Extent of over-expression of hepatocyte growth factor receptor in colorectal tumours is dependent on the choice of normaliser.

For reliable results from quantitative RT-PCR, the starting quantity of total RNA and other parameters need to be controlled. Most studies do this by normalising their results to a single reference gene. This study quantified the mRNA expression of three putative reference genes (ubiquitin C, cyclophilin E, and porphobilinogen deaminase) and the target gene hepatocyte growth factor receptor (HGFR) in matched colorectal tumour and normal mucosa samples. Each of the putative reference genes was found to be significantly over-expressed in the tumour samples compared to the normal samples. When HGFR expression was normalised to each of these reference genes using the 2 (-DeltaDeltaC(T)) method of relative quantification, the number of tumour samples in which HGFR was found to be over-expressed varied from 30% to 63% depending on the reference gene chosen for normalisation. This shows that normalising to a single reference gene without prior validation is inappropriate.