phantom non specific amplification at high ct values - (Nov/13/2007 )
Hi everyone!
I was wondering if anybody could help me? I'm at my wits end with a real time PCR that will not behave. I'm running a Taqman based assay over a period of 40 cycles. For my target bug, it's wonderful. However, I'm now testing against genetic near neighbour organisms. From the genetic data available on line, there should be no cross reaction. Indeed when testing small values such as 10pg- 100fg, there is no problem at all. However, when I tested against more concentrated samples (5ng +), I get positive values at around 36 cycles (corresponding to 10fg of my target). My no template controls are clean each time and so it's not something in either the mix or something in the air. I've tried changing the annealing temperature (increasing all the way to 68oC) and changed the MgCl2 concentration to hopefully make the reaction more specific. However, this has been of no use. I've tried Southern Blotting of the extracts to see if the target is present, but this has proved unsuccessful. I thought that our samples were contaminated so I've got fresh DNA from a lab that has never isolated my target bug. However, this too is amplifying quite well. To make things worse, I've now tried a different target and am seeing the same problems 
.
I feel like I'm going round in circles and I know that the easiest way out is to cut at 35 cycles but I want to know what is going on!!! I've digested the DNA (with EcoRV) and used DMSO to see if it was a structural thing, but that has failed too! Can anyone explain what is going on or is it that the gods of PCR are angry with me!
Many thanks!
Stand free 1983
Okay so a little hand waving for you...
I have seen in the past something which I have called "template dependent primer dimers" ie: there is a primer dimer in samples that are negative for the target gene but contain DNA while there are no primer dimers in say the water control. (Thus the name) I cannot say that this is a well vetted phenomenon, but I can imagine a scenario where some non-specific sequence potentiates the interaction of two primers that do not interact in the no-template sample, it could be a sequence or it could just be that by adding additional molecules to the reaction you have set up a condition (perhaps concentration dependent) where primer primer interactions are more favored than they are in the absence of this DNA. It is possible that this is what you are seeing?
To test my theory you could add into the PCR a DNA that is non-specific (ie: an unrelated plasmid construct) and see if at high concentration of plasmid the interaction is produced. If the interaction is dependent on some sequence in the genome/sample you are using (more complex so I am presuming less likely but you know Murphy... 
) then the plasmid, even at very high concentrations, will not produce the high Ct amplification... so, this experiment wouldn't definitively rule out non-specific interactions with the sample but it may tell you if there is a sequence-independent/concentration-dependent non-specific interaction with other components of the PCR.
I have seen in the past something which I have called "template dependent primer dimers" ie: there is a primer dimer in samples that are negative for the target gene but contain DNA while there are no primer dimers in say the water control. (Thus the name) I cannot say that this is a well vetted phenomenon, but I can imagine a scenario where some non-specific sequence potentiates the interaction of two primers that do not interact in the no-template sample, it could be a sequence or it could just be that by adding additional molecules to the reaction you have set up a condition (perhaps concentration dependent) where primer primer interactions are more favored than they are in the absence of this DNA. It is possible that this is what you are seeing?
To test my theory you could add into the PCR a DNA that is non-specific (ie: an unrelated plasmid construct) and see if at high concentration of plasmid the interaction is produced. If the interaction is dependent on some sequence in the genome/sample you are using (more complex so I am presuming less likely but you know Murphy...
) then the plasmid, even at very high concentrations, will not produce the high Ct amplification... so, this experiment wouldn't definitively rule out non-specific interactions with the sample but it may tell you if there is a sequence-independent/concentration-dependent non-specific interaction with other components of the PCR.Dear Beccaf22,
you've certainly worked some magic! I tried adding titrations of three different mini-prep plasmid constructs and they too are amplifying (at the higher concentrations) although there is no sequence homology. Although this is not ideal, at least I can rest easy! However, does this mean that I should redesign the whole reaction around this target? If so, will this be hit or miss (I used software that said the primers/ probe were the best combination for this target)? Many thanks again. It's so nice to know that I'm not just an isolated case
!Best wishes
Stand Free
Sorry for the delay in getting back to you, I am really glad this worked for you! Very cool and helps to validate something I believed may happen but previously had no evidence for. Thank you!
As to your question, I think this means your reaction is good with the primers/probes/conditions you are using. Maybe you want to qualify with something like you must use amounts of input DNA less than 5ng (or even less than 1ng to be on the safe side) but with that qualification, (and assuming you are using samples where your target is detectable in these lower amounts of input) I think that you have a working assay! Congrats and good luck!
As to your question, I think this means your reaction is good with the primers/probes/conditions you are using. Maybe you want to qualify with something like you must use amounts of input DNA less than 5ng (or even less than 1ng to be on the safe side) but with that qualification, (and assuming you are using samples where your target is detectable in these lower amounts of input) I think that you have a working assay! Congrats and good luck!
"template dependent primer dimers", this's cool. I also met the same problem in my work. And i redesigned the primers and probe.
but how could primer dimers be detected in a TaqMan assay? there is no probe hydrolysis in primer dimers ..... ?
We have used the term "primer dimers" but this is being used to include all component interactions including primer-primer and probe-probe and with primers and probes...
but this should not make a difference. In a TaqMan Assay just the formation of primer-primers or probe-probes or primer-probes is not sufficient to generate a signal because there is no probe hydrolysis due to the 5´->3´ nuclease activity of the enzyme.