Competing plasmids in transfections, more details

John Ladasky ladasky at
Wed Mar 20 12:13:24 EST 2002

Thanks to everyone who has responded to my questions about
transfection of multiple plasmids.  Let me follow up with some
additional details and remarks.

First, concerning the transfection method: I have tried two

In the simpler case, I perform a double-transient transfection with
both the CFP and YFP plasmids simultaneously.  I transfect HeLa cells
on sterilized cover slips in 35 mm culture dishes.  I use one
microgram of each plasmid, and I use six microliters of FuGene6 lipid
transfection reagent from Roche.  The cells are at about 5% to 10%
confluency at the time of transfection, lower than usual, because I
know that it takes about 48 hours for the proteins to reach maximum
expression.  Between transfection and analysis, HeLa cells will divide
approximately twice.  I want the cells to be reasonably sparse and
spread out when I fix them, so that I have a nice flat cytoplasm in at
least some of the cells.

Using FuGene, the transient transfection efficiency is around 5% for
each plasmid.  I have conducted a pilot electroporation of HeLa cells
and found that I can obtain up to 25% positives.  I'll be looking into
electroporation in an attempt to improve my odds.

I have also prepared cell lines which stably express the YFP
construct.  So, 100% of these are YFP-positive, although the
fluorescence intensity from cell to cell can vary by as much as a
decade.  I have kept these in culture for a few months at a time, with
no observable decline in fluorescence intensity.  When I transfect
these cells transiently with one microgram of a plasmid containing a
CFP fusion protein, again I get about 5% CFP expression -- but,
invariably, I find that the cells which are CFP-positive are the
dimmest by YFP.  I have tried two entirely different CFP fusion
proteins, so I do not believe that the problem is the formation of a
complex between the CFP and YFP fusion products which are rapidly
degraded, as suggested by Peter Cerepanov.  Also, we have no trouble
co-immunoprecipitating the native protein dimers which I'm trying to

I will need to check this out further, though.  Supposedly, one our
lab's former grad students co-transfected CFP- and YFP-class I MHC
fusion proteins into HeLa cells, and did NOT observe a negative
correlation between the two products.  At least, she didn't complain
about this problem to the boss when she was trying to do FRET. 
Perhaps there's something specific about one of the proteins I'm

I entertained the possibility that my FRET was actually so great that
almost all of the CFP signal might be mopped up by FRET, when YFP
expression was high.  This does not appear to be the case.  When I
look at the double transfections, I never see strong dequenching of
CFP in an unexpected, YFP-high cell.

I have considered using different promoters for one of the genes, as I
mentioned in my first post.  Ian York, Wolfgang Schechinger, and
Pamela Norton all seem to  think that this may be a reasonable way to
go, and have even suggested some of the approaches that I have

I have Invivogen's pMG plasmid, which contains two multiple cloning
sites, one driven by a variant of pCMV-IE, and the other by a
synthetic promoter containing bits of the HIV LTR.  As I recall, this
promoter is also found in the vector pBJ1-neo.  (I'm posting from home
right now, away from my notes.)  pMG uses an IRES element to drive
expression of the NeoR gene from one of the promoters.  I prefer this
approach to using the IRES to express my second gene of interest,
because it is known that IRES elements are substantially less than
100% efficient.  I might try cloning both of my genes into pMG. 
Alternately I may be able to get expression of just one product from
the HIV LTR promoter, without having to move the other one out of pN3.

I also have Clontech's pBI, which contains a multiple cloning site on
each side of a tetracyline-responsive element (TRE) promoter.  The
transcription factor which binds TRE is called the tetracycline
transactivator (tTA), which is a synthetic protein encoded on another
Clontech plasmid.  There are several options here, as with pMG.  I
might perform transient or stable transfections.  I might insert one
or both genes into pBI.  It is difficult to know how to proceed.

I would love to use a fluorescence-activated cell sorter (FACS) to
select for stable cell lines expressing high levels of both YFP and
CFP.  However, I don't have ready access to a FACS machine with a
violet laser.  That's a pretty rare item!  I don't think that making
double-stable lines would work well, anyway, until I manage to address
the negative correlation issue.

Of course, I have no relation to any of the companies I've mentioned
in this post.

Thanks again for your interest and advice.

John J. Ladasky Jr., Ph.D.
Department of Biology
Johns Hopkins University
Baltimore, MD 21218

More information about the Methods mailing list