Random chuck of nonsense in middle of PCR - (Feb/26/2015 )

Hi, i'm having some trouble cloning a gene into a plasmid and I was wondering if anyone would be able to help?

I've managed to troubleshoot my problem and pinned it down to some incorrect PCR.  I'll let you know what I did first of all:

• Double digest recipient plasmid, run on gel, purify, nanodrop.  
• Design primers which are have ~10 bases 5' homologous for my plasmid and 3' 18 bases homologous for my gene of interest.
• PCR of my gene of interest using 300ng of plasmid template and two different (two different reactions that is, not 2 in 1!) Taq polymerases with proofreading.  I used either 0.4 or 0.5 uM primer as recommended by each enzyme protocol.  Cycles were according to each enzyme protocol.
• Gel purification of my PCR products
• Homologous recombination of my gene into the linear plasmid - two separate reactions, one for each polymerase.
• Transforming into competent E. coli recommended by HR kit.
• Pick 3 colonies from each plate growth, grow up in broth, make glycerol stock, miniprep the rest.
• Send some of miniprep product away for sequencing.

I've just got my sequencing back and it looks like my PCR has gone really strange.  The first 14 bases of my gene are correct (14 out of 18 match with FOW primer), then I have 370 bases of nonsense that i've run an NCBI blast for and there is no known match, then my sequence resumes again from about 65 bases later and runs perfectly until the end of the gene.  So 65 bases go missing and are replaced with 370 of random junk - even cutting out a section that the primer was binding to.  My starting sequence and the point at which it resumed do have a small amount of homology (start = ATGGCCGCA, resume point = ATGCCTCC) but thats it.

Does anyone have any idea what went wrong with the PCR?  It's strange that its happened with two separate reactions.  One was an all-in-one mix and the other made up of separate components.

Thanks for reading

You are using FAR too much template in your PCR reaction. 300 ng of DNA is about 1000x too much, and will leave huge quantities of unchanged plasmid DNA in your PCR product. I suspect this is the source of your difficulty, but of course it could be something else. Also, 10 bp of homology sounds quite low for good recombination. What is the kit you are using?

phage434 on Thu Feb 26 21:14:35 2015 said:

You are using FAR too much template in your PCR reaction. 300 ng of DNA is about 1000x too much, and will leave huge quantities of unchanged plasmid DNA in your PCR product. I suspect this is the source of your difficulty, but of course it could be something else. Also, 10 bp of homology sounds quite low for good recombination. What is the kit you are using?

Hi, thanks for replying.

300ng is actually exactly 1000X too much, I got the wrong units!  I'm using 300 fg template (3 in 597 of 30ng/uL - I prefer to pipette as large quantity as I can to reduce error, probably just superstitious though!)

The kit is Clontech In-Fusion HD, I'm using the primer set that was recommended using the tool on the website.  I'm pretty confident the recombination is correct because sequencing shows the gene in the right position, its just this random error in the PCR. I've just realised i've been really dumb - when I ran the PCR product on the gel it did actually come up a little too high, about 200-300 bases too high (I'm more used to running protein gels which do (in my experience) often come up slightly off weight).  I think its clear that my PCR isn't amplifying for some reason, could you tell me what you'd recommend please?

My FOW primer (which is where the error is located) is quite GC right in the 3' end - 7 of the last 9 bases are GC, and overall the section for the gene is 72% GC - but there isn't much I can do about that because I need the whole gene from start to finish. I think I could design a primer that starts just before my gene on the donor plasmid, one that isn't so GC rich .. as long as ATG is in there to start and I don't leave too big a gap between promoter and gene it should be OK. Is that correct?

Or is there another way?  Perhaps if I increase the Mg²⁺ conc slightly, at the moment i'm just using supplied premix but I could add some MgCl₂ to it.  Because if the DNA is binding more tightly then maybe it will be more specific.

It sounds as if your forward primer is binding to an alternate site. This is strange, since normally shorter fragments are favored.  I would add 5% of a 5M betaine solution (Sigma B0300) to your PCR reaction, which helps a great deal in high-GC reactions. I would do a gradient PCR changing the annealing temperature (mostly higher).  When you look at the sequence of the "wrong" product, is the desired primer binding site present in the sequence data? Does it really match your primer sequence? Perhaps the sequence you based your primer design on was wrong.

Also, you really used 300 pg of DNA, not 300 fg, which would be 1000x less.

I'm really not having much luck with my units today ...

I'll try the Betaine, thanks.

So when I sequenced insert on the plasmid after recombination I had: few bases of plasmid -> restriction site -> first 14 bases of gene correctly (primer is 18 long) -> ~300 bases of nonsense -> gene resumes ~50 bases ahead at vaguely similar site -> rest of gene followed by restriction site and back to plasmid. Weird. I've checked the primer sequence 10s of times and redesigned it with several tools, they all give the same result (bar adding or removing a 3' base occationally), its definitely the right sequence for the gene.

What evidence do you have that the template sequence is what you think it is?

I took it from a miniprep of the e coli we use for overexpression -> purification of our protein of interest. Someone else from my group sequenced all of our stocks as part of housekeeping recently and everything came back good, but I'll ask him for a copy of the sequencing to check myself and see if I can identify the random sequence.