strange transformation results

D.K. dk at
Sat Apr 27 14:46:53 EST 2002

sun at (William Sun) wrote:
>I have sometimes observed a strange phenomenon while subcloning and would
>like to get some input.  When I subclone a piece of DNA into a cloning
>vector (I usually subclone using two different restr. enzymes) and transform
>into bacteria, I include a negative control of just the cut vector without
>insert.  Once in a while, I would get more colonies on my negative control
>than experimental plate.  But when I screen colonies from my experimental
>plate, most of them are positive!  Can anyone offer an explanation?

IMHO, the explanation is trivial. I always tell about this to very new
student in the lab and explain why he/she should feel very optimistic 
doing minipreps when sample has same or lesser number of colonies
than control (as long as the total number is not outrageously high).  
Here is why:

vector composition: 

a)                   1---------------------------------0 
b)                   1------------------------------------1
c)                   0------------------------------------0 ,                      

where b+c are the only guys that contribute to the background
signal (nerative control, vector only). They are _always_ present
at ~ 0.1-1% of the total. 

insert:                      0---------------1

If complete ligation of insert into vector is not efficient enough
(say, one of the sites, 0 or 1, ligates a lot worse - be it incomplete 
cut or imprecise cut or one base overhang), then with an excess
of insert you are going to get mainly either of these ligation 


None of them will give rise to any colonies, but as a result in
vector+insert ligation there will be fewer colonies because there
will be less vector to self-ligate available!

A summary of all of the above can be this: in directional cloning, 
when ligation mix is far from what it is supposed to be in ideal 
case (perfect and 100% complete cuts), the excess of the insert 
inhibits vector self-ligation by ligating to one, but not to another 
end of the vector. 

Have fun with cloning!


More information about the Methods mailing list