Tween-20 is very slow to remove by dialysis because the micelles are large and the critical micellar concentration (CMC) is low (60 mg/liter, about 50 µM). Only the non-micellar detergent will pass through the tubing. The micelles are in equilibrium with the monomer, so the micelles will gradually decrease in concentration as the monomers dialyze out, but this will take a long time.

You didn't say what the volume of the sample was going to be, so I'll assume it is 1 ml. The Tween concentration would be 1600 mg/L, way over the CMC. To get below the CMC, the protein sample volume would have to be about 27 ml. In principle, that would allow the detergent to dialyze out at the same rate as any similarly-sized molecule. Afterwards, you could reconcentrate the protein using ultrafiltration.

Trehalose, being a low-molecular-weight sugar, will dialyze out quickly.

Tween-20 is very slow to remove by dialysis because the micelles are large and the critical micellar concentration (CMC) is low (60 mg/liter, about 50 µM). Only the non-micellar detergent will pass through the tubing. The micelles are in equilibrium with the monomer, so the micelles will gradually decrease in concentration as the monomers dialyze out, but this will take a long time.

You didn't say what the volume of the sample was going to be, so I'll assume it is 1 ml. The Tween concentration would be 1600 mg/L, way over the CMC. To get below the CMC, the protein sample volume would have to be about 27 ml. In principle, that would allow the detergent to dialyze out at the same rate as any similarly-sized molecule. Afterwards, you could reconcentrate the protein using ultrafiltration.

Trehalose, being a low-molecular-weight sugar, will dialyze out quickly.

As an alternative to Adam's suggestion, if there is a resin to which your protein would bind reversibly (e.g. ion exchange, lectin, metal affinity etc.), you could try to bind the protein to it, wash out the detergent + trehalose and then elute it with an appropriate buffer (e.g. high salt, carbohydrate, imidazole for the resins mentioned above). Then you could adjust pH and salt concentrations by dialysis against the buffe of your choice

Dear Dr. Shapiro and Dr Beguin.

The protein concentration is high enough that dilution is not a hurdle, but in the Thermo-Scientific tips for removing detergents, it is stated that Tween 20 is not dialyzable. Is it not dialyzable because the micelles are stable?or the formed micelles   would decompose by dilution?

Thank you for your enlightening comments

I think the reason Tween-20 is considered non-dialyzable is that the critical micellar concentration is so much lower than the concentration of the detergent typically used for protein purification. The micelles themselves are too large to dialyze. Only the tiny proportion of monomeric detergent will pass through the tubing. The result is that it takes forever to get rid of it this way. In principle, the micelles would dissolve if the concentration of detergent were reduced below the CMC, making the detergent dialyzable. That said, I've never tried it, so I can't say for sure it will work.

Are you concerned that the protein will aggregate or precipitate if the detergent and trehalose (a stabilizer) are removed?

The protein I am dealing with is an IgG antibody, and I am going to use it for immnunoprecipitation. The presence of this much detergent in the solution would compromise the polarity of the protein and would interfere with binding of the IgG to the support. I think aggregation would be a problem too, I should think of that as well.

>in the Thermo-Scientific tips for removing detergents, it is stated that Tween 20 is not >dialyzable

I would treat statements from companies (especially Fisher) with great care, b.c. in the next paragraph they happily offer to sell you their resin to remove tween-20.