your sample before loading the column may be too concentrated. This may lead to a 'forced' dimer that may not be natural. It may still be reversible as you noticed by adding substrate. If your initial extraction is in a higher volume you may not get any dimers, or a lot less. Whether or not it is S bridges at that point is unclear, it may be. You may add DTT before loading your column and try to dissociate your dimer before SDS. That is one thing. The second thing is that you need to have enough SDS present to solubilize the amount of protein. But most likely what is happening is that when heating your sample, hydrophobic interactions (membrane domains for example) prevail. Hydrophobic interactions as I understand tend to be stronger the higher the temperature gets. And since in the dimer fractions the protein is already positioned close together, this may be the reason why you only see dimers there due to these hydrophobic interactions.

Hello Tomas,

some membrane proteins aggregate in SDS sample buffer when boiled. Cytochrome oxidase subunit 1 does this. To see it on SDS-PAGE gels I do not heat the sample and load it directly after adding the sample buffer. Give this a try.

Gavin

Hi, thanks for quick reply.

Yep, when I tried keeping it for about 10 min @ RT, I didn't get any dimer band (at least not much). But that was without reducing agent, so I tried more of them, kept it @ 37°C for 30 minutes and I've got nice dimer band.

And as I said, it's really only in the fractions, where the dimer should be. If I didn't see it with my own two eyes, I wouldn't probably believe it either...

Another weird protein is Atp9 of ATP synthase. I forms a multimeric ring that cannot be dissociated by SDS. Some membrane proteins act very strangely in SDS-PAGE. Maybe this dimer form is in an ultra stable conformation that can't be dissociated by SDS, only by your substrate. Does your protein have a lot of trans-membrane domains?

No, it should have only one membrane spanning helix.

Or your reagent are old. Try fresh denaturing buffer, or more bME

hello.I know some membrane proteins aggregate in SDS sample buffer when boiled.

As Gavin I expirienced that some proteins when boiled get aggregated. Have you tried seeding the same sample before and after boiling? On the other hand other bonds like dityrosine are resistant.

yep, when left for about 10' at RT, there wasn't much, when kept at 37°C for 30', the dimer was there. But the thing is, that I don't see the dimer/aggregate in fractions with monomer after size exclusion chromatography even when boiled.

Do you have something about the dityrosine bonds?

i saw the same thing during a purification of one protein. Seems to only happen when a certain protocol was used to purify, but not with another protocol. Im thinking that there must be some covalent bonding taking place and not sulphydral. I did not only see double sizes when purified, but also tripple and quadruple size bands and nothing else on the gel, nothing below the target size.

Did you also see both (or all) versions in one fraction? It seems weird, that it can survive the bME treatment to some level, yet most of it is monomerized...

your sample before loading the column may be too concentrated. This may lead to a 'forced' dimer that may not be natural. It may still be reversible as you noticed by adding substrate. If your initial extraction is in a higher volume you may not get any dimers, or a lot less. Whether or not it is S bridges at that point is unclear, it may be. You may add DTT before loading your column and try to dissociate your dimer before SDS. That is one thing. The second thing is that you need to have enough SDS present to solubilize the amount of protein. But most likely what is happening is that when heating your sample, hydrophobic interactions (membrane domains for example) prevail. Hydrophobic interactions as I understand tend to be stronger the higher the temperature gets. And since in the dimer fractions the protein is already positioned close together, this may be the reason why you only see dimers there due to these hydrophobic interactions.

Tomas, yes, I would see the target size, double the size, triple the size, quadruple all in one lane, also aggregates right at the stacking/running interface, indicating that there could be even more than quads. Also, that protocol if I remember correctly started with a high concentration of crude lysate onto a Ni-NTA column using a His tag.

This is an interesting discussion. Actually we observed a similar phenomenon with a TM spanning protein (single TM sequence) independent from boiling or not boiling.

- Ultimately it could only be monomerized by treatment with bME, TCEP, and dithionite. So one possibility is that there is a disulfide bridge which is for some reason (e.g. shielding towards solution) much more resistant to bME or DTT than usual.

- for our case we could rule out a covalent dityrosine bridge but this would be bME-resistant too and can be induced by Cu or other radical reactions

(for ref. see http://www.ncbi.nlm.nih.gov/pubmed/14717612)

- another possibility could be a metal-complex formation with Cys and Met residues which can lead to extremely stable complexes (see http://pubs.acs.org/doi/abs/10.1021/bi0478392) Are there Cys and/or Met rich regions in your protein (inside or outside the TM sequence) which could take part in a metal-coordination?

My feeling is that this phenomenon happens more often but tends to be disregarded as an SDS-PAGE artifact because it does not fit into the expected pattern of disulfid-bridge behavior. But maybe it reflects a more complex way (literally) of protein-protein interactions?!

But as usual there is most likely not one single truth for all proteins. I agree with the previous posts that the sample preparation/treatment, the detergents used (especially for membrane proteins), and other factors are definitely crucial for the outcome of the SDS-PAGE. The problem is always: When you see a dimer with one condition and not with another... what is "real" and what is the "artifact"?

Very nice

Here is a link to a paper showing how SDS micelles can mimic the membrane environment. http://pubs.acs.org/doi/abs/10.1021/bi9013819

So at least for some proteins, you can observe protein-protein interactions during electrophoresis that correspond to interactions in biological membranes.

We have been studying a membrane-bound protein (with a single TMD). We have been overexpressing this protein and truncated forms in mammalian tissu culture cells. However when we express one particular truncated form, which lacks a portion of the protein's cytoplasmic domain, we see the appearance of a weak band by western blot from SDS-PAGE that corresponds to a dimer. We estimate that this putative dimeric form represents a small fraction of the total protein (

I think this will work

what about the covalent bonds? The covalent bonds ,that form between the monomers to give dimer , trimer...., can resist the SDS.

yeah, but what kind of covalent bonds? Disulfide bridges should be broken by DTT or bME

my protein for example can form c-c covalent bond between tyrosine residues and give me dimer that resisted the SDS.

I saw something similar a lot of times, but in my case with trimers. Our fibrilar proteins make trimers and they are resistent to SDS. We usually heat the protein for at least ten min to get the monomer. For us is just another way to be sure that we have our protein and not other one with the same molecular weight: running in the same gel boiled and unboiled samples.

So, it's possible even if it's not a membrane protein. If you want to check if the dimer is a real dimer, reviewers will ask for analytical centrifugation, cross-linking or DLS. The easiest one is cross-linking, in two hours you will be sure if your protein is a real dimer or not!

Sometimes increased levels of SDS are needed. We had a case where we needed to double the amount of SDS in order to fully disrupt all quaternary interactions in a certain protein.

95ºC, almost boiling...

it is well known that hydrophobic interactions can be so strong in certain proteins, ie GPCRs (7 m pass) and others, that SDS has difficulty to break them. additionally heat at 95-100 promotes oligomer formation for some of these, so stay away from 95-100, go w 70-80C.

Instead of resuspending my sample in 1X Laemli, I sometimes use a final concentration of 2X, and it helps when I have more concentrated samples that give me oligomers on SDS-PAGE.

My protein got two bands in SDS-PAGE_dimer and monomer. I tried DTT and 2ME and dimer was still there. I also did Mass-spec and LC/MS. The dimer even survived under harsh treatment in LC/MS. What a strong interaction in my dimer!

yes it is possible to see the oligomers in SDS PAGE. I have seen hexamer formation of my protein in SDS PAGE. AND I have checked the presence of hexamer by western blotting.The best way is to boil your sample at 95C for 10 min which will give you only monomers.

I also observe the a dimeric form of my protein when running a reducing SDS-PAGE. It's a cytosolic protein, but I deleted a low complexity region. I thought that the interactions may have been hydrophobic, due to the deletion, so I ran my size-exclusion with5 mM n-butanol to disrupt the hydrophobic interactions. You can also try that perhaps. Otherwise, it could be that the protein is not being completely reduced, since technically disulphide bonds are the only covalent bonds you get. However, some ionic salt bridges (electrostatic interactions) can also approach covalent bonds strength.

Hi Tomas,

thanks for raisning this issue. I also didnt know whether we can see dimer or oligomer in SDS/PAGE , but now there is no doubt, now I am struggling with this problem how to get rid of these dimer oligomers from western analzsis to check wht is the exact molecular weight of mz protein. I think this phenomeno depends on the nature of the protein. Its verz difficult to make monomers for some proteins like channel protein I am working on. I tried denaturing the protein and also boiling in the Urea buffer but its not going... Well I think its due to hzdrophibic interactions which is verz strong in the channels proteins as the are integral membrane proteins .....

So if I want to study the heterodimer ex. EGFR-HER2 heterodimer, I can use the denature method containing SDS to do the immuonoprecipitation and then blotting? Is there probleme?

I totally agree the idea that it depends on the nature of the proteins. Membrane raft proteins and channel proteins are not easy to be isolated as well. 

Try increasing the concentration of BME

Similar dimerization of protein in SDS-PAGE was also observed with recombinant CENP-A protein (Reference: Developmental cell 32: 589-603 2015). Like membrane protein (as in your case), CENP-A is also has hydrophobic domains.

I also encountered similar dimerization with one of the protein I am working with but increasing the beta-ME resolved the issue.

Opposite to what is usually assumed, heating the sample doesn't always yield the best results. Indeed, many proteins (particularly transmembrane proteins) aggregate upon heating at high temperatures. Moreover, in some cases, the more you boil your sample the more your protein aggregate, even including chaotropic agents such as urea in your denaturizing buffer. I have seen that effect for several transmembrane GPCRs and enzymes.

I have observed something similar, heating removes the oligomer, but not reducing agents. Can anyone provide a literature reference supporting this fact?

Prarthna Saraswat, please go through all the answers. Very enlightening.

I had the same problem with my protein, i tried all the things suggested in this discussion, but the problem still there, am not sure yet about which type of interaction that i have in my protein is thers is any other suggestions