[HTML][HTML] Relationships linking amplification level to gene over-expression in gliomas

N Vogt, A Gibaud, A Almeida, I Ourliac-Garnier… - PloS one, 2010 - journals.plos.org
N Vogt, A Gibaud, A Almeida, I Ourliac-Garnier, M Debatisse, B Malfoy
PloS one, 2010journals.plos.org
Background Gene amplification is thought to promote over-expression of genes favouring
tumour development. Because amplified regions are usually megabase-long, amplification
often concerns numerous syntenic or non-syntenic genes, among which only a subset is
over-expressed. The rationale for these differences remains poorly understood.
Methodology/Principal Finding To address this question, we used quantitative RT-PCR to
determine the expression level of a series of co-amplified genes in five xenografted and one …
Background
Gene amplification is thought to promote over-expression of genes favouring tumour development. Because amplified regions are usually megabase-long, amplification often concerns numerous syntenic or non-syntenic genes, among which only a subset is over-expressed. The rationale for these differences remains poorly understood.
Methodology/Principal Finding
To address this question, we used quantitative RT-PCR to determine the expression level of a series of co-amplified genes in five xenografted and one fresh human gliomas. These gliomas were chosen because we have previously characterised in detail the genetic content of their amplicons. In all the cases, the amplified sequences lie on extra-chromosomal DNA molecules, as commonly observed in gliomas. We show here that genes transcribed in non-amplified gliomas are over-expressed when amplified, roughly in proportion to their copy number, while non-expressed genes remain inactive. When specific antibodies were available, we also compared protein expression in amplified and non-amplified tumours. We found that protein accumulation barely correlates with the level of mRNA expression in some of these tumours.
Conclusions/Significance
Here we show that the tissue-specific pattern of gene expression is maintained upon amplification in gliomas. Our study relies on a single type of tumour and a limited number of cases. However, it strongly suggests that, even when amplified, genes that are normally silent in a given cell type play no role in tumour progression. The loose relationships between mRNA level and protein accumulation and/or activity indicate that translational or post-translational events play a key role in fine-tuning the final outcome of amplification in gliomas.
PLOS