Better to sequence from plasmid? - (Nov/01/2010 )

I'm about to do some sequencing and it could be very, very bad further down the line to have relied on a faulty sequence. We were originally going to sequence directly from PCR using FastTaq and then again on products from a proof-reading Taq. My question is whether sequencing from plasmid clones is better than sequencing straight from a cleaned-up PCR, provided that a proof-reading taq is used?

If it is better, would the reason be that bacteria have better mechanisms for error-free amplification of the sequence than does a simple proof-reading taq?

both are ok and it's more the matter of facilitating the sequencing reaction than of error-free amplification
Even if Taq introduces mutation, you have to remember that amplicon you're sending for seqencing consists of milions of DNA copies, therefore a single SNP won't change the nucleotide read by the sequencer. And you'll distinguish it on the chromatogram by the peak size
As for plasmids, there's always a risk of them being supercoiled, what inhibits amplification and sequencing, but if you send enough template (let's say 200ng of plasmid vs 50 ng of PCR product) all shoud be fine. You can also linearize plasmid to make sure it's relaxed.
Anyway, if it's an important construct I'd sequence the plasmid in order to check not only insert, but also regulatory sequences of the vector itself
G.

gangut on Mon Nov 1 12:22:37 2010 said:

Ok, so I guess you've amplified the genomic DNA fragment and then ligated it into the vector, pGEM or sth? If so, I'd rather sequence the product of direct genomic amplification, and if it's impossible (because of dimers, other products in the PCR), I'd definetely sequence more than one such plasmid. That's because I'd rather be afraid of mutations introduced by polymerase on the stage of genomic PCR. if there was a SNP introduced by Taq to a single PCR copy and you had enough bad luck to have this single copy cloned into pGEM and sequenced, you'll get false results.

This depends a lot on what you want to do next. If you simply need the sequence of an existing piece of DNA, then pcr amplification is fine, and you need not even both with a low error rate enzyme. Just clean up the reaction and send it in.

If you want to use the DNA as part of a construct you are making, then you really need to clone it and pick single colonies to sequence. Errors rates that are inconsequential in a sequencing reaction can be very important if you select a single molecule to clone into a single cell and then propagate. You can't rely on statistical averages to help you when you are selecting a single molecule.

What about the Taq? Would you only use a proof-reading polymerase or is it okay to do multiple reactions with a normal Taq?