Strategy for finding RT-PCR housekeeping gene

Chang Zhu czhu at changbioscience.com
Wed May 19 23:51:56 EST 2004


Or maybe housekeeping genes are not necessarily stable genes. See
Lee PD, Sladek R, Greenwood CM, Hudson TJ., Control genes and
variability: absence of ubiquitous reference transcripts in diverse
mammalian expression studies. Genome Res. 2002 Feb;12(2):292-7.

I've searched through thousands of microarray experiments, the
commonly used standard such as actin frequently has greater than 2
fold variations across samples. So if your treatment effects are
modest two-fold changes, it will be very difficult to demonstrate that
at the RNA level. Will it be possible for you to measure at protein or
activity level. I guess a long-lived protein might be a better
standard.

Chang


Tom Anderson <ucgatan at ucl.ac.uk> wrote in message news:<Pine.GSO.4.58.0404271624360.4177 at socrates-a.ucl.ac.uk>...
> On Mon, 26 Apr 2004, Paul K. Farmer wrote:
> 
> > We're performing RT-PCR on RNA extracted from cultured cells treated
> > with several potent reagents (alone and in combination with one
> > another). Our gene-specific primers work great (good specificity,
> > efficiency, reproducibility, etc).
> 
> Okay.
> 
> > Unfortunately, our treatment effects are modest two-fold changes.
> 
> That's one-fold good enough!
> 
> > Furthermore, we have not been able to identify any primers corresponding
> > to a "housekeeping gene" that provides constant mRNA levels in response
> > to cell treatments. We've tried several primer pairs used by others in
> > different systems, but the abundance of these transcripts have also
> > varied with cell treatments.
> 
> Perhaps your treatments are having some effect on overall transcription
> levels, eg killing the cells or making them sick. Have you tried the
> classic housekeeping genes, like GAPDH, actin and tubulin? If you're
> seeing similar variation in all of those, then this would seem likely.
> 
> In that case, you might simply use the levels of the housekeeping genes to
> normalise the levels of the other genes. However, you do have to worry
> about the nonspecificity of the treatments.
> 
> HTH.
> 
> tom



More information about the Methods mailing list