Mitsutoshi Nakamura

p14ARF deletion and methylation in genetic pathways to glioblastomas.

The CDKN2A locus on chromosome 9p21 contains the p14ARF and p16INK4a genes, and is frequently deleted in human neoplasms, including brain tumors. In this study, we screened 34 primary (de novo) glioblastomas and 16 secondary glioblastomas that had progressed from low‐grade diffuse astrocytomas for alterations of the p14ARF and p16INK4a genes, including homozygous deletion by differential PCR, promoter hypermethylation by methylation‐specific PCR, and protein expression by immunohistochemistry. A total of 29 glioblastomas (58%) had a p14ARF homozygous deletion or methylation, and 17 (34%) showed p16INK4a homozygous deletion or methylation. Thirteen glioblastomas showed both p14ARF and p16INK4a homozygous deletion, while nine showed only a p14ARF deletion. Immunohistochemistry revealed loss of p14ARF expression in the majority of glioblastomas (38/50, 76%), and this correlated with the gene status, i.e. homozygous deletion or promoter hypermethylation. There was no significant difference in the overall frequency of p14ARF and p16INK4a alterations between primary and secondary glioblastomas. The analysis of multiple biopsies from the same patients revealed hypermethylation of p14ARF (5/15 cases) and p16INK4a (1/15 cases) already at the stage of low‐grade diffuse astrocytoma but consistent absence of homozygous deletions. These results suggest that aberrant p14ARF expression due to homozygous deletion or promoter hypermethylation is associated with the evolution of both primary and secondary glioblastomas, and that p14ARF promoter methylation is an early event in subset of astrocytomas that undergo malignant progression to secondary glioblastoma.

Loss of Heterozygosity on Chromosome 10 Is More Extensive in Primary (De Novo) Than in Secondary Glioblastomas

Glioblastomas develop de novo (primary glioblastomas) or through progression from low-grade or anaplastic astrocytoma (secondary glioblastomas). There is increasing evidence that these glioblastoma subtypes develop through different genetic pathways. Primary glioblastomas are characterized by EGFR and MDM2 amplification/overexpression, PTEN mutations, and p16 deletions, whereas secondary glioblastomas frequently contain p53 mutations. Loss of heterozygosity (LOH) on chromosome 10 (LOH#10) is the most frequent genetic alteration in glioblastomas; the involvement of tumor suppressor genes, other than PTEN, has been suggested. We carried out deletion mappings on chromosome 10, using PCR-based microsatellite analysis. LOH#10 was detected at similar frequencies in primary (8/17; 47%) and secondary glioblastomas (7/13; 54%). The majority (88%) of primary glioblastomas with LOH#10 showed LOH at all informative markers, suggesting loss of the entire chromosome 10. In contrast, secondary glioblastomas with LOH#10 showed partial or complete loss of chromosome 10q but no loss of 10p. These results are in accordance with the view that LOH on 10q is a major factor in the evolution of glioblastoma multiform as the common phenotypic end point of both genetic pathways, whereas LOH on 10p is largely restricted to the primary (de novo) glioblastoma.

Promoter hypermethylation and homozygous deletion of the p14ARF and p16INK4a genes in oligodendrogliomas.

Abstract. The INK4a/ARF locus on chromosome 9p21 encodes two gene products that are involved in cell cycle regulation through inhibition of CDK4-mediated RB phosphorylation (p16INK4a) and binding to MDM2 leading to p53 stabilization (p14ARF). The locus is deleted in up to 25% of oligodendrogliomas and 50% of anaplastic oligodendrogliomas, but little is known on the frequency of gene silencing by DNA methylation. We assessed promoter hypermethylation of p14ARF and p16INK4a using methylation-specific PCR, and homozygous deletion of the p14ARF and p16INK4a genes by differential PCR in 29 oligodendrogliomas (WHO grade II) and 20 anaplastic oligodendrogliomas (WHO grade III). Promoter hypermethylation of the p14ARF gene was detected in 6/29 (21%) oligodendrogliomas and 3/20 (15%) anaplastic oligodendrogliomas. None of the oligodendrogliomas and only 1 out of 20 anaplastic oligodendrogliomas showed hypermethylation of p16INK4a. Homozygous deletion was not detected in any of the WHO grade II oligodendrogliomas but was present in 5/20 (25%) anaplastic oligodendrogliomas and always affected both genes. In one tumor containing distinct areas with and without anaplasia, p14ARF hypermethylation was detected in the WHO grade II area, while homozygous co-deletion of p14ARF and p16INK4a was found in the region with anaplastic features (grade III). These data suggest that aberrant p14ARF expression due to hypermethylation is the earliest INK4a/ARF change in the evolution of oligodendrogliomas, while the presence of p14ARF and p16INK4a deletions indicates progression to anaplastic oligodendroglioma.

Reduced expression of the Aα subunit of protein phosphatase 2A in human gliomas in the absence of mutations in the Aα and Aβ subunit genes

Protein phosphatase 2A (PP2A) consists of 3 subunits: the catalytic subunit, C, and the regulatory subunits, A and B. The A and C subunits both exist as 2 isoforms (α and β) and the B subunit as multiple forms subdivided into 3 families, B, B′ and B″. It has been reported that the genes encoding the Aα and Aβ subunits are mutated in various human cancers, suggesting that they may function as tumor suppressors. We investigated whether Aα and Aβ mutations occur in human gliomas. Using single strand conformational polymorphism analysis and DNA sequencing, 58 brain tumors were investigated, including 23 glioblastomas, 19 oligodendrogliomas and 16 anaplastic oligodendrogliomas. Only silent mutations were detected in the Aα gene and no mutations in the Aβ gene. However, in 43% of the tumors, the level of Aα was reduced at least 10‐fold. By comparison, the levels of the Bα and Cα subunits were mostly normal. Our data indicate that these tumors contain very low levels of core and holoenzyme and high amounts of unregulated catalytic C subunit.

Methylation of the p73 gene in gliomas.

Abstract. The p73 gene encodes a protein structurally and functionally homologous to TP53, and maps to chromosomal band 1p36.33, where loss of heterozygosity has been observed in up to 90% of oligodendrogliomas and in 10–25% of diffuse astrocytomas. We assessed the methylation status of the CpG islands in the promoter region of the p73 gene by methylation-specific PCR in 117 glioma biopsies. Methylation was detected in 5 out of 28 (18%) glioblastomas and in 4/26 (15%) anaplastic oligodendrogliomas (WHO grade III) but not in grade II oligodendrogliomas, low-grade diffuse astrocytomas (grade II), and anaplastic astrocytomas (grade III). To assess whether p73 methylation leads to loss of expression, we carried out reverse transcription-PCR and methylation-specific PCR in 10 glioblastoma cell lines. Two lines (U87MG and T98G) showed p73 methylation. U87MG had no unmethylated p73 (complete methylation), and showed loss of expression. T98G had methylated and unmethylated p73 (partial methylation), and retained p73 expression. A third cell line (LN-308) showed loss of p73 expression without p73 methylation. These results suggest that complete p73 methylation is associated with loss of expression, but that additional mechanisms may cause loss of p73 expression. Analysis of a polymorphic site in exon 2 further showed that p73 was mono-allelically expressed in 6 out of 7 primary gliomas with heterozygous GC/TA polymorphism.

APO2L/TRAIL expression in human brain tumors

Abstract APO2 ligand (APO2L)/TRAIL is a novel member of the tumor necrosis factor cytokine family and a potent inducer of apoptosis in tumor cell lines. We recently reported that APO2L is consistently expressed in low-grade astrocytomas, anaplastic astrocytomas, glioblastomas, and cell lines derived thereof, and that malignant glioma cell lines are susceptible to APO2L-induced apoptosis. In this study, we investigated whether APO2L is expressed in medulloblastoma or neuroblastoma cell lines and whether these cells are sensitive to APO2L-induced apoptosis. Immunoblot analyses revealed full-length APO2L protein expression in one (DAOY) of three medulloblastoma cell lines but not in two neuroblastoma cell lines (SKN-BE and SKN-LE). Viability assay performed after exposure to soluble APO2L for 16 h showed that DAOY medulloblastoma cells were the most sensitive and that apoptosis induced by APO2L was greatly enhanced when protein synthesis was inhibited by cycloheximide. Neuroblastoma cell lines were almost completely resistant to APO2L-induced apoptosis. We also carried out APO2L immunohistochemistry in a total of 115 tumors of the nervous system with different histogenesis and biological behavior. In all 9 pilocytic astrocytomas, the areas of dense fibrillary network showed diffuse and strong APO2L expression. In oligodendrogliomas, APO2L expression was observed in areas with a significant admixture of astrocytic cells, but was absent in neoplastic oligodendrocytes. In 13 of 14 ependymomas, APO2L was expressed in perivascular pseudorosettes. In all 12 medulloblastomas, strong APO2L expression was observed in intra-tumoral-reactive astrocytes, but neoplastic cells did not show APO2L immunoreactivity. Thus, the pattern of APO2L expression was largely similar to that of glial fibrillary acidic protein (GFAP), except for choroid plexus tumors and 3 of 8 anaplastic meningiomas, in which APO2L was focally expressed without concomitant GFAP expression. APO2L expression was absent in meningiomas, neurocytomas, and schwannomas. Thus, there is considerable heterogeneity of APO2L expression and susceptibility to APO2L-induced apoptosis among human brain tumors.

Mutation analysis of hBUB1, hBUBR1 and hBUB3 genes in glioblastomas

Abstract. Glioblastomas, the most malignant human brain tumors, are characterized by marked aneuploidy, suggesting chromosomal instability which may be caused by a defective mitotic spindle checkpoint. We screened 22 glioblastomas for mutations in the mitotic spindle checkpoint genes hBUB1, hBUBR1 and hBUB3. DNA sequencing revealed a silent mutation at codon 144 of hBUB1 (CAG→CAA, Gln→Gln) in one glioblastoma, a silent mutation at codon 952 of hBUBR1 (GAC→GAT, Asp→Asp) in another glioblastoma, and a silent mutation at codon 388 of the hBUBR1 gene (GCG→GCA, Ala→Ala) in 8 glioblastomas. We also observed a known polymorphism at hBUBR1 codon 349 (CAA/CGA, Gln/Arg), with an allelic frequency of 0.75 for Gln and 0.25 for Arg, which is similar to that among healthy Caucasian individuals (0.73 vs 0.27). The coding sequence of the hBUB3 gene did not contain any mutation, but in 4 glioblastomas (18%), a C→T point mutation was detected at position –6 (6 nucleotides upstream of the ATG initiator codon). Analysis of blood DNA of these patients showed identical sequence alterations, indicating that this is a polymorphism. Again, the frequency in glioblastomas was similar to that in healthy Caucasians (15%). We further screened hBUB1 in 18 cases of giant cell glioblastoma, a variant characterized by a predominance of bizarre, multinucleated giant cells. There were no changes, except for a silent mutation at codon 144 in two cases. These results suggest that mutations in these mitotic spindle checkpoint genes do not play a significant role in the causation of chromosomal instability in glioblastomas.