Quantitative Assay for DNMT1 activity... - need help please (Feb/07/2007 )

Hi fellow Methylvestigators

First off thanks for this forum, I found a lot of information here while working on my MSc thesis where I used HPLC for methylation studies.

Now for my questions:

I am trying to find/develop an assay that detects the activity of DNMT1.
I need this as I am trying to create a line of DNMT1 overexpressing mice (conditionally by Cre-lox recombination).

I have found many assays that use Tritium (H3+) incorporation via SAM, but we don't have the capacities here for radioactive assays.
I have also found antibody kits from EpiGentek and others, but these would only allow quantiation of the enzyme, and not of its actual activity.

For my assay I was thinking something in the terms of an oligonucleotide substrate, to which I add the isolated protein.
Then either a methylation sensitive restriction digest or a combination of bisulphite treatment and bisulphite specific realtime PCR.

Here are my problems/questions:

1.) Do the people here think my approch could work, or have I overlooked any major problems?

2.) As I am studying DNMT1, I suppose I would need a Hemi-methylated dsDNA Substrate. How could I create this in a reproducible manner? Or would an unmethylated oligo work as well?

3.) Does anyone have any other ideas on how this measurement may be achieved?

Thanks for your help!

Goofy

Hi Goofy,

I'm just curious - why do you need the activity of DNMT-1? Most studies that I am aware of tested the gene expression or amount of protein, which correlates well with the amount of hypermethylation in many cancer tissues (but does not crucially). And many studies focussing on pathways mediating hypermethylation found changes in the gene activity, i.e. the expression of DNMT-1 gene. Do you have any information on the activity? Are there posttranslational modifications?

Krümel

Hi Krümel,

thanks for your interest!
Basically you are right - I could do a western with DNMT1 AB to check the gene expression, but this
would only give me an insight to the amount of protein - still leaving open if the excess in DNMT actually
results in a higher activity of the enzyme.


A DNMT activity assay would exclude rate limiting factors such as the de novo methylation by DNMT3,
which is necessary to create the hemimethylation as a basis for DNMT1 activity.

I hope this makes things clearer for you?

thanks again for any answers!

Goofy

Hi Goofy,

that makes it clearer. Maybe this paper offers a method you can adopt? Pubmed-ID: 17263334

K.

thanks - will look into it right away !

well, thought a little more about it - probably you will need to invert the assay - incorporate a HpaII cleavage site (5'-CCGG-3') and use HpaII to cleave the dsDNA. Only unmethylated dsDNA will be cleaved and set the fluorophore free.
But the need for hemimethylated DNA lasts ... I just checked on the websides of operon and eurogentec - they don't offer custom CpG methylation (but maybe they do if you ask them?) This could be the easiest way - optain fluorescent hairpin oligos with a CCGG in it, that is already hemimethylated. Than you can real-time monitor the activity of DNMT-1.

K.

Hey Krümel,

Thank you very very much for the paper - it's exactly what I was looking for !
The modification with HpaII was exactly what came to my mind, thank you for the confirmation.
(This gives a nice control via MspI digest, as I am sure you are aware of)
I am working on the exact modifications basically right now as I am typing.

One question came to me while reading the paper that you posted:
The Dam MTase in the paper seems to be sensitive to the bases next to the cleavage site of the enzyme.
I have never heard of this phenomenon for methyl-sensitive restriction enzymes.
Does HpaII also depend on neighbouring bases?

As for the hemimethylation, this seems to be less of a problem.
Other publications have used poly[d(i-C)d(I-C)] as a substrate.
Poly [d(I-C)d(I-C)] is not methylated at all, but they still had success, so I will try the assay with an unmethylated hairpin oligo to start off with.
If I run into problems I will see if I can order a hemi-methylated oligo somewhere. But I will cross that bridge when I get there

Danke vielmals

Goofy

Hi,

good idea to use poly[d(i-C)d(I-C)] as a substrate. I guess that this should work well.
Regarding HpaII - there has been a discussion somewhere in the forum if it matters if the outer C of the CCGG motive was also methylated and could then interfere with HpaII cleavage. That is the thing that comes to my mind when thinking about neighbouring bases. We use HpaII/MspI digestion for analysis of global DNA methylation and haven't had any problems...

K.

Hello Krümel,

I will post my progress here so other people can follow in future if they need to

A little misunderstanding there - the poly[d(i-C)d(I-C)] was only an example for an unmethylated substrate,
we have already ordered a hairpin oligo similar to the one in the paper, but with a HpaII/MspI cleavage site.

Can't wait to test it - will post here when I have made progress.

Thank you agian very much for your help Krümel

Goofy

It worked!

I adapted the hairpin and a few other parameters/buffers and read out the signals in a realtime PCR machine and it seems to work fine

thanks again for the idea Krümel, I owe you a favor!