I haven't done Northern/Southern blots for over a decade, and do not have extensive experience with non-radioactive labelling, so my advice may not apply.

In general, the longer the probe, the better the sensitivity, but the longer it takes for hybridization to complete, to do the washes, etc. Specificity may be lower, background also tends to be higher.

With shorter probes (including oligonucleotides labelled with PNK or TdT) hybridization/wash times are shorter and if you design the experiment carefully, you can distinguish between single-nucleotide differences. The disadvantage is that sensitivity is lower.

You don't mention how experienced you or your lab are at doing Northern blots, but I would start by running a few Southern blots to troubleshoot any issues stemming from poor transfer, poor probe design, mistakes while performing the procedure, etc. without dealing with the potential hurdle of sample degradation or simply lack of expression of the target mRNA (as you would in a Northern blot).

Out of curiosity, why haven't you gone the route of RT-qPCR?

Cheers!

@ Alejandro Martin Thank you so much for your advice. My lab or me has never done northern blots ever. It's the first time hence we are trying to optimize the results. My PI doesn't like qPCR because we haven't been able to get good error bars from the data generated. I would like to ask you about your hybridization steps as to what buffers do you use ? I am using perfect Hyb plus from sigma and I'm not getting alot of signal. However I think that could be due to the fact that my probe is just 100 bp long. Do you have suggestions to reduce probe trapping by rRNA ?. Thank you so much for all of your help

I could search the exact composition of the buffers if you want but I'm sure PerfectHyb plus is way better. Things you could try to clear things up:

• First make sure the detection system and the hybridization step is working by dot-blotting a standard curve of the target and then probing it. This will also give you an idea of what the limit of detection is under your conditions.
• Then troubleshoot the transfer step. Do a Southern blot (this doesn't have to be with genomic DNA; just decreasing concentrations of restriction digests of a plasmid with the target sequence), then take a look at the wet membrane, after transfer, under UV (you should see the EtBr-stained transferred DNA fluorescing for the most concentrated samples) or stain it with methylene blue. I tried several variations of semidry transfers, etc. but old-fashioned capillary transfers were always better, by the way. Just to make sure everything is allright, probe this one too.
• And then do the Northern blots. If the specificity of the probe is good, you could even use normal agarose gels instead of denaturing ones, after ensuring that the probe only binds the target sequence.

Hope it helps!

Hello, Nikesh Kunder,

I implemented Northern blot in my lab so I can tell you what have I done. Hopefully you can follow the same protocol or get ideas for your experiments

I produced biotinylated riboprobes of approx. 400 nt to detect 3000 nt RNA. I have chosen long probes to increase the amount of inserted biotin which in turn increased the detection signal.

To produce the riboprobe, first I PCR-amplified the desired part of the target gene. Next, by restriction digestion, I inserted PCR product into a vector and then, by in vitro transcription in the presence of biotinylated UTP, I obtained biotinylated riboprobe.

Hybridization: background can result from a membrane. Which membrane do you use?

Do you block it before hybridization? Blocking should prevent a probe from getting physically trapped in abundant rRNA. After hybridization, I used twice low (2xSSC, 0.1% SDS) and twice high (0.1xSSC, 0.1% SDS) stringency buffer, each 15 min incubation. Then I proceeded to the kit Ultrahyb ultrasensitive hyb buffer (thermo fisher).

If you have low signal, I agree with @Alejandro Martin to perform Dot-blot to find optimal probe concentration. I used 800 ng probe per approx. 4x9 cm membrane in 10 ml hyb buffer. If you think you have enough probe, pay attention to stringent post-hyb washing. I found it very important in order to reduce background and nonspecific binding.

How to be sure your probe binds the target?

In my experiments was important to show orientation specificity of the probe (that a probe binds to antisense, but not sense transcript). To show it, I had hybridization controls. I in vitro transcribed (without biotinylated UTP) sense and antisense strand of the probe (the same process as described above: PCR then restriction digestion to insert a sequence in the desired orientation=the opposite to the final riboprobe). These unlabeled sequences were loaded on the gel, transferred to the nylon membrane and hybridized with sense riboprobe. Hopefully you can perform something similar, maybe produce nonlabeled hybridization controls of similar genes to show that your probe binds sequence-specifically to a certain RNA, and not to similar ones.

Check our paper for NB results: https://link.springer.com/article/10.1007%2Fs00018-017-2731-6

Another thing that just came into my mind. How much RNA do you load on the gel? This ofc depends on the amount of your target in the total RNA. Sometimes it means you need 20 or more micrograms! of total RNA per lane to obtain some signal in the end.

Greetings,

Mirjana

Alejandro Martin & Mirjana Polovic Thank you so much for your valuable inputs. I will try all the recommendations that you'll have suggested.

Alejandro Martin Mirjana Polovic I have attached some of the northern's that I ran as controls to see if my probes work and I am finding great difficulty in pin pointing the problem. Any suggestions would be of great help. The steps that I did are as follows:

The GAPDH probe used was 120bp long and my target size is 1.2kb

The Fam155a probe used was 500bp long and my target size is 3.3kb

Based on my GAPDH blot I am getting hybridization, I would just like a cleaner blot and sharper bands. As for my FAM155a blot, I understand it. This transcript is abundant in the brain hence I should see a band for it. Instead I see alot of non - specific binding. Any suggestions from you'll would be really helpful. Thank you'll very much.

Do a Southern blot. If you don't get the band sharpness/smudged appearance problem then you know either the samples are degraded or there is something wrong with the formaldehyde gel. Did you take a look at the gel? The rRNA bands, did they look sharp?

Hi, Nikesh Kunder.

Your protocol looks good.

One suggestion: I added SERVA DNA Stain G dye in the gel mixure before it polymerizes. It helps with visualizing your RNA during electrophoresis and after electrophoresis you can really quickly put a gel under UV to visualize and/or make a photo of your gel. Then you can judge RNA degradation.

You have quite fair signal (except the nonspecific smear) with 10 ug RNA, so maybe you shouldnt go above it (because then the nonspecific signal increases).