Protocol Online logo
Top : New Forum Archives (2009-): : DNA Methylation and Epigenetics

Validating MeDIP by qPCR - (Sep/11/2011 )

Hi all
I am carrying out MeDIP which may be followed by sequencing on SOLiD platform. Before taking it to SOLiD platform, as is necessary, I would be carrying out qPCR to validate and calculate the enrichment.
The question is, what should be the way to do this validation. What I know is that (similar to ChIP) we do delta delta Ct calculation. I am a little confused as to should this be between Input and MeDIP samples or between MeDIP and mock IP?
Also, is there any method we can quantify the percentage of DNA recovered after IP (except nanodrop)
A brief introduction to ACTUALLY what should we and what we can analyze after MeDIP through qPCR would be really appreciated

-Muhammad Umer-

You need to find a gene which you know that is hypermethylated (for example GSTP1 for prostate cancer) and one that you know that is not hypermethylated. Design DNA primers for both genes and use for example SYBRgreen to measure Ct threshold for both input and enriched samples. You need to use diluted Sss1 methylated DNA (with WGA DNA) as a standard curve to normalise. This way you get fold enrichment of your MeDIP estimated by GSTP1 enrichment. Secondly you can pyrosequence a known hypermethylated gene before and after enrichment, which should give you an approximately three fold increase (in our experience). Hope this helps,

Bram

-ForEpi-

Hi
Thanks for the reply and useful suggestions. I have a panel of genes/loci which I will be using for validation. RASSF1a, H19, PWS-SRO and GAPDH as "negative control". My question is, what will be the suitable calculation methods after carrying out this analysis. The matter which is confusing me is that should the starting quantity of DNA for qPCR be same for both input and MeDIP fractions? If not, as is suggested in the protocol I have, how can we normalize the MeDIP fraction with input (which is way higher in starting quantity than MeDIP fraction).
My understanding regarding what we ACTUALLY need to prove by this validation is, that firstly we have to prove that MeDIP is specific for methylated fragments. This will be proved by amplification/non-amplification of methylated and non-methylated fragments. Secondly, we need to prove it sensitivity. Say we have a region which is almost 100% methylated and then there is a region where around 20% of CpGs are methylated in the target fragment. We need to show that MeDIP successfully, though may be differentialy, pulls down 100% and 20% methylated regions. I am stuck, how can we do that.

Umer

-Muhammad Umer-

Hi,

be very careful using GAPDH as a negative control. We found it to be partially methylated (and enriched) in CaP using MeDIP.

About the starting quantity: yes it should be the same (we just used NanoDrop).

About the sensitivity: in the literature you can find examples that will tell you (Weber et al, Nature 2005 I think and subsequent papers) how well MeDIP pulls down specific methylated fragments in relation to the CpG content. I have not seen specific assays to validate the MeDIP procedure for that though.

You could use PCR fragments with known CpG content, methylate them, enrich them and QPCR them. Just a thought.

Bram

-ForEpi-