miRNA detection problem by Northern blot - (Dec/09/2007 )

Hi,
I have problems regarding detection of microRNA. Everyone said that it is quite simple, but it's not...

I isolate total RNA using Trizol from Hela cells. Is it possible that I lose small RNAs?
What is the maximum amount of RNA that I can load on a gel? I usually put 50ug.
I control almost every step of Northern: electrophoresis by staining with EtBr, transfer and crosslinking by staining membrane with methylene blue.

I don't have a good control of isolation RNA and hybridization. Even if my probe is complement to let-7 (which supposed to be abundant in Hela cells) the signal is barely seen.I use a standard hybridization buffer: formamide, Denhardt solution , SSCE, SDS. My probe is oligoDNA of a complement sequence and length to the miRNA I want to detect. DNA is labeled at 5’ end by PNK with 32P. Is Optikinase so much better?
Does anyone has experience with isotope from Hartmann Analytic or other Amersham, Perkin Elmer? I have heard that from ICN is the best...

If anyone can help me reveal what I can do wrong with Northern, please write…

I've never used trizol, but it should be all right for small RNAs. If I remember correctly, it involves at least one isopropanol precipitation. Maybe that can be adapted a little, increasing conc, incubation time and lowering the temperature. For instance, 1 vol and incubation at 4 °C for 1 h is certainly more effective than 0.8 vol and a ten min incubation at room temperature. However, this may lead to decreased purity, and so a compromise should be adopted. Siliconized tubes may increase yield.
As for the northern, I would stick to some protocol well established. You can download a good one from the Bartel lab web site (just goggle Bartel lab). I've read somwhere that not all the membranes work well with small rna. I would use one of those commonly cited like zeta probe biorad or hybond amersham. Personally, I use a roche nylon membrane positively charged that has never been cited, but I've been lucky and it works fine. I cross link by UV followed by baking.
I don't personally prepare the probes, nor do the hyb, so I can't help you on that side.

Thanks for the reply.
I followed the protocol for short miRNA northern from many labs.
It seems that the method works fine, but in my case the amount of miRNA is propably to low be be detected.

Does anyone know if is it possible to use gels (separating -15% and stacking -5%) for Northern blots similar as are used for SDS-gels?

If the expression is too low, I would rather try an LNA probe. And if it still doesn't work, a PCR approach. One more option is the alternative fixation with carbodiimide (Pall et al 2007, NAR 35).

I have tried the carbodiimide fixation but RNA was poorly transferred to the membrane.

It doesn't work as good as it was shown in the publication.