Can multiple plasmids comete in transfections?

D.K. dk at
Tue Mar 19 21:55:10 EST 2002

ladasky at (John Ladasky) wrote:
>Hello again,
>A few days ago I posted that I was trying to make a cyan fluorescent
>protein (CFP) fusion construct out of a yellow fluorescent protein
>(YFP) fusion construct.  Well, I did it.
>I'm using various CFP and YFP chimeras in fluorescence resonance
>energy transfer (FRET) studies of protein interactions in vivo.  I
>have returned to the microscope after a few weeks at the molecular
>biology bench -- and I've rediscovered an old problem.
>I am transfecting the human cell line HeLa.  I have performed the
>double transfection in two ways: either as a transient transfection
>with both the CFP and YFP plasmids simultaneously; or, by establishing
>stable lines with the YFP plasmid, then transfecting these transiently
>with CFP.  In both cases, I've observed an extremely annoying pattern:
>those cells which are CFP-bright are YFP-dim, and vice versa.  This is
>making it extremely hard for me to do FRET!  I do want variations in
>the expression levels of the two proteins from cell to cell, but I
>need them to be somewhat uncorrelated -- positive correlation I might
>be able to work with, but negative correlation is a disaster.
>All of my plasmids are built from Clontech's pN3 vector.  Therefore I
>have the same eukaryotic promoter in each plasmid, the ever-popular
>human cytomegalovirus immediate early promoter (pCMV-IE).  One of my
>colleagues suggests that a transcription factor which binds pCMV-IE
>may be in limited supply.  Thus, when I succeed in obtaining a high
>level of expression from one promoter, it comes at the expense of the
>other promoter.  This seems plausible to me.  

Does not to me. It does not really explain mutual incompatibility of two
constructs you report. If titrating out some factor needed for expression 
were the case, you'd expect to see lower (compared to a single construct)
expression of both genes. Unless the factor(s) bind essentially irreversibly 
to one plasmid or have memory as to what particular plasmid they were 
bound before... Sounds too strange to my taste. 

The first thing that comes to my mind is that cells do not tolerate well
the event of your two proteins interacting [at elevated levels]. The obvious
control is co-transfecting just CFP and YFP. Presumably you have done 
this as part of controls for FRET signal, etc. Do you observe a negative
correlation in this case too?


More information about the Methods mailing list