Question on basic cloning theory

Simon Andrews simon.andrews at
Wed May 2 04:56:52 EST 2001

rocky wrote:
> Hi
> Can anyone tell me:
> Is there only one plasmid goes into single competent cell?

Not necessarily.
> If no, how can I make sure I get a unique clone from the single
> bacterial colony?

Bacteria have inbuilt mechanisms for regulating the copy number of a
particular plasmid they carry.  (Look for information on partitioning
systems).  Basically when you first transform a plasmid then a single
bacterium may contain more than one variant of the plasmid.  When that
bacterium divides though the plasmids will be split between the two
offspring.  This provides a chance for the different plasmids to be
separated (just by chance).  Once a bacterium has a pure plasmid
population it will never become contaminated from outside (unless you
make it competent again) so the purification of plasmids is a one way

Under normal conditions you will find that the partitioning systems for
the sort of plasmids used for cloning will produce a pure plasmid
population (one clone per bacterium) within a few divisions.  This is
why you leave your transformed culture to grow in liquid for a few
doubling times before you plate them out.  If you plated them straight
from transformation your could end up with mixed populations (although
if you streak the colony out to single colonies and picked one then that
should be pure).

> (  Note: I carry out an PCR for the target using mini-prep rised from
> single colony but find 2 products with ~100 bp differences. Not sure
> the 2nd product is non-specific or not since sequence analysis not yet
> done! )

If you're only using one vector (different plasmids can have independent
partitioning systems) then this shouldn't be a problem.  If your culture
appears to be mixed then streak out to single colonies and regrow.

It is possible that the larger product is toxic (maybe only mildly) to
the cell so there is an evolutionary pressure to delete part of this. 
We have had clones which were unstable for this reason, and gave bands
of varying size when prepped.  Try using a vector with transcriptional
control if this might be a problem.


More information about the Methods mailing list