PDB Community Focus: Brian W. Matthews
Brian W. Matthews is a long-time PDB depositor. From 1982 through the present time, he has contributed approximately 500 sets of coordinates to the archive -- more than any other single author. A Professor of Physics and a member of the Institute of Molecular Biology at the University of Oregon, Matthews' numerous distinctions and achievements also include an appointment as a Howard Hughes Medical Institute investigator and membership in the U.S. National Academy of Sciences. He is involved in studying some of the fundamental problems in biology, and his X-ray studies have always played a critical role in his research. The PDB recently interviewed Prof. Matthews on his experiences in this field:
PDB: Over the years you have been an integral part of the tremendous technical revolution that has taken place in protein crystallography. How long did it take you to solve your first structure and what methods did you use? How long would it take to do the same structure today and why?
Prof. Matthews: The first structure that I solved was of a "small molecule" of 14 atoms including a sulfur that was used as a "heavy atom". It took me over a year in part because all of the calculations had to be done by hand and also because the Patterson function gave misleading information regarding the location of the sulfur. Today such a structure would be solved in hours if not minutes.
The first protein structure determination with which I was involved was that of alpha-chymotrypsin. David Blow had been working on the project at the MRC Lab in Cambridge and he and Michael Rossmann had used the structure as a test case during their early development of the molecular replacement technique. Their calculations suggested that the two molecules in the asymmetric unit were probably related by a local two-fold axis. Shortly after I arrived in Cambridge, Barbara Jeffries, a technician working on the project, found that the inclusion of 2% dioxane in the crystallizing medium eliminated persistent twinning.
This made it possible to grow large, untwinned crystals and to make a serious attempt at high-resolution data collection. Data collection was by precession photography. A single data set necessitated 20-30 film packs, took perhaps three months to collect and a similar amount of time to process. Paul Sigler joined the project about six months after I had arrived and Richard Henderson joined the group later as a starting graduate student. From the time that I joined the project until the structure was solved took over three years, but as I mentioned, the project had already been underway for several years. We used the method of isomorphous replacement following the pattern set with myoglobin and lysozyme. In the case of alpha-chymotrypsin we could average the density for the two molecules in the asymmetric unit and this substantially improved the quality of the electron density map. We also included information from anomalous scattering measurements incorporating ideas that were being developed around that time. Given high quality crystals the structure could probably be solved today within a month or so. As I mentioned, however, the initial crystals were subject to frequent twinning. Without the introduction of dioxane by our technician, Barbara Jeffries, the structure might still remain unsolved even today.
PDB: How has your research program evolved over the years? What role has crystallography played in that evolution?
Prof. Matthews: Because of my involvement with the alpha-chymotrypsin project, I had an early interest in proteolytic enzymes. Therefore it was natural for me, when I started my own laboratory at the University of Oregon, to work on the thermostable protease thermolysin. Because it was a zinc peptidase we were curious as to whether the structure might be related to that of carboxypeptidase which by then had been determined by the Lipscomb group at Harvard. Also I had begun to develop an interest in protein stability and folding.
Crystallography has been central to all of the subsequent work that we have done. The structure of the Cro protein immediately suggested how it might bind to DNA. The introduction of site-directed mutagenesis, coupled with crystallography, led to our exploitation of the T4 lysozyme system as a way to try to understand the structural basis of protein stability. It also contributed to what has been described as our "pollution" of the Protein Data Bank with structures of mutant lysozymes.
PDB: You deposited your first structure approximately 20 years ago - what do you think crystallography will be like for someone depositing their first structure 20 years from now?
Prof. Matthews: I was extremely excited when we were able to determine our first structure in my own lab. This was thanks to the outstanding contributions of Peter Colman, my first postdoc, together with Hans Jansonius who spent 1971 in Eugene as a sabbatical visitor. I am pleased to see that students and postdocs who determine their first structure in my group still share that same excitement. I certainly hope that this will remain 20 years in the future. At the time that I started my career in structural biology it was sufficient to be just a crystallographer. Now knowledge of crystallography is just one tool in the repertoire of skills that are necessary. I like to think that crystallography remains as the most powerful tool that we have and hope that it will remain so in the future.
PDB: The PDB plans to be there and to be ready - do you have any advice for us based on your experience?
Prof. Matthews: Based on my own experience I have nothing but good things to say about the PDB. On a technical level the PDB has adapted extremely well to handle the ever-increasing number of depositions. I have also found the PDB willing to consider suggestions for improvement. Also the PDB has played an essential role in helping to ensure that key structural information are preserved and made available in a timely fashion to the community at large. Keep up the good work!