Dear all, I'm doing RNA pulldown experiments to identify RNA-binding proteins which interact with my RNA. I biotinylated my RNA using biotin RNA labeling mix (roche) and PAGE purified the RNA. For binding, I overnight incubated 5 ug (~80 pmoles) of biotinylated RNA with 1.6 mg of Hela nuclear extract in the presence of 0.1 ug/ul tRNA and 1U/ul RNasin. For pulldown, I added 30 ul of streptavidin agarose beads (Thermo) to the binding reaction and incubated for 1 hr. After extensitve washing, I eluted captured proteins in SDS sample buffer and performed SDS-PAGE. Following coomassie staining, I could observe several bands specifically enriched in the RNA(+) sample, but the problem is that there are too many non-specific proteins appearing in both beads-only and RNA(+) samples to precisely excise out bands of interest without contamination. I tried pre-clearing my extract by incubating it with 60 ul of streptavidin beads before binding, but this did not significantly improve specificity. Can anybody help me reduce non-specific protein binding on beads? Any comments would be greatly appreciated.
Non-specific protein binding to beads is a common IP problem, and has nothing to do specifically with your experiment. Some things to keep in mind when troubleshooting. You need to increase the stringency of the buffer solution, but not too much or else you will lose the specific binding you are looking for. Since biotin-streptavidin is incredibly high affinity you don't need to worry about that one. But you do need to worry about RNA binding to whatever you are looking for. Hopefully you have a positive control that you can validate your assay with (ideally via western blotting). Protein binds to the beads fairly regularly, to combat this I tend to increase the salt concentration of the buffer to 250 mM, and in some cases I will add 0.001% Tween20. If this isn't sufficient, then keep increasing the salt concentration, but keep in mind that you need to validate this with a positive control...so if you lose signal in your control, you have likely gone too far. I suspect 250 NaCl and Tw20 will get you where you need to be though.
My suggestion is to bind your biotinylated RNA to your Streptavidin beads first, then berform a free streptavidin wash. Then incubate your lysate to the mix. The reason for this adding the streptavidon wash is to make sure you block any access streptavidin site on the agarose which has no RNA bound to them. By doing this ypu elimnate any access room for natuarlly occuring biotinylated proteins to bind to your streptavidin-agarose, thus increasing your specificity.
Bassam, thank you for your reply. I'm a little confused, what exactly do you mean by 'streptavidin wash'? According to your description, it sounds like some kind of blocking step, but block the beads with what? I just googled 'streptavidin wash buffer' and found nothing special: Tris buffer plus 0.1% mild detergent, which is essentially the same I use for washing.
Sorry my bad, i made a mistake here. What you need to do is a Biotin wash step to your conjugated gel mix after binding the biotinylated RNA. Suspend biotin in PBS buffer at the same concentration you have used for your binding reaction. Then washStreptavidin conjugated beads with biotinylated RNA two times with biotin wash to block unbound Streptavidin sites before you apply your lysates. You can incubate the biotin wash with your conjugated beads for 2 mins at room temperature couple of times. Make sure to remove the biotin wash before moving forward, centrifugation will do the trick. Does this clarify the process further to you?
Non-specific protein binding to beads is a common IP problem, and has nothing to do specifically with your experiment. Some things to keep in mind when troubleshooting. You need to increase the stringency of the buffer solution, but not too much or else you will lose the specific binding you are looking for. Since biotin-streptavidin is incredibly high affinity you don't need to worry about that one. But you do need to worry about RNA binding to whatever you are looking for. Hopefully you have a positive control that you can validate your assay with (ideally via western blotting). Protein binds to the beads fairly regularly, to combat this I tend to increase the salt concentration of the buffer to 250 mM, and in some cases I will add 0.001% Tween20. If this isn't sufficient, then keep increasing the salt concentration, but keep in mind that you need to validate this with a positive control...so if you lose signal in your control, you have likely gone too far. I suspect 250 NaCl and Tw20 will get you where you need to be though.
Thanks Colin for helpful comments. I used 100 mM KCl plus 0.1% NP-40 for binding/washing and I can consider increasing KCl concentration to 250 mM. Unfortunately, I have no idea what proteins bind to my RNA because the purpose of this experiment is to identify them.
The salt concentration is definitely the problem....100 is far too low. Even going up to 150 will help quite a bit. Try 4 conditions 100, 150, 200, 250 mM salt in your next experiment. might even try 500 and 1M as negative control. In biochemistry there are 2 things you need to demonstrate to prove that an interaction is specific: 1) that it is reversible, and 2) that it is saturable.
I would also encourage you to validate a new protocol with a positive control before moving on to unknowns. Even if, in your case, that means using a different RNA molecule.
First of all, I would figure out the interaction basics without the beads. Have you tried UV crosslinking? That is the best method to give you an idea of the number and size of proteins you are looking for (sometimes even guess what they are!) and you can play with the buffers and competitors to minimise background. Then apply the same conditions to your pull-down.
Pre-clearing the extract on beads rarely works, I tried it too.
As far as actual buffer conditions, I second the salt suggestion, but would stick to ~150 mM (physiological). You can also include a low concentration of an ionic detergent (I use Sodium Deoxycholate). Pre-blocking the bead-RNA complex with pure protein such as BSA also helps - unless your protein of interest is of the same size.
Hope it makes sense. If you need more details, let me know!
I don’t agree with the affirmation that a specific interaction has to be reversible. Irreversible interactions through covalent bond formation are also specific. There are examples. Sometimes there is no covalent bond formation but practically they behave as if they were irreversible. The biotin-streptavidin interaction is an example.
I do agree that you may increase the salt concentration together with non-ionic detergent; It will even favor Biotin-streptavidin binding but can affect negatively the interaction of the proteins you are interested in with your RNA. Maybe you should try Streptavidin magnetic beads which yield less non-specific binding than Agarose beads.
You could try PAR-CLIP technology. Whatever you do don’t forget as advised by Colin about controls.
Thank all of you for valuable comments. Dear Anna, UV crosslinking seems to be a great preliminary experiment I can try to do. I like it. Dear Abdelhalim, CLIP is not applicable to my case as it aims to identify RNA sequences with which a given protein interacts.
But PAR-CLIP allows you to map the sequence of your RNA with which a protein is intetacting
Hey!
i think you should do SILAC. Where LIGHT probe is control (unspecific RNA sequence or just beads) and Heavy is sample. Than you can discard unspecific binding. If you do MS. Or isobaric labbeing... And just on-bead digestion, without gel. But dont forget to use OG instead of Triton or NP-40.
To reduce nonspecific binding, add 0.1% SDS, 1% NP-40 or 0.5% sodium deoxycholate to the buffer
Hello Lee
I hope you already solved your problem but before performing SDS page you should check the reterival % of your traget RNA by RT-qPCR then you can pellet your protein and check with MS if the rtterival% over than 15% considered okay
I hope you have solved this problem of non specific binding. I am trying the same experiment but I was first binding the RNA to beads and then incubating beads with the lysate. Does your way work better ? And I see that you are using milligrams of protein. Do we need to use such high amount ? I m still struggling with the experiment.
Are there any biotinylated proteins in the extract? If so - these could bind to the beads in an RNA independent manner.
First, the final concentration of extract should be considered -not just the quantity of HeLa extracts (1.6mg). It should be sufficiently diluted to limit non-specific interactions (I would use 320 microL or in this range).
Second, I would add:
-some detergent (0.05-0.1 % (final) NP40 or Triton) to reduce unspecific interactions (sticky one's)
-some digitonin (0.05-0.1 % final)
-some BSA (0.1mm/ml) to compete with unspecific interactions
I would also make every effort to avoid drying my pellets during wash steps. Any dry sample becomes sticky (hydrophobic interactions).
Good luck
Hi all,
I am really happy to see that this old thread is now full of constructive comments and suggestions. For those who are doing similar experiments, I recommend to follow the protocol developed by the Caceres Lab (RNA. 2010 Aug;16(8):1673-8), with which I successfully identified the proteins bound to my RNA of interest with no or little nonspecific contaminants. This protocol utilizes RNases for elution and does not require high strigency for specificity (i.e., you can perform elution at physiological conditions), which may lead to the loss of relatively weak but still important RNA-protein interactions.
Thank you all of you for enriching this thread.
- Badges
- Science topic
- Similar topics
- Biological Science
- Molecular Biology
- Biomolecules
- Nucleic Acids
- RNA