Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Frequent loss of chromosome 9, homozygous CDKN2A/p14ARF/CDKN2B deletion and low TSC1 mRNA expression in pleomorphic xanthoastrocytomas


The molecular pathogenesis of pleomorphic xanthoastrocytoma (PXA), a rare astrocytic brain tumor with a relatively favorable prognosis, is still poorly understood. We characterized 50 PXAs by comparative genomic hybridization (CGH) and found the most common imbalance to be loss on chromosome 9 in 50% of tumors. Other recurrent losses affected chromosomes 17 (10%), 8, 18, 22 (4% each). Recurrent gains were identified on chromosomes X (16%), 7, 9q, 20 (8% each), 4, 5, 19 (4% each). Two tumors demonstrated amplifications mapping to 2p23–p25, 4p15, 12q13, 12q21, 21q21 and 21q22. Analysis of 10 PXAs with available high molecular weight DNA by high-resolution array-based CGH indicated homozygous 9p21.3 deletions involving the CDKN2A/p14ARF/CDKN2B loci in six tumors (60%). Interphase fluorescence in situ hybridization to tissue sections confirmed the presence of tumor cells with homozygous 9p21.3 deletions. Mutational analysis of candidate genes on 9q, PTCH and TSC1, revealed no mutations in PXAs with 9q loss and no evidence of TSC1 promoter methylation. However, PXAs consistently showed low TSC1 transcript levels. Taken together, our study identifies loss of chromosome 9 as the most common chromosomal imbalance in PXAs and suggests important roles for homozygous CDKN2A/p14ARF/CDKN2B deletion as well as low TSC1 mRNA expression in these tumors.


Pleomorphic xanthoastrocytoma (PXA) is a rare astrocytic tumor that typically presents as a well-circumscribed, superficially located lesion in the cerebral hemispheres of children and young adults, often after a long-standing history of epileptic seizures (Kepes et al., 2000). Characteristic histologic features include pleomorphic and lipidized astrocytic cells with expression of glial fibrillary acidic protein, a dense stromal reticulin network and prominent perivascular lymphocytic infiltrates (Kepes et al., 2000). Most cases of PXAs grow slowly and correspond histologically to World Health Organization (WHO) grade II. However, a subset of PXAs shows histologic features of anaplasia, such as increased mitotic activity and/or necrosis. These tumors are designated as PXA with anaplastic features (Kepes et al., 2000). In comparison to the diffusely infiltrating astrocytic gliomas, PXAs show a more favorable prognosis as indicated by a 10-year survival rate of approximately 70% (Giannini et al., 1999).

The etiology of PXAs is unknown. Rare cases of PXA have been reported in patients with neurofibromatosis type 1 but most tumors arise sporadically without evidence of a genetic predisposition (Kepes et al., 2000). To date, only few cytogenetic and molecular genetic studies have been performed on these tumors. Cytogenetic aberrations found in one PXA each were (a) trisomy 3 and 5 as well as monosomy 22 (Sawyer et al., 1991, b) telomeric associations resulting in subclones with ring chromosomes (Sawyer et al., 1992), and (c) translocations at band 1q42 in both homologs of chromosome 1 (Li et al., 1995). Comparative genomic hybridization (CGH) analysis of three PXAs detected gain on chromosome 7 and loss on 8p as recurrent imbalances (Yin et al., 2002). Molecular genetic studies revealed that mutations in the TP53 tumor-suppressor gene are restricted to a small fraction of cases (Paulus et al., 1996; Giannini et al., 2001; Kaulich et al., 2002). Thus, to date no specific genetic aberration has been linked to PXAs.

Here, we report on the results of a genome-wide screening for chromosomal imbalances in 50 PXAs using CGH analysis (Kallioniemi et al., 1992; du Manoir et al., 1993). The CGH studies were supplemented by high-resolution array-based CGH (array-CGH) experiments (Solinas-Toldo et al., 1997; Pinkel et al., 1998), interphase fluorescence in situ hybridization (FISH) to tissue sections, and molecular genetic analyses of selected genes. The results of this integrated approach indicate loss of chromosome 9 as the most common chromosomal alteration in PXAs. Furthermore, we identified homozygous deletions involving the CDKN2A/p14ARF and CDKN2B tumor-suppressor gene loci at 9p21.3 in a significant fraction of these tumors and found that PXAs consistently show low expression of transcripts from the TSC1 gene at 9q34.


Genomic imbalances detected by chromosomal CGH

CGH analysis revealed chromosomal imbalances in 38 of 50 PXAs (76%), with a range of 0–10 alterations per tumor (mean±s.e.m.: 1.58±0.27). PXAs with anaplastic features (n=14) showed a slightly higher mean number of alterations per tumor as compared to WHO grade II PXAs (n=36) (mean±s.e.m.: 1.64±0.65 versus 1.55±0.28). The most common alteration identified in 25/50 PXAs (50%) was loss on chromosome 9 (Table 1, Figure 1). In most instances, both arms of chromosome 9 were lost. However, a subset of tumors showed losses restricted to 9p (n=7) and one PXA displayed a loss of 9q only. Less common losses affected chromosome 17 (5/50; 10%). The most frequent gains were on chromosome X (8/50; 16%), 7, 9q and 20 (4/50; 8% each) (Figure 1). Six different high-level amplifications were found in two PXAs and mapped to 2p23–p25, 4p15, 12q13, 12q21, 21q21 and 21q22, respectively.

Table 1 Summary of selected clinical data, results of immunohistochemical analysis and chromosomal comparative genomic hybridization in 50 pleomorphic xanthoastrocytomas
Figure 1

Imbalances detected by chromosomal CGH analysis of 50 PXAs (a) and 19 diffuse astrocytomas (b). Losses are given by lines on the left side of a chromosome ideogram; gains by lines on the right side; high-level amplifications by bold lines on the right side.

Comparison of chromosomal CGH results in PXAs versus diffuse astrocytomas

Chromosomal CGH results obtained for the 50 PXAs were compared with those from 19 diffuse astrocytomas of WHO grade II (nine cases included in this study, 10 cases reported by us previously (Weber et al., 1996)) (Figure 1). The alteration found most frequently in PXAs, that is, loss on chromosome 9, was not detected in any of the diffuse astrocytomas (25/50 versus 0/19, P=0.0001). The most common gains and losses identified in diffuse astrocytomas were significantly less frequent in PXAs: gain on 7 (7/19 versus 4/50, P=0.007), gain on 8 (5/19 versus 0/50, P=0.001), gain on 12 (4/19 versus 2/50, P=0.04), loss on 11p (4/19 versus 0/50, P=0.005), loss on Xp (4/19 versus 0/50, P=0.005).

Genomic imbalances identified by array-CGH

Imbalances were detected in eight of 10 PXAs analyzed by array-CGH (Figure 2) with a range of 0–6 (mean±s.e.m.: 2.9±0.68) alterations per tumor. The most frequent alteration was complete or partial loss of chromosome 9 detected in 7/10 cases (70%). The other recurrent alterations were combined gains of chromosomes 5 and 7 (3/10), and gain of 20 (2/10). In five of 10 tumors, the array-CGH pattern was consistent with a complete or partial loss of one chromosome 9 and a homozygous deletion involving the CDKN2A/p14ARF and CDKN2B loci at 9p21.3 (Figure 2a–e). In addition, two PXAs were found to have deletions of CDKN2A/p14ARF and CDKN2B without a loss of an entire chromosome 9 (Figure 2f and g). The breakpoints of the partial losses on 9p, 18p and 19p detected in PXA51 were mapped to 9p13.3, 18p11.21 and 19p12–p13.11 (Figure 3).

Figure 2

Genomic profiles of the 10 PXAs analyzed by array-based CGH. Midpoints of all clones are plotted in genomic order from 1p to Yq on the X axis against their normalized log2 test to reference ratio on the Y axis. The following tumors are shown: (a) PXA52. (b) PXA47. (c) PXA53. (d) PXA85. (e) PXA51. (f) PXA48. (g) PXA43. (h) PXA71. (i) PXA46. (j) PXA49. Profiles (a–e) show losses of chromosome 9 (a–d) or terminal loss of 9p (e) and a more pronounced loss of clones at 9p21.3 including RP11-149I2 (encompassing the CDKN2A/p14ARF and CDKN2B loci) pointed out by an arrow. Profiles (f) and (g) show losses at 9p21.3 including clone RP11-149I2 (arrow) without loss of chromosome 9. Profiles (h–j) show neither chromosome 9 loss nor diagnostic losses of clone RP11-149I2 (arrow).

Figure 3

Array-CGH identification, fine-mapping and interphase-FISH verification of partial losses on 9p and 18p in tumor PXA51 (a–g) and of homozygous deletions of clone RP11-149I2 containing the CDKN2A/p14ARF and CDKN2B loci in PXA53 (h–j) and PXA43 (k–m). Chromosomal CGH profiles are given for comparison. Array-CGH profiles are given as log2 test to reference ratios (Y axis) against mbp positions of clone midpoints (X axis). Interphase-FISH was performed to tumor tissue sections. (a) Array-CGH profile of chromosome 9. (b) Chromosomal breakpoint at 9p13.3 with realistic clone lengths. (c) Interphase-FISH showed that RP11-182N22 (9p13.3) was deleted in 42% of cells. (d) Array-CGH profile of chromosome 18. (e) Chromosomal breakpoint at 18p11.21. (f and g) Interphase-FISH showed that RP11-703I16 (18p11.21) was deleted in 40% of cells. (h) Array-CGH profile of chromosome 9. (i) Detailed view of RP11-149I2. (j) Interphase-FISH showing that 83% of cells had homozygous deletion of RP11-149I2 (9p21.3) and one copy of clone RP11-3J11 (9q31.3). (k) Array-CGH profile of chromosome 9. (l) Detailed view of RP11-149I2. (m) Interphase-FISH showing that 32% of cells had homozygous deletion of RP11-149I2 (9p21.3) and two copies of RP11-3J11 (9q31.3). (n) Chromosomal CGH profile of chromosome 9 (PXA51). (o) Chromosomal CGH profile of chromosome 18 (PXA51). (p) Chromosomal CGH profile of chromosome 9 (PXA53). (q) Chromosomal CGH profile of chromosome 9 (PXA43).

Verification of array-CGH breakpoints and homozygous deletions by interphase FISH to tumor tissue sections

The breakpoints detected by array-CGH were confirmed by interphase-FISH to tumor sections (Figure 3a–g). In addition, three of seven PXAs with loss at 9p21.3 were analyzed by interphase-FISH to tumor sections (Figure 3h–m). In each case, homozygous deletions of the clone RP11-149I2, that includes the CDKN2A/p14ARF and CDKN2B loci, were detected in subsets of cells. In PXA53, the signal pattern of clones from 9p21.3, 9q31.3 and 4p12–p13 (Figure 3h–j) was consistent with a diploid karyotype containing only one copy of chromosome 9 and a homozygous deletion of the CDKN2A/p14ARF and CDKN2B loci in 20/24 cells (83%). In four cells, two signals were found for each clone. PXA52 showed the same aberrant signal pattern in 5/6 cells. In PXA43 (Figure 3k–m), the signal pattern was consistent with a diploid karyotype containing two copies of chromosome 9 that both have a deletion of the CDKN2A/p14ARF and CDKN2B loci in 6/19 cells (32%). In 13 cells, two signals were found for each clone. As the array-CGH patterns of five PXAs (profiles a–e in Figure 2) are indicative of homozygous CDKN2A/p14ARF and CDKN2B deletions (confirmed by FISH-analysis in two cases) and the loss at 9p21.3 in PXA43 was shown to correspond to a homozygous deletion by FISH-analysis (Figure 3k–m), at least six of the 10 PXAs investigated by array-CGH contain tumor cells with biallelic 9p21.3 deletions involving the CDKN2A/p14ARF and CDKN2B loci.

Molecular analysis of the PTCH tumor-suppressor gene

Single-strand conformation polymorphism (SSCP) analysis of the PTCH coding sequence revealed no evidence for mutations in 18 PXAs, including 16 tumors with losses on chromosome 9 including the PTCH gene locus. Real-time reverse transcription–polymerase chain reaction (PCR) analysis revealed increased PTCH transcript levels in PXAs relative to the reference brain tissues (mean 4.3, s.d. 4.1). There was no evidence for loss of PTCH mRNA expression in any of the PXAs. However, PTCH transcript levels were lower in PXAs as compared to diffuse astrocytomas (mean 21.9, s.d. 20.3, P<0.001) and anaplastic astrocytomas (mean 29.5, s.d. 32.0, P<0.001), but not as compared to giant cell glioblastomas (mean 3.0, s.d. 5.0, P=0.117).

Molecular analysis of the TSC1 tumor-suppressor gene

Mutation analysis of the TSC1 gene did not identify any somatic mutation in the 18 selected PXAs, 16 of which had loss on 9q as determined by CGH-analysis. Real-time reverse transcription–PCR analysis showed TSC1 transcript levels reduced to less than 20% relative to non-neoplastic brain tissue in 9/9 PXAs investigated (mean 0.12, s.d. 0.06) (Figure 4). TSC1 mRNA levels were significantly lower in PXAs as compared to diffuse astrocytomas (mean 0.93, s.d. 0.55, P<0.0001), anaplastic astrocytomas (mean 0.73, s.d. 0.61, P<0.001) and giant cell glioblastomas (mean 0.27, s.d. 0.13, P<0.01). Methylation analysis of the TSC1 promoter region using direct sequencing of sodium bisulfite modified DNA showed no evidence for methylation at any of the investigated 22 CpG sites in 11 PXAs, including the nine tumors with reduced TSC1 transcript levels.

Figure 4

(a) Demonstration of lower TSC1 transcript levels in PXA47 as compared to non-neoplastic brain tissue (NB) and an anaplastic astrocytoma (AA143). Shown are results obtained by real-time reverse transcription–PCR. X axis, cycle number; Y axis, relative amount of PCR product (RFU, relative fluorescence units). Note that the curves for PXA47, AA143 and NB are nearly identical for the reference transcript ADP-ribosylation factor 1 (ARF1). In contrast, the TSC1 curve of PXA47 crosses the threshold value (t) on the right side of the TSC1 curves of AA143 and NB, indicating lower expression of TSC1 in PXA47. The calculated TSC1 mRNA level in PXA47 was 12% of the level in NB. (b) Mean expression levels of TSC1 transcripts in nine PXAs, eight diffuse astrocytomas WHO grade II (AII), eight anaplastic astrocytomas WHO grade III (AAIII), and nine giant cell glioblastomas WHO grade IV (GC-GBIV). The TSC1 mRNA level was normalized to the ARF1 mRNA level in each tumor and calculated in relation to the normalized TSC1 mRNA level in non-neoplastic brain tissue. Note that the mean TSC1 transcript level was significantly lower in PXAs as compared to AII, AAIII and GC-GBIV.


In contrast to the common diffuse astrocytic gliomas, relatively little information is available about the genetic changes associated with PXAs, which constitute a rare type of astrocytic glioma in children and young adults. In a previous study, we reported that the genetic alterations commonly found in diffuse astrocytomas are rare or absent in PXAs (Kaulich et al., 2002). In particular, only 5% of PXAs carried a TP53 mutation and the CDK4, MDM2 and EGFR proto-oncogenes showed no amplification in any of 62 PXAs investigated (Kaulich et al., 2002). However, neither our previous data nor studies from other groups (Paulus et al., 1996) indicated any particular genetic alteration as being typically associated with PXAs. Therefore, we used genome-wide screening methods, that is, chromosomal and array-based CGH, to perform a comprehensive genomic profiling of PXAs.

Chromosomal CGH analysis of 50 PXAs revealed chromosome 9 loss as the hallmark alteration identified in 50% of the cases. All other chromosomal imbalances were restricted to less than 20% of the tumors. The pattern of genomic imbalances detected in PXAs was markedly different from that identified in diffuse astrocytomas (this study; Weber et al., 1996). In line with the generally distinct chromosomal aberration patterns, the amplicons identified in two cases each mapped to different chromosomal arms, that is, 2p, 4p, 12q and 21q in PXAs as compared to 12p and 13q in diffuse astrocytomas.

Our array-CGH experiments extend the results obtained by chromosomal CGH analysis in two ways. Firstly, imbalances of small chromosomal regions could be additionally identified and fine mapped by array-CGH. Secondly, imbalances of entire chromosomes could be detected by array-CGH that did not reach the diagnostic threshold in the chromosomal CGH analysis. When verifying array-CGH data by interphase-FISH to tumor sections, the alterations were confirmed but only detected in 30–80% of cells on the tissue section. This can be explained either by the fact that the aberrations were restricted to a subpopulation of tumor cells and/or by an admixture of non-neoplastic cells, for example, lymphocytic infiltrates, microglial cells and reactive astrocytes. Taken together, these results indicate that in addition to increasing the resolution of detectable imbalances, array-CGH allows the detection of copy number changes in cell subpopulations. This assumption is supported by a study on chronic lymphocytic leukemia, which noted that genomic imbalances present in less than 25% of cells, as ascertained by interphase-FISH, could be scored correctly by array-CGH (Schwaenen et al., 2004).

Our chromosomal CGH data suggested that the short arm of chromosome 9 likely harbors a tumor-suppressor gene of importance in the pathogenesis of PXAs. Array-CGH then allowed the identification of a critical region at 9p21.3 by showing pronounced losses of clones from this chromosomal subband in seven of 10 PXAs investigated. In each case, these circumscribed losses included the RP11-149I2 clone, which contains the CDKN2A/p14ARF and CDKN2B loci. When visualized at the cellular level by interphase-FISH to tumor sections, homozygous deletions encompassing the CDKN2A/p14ARF and CDKN2B loci were detected in 30–80% of the cells present on the respective tumor sections. In contrast to these findings, previous analyses of the same 10 tumors using duplex-PCR assays did not reveal any PXA with clear evidence of a homozygous CDKN2A, CDKN2B and p14ARF deletion, that is, none of the cases showed a CDKN2A, CDKN2B or p14ARF gene dosage of less than 30% relative to constitutional DNA (Kaulich et al., 2002). The reduced gene dosages found in six PXAs were judged as being compatible with hemizygous deletion (Kaulich et al., 2002). Thus, duplex-PCR may miss homozygous deletions restricted to a subpopulation of tumor cells or in cases with significant non-neoplastic cell contamination. Taken together, our new data clearly indicate a role for homozygous deletion of the CDKN2A, CDKN2B and p14ARF tumor-suppressor genes in PXAs. This finding is of interest since so far homozygous deletions of these genes were detected predominantly in anaplastic gliomas and glioblastomas (Reifenberger and Collins, 2004). The respective gene products are crucial regulators of the p53 and pRb1 tumor-suppressor pathways, with the CDKN2A and CDKN2B gene products p16INK4a and p15INK4b functioning as regulators of G1/S-phase transition by inhibiting the activity of cyclin-dependent kinases Cdk4 and Cdk6 (Pei and Xiong, 2005). The p14ARF protein inhibits Mdm2-mediated degradation of p53 (Harris and Levine, 2005). Thus, our findings suggest that a significant percentage of PXAs contain tumor cells with impaired function of the p53 and pRb1 pathways due to biallelic deletions of the CDKN2A, p14ARF and CDKN2B genes.

The findings that most chromosome 9 deletions involved both chromosomal arms and that one deletion was restricted to 9q suggest that another PXA-associated tumor-suppressor gene is located on 9q. To address this issue, we took a candidate gene approach by performing molecular analyses of two known tumor-suppressor genes located on 9q, that is, the PTCH and TSC1 genes. Germline mutations in the PTCH gene at 9q22.3 cause nevoid basal cell carcinoma syndrome (Gorlin syndrome) (Hahn et al., 1996), a rare autosomal dominant disorder that predisposes individuals to cutaneous basal cell carcinomas and other tumor types, including cerebellar medulloblastomas. Mutations in the TSC1 gene (9q34) are found in patients with tuberous sclerosis complex (van Slegtenhorst et al., 1997), a dominantly inherited disease characterized by the widespread development of hamartomas in multiple organ systems as well as a predisposition to certain tumor types, in particular subependymal giant cell astrocytomas. Mutational investigation of these two candidate genes in PXAs with 9q deletions did not reveal any tumor-associated sequence alterations. However, while PTCH transcript levels were upregulated relative to non-neoplastic brain tissue in PXAs and other astrocytic gliomas, TSC1 mRNA was consistently downregulated in PXAs relative to non-neoplastic brain tissue. Furthermore, TSC1 transcript levels were significantly lower in PXAs as compared to diffuse and anaplastic astrocytomas as well as glioblastomas. In contrast to recent findings in breast carcinomas (Jiang et al., 2005), we found no evidence for TSC1 promoter hypermethylation in PXAs. Thus, the precise molecular mechanisms causing reduced TSC1 mRNA expression in PXAs remain to be elucidated. Furthermore, it needs to be investigated whether transcriptional downregulation of TSC1 in PXAs is associated with increased activity of the Rheb/mTOR pathway, which is inhibited by the TSC1/2 protein complex (Long et al., 2005). A study on bladder cancer suggested a role of TSC1 haploinsufficiency in tumor pathogenesis (Knowles et al., 2003), which might also be involved in PXAs carrying 9q deletions.

In summary, we performed a comprehensive analysis of genomic alterations in PXAs and identified chromosome 9 loss as the most common chromosomal alteration in these tumors. Furthermore, a significant fraction of PXAs was found to carry tumor cells with 9p21.3 homozygous deletions involving the CDKN2A/p14ARF and CDKN2B tumor-suppressor gene loci. Molecular analysis of candidate genes from 9q showed no evidence for PTCH mutation or loss of expression in PXAs. The TSC1 gene also lacked detectable mutations and promoter hypermethylation in PXAs but was consistently downregulated at the transcript level. Taken together, these data suggest homozygous deletion of CDKN2A/p14ARF and CDKN2B as well as reduced expression of TSC1 as important molecular alterations in PXAs.

Material and methods

Tumor samples

Chromosomal CGH analyses were performed on PXAs from 50 patients (24 female, 26 male; mean age at operation: 20. 3 years; Table 1). All cases were included in the studies of Kaulich et al. (2002) and Reifenberger et al. (2003). The tumor material was collected and analyzed as reported in these papers and as approved by the local institutional review board at the University of Bonn Medical Center (study no. 163/99). All tumors were histologically classified according to the WHO classification of tumors of the nervous system (Kepes et al., 2000). Thirty-six tumors corresponded to classic PXA (WHO grade II) and 14 tumors to PXA with anaplastic features. Formalin-fixed and paraffin-embedded tumor tissue samples were collected from all 50 patients. Unfixed frozen tumor samples were additionally available from 10 of the patients. For comparison of chromosomal imbalances, nine diffuse astrocytomas of WHO grade II (AII) were analyzed by CGH. For comparison of candidate gene expression, tumor material from eight patients with AII, eight patients with anaplastic astrocytoma WHO grade III (AAIII) and nine patients with giant cell glioblastoma WHO grade IV (GC-GBIV) was analyzed. As reference tissue for the mRNA expression studies, we used non-neoplastic brain tissue samples obtained at autopsy from two adult patients.

Nucleic acid extraction

High molecular weight DNA and RNA were extracted from frozen tissue samples of 10 PXAs, 25 other astrocytic gliomas, and two samples of non-neoplastic brain tissue by ultracentrifugation as described elsewhere (van den Boom et al., 2003). In the remaining 40 PXAs, DNA was extracted from formalin-fixed paraffin-embedded tumor specimens (Reifenberger et al., 1996; Kaulich et al., 2002). As reference DNA for CGH analyses, high molecular weight DNA was extracted from peripheral blood lymphocytes of normal individuals according to standard procedures (Sambrook et al., 1989).

Chromosomal CGH

Metaphase spreads from a normal healthy male subject, prepared by standard procedures, were obtained from stimulated peripheral blood lymphocytes. CGH was carried out as described previously (Lichter et al., 1995). Image capture and processing for CGH analysis were performed using the Isis CGH imaging system (MetaSystems, Altlussheim, Germany). The diagnostic thresholds used to score losses, gains and high-level amplifications were 0.75, 1.25 and 2.0, respectively.

Array-based CGH

Ten PXAs for which high molecular weight DNA could be extracted from frozen tumor samples were analyzed by array-CGH using genomic DNA microarrays with more than 8000 large insert clones. Except for the addition of 2000 region specific clones from the RPCI (RZPD, Berlin, Germany) and CalTech (Invitrogen, Karlsruhe, Germany) BAC libraries, the array was published previously (Zielinski et al., 2005). Array assembly, hybridization and analysis were essentially performed as described previously (Zielinski et al., 2005) with the following minor modifications: slides were spotted in triplicate using an Omnigrid microarray spotter (GeneMachines, San Carlos, CA, USA). The reference DNA pool was from 10 healthy donors and sex matched. Preannealing was performed using 120 μg COT Human DNA (Roche, Basel, Switzerland) for 1 h. Hybridization was carried out in a HybArray 12 hybridization station (Perkin-Elmer, Beaconsfield, UK) at 37°C for about 70 h.

Interphase FISH to tumor tissue sections

Formalin-fixed paraffin-embedded tissue sections of 6 μm thickness were deparaffinated, pretreated with 85% formic acid/1% hydrogen peroxide solution followed by 1 M sodium thiocyanate and 1 mg/ml pepsin in 0.02 N. HCl, fixed with 1% formaldehyde and denaturated. DNA isolated from BAC or PAC clones was labeled with digoxigenin-11-dUTP or biotin-16-dUTP (Roche) by nick translation. FISH was carried out as described previously (Lichter et al., 1995).

Mutation analyses

All 50 PXAs had been investigated before for the presence of homozygous deletions of the CDKN2A tumor-suppressor gene at 9p21 using duplex-PCR analysis (Kaulich et al., 2002). In the same study, no point mutations, promoter hypermethylation or complete loss of mRNA expression of CDKN2A/p14ARF and CDKN2B had been detected in 10 PXAs analyzed (Kaulich et al., 2002). Eighteen PXAs, including 16 tumors with loss of 9q as demonstrated by CGH analysis, were screened for mutations in all coding exons of the TSC1 and PTCH tumor-suppressor genes by using SSCP analysis. PCR amplification, SSCP analysis and DNA sequencing were performed as reported (Becker et al., 2001; Reifenberger et al., 2005).

Expression analyses

Nine PXAs and 25 diffuse astrocytic gliomas were studied for the expression of PTCH and TSC1 transcripts using real-time reverse transcription–PCR and the ABI PRISM 5700 sequence detection system (Applied Biosystems, Foster City, CA, USA). Continuous measurement of the PCR product was enabled by incorporation of SYBR-Green fluorescent dye into the double-stranded PCR products. The transcript levels of each gene were normalized to the transcript level of ARF1 (ADP-ribosylation factor 1, NCBI GenBank Accession-No. M36340) and calculated in relation to the mean normalized transcript value obtained for the non-neoplastic reference tissue samples. The respective primer sequences were as follows: TSC1-TaqF: 5′-IndexTermcctcacaacaggcgtcttgg-3′, TSC1-TaqR: 5′-IndexTermtggcatggagatggacgag-3′; PTCH-TaqF: 5′-IndexTermgccagcggctacttactcatg-3′, PTCH-TaqR: 5′-IndexTermgccactgacagtgcaaccag-3′; ARF1-TaqF: 5′-IndexTermgaccacgatcctctacaagc-3′, ARF1-TaqR: 5′-IndexTermtcccacacagtgaagctgatg-3′.

TSC1 promoter methylation analysis

The nine PXAs with reduced TSC1 mRNA expression as well as two additional PXAs were analyzed for TSC1 promoter hypermethylation using direct sequencing of sodium bisulfite-modified genomic DNA as reported (Möllemann et al., 2005). PCR was performed for 40 cycles with oligonucleotide primers (sense: 5′-IndexTermgttgcgaagagattttgtgattt-3′ and antisense: 5′-IndexTermcaaaacgctccaaccacacccaaa-3′) spanning a fragment of 175 base pairs within the TSC1 promoter region (nucleotides 188701–188877, GenBank accession no. AL445645). This sequence includes 22 CpG sites located 5′ of the transcription start site. Purified PCR products were sequenced using the BigDye Cycle Sequencing Kit and an ABI PRISM 377 semiautomated DNA sequencer (Applied Biosystems, Foster City, CA, USA).

Statistical analyses

Comparisons between groups were performed by Fisher's Exact Test or Student's t-test. A P-value of less than 0.05 was considered statistically significant.

Accession codes




  1. Becker AJ, Lobach M, Klein H, Normann S, Nothen MM, von Deimling A et al. (2001). Mutational analysis of TSC1 and TSC2 genes in gangliogliomas. Neuropathol Appl Neurobiol 27: 105–114.

    CAS  Article  Google Scholar 

  2. du Manoir S, Speicher MR, Joos S, Schrock E, Popp S, Dohner H et al. (1993). Detection of complete and partial chromosome gains and losses by comparative genomic in situ hybridization. Hum Genet 90: 590–610.

    CAS  Article  Google Scholar 

  3. Giannini C, Hebrink D, Scheithauer BW, Dei Tos AP, James CD . (2001). Analysis of p53 mutation and expression in pleomorphic xanthoastrocytoma. Neurogenetics 3: 159–162.

    CAS  Article  Google Scholar 

  4. Giannini C, Scheithauer BW, Burger PC, Brat DJ, Wollan PC, Lach B et al. (1999). Pleomorphic xanthoastrocytoma: what do we really know about it? Cancer 85: 2033–2045.

    CAS  Article  Google Scholar 

  5. Hahn H, Wicking C, Zaphiropoulous PG, Gailani MR, Shanley S, Chidambaram A et al. (1996). Mutations of the human homolog of Drosophila patched in the nevoid basal cell carcinoma syndrome. Cell 85: 841–851.

    CAS  Article  Google Scholar 

  6. Harris SL, Levine AJ . (2005). The p53 pathway: positive and negative feedback loops. Oncogene 24: 2899–2908.

    CAS  Article  Google Scholar 

  7. Jiang WG, Sampson J, Martin TA, Lee-Jones L, Watkins G, Douglas-Jones A et al. (2005). Tuberin and hamartin are aberrantly expressed and linked to clinical outcome in human breast cancer: the role of promoter methylation of TSC genes. Eur J Cancer 41: 1628–1636.

    CAS  Article  Google Scholar 

  8. Kallioniemi A, Kallioniemi OP, Sudar D, Rutovitz D, Gray JW, Waldman F et al. (1992). Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors. Science 258: 818–821.

    CAS  Article  Google Scholar 

  9. Kaulich K, Blaschke B, Numann A, von Deimling A, Wiestler OD, Weber RG et al. (2002). Genetic alterations commonly found in diffusely infiltrating cerebral gliomas are rare or absent in pleomorphic xanthoastrocytomas. J Neuropathol Exp Neurol 61: 1092–1099.

    CAS  Article  Google Scholar 

  10. Kepes JJ, Louis DN, Giannini C, Paulus W . (2000). Pathology and genetics of tumours of the nervous system. In: Kleihues P, Cavenee WK (eds). World Health Organization Classification of Tumours. IARC Press: Lyon, pp. 52–54.

    Google Scholar 

  11. Knowles MA, Habuchi T, Kennedy W, Cuthbert-Heavens D . (2003). Mutation spectrum of the 9q34 tuberous sclerosis gene TSC1 in transitional cell carcinoma of the bladder. Cancer Res 63: 7652–7656.

    CAS  PubMed  Google Scholar 

  12. Li YS, Ramsay DA, Fan YS, Armstrong RF, Del Maestro RF . (1995). Cytogenetic evidence that a tumor suppressor gene in the long arm of chromosome 1 contributes to glioma growth. Cancer Genet Cytogenet 84: 46–50.

    CAS  Article  Google Scholar 

  13. Lichter P, Bentz M, Joos S . (1995). Detection of chromosomal aberrations by means of molecular cytogenetics: painting of chromosomes and chromosomal subregions and comparative genomic hybridization. Methods Enzymol 254: 334–359.

    CAS  Article  Google Scholar 

  14. Long X, Lin Y, Ortiz-Vega S, Yonezawa K, Avruch J . (2005). Rheb binds and regulates the mTOR kinase. Curr Biol 15: 702–713.

    CAS  Article  Google Scholar 

  15. Möllemann M, Wolter M, Felsberg J, Collins VP, Reifenberger G . (2005). Frequent promoter hypermethylation and low expression of the MGMT gene in oligodendroglial tumors. Int J Cancer 113: 379–385.

    Article  Google Scholar 

  16. Paulus W, Lisle DK, Tonn JC, Wolf HK, Roggendorf W, Reeves SA et al. (1996). Molecular genetic alterations in pleomorphic xanthoastrocytoma. Acta Neuropathol (Berlin) 91: 293–297.

    CAS  Article  Google Scholar 

  17. Pei XH, Xiong Y . (2005). Biochemical and cellular mechanisms of mammalian CDK inhibitors: a few unresolved issues. Oncogene 24: 2787–2795.

    CAS  Article  Google Scholar 

  18. Pinkel D, Segraves R, Sudar D, Clark S, Poole I, Kowbel D et al. (1998). High resolution analysis of DNA copy number variation using comparative genomic hybridization to microarrays. Nat Genet 20: 207–211.

    CAS  Article  Google Scholar 

  19. Reifenberger G, Collins VP . (2004). Pathology and molecular genetics of astrocytic gliomas. J Mol Med 82: 656–670.

    CAS  Article  Google Scholar 

  20. Reifenberger G, Kaulich K, Wiestler OD, Blumcke I . (2003). Expression of the CD34 antigen in pleomorphic xanthoastrocytomas. Acta Neuropathol (Berlin) 105: 358–364.

    CAS  Google Scholar 

  21. Reifenberger J, Ring GU, Gies U, Cobbers L, Oberstrass J, An HX et al. (1996). Analysis of p53 mutation and epidermal growth factor receptor amplification in recurrent gliomas with malignant progression. J Neuropathol Exp Neurol 55: 822–831.

    CAS  Article  Google Scholar 

  22. Reifenberger J, Wolter M, Knobbe CB, Kohler B, Schonicke A, Scharwachter C et al. (2005). Somatic mutations in the PTCH, SMOH, SUFUH and TP53 genes in sporadic basal cell carcinomas. Br J Dermatol 152: 43–51.

    CAS  Article  Google Scholar 

  23. Sambrook J, Fritsch EF, Maniatis T (eds) (1989). Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press: Cold Spring Harbor, NY.

    Google Scholar 

  24. Sawyer JR, Roloson GJ, Chadduck WM, Boop FA . (1991). Cytogenetic findings in a pleomorphic xanthoastrocytoma. Cancer Genet Cytogenet 55: 225–230.

    CAS  Article  Google Scholar 

  25. Sawyer JR, Thomas EL, Roloson GJ, Chadduck WM, Boop FA . (1992). Telomeric associations evolving to ring chromosomes in a recurrent pleomorphic xanthoastrocytoma. Cancer Genet Cytogenet 60: 152–157.

    CAS  Article  Google Scholar 

  26. Schwaenen C, Nessling M, Wessendorf S, Salvi T, Wrobel G, Radlwimmer B et al. (2004). Automated array-based genomic profiling in chronic lymphocytic leukemia: development of a clinical tool and discovery of recurrent genomic alterations. Proc Natl Acad Sci USA 101: 1039–1044.

    CAS  Article  Google Scholar 

  27. Solinas-Toldo S, Lampel S, Stilgenbauer S, Nickolenko J, Benner A, Dohner H et al. (1997). Matrix-based comparative genomic hybridization: biochips to screen for genomic imbalances. Gene Chromosome Canc 20: 399–407.

    CAS  Article  Google Scholar 

  28. van den Boom J, Wolter M, Kuick R, Misek DE, Youkilis AS, Wechsler DS et al. (2003). Characterization of gene expression profiles associated with glioma progression using oligonucleotide-based microarray analysis and real-time reverse transcription-polymerase chain reaction. Am J Pathol 163: 1033–1043.

    CAS  Article  Google Scholar 

  29. van Slegtenhorst M, de Hoogt R, Hermans C, Nellist M, Janssen B, Verhoef S et al. (1997). Identification of the tuberous sclerosis gene TSC1 on chromosome 9q34. Science 277: 805–808.

    CAS  Article  Google Scholar 

  30. Weber RG, Sabel M, Reifenberger J, Sommer C, Oberstrass J, Reifenberger G et al. (1996). Characterization of genomic alterations associated with glioma progression by comparative genomic hybridization. Oncogene 13: 983–994.

    CAS  PubMed  Google Scholar 

  31. Yin XL, Hui AB, Liong EC, Ding M, Chang AR, Ng HK . (2002). Genetic imbalances in pleomorphic xanthoastrocytoma detected by comparative genomic hybridization and literature review. Cancer Genet Cytogenet 132: 14–19.

    CAS  Article  Google Scholar 

  32. Zielinski B, Gratias S, Toedt G, Mendrzyk F, Stange DE, Radlwimmer B et al. (2005). Detection of chromosomal imbalances in retinoblastoma by matrix-based comparative genomic hybridization. Gene Chromosome Canc 43: 294–301.

    CAS  Article  Google Scholar 

Download references


This work was supported by the Deutsche Krebshilfe (10-1639-Re3; 70-3163-Wi3), the German Ministry for Education and Research (National Network for Genome Research, NGFN-2) and the BONFOR program of the Medical Faculty, Rheinische Friedrich-Wilhelms-University, Bonn (O-149.0058).

Author information



Corresponding author

Correspondence to R G Weber.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Weber, R., Hoischen, A., Ehrler, M. et al. Frequent loss of chromosome 9, homozygous CDKN2A/p14ARF/CDKN2B deletion and low TSC1 mRNA expression in pleomorphic xanthoastrocytomas. Oncogene 26, 1088–1097 (2007).

Download citation


  • pleomorphic xanthoastrocytomas
  • comparative genomic hybridization
  • molecular genetics
  • tumor-suppressor gene
  • DNA-microarray

Further reading


Quick links