Bob:
I would personally prefer, if possible, to avoid synthesizing a
150-mer. The final goal could be achieved by an alternative
method, such as stepwise annealing and extending 6 overlapping oligos of
approx. 40 bases each (3 upper/forward primers, 3 lower/reverse, both
strands), followed by PCR amplification with end primers.
Alternatively, a 150-mer can be synthesized and PAGE-purified. This is
the preferred way for most customers, because they do not want to mess
around with multistep annealing-extension experiments.
Invariably all these long oligos are used for cloning, and I always
insist on warning the customers of the biggest problem with the
very long oligos, notably the internal one-base deletions, caused by a)
incomplete capping; b) incomplete deblocking; c) "fast"
protocols; d) other factors. Inadequate protocols and incomplete
deblocking can be greatly ameliorated by designing custom
programs for very long oligo synthesis. Unfortunately, little can be
done to avoid the incomplete capping. In addition, G duplications lead
to a (n+1) phenomenon and can cause (n-1) mutants to migrate in PAGE as
normal full-size oligos.
We do several things when synthesizing longer than 100 bases oligos:
a) use low load 2000A CPG (1000A CPG is for 50-120 mers; 2000A for
100-200 mers);
b) use larger scale synthesis in order to compensate for the low load
CPG and the high percentage of aborted products;
c) use custom programs for slow and more accurate synthesis;
d) perform or strongly suggest PAGE purification;
e) prefer or suggest analytical PAGE runs (instead of preparative runs),
followed by NARROW full-size band excision and subsequent PCR
amplification (if double-stranded DNA is acceptable), in order to
decrease the chances of contaminating our wt sequence with n-1 products.
e) warn the customer that he/she needs to select several clones for
further verification, not just one or two; and
f) pray to Jesus, Yahweh, Allah, David Koresh; when it comes to very
long oligos we also touch wood, and do not let black cats cross our
way...
There have been a couple of suggestions about mild deprotection. I
would personally caution against the mainstream in this respect. Full
detritylation (complete deblocking) is a must for high fidelity
synthesis of long oligos, and in that respect TCA has its advantages
over the DCA. We used to have a DCA-based chemistry on an old machine
that we scrapped years ago (Biosearch 8700), and fidelity on that
machine was generally worse that the fidelity of our new machines, which
are all based on TCA. I believe PerSeptive did extensive research prior
to switching from DCA to TCA, and TCA showed better results (there are
other opinions on the subject DCA vs. TCA, and I admit that I might be
wrong in this respect).
Best regards,
victor
This archive was generated by hypermail 2b29 : Fri Jun 30 2000 - 07:51:55 EDT