## Relative or absolute quantification

Relative quantification is discussed by Pfaffl in Chapter 3. In its simplest form, this is the comparison of the amplicon of interest to a control ampli-con. For a mtDNA assay, either the mtDNA region (as above) or nDNA can act as a control. When selecting a nDNA region it is useful to select a single copy gene to allow for your experimental data to be easily adjusted to a value per nuclear gene copy.

Relative quantification of an mtDNA reference gene/region to a sample mtDNA gene/region allows for the calculation of the percentage of one species to another. This in essence represents the percentage heteroplasmy of a mtDNA sample. For the design of such experiments it is important that both amplification reactions are linear over the expected experimental range of DNA concentrations. Ideally, when analyzed over a DNA dilution series, both reactions should generate equations for the line that are the same. From the slope (gradient) of this line it is possible to show that the efficiency of amplification for both reactions is the same. The intercept value of the equation shows whether for the same number of copies the reactions generate the same Ct. Perhaps the more important of these two pieces of data is the slope, or PCR amplification efficiency. Assuming that a PCR reaction is 100% efficient, each molecule should be copied once each cycle, leading to a doubling of the copy number each cycle. Over a log dilution series this equates to a 3.3 cycle difference between log dilutions. The slope of the line through the dilution series should therefore be as close to -3.3 as possible. Slopes less than -3.4 indicate a reaction that is less than 100% efficient. Conversely gradients greater than -3.2 indicate an efficiency of more than 100%, most likely due to primer-dimers formation or non-specific product formation. Consistency in the slope indicates that the reaction is optimized, slope values between -3.2 and -3.4, show a well optimized reaction. If the two reactions have the same slope this shows that, over the linear range, the two reactions are consistent. If the intercept is the same, then they are consistently the same. If the intercept is different but the slope the same, and this is a reproducible difference, the two reactions can still be compared and this difference can be accounted for mathematically in the post-run data analysis.

Comparing the quantity of mtDNA to a nDNA reference allows correction for the difference in the number of mononuclear cells present in each sample. As we are not performing a gene expression study and simply a DNA investigation, it does not matter if there is more than one copy of the gene, or if it is differentially expressed in different cells. Any region or gene can therefore be used. The same numbers of copies of each piece of nDNA are present in every cell, so the fluorescent signal will be consistent between samples containing the same number of cells. This is true in most circumstances, however not in cancer-derived cell lines, for example. Selection of any region will allow the normalization of your mtDNA data. However, it is more beneficial and we would recommend the selection of a single copy gene as the nDNA reference target. This allows for the quantity of mtDNA in the sample to be expressed as copies per diploid genome without the need for an mtDNA copy number standard curve. A mtDNA : nDNA ratio in this manner is of most use when investigating DNA extracted from bloods or homogenate tissue samples, where the number of cells present is unknown. If the study is on individual micro-dissected cells, then comparison to a nuclear target may not be necessary.

Unlike relative quantification methods, the absolute method requires a known standard curve to generate meaningful data. This means, therefore, that a series of standards needs to be generated. There are several means for doing this and considerations that need to be taken into account for each.