Andy Phillips andy.phillips at
Wed Feb 7 19:24:01 EST 1996

Joerg Hauf wrote:
> Hi everybody,
> I want to perform a one hybrid screen, for determining a DNA-binding
> protein. Therefore I have to clone my putative binding site tandemly four
> or five times (same orientation) into a vector in front of a reporter
> gene.
> Does anybody know an effective method to achieve this aim?
> Many thanks in advance!
> Joerg

This is the method we used: 

If a double-stranded oligo ligates head-to-head, you get a palindrome. If 
you're lucky, the centre of this palindrome, which is derived from the end 
of your sequence, will be the cleavage site of a restriction enzyme that 
doesn't cut elsewhere in the oligo.


Head to head dimers have the sequence ...XXGATCXX... in the middle and so 
are cleaved by Sau3A. Tail-to-tail dimers have the sequence ...XXGGCCXX... 
in the middle and so are cleaved by HaeIII. phosphorylate and anneal your two oligos, self ligate into 
concatamers, then digest with Sau3A and HaeIII. Run the products on a gel 
(hint: we got much better resolution of oligomers on a 6-8% polyacrylamide 
gel than we did on Nu-sieve), elute the multimer that you need (we saw 
monomers up to 8-mers quite clearly) and ligate into a blunt-cut vector. 
Transfer into your reporter construct.

If the ends of your oligo aren't so felicitous, and don't form a 
restriction site, you'll have to add a few bases at one or both ends to 
create a site.


can be 'cured' of head-to-tail oligomers by digestion with EcoRI & BamHI.

Hope this helps.


 Email   : andy.phillips at    : University of Bristol
 Home    : (not working!)               : IACR Long Ashton Research Station 
 Phone   : +44-275-392181 ext:257       : Long Ashton
 Fax     : +44-275-394281               : Bristol, BS18 9AF, UK

More information about the Methods mailing list