Supramolecular assemblies – formed by the self-assembly of hundreds of protein subunits – are part of bacterial nanomachines involved in key cellular processes. Important examples in pathogenic bacteria are pili and type 3 secretion systems (T3SS) that mediate adhesion to host cells and injection of virulence proteins. Structure determination at atomic resolution of such assemblies by standard techniques such as X-ray crystallography or solution NMR is severely limited: Intact T3SSs or pili cannot be crystallized and are also inherently insoluble. Cryo-electron microscopy techniques have recently made it possible to obtain low- and medium-resolution models, but atomic details have not been accessible at the resolution obtained in these studies, leading sometimes to inaccurate models.I propose to use solid-state NMR (ssNMR) to fill this knowledge-gap. I could recently show that ssNMR on in vitro preparations of Salmonella T3SS needles constitutes a powerful approach to study the structure of this virulence factor. Our integrated approach also included results from electron microscopy and modeling as well as in vivo assays (Loquet et al., Nature 2012). This is the foundation of this application. I propose to extend ssNMR methodology to tackle the structures of even larger or more complex homo-oligomeric assemblies with up to 200 residues per monomeric subunit. We will apply such techniques to address the currently unknown 3D structures of type I pili and cytoskeletal bactofilin filaments. Furthermore, I want to develop strategies to directly study assemblies in a native-like setting. As a first application, I will study the 3D structure of T3SS needles when they are complemented with intact T3SSs purified from Salmonella or Shigella. The ultimate goal of this proposal is to establish ssNMR as a generally applicable method that allows solving the currently unknown structures of bacterial supramolecular assemblies at atomic resolution.