Pittsburgh diffraction conference

The program of the 51^st Pittsburgh Diffraction Conf., held in Valley Forge, PA, USA on Nov. 3‑5, 1993 consisted of symposia on Knowledge Based Molecular Design, Growth Factors and their Receptors and Area Detector Data Collection for Small Molecules.

Papers describing a method of calculating the association energy of quaternary complexes from their atomic coordinates

(M. Lewis, Johnson Research Foundation) and an analysis of the energetics and the role of water in protein ligand interaction based on analyzing over 400 protein ligand complexes in the Protein Data Bank (J.J. Keske, SmithKline Beecham Pharm.) set the stage for presentations of structure based ligand design based upon crystallographic studies. Detailed summaries were presented of recent applications of the technique to the streptavidin biotin complex (F.R. Salemnic, 3-Dimensional Pharm., Inc.), Thymidylate Synthase inhibitor complexes (D. Matthews, Agouron Pharm.), Purine Nucleoside Phosphorylase (C.E. Bugg, U. of Alabama,), Lipase structures (B. Rubin, Sterling Winthrop Pharm. Res.), the FK-506 Binding Protein Calcineurin System U.W. Becker, Merck Res. Lab.) and the Lec-EGF growth factor form of E-selectin (B.J. Graves, Hoffman LaRoche, Inc.). Recurring themes in the presentation were the need to study complexed and uncomplexed forms of the proteins in order to elucidate flexibility and specific interactions, the importance of iterative cycles of ligand design, synthesis, biochemical evaluation and crystallographic analysis; and the importance of high resolution studies to gain reliable data on subtle conformational changes and the movement of ordered water. Charlie Bugg described making 60 potential inhibitors to secure 6 patents and net two promising candidates in clinical trial. There was a good deal of flap about flaps. Many of the structures under discussion, including the lipases and the proteases have highly flexible flaps that undergo significant movement from one crystal form to another. Does this demonstrate the power of crystal packing forces? Does it indicate the inherent flexibility of the system? Is it related to the dynamic states postulated to satisfy NMR data? Does it relate directly to physiological functions? All of the above? Don't miss the next exciting episode. (W.L. Duax)

The session on Growth and Differentiation Factors and their Receptors focused on the nature of interactions between growth factors and their receptors and the design of therapeutically relevant agonists and antagonists of the ligands. S. Abdel-Meguid presented a description of the anti-parallel four-helix bundle, its connectivity and handedness. The up-up-down-down connectivity of the anti-parallel four-helix bundle is a structural motif common to many cytokines and growth factors. F. Basan of the DNAX Research Inst. described the evolutionary kinship of many of the cytokine receptors and their ligands, despite little amino acid sequence homology. He also analyzed the similarities and differences between the structures of the ligands, and discussed their mechanism of receptor engagement. B. de Vos of Genentech described the interaction between growth hormone and its receptor. He and his colleagues have determined the structures of the 1: 1 and 1:2 ligand/receptor complexes. These structures, and the structure of unliganded human growth hormone, have formed the basis for understanding of signal transduction at the molecular level. P. Young (SmithKline Beecham) describe mapping the receptor binding site on the surface of Interleukin-1B (IL-1b) using deletion and site specific mutagenesis, as well as partial proteolysis of ligand-receptor complexes. His work was aided by analysis of the crystal structures of IL-1b and IL-1a. Finally, A. Wlodawer of the National Cancer Inst. compared his laboratory's work on the IL-4 structure with other IL-4 structures determined by crystallographic and NMR techniques and presented a model for the interaction of IL-4 with its receptor, based on the structure of growth hormone ligand/receptor complex. (S. Abdel-Meguid, Chair and Reporter)

Protease-inhibitor crystal structures provided the basic information to design more potent inhibitors, to understand specificity differences among proteases, and to explain the development of drug resistance with potential uses for a variety of diseases from emphysema to AIDS. G. Hanson gave a clear description of how different parts of a promising renin inhibitor were altered in order to improve solubility and specificity. A P3 'auxiliary site' deduced from the crystal structures was important for increasing the inhibitor specificity. Small orally active inhibitors were developed for testing as anti-hypertensive agents. He also deduced the interactions of renin with its substrate by models based on the renin-inhibitor structures. Inhibitors of human neutrophil clastase were reported by J. Darnewood and C. Ceccarelli. In this case porcine pancreatic elastase was used as a model for the human enzyme. Crystal structures at 2A resolution with three different inhibitors were described with many details of the crystallography. The crystal structures, free energy of perturbation calculations, and molecular dynamics simulations were used to develop more potent orally active inhibitors for treatment of emphysema The best in vivo inhibitors did not show the best in vitro inhibition. C. H. Chang showed crystal structures of HIV-1 protease with a novel inhibitor that was designed to replace the conserved water molecule with a carbonyl oxygen of a cyclic urea. The carbonyl oxygen of the inhibitor was in an almost identical position to that of the water molecule seen in other HIV protease-inhibitor crystal structures. These crystals diffract to 1.8A, but due to the internal symmetry of the inhibitor in the protease dimer, it was not clear whether the space group was P6, or P6, 22. One problem for developing antiviral agents effective against AIDS was portrayed by J. Erickson who reported the structure of a drug-resistant mutant of HIV-I protease (V82A) in complex with the inhibitor. The crystal structure and models of other resistant protease mutants suggested that the inhibitors bind less tightly to the mutant proteases because of unfavorable interactions of the inhibitor with the mutated amino acids.

The specificity of the long-studied trypsin family of serine proteases was elucidated by J. Birktoft from the structure of a glutamic acid specific serine protease. The specificity for glutamic acid is provided by hydrogen bonds with a histidine and two serine side chains. The charge on the glutamic acid may be stabilized by a novel histidine triad. This was contrasted with trypsin which recognizes basic amino acids by interactions with an aspartic acid. A. Tulinsky discussed the different binding sites on thrombin for its substrate, fibrinogen, and the inhibitor hirudin. Modeling, NMR, and crystallography have led to designs for inhibitors that mimic peptide beta-turns. The structure of thrombin with the fibrinopeptide A suggested an improvement in the bicyclic beta-turn mimetic inhibitor that would make it more similar to the peptide conformation. Thrombin inhibitors were also discussed by P. Weber. Arginine in a boronic acid containing peptidic inhibitor binds in the specificity pocket of thrombin. Crystal structures of thrombin with a series of inhibitors in which the arginine was substituted by lysine, homolysine or ornithine were analyzed to understand the differences in affinity. A hydrophobic binding pocket in the enzyme was also utilized in the inhibitor design to make a smaller molecule that was easier to synthesize. Serine proteases were among the first enzymes to be studied crystallographically, and this session demonstrated that technological advances have made it possible to determine many. crystal structures of enzymes with a variety of different inhibitors. This is a valuable tool for drug design. (Irene T. Weber, Chair and Reporter)

Bob Sweet opened a half day session on 'Experiences with Small Molecule Area Detector Data Collection'. with a general talk on the use of area-sensitive detectors currently available for small molecules. Experiences with the MA image Plate Scanner using both sealed and synchrotron radiation and with a Rigaku RAXIS image plate system were described by Z. Dauter and J. Calabrese respectively. C. Chidester described benchmarking both systems. The FAST system was discussed by H.L. Carrell and R. Scheidt. R. Sparks described a new area detector specifically designed by Siemens to be efficient for short wavelength radiation. The first data sets collected with the new system were detailed on a poster by H.D. He, et al. In all, the reports were quite positive and all speakers were pleased with their new systems.

Sidhu Awards for young crystallographers were presented to T.O. Yeates (UCLA) and M.R. Pressprich (SUNY at Buffalo). In his award lecture 'Molecular Replacement in High Symmetry', Yeates described the use of analysis based upon Dirichlet domains to resolve problems in structure determination used by high symmetry. In his talk 'Geometry and Charge Density of an Excited Electronic State by X-ray Crystallography: The Study of Na Nitroprusside' Pressprich described detection of a lengthening of the Fe-N bond and a relaxation of the cis C-Fe-N angle of sodium nitro prusside in the excited state.