Hi, it is not uncommon that membrane proteins run aberrantly on SDS page. If we assume that its not due to modifications (glycosylation, phosphorylation etc) and its due to the protein itself, which often occurs with membrane proteins and other random proteins. It can often be explained if one considers a more realistic model of protein SDS-complexes. Unlike the simple view of one SDS monomer bound per so many amide bonds the actual situation is more like beads of SDS on a string of protein (there is a lot of neutron scattering and other data supporting this idea). For a typical protein the average holds and you get migration that depends upon the protein Mw, but for some proteins this is not the case. For example, with short peptides you can get sharp jumps in the migration behavior as a result of the limited and quantized number of SDS micelles bound. As the peptide gets longer the quantization effect tends to be lost and you get migration that is roughly dependent upon the size of the protein. Any protein that either effects the size of the SDS micelle or the number of micelles bound can run aberantly. For example, the GroES monomer runs at 14-15 kDa, which is much larger than its actual molecular weight. This is likely due to its binding of SDS micelles. Alternatively a number of other proteins run much faster than one expects as they either are not completely denatured or equally likely do not bind SDS micelles well. Membrane protein transmembrane helices certainly have the potential for this behavior. Last, another issue with some membrane proteins is the tendency to oligomerize in the SDS gel. We observed this with glycophorin A where the TM was stable to SDS and its native dimeric state was observed on gels. This is also true for the isolated bacteriorhodopsin helices where you actually see multimers on an SDS page gel. In short, SDS page does not really measure the molecular weight unless the protein conforms to typical "SDS-binding" profile. The fact that so many do is great, but when it does not and you can rule out modification and oligomerization it most often is the result of differences in its interaction SDS micelles. While interesting, it probably is not what you want to study.

Obviously! membrane proteins are covalently attached to carbohydrates or lipid moiety, most of the times. so its generally expected that the molecular weight is always higher than theoretical molecular weight

Most companies will tell you if there is a discrepancy in predicted and expected molecular weight when purchasing primary ab

A slight discrepancy between the theoretical MW and the MW observed during SDS-PAGE is not uncommon. As mentioned above, incomplete unfolding might influence electrophoretic mobility. Also, the observed MW of a protein might differ from the theoretical value when the protein binds more or less SDS than it should on average due to amino acid composition, thus appearing to be larger or smaller than expected.

Another cause (in addition to those listed by Michael above) is the presence of Pro-rich regions, which lend rigidity to the protein backbone.

Also addition of some post translational modification like phosphate or glucose moiety may lead to conformational change in the protein leading to abnormal behaviour on the SDS PAGE. to see whether phosphorylation is the cause trat your protein with calf alkaline phosphatase and see for yourself what happens. lower molecular weight could be due to truncation of the protein as some times some proteins carry protease sensitive motifs and may get cleaved off.

In my experience, SDS PAGE is a separation technology approximately based on protein size. I have experienced:

A 100kDa protein which, when the cofactor (200Da approx) was covalently attached by reduction, resulted in a change in migration by +10kDa to approx. 110kDa equivalent

A beta barrel that should have migrated at 21kDa but which runs really nicely at the position equivalent to 15kDa.

In other words, don't get over-excited about the relationship between migration position and molecular weight. SDS binding, covalent changes that introduce deviations from a linear primary structure, inability to completely unfold can all conspire to break that relationship. Healthy scepticism is the order of the day.

And remember that low temperature SDS (no reductant) is a standard method for the first step in preparation of gelatin zymograms. I think the expectation here is a degree of structural integrity or refolding ability, and 'aberrant' migration might be expected.

Suspect glycoprotein - use a series of enzymes to cleave it off?

You didn't mention whether you had a reducing agent in the SDS buffer

From my experience almost 90% of helical MPs without post-translational modifications run faster on SDS-PAGE (i.e. show lower apparent MW). The major reason is that SDS for most of them are almost a native-like environment. In part because of that, as Rob wrote, MPs often can not be completely unfolded in SDS even after extensive heating.

Here is a good paper that discusses anomalous SDS-PAGE migration.

http://www.pnas.org/content/106/6/1760.short

Besides all the possibilities suggested by others, it also depends on what you mean by theoretical molecular weight. If you mean the calculated size starting from the Methionine where translation starts, and your protein is on the membrane, there will be a co-translational modification as the signal peptide is removed. The website at Expasy.org has programs that can predict where the signal peptidase cuts, then beyond that, there might also be further oroteolytic cleavage as a protein precursor undergoes further processing.

Molecular weight determined by SDS -PAGE also depends on the nature of membrane protein i.e. peripheral / integral / transmembrane. Some of the membrane proteins show aggregation in SDS too and heating for a long time may be helpful to dissociate and unfold them. Several factors as described above may affect the electrophoretic mobility of the protein on the gel.

Hi, it is not uncommon that membrane proteins run aberrantly on SDS page. If we assume that its not due to modifications (glycosylation, phosphorylation etc) and its due to the protein itself, which often occurs with membrane proteins and other random proteins. It can often be explained if one considers a more realistic model of protein SDS-complexes. Unlike the simple view of one SDS monomer bound per so many amide bonds the actual situation is more like beads of SDS on a string of protein (there is a lot of neutron scattering and other data supporting this idea). For a typical protein the average holds and you get migration that depends upon the protein Mw, but for some proteins this is not the case. For example, with short peptides you can get sharp jumps in the migration behavior as a result of the limited and quantized number of SDS micelles bound. As the peptide gets longer the quantization effect tends to be lost and you get migration that is roughly dependent upon the size of the protein. Any protein that either effects the size of the SDS micelle or the number of micelles bound can run aberantly. For example, the GroES monomer runs at 14-15 kDa, which is much larger than its actual molecular weight. This is likely due to its binding of SDS micelles. Alternatively a number of other proteins run much faster than one expects as they either are not completely denatured or equally likely do not bind SDS micelles well. Membrane protein transmembrane helices certainly have the potential for this behavior. Last, another issue with some membrane proteins is the tendency to oligomerize in the SDS gel. We observed this with glycophorin A where the TM was stable to SDS and its native dimeric state was observed on gels. This is also true for the isolated bacteriorhodopsin helices where you actually see multimers on an SDS page gel. In short, SDS page does not really measure the molecular weight unless the protein conforms to typical "SDS-binding" profile. The fact that so many do is great, but when it does not and you can rule out modification and oligomerization it most often is the result of differences in its interaction SDS micelles. While interesting, it probably is not what you want to study.

We have never had a protein than runs at its theoretical molecular weight. There are many reasons for this. Many membrane proteins do not bind SDS to the same degree as other proteins. Since proteins on a gel run by charge:mass ratio any anomolies in SDS binding will affect how your protein runs. Also, don't forget the possibilities of post-translational modifications that can also affect how a protein runs.

Make sure that there is no degradation of your protein

Hi everyone...

This is not an answer but I was wondering if there is a way to fix this problem of proteins appearing at lower molecular weights? I am having the same problem. Should I boil my samples for longer than 5 minutes? Thank you.

Toni, it depends on your protein.  Certainly membrane proteins should never be boiled.  37C for 5 min is more than sufficient.  The theory behind SDS gels and molecular weight is that SDS binds uniformly to all proteins to give every protein the same charge:mass ratio.  While this occurs in theory, in practice it is rarely the case.  Add proteases to your samples when you make them, keep samples at 4C when processing, can you rule out post translational cleavage?

Hi Neil. Thank you. Can you perhaps head over to my profile and see the question I asked today? This is happening with my pAkt, pJAK and SOD1.

Do you think also the protein PI will affect? So for example if the protein has PI of 9.9? Would that affect the migration?

pI should not drastically affect the migration, as the SDS shall provide sufficient negative charge. There should be 2 molecules per amino acid.

Of course 9.9 is quite high and you have probably many charged amino acids. That could potentially disrupt some binding of SDS.

My actual protein band size is 43 kDa, but I have got a single clear band at 25 kDa. Is it possible to have such a large difference in shifting of bands, i.e., 18 kDa?