Assembly PCR, combinatoric library

[John Ladasky]

Hi, folks,

I'm interested in overexpressing a mammalian protein in E. coli -- but
not just the wild type, I want to sort through a library of variants
for improved function.

When moving from mammals to bacteria, it is desirable to switch to
those codons favored by E. coli, so that expression levels are high.
Therefore I expect to re-synthesize the entire gene, using assembly
PCR and overlapping oligonucleotides.  I recently found this
reference:

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=12000848&dopt=Abstract

The authors have ported their computer program to a public web
interface here:

http://helixweb.nih.gov/dnaworks/

My gene is not especially long -- 441 nucleotides.  For the library, I
want to make a broad range of coding substitutions, for the must part
substituting hydrophobic residues for other hydrophobic residues, at
16 codons scattered throughout the gene.

Would it be practical to assemble my combinatoric library using the
same assembly PCR method that one would use to prepare the wild-type
gene?  Although one could clearly have problems with reduced annealing
efficiency and mis-priming, this seems like such an obvious thing to
do.  Has anyone tried?  Are there recommendations and rules of thumb
for making it work?  Is another method favored?  (I *could* start with
a wild-type template and make several megaprimers, then finish the
process using sexual PCR -- it seems like a tedious approach, though!)

I forwarded a general description of my needs to a commercial DNA
synthesis company.  They estimated it would cost them $50,000 to
prepare my library.  That's an awful lot, and I'd like to see if I can
get the cost down!

I have been looking for a reference which describes the generation of
a combinatoric library in the manner I have described here -- so far
without success.  I would appreciate your input!  Thanks.


+-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-+
|   Ladasky Home Solar, Inc.:  blowing sunshine up your   |
|       power grid since March 24, 2005.  Fiat lux!       |
+-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-+
|   Uptime    Downtime    kWh generated     kWh consumed  |
|   737 days  6.5 hours       12764            14195      |
+-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-+