Association of loss of heterozygosity with shorter survival in primary glioblastoma patients.

Loss of heterozygosity (LOH) co-deletion 1p/19q, MGMT promoter methylation and/or IDH1 mutation generally signify a better prognosis for patients with glioma. However, the influence of 1p/19q co-deletion and the LOH on other chromosomes in primary glioblastoma on survival is still debatable. The aim of our study was to identify LOH on chromosomes 1p, 19q, 9p, 10q, 13q, and 17p, and evaluate their impact either alone or 1p/19q co-deletion or by groups of LOH on the overall survival of 42 primary glioblastoma patients without an oligodendroglial component. These patients were additionally molecularly characterized for EGFR amplification, IDH1 mutations and TP53 mutations. We assessed their influence on the overall survival of glioblastoma patients. LOH in at least one of the loci on all examined chromosomes was detected in 65% of cases and was significantly associated with shorter overall survival (hazard ratio 3.07; 95% CI: 1.29-7.31, p = 0.006). 1p/19q co-deletion was infrequent (7.14%) and had no impact on overall survival. Our results indicate that in primary glioblastoma a specific LOH group analysis may be important for the prognosis. LOH 1p/19q co-deletion is rare in glioblastoma without an oligodendroglial component and has no impact on patient survival.

Glioma cells showing IDH1 mutation cannot be propagated in standard cell culture conditions.

BACKGROUND: It has recently been reported by several sources that original (i.e., present in vivo) glioma cell phenotypes or genotypes cannot be maintained in vitro. For example, glioblastoma cell lines presenting EGFR amplification cannot be established. METHODS AND RESULTS: IDH1 sequencing and loss of heterozygosity analysis was performed for 15 surgery samples of astrocytoma and early and late passages of cells derived from those and for 11 archival samples. We were not able to culture tumour cells presenting IDH1 mutations originating from currently proceeded 10 tumours; the same results were observed in 7 samples of archival material. CONCLUSION: The IDH1 mutation is expected to be almost mutually exclusive with EGFR amplification, so glioma cells with IDH1 mutations seem to represent a new group of tumour cells, which cannot be readily analysed in vitro because of their elimination. The reasons for this intriguing phenomenon should be investigated since its understanding can help to define a new therapeutic approach based on simulating in vivo conditions, responsible for tumour cells elimination in vitro. Moreover, a new model for culturing glioma cells in vitro should be designed since the current one does not provide conditions corresponding to in vivo growth.

BCR expression is decreased in meningiomas showing loss of heterozygosity of 22q within a new minimal deletion region.

Neurofibromin 2 (NF2), located on chromosome arm 22q, has been established as a tumor suppressor gene involved in meningioma pathogenesis. In our study, we investigated 149 meningiomas to determine whether there are additional tumor suppressor genes localized on chromosome 22q, apart from NF2, that might be involved in meningioma pathogenesis. The LOH analysis on chromosome 22q identified two regions of deletion: the first one, which is limited to the NF2 gene locus, and the second one, which is outside this location. The new minimal deletion region (MDR) included the following genes: BCR (breakpoint cluster region), RAB36 (a member of RAS oncogene family), GNAZ [guanine nucleotide binding protein (G protein), alpha-z polypeptide], and RTDR1 (rhabdoid tumor deletion region gene 1). The expression levels of all these genes, including NF2, were subsequently analyzed by quantitative real-time polymerase chain reaction. We observed a significantly lowered expression level of NF2 in meningiomas with 22q loss of heterozygosity (LOH) within NF2 region compared to the one in meningiomas with 22q retention of heterozygosity (ROH, P<0.05). Similarly, BCR showed a significantly lowered expression in meningiomas with 22q LOH within the new MDR compared to cases with 22q ROH (P<0.05). Our data, together with the already published information considering BCR function suggest that BCR can be considered as a candidate tumor suppressor gene localized on chromosome 22q which may be involved in meningioma pathogenesis.

Elimination of wild-type P53 mRNA in glioblastomas showing heterozygous mutations of P53.

We screened 50 glioblastomas for P53 mutations. Five glioblastomas showed heterozygous mutations, while three were putatively heterozygous. Six of these eight glioblastomas showed elimination of wild-type P53 mRNA. These results strongly suggest that some sort of mechanism(s) favouring mutated over wild-type P53 mRNA exists in glioblastoma cells with heterozygous mutations of this gene.

The infrequent simultaneous genetic alterations in glioblastoma multiforme (LOH 10, 17, 19q, TP53 mutation and EGFR amplification) with short clinical course.

We described the case of an unusual, complex genetic alteration in 57 year-old male patient with glioblastoma multiforme (GBM) with short survival (6 and half months). Alterations consisted of p53 mutation, LOH 10, LOH 17, LOH 19q and EGFR amplification. LOH1p, LOH 9 and LOH 13 were negative. Immunohistochemical study did not correlate with molecular results. The overexpression of TP53 protein and RB protein was detected only in small percentage of cells and interestingly the overexpression of EGFR was present only focally. Immnunostainings for PTEN, P16, PI3-K were negative. Additionally, we observed an overexpression of IGFB2 protein. This case indicates the accumulation of molecular changes in glioblastoma multiforme in patient with short survival.

GADD45A and EPB41 as tumor suppressor genes in meningioma pathogenesis.

Deletions of 1p occur in approximately 30% of meningiomas. Based on loss of heterozygosity (LOH) analysis, two regions on 1p have been suspected to be carriers of tumor suppressor genes. We chose the GADD45A and EPB41 genes as tumor suppressor candidates based on their function and chromosomal localization. We analyzed 19 cases of meningioma with LOH of 1p by means of sequencing of the GADD45A gene and Western blotting of the GADD45a protein. Twenty cases of meningioma without 1p LOH were also analyzed by Western blotting to find out if changes of the GADD45a protein expression occurred. Nineteen samples with 1p LOH (12 grade I; 7 grade II, WHO classification) and 20 samples without 1p LOH (18 grade I; 2 grade II) were also analyzed by means of real-time polymerase chain reaction to find abnormalities in EPB41 mRNA levels in meningioma. LOH analysis was performed using seven microsatellite markers: D1S508 (1p36.2), D1S199 (1p36.1) D1S2734 (1p36.1), D1S2720 (1p34), D1S197 (1p32), D1S162 (1p32), D1S429 (1p11). LOH analysis confirmed previously described localization of putative tumor suppressor genes on 1p and involvement in meningioma pathogenesis (1p36 and 1p32). The open reading frame of GADD45A and intron splicing sites showed neither mutations nor polymorphisms. GADD45a protein molecular weight and expression level were unaltered in meningiomas with and without 1p LOH. We conclude that the GADD45A gene is not involved in meningioma tumorigenesis. EPB41 gene expression was unchanged in all analyzed meningiomas. This suggests that involvement of the EPB41 gene (4.1R protein) in meningioma pathogenesis should be reconsidered.

Abnormalities of the P53, MDM2, BCL2 and BAX genes in acute leukemias.

Abnormalities of the P53 network have been implicated in the pathogenesis of acute lymphoblastic leukemia (ALL) and acute myeloblastic leukemia (AML). The purpose of this study was to define P53 gene mutations, to detect MDM2 gene amplification and to estimate mRNA expression of P53, MDM2, BCL2 and BAX genes in patients with ALL and AML. Twenty-five patients with ALL and 65 patients with AML, both recently diagnosed, were included into this study. Exons 5-8 of the P53 gene with flanking intronic sequence were amplified by the polymerase chain reaction (PCR) method and subjected to mutation screening by single-strand conformation polymorphism analysis (SSCP). Mutation of the P53 gene was found in one patient of the 25 with ALL and in five patients of the 65 with AML. Sequence analysis was subsequently performed. One mutation in intronic sequence in ALL and four missense mutations and one silent nucleotide substitution in AML were identified. Amplification of MDM2 gene was detected by multiplex-PCR analysis in only one sample from patient with ALL, but was not observed in any case of AML. To gain further insight into the role of P53 network in the evolution of acute leukemias, the P53, MDM2, BCL2 and BAX mRNAexpressions in portion samples from patients with ALL and AML were analyzed using multiplex RT-PCR. Although a low frequency of molecular disturbances of the P53 and the MDM2 genes was detected in this study, there was a high percentage of cases with increased mRNA level of P53 and MDM2. A high frequency of BCL2 mRNA overexpression and a relatively low frequency of BAX mRNA overexpression detected in both analyzed leukemias in this study, indicate that altered transcription of these genes may be involved in leukemogenesis.

Microsatellite instability and expression of DNA mismatch repair genes in malignant astrocytic tumors from adult and pediatric patients.

Microsatellite instability (MSI) is used as a molecular marker for defective DNA mismatch repair (MMR) genes. We report here alterations of MSI in 15 malignant astrocytomas (WHO grade III) and glioblastomas (GBM; WHO grade IV) of pediatric patients (2 - 21 years) and 12 GBM from adults (44 - 68 years) by comparative analysis of BAT25/BAT26 loci and 10 other microsatellite markers. High-level microsatellite instability (MSI-H) occurred in 4 of the 15 pediatric cases (26.7%) and in 1 of the 12 adult GBM cases (8.3%). Low-level microsatellite instability (MSI-L) was observed in 6 pediatric cases (40%) and 8 adult GBM (66.7%). Unstable BAT-25 locus was found in 1 of the MSI-H pediatric cases. Thus, 2 unstable cases showed no instability of this marker. For BAT-26, such a discordance was even more profound: in 1 of MSI-H cases, we obtained no PCR product and the remaining 3 showed no alterations of this marker. MSH2 (Human MutS, Homologue2) protein was detected in all but 3 pediatric cases (1 highly unstable and 2 low-level unstable) and in all adult cases. MLH1 (Human MutL, Homologue 1) protein was detected in all but 2 pediatric cases (1 highly unstable and 1 low-level unstable). Thus, 2 highly unstable pediatric cases showed no detectable MLH1/MSH2 proteins. Our data support earlier observations that MSI occurs predominantly in malignant astrocytic tumors of young patients, which lends support to the hypothesis of different molecular mechanisms of pediatric brain tumors. Surprisingly, we found no significant correlation between the status of 10 microsatellite markers and that of either BAT25 or BAT26 loci or with the expression of MMR genes.