Allelic loss on chromosome 10 is a frequent event in high grade gliomas. Earlier studies have shown that in most cases a complete copy of chromosome 10 is lost in the tumor. To define more accurately and specifically the region of common deletion on chromosome arm 10p, we have screened a large series of gliomas for allelic losses that exclusively affect this part of the chromosome. Allelic loss profiles were determined for 127 gliomas, including 118 astrocytomas of various malignancy grades. Seventeen tumors displayed loss of part of chromosome 10. In three of these, only chromosome arm 10p sequences were lost. The interval between loci D10S559 and D10S435 in 10p15, with a length of approximately 800 kilobase pairs, was commonly deleted in the latter tumors, suggesting that this region may harbor a tumor suppressor gene important in glioma tumorigenesis. Comparison of the allelic loss profiles in the low and high grade astrocytomas revealed that astrocytoma progression is associated with increased loss of chromosome 10 sequences.
Ependymomas are glial tumours of the brain and spinal cord. The most frequent genetic change in sporadic ependymoma is monosomy 22, suggesting the presence of an ependymoma tumour suppressor gene on that chromosome. Clustering of ependymomas has been reported to occur in some families. From an earlier study in a family in which four cousins developed an ependymoma, we concluded that an ependymoma-susceptibility gene, which is not the NF2 gene in 22q12, might be located on chromosome 22. To localize that gene, we performed a segregation analysis with chromosome 22 markers in this family. This analysis revealed that the susceptibility gene may be located proximal to marker D22S941 in 22pter–22q11.2. Comparative genomic hybridization showed that monosomy 22 was the sole detectable genetic aberration in the tumour of one of the patients. Loss of heterozygosity studies in that tumour revealed that, in accordance to Knudson’s two-hit theory of tumorigenesis, the lost chromosome 22 originated from the parent presumed to have contributed the wild-type allele of the susceptibility gene. Thus, our segregation and tumour studies collectively indicate that an ependymoma tumour suppressor gene may be present in region 22pter–22q11.2.
We investigated the dynamics of the genetic changes that are associated with two types of glioma recurrence, that is, progression from a lower-grade to a high-grade tumor (7 cases) and development of a same high-grade recurrence (15 cases). Each pair of tumors was analyzed for TP53 mutation, EGFR amplification, and loss of heterozygosity for tumor suppressor genes (TP53, RB1, CDKN2A, PTEN, DMBT1) and tumor suppressor gene regions (1p36, 19q13, 11p15, 10p15) known to be frequently implicated in glioma tumorigenesis. By comparing the genetic changes in the primary and corresponding secondary tumors, we found that additional loss of CDKN2A and/or RB1, encoding important components of the cell cycle regulatory pathway, was the most frequent genetic change in both types of recurrence development (10 of 22 cases, 45%). Additional loss of heterozygosity for the 10p15 region, for PTEN, and/or for DMBT1 in the recurrent tumor was noted in 7 of 22 cases (32%), suggesting that additional inactivation of tumor suppressor genes on chromosome 10 is another important feature of glioma relapse. Less frequent additional losses were detected for chromosome regions 11p15 and 19q13 (3 of 22 cases, 14%, each). We conclude that glioma recurrences are characterized by an increased involvement of tumor suppressor genes, even in those cases in which the primary and secondary tumor are of the same high malignancy grade.
