NVK Nederlandse Vereniging voor Kristallografie, Werkgroep Eiwitkristallografie

Report on the 6th Protein

Delft, 26-9-2003
On September 26th the NVK protein section gathered with 52 attendants at the Bruker-Nonius site in Delft for an interesting program. The scientific program was organized from Amsterdam, but all the local arrangements were thanks to the Bruker/Nonius people, who also arranged an interesting tour of the facilities after lunch. Support from Bruker/Nonius and Organon are gratefully acknowledged.

The scientific program was started off by a talk by Shizuko Tsuji (Utrecht), who spoke on the activation of von Willebrand factor (vWF). Based on the crystal structure of the complex of vWF A1 domain with glycoprotein 1bα (Gp1b) a mechanism has been proposed that involves the relative arrangement of the C- and N-terminus with respect to eachother and the protein. By analysing the A1 domain with longer N- and C-terminal domains this hypothesis has now been tested. The next speaker was Lieven Buts (Brussel), who spoke on the lectins in the pili of bacteria and the strand exchange that is involved in intersubunit interactions.

His talk was followed by a presentation by Anita Coetzee (Bruker/Nonius) who showed data on small crystals and the effect of a higher brilliance beam. There is an impressive effect on I/sigma for the weaker reflections. She also presented a highly redundant data set on thaumatin that was sufficient to solve the crystal structure on sulphurs alone. After lunch the first guest speaker was Gert Vriend (Nijmegen) who went into detail on the need for validation and the kinds of errors that are found in crystal structures. The main message was that of course we should validate ourselves, prior to submission to the PDB, the second that people should really use the Whatif output for improving their structures. Steven Ness (Leiden) discussed an integrated package that should be able to do automated structure solution. The integration is over the packages Crunch2 (heavy atom search), Solomon (density modification), BP3 (maximum likelihood phasing) and Arp/Warp (model building). The authors of the individual programs were all in the room indicating that there is still a substantial amount of crystallographic software development in the Netherlands. The new combined package will be implemented in CCP4, and thus be very easily accessible to everyone soon. Patrick Celie (Amsterdam) described the progress in ligand-binding to Acetylcholine binding protein. Binding of nicotine and carbamylcholine correspond well with biochemical data on the nicotinic receptors, but also explain specific features in the protein. After tea Thomas Barends (Groningen) described his problems with a twinned crystal as an entertaining detective story. It pointed at the difficulty in recognizing twinning in the presence of NCS. Finally, Gerard Kleijwegt (Uppsala), as special guest, described the EDS electron density server, and its function in validating structures. This very worthwhile initiative fulfills a great need, not in the least since it points out simple errors that were not realized by the authors, as became clear in the discussion. The combination of the two guest speakers was somewhat fortuitous, but really drove the point home that validation of data is critical and that protein crystallographers should really take good care in analysing the validation data. Some discussion arose that it were primarily the Dutch who enjoy pointing out errors in other peoples work. It should be noted that this is not only true for macromolecules but also extends to the small molecule analysis. Hence it was fitting to have these speakers at the NVK meeting.