How Bisulfite Pyrosequencing Works

Written by: Kirsten Hogg

last updated: October 7, 2021

Bisulfite pyrosequencing is becoming a routine technique in molecular biology labs as a method to precisely measure DNA methylation levels right down to the single base. The technique allows for detailed and high resolution analysis of DNA methylation at specific genomic regions.

How to detect the 5th base?

Methylation of any of the four nucleotides that make up the DNA strand (adenine [A], cytosine [C], guanine [G] and thymine [T]) can occur in various dinucleotide pairs. By far, the most common is methylation of the cytosine residue (mC) in the context of a CpG dinucleotide (adjacent C and G bases linked by a phosphate bond); aka the 5th base. But how do we distinguish this 5th base (methylated C) from unmethylated C?

If an ‘epigenetic mark’ is converted into a ‘genetic mark’ then this can be measured by DNA sequencing.

A DNA sample is incubated with sodium bisulfite, a chemical compound that converts unmethylated C into U (uracil), but leaves methylated C unmodified (bisulfite conversion can be performed using commercial kits).  This generates a DNA strand that is differentiable upon subsequent sequencing.

Amplify the bisulfite-converted DNA

Now that we have converted our DNA with bisulfite, it’s time to design PCR primers that are specific to the modified DNA strand. Any C not followed by a G is now treated as a T (since uracil will be converted to thymine during PCR) and one primer (forward or reverse) is tagged at the 5’ end with a biotin label. During pre-pyrosequencing set-up steps, streptavidin beads are combined with the PCR product to select out the biotin labeled product, and the DNA is denatured to produce single stranded molecules for sequencing. Ideally PCR products should be 150-250 base pairs in size and primers should not overlap CpG sites as this could introduce PCR bias. Free online software such as MethPrimer is useful for methylation-specific primer design.

Let there be (pyro) light   

What distinguishes pyrosequencing from other traditional DNA sequencing methods such as Sanger sequencing, is the use of a chemical light reaction to detect the incorporation of deoxynucleotide triphosphates (dNTPs) into the synthesizing strand. It goes something like this: 1)  The sequencing primer hybridizes to single stranded, biotin labeled DNA molecules. DNA polymerase synthesizes the strand in the 5’-3’ direction: dNTPs (dATP, dCTP, dGTP, dTTP) are added in a specific order. 2)  If complementary, the dNTP is incorporated into the strand and a molecule of pyrophosphate (PPi) is released. 3)  In the presence of 5’phosphosulpfate (APS), ATP sulphurylase converts PPi to ATP. 4)  ATP drives a light reaction through which luciferin is converted to oxyluciferin by luciferase. 5)  The light generated is directly proportional to the amount of ATP present. Light is captured by a camera and the signal converted into a peak which is visualized graphically. 6)  Unincorporated dNTPs and left over ATP are degraded by apyrase, so that the next dNTP may be dispensed.
At each CpG site to be measured, a dTTP and dCTP are added sequentially. The relative peak height of T versus C at that CpG site is measured to give percent methylation.  Bear in mind that at each DNA strand the CpG site can only be methylated or unmethylated (0 or 100%). So the percent methylation is the proportion of methylation at that site in DNA from all the cells in the sample.

Some technical tips:

  1. Include high and low methylated DNA samples as controls for assay accuracy (these are commercially available), in addition to the usual no template, and water only controls.
  2. Randomize samples across multiple runs. There may be inter-plate technical noise of up to 5%, so ensure that control and case samples are not isolated on different plates.
For more tips and advice on the design and analysis of DNA methylation data have a look at my previous bitesize article ‘Getting the most out of your human DNA methylation studies’

Kirsten gained a PhD in Reproductive Biology from the University of Edinburgh.

More 'Genomics and Epigenetics' articles