optimisation of reaction - extra peak in melting curve - (Aug/13/2007 )

I have been trying to amplify a piece of cDNA from a virus gene but I keep getting a 2 nd peak in my melting curve (see attached file). I have tried varying the Mg concentration from 1 - 5 mM but this does not alter the peak. I have varied the annealing temperature from 45 oC - 62 oC but the 2nd peak remains. My primer concentration is at 0.4 uM, but it is unlikely to be primer dimer as there is no peak in my negative control. Is there anything else I can do?

---

I often see multiple leaks like that.

It is my understanding that the peak comes from the SYBR that is released as the strands melt open. I have wondered if the second peak might simply be a portion of the double-strand that is melting earlier, and the main peak represents when the rest of the strand melts open. I don't know if this is the case, though, but if it is it should be a constant fraction of the main peak.

I'll be curious if anyone has any other ideas.

---

The Tm for all PCR products from a particular primer pair should always be the same and there should be one peak only, so I am not sure if I agree that it is a portion of the double strand coming apart first, (not saying that the other comment is wrong, just that I am thinking something different and I don't understand it kinetically)
The curve you look at is a negative first derivative of relative fluorescence versus time, so it goes up with folding, peaks at Tm, and down with unfolding.
The question is do you do RT-real time or real time straight from your cDNA's? Because if you do a two step reaction, (real time PCR directly from cDNA's no reverse transcriptase) , your negative control is a rxn mix with no template .... and you are getting the second peak only in reactions where there is template, then it is probably a non specific amplification product.
Tm depends on base content, but also size for short molecules. Have you run your reaction products out on an agarose gel? If not you should.
If you do a single step RT-real time PCR and your neg. control is leaving out RT, this is probably not the case.

Hope this helps.

---

avalon:
Try to check it on a agarose gel with EtBr, you will see if it's a dimer or what.

---

After running the pcr product on a gel I can only see one band, and no primer dimer - even after overexposing the gel to show up any weak bands which might be there. So It looks like there is only one product, but I guess if there were 2 products which were very similar in size, the bands might not show up as separate bands on a gel - eg its an RNA virus so any changes in sequence might lead to similar sized products????? Does anyone think thats a possibility? The aim of my experiment is to monitor appearance of viral RNA over time - if the gel is only showing one product but the melting curve shows two do I need to start from scratch with new primers?

---

I still believe that your product is melting in two bits. This happens in DGGE, for example.

If you have an AT rich segment in your product, and a GC rich segment in another part of your product, they will melt at different temperatures.

I routinesly see, with some amplicons, that the Tm can be different depending on the specific sequence that I am amplifying. Again, a GC rich amplicon will melt differently than an AT rich amplicon.

Can you post a picture of some of your melt profiles? Preferably with different strengths of amplification (ie different peak heights between samples.) I am curious if the secondary peak is always a constant fraction of the primary peak.

---

When I performed a standard curve using decresing amounts of cDNA the 1st peak became larger in comparison the the second.....

---