NVK Nederlandse Vereniging voor Kristallografie, Werkgroep Eiwitkristallografie

Report on the 5th Protein meeting

Leiden, 27-9-2002
The meeting took place in the Gorlaeus Laboratory, hosted by Jan Pieter Abrahams and his group. Approximately 50 people took part from the various crystallography groups in Holland and Belgium, as well as a delegation from the Dubble beamlines. The meeting was opened by Eric Huizinga of the Utrecht group, who described the work on the interaction between platelets and vessel wall, by Gp1b and von Willebrand factor. In this fascinating complex, initial contacts are built up of electrostatic interactions between the proteins, followed by more extensive close contacts that include a conformational change of a disordered region to a beta-hairpin. This observed conformational change from free to bound Gp1b beautifully explains the effects of a number of gain-of-function mutants that cause a bleeding disorder. Other classes of mutants could also be explained but the most surprising ones, are those that begin to provide an answer to the question why this interaction is partially dependent on shear-stress. A model was proposed in which pulling on the N- and C- terminus by the rest of the extenden vWF protein is necessary to open up its interaction site for Gp1b binding.

Subsequently Julie Bouckaert (Brussel) described their interesting work on a fimbrial protein FimH, that is responsible for adhesion of bacteria involved in urinary tract infections. She described a mannose binding site in this protein, that could already be observed at lower resolution, but was confirmed at 1.4 a to be a superbly bound molecule with very many hydrogen bonds, which explains its nanomolar affinity. Almost every mutation in this binding site relieves binding , and prevents infection by the bacteria. Interestingly she found a ligand bound in the binding site that was not expected, presumably from the culture media. This ligand had higher affinity, because of a hydrophobic tail that was bound to the protein, and is suggestive for drug design.

Bram Schierbeek (Bruker-Nonius) described a series of experiments on weakly diffracting MutS crystals on the CCD detector. Images on their system could be taken in an impressive 30 seconds on a home source, which allows faster testing of crystals. However, there was no good comparison with a modern system with good mirrors available yet. Some discussion occurred on the issue of the sensitivity of a CCD detector versus an image plate which was inconclusive. A possible advantage of such a short data collection time is that it allows impressively redundant data to be collected in an acceptable time period. In collaboration with the Sheldrick group they had worked on a trypsin data set that was collected to 1.4 angstrom and phased with the new Shelx modules, using on the sulfur and calcium signals, with stunning phases. He also showed the Bruker-Nonius caps that can replace cryotongs and their automated data-collection robot.

In the break a report from the Dubble beam line was made. A team of three people is now actively working on making the protein station workable, including Klaas Decanniere (Brussel) and Maxim Kuil (Leiden). They will visit all the Flemish and Dutch groups in the upcoming period to come with a list of issues to address.

After lunch Rene de Jong (Groningen) described an unusual enantionselective halohydrin dehalogenase HheC from Agrobacter. It has a number of interesting products, both for the forward and for the backward reaction, which are difficult to synthesize in enantiopure form. The direction of the reaction is strongly dependent on the pH and this could be explained by a proton pathway that was visible in the crystal structure. Bauke Dijkstra (Groningen) described the structure of quinohemoprotein alcohol dehydrogenase - that had proven particularly difficult to solve. This protein contains two cofactors, a heme and a PQQ and is a redox enzyme in bacteria, involved in catabolic pathways. Crystals had already been grown by Eric Huizinga 15 years ago, who because of the spectroscopic properties of the heme was able to deduce the relative orientation of the heme groups. A MAD dataset on the Fe in the heme gave the Fe positions accurately, but was not sufficient to phase the 752 amino acid protein, but together with the information about their orientation Henriette Rozenboom was able to place the Heme and phase the protein. It consists of a beta-propeller domain that binds the PQQ and a cytochrome C domain that binds the heme. A tetrahydrofuran-2-carboxylic acid molecule was found unexpectedly in the binding site, again likely to be a remnant from the medium. This ligand allowed speculation on the redox pathway between the two cofactors. The stacking of a vicinal disulfide on the PQQ is likely to be relevant for the long range transfer of electrons from the PQQ to the heme, while a water channel is likely to be involved in the protein transfer.

Mark van Raaij (Leiden) presented the structures of two overlapping proteolytic fragments of the trimeric protein gp12 that forms the short tail fibers of bacteriophage T4. The LPS-binding domain located at the tip of the short-tail fibre has a remarkable fold that is devoid of regular secondary structure and is stabilized by a Zn2+ ion located at a three-fold axis and coordinated by six histidine side chains. As the final speaker of the day. Finally, Mark Hilge (Leiden) presented the NMR structure of the nucleotide-binding domain of the sodium pump Na,K ATPase. Analysis of structures with and without ATP reveal conformational changes in the N- and C-termini. The termini connect the nucleotide binding domain to the phosphorylation domain thus providing a pathway of cummunication between the two domains.

The day was finished with a borrel, hosted by the Leiden group.