Mass General Home


Project 1: Development of yeast functional genomic approaches to identify proteins directly delivered into host cells by type IIII secretion systems

The vast array of strategies that bacterial pathogens utilize to survival within hosts and ultimately cause disease is staggering. Many gram-negative pathogens encode specialized secretion systems that can deliver on the order of tens of effector proteins (effectors) into host cells. Each pathogen delivers its own unique set of effectors. The identification and characterization of these proteins is often challenging and is a major limitation in furthering our understanding of how these pathogens cause disease. For example, pathogens that no longer express individual effectors often do not exhibit phenotypes presumably due to functional redundancy. Similarly, most of the effectors lack homology with proteins of known function and do not exhibit a detectable phenotype when expressed in mammalian cells.

We have expressed close to 2,000 bacterial proteins in yeast. Remarkably, we have found that bacterial housekeeping proteins rarely inhibit yeast growth. In contrast, close to half of proteins that are injected into host cells during the course of an infection inhibit growth when expressed in yeast. Presumably these proteins target cellular processes conserved from yeast to mammals. In addition, many of the virulence proteins localize to distinct subcellular compartments and/or alter yeast morphology- phenotypes that have provided novel insights into conserved host cell processes targeted by these virulence proteins.

Project 2: Development of yeast functional genomic approaches to identify conserved host cell processes targeted by bacterial virulence proteins.

There is ample evidence from our laboratory and others that yeast growth inhibition due to the expression of individual effector proteins is due to conserved targeting of cellular processes. In order to identify these processes we have developed the technology to systematically express individual effector proteins into the collection of >6,000 strains that overexpress one of the annotated yeast open reading frames as well as a collection of ~4,500 yeast haploid deletion strains that no longer express one of the annotated yeast non-essential genes. We then systematically screen for those yeast strains that are particularly sensitive (hypersensitivity or synthetic lethal screens) or resistant (suppressor screens) to expression of the effectors. In parallel with these functional genomic approaches we conduct genome-wide mRNA profiling experiments and biochemical screens to provide parallel insights into conserved cellular processes. We then use gene ontology enrichment studies and network analyses to identify candidate roles for the effector proteins in pathogenesis that we then test in physiologic models of disease.

This approach proved to be particularly fruitful in identifying a function for OspF in pathogenesis. Our functional genomic studies in yeast led us to determine that OspF inhibits yeast growth by targeting the cell wall integrity pathway, a highly conserved MAPK signaling pathway. Studies in yeast and mammalian cells led us to demonstrate that OspF inhibits MAPK signaling cascades by inhibiting phosphorylation of the terminal MAPKs in these highly conserved signaling cascades. Furthermore, we demonstrated that this down-regulation of MAPK signaling leads to a down-regulation of the host innate immune response during the course of a Shigella infection in a mouse model of infection. Notably, while biogenesis of the yeast cell wall and the innate immune system appear to have nothing in common, they are both regulated, at least in part, by MAPK signaling cascades. Thus, other effectors that target "complex" mammalian cellular processes like innate immunity are also amenable to initial analyses in yeast.

Project 3: Targeted approaches for identifying host cell processes targeted by bacterial virulence proteins in yeast and mammalian cells.

We have a collection of over 200 type III secreted effector proteins in the laboratory in the Gateway site-specific recombination system. We are currently comparing the behavior of these proteins when expressed in a variety of cell types including yeast and mammalian cells in the presence and absence of a variety of reporter plasmids that monitor cellular processes and signaling pathways involved in pathogenesis. Processes of interest include regulation of MAPK signaling cascades, apoptosis and autophagy and other components of the host innate immune response.

Project 4: Development of Protein Interaction Platforms a new yeast-based visualization assay for identifying interacting proteins.

We have developed a visualization assay for identifying interacting proteins. This assay is based on the ability of a viral protein, µNS, to form functional inclusions when expressed in yeast. Proteins fused to µNS are recruited to these inclusions, which we refer to as platforms. These platforms, which display the chosen protein, protein A, also serve as platforms to recruit proteins that interact with protein A. If proteins that interact with protein A are fused to a fluorescent protein like GFP, then fluorescent green inclusions or PIPs are visible by fluorescence microscopy.

To initially determine the utility of the PIP assay for detecting binary protein interactions, we compared the ability of PIP and the conventional yeast two-hybrid (Y2H) assay to detect interactions between Shigella flexneri chaperones and their type III secreted effector proteins. Remarkably, the PIP assay outperformed the Y2H and furthermore. One reason for this is that in our assay, we regulate expression of the GFP and muNS fusion proteins, so we are able to screen for interactions with toxic proteins. Furthermore, even with the non-toxic effectors, we detect additional interactions with PIP. Presumably this is because a positive readout in our assay just requires an interaction between the two proteins, while an interaction in the two-hybrid assay requires that the interacting proteins be oriented such that the reconstituted GAL4 transcription factor is functional.

Project 5: Investigations into the nature of chaperone-effector interactions.

Using the PIP assay we have now determined that Spa15, the Shigella flexneri "super chaperone" interacts with and is required for the secretion of at least 9 effectors into host cells. We are currently investigating the molecular nature of the chaperone-effector interactions as well as determining how Spa15 interacts with type III secretion machinery in order to facilitate the delivery of effector proteins into host cells. We are particularly interested in determining how the bacterial determine which of their thousands of proteins to deliver into host cells via the type iii secretion system.

Project 6: Development of yeast chemical genomic approaches to identify novel anti-virulence compounds.

Traditionally, screens for antimicrobial agents have focused on the identification of compounds that inhibit bacterial growth. A powerful alternative approach is to target processes required for virulence. Indeed, antivirulence molecules are potentially more powerful as there is less selective pressure for resistant. In collaboration with the laboratory of Dr. Stephen Lory at HMS, we recently exploited the yeast system to identify two lead compounds for the treatment of Chlamydia pneumonia. These molecules were identified in a screen for compounds that alleviate the yeast cell cycle arrest conferred by expression of CopN, a Chlamydia effector. Remarkably, both of these molecules inhibit replication of C. pneumonia within host cells. We plan to expand this approach to screen for inhibitors of effectors that inhibit yeast growth that play a major role in virulence.

Last update: Wed April 22 10:30 EST 2009