Daniel,

As the others have commented this is a common yet poorly reported issue and one that is most certainly frustrating. We have attempted to use flourescent dyes for accurate determination of template input into qPCRs but abandoned this due to very similar issues that you mention. We have also seen that there appears to be a batch to batch affect of the dye preparations (however convincing the suppliers of this is another problem!).

A couple of suggestions. We generally find that if we do a dual concentration measurement using a UV-spec measurement (we routinely use a nanodrop instrument) and your flourescent system AND you get good correlation of the concentration measurements you can have confidence in your calculations. As we use the Qubit system a significant proportion of the issues over linearity of the measurements is taken care of by the instrumentation and the calibrations you need to do with this prioor to measurements of unknowns (in other words if the 'system' is mis behaving the instrument rejects the standards and requires you to re run these). However we have seen a lot of drift in measurements with incubation times.

As the others have commented I suspect what is happening is a snap shot of the binding kinetics of the dye to the sample. I also suspect that you may also be seeing both dye and DNA interacting with the surfaces of your tubes/plates. With low concentration DNA samples the binding of samples to the surface of your plasticware is a major problem and moving to low-bind tubes may assist.

Finally I think you have moved to the solution I was going to suggest in that you standardise the incubation times prior to measurement dependandt on the linear range that you wish to measure.

I hope this is useful....

Daniel, I noticed some major discrepancies myself as well with picoGreen vs. raw spec. gDNA [bacterial gDNA] measurements... this is indeed troubling. Re-do everything you have done until you figure this out - I find it a major problem as well.

This is of massive concern when it comes to pre-diluting and then loading an "equal" amount of DNA sample for qPCR purposes. I really appreciate you bringing this point up as it impacts many experiments predicated on believing that an accurate pre-assessment has been made with fluorometric assays which purport to have precisely-measured template material. It is maddening.

Daniel, what you see could be slow rearrangements of the dye on the DNA or alternatively the dye going from being stacked onto itself (dimers or higher) in solution and then binding to DNA. We have plenty examples of rearrangements of molecules on DNA that could take hours/days. This might not be the correct answer to your problem but are palusible scenarios. To avoid this you need to equilibrate your DNA-dye-samples for a longer time. One problem with waiting longer could be that dyes sometimes also have a tendency to stick to surfaces with time. Good luck!

Fluorescence increased with the incubation time for all concentrations (0.02 ~ 2 ng/ul). The incubation time did affect linearity of standard curve. As time goes (5, 10, 20, 30, 45 min), the linearity become better. I got a perfect standard curve (0.02~ 1.0 ng/ul) after incubation of 30 mins.

I found incubation time also affected the linear range. For example, 10 or 15 min is enough to get a good line for 0.02~0.5 ng/ul and 30 mins is enough for 0.02~ 1.0 ng/ul but not enough for 2.0 ng/ul. Longer incubation time seems to improve the linearity for 0.02 ~2.0 ng/ul, but not as perfect as the narrow range.

Daniel,

As the others have commented this is a common yet poorly reported issue and one that is most certainly frustrating. We have attempted to use flourescent dyes for accurate determination of template input into qPCRs but abandoned this due to very similar issues that you mention. We have also seen that there appears to be a batch to batch affect of the dye preparations (however convincing the suppliers of this is another problem!).

A couple of suggestions. We generally find that if we do a dual concentration measurement using a UV-spec measurement (we routinely use a nanodrop instrument) and your flourescent system AND you get good correlation of the concentration measurements you can have confidence in your calculations. As we use the Qubit system a significant proportion of the issues over linearity of the measurements is taken care of by the instrumentation and the calibrations you need to do with this prioor to measurements of unknowns (in other words if the 'system' is mis behaving the instrument rejects the standards and requires you to re run these). However we have seen a lot of drift in measurements with incubation times.

As the others have commented I suspect what is happening is a snap shot of the binding kinetics of the dye to the sample. I also suspect that you may also be seeing both dye and DNA interacting with the surfaces of your tubes/plates. With low concentration DNA samples the binding of samples to the surface of your plasticware is a major problem and moving to low-bind tubes may assist.

Finally I think you have moved to the solution I was going to suggest in that you standardise the incubation times prior to measurement dependandt on the linear range that you wish to measure.

I hope this is useful....

If binding to the tube is a problem, could adding BSA be a solution? We have used this technique when preparing samples for qualitative PCR.

EDIT: That should have been QUANTITATIVE PCR, of course.

Yes - BSA can assist in blocking the non-specific binding of the DNA to the tube surfaces. Probably the easiest thing to do if the budgets are tight is to pop a beaker of tubes into a vaccuum jar along with a small beaker of dichlorosilane and apply some vaccuum for 30 mins to about an hour to coat the tubes with a layer of silane. If budgets arent too tight both Axygen and Eppendorf supply low bind tubes - the eppendorf tubes are better though considerably more pricey

Well, the GelGreen protocol lists 30 min incubation for Agarose-gel staining. Clearly, that's quite a different application, but may give a hint. Although the picoGreen protocol (a.k.a. Quant-iT reagent) only lists 2-5 min incubation.

Biotium, the company that sells GelGreen, lists among their products a dye called AccuBlue for a plate reader test for nucleic acid quantification (maybe that has a better behavior?), with also only 5 min incubation (ok, unlikely).

But for sure the problem is common - our Qubit results drift was well (not very much) - which is a known effect. We are using TE buffer as reference and DNA in TE.

Preventing from light is not that critical for most of the new DNA dyes (GelGreen, PicoGreen, GelRed, etc), as they are made much more light-stable than old dyes. Also, it could not explain why fluorescence increase over the time. Binding to the surface might be a problem, but it is also hard to explain why fluorescence increase and why for some concentration the fluorescence increase more. I think salinity and pH of the buffer have a critical impact. What buffer system do you use for the DNA fluorometric measurement?

Dear all

I am looking for a reference on this topic. I want to quantify FAM labelled DNA attached to a surface by measuring the fluorescence.