Superficial bacterial folliculitis of dogs (pyoderma) caused by Staphylococcus pseudintermedius is one of the most common diseases seen in small animal veterinary practice, worldwide. Treatment of canine pyoderma often requires aggressive antibiotic therapy and is subject to common relapse of infection. Worringly, methicillin-resistant S. pseudintermedius has recently emerged as a major problem in veterinary clinics worldwide. Although rare, several episodes of life-threatening infections of humans by S. pseudintermedius have been reported with the typical route of transmission being through dog bite wounds. Gram-positive bacteria cell-wall anchored (CWA) protein play a key role in host colonisation and disease pathogenesis and represent targets for therapeutic intervention. For example, the fibrinogen-binding protein ClfA made by S. aureus has been a major focus of antibody-based therapeutics. Very recently, in a team led by our collaborator Prof. Magnus Hook, we have published the structure for ClfA binding to its major ligand fibrinogen, resulting in the design of peptides which block ClfA function. These peptides are currently under investigation as potential therapeutics for treating staphylococcal disease. We have recently sequenced the whole genome of a clinical isolate of S. pseudintermedius isolated from a case of canine pyoderma (unpublished data). Importantly we have identified genes encoding 18 predicted CWA proteins, each of which is unique to S. pseudintermedius. Through the expression of purified recombinant proteins and the use of a heterologous expression host, Lactococcus lactis, we have discovered 2 S. pseudintermedius CWA proteins which mediate binding to both fibrinogen and fibronectin (SpsD and SpsL) and cytokeratin (SpsD) and which are encoded by all strains examined. These data suggest that they play a critical role in host-pathogen interactions and could represent excellent targets for novel therapeutics. In the current proposal we will characterise the molecular interaction between the CWA proteins SpsD and SpsL with their ligands in order to facilitate the design of novel therapeutic approaches for the control of canine pyoderma. 1) Examination of the allelic variation and expression of SpsD and SpsL among S. pseudintermedius strains. Sequencing of genes encoding SpsD and SpsL from 20 S. pseudintermedius strains representing the breadth of species diversity will allow us to determine how conserved the proteins are among the common pathogenic clones. In addition, surface expression of the proteins by the same panel of strains will be examined by ELISA-type assays with specific IgY antibodies. 2). Identification of the bacterial protein domains and host ligand domains required for binding. This phase of the project will involve cloning and recombinant expression of selected sub-domains of SpsD and SpsL in order to test their ligand-binding activity in ELISA-like assays. Using similar assays, recombinant or protease-cleaved fragments of ECM proteins, fibrinogen, fibronectin and cytokeratin will be employed to determine which region of each host protein is the target of adherence. Recombinant proteins will then be used to block the interaction of L. lactis strains expressing each protein or S. pseudintermedius wild type strains, to ECM proteins. 3) Identification of peptides which block the function of SpsD and SpsL. Structural modeling of SpsD and SpsL will be carried out using the programme PyMol based on the crystal structure of ClfA and other related CWA proteins of S. aureus in order to predict amino acid residues which may be involved in ligand binding. Based on this analysis, we will design short peptides which we predict could interfere with ligand binding. In addition, antibodies specific for the ligand-binding domains will be raised and the ability of each of these molecules to block function and inhibit binding to ECM proteins will be tested.

Our sense of taste is our primary means of detecting nutrients and toxins in food, and its function is important for our health and well-being. The principles of gustation are shared in mammals and insects making it possible to use insects as model organisms to understand how chemical information is detected and encoded by the gustatory system. When animals ingest food, nutrients like sugars are detected by cells in taste buds on the tongue, and in insects by neurons in chemosensory sensilla. In general, the gustatory system is organized such that a subset of gustatory cells or neurons is excited by sugars, whereas others are excited by toxic (bitter) compounds, by amino acids, by salts, or by water. In insects, gustatory receptors (Grs) on the membranes of taste neurons selectively bind to classes of chemical compounds (e.g. sugars) and their activation indicates which compounds are present in food. Animals may have as many as a few hundred Grs, but millions of potential ligands exist. We do not know how Gr diversity affords the detection and classification of chemical compounds. Decoding gustation, therefore, first requires the identification of the nutrients or toxins that activate specific Grs in one animal species. This could then be related to the response properties of its Gr neurons and to its taste perception and acuity. The research proposed here will develop the honeybee as a model system for understanding the logic of the gustatory code. The honeybee has only 10 Gr genes - the least reported from insects with sequenced genomes. Based on sequence homology with Drosophila and what we know about the structure of the bee's Gr genes, we predict it has less than 20 functional Grs. For this reason, it would be possible to identify the chemical ligands for the Grs produced by these genes with the aim of using the bee as a model to understand the principles of gustatory coding. Having few Gr genes makes the bee a tractable model system in contrast to Drosophila with its 60 genes. Ligands have been determined for only 13 Drosophila Grs. This proposal describes a project that will use two approaches to identify the ligands for the receptors associated with the honeybee's Gr genes. Using a 'gain-of-function' approach, we will employ a newly-developed transgenic fruit fly line in which all of the putative genes for sugar receptors have been knocked out. Each of the bee's Gr genes will be expressed in this line. Flies from each bee Gr line will be assayed using calcium imaging of their tarsal gustatory neurons. By stimulating with a series of ligands, we will be able to identify whether functional receptors are produced by the expression of single Gr genes and to identify their Grs' ligands. Based on what we know about fruit fly sugar receptors and their bee homologues, we will also test whether expression of multiple Gr genes that encode sugar receptors is necessary to form functional Grs. We do not know if several Gr genes must be expressed to form functional receptors for the detection of compounds other than sugars. For this reason, we must also use a 'loss-of-function' approach in which we knock down expression of each Gr gene in vivo in the bee using small-interfering RNA (siRNA). Using this method, we will knock down expression of each Gr gene and assay the bee's taste neurons using electrophysiology and behaviour. We will test a suite of nutrients and toxic compounds that includes common pesticides encountered by bees in flowering crops. In spite of the fact that bees have only 10 Gr genes, they are still able to detect some toxins and to regulate their intake of nutrients like sugars and amino acids that are detected by Grs. The experiments proposed here will reveal how the bee's few Gr genes translates into the spectrum of what it can taste and will lead to future work that identifies how populations of Gr neurons encode information about the chemical nature and complexity of food.

A fundamental step for the survival and replication of intravacuolar bacterial pathogens is the establishment of a replicative niche inside host cells. This is achieved by secreting bacterial effector proteins in the cytoplasm of the infected cells by specialised secretion systems. Bacterial effector proteins interact with eukaryotic proteins and lipids to manipulate host signalling pathways, thus allowing to escape the host degradative pathway and converge nutrients required for intracellular replication of bacteria. Understanding the host/pathogen interactions that regulate these processes is therefore of prime importance to counter bacterial infections and identify candidate targets for the development of host-targeted antimicrobials, as an alternative to antibiotics. Against the background of host/pathogen interactions, phosphoinositides (PIs) are emerging as targets for a growing number of bacterial effector proteins. These lipids are key player in eukaryotic cell homeostasis, which define the identity of intracellular membranes and serve as regulators of eukaryotic signal transduction. Coxiella burnetii is a class 3 pathogen responsible for the zoonosis Q fever, a debilitating disease with severe health and economic impact. The high infectivity and resistance to environmental stress make Coxiella a potential threat for bioterrorism applications. Key to Coxiella virulence is the Dot/Icm-dependent secretion of bacterial effector proteins that coordinate the biogenesis of a large compartment, the Coxiella-Containing vacuole (CCV). Initially defined as expanded autolysosomes, our recent characterisation of the lipid composition of CCVs suggests that several membrane trafficking pathways of the infected cell are subverted for the biogenesis of these compartments. Importantly, we have observed that perturbing the capacity of Coxiella to manipulate PI metabolism for CCVs biogenesis has in vivo relevance. With this project, we aim at a global characterisation of the Coxiella/PIs interactions with the double aim of 1) characterising the molecular mechanisms regulating the biogenesis of these compartments and 2) target PI metabolism in infected cells to affect Coxiella virulence in vivo. To this aim, we will define the comprehensive lipid profile of CCVs by integrating state-of-the-art microscopy on Coxiella infected cells with lipidomics approaches on isolated CCVs. In parallel, we will identify PI-binding Coxiella effector proteins (PIEs) using PI- and lipid-coated agarose beads. PI/protein interactions will be further investigated using biomimetic membranes of specific lipid composition. The role of Coxiella PIEs in infection will be investigated using bioinformatics approaches coupled to multi-parametric phenotypic screens, taking advantage of our previously generated library of Coxiella mutants. Finally, the in vivo relevance of PIE mutants as well as that of inhibitors of PI metabolism will be established using three independent in vivo models for Coxiella infections. This novel integrative approach will help us drawing a comprehensive map of the PI/Coxiella interactome that orchestrates CCVs biogenesis and identify key interaction hubs for the development of new, tailored antimicrobials targeting the pathogen as well as the host. Of note, our approach developed using C. burnetii as a model pathogen, will serve as a strategic roadmap for the study of other intravacuolar bacterial pathogens.