pET and ampicillin

Paul A Bucciaglia bucc0003 at
Tue Nov 15 20:25:51 EST 1994

tmiller at (Todd Miller - Pharmacology) writes:

>	Our lab is using the pET system to express eukaryotic pro-
>teins.  We seem to be experiencing a problem with the BL21 host cells
>losing their plasmid.  Does anyone have suggestions on how to avoid
>this?  We're thinking of trying carbenicillin but its about 10 times
>as expensive as ampicillin.  Also, have inhibitors of beta lactamase
>ever been used (like clavulanic acid)?

>	I'm aware that overnight cultures should be highly diluted and
>we've also played around a bit with the pLys S and E containing hosts.
>We're trying to use this system in a fermentor, but I fear that if we
>grow to the high densities that can be achieved in a fermentor, none
>of the bacteria will have the plasmid when we induce expression.

>Thanks for any suggestions.

>Todd Miller
>tmiller at

The people at Novagen reccomended to me not to grow overnight cultures if 
your construct encodes something unstable in ecoli.  They reccomend 
growing cultures to the point where bacteria become visible and then 
storing this overnight at 4 C.   Use this to innoculate your culture the 
next day.  It won't help to dilute an overnight that has been overrun by 
bugs which have lost the plasmid or found a way not to express your protein.

I assume you have tried using Bl21(DE3)pLysS , which has a plasmid 
expressing T7 lysozyme, a bifunctional protein which inhibits the  T7 RNA 
polymerase, which can increase the stability of some plasmids.  Another 
thing to try is using glucose (0.4%) when plating out the BL21 
transformants, when growing starter cultures and perhaps in cultures to 
be induced (its generally gone by the time you add IPTG to induce). Like 
everything else in protein expression, its very empiracle :-(

If all else fails try the newer Novagen vectors which carry a Kanamycin 
resistance gene and have the lac operator downstream of the T7 
promoter; next to using a phage to deliver the polymerase this is as 
tight as it gets.

Hope things work out,

paul bucciaglia

More information about the Methods mailing list