Molecular genetic analysis of medulloblastomas(Prof. V. Peter Collins)
3 year project with a grant of £138,150.00
2001 - 2004
Why medulloblastomas develop in children is unknown. The current ways of treating medulloblastomas are long from optimal. Where therapy is successful in eradicating the tumour the developing brain is frequently damaged. This becomes evident when treated children reach school age. We therefore need to understand how the tumours arise so that we can target the tumour cells specifically and avoid damaging the normal brain. Hopefully we will identify tumour-specific cellular mechanisms that are vulnerable to innovative molecular treatment forms. This project utilizes the latest molecular genetic techniques to identify abnormalities of the genetic material in medulloblastoma tumour cells.
Our analysis of the expression of the EGF receptor family (4 members) and its ligands (15 in total) in 31 medulloblastomas was completed in the spring of 2002. The data showed wide variations in expression levels of these 19 genes. Key findings were confirmed by repeated experiments during mid and late 2003. No clear pattern emerged (see Figures 1 & 2 below). However, it is clear that both receptors and a number of ligands are expressed by the tumour cells indicating the presence of autocrine growth stimulatory loops. In late 2003 we carried out a preliminary survival analysis of the 31 patients and attempted to correlate survival to the expression pattern of the EGF receptor family and their ligands. The preliminary data do not identify any correlation between high expression of erbB4 (and erbB2) with poor patient outcome. This contradicts the only report in the literature on this matter.
We previously reported that we had identified a hitherto unrecognised splice variant of the erbB4 gene in medulloblastomas and cloned this. We have now identified yet another splice variant from a medulloblastoma, bringing the total number of transcripts from the erbB4 gene to 10. The function of these splice variants is completely unknown and all are expressed in medulloblastomas. The data shown in Figures 1 & 2 on the expression only shows the total transcript expression from each gene. We do not know as yet which individual transcripts are expressed. It is expected that many genes will express multiple transcripts increasing the complexity of the protein products and their functions. The presence or absence of expression of these splice variants is being documented in normal tissues and a spectrum of brain tumours. In addition we have now cloned all the 10 splice variants of erbB4. We plan to transfer these sequences into expression vectors under the CMV promoter during the spring of 2004 and later in the year to initiate functional studies in the CHO cell line which we have shown to lack expression of all erbB receptors with the exception of a very low level of erbB2 expression.
During the last 6 months we have developed our 10MB CGH array and have successfully applied this to our medulloblastoma series. The findings confirm previous studies but also indicate new regions of the genome where consistent abnormalities are found. We are presently producing a 1MB CGH array and expect to have completed it by Feb. 2004 when we will start to use it to in greater detail than ever before analyse the genome of medulloblastomas. This will be carried out during the spring/summer of 2004 with analysis and writing up in the summer/autumn. Tiled arrays give even more detailed information with a resolution of a few hundred kilobases or less. We have completed the collection of data using our tiled arrays for chromosomes 6 and 22. The findings from the chromosome 22 study show monosomy 22 in about 15% of the cases with no small regions of loss identified making the identification of the genes targeted very difficult. However chromosome 6 has proven more interesting with at least one small homozygous deletion and a number of small hemizygous deletions as well as one case that has a number of amplified regions on chromosome 6. These are now being studied using other techniques to define in detail the sequences lost and thereby the targeted genes. Our chromosome 10 tiling path array is now operational and will be used for the medulloblastoma studies during 2004. Extension of these studies to other arrays will be carried out as the new arrays become available.
Click here to go back to full list of completed research projects.
