Contamination from an affinity column: an encounter with a new villain in the world of membrane-protein crystallization.

Attempts to crystallize AtNTT1, a chloroplast ATP/ADP transporter from Arabidopsis thaliana, revealed an unexpected contaminant, Strep-Tactin, a variant of streptavidin that was used during purification of the protein. Although it was present in very small amounts, crystals of Strep-Tactin were reproducibly grown from the AtNTT1 solution. AtNTT1 was overexpressed in Escherichia coli and purified from detergent-solubilized membrane fractions using Strep-Tactin affinity chromatography based on an engineered streptavidin. The contamination of protein solutions purified on Strep-Tactin columns has never been described previously and seems to be specific to membrane proteins solubilized in detergents. Trace amounts of Strep-Tactin were observed to be eluted from a Strep-Tactin column using several routinely used detergents, illustrating their possible role in the contamination. This finding raises an alarm and suggests caution in membrane-protein purification using Strep-Tactin affinity columns, where detergents are essential components. The small crystals of contaminant protein led to the structure at 1.9 Å resolution of Strep-Tactin in complex with desthiobiotin.