Hello colleagues,
I have one dubious pair of primers for a snail gene which efficiency is very unstable and can oscillate from 80 to 110% in different reactions (standards are not to blame, because other primers demonstrate stable efficiency with the same standards).
In one run I noticed that melt curves of concentrated and diluted standards are slightly different, the peak in diluted one is shifted about 1-2 degrees to the left. It can't be primer dimers since in the negative control there's no such peak (there is very little dimer peak sometimes, but it's about 6 degrees to the left from target peak). Moreover, there was one particuarly bad sample for which 3 replicates had 3 completely different melt curves: the first was normal, the second with only dimer peak (probably I just didn't add the matrix to this well), and the third had two peaks: one was dimer and another was this unusual slightly left-shifted target peak. I attach here the pictures of positive and negative controls, of bad sample replicates ("sample 1") and normal sample replicates ("sample 2").
Is it possible for melt curve to be variative depending on sample concentration, or is it alternative amplification product? My target should be about 150 bp, and dimers (if present) about 50 bp.
Difficult to answer without more detail, but there are a number of possible explanations.
1) small differences in melt peak are not uncommon: most default melt curve programs use steps of 0.5-1 degree and usually hold at that temperature for a very brief time (just long enough to collect fluorescence data). I would generally accept variation of +/- 1 degree as simply measurement noise: the melt peak isn't changing, but the accuracy of your measurement may be, in essence.
2) salt contributions: more concentrate standard stocks may bring along a higher level of contaminating salts: these salts may be of minimal concern to reaction efficiency etc, but can alter the measured melting point of your amplicon.
3) point mutations/heterozygosity: it is not impossible that your region of interest contains a SNP that alters melting properties (a G>T switch reduces strength of base interaction, for example). If the majority of your population has one allele this will preferentially be amplified and your melt curve will reflect this. If you dilute your samples down you may reach a point where (through random stochastic partitioning) some wells contain only the rare allele, and the melt curve may this reflect this. For your triplicates in the second image, this may be the case (with the no-amplification well simply containing no template at all: a common problem at high dilutions).
Furthermore, if you amplify both alleles to a similar level, the melt curve observed will appear to be a combination of the two (so somewhere in the middle).
Since I don't know how your prepare your standards (or what your standards are) I cannot really advise further.
Have you run these out on a gel to see what size they are? You could also clone the products and sequence them, potentially.
The meelting temperature is influenced by many factors one of them is the dilution buffer and the final salt concentration. Your diluted samples has a different ratio sample / buffer from the original sample. Your finding is quite usual.
best regards
Difficult to answer without more detail, but there are a number of possible explanations.
1) small differences in melt peak are not uncommon: most default melt curve programs use steps of 0.5-1 degree and usually hold at that temperature for a very brief time (just long enough to collect fluorescence data). I would generally accept variation of +/- 1 degree as simply measurement noise: the melt peak isn't changing, but the accuracy of your measurement may be, in essence.
2) salt contributions: more concentrate standard stocks may bring along a higher level of contaminating salts: these salts may be of minimal concern to reaction efficiency etc, but can alter the measured melting point of your amplicon.
3) point mutations/heterozygosity: it is not impossible that your region of interest contains a SNP that alters melting properties (a G>T switch reduces strength of base interaction, for example). If the majority of your population has one allele this will preferentially be amplified and your melt curve will reflect this. If you dilute your samples down you may reach a point where (through random stochastic partitioning) some wells contain only the rare allele, and the melt curve may this reflect this. For your triplicates in the second image, this may be the case (with the no-amplification well simply containing no template at all: a common problem at high dilutions).
Furthermore, if you amplify both alleles to a similar level, the melt curve observed will appear to be a combination of the two (so somewhere in the middle).
Since I don't know how your prepare your standards (or what your standards are) I cannot really advise further.
Have you run these out on a gel to see what size they are? You could also clone the products and sequence them, potentially.
Thank you everybody.
Salt concentrations are indeed a little different in different standards: I prepare serial dilutions of reverse transcripltion product with MQ water, then add 2 ul of each dilution to 10 ul master mix.
Unfortunately I can't now check these particular samples on gel, but I checked the product of these primers before.
The standards are prepared from RNA extracted from a mixture of about 10 snail brains, so they possibly contain more different alleles. The samples are from individual snail brains.
I reverted, amplified, cloned and sequenced the target product myself (using one snail) before designing primers. But since there is no RefSeq-quality sequence, of course there may be SNPs I don't know about. I don't even know where exons are on my snail mRNA! I remember now that once I did double-check the sequence of RT-PCR product and there was one letter mismatch in the middle of it compared to the initial sequence that I cloned earlier.
It also bothers me that the standards and samples can be a little degraded - I aliquoted the standards and store all cDNA at -70, but I noticed that Ct for the same standard is gradually geting higher in later experiments (in a year it increased for 1-1,5 cycles for all primers I use).
- Badges
- Science method