Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Structural characterization of bacterial type III secretion system components Yip, Calvin K. 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2006-200230.pdf [ 19.62MB ]
Metadata
JSON: 831-1.0092926.json
JSON-LD: 831-1.0092926-ld.json
RDF/XML (Pretty): 831-1.0092926-rdf.xml
RDF/JSON: 831-1.0092926-rdf.json
Turtle: 831-1.0092926-turtle.txt
N-Triples: 831-1.0092926-rdf-ntriples.txt
Original Record: 831-1.0092926-source.json
Full Text
831-1.0092926-fulltext.txt
Citation
831-1.0092926.ris

Full Text

S T R U C T U R A L C H A R A C T E R I Z A T I O N O F B A C T E R I A L T Y P E III S E C R E T I O N S Y S T E M C O M P O N E N T S by C A L V I N K . Y I P B . S c , The University of British Columbia, 2001 A THESIS S U B M I T T E D I N P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F D O C T O R O F P H I L O S O P H Y i n T H E F A C U L T Y OF G R A D U A T E S T U D I E S (Biochemistry and Molecular Biology) T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A October 2006 ©CalvinK. Yip, 2006 11 ABSTRACT The virulence-associated type III secretion system (T3SS) mediates the direct translocation of bacterial proteins known as effectors into the cytoplasm of eukaryotic cells, a process essential to the pathogenesis of many Gram negative pathogens. In this thesis, the molecular architecture of the T3SS was investigated through the biochemical and structural characterization of four representative components: EspA, EscJ, and EscC from enteropathogenic Escherichia coli (EPEC) , and PrgH from Salmonella typhimurium. EspA is a component o f the E P E C T3SS that assembles into extracellular filaments believed to be the molecular conduit for protein translocation. Results from biochemical analysis showed that EspA alone is sufficient to form filamentous structures and that an intact C-terminal coiled coil segment is required for oligomerization. CesA, the EspA-specific chaperone, was found to trap EspA in a monomelic state. Crystallographic analysis of the heterodimeric CesA-EspA complex at 2.8A revealed that EspA contains two long a-helices, which are engaged in extensive coiled coil interactions with CesA. E P E C EscJ is a member of the highly conserved YscJ family of proteins believed to form the inner membrane ring substructure of the T3SS. Sucrose gradient experiments on E P E C membranes showed that EscJ localizes to the inner membrane. The crystal structure of EscJ, refined to 1.8A, revealed repetitive and extensive intersubunit packing in the crystal lattice, which allowed the construction of a 24-subunit ring model. Data e from stoichiometric and surface mapping analyses of Salmonella typhimurium PrgK validated this model, which possesses surface features indicative o f a role as an assembly platform. Salmonella typhimurium PrgH and its distant orthologues represent the second major component of the inner membrane ring complex. Detergent extraction experiments I l l confirmed that PrgH is a membrane protein. A putative transmembrane segment appears to target this protein to the membrane as the predicted cytoplasmic and periplasmic regions of PrgH could exist as soluble, independent folding domains. The crystal structure of PrgH( 170-362), the core periplasmic domain, refined to 2.3A resolution, showed that it possess two EscJ-like domains, which may be involved in intersubunit interactions. The outer membrane channel elaborated by the secretin (YscC) family of proteins represents the second major ring complex in the T3SS. Limited proteolysis and protein expression studies of EscC, the E P E C T3SS secretin, confirmed the modular nature predicted for this family. Surface electrostatic analysis of the crystal structure of the N-terminal domain of EscC, EscC(22-174), determined to 2.2A, revealed two unique charge patches. Mutagenesis and complementation experiments demonstrated that one of these patches is required for proper function o f EscC. iv TABLE OF CONTENTS ABSTRACT ii TABLE OF CONTENTS iv LIST OF TABLES viii LIST OF FIGURES ix ABBREVIATIONS..... , xi ACKNOWLEDGEMENTS xiv CO-AUTHORSHIP STATEMENT xv CHAPTER 1 - INTRODUCTION 1 1.1 M E M B R A N E FUNCTION A N D TRANSPORT 1 1.2 PROTEIN TRANSPORT ACROSS M E M B R A N E S 2 1.2.1 The general secretory / Sec-dependent pathway 2 1.3 PROTEIN TRANSPORT IN G R A M NEGATIVE BACTERIA 4 1.3.1 Type I secretion system (T1SS) 6 1.3.2 Type II secretion system (T2SS) 7 1.3.3 Type III secretion system 8 1.3.4 Type TV secretion system (T4SS) 9 1.3.5 Type V secretion system (T5SS) 11 1.3.6 Type V I secretion system (T6SS) 12 1.4 T H E BACTERIAL F L A G E L L A 13 1.4.1 Overall structure 13 1.4.3 Assembly mechanism 16 1.5 T H E VIRULENCE-ASSOCIATED TYPE III SECRETION S Y S T E M ( T 3 S S ) 18 1.5.1 Prevalence and distribution 19 1.5.2 Genetic organization 20 1.5.3 Translocated effectors 22 1.5.4 Secretion signal and type III secretion chaperone (T3SC) 24 1.6 T H E TYPE III SECRETION APPARATUS 27 1.6.1 The needle complex (NC) 27 V 1.6.2 Assembly 29 1.6.3 Extracellular structures - needle, needle extension, and translocation pore 30 1.6.4 Outer membrane structure - secretin and pilotin 32 1.6.5 Inner membrane structure 33 1.6.6 Export apparatus 35 1.7 OBJECTIVES OF THESIS 35 CHAPTER 2 - STRUCTURAL AND BIOCHEMICAL CHARACTERIZATION OF ESPA FROM ENTEROPATHOGENIC ESCHERICHIA COLI 38 2.1 I N T R O D U C T I O N 38 2.2 M E T H O D S .40 2.2.1 Cloning, protein expression, and purification 40 2.2.2 Multiangle light scattering 42 2.2.3 In vitro protein cross-linking 42 2.2.4 Analytical gel filtration 43 2.2.5 Transmission electron microscopy (TEM) 43 2.2.6 Circular dichroism (CD) spectroscopy 43 2.2.7 Crystallization and structure determination 44 2.3 R E S U L T S 46 2.3.1 Purification and characterization of recombinant EspA 46 2.3.2 Mutagenesis of EspA 47 2.3.3 Biochemical characterization of CesA 48 2.3.4 CesA traps EspA in a monomelic state 51 2.3.5 Structure determination and overall architecture of the CesA-EspA complex 53 2.4 DISCUSSION 60 2.4.1 Role of coiled coils in EspA filament assembly 60 2.4.2 Secretion of EspA to bacterial surface 62 2.4.3 CesA is unique amongst T3SCs 63 2.4.4 Regulation of EspA polymerization in bacterial cytoplasm 65 CHAPTER 3 - STRUCTURAL INVESTIGATION OF ESC J FROM ENTEROPATHOGENIC ESCHERICHIA COLI 67 3.1 I N T R O D U C T I O N 67 3.2 M E T H O D S 70 VI 3.2.1 Membrane localization of EscJ VO 3.2.2 Cloning, protein expression, and purification 70 3.2.3 Surface entropy reduction mutagenesis... 70 3.2.4 Crystallization 71 3.2.5 Complementation and secretion assays 71 3.2.6 Data collection and structure determination 72 3.2.7 Molecular Modeling 72 3.3 R E S U L T S 73 3.3.1 Localization of EscJ in EPEC 73 3.3.2 Purification of EscJ ...75 3.3.2 Surface entropy reduction mutagenesis of EscJ 75 3.3.3 Crystal structure of EscJ 79 3.3.4 Modeling the EscJ ring 82 3.4 D I S C U S S I O N ...... 88 3.4.1 Localization of YscJ proteins 88 3.4.2 Oligomerization of YscJ proteins 89 3.4.3 Function of YscJ proteins 92 CHAPTER 4 - STRUCTURAL AND BIOCHEMICAL CHARACTERIZATION OF PRGH FROM SALMONELLA TYPHIMURIUM 94 4.1 I N T R O D U C T I O N 94 4.2 M E T H O D S 97 4.2.1 Cloning, protein expression, and purification 97 4.2.2 Detergent extraction 98 4.2.3 Nickel pull-down assays 98 4.2.4 Crystallization of PrgH periplasmic domain 99 4.2.5 Data collection and structure determination 100 4.2.5 Structural alignment and surface electrostatics analysis 101 4.3 R E S U L T S 102 4.3.1 Membrane localization of full length PrgH 102 4.3.2 Purification and biochemical characterization of putative domains of PrgH 103 4.3.3 Crystallization and structural determination of PrgH periplasmic domain 105 4.3.4 Architecture of PrgH(170-362) and surface electrostatic analysis 108 4.4 D I S C U S S I O N 112 Vll 4.4.1 Topology of PrgH and the proposed role : 112 4.4.2 The periplasmic domain and PrgH oligomerization 113 CHAPTER 5 - STRUCTURAL AND BIOCHEMICAL CHARACTERIZATION OF ESCC FROM ENTEROPATHOGENIC ESCHERICHIA COLI 116 5.1 I N T R O D U C T I O N 116 5.2 M E T H O D S 119 5.2.1 Cloning, protein expression, and purification 119 5.2.2 Limited proteolysis 119 5.2.3 Planar lipid bilayer experiment 120 5.2.4 Analytical gel filtration 120 5.2.5 Mutagenesis, complementation, and secretion assays 120 5.2.6 Crystallization of EscC(22-174) 121 5.2.7 Data collection and structure determination 121 5.2.8 Structural alignment and surface electrostatic analysis 122 5.3 R E S U L T S 123 5.3.1 EscC devoid of its signal peptide is soluble and monomelic 123 5.3.2 EscC is a modular protein 124 5.3.3 Crystal structure of EscC(22-174) 127 5.3.4 Surface electrostatic and preliminary mutation analyses 131 5.4 D I S C U S S I O N . . . . 133 5.4.1 Biogenesis of the EscC oligomer : 133 5.4.2 Function of the N-terminal domain of T3SS secretin 134 CHAPTER 6 - CONCLUSIONS AND FUTURE DIRECTIONS 137 6.1 S U M M A R Y A N D S I G N I F I C A N C E OF R E S U L T S 137 6.2 F U T U R E D I R E C T I O N S 140 REFERENCES 145 APPENDIX - PUBLICATIONS ARISING FROM GRADUATE WORK 171 Vlll LIST OF TABLES Table 1.1 Components of the bacterial flagellar and the virulence-associated type I I I secretion systems 28 Table 2.1 Data collection and structure refinement statistics for the CesA-EspA complex 54 Table 3.1 Data collection and structure refinement statistics for EPEC EscJ(21-190) (E62A/K63A/E64A) 77 Table 4.1 Data collection statistics for PrgH( 170-392) 106 Table 4.2 Data collection and structure refinement statistics for PrgH( 170-362) 107 Table 5.1 Data collection and structure refinement statistics for EscC(22-174) 128 Table 5.2 Effects of EscC mutation on type III secretion 132 IX LIST OF FIGURES Figure 1.1 The five major protein secretion pathways in Gram negative bacteria 5 Figure 1.2 Schematic representation of the bacterial flagelium and the virulence-associated type III secretion system (T3SS) 14 Figure 1.3 Assembly mechanism of the bacterial flagelium 16 Figure 1.4 Genetic organization of the Locus of Enterocyte Effacement (LEE) pathogenicity island of enteropathogenic Escherichia coli (EPEC) 21 Figure 1.5 High resolution structures of type III secretion chaperones (T3SCs) 26 Figure 1.6 Assembly mechanism of the T3SS 29 Figure 2.1 Full length recombinant EspA forms filamentous oligomers 46 Figure 2.2 EspA(l-141) exists as limited oligomers 48 Figure 2.3 CesA is dimeric and helical 50 Figure 2.4 CesA traps EspA monomer : 52 Figure 2.5 Overall architecture of the CesA-EspA complex ; 56 Figure 2.6 Interaction interface between CesA and EspA 58 Figure 2.7 Arrangement of secondary structures at the terminal regions of flagellin and EspA. ..62 Figure 2.8 Regulation of EspA by CesA in the bacterial cytoplasm ...66 Figure 3.1 Sequence alignment of EPEC EscJ with members of the YscJ family and flagellar FliF. : 73 Figure 3.2 EPEC EscJ is anchored to the inner membrane in vivo 74 Figure 3.3 EPEC EscJ(21-190) exists as monomer in solution 75 Figure 3.4 The surface mutation (E62A/K63A/E64A) does not affect EscJ function 78 Figure 3.5 EscJ structure and intermolecular interactions 80 Figure 3.6 EscJ packs into superhelical structures in the crystal lattice 83 Figure 3.7 Modeling and surface electrostatics analysis of the EscJ ring 85 Figure 3.8 Surface mapping of Salmonella typhimurium NC by limited biotinylation and M A LDI-TOF mass spectrometry 86 Figure 3.9 EscJ ring model in context of the needle complex 92 Figure 4.1 Domain organization of Salmonella typhimurium PrgH 95 Figure 4.2 Full length recombinant PrgH is a membrane protein 102 Figure 4.3 PrgH contains two discrete domains 104 Figure 4.4 Improvement of electron density map 108 Figure 4.5 Architecture of PrgH( 170-362) 110 X Figure 4.6 Surface electrostatic analysis of PrgH(l 70-362) 111 Figure 5.1 EPEC EscC(22-512) exists as a monomer in solution 123 Figure 5.2 EscC is a modular protein 126 Figure 5.3 Structure determination and architecture of EscC(22-174) 129 Figure 5.4 Surface electrostatic analysis of EscC(22-174) 132 XI ABBREVIATIONS 3D A / E A B C A L S A T P C D C H A P S c fyoEM C U D D M D M D M E M D M S O D N A DSP E G S E H E C E M E P E C E R G A P G E F GSP H E G A - 8 H E P E S IPTG L B three-dimensional attaching and effacing ATP-binding cassette Advanced Light Source adenosine triphosphate circular dichroism [3-[(3-Cholamidopropyl)-dimethylammonio]-l-propane sulfonate / N,N-Dimethyl-3-sulfo-N-[3-[(3a,5P,7a,12a)-3,7,12-trihydroxy-24-oxocholan-24-yl]amino]propyl]-l-propanaminium hydroxide, inner salt] electron cryomicroscopy chaperone/usher pathway n-dodecyl-P-D-maltoside n-decyl-P-D-maltoside Dulbecco's modified essential medium dimethyl sulfoxide deoxyribonucleic acid dithiobis(succinimidylpropionate) ethylene glycol bis(succinimidylsuccinate) enterohemorrhagic Escherichia coli electron microscopy enteropathogenic Escherichia coli endoplasmic recticulum GTPase-activating protein Guanine nucleotide exchange factor general secretory pathway octanoyl-n-hydroxyethylglucamide 4-2-hydroxyethyl-l-piperazineethane sulfonic acid isopropyl-P-D-thiogalactopyranoside Luria Bertoni media L D A O N,N-dimethyldodecylamine-N-oxide L E E Locus of Enterocyte Effacement L P S lipopolysaccharide M A L D I - T O F matrix-assisted laser desorption ionization time-of-flight M S ring membrane-supramembrane ring M T B main terminal branch N A P S Nucleic A c i d and Protein Service N C needle complex N M R nuclear magnetic resonance Oca oligomeric coiled-coil adhesins O G n-octyl-P-D-glucoside P A I pathogenicity island P C R polymerase chain reaction P D B Protein Databank P E G polyethylene glycol r.m.s.d. root mean squared deviation R N A ribonucleic acid S A D single anomalous diffraction SeMet selenomethionine S D S - P A G E sodium dodecyl sulfate polyacrylamide gel electrophoresis SPase signal peptidase SPI Salmonella pathogenicity island SRP signal recognition particle T1SS type I secretion system T2SS type II secretion system T3SC type III secretion chaperone T3SS virulence-associated type III secretion system T4SS type IV secretion system T5SS type V secretion system T C A trichloroacetic acid T E M transmission electron microscopy Xlll Tris 2-amino-2-(hydroxymethyl)-l ,3-propanediol XIV ACKNOWLEDGEMENTS I would like to first of all thank my parents and my brother for their support of which 1 have often taken for granted. I would never have been able to pursue this dream work without their loving care over the years. I want to thank my supervisor Dr. Natalie Strynadka for giving me the interesting and challenging project on bacterial type III secretion, providing a resourceful and stimulating environment to carry out this research, and allowing me to attend various workshops, courses, and conferences. None of this work would have been possible without her faith in my ability and her encouragement especially during those troubling days in the beginning stages of the project. I would like to extend my thanks to members of my supervisory committee Dr. Brett Finlay and Dr. Lawrence Mcintosh for their advice and support. A n d to Dr. N ikh i l Thomas and Dr. Wayin Deng in the Finlay lab, Dr. Tyler Kimbrough and Dr. Sam Mi l l e r and his lab for their excellent collaborations on many of the projects. I have benefited greatly from the quality members of the Strynadka lab over the years, more specifically L iza deCastro for maintaining a neat and ordered lab, Dr. Daniel L i m and Dr. Y u Luo for being great role models, Dr. Michela Bertero, and Dr. Andrew Lovering for teaching me protein crystallography, Richard Pfuetzner for discussions on protein purification, Dr. Elizabeth Frey and Cynthia Hou for helping out on cloning and purification. O f course, special thanks should go to Marija Vuckovic, the cheerful and optimistic "sister" who really pushed my graduate research on track with her tremendous help and was never scared of trying out my wi ld ideas, and to Angel Y u for giving me a great deal of support in the past two years. I am extremely grateful for the financial support from the National Science and Engineering Research Council , and the Michael Smith Foundation for Health Research throughout my graduate studies. However, I would have never received these.fellowships without the excellent and timely reference letters from Dr. Michael Murphy and Dr. Brett Finlay. I am also deeply honored to have received the Dr. Lionel McLeod Health Research Scholarship from the Alberta Heritage Foundation for Medical Research. A n d finally to M r . Hallette, a high school teacher who inspired me about biology and scientific research. XV CO-AUTHORSHIP STATEMENT Chapter 2 and 3 of this thesis describe work that were previously published. A l l manuscripts were written by myself and revised by my supervisor Dr. Natalie Strynadka. Chapter 2 contains portions from a manuscript published in Nature Structural and Molecular Biology [Yip, C . K . , Finlay, B . B . , and Strynadka, N . C . J . Structural characterization of a type III secretion system filament protein in complex with its chaperone. Nature Structural and Molecu la r Biology. 2005 12(1 ):75-81.] Dr. Brett Finlay provided initial unpublished data on the EspA chaperone CesA. I designed and performed all the biochemical experiments as well as determined and refined the structure of the CesA-EspA complex myself. A l l text and figures in this article were prepared by myself. Chapter 3 contains portions from a manuscript published in Nature [Yip, C . K . , Kimbrough, T .G. , Felise, H . B . , Vuckovic, M . , Thomas, N . A . , Pfuetzner, R . A . , Frey, E . A . , Finlay, B . B . , Mi l le r , S.I., and Strynadka, N . C . J . Structural characterization of the molecular platform for type III secretion system assembly. Nature. 2005 435(7042):702-707.] The original cloning of EscJ was done by Dr. Elizabeth Frey, a former postdoctoral fellow in the Strynadka lab. The purification procedure of tagless EscJ was developed by Richard Pfuetzner, a former technician in the Strynadka lab. I modified the original EscJ expression construct, more specifically deleting the N-terminal Cysteine. I also designed and made all five mutant EscJ constructs for testing the surface entropy reduction approach. Protein purification and crystallization of EscJ mutants were assisted by Marija Vuckovic , a technician in the Strynadka lab. Data collection, structure determination, and refinement were completed by myself. I constructed the ring model based on the crystal structure and performed surface electrostatics analysis as well . Sucrose gradient xvi experiment on E P E C membranes was performed by Dr. N i k h i l Thomas, a postdoctoral fellow in Dr. Brett Finlay's lab at the University of British Columbia. Stoichiometric analysis and surface mapping of PrgK by limited biotinylation and mass spectrometry were performed by Dr. Tyler Kimbrough, a former graduate student in Dr. Sam Mil le r ' s lab at the University of Washington. Imade the constructs for complementation experiment, and Dr. N ikh i l Thomas generated the respective E P E C strains and performed the secretion assays. A l l figures were prepared by myself with the exception of Figure 3.2, 3.4 and 3.8 which were contributed by Dr. N ikh i l Thomas and Dr. Tyler Kimbrough respectively. Chapter 4 and 5 of this thesis describe work that w i l l appear in two separate manuscripts in the future. Chapter 4 describes the characterization of Salmonella typhimurium PrgH. I made the expression construct for his-tagged full length PrgH, PrgH(170-392), PrgH( 170-362), PrgK(l9-200), as well as tagless PrgH(l 70-392) and P rgK( l 9-200). The expression construct for his-tagged PrgH(l-138) was obtained from Dr. Tyler Kimbrough in Dr. Sam Mi l le r ' s lab. Detergent extraction of full length PrgH was carried out by Marija Vuckovic under my supervision. I performed the binding experiments on PrgH and PrgK. Purification of the various PrgH domains was performed by myself, Marija Vuckovic and Angel Y u , a technician in the Strynadka lab. The original crystallization condition o f PrgH(l 70-392) was determined by me. Crystallization of PrgH(l 70-362) was carried out by Angel Y u and she refined the conditions for growing the orthorhombic crystals. I determined the condition for growing the trigonal crystal form. The orthorhombic SeMet xvn crystals were prepared by Angel Y u . Data collection, structure determination, refinement, and analysis were completed by myself. Chapter 5 describes the characterization of E P E C EscC. The original cloning o f EscC(22-512) was done by Dr. Elizabeth Frey. Purification of full length EscC was initially performed by Richard Pfuetzner, and Y u Luo, a former postdoctoral fellow in the Strynadka lab, and subsequently by Marija Vuckovic . I designed and made the expression constructs for EscC(22-174) and EscC(238-512). Purification of EscC(22-174) and EscC(238-512) were assisted by Marija Vuckovic . Initial crystallization condition for EscC(22-174) was determined by Marija Vuckovic , and I refined this condition as well as prepared SeMet crystals. Data collection, structure determination, refinement, and analysis were completed by myself. I made the constructs for complementation and designed several mutants based on surface electrostatics analysis of the crystal structure. Complementation experiments and secretion assays were performed by Dr. N ikh i l Thomas in Dr. Brett Finlay's lab. Planar l ipid bilayer experiment was assisted by Manjeet Bains in Dr. Robert Hancock's lab. , 1 CHAPTER 1 - Introduction 1.1 Membrane function and transport The cytoplasmic or cell membrane encompasses the entire surface o f a cell and separates the cytoplasm from the environment. This outer layer of the cell consists of a bilayer of phospholipids, with the hydrophobic fatty acid tails projecting inward and the hydrophilic phosphate-containing head groups facing the aqueous environment. The hydrophobic nature o f lipids and their bilayer arrangement result in a selectively permeable barrier in which only small, non-polar molecules can pass through, effectively preventing leakage of cell contents and entry of damaging substances to the interior. In addition to its vital role as a barrier, the cell membrane is a cell 's gateway to the outside world. To establish homeostasis, a cell constantly interacts with its surroundings to acquire nutrients, export waste products, and monitor environmental changes. These activities require the import and export o f molecules that are otherwise not diffusible across the l ipid bilayer. Many of the proteins and protein complexes embedded in the cell membrane are responsible for shuttling specific ions and molecules into and out of the cell. In spite of the diversity of molecules moved by these membrane proteins and protein complexes, the mechanisms of transport fall into two broad categories: facilitated diffusion and active transport. In facilitated diffusion, which does not require energy input, the transport protein creates water-filled pores or channels which provide a path for specific ions or hydrophilic molecules such as sugars to move down their concentration gradients and into the cytoplasm. In active transport, the transport protein utilizes energy either directly from hydrolysis of adenosine triphosphate 2 (ATP) or indirectly from the downhill flow of an ion to "pump" molecules against their concentration gradients. 1.2 Protein transport across membranes While channels and pumps effectively move various ions and small molecules into and out o f the cytoplasm, cells also need to transport macromolecules, more specifically proteins, across their membranes. Protein transport in general is a vital cellular process because the various proteins synthesized in a cell must localize to the correct compartments where they perform their respective functions. For eukaryotic cells, as protein synthesis occurs only in the cytoplasm and the mitochondria, proteins must be delivered to other membrane-bound organelles such as the endoplasmic recticulum (ER) and the nucleus. In bacterial cells, enzymes and toxins which exert their functions outside the cytoplasm have to be secreted and membrane proteins need to be integrated into the l ipid bilayer. Similar to ions and small molecules, biological membranes represent a major impediment to protein trafficking in a cell. Because of their large sizes, transport o f proteins across cellular membranes requires specialized machineries or pathways that function more than just generating an opening in the lipid bilayer. 1.2.1 The general secretory / Sec-dependent pathway Found in all kingdoms of life, the general secretory pathway (GSP) or the Sec-dependent system represents the major pathway for transporting proteins across biological membranes (Cao and Saier, 2003). Prokaryotic cells use this pathway for translocating proteins across their cytoplasmic membrane, whereas eukaryotic cells use it for shuttling proteins from the cytoplasm into the E R lumen. Precursor proteins destined for translocation by the Sec-system are distinguished from cytoplasmic proteins by an 3 amino-terrninal targeting sequence known as the signal peptide, which is proteolytically-cleaved by signal peptidase (SPase) during or shortly after protein translocation. The structural and physiochemical characteristics o f the signal peptide are conserved in all organisms: a positively charged N-terminal region, followed by a central hydrophobic core of 6 to 20 residues, and finally a polar C-terminal segment containing the recognition site for SPase (reviewed in Stephenson, 2005). The core molecular machinery of the Sec-system is the S e c Y E G or Sec61ayP protein complex, which is localized to the cytoplasmic or E R membrane (reviewed in Osborne et al., 2005). Also known as the Sec-translocase, this heterotrimeric membrane protein complex generates a translocation channel or translocon through which polypeptides are transferred across or integrated into the lipid bilayer. Recent X-ray crystallographic studies of the archeaon Methanococcus jannaschii SecY translocase revealed that the actual channel is probably located at the centre o f a single copy o f the SecY complex, and that the channel is plugged by a short a-helix during its closed or inactive state but may become open when it interacts with the signal peptide (Van den Berg et al, 2004). The SecYEG/Sec61ayp complex, however, is only a passive conduit and other components of the Sec-system must provide a driving force for translocation. Depending on the components involved, the protein "cargo" is targeted to the translocon either in a co-translational or post-translational manner. The co-translational mode of targeting, which is found in all organisms and is utilized for the integration of most membrane proteins, involves 3 additional protein complexes: the signal recognition particle (SRP), the SRP receptor, and the ribosome (reviewed in Halic and Beckmann, 2005; Luirink and Sinning, 2004). The SRP binds to 4 the signal peptide o f the growing nascent polypeptide chain protruding from the ribosome and through interaction with the SRP receptor directs the ribosome/nascent polypeptide complex to the membrane. The ribosome then binds the Sec-translocase, releasing the SRP and allowing the elongating polypeptide to move into the translocation channel and across the membrane. The mechanism of post-translational targeting is slightly different between prokaryotes and eukaryotes. In prokaryotes, the cytosolic chaperone SecB prevents folding of the precursor protein and presents it to the molecular motor SecA (de Keyzer et al, 2003). SecA undergoes conformational changes coupled to its ATPase cycle to push the polypeptide cargo through the Sec-translocase in a step-wise manner (Osborne et al, 2004). In eukaryotes, the precursor protein is maintained in an unfolded state by cytosolic chaperones, but is targeted to the Sec-translocase embedded in the E R membrane by the Sec62/63 complex (Plath and Rapoport, 2000). The Bip protein located in the E R lumen then acts as a "molecular ratchet" to pull the precursor across the E R membrane in an ATP-dependent manner (Matlack et al, 1999). 1.3 Protein transport in Gram negative bacteria In additional to their cytoplasmic membrane, Gram negative bacteria are surrounded by a cell wall consisting of peptidoglycan as well as an outer membrane. The outer membrane can be considered another lipid bilayer, but the composition of this membrane differs significantly from the inner membrane in that its outer leaflet contains predominantly lipopolysaccharide (LPS) whereas its inner leaflet is phospholipid-rich. Even though the peptidoglycan layer and the outer membrane provide mechanical support to maintain cell shape against internal turgor pressure and protection against toxic 5 agents in the environment, they pose enormous challenges to the protein secretion process since the GSP only mediates protein translocation across the cytoplasmic or inner bacterial membrane. Gram negative bacteria have developed specialized secretion pathways or systems which allow them to efficiently transport proteins from the cytoplasm to the extracellular environment (reviewed in Thanassi and Hultgren, 2000b). These pathways, designated I to V , differ not only in their secretion mechanisms but also in the types and number of proteins involved (Figure 1.1). It is important to note that the GSP is still involved in some aspects of these specialized secretion systems, more specifically in mediating the translocation of protein substrates (type II and type V) and structural components (type I, II, III, and IV) across the bacterial inner membrane. Figure 1.1 The five major protein secretion pathways in Gram negative bacteria. 6 The various compartments are labeled (left) as follows: extracellular space (E), outer bacterial membrane (OM), periplasm (P), inner bacterial membrane (IM), and bacterial cytoplasm (C). The type I, type II, type III, type IV, type V systems are represented by the E. coli HlyB-HlyD-TolC complex, the Klebsiella oxytoca pullulanse secretion machinery, the Yersinia spp. Ysc system, the Agrobacterium tumefaciens VirB system, and the Neisseria meningitidis NalP autotransporter respectively in this schematic. The Sec system or the general secretory pathway (GSP) is represented by the SecYEG protein-conducting channel embedded in the inner membrane and the motor SecA. The GSP is used for transporting the export substrate to the periplasm (type II and type V), and for delivering components of the secretion apparatus across the inner membrane (types I, II, III, and IV). For the type I, type III, and type IV systems, the export substrate is secreted across the inner and outer bacterial membranes in a Sec-independent manner without a periplasmic intermediate. All secretion systems, with the exception of type V, require energy in the form of ATP hydrolysis, which occurs proximal to or right at the inner membrane. This schematic is modified from (Thanassi & Hultgren, 2000b). 1.3.1 Type I secretion system (T1SS) The type I secretion system is used by many Gram negative bacterial species to secrete toxins, proteases, and lipases to their environment, with the secretion of a-hemolysin by E. coli being the most well studied model system. The type I secretion machinery consists of three proteins: a pore-forming outer membrane protein, an inner membrane-anchored periplasmic membrane fusion protein, and finally an inner membrane-embedded ATP-binding cassette ( A B C ) protein which provides energy for the secretion process (reviewed in Binet et al., 1997). Proteins targeted for export by this pathway do not possess an N-terminal signal sequence, but instead contain a C-terminal ~ 60 residue non-cleavable secretion signal (Binet et al., 1997). The secretion of substrates across the inner and outer bacterial membranes bypasses the Sec system and occurs in a single step without the presence of a periplasmic intermediate. 7 1.3.2 Type II secretion system (T2SS) The type II secretion system (T2SS) is also known as the main terminal branch ( M T B ) of the general secretory pathway (GSP). First discovered in Klebsiella oxytoca for the secretion of the starch-hydrolyzing enzyme pullulanse (d'Enfert et al, 1987), the type II pathway has since been shown to be responsible for the targeted delivery of toxins, proteases, cellulases, and lipases by various other Gram negative bacterial species (reviewed in Cianciotto, 2005). The protein secretion process of the type II pathway occurs in two discrete steps with two independent membrane translocation events. The first step involves the Sec system which translocates the substrate across the inner membrane to the periplasm followed by cleavage o f the N-terminal signal peptide by the SPase. The second step, involves the recognition and translocation of substrate across the outer membrane by core components of the T2SS. The secretion machinery of the T2SS consists of approximately 12 to 15 protein components, designated A to O respectively (reviewed in Fil loux, 2004). It has been proposed that these proteins assemble into a macromolecular complex spanning the two bacterial membranes and the periplasm, although such complex has yet to be purified and visualized experimentally (Johnson et al, 2006). The most well characterized components of the T2SS are protein D or GspD and protein S or GspS. GspD belongs to the secretin superfamily, which is also involved in type III secretion, filamentous phage assembly, and type IV pi l i biogenesis (Linderoth et ai, 1997). Biochemical and electron microscopic studies of PulD, the prototypical secretin from Klebsiella oxytoca have demonstrated that this protein family forms a large, homomultimeric annular complex of approximately 12 subunits in the outer membrane (Nouwen et al, 1999). The channel of 8 the PulD complex has an estimated diameter of 7.6nm, suggesting that substrates could potentially move across the outer membrane in a fully folded state. GspS, on the other hand, plays an important role in the biogenesis of the secretin complex. Also known as pilotin, GspS and other members of this family bind to the C-termini of their cognate secretins, protecting them from proteolysis and promoting their proper insertion into the outer membrane (Daefler et al, 1997; Hardie et al, 1996). Although the T2SS secretion apparatus functions mainly to transport proteins across the outer membrane, its core components apart from GspD and GspS are integral inner membrane proteins or are peripherally-associated with the inner membrane. The current belief is that the peripheral and integral membrane proteins provide energy for assembly of the secretion apparatus rather than directly mediating the outer membrane translocation process (Johnson et al, 2006). The T2SS is closely related to the molecular machinery used for assembling the type IV pi l i , surface structures which are important to the pathogenesis of many bacterial pathogens (Nunn, 1999). Interestingly, several components of the T2SS encode pilin-like subunits. Recent studies have shown that these "pseudopilins" could assemble into multimeric pilus structures (Kohler et al, 2004). It has been proposed that these "secretion p i l i " act as a piston to mediate the opening of the GspD pore as well as to "push" substrates across the outer membrane. 1.3.3 Type III secretion system First discovered in Yersinia spp., the type III secretion system initially referred to a specialized pathway used by several Gram negative bacterial pathogens to translocate proteins known as effectors into the cytoplasm of eukaryotic cells in a contact-dependent 9 manner (Rosqvist et al, 1994; Sory and Cornells, 1994). It was later realized that this is the same pathway utilized by these and many other species to secrete structural components to the bacterial surface for assembly of the type III secretion apparatus and the flagelium. Although the definition of the type III pathway has been revised, the abbreviation "T3SS" generally refers to the virulence-associated type III secretion system, whereas the homologous machinery in bacterial flagella is usually called the "flagellar export system" (Desvaux et al, 2006). Substrates of the type III pathway, effectors or structural proteins, do not possess canonical N-terminal signal peptides, and they are believed to move from the bacterial cytoplasm to the external milieu or the eukaryotic cytoplasm in a single step without a periplasmic intermediate and proteolytic processing (reviewed in Galan and Collmer, 1999). The secretion process requires over 20 unique protein components, thus making the type III pathway the most sophisticated secretion system in Gram negative bacteria. 1.3.4 Type I V secretion system (T4SS) The type IV secretion system, which is ancestrally related to the IncP and IncN bacterial conjugation systems, is used by several species for the intercellular transport of D N A , nucleoprotein complexes, and proteins (reviewed in Christie et al, 2005). This system has been shown to play central roles in the pathogenesis of Bordetella pertussis (Weiss et al, 1993), Legionella pneumophila (Coers et al, 2000; Nagai and Roy, 2001), Brucella spp. (Delrue et al, 2001), Bartonella henselase (Schmid et al, 2004), and Helicobacter pylori (Censini et al, 1996). The prototypical T4SS is the V i r B system from Agrobacterium tumefaciens, which catalyzes the transfer of oncogenic T - D N A as 10 well as several proteins (including V i r D 2 , V i r E 2 , and Vi rF) into plant cells (reviewed in Christie and Cascales, 2005). The type IV secretion machinery consists of 12 proteins, designated V i r B l - 1 1 and V i r D 4 , many of which are similar in sequence to conserved components of the mating pair formation complex of the bacterial conjugation machinery. These proteins form a multi-subunit complex that spans the inner and outer membranes and the bacterial surface. More specifically, V i r B l is a specific transglycosylase, likely to be involved in digesting regions of the peptidoglycan cell wall for insertion of the apparatus across the periplasmic space (Mushegian et al, 1996). V i r B 4 , V i r B l 1, and V i r D 4 , on the other hand, are NTPases that are cytoplasmic but closely associated with the inner membrane (Atmakuri et al, 2004). V i r B 8 and V i r B l O are periplasmic proteins that form a complex with the inner membrane protein V i r B 6 (Judd et al, 2005), whereas V i r B 9 and the lipoprotein V i r B 7 form a complex in the outer membrane (Spudich et al, 1996). Finally, V i r B 2 and V i r B 5 are the major and minor components of the extracellular pilus (Schmidt-Eisenlohr et al, 1999). Despite a wealth of biochemical and structural data on the secretion apparatus, the secretion mechanism of the type IV pathway is not fully understood. The translocation of protein-DNA complex into plant cells by A. tumefaciens and CagA into epithelial cells by Helicobacter pylori are believed to occur in a single continuous step from the bacterial cytoplasm to the cytoplasm of the target cell (Odenbreit et al, 2000). In contrast, the secretion of pertussis toxin by the Bordetella pertussis Ptl system occurs in a 2-step manner reminiscent of the T2SS and other terminal branches of the GSP . More specifically, individual subunits of the toxin are transported across the bacterial inner 11 membrane by the Sec system. Following signal peptide cleavage and assembly in the periplasm, the hetero-oligomeric toxin is targeted for transport across the outer membrane by the core type IV machinery (Farizo et al, 2000). 1.3.5 Type V secretion system (T5SS) The type V secretion system (T5SS) represents the simplest and most prevalent terminal branch of the GSP . It is responsible for transporting adhesins to the bacterial surface as well as a variety of enzymes to the extracellular environment (reviewed in Henderson et al, 2004). Unlike other Gram negative secretion systems described so far, the transportation of substrates across the outer membrane by this pathway occurs without harnessing energy from the inner membrane. The canonical T5SS is the autotransporter protein I g A l protease from Neisseria gonorrhoeae (Pohlner et al, 1987). A l l autotransporters encode, from the amino to the carboxyl terminus, four functional domains: a signal sequence, a passenger domain, a linker region, and a p-domain. The export substrate is first synthesized in the cytoplasm and then transported across the inner membrane by the Sec system. Upon signal peptide cleavage, the (3-domain inserts into the outer membrane as a biophysically-favored [3-barrel structure, generating a pore in the outer membrane (Oomen et al, 2004). The passenger domain is then translocated through this pore to the cell surface where it may or may not undergo further processing. The two-partner secretion pathway and the oligomeric coiled-coil adhesins (Oca) pathway are subfamilies of the type V pathway. The secretion mechanisms of these pathways largely resemble the autotransporters except that for the two partner pathway, the P-domain and the passenger domain are translated as separate proteins (Jacob-12 Dubuisson et al, 2001), and for the Oca system, the formation of the (3-barrel pore in the outer membrane requires trimerization of the protein (Cotter et al, 2005). The chaperone/usher (CU) pathway, which is a branch of the GSP used by several species to assemble virulence adhesive structures on their surfaces, represents a slight variation of the type V pathway. The canonical example of the C U pathway is the assembly of the P and type I pi l i by uropathogenic E. coli (reviewed in Thanassi and Hultgren, 2000a). Secretion of substrates across the outer membrane requires two components: a periplasmic chaperone, which stabilizes the transport substrate, and an outer membrane usher, which provides a translocation channel through the outer membrane. The pilus subunit is first transported across the inner membrane by the Sec system and, upon cleavage of the signal peptide by the SPase, binds the periplasmic chaperone which mediates the proper folding of this protein and prevents premature intersubunit interactions (Kuehn et al, 1991). The chaperone-subunit complex is then targeted to the outer membrane usher, which triggers chaperone dissociation, exposes subunit-assembly surfaces, and promotes incorporation of the subunit into the pilus fiber (Sauere^a/., 2002). 1.3.6 Type VI secretion system (T6SS) The V A S ("virulence associated secretion") genes of Vibrio cholera strain V 5 2 has recently been shown to encode a unique secretion apparatus that transports protein to the exterior of bacterial cells by a mechanism that does not require .an N-terminal signal sequence (Pukatzki et al, 2006). This secretion system has been designated the type V I secretion system (T6SS) but further work is needed to characterize the secretion machinery and to elucidate the secretion mechanism. 13 1.4 The bacterial flagella Flagella are long, thin appendages free at one end and attached to the bacterial cell at the other end. A s the principle organelle responsible for motility in prokaryotes, the flagelium converts electrochemical energy from proton motive force into rotational motion of an extracellular helical filament. Rotation of this filament, in turn, converts torque into thrust, allowing the bacterial cell to move or propel. 1.4.1 Overa l l structure The bacterial flagelium, which spans the cytoplasm, the inner membrane, the periplasm, the outer membrane, and the extracellular space, consists of over 20 unique protein components and represents one of the most sophisticated macromolecular assemblies observed experimentally. Although the fully-assembled flagelium exists as a single "nanomachine", it actually consists of several functionally-distinct elements: the basal body, the motor, the switch, the hook and junction, the filament, and the export apparatus (reviewed in Macnab, 2003) (Figure 1.2). The basal body is a passive structural element that receives torque from the motor and transmits it to the extracellular structures. It also houses the export apparatus that directly mediates the secretion process at the inner membrane. Structurally, the basal body represents a collection of multimeric protein complexes in the bacterial envelope, which includes the inner membrane-embedded membrane-supramembrane (MS) ring (Ueno et al, 1992), a rod that traverses the periplasm (Minamino et al, 2000b), the P ring in the peptidoglycan, and finally the L ring embedded in the L P S of the outer membrane (Jones et al., 1987). On the other hand, the flagellar motor, which consists of the stator and the rotor proteins and operates by a rotary mechanism, is responsible for 14 torque generation (Lloyd et al, 1996). The stator proteins MotA and MotB form an integral membrane structure around the basal body anchored to the peptidoglycan cell wall (Garza et al, 1995), whereas the rotor protein FliG oligomerizes into a circular structure non-covalently attached to the MS ring made of FliF (Francis et al, 1992). The direction of the rotation of the motor is controlled by the switch complex, a fairly large structure consisting of the rotor protein FliG as well as F l i M and FliN (Francis et al, 1994). F l iM and FliN form a cytoplasmic cup-like structure known as the C-ring. F l iM receives input from the cellular machinery involved in sensing environmental conditions (Welch eta!., 1993). filament hook L ring P ring rod MS ring C ring Flagellar hook-basal-body YopB & YopD LcrV/EspA/SseB (FliC?) YscJ - (FliF) W M M M M M M N H cytoplasm (host) host membrane outer membrane periplasm 1 inner membrane cytoplasm (bacterium) Virulence-associated type III secretion system Figure 1.2 Schematic representation of the bacterial flagellum and the virulence-associated type III secretion system (T3SS). The overall morphologies of the two supramolecular complexes are highly similar. Each consists of inner membrane (green and brown) and outer membrane (red) ring structures, a membrane-associated ATPase, and extracellular helical structures (gray). The bacterial flagellum, which consists of several distinct structural elements (left labels), serves two major functions: (1) a 15 secretion system for exporting structural components responsible for constructing the extracellular structures during the assembly process, and (2) a motor to propel bacterial motion after assembly of the entire complex. The T3SS also export structural components (needle, needle extension, and translocation pore subunits) to the extracellular compartment during assembly. However, upon assembly, the T3SS translocates bacterial proteins directly into the cytoplasm of eukaryotic cells (shown by the arrow), and this process requires the needle extension (in blue) and a translocation pore (in yellow) formed by two proteins on the host cell membrane. Also shown in the diagram are key T3SS protein components, referred by their Yersinia family names, with their counterparts in the flagellar system in parentheses. This schematic is adopted from (Yip & Strynadka, 2006). The hook and junction function to connect the filament to the motor. This structure consists of F lgE , which forms the hook (Samatey et ai, 2004), F lgD, which caps the hook (Ohnishi et al, 1994), and two hook-associated protein or hook-filament junction proteins F l g K and F l g L (Homma et al, 1990). The filament is an extracellular helical structure that rotates and functions like a propeller. It is made from multiple copies o f a single protein F l i C (Yonekura et al., 2003), and is capped by the pentameric F l i D at its distal end (Yonekura et al., 2000). Finally, the export apparatus represents the core machinery of the type III secretion system that directly mediates transport of structural components to the bacterial surface. It consists of six integral membrane proteins (FlhA, F lhB , F l i O , F l iP , F l i Q , F l iR) (Minamino and Macnab, 1999), believed to localize within a pore of the M S ring, as well as a number of cytoplasmic or peripherally-associated components including the general chaperone Fl iJ (Minamino et al., 2000a), the ATPase F l i l (Fan and Macnab, 1996) and its regulator F l i H (Minamino and MacNab, 2000). 16 1.4.3 Assembly mechanism Biogenesis of the flagelium has been shown to follow an ordered and tightly-regulated mechanism involving sequential and linear addition of substructures from the inner membrane all the way to the bacterial surface in a "bottom-to-top" fashion (Kubori et al, 1992) (Figure 1.3). Assembly of the M S ring (FliF) and the integral membrane components o f the export apparatus (FlhA, F lhB, F l i O , F l iP , F l i Q , F l iR) represent the very first event in this process (Jones and Macnab, 1990). The Sec system participates extensively at this stage to mediate transport and membrane integration of these proteins. Completion of the inner membrane complex is followed by formation of two peripheral ring structures, the F l i G ring and the C-ring made o f F l i M and F l i N , which are mounted non-covalently onto the M S ring. The stator proteins M o t A and M o t B , which work closely with F l i G , may also integrate with the inner membrane assembly during this time. FliF FlhA FlhB FliO FliP FliQ • FliG FliM FliN MotA MotB ( j < > MS RING FliR Vajfr-p" FISH Flil FliJ EXPORT FlgJ APPARATUS Q MOTOR/SWITCH FliE FlgJ FlgB FlgC FlgF FlgG FIgl FlgH FlgE FliK P RING L RING FULL-LENGTH HOOK PROXIMAL ROD DISTAL ROD FlgK FliD FlgL FliC "FULL-LENGTH" FILAMENT Figure 1.3 Assembly mechanism of the bacterial flagelium. 17 Construction of the flagelium begins at the inner membrane with the assembly of the MS ring, export apparatus, and other peripheral ring structures, although the exact order of these processes is not fully understood (in parentheses). The Sec system is used at this stage for insertion of integral membrane components and transportation of several components to the periplasmic space. The fully assembled inner membrane complex functions as a secretion system to deliver components to the periplasm for constructing the rod. Completion of the rod allows the formation of the P and L ring structures, whose subunits are transported to the periplasm by the Sec system. The hook subunits, hook-associated proteins, and flagellin are sequentially delivered via the type III pathway for assembling their respective extracellular structures: the hook, the hook-filament junction, and the filament. The various compartments are designated as follows: outer membrane (OM), periplasm (P), and inner membrane (IM). This schematic is modified from (Macnab, 2003). The fully-assembled inner membrane complex functions as a protein secretion system. It takes over the assembly process from the Sec system and begins exporting components distal to the inner membrane sequentially from the cytoplasm. Although these components are added on top of each other, a central channel of at least ~20A is maintained throughout the new substructures to allow for transportation o f more distal components in a partially unfolded or completely unfolded state. The first substrate delivered by the type III pathway is F l i E , which forms a polymeric junction or adaptor between the M S ring and the rod (Kubori et al, 1992; Minamino et al, 2000b). Export of the axial proteins, which include subunits of the rod (FlgB, F lgC, FlgF, and FlgG), hook (FlgE), and filament (Fl iC) , then follows in sequence. These components share similar sequence and structural properties (Saijo-Hamano et al, 2004), and the individual subunits are added in a prqximal-to-distal fashion, with the newest subunits incorporated at the distal location and often under the control of a capping structure. The rod is the first assembled axial structure, and completion of this structure is required for construction of the periplasmic P ring and the outer membrane L ring. FIgl 18 and FlgH, components of these two rings, are exported to the periplasm by the Sec system rather than the type III pathway (Schoenhals and Macnab, 1996). These proteins remain monomeric in periplasm and would only associate into circular oligomers upon completion of the rod. The hook is the next structure added. It has an associated cap made from FlgD (Ohnishi et al, 1994), and has a precise length of 55nm (Hirano et al, 1994). Regulation of hook length has been shown to be carried out by the export apparatus component FlhB and a secreted protein Fl iK, but the molecular mechanism of how these two proteins control the number of FlgE subunits incorporated into the hook remains controversial (reviewed in Minamino and Pugsley, 2005). After the hook reaches its mature length, the hook cap is displaced and a number of hook-associated or hook-filament-junction proteins are added. The final stage of flagellar morphogenesis is assembly of the filament. A single filament can contain up to 20,000 subunits and incorporation of new FliC subunits into the existing helical structure is mediated by the filament cap, a pentameric structure made from FliD (Yonekura et al, 2000). 1.5 The virulence-associated type III secretion system (T3SS) Similar to the bacterial flagellar system, the virulence-associated type III secretion system (T3SS) is a macromolecular complex that allows Gram negative bacteria to export protein components for assembling surface structures (Figure 1.2). However,.instead of participating in bacterial motility, the main biological function of the T3SS is to translocate effector proteins from the bacterial cytoplasm into eukaryotic cells (reviewed in Galan and Collmer,T999). The type III protein translocation process is found to be critical to the pathogenesis of a large number of Gram negative pathogens because many 19 of the translocated effectors resemble eukaryotic factors and are capable of altering basic eukaryotic cellular processes (reviewed in Stebbins and Galan, 2000). B y "reprogramming" its host cell, the infecting bacterium can directly modulate the host environment to its benefit. 1.5.1 Prevalence and distribution First discovered in the early 1990s, the T3SS have been identified and extensively studied in the Gram negative animal and human pathogens Yersinia spp., Salmonella typhimurium, Shigella flexneri, enteropathogenic Escherichia coli (EPEC) , enterohemorrhagic E. coli ( E H E C ) as well as the plant pathogen Pseudomonas syringae. Recent data from various microbial genome sequencing projects showed that the T3SS is more widely distributed than initially envisaged (Pallen et al, 2005b). The number of species known to encode T3SS continues to increase and now includes species of Chlamydia, Burkholderia, and Bordetella, Pseudomonas aeruginosa, Edwardsiella tarda, Aeromonmas salmonicida, Vibrio parahaemolyticus, Photorhabdus luminescens, and Chromobacterium violaceum (reviewed in Troisfontaines and Cornells, 2005). Interestingly, some species are found to harbor more than one T3SSs in their genomes (Pallen et al, 2003). While some of these newly discovered T3SSs have been shown to be functional and contribute to virulence, the exact roles for many others remain to be determined. Although T3SSs are only found in species that interact with eukaryotes, some T3SSs facilitate symbiotic rather than pathogenic interactions. For example, various species of Rhizobium have been shown to be important symbionts of plants by inducing formation of nodules on their roots (Marie et al, 2001). 20 1.5.2 Genetic organization The 20 to 25 genes encoding the secretion, translocation, and regulatory machinery of the T3SS are found clustered in the bacterial chromosome in what are often referred to as pathogenicity islands (PAIs) (reviewed in Collazo and Galan, 1997) (Figure 1.4). Interestingly, several o f these PAIs are encoded on plasmids, including the wel l -characterized Shigella Mxi -Spa and Yersinia Y s c T3SS. PAIs generally have different (G + C) % content from their resident genomes and their borders are often flanked by repeat sequences or insertion elements, suggesting that they have originated in other species but were subsequently acquired by the current bacterial host via horizontal gene transfer (Pallen et al, 2005b). This hypothesis is largely supported by phylogenetic analyses, which revealed that the evolutionary tree constructed based on T3SS genes in the PAIs is drastically different from that constructed based on 16S r R N A sequences encoded outside the PAIs (Troisfontaines and Cornells, 2005). Based on phylogenetic analysis of the protein components and gene organization within the PAIs, the T3SSs from different species have been categorized into 5 families or subgroups: Ysc , Inv-Mxi-Spa, Ssa-Esc, Hrc-Hrp 1, and Hrc-Hrp 2 (reviewed in Cornells, 2002a). The Y s c family, named after the plasmid-encoded prototypical Ysc T3SS from Yersinia spp. (Cornells, 2002b), includes the Psc system of Pseudomonas aeruginosa (Yahr et al, 1996), the Lsc system of Photorhabdus luminescens (Ffrench-Constant et al, 2000), the Asc system of Aeromonas spp. (Burr et al, 2002), and the Vsc system of Vibrio parahaemolyticus (Park et al, 2004). These T3SSs confer resistance to the host innate immune response by enabling these pathogens to avoid phagocytosis and initiate apoptosis in macrophages. The Inv-Mxi-Spa family is named after the Inv-Spa 21 system oX Salmonella enterica and the Mxi -Spa system of Shigella spp., both of which mediate the triggering o f bacterial uptake by nonphagocytic cells (Collazo and Galan, 1997; Menard et ai, 1994). Interestingly, the intracellular endosymbiont of the tsetse fly, Sodalis glossinidius, encodes two separate T3SSs of this type in its genome (Dale et al, 2001). The Ssa-Esc family is named after the T3SS encoded by the SPI-2 P A I of Salmonella enterica and the Esc T3SS of E P E C and E H E C (McDaniel et al, 1995; Shea et al, 1996). These T3SSs, which usually contain a unique extracellular filament, mediate survival within host cells and adherence to host cells. The T3SSs of all plant pathogens fall into either the Hrc-Hrp 1 or the Hrc-Hrp2 families. Both of these systems are distinguished by a long flexible extracellular pilus which is involved in traversing the thick plant cell wall (reviewed in He and Jin, 2003). Okb 10 20 30 40 I I I I I LEE1 LEE2 LEE3 LEE51 Tir LEE4 * »- 4 *• >• esc RSTU C J V N D F esp G H A D B F grIRgrIA sepD sepZ sepQ map tir eae sepL ces A D F T D2 orf/rorf r1 2 45 r3 r6 rS 12 1516 r10 29 ^ Type III secretion apparatus genes ^ Secreted / translocated proteins ^ Translocator proteins chaperones ^ Transcriptional regulators ^ Intimin (EPEC adhesin) Unknown function Figure 1.4 Genet ic organizat ion of the L o c u s of Enterocyte Effacement (LEE) pathogenicity is land of enteropathogenic Escherichia coli (EPEC) . Genes encoding components of the virulence-associated type III secretion system (T3SS) are usually clustered in the bacterial genome in pathogenicity islands. This schematic illustrates the 22 genetic organization of a pathogenicity island found in EPEC known as the LEE. The LEE contains 41 open reading frames organized into at least 5 operons, designated LEE1, LEE2, LEE3, LEE4, and LEE5/Tir respectively. In addition to the esc genes (red arrows), which encode structural components of the T3SS, several translocated effector proteins (green arrows) are found in this gene cluster. The needle extension and translocators are encoded by three esp genes (orange arrows). The presence of accessory factors including chaperones encoded by the ces genes (grey arrows) and transcriptional regulators (purple arrows) illustrate the functional independence of this gene cluster. Intimin (blue arrow), the EPEC adhesin which plays a central role in bacterial attachment of host cell by acting as the receptor for the type III translocated protein Tir, is also encoded within the LEE. This schematic is adopted from (McDaniel and Kaper, 1997). 1.5.3 Translocated effectors The T3SSs of animal and plant pathogens deliver a broad range of effectors that have the capability to stimulate or interfere with host cellular functions including signal transduction, cytoskeletal rearrangement, membrane trafficking, and cytokine gene expression (reviewed in Cornells and Van Gijsegem, 2000). Unlike the genes encoding the secretion apparatus, effectors are divergent in sequence and function across species. This observation is consistent with the different phenotypes associated with different systems, and likely reflects adaptation by each T3SS to carry out specific functions in its respective bacterial pathogen. Some effectors are encoded within the same PAIs as their cognate T3SSs (Figure 1.4), but many localize to other regions of the bacterial chromosome. Interestingly, non-PAI-encoded effector genes are often flanked by phage D N A sequences or are actually encoded within prophage genomes, suggesting that they were acquired by horizontal gene transfer events at a different time from the acquisition of the T3SS (Pallen et ai, 2005b). The biological function and the eukaryotic target of many effectors have been identified. The actin cytoskeleton and its associated regulatory machinery, such as the 23 monomelic GTPases from the Rho family, represent major targets of effectors from animal and human pathogens. Wel l characterized examples include SopE, SptP, SipC, and S ipA from Salmonella spp., Y o p E , Y o p H , and Y o p T from Yersinia spp., ExoS and ExoT from P. aeruginosa, and Tir from E P E C . SopE is a guanine nucleotide exchange factor (GEF) that promotes G D P to G T P nucleotide exchanges, and stimulates activities of the small GTPases Rac-1 and Cdc42 (Hardt et al, 1998). In turn, these small GTPases activate actin assembly through the heptameric Arp2/3 complex, the actin nucleation center o f mammalian cells. In contrast, SptP is a GTPase-activating protein (GAP) that inhibits Rho GTPases by promoting G T P hydrolysis, resulting in downregulation of actin assembly (Fu and Galan, 1999). SipC is membrane-associated protein but can nucleate actin assembly (Hayward and Koronakis, 1999), whereas S ipA binds directly to actin filaments and inhibits their depolymerization (Zhou et al, 1999). Y o p H is a powerful phosphotyrosine phosphatase that dephosphorylates proteins from focal adhesions, leading to disassembly and reorganization of the cytoskeleton (Persson et al, 1997). Similar to SptP, Y o p E acts as a G A P , switching RhoA, Rac, and Cdc42, to their off-state by accelerating G T P hydrolysis (Von Pawel-Rammingen et al, 2000). Y o p T is a cysteine protease that inactivates RhoA, Rac, and Cdc42 by cleaving these proteins close to their carboxyl terminus and releasing them from their membrane anchor (Shao and Dixon, 2003). ExoS and ExoT share significant sequence homology with each other, and both proteins contain an N-terminal R h o - G A P domain, which inactivate RhoA, Rac, and Cdc42 in a similar manner as Y o p E and SptP (reviewed in Barbieri and Sun, 2004). Finally, E P E C and E H E C translocates Tir, which not only functions as a receptor for their cell surface adhesin intimin, but also induces rearrangement of actin into pedestal 24 structures (Kenny et al, 1997). However, rather than modulating the activity of the Rho GTPases, Tir stimulates N-WASP, a key regulator of the actin cytoskeleton that promotes actin filament nucleation and branching by directly activating the Arp2/3 complex (Gruenheid et al, 2001). 1.5.4 Secretion signal and type III secretion chaperone (T3SC) The exact molecular determinants that distinguish substrates of the T3SS (effectors and several secretion apparatus components) from other bacterial proteins remain obscure. It is generally accepted that the type III secretion signal is encoded within the first 15 to 20 residues of the N-terminal region of the export substrate (reviewed in Ghosh, 2004). Extensive studies on the putative secretion signals of effectors from Yersinia, Salmonella, Pseudomonas, EPEC, Shigella, and Xanthomonas have demonstrated an enrichment of lie, Ser, Asn, and Thr and an absence of Cys and Trp, but a consensus sequence could not be deduced (Lloyd et al, 2002). Furthermore, due to the high mutability of the putative signal, there is some debate as to whether the molecular composition of this signal is mRNA or polypeptide in nature (Anderson and Schneewind, 1997). In addition to the N-terminal signal, effector translocation often depends on a specialized family of small, acidic proteins known as type III secretion chaperones (T3SC) (Wattiau et al, 1994). T3SCs, which associate with their cognate effectors through protein-protein interactions, have been demonstrated to play a variety of roles including stabilizing their cognate effectors, targeting them to the secretion apparatus, establishing a secretion hierarchy, and in a few cases, regulating the transcription of effector genes (reviewed in Feldman and Cornells, 2003; Parsot et al, 2003). Despite the lack of 25 sequence identity, T3SCs are remarkably similar in structure (Birtalan and Ghosh, 2001; Luo et al, 2001; Stebbins and Galan, 2001) (Figure 1.5), indicating a common mechanism of action may be in place. More specifically, dimerization of the T3SCs generates an extended surface for binding a relatively unstructured region near the N -terminus of their cognate effectors. It has been proposed that the T3SC in complex with the chaperone binding domain o f the effector, constitutes a conserved three-dimensional (3D) signal, which is recognized by the type III secretion and translocation machinery (Birtalan et al, 2002). 26 InvB-SipA Figure 1.5 High resolution structures of type III secretion chaperones (T3SCs). Ribbon representations of the structures of two T3SCs determined alone are shown on top: Salmonella typhimurium SigE (Protein databank accession code 1K3S), Yersinia spp. SycT (2BSH). Four structures of T3SCs determined in complex with fragments of their cognate effectors are shown in the bottom: Yersinia spp. SycE-YopE (1L2W) and SycN-YscB-YopN (1XKP), S. typhimurium SicP-SptP (1JYO) and InvB-SipA (2FM8). In spite of the low level of sequence identities, T3SCs demonstrate very similar 3-dimensional structures. All T3SCs exist as dimers, which generate an extended surface for binding the N-terminal region of their cognate effectors. 27 1.6 The type III secretion apparatus The cellular machinery that directly mediates the type III protein secretion and translocation process is a macromolecular complex that spans the inner bacterial membrane, the periplasmic space, the peptidoglycan layer, the outer bacterial membrane, the extracellular space, and the host cellular membrane. Also known as the injectisome, the type III secretion apparatus is believed to provide a continuous and direct path for effectors to move from the inside of the bacterium to the cytoplasm of the host (Ghosh, 2004) (Figure 1.2). Approximately 20 to 25 unique proteins are required for assembling this complex. Ten of these proteins are conserved across species, and eight of them show similarity in sequence to proteins o f the bacterial flagellar system (reviewed in Ghosh, 2004) (Figure 1.2 and Table 1.1). For historical reasons, the T3SS genes from different species have been given different names by separate research groups. A Yersinia " Y s c " -based naming scheme has been developed with the hope of standardizing the rather confusing nomenclature (Bogdanove et al, 1996) (Table 1.1). 1.6.1 The needle complex (NC) The purification of the core type III secretion apparatus, termed the needle complex (NC) because of its characteristic shape, was a landmark achievement (Kubori et al, 1998). First isolated from S. typhimurium, the N C s of several other species have since been purified and visualized by electron microscopy (EM) (Blocker et al, 2001; Sekiya et al, 2001; Tamano et al, 2002a). The N C s from the different species show very similar gross morphologies: a relatively rigid and extended helical structure (the needle) anchored to a base that spans the inner and outer bacterial membranes and the periplasmic space. The base itself consists of two sets of concentric ring complexes 28 embedded in each of the two bacterial membranes. Despite being smaller in size, the overall morphology of the N C resembles the flagellar hook-basal body complex. Table 1.1 Components of the bacterial flagellar and the virulence-associated type III secretion systems. The T3SS genes demonstrating sequence similarities with components of the flagellar system are colored purple whereas the T3SS genes conserved across species are colored red. Components absent in the flagellar system are designated (N/A). Flagella Yersinia Salmonella Shigella EPEC (SPI-1) Export apparatus FliP YscR SpaP Spa24 EscR FliQ YscS SpaQ Spa9 EscS FUR YscT SpaR Spa29 EscT FlhB YscU SpaS Spa40 EscU FlhA LcrD InvA MxiA EscV ATPase Flil YscN InvC Spa47 EscN A TPase regulator FliH YscL ? ? Orf5? Inner membrane base FliF YscJ PrgK MxiJ EscJ FliG? YscD PrgH MxiG EscD FliM, FliN YscQ SpaO Spa33 SepQ Inner rod N/A Yscl PrgJ Mxil rOrf8 Outer membrane secretin N/A YscC InvG MxiD EscC Pilotin N/A YscW InvH MxiM ? Needle FlgE(hook YscF Prgl MxiH EscF subunit) ? Needle length regulator FliK (hook length YscP InvJ Spa32 Orf16? regulator) Translocators FliC(flagellin)? LcrV ? ? EspA N/A YopB SipB IpaB EspD N/A YopD SipC IpaC EspB 3D reconstructions have been calculated from low-dose E M images of negatively-stained Shigella flexneri N C s (Blocker et al, 2001). The resulting E M map revealed a central channel of approximately 20 to 30 A that extends from the bottom set of rings all the way to the tip of the extracellular needle, suggesting that the secretion apparatus 29 generates a molecular passageway for protein transport. However, this channel is too narrow for most globular proteins to pass through, and effectors most l ikely need to be partially i f not completely unfolded prior to translocation. 1.6.2 Assembly The assembly process of the T 3 S S is believed to proceed in a "bottom-to-top" fashion as the bacterial flagellum (Figure 1.6). Genetic and biochemical analyses of the S. typhimurium SPI-1 T 3 S S revealed that the inner and outer membrane rings are the first oligomeric structures built (Kimbrough and Mi l ler , 2000; Kubori et al, 2000; Sukhan et al, 2001). Construction of these complexes requires the Sec system but not other T 3 S S components. Completion of these ring structures is followed by integration of the export apparatus within the inner membrane ring. In conjunction with the inner membrane-associated ATPase component, this inner membrane complex functions as a secretion system to export proteins for assembling the periplasmic inner rod as well as surface structures including the needle, the needle extension, and the translocation pore in the eukaryotic cell membrane. Just as in the flagellar export system, a central channel of at least 20A is maintained throughout the newer substructures to ensure the more distal components could be transported to their destinations. a • • • « CM YscR YscJ YscD? YscC YscW YscT ; YscU t~> YscF YscI c , n YscP LcrV OM YopB W p YopD crrix • C f—) D 'S/J IM T SCI I S C f LcrD YscN YscL Inner membrane ring Outer membrane ring Export apparatus Inner rod Needle Needle extension Translocation pore Figure 1.6 Assembly mechanism of the T3SS. 30 The initial event is the Sec-dependent assembly of two ring-like structures in each of the two bacterial membranes. Genetic and biochemical analyses suggested that the inner membrane ring is likely to be constructed first. The next stage of assembly involves integration of the various components of the export apparatus within the inner membrane ring. The inner membrane complex then functions as a secretion system to export components to compartments distal to the inner membrane. The inner rod component is delivered first but is not clearly understood if the completion of the rod is required for the inner and outer membrane ring structures to be connected. Completion of the basal structures at the bacterial membranes is followed by the secretion and construction of the extracellular needle, needle extension, and finally the translocation pore on the eukaryotic host membrane. The various compartments are labeled as in Figure 1.3 except that CM represents the eukaryotic cell membrane. 1.6.3 Extracellular structures - needle, needle extension, and translocation pore The extracellular portion of the T3SS can be divided into three main parts: the needle, the needle extension, and the translocation pore (Figure 1.2). The needle, which is a helical polymer made from a few hundred copies of a single protein of the Y s c F family (Cordes et al, 2003) and extends from the periplasm to the extracellular space, is the most distinctive structural feature o f the T3SS. Like the flagellar hook and filament, the needle contains a hollow interior of -25A in diameter and is believed to function as a molecular conduit for type III protein translocation (Blocker et al, 2001). E M analysis of T3SS needles isolated from Shigella (Mx iH) revealed that the helical parameters of these structures (5.6 units per turn, 4.2 A axial rise per subunit, 24 A helical pitch) are strikingly similar to those o f the flagellar hook and filament (Cordes et al, 2003). This finding is quite surprising considering that members of the Y s c F family is significantly smaller in size (usually ~ 9 kDa) and does not show appreciable sequence identity to components of the flagellar extracellular structures. The length of the needle for many species appears to be tightly controlled within a narrow range of ~ 60A. Genetic knockouts studies have identified Salmonella InvJ, Shigella Spa32, and Yersinia YscP as 31 key regulators of needle length (Journet et al, 2003; Russmann et al, 2002; Tamano et al, 2002b). Recent deletion and insertion experiments on Yersinia YscP revealed that the needle length is proportional to the number of residues in the middle region of YscP (Journet et al, 2003). The current hypothesis is that YscP acts as a "molecular ruler" in which two globular terminal domains attach to the base and the growing tip of the needle respectively and span an unstructured and extended ruler domain. When the needle reaches its desired length, the ruler domain wi l l become fully stretched and wi l l signal, through the base-associated domain, to stop further export o f the needle subunit (Agrain etal, 2005). The needles from T3SSs of many species are topped with sequence-divergent proteins termed needle extensions. The structures elaborated by these proteins may play a role in connecting to the host cell as well as mediating the formation of the translocation pore. L c r V , a protein secreted by the Yersinia T3SS, has been shown to form a bell-shaped complex at the tip of the needle, and this tip structure is needed for generating the translocation pore (Mueller et al, 2005). The EspA filament, on the other hand, is found in E P E C , E H E C , and other attaching and effacing (A/E) animal pathogens (Knutton et al, 1998). A n analogous structure has also been recently observed for the SPI-2 T3SS o f S. typhimurium (Chakravortty et al, 2005). The EspA filament is a helical polymer made from multiple copies of EspA, a protein which does not show sequence homology to the needle component. The EspA filament is firmly attached to the needle, but the needle part appears to be shorter than typically observed in the T3SS from other species (Sekiya et al, 2001). On the contrary, the lengths of the EspA filaments vary considerably and can extend to more than 600nm. The EspA filament might be a necessary adaptation to allow 32 the T3SS to penetrate the thick glycocalyx layer that covers the surfaces of the intestinal epithelium which these attaching and effacing (A/E) pathogens infect. The E M structure of the EspA filament at 26A revealed a central channel o f 25A in diameter analogous to those observed in the E M structures of the N C s and the Shigella needle, suggesting that this needle extension also serves as a molecular conduit for translocation (Daniell et al, 2003). Furthermore, as with the Shigella needle, the helical parameters of the EspA filament (5.6 subunits per turn, 4.3A axial rise per subunit) resemble those of the flagellar extracellular structures (Daniell et al, 2003). So, despite the sequence divergence in the individual components, the extracellular structures of the T3SSs and flagella are probably assembled in a similar fashion. Finally, the translocation pore consists of two proteins from the Y o p B and Y o p D families that hetero-oligomerize into a pore or channel in the eukaryotic cell membrane, presumably allowing effectors to breach the host cell membrane barrier in the translocation process (reviewed in Tardy et al., 1999). The pore structure has been visualized by atomic force microscopy from membranes of red blood cells infected with an atypical E P E C strain (Ide et al, 2001), but the exact stoichiometry of the two constituents is unknown, and it remains unclear i f or how this pore is connected to the needle or the needle extension. 1.6.4 Outer membrane structure - secretin and pilotin The outer membrane substructure of the T3SS is a single ring complex composed of one major protein component - the secretin. L ike PulD, the prototypical secretin from the T2SS, secretins of the T3SS (also known as the Y s c C family of proteins) associate into stable 12 to 14 subunit ring-like oligomers with a central channel of 5 to 10 nm in 33 diameter (Crago and Koronakis, 1998; Koster et al, 1997). This channel, which encompasses the needle, serves as a portal for the export substrate to cross the outer membrane barrier. Interestingly, the opening of this channel is tightly regulated prior to completion of the secretion apparatus. E M studies on the Salmonella N C revealed that the outer membrane channel is closed by a "sepfum"-like structure prior to assembly o f the needle (Marlovits et al, 2004). Furthermore, overexpression of Yersinia enterocolitica Y s c C in E. coli did not lead to increased permeability of the outer membrane (Burghout etal, 2004b). Biogenesis of the secretin ring begins from protein synthesis in the bacterial cytoplasm and export of the individual monomers to the periplasm by the sec-dependent pathway. Subsequent folding and insertion into the outer membrane are critical next steps to ensure proper oligomerization and channel formation. Pilotins are a sequence-divergent family of small lipoproteins which bind to the C-termini of their cognate secretins. Although not present in purified N C and not absolutely needed for protein secretion and translocation, pilotins of the T3SS, which include Salmonella InvH, Yersinia Y s c W , and Shigella M x i M , have been shown to promote outer membrane localization of their respective secretins InvG, Y s c C , and M x i D (Burghout et al, 2004a; Crago and Koronakis, 1998; Schuch and Maurell i , 2001). 1.6.5 Inner membrane structure The inner membrane ring structure of the T3SS encompasses the inner membrane and part of the periplasm. Similar to its counterpart in the flagellar system, this ring complex likely provides an early platform for localizing various structural components in 34 the assembly process as well as to house the export apparatus that directly mediates protein secretion and translocation. Genetic and biochemical analyses of Salmonella N C indicated that PrgK and PrgH are the major components of the inner membrane ring (Kimbrough and Miller, 2000). These proteins are capable of forming circular oligomers in the inner bacterial membrane in the absence of other T3SS proteins. PrgK belongs to the highly conserved YscJ family of proteins, which, despite smaller in size, is similar in sequence to a central region of the flagellar MS ring protein FliF (Suzuki et ai, 1998) (Table 1.1). These proteins are anchored to the bacterial membrane by a lipid covalently-linked to their N -termini after signal peptide cleavage and, for some members, a single transmembrane segment at the C-terminal region (Allaoui et al, 1992). In contrast, PrgH is not related to flagellar proteins, but has been shown by PSI-BLAST to be a distant homologue of the YscD family of proteins (Pallen et al., 2005a). Topologies predicted for PrgH resemble that predicted for Yersinia YscD: a small N-terminal cytoplasmic domain, followed by a single transmembrane segment and a slightly larger C-terminal periplasmic domain (Ghosh, 2004). Finally, the YscQ family of proteins exhibits sequence identities at its C-terminal region to the flagellar C-ring proteins Fl iN and F l i M (Fadouloglou et al., 2004) (Table 1.1). The "cytoplasmic bulb" observed by E M in osmotically-shocked Shigella has been hypothesized to be the C-ring in T3SS (Blocker et al, 1999). However, this structure is absent in purified NC, and it remains to be confirmed whether the YscQ proteins constitute a significant portion of the inner membrane structure. 35 1.6.6 Export apparatus The integral membrane components of the T3SS Y s c V , Y s c U , YscR , Y s c S , and Y s c T are not only highly conserved across T3SSs, but are also related in sequence to the flagellar proteins F l h A , F lhB, F l iP , F l iQ , and F l i R which form the core of the export apparatus (reviewed in Ghosh, 2004) (Table 1.1). Although the flagellar export apparatus proteins have been shown by biochemical means to associate with the flagellar basal body, the related T3SS proteins have not been found to co-purify with the N C s . Similar to flagellar F l i l , the Y s c N family o f proteins, which is highly conserved and demonstrates extensive primary sequence similarity to the P-subunit of FoFi-ATPases, is an essentially component of the T3SS (Woestyn et al, 1994) (Table 1.1). These proteins are peripherally-associated with the inner membrane, and hexamerization is required for hydrolyzing A T P (Pozidis et al, 2003). Exactly how energy harnessed from A T P hydrolysis is used for protein transport in the flagellar system and the T3SS is not clearly understood. It was reported that the Salmonella Y s c N protein, InvC displaces, in an ATP-dependent manner, the type III secretion chaperone (T3SC) from the T 3 S C -effector complex to facilitate effector loading onto the secretion machinery (Akeda and Galan, 2005). 1.7 Objectives of thesis The precise molecular mechanism of protein secretion and translocation by the T3SS is not fully understood despite a wealth of knowledge on the identity and function of the translocated effectors. Important insights could be gained from understanding the structure and function of the molecular machinery that directly mediates the protein export process. The objective of this project is to characterize the T3SS through v 36 examining the biochemical and structural properties of several representative components. At the start of this thesis investigation, there were very few detailed biochemical analyses on the individual components constituting the T3SS, and no high resolution structures of any of these proteins. Also , despite the availability of a 3D reconstruction o f the Shigella N C (Blocker et al, 2001), little is known about the organization o f the different proteins within the complex and the molecular forces that enable these proteins to interact with themselves and others to form larger assemblies. Chapter 2 describes the biochemical and structural characterization of EspA from E P E C , the needle extension protein that forms a filament on top of the needle of the T3SS. Mutagenesis studies on full length EspA protein as well as crystallographic and biophysical data on EspA in complex with its specific chaperone CesA are presented. Most of this work has been published in Nature Structural and Molecular Biology (Yip et al, 2005a). Chapter 3 describes the structural investigation of the inner membrane ring component EscJ from E P E C . Crystallographic analysis of this protein and data from modeling studies and localization experiments are presented. Most o f this work has been published in Nature (Yip et al, 2005b). Chapter 4 describes the structural characterization of the second major inner membrane ring component PrgH from Salmonella typhimurium. Biochemical analysis of full length PrgH and structural determination of a C-terminal periplasmic fragment of this protein are presented. This manuscript is in preparation. Chapter 5 describes the biochemical and structural characterization of the E P E C outer membrane secretin EscC. Biochemical analysis of the full length protein and structural determination of an N-terminal cytoplasmic fragment are presented. The manuscript is in preparation. 38 CHAPTER 2 - Structural and biochemical characterization of EspA from enteropathogenic Escherichia coli1 2.1 I N T R O D U C T I O N EPEC is a major cause of infantile diarrhea and child mortality worldwide (reviewed in Clarke et al, 2003). This pathogen encodes a T3SS in a 35-kb pathogenicity island known as the locus of enterocyte effacement or LEE, which is absolutely required for its tight adherence to the intestinal surface during infection (McDaniel and Kaper, 1997). In addition to escF, which encodes the structural subunit of the needle (Wilson et al, 2001), three other genes in the L E E encode extracellular components of the EPEC T3SS. Initial genetic studies showed that non-polar deletion mutation of either espA, espB, or espD eliminates EPEC's ability to translocate effectors into host cell cytoplasm, but maintains its potential to secrete proteins to the medium (Donnenberg et al, 1993; Kenny and Warawa, 2001; Lai et al, 1997). It was discovered later that EspA polymerizes into an extracellular filament that is associated with the needle of the EPEC T3SS, while EspB and EspD, the so-called translocator proteins (Knutton et al, 1998), hetero-oligomerize into pore-forming complexes on the surface of the host cell (Ide et al, 2001). The EspA filament has been proposed to be the molecular conduit in which effectors move along during type III protein secretion and translocation. This hypothesis was confirmed by a low resolution E M structure of sheared EspA filaments, which revealed that the filament contains a hollow centre of approximately 25 A analogous to the one observed for the Shigella T3SS needle (Daniell et al, 2003). Thus, understanding ' A version of this chapter has been published. Yip , C.K. , Finlay, B .B . , and Strynadka, N.C. (2005) "Structural characterization of a type I I I secretion system filament protein in complex with its chaperone". Nat Struct Mol Biol 12, 75-81. 39 the detailed molecular architecture o f this filamentous polymer wi l l provide direct insight into the molecular basis of the type III secretion and translocation processes. In this study, recombinant EspA in the presence and absence o f CesA, a recently identified EspA secretion chaperone (Creasey et al, 2003b), was characterized. Gel filtration and transmission electron microscopy ( T E M ) was used to evaluate EspA ' s ability to polymerize in the absence of other E P E C proteins. Substitution mutation of selected residues in the putative C-terminal coiled coil segment and systematic deletion mutation of this C-terminal region were carried out. The effects of CesA on the ability of EspA to form higher ordered oligomers were assessed by multiangle light scattering, in vitro protein cross-linking, and melting curve analysis. Finally, the structure of EspA in complex with CesA was determined by X-ray crystallographic methods and refined to 2.8A. Results from this study indicated that recombinant EspA can spontaneously assemble into filamentous multimers resembling the physiological structure and that the C-terminal putative coiled coil segment is required for this process. However, CesA could trap EspA in a monomeric state by binding via three sets of coiled coil interactions. This likely represents the molecular mechanism by which E P E C regulates the activity of this naturally polymeric protein in the cytoplasm prior to export to the bacterial surface. 40 2.2 M E T H O D S 2.2.1 Cloning, protein expression, and purification For purification of recombinant wi ld type EspA, the expression construct pEspA was generated by PCR-amplifying espA from E P E C E2348/69 genomic D N A and cloning the amplified D N A fragment into the Ndel/Xhol sites of the pET-41(a) expression vector (Novagen). E. coli BL21 (A.DE3) transformed with pEspA was grown to mid-exponential phase at 30 °C in Luria Bertani (LB) media containing 30 u.g mL" 1 o f kanamycin and induced with 0.1 m M isopropyl-P-D-thiogalactopyranoside (IPTG). Cells were harvested after a 3-hour incubation at 30 °C, resuspended in buffer (20mM Tris, pH 8.0), lysed with a pressurized homogenizer (Avestin), and centrifuged at 25,000 x g for 35 minutes. The supernatant was loaded onto Q-sepharose (Amersham) and the bound EspA was eluted with a linear gradient of increasing N a C l concentration. The eluted protein was dialyzed overnight, further purified on a MonoQ column (Amersham) and a Superdex 200 H R 10/30 gel filtration column (Amersham) in an A K T A F P L C system (Amersham). In vitro mutagenesis of espA was performed using the megaprimer-based P C R strategy on the plasmid pEspA. This strategy requires two rounds of P C R amplification with two flanking primers and two internal mutagenic primers. The resulting espA fragments containing the single-point mutations were digested with Ndel/Xhol and ligated to the corresponding sites of pET-28(a) (Novagen). Two rounds of this procedure were carried out for generating the double mutants. Expression and purification of the six mutant proteins (L149R, S156R, M163R, L149R/S156R, L149R/M163R, and S156R/M163R) follow essentially the same procedure as that for wi ld type EspA. 41 For purification of his-tagged CesA, the expression construct pCesA was generated by PCR-amplifying the cesA (previously known as or/3) open reading frame from genomic D N A and cloning it into the Nhel/HindlU. sites of the pET-28(a) expression vector. E. coli BL21 (A.DE3) transformed with pCesA was grown to mid-exponential phase at 37 °C in L B broth containing 30 u.g mL" 1 o f kanamycin and induced with I m M IPTG. Cells were harvested after a 3-hour incubation at 37 °C, resuspended in buffer (20 m M Tris + 500 m M N a C l , p H 7.5), lysed using a pressurized homogenizer, and centrifuged at 25,000 x g for 35 minutes. His-tagged CesA was purified from the soluble fraction using cobalt-chelating sepharose (Amersham). Protein eluted by buffer (20 m M Tris + 500 m M N a C l + 300 m M imidazole, p H 7.5) was further purified by gel filtration chromatography using a Superdex-75 H R 10/30 column. For preparation of the full length CesA-EspA complex, the espA open reading frame was amplified and inserted into the HindUl/BamRl sites (downstream of cesA) in pCesA to generate a bicistronic co-expression plasmid. Overexpression and purification procedures of this protein complex are essentially the same as CesA, except that column buffers contain 150 m M instead of 500 m M of N a C l and the his-tag was removed by addition of thrombin upon elution. Further purification was achieved by MonoQ 10/10 and Superdex 200 HR10/30 columns. For preparation of the tagless CesA( l -95) -EspA complex, a second bicistronic co-expression was constructed by first cloning cesA into the Ncol/HindUl sites of pET-28(a) (the N-terminal plasmid-encoded sequence G S H M A S was removed, and Ser2 was replaced by a glycine in this process) and then inserting espA into the HindUVBamHl sites. This protein complex was overexpressed using the same conditions as above. Cells 42 were harvested, resuspended in buffer (20 m M Tris + 50 m M N a C l , pH 7.5) and lysed. The soluble fraction was loaded onto SP-sepharose (Amersham) and the bound complex was eluted with a linear gradient of increasing N a C l concentration. Fractions containing the complex were pooled and dialyzed at 4 °C against buffer (20 m M Tris + 50 m M N a C l , p H 7.5) for two days during which the C-terminal 12 residue-fragment of CesA would be cleaved by an endogenous protease co-purified with the complex. The cleaved complex was further purified by a MonoS 5/5 column (Amersham) and a Superdex 200 H R 10/30 gel filtration column. 2.2.2 Multiangle light scattering Purified protein (1 to 2 mg mL" 1 ) was loaded onto either a Shodex protein K W -803 gel filtration column (Shoco) or a Superdex 75 H R 10/30 gel filtration column, equilibrated with buffer (20 m M H E P E S + 150 m M N a C l , pH 7.5), and connected in line with a m i n i D A W N multiangle light scattering equipment coupled to an interferometric refractometer (Wyatt Technologies). Data collection was carried out real time using the A S T R A software package (Wyatt Technologies) and molecular masses were calculated by the Debye fit method with the A S T R A software. 2.2.3 In vitro protein cross-linking Purified proteins were exchanged into reaction buffer (20 m M H E P E S + 150 m M N a C l , pH 7.5). Cross-linking agents dithiobis[succinimidylpropionate] (DSP) or ethylene glycol bis[succinimidylsuccinate] (EGS) (Pierce) were solubilized in dimethyl sulfoxide ( D M S O ) and added to the protein sample to yield a range o f final concentrations (0.1 to 2.0 m M ) . After 30 minutes of incubation at room temperature, the cross-linking reaction was quenched by addition of Tris pH 8.0 (final concentration of 100 m M ) and samples 43 were subsequently analyzed by sodium'dodecyl sulfate-polyacrylamide gel electrophoresis ( S D S - P A G E ) . 2.2.4 Analytical gel filtration Purified EspA (250 ug), CesA (800 ug), or EspA (250 pg) in complex with CesA (800 ug) were loaded onto a Superdex-200 HR10/30 column equilibrated with buffer (20 m M Tris + 150 m M N a C l , p H 7.5). Peak fractions were analyzed by S D S - P A G E . 2.2.5 Transmission electron microscopy (TEM) Purified EspA (~1 mg mL" 1 ) was applied onto carbon-collodion-coated copper grids and negatively stained with 1% uranyl acetate. Specimen was viewed using a Hitachi H7600 transmission electron microscope at the U B C Biolmaging Facility. 2.2.6 Circular dichroism (CD) spectroscopy C D spectra were recorded by a Jasco spectropolarimeter (Model J-810) equipped with a Pelletier device. A l l spectra are averages of four scans and quartz cells of 0.2-cm optical path-length and concentrations of 9.9 u M CesA or 3.6 u M CesA-EspA in buffer (0.1 m M Tris, p H 7.5) were used. For thermal denaturation experiments, melting curves were determined by monitoring the changes in dichroic density at 222 nm as a function of temperature in the range of 5 - 95 °C and a heating rate of 1°C per minute. The thermodynamic parameters associated with temperature-induced denaturation were obtained by non-linear, least-squares analysis o f the temperature dependence of C D and a two-state denaturation process was assumed during curve fitting analyses (Rosell et ai, 2005). 44 2.2.7 Crystallization and structure determination Purified full-length CesA-EspA complex (at 10 mg mL" 1 ) crystallizes into thin plates (thickness less than 0.01 mm) from a variety of conditions containing polyethylene glycol (PEG) 4000. The best crystals diffracted to 3.5 A only after very long exposures (30 seconds) at synchrotron sources, and numerous attempts to determine the structure from these crystals were unsuccessful. A new rod shaped crystal form was obtained from the purified CesA( l -95 ) -EspA complex (at 15 mg mL" 1 ) . Crystals were grown at room temperature using the hanging drop vapor diffusion method by mixing 0.5 u L of protein with the same volume o f reservoir solution (30 to 32% (w/v) P E G 1500 + 100 m M Tris pH 8.5). Small crystals were observed 1 to 2 weeks after initial setup and streak seeding was used to increase the size and quality of these crystals. Single crystals were cryoprotected by soaking in a solution containing 35% (w/v) P E G 1500, flash-frozen in liquid nitrogen, and dried in a pre-frozen but fortuitously leaking dewar for over two days. Extensive crystal screening and all subsequent data collection were performed at beamline 8.2.2. of the Advanced Light Source ( A L S ) . Data were reduced and scaled using M O S F L M (Leslie, 1992) and S C A L A (Evans, 1993). For phasing, crystals were grown from protein containing incorporated selenomethionine (SeMet) by streak seeding from native crystals. Positions of four selenium sites out of a possible nine were found and refined by S O L V E (Terwilliger and Berendzen, 1999). Phase calculation and solvent flattening were performed using R E S O L V E (Terwilliger, 2000), which already resulted in a very interpretable map. The model was built using Xfi t (McRee, 1999), and refinement was performed using C N S (Brunger et al, 1998). Other than the processed N-terminal methionine, residues Ser86 to 45 Thr95 of CesA are not observed. The following regions are not observed for EspA: M e t l to Asp30, Ser60 to A s n l 4 7 , G l y l 9 1 and L y s l 9 2 . 46 2.3 R E S U L T S 2.3.1 Purification and characterization of recombinant E s p A Studying the interactions that govern assembly of polymeric proteins such as EspA is technically challenging because these proteins are designed to readily associate into higher ordered structures. It was thus surprisingly to observe that in the absence of affinity tags, EspA could be overexpressed at high levels in the E. coli expression strain BL21()X)E3) in an apparently soluble form. Using a two-step anion exchange procedure, this recombinant protein was purified to approximately 85% purity (Figure 2.1a). (a) (c) (kDa) ifcST- "HP i (b) Volume (mL) Figure 2.1 Full length recombinant EspA forms filamentous oligomers. (a) Purification of EspA with Q-sepharose resin. In the absence of affinity tags, EspA expresses in an apparently soluble form and could be purified to almost 85% purity using anion exchange resins, (b) Purified EspA is polydisperse. Multiangle light scattering analysis of purified EspA is represented by a molar mass versus volume plot overlaid with a gel filtration elution profile. The calculated molecular mass across the elution peak ranges from approximately 2 to 6 MDa. (c) 47 Transmission electron microscopy (TEM) of negative-stained purified EspA. TEM showed that EspA exists as filamentous oligomers. Unfortunately, the purified sample consistently eluted in the void volume of the Superdex 200 gel filtration column, and subsequent analysis by multiangle light scattering showed that recombinant EspA is a heterogeneous mixture of polymers ranging in size from approximately 2 to 6 MDa (Figure 2.1b). Transmission electron microscopy (TEM) further revealed that these polymers are filamentous structures of varying lengths (Figure 2.1c), proving that EspA alone is sufficient to form filaments in the absence of other LEE pathogenicity island-encoded proteins. As judged by gel filtration on a Superdex 200 column, varying the buffer pH and addition of various. detergents and denaturants at different concentrations did not significantly reduce the heterogeneity of the purified EspA sample (data not shown). 2.3.2 Mutagenesis of E s p A EspA is predicted to contain a coiled coil segment near its C-terminus (Lupas et al, 1991). Delahay et al. showed that non-conservative single and double mutations to the putative "a" residues of the heptad repeats of this segment disrupted filament formation in vivo (Delahay et al, 1999). Expression vectors for six of these mutants (L149R, S156R, M163R, L149R7S156R, L149R/M163R, and S156R/M163R) were constructed, and the mutant proteins purified. Similar to the wild type protein, all six purified EspA mutants eluted in the void volume of the Superdex 200 column and did not show deficiencies in forming higher ordered oligomers (data not shown). Systematic truncations of EspA were also carried out to define the role of the C-terminal region in intermolecular interactions. Removal of 20 or more residues from the 48 C-terminus of EspA, although not completely abolishing EspA-EspA interactions, dramatically reduced the oligomeric state (data not shown). In particular, repeated multiangle light scattering experiments demonstrated that an EspA mutant with its entire C-terminal putative coiled coil deleted forms limited pentameric / hexameric assemblies (Figure 2.2). These results suggested that while an intact C-terminal region of EspA is essential to the formation of higher ordered oligomers, there exists a second region in EspA that mediates more limited intermolecular interactions, (a) (b) Figure 2.2 EspA(1-141) exists as limited oligomers. (a) Purification of EspA(1-141). EspA(1-141) is soluble when overexpressed and could be purified to near homogeneity using anion exchange resins such as Q-sepharose. (b) Multiangle light scattering analysis of purified EspA(1-141). A molar mass versus volume plot overlaid with a gel filtration elution profile is shown. The molecular mass calculated from light scattering analysis revealed that EspA(1-141) exists as putative pentamer/ hexamer in solution and suggested that an N-terminal region of EspA is also involved in mediating intersubunit interactions. 2.3.3 Biochemical characterization of CesA Previously known as Orf3, CesA was identified as a potential binding partner of EspA in a systematic yeast two-hybrid analysis of proteins encoded in the L E E (Creasey et al, 2003a). It was later shown that CesA (also called CesAB) is a specific chaperone for EspA, promoting its secretion to the EPEC surface via the type III pathway (Creasey 8 0 10.0 Volume (ml) -76 kDa 12.0 14.0 49 et al, 2003b). Intriguingly, unlike the majority o f T3SCs, which are encoded next to or in close proximity to their cognate substrates, the cesA gene is located approximately 25 kb upstream of espA in the first operon of the L E E pathogenicity island (Elliott et al, 1998). Sequence analysis indicated that CesA, despite being small (-12 kDa) and cytoplasmic, does not possess several features shared amongst previously characterized T3SCs. While the latter are generally found to be acidic, the sequence of CesA is enriched in positively-charged residues with an estimated p i of 9.5. The C O I L S server predicted that CesA contains two unique coiled coil regions (an N-terminal and central sequence), which are not present in other characterized T3SCs (Lupas et al, 1991). Finally, no significant sequence similarity could be deduced between CesA and other T3SCs, and the CesA sequence could not be threaded onto any o f the available T3SC structures. To gain a deeper understanding of CesA ' s function, recombinant CesA was purified and its biochemical properties were examined. Consistent with its predicted p i , purified CesA binds to various cation exchange resins at physiological p H (data not shown). Multiangle light scattering analysis showed that CesA is monodisperse with an estimated molecular weight of 29 kDa, suggesting that CesA is a molecular homodimer in solution (Figure 2.3a). Subsequent in vitro chemical cross-linking experiments confirmed that CesA forms a homodimer in solution at physiological p H (Figure 2.3b). In agreement with the secondary structure prediction, the far U V circular dichroism (CD) spectrum of CesA resembled that of all a-helical proteins such as cytochrome c (Figure 2.3c). 50 (a) (b) • 10' E o 5 I -29 kDa DSP EGS 0 0.1 0.2 0 0.1 0.2 (mM) (kDa -(CesAfc - CesA 3 0 100 120 Volume (ml) (C) 200 220 240 260 280 300 Wavelength (nm) Figure 2.3 CesA is dimeric and helical. (a) CesA exists as a dimer in solution. Multiangle light scattering analysis of purified CesA is presented as a molar mass versus volume plot overlaid with a gel filtration elution profile. The calculated molecular mass of 29 kDa showed that purified his-tagged CesA is a molecular dimer in solution, (b) In vitro chemical crosslinking of CesA. One major cross-linked species, with an apparent molecular weight of -29 kDa was observed in both sets of reactions in a Coomassie-stained SDS-PAGE gel, confirming the inherent dimerization of CesA. (c) Circular dichroism (CD) spectra of CesA. The far UV CD spectrum (190 - 300 nm) of CesA at 20 °C showed the negative bands at 222nm and 209nm characteristic of helical structures. Increasing the temperature to 95 °C resulted in reduction in overall CD. 51 2.3.4 CesA traps EspA in a monomelic state Biogenesis of the extensive EspA filament requires the export of thousands of EspA molecules to the bacterial surface. The polymeric nature of EspA together with the recent finding that surface EspA expression is controlled at the post-transcriptional level illustrates the need to prevent premature EspA-EspA interactions upon synthesis of this protein (Roe et ai, 2003). To determine whether CesA plays a role in regulating EspA polymerization in the cytoplasm, his-tagged CesA and tagless EspA were co-expressed in the same expression host (E. coli BL21) . Using cobalt-chelating sepharose, a stable CesA-EspA complex was purified, confirming that the CesA-EspA interaction does not require other LEE-encoded proteins. Subsequent multiangle light scattering analysis demonstrated that this protein complex is monodisperse with an estimated molecular weight of 34.2 kDa, which corresponds to a 1: 1 molar ratio of the two proteins (Figure 2.4a). This 1:1 stoichiometry was further substantiated by chemical cross-linking experiments, as only one major band (approximately 33 kDa) was obtained when the homo-bifunctional chemical cross-linking agents D S P or E G S were used (Figure 2.4a). These results define CesA as a potent inhibitor of EspA polymerization, suppressing EspA-EspA interactions by physically binding and trapping EspA in a monomeric state. C D spectroscopy showed that the CesA-EspA complex is also predominantly helical (Figure 2.4b), and subsequent thermal denaturation experiments showed that this heterodimeric complex has an estimated melting temperature of 68.4°C, significantly higher than the 48.1 °C estimated for the CesA dimer (Figure 2.4c). 52 Figure 2.4 CesA traps EspA monomer. (a) CesA forms a heterodimer with EspA. Multiangle light scattering analysis indicated that purified CesA-EspA complex is relatively monodisperse with a mass corresponding to a 1:1 molar 53 ratio. In vitro chemical crosslinking and SDS-PAGE analysis showed that one major cross-linked species with an apparent molecular weight of -33 kDa was obtained in both reactions, (b) CD spectra of the CesA-EspA complex. The far UV CD spectrum (190 - 300 nm) of the CesA-EspA at 20 °C resembled that of CesA. Increasing the temperature to 100 °C resulted in substantial reduction in overall CD including the negative bands at 222nm and 209nm. (c) Thermal denaturation of CesA and the CesA-EspA complex. Melting curves were determined by monitoring the changes in dichroic density at 222nm for both proteins. To simplify comparison, the extent of protein denaturation is presented as the fraction of unfolded protein. As shown, the heterodimeric CesA-EspA complex has a significantly higher estimated melting temperature compared to the homodimeric CesA. (d) Analytical gel filtration analysis. Incubation with CesA did not change the elution volume of EspA, whose size exceeded the exclusion limit of this column (-200 kDa), indicating that CesA cannot dissociate polymerized EspA. The ability of CesA to function as an EspA-depolymerization agent was examined by an analytical gel filtration assay. The elution profiles and accompanying S D S - P A G E analyses showed that CesA does not bind to polymerized EspA and does not apparently alter the oligomeric states of polymerized EspA, suggesting that CesA cannot dissociate EspA once it is polymerized (Figure 2.4d). 2.3.5 Structure determination and overall architecture of the CesA-EspA complex The identification of CesA as an inhibitor of EspA polymerization enabled the structural analysis of monomelic EspA by X-ray crystallography. The full-length CesA-EspA complex was initially crystallized, but the resulting crystals were poorly-diffracting. Crystals suitable for structure determination were obtained only after the C-terminal 12 residues and the N-terminal polyhistidine tag with plasmid-encoded sequence were removed from CesA. Multiangle light scattering analysis showed that the introduced C -terminal truncation does not apparently disrupt CesA's association with EspA (data not shown). The CesA( l -95) -EspA complex yielded crystals that diffracted to approximately 2.8 A at synchrotron sources. The structure was solved by the single anomalous 54 diffraction (SAD) method, using crystals of selenomethionine-substituted proteins (Table 2.1). Table 2.1 Data collection and structure refinement statistics for the CesA-EspA complex. X-ray crystallographic data Spacegroup Unit cell dimensions (A) Wavelength (A) Resolution (A) Unique reflections Average multiplicity Completeness (%)1 <l/al>1 Rmerge 1,2 P2 12 12 1 35.4 x 72.4 x 95.7 0.980 2.80 6480 13.1 100.0(100.0) 21.5(4.1) 0.123 (0.507) Crystal structure refinement Rwork/Rfree 0.236/0.265 R.m.s deviations bond lengths (A) 0.014 angles (°) 1.5 Average B factor 70.2 Ramachandran plot most favorable 95.9% allowed region 4.1% disallowed region 0.0% Values in parentheses refer to values in the highest resolution shell. 2Rmerge= l^Ohki) - < l > | / O^hki), where l h w is the integrated intensity of a given reflection. 3R-factor = (I |F o b s - Fc a k :|) / I|F o b s|. The R f r e e is calculated using a randomly selected 5% of reflections not used throughout refinement. 55 The crystallographic asymmetric unit contains one heterodimeric CesA-EspA complex, in complete agreement with the 1:1 binding ratio obtained from biophysical and biochemical analyses. Structurally, CesA adopts an extended all-helical topology with an overall hairpin shape of approximately 66 A in length (Figure 2.5a). A long and extensive N-terminal a-helix a l (Ile3-Lys46) is connected by a very tight 3-residue turn to two shorter a-helices, a2 (Gln50-Glu60) and a3 (Glu67-Glu83), which are separated by a loop and are positioned approximately 120° against each other. The surface-exposed face of each helix is decorated with primarily charged residues, whereas the inner facing residues are predominantly nonpolar, generating a hydrophobic groove for EspA binding and potential dimerization. In addition to the C-terminal 12 residues (Met96-Val l07) , which are absent from the protein used in crystallization, the region spanning Ser86 through Thr95 was also not observed in the electron density map, suggesting that the C-terminus of CesA may be quite flexible and possibly unstructured. Although the crystals contain full length EspA in complex with CesA (as confirmed by S D S - P A G E analysis of the crystals), the experimental electron density map only showed clear density for the N -terminal region (Asp31-Leu59) and the C-terminal region (Leu l48-Leu l90) of EspA. Each of these segments consists of a relatively long a-helix, and these two helices, designated a l (Phe36-Ser57) and a l l (Leul49-I le l88) , constitute the CesA-binding domain of EspA (Figure 2.5b). 56 Figure 2.5 Overall architecture of the CesA-EspA complex. (a) Ribbon representation of CesA based on the structure of the CesA-EspA complex. The structure of CesA is clearly distinct from other secretion chaperones. (b) Structure of CesA (blue) in complex with EspA (red). The flexible central region of EspA is marked with a dashed line in red. Two extensive a-helices of EspA interact intimately with CesA in the protein complex, (c) Ribbon representation of the FliS-FliC(464-518) complex (PDB code 10RY) and the SycE-YopE(17-85) complex (1L2W). The chaperones and their cognate substrates are colored blue and red respectively. FliS, SycE and other T3SCs bind their targets in an extended conformation. The overall structure of the CesA-EspA complex resembles a four-helix bundle completely distinct from previously solved structures of chaperone-substrate complexes 57 of the T3SS and flagellar export system (Figure 2.5c). The four helices involved (CesA a l and a3, EspA a l and a l l ) differ in length, and there is an extra helix in CesA (a2) which is not part of the intermolecular interface. The bundling of helices from CesA and EspA results in 4,811 A 2 o f solvent accessible area being buried. Although EspA has an overall negative charge complementary to that of the basic CesA, the interface between the two is largely nonpolar. The dominant feature of the interface is a long parallel coiled coil interaction spanning residues Ile3 to Val38 of CesA a l and residues V a i l 53 to Ilel 88 of EspA a l l (Figure 2.6a and Figure 2.6b). The backbone topology of these regions resembles the dimeric G C N 4 "leucine zipper", and the interacting sidechains show an interdigiting, "knobs-into-holes" pattern, characteristic of classical coil coils (O'Shea et al, 1991). This long coiled coil also contains a number of polar and charged residues, which provide binding specificity and additional stability by forming intra-chain hydrogen bonds and inter-chain hydrogen bonds, such as those between A r g l 74 and Gln l81 o f EspA a l l and Glu20 and Glu30 of CesA a l respectively (Figure 2.6c). Interestingly, both CesA a l and EspA a l l are not completely amphipathic, and both sequences are found interspersed with occasional nonpolar residues in the hydrophilic face (Figure 2.6b). These nonpolar residues appear to play important roles in mediating interactions in the adjacent faces of these two helices. More specifically, Leu l69 , A l a l 7 3 , A s p l 7 6 , and Leu l80 of EspA a l l , mediate a short anti-parallel coiled coil interaction with CesA a3 (Figure 2.6a). In the same vein, CesA a l , through residues A r g l 8 , Ile21, Lys25, and Ile28, engages in an anti-parallel coiled coil interaction with EspA a l (Figure X ) . Finally, although EspA a l does not form the two-faced coiled coil interactions as EspA a l l or CesA a l , it contributes three bulky hydrophobic side chains (Phe42, Phe49, 58 and Tyr53) to the interior core, suggesting that it has an important role in maintaining the stability o f the helix bundle (Figure 2.6a). (a) Figure 2.6 Interaction interface between CesA and EspA. 59 (a) Extensive coiled coil interactions between different helices in the CesA-EspA complex. Coiled coil interactions are observed in three out of four faces of the complex, (b) Helical wheel representation of CesA oc1 and EspA all interaction. As in other dimeric parallel coiled coil, a and a' as well as d and d' residues are side-by-side in the interaction interface. Residues marked with an asterisk(*) are those involved in coiled coil interactions in the adjacent face, (c) Interchain hydrogen bonding between Glu30 and Gln181 as well as Glu20 and Arg174 provide binding specificity for the core coiled coil between.CesAal and EspAall. (d) End-on view of the CesA-EspA complex. Bulky hydrophobic residues from EspA al mediate hydrophobic packing of the complex. 60 2.4 D I S C U S S I O N 2.4.1 Role of coiled coils in E s p A filament assembly Due to their inherent property to associate into polymeric structures, especially at high protein concentrations, filamentous proteins pose enormous challenges to conventional biochemical and structural analyses. The heterogeneity and polymeric properties of purified recombinant EspA demonstrated by initial multiangle light scattering and T E M analyses clearly showed that EspA falls into this category o f difficult proteins. While this technical issue could be circumvented by examining fragments rather than the entire filament-forming subunit (such as by truncation of terminal regions), significant amount of information regarding intersubunit interactions are often lost in the process. The identification of CesA as the inhibitor of EspA polymerization enabled one to examine a filament component of the T3SS in its monomeric state without denaturation or deletion. CesA inhibits EspA polymerization likely by mimicking the types of interactions EspA makes with other subunits during filament assembly. The crystal structure of the CesA-EspA complex revealed that the C-terminal region of EspA indeed contains a coiled coil motif as predicted from sequence analyses. The extensive and specific contacts CesA makes with this coiled coil indicates that it is very critical for EspA-EspA association. This observation agrees with results from systematic deletion mutagenesis and gel filtration analyses which showed that removal of 20 or more C-terminal residues completely abrogates EspA's ability to form high molecular weight oligomeric structures. The extensive nature of this coiled coil also explains why initial single and double substitution mutations to the heptad repeats failed to disrupt the ability o f recombinant 61 full length EspA to oligomerize. The presence of an N-terminal coiled coil in EspA, although unexpected and not predicted by initial sequence analysis, confirms the light scattering data which showed that an EspA truncation mutant devoid of its C-terminal coiled coil , EspA( l -141) , forms putative pentamers/hexamers. Future crystallization and structure determination of these limited oligomeric forms of EspA, together with docking of these structures into a higher quality E M map of the physiological filament should enable one to further understand the role of the N-terminal region in EspA-EspA interaction and filament formation. The arrangement of coiled-coil motifs in EspA is reminiscent of the structural organization of F l i C , which contains two long terminal helices (Samatey et al, 2001; Yonekura et al, 2003) (Figure 2.7). High resolution c r y o E M analysis of F l i C showed that these terminal helices form an intramolecular coiled coil and generate the DO domain lining the central channel of the flagellar rod (Yonekura et al, 2003). The DO domain packs against adjacent subunits via hydrophobic interactions, ultimately leading to filament formation. Based on similarity in organization of secondary structures and gross structural similarity between the EspA filament and the flagellar rod, it is conceivable to believe that the two terminal regions of EspA would also form an intramolecular coiled coil to generate an analogous "DO domain". On the other hand, the central region of EspA, which shows the least sequence similarity amongst three related pathogens ( E P E C , E H E C , and Citrobacter rodentium) and which is not observed in the electron density due to disorder, likely makes up the surface-exposed domain that confers to the structural polymorphism of EspA filaments from different species (Crepin et al, 2005b). 62 Flagellin DO domain EspA-CesA complex Figure 2.7 Arrangement of secondary structures at the terminal regions of flagellin and E s p A . Ribbon representation of flagellin determined by X-ray crystallography and electron cryomicroscopy (PDB code 1UCU) is shown. This 4.5A EM structure showed that the terminal regions of flagellin (red and blue helices) are engaged in an intramolecular coiled coil interaction. The DO domain generated by this coiled coil mediates intersubunit interaction and is actually lining the inner surface of the central channel of the flagellar filament. The crystal structure of CesA-EspA (1XOU) illustrates a similar arrangement of the terminal regions of EspA. CesA probably prevents EspA multimerization by blocking an intramolecular coiled coil of EspA. For clarity, the color of CesA in this ribbon representation is changed to light grey. 2.4.2 Secretion of E s p A to bacterial surface Although the T3SS was discovered more than a decade ago, little is known about the secretion mechanism of the extracellular structural components (needle extension and translocator proteins). The data presented here also sheds light on this particular process. 63 First o f all , the 1:1 molar ratio of the CesA-EspA complex indicates that EspA monomers, not limited oligomeric forms, are secreted to the bacterial surface. Secondly, monomelic EspA is probably secreted with its key secondary structural elements (two long helices important for EspA-EspA interaction) preserved, rather than as a completely unfolded random coil . Unlike the translocated effectors, which are targeted to the host cytoplasm and whose folding could be assisted by chaperonins in the host cell, EspA is targeted to the extracellular space and has to efficiently adopt its final conformation upon reaching the bacterial surface. B y "priming" it prior to secretion and maintaining the necessary helical conformation, EspA can readily interact with other subunits and quickly assemble into the extracellular filament. It is reasonable to believe that other extracellular structural components would also follow a similar secretion mechanism. 2.4.3 CesA is unique amongst T 3 S C s T3SCs are a sequence-divergent but structurally-similar family of small, acidic, and cytoplasmic proteins which mediate type III secretion process by stabilizing the secreted or translocated substrates, preventing premature protein-protein interactions, imposing a secretion hierarchy, as well as regulating transcription of effector genes (reviewed in Feldman and Cornelis, 2003; Parsot et al, 2003). The data presented here demonstrate that CesA facilitates secretion of the E P E C T3SS filament protein EspA by preventing premature EspA polymerization in the bacterial cytoplasm. In line with this unique anti-polymerization function, the overall architecture of CesA is completely distinct from the compact, mixed a/(3 sandwich arrangement observed in several other published T3SCs (Figure 2.5c). The highly extended nature of CesA provides an elongated yet highly specific platform for binding EspA. However, 64 adoption of this relatively non-globular structure would probably pose difficulty for CesA to remain monomelic in solution. CesA dimerization in solution is l ikely a means for this protein to maintain its stability in the absence of EspA, as opposed to the generation of , extended hydrophobic surfaces for substrate binding as have been observed in the crystal structures of the T3SCs Salmonella SicP and Yersinia SycE in complex with the minimal binding fragment of their respective substrates SptP and Y o p E (Birtalan et ai, 2002; Stebbins and Galan, 2001). In retrospect, the formation of a complex with EspA not only prevents EspA self-association but also completes a necessary protein-protein interaction for CesA. Although lacking sequence identity, CesA appears to be functionally analogous to chaperones of the flagellar export system such as F l iS , the chaperone which inhibits polymerization of flagellin (also known as F l i C , the structural subunit of the bacterial flagellar rod) (Auvray et al, 2001; Ozin et al, 2003). Interestingly, the recently solved structure of Aquifex aeolicus F l iS in complex with a small binding fragment of F l i C (residues 464-518) shows a 1:1 binding stoichiometry like that observed in CesA-EspA and the preservation of helical secondary structure in the bound F l i C fragment (Figure 2.5c) (Evdokimov et al, 2003). However, the structure and mode o f substrate binding appear to be significantly different between these functionally similar chaperones in that Fl iS adopts a compact four-helix bundle structure as opposed to the highly extended 3-helix hairpin of CesA, and instead of engaging in extended coiled coil interactions with its substrate, F l iS provides a meandering hydrophobic surface which allows the helical F l i C fragment to wrap around the chaperone (Evdokimov et al, 2003). Unfortunately, the size of the F l i C fragment used in this particular study was not sufficient to understand the 65 molecular role o f F l iS in preventing F l i C polymerization. Structures of intact F l i C in complex with Fl iS from A. aeolicus and more well characterized model systems such as Salmonella w i l l be important to fully understand the similarities or differences in the regulation of polymerization in the flagellar and T3SS filaments by these seemingly functionally-analogous chaperones. 2.4.4 Regulation of EspA polymerization in bacterial cytoplasm The observation that CesA binds EspA in a 1:1 molar ratio has raised questions concerning the regulatory mechanism since CesA appears to be inherently dimeric and cesA and espA are located in different operons and are not co-transcribed nor co-translated. Thermal denaturation experiments hinted that the homodimeric CesA probably converts into the thermodynamically more stable heterodimeric CesA-EspA via a "partner exchange" process in which monomelic EspA displaces one molecule from the CesA dimer (Figure 2.8). This mechanism would allow effective anti-polymerization to take place at the post-translational level as dimeric CesA can readily "scavenge" any newly synthesized monomelic EspA which is prone to premature self-association and interaction with the translocator protein EspB in the bacterial cytoplasm (Hartland et al, 2000). 66 CesA dimerization CesA binding CesA synthesis displacement 7 D EspA export EspA & EspB synthesis LEE1 LEE2 LEE3 LEE51 Tir LEE4 Figure 2.8 Regulation of EspA by CesA in the bacterial cytoplasm. CesA (blue arrow) and EspA (red arrow) are encoded on different operons and are transcribed and translated separately. CesA is located in LEE1 which encodes the positive regulator Ler (LEE-encoded regulator) and is generally believed to be the first transcript synthesized upon activation by signals in the environment. CesA (blue circle) dimerizes immediately after translation, and this dimeric protein accumulates in the bacterial cytoplasm. Ler subsequently activates transcription of several operons including LEE4, which encodes EspA. The circulating pool of CesA dimer scavenges newly-synthesized EspA (red rectangle) by direct protein-protein interaction. Formation of the CesA-EspA complex results in displacement of one CesA molecule from the dimer interface. The trapped EspA could not interact with other EspA molecules as well as the translocator EspB (yellow arrow and yellow hexagon). CesA also promotes EspA secretion to the bacterial surface, likely through targeting it to the T3SS. 67 CHAPTER 3 - Structural investigation of EscJ from enteropathogenic Escherichia coli2 3.1 I N T R O D U C T I O N Although the needle and the EspA filament play important roles in contacting host cell and mediating protein translocation across the host cell membrane, assembly of these extracellular structures relies on the basal elements of the T3SS. The base of the T3SS consists of two sets of ring complexes situated in each of the two bacterial membranes (Kubori et al, 1998). In particular, the inner membrane complex, which spans part of the periplasmic space and the inner membrane, together with the export apparatus, is believed to directly mediate secretion in a manner similar to its counterpart in the flagellar system. The inner membrane ring of the T3SS can be assembled independent o f other proteins and is generally considered the first oligomeric structure elaborated in the T3SS biogenesis process (Sukhan et al, 2001). Thus, in addition to its important role in secretion, this complex might serve as a molecular scaffold or foundation for constructing other parts of this secretion machinery. Assembly of the prototypical SPI-1 pathogenicity island-encoded T3SS of S. typhimurium has been dissected by genetic and biochemical approaches. These studies unambiguously identified two proteins, P rgK and PrgH, as the major structural components of the inner membrane ring complex (Kimbrough and Mil ler , 2000; Kubori et al, 2000; Sukhan et al, 2003). More specifically, P rgK belongs to the YscJ family of 2 A version of this chapter has been published. Yip, C.K., Kimbrough, T.G., Felise, H.B., Vuckovic, M . , Thomas, N .A. , Pfuetzner, R.A. , Frey, E.A. , Finlay, B .B . , Miller, S.I., Strynadka, N.C. (2005). "Structural characterization of the molecular platform for type III secretion system assembly". Nature 435, 702-707. 68 proteins which is highly conserved across species. Members of the Y s c J family possess a canonical N-terminal signal peptide, and are transported to the periplasm by the Sec system. Upon processing by signal peptidase, an N-terminal cysteine residue is believed to be covalently-linked to a lipid molecule and would anchor the protein to the membrane (Allaoui et al, 1992). Although the YscJ proteins share sequence identity with a domain of flagellar F l iF , a protein which oligomerizes into the 26-subunit M S ring in the inner membrane (Suzuki et al, 2004), similar oligomerization has not been experimentally observed for the YscJ proteins, and the precise stoichiometry o f these proteins in the fully assembled T3SS is not known (Blocker et al, 2001). Furthermore, there has been wide debate regarding whether the lipid anchor of this protein is localized to the inner leaflet of the outer membrane or the outer leaflet of the inner membrane. This chapter presents the biochemical and structural analysis of a representative member of YscJ family, EscJ from E P E C . Membranes from E P E C cells expressing T3SS were separated on a sucrose gradient to determine the localization of EscJ in vivo. A triple mutant of EscJ that produced X-ray diffracting crystals was identified from a pool of mutants generated using the surface entropy reduction strategy (Derewenda, 2004). Secretion assays were carried out to evaluate the effects of the surface mutation on the function of EscJ. Finally, a crystal structure of this mutant EscJ, devoid of its signal peptide and N-terminal cysteine, was determined by X-ray crystallography and refined to 1.8 A. The structural and biochemical data presented here definitively showed that EscJ is anchored to the inner bacterial membrane. Crystal packing and computer modeling suggested that EscJ could form a 24-subunit ring-like oligomer. Predictions made from the modeled EscJ ring agree with data obtained from stoichiometric analysis and mass 69 spectrometry-based surface accessibility mapping performed on S. typhimurium PrgK. The original cloning and purification of EscJ were performed by Dr. Elizabeth Frey and Richard Pfuetzner. Modification of the expression plasmid, site-directed mutagenesis, protein purification, crystallization, structural determination, and computer modeling were performed by the author with assistance from Marija Vuckovic . Localization and complementation experiments were performed by Dr. N i k h i l Thomas in Dr. Brett Finlay's group at the University of British Columbia. Data on S. typhimurium PrgK was contributed by Dr. Tyler Kimbrough in Dr. Sam Mil le r ' s group at the University of Washington. 70 3.2 METHODS 3.2.1 Membrane localization of EscJ Total membranes were isolated from stationary cultures of E P E C strain E2348/69 grown in Dulbecco's Modified Essential Medium ( D M E M ) for approximately 5 hours. Separation of the inner and outer membranes was performed using sucrose gradients from 0.5 to 2.0 M as described elsewhere (Thomas et al, 2004). Fractions were collected, precipitated by trichloroacetic acid (TCA) , and subsequently analyzed by Western blotting with polyclonal anti-EscJ antibodies. 3.2.2 Cloning, protein expression, and purification Expression construct pET-41EscJ(21-190) was generated by PCR-amplifying escJ from E P E C E2348/69 genomic D N A and cloning the amplified D N A fragment into the NdellBamWl sites of the pET-41(a) expression vector (Novagen). E. coli BL21(A.DE3) transformed with this construct was grown in L B media containing 50 ug mL" 1 kanamycin at 37 °C to an OD 6 0 o of 0.6 to 0.8, induced with 0.5 m M of I P T G overnight at 20 °C. Cells were harvested, resuspended in 2 0 m M H E P E S + 150mM N a C l , pH 6.8, and lysed using a pressurized homogenizer. The soluble fraction was loaded onto a DE-52 column and the flowthrough which contained the recombinant protein was immediately fractionated with 20 to 30% ammonium sulfate. Precipitated proteins were resuspended, dialyzed overnight, and purified further using a Mono-Q column and a Superdex-200 column (Amersham). 3.2.3 Surface entropy reduction mutagenesis Mutations targeting different regions of EscJ were designed based on the surface entropy reduction strategy as outlined (reviewed in Derewenda, 2004). 5 sets of mutants 71 (E30A/K31A/E32A, K49A/E50A, E62A/K63A/E64A, K 8 0 A / K 8 1 A / K 8 2 A , and K I 1 O A / E l 11 A) were generated using the Quikchange method (Stratagene) with p E T -41EscJ(21-190) as the template. 3.2.4 Crystallization A l l crystallization trials were performed using the hanging drop vapor diffusion method by mixing 1 uL of protein solution (12 mg mL"1) with lu l of reservoir solution. Wild type EscJ(21-190) crystallizes in several conditions containing PEG3350, but none of these crystals diffracted X-rays. O f the five surface mutants generated, only (E62A/K63A/E64A) yielded crystals suitable for structure determination. This mutant crystallizes in 0.2 M di-ammonium hydrogen phosphate, and the crystals grew to maximal size over two to three days at 18 ° C . 3.2.5 Complementation and secretion assays Full length, wild type escJ and escJ harboring the triple mutation (E62A/K63A/E64A) were amplified and cloned separately into the BamUl site of the p A C Y C - 1 8 4 vector. These plasmids were transformed into E P E C AescJ, which contains a non-polar and in-frame deletion of esq, to generate two complemented strains. Secretion assays were performed on wild type E P E C and the two complemented strains as previously described (Kenny and Finlay, 1995). Briefly, 40uL of overnight standing cultures were inoculated into 2mL of D M E M . After 6 hours of growth at 37 °C in a 5 % CO2 incubator, l m L of cultures was harvested by centrifugation. The supernatant was filtered, precipitated with T C A , and analyzed by S D S - P A G E , and gels were stained with Coomassie blue. 72 3.2.6 Data collection and structure determination EscJ crystals were harvested by sequentially soaking in mother liquor containing 17.5% and 35% glycerol, and then flash-frozen to 100 K . A l l data were collected at beamline 8.2.1 of the Advanced Light Source ( A L S ) . Datasets were collected from a native crystal as well as a derivatized crystal, which was soaked in 5 m M of para-chloromercuribenzoate ( P C M B ) in reservoir solution for 32 hours, near the mercury peak wavelength. Data were processed with M O S F L M (Leslie, 1992) and scaled with S C A L A (Evans, 1993). Heavy atom positions were found and refined using S O L V E (Terwilliger and Berendzeri, 1999), and a high quality experimental map was obtained after density modification with R E S O L V E (Terwilliger, 2000). The model was built manually using Xfi t (McRee, 1999). Refinement was performed with C N S (Brunger et al, 1998) and finally with Refmac5 (Murshudov, 1997). The refined model is complete except for (92-97 and 134-140) in all 4 molecules o f the asymmetric unit and the C-terminal 5 to 6 residues in 3 out of 4 molecules of the asymmetric unit which were disordered. 3.2.7 Molecular Modeling The rotation and translation parameters of one monomer to the other three molecules in the EscJ tetramer were determined using L S Q K A B (Kabsch, 1976). A "flattened" tetramer was generated by setting the z-translation to zero, and a 6-fold rotation was then applied. The resulting model was refined using C N S with manual inspection. The final model contains no unfavorable molecular contacts. Surface calculation were performed using G R A S P (Nicholls et al, 1991). 73 3.3 R E S U L T S 3.3.1 Localization of EscJ in E P E C Alignment of YscJ proteins from several species revealed that EscJ from EPEC, despite demonstrating extensive sequence identity and possessing the same N-terminal organization as its orthologues, is slightly shorter and lacks the C-terminal stop transfer signal present in other members of the YscJ family (Figure 3.1). 9—FPTV EscJ MKKIli KNLFLLAAICLTV PrgK MIRRYLYTFLLVMTLAGlKD MxiJ MIRYKGFILFLLLMLIGlEQRE Y S C J MKVKTSLSTLILILFLTGT PscJ MRRTVKGLSCMALLALVLALGGB—KVEl F l i F TI KEOJYTGLTEKEA1 MOAL|L S N DVNV S|EM D KSG !-:D|LKGLDQE:QA|EV:AV|QMHNIFAN|JDSGKL ERHNITARIVDGGKQ RQEGLSADjEPDKDG •QADKDG FSNLSDQDGGAIVAQLTQMNIPYRFANGSG-I S N L S Q R Q A l E I I S TJIYTGISQKEGIEMLAL) YTGTSQKEGT EMLALfRSEGVSA -QZZsH > -€ED QI o3 EscJ PrgK MxiJ YscJ PscJ 60 NMTLS|EKEDFVR GYSITIAEPDFT GISVQJEKGTFAS KIKLLJEESDVAQJ -VTLRIEQSQFIN! R I' A V N V ITILNNNGFj YWIKTYQL DLMRMYDLlNPERVDISQI IDILKRKGY ELLRLNGY KKKFADIEVI PRPRVEIAQMI HESFSTLQDKj IHRQFTTAD: F l i F —AIEiPADKVHELRLRLAQQGLlKGGAVGFELLDQEK-FGliQFSiQVNYQRALEGELA jSQBNAKINYLKE DIE fKARLYSAIF, RLE :KARLYSAIE RLE jLARLNYAKA EIS JQQKINFLKE RIE t LZM=^> - e r IDCSVSLNV NNNESQPSSAAVLVTSSPEVNLAPSVI LSARVHISYDIDAGENGRPPKPVHLSALAVYERGSPLAHQIS ISAKIHVSYDLEEK—NISSKPMHISVIAIYDSPKESELLVS •VARVHVVLPEEQNNKGKKGVAASASVFTKHAADIQFDTYIP VARVHVVLPEERDGLGRKSSPASASVFIKHAADVQLDAYVP1 KSARVHLAMPKPSLFVREQKSP-SASVTVTLEPGRALDEGQISAVVHLVSS EscJ PrgK MxiJ YscJ PscJ F l i F VDDLKLENI FADVDYDNJ FSDVKYEN IEGLAYDRj IEGLSYDRS JS6 18C VIKSSSGQDG VLSERSDAQLQAPGTPVK RNSFATSHIVLIILLSVMSAGFGVWYY ILTPKEEYV-YTNVQPVKE-VKSEFLTNEVIYLFLGMAVLVVILLVWAF liVPSVDVR-QSSHLPRNTSILSIQVSEESKGRLIGLLSLLILLLPVTN VLVPSAGVR-QVPLAPRFESVFSIQVAEHSRGRLLGLFGLLLALLLASN VAGLPPGNVTLV EscJ PrgK KNH-YARNKKGITADDKAKS SNE MxiJ KTGWFKRNKI YscJ LAQYFWLQRKK PscJ LAQFFWHRQRG Figure 3.1 Sequence alignment of EPEC EscJ with members of the YscJ family and flagellar FliF. 74 Other proteins aligned include PrgK from Salmonella typhimurium, MxiJ from Shigella flexneri, YscJ from Yersinia pestis, PscJ from Pseudomonas aeruginosa, and residues 50 to 200 of Salmonella typhimurium FliF. Identical residues are highlighted in red while similar residues in yellow. The location of the putative signal peptidase cleavage site is shown by the dark arrow, followed by a conserved cysteine residue believed to be the lipid modification site. The red arrows indicate residues mutated by the surface entropy mutagenesis experiment. The grey box shows the putative C-terminal transmembrane segment of PrgK. Secondary structural elements based on the crystal structure of EscJ are shown on top of the alignment, with a-helices represented as grey rods and (3-strands as arrows. So, unlike most YscJ proteins, the localization of E P E C EscJ to the bacterial membrane relies solely on its N-terminal lipid. E P E C membranes were isolated and fractionated on a sucrose gradient by Dr. Nikh i l Thomas, a postdoctoral fellow in Dr. Brett Finlay's laboratory. Subsequent Western blotting analysis revealed that EscJ partitions exclusively to the inner membrane but not the outer membrane fractions (Figure 3.2). This clearly suggested that the lipidated N-terminus o f EscJ is positioned in the outer leaflet of the inner membrane rather than the inner leaflet of the outer membrane. 1 2 3 4 5 6 7 8 9 10 11 12 inner outer membrane membrane [sucrose] (M) 0.5 2.0 Figure 3.2 EPEC EscJ is anchored to the inner membrane in vivo. Total membranes were extracted from EPEC and fractionated using a sucrose density gradient. As shown by Western blotting of the different fractions, EscJ localizes to the inner membrane while the bacterial surface adhesin, intimin, is found exclusively in the outer membrane. cx-lntimin a-EscJ 3.3.2 Purification of EscJ To gain further insight into its biochemical properties, EscJ was cloned and overexpressed. In the absence of its N-terminal signal sequence and cysteine, EscJ is soluble and expresses at a high level in E. coli BL21. The recombinant protein, with a predicted pi of 4.85, is relatively acidic and binds to various anion exchange resins at neutral pH (Figure 3.3a). Multiangle light scattering showed that the purified protein has an estimated molecular mass of 20 kDa, indicating that non-lipidated EscJ exists as a monomer in solution (Figure 3.3b). (a) (b) (kDa) 66 45 31 22 (sais«... " - 1 1 fc,, -mil- liiiilfi 12.0 140 160 Volume (ml) Figure 3.3 EPEC EscJ(21-190) exists as monomer in solution. (a) Purification of EPEC EscJ(21-190). Recombinant EscJ(21-190) is soluble when overexpressed and could be purified by anion exchange resins such as Q-sepharose. (b) Multiangle light scattering analysis of purified EscJ(21-190). The data is presented as a molar mass versus volume plot overlaid with a gel filtration elution profile. The calculated molecular mass of-21 kDa showed that EscJ(21-190) is monomeric. 3.3.2 Surface entropy reduction mutagenesis of EscJ Purified EscJ readily crystallizes in a variety of conditions, but extensive screening showed that these hexagonal-shaped crystals do not diffract X-rays to sufficient resolution for structural determination. Surface entropy reduction mutagenesis is a method that has been shown to extend the diffraction limits of crystals and in some cases 76 to even generate X-ray quality crystals from proteins that were recalcitrant to crystallization (reviewed in Derewenda, 2004). This method, which involves replacing large flexible side chains on protein surfaces with alanines, is aimed at generating conformationally-homogeneous surface patches that would promote or improve intermolecular contacts within the crystal lattice. Because of their prevalence on protein surfaces, lysines and glutamates represent the two residues most commonly targeted for this type of mutagenesis (Derewenda, 2004). Using this strategy, five surface mutants of EscJ ( E 3 0 A / K 3 1 A / E 3 2 A , K 4 9 A / E 5 0 A , E 6 2 A / K 6 3 A / E 6 4 A , K 8 0 A / K 8 1 A / K 8 2 A , and K I 10A/E111 A ) were designed and their expression plasmids constructed. The surface mutations did not appear to affect the overall stability of EscJ as all five mutants could be expressed at high levels and purified to similar yields as the wi ld type protein using the same procedures (Figure 3.4a). Although four of the five mutants could be crystallized, only the triple mutant (E62A/K63A/E64A) produced X-ray diffracting crystals. The structure o f EscJ (E62A/K63A/E64A) was determined by the single anomalous diffraction (SAD) method from a mercury-derivatized crystal and refined to 1.8A (Table 3.1). To evaluate the effects of the triple mutation on EscJ, Dr. N i k h i l Thomas performed secretion assays on wi ld type E P E C as well as the AescJ strain complemented with plasmid encoding either escJ or escJ (E62A/K63A/E64A) . When E P E C is grown in D M E M , type III secretion is upregulated and the level of secreted proteins in the culture supernatant provides a simple and reliable test for T3SS function (Kenny and Finlay, 1995). The two complemented strains showed similar levels of type III secretion as wild 77 type, strongly suggesting the function of EscJ is largely preserved and not compromised by the introduced mutation (Figure 3.4b). Table 3.1 Data collection and structure refinement statistics for EPEC EscJ(21-190) (E62A/K63A/E64A). X-ray crystallographic data Dataset Spacegroup Unit cell (A) Wavelength (A) Resolution (A) Total reflections Unique reflections Completeness Redundancy1 <l/ol>1 Rmerge (%) Crystal structure refinement Rwork/Rfree ' ' 0.184/0.207 Average B-factor Main chain 21.3 Side chains with water 25.7 R.m.s. deviation Bond lengths (A) 0.013 Angles (°) 1.33 Ramachandran plot most favorable 96.2% allowed region 3.8% disallowed region 0.0% Values in parentheses correspond to the highest resolution shell. 2Rmerge=^ l(lhki)-<l>|/^ (lhki)> where lhki is the integrated intensity of a given reflection. 3Rwork=(£|Fo-Fc|)/(IF0), where F 0 and F care observed and calculated structure factors 4 R f r e e was calculated from 10% of reflections excluded from refinement. PCMB Native P6 5 P 6 5 164.7x164.7x66.5 164.8 x 164.8x67.2 0.980 1.000 1.74 1.80 2070092 1349839 105472 96428 99.8 (99.8) 99.9 (99.9) 9.6 (4.6) 5.6 (5.6) 21.9(2.9) 14.1 (4.6) 0.062 (0.368) 0.082 (0.313) 78 Figure 3.4 The surface mutation (E62A/K63A/E64A) does not affect E s c J funct ion. (a) Multiangle light scattering analysis of purified EscJ(21-190)(E62A/K63A/E64A). A molar mass versus volume plot overlaid with a gel filtration elution profile is shown. The purified triple mutant is relatively monodisperse and exhibits the same molecular mass as the wild type protein, (b) Complementation and secretion assays of EscJ mutant. An EPEC strain deficient in EscJ was complemented in trans with plasmid-encoded esq or esq triple mutant to generate the strains Aescj/pesq and Aesc//pesc/'(E62A/K63/E64A). To evaluate the function of EscJ, type III secreted proteins were obtained from supernatants of bacterial cultures grown in DMEM. After TCA 79 precipitation, these proteins were analyzed by 12% SDS-PAGE. Whole cell lysates were analyzed by Western blotting using anti-EscJ polyclonal antibodies to monitor EscJ expression. Both complemented strains showed equal levels of secreted proteins as the wild type strain, suggesting that the triple mutation does not affect the function of EscJ. (c) Localization of the mutation site in the crystal structure of EscJ. Glu62, Lys63, and Glu64 are located in a solvent accessible loop that does not participate in intramolecular interactions. The mutations do not appear to affect the overall structure of EscJ, and the substituted alanines are involved in crystal contacts in two out of four molecules of the asymmetric unit (colored in gold). Furthermore, examination of the crystal structure revealed that the triple mutation localizes to a surface-accessible loop, and does not interfere with protein folding and organization of secondary structural elements (Figure 3.4c). Interestingly, this loop region is involved in intersubunit interaction of two out of four molecules in the asymmetric unit. 3.3.3 Crystal structure of EscJ EscJ crystallized in the space group P65 with four molecules in the crystallographic asymmetric unit. The refined model shows that the EscJ monomer is a flat, triangular-shaped molecule consisting o f two mixed a/p domains with similar and novel topologies (Figure 3.5a). The two globular regions are connected by a relatively extended linker region (Gly77 to Ala84) that may serve to fine-tune the angular orientation of the two domains. The four EscJ molecules in the asymmetric unit, which are positioned in tandem and pack into a compact arc-shaped tetramer, adopt highly similar conformations, with the matched 153 C a atoms superimposing with root mean squared deviation (r.m.s.d.) values in the range of 0.5 to 1.0 A (Figure 3.5b). Figure 3.5 E s c J structure and intermolecular interactions. (a) Ribbon representation of EscJ. The triangular-shaped monomer consists of two topologically-similar mixed ct/p domains, (b) Arc-shaped EscJ tetramer in the crystallographic asymmetric unit. 81 The four molecules are viewed with their N-termini projecting into the page, (c) Interface between domain 1 of two EscJ monomers. The two monomers are colored green and blue respectively, and the residues involved in hydrogen bonding interactions were highlighted with dotted lines depicting potential hydrogen bonds, (d) Interface between domain 2 of two EscJ monomers. Residues from two a-helices of the green monomer interact with residues from the (3-sheet of the blue monomer, (e) Surface representation of tetrameric EscJ colored by electrostatic potential. The orientation of the molecules is approximately the same as in (b). Positively-charged patches (colored blue) are observed near the top of the outer face of the tetramer. (f) Molecular surface of tetrameric EscJ on the inner face. The surface in (c) is rotated 180° and shown with the same coloring schemes (positive = blue, negative = red). A number of notable features were observed when examining the interactions between the EscJ monomers. First, each EscJ monomer is involved in extensive interactions with its two neighboring molecules over its entire length, resulting in 35% of total solvent accessible surface or approximately 3500 A 2 being buried. Second, the EscJ-EscJ interface contains a significant number of charged residues, and several of them are involved in hydrogen-bonding interactions. The majority of these interactions are repeated between any two monomers. More specifically, the side chains of Glu30 and Lys49 in domain 1 hydrogen bond with the backbone amide and carbonyl of residues Thr26' and Gln23' o f the adjacent chain, while the side chains of Glu35 and Gln37 hydrogen bond to the side chains of Asn74' and Glu22' respectively (Figure 3.5c). Similarly, the side chains of L y s l 10, A r g l 16, and G l n l 12 of the slightly larger domain 2 make hydrogen bond contacts with the backbone carbonyls and amide o f L e u l 3 1 ' and Ala84', while the side chains o f L y s l 2 0 of helix a3, Serl58, G l n l 6 1 , and Asn l68 of helix GI4 hydrogen bond with residues Aspl26 ' , Serl28', Lys l83 ' , and Serl79' o f the 3-stranded (3-sheet in the adjacent chain (Figure 3.5d). A consequence of the burial of charged residues and formation of these hydrogen bonds at the oligomeric interface is a dramatic 82 overall change in disposition of electrostatic charge on the molecular surface. While charges appear to be scattered randomly along the surface of the EscJ monomer, two charge patches (one positive and one negative) accumulate on opposite faces (referred to as the outer face and the inner face) of the EscJ tetramer (Figure 3.5e and Figure 3.5f). Third, the tetramer surface at the inner face consists of four deep cavities lined with several lysine and arginine residues (Figure 3.5f). Finally, in addition to the hydrogen bonding interactions, there are many van der Waals and hydrophobic contacts dispersed across the interaction interface. The extensive and specific nature of these intermolecular contacts indicates that EscJ could potentially form higher ordered structures. 3.3.4 Modeling the EscJ ring Further analysis of the molecular packing in the EscJ crystal revealed that symmetry-related tetramers pack into a super-helical structure via similar interactions as detailed above (Figure 3.6a), indicating that EscJ could form a much larger assembly. This superstructure, whose helical axis is parallel to the crystallographic 65 screw axis, contains 24 EscJ monomers per helical turn with a pitch equal to the length of the crystallographic c axis (67 A). The center of this super-helix, with an estimated width of 75 A, is solvent-filled and completely devoid of protein densities (Figure 3.6b). The N-termini of all the EscJ molecules in the crystal project in the same direction with respect to the super-helical axis, and when viewed along this axis, this super structure resembles a stack of tilted rings.(Figure 3.6c). In the crystallization environment, the non-lipidated nature of the protein and the absence of a membrane would permit axial movements of individual molecules. However, such translations are severely restricted in the physiological environment because the N-terminal lipid anchors the protein to the 83 inner membrane and constrains EscJ molecules to only planar movement. Thus, EscJ is more likely to form flat, circular oligomers than superhelices in the physiological setting. Using a modeling procedure that involves projecting individual molecules in one direction without altering existing orientation of the individual molecules, the crystallographic superhelical superstructure is collapsed into a flat "ring". Because of the large number of molecules involved, the translation of individual molecules is small, and several of the favorable intermolecular contacts observed amongst different EscJ molecules in the crystal structure are preserved in the model. Figure 3.6 EscJ packs into superhelical structures in the crystal lattice. 84 (a) One turn of the EscJ superhelix. The EscJ tetramer generates a superhelix by repeating itself six times in a circular fashion along the crystallographic 6 5 screw axis. Ribbon representation of one full turn of the superhelix, which contains 24 EscJ molecules, is shown, (b) Molecular packing of EscJ viewed above the 6 5 screw axis. For clarity purposes, only the Ca-backbone of the molecules is shown. The superhelical arrangement extends along the entire length of the crystal, resulting in long solvent-filled central channels, (c) Molecular packing of EscJ viewed in projection along the 6 5 axis of the crystal. The superhelices in this view resemble stacks of tilted rings. The refined EscJ ring model has an overall diameter of 18uA and height of 52 A (Figure 3.7a). These dimensions match the values previously estimated for the inner membrane ring from E M images o f purified E P E C N C s (Sekiya et al, 2001). This model maintains the 24-fold symmetry observed in the crystal, and this subunit number largely agrees with the stoichiometry of PrgK (22.0 + 1.7) obtained from labeling experiments by Dr. Tyler Kimbrough, a former graduate student in collaborator Dr. Sam Mi l le r ' s laboratory, on the Salmonella typhimurium N C (Kimbrough, 2002). The N-termini of the individual subunits localize to the wider face of the super-molecular complex, suggesting that the ring is anchored to the inner membrane at this end with the opposite face extending towards the periplasm (Figure 3.7a). The central channel of the ring constricts from -120 A at the membrane face to -73 A at the periplasmic face. The overall dome-shaped morphology and the two-layered exterior appearance o f the EscJ ring are strikingly similar to the recent 22 A c ryo-EM structure of the flagellar F l i F ring (Suzuki et al, 2004). The interior of the ring is characterized by a prominent negatively-charged "ridge" and is fully circumscribed by the unusual positively-charged cavities observed in the tetramer structure (Figure 3.7b). The periplasmic opening of the central channel, on the other hand, is surrounded by a deep and negatively charged "trench", approximately 32 A in height and 11 A in width (Figure 3.7b). 85 (a) (b) Figure 3.7 Modeling and surface electrostatics analysis of the EscJ ring. (a) Ribbon and surface representation of the modeled 24-subunit EscJ ring. The side view shows the two-layered exterior structure highly analogous to the FliF ring structure determined by cryoEM. The N-termini for all subunits are located at the wide face of the ring, with two of them labeled "N" in the ribbon diagrams, (b) Surface electrostatics of the EscJ ring. The periplasmic face and the inner membrane face of the channel are shown, and the ring is slightly tilted to allow better visualization of the depth. The trench region surrounding the central channel at the periplasmic face is quite negatively charged. The spacious interior of the ring showed the same set of cavities observed in the tetramer. This model is further validated by data from the biotinylation coupled to mass spectrometry surface accessibility analysis of PrgK, obtained from Dr. Tyler Kimbrough in Dr. Sam Mil le r ' s lab (Figure 3.8a) (Kimbrough, 2002). More specifically, the lysine 86 residues of PrgK inaccessible to biotinylation (Lys 4, 8, 31, 56) correspond to residues buried at the intermolecular interfaces in the EscJ model (Figure 3.8b). (a) Purified NC A Biotinylation R e p u r i f i c tion B SDS/PAGE •fr-i t Trypsin | » | digest MALDI-TOF Mass Spec. m/z (b) ( . ' K D K D I . . I K G l , l ) Q E g A N K V I A V l . y M H N I E A N K I l » . S < i K l . < ; V . S r r V A F P n h T A A V ' V W I K T Y Q I P P R P R V F . I A Q M F P A D S I V S S P R A F k ' A H I . Y S A I I ORl . lCQSI . y l ,MI( . \ I,SAK\ INSn>II>A<, l NGRPPkPS I1I.SAI. A Y M R C S P I .AIIQIS1MKR1 l § N S I ' A ) ) V I ) y D N I S V V I . S E R S I > A Q I X > A K ; T P V S R N S K \ T S ^ l . M « l l » ™ i k ^ W i l i m m » ^ K N H V A R N « f c : | - r A I > l * A f e S N I . -a > u> CKBKIM .l .KGI.BQFQA .NKVi.W Q K y m > : y s i QTMKA I.S\R\ IIMI>II)A<;F.\<;RPPBP\IHSAI A\AF.Rf.spi.AHQISDI|RFI|N$FAD\'DYDNISVVI.SERSDAQI.A)AP«TPV i° | R . \ S F A | s ttq»i^tfal»t'l |>f»m'NHVARN||| |<;llAI)ll | |A|| |^SNF (c) periplasm inner membrane Figure 3.8 Surface mapping of Salmonella typhimurium NC by limited biotinylation and MALDI-TOF mass spectrometry. (a) Schematic of NC surface mapping. Purified NCs were treated with biotinylation reagent. Protein components of biotinylated NCs were then separated by SDS-PAGE and visualized by Coomassie Blue staining. Biotinylated protein band corresponding to PrgK was excised and digested in situ with trypsin. Tryptic peptide fragments were eluted and their masses determined 87 by MALDI-TOF mass spectrometry. The resulting mass spectra were then analyzed using the ExPASy proteomics tool (FindPept program, http://www.expasy.ch/tools/findpept.html) to identify tryptic peptides containing biotinylated lysine residues, (b) Graphical representation of MALDI-TOF mass spectrometric peptide mapping analysis of biotinylated peptide fragments from PrgK after in situ trypsin digestion and streptavidin affinity purification. Purified NC was either biotin-labeled directly (-SDS) or denatured by boiling in 2% SDS (+SDS) prior to labeling. Biotin-modified lysine residues are highlighted in red. Residues corresponding to the single transmembrane segment of PrgK are highlighted in grey, (c) Surface lysines of PrgK are inaccessible to biotinylation. Three surface-exposed lysines of PrgK are mapped onto a surface representation of the EscJ ring model in grey, with the corresponding residues in EscJ labeled in parentheses. Inaccessible residues are colored blue whereas accessible or biotinylated residues red. 88 3.4 DISCUSSION 3.4.1 Localization of YscJ proteins Ambiguity in topology was a major hindrance towards understanding the biological function of YscJ proteins. It is generally accepted that members of the YscJ family are lipidated at their processed N-termini, and that many of these proteins possess a C-terminal transmembrane segment spanning the inner membrane. The major area o f debate or uncertainty concerns the localization o f this lipoprotein. Based on analysis of the signal sequence of Shigella M x i J , Al laoui et al. first proposed that the YscJ proteins span the entire periplasmic space with their acylated N-termini projected to the outer membrane and a C-terminal transmembrane segment embedded in the inner membrane (Allaoui et al, 1992). This model, however, has been challenged by several lines of evidence. First of all , no bacterial lipoproteins characterized to date are known to adopt this unusual topology. Secondly, more detailed sequence analysis revealed that the N -terminal lipoprotein sorting signal of YscJ proteins is actually not conserved, and thus the typical lipoprotein-sorting model is actually not suitable for predicting the membrane localization of this l ipid (Blocker et al, 2001). In this study, the key issue concerning the N-terminal lipid was addressed by determining the membrane localization of EscJ from E P E C . EscJ is a member of the YscJ family that lacks the C-terminal transmembrane segment, and its localization to the membrane is dictated by its lipidated N-terminus. The observation that EscJ partitioned only to the inner membrane fractions in the sucrose gradient conclusively showed that the N-terminal l ipid of this and other members of the YscJ family is anchored to the outer leaflet of the inner membrane rather than the inner leaflet of the outer membrane. So, 89 instead of spanning the entire periplasmic space, the core of EscJ and other YscJ proteins is perched on top of the inner membrane with the terminal regions anchored to or embedded in the l ipid bilayer. The reason why some members of the YscJ family such as EscJ do not possess a C-terminal membrane anchor is not known, but this property may explain why EscJ is easily dissociated during the detergent treatment step in the E P E C N C preparation procedure (Ogino et al, 2006). 3.4.2 Oligomerizat ion of Y s c J proteins Members of the Y s c J family, which share sequence homology with an N-terminal domain of F l iF , have been hypothesized to be structurally similar to this flagellar protein (Suzuki et al.,, 1998). F l iF spontaneously oligomerizes into ring-shaped complexes in the inner membrane with an estimated stoichiometry of 26 (Suzuki et al, 2004). However, similar self-association has not been observed for YscJ proteins. A recent systematic yeast 2-hybrid analysis o f E P E C T3SS proteins, despite uncovering novel interactions between several L E E proteins, did not show binary interaction of EscJ with itself (Creasey et al, 2003a). In the same vein, EscJ devoid of its signal peptide and N-terminal l ipid was found to exist as a monomer in solution. One possible explanation for this phenomenon is that the intermolecular interactions required for maintaining higher ordered structures might be too weak to be detected using conventional biochemical approaches. In the crystallization condition, in which the protein is present in a supersaturated environment, weaker interactions are sometimes preserved and occasionally enhanced by molecular packing. These effects were indeed observed in the EscJ crystal structure. 90 The extensive and repetitive nature of the subunit interface strongly suggests that the ordered arrangement of EscJ subunits within the superhelical structure is unlikely to be a pure crystallographic artifact. Rather, it reflects the inherent potential of EscJ and possibly other YscJ proteins to associate into .larger oligomers. The observed EscJ-EscJ interface in the crystal structure is dominated by highly specific but short range hydrogen bonding interactions. Helical packing made these interactions possible by bringing different EscJ molecules in close contact with their optimal interaction interface oriented towards neighboring subunits. Such packing also appears to fix the position of the linker region connecting the two domains in EscJ, which has been shown by nuclear magnetic resonance ( N M R ) spectroscopy to be highly flexible (Crepin et al, 2005a). B y restricting the conformational freedom of this linker, EscJ is trapped in a state competent for self-association and is prevented from adopting other conformations that may hinder oligomerization. In the physiological environment, membrane anchoring by the terminal regions of EscJ and other YscJ proteins may promote intersubunit interaction and subsequent oligomerization in a similar manner as crystal packing. Constraining the protein in a planar environment dramatically increases the local concentration and effectively reduces the distances between individual molecules. Tethering o f the termini to the membrane reduces conformational freedom of the two domains and the flexible linker connecting these regions, maintaining these proteins in an orientation optimal for self-association. While further work is needed to address whether anchoring EscJ and other members of the Y s c J family to a surface would result in formation of large oligomers, an arrangement that these proteins likely adopt, based on the crystallographic superhelical 91 structure of EscJ, is the 24-subunit flat ring. The procedure used for modeling the EscJ ring is gentle since each individual molecule is not rotated but is translated only by a small distance in one direction. A similar method has been previously used to model the hexameric ring of T4 helicase, a protein which packs into superhelical structures in the crystal, and this model did not deviate much from the actual hexameric ring whose crystal structure was determined later (Sawaya et al, 1999; Singleton et al, 2000). Several lines of evidence suggest that the EscJ model is valid and similar to the physiological structure. Firstly, the N-termini of all subunits project to the same direction, a requirement for the physiological structure. Secondly, the stoichiometry o f EscJ in the model is very close to that estimated for PrgK within the S. typhimurium SPI-1 N C , and the EscJ ring model could be docked into a region of the S. typhimurium N C E M map corresponding to the inner membrane ring (Marlovits et al, 2004) (Figure 3.9). Although a small range o f oligomeric states has been shown to exist in the physiological setting (Marlovits et al, 2004), the relatively large number of subunits constituting the ring means that adding or subtracting a few molecules would not dramatically alter the overall dimensions and structural properties. Finally, surface accessibility of P rgK mapped out experimentally agrees with that predicted from the EscJ model. Can one learn anything about the oligomerization of F l iF from EscJ? The E M structure of F l iF ring has been determined to sufficient quality to distinguish subunit boundaries. However, the limited resolution has prevented a detailed understanding of the molecular basis o f the intersubunit interactions. It is conceivable to hypothesize that the domain of F l iF mediating M S ring formation probably lies in regions conserved with the Y s c J family of proteins, while residues present in F l iF but not P rgK proteins may be 92 responsible for forming the rod-like extension that protrudes from the M S ring into the bacterial periplasm as well as the integral membrane domains. Figure 3 .9 EscJ ring model in context of the needle complex. The EscJ ring model is manually docked into the 17A cryoEM map of Salmonella typhimurium needle complex (3D-EM Database accession code: emd1100) using Pymol. The side view shows that the ring model occupies significant portion of the cryoEM map in a region corresponding to the inner membrane ring. The top view shows the relative good fit of the ring model to the cryoEM map despite slight differences between the symmetry of the EscJ ring model (24) and the subunit number used for 3D reconstruction in the cryoEM structure (21). 3.4.3 Function of Y s c J proteins Flagellar FliF plays a structural role in the flagellar system by forming the critical M S ring structure that houses the export apparatus and acting as a mounting platform for the rotor proteins and the switch complex. Analysis of the EscJ ring model revealed features similar to the M S ring, indicating that EscJ as well as other YscJ proteins may play an analogous role as a molecular platform for T3SS assembly. The large interior together with the regularly-spaced and highly polarized electrostatic features of the dome-shaped EscJ ring provide sufficient room and specific interaction sites for anchoring the periplasmic domains of the integral membrane components of the export apparatus including EscV, EscR, EscS, EscT, and EscU and their equivalences in other 93 species. The prominent negatively charged trench in the periplasmic opening of the EscJ ring, which is inaccessible based on the PrgK biotinylation study (Figure 3.8c), may serve as the critical adaptor region for binding the central rod, a helical polymeric structure that connects the inner and outer membrane rings. Interestingly, data from surface mapping experiment on PrgK also showed that the predicted outer surface of the ring is relatively inaccessible and likely buried (Figure 3.8c). This, together with the fact that the EscJ ring model could not account for all the densities corresponding to inner ring in the E M map, imply the presence o f an additional protein component. Previous E M studies have shown that overexpression o f S. typhimurium PrgK and PrgH but not PrgK alone in E. coli results in formation of circular complexes in the inner membrane (Kimbrough and Mil ler , 2000). Although not broadly conserved across species, PrgH, which is believed to be the functional equivalence of the Y s c D proteins, may function to anchor the relatively large ring constructed by the YscJ proteins in the dynamic membrane environment. The biochemical and structural characteristics of PrgH w i l l be described in the following chapter. 94 CHAPTER 4 - Structural and biochemical characterization of PrgH from Salmonella typhimurium 4.1 I N T R O D U C T I O N PrgH is generally considered the second major component of the inner membrane ring of the S. typhimurium SPI-1 T3SS since overexpression of this protein and PrgK, a member of the YscJ family, results in formation of ring-shaped complexes in the bacterial membrane (Kimbrough and Mil le r , 2000). It has been proposed that PrgH forms an oligomer that encompasses and stabilizes the multimeric structure formed by PrgK in the inner membrane (Kimbrough, 2002) (Figure 4.1). In spite of this important and likely conserved function, PrgH does not resemble any components of the flagellar system (Kubori et al., 1998). Even more intriguing is the fact that Shigella T3SS protein M x i G is the only protein from any species to demonstrate a significant level of identity (24%) at the protein sequence level to PrgH (Blocker et al., 2001). Such divergence is in sharp contrast to the high level ofconservation observed for proteins of the YscJ family. Members of the Y s c D family have recently been suggested to play an equivalent functional role to PrgH in other T3SSs. This hypothesis is based on similarities in their predicted topologies: a small N-terminal cytoplasmic domain followed by a single transmembrane region and a large domain protruding to the periplasm (Ghosh, 2004) (Figure 4.1). In agreement with this prediction, N-terminal sequencing of components of the S. typhimurium SPI-1 N C showed that PrgH is neither lipidated nor processed by SPase (Kubori et al., 1998). The topology o f PrgH, however, has not been verified in vitro or in vivo, and exactly how PrgH binds or stabilizes the oligomeric complexes formed by YscJ proteins remains unclear. 95 1 141 165 392 Cytoplasmic domain TM Periplasmic domain Figure 4.1 Domain organization of Salmonella typhimurium PrgH. PrgH has one predicted transmembrane segment from residues 142 to 164 (labelled TM), with an N-terminal domain projecting to the cytoplasm and a C-terminal domain to the periplasm. PrgK, a member of the YscJ family, is anchored to the inner membrane by a lipid covalently-linked to its processed N-termini, and a C-terminal transmembrane segment. PrgH likely encompasses the outer surface of the oligomeric ring structure elaborated by PrgK on top of the inner membrane. A s a first step towards resolving some of these outstanding issues, the structural and biochemical properties o f PrgH were analyzed. Ful l length PrgH was overexpressed and the ability of different detergents to extract this protein was examined. Binding between the periplasmic domains of PrgH and PrgK was investigated by nickel pull-down assays. Finally, the structure of the core periplasmic domain of PrgH, PrgH(170-362), was determined by X-ray crystallographic methods and refined to 2.3A. The data presented in this chapter proved that PrgH indeed localizes to the bacterial membrane. However, this protein does not interact with the periplasmic domain of PrgK in the absence o f membrane anchoring. The structure o f the periplasmic domain o f PrgH revealed the presence of two EscJ-like domains, suggesting that this protein may have the 96 potential to oligomerize in a similar manner as YscJ proteins. The majority of the work in this study was performed by the author. Detergent extraction and purification of full length PrgH were carried out by Marija Vuckovic . Purification and crystallization of PrgH (170-362) were assisted by Angel Y u . 97 4.2 METHODS 4.2.1 Cloning, protein expression, and purification For full length PrgH or PrgH(l-392), the periplasmic domains PrgH(170-392) and PrgH( 170-362), expression constructs were generated by PCR-amplifying the corresponding prgH sequence from S. typhimurium genomic D N A and cloning the amplified D N A into the NdeVBamHl sites of the pET-28(a) vector. The expression construct for PrgH(l-138) was provided by Dr. Sam Mil le r ' s lab at the University of Washington. This pET-15-based plasmid encodes full length prgH with the codon of residue 139 mutated to a stop codon. E. coli BL21(A,DE3) transformed with these constructs were grown to mid-exponential phase at 37 °C in L B broth containing 50 ug mL" 1 o f kanamycin, induced with 0.5 m M IPTG, and incubated for an additional 16 hours at 20 °C before harvesting. A l l overexpressed proteins possess N-terminal hexa-histidine tags. For purification of PrgH(l-138), induced cells were first resuspended in buffer (20mM H E P E S + 150mM N a C l , p H 6.8), lysed with a pressurized homogenizer, and centrifuged at 25,000 x g for 35 minutes. The protein was purified from the soluble fraction using nickel-charged chelating sepharose. Thrombin was added to the eluted protein and the sample was dialyzed overnight against 20mM H E P E S , p H 6.8, and further purified by Mono Q 5/5 and Superdex-75 H R 10/30 columns. For purification of PrgH(170-392) and PrgH( 170-362), induced cells were first resuspended in buffer (20mM H E P E S + 150mM N a C l , pH 6.8), lysed with a pressurized homogenizer, and centrifuged at 25,000 x g for 35 minutes. The protein was purified from the soluble fraction with cobalt-charged chelating sepharose. Thrombin was added 98 to the eluted protein and the sample was dialyzed overnight against 20mM H E P E S , p H 6.8, and further purified by Mono S 5/5 and Superdex-75 H R 10/30 columns. 4.2.2 Detergent extraction Cells induced to overexpress full length PrgH were resuspended in buffer (20mM H E P E S + 150mM N a C l , pH 6.8), lysed by sonication, and centrifuged at 15,000 x g for 20 minutes. The pellet or insoluble fraction was subjected to extraction with resuspension buffer containing different types o f detergents (Anatrace): Anzergent 3-14 ( L 6 m M or 10 x C M C ) , C H A P S (16mM or 2 x C M C ) , n-decyl-P-D-maltoside (DM) (18mM or 10 x C M C ) , n-dodecyl-p-D-maltoside ( D D M ) (1.7mM or 10 x C M C ) , Fos-choline-12 (15mM or 10 x C M C ) , n-octyl-p-D-glucoside (OG) (20mM or 1 x C M C ) , N , N -dimethyldodecylamine-N-oxide ( L D A O ) (15mM or 10 x C M C ) . The extraction was allowed to proceed for 1 hour at 4 °C, and after centrifugation at 14,000 x g for 15 minutes, the supernatant containing the extracted proteins was analyzed by S D S - P A G E . 4.2.3 Nicke l pull-down assays The expression constructs pET28PrgK(20-200), which encodes the periplasmic fragment of PrgK with an N-terminal his-tag, and pET21PrgK(20-200), which encodes a tagless version of the same protein fragment, were generated by PCR-amplifying the corresponding region ofprgK from S. typhimurium genomic D N A and cloning the amplified D N A into the Ndel/BamRl sites of pET-28(a) and pET-21(a) respectively. The expression construct pET21PrgH(l 70-392) was generated by subcloning the insert of pET28PrgH( 170-392) into the Ndel/Bamm sites o f pET-21(a). Two different expression strains were generated by co-transforming E. coli B L 2 1 ( m E 3 ) with either pET28PrgK(20-200) and pET21PrgH( 170-392) or 99 pET21PrgK(20-200) and pET28( l 70-392). For co-expression of proteins, 50mL of cultures were grown in L B broth containing 50 p.g mL" 1 o f kanamycin and 100 ug mL" 1 of ampicillin to mid-exponential phase at 37 °C, induced with I m M IPTG, and incubated overnight at 20 °C. Cells were then harvested, resuspended in 2mL of buffer (20mM H E P E S + 150mM N a C l , pH 6.8), and lysed by sonication. After centrifugation at 14,000 x g for 20 minutes, the supernatant was collected and incubated with 100 uL of cobalt-charged chelating sepharose on ice for 10 minutes. The resin was subsequently washed twice with buffer containing 50mM imidazole, and bound proteins were eluted with buffer containing 300mM imidazole and analyzed by S D S - P A G E . 4.2.4 Crystal l izat ion of P r g H periplasmic domain A l l crystal trials were performed using the hanging drop vapor diffusion method by mixing 1 uL o f protein solution (12 to 24 mg mL" 1 ) with 1 (J.L o f reservoir solution. PrgH(l70-392) crystallizes in the condition (16% P E G 3350 + 0.2 M tri-ammonium citrate + 0 .1M Tris p H 8.0) at room temperature over 10 to 14 days. The native crystals, cryo-protected with solution (20% P E G 3350 + 0 .2M N a C l + 0.1 H E P E S 7.5 + 10% M P D ) , diffracted to 2.9A at beamline 8.2.2 of the A L S . PrgH(l 70-362) initially crystallized in a variety o f conditions containing salts or P E G but these crystals did not diffract beyond approximately 8 A . Further screening identified two conditions that produce two different crystal forms suitable for X-ray analysis. Condition 1 (2.8M N a C l + 0 .1M Tris p H 8.5 + 0.01M cupric chloride) gave orthorhombic crystals, whereas condition 2 (30% P E G 1500 + 0 .1M lithium sulfate + 0.1M Tris 8.5 + 0.109M octanoyl-n-hydroxyethylglucamide or H E G A - 8 ) gave trigonal 100 crystals. Only condition 1 but not condition 2 produced crystals from SeMet-substituted proteins. 4.2.5 Data collection and structure determination A l l diffraction data were collected at beamline 8.2.2 of the A L S . A redundant S A D dataset was collected at the selenium peak wavelength from an orthorhombic SeMet crystal, cryo-protected in mother liquor containing 18% ethylene glycol, and flash-frozen to 100K. A complete dataset was collected from a native trigonal crystal already cryo-protected by the mother liquor. Data were processed with M O S F L M (Leslie, 1992) and scaled with S C A L A (Evans, 1993). For structure determination, selenium positions (6 out of 8 expected sites) were initially found by S H E L X - C / D (Schneider and Sheldrick, 2002) in the C C P 4 suite of programs (Collaborative Computational Project, Number 4, 1994) . After refinement with S H A R P (Bricogne et al, 2003), the sites were input into S O L V E for protein phasing and then R E S O L V E for density modification (Terwilliger, 2000; Terwilliger and Berendzen, 1999). A n initial model was manually-built from the experimental map using Xfi t (McRee, 1999) and refined with C N S (Brunger et al, 1998). This model was only partially complete and the first 50 residues could not be built due to poor quality of the map near the N-terminus. Molecular replacement was then performed on a high resolution trigonal native dataset using the partially-built structure as a search model. A clear solution was found by the program P H A S E R (McCoy et al, 2005), and this model was put through successive cycles of rebuilding and refinement using Xfit (McRee, 1999), Coot (Emsley and Cowtan, 2004), C N S (Brunger etal, 1998), and Refmac (Murshudov, 1997). The refined structure is essentially complete but the N -101 terminal 2 to 13 residues and several loops near this region could not be accurately modeled for each of the four molecules in the asymmetric unit due to disorder. 4.2.5 Structural alignment and surface electrostatics analysis Structural alignment of PrgH(l70-362) with EscJ (PDB code: 1YJ7) was performed using the programs Coot (Emsley and Cowtan, 2004) and SwissPdb Viewer (Guex and Peitsch, 1997). Surface electrostatics of PrgH( 170-362) were calculated and analyzed using the A P B S plugin in Pymol (Baker et al., 2001). 102 4.3 R E S U L T S 4.3.1 Membrane localization of full length P r g H Full length N-terminal his-tagged PrgH, when overexpressed in E. coli, sequestered to the insoluble traction after cell lysis. However, significant quantities of this recombinant protein could be recovered when the pellet was subjected to extraction by detergents, suggesting that PrgH is localized to the membrane compartment (Figure 4.2). O f the seven different detergents tested, the zwitterionic detergent Fos-choline-12 was the most effective in extracting PrgH. Interestingly, the detergent types (zwitterionic, nonionic, or ionic) do not appear to be a reliable prediction factor of extraction efficiency. The nonionic detergent n-dodecyl-P-D-maltoside (DDM), for example, could not extract any PrgH from the membrane, but a very similar detergent n-decyl-P-D-maltoside (DM) was amongst the best and most effective. Figure 4.2 Full length recombinant PrgH is a membrane protein. Full length PrgH expresses at relatively high level in E. coli BL21 but partitions to the insoluble fraction after cell lysis. The pellet was subjected to extraction by buffers containing various types of detergents and the samples analyzed by SDS-PAGE. Lane 1 represents buffer containing no detergent, whereas lanes 2 to 8 represents buffer containing the following detergents: 1.6mM Anzergent 3-14, 16mM CHAPS, 18mM n-decyl-P-D-maltoside (DM), 1.7mM n-dodecyl-P-D-maltoside (DDM), 15mM Fos-choline-12, 20mM n-octyl-P-D-glucoside (OG), and 15mM N,N-dimethyldodecylamine-N-oxide (LDAO). Significant quantities of recombinant PrgH could be recovered by detergent extraction, suggesting that this protein localizes to the bacterial membranes. 103 4.3.2 Purification and biochemical characterization of putative domains of PrgH Both the TMpred and T M H M M servers predicted that the region spanning residue 142 and 164 has a high propensity to form a transmembrane helix (Krogh et al, 2001). Thus, the cytoplasmic domain of PrgH likely corresponds to the region from residue 1 to 141, whereas the periplasmic domain spans from residue 165 to the C-terminal end (residue 392) o f the protein. These two putative domains, designated PrgH(l-138) and PrgH(l 70-392) respectively, were cloned and overexpressed. Both of these PrgH fragments are relatively soluble and could be purified to 95% pure using nickel or cobalt chelating sepharose (Figure 4.3a and Figure 4.3c). PrgH(l-138), which has a predicted p i of 4.6, binds to an anion exchange resin (MonoQ) at neutral p H . In contrast, PrgH(170-392) is relatively basic (predicted p i o f 8.9) and binds to a cation exchange resin (MonoS). Gel filtration and multiangle light scattering analyses suggested that these two proteins are monodisperse and exist as monomers in solution (Figure 4.3b and Figure 4.3d). The periplasmic domain of PrgH has been postulated to mediate formation of an oligomer that encompasses the ring complex formed by PrgK as well as to stabilize the PrgK oligomer via protein-protein interactions. To determine i f PrgH(l 70-392) could interact with PrgK, his-tagged PrgH( 170-392) was co-expressed with the periplasmic domain of PrgK, PrgK(20-200), which lacks its N-terminal signal peptide and C-terminal transmembrane segment. Although both proteins were soluble when overexpressed, only his-tagged PrgH(l70-392) could be precipitated by nickel resin upon incubation with the lysates (Figure 4.3e). The placement of the affinity tag likely did not affect binding as no 104 significant PrgK-PrgH interaction could be detected when the his-tag was switched to the N-terminus of PrgK(20-200). (a) (b) Volume (ml) (d) pET28aPrgK(20-200) + pET28aPrgH(l 70-392) + pET21 aPrgH( 170-392) pET21 aPrgK(20-200) Hi: His-PrgHd 70-392) ft •§ ff - A— PrgK(20-200) % l l Figure 4.3 PrgH contains two discrete domains. (a) Purification of PrgH(1-138). PrgH(1-138) is soluble and stable when overexpressed. His-tagged PrgH(1-138) could be purified by nickel-chelating sepharose from cell lysates. (b) Multiangle light scattering analysis of purified PrgH(1-138). The data is presented as a molar mass versus volume plot overlaid with a gel filtration elution profile. Purified PrgH(1-138) is 105 relatively monodisperse and exists as a monomer in solution, (c) Purification of PrgH(170-392). PrgH(170-392) is soluble and stable when overexpressed. His-tagged PrgH(170-392) could be purified by cobalt-chelating sepharose from cell lysates. (d) Multiangle light scattering analysis of purified PrgH(170-392). A molar mass versus volume plot overlaid with a gel filtration elution profile is shown. Purified PrgH(170-392) is monodisperse and exists as a monomer in solution, (e) Nickel pull-down assay. Interaction between PrgH(170-392) and PrgK(20-200) was analyzed by co-expressing the two proteins in the same host and incubating the lysates with nickel resin. No complex was detected in the elutions from the resin. The various samples are designated as follows: L=lysates, F=flowthrough, W1=wash 1, W2=wash 2, E=elution. 4.3.3 Crystallization and structural determination of PrgH periplasmic domain To gain further understanding of the structural properties of the PrgH periplasmic domain, attempts were made to crystallize and determine the structure of PrgH(l 70-392). When the his-tag was removed, this fragment of PrgH could be crystallized, with the best crystals diffracting to a maximum resolution of 2.9A. Unfortunately, the structural determination process was hampered by a series of technical issues. First of all, the unit cell o f this orthorhombic crystal form has one relatively long axis (-306A) (Table 4.1). This, together with inherent high mosaicity of these crystals, caused serious spot overlaps in the diffraction pattern and posed significant challenges in collecting quality data even with the larger 3x3 C C D detector on beamline 8.2.2 at the A L S . Secondly, SeMet-substituted PrgH(l70-392) proteins did not yield crystals suitable for X-ray analysis. Finally, extensive soaking and screening experiments failed to generate a new crystal form and heavy atom derivatives needed for phasing. The C-terminal 25 to 30 residues are predicted to be unstructured based on sequence analysis by Jpred and other secondary structure prediction servers. To determine i f removal of this segment would facilitate crystallization and structure determination, a slightly shorter version o f the periplasmic domain, PrgH( l 70-362), was 106 generated. PrgH(l 70-362) could be overexpressed and purified in a similar manner as PrgH( 170-392). Two types of PrgH(170-362) crystals were obtained after a new series of screening. The orthorhombic form and the trigonal crystal form diffract to maximum resolutions of 2.8A and 2.1 A respectively. Interestingly, only orthorhombic SeMet crystals could be prepared, and initial phases were obtained by the S A D method using a redundant dataset collected from one o f these crystals (Table 4.2). Table 4.1 Data collection statistics for PrgH(170-392). X-ray crystallographic data Dataset Spacegroup Unit cell (A) Wavelength (A) Resolution (A) Total reflections Unique reflections Completeness (%)1 Redundancy1 <l/al>1 1,2 R, merge (%)' Native C222! 79.6 x 188.7 x 305.6 1.24 2.80 1175870 55405 96.8 (81.4) 11.5(5.2) 27.5 (4.2) 0.069 (0.293) Values in parentheses correspond to the highest resolution shell. 2Rmerge-^ l(lhki)-<l>|/^ (lhki), where l h k ! is the integrated intensity of a given reflection. The initial experimental map from S A D phasing was of decent quality (Figure 4.4a) and about 65% of the residues could be built with confidence for each of the two molecules in the crystallographic asymmetric unit. However, the electron density corresponding to the N-terminal 50 residues as well as several loop regions remained' 107 poorly defined even after further refinement and phase combination. To complete the model, molecular replacement was performed on a high resolution dataset collected from a trigonal crystal (Table 4.2), using the partially-built structure as a search model. A correct solution was readily determined and upon refinement, the map was significantly improved, and the majority of the missing regions could be sequentially built (Figure 4.4b). Table 4.2 Data collection and structure refinement statistics for PrgH(170-362). X-ray crystallographic data Dataset SeMet Native Spacegroup P2 12 12 P3, Unit cell (A) 87.0 x 106.6x55.5 53.5 x 53.5 x 282.5 Wavelength (A) 0.980 0.980 Resolution (A) 2.80 2.10 Total reflections 183987 150468 Unique reflections 13269 49331 Completeness (%)1 100 (100) 94.1 (71.0) Redundancy1 13.9(14.4) 3.1 (2.0) <l/ol>1 29.4 (9.9) 20.6 (5.5) Rmerge(%) 0.074 (0.275) 0.037 (0.137) Crystal structure refinement (trigonal) R /R a ' 4 rework' "Mree 0.206/0.255 Average B-factor 23.5 R.m.s. deviation Bond lengths (A) 0.013 Angles (°) 1.33 Ramachandran plot most favorable 92.5% allowed region 7.5% disallowed region 0.0% 108 Values in parentheses correspond to the highest resolution shell. 2Rmerge= l^(lhki)-<l>|/^ (lhki), where lhki is the integrated intensity of a given reflection. 3Rwork=(£|F0-Fc|)/(IF0), where F 0 and F care observed and calculated structure factors 4Rfree was calculated from 10% of reflections excluded from refinement. Figure 4.4 Improvement of electron density map. (a) Sigmaa-weighted 2F 0 -F cmap of PrgH(170-362) crystallized in the orthorhombic spacegroup $1^2. The map shown here is contoured at 1 .Oo and centered at Arg220 of one of the molecules in the asymmetric unit. The poor quality of the density map prevented modeling of the N-terminal region (residues 170 to 219). (b) Sigmaa-weighted 2F 0 -F cmap of PrgH(170-362) crystallized in the trigonal spacegroup p3-,. This map shown here is contoured at 1.0o and centered at Arg220 of one of the molecules in the asymmetric unit. Improved densities allowed model building for most of the N-terminal region. 4.3.4 Architecture of PrgH(l70-362) and surface electrostatic analysis The asymmetric unit of the trigonal crystal form (spacegroup P 3 1 ) contains four molecules that form a closely-packed tetramer (Figure 4.5a). This propeller-shaped tetramer is actually a dimer of two head-to-tail dimers, which are related by a 2-fold symmetry and maintained by interactions between two 3-stranded P-sheets contributed by one molecule in each o f the two dimers. The four molecules in the asymmetric unit adopt higher similar conformations, as one can superimpose 145 matched C a atoms with r.m.s.d. values in the range o f 0.6 to 1.0 A. 109 Each PrgH(l 70-362) monomer, which is relatively compact and has an overall "boot" shape, is made of three discrete a/p domains (domains I, II, and III) connected in tandem by two short strand-to-helix linkers (Figure 4.5b). Interestingly, all three domains of the monomer are topologically-similar. Each of these domains consists of a three-stranded p-sheet overlaid by one (domain I) or two (domains II and III) a-helices of approximately the same length as each P-strand. Results from the D A L I server, which performed 3D comparison with other entries in the Protein Databank (PDB) (Holm and Sander, 1996; Holm and Sander, 1997; Holm and Sander, 1998), indicated that although PrgH(l 70-362) does not appear to fall into a specific family, it is structurally similar to EscJ (despite no detectable sequence similarity), with a reported Z score of 6.3. Further structural alignment analyses revealed both domains II and III of PrgH(170-362) could be overlapped with domain 2 of EscJ (Figure 4.5c and Figure 4.5d). However, some small differences, particularly the orientation of the p-sheet with respect to the two a-helices, exist between the domains of these two proteins. Despite the overall basic nature of PrgH( l 70-362), surface electrostatic analysis showed that the bulk of the charges localizes to discrete patches on its solvent accessible surface (Figure 4.6). The most distinctive of these surface features are two charged grooves, one positive and one negative, located on opposite sides of the monomer respectively. Further examination identified A s p l 7 4 and Glu225 as the major residues responsible for the negative charge whereas the accumulation of charges at the positive patch is conferred primarily by the side chains of Arg247 and Arg353 (Figure 4.6). 110 Figure 4.5 Architecture of PrgH(170-362). (a) Propeller-shaped tetramer in the crystallographic asymmetric unit. This ribbon representation is shown in stereo. This tetramer is formed by dimerization of two head-to-tail dimers, related by a 2-fold symmetry. The major interaction involves two 3-stranded 3 sheets from domain III of each of the two dimers. (b) Ribbon representation of PrgH(170-362). The "boof'-shaped monomer consists of three mixed a/3 domains (I, II, and III) connected by short linkers. The N-terminus and C-terminus of the model are labeled "N" and "C" respectively, (c) Alignment of PrgH(170-362) I l l with EPEC EscJ. Domain III of PrgH(170-362) and domain 2 of EPEC EscJ (PDB code 1YJ7) were overlapped using the program Coot, (d) Alignment of PrgH(170-362) with EPEC EscJ. Domain II of PrgH(170-362) and domain 2 of EPEC EscJ were overlapped using the program Swiss-Pdb Viewer. Figure 4.6 Surface electrostatic analysis of PrgH(170-362). (a) Negatively charged groove of PrgH(170-362). In spite of the net overall positive charge of PrgH(170-362), a prominent negatively charged groove is present on its solvent-accessible surface (red = negative, blue = positive) and is conferred primarily by the carboxylate side chains of Asp174 and Glu225. (b) Positively charged groove of PrgH(170-362). The surface in (a) is rotated approximately 180° and shown with the same coloring scheme. The positively charged groove is located at the mid-region of the molecule and is conferred primarily by the side chains of Arg247 and Arg353 with Lys230 and Arg348 contributing to the charge in the region peripheral to the groove. (a) (b) 112 4.4 DISCUSSION 4.4.1 Topology of PrgH and the proposed role PrgH is predicted to be a monotopic membrane protein with two separate domains protruding to the bacterial cytoplasm and periplasm respectively. Work described in this chapter provided experimental evidence supporting this topology. First o f all, overexpression and detergent extraction confirmed that PrgH localizes to the bacterial membrane. PrgH is most likely to be embedded in or anchored to the inner membrane since it is not transported to the periplasm nor processed by SPase (Kubori et al., 1998). Secondly, fragments of PrgH corresponding to the predicted cytoplasmic and periplasmic regions exist as independent folding domains in solution, suggesting that membrane-targeting of this protein arises primarily from the predicted transmembrane segment. One o f the most perplexing question surrounding PrgH initially was that while this protein is undisputedly one of the most important and abundant proteins in the prototypical S. typhimurium SPI-1 T3SS, orthologues of this putative inner membrane component could not be identified in other T3SSs at the protein sequence level (except for Shigella). The topology of PrgH has allowed the putative identification of the Y s c D family of proteins, which is the only group o f T3SS proteins predicted to adopt this domain arrangement, as the potential analogues of PrgH in other T3SSs (Ghosh, 2004). Although it remains to be confirmed experimentally, this hypothesis is largely supported by recent bioinformatic analysis, more specifically P S I - B L A S T , which revealed distant homologies between the cytoplasmic and periplasmic domains of PrgH and the Y s c D proteins (Pallen et al, 2005b). 113 With prominent domains situated on either side o f the inner membrane, the topology of PrgH and Y s c D proteins resembles that o f molecules involved in transmitting signals across cellular membranes. One might speculate that the periplasmic domain o f PrgH and Y s c D would act as a receptor for sensing environmental changes, and transfer the signal downstream via the transmembrane and cytoplasmic domains to regulate T3SS-related processes such as substrate switching. Interestingly, recent P S I - B L A S T analysis revealed that the cytoplasmic domain of Y s c D proteins is homologous to the forkhead-associated ( F H A ) domain, a phosphoserine / phosphothreonine recognition domain found in many bacterial and eukaryotic regulatory proteins involved in diverse cellular processes including signal transduction, protein degradation and D N A repair (Pallen et al, 2002). While it is currently not known i f reversible Ser/Thr phosphorylation has a role in modulating T3SS activities, one could subject the purified cytoplasmic domain of PrgH or Y s c D proteins to an interaction screen against phosphopeptide libraries to test this hypothesis in vitro. 4.4.2 The periplasmic domain and PrgH oligomerization Although the periplasmic domain of PrgH may be involved in sensing, the primary roles of this domain are probably to mediate the formation of PrgH oligomers and to stabilize the PrgK ring structure (Kimbrough, 2002). The observation that the periplasmic domains o f PrgK and PrgH do not apparently interact in solution was not entirely unexpected. It was evident from work on EscJ (Chapter 3) that membrane anchoring is l ikely required for oligomerization and protein-protein interactions of these inner membrane ring components. Indeed, recent unpublished data from collaborators in Dr. Sam Mil le r ' s laboratory at the University of Washington have shown that full length 114 PrgH co-purified with mature PrgK in membranes isolated from Salmonella typhimurium. Unfortunately, the crystal structure o f the core periplasmic domain of PrgH, PrgH(170-362), in spite of its high resolution, could not offer much insight into the molecular basis of the potential PrgH-PrgK interaction. More specifically, from the structure alone, one cannot unambiguously deduce the orientation of the periplasmic domain of PrgH with respect to PrgK within the fully assembled T3SS. The unique surface electrostatics of the PrgH monomer, however, provide a template for designing rational mutagenesis experiments aimed at probing the potential function of PrgH and mapping the interaction interface with PrgK. Unlike EscJ, the packing of PrgH(l 70-362) in the crystal does not offer much clues regarding how PrgH might form higher-ordered oligomeric structures. Despite the presence of an EscJ-like domain in this protein fragment, the observed interaction between molecules in the crystal occurs in a non-parallel fashion and lacks the extensive and repetitive characteristics observed in the helical packing of molecules in the EscJ crystal. While it is possible that the packing of molecules in the crystal could not mimic the actual interactions amongst PrgH molecules in the physiological environment, another explanation is that PrgH lacks the capability to form large ring-like oligomers on its own. In the flagellar system, assembly of the F l i G ring and the C-ring proceeds after the generation of the M S ring (reviewed in Macnab, 2003). Similarly, the circumferential oligomerization of PrgH may occur in conjunction with PrgK or after completion of the PrgK ring. It is important to note, however, that there are differences between the basal structures of the flagellar system and the T3SS. Although PrgH is required for the stability o f the PrgK oligomer, the M S ring in the flagellar basal substructure is very 115 stable and the peripheral rings including the F l i G ring and the C-ring function exclusively in the rotation and switching processes. The well-defined tetramer observed in the crystallographic asymmetric unit was initially thought to be a crystallographic artifact since gel filtration analysis showed clearly that PrgH(l 70-362) is monomelic in solution. Previous studies by Kimbrough et al., however, showed that an apparently tetrameric oligomer o f full length PrgH could be isolated from bacterial membranes when PrgH was overexpressed in the absence of PrgK, indicating that there might be a molecular basis for the crystallographic tetramer (Kimbrough and Mil le r , 2000). Whether the PrgH tetramer has any real biological function, such as serving as a precursor for oligomerization, remains to be discovered. 116 CHAPTER 5 - Structural and biochemical characterization of EscC from enteropathogenic Escherichia coli 5.1 I N T R O D U C T I O N In addition to the substructure in the inner membrane, the T3SS contains a second ring complex localized to the outer membrane. The organization of this outer membrane complex is relatively simple as it consists of only one major protein component - the secretin. Although they bear neither sequence nor structural homology with the outer membrane components of the flagellar system, secretins belong to a large protein family whose members are widely dispersed across many species of Gram negative bacteria. Not only are they found in the T3SS, secretins participate in a variety of macromolecular transport processes including type II secretion, type IV pilus biogenesis, and filamentous phage release (Linderoth et al, 1997). E M studies on secretins of various secretory pathways, including Salmonella InvG and Yersinia Y s c C from their respective T3SSs, indicated that individual monomers associate into stable 12 to 14 subunit ring-like oligomers (Crago and Koronakis, 1998; Koster et al, 1997). The central channel, which has an estimated diameter of 5 to lOnm, is believed to be a large (3-barrel generated by |3-strands in the C-terminal region of individual subunit (Chami et al, 2005). Given its size, the secretin central channel likely functions as a portal for the export substrate to move across the outer membrane. The secretins in the T3SS, also known as the Y s c C family o f proteins, appear to play an additional role to anchor and stabilize the needle which propagates from the periplasmic space. Interestingly, the 17A c r y o E M structure o f the N C from S. typhimurium showed that the outer membrane channel generated by the secretin is closed by a "septum"-like 117 structure prior to assembly of the needle (Marlovits et al, 2004). Indeed, overexpressing Yersinia enterocolitica Y s c C in E. coli did not lead to increased permeability of the outer membrane, confirming that the secretin channel is likely closed in the absence of the needle (Burghout et al, 2004a). The recently reported 17A c ryo -EM structure of PulD, the canonical secretin from the T2SS of Klebsiella oxytoca, showed that a domain located at the N-terminal region of this protein may be directly involved in regulating the opening of the central channel by acting as a molecular "plug" (Chami et al, 2005). In spite o f the extensive E M studies on the secretin family, no high resolution structural data is available for any o f its members. Also , other than the C-terminal |3-rich region important for oligomerization, the organization and function of the different domains of this protein are not clearly understood. In this chapter, the biochemical and structural characteristics of EscC, a member of the secretin family from the E P E C T3SS, were examined. Ful l length EscC, devoid o f its signal sequence, was cloned, overexpressed, and purified. The purified protein was subjected to limited proteolysis, which allowed the identification of an N-terminal stable fragment. The N-terminal and C -terminal domains of EscC were subsequently cloned, overexpressed, and purified. Gel filtration was used to analyze potential interactions between these two fragments. The N -terminal fragment o f EscC was crystallized and its structure determined by X-ray crystallographic methods. Based on surface electrostatics analysis of this structure, several mutants were designed and examined by complementation and secretion assays. The original cloning and purification o f full length EscC were performed by Richard Pfuetzner, Elizabeth Frey, and Y u Luo. Complementation experiments and secretion assays were performed by N i k h i l Thomas in Dr. Brett Finlay's lab. Purification of EscC 118 fragments was assisted by Marija Vuckovic, and planar l ipid bilayer experiment was assisted by Manjeet Bains in Dr. Robert Hancock's lab. 119 5.2 M E T H O D S 5.2.1 Cloning, protein expression, and purification Expression constructs pET-EscC(22-512), pET-EscC(238-512), pET-EscC(22-174), which encode full length EscC, the C-terminal domain, and the N-terminal domain respectively (all devoid of the signal peptide), were generated by PCR-amplifying appropriate regions of escC from E P E C E2348/69 genomic D N A and cloning the amplified D N A into the Ndel/BamUl sites o f pET-28(a). E. coli BL21(1DE3) was transformed with these constructs, grown in L B containing 50 pg mL" 1 o f kanamycin at 37 °C to an O D 6 o o of 0.6 to 0.8, induced with 0 .5mM IPTG, and incubated overnight at 20 °C. Cells were harvested, resuspended in buffer (20mM H E P E S + 150mM N a C l , p H 6.8), lysed with a pressurized homogenizer, and centrifuged at 25,000 x g for 35 minutes. A l l three proteins were purified from the supernatant with cobalt-charged or nickel-charged chelating sepharose. EscC(22-512) was further purified by MonoS 5/5 and Superdex 200 HR10/30 columns, whereas EscC(238-512) and EscC(22-174) by MonoQ 5/5 and Superdex 75 HR10/30 columns. 5.2.2 Limited proteolysis 2.5 uL of purified EscC(22-512) at 8 mg mL"' was mixed with 5 uL of V 8 protease, trypsin, and thermolysin at various concentrations (1.0, 0.1, and 0.01, mg mL" 1 ) . These reaction mixtures were incubated at 22°C or on ice for 30 minutes, before being analyzed by S D S - P A G E . Gels were either stained with Coomassie blue or transferred to a P V D F membrane. Bands were excised from Coomassie blue-stained membrane for N -terminal sequencing at the Nucleic A c i d and Protein Service (N A PS) Unit of the University o f British Columbia. 120 5.2.3 Planar lipid bilayer experiment 1 M K C l solution was placed in a Teflon cell containing two compartments separated by a 0.1 -mm circular hole. The hole was covered with a lipid layer made from a solution of 1.5% oxidized cholesterol in «-decane. Electrodes were inserted into the K C l solutions in each compartment and a voltage of 50 m V was applied. Purified EscC at approximately 1 mg mL"1 was diluted in 0.1% Triton X-100, added to one o f the compartments, and conductance across the chambers was monitored. 5.2.4 Analytical gel filtration Purified EscC(22-174) (250ug) was mixed with purified EscC(238-512) (50u.g) and loaded onto a Superdex-75 HR10/30 column equilibrated with buffer (20 m M Tris + 150 m M N a C l , pH 6.8). The two proteins were also loaded individually onto the column in separate runs to determine the elution volume. 5.2.5 Mutagenesis, complementation, and secretion assays Full length escC and escC(l-174) were amplified from E P E C E2348/69 genomic D N A and cloned separately into the BamWVSall sites of p A C Y C - 1 8 4 , generating p A C Y C E s c C and p A C Y C E s c C ( l - 1 7 4 ) respectively. Single substitution mutants (K85A, K 8 5 D , E154A, E l 54K, E l 69A, and E169K) were constructed by the Quikchange method using p A C Y C E s c C as the template. The same method was also used to make the construct p A C Y C E s c C A(22-174), which contains a deletion of the region encoding residues 24 to 174. These plasmids were transformed into the E P E C AescC strain, which possesses a non-polar, in-frame deletion of the escC gene, to generate several different complemented strains. Secretion assays were performed on these complemented strains and wi ld type E P E C as previously described (Kenny and Finlay, 1995). Briefly, 40uL o f 121 overnight standing cultures were inoculated into 2mL of D M E M . After 6 hours of growth at 37 °C in a 5 % CO2 incubator, l m L of culture was harvested by centrifugation. The supernatant was filtered, precipitated with T C A , and analyzed by S D S - P A G E and gels were stained with Coomassie blue. 5.2.6 Crystallization of EscC(22-174) Crystallization trials were performed using the microbatch method by mixing 0.2 or 0.5 uL of protein solution (7 to 18 mg m L ' 1 ) with same volume of crystallization solution either manually or automatically with the Oryx-6 robot (Douglas Instruments). A layer of paraffin oil was overlaid on top of the wells holding the crystallization droplets, and the entire setup was incubated at 18 °C. Hexagonal-shaped crystals grew to maximum size in the condition (2.0M Ammonium sulfate + 0 .1M sodium acetate pH4.5, or Bis-Tris pH 6.5) in approximately 15 to 20 days. The.same type o f crystals could also be grown from SeMet-substituted proteins using essentially the same condition and method. 5.2.7 Data collection and structure determination EscC(22-174) crystals were harvested by soaking in mother liquor containing 35% glycerol, and then flash-frozen to 100 K . A highly redundant dataset was collected from a SeMet crystal near the peak wavelength at beamline 8.3.1 of the Advanced Light Source ( A L S ) . Data were processed with M O S F L M (Leslie, 1992) and scaled with S C A L A (Evans, 1993). The positions of one selenium and one disulfide were found and refined using S O L V E , and a high quality experimental map was obtained after density modification by R E S O L V E (Terwilliger, 2000; Terwilliger and Berendzen, 1999). About 95% of the model could be automatically built using A r p / w A R P (Morris et al, 2002), 122 and the missing regions were added manually using Xf i t (McRee, 1999). Refinement was performed using C N S (Brunger et al, 1998) and Refmac5 (Murshudov, 1997). Due to disorder, the refined model is missing the N-terminal 4 residues introduced from the vector, the C-terminal 2 residues, as well as Thr l28 and Ilel29. 5.2.8 Structural alignment and surface electrostatic analysis Structural alignment of EscC(22-174) with EscJ (PDB code: 1YJ7) was performed using the programs Coot (Emsley and Cowtan, 2004) and SwissPdb Viewer (Guex and Peitsch, 1997). Surface electrostatics of EscC(22-174) were calculated and analyzed using the A P B S plugin in Pymol (Baker et al, 2001). 123 5.3 R E S U L T S 5.3.1 E scC devoid of its signal peptide is soluble and monomeric EscC devoid of its signal peptide, EscC(22-512), was cloned and overexpressed. Recombinant EscC(22-512) could be expressed in a soluble form as long as the post-induction temperature was maintained at 20 °C or lower, and could be readily purified using nickel-charged chelating sepharose (Figure 5.1a). In agreement with its predicted pi of 8.2, this recombinant protein binds to a cation exchange resin (MonoS). Gel filtration and multiangle light scattering suggested that purified EscC(22-512), even at concentrations as high as 40mg mL" 1 , exists as a monomer in solution (Figure 5.1b). (a) (b) Volume (ml) (c) V8 Protease Trypsin Thermolysin ICE KT ICE RT ICE RT C 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 C Figure 5.1 EPEC EscC(22-512) exists as a monomer in solution. 124 (a) Purification of EPEC EscC(22-512). EscC(22-512) is relatively soluble when overexpressed in E. coli BL21 at low temperature. The his-tagged recombinant protein could be purified using nickel-charged chelating sepharose. (b) Multiangle light scattering analysis of EscC(22-512). The data presented is a molar mass versus volume plot overlaid with a gel filtration elution profile. The purified recombinant protein is monodisperse, and the calculated molecular mass of -58 kDa suggested that this protein is monomeric in solution, (c) Limited proteolysis of EscC(22-512). The purified protein (lane C, top arrow) was subjected to proteolysis by three different proteases (V8 protease, trypsin, and thermolysin) at three different dilutions: 1.0, 0.1, and 0.01 mg mL"1 (lanes 1,2,3 respectively). The various reactions were allowed to proceed on ice or at room temperature (RT) for 30 minutes before being analyzed by SDS-PAGE. A stable fragment of -18 to 20 kDa (bottom arrow) was obtained from both the trypsin and thermolysin digests. Planar lipid layer experiments were carried out on the purified protein to determine its ability to generate channels in a lipid layer. EscC(22-512) was diluted in buffer containing 0.1% Triton X - l 00 and added to one of the two chambers connected by a hole coated with oxidized cholesterol. Movement of ions across the l ipid layer was monitored by a voltmeter. N o conductance could be detected even when very high amounts o f purified protein was added (data not shown), indicating that EscC(22-512) could not readily integrate into an artificial l ipid layer on its own. 5.3.2 E s c C is a modular protein Secretins from the.T3SS are believed to consist o f discrete domains according to multiple sequence alignment and secondary structure prediction. To test this hypothesis, purified EscC(22-512) was subjected to proteolysis by three different proteases. A prominent fragment of approximately 18 to 20 kDa in size was obtained from the trypsin and thermolysin digestions (Figure 5.1c). N-terminal sequencing showed that this stable fragment possesses the same sequence as EscC(22-512), indicating that the N-terminal region is resistant to proteolytic cleavage. Further analysis by M A L D I - T O F mass spectrometry revealed that the proteolytic cleavage occurs at approximately residue 180, 125 which is close to one of the domain boundaries predicted for EscC. These data indicated the presence o f a stable domain within the N-terminal 180 residues of EscC. To determine i f this proteolytically-stable fragment could interact with other parts of EscC, two separate fragments o f EscC were designed based on the predicted domain boundaries (Figure 5.2a). A n N-terminal fragment corresponding to residue 22 to 174 of EscC, EscC(22-174), and a C-terminal fragment spanning residue 238 to 512, EscC(238-512), were cloned separately and overexpressed. Although not as highly expressed as EscC(22-512), these EscC fragments could be purified to homogeneity by a combination of nickel chelating sepharose, ion exchange, and gel filtration chromatography (Figure 5.2b and Figure 5.2d). Multiangle light scattering indicated that just as their full length counterpart, EscC(22-174) and EscC(238-512) exist as stable monomers in solution (Figure 5.2c and Figure 5.2e). When EscC(22-174) and EscC(238-512) were mixed and loaded onto a Superdex 200 HR10/30 column, two distinctive peaks corresponding precisely to the elution volume of each individual domain were obtained, indicating the two fragments do not interact (Figure 5.2f). 126 (a) signal T3SS-specific Secretin-N Secretin-N peptide domain domain domain Secretin domain Figure 5.2 EscC is a modular protein. (a) Schematic representation of EscC. Sequence analysis of EscC revealed the presence of multiple domains. The core secretin domain is located at the C-terminal region and is conserved 127 across all members of the secretin family. Two secretin-N domains, which are found primarily in secretins of the T2SS and T3SS, are located in the middle and N-terminal regions. The most N-terminal region of EscC contains the signal peptide, which is cleaved upon transport of the protein to the periplasm, and a T3SS-specific short domain. Based on results from limited proteolysis and sequence analysis, two different constructs spanning different regions of EscC (22 to 174, and 238 to 512) were designed, (b) Purification of EscC(22-174). EscC(22-174) is relatively soluble when overexpressed in E. coli BL21. The his-tagged recombinant protein could be purified using cobalt-chelating sepharose as shown, (c) EscC(22-174) is monomeric in solution. Multiangle light scattering analysis, which is represented by a molar mass versus volume plot overlaid with a gel filtration elution profile, showed that purified EscC(22-174) is monodisperse with a calculated molecular mass corresponding to the monomer size, (d) Purification of EscC(238-512). EscC(238-512) overexpressed at relatively low levels in E. coli BL21. The his-tagged recombinant protein, however, could be purified using cobalt-chelating sepharose, and concentrated on a cation exchange column as shown by SDS-PAGE. (e) EscC(238-512) is monomeric in solution. Multiangle light scattering analysis, which is represented by a molar mass versus volume plot overlaid with a gel filtration elution profile, showed that purified EscC(238-512) is monodisperse with a calculated molecular mass corresponding to the monomer size, (f) Analytical gel filtration analysis. Addition of EscC(238-512) did not change the elution profile of EscC(22-174), indicating the lack of interaction between the separate regions of EscC. 5.3.3 Crystal structure of EscC(22-174) Purified EscC(22-174) readily crystallized in several conditions, including a hexagonal form that diffracted beyond 2.5A. Although there is no inherent methionine between residue 22 and 174 of EscC, a methionine residue directly before residue 22 was introduced due to a cloning artifact and was retained in the purified protein. The high symmetry of the hexagonal spacegroup and the inherent low disorder of the crystal prepared from single-substituted SeMet protein, allowed the collection o f an accurate and highly redundant dataset (Table 5.1). The positions of the single N-terminal methionine and a sulfur cluster (disulfide bond) could be readily located from this dataset using the S A D method, and quite astonishingly, these were sufficient to provide initial phases to generate a relatively high quality electron density map (Figure 5.3a). Table 5.1 Data collection and structure refinement statistics for EscC(22-174). X-ray crystallographic data Dataset Spacegroup Unit cell (A) Wavelength (A) Resolution (A) Total reflections Unique reflections Completeness Redundancy1 <l/al>1 Rmerge (%)1'2 Crystal structure refinement R /R a ' 4 rework' rvree 0.215/0.251 Average B-factor 37.9 R.m.s. deviation Bond lengths (A) 0.032 Angles (°) 2.10 Ramachandran plot most favorable 88.8% allowed 11.2% disallowed 0.0% Values in parentheses correspond to the highest resolution shell. 2Rmerge= l^(lhki)-<l>|/^ (lhki), where l h w is the integrated intensity of a given reflection. 3Rwork=(I|F0-Fc|)/(IF0), where F 0 and Fcare observed and calculated structure factors 4Rfr e e was calculated from 10% of reflections excluded from refinement. SeMet P6522 90.6 x 90.6 x 133.4 0.980 2.05 436787 20991 99.9 (99.9) 20.9(19.6) 31.9 (5.9) 0.072 (0.501) 129 (a) (b) (d) Figure 5.3 Structure determination and architecture of EscC(22-174). (a) Experimental electron density map. The solvent-flattened map obtained from RESOLVE after SAD phasing from SOLVE is shown contoured at 1 .Oo here. Helical features could be readily distinguished in this map. (b) Ribbon representation of EscC(22-174). The monomer contains two structurally-distinct domains (I and II) connected by a short helix (a4). The N-terminus and C-130 terminus of the model are labeled "N" and "C" respectively, (c) Crystallographic dimer of EscC(22-174). Two molecules related by the crystallographic 2-fold axis appear to form a dimer. (d) Structural alignment between EscC(22-174) and EPEC EscJ (PDB code 1YJ7). Domain II of EscC(22-174) (in green) was overlapped with domain 1 of EscJ (in blue) using Coot. Both domains contain two a-helices positioned on top of a three-stranded B-sheet. EscC(22-174) crystallizes in the spacegroup P6s22 with one molecule in the asymmetric unit. The model, refined to a final resolution of 2.05A, showed that EscC(22-174) is compact and possesses two discrete and structurally-dissimilar domains (Figure 5.3b). More specifically, domain 1 (Ser22 to Lys l02) has an overall "wheel" shape, highlighted by two (3-sheets, one 2-stranded and one 3-stranded, sandwiching two a -helices. On the other hand, domain 2 (Ilel 16 to Serl72) contains a 3-stranded anti-parallel p-sheet overlaid by two a helices. Interestingly, this domain shared structural homologies with domain 1 of EscJ, with a Z score of 4.5 and r.m.s.d. of 2.5A according to the D A L I server (Holm and Sander, 1997) (Figure 5.3d). The two domains of EscC(22-174) are connected in tandem by a short helical linker region (Thrl03 to Glu l05) . Further examination of the molecular packing in the crystal suggested potential dimerization of this molecule (Figure 5.3c). In particular, two molecules, which are related to one another by the crystallographic two-fold axis, appear to form an intertwined dimer. The packing of these two molecules results in the burial of 11.7% or approximately 1868A 2 o f total solvent accessible surface. The two molecules are held together by hydrogen bonding interactions between the tip o f helix oi6 o f domain II and a mixture of hydrogen bonding and hydrophobic packing interactions at one side of domain I o f each monomer. 131 5.3.4 Surface electrostatic and preliminary mutation analyses Analysis of the solvent accessible surface revealed the presence of two charge patches located on opposite sides of the monomer. Notably, a positively charged surface (surface A ) is found near the mid section of the molecule extending from domain II to domain I (Figure 5.4a), while the negatively charged surface (surface B) surrounds the C -terminal tip of domain 2 (Figure 5.4b and Figure 5.4c). Further computer analysis suggested that mutating K85 in a.3 to alanine or an aspartic acid could eliminate the positive charge on surface A . In contrast, two mutations E154A and E154K could significantly reduce negative charges on one face while two other mutations E169A and / E169K on the other face o f surface B . To assess the effects of altering the surface electrostatic properties, Dr. N ikh i l Thomas in Dr. Finlay's lab carried out secretion assays on an escC-deficient E P E C strains complemented with plasmid-encoded escC harboring these mutations. Analysis of whole cell lysates prepared from these strains demonstrated that the introduced point mutations did not affect expression of EscC. Two mutations targeting surface B (E154K and E169K) abolished type III secretion, but mutation of the same residues to alanine had no apparently effects (Table 5.2). Intriguingly, in spite of the dramatic disruption o f the positive charge, mutations targeting surface A ( K 8 5 A and K85D) did not affect type III protein secretion to any extent (Table 5.2). 132 (a) (b) (c) Figure 5.4 Surface electrostatic analysis of EscC(22-174). (a) Positively charged patch of EscC(22-174). The positive charged patch localizes to the mid section of the molecule. Computer modeling based on the crystal structure suggested that Lys85 is primarily responsible for this charge, (b) Negatively charged patch of EscC(22-174). The surface electrostatics of EscC(22-174) is shown with the molecule rotated 180° from its orientation in (a). The charge near the mid section of the molecule is probably conferred by Glu154. (c) Second negatively charged patch of EscC(22-174). The surface electrostatics of EscC(22-174) is shown again with the molecule rotated approximately 90° from its orientation in (b) . The charge near the tip of the molecule is likely contributed by Glu169. Table 5.2 Effects of EscC mutation on type III secretion. Strain EscC expression1 Type III secretion wild type + + AescC/pEscC(1-174) + -AescC / pEscC(1 -512)A(22-174) + -AescC / pEscC(1-512)K85A + + AescC / pEscC( 1 -512)K85D + + AescC / pEscC(1-512)E154A + + AescC / pEscC(1-512)E154K + -AescC / pEscC(1 -512)E169A + + AescC / pEscC(1-512)E169K + -1 EscC expression as detected by polyclonal anti-EscC antisera in Western blotting. 133 5.4 DISCUSSION 5.4.1 Biogenesis of the EscC oligomer The purification of soluble EscC(22-512) allowed the characterization, in part, o f a member of the secretin family in its monomeric state for the first time. However, the self-association process essential to the function of secretin obviously could not be examined because the purified protein is unable to assemble a functional oligomer in vitro. The observation that EscC(22-512) remains monomeric even at relatively high protein concentrations and that it could not form conducting channels in an artificial membrane imply that additional assembly factors might be required for insertion into the outer membrane and oligomerization. These factors, are most likely to be present in the periplasm and the outer membrane and not accessible to EscC(22-512), which, despite being synthesized in its final processed form, lacks the signal peptide and accumulates in the bacterial cytoplasm. Indeed, the SDS-resistant oligomers observed previously for Salmonella InvG and Yersinia Y s c C have all been extracted from the bacterial outer membrane (Crago and Koronakis, 1998; Koster et ai, 1997). What could be the potential periplasmic or outer membrane assembly factors o f EscC? DsbA, a periplasmic disulfide oxidoreductase, has been shown to improve Y s c C stability and mediate its oligomerization by promoting disulfide formation near the C -terminal region (Jackson and Piano, 1999). This, however, does not seem to apply to EscC which lacks the corresponding C-terminal cysteines present in its Yersinia counterpart. Another potential assembly factor is the pilotin, a sequence-divergent family o f small lipoproteins localized to the outer membrane. Found in several T2SS and T3SS, these proteins directly bind to their cognate secretins and mediate their outer membrane 134 insertion and oligomerization processes (Burghout et al, 2004a; Crago and Koronakis, 1998; Schuch and Maurel l i , 2001). Intriguingly, EscC lacks the C-terminal extension which serves as the pilotin-binding domain for many T3SS secretins, raising the possibility that EscC may not actually possess a cognate pilotin. Omp85/YaeT and associated proteins is a recently discovered protein complex believed to function as a general assembly factor for outer membrane proteins in Gram negative bacteria (Voulhoux et al, 2003; W u et al, 2005). Although its molecular mechanisms of action remain poorly understood, this protein complex is required for insertion and/or multimerization of every outer membrane protein examined so far including the Neisseria menigitidis secretin P i l Q (Voulhoux et al, 2003; W u et al, 2005). It is conceivable to believe that outer membrane insertion of other secretins including EscC would require the Omp85/YaeT complex. Finally, recent work on both the E P E C T3SS and the Shigella T3SS suggested that other components o f the T3SS may be required for the biogenesis of the secretin oligomer. More specifically, it was observed that E P E C EscC accumulates in the periplasm and could not be properly targeted to the outer membrane in the absence of the export apparatus component EscV and the inner membrane-associated ATPase EscN (Gauthier et al, 2003). For the Shigella T3SS secretin M x i D , the inner membrane component M x i J (a member of the Y s c J family) is required for proper secretin insertion and oligomerization in addition to the cognate pilotin M x i M (Schuch and Maurelli , 2001). 5.4.2 Funct ion of the N- terminal domain of T3SS secretin Results from limited proteolysis and analytical gel filtration confirmed the modular nature of EscC that has been predicted for the T3SS secretin: a C-terminal domain constituting the core outer membrane structure and an N-terminal domain 135 projecting into the periplasmic space. Although the major function o f the secretin is to generate an outer membrane opening, mutagenesis data presented in this chapter showed that the N-terminal domain, which is not directly involved in constructing the outer membrane channel, is indispensable for type III secretion. What, then, is the role of this domain? Based on the refined E M structure of the T2SS secretin PulD, Pugsley and colleagues hypothesized that a part of the N-terminal region folds back to plug the channel generated by the C-terminal core secretin domain (Chami et al, 2005). Interestingly, sequence analysis identified a so called "secretin-N short domain" that is shared amongst secretins of the T2SS and T3SS. It is possible that the N-terminal domain o f EscC encodes a structurally-conserved element which mediates gating o f the secretin channel through similar mechanisms as its counterpart in the T2SS. The second putative function for the N-terminal domain is to connect and mediate cross talk between the outer membrane and the inner membrane structures of the T3SS. Both systematic yeast-2-hybrid analysis and GST-pul l down assays have shown that EscC binds to the putative inner membrane component EscD, a member of the Y s c D family (Creasey et al, 2003a; Ogino et al, 2006). Because of its periplasmic localization, the N-terminal domain is arguably the most logical part of EscC to engage in this important interaction. Thirdly, the N-terminal domain may be required for proper targeting of EscC to the outer membrane. Preliminary fractionation data from Dr. N ikh i l Thomas on the strain AescC I pEscC(l-512)A(22-174) showed that EscC missing residues 22 to 174 could not integrate into the outer membrane despite being properly transported to the periplasm. Finally, as revealed by the unexpected structural similarity between the crystal structure of EscC(22-136 174) and EscJ, the N-terminal domain may play a role in mediating EscC-EscC oligomerization upon integration into the outer membrane. 137 CHAPTER 6 - CONCLUSIONS AND FUTURE DIRECTIONS 6.1 S U M M A R Y A N D S I G N I F I C A N C E O F R E S U L T S The main goal of this thesis investigation was to improve the current. understanding of the molecular architecture of the T3SS. To this end, the structural and biochemical properties of four representative components from different parts of the E P E C and Salmonella T3SSs have been examined. Collectively, these data have shed new light into the T3SS assembly process as well as allow one to understand how at the molecular level these structural components assemble to mediate the protein-protein interactions necessary to construct the structural elements of this highly sophisticated macromolecular complex. The extracellular needle and needle extension of the T3SS play important roles in mediating protein translocation as well as host sensing. EspA is a needle extension protein of the E P E C T3SS that assembles into filaments believed to be the molecular conduit for protein translocation. The biochemical and structural characterization of this protein, both on its own and in complex with its chaperone CesA, provides new insights into the molecular basis of EspA polymerization as well as its regulation within the bacterial cytoplasm (Chapter 2). First o f all , EspA could spontaneously polymerize in the absence o f other E P E C proteins, a property that is l ikely shared with other extracellular components of the T3SS. Secondly, the crystal structure of EspA in complex with its specific chaperone CesA, together with deletion mutagenesis, showed that two discrete coiled coil motifs located at the termini of EspA are involved in filament formation. Similarities in organization of their secondary structures between EspA and the flagellar rod subunit F l i C implied that these coiled coils l ikely engage in an intramolecular 138 interaction to generate a module for packing against neighboring subunits within the filament. Lastly, the CesA chaperone protein prevents premature self-association of the inherently polymeric EspA in the E P E C cytoplasm by forming a stable heterodimeric CesA-EspA complex. This process is thermodynamically favorable and involves the displacement of one monomer from the dimeric CesA, which is transcribed and translated on an operon different from EspA. The inner membrane ring complex houses the export apparatus and together they function as a secretion system to deliver components such as EspA to locations beyond the inner membrane and the bacterial surface. The highly conserved YscJ family of T3SS proteins is believed to be one of the two major components involved in constructing the actual ring substructure. Results from biochemical and structural studies of E P E C EscJ have confirmed this hypothesis (Chapter 3). Notably, the finding that E P E C EscJ partitioned exclusively to the inner membrane fractions in a sucrose gradient definitively proved that the N-terminal l ipid anchor of YscJ proteins is positioned in the outer leaflet of the inner membrane with the core domain located peripheral to the inner membrane in the periplasm. The potential of YscJ proteins to form higher ordered oligomers was illustrated by the repetitive and extensive intersubunit packing observed in the EscJ crystal structure. Based on the crystal structure, a 24-subunit EscJ ring model was generated, and this model represents the first atomic snapshot of a major substructure of the T3SS. This dome-shaped ring, which was validated by results from stoichiometric and surface mapping analyses of Salmonella PrgK, possesses unique surface and electrostatic features indicative of a role as a platform for the docking of other structural components. 139 The second major component of the inner membrane ring complex is believed to be Salmonella PrgH and its distant orthologues. Biochemical characterization of PrgH (Chapter 4) confirmed the unique topology adopted by this class of proteins, by showing that overexpressed full length protein sequestered to the inner membrane and that the predicted cytoplasmic and periplasmic regions exist as stable domains. The crystal structure of the core periplasmic domain of PrgH, although not showing the repetitive intersubunit packing as observed in the EscJ crystal, provides a template for designing mutagenesis studies to elucidate the molecular basis of oligomerization as well as potential interaction with the PrgK oligomer. The outer membrane channel elaborated by the secretin (YscC) family of proteins represents the second major ring complex of the T3SS. This channel not only serves as a portal for the export substrate to cross the outer membrane barrier but also acts as an anchor to stabilize the extracellular needle of the T3SS. Biochemical analysis of the monomelic form of E P E C EscC illustrated the modular nature predicted for members of the secretin family (Chapter 5). The crystal structure of the N-terminal domain of EscC, on the other hand, represents a major first step towards achieving a high resolution structure of the full length secretin and provides a framework for understanding the connection between the outer membrane and inner membrane ring complexes of the T3SS. Although the four components examined in this thesis differ in localization and exhibit distinct biochemical and structural properties, several themes regarding the T3SS assembly have emerged. First of all, interactions governing intersubunit packing within the T3SS are quite diverse ranging from coiled coils to hydrogen bond-mediated helix-to-140 sheet interactions. Next, the components responsible for constructing the larger substructures are relatively compact in structure. Lastly, interaction between different subunits and/or different components is tightly regulated either by accessory proteins such as chaperones or by its own design in which protein-protein binding will only occur in the presence of a membrane or upon binding other components or substructures. 6.2 FUTURE DIRECTIONS In spite of the wealth of biochemical and structural data presented in this thesis, obviously much more still has to be learned about these four T3SS components. For EspA, there is an urgent need to clarify the role of coiled coils in self-association. Crystallization and structure determination of the putative pentameric/hexameric form of EspA generated by the C-terminal coiled coil deletions should enable one to understand the precise role of the N-terminal coiled coil segment in the EspA-EspA interaction. Recent advances in cryoEM techniques and computer algorithms have resulted in dramatic improvement in the quality of 3D reconstructions from E M images (Jiang and Ludtke, 2005). One could take advantage of these new technologies and try to obtain a higher quality E M structure of the EspA filament. With improved resolution, such structures would not only reveal intersubunit boundaries in the physiological polymer, but also provide a molecular envelope for potential docking of crystal structures of the limited oligomers of EspA. Knowledge on the regulation of EspA polymerization by CesA should be applied to understand the regulatory mechanisms of other extracellular T3SS components such as the needle subunit. Quinaud et al. have recently identified a heterodimeric chaperone in the Pseudomonas aeruginosa T3SS that binds the needle subunit to form a heterotrimeric complex (PscE-PscG-PscF) (Quinaud et al, 2005). 141 Initial bioinformatic and biochemical analysis suggested that an analogous chaperone complex, Orf2-Orf29, in the E P E C T3SS binds to the needle subunit EscF. Future experiments should focus on examining the molecular and structural details of this novel interaction. For the inner membrane ring component EscJ, additional work is needed to confirm its ability to form a circular oligomer when constrained in a planar environment. One possible approach is to generate unilamellar l ipid vesicles or liposomes containing nickel chelating lipids, which have specific affinity towards his-tagged proteins. B y adding purified N-terminal his-tagged protein to these vesicles, EscJ could be concentrated on a surface, and protein cross-linking, density gradient centrifugation, and electron microscopy could then be used to evaluate potential oligomerization. More experiments are absolutely needed to address the biological function of PrgH, the second major inner membrane ring component. The crystal structure of PrgH( 170-362) provides an excellent template for probing the function of the periplasmic domain of PrgH. Based on this structure, one could construct mutants of PrgH with altered surface electrostatics and evaluate them by complementation and well-established functional assays such as secretion profiling. The ability of the periplasmic domain of PrgH to form oligomers and interact with the periplasmic domain of PrgK on a surface could be assessed using l ipid vesicles as detailed above. Another approach to examine the interaction between PrgH and PrgK would involve purifying a PrgH-PrgK complex from whole needle complex preparations, a procedure recently developed by collaborators in Dr. Sam Mil le r ' s group. This purification method still needs to be further optimized to increase the yield and quality of the purified proteins. However, once a stable PrgH-PrgK complex becomes 142 available, protein cross-linking together with mass spectrometry should enable one to map out the interaction interface. While the periplasmic domain has received most of the attention so far, efforts should also be put into investigating the role o f the cytoplasmic domain. A n obvious first step would be to determine the structure of this small domain by X-ray crystallography or N M R . Structural analysis should also be performed on members o f the Y s c D family such as E P E C EscD using similar approaches as PrgH. It is anticipated that these structures would enable one to understand the unusual sequence divergence observed between the probably functionally-analogous PrgH and the Y s c D proteins. A s for the outer membrane secretin, the quest for a crystal structure of the full length protein should continue to be a priority. Although E P E C EscC(22-512) could be purified to homogeneity, attempts to crystallize this protein have been unsuccessful so far. A n approach that has become increasingly popular in the structural analysis o f membrane proteins is homologue screening. Secretins from T3SS as well as other secretory pathways could be cloned from E P E C and other Gram negative bacterial species, with the hope of identifying a candidate more amenable to crystallization. One major drawback of the current purification strategy of EscC(22-512) is that the recombinant protein is expressed in a non-native environment and the expected oligomerization does not occur. Indeed, the majority of outer membrane proteins used for crystallization and structural determination have been extracted from the outer membrane or refolded from inclusion bodies. Attempts should therefore be made to develop methods to express EscC and other secretins in their functional oligomeric forms in the bacterial outer membrane. A logical and simple first approach would be to fuse the coding region of EscC(22-512) to a PelB 143 J or OmpA signal sequence, which would enable the efficient transport of EscC to the periplasmic space. Density gradient centrifugation could then be used to determine i f the expressed protein is properly targeted to the outer membrane. In addition to the typical E. coli expression strain BL21(ADE3), one should also try to express the PelB or OmpA-fused EscC in its natural host E P E C by creating an expression construct driven by an arabinose promoter. This would allow EscC access potential species-specific assembly factors in the periplasm. In addition to the full length protein, further experiments should be carried out to characterize the function of the N-terminal domain. A first approach would be to investigate what causes the type III secretion defects observed for the point mutants E154K and E169K. Sucrose gradient or differential detergent extraction could be used to evaluate whether these mutants are properly processed and integrated into the outer membrane. With the purified recombinant EscC fragments, the domain requirement of interaction between EscC and other T3SS components such as EscD could be examined by proven methods such as GST-pul l down assays. Significant progress has been made on understanding the structural properties o f the T3SS since the beginning of this thesis investigation. In addition to the structural and biochemical analyses of several T3SS components including the four described here, intermediate resolution E M maps of the Salmonella N C and extracellular needle and filament have recently become available (Cordes et al, 2003; Daniell et al, 2003; Marlovits et al, 2004). While on-going studies w i l l continue to improve the knowledge on the core T3SS components and the substructures constructed by these proteins, there is an urgent need to integrate the wealth of available structural and biochemical data into a more unified picture. The E M map of the Salmonella N C at 17A determined by Marlovits 144 et al. could be used as a molecular envelope for docking the currently available high resolution crystal structures (Marlovits et al., 2004). Although the resolution of this map is somewhat limiting in terms of unambiguous localization of individual components and assignment o f secondary structural elements, techniques such as antibody-labeling could potentially aid in locating specific components within the large complex. Development of more advanced protein cross-linking and mass spectrometry-based methods to map out the underlying protein-protein interaction network within the N C would also greatly facilitate this process. 145 REFERENCES Agrain, C , Callebaut, I., Journet, L . , Sorg, I., Paroz, C , Mota, L . J., and Cornells, G . R. (2005). "Characterization of a Type III secretion substrate specificity switch (T3S4) domain in YscP from Yersinia enterocolitica". M o l Microbiol 56, 54-67. Akeda, Y . , and Galan, J. E . (2005). "Chaperone release and unfolding of substrates in type III secretion". Nature 437, 911-915. Allaoui , A . , Sansonetti, P. J., and Parsot, C. (1992). " M x i J , a lipoprotein involved in secretion o f Shigella Ipa invasins, is homologous to YscJ , a secretion factor of the Yersinia Y o p proteins". J Bacteriol 174, 7661-7669. Anderson, D . M . , and Schneewind, p . (1997). " A m R N A signal for the type III secretion of Y o p proteins by Yersinia enterocolitica". Science 278, 1140-1143. Atmakuri, K . , Cascales, E . , and Christie, P. J. (2004). "Energetic components V i r D 4 , V i r B l 1 and V i r B 4 mediate early D N A transfer reactions required for bacterial type IV secretion". M o l Microbiol 54. 1199-1211. Auvray, F., Thomas, J., Fraser, G . M . , and Hughes, C. (2001). "Flagellin polymerisation control by a cytosolic export chaperone". J M o l B i o l 308, 221-229. Baker, N . A . , Sept, D . , Joseph, S., Hoist, M . J., and McCammon, J. A . (2001). "Electrostatics of nanosystems: application to microtubules and the ribosome". Proc Natl Acad Sci U S A 98. 10037-10041. Barbieri, J. T., and Sun, J. (2004). "Pseudomonas aeruginosa ExoS and ExoT". Rev Physiol Biochem Pharmacol 152, 79-92. Binet, R., Letoffe, S., Ghigo, J. M . , Delepelaire, P., and Wandersman, C . (1997). "Protein secretion by Gram-negative bacterial A B C exporters—a review". Gene 192, 7-11. Birtalan, S., and Ghosh, P. (2001). "Structure of the Yersinia type III secretory system chaperone SycE". Nat Struct B io l 8, 974-978. Birtalan, S. C. , Phillips, R. M . , and Ghosh, P. (2002). "Three-dimensional secretion signals in chaperone-effector complexes of bacterial pathogens". M o l Cel l 9, 971-980. Blocker, A . , Gounon, P., Larquet, E . , Niebuhr, K . , Cabiaux, V . , Parsot, C , and Sansonetti, P. (1999). "The tripartite type III secreton of Shigella flexneri inserts IpaB and IpaC into host membranes". J Cel l B i o l 147, 683-693. Blocker, A . , Jouihri, N . , Larquet, E . , Gounon, P., Ebel, F., Parsot, C , Sansonetti, P., and Al laoui , A . (2001). "Structure and composition of the Shigella flexneri "needle complex", a part of its type III secreton". M o l Microbiol 39, 652-663. Bogdanove, A . J., Beer, S. V . , Bonas, U . , Boucher, C. A . , Collmer, A . , Coplin, D . L . , Cornelis, G . R., Huang, H . C , Hutcheson, S. W. , Panopoulos, N . J., and Van Gijsegem, F. (1996). "Unified nomenclature for broadly conserved hrp genes of phytopathogenic bacteria". M o l Microbiol 20, 681-683. Bricogne, G . , Vonrhein, C , Flensburg, C , Schiltz, M . , and Paciorek, W . (2003). "Generation, representation and flow of phase information in structure determination: recent developments in and around S H A R P 2.0". Acta Crystallogr D B i o l Crystallogr 59, 2023-2030. Brunger, A . T., Adams, P. D . , Clore, G . M . , DeLano, W . L . , Gros, P., Grosse-Kunstleve, R. W. , Jiang, J. S., Kuszewski, J., Nilges, M . , Pannu, N . S., et al. (1998). "Crystallography & N M R system: A new software suite for macromolecular structure determination". Acta Crystallogr D B io l Crystallogr 54 ( Pt 5), 905-921. 147 Burghout, P., Beckers, F., de Wit , E . , van Boxtel, R., Cornelis, G . R., Tommassen, J., and Koster, M . (2004a). "Role of the pilot protein Y s c W in the biogenesis of the Y s c C secretin in Yersinia enterocolitica". J Bacteriol 186, 5366-5375. Burghout, P., van Boxtel, R., Van Gelder, P., Ringler, P., Muller, S. A . , Tommassen, J., and Koster, M . (2004b). "Structure and electrophysiological properties of the Y s c C secretin from the type III secretion system of Yersinia enterocolitica". J Bacteriol 186, 4645-4654. Burr, S. E . , Stuber, K . , Wahl i , T., and Frey, J. (2002). "Evidence for a type III secretion system in Aeromonas salmonicida subsp. salmonicida". J Bacteriol 184, 5966-5970. Cao, T. B . , and Saier, M . H . , Jr. (2003). "The general protein secretory pathway: phylogenetic analyses leading to evolutionary conclusions". Biochim Biophys Acta 1609,115-125. Censini, S., Lange, C , Xiang, Z . , Crabtree, J. E. , Ghiara, P., Borodovsky, M . , Rappuoli, R., and Covacci, A . (1996). "cag, a pathogenicity island ofHelicobacter pylori, encodes type I-specific and disease-associated virulence factors". Proc Natl Acad Sci U S A 93, 14648-14653. Chakravortty, D . , Rohde, M . , Jager, L . , Deiwick, J., and Hensel, M . (2005). "Formation of a novel surface structure encoded by Salmonella Pathogenicity Island 2". Embo J 24, 2043-2052. Chami, M . , Guilvout, I., Gregorini, M . , Remigy, H . W. , Muller, S. A . , Valerio, M . , Engel, A . , Pugsley, A . P., and Bayan, N . (2005). "Structural insights into the secretin PulD and its trypsin-resistant core". J B i o l Chem 280, 37732-37741. 148 Christie, P. J. , Atmakuri, K . , Krishnarnoorthy, V . , Jakubowski, S., and Cascales, E . (2005). "Biogenesis, architecture, and function of bacterial type IV secretion systems". Annu Rev Microbiol 59, 451-485. Christie, P. J . , and Cascales, E . (2005). "Structural and dynamic properties of bacterial type IV secretion systems (review)". M o l Membr Biol 22, 51-61. Cianciotto, N . P. (2005). "Type II secretion: a protein secretion system for all seasons". 'Trends Microbiol 13, 581-588. Clarke, S. C , Haigh, R. D. , Freestone, P. P., and Williams, P. H . (2003). "Virulence of enteropathogenic Escherichia coli, a global pathogen". Clin Microbiol Rev 16, 365-378. Coers, J. , Kagan, J. C , Matthews, M . , Nagai, H . , Zuckman, D. M . , and Roy, C . R. (2000). "Identification of Icm protein complexes that play distinct roles in the biogenesis of an organelle permissive for Legionella pneumophila intracellular growth". M o l Microbiol 38.719-736. Collazo, C . M . , and Galan, J. E . (1997). "The invasion-associated type-Ill protein secretion system in Salmonella—a review". Gene 192, 51-59. Collaborative Computational Project, Number 4 (1994). "The C C P 4 Suite: programs for protein crystallography". Acta Crystallogr D 50, 760-763. Cordes, F . S., Komoriya, K . , Larquet, E . , Yang, S., Egelman, E . H . , Blocker, A . , and Lea, S. M . (2003). "Helical structure of the needle of the type III secretion system of Shigella flexneri". J Biol Chem 278. 17103-17107. Cornells, G . R. (2002a). "The Yersinia Ysc-Yop 'type IIP weaponry". Nat Rev M o l Cell Biol 3, 742-752. Cornells, G . R. (2002b). "The Yersinia Y s c - Y o p virulence apparatus". Int J Med Microbiol 291, 455-462. Cornells, G . R., and Van Gijsegem, F. (2000). "Assembly and function of type III secretory systems". Annu Rev Microbiol 54, 735-774. Cotter, S. E . , Surana, N . K . , and St Geme, J. W. , 3rd (2005). "Trimeric autotransporters: a distinct subfamily of autotransporter proteins". Trends Microbiol 13, 199-205. Crago, A . M . , and Koronakis, V . (1998). "Salmonella InvG forms a ring-like multimer that requires the InvH lipoprotein for outer membrane localization". M o l Microbiol 30, 47-56. Creasey, E . A . , Delahay, R. M . , Daniell, S. J., and Frankel, G . (2003a). "Yeast two-hybrid system survey of interactions between LEE-encoded proteins of enteropathogenic Escherichia coli". Microbiology 149, 2093-2106. Creasey, E . A . , Friedberg, D. , Shaw, R. K . , Umanski, T., Knutton, S., Rosenshine, I., and Frankel, G . (2003b). "CesAB is an enteropathogenic Escherichia coli chaperone for the type-Ill translocator proteins EspA and EspB". Microbiology 149, 3639-3647. Crepin, V . F., Prasannan, S., Shaw, R. K . , Wilson, R. K . , Creasey, E. , Abe, C. M . , Knutton, S., Frankel, G . , and Matthews, S. (2005a). "Structural and functional studies o f the enteropathogenic Escherichia coli type III needle complex protein EscJ". M o l Microbiol 55. 1658-1670. Crepin, V . F., Shaw, R., Knutton, S., and Frankel, G . (2005b). "Molecular basis of antigenic polymorphism of EspA filaments: development of a peptide display technology". J M o l B io l 350. 42-52. 150 Daefler, S., Guilvout, I., Hardie, K. R., Pugsley, A . P., and Russel, M . (1997). "The C-terminal domain of the secretin PulD contains the binding site for its cognate chaperone, PulS, and confers PulS dependence on pIVfl function". Mol Microbiol 24, 465-475. Dale, C , Young, S. A. , Haydon, D. T., and Welburn, S. C. (2001). "The insect endosymbiont Sodalis glossinidius utilizes a type III secretion system for cell invasion". Proc Natl Acad Sci U S A 98. 1883-1888. Daniell, S. J., Kocsis, E., Morris, E., Knutton, S., Booy, F. P., and Frankel, G. (2003). "3D structure of EspA filaments from enteropathogenic Escherichia coli". Mol Microbiol 49,301-308. de Keyzer, J., van der Does, C., and Driessen, A . J. (2003). "The bacterial translocase: a dynamic protein channel complex". Cell Mol Life Sci 60, 2034-2052. Delahay, R. M . , Knutton, S., Shaw, R. K., Hartland, E. L., Pallen, M . J., and Frankel, G. (1999). "The coiled-coil domain of EspA is essential for the assembly of the type III secretion translocon on the surface of enteropathogenic Escherichia coli". J Biol Chem 274, 35969-35974. Delrue, R. M . , Martinez-Lorenzo, M . , Lestrate, P., Danese, I., Bielarz, V . , Mertens, P., De Bolle, X . , Tibor, A. , Gorvel, J. P., and Letesson, J. J. (2001). "Identification of Brucella spp. genes involved in intracellular trafficking". Cell Microbiol 3, 487-497. d'Enfert, C , Ryter, A. , and Pugsley, A. P. (1987). "Cloning and expression in Escherichia coli of the Klebsiella pneumoniae genes for production, surface localization and secretion of the lipoprotein pullulanase". Embo J 6, 3531-3538. 151 Derewenda, Z . S. (2004). "Rational protein crystallization by mutational surface engineering". Structure (Camb) 12, 529-535. Desvaux, M . , Hebraud, ML, Henderson, I. R., and Pallen, M . J. (2006). "Type III secretion: what's in a name?" Trends Microbiol 14, 157-160. Donnenberg, M . S., Y u , J., and Kaper, J. B . (1993). " A second chromosomal gene necessary for intimate attachment of enteropathogenic Escherichia coli to epithelial cells". J Bacteriol 175, 4670-4680. Elliott, S. J., Wainwright, L . A . , McDaniel , T. K . , Jarvis, K . G . , Deng, Y . K . , La i , L . C. , McNamara, B . P., Donnenberg, M . S., and Kaper, J. B . (1998). "The complete sequence of the locus of enterocyte effacement ( L E E ) from enteropathogenic Escherichia coli E2348/69". M o l Microbiol 28, 1 -4. Emsley, P., and Cowtan, K . (2004). "Coot: model-building tools for molecular graphics". Acta Crystallogr D Bio l Crystallogr 60, 2126-2132. Evans, P. R. (1993). "Data reduction". Proceedings of C C P 4 study weekend on data collection and processing, 114-122. Evdokimov, A . G . , Phan, J., Tropea, J. E . , Routzahn, K . M . , Peters, H . K . , Pokross, M . , and Waugh, D . S. (2003). "Similar modes of polypeptide recognition by export chaperones in flagellar biosynthesis and type III secretion". Nat Struct B i o l 10, 789-793. Fadouloglou, V . E . , Tampakaki, A . P., Glykos, N . M . , Bastaki, M . N . , Hadden, J. M . , Phillips, S. E . , Panopoulos, N . J., and Kokkinidis , M . (2004). "Structure of H r c Q B - C , a conserved component of the bacterial type III secretion systems". Proc Natl Acad Sci U S A 101, 70-75. 152 Fan, F. , and Macnab, R. M . (1996). "Enzymatic characterization of F l i l . A n ATPase involved in flagellar assembly in Salmonella typhimurium". J B i o l Chem 271, 31981-31988. Farizo, K . M . , Huang, T., and Burns, D . L . (2000). "Importance o f holotoxin assembly in Ptl-mediated secretion of pertussis toxin from Bordetella pertussis". Infect Immun 68, 4049-4054. Feldman, M . F. , and Cornells, G . R. (2003). "The multitalented type III chaperones: all you can do with 15 kDa". F E M S Microbiol Lett 219, 151-158. Ffrench-Constant, R. H . , Waterfield, N . , Burland, V . , Perna, N . T., Daborn, P. J., Bowen, D . , and Blattner, F. R. (2000). " A genomic sample sequence of the entomopathogenic bacterium Photorhabdus luminescens W14: potential implications for virulence". A p p l Environ Microbiol 66, 3310-3329. Fil loux, A . (2004). "The underlying mechanisms of type II protein secretion". Biochim Biophys Acta 1694, 163-179. Francis, N . R., Irikura, V . M . , Yamaguchi, S., DeRosier, D . J., and Macnab, R. M . (1992). "Localization o f the Salmonella typhimurium flagellar switch protein F l i G to the cytoplasmic M-r ing face of the basal body". Proc Natl Acad Sci U S A 89, 6304-6308. Francis, N . R., Sosinsky, G . E . , Thomas, D . , and DeRosier, D . J. (1994). "Isolation, characterization and structure of bacterial flagellar motors containing the switch complex". J M o l B i o l 235. 1261-1270. Fu, Y . , and Galan, J. E . (1999). " A Salmonella protein antagonizes Rac-1 and Cdc42 to mediate host-cell recovery after bacterial invasion". Nature 401, 293-297. 153 Galan, J. E . , and Collmer, A . (1999). "Type III secretion machines: bacterial devices for protein delivery into host cells". Science 284, 1322-1328. Garza, A . G . , Harris-Haller, L . W. , Stoebner, R. A . , and Manson, M . D . (1995). "Motil i ty protein interactions in the bacterial flagellar motor". Proc Natl Acad Sci U S A 92, 1970-1974. Gauthier, A . , Puente, J. L . , and Finlay, B . B . (2003). "Secretin of the enteropathogenic Escherichia coli type III secretion system requires components of the type III apparatus for assembly and localization". Infect Immun 71, 3310-3319. Ghosh, P. (2004). "Process of protein transport by the type III secretion system". Microbiol M o l B i o l Rev 68, 771-795. Gruenheid, S., DeVinney, R., Bladt, F., Goosney, D . , Gelkop, S., Gish, G . D . , Pawson, T., and Finlay, B . B . (2001). "Enteropathogenic E. coli Ti r binds Nek to initiate actin pedestal formation in host cells". Nat Cel l B i o l 3, 856-859. Guex, N . , and Peitsch, M . C . (1997). " S W I S S - M O D E L and the Swiss-Pdb Viewer: A n environment for comparative protein modeling". Electrophoresis 18, 2714-2723. Halic, M . , and Beckmann, R. (2005). "The signal recognition particle and its interactions during protein targeting". Curr Opin Struct B i o l 15, 116-125. Hardie, K . R., Lory, S., and Pugsley, A . P. (1996). "Insertion of an outer membrane protein in Escherichia coli requires a chaperone-like protein". Embo J 15, 978-988. Hardt, W . D. , Chen, L . M . , Schuebel, K . E . , Bustelo, X . R., and Galan, J. E . (1998). "S. typhimurium encodes an activator of Rho GTPases that induces membrane ruffling and nuclear responses in host cells". Cel l 93, 815-826. 154 Hartland, E. L. , Daniell , S. J . , Delahay, R. M . , Neves, B . C , Wallis , T., Shaw, R. K. , Hale, C , Knutton, S., and Frankel, G . (2000). "The type III protein translocation system of enteropathogenic Escherichia coli involves EspA-EspB protein interactions". M o l Microbiol 35. 1483-1492. Hayward, R. D. , and Koronakis, V . (1999). "Direct nucleation and bundling of actin by the SipC protein of invasive Salmonella". Embo J 18, 4926-4934. He, S. Y . , and Jin, Q. (2003). "The Hrp pilus: learning from flagella". Curr Opin Microbiol 6. 15-19. Henderson, I. R., Navarro-Garcia, F., Desvaux, M . , Fernandez, R. C , and Ala'Aldeen, D. (2004). "Type V protein secretion pathway: the autotransporter story". Microbiol M o l B i o l Rev 68, 692-744. Hirano, T., Yamaguchi, S., Oosawa, K. , and Aizawa, S. (1994). "Roles of F l i K and F l h B in determination of flagellar hook length in Salmonella typhimurium". J Bacteriol 176, 5439.5449. H o l m , L. , and Sander, C. (1996). "The FSSP database: fold classification based on structure-structure alignment of proteins". Nucleic Acids Res 24, 206-209. H o l m , L., and Sander, C. (1997). "Dali/FSSP classification of three-dimensional protein folds". Nucleic Acids Res 25, 231 -234. H o l m , L. , and Sander, C. (1998). "Touring protein fold space with Dali/FSSP". Nucleic Acids Res 26.316-319. Homma, M . , DeRosier, D. J . , and Macnab, R. M . (1990). "Flagellar hook and hook-associated proteins of Salmonella typhimurium and their relationship to other axial components of the flagellum". J M o l B i o l 213, 819-832. 155 Ide, T., Laarmann, S., Greune, L . , Schillers, H . , Oberleithner, H . , and Schmidt, M . A . (2001). "Characterization o f translocation pores inserted into plasma membranes by type Ill-secreted Esp proteins of enteropathogenic Escherichia coli". Ce l l Microbiol 3, 669-679. Jackson, M . W. , and Piano, G . V . (1999). "DsbA is required for stable expression o f outer membrane protein Y s c C and for efficient Y o p secretion in Yersinia pestis". J Bacteriol 181, 5126-5130. Jacob-Dubuisson, F. , Locht, C , and Antoine, R. (2001). "Two-partner secretion in Gram-negative bacteria: a thrifty, specific pathway for large virulence proteins". M o l Microbiol 40. 306-313. Jiang, W. , and Ludtke, S. J. (2005). "Electron cryomicroscopy of single particles at subnanometer resolution". Curr Opin Struct B i o l 15, 571-577. Johnson, T. L . , Abendroth, J., Ho i , W . G. , and Sandkvist, M . (2006). "Type II secretion: from structure to function". F E M S Microbiol Lett 255, 175-186. Jones, C . J., Homma, M . , and Macnab, R. M . (1987). "Identification of proteins of the outer (L and P) rings of the flagellar basal body of Escherichia coli". J Bacteriol 169, 1489-1492. Jones, C. J., and Macnab, R. M . (1990). "Flagellar assembly in Salmonella typhimurium: analysis with temperature-sensitive mutants". J Bacteriol 172, 1327-1339. Journet, L . , Agrain, C , Broz, P., and Cornelis, G . R. (2003). "The needle length of bacterial injectisomes is determined by a molecular ruler". Science 302, 1757-1760. 156 Judd, P. K . , Kumar, R. B . , and Das, A . (2005). "Spatial location and requirements for the assembly o f the Agrobacterium tumefaciens type IV secretion apparatus". Proc Natl Acad Sci U S A 102, 11498-11503. Kabsch, W . (1976). Acta Crystallogr A 32, 922-923. Kenny, B . , DeVinney, R., Stein, M . , Reinscheid, D . J., Frey, E . A . , and Finlay, B . B . (1997). "Enteropathogenic E. coli (EPEC) transfers its receptor for intimate adherence into mammalian cells". Cel l 91, 511-520. Kenny, B . , and Finlay, B . B . (1995). "Protein secretion by enteropathogenic Escherichia coli is essential for transducing signals to epithelial cells". Proc Natl Acad Sci U S A 92,7991-7995. Kenny, B . , and Warawa, J. (2001). "Enteropathogenic Escherichia coli (EPEC) Tir receptor molecule does not undergo full modification when introduced into host cells by EPEC-independent mechanisms". Infect Immun 69, 1444-1453. Kimbrough, T. G . (2002) Contribution of the prgHKJK operon to the assembly of the SPI1 TTSS of Salmonella typhimurium, University of Washington, Seattle. Kimbrough, T. G . , and Mil le r , S. I. (2000). "Contribution of Salmonella typhimurium type III secretion components to needle complex formation". Proc Natl Acad Sci U S A 97, 11008-11013. Knutton, S., Rosenshine, I., Pallen, M . J., Nisan, I., Neves, B . C , Bain, C , Wolff, C , Dougan, G . , and Frankel, G . (1998). " A novel EspA-associated surface organelle of enteropathogenic Escherichia coli involved in protein translocation into epithelial cells". Embo J 17, 2166-2176. 157 Kohler, R., Schafer, K . , Muller, S., Vignon, G . , Diederichs, K . , Philippsen, A . , Ringler, P., Pugsley, A . P., Engel, A . , and Welte, W . (2004). "Structure and assembly o f the pseudopilin PulG" . M o l Microbiol 54, 647-664. Koster, M . , Bitter, W. , de Cock, H.., Al laoui , A . , Cornells, G . R., and Tommassen, J. (1997). "The outer membrane component, Y s c C , o f the Y o p secretion machinery of Yersinia enterocolitica forms a ring-shaped multimeric complex". M o l Microbiol 26, 789-797. Krogh, A . , Larsson, B . , von Heijne, G . , and Sonnhammer, E . L . (2001). "Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes". J M o l B io l 305, 567-580. Kubori , T., Matsushima, Y . , Nakamura, D . , Ural i l , J., Lara-Tejero, M . , Sukhan, A . , Galan, J. E. , and Aizawa, S. I. (1998). "Supramolecular structure of the Salmonella typhimurium type III protein secretion system". Science 280, 602-605. Kubori , T., Shimamoto, N . , Yamaguchi, S., Namba, K . , and Aizawa, S. (1992). "Morphological pathway of flagellar assembly in Salmonella typhimurium". J M o l B io l 226, 433-446. Kubori , T., Sukhan, A . , Aizawa, S. I., and Galan, J. E . (2000). "Molecular characterization and assembly of the needle complex of the Salmonella typhimurium type III protein secretion system". Proc Natl Acad Sci U S A 97, 10225-10230. Kuehn, M . J., Normark, S., and Hultgren, S. J. (1991). "Immunoglobulin-like PapD chaperone caps and uncaps interactive surfaces of nascently translocated pilus subunits". Proc Natl Acad Sci U S A 88, 10586-10590. 158 La i , L . C , Wainwright, L . A . , Stone, K . D . , and Donnenberg, M . S. (1997). " A third secreted protein that is encoded by the enteropathogenic Escherichia coli -pathogenicity island is required for transduction of signals and for attaching and effacing activities in host cells". Infect Immun 65, 2211-2217. Leslie, A . G . W . (1992). "Recent changes to the M O S F L M package for processing film and image plate data". Joint C C P 4 + E S F - E A M C B Newsletter on Protein Crystallography 26. Linderoth, N . A . , Simon, M . N . , and Russel, M . (1997). "The filamentous phage pIV multimer visualized by scanning transmission electron microscopy". Science 278, 1635-1638. Lloyd , S. A . , Sjostrom, M , Andersson, S., and Wolf-Watz, H . (2002). "Molecular characterization of type III secretion signals via analysis of synthetic N-terminal amino acid sequences". M o l Microbiol 43, 51-59. Lloyd , S. A . , Tang, FL, Wang, X . , Bil l ings, S., and Blair , D . F. (1996). "Torque generation in the flagellar motor of Escherichia coli: evidence of a direct role for F l i G but not for F l i M or F l i N " . J Bacteriol 178, 223-231. Luirink, J., and Sinning, I. (2004). "SRP-mediated protein targeting: structure and function revisited". Biochim Biophys Acta 1694, 17-35. Luo, Y . , Bertero, M . G . , Frey, E . A . , Pfuetzner, R. A . , Wenk, M . R., Creagh, L . , Marcus, S. L . , L i m , D . , Sicheri, F. , Kay, C , et al. (2001). "Structural and biochemical characterization of the type III secretion chaperones CesT and SigE". Nat Struct B i o l 8, 1031-1036. 159 Lupas, A . , Van Dyke, M . , and Stock, J. (1991). "Predicting coiled coils from protein sequences". Science 252, 1162-1164. Macnab, R. M . (2003). "How bacteria assemble flagella". Annu Rev Microbiol 57, 77-100. Marie, C , Brought on, W . J., and Deakin, W . J. (2001). "Rhizobium type III secretion systems: legume charmers or alarmers?" Curr Opin Plant B io l 4, 336-342. Marlovits, T. C , Kubori , T., Sukhan, A . , Thomas, D . R., Galan, J. E . , and Unger, V . M . (2004). "Structural insights into the assembly of the type III secretion needle complex". Science 306, 1040-1042. Matlack, K . E . , Misselwitz, B . , Plath, K . , and Rapoport, T. A . (1999). "BiP acts as a molecular ratchet during posttranslational transport of prepro-alpha factor across the E R membrane". Ce l l 97, 553-564. M c C o y , A . J., Grosse-Kunstleve, R. W. , Storoni, L . C , and Read, R. J. (2005). "Likelihood-enhanced fast translation functions". Acta Crystallogr D B i o l Crystallogr 61, 458-464. McDanie l , T. K . , Jarvis, K . G. , Donnenberg, M . S., and Kaper, J. B . (1995). " A genetic locus o f enterocyte effacement conserved among diverse enterobacterial pathogens". Proc Natl Acad Sci U S A 92, 1664-1668. McDanie l , T. K . , and Kaper, J. B . (1997). " A cloned pathogenicity island from enteropathogenic Escherichia coli confers the attaching and effacing phenotype on E . coli K-12" . M o l Microbiol 23, 399-407. McRee, D . E . (1999). "XtalView/Xfi t—A versatile program for manipulating atomic coordinates and electron density". J Struct B io l 125, 156-165. 160 Menard, R., Sansonetti, P., and Parsot, C. (1994). "The secretion of the Shigella flexneri Ipa invasins is activated by epithelial cells and controlled by IpaB and IpaD". Embo J 13, 5293-5302. Minamino, T., Chu, R., Yamaguchi, S., and Macnab, R. M . (2000a). "Role o f F l i J in flagellar protein export in Salmonella". J Bacteriol 182, 4207-4215. Minamino, T., and Macnab, R. M . (1999). "Components of the Salmonella flagellar export apparatus and classification of export substrates". J Bacteriol 181, 1388-1394. Minamino, T., and MacNab, R. M . (2000). " F l i H , a soluble component of the type III flagellar export apparatus of Salmonella, forms a complex with F l i l and inhibits its ATPase activity". M o l Microbiol 37, 1494-1503. Minamino, T., and Pugsley, A . P. (2005). "Measure for measure in the control of type III secretion hook and needle length". M o l Microbiol 56, 303-308. Minamino, T., Yamaguchi, S., and Macnab, R. M . (2000b). "Interaction between F l i E and F lgB , a proximal rod component of the flagellar basal body of Salmonella". J Bacteriol 182, 3029-3036. Morris , R. J., Perrakis, A . , and Lamzin, V . S. (2002). " A R P / w A R P ' s model-building algorithms. I. The main chain". Acta Crystallogr D Bio l Crystallogr 58, 968-975. Mueller, C . A . , Broz, P., Muller , S. A . , Ringler, P., Erne-Brand, F. , Sorg, I., Kuhn, M . , Engel, A . , and Cornells, G . R. (2005). "The V-antigen of Yersinia forms a distinct structure at the tip of injectisome needles". Science 310, 674-676. Murshudov, G . N . (1997). "Refinement of macromolecular structures by the maximum-likelihood method". Acta Crystallogr D B i o l Crystallogr 53, 240-255. Mushegian, A . R., Fullner, K . J. , Koonin, E . V . , and Nester, E . W . (1996). " A family of lysozyme-like virulence factors in bacterial pathogens of plants and animals". Proc Natl Acad Sci U S A 93, 7321-7326. Nagai, H . , and Roy, C. R. (2001). "The DotA protein from Legionella pneumophila is secreted by a novel process that requires the Dot/Icm transporter". Embo J 20, 5962-5970. Nicholls, A - , Sharp, K . A . , and Honig, B . (1991). "Protein folding and association: insights from the interfacial and thermodynamic properties of hydrocarbons". Proteins 11,281-296. Nouwen, N . , Ranson, N . , Saibil , H . , Wolpensinger, B . , Engel, A . , Ghazi, A . , and Pugsley, A . P. (1999). "Secretin PulD: association with pilot PulS, structure, and ion-conducting channel formation". Proc Natl Acad Sci U S A 96, 8173-8177. Nunn, D . (1999). "Bacterial type II protein export and pilus biogenesis: more than just homologies?" Trends Cel l B i o l 9, 402-408. Odenbreit, S., Puis, J., Sedlmaier, B . , Gerland, E . , Fischer, W . , and Haas, R. (2000). "Translocation of Helicobacter pylori CagA into gastric epithelial cells by type IV secretion". Science 287, 1497-1500. Ogino, T., Ohno, R., Sekiya, K . , Kuwae, A . , Matsuzawa, T., Nonaka, T., Fukuda, H . , Imajoh-Ohmi, S., and Abe, A . (2006). "Assembly of the type III secretion apparatus o f enteropathogenic Escherichia coli". J Bacteriol 188, 2801-2811. Ohnishi, K . , Ohto, Y . , Aizawa, S., Macnab, R. M . , and lino, T. (1994). "FlgD is a scaffolding protein needed for flagellar hook assembly in Salmonella typhimurium". J Bacteriol 176, 2272-2281. 162 Oomen, C . J., van Ulsen, P., van Gelder, P., Feijen, M . , Tommassen, J., and Gros, P. (2004). "Structure of the translocator domain of a bacterial autotransporter". Embo J 23, 1257-1266. Osborne, A . R., Clemons, W . M . , Jr., and Rapoport, T. A . (2004). "A large conformational change of the translocation ATPase SecA". Proc Natl Acad Sci U S A 101, 10937-10942. Osborne, A . R., Rapoport, T. A . , and van den Berg, B . (2005). "Protein translocation by the Sec61/SecY channel". Annu Rev Cell Dev Biol 21, 529-550. O'Shea, E . K . , Klemm, J. D. , K i m , P. S., and Alber, T. (1991). "X-ray structure of the G C N 4 leucine zipper, a two-stranded, parallel coiled coil". Science 254, 539-544. e Ozin, A . J., Claret, L . , Auvray, F. , and Hughes, C . (2003). "The FliS chaperone selectively binds the disordered flagellin C-terminal DO domain central to polymerisation". F E M S Microbiol Lett 219, 219-224. Pallen, M . , Chaudhuri, R., and Khan, A . (2002). "Bacterial F H A domains: neglected players in the phospho-threonine signalling game?" Trends Microbiol 10, 556-563. Pallen, M . J. , Beatson, S. A . , and Bailey, C . M . (2005a). "Bioinformatics analysis of the locus for enterocyte effacement provides novel insights into type-Ill secretion". B M C Microbiol 5, 9. Pallen, M . J. , Beatson, S. A . , and Bailey, C . M . (2005b). "Bioinformatics, genomics and evolution of non-flagellar type-Ill secretion systems: a Darwinian perspective". F E M S Microbiol Rev 29, 201-229. Pallen, M . J. , Chaudhuri, R. R., and Henderson, I. R. (2003). "Genomic analysis of secretion systems". Curr Opin Microbiol 6, 519-527. Park, K . S., Ono, T., Rokuda, M . , Jang, M . H . , Okada, K . , Iida, T., and Honda, T. (2004). "Functional characterization of two type III secretion systems of Vibrio parahaemolyticus". Infect Immun 72, 6659-6665. Parsot, C , Hamiaux, C , and Page, A . L . (2003). "The various and varying roles of specific chaperones in type III secretion systems". Curr Opin Microbiol 6, 7-14. Persson, C , Carballeira, N . , Wolf-Watz, H . , and Fallman, M . (1997). "The PTPase Y o p H inhibits uptake of Yersinia, tyrosine phosphorylation of p l30Cas and F A K , and the associated accumulation of these proteins in peripheral focal adhesions". Embo J 16, 2307-2318. Plath, K . , and Rapoport, T. A . (2000). "Spontaneous release of cytosolic proteins from posttranslational substrates before their transport into the endoplasmic reticulum". J Cel l B io l 151, 167-178. Pohlner, J., Halter, R., Beyreuther, K . , and Meyer, T. F. (1987). "Gene structure and extracellular secretion of Neisseria gonorrhoeae IgA protease". Nature 325, 458-462. Pozidis, C , Chalkiadaki, A . , Gomez-Serrano, A . , Stahlberg, H . , Brown, I., Tampakaki, A . P., Lustig, A . , Sianidis, G . , Politou, A . S., Engel, A . , et al. (2003). "Type III protein translocase: H r c N is a peripheral ATPase that is activated by oligomerization". J B i o l Chem 278, 25816-25824. Pukatzki, S., M a , A . T., Sturtevant, D. , Krastins, B . , Sarracino, D . , Nelson, W . C , Heidelberg, J. F. , and Mekalanos, J. J. (2006). "Identification o f a conserved bacterial protein secretion system in Vibrio cholerae using the Dictyostelium host model system". Proc Natl Acad Sci U S A 103, 1528-1533. Quinaud, M . , Chabert, J . , Faudry, E . , Neumann, E . , Lemaire, D . , Pastor, A . , Elsen, S., Dessen, A . , and Attree, I. (2005). "The PscE-PscF-PscG complex controls type III secretion needle biogenesis in Pseudomonas aeruginosa". J B i o l Chem 280, 36293-36300. Roe, A . J . , Y u l l , FL, Naylor, S. W. , Woodward, M . J . , Smith, D . G . , and Gaily, D . L . (2003). "Heterogeneous surface expression of EspA translocon filaments by Escherichia coli 0157:H7 is controlled at the posttranscriptional level". Infect Immun 71, 5900-5909. Rosell, F. I., Mauk, M . R., and Mauk, A . G . (2005). "pH- and metal ion-linked stability of the hemopexin-heme complex". Biochemistry 44, 1872-1879. Rosqvist, R., Magnusson, K . E. , and Wolf-Watz, H . (1994). "Target cell contact triggers expression and polarized transfer of Yersinia Y o p E cytotoxin into mammalian cells". Embo J 13, 964-972. Russmann, H . , Kubori , T., Sauer, J . , and Galan, J. E . (2002). "Molecular and functional analysis of the type III secretion signal of the Salmonella enterica InvJ protein". M o l Microbiol 46, 769-779. Saijo-Hamano, Y . , Uchida, N . , Namba, K . , and Oosawa, K . (2004). "In vitro characterization of F lgB, F lgC, FlgF, F lgG , and F l i E , flagellar basal body proteins of Salmonella". J M o l B io l 339, 423-435. Samatey, F. A . , Imada, K . , Nagashima, S., Vonderviszt, F. , Kumasaka, T., Yamamoto, M . , and Namba, K . (2001). "Structure of the bacterial flagellar protofilament and implications for a switch for supercoiling". Nature 410, 331-337. 165 Samatey, F . A . , Matsunami, FL, Imada, K . , Nagashima, S., Shaikh, T. R., Thomas, D . R., Chen, J. Z . , Derosier, D . J., Kitao, A . , and Namba, K . (2004). "Structure o f the bacterial flagellar hook and implication for the molecular universal joint mechanism". Nature 431, 1062-1068. Sauer, F . G . , Pinkner, J. S., Waksman, G . , and Hultgren, S. J. (2002). "Chaperone priming of pilus subunits facilitates a topological transition that drives fiber formation". Cel l 111, 543-551. Sawaya, M . R., Guo, S., Tabor, S., Richardson, C . C , and Ellenberger, T. (1999). "Crystal structure of the helicase domain from the replicative helicase-primase of bacteriophage T7". Ce l l 99, 167-177. Schmid, M . C , Schulein, R., Dehio, M . , Denecker, G . , Carena, I., and Dehio, C. (2004); "The V i r B type IV secretion system of Bartonella henselae mediates invasion, proinflammatory activation and antiapoptotic protection of endothelial cells". M o l Microbiol 52, 81-92. Schmidt-Eisenlohr, FL, Domke, N . , Angerer, C , Wanner, G . , Zambryski, P. C , and Baron, C . (1999). " V i r proteins stabilize V i r B 5 and mediate its association with the T pilus of Agrobacterium tumefaciens". J Bacteriol 181, 7485-7492. Schneider, T. R., and Sheldrick, G . M . (2002). "Substructure solution with S H E L X D " . Acta Crystallogr D B i o l Crystallogr 58, 1772-1779. Schoenhals, G . J., and Macnab, R. M . (1996). "Physiological and biochemical analyses of F l g H , a lipoprotein forming the outer membrane L ring of the flagellar basal body of Salmonella typhimurium". J Bacteriol 178, 4200-4207. 166 Schuch, R., and Maurell i , A . T. (2001). " M x i M and M x i J , base elements of the Mxi -Spa type 111 secretion system of Shigella, interact with and stabilize the M x i D secretin in the cell envelope". J Bacteriol 183, 6991-6998. Sekiya, K . , Ohishi, M . , Ogino, T., Tamano, K . , Sasakawa, C , and Abe, A . (2001). "Supermolecular structure of the enteropathogenic Escherichia coli type III secretion system and its direct interaction with the EspA-sheath-like structure". Proc Natl Acad S c i U S A 98, 11638-11643. Shao, F., and Dixon, J. E . (2003). "YopT is a cysteine protease cleaving Rho family GTPases". A d v Exp M e d B i o l 529, 79-84. Shea, J. E . , Hensel, M . , Gleeson, C. , and Holden, D . W . (1996). "Identification o f a virulence locus encoding a second type III secretion system in Salmonella typhimurium". Proc Natl Acad Sci U S A 93, 2593-2597. Singleton, M . R., Sawaya, M . R., Ellenberger, T., and Wigley, D . B . (2000). "Crystal structure o f T7 gene 4 ring helicase indicates a mechanism for sequential hydrolysis of nucleotides". Ce l l 101, 589-600. Sory, M . P., and Cornells, G . R. (1994). "Translocation o f a hybrid YopE-adenylate cyclase from Yersinia enterocolitica into HeLa cells". M o l Microbiol 14, 583-594. Spudich, G . M . , Fernandez, D . , Zhou, X . R., and Christie, P. J. (1996). "Intermolecular disulfide bonds stabilize V i r B 7 homodimers and V i r B 7 / V i r B 9 heterodimers during biogenesis of the Agrobacterium tumefaciens T-complex transport apparatus". Proc Natl Acad Sci U S A 93, 7512-7517. 167 Stebbins, C. E . , and Galan, J. E . (2000). "Modulation of host signaling by a bacterial mimic: structure of the Salmonella effector SptP bound to R a c l " . M o l Ce l l 6, 1449-1460. Stebbins, C. E . , and Galan, J . 'E. (2001). "Maintenance of an unfolded polypeptide by a cognate chaperone in bacterial type III secretion". Nature 414, 77-81. Stephenson, K . (2005). "Sec-dependent protein translocation across biological membranes: evolutionary conservation of an essential protein transport pathway (review)". M o l Membr B i o l 22, 17-28. Sukhan, A . , Kubori , T., and Galan, J. E . (2003). "Synthesis and localization of the Salmonella SPI-1 type III secretion needle complex proteins PrgI and PrgJ". J Bacteriol 185. 3480-3483. Sukhan, A . , Kubori , T., Wilson, J., and Galan, J. E . (2001). "Genetic analysis of assembly of the Salmonella enterica serovar Typhimurium type III secretion-associated needle complex". J Bacteriol 183. 1159-1167. Suzuki, H . , Yonekura, K . , Murata, K . , Hira i , T., Oosawa, K., and Namba, K . (1998). " A structural feature in the central channel of the bacterial flagellar F l iF ring complex is implicated in type III protein export". J Struct B i o l 124, 104-114. Suzuki, H . , Yonekura, K . , and Namba, K . (2004). "Structure o f the rotor of the bacterial flagellar motor revealed by electron cryomicroscopy and single-particle image analysis". J M o l B i o l 337, 105-113. Tamano, K . , Aizawa, S., and Sasakawa, C. (2002a). "Purification and detection of Shigella type III secretion needle complex". Methods Enzymol 358, 385-392. Tamano, K . , Katayama, E . , Toyotome, T., and Sasakawa, C. (2002b). "Shigella Spa32 is an essential secretory protein for functional type III secretion machinery and uniformity of its needle length". J Bacteriol 184, 1244-1252. Tardy, F. , Homble, F., Neyt, C , Wattiez, R., Cornells, G . R., Ruysschaert, J. M . , and Cabiaux, V . (1999). "Yersinia enterocolitica type III secretion-translocation system: channel formation by secreted Yops". Embo J 18, 6793-6799. Terwilliger, T. C. (2000). "Maximum-likelihood density modification". Acta Crystallogr D B i o l Crystallogr 56 ( Pt 8). 965-972. Terwilliger, T. C , and Berendzen, J. (1999). "Automated M A D and M I R structure solution". Acta Crystallogr D B i o l Crystallogr 55 ( Pt 4), 849-861. Thanassi, D . G. , and Hultgren, S. J. (2000a). "Assembly of complex organelles: pilus biogenesis in gram-negative bacteria as a model system". Methods 20, 111-126. Thanassi, D . G. , and Hultgren, S. J. (2000b). "Multiple pathways allow protein secretion across the bacterial outer membrane". Curr Opin Cel l B io l 12, 420-430. Thomas, J., Stafford, G . P., and Hughes, C. (2004). "Docking of cytosolic chaperone-substrate complexes at the membrane ATPase during flagellar type III protein export". Proc Natl Acad Sci U S A 101, 3945-3950. Troisfontaines, P., and Cornells, G . R. (2005). "Type III secretion: more systems than you think". Physiology (Bethesda) 20, 326-339. Ueno, T., Oosawa, K . , and Aizawa, S. (1992). " M ring, S ring and proximal rod of the flagellar basal body of Salmonella typhimurium are composed of subunits of a single protein, F l i F " . J M o l B i o l 227, 672-677. 169 Van den Berg, B . , demons, W . M . , Jr., Collihson, I., Modis, Y . , Hartmann, E . , Harrison, S. C , and Rapoport, T. A . (2004). "X-ray structure o f a protein-conducting channel". Nature 427, 36-44. V o n Pawel-Rammingen, U . , Telepnev, M . V . , Schmidt, G . , Aktories, K . , Wolf-Watz, H . , and Rosqvist, R. (2000). " G A P activity of the Yersinia Y o p E cytotoxin specifically targets the Rho pathway: a mechanism for disruption of actin microfilament structure". M o l Microbiol 36, 737-748. Voulhoux, R., Bos, M . P., Geurtsen, J., Mols , M . , and Tommassen, J. (2003). "Role o f a highly conserved bacterial protein in outer membrane protein assembly". Science 299, 262-265. Wattiau, P., Bernier, B . , Deslee, P., Michiels, T., and Cornells, G . R. (1994). "Individual chaperones required for Y o p secretion by Yersinia". Proc Natl Acad Sci U S A 91, 10493-10497. Weiss, A . A . , Johnson, F. D . , and Burns, D . L . (1993). "Molecular characterization of an operon required for pertussis toxin secretion". Proc Natl Acad Sci U S A 90, 2970-2974. Welch, M . , Oosawa, K . , Aizawa, S., and Eisenbach, M . (1993). "Phosphorylation-dependent binding of a signal molecule to the flagellar switch of bacteria". Proc Natl Acad Sci U S A 90/8787-8791. Wijson, R. K . , Shaw, R. K . , Daniell, S., Knutton, S., and Frankel, G . (2001). "Role of EscF, a putative needle complex protein, in the type III protein translocation system of enteropathogenic Escherichia coli". Cel l Microbiol 3, 753-762. Woestyn, S., Al laoui , A . , Wattiau, P., and Cornells, G . R. (1994). " Y s c N , the putative energizer of the Yersinia Y o p secretion machinery". J Bacteriol 176, 1561-1569. W u , T., Malinverni, J., Ruiz, N . , K i m , S., Silhavy, T. J., and Kahne, D . (2005). "Identification of a multicomponent complex required for outer membrane biogenesis in Escherichia coli". Cel l 121, 235-245. Yahr, T. L . , Goranson, J., and Frank, D . W . (1996). "Exoenzyme S of Pseudomonas aeruginosa is secreted by a type III pathway". M o l Microbiol 22, 991-1003. Y i p , C . K . , Finlay, B . B . , and Strynadka, N . C. (2005a). "Structural characterization of a type III secretion system filament protein in complex with its chaperone". Nat Struct M o l B io l 12. 75-81. Y i p , C . K . , Kimbrough, T. G. , Felise, H . B . , Vuckovic , M . , Thomas, N . A . , Pfuetzner, R. A . , Frey, E . A . , Finlay, B . B . , Mi l le r , S. I., and Strynadka, N . C . (2005b). "Structural characterization of the molecular platform for type III secretion system assembly". Nature 435, 702-707. Y i p , C . K . , and Strynadka, N . C . (2006). "New structural insights into the bacterial type III secretion system". Trends Biochem Sci 31, 223-230. Yonekura, K . , M a k i , S., Morgan, D . G. , DeRosier, D . J., Vonderviszt, F., Imada, K . , and Namba, K . (2000). "The bacterial flagellar cap as the rotary promoter of flagellin self-assembly". Science 290, 2148-2152. Yonekura, K . , Maki-Yonekura, S., and Namba, K . (2003). "Complete atomic model of the bacterial flagellar filament by electron cryomicroscopy". Nature 424, 643-650. Zhou, D . , Mooseker, M . S., and Galan, J. E . (1999). "Role of the S. typhimurium actin-binding protein S ipA in bacterial internalization". Science 283, 2092-2095. Appendix - Publications arising from graduate work First author publications arising from Ph.D. studies 2006 Yip , C . K . & Strynadka, N . C . New structural insights into the bacterial type III secretion system. Trends Biochem. Sci. 2006, 31(4):223-230. 2005 Yip , C . K . , Kimbrough, T . G . , Felise, H .B . , Vuckovic, M . , Thomas, N . A . , Pfuetzner, R . A . , Frey, E . A . , Finlay, B .B . , Miller, S.I., & Strynadka, N . C . Structural characterization of the molecular platform for type III secretion system assembly. Nature. 2005, 435(7042):702-707. 2005 Yip , C . K . , Finlay, B.B. , & Stynadka, N . C . Structural characterization of a type III secretion system filament protein in complex with its chaperone. Nature Struct. Mol. Biol. 2005, 12(1):75-81. Additional publications during Ph.D. studies 2005 Thomas, N . A . , Deng, W. , Puente, J .L. , Frey, E . A . , Y ip , C . K . , Strynadka, N . C , & Finlay, B .B . CesT is a multi-effector chaperone and recruitment factor required for the efficient type III secretion of both L E E - and non-LEE-encoded effectors of enteropathogenic Escherichia coli. Mol. Microbiol. 2005, 57(6): 1762-1779. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0092926/manifest

Comment

Related Items