Tissue disruption - for RNA isolation (Jul/27/2011 )

Hi,
I'm trying to isolate qPCR grade RNA from human lymphoid tissue (lymphoma) and I use QIAGEN kits for RNA isolation (RNeasy) and lysate homogenisation (QIAshredder). But I'm not sure about the first step, the tissue (frozen in RNAlater) disruption to single cells.

The pieces are about 3x3 mm and currently I'm sonicating them thawed in RNAlater, but that requires at least 10 times few-seconds of sonication pulse and still some small part is present, so I usually just take the remnant out, which is decreasing the yield, because I'm affraid I would damage the cells if I sonicate them too long. Then I pelet cells with centrifugation. I got relatively good RNA but I could use higher yield and sonication takes to much time (10-15 min for one sample) so I'm looking for some better option.

Now some koleagues use liquid nitrogen and classic ceramic big mortar and pestle, but I don't like this since it has porous surface which is difficult to clean. I found out that it's possible to buy micropestile, that could disrupt the nitrogen-frozen tissue right in the tube which looks like a good idea, but I don't know if it really works well.

Does anyone have experience with micropestle, sonication or some other tissue disruption method (without use of machine homogenisers) that could be used for RNAlater soaked tissues?

before freezing we grinded the tissue with a blade - stored it in RNAlater or in nitrogen.
then we put the frozen tisse directly into the Trizol and homogenized it by using a syringe with a 27G needle. fill and drain the syringe 10 times while stirring the homogenate.

Thanks for an idea. We have some 27G needles so I can try it too.

So far I tried to grind the frozen tissue in the ceramic mortar cleaned with RNaseZap but the samples were quite hard to crush, I ordered micropestles but I have doubts if they would be hard enough to grind sample in the tube and if the tube would withstand such treatment while frozen. I then lysed grinded tissue in RLT buffer and shredder from the kit or Trizol and got more RNA from Trizol (but as the mortar had been cooled the Trizol froze in it and it took several minutes to change from sorbet back to liquid). Alternatively I tried to just put the sample to Trizol directly, but some chunks appeared despite repeated pipetting and I had to centrifuge them out. Now it seems Trizol is better in lysis and homogenisation, grinded tissue homogenized nicely in it, but not in RLT buffer and part was lost on shredder, but I used a mixed protocol, where the aqueous phase from Trizol is transfered to RNeasy columns.
The yield was better than from sonication, but gel check showed the 28S:18S ratio to be more like 1:1 than 2:1 so I have mixed feeling about that.

I was really trying hard to keep samples completely frozen (in nitrogen) in the mortar to keep the RNA integrity, which was kind of difficult as the nitrogen is constantly boiling and sample pieces jump around. But this tissue was frozen in RNAlater and manual says it's OK to thaw them several times so I think I will try to grind them in tube to smaller pieces even when they start to thaw, because they're still protected and then lyse them and homogenise in Trizol without messy nitrogen splashing all over the place.