I work with Lys49-PLA2s from snake venom and we observed two dimer models, one with PEG in the hydrophobic channel and the other without it. The presence of PEG made some changes in the quaternary structure and in the space group of the crystal. Thus, in our case, the impurity (PEG) helped us to find a different state of the protein.

Basically, salts and organic modifiers in your crystallization screens affect the relative strength of hydrophobic and electrostatic contribution to crystal contacts, while the pH directly affects the charge of your protein. Ions, buffers, detergents and so on can bind to your protein. Both crystal contacts and reagents used for purification and crystallization can influence the conformation of flexible regions of your protein. In a protein that assumes multiple conformational states in the course of its function, they can stabilize one state relative to an other state.

Dear Ravi,

It is difficult to answer this question basically because for me it is not clear. Saying that, the interpretation of a 3D model will be harder as the resolution limit is lower. There is a direct relationship between resolution limit and crystal quality and crystal quality depends on impurities. Now, being here, other contaminant proteins, un-proper folded molecules or different aggregation states of your protein can be the most dangerous impurities affecting crystal quality.

Now, if you are referring to the interpretation of the 3D model influence by the presence of impurities, following Annemarie answer, this is an issue going on and on regarding structures obtained from x ray crystallography or NMR. This could be in principle addressed by obtaining two different polymorph of the same protein.

Some times the right type of impurities can work magic in crystallization - already Perutz found that you need a whiff of toluene to get really good crystals of hemoglobin, and Blundell told us an about insulin crystals that could not be reproduced until they found out that for the initial crystals, tap water was used which fortuitously contained just the right trace of zinc ions needed - when distilled water was used, crystallization failed. Nowadays crystallization screens explore a huge number of additives, but predicting which ones will help and which ones will hinder crystallization of a particular protein still eludes us.