Prok. expression vectors; anything that is not based on Lac operon and Amp R?

Mikhail Alexeyev malexeyev at
Fri Mar 8 10:09:53 EST 1996

In article <4hnn8q$alr at>, Hedrick Lab
<lkane at> wrote:

>  You say that 
> you MUST use another origin of replication, yet I've heard from various 
> sources that it is possible to get both plasmids in if you simply 
> maintain the necessary drug selection.  Is this efficiency just very 
> low?  

The efficiency of introduction of the second plasmid with the same
replicon (if I understood correctly what you meant) is quite low, indeed,
but still is in the practical range. The trouble is that at least
sometimes (my gut feeling tells me that this should happen quite often)
strain with two plasmids that have ori from the same incompatibility group
grow poorly and form small colonies (that is if they forced to maintain
both plasmids). I think that this should be especially true for multicopy
replicons. I am aware of at least one report when incompatibility has been
used to integrate one of the two plasmids into the chromosome (those were
low-copy-number plasmids). This is just to demonstrate how (allegedly)
unhappy E. coli is about carrying two 'incompatible' plasmids. I think
that it is a good idea not to stretch your luck if you want to get a
decent expression level of your protein in a strain that has two
{multicopy?} plasmids. Therefore, I should admit that my 'MUST' was kind
of too emphatic, but the truth is that even though one can make two
'incompatibe' plasmids coexist in one cell,  expression of protein under
these circumstances is (in my humble opinion) too much to ask.
> Secondly, which of the plasmids that you mentioned (i.e. with a replicon 
> distinct from ColE1) are high copy number?  I'd like to get decent 
> expression in the bugs, so I'd assume that having the high copy number 
> would help.

I agree that high-copy-number replicon is a good place to start (although,
at the end it could become clear that low-copy-number one would do the
best job). I would suggest R6K gamma-ori (about 420 bp). It requires
product of pir gene for replication. Same plasmid can be either moderate
(appr 15 copies/chromosome equivalent) or high (appr. 250 copies/chr.eq.)
depending on which allele of the pir gene is used. Since pi protein
(product of the pir gene) is active in trans, you may propagate the same
construct at either moderate or high copy number. Here is a nifty trick:
you can place wt or mutant allele of the pir gene on your  colE1-based
plasmid and select for R6K-based plasmid marker to maintain both plasmids
(Col E1-based plasmid will be maintained because pir gene cloned there is
required for replication of R6K-based plasmit that is being selected for).

> Thanks again for all of the tips.
> Craig Walsh

Good luck,

M. Alexeyev

More information about the Methods mailing list