A comparison between AgeI and BshTI

[Trond Erik Vee Aune]

Hi again

I've been trying to cut a PCR product (actually a PCR library) with both 
AgeI and SacI for cloning into my expression vector (which is cut with 
the same enzymes). The results have been fra from satisfying since I 
need lots of tranformants, preferably 10 000+. Thank to a hint from Dr. 
Peter Barrett I tried the isoscizomer BshTI instead of AgeI. These 
results may be on interest, at least they were to me.

Here is what I tried to clone:

1) Insert (cut with BshTI/SacI) : vector (AgeI/SacI)
2) Insert (BshTI/SacI)          : vector (BshTI/SacI)
3) Insert (AgeI/SacI)           : vector (AgeI/SacI)
4) Insert (AgeI/SacI)           : vector (BshTI/SacI)
5)                              : vector (AgeI/SacI)
6)                              : vector (BshTI/SacI)

When cutting with both SacI and BshTI, I first cut with SacI for 3 hours 
then I added more NaCl and Tris-HCl to optimal BshTI buffer conditions, 
and added BshTI for 3 hours further cutting. When cutting with SacI and 
AgeI, I cut for 3 hours with SacI and then added AgeI for 3 hours 
further cutting. This ensured that the results were comparable.

The results were as follows, I did have blue/white selection (alpha 
complementation) on my vector (LITMUS28):

           Blue             White
1)    many hundred   3 times less colonies
2)       ca 14            ca 2000 (!)
3)    many hundred   5-6 times less colonies
4)         8               ca 432
5)      ca 2000               0
6)        12                  3

As you can see it is obvious that BshTI cuts much better than AgeI, this 
is especially apparent on the re-ligated vectors (nr 5 and 6).

And for me personally the results where I cut both insert and vector 
with BshTI (nr 2) were very welcomed, ca 2000 white colonies and very 
few blue :) This means that I now have a method to cut my PCR library 
and clone it in with high efficiency, and extrapolating from this 
results I end up with a library of around 10 000 colonies after one 
error-prone PCR reaction.

To lab veterans this may not be very surprising, but to a person without 
many years of experience working with restriction enzymes, it was quite 
shocking. There should be a database for enzymes which shows the 
activity of different isoscizomers compared. I'll never use AgeI again 
(even though it works fairly ok when one only needs one transformant, 
but taking care of a library is something else completely!)...

Regards,
Trond Erik