I've had success using a Nylon membrane (normal blotting nitrocellulose does not immobilize DNA). You'll also need to crosslink the DNA to the membrane after filtration using UV (UV stratalinker, etc). This paper has the method (just ignore the part about a double-membrane assembly):

http://www.ncbi.nlm.nih.gov/pubmed/15620896

Afraid I have no expertise in the infrared detection.

Nylon N+ membrane work fine. As per the amount of DNA, take into account that Me-Cy are everywhere in the genome so you will detect it quite easily, so make sure you use various amounts of DNA...I detected Me-Cy with only 50 ng! For blotting, as I mentioned before use Nylon N+ (if you have a slot-blot device much better as the bands are sharper than a dot blot...otherwise, dot should do ok), take say 50 to 500 ng of gDNA and add TE pH:8 to 100 uL final volume (or more if your gDNA stock concentration is too diluted for some of them). It does not really matter much how much difference in volume you have between samples as long as you have the same amount of gDNA). After that, you heat the sample to 95deg (bath or block) for 10' and put them immediately on ice after this. Incubate samples for 5' and three volumes of of 2M NH4Ac (Amonium acetate) pH:7, ice cold. Mix well, and load on your blot as per your method...slot/dot blot. Before disassembling the blotting device, wash each slot/blot with 500 uL of 2M NH4Ac (Amonium acetate) pH:7.

Take the membrane, incubate for 10' is SSC 5X, wash in distilled water and let dry. Cook 80 degrees for 2 hs or 15' in laminar flow (or alternative cross-linking device you have).

TIP: before blocking the membrane wash it 1 h (the longer the better up to 24 hs) with 0.1% SDS...as per Li-Cor theoretically gets rid of background....it actually does.

For Odyssey, both green and red work fine, but red was always stronger for me.

Hope this helps.

Good luck!

Thanks a lot Luciano, this helps me a lot! But two further questions: did you use any kind of DNA stain to control the blotted amount of DNA, or did you count on the measured sample concentrations and the reproducibility of the actual blotting? And last, may I ask which anti-methylcytosine antibody did you use?

Ah, maybe a third one: can you tell me how small differences in the global methylation levels could you measure this way? Did you do some kind of standard series to check this?

ooooh...so sorry, I knew I forgot to answer something...yes I did...very simple....once you finish up with your southern blot, and you got your nice scanning and you are happy with them, put your membrane in 1-5x SYBR Gold (NO SYBR Safe....doesn't work) for 10' and they use an UV or blue light imager, the same you use for your gels to take a photo. That is how is usually done and I did, works very well.

Another tip, in your empty wells, surrounding your blotted gDNA add some NH4Ac with loading buffer (blue) to help you find where to cut the membrane and also to see how the blotting went, if it looks anything but a dot you need to tighten more the device or add more whatman filter.

For reproducibility I tried to do at least 3 replicates (technical or biological) and get the average for quantification purposes. I used the Odyssey software to do so. I can't remember off the top of my head how to do it.

Anty-Me-Cy I use one from Epigentek...it was very good.

The global methylation measured in this way is not an exact science, it's all relative measurements so I would say I could detect up to 20% between a control sample and another treated with Aza dC for 3 days.

The standard series would be good if you have those standards...I didn't have them so I couldn't, but a colleague of mine bought hyper-methylated gDNA and unmethylated gDNA and used those too to set up all his experiments.

Thanks again!

I have a somewhat similar set-up using the licor odessey to quantify repetitive DNA with flourescent probes on dot-blotted genomic DNA.

What charged Nylon membrane do you recommend. I am using the BIORAD Zeta-Probe right now, but it seems to autoflouresce with UV and the blue spectrum, which makes it impossible to use SYBR gold .

How to stain a dot blot of genomic DNA as a loading control?

The DNA (ng range) to stain is most probably denaturalized ssDNA (Urea/SDS backextraction from the TRIZOL inter-phase) and it's blotted on a positively charged Nylon membrane.

A dsDNA antibody recognizes other samples, but not this particular dots, while 5mC-antibody gives good signal.

I already tried methylene blue, but the staining is not sensitive enough and gives high background.

Next I tried 1xSYBR GOLD in 1xTAE. There are controversial comments out there in the web. To some it seems to stain well, while others comment that it doesn't work.

It seems that it works well for southern blot using the DNA previously stained in the gel before transfer, but not incubating the membrane (that stains entirely in my hands).

There is info about SYBR DX stain specifically for membranes (from Molecular Probes/Lifetechnologies) but it seems to be discontinued.

Does anyone have tips to get SYBR Gold working or any alternative method/protocol?

 Dear all,

Thank you a lot for all your tips, it helped me a lot. 

I have another question related to this method, I would like to detect 5-mC on RNA. Does the blotting procedure is the same??? 

Thank you in advance for your help,

best, F

In prinicple a similar plotting strategy should work. The only thing to take in account may be to use a formamide-containing solution when blotting the the RNA. This avoids pledging of the RNA in secondary structures. Just have a look on oldschool northern blot protocols. Good luck!

Has anyone tried blotting nucleosomal DNA and then probing with 5mC antibody?

I plotted genomic DNA in a standard dot-blot south-western procedure (fixation by baking) and hybridized with a 5hmC antibody. Happy to take questions!

Stephan

Did it work? What membrane did you use? How did you quantitate the amount of DNA slotted on each well or do you do this before dotting? Thanks very much for the quick answer.

Dennis

South western:

Quantify DNA and put equal amounts in 20-50 ul H2O or TE. (for tiny amounts you may choose Qubit quantification.)

Dot/Slot blot straight to the NYLON membrane (Amersham Hybond N+, GE Gealthcare) – I use a vacuum blot (Biorad). Do not heat the sample before loading. (If you boil you may get DNA denaturation resulting in unknown amount of ssDNA stretches, since I use dsDNA antibody as loading control heating would potentially alter the result.

I tried UV crosslinking, of the membrane and baking worked better. I boiled samples before blotting and got worse signals. I blotted with SCC-based buffers and they worked worse in my hands then straight blotting in water.

Fix by baking at 80ºC for 2h

Block directly with 5% milk in TBS-Tween for 2h

In the following I combine HRP/ECL detection in combination with fluorescent secondary antibodies so I mix mouse dsDNA antibody and rabbit 5hmC primary antibodies for incubation (o/n 4ºC). So I get the loading ctrl of the same spot.

The same way, I incubate together a mix of secondary antibodies (anti-rabbit-HRP (1:5000/1:10000) and anti-mouse-Alexa flour 680 (1:5000) for 1-2h. Then, I first (!!!) do fluorescent detection in a Licor Odyssey or camera-based system to detect dsDNA signal strait with the washed wet membrane and then I incubate the blot with standard ECL. (The fluorescent signal is easily quantified (linear signal) – ECL is more sensitive as signal is accumulative on the x-ray film and may be exposed longer at will. ECL solutions may screw up fluorescent secondary signal detection if you do the other way round.

I use a mouse anti dsDNA antibody (ABCAM HYB331-01, ab27156) as loading control (1:10000) and 5hmC antibody (1:1000) Active Motive. Cat. Nº 39769. I also tried a mouse Methyl-cytosine antibody once and it worked fine.

I hope this helps, good luck!

My very best,

Stephan

SYBR Gold could work if you pre-stain DNA instead of staining the whole membrane. I used 0.1X with total volume of 100uL loaded in the well on a Hybond membrane. At first, I saw signal on my negative control (no DNA), but overnight washing with TBST at RT can completely get rid of the nonspecific. My question now is how to accurately quantify it. Because SYBR Gold is so sensitive, it is hard to tell the difference in DNA quantity just by eyeballing it.