I've tried dialysis and I lost a lot of protein (~70% loss). I'd like to know if anyone has a suggestion of another desalinization method other than dialysis. My protein is ~6kD and is negatively charged.
You most likely had a membrane with large pores, you can run ultrafiltration with amicon filters with MWCO of 3000 Da, you will still lose some protein. Don't know if you can buy filters with lower MWCO. You can also try gel filtration which is expensive if you want to process a lot of sample. If your sample is small gel filtration will work.
It looks like your protein escaped from the dialysis bag because the latter had a too high Mr cut-off. Use a low cut-off dialysis membrane (e.g. 2 kDa), if available. Alternatively, you could use an ultrafiltration cartridge with a similarly low cut-off to concentrate your protein, dilute it in the final buffer and repeat the operation a few times until the original buffer has been replaced to a sufficient extent. You could also use a gel filtration column that will exclude your protein, e.g. Sephadex G-25 or equivalent.
I think that gel filtration using Sephadex G25 could not be a good idea. The size exclusion limits of this gel is 5KDa and the protein has a MW of approx 6KDa so it is too close and could result in an inefficient separation from the salts. On the other hand, are sure that are losing proteins because they are flushing through the membrane? On the other hand, are sure that are losing proteins because they are flushing through the membrane? Do you detect your protein in the eluate from the dialysis? Is well estimated the protein concentration before dialysis? Sometimes some substances cause positive interferences and overestimate the real content and of course, after dialysis they are removed and lower protein concentration is obtained but real. Anyway both previous comments are correct and you can try ultrafiltration using a low cut-off dialysis membrane (1KDa).
You could run the protein through a reverse-phase HPLC column (probably C8) eluted with acetonitrile/water/TFA. The salt will come out in the void volume. The eluted protein can then be lyophilized to remove the solvent. This assumes your protein will tolerate such treatment.
To suggest you the best alternative is important to undertand whats happened in your dyalisis:
The Protein lost was associated to sample precipitation? to membrane damage? to protein adhesion to the membrane? Which is in your opinion the cause of that lost? Which were the starting and the final buffers?
However If you do not have high sample volumes or you can concentrate your protein until low volumes (1-2ml) you can certanilly try to use desalting as approach.
The most used prepacked coloumns (gravity flow - you do not need any insturmentatiob) for this purpose are the PD-10 (GE) that are based on G-25 supedex resin and they can support up to 2.5ml of sample for each runs. Is it true that the MW of your protein is close from the coloum exclusion limit (6KDa vs 5KDa) but i used it in the past for a protein of 7Kda and it worked quite well. It wil depends from the shape of your protein. If you have protein sample to spend for trials and you can bu the colouns is it a very fast and simple test.
Alternativelly you can try the PD MidiTrap G-10 (Ge) that are based on the G-10 resin.