Name: Stefan Knight
PI: Eklund (SLU)
Project: Structure/function studies of virulence associated adhesion organelles from pathogenic Gram negative bacteria
Gender: M
Actual starting date: 1 January, 1997
Administrative starting date: 1 January, 1997
Structure determination of chaperones and chaperone-inhibitor complexes
The structure of the E. coli SfaE chaperone involved in S pili biogenesis has been determined using MIR techniques. The structure solution of SfaE had been held up by lack-of-isomorphism problems, but eventually two common Pt sites could be identified through averaging of difference Patterson maps from a number of non-isomorphous Pt soaked crystals. The resulting low resolution phases allowed me to identify the positions of Se atoms in crystals of selenomethionylated SfaE from the isomorphous differences between SeMet and native SfaE data collected at room temperature in the lab. Refinement of the platinum and selenium parameters in SHARP, followed by solvent flattening using Solomon, resulted in an excellent starting MIR map. Two molecules of SfaE have been traced in the P212121 asymmetric unit and the first round of modelling was completed yesterday. Refinement will be initiated shortly (as soon as I'm done writing this report!). Although the room temperature data only extends to 2.8 Å, excellent cryo data to 2.0 Å are available so that a well refined decent-resolution structure of SfaE should become available early next year.
X-ray data to high resolution have earlier been collected from SfaE crystallised in the presence of two compounds that were predicted to bind in the cleft of PapD-like chaperones, based on computational screening of a chemical database. When a refined model of SfaE is at hand, it will be very interesting to look at these structures to see if either of these compounds bind in the cleft of SfaE as predicted. If indeed they do, the binding of these compounds will form the basis for design and/or identification of improved chaperone inhibitors.
The SfaE structure also provides a much better molecular replacement search model for the FimC chaperone than previously available. FimC and PapD are only about 30% identical whereas FimC and SfaE share about 60% identity. Furthermore, given two different chaperone structures (PapD and SfaE), the definition of a chaperone core structure becomes much more precise. It might thus become possible to solve the FimC structure using molecular replacement techniques.
Structure determination of chaperone/subunit complexes -- FimC/FimH
Over the year, a large number of FimC/FimH data sets have been collected at the synchrotron at Grenoble to try to identify useful heavy atom derivatives. Two MAD data sets from SeMet FimC/FimH crystals have also been collected, as well as a new native data set extending the resolution to 2.5 Å. So far no useful heavy atom derivative has been found although several of the difference Pattersons show some indication of heavy atom binding. I am currently investigating whether a similar approach of Patterson averaging as that used successfully for SfaE could be applied to the FimC/FimH problem.
One of the reasons we have had problems finding heavy atom derivatives of the FimCH complex may be that these crystals grow from 1 M ammonium sulphate solutions at high pH. We have now established a procedure for transferring the crystals to Li sulphate at neutral pH where heavy atom binding might be more amenable to occur. An excellent native cryo data set has been measured at home confirming that the transferred crystals are isomorphous to the original ones, and we are currently using Li sulphate soaked crystals to screen for heavy atom derivatives.
Latest update at 9 March, 1998.