Difficult PCR containing triplet repeats and Frt sites - (Jul/13/2013 )

Hi all.

I'm having a HUGE problem since I can't PCR a product no matter how hard I tried, and this has been driving me nuts for the past 2 months.

I'm trying to do gene targeting in human stem cells for a knockout mice. My gene of interest contains of triplet repeats (GAA) but the length of the repeats are different in both alleles. It's about 300 repeats in one allele and 450 repeats in another alleles. Therefore, a PCR that amplifies the GAA region would give 2 bands, one band containing the 300 repeats and the other the 450 repeats.
My goal is to target either band, using electroporation as a method, to replace one allele with a shortened GAA repeat length (around 10). my "targeting template"contains of a neomycin cassette that is flanked by Frt sites. so my template looks like this,

-----Frt----Neo-----Frt----short GAA repeats.

If the template correctly replaces the long GAA repeat allele, I should expect to see 2 bands, either
1. the 450 repeat band and the Neomycin/short GAA band
or
2. the 300 repeat band and the neomycin/short GAA band.

Here's the wierd thing. After antibiotic selection, I have 2 clones left, and when I do the PCR,
one only shows 1 band, the 450 repeat band
the other only shows 1 band, the 300 repeat band.

No matter how hard I try , the PCR result is always the same. I've tried 3 different polymerases (NEB Onetaq, Phusion, Takara) and they all show the same thing--only one band.

By doing RT-PCR to look for mRNA from this gene, the gene expression changes, and it is compatible with the assumption that the neomycin/short GAA band is inserted in that area. (a short GAA length will have a different gene expression with a long GAA length) However, I simply can't prove that by PCR.

My theory is the Frt-Neo-Frt cassette is a hard template to PCR, since Frt is palindromic by itself, it might have secondary structures. Also, the GC content for the Frt-Neo-Frt-short GAA is significantly higher than the long GAA band (57% versus 30%). Our PCR conditions were optimized for the long GAA band (It itself is already pretty hard to amplify and will only work under specific conditons). So it might not be optimal for amplifying the Frt-Neo-Frt band.

I've tried changing DMSO content (up to 3%), increasing melting time, etc. to no avail. Gradient PCR and changing template amount didnt work either.

Let me list a cndition that works for the long GAA band
(NEB Phusion polymerase)
98 degrees 5 mins
98 degrees 5 secs
70 degrees 15sec
72 degree 90 sec
72 degree 5 min

35 cycles
primer Tm: 68

Thanks for reading this very long narrative and I would be entirely grateful for any suggestions

Best,

J

I assume you have the Frt - neo - Frt short GAA cassette available as a plasmid or PCR product. You should check that it can be successfully amplified by your primers. Where do the primers bind? Have you tried lowering the annealing temperature? 70 is quite high, and I would definitely investigate lowering this to 60 or so.