Nonspecific PCR problems with viral integration site

Go to the profile of kjoseph3278
Jul 24, 2017


I'm trying to amplify the host virus junction in human genomic dna of a specific provirus. The ratio of cells containing a provirus is very small (1 in 1000 cells contain a provirus).

I have determined a blend of 6 restriction enzymes that allow me to selectively shear human gDNA down to a mean of ~ 2,000 bp. The length of the provirus fragment I am trying to amplify before hitting the host junction is ~ 650bp, so I am trying to generate an amplicon that is between 650 bp and 2,650 bp-ish. (650bp of virus + distance to first cut site in human DNA).

I have screened these enzymes and am confident they do not cleave the provirus upstream of my nested PCR priming sites. Our viral primers are very specific and work very well for end point PCR (single genome sequencing) when paired with another nested viral primer set. However, since I am trying to amplify an integration site, I can’t ever expect to know the sequence on the human portion of the DNA.

I’ve been trying to utilize a dsDNA adapter that has the forward priming sites missing on the anti-sense strand, and the 3’ end of the anti-sense strand blocked with a C3 spacer so it cannot be extended. In addition, I have tried having a 5’P on the anti-sense strand of the linker and not having it (not having it helps prevent adapter dimer, and I don’t care if the anti-sense strand of the adapter ligates because it doesn’t contain the priming site sequence anyway. The goal is to have the internal proviral primer extend during PCR and regenerate a full length complementary sequence of the sense strand of the adapter, which will then allow priming of the adapter only on fragments that also contain the proviral DNA priming sites.

When I try this, I get a lot of nonspecific amplification, mostly smears by gel but on 12K labchip I often get specific sized bands. These are negative for the provirus by a very specific qPCR assay I have, which I use to screen amplicons for presence of provirus.

Is there a way I can uni-directionally ligate an adapter onto fragments that only contain my proviruses, since they are in such small numbers? OR a way I can block nonspecific amplification caused by the adapters? I've run PCR controls with various combinations of primers and i get amplification even with just adapter primer and nothing else. I perform a 0.6X SPRI purification prior to nested PCR which should clean out most, if not nearly all, of any 100 bp dimers I might be getting.

I’ve also tried hybridization pull-down of provirus using the Gnirke et al. method for exome sequencing, an can pulldown extremely small numbers of provirus efficiently, and detect them via qPCR. But once I have them on streptavidin beads, I cannot seem to get an adapter on their 5’ end for amplification. I've tried extending my RNA probes with polymerases but it doesn't seem to work. I've tried Ranger polymerase, Large Klenow, and E. coli DNA polymerase I. If i could get my RNA probes to extend and allow formation of a dsDNA end, I think I'd be golden.

Can anyone help?


Go to the profile of OldCloner
OldCloner 7 months ago

This is a tricky one. First, I could use some additional info on how you ligate the adapters- Do the enzymes in the 6-enzyme blend (6-EB) create blunt or sticky ends (both?). After you digest the genomic DNA with your 6-EB, do you polish the ends with Klenow to make them blunt? Are the adapters blunt? Or are you making adapters with sticky ends to match the termini that are produced by the digestion? Exactly what kind of DNA ligase are you using? (T4, T7 etc.) Are you making your own adapters by annealing two oligos?