Unit: Macromolecular crystallisation and X-ray diffraction facility (Platform)
Director: ALZARI, Pedro M.
The goal of the platform is to provide the research teams working on macromolecular crystallography at Pasteur with the equipment and technology required for automatic crystallogenesis, computing and X ray diffraction, as well as to develop high-throughput methodologies for gene cloning, protein production and crystallization, by automating the different steps of the process. During the last eighteen months, over 100 proteins have been submitted to automatic crystallization assays, with a success rate (diffracting crystals or exploitable hints) of roughly one every three proteins. Most of the proteins originated from the project on the structural genomics of mycobacteria (about 60%) described below, while the rest came from the different labs of protein crystallography or direct collaborations of the PF6 with non-crystallographers.
Structural Genomics of Mycobacteria
With the availability of the complete genome sequences of M. tuberculosis, M. bovis, M. bovis BCG and M. leprae, we have unparalleled amounts of information and knowledge about the bacillus, and direct access to the complete gene set and the entire proteome. Three years ago, we started a structural genomics initiative, involving several Pasteur laboratories with complementary expertise in different aspects of structural biology and mycobacterial research. The long-term objective of the project was to use an integrated, multidisciplinary approach combining proteomics with structural and functional genomics, with the explicit aim of identifying and validating novel drug targets. The project was funded by grants from the Institut Pasteur, the National Genopole Network, the Ministery of Research, and two projects from the European Commission: X-TB (structural and functional genomics of M. tuberculosis) and SPINE (structural proteomics in Europe). Target selection included both known proteins, selected for their potential as drug targets, as well as hypothetical proteins conserved in (and restricted to) mycobacteria and/or actinomycetes for which no functional annotation is currently available. Indeed, functional information is lacking for about 40% of the proteins predicted in M. tuberculosis, and many of these may participate in novel metabolic pathways or confer unique properties like the ability to persist indefinitely in infected tissue. A comparative analysis of the deduced proteomes from M. tuberculosis and M. leprae led to the identification of a set of ~300 proteins that are exclusively found either in mycobacteria or in actinomycetes, but which have no counterparts in other organisms. The rationale behind this approach was that, since these genes are conserved in the degraded genome of M. leprae (which has lost as many as 60% of the M. tuberculosis functional genes), their protein products might presumably be important for survival and could provide potential drug targets.
As of September 2004, 356 mycobacterial genes have been cloned into bacterial expression vectors using Gateway or related technologies, 115 gene products were expressed as soluble proteins and 65 of them purified to homogeneity and subjected to crystallization trials. As expected, a major bottleneck of the pipeline was protein solubility, since less than one third of the proteins can be expressed in a soluble form. We are using different approaches to address the solubility problem, including the cloning of orthologous genes from other mycobacterial genomes (M. leprae, M. smegmatis) or the optimization of expression conditions using cell-free systems before cloning into a bacterial vector. Suitable crystals or exploitable hints have been obtained for roughly one third of the 65 mycobacterial proteins subjected to crystallization trials and we have determined so far 11 crystal structures, mostly using SAD or MAD techniques on SeMet-labeled proteins of unknown function (Figure 1), and diffraction data has been collected for 3 other proteins, for which structure determination is in progress. The up-to-date status of the project is described in the dedicated Web site "Structural Genomics of Mycobacteria" (http://www.pasteur.fr/SGM).
Other ongoing projects
During the last year, the PF6 started to participate in other research projects involving crystallographic studies of single proteins and protein complexes. These structural studies include several proteins involved in Trypanosoma cruzi metacyclogenesis (an Amsud-Pasteur project), secreted Chlamydia proteins of unknown function that could be involved in pathogenicity (coll. A. Dautry, IP), protective anti-cholera antibodies specific for the LPS antigen (coll. J.M. Fournier, IP), human UMP-CMP kinase and complexes with inhibitor (coll. D. Deville-Bonne, IP), P. falciparum serine-proteases (coll. J.C. Barale, PTR, IP), mycobacterial TMPK-inhibitor complexes (coll. H. Munier, IP), and Yak1 kinase from pathogenic Candida (coll. G. Janbon, PTR, IP).
Figure 1: Three proteins of unknown function from M. tuberculosis.
Keywords: structural genomics, gene cloning, protein purification, crystallogenesis, X-ray diffraction, LIMS