I think your analysis of the problem is correct, there is probably some degradation during transfer. The transfer time does not seem to be a problem, especially when you  observe empty gel after transfer (and all marker bands).

You can try DEPC treatment of your water, it is cheaper than using MB-grade water for buffers. But the source of your contamination can be anywhere, so perhaps prepare the buffers fresh as well, treat the transfer aparatus with some RNase inhibitors etc.

Good luck!

what i feel is you should keep the x ray film for more time as when week signals are there with a positive control, it takes longer to expose fore the sample, expose the film for a longer time in fact few days, i am sure you will get it.

I agree with Jiri, I think you need to control for RNases at all steps in your assay. In addition to using DEPC treated water, you may also consider autoclaving the water and using some of the commercially available RNAse wash solutions to clean your workspace. One other suggestion, albeit unlikely, is it possible that you are overtransferring? I wouldn't expect it, though, you didn't mention seeing an RNA ladder and it may be possible that you are running your transfer too long.

Hope this helps and good luck.

Chris

Thank you Jiri, Chris, and Monica

We have a lot of molecular grade water at hand so I'll use this to make up new buffers & RNaseZap my workbench/apparatus to see if I can minimize RNA degradation.

Chris, I did consider over transferring, however I do get a crisp exposure of my RNA ladder.

Monica, the detection kit that I am using has a very limited emission period (within the first 3 hours). I usually expose my film for 15mins then for an hour if I don't see anything, then finally overnight. Surprisingly my overnight exposure is dimmer than my 1 hour exposure, so I am not sure if leaving it a couple of days will change. However with your suggestion, I am now considering leaving it for 3 hours without the prior short exposures to "catch" all of the emission that I may be losing in the beginning.

Thank you all for your suggestions.

Did you test the transfer efficiency by detecting RNA on the membrane? A very simple method is to stain the membrane with methylene blue (MB). You can do this even after the chemiluminescent detection step, although the background is worse. You should be able to see the rRNA bands with MB if you loaded 1-10 micrograms of total RNA. I find that MB staining is good for locating the RNA ladder bands when you want to size the transcript detect by chemiuminescence.

If the RNA is getting degraded during the transfer, then you won't see the rRNA bands. I suggest this as a trouble-shooting technique, and certainly wouldn't recommend staining with MB before biotin detection. Another way to check transfer is by UV, since the RNA will fluorescence even when attached to the membrane, albeit rather weakly. 

Lastly, reconsider the transfer protocol. I am following the KPL kit protocol in which transfer is complete in 3h or less using 5x SSC/10 mM NaOH. They specifically advise against prolonged transfer (perhaps because of RNA degradation?).

Hope this helps.

Spot blotting is not a good control because it condenses all hybridization into a tiny area  - even non-specific binding. For this reason you can sometimes observe hybridization even when no bands appear in a traditional Northern. I recommend you use a known highly expressed gene as a positive control. 

Could you please confirm the nature of your probe? For northerns, your probe must be antisense RNA, and you need to confirm that it is labeled with high specific activity. Although many kits say that DNA probes can be used, in my experience they only work for detecting very abundant transcripts. 

Philip, My probe is anti-sense of my target transcript and it is a DNA probe with a biotin label.

I recently ran a test by dot blotting my positive control samples and hybridized with my anti-sense and my sense probe. I only got a signal from my anti-sense probe, so I am quite convinced that my probe is working properly.

Mark, I understand your concern that a spot blot may not be a good control, and I will look into making a Gapdh probe.

Therein lies your problem. Design a PCR primer pair to amplify a sequence

Thank you for your advice Philip,

The Gapdh probes worked, so my northern protocol should be working well. I agree that the probes are the problem, and I will be design RNA probes next.Cheers,

Yuka

Maybe your target gene is expressed at a very low level. Did you use the probe for a genomic Southern?