Protein crystallization plays an important role in drug design and delivery as well as in other applications. Researchers therefore have devoted considerable time and energy to understanding the process of protein self-assembly. In particular, they have sought solution conditions that are conducive to crystallization. Significantly, these studies have shown that weakly attractive protein interactions correlate strongly with conditions that are conducive to protein crystallization, whereas repulsive and strongly attractive protein interactions do not.Investigators have characterized weak protein interactions in terms of slightly negative values of the osmotic second virial coefficient, B22; this range of values sometimes is referred to as “the crystallization slot.” However, because few methods are available for efficient measurement of B22, they have found it difficult to use the values as a guide for protein crystallization. In the March 12 issue of the Journal of the American Chemical Society, researchers at the University of Delaware in Newark reported an alternative technique. With this method, dubbed self-interaction nanoparticle spectroscopy, proteins are adsorbed on the surface of gold nanoparticles, and the conjugates are added to solutions of interest for crystallization. The optical properties of the nanoparticles then reveal different types of interactions in the solutions. “The important thing about nanoparticles is that their optical properties are dependent on separation distance,” said Peter M. Tessier, the first author of the study, now with Rensselaer Polytechnic Institute in Troy, N.Y. “For this reason, they are attractive for characterizing protein-protein interactions.” The investigators found that the optical properties correlate with B22 values and, serendipitously, that those values associated with weakly attractive interactions correspond with the maximum change in color in the nanoparticles. Thus, the new technique can help determine solution conditions that are conducive to protein crystallization. The researchers demonstrated the method by adding protein-gold nanoparticle conjugates to various solutions and monitoring the aggregation behavior with an ultraviolet-visible spectrophotometer made by PerkinElmer Inc. of Waltham, Mass. They performed dynamic light scattering and transmission electron microscopy to confirm the behavior. They used a Brookhaven Instruments Corp. light-scattering apparatus outfitted with an argon-ion laser and a correlator for the former, and an FEI Co. transmission electron microscope equipped with a multiscan CCD camera for the latter. Researchers have reported a technique with which to identify protein solutions that are conducive to protein crystallization. Previous studies have shown that weakly attractive protein interactions (represented in the center schematic) -- as opposed to repulsive (left) or strongly attractive (right ) — correlate strongly with conditions that are conducive to crystallization, and the new technique can detect such interactions based on color change in conjugates of protein and gold. Reprinted with permission of the Journal of the American Chemical Society.The experiments demonstrated the high efficiency of self-interaction nanoparticle spectroscopy, and they showed that the technique is comparable to empirical crystallization screens in terms of protein and time consumption. (Currently, researchers mostly use such screens to identify solutions conducive to protein crystallization. Empirical crystallization screens often fail, however. Also, because they are empirical, they give no measure of how close to ideal the conditions are.) Tessier further noted that the technique offers the advantages of being relatively simple and of allowing visual detection. Other methods recently have been proposed for measuring B22, including self-interaction chromatography and updated forms of static light scattering and size-exclusion chromatography. Indeed, some of these provide significantly improved efficiency over the more conventional techniques. However, none offers the ability to assay protein self-assembly in a parallel format, which would improve efficiency significantly. The investigators noted another method for assaying protein self-association in a parallel format: developing microfluidic versions of established methods of measuring B22. With this approach, parallelization could be achieved using lab-on-a-chip technologies. Chromatography-based techniques such as self-interaction and size-exclusion chromatography hold promise, for example.In fact, researchers already have reported a miniaturized version of self-interaction chromatography — a technique that offers the highest efficiency for measuring B22. However, obstacles such as complex fabrication and assembly still must be overcome before quantitative, parallel microfluidic systems based on such techniques are possible. There are challenges to be addressed with self-interaction nanoparticle spectroscopy as well. For instance, the researchers are not certain whether it can be applied to proteins with different physical and chemical properties. Both the gold nanoparticles and the proteins that were used in the study are negatively charged, Tessier explained. However, he said, “depending on their pH, some proteins are going to be positively charged. When we added such proteins to the solutions, they aggregated to the gold nanoparticles nonspecifically.” They are addressing this problem by tailoring the properties of the protein surface. Also, the technique does not offer quantitative measurement. Although studies have designated a specific range of values for B22 that are conducive to protein crystallization, such specificity is not possible with self-interaction nanoparticle spectroscopy because the color change upon which the technique relies depends on a variety of factors, including the concentrations of proteins and nanoparticles, the ratio of the sizes of the proteins and nanoparticles, and the aggregation time. The researchers noted, though, that the qualitative nature of the technique should be considered in the context of its high efficiency and the ease of parallelization.Contact: Peter M. Tessier, Rensselaer Polytechnic Institute, Troy, N.Y.; e-mail: tessier@rpi.edu.