You are assuming that the aggregated protein is due to intermolecular disulfide bonds. This may not be the reason for the aggregation. It is common for overexpressed proteins to have some aggregation, which is why gel filtration/size exclusion chromatography is a good last step, to get rid of the aggregated protein.

Even if the aggregation is caused by intermolecular disulfide bonds, you should be able to reduce the disulfide bonds by including a reducing agent (dithiothreitol, TCEP, or 2-mercaptoethanol), as long as the protein does not also contain intramolecular disulfide bonds that must remain in the oxidized state.

Instead of treating the protein with N-ethylmaleimide to block the cysteine residues, which might adversely affect the activity and/or solubility of the protein, you should include the reducing agent throughout the purification procedure and in the storage buffer to keep the cysteine residues in the reduced state.

Adam B Shapiro Thank you for your reply. Yes, my protein contains intramolecular disulphide bridges, and it is necessary they remain oxidized because they stabilise my protein (and I need to check how stable they are). I think that for this reason, I cannot use any reducing agent, right?

PS: how can NEM affect the solubility of a protein?

Just a suggestion, the oxidized form of Dithiothreitol (DTT), is highly soluble (in both its oxidized and reduced forms) and generally considered non-toxic. By thiol exchange it can readily form mixed disulfides with free protein thiols (the reaction is completely reversible), in fact thinking about it perhaps, the oxidized form of alpha-thioglycerol might be better than DTT disulfide, as it has all the afore mentioned properties, and won't as readily self cleave (thiol exchange reactions are very fast and intramolecular ones will occur essentially instantaneously).

Philip G Penketh Thank you for your suggestion. Is there a risk in using DTT that not only the free cysteines but all cysteines in the protein will be blocked? Do you have some papers that I can read to know better about the use of DTT and alpha-thioglycerol to block free cysteines? I cannot find anything.

They (reduced DTT & alphathioglycerol) will only react with oxidized thiols/cysteines, i.e. cystines not cysteines. It is a long time since I worked in this area but we were interested in this reaction as a strategy to design thiol activated anticancer, and antiparasitic agents see

Article Thiolysable Prodrugs of 1,2-Bis(methylsulfonyl)-1-(2-chloroe...

Article Synthesis and Evaluation of 1-Acyl-1,2-bis(methylsulfonyl)-2...

We envisioned using the thiol/disulfide reaction to selectively activate antitrypanosomal agents in mice. The rational was going to be trypansomes contain the dithiol trypanothione and the one of it thiol groups would react via Michael addition with our drug, this would lock the molecule in the cell, then its second thiol group via intramolecular reaction would release the short-lived toxicophore. This later reaction was to be achieved by thiolysis or a disulphide bond reduction/thiol exhange reaction. However, this project/idea/strategy was abandoned and never pursued. It is possible that some of the references contained in those papers (now 20-30 years old) may be helpful, but a more recent and relevant review of such reactions and their biochemical relevance such as

Article Kinetics and Mechanisms of Thiol–Disulfide Exchange Covering...

Will undoubtedly contain the information you require either directly or cited therein.

Good luck.

Alkylating free thiols on the surface of the protein with NEM changes the surface property of the protein (see the reaction here: https://en.wikipedia.org/wiki/N-Ethylmaleimide). If enough of a change is made, the protein's solubility could be affected

Try TCEP. It doesn't permeate cell membranes.