If a protein has at least 15 free Cysteine residues, is getting expressed but is difficult to purify; how can the purification procedure be optimized to get augmented amounts of stable protein?
I'm presuming you have already tried adding reducing agents during the purification? I normally prefer using TCEP as it is stable in solution for a long time. Have you controlled that the protein is produced in a soluble "happy" state, or is it produced into inclusion bodies?
Besides reducing the cysteines, and thus prevent any disulfide-bond formation leading to aggregation, you're down to standard techniques for trying to optimise the purification (using various purification tags, changing pH, salt concentration etc).