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

MSP data presentation - What would be the best way to show MSP data (Apr/20/2005 )

Hi everyone, here's a question, I was wondering what the best way to report MSP data?
I am currently looking at CpG islands in various cell lines. As of now I just have a chart showing the cell lines and methylation primers with checks if I get a product. The papers I've seen correlate MSP data with the frequency observed in different tissues or show a picture of their products on an agarose gel. I also want to add sequence data which should make things more comprehensive but with just the MSP does anyone have any recommendations for data reporting(or a paper for reference)?

Second question, in MSP are we almost guaranteed to get a band with our unmethylated primer (all C converted to T including methylated C) set since it is not as specific? and if so would the absence of a band with the unmethylated primers and presence of a band with methylated primers mean that our sample is totally methylated?

Hope I didn't confuse anyone.Thank you very much in advanced. biggrin.gif


Hi Gungrave,

I think a agarose gel picture of the presence of a band or no band is the best and most conclusive way to dsplay MSP data. It would be very good to add bisulfite sequencing to the data as a way of confirming your MSP result.

Most MSP work I have come across show this and also add the same result post 5-azadC treatment.

If you don't see a band in the unmethylated primer set but one in the methylated primer set then I would say that particular MSP reaction of that region is methylated.

You weren't confusing at all.

Nick cool.gif


Thanks for the help. Follow up question
In a situation where we have both methylated and unmethylated bands show up,would that signify partial methylation???

What is the significance of partial methylation and does it mean we report methylation as a percentage since it's not all or nothing??

Anyone know any good literature regarding partial methylation as most papers I've come across seem to dance around it. Thanks again laugh.gif


There is never an all or nothing result.

Most people are looking at a population of cells containing variations of methylation status at a particular region. thus with PCR you would be getting a heterogeneous population of amplicons. Because of this, bisulfite PCR and seqeuncing requires the sequence many clones from the amplicons to get a representation of the population. Many people are now directly seqeuncing the amplicon and this can give you a true measure of the proportion of the population where that particular CpG is methylated or not.

I know Clark's laboratory (eg: Stirzaker et al 2004 Caner Research) uses this strategy.

Like you I have not seen one paper on this subject, in review articles it is usually one pass off line about it.



I am trying to check methylation status of a human promoter. My problem is although I can PCR out the fragment with a set of primer without C (they should PCR out both converted and unconverted template), I always got unconverted sequence after sequencing. if I use BS-converted specific primer, I simply can not get any product. (these primer worked fine with BS-treated plasmid) Is that possible that my genomic DNA was not converted? I digested the DNA and then used chemicon- methylation kit and let the reaction go overnight at 50 degree.

also are there rules regarding the primer design for BS specific primer? highly Tm? and how much DNA should I start with? I used 1ug or 2ug, that did not work.
any suggestion is highly recommended!



Hi Bahtigo,

For bisulfite genomic sequencing PCR you should only use primers that can tell apart converted from non-converted DNA, i.e. there must be multiple non-CpG 'C's in the primers. These 'C's are best at the 3' end of the primer and in a stretch such as 'cccc'.

Here you can find additional rules

Make sure your DNA is very pure otherwise it may end up with uncomplete conversion. 1-2 ug of DNA is OK.