Bisulfite-PCR followed by restriction is a rapid and semi-quantitative method of analyzing DNA methylation. The PCR products are also suitable for either direct sequencing or cloning and sequencing. The most important step here is primer selection.

Generally, we prefer to be as close to the transcription start site of the gene as possible. This is also dictated by the availability of restriction enzyme sites. Here is a general strategy:

• Identify the region of interest/CpG island. Make a restriction map of the area (all enzymes). This is the unconverted map.

• Copy sequence and paste in a text editor.

• Convert all C to T except for CG. First convert all CG to XG. Then convert all C to T. Then convert all X to C. Et voila! Make a restriction map of this converted sequence (methylated map).

• Convert all remaining C to T. Make a restriction map of this converted sequence (unmethylated map).

• Find restriction enzyme sites that are unique to the methylated map (not in the unconverted or unmethylated map). These are the best to use. If none is available, find restriction sites that are present in the methylated map but absent in the unmethylated map.

• Design primers for PCR, aiming for a 200 bp region that contains one or several usable enzymes, and that is close to the transcription start. Use the methylated/converted sequence to design primers, but try to avoid having C in the sense primer or G in the anti-sense primer. If you don't find suitable primers, you can include up to one C in each primer, but make sure they are in the 5' end of the primer and, when you synthesize them, synthesize either C or T (sense strand) and either G or A (antisense strand) instead of simply C or G.

• Amplify using the calculated annealing temperature. We use a special buffer that, in our hands, works better than standard buffers. Optimize the reaction as you would any PCR.

• To study methylation, restrict the PCR product with the enzyme of interest (usually 2 hours are enough) and run on a high resolution gel (3% agarose or 6-8% acrylamide). Using this strategy, anything that cuts down reflects methylation.

Sequencing Bisulfite-PCR products gives you more information than restriction. You can  clone the PCR products (TA cloning, for example) and sequence multiple clones (automated sequencing works well here). Any C in the sequence reflects methylation (or, rarely, an artifact!). Alternatively, you can use direct sequencing: Design nested primers that include sequencing primers, PCR using the nested primers first, then sequence.