I am getting crazy about this. I am having less than perfect melting curves, with different primers. The funny thing is that while primers give a main peak at different temperatures, as it is to be expected, the shoulders associated to the main peaks appear always at the same temperature, around 71. I tried with a second different master-mix, and I got even taller shoulders, still melting at a constant temperature, this time always at 76. NTC are clear. Gel 2% does not shows any unspecific band, but amplicon is 74, so it could be difficult to resolve 2 distinct bands.
The "shoulder" on the peak is likely due to a region of your PCR product that has a lower GC content than the rest. As long as the height of the main peaks are distinct from each other, you can use your assay. I used to do a lot of melt-curve based genotyping of T-DNA transformed Arabidopsis plants and this sort of result was very common.
Thank you Katie, it is possible that those shoulders are lower GC content which melt a lower temperature, however I am talking about a number of like 6 targets...6 different genes... and it seems strange that when I designed their respective primers I got amplicons which melt in 2 stages in all of the cases, and that the first stage, the shoulder, it melt always at the same temperature.... I am puzzled
- Agree with Katie: This is often due to variation in local G/C and A/T content in your target region and consequent secondary structure
- You could try designing new primers to a different target region but in truth your signal to noise is strong and specific and as you say you see nothing non specific on an agarose gel
- Therefore do not worry too much about technicalities and go ahead with qPCR calculations: I think people get too hung up on shoulders on strong peaks: Your data if 'fit for purpose'
Thanks Laurence, I suppose the take away lesson is: Melting curve and control are ok, go ahead; melting curve not perfect, run a gel and if gel shows no problem, and signal to noise is strong, go ahead too.
- In essence Yet:
- Sometimes you get perfect melt curves sometimes you dont; Its idiosynctatic and as mentioned by Katie Target/sequence dependent and therefore in some ways an inherent limitation of sybr green qPCR
- But that isn't to say with strong specific melt curves and lack of obvious secondary bands on gels you cannot use the data; On the contrary you can
- I routinely add 5% DMSO to my sybr green qPCr reactions and I beleive this helps
Are you sure of the Tm melting and the concentration of the primer?
I suggest to perform the Tm in the PCR (with a gradient temperature) and after pass on the qPCR. In case try the elongation step at 76°C.
How was your thermoprofile of qPCR?
Giorgia, I did the gradient before, all targets were running together as the best Ta for the primers was 60 or 61. The profile was 10min @95 initial denaturation, then 40 cycles 95-61-72,each for 15 sec, reading at 72 after elongation. Then directly (without a 95C step of denaturation) 65 to 95 of melting curve. Interesting the elongation a 76, could you explain me what would be the effect? I know you can set the reading higher, to melt unspecific product before the reading
At 76°C is possible to avoid dimer primer.
Do you try the same conditions but load less gDNA?
In the gel, did you see the smear in the samples?
Is possible to see the melting curve of the NTC alone?
yes, I run dilution curves, and nothing changed in term of shape, from undiluted to 10.000 folder dilution, no smear in the gel. I have the NTC at work, I can post it tomorrow but it is completely flat
For the NTC I want to see the melting curve and no the amplification plot.
I added complete curves, including NTC, for the RPLO reference gene. The "shoulder" here is even more evident, I have used here a different MasterMix, hence slightly different melting temperatures
I updated with different results from gradient run and dilution series run. In dilution curve I excluded the undiluted sample, as was giving some inhibition and pushing efficiency to 110
- Badges
- Science method