The type I polyketide synthases (PKS) are giant, multifunctional enzymes. They are responsible for the biosynthesis of a full variety of natural compounds, the so-called polyketides. These compounds are involved in important biological processes such as the cell wall biogenesis or pathogenicity of various bacteria. Some polyketides also exhibit very interesting and widely used pharmacological properties (antifongal, antimicrobial, immunosuppressive, antitumoral') The polyketide synthases catalyse the elongation of a substrate using several catalytic domains and a wide range of starter and extender units. The number and the nature of the domains present on a PKS will determine which type of polyketide will be synthesized. Such a concept that relies both on molecular recognition and iterative processes may be called molecular programming. The aim of this project is to establish the molecular and structural bases that are responsible for the molecular programming of the PKS. To achieve this goal, three teams of the « Institut de Pharmacologie et de Biologie Structurale » decided to share their scientific expertise within a consortium. These teams have complementary skills in molecular biology and protein engineering (Team C. Guilhot), analytical and structural biochemistry (Team M. Daffé), and structural biophysics and biology (Team L. Mourey). A coherent set of PKS produced by Mycobacterium tuberculosis has been selected for this project. Deciphering the structure-function relationships of complex enzymatic machineries constitutes a fantastic challenge. It should be noticed that, to date, no structure of such a PKS has been solved. Several complementary strategies have been set up to study these enzymes. We propose to solve the structures of the full-length proteins using X-ray crystallography. To overcome eventual difficulties in crystallizing these large enzymes, we also intend to work on the various catalytic domains or fragments of the proteins, whose limits will be determined using several approaches. Thus, we will use a hybrid approach where structural information obtained at low resolution for the entire proteins will be combined with the high-resolution crystal structures of individual domains. The structural information will be combined with data deduced from complementary enzymology and molecular biology studies that comprise the elaboration of enzymatic assays and the study of ponctual mutants of the selected enzymes. This original and ambitious project will generate new and important results with respect to the structural organization and molecular programming of complex enzyme systems, paving the way for the rational modification of these enzymes. The complementarity and the synergy of the three implied teams are attested by a set of joint publications and contracts of collaboration. In addition, preliminary experiments and validations were realized on some of the selected PKS.