Is the protein loosing the tag or might be not 100% of the prot stabled tagged, try to confirm by precipitating some untagged control

I would go to magnetic beads like Dynabeads. Just mix in your antibody with them for 15-30 min at room temp in PBS (I use 5ug antibody per 50 uL beads in 250uL). Wash 3 x 500 uL, then add that to your sample. You will need to get a little magnetic stand, but overall magnetic beads are light-years ahead of anything agarose in terms of purity and reproducibility.

Daniella, Decribe what you have in your negative control.

I also prefer magnetic beads with streptavidin and labeled antibosy with biotin, usually the signal/noise is perfect.

^ That's a good thing to do. Also, check your protein concentration and make sure that you don't exceed the myc-bead binding capacity.

What MW is the co-precipitating protein?

I have had some success doing shorter IPs (1h while rotating at 4C) and then increasing wash time - rotate the beads 5-10min x 5 washes in lysis buffer (1%TX100, 150mM NaCl, 50mM Tris, 0.1%SDS, 0.5% Deoxycholate).

So, the co-purified protein runs around 50 kDa. I use as negative control beads+antibody with lysate (without myc-tagged protein) and also with Lysate containing co-purified protein (transfected).

It sounds like you might have a sticky protein but definitely depends on a few things. Do you see this problem in your negative control if you use just Protein A or Protein G beads alone? If yes, are you using sepharose or agarose? I find that a lot of proteins display differential non-specific binding dependent on your matrix so try switching. If you do not see non-specific binding because of the beads perhaps you need to rethink the antibody you are using for the myc pull-down... If you want to try and wash off the negative binding, this can become tricky as well mainly because if you are going to try and co-IP interaction complexes from this, you might not be able to isolate them as easily with these type of washes. If thats not an issue then its something you can definitely try!

David I think you are totally right. The co-purified protein binds the beads (agarose) and even more when it is transfected. Also I think I am not enriching too much the myc-tagged protein. So I am thinking to change antibody and change the beads. Someone suggested to use the magnetic beads. I've never tried them. So I do not know.

As I understand, there may be a problem with the sample preparation. It is possible that your solution is not homogeneous, with precipitated proteins. Can you tell us what protocol you used to solubilize the proteins?

Two other tricks you can try are 1) Pre-incubate the beads with BSA to block non-specificity from your matrix or tritrate BSA into the IP reaction to reduce non-specific binding in your lysate (adjusting buffer concentration of the lysate pre-IP also accomplishes this objective) or 2) Titrate a very small amount of SDS (0.05%-0.5%) into the IP reaction - this will reduce total recovery, but improve specificity.

This may not be applicable in this application (anti-myc-bound Protein A), but with antibodies one can pre-clear the antibody mixture with the negative control and use the supernatant in the target IP incubation. This removes antibodies (in a polyclonal setting) that are non-specific or have cross-reactivity with other proteins in the lysate of interest.

I perfomed multiple IP with different antibodies and different beads and they worked always. So I don't think that my lysis buffer is a problem. I got in trouble only with this specific experiment: IP=Myc (beads already conjugated or ProteinA-G plus antiMyc).

I'm attaching a file of my blot.

So on the left are the input and on the right the IP:MYC. Between right and left samples there is an empty lane. So, the first lane coming from the IP (right after the empty lane)should be the negative control (transfected cells with X protein, without MYC TAG) and as you can see this protein is co-purified even without myc-tagged protein!!!

Hi Daniela,

Do you have a negative control where you have just beads and your lysates without addition of antibody? The reason I am asking is that your co-purified protein runs at the same MW as IgG heavy chain under reduced conditions. It may be possible that you are picking up the IgG and not the protein you are trying to co-IP. If you are trying to detect your co-IP protein by western blot using a secondary that is from the same species (ie, IP with a rabbit polyclonal and then western blotting with the same rabbit polyclonal) then you run the risk of detecting the 50 kDa IgG band. One way we get around that is to run our samples under non-reducing conditions (ie., leave out the reducing agent in our SDS-PAGE sample buffer) to prevent separation of the IgG heavy and light chain. The MW of IgG is approx 150 kDa under non-reduced conditions. Also, if you have an antibody to your co-IP protein that is from a different species that should also help. For example, if you IP with a rabbit polyclonal, and western blot with a mouse monoclonal, the mouse secondary shouldn't pick up the rabbit IgG used to perform the IP.

If it does turn out that your co-IP protein is actually sticking to your beads, I agree with Brock that you may want to try blocking your beads first with BSA (we do the first wash of our beads in lysis buffer containing up to 5% BSA followed by 2 more washes in lysis buffer without BSA) prior to adding them to your lysates.

I do, I usually block the beads with a buffer containing BSA and regarding IgG, than band can not be the IgG because the antibody that I use to detect my co-purified protein is coming from another specie.

There are a lot of possible explanations! If precipitation is not a problem and none of the suggested above did work then may be your myc antibody is detecting endogenous Myc protein. I couldn’t open your tiff image maybe you could post it as jpg. To test this possibility you should buy myc peptide and use it as competitor.

What Tina is suggesting should not be the case because the myc antibody is covalently linked to protein A and protein G or that what I understood from what you were saying.

Can you try to open as .png file?

Hi Daniela! Thanks for the png image. I definitely think you should first check if your Myc antibody detects endogenous Myc from your supplier. I am confident that this is your problem. It is well known that all myc antibodies I know recognize the endogenous c-myc. You could change the tag if necessary.

but I don't think the size is that

What is the size of this protein on your gel?

around 50 kDa

My boss thinks that probably I am not pulling down anything, that's why the co-IP proteins are exactly as in the INPUT.

Could you put a legend to your western and post it as png image. I would like to have a look to it tomorrow.

Normally C-myc migrates around 60 Kd. Please provide these data. I guess you keep 10% of input. In this case how much do you load on the gel (from the 10%)? And how much do you load from the IP (what is the total volume of the IP)?

c-Myc 48.8 kDa (migrates around 50 kDa)

http://www.uniprot.org/uniprot/P01106

Since its a Myc antibody, it could be c-Myc gets coprecipitated. But this protein in the first lane of IP is not there in equal amounts in IP lanes 2 and 4 hence it could be that, either it competitively binds with your Myc tagged protein (if it is a Myc Western blot). You need to label the lanes (let it be protein -x or antibody -y) for interpretation. I am surprised you dont have heavy chain or light chain bands in IP lanes (note the heavy chain should be around 55kda and light chain should be around 25kda (some antibodies may have either one of the chains less or more but shouldnt be lacking both chains). If it is antibody heavy chain band then the beads are not in equal quantity between lanes (since you mentioned the beads are preconjugated with antibody). You need to provide more details (of course dont reveal the protein name but need the protocol how you did) for better troubleshooting.

Hey Daniela,

please check your myc antibody with the lysate....if it give more than the desired band and its intensity is sharp, then you have to low temperature binding protocol, and lysate must be pre cleared....hope you get you desired band...

Good luck with any protein around 25 or 50 Kd. They are similar size for light/heavy chain of IgG. We usually try ImmunoCruz (ExactaCruz) to resolve the problem and avoid using too much antibody (>1 mcg) for the pull-down which will make it worse even more.

Hey Daniela I really recommend to skip myc and flag tags for IP as the beads can bind unspecif proteins very often. My experience is on plant proteins (very difficult) and I did many many co-IPs so I strongly suggest you to use GFP (25KDa) as a tag to pulldown (there are fantastic agarose beads commercially available) and HA tag to detect for the interactor if you want to do co-IP. Perfect monoclonal antibodies are available for both tags which also helps a lot to get nicer Western blots. Good luck and let me know if you need more info!

Thank you very much to everybody for helping me. With my boss we think that probably the IP is not working at all, so what we saw in the bound fraction is actually a picture of the inputs.

I need to check the antyMYC which epitope recognizes and see if Myc-tagged protein contains that peptide.

To answer some of you: it can not be the HEAVY chain coming from the antiMYC because the secondary antibody that I used to detect my protein is from another species.

Moreover, I have to use a reducing buffer because the two proteins interact likely through disulfide bonds.

Hi Daniela,

I agree with Bastian that even if you use a different species antibody in the Western, this can still pick up your IP-antibody. Our apraoch is to use magnetic beads (Pierce) for the IP (quicker protocol and less background then agarose beads, because they are not porous) and use a protocol where we crosslinkthe antibody to the beads so they don't come off, and don't end up on your Western. These beads can then also be reused several times, saving money. I also agree that a FLAG or a his-tag will probably cause less trouble with precipitating endogenous proteins.

Good luck,

Rob

Rob can you send me a working protocol with the magnetic beads (pierce) and how to refresh them?

Thank you so much

Hi Daniela,

I'm not in the lab just now, but will look up my protocol tomorrow

Hi Daniela,

Just did a quick web search. Both Pierce and Invitrogen do very good crosslinkable magnetic beads. Both have detailed protocols on their web site. last time we used the Inviutrogen dynabeads ones. I've copied their crosslinking protocol here. it is really rather straightforward. We eluted the captured proteins from the beads using a pH3 citrate buffer.

Cross-linking protocol for IgGs immobilized to Dynabeads Protein A or Protein G

Prepare 100 mM BS3 in Conjugation Buffer (Stock Solution). Proceed to making a 5 mM solution by diluting in Conjugation Buffer, 250 µl is required per sample.

Note: BS3 Stock and Conjugation solutions must be freshly made prior to use!

Wash the Ig-coupled Dynabeads Protein A or Protein G twice in 200 Conjugation Buffer.

Place on magnet and discard supernatant.

Resuspend the Dynabeads in 250 µl 5 mM BS3.

Incubate at room temperature for 30 min with tilting/rotation.

Quench the cross-linking reaction by adding 12.5 µl Quenching Buffer

Incubate at room temperature for 15 min with tilting/rotation.

Wash the cross-linked Dynabeads three times with 200 µl PBST (or IP buffer of your choice). Place on magnet and discard supernatant.

Proceed with your IP and antigen elution (starting from step 2.3 of the Immunoprecipitation protocol).

BS3 Conjugation Buffer: 20 mM Sodium Phosphate, 0.15M NaCl (pH 7-9)

BS3 Quenching Buffer: 1M Tris HCl (pH 7.5)

Cross-linking reagent: Bis(sulfosuccinimidyl)suberate (BS3), f.ex. Cat. # 21580 from Thermo Fisher Scientific Inc.

One comment on Carmelo's answer.....boiling does not break disulfide bonds within or between proteins. Only a reducing agent is effective at performing this.

I do agree with Bastian that Myc epitopes as affinity tags can introduce a substantial amount of crossreactivity with other irrelevant proteins and confound your results. I have found the HA epitope (and even a 3X HA epitope) to be a much "cleaner" affinity tag for recombinant proteins.

Daniela, there are some good tips and advice in this thread but what I have read so far makes me wonder if some of the answers are a bit misguided because of what exactly is your situation regarding the interaction you are investigating. I have extensive experience with IPs and Westerns and would be happy to help you out but would need to know a little more of the specifics regarding your interaction system (ie. how are your proteins found in their native state). From a purely professional stand point I encourage you to contact me directly if you still are unsure of how to solve your dilemma. Best of luck to you!!

maybe you can change a more strict IP buffer, and increase the wash time, decreasing the binding time for beas and proteins.

Hi Daniela,

I agree

- with David (try different gel matrix: agarose, sepharose, magnetic, ...),

- with Sergio (if you have some protein aggregation, during the centrifuge step it will precipitate by g force, but it will appear as precipitated by the Ab/resin complex, unfortunately you cannot distinguish them. Magnetic beads should solve this kind of problem (during washes, beads on the wall of the vial, and most of the unbound-aspecific proteins, in solution or on the bottom of the vial)),

- and with Tina :)

(also covalently bound antibody-beads can produce an aspecific loss of IgG during the sds-reductant boiling step before the loading (it depends on the bound protein, see M2 a-flag agarose (cross-linked) by sigma. e.g.). so, an additional solution to the non-reducing condition run, is to dissolve your ip -first- in a sample buffer without reducing agent (dtt or b-mercapto), then collect the supernatant, add the dtt/b-mercapto, and load (spin down and boil, in each step).

About the secondary antibody, check if that's pre-adsorbed on the species of your ip antibody; aspecific binding is not only related to the same ip/wb primary antibody species.

you could try also protein-A HRP, protein-G HRP, or trueblot (as already proposed).

good luck. and keep us updated on your results.

Giuseppe

Could you try the monoclonal anti-myc antibody (not the conjugated) to IP the samples after the pre-clearing?

You can try blocking the beads before binding the target protein with a nonspecific protein such as BSA, including 0.05% Tween-20.

I have the same problem. My two proteins are membrane proteins, and I tagged them with myc and GFP, respectively. If I use total protein, I couldn't pull down each other. However, if I use membrane fraction and IP use the same condition, I can pull down the negative control as well. I already tried different system, GFP trap, Ab290 with Dynabeads, and the Miltenyi Biotec GFP, or Miltenyi Biotec myc. The results were same.

 Dear Sabarish,

Got to know from your reply that you have used Milteny beads for Co-iPs. I have recently started a PDF in Durham university and need to use iP for my work. The lab also use Milteny beads with magnetic stand and columns (just as you mentioned). However, the issue is the experiment is not working my hands. I am not able to get enriched proteins in my IP. I feel that I need to shake the beads vigorously (but other labmates said is not that critical. Also, I am not getting very clean IP, getting bands around  slightly above and below 25 kD, which earlier was thinking of light chain (but you mentioned that antibody does't come down). So now I am really wondering what is that. This also makes me think that may be some other protein is competing for binding with beads thereby affecting binding of my protein of interest. But another issue is I am observing an additional band below 25 kD even for empty GFP (the IP is very clean but less enrichment).

Kindly help and also if possible kindly share your detailed protocol. If required I can attach the pics of my blots