JMILLER%VXBIO.SPAN@STAR.STANFORD.EDU (02/15/91)
I was looking into doing what you describe about a year ago, and similar schemes using RNA ligase. I don't think it would be that useful for normal oligos. It WOULD be VERY USEFUL if you could get the reaction to work with trinucleotides instead of mononucleotides, for the following reasons: There is a lot of mutagenesis work on proteins these days that utilizes degenerate oligos to introduce a spectrum of protein substitutions into one or more sites on the protein. The problem with these methods is that a)it is impossible to recover all 20 amino acids at a site, even if you make a ACGACNNNACGACG type oligo, and b)It is impossible to introduce a specific subset of amino acids at a particular site. In a protein engineering project, one might already know that only 4 of the 20 amino acids should be allowed at one site, 7 at another, and so on. The key to doing this would be to synthesize oligos using trinucleotides. You would need about 30 codons. At each step, you would put in a specific subset, precisely restricting variability at that site to the desired amino acids. Unfortunately, this would be very difficult using phosphoramdite chemistry. It could be done using triester chemistry, but nobody uses it anymore, and it would be extremely expensive to make the protected trinucleotides. But a enzymatic method might be practical. One would make the set of trinucleotides using conventional oligo synthesis, and add on a 3' phosphate. These would be the substrates for the reaction. Any enzyme capable of adding on the trinuclotides would make a very interesting methodology. RNA ligase might be a better bet than terminal transferrase. Comments? -Peter Markiewicz