Nested Deletion paradox

[Klaus.Matthaei at ANU.EDU.AU]

>Dear Netters:
>
>I've been sequencing a couple cDNAs using nested deletions (made with the
>Promega Erase-A-Base kit) and have consistently run into something that
>puzzles me. After digesting the DNA with Exo III, blunt ending with S1
>followed by Klenow treatment, I will take aliquots of each time point in
>the deletion series to analyze by gel, as recommended by the kit. In the
>meantime, I set up ligations that I leave overnight at room temp. Assuming
>the restriction digests I used to prepare the DNA worked well, I usually
>get a very nice ladder of tight bands ranging in size from my cut vector
>plus insert to the empty cut vector. I then transform JM101 with the
>products of the ligation reactions. (I should mention that I EtOH ppt, wash
>w/70% EtOH and resuspend the ligated DNA in water to cut down on salts
>before electroporation.)
>
>The puzzle comes when I pick several transformants from each time point and
>analyze miniprep DNA for clones of the right size. Each time point always
>reveals a radical range of differently sized clones with little
>correspondence between the timepoint and the size of the deletion clones
>observed. As a result I'm forced to pick, miniprep and analyze lots of
>clones in order to get a series of deletions spanning my cDNA. I'm not
>griping so much as I'm curious why there isn't a better correspondence
>between timepoint and resultant size of deletions. My colleagues have told
>me this result is typical. The poor Promega tech said he hadn't heard of
>this happening and would check into it (he still hasn't come back to me.)
>My best explanation is that the products of Exo III deletions at each
>timepoint are a range of sizes based on the fact that not all of the DNA is
>bound by Exo III at the same time. If the products at each time timepoint
>were plotted as size of DNA vs frequency, I would expect to see a skewed
>left distribution, as the size distribution would tail out to those
>molecules that had been barely digested.
>
>According to this thinking, it would seem that the concentration of Exo III
>is limiting, and that I would probably get a smaller range of clones at
>each timepoint if I increased the Exo III concentration.
>
>Please comment on whether this result is typical; and, if not, how I can
>rectify the situation. I'm tired of all these minipreps.
>
>P.S. I've also found that I need to transform E. coli with the ligated DNA
>immediately if I'm to get a decent number of transformants. The products of
>the ligation reactions do not seem to store well, even as pellets under 70%
>EtOH. Why is this?
>
>-- 
>Casey M. Finnerty
>Boyce Thompson Institute at Cornell University
>cmf5 at cornell.edu


The number of different size inserts I think can occur for a number of
reasons including differential cutting by the two restriction enzymes used
i.e. the blocking enzyme and that which allows Exo III to 'erase'.  This
must be checked with approriate gels after each cutting step.  But to help
prevent MiniprepRSI try the following 'Cracking" procedure that we have
been using for years to size inserts


"CRACKING for insert size"
Use a sterile toothpick to replicate an individual colony (approx. 1 mm or
larger) onto a grid pattern on a fresh plate of LB agar/ampicillin (50
ug/ml), then swirl the toothpick in 25 ul of fresh 'cracking' solution:

        835 ul distilled water
        25 ul 2 M NaOH
        25 ul 20% SDS
        10 ul 0.5 M EDTA
        100 ul glycerol
        5 ul 2% bromocresol green

Do not leave toothpicks standing in the solution since they absorb it,
thereby concentrating the (chromosomal) DNA which causes both difficulty in
sample loading and poor electrophoretic resolution.

Repeat for a reasonable number of colonies (usually 10-50, depending on
background) and for supercoiled plasmid standards (vector alone, and a
plasmid approximating the size of the desired recombinant).

Incubate the replica plate at 37oC til' clones visible again.

Heat all tubes at 65oC for 30 min, chill on ice, centrifuge very briefly
and load into narrow slots in a 1% agarose minigel containing ethidium
bromide (0.5 ug/ml).  Electrophorese at 60-100V until the marker dye has
run 3-4 cm.

Supercoiled plasmids are clearly visible under u.v. illumination; desired
recombinants are readily identifiable by their size.

Colonies containing desired recombinant plasmids are picked from the
replica plate and inoculated into 2 ml of fresh LB broth supplemented with
0.2% glucose and 50 ug/ml ampicillin.  The cultures are aerated vigorously
overnight at 37oC, and plasmids isolated by alkaline/SDS miniprep or
whatever you like.


With this you can crack as many as your gel system allows.  With
erase-a-base I crack about 4/timepoint and miniprep the most likely ones
from size although I generally make single stranded DNA since I use a
phagemid.

Have fun

Cheers, Klaus
--------------------------------------------------------------------------
Klaus Matthaei
Gene Targeting
The John Curtin School of Medical Research
The Australian National University
PO Box 334, Canberra, ACT 2600, Australia
E-mail: Klaus.Matthaei at anu.edu.au

"I'd rather a full bottle in front of me than a full frontal lobotomy"
--------------------------------------------------------------------------