Protocol Online logo
Top : Forum Archives: : Real-Time PCR

Strange results with beta actin melt curve - Please Help... (Apr/15/2006 )


I am getting strange results with a primer for beta actin that I received from a collaborator of mine. He assured me that the primer has been properly validated, and it worked fine. However the last time I did a run, I noticed something strange with the melt curve for my serial dilutions.

The first two (highest concentrations of cDNA) produced two products during the amplification, as verified through the melt curve i.e. two peaks, one where I expected beta actin to be, and the other past (higher than) the beta actin melt temperature (it was about 92, vs. 82 or so for beta actin). All the rest of the dilutions produced the single expected product. For some reason though, this problem has not happened before in other runs that I have done with this primer.

This was a trial run that I was doing just before starting to work on some irreplaceable samples. Furthermore, my PI needs the data from these samples by June.

One of the PI's that I am collaborating with told me that I need to characterize the second product and find out what exactly the other product is. While I wouldn't mind doing this in theory, due to my time limit, I am considering trying to get another primer sequence to work with.

I have considered using GAPDH (which I have available as well) but I had heard that estrogen treatment (that was used on the rat tissue) does effect GAPDH, whereas beta actin is unaffected by estrogen.

So I have three questions:

1. What could possibly be causing the second peak/product on my melt curve? I could redo the melt curve, but my fear is that I might get two products when I start working with my experimental samples.

2. Why is it only coming up with the first two concentrations? Could I have some contamination that I am diluting away?

3. Does anyone know of a good beta actin sequence (i.e. previously validated and published)? Either the sequence itself or a link to one that is commercially available and usable with QPCR. I had looked a few of them online, and all the ones that I found had products around 280-310 bp long, which would be too inefficient for QPCR. All of my other products are between 90-120 bp.

I would ask the person who gave me the sequence, but he left to go to med school in another country in 2004. I don’t know how to contact him.



I just spoke to my PI, and we decided that I am going to try to replicate the run. This will eliminate the chances that it was just a bad batch of cDNA.

I will post the results when I have them.


Pure speculation, but could your sample be heavily loaded with genomic DNA? Such DNA might have such a high melting temp as 92C, I suppose.
Also, have you examined your PCR products on gel? Might give a clue.
Finally, if your second amplicon has a normal length (less than 400bp, say), it would need to be GC-rich to have such a high Tm.


I have not yet run it out on a gel, but I will do so. Getting an idea of the lenght of the products will no doubt help me.

As for genomic DNA, I used Trizol for the extraction. I was told that the Trizol procedure takes care of separating genomic DNA

As for the G-C rich region, that makes perfect sense.