The molecular weight of IgG is indeed 150,000, but it is made up of smaller subunits of about 50 kDa and about 25 kDa (heavy and light chains, respectively). Therefore, you would not expect to see IgG at 150 kDa on a Western blot.

To avoid picking up a signal from the IgG used for the immunoprecipitation, use a primary antibody for the Western blot from a different species than the one used for the IP.

I agree with Adam. It seems weird that you get a 150 kD band, unless you are running a non-reducing gel? 

Alternatively, you could also try crosslinking your antibody to your beads. This retains the antibody on the beads when eluting, and allows for a more stringent wash, reducing possible background-binders. 

I know that there are several commercial kits available. 

Both answers above are great - I would only add that if you want to confirm that your 150 kDa band is IgG, the simplest way is to repeat your WB, but use HRP-secondary antibody only. If it still reacts, it's IgG and you'll need to do one of the two things suggested by Adam or Daniel. Otherwise, you have a different protein coming down, and will need to increase the specificity of your IP, either by increasing stringency (higher detergent concentration in your IP buffer) or the specificity of the antibody by affinity purification. 

Adam, my IP primary antibody is mouse, my WB primary is rabbit. Which makes the 150kDa band a mystery. 

Daniel, it's a standard SDS polyacrilamide gel. I will look in to cross-linking kits (or perhaps do my own if I can, with formalin). 

David, I will definitely try that to rule out IgG binding.

To all, I should add, I left out a very important detail. My negative control (HEK293 cells which are negative for my protein of interest) also showed bands at both positions. 

Thank you all for  your feedback!

Oftentimes, Abs are not truely specific, meaning they will recognize more than one protein.  That being said, it doesn't really matter for western purposes as long as you know which band is your protein of interest.  Alternatively, the second band may represent post translational modifications of your protein.  If this is of interest, knockdown your protein and see if the band(s) go away.  This will confirm that the band(s) are your protein.

Try to purify using an affinity matrix by chromatography. There are several commercial activated beads for immovilization of proteins as immunoglobulins. After covalent binding no releasing of material is done.

In my case, when using protein A/G agarose beads, I always have problem with heavy and light chain. So now I am using ImmunoCruz IP/WB Optima D system. As my experience, when I performed IP by rabbit, detect WB by mouse, just the specific interacting band appeared.

If you are looking to obtain a relatively clean band that is your protein of interest only, a good strategy maybe to over express a 6xHis tagged version and purify using Ni-NTA. If you are using reducing agent in your loading (BME, DTT) you will not see 150 kD IgG. Therefore, the presence of your band in negative control lysate indicates unspecific, though high affinity binding to another protein.

Dear Terra Elizabeth Wolfe ,

I agree with Michael that a knock down experiments would show which bands are specific in the IP, but make sure before doing IP that the knock down worked efficiently enough. Also, it would be good to include a sample from the lysate used for IP in the WB analysis - in order to verify that the IP is indeed enriching for a specific band. You need to do calculations to apply the corresponding amount of lysate to be able to directly compare band intensities. Please don't hesitate to write to me if you need help! You mentioned that 293 cells are negative for your protein of interest. How do you know that? Is UGT2B15 not a ubiquitous protein? Is it excluded  that the antibody also reacts with UTP glucosyl transferase (appr. 150 kDa) because of some sequence similarities?

I do not think these bands have to do with crossreacting Ig..

Ingrid, UGT2B15 is not ubiquitous. It has been previously established that the kidney cells do not express it.

You make a good point about UTP glucosyl transferase cross reactivity. I'll look in to that. 

Thanks!

As others have mentioned, in WB of IPs there should not be 150kDa band in reducing SDS-PAGE. Using primary and secondary antibodies of different species minimises the detection of heavy and light chains of IgG. In some IP, I use a secondary antibody against the light chain of the primary antibody, which also helps remove background bands. Detection of 61kDa band in IP from 293T cells may suggest that these cells express low levels of the target protein. In our experience, some endogenous proteins in 293T cells are undetectable by WB of cytosolic lysates, but weak bands can be observed following IP or co-IP procedures, suggesting low cellular expression. Furthermore, an antibody may preferentially recognize a protein in its native form as we have experienced in our work.

Thanks Nasim.

I am  using different species for my IP vs WB.

Good point about IP/coIP detection vs. WB, I'll keep that in mind. 

Hi Terra,

In agreement with the former colleagues I also do not believe that your additional bands are due to ppt antibodies. Instead, I thought of "specific" ppt or co-IP of intracellular proteins with possibly similar epitopes - for instance UDP-glucose:glycoprotein glucosyltransferase, which has an apparent MW of 150.000 kD. Is this possible? Are all bands detected enriched by IP? A possibility to check for that is to include a sample of the lysate used for IP in the analysis by WB. If you do not observe an enrichment, the antigen-binding capacity of the ppt ab may be saturated. In this case, you could use less of the lysate to obtain the same amount of precipitated material (this also has the advantage that it lowers unspecific background). I agree also with Nasim on the possible presence of your  protein of interest in 293T cells.

Good luck!

Ingrid, I have the 293 lysate that I'm adding to the WB as well as a know positive control (A-431 IP lysate from liver). I can also include a sample lysate.

I do think that the similar epitope explanation makes the most sense thus far. UDP-glucose:glycoprotein glucosyltransferase is a strong possibility. 

Thanks!

Thanks again for all of your feedback! As it turns out, the HEK293 cells do have a low level of ugt2b15 expression. 

I'm thinking that the band at 125kDa may be something that the ugt2b15 is binding to endogenously. Thoughts? B15 is 61kDa so it would be a protein ~ 64kDa. Am I interpreting this correctly? 

Non-covalently bound proteins are separated by SDS-PAGE.