBT1], cyclin-dependent kinase inhibitor 2A [CDKN2A], tumor protein [TP] 53, and retinablastoma [RB]).
MATERIALS AND METHODS
Tumor samples were obtained from 32 patients with GBM who underwent surgery at the Departments of Neurosurgery of the General Hospital of Tianjin Medical University, Tianjin, People's Republic of China (22 patients); Hospital de Navarra, Pamplona, Spain (eight patients); and Hospital Miguel Servet, Zaragoza, Spain (two patients). A case of pilocytic astrocytoma (grade I) was included as a negative control. Genomic DNA from all tumors was extracted by standard methods. CGH analysis was performed according to the standard protocol, with slight modifications [14]. Differential polymerase chain reaction (PCR) assays were designed to test for homozygous deletions at the loci CDKN2A (9p21), PTEN (10q23.31), and DMBT1 (10q25.3-q26.1) [15] and for gene amplification at the locus CDK4 (12ql4) [3,16]. A methylation-specific PCR strategy to detect CDKN2A promoter hypermethylation also was applied [17]. Statistical analysis to study the hypothetical correlation between frequencies of lesions at any pair of damaged loci was by the chi-square test, including Fisher's exact test.
RESULTS AND DISCUSSION
Genetic Profile of GBM
The chromosomal regions with DNA copy number alterations (losses and gains) identified in all 31 GBMs by CGH are illustrated in Figure 1 and Table 1. Overall, gains occurred slightly more frequently than deletions (59 of 109 total aberrations). The most common gains occurred on chromosome 7 (65%). The next most common relative gains in genetic material involved chromosomes 20 (39%), 8 (35%), and 9 (26%). Finally, chromosomes 1 and 5, showing gains in 23% of GBMs, were among the chromosomes on the borderline of 25% gains. The most common losses in this series of GBMs involved chromosome 10, with 58% of cases affected. Regions of chromosomes 13 (35%) and 6 (26%) also were lost in these tumors.
Gains of genetic material
The frequency of gains of genetic material at 7p has been reported to range from 25% to 70% in different studies [11,12]. Candidates for oncogenes on 7p include EGFR, which is assigned to 7pl2. The EGFR gene is the most frequently amplified oncogene in astrocytic tumors; it is amplified in about one-third of GBMs and in a few anaplastic astrocytomas [4].
We found gains on chromosome 20 in 36% of GBM cases. Similar results have been published elsewhere
[13]. Candidate oncogenes on 20q are hematopoietic cell kinase (20qll); AIB1, a steroid receptor co-activator; and BTAK, a serine/threonine kinase that has been found to be amplified and overexpressed in breast cancer. The PTPN1 gene, located at 20ql2, is a nonreceptor tyrosine phosphatase involved in growth regulation and recently has been reported to be overexpressed in 72% of breast carcinomas. Another candidate gene is MYBL2, at 20ql3, which encodes a transcription factor and plays an important role in cell-cycle progression. Furthermore, the human cellular apoptosis susceptibility gene (CAS) has been mapped to this same region [18].
It has not been reported that any of these oncogenes is amplified in gliomas, which suggests the presence of other oncogenes amplified at 20q that have yet to be identified. PCNA [19] and E2F1 [20], both of which act as proliferating factors in numerous tumors, also map to chromosome 20. The next most frequent gains were noted at chromosome 8. The main candidate oncogene on 8q is myc (8q24). Amplification of myc in gliomas has been reported [21]. Mohapatra et al. [22] found five GBMs with gains at 9q. One candidate oncogene at 9q is TRKB (9q22.1). This oncogene has been found to be expressed in gliomas [23].
Losses of genetic material
The most frequent loss found in this study was on chromosome 10 (58%). This finding agrees with those of previous studies [22]. At least three separate regions on chromosome 10 have been implicated in GBM [5]: one in the telomeric region of lOp and the others in the telomeric and centromeric regions of lOq. Our results showed that the common region of loss on chromosome 10 was 10q23-qter, which suggests inactivation of tumor-suppressor genes in the telomeric region of lOq. It has been found that the PTEN gene, on 10q23.31, is mutated in various malignant tumors, including GBM [24,25]. DMBT1, on 10q25.3-q26.1, has been reported to carry intragenic homozygous deletions in about 23% of GBMs [26].
Other losses were seen frequently on chromosomes 13q and 6q. Inactivation of the retinoblastoma gene on 13q in gliomas has been reported previously [6]. The loss on 9p probably is related to the genes encoding pl5 and pl6 [7]. Structural rearrangements involving 9p are believed to target homozygous deletion of pl6 and/or pl5 in 35-40% of high-grade GBMs [7]. Only 23% of GBMs in our series had loss on 9p. This low frequency may be related to the limited sensitivity of CGH, which depends on both the copy number and the size of the region, whereas other methods, such as differential PCR, depend only on copy number. If the size of the region at 9p is below the sensitivity of CGH, a deletion may still be detected by other methods.
Figure 1. Summary of all DNA copy number changes detected by CGH in 32 glioblastomas. Each line illustrates the affected chromosome region in a single tumor sample. The vertical lines on the right side of the chromosome ideograms indicate gains, and the lines on the left represent losses.
26
7p GAIN/lOq LOSS CORRELATION IN GLIOBLASTOMA 9
Table 1. Genetic Profiles of 32 Patients With Glioblastoma
No.	Regions of DNA copy number changes		Number of chang		es
	Loss	Gain	Losses	Gains	All
GB5		11q13, 20q, 22q	0	3	3
GB6		8q23-qter, 9q22-qter, 15, 16, 17, 19,	0	8	8
		20, 22q			
GB7	4	1q, 7p, 8q, 15, 22q	1	4	5
GB9			0	0	0
GB10			0	0	0
GB12	3p22-qter, 6q, 10q, 11,	4p14-pter, 5p, 5q14-qter, 6p, 7, 8p,	7	9	16
	13, 16q, 18	9q22-qter, 15, 16p, 20			
GB14	4q, 21	2p21-pter, 7,8, 11p11-p12, 13q11-q21,20	2	6	8
GB15		8p21-pter, 12q13-qter	0	2	2
GB16	6q, 10	1, 2p21-pter, 7, 9q22-qter	2	4	6
GB17	3p13-q23, 6, 10, 13	3p14-pter, 3q25-qter, 4q27-qter, 7	4	4	8
GB19			0	0	0
GB20	4q, 13	1, 2p14-pter, 2q14-qter, 2q35-qter,	2	14	18
		5p14-pter, 5q31-qter, 6p, 6q25-qter, 8,			
		12p, 12q15-qter, 15q14-q21,			
		15q24-qter, 16, 20			
GB21	6, 9p, 10, 13	3p14-pter, 3q26-qter, 4p15-pter, 7,	4	7	11
		9q, 11 p12-pter, 12q24-qter			
GB22	10	2, 5, 7, 8p, 12, 15	1	6	7
GB24	3p11-pter, 9p, 10q24-qter,	3q12-qter, 5p, 6p, 6q24-qter, 7p, 8, 9q, 11,	5	11	16
	13, 15q11-q24	12q23-qter, 14, 16p, 17q, 18q, 19, 21			
GB25	3p14-q13, 4q27-qter, 9p	1q21-q23, 6p, 8p21-pter, 8q23-qter,	7	6	13
	13-pter, 10q23-qter, lip.	12p12-pter, 16p			
	12q11-q24, 13q11-q32				
GB27	6q11-q23	12	1	1	2
GB28	6p12-q21, 11p13-qter, 13	1 p22-pter, 7	3	2	5
GB29			0	0	0
GB30	3pter-q26, 5q31-qter, 10, 21	2p22-pter, 3q27-qter, 5q31-pter, 7, 8, 9, 20	4	5	9
GB32	2q32-qter, 4p14-q27,	1, 3q, 5p, 5q31-qter, 6p, 7p, 7q21-qter, 9,	4	9	13
	11q21-qter, 13	12q22-qter, 17q, 20			
GB33	10p, 13	1, 2, 3, 5q31-qter, 6p21-pter, 7, 8q23-qter,	2	11	13
		9q, 17q, 20, 22			
HN11	1p32.2-pter, 8q22.3-q23.2,	4q24-qter, 7, 14q31-qter, 19, 20	4	5	9
	9p22, 10				
HN15	1q11-q25, 9p, 10	7pter-q11.22, 19,	3	2	5
HN16	6q16-q22, 9p13-pter, 10, 22q	7, 19p 13.2-qter, 20	4	3	7
HN19	10, 13q14-q31	7	2	1	3
HN20	6, 8q22-q24, 9q12-13, 10,	5p15-pter, 17q23-qter, 18pter-q11,	5	6	11
	13q, 14q	1821.3-qter, 20, 21, 22			
HN26			0	0	0
HN51	10	7p14-q1 1.22, 20, 19	1	3	4
HN53	9p13-p23, 10, 13q,	1q21-q23, 7	4	2	6
	15q15-q25				
Z3	10	4q11-q13.1, 7p15-qter	1	1	2
Z6	1p34-pter, 10, 14q21-qter,	4q11-q13, 7	5	2	7
	19p13-pter, 22q				
Total			50	59	109
Differential PCR Versus CGH to Detect Deletions or Amplifications at Genomic Loci
To compare the very sensitive molecular technique differential PCR [3,15], which aims to show deletion or amplification of genomic material at a particular chromosomal locus, with CGH [27],
which shows a panoramic overview of all chromosomal loci, presenting losses and gains in a single experiment, we decided to examine by both methods the 22 GBMs from the General Hospital of Tianjin Medical University. First, tumor DNA was subjected to differential PCR as previously described [3,15]. Cases with homozygous deletions of CDKN2A,
INDA ET AL.
10
PTEN, and DMBT1 and amplification of CDK4 are shown in Table 2. Then CGH was performed on the same 22 cases. As shown in Table 2, CDKN2A showed homozygous deletions in six of 22 (27%) GBMs. Two of the 22 cases (9%) (GB6 and GB28) were found to contain pi 6 promoter hypermethylation when the methylation-specific PCR technique was performed [15]. Hypermethylation of the CDKN2A promoter can explain a loss of function of the pl6 protein in the absence of deletions of the chromosomal locus. In contrast, in samples GB15, GB17, GB21, GB24, GB30, and GB33, the CDKN2A locus was found to be homozygously deleted, which caused an obvious loss of function of pl6.
Only two of the six samples (GB21 and GB24) showed chromosomal loss after CGH analysis. Samples GB15, GB17, GB30, and GB33 might have suffered a small deletion in the CDKN2A locus not detected by CGH. CGH revealed a loss of 9pl 3-pter in sample GB25, but concordance could not be estab-
Table 2. Deletions and Amplifications by PCR and CGH
Differential
PCR CGH
Deletions of CDKN2A
GB21 + +
GB24 + +
GB25 - +
GB15 +
GB17 +
GB30 +
GB33 + GB6-met* GB28-met*
Deletions of PTEN
GB16 + +
GB30 + +
GB12 - +
GB17 - +
GB21 - +
GB22 - +
GB25 - +
Deletions of DMBT1
GB16 - +
GB30 - +
GB12 - +
GB17 - +
GB21 - +
GB22 - +
GB24 - +
GB25 - +
Amplification of CDK4
GB15 + +
GB27 + +
GB28 +
*met, CDKN2A promoter methylation, as detected by a methylation-specific PCR technique.
lished with the molecular study. This sample, in fact, might have harbored loss of heterozygosity (LOH) at the locus CDKN2A (shown by CGH but not by differential PCR). It would have been necessary to conduct an LOH study to clarify the genetic lesion, but we did not have blood DNA from the patients, which is required for LOH.
Table 2 shows the results for PTEN. Two of 22 cases of GBMs (9%)—GB16 and GB30—showed homozygous deletions of PTEN, but seven of 22 (32%) GBMs showed deletions by CGH. DMBT1 results also are presented in Table 2, showing no concordance at all between the molecular and the cytogenetic techniques in eight cases of GBM. It seems from the PTEN and DMBT1 results that lOq was a frequent target of deletions in GBM (eight of 22, or 36%), but the corresponding pattern of homozygous deletions of PTEN and/or DMBT1, two important tumor-suppressor genes, did not occur. This might indicate that a loss at lOq, detected by CGH and perhaps corresponding to a lOq— cytogenetic event, is more common than homozygous deletion itself. This could mean that the second allele at the PTEN or DMBT1 locus (or both) might undergo a different genetic event to fulfil Knudson's two-hit hypothesis. The second lesion might correspond to point mutations or even promoter hypermethylation of those genes.
Table 2 also shows that only three of 22 cases had amplification of CDK4 [3] at chromosome 12q. In two cases (GB15, and GB27), there was a corresponding CGH result. After comparison of differential PCR and CGH to detect gains and losses at specific loci, we concluded that losses at the chromosome 9 CDKN2A locus corresponded to small interstitial deletions not usually detected by CGH. CGH proved to be a more appropriate technique for the detection of deletions on chromosome 10 than differential PCR, although this might have been due to a higher frequency of LOH versus homozygous deletions at lOq in GBMs.
Gains or Losses at 16 Selected Loci in GBM
Once we had results from the analysis of 32 GBMs by the CGH technique (Table 1 and Figure 1), we studied alterations in 16 different chromosomal loci selected on the basis of their known or supposed role in gliomagenesis. Eleven corresponded to oncogene-like loci (EGFR, PDGFR, myc, CDK4, and such proliferation markers as PCNA and the E2F transcription factor family of genes, E2F1-E2F6) and five to tumor-suppressor genelike loci (PTEN, DMBT1, CDKN2A, TP53, and RB). The selected loci placed at their specific chromosomes and the typical lesion by which they are affected are shown in Figure 2. A summary of gains and losses at those specific loci in the different patients is presented in Table 3.
E2F2	III
	-
	■
	
	
	
	■
	
	
1	
	
	
=E2F5=	
	■
: MYC:	
CJUUatTlJCJ
UJlllliLLFitjr'LLI
BUBBD : B
PDGFR
Sojftjflilijiil
:E2F3:
£GFR-C
ll
lY l rilfM lY l i-EJ--11 J^ J ^ ^ J lyil^l'l W|-
UUaCWN IITIITI
CDKN2AI
8
ll
■ -3 Y*. V* -> 7
CDK4:
I
12
o
-Q r-O K>
8
g
si O
5
CD
S
i/i
3
J n.T.ifir-1(-+-i3T■:J
1=RB=
E2F4I
£TP53:
r-kTI.' rj-r.,-"- T 11.1 , £ Ljl; 1,1; ftft | f.v *-^- * :
:E2F6
Figure 2. Summary of DNA copy number changes detected by CGH in 32 glioblastomas. Only 16 selected loci from 13 different chromosomes were considered for analysis. Each line illustrates the affected chromosome region in a single tumor sample. The vertical lines on the right side of the chromosome ideograms indicate gains, and the lines on the left represent losses.
86
Table 3. Gains and Losses Determined in 32 Glioblastoma at 16 Different Loci From 13 Chromosomes*
E2F2 PDGFR E2F3 EGFR 1p36     5q31-q32     6p22 7p12
E2F5 myc CDKN2A      PFEN DMBF1 CDK4 RB
8q21.13     8q24.12-q24.13      9p21       10q23.3      10q25.3-q26.1       12q14 13q14.2
E2F4 FP53 PCNA E2F1 E2F6
15q21-q22     17p13.1     20pter-p12     20q11.2      22q11
GB5	—	—
GB6	—	—
GB7	—	—
GB9	—	—
GB10	—	—
GB12	—	G
GB14	—	—
GB16	—	—
GB16	G	—
GB17	—	—
GB19	—	—
GB20	G	G
GB21	—	—
GB22	—	G
GB24	—	—
GB25	—	—
GB27	—	—
GB28	G	—
GB29	—	—
GB30	—	G
GB32	G	G
GB33	G	G
HN11	G	—
HN15	—	—
HN16	—	—
HN19	—	—
HN20	—	—
HN26	—	—
HN51	—	—
HN53	—	—
Z3	—	—
Z6	L	—
G G
G L
G G
G G
G G
G G
G G G
G G G G G G G
G G G G
G L
G G
G G
L G
L G
G G
G G G G
G
G
G
G G
G G
G G G
G G G G
G
G
G
L G
*G = gain; L = loss; — = no gain or loss. HN20, secondary glioblastoma. HN26, pilocytic astrocytoma (grade I).
7p GAIN/lOq LOSS CORRELATION IN GLIOBLASTOMA
13
Only gain of chromosome 7p (containing EGFR) significantly associated with lOq deletions (containing PTEN and/or DMBT1 among the most frequent targets of deletion; p = 0.003). In our study, 75% of cases with a gain of 7p had a simultaneous loss of chromosome 10 and 88% of cases with loss of chromosome 10 had gains at 7p. Other studies showed that EGFR amplification occurs at similar frequencies in GBMs with or without homozygous deletions or mutations of the PTEN gene, located on 10q23 [8], although about 20% of primary GBMs show both EGFR amplification and PTEN mutations [9]. This might imply that other oncogenes on 7p have some kind of regulatory association with suppressor genes on chromosome 10.
In conclusion, our data showed that CGH was a comprehensive and rapid approach for the analysis of global and locus-specific genomic alterations in glial tumors. Therefore, it may serve as an additional tool to define subsets of human GBMs exhibiting different genetic pathways. Nevertheless, owing to its sensitivity, CGH should be combined with other, more sensitive techniques, such as differential PCR, to guarantee detection of small deletions.
ACKNOWLEDGMENTS
We thank Oskar Sáez for his help and advice with the artwork. This research was supported in part by grants from the Departamento de Salud del Gobierno de Navarra, Fondo de Investigación Sanitaria (FIS 99/1265), Fundación Científica de la AECC, and Ministerio de Ciencia y Tecnología (PROFIT), Madrid, Spain. Maria del Mar Inda and Jorge Mufioz are predoctoral fellows from the Ministerio de Educación, Cultura y Deporte, Madrid, Spain. Xing Fan received funds from the Agenda Espafiola de Cooperación Internacionál, and Ministerio de Ciencia y Tecnología, Madrid, Spain.
REFERENCES
1. Davies MP, Gibbs FE, Halliwell N, et al. Mutation in the PTEN/ MMAC1 gene in archival low grade and high grade gliomas. Br J Cancer 1999;79:1542-1548.
2. Fan X, Furnari FB, Cavenee WK, Castresana JS. Non-isotopic silver-stained SSCP is more sensitive than automated direct sequencing for the detection of PTEN mutations in a mixture of DNA extracted from normal and tumor cells. Int J Oncol 2001;18:1023-1026.
3. Fan X, Gomez L, Nistal M, Sierrasesumaga L, Castresana JS. Differential polymerase chain reaction: A technical comparison of three methods for the detection of CDK4 gene amplification in glioblastomas. Int J Oncol 1998;13:963-966.
4. Wong AJ, Bigner SH, Bigner DD, Kinzler KW, Hamilton SR, Vogelstein B. Increased expression of the epidermal growth factor receptor gene in malignant gliomas is invariably associated with gene amplification. Proc Natl Acad Sei USA 1987;84:6899-6903.
5. Karlbom AE, James CD, Boethius J, et al. Loss of heterozygosity in malignant gliomas involves at least three distinct regions on chromosome 10. Hum Genet 1993;92: 169-174.
6. Henson JW, Schnitker BL, Correa KM, et al. The retinoblastoma gene is involved in malignant progression of astrocytomas. Ann Neurol 1994;36:714-721.
7. Walker DG, Duan W, Popovic EA, Kaye AH, Tomlinson FH, Lavin M. Homozygous deletions of the multiple tumor suppressor gene 1 in the progression of human astrocytomas. Cancer Res 1995;55:20-23.
8. Liu W, James CD, Frederick L, Alderete BE, Jenkins RB. PTEN/ MMAC1 mutations and EGFR amplification in glioblastomas. Cancer Res 1997;57:5254-5257.
9. Tohma Y, Gratas C, Biernat W, et al. PTEN (MMAC1) mutations are frequent in primary glioblastomas (de novo) but not in secondary glioblastomas. J Neuropathol Exp Neurol 1998;57:684-689.
10. Mao X, Hamoudi RA. Molecular and cytogenetic analysis of glioblastoma multiforme. Cancer Genet Cytogenet 2000; 122:87-92.
11. Kim DH, Mohapatra G, Bollen A, Waldman FM, Feuerstein BG. Chromosomal abnormalities in glioblastoma multiforme tumors and glioma cell lines detected by comparative genomic hybridization. Int J Cancer 1995; 60:812-819.
12. Jung V, Romeike BF, Henn W, et al. Evidence of focal genetic microheterogeneity in glioblastoma multiforme by area-specific CGH on microdissected tumor cells. J Neuropathol Exp Neurol 1999;58:993-999.
13. Nishizaki T, Ozaki S, Harada K, et al. Investigation of genetic alterations associated with the grade of astrocytic tumor by comparative genomic hybridization. Genes Chromosom Cancer 1998;21:340-346.
14. Fan X, Paetau A, Aalto Y, et al. Gain of chromosome 3 and loss of 13q are frequent alterations in pituitary adenomas. Cancer Genet Cytogenet 2001;128:97-103.
15. Fan X, Muhoz J, Sanko SG, Castresana JS. PTEN, DMBT1 and p16 alterations in diffusely infiltrating astrocytomas. Int J Oncol 2002;21:667-674.
16. Fan X, Gomez L, Nistal M, Sierrasesumaga L, Castresana JS. Lack of gene amplification as a mechanism of CDK4 activation in human neuroblastoma. Oncol Rep 1999;6: 647-650.
17. Fan X, Inda MM, Tuhön T, Castresana JS. Improvement of the methylation specific PCR technical conditions for the detection of p16 promoter hypermethylation in small amounts of tumor DNA. Oncol Rep 2002;9:181-183.
18. Knuutila S, Björkqvist AM, Autio K, et al. DNA copy number amplifications in human neoplasms: Review of comparative genomic hybridization studies. Am J Pathol 1998; 152:1107-1123.
19. Mottolese M, Natali PG, Coli A, et al. Comparative analysis of proliferating cell nuclear antigen and epidermal growth factor receptor expression in glial tumours: Correlation with histological grading. Anticancer Res 1998;18:1951-1956.
20. Dirks PB, Rutka JT. Current concepts in neuro-oncology: the cell cycle: A review. Neurosurgery 1997;40:1000-1013.
21. Galanis E, Buckner J, Kimmel D, et a I. Gene amplification as a prognostic factor in primary and secondary high-grade malignant gliomas. IntJ Oncol 1998;13:717-724.
22. Mohapatra G, Bollen AW, Kim DH, et al. Genetic analysis of glioblastoma multiforme provides evidence for subgroups within the grade. Genes Chromosom Cancer 1998;21:195— 206.
23. Wang Y, Hagel C, Hamel W, Muller S, Kluwe L, Westphal M. Trk A, B, and C are commonly expressed in human astrocytes and astrocytic gliomas but not by human oligodendrocytes and oligodendroglioma. Acta Neuropathol 1998;96:357-364.
14
INDA ETAL.
24. Li J, Yen C, Liaw D, et al. PTEN, a putative protein tyrosine phosphatase gene mutated in human brain, breast, and prostate cancer. Science 1997;275:1943-1947.
25. Steck PA, Pershouse MA, Jasser SA, et al. Identification of a candidate tumour suppressor gene, MMAC1, at chromosome 10q23.3 that is mutated in multiple advanced cancers. Nature Genet 1997;15:356-362.
26. Mollenhauer J, Wiemann S, Scheurlen W, et al. DMBT1, a new member of the SRCR superfamily, on chromosome 10q25.3-26.1 is deleted in malignant brain tumours. Nat Genet 1997;17:32-39.
27. Kallioniemi A, Kallioniemi OP, Sudar D, et al. Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors. Science 1992;258:818-821.