RNase A is indeed the "industry standard", BUT you have to titrate it for every experimental set up as RNase A will only digest single stranded exposed RNA. Assuming that your samples are not crosslinked, increasing the temperature of the digest to reduce local secondary structure might cost you the IP. Even 37oC might be too high to maintain native protein-protein interactions. BEWARE: An over-digestion may lead to conformational changes of some proteins that can result in non-specific binding, in particular to sepharose/agarose beads. Monitoring rRNA digestion is a good suggestion, assuming your input is total cell lysate. Not too good if you are using nuclear lysate.

Orthogonal ideas:

1. micrococcal nuclease (MNase) is a much milder nuclease, digesting both DNA and RNA if they are unprotected. It can be activated and deactivated by manipulating Ca+2 levels and chelation by EGTA, respectively. I am not aware of over-digestion issues.

2. If you can express a tagged version of your pet protein, you can create truncation or point mutants of the RRM domain.

3. How about performing the IP in cells under conditions inhibiting RNA polymerase activity? The most physiological one is synchronising the cells in mitosis, where most transcription is arrested, but to the best of my knowledge Pol II is happily swimming around. There is also a wide range of reagents, some of which reduce the level of Pol II and some that do not. I cannot remember off the top of my head which is which, but Actinomycin D, alpha Amanitin and Triptolide spring to mind. Possibly a metabolic starvation of some sort could do the same.

4. Extract and characterise the nucleic acid from the IP if you are using anti-pet-protein antibody. Anti-PolII will give you... well... lots of RNA and DNA :-)

MST sounds cool, but possibly over the top - although I will definitely look it up, I am such a sucker for cool methods...

Back to the question, I am not sure what strength of statement you would like to be able to make, but a combination of the above should be a good start. I would use two independent methods if possible.

Hope some of it helps and good luck!

Anna

You might want to employ MicroScale Thermophoresis (MST) to investigate the interaction between your protein, RNA Pol II and the RNA!

MST allows you to measure interactions not only for binary but also for multiple component interactions. You could also study the interaction of your protein to RNA directly.

One of the binding partners would be labeled fluorescently and added in constant concentration to the dilution series of the interactant. You only need little amounts of your sample (4µl per titration point) with nM concentrations of the fluorescent molecule.

So it is very easy to set up an MST experiment, the instrument is straightforward to use and the measurement is really fast (it takes approx. 10min to obtain a Kd using 16 titration points)!

For further information, you could have a look at our homepage:

http://www.nanotemper-technologies.com//home/

And please feel free to contact me via [email protected]!

Best regards,

Heidi

I would go with the RNase treatment too

Splitting your coIPs in two and adding RNAse A to one is the standard protocol, and the one we always used in my old lab. If you only see the proteins coimmunoprecipitating in the sample without the RNAse, you can say that the interaction is RNA-dependent

RNAse A has a very broad range of activity conditions, so should work in most - to be sure, I would add some RNAse to RNA in your coIP buffer and run it out on a gel to make sure you can see degradation of prominent RNA bands (e.g. rRNA).

RNase treatment is the gold standard in those cases. You can immunoprecipitate your protein of interest ; split your beads in two different tubes and treat one with RNAse A 5 to 10 min at 37°C; pellet your beads save the supernatants. Then analyze what was on your beads and in the supernatants. If your interaction is RNA mediated, then the RNA pol II should be predominently in the supernatant of the RNAse A treated samples.

Do not forget to have a positive and negative control...

RNase A is indeed the "industry standard", BUT you have to titrate it for every experimental set up as RNase A will only digest single stranded exposed RNA. Assuming that your samples are not crosslinked, increasing the temperature of the digest to reduce local secondary structure might cost you the IP. Even 37oC might be too high to maintain native protein-protein interactions. BEWARE: An over-digestion may lead to conformational changes of some proteins that can result in non-specific binding, in particular to sepharose/agarose beads. Monitoring rRNA digestion is a good suggestion, assuming your input is total cell lysate. Not too good if you are using nuclear lysate.

Orthogonal ideas:

1. micrococcal nuclease (MNase) is a much milder nuclease, digesting both DNA and RNA if they are unprotected. It can be activated and deactivated by manipulating Ca+2 levels and chelation by EGTA, respectively. I am not aware of over-digestion issues.

2. If you can express a tagged version of your pet protein, you can create truncation or point mutants of the RRM domain.

3. How about performing the IP in cells under conditions inhibiting RNA polymerase activity? The most physiological one is synchronising the cells in mitosis, where most transcription is arrested, but to the best of my knowledge Pol II is happily swimming around. There is also a wide range of reagents, some of which reduce the level of Pol II and some that do not. I cannot remember off the top of my head which is which, but Actinomycin D, alpha Amanitin and Triptolide spring to mind. Possibly a metabolic starvation of some sort could do the same.

4. Extract and characterise the nucleic acid from the IP if you are using anti-pet-protein antibody. Anti-PolII will give you... well... lots of RNA and DNA :-)

MST sounds cool, but possibly over the top - although I will definitely look it up, I am such a sucker for cool methods...

Back to the question, I am not sure what strength of statement you would like to be able to make, but a combination of the above should be a good start. I would use two independent methods if possible.

Hope some of it helps and good luck!

Anna

Heide, MST I guess would need purified components. Studying the RNA here might be difficult, as Ahn might not know what RNA species it is. Besides, her protein may or may not interact with Pol II directly (it is unclear from the description of the problem) and the connection may be bridged by any/many of the trillions of components that the Pol II holoenzyme is made up or interacts with (OK, that was a joke, not trillions, I take back). So, how to do MST for such huge range of possible interactants? Good answer, Anna; I would add starvation and serum-release to create transcriptionally quiescent and proficient conditions. I would also suggest doing parallel IPs: (1) lysate pretreated with RNase and (2) IP'ng first and then treating with RNase. This is because RNase will not digest protein-protected regions of RNA.

I would go with Benzonase treatment, which is an endonuclease that chopps both DNA and RNA. First i would add Benzonase to the lysis buffer (make sure it contains Mg). Alternatively, you could do benzonase-treatment of IP-beads, as suggested by Daniel Scott....and- did you see RNA-"contamination" of your IP´s when running part of your IP-samples on agarose/denaturing formamide gels and ethidiumbromide staining? good luck!

Anil, basically you can also perform MST measurments in cell lysate. If your protein of interest is expressed as a fluorescent fusion protein (such as GFP, YFP fusion), you would not have to purify the protein. To address the question from Ahn, it would be difficult since I guess that the RNA binding is not sequence specific and the protein would just bind the RNA present in the lysate.

To just get an answer if the binding involves RNA or not the RNAse assay seems to be the easiest one, I agree. If, howeve,r you would like to obtain any additional information on the binding events (such as affinity, stoichiometry etc.) MST could be very helpful.

I Know That one of the most methods used to study the interaction RNA-protein is Electrophoresis Mobility Shift Assay (REMSA) which is based on the different speed of migration of the RNA-protein complex during gel electrophoresis.

Individual-nucleotide resolution Cross-Linking and ImmunoPrecipitation, and electrophoretic mobility shift assay (EMSA) can be used to study RNA-protein interaction

We have gathered nice views. Now, what if a short chunk of RNA is trapped between the PolII/associates and Ahn's protein of interest and it is not accessible to the nuclease (that is, the nuclease will stop at the point it cannot touch the RNA any longer)? If so, the IP result would appear as if RNA has no role to bridge the two components, which may not be correct. One way to go forward is to follow up IP (from x-linked cell lysate) with RNA-seq to identify the RNA species. If it is specific, it can be knocked down and the protein-protein interection can be interrogated. But, what if it is any piece of the nascent RNA from any part of the transcribing genome?

I agree with all above me about doing RNase assay. I actually used a mixture of RNase A & T, at r.t. for 15-30min.

You can also try RIP with your protein and see if you IP only mature mRNAs (which would suggest it binds the RNA posttranscriptionally) or only/also nascent mRNA (unspliced or non-adenylated etc.. though its not definite proof since both forms may also be cytoplasmic e.g. NMD or normal mRNA decay).

a different approach is to knock-down the RNA binding ability (mutate the RRM) and repeat the co-IP with this mutant.

RNA ip