ExoIII is a Dinosaur. Try DNAse 1 or TN1000

Roger F. Anderson rander at cco.caltech.edu
Thu Sep 16 14:32:39 EST 1993

Dear Netters,

Recently there have been several posts regarding EXO III nested deletions.
Years ago I used to do them and I found the proceedure time consuming
and inefficient. Once you got past 3 Kb you had trouble. That was IF you
could find the right protecting enzyme sites. There was some methylation
method but that was also trouble. Fortunately there are two new, faster and 
better ways of getting a "nested set" of sequencing clones. Both of these
methods have been used in our lab with great success. Both generate sets
on a s few as 2 plates! Not the 10 - 30 plates used for EXO-S1. The kit 
from Promega is nice but I made the same thing for less than half before.

Method 1:
PNAS vol 88 pp1247-1250 feb. 1991. The materials are now available from
GOLD Bio Technology (800) 248-7609, in Mo. (314) 647-2509.
I do not work for them and I will not supply the materials because they have
been upgraded and because GOLD has the license. I don't know the cost either.

This method is based on the random insertion of a transposon. Each resulting
colony has the transposon in a different place. The vector is constructed
to select against insertions outside of the "insert". This works 95% of
the time. To determine the insert location PCR primers are made to sites
in the vector polylinker and to the inverted repeat at the ends of the 
transposon. There are unique sites on each end of the transposon for use as
sequencing primers later on. Using PCR amplification with one end primer
and the IR primer and a second reaction fron the other vector primer and
the IR you can determine the distance to the insertion site. Then by 
selecting the apropriate set of colonies you can sequnce the whole thing 
in both directions. I did a 6.5 Kb clone in 3 weeks from start of mating
to finished Sanger sequencing. Reading took longer but that's another

There are some probblems encountered with this method. One you have to have
threee strains. This is trivial but adds two days over what I will describe
next. The only other trouble is that occasionally we got expression of
the insert which was toxic to the bugs. We know this because we removed
the first part of the coding region and things were fine. I like this
method alot and I have used it for 2 years. I will never do an EXO III
deletion agaain to generate sequencing sets.

Method 2:

Biotechniques Vol 14 No. 3 1993 pg 408 -411

DNAse 1 cuts both strands in the presence of Mn2+. By allowing random cuts
on a frequency of ~1 cut per insert molecule  and having different ends
cut to leave a 5' overhang two sets of nested delteions are generated.
The paper is very clear and so I will not go into all of the details.
Suffice to say that no special strains are needed and not matings have to be 
performed. What you get is a nested set which is mapped in a similar way to
the transposon method but you only need polylinker primers.

Drawback: you have to isolate the insert.
Variant: use 4 hitters which leave blunt ends.

This method was submitted by James Eberle from the Dept. of Bio Syracuse
University. A tip of my hat to him and/or those who helped design this
method. I think it is the fastes most straightforward of the methods we
have tried. You also end up with some potentially useful deletion clones.

I would ask that any questions be held until you have read the papers or
contacted GOLD Biotech. I posted this information because I spent months
screening EXOIII sets before I could sequence and I know what a pain
that can be.

Best of Luck to All
Roger Anderson

More information about the Methods mailing list