PROTEIN CRYSTALLOGRAPHY. Michael G. Rossmann, Department of Biological Sciences, Purdue University, West Lafayette, IN 47907-1392

It was not clear in the early 50's that proteins would have unique structures rather than be colloidal aggregates. Nevertheless, early attempts at determining the structures of proteins included studies of lysozyme (Linus Pauling), insulin (Dorothy Hodgkin) and hemoglobin (Max Perutz), stimulated by the availability of diffracting crystals. W.T. Astbury, J.D. Bernal, I. Fankuchen and others were starting to elucidate some features of proteins, such as the [[alpha]]- and [[beta]]-structures in fibers and even the structure of the rod-shaped tobacco mosaic virus.

Perutz had demonstrated that hemoglobin contained rod-like structures and that these rods were probably Pauling's [[alpha]]-helices. He had also tried a variety of innovative techniques to solve the phase problem using the shrinkage of cell dimensions produced by drying and the changes in structure amplitudes caused by alterations in salt concentration.

Bragg had intuitively used the isomorphous replacement method in solving the structure of Na and KCl. J.M. Robertson had demonstrated the possibility of isomorphous replacement in the study of phthalocyanines. Perutz, in the face of skepticism, showed that heavy atoms could diffuse into crystals and produce measurable differences in the diffraction pattern. In 1953, he was able to produce a map of horse hemoglobin in projection, having phased the centric monoclinic (h0l) reflections. It was to take another 6 years before a three-dimensional 5.5 Å map was possible. That required a group of young assistants to measure with rulers hundreds of maxima representing Bragg reflections. All these had to be sorted and used in a home-built computer having the equivalent of only 10Kbytes of memory. In the meantime, John Kendrew had made progress in the structure determination of myoglobin. In 1959, it was possible to recognize the extraordinary evolutionary conservation of the globin fold in comparing the three-dimensional structures of myoglobin and hemoglobin, as well as seeing the atomic details of Pauling's [[alpha]]-helix. The techniques developed for these first protein structure determinations are still the essence of today's technology.