ExoIII Deletions of Long Inserts???

[William R. Morgan]

mmelan at marcie.wellesley.edu asks...

>        I am currently having a problem trying to do ExoIII deletions of a
>5kb insert fragment in Bluescript.  I looks as though the ExoIII does a fine
>job of chewing halfway through the insert (i.e. reactions are beautiful
>reducingthe 8kb linear DNA down to about a 5kb fragment) but further reactions
>(later
>timepoints) show no DNA in the lanes.  I am starting with 3-4 ug of DNA in the
>reaction and using 300 units of enzyme.  Has anyone else had a similar problem?
>Am I not starting with enough DNA? Is there a S1 nuclease problem? Could this
>be some strange problem specific to this piece of DNA?  I am using the original
>Henikoff procedure.  Is there a better method?


Dear Melissa Melan,

I have seen similar problems working with fragments > 3 kb.  So generally I
try to use one or more fragments each < 3 kb in length.  Sorry I don't have
any better suggestion than that or an explanation.  

One other possibility is to think about some other methods for generating a
nested set of deletion.  Personally I do like Henikoff's method, but others
have suggested other procedures when I discussed this previously.  You can
search the methods-and-reagents archives using Gopher for a more thorough
discussion of these alternatives, but I've included a particularly relevant
posting below.  I actually tried the transposon-mediated procedure when it
was first marketed by GBT but I found it difficult to obtain certain
deletions and therefore of limited utility.  Their new system may have
overcome this deficiency.  I don't know anything about the DNAse I method. 


Good luck,
Bill Morgan

Date: Thu, 16 Sep 1993 19:32:39 +0000 (GMT)
From: rander at cco.caltech.edu (Roger F. Anderson)
Subject: ExoIII is a Dinosaur. Try DNAse 1 or TN1000
To: methods-and-reagents at net.bio.net
Content-transfer-encoding: 7BIT
Nntp-Posting-Host: sandman.caltech.edu

{message deleted to save space, but header included so it can be quickly
found using Gopher in the methods-and-reagents archives at net.bio.net}


William R. Morgan
The College of Wooster
Department of Biology
931 College St.
Wooster, OH  44691
Phone:  216-263-2026
FAX:    216-263-2378
E-mail: wmorgan at acs.wooster.edu