Mapping random mutagenesis

Greg Tobin tobin at
Mon Feb 26 16:58:08 EST 1996

In article <4gt4em$9mt at> Peter Wang <plw at> writes:
>Benjamin Braun (bbraun at wrote:
>>I am thinking of a project that would entail random mutagenesis of
>>a 300-1500 bp promoter fragment.  My question is, is there any way
>>to identify where a mutation lies to within a few base pairs short
>>of sequencing the construct?  That way I could assess an entire
>>population of mutants in one reaction - to assess the randomness
>>of mutation, and after a genetic selection to look for clusters of
>>mutations.  For example, I was considering using RNase protection
>>against a library of mutants.  In an ideal mutagenesis I would get
>>cleavage at each base with equal frequency, but hot spots would
>>show greater cleavage frequency, and therefore darker bands.
>>Linker insertion would be great, but I don't know of an easy way 
>>to get good distribution of the linkers.
>>Any ideas out there?  I'd love to hear some!
>My disclaimer for the ideas below is that I don't have any personal 
>experience with the methods, but have been looking into them for uses 
>similar to what you have in mind (mapping mutations in selected 
>populations of DNAs).
>Similar to your idea of RNAse protection (RP) of a selected population, 
>you might look into chemical cleavage of mismatches (CCM) which probably 
>would work better than RP for what you have in mind.  You can look for 
>reviews by Richard G. H. Cotton (I think the original reference is PNAS 
>85:4397 (1988), but there is probably more recent stuff too).  
>Heteroduplexes between the wildtype DNA and mutant DNA (made by PCR) are 
>cleaved by chemical reagents like hydroxylamine that specifically attack 
>unpaired bases, and the location of the mutation can be estimated by 
>sizing the fragments on a denaturing gel (DNAs are radioactively 
>A very elegant way to use CCM, if you have access to an ABI automated 
>sequencer, is fluorescence-assisted mismatch analysis (FAMA) (PNAS 
>91:1873 1994).  I heard a very nice talk by Tommy Meo about this and he 
>would be very helpful if you were interested, I think.
>Similarly, there are a couple of papers on using enzymes to cleave at 
>mismatches (Nature Genetics 9:177 (1995) and PNAS 92:87 (1995)), though 
>I think CCM is the gold standard.
>- Peter
>Peter Wang, M.D., Ph.D.
>MRC Centre for Protein Engineering,
>Hills Road, Cambridge, CB2 2QH, England
>Tel (01223) 402104  (international calls +44-1223-402104)
>Fax (01223) 402140  (     "          "   +44-1223-402140)

Another method came out of Dan Nathan's lab in the mid-80s.  They digested
a plasmid containing the SV40 genome to statistically nick the plasmid once.
The population of molecules representing a nick at every nt was opened with
S1, and an 11-mer was ligated to throw off the reading frame.  The re-circularized DNA can now be used to transform bacteria.  Probe colony lifts with end-
labeled oligos that span the region that you want mutated. and probe a 
replicate with an oligo that spans another region of interest.  If the 
inserted oligo is at that site, the probe will not hybridize. So you pick
colonies that hybridize to one, but not the other oligo.
This worked very well and was shared among many labs working on SV40.


DPeebles nicked a plasmid containing the SV40 genome such that h had a population of molecules with one nick each.  This was then opened
for SV40.  Peebles nicked a\

More information about the Methods mailing list