100 bp long DNA is a piece of cake to synthesize and clone. You do not
really need to go with shorter oligos, annealing, extension, ligation etc.
You do not even need to PAGE purify. I would suggest:
1) Make one 100-mer, corresponding to the full length of your DNA. It may
contain more "aborted" products than full-size oligos, but there is no need
to purify it.
2) Make one 20-mer, corresponding to the 5'-end of the 100-mer (forward
orientation, i.e. identical to the 5'-end of the 100-mer). It would help to
have a restriction enzyme site incorporated as an add-on to the 5'-end of
the 20-mer. If that is the case, the 20-mer may become a 26- or a 30-mer.
3) Make one 20-mer, corresponding to the 3'-end of the 100-mer (reverse
orientation, i.e. reverse complement to the 3'-end of the 100-mer). Again,
you can introduce a restriction site (depending of the cloning vector, but
preferably one that will give you a cohesive end different from that at the
5'-end). If a site is introduced, the 20-mer will become a bit longer (up
to 30-mer).
4) Use the 100-mer as a template for PCR and amplify it with the short 5'-
and 3'-end primers. This way you will greatly reduce the relative amount of
all "aborted" products from the 100-mer synthesis (only the full-size
100-mer synthesis products will be amplified exponentially by both primers).
It is better to use high-fidelity thermostable polymerase (such as Pfu from
Stratagene) instead of the regular Taq polymerase, but we also worked with
regular Taq and were able to get good results.
5) Digest the PCR product with the two restriction enzymes (if you
introduced those at steps "2" and "3"), and clone it in your cloning vector.
6) SELECT SEVERAL CLONES FOR VERIFICATION BY SEQUENCING. Bear in mind, even
PAGE-purified oligos will often contain some "mutations" (synthesis errors),
most often one-base "deletions". This is well explained in the article
Error analysis of Chemically Synthesized Polynucleotides. Biotechniques,
1998, 24(2), 256-260. The authors synthesized and cloned a 123- and a
126-mer and they had 50% of the clones containing deletions.
Our experience with thousands of clients from 5 continents (including
Columbia University) over the last 3 years also shows that one-base
deletions are the most common problem with the long oligos, which is easily
surmountable if you are aware of it.
Best regards,
victor, Alpha DNA
www.alphadna.com
-----Original Message-----
From: Shalom David Goldberg <sdg11@columbia.edu>
To: Recipients of ABRF List <abrf@aecom.yu.edu>
Date: June 20, 1999 1:11 AM
Subject: gene synthesis
>
>Hi All,
>
>I need to synthesize a gene that is about 100 bp long. I'm planning to
>use a PCR-based method, using overlapping oligos. I would really like to
>know if there's anyone who has done a PCR-based gene synthesis, especially
>of a long gene, how well it worked, and suggestions for construction of
>the oligos (ie length of overlap, oligo size, etc.)
>
>Thanks,
>Shalom Goldberg