Strange PCR problem - (Sep/07/2013 )

I have a rather odd PCR problem that has plagued me for a while. I was wondering if any of you have encountered a similar problem and, if so, what have you done to fix it?

The problem:

I have constructed a synthetic gene library that contains both random and non-random sequences within the gene. The library was constructed from overlapping oligonucleotides using PCR to join segments of the gene library together. The final step of assembly is to amplify the full length library with oligos that bind to the extreme 5’ and 3’ ends and gel purify the full length product, which is approximately 600 bp. The full-length product is a mixture of sequences due to the random regions in the library.

Here’s the strange part. If I try to amplify the full-length, gel-purified DNA with the oligos that I used to create the full-length library in the first place, I get smaller distinct fragments…often, but not always, without any full length product showing up. I see a similar thing if I transcribe the library to RNA, then perform RT-PCR on the resulting mRNA: smaller distinct fragments that are the same size as those when I try to amplify the gel-purified DNA.

I have cloned and sequenced both the full-length DNA and some of the smaller products. It appears as if the downstream oligo is mis-priming at non-random sequences where there is  4 nt complementarity to the 3’ end of the primer. I am not getting mis-priming in the random regions. If I perform PCR on cloned DNA (which is a single sequence), I do not see this mis-priming and get a nice clean band of the correct size.

The PCR reactions are performed at a relatively high annealing temp (65°C), so I don’t understand why I would get preferential priming at the 4 nt sites when I’m using an oligo that is 35 nt long. My primers have been PAGE purified, so I am confident of their purity. I have also made an alternative, shorter oligo that does not contain the offending 4 nts on the 3’ end, but I still get mis-priming at other sites.

I am completely baffled as to why I can assemble the full-length gene library by PCR, yet when I try to re-amplify the purified assembled library with the same oligos, I usually only get smaller pieces back.

Any help or suggestions will be welcome. Thanks!

PCR preferentially amplilfies small fragments, so although your mixture may contain mostly longer fragments, the smaller ones amplify better, and you end up with more of the small fragments. Once this starts, it continues, since there are now many many more of the small fragments. I'd suggest raising your annealing temperature or going to a two-step cycle.

Thank you for your reply, but that does not resolve the issue. I think the most perplexing part of this problem is that I get mis-priming in non-random regions when I have a mixture of sequences (i.e. from the library pool), yet I do not get mis-priming from individual genes that have been cloned from the library. That doesn't make sense to me.