How to Optimize qPCR Analysis Using a Melt Curve

When choosing a qPCR procedure there are two basic methods: the use of probes or intercalating dyes.  The use of intercalating dyes such as SYBR® Green only uses primers and binds to any double stranded DNA.  The dye absorbs blue light and emits green, which is then detected by the instrument.  The use of probes and primers introduces a third sequence and adds specificity to the reaction.  Although the use of intercalating dyes generates signal from non specific PCR products and primer dimmers, it is more widely used because it is far more affordable and can be used with a wide range of primers without the need of a specific probe per assay.  The use of a melt curve and controls can be implemented in quality control procedures to insure proficiency of quantitative PCR experiments using intercalating dyes.


 A melt curve is used after the amplification cycles have been completed.  The temperature is incrementally increased usually around 0.5°C per cycle starting at 60 to 65°C.  As the temperature is increased, the fluorescence will gradually decrease evenly as the dye is pulled off the double stranded DNA.  There will be a sharp drop off of fluorescence when dissociation of the double stranded DNA occurs.  This slope is recorded as Tm, or melting temperature.  It is interesting to note that a melt curve will not work with qPCR experiments that use probes because the initial fluorescence recorded by the instrument is not due to the fluorophore binding to the amplicon, but rather the probe being cleaved after amplification.  Typically, the presence of one Tm indicates specific amplification. However, further investigation is required to validate this assumption.


The presence of more than one Tm is not exclusively the result of multiple amplifications.  The best way to validate the results is to perform an agarose gel analysis.  If only one band is visible on the gel but there is more than one Tm, other variables can explain this contradicting result.  Typically, scientists classify the melt curve producing two stages: double stranded DNA and single stranded DNA.  If this were true, there would always be one distinct Tm.  However, some amplicons will produce multiple melt events depending on nucleotide patterns.  A/T dense regions of an amplicon will dissociate at a lower temperature than G/C dense regions because they are less stable.  Therefore, the melting temperature does not occur at one specific temperature.


There are software programs that serve as an alternative to confirming selective amplification on a gel. UMelt℠, a free melting curve prediction software developed by The University of Utah, uses algorithms to predict the shape of melting curves and the complicated melting transitions that occur during dissociation.  The interface provides a textbox for entering amplicon sequence and pull down menus to enter other concentrations that may affect dissociation.  The scientists at The University of Utah have found that the software is extremely reliable in predicting the shape and number of peaks although the melting temperature varies.  However, the melting temperature is irrelevant.  The goal of the software is to compare only the shape and the number of melting events to the predicted results.

If the melting curve has more than one Tm that is validated by either gel analysis or inconsistent with prediction software, there are troubleshooting steps a scientist can explore.  If there is a primer dimmer, decreasing the primer concentration or increasing the annealing temperature could eradicate the issue.  Multiple amplification points outside of primer dimmers are usually indicative of contamination.  A non template control (NTC) should always be included in qPCR experiments.  It is common for amplification of a NTC to occur, but melt curve analysis could reveal that the Tm is consistent with what would be expected for with a primer dimmer.  A negative reverse transcriptase control (-RT) will monitor genomic DNA contamination when the target is cDNA.  DNase treatment of cDNA can also reduce genomic contamination.  When working with several samples, it is expected to see a difference in Tm of up to two degrees Celsius.