Protocol Online logo
Top : Forum Archives: : DNA Methylation, Histone and Chromatin Study

Methylation difficulties - (Oct/19/2004 )

HI , I am new to this group and would really appreciate some help with a couple of the methylation problems that I have encountered.

The first lies with the reproducability of the bisulfite modification process. I am currently using the original sodium bisulphite protocol to modify tissue and body fluids. I have since analysed over 500 samples by MSP with published primers designed against 6 different genes. However I have found that whilst the data is reproducible between MSP (PCR performed in triplicate) its is not consistent between bisulfite modification batches eg a sample of DNA is modified in two seperate batches performed a day apart. The DNA sample is methylation negative for one modification batch (having the same MSP result 3 times) whilst the second batch of that DNA is meth negative. Protocol conditions are identical between the modification batches and unmethylated primers show that the DNA has been modified in both batchs of DNA. So I am not sure where the problem lies. Has anyone else encountered this problem?

Secondly I am also trying to identify novel genes that may be aberrantly method in my cancer samples using AP-PCR. I have two genes that are potentially of interest however the Sequence from the fragments that I isolated is localised in a transcribed but untranslated exon (first exon). It is contained within a large CpG island as determined by CpG plot. Is this a real result?

Thanks for your help in advance.


Qustion 1.

Two things may contribute to the inconsistence between batches of modification, one is loss of DNA during purification steps and thus different amount of template may give different results because the amplifying efficiency of M and U pair differ; the other is incomplete modification which is not uncommon and will cause biased amplification.

Question 2.

Some CpG islands extend from the promoter region well into 1st exon, methylation of which has been reported in a lot of studies to be responsible for gene silencing. So what you have found could be potential methylation hot spots.


Thanks for your reply.

I had thought that the efficiency of the modification process may be the problem. I am performing the modification at 55C for 16 hrs (to maximise conversion and minimise DNA degradation). Do you think that using a kit (like those I have read about on this site) may alleviate this problem?


Yes a kit may help, at least save you some time.

Before I switched to DNA modification kits long time ago, I did modification by preparing my own reagents and purification using Promega's wizard DNA clean up kit, sometimes I lost my DNA at the end. Results have been consistent since kits were used in terms of DNA recovery.

The time and temperature you used sounds OK to me.


Just a word of warning regarding switching to kit modification. We have tried this and ended up having to extensively purify our DNA after modification in order to get relible PCR amplification even though the kit was supposed to eliminate the need for downstream purification.
In the end the cost of the kit and the time spent purifying our DNA afterwards made it unadvantageous.
Instead, perhaps you could try heating your sample during bisulfite modification to maintain it in a single stranded conformation.

ie instead of heating at 55C for 16hr -cycle in a thermal cycler - 95C for 2min , 55C for 30min 8x.

The only advantage we found to kit modification was that the DNA was more stable on storage.



were the difficulties in purification that you experienced observed with any particular kit or any kit? I need to use the modified DNA for both bisulphite sequencing and MSP so I need it to be reliably converted and PCR amplifiable.

I am also concerned that the quality of my starting DNA (extracted by a salting out procedure) maybe less then desirable for bisulphite modification (although it is suitable for standard PCR and cloning procedures.



We have tried the Methyl Easy Kit and it is very convenient but we were unable to get satifactory results with it.

We also found the quality of the starting DNA was extremely important but cleaning up the genomic DNA by phenol:chloroform, ethanol precipitation was enough to improve the efficiency of our MSP and Bis sequencing.