Every lab I know does co-transfect an internal control, and a couple that
used not to told me they were told to do so by reviewers when they
submitted their papers. In my experience, even models where the variance
in raw RLUs is small and allows to see nice significant effects do
benefit from an internal control. The other models where a larger
variance seem to be intrinsic also certainly benefit. At the very least,
an internal control will allow you to use results when something did go
less than optimally in your experiment.
When I first started transfecting, as a post-doc, someone did voice to me
the argument about competing away the transcription factors required for
good expression of the test plasmid, especially when using a strongly
expression internal control. I have transfected several cell types over
several years, and have never encountered a problem. This does not mean
that such a competition did not occur (I never tested with and without
the internal control in parallel), but I was always able to see my
reporter and, since the same amount of control plasmid is present, the
competition would be even throughout the experiment and the results
valid. (Anyway, more valid than without an internal control, IMHO.)
I would also encourage all others who have an opinion to voice it.
On 9 Oct 1996, Joe Miano wrote:
> Date: 9 Oct 1996 03:02:56 GMT
> From: Joe Miano <jmiano at post.its.mcw.edu>
> To: methods at net.bio.net> Subject: Re: Best way to express Lucif activity?
>> Thanks for the note. I have always wondered with suspicion about
> contransfections. Does this not introduce squelching and perhaps an
> under-representation of the true promoter activity? I have seen some labs
> not use an internal control plasmid but I think you're right if you say a
> majority of labs do.
>> What I typically do is transfections in triplicate (6 well dishes) and I do
> this at least 3 times. I am finding no significant difference between
> experiments when I express the normalized RLUs (normailzed to protein) as a
> percent of SV40 or as a fold increase over pGL3basic (which bythe way gives
> high background in many cells!). I do not see any profound argument
> against not using an interbnal standard although I can imagine some hostile
> reviewers wanting one to do it the "correct way."
>> Is there really a correct way of expressing the data?
>> I would encourage more people to comment on this issue that I have debated
> for some time know. I really do not know the answer.
> The Love you take is equal to the Love you make--P. McCartney
Dr. Louis H. Ferland
Centre de Recherche, Hotel-Dieu de Montreal
Dept de Nutrition, Universite de Montreal
Phone: (514) 843-2757 FAX: (514) 843-2719