I was wondering if anything is known about the indigenous E.coli (or other
bacterial) proteins that are expressed in response to expression of
(poisonous) recombinant proteins.
Here's my case. I am expressing a 6his-tagged yeast DNA-binding protein in
E. coli (BLR, a RecA- strain) from a T7 promoter-driven expression vector
(pET-11). The protein is a homologue of RecA and that is why expression in
RecA- E.coli is important to prevent crosscontamination of the target
protein. In addition, overexpression of the recombinant protein is highly
toxic for E.coli, that is why I am using lambda CE6 phage to induce
expression in the non-DE3 strain (strain with pLysS shows same result in
terms of what described below).
On Westerns with primary antibodies against the recombinant protein, I
always see 2 bands, one of the target protein (~37kDa) and another of some
unknown protein (~34kDa) that happens to bind same antibodies. That 2nd
protein is present both in non-induced, and induced cells, as well as it is
present in the non-transformed cells (i.e., the ones not bearing the
expression vector, ! ). Thus, the protein #2 is indigenous E.coli protein.
Interestingly, the level of that protein increases upon induction of
expression of the recombinant protein, almost to the same level as the
recombinant protein, as seen on SDS-PAGE and on Westerns. Since the protein
#2 is present in non-transformed cells, its appearance in transformants and
coinduction with the target protein cannot be explained by proteolysis of
the latter or secondary translation initiation site on the plasmid
construct. RecA- strains still have E. coli RecA homologues DnaB and RadA,
but both proteins have MW higher then 40kDa (vs. 34kDa of the protein #2)
I am trying to understand what is the nature of that protein #2 and is it
possible that it is a certain defense agent synthesized by the host cells to
neutralize the poisonous recombinant protein? It could be that this
neutralizing agent binds the recombinant protein and inactivates somehow,
i.e., some sort of bacterial antibody (SOS response, etc.). The binding
hypothesis could explain the crossreactivity of antibodies. I am sure that
everybody noticed that almost in every case of recombinant expression, one
sees a number of other protein bands that become more intensive on stained
SDS-PAGE gels in induced cells vs. non-induced, whereas the majority of the
bands has the same intensity.
Any comments, ideas, references would be much appreciated.