Does anybody have experience with the RRBS protocol?

[Simone78]

Dear all,
we want to do methylome sequencing after enrichment using Solid4.
Our approach is based on the RRBS paper(s) (Meissner et al., Nature, 2008; Smith et al., Methods 2009 and Gu et al., Nat. Methods, 2010) with some modifications.
I encountered HUGE problems so far, starting from step 1! Here are just 2:

- in the original paper they used 40 (and later 20) units of MspI to cut the gDNA, starting from as little as 1 ug (later just 300 ng). When I did the same I saw (Bioanalyzer) that the overwhelming majority of the DNA is still uncut. Only when I increase the amount of MspI to 200-300U (!!!) short fragments appear...probably generated by star activity (?)
- we decided to perform size selection with dSPRI using AMPure beads but I found that the yield is not 90-95% as declared on the BC's website but as low as 10-15%. Furthermore, the range appears to be poorly reproducible (and I want to select the range 100-300 bp) and after the second SPRI I still see long fragments that should have been removed after the first SPRI...but I guess they come from using 2 ug gDNA and 40 ul beads. According to Lundin et al. (PlosOne, 2010) it seems that 10 ul of beads can bind 0.5 ug DNA. I'm now trying with less DNA.

It would be nice if anyone out there has some comments/suggestions!
Thanx a lot!

[sci_guy]

Hi Simone,

I use MspI quite a bit, 200 - 300 U sounds excessive. Try Phenol/Chloroform/Isoamyl alcohol treating your DNA to make sure it is clean. You could also try a commercial (or some plasmid) DNA to see if the enzyme is OK.

Also, are you aware of the bioinformatics issues in working with bisulphite in colorspace? Horrible. I've had to deal with it the past year. If you have a choice then definately use Illumina sequencing.

[Simone78]

thank you for your reply.
yes, I tried different amount of MspI and I think that 100U/ug is the threshold. After that the graph at the Bioanalyzer looks like if there is unspecific cutting (the microsatellite peaks, as predicted by in silico digestion, disappear).
I wanted to try to do phe/chl purification - size selection on agarose gel - gel extraction with ethanol precip because of the huge losses of DNA. I thought that doing size selection AND cleaning up with AMPure beads would solve the problem, but what I get it's not a purification, rather an enrichment.
Unfortunately we can't use Illumina. Our institute got 10 Solid machines from Life Tech...no other choice but Solid!