fresh transformations for protein expression??

davesmith at davesmith at
Mon Jan 8 16:20:05 EST 1996

Torsten Boerchers <borcher at> wrote:

>Dear colleagues

>for quite some time we are successfully overexpressing
>recombinant proteins in bacteria,mostly with the pET
>system and mostly simple, nice, soluble proteins.

>The more we were surprised that we now on two occacions
>with similar proteins encountered problems (i.e. no
>expression) which were solved by freshly transforming
>the E.coli (BL31(DE3)) with the expression vector instead
>of starting from a plate. So fine, that works, but I would
>nevertheless like to know whether others had similar 
>experiences and whether there is an explanation for this 
>behaviour available. I heared from some other folks routinely
>doing fresh transformations but could not get a rationale.

>Looking forward to some ideas


We routinely use BL21 (DE3) harboring various constructs, all of which
are pET11a derivatives of one sort or another.  All of our expressions
with this system to date have been membrane proteins (lysis proteins
from bacteriophage).  We have seen the effect you mentioned
consistently since first using the system.  We rationalize this effect
to 1) the toxicity or nature of the protein being expressed and 2) the
dual layer of transcriptional control (repression at the promoter for
T7 RNAP and at the hybrid promoter on the vector), although much
tighter than standard cloning and expression vectors, IS LEAKY.  In
other words, in our experience, E. coli has difficulty with membrane
protein overproduction, and the pET system is never OFF.  Thus, any
colonies on a plate are almost guaranteed to have a low background
level of plasmid-encoded protein.
Further complications are due to plasmid copy number effects....a
fresh transformant will have FEWER plasmids per cell and presumably
lower background levels of protein expression as compared to an aged
cell on a plate where copy number is maximized due to stationary
growth phase effects.
Is anyone aware of other problems which we haven't considered?
Does anyone have any suggestions on how to alleviate the need to use
fresh transformants for every experiment?  I'd sure like life to be

Dave Smith

More information about the Proteins mailing list