Principles of overlap PCR - (Oct/05/2011 )
I need to clone a 15k genome so I want to run overlap PCR with 4-5 fragments. I want to sequentially perform the overlap PCR and produce the full-length genome. Has anybody done this before?
I can't find any info on protocol-online.org, and the Wikipedia instructions are not enough.
I'm assuming this is a 15kb viral genome, since there's not much else out there in that range. I have not attempted to perform a 15 kb overlapping pcr before, but I have had experience peforming PCR based recombination on large PCR products which should be a similar technique. I will tell you that I have also cloned a few viral genomes in the range of 7-12 kb, and I have always been able to generate 3-6 larger PCR products and then assemble them by restriction digest and cloning. The major problem I see doing this by PCR is that even with a proofreading polymerase, you will be almost guaranteed to generate mutations after the multiple sequential rounds of PCR that would be required. You would have to sequence your final clone, and if you find mutations, you will not be able to tell at which stage you acquired them and will have to start all over with the 1st PCR. At least by dividing up the genome into 3-4 segments, you can subclone and sequence each one, and then assemble them into a final clone without worrying about mutations.
If you do decide to go the PCR route, I can solidly recommend Phusion polymerase by NEB. I personally tested about 8 different polymerases for long overlapping PCRs and Phusion was the cleanest and cheapest. My overlapping PCR protocol basically requires gel purifying each of the semi-overlapping 1st step PCR products, and using 15 ng of each cleaned PCR product in the 2nd step PCR with only the outside primer pair. For you, this project would probably require 3 phases (if you assume 4 initial PCR reactions), gel purifying and quantitating the product from each step:
1) PCR #1, #2, #3, #4
2) Overlapping PCR #1-2, Overlapping PCR #3-4
3) Full-length Overlapping PCR #1-4 (using #1-2 and #3-4 products)
Best of Luck.
You are absolutely right. It is an NDV viral genome. I wish I could have a look at your articles. Can you PM me a link?
The problem is that we do not know the exact sequence of our viral genome. All the genes of the genome were sequenced and published 10 years ago, but the rest of the genome of our NDV strain is unknown. specially the leader, trailer and the intergenic sequences.
First, I went ahead with the Restriction Enzyme method. I mean I designed primers with the RE overhangs to sequentially ligate the PCR fragments, but as you know it is very risky to do that because the sequence of the full-length genome changes from inoculation to inoculation, not to mention that the RE sites in other strains are also very different.
So, I compared the intergenic sequences to pick the REs which are very unique to my strain. However, as I posted last week on the Bioforum my PCR for the first fragment failed because the primers had 20C+ difference in Tm. So I screwed up big time, and now I want to go ahead with overlap PCR to be less risky.
With overlap PCR I don't have to worry about the any RE site. isn't it? what do you recommend? I think it would be better to sequence the genome first before doing the RE method. We just need a RACE kit.
oh and...I use pfx DNA polymerase from Invitrogen, but I do agree with you that even proofreading polymerase doesn't guarantee no mutations
You are absolutely right about the potential for viral sequences to evolve over passages and acquire mutations, but any time you are making a full-length clone, you have to make the concession that you are cloning a "snapshot" or an "average sequence" of a mixed viral population. Even within a single virus prep that you are cloning from, you likely have hundreds of thousands of genomes with point mutations from the consensus, which is what you see when you PCR, clone, and sequence. This will always be a risk you take when you do this work, but most of the time it doesn't become too much of an issue.
Second, you are definitely right that what you need first is a 5'/3' RACE kit to sequence your Leader and Trailer, but that will also give you a head start on the clone because you will end up with plasmid clones of your 5' and 3' ends. What I would recommend is finding some unique or Semi-unique restriction sites that you can identify, spaced out throughout the genome, and design PCR primers outisde those regions, so you have an overlapping set of PCR reactions. But instead of Re-PCR amping in sequential steps, just clone the PCR products into a subcloning vector with blue/white capability (any pGEM cut with EcoRV and phosphatased is my preference). Once you have those clones, sequence them to verify that you have what you want, and then you can sequentially cut each piece out and assemble it into the full lenth. Doing it this way, you shouldn't have to engineer primers with RE tails, you just use the natural RE sites in the genomic sequence. But yes, you may have to change strategies for each new strain if the RE sites are significantly different (as you will see from Refs below). If this is an issue, the PCR method may be a better bet, but I still think cloning fragments and assemling them is a more dependable strategy.
Despite having done this protocol, I don't have a really good reference of my own for assembly I can point you to, but just for a few examples:
Best of Luck.
thank you so much allynspear
you should also watch out for the Gibson assembly ...see this link!
I was told the overlapping sequences need to have 60C Tm. why is that?