epitope tag: two or one?

Frank O. Fackelmayer fof1 at chclu.chemie.uni-konstanz.de
Wed Jun 24 05:27:35 EST 1998


Chris Boyd wrote:
> 
> Frank O. Fackelmayer (fof1 at chclu.chemie.uni-konstanz.de) wrote:
> : Chris Boyd wrote:
> : >
> 
> <snip -- use of 132 nt long oligos for epitope tagging>
> 
> : > They're still usable for PCR though, as many published papers testify.
> 
> : Yes, they can be used for PCR, but that does not say they have the correct
> : sequence. Have some 10 correct bases at the 3´ end, and viola, your PCR will
> : work.
> 
> Thanks for that, but I had already grasped the principle of PCR.

I didn´t want to offend you, sorry. 
The point I wanted to make is that it is NOT AT ALL safe to use very long
oligos. From my experience, many people are NOT aware that oligo synthesis is
a complicated technique with intrinsic problems. Instead, they believe that
ordering an oligo is like ordering some lab chemical ("what you order is what
you get"). 

> 
> Nowhere have I claimed that when you order a batch of 132-mers they are
> all of the correct sequence, and I fully realise such long syntheses
> are pushing the current technology to the limit.  I was offering an
> extreme example to illustrate a point -- in reality, it would be safer
> to add no more than two or three tags by a single PCR.

agreed

> 
> However, unless you or someone else can convince me otherwise by
> reference to published data, I will still cherish the notion that a
> usable fraction of a 132-mers batch (if HPLC-purified) has a sufficient
> proportion of oligos of correct sequence to make their use as I've
> outlined a viable option.

I cannot point to published data, but I have talked to several companies some
two years back, when we wanted to construct an artifical protein domain by
cloning (without PCR) complementary oligos. The "coding region" was planned to
be 150bp long, but not a single company felt able to synthesize such oligos
with an acceptable error rate. We resorted to annealing/ligating six 50mers
(with overhangs), but even then sequence analysis showed that none of more
than twenty clones was correct. In most cases, there were deletions or missing
bases, and some clones could not be explained at all. Project cancelled.
Since then, I have only once used an oligo >50bases (56nt, no problems in that
case), but generally try to avoid those big monster oligos. 

Best wishes,
Frank



More information about the Methods mailing list