Hi Tiwa Ogunleye . Can you provide more detail? Is the protein precipitating? There are additives that you can have in buffers that stabilize proteins.

How was the protein prepared? Is it still in the membrane, or was it detergent-solubilised? What dialysis membrane did you use, how did you prevent the protein from sticking to it? Is your protein biologically active in dd water? Usually you require a buffer (10 mM HEPES or Tris, pH around 7-7.5 should do the trick). Extracellular proteins require high Na + Ca (e.g., PBS), intracellular proteins high K + Mg for optimal stability (for example 5:1 buffer: 82 mM KCl, 17 mM NaCl, 10 mM HEPES-KOH pH 7, 1 mM MgCl2, 0.2 mM CaCl2). Cytosol is also reducing (2 mM DTT), extracellular fluid oxydising. 1 mM PMSF (freshly prepared from 200 mM stock in isopropanol, in water it decomposes within 20 min. Careful: very toxic) will kill many proteases, but more complex protease inhibitor cocktails are often required. Total osmolarity should be about 380 mosm, add glycerol or sucrose as osmolyte. These are basic recipes, some proteins require their ligands or other substances for optimal stability.

Membrane proteins require a detergent to keep them in solution. If dialysis dilutes the detergent that was used in the purification too much, the protein will precipitate. Some detergents dialyze poorly because the micelles are too large to pass through the pores of the dialysis membrane. Other detergents are easily removed by dialysis. Putting the detergent in the dialysis bath may then be necessary.

Dialysis also may cause the sample volume to increase due to osmotic pressure. If no detergent is in the dialysis bath, the detergent concentration in the sample will decrease due to dilution. If it falls too low, the protein will precipitate.

You should always have a pH buffer in the dialysis bath to maintain the pH of the protein solution at its optimum for protein stability. The only exception would be if your goal is specifically to get the protein into distilled water.

Thanks @Engelbert protein was grown in E-coli, still in the membrane and inactive in water.

Thanks @Shapiro, I will try dialysing in HEPES buffer with a low detergent concentration