dilution dependent melt temp variation

Go to the profile of molbionerd
Mar 27, 2017

Hello everyone,

I'm new to the forum and have done quite a few qPCR experiments in my career. I am here today though, because I have never seen something like this before and don't know how to interpret.

I am doing a primer optimization for a Sybr green based qPCR experiment. Ran a dilution series of my cDNA with Rox as a reference dye and for detecting product. My primers are the same in each well. I was looking at my melt curve analysis with my run to make sure I was getting only a single product during the amplification. I have attached an image of it.

Each peak represents the average of technical duplicates in the dilution series. The thing that is particularly weird is that the melt temp appears to be dependent on the dilution in the series. So in the picture the far left, yellowish asterisk, peak is the 1 x10^-4 dilution, gray diamonds is 10^-3, green triangles is 10^-2, red squares is 10^-1, and blue circles is 10^-0. Each sample is made from the same starting cDNA. The amplification curves are good (as far as their shape and when they appear, mid-20s for the first sample). While it is not a perfect effeciency (approximately 140%, R-squared of .92) it didn't seem that major other products are being detected as their is a single point within each duplicate.

Has anyone ever seen anything like this? Any insight on this issue would be much appreciated. And if there is any other information that you need to help analyze, just let me know.


Go to the profile of Roberto Rosati
Roberto Rosati 6 months ago

I took the liberty of adding the image inline as jpg, instead of a downloadable .zip (sorry if it was too big in these last 5 minutes; I've scaled it down).

Go to the profile of Roberto Rosati
Roberto Rosati 6 months ago

Hi... It never happened in my lab. But you comment on efficiency being 140%, wouldn't you consider this suspiciously high, as in your template solution carrying possible PCR inhibitors?
It's funny, these are very sharp bands and don't look like what a smear would generate. Yet, you can't possibly be generating 5 different products at 5 different concentrations, or well, I'd be very amazed to see something of the like happening.
Have you ran these reactions on a gel? Sometimes it's the best option to shed light into complex situations, and with SYBR, running a gel at least once to make sure that there's no unspecific amplification is the safest way.