How to randomly cut DNA yielding blunt ends?

[Paul Digard]

o. van-ham wrote:
> 
> I'm developing a protocol that includes a step where DNA needs to be cut
> randomly, leaving blunt ends.  I initially digested the DNA with DNase I
> and cut out the band corresponding to linear DNA.  When I sequenced my
> final clones, it turned out that many of the termini had been chewed
> back.
> 
> 
> Does anyone have any suggestions?  If so, please let me know.


How about partial digestion with four-base cutters like HaeIII, ThaI or 
AluI?  Should give you a pseudo-random set of fragments.  If you're using 
supercoiled plasmid DNA as the starting point, ethidium bromide works 
well as a means of controlling the amount of digestion. [Parker, R.C. 
(1980). Conversion of circular DNA to linear strands for mapping. Methods 
in Enzymology 65, 415-426.].  
Otherwise, sonication followed by end repair is what people use (or used 
to use?) for sequenceing projects.

Paul