CAT assays-reproducability

Tue Jul 6 06:14:00 EST 1993

  Sorry if this is repeating myself, but I sent this message out over a
month ago and got no reply-I may have screwed up at this end, so I am trying
again! Again sorry if everyone has had this already. Also sorry if this is
too long-is it convention to keep messages short?
>         Are there any old fashioned CAT assayers out there?
>     I am engaged on a project in which I am attempting to analyse various
>putative enhancer elements for response to oncogenes (eg src, int, hst etc).
>We are currently using a reporter system based on the HSVtk promoter ie 
>pBLCAT2. The big problem has been reproducability of the response. I seem to be
>able to get everthing from 30-fold induction down to nothing from experiment
>to experiment. Is this level of variation unusual? If so, any ideas on
>minimising it? 
>  Another problem is that the negative control ie pBLCAT2 containing no added
>enhancer, appears to respond up to 10x to oncogene transactivation, minimising
>any putative enhancer effects seen on the elements under test. 
>  Several other groups use this reporter system with no such apparent problems.
>  We have tried controlling many perceived sources of variation with little
>lasting success, I can go into more detail with any respondents.
>Graham Atherton
Hi Graham:
        I do not know exactly what you are doing.  But I think if you use
an internal control such as a GUS fusion, you may get what you want.
        Good luck.
Jim Hu
Thanks Jim,
    I assume that you mean the variation between experiments in my work 
is due to fluctuations in DNA uptake. The difficulty in introducing an
internal control is that the internal control promoter can respond to 
the cotransfected oncogene in my experiments, making within experiment
variation difficult to control but presumably not between experiment
variation, as long as I am comparing like with like. Do you know if
any GUS fusion constructs exist which may be insensitive to oncogenes src
 int, hst?
I did not suspect uptake to be the problem as for no apparent difference in
basal levels of expression from the reporter(ie in the absence of the 
transactivating oncogene) does not change noticably in an experiment which
shows a large response compared with one that shows little response. If
DNA uptake were to blame, the former would be higher than the latter?

Throwing out a wider question, where does the variation in DNA uptake as
assessed by transient transfection of reporter constructs come from? My
guess is changes in the growth state of the recipient cells, or differences
in the composition of the CaPO4 ppt caused by temperature variation between
experiments as the ppt is formed?.


`Fraid I'm going to have to be tiresome and ask YOU for detail.  What
method of DNA introduction are you using?  It makes a big difference if
you're using lipofectin, calcium or electroporation.  Lipofectin is
supposed to be the most reproducible.  I know myself that 3 to 4 fold
variation with electroporation is not uncommon.  Calcium phosphate can go
sour very quickly if your reagents are compromised.  

You say that you're using a reporter with a promoter for a negative
control.  Why?  If you want something that will just show background
without any response you need something like pCATbasic or pSV0CAT.  If
you're just comparing the response to one enhancer to one oncogene to the
response of the same plasmid without the enhancer then what's the problem?

E-mail: mhollowa at (mail to freenet is forwarded)
phone:  (516)444-3612

Thanks Mike
           I am currently using CaPO4-mediated transfection, I keep all of the
reagents at -20 degrees centigrade until thawed for use. Exactly which of the
reagents goes sour? DNA uptake seems to vary up and down from experiment to
experiment, there is no sign of a progressive reduction in uptake as would be
envisaged if these reagents were going `off` over a period of time. These 
reagents are stored for periods of six months or more, freezing and thawing at
each use - perhaps I should keep small aliquots if this is a problem.
  I do know that electroporation into A4 haemopoetic stem cells gives very
high basal (ie no oncogene cotranfected) levels of expression, obliterating
the oncogene response, presumably caused by influx of calcium ions. Thus we
do not want to use this method for transfections into the NIH3T3`s I am 
currently using. The lipofectin idea is good but expensive, well worth a go
on your recomendation!
  I take your point on the negative control and as you say I am comparing
pBLCAT2 + enhancer with pBLCAT2 alone, with and without oncogene. The problem
is that I get a 10x response with pBLCAT2 alone, and perhaps 11/12x response
with pBLCAT2+enhancer. The response is thus poor and would be clearer if the
background response could be reduced. The greater problem is the lack of
reproducability of the responses between experiments.
  Another suggestion I have had is to treat my cells after transfection
with EGTA to reduce the influence of calcium-dependant (ie well studied)
signal transduction pathways in the hope of isolating other less well
studied pathways. Any comment?

More information about the Methods mailing list