Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

On the quaternary structure and gating of the bacterial protein translocation channel Dalal, Kush 2012

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

Item Metadata

Download

Media
24-ubc_2012_spring_dalal_kush.pdf [ 19.44MB ]
Metadata
JSON: 24-1.0072537.json
JSON-LD: 24-1.0072537-ld.json
RDF/XML (Pretty): 24-1.0072537-rdf.xml
RDF/JSON: 24-1.0072537-rdf.json
Turtle: 24-1.0072537-turtle.txt
N-Triples: 24-1.0072537-rdf-ntriples.txt
Original Record: 24-1.0072537-source.json
Full Text
24-1.0072537-fulltext.txt
Citation
24-1.0072537.ris

Full Text

ON THE QUATERNARY STRUCTURE AND GATING OF THE BACTERIAL PROTEIN TRANSLOCATION CHANNEL  by Kush Dalal  B.Sc., Simon Fraser University, 2002 M.Sc., Simon Fraser University, 2006  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  Doctor of Philosophy in THE FACULTY OF GRADUATE STUDIES (Biochemistry and Molecular Biology)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  January 2012 © Kush Dalal, 2012  Abstract The SecY protein-conducting channel associates with different cytosolic partners to drive the translocation of preprotein substrates across the bacterial inner membrane. In this thesis, several outstanding questions regarding the structure and function of the SecY channel are addressed. Our first study is motivated by the poorly defined interactions between the channel and its binding partners. We characterize the binding mode and stoichiometry of two SecY interacting proteins, the SecA ATPase and Syd, which each form 1:1 complexes with the channel. In the second study, we isolate the SecY dimer (i.e. two SecY channels), which is shown to be essential to activate the SecA ATPase activity and support protein transport. Analysis of SecY dimers in vivo further demonstrates that each constituent SecY copy has a different role in the translocation reaction. Finally, we discover that the SecY channel, in addition to transporting preprotein substrates, is also highly specific for monovalent anions. This selective conductance explains why translocation does cause a general membrane permeability and cell death. Our findings are discussed in the broader context of genetic, biochemical and structural information on the SecY channel and other translocation components.  ii  Preface The following describes several publications that were used as a basis for the text and figures in this thesis. Where appropriate, the contributions of supporting authors in each publication are described. Permission was obtained from the respective journals to reuse information.  Chapter 1: Introduction Chapter 1 is based on a review published in Trends in Cell Biology in 2011.  •  Dalal, K., and Duong, F. (2011). The SecY complex: conducting the orchestra of protein translocation. Trends Cell Biol. In press.  I wrote the text jointly with my supervisor. I conceived and produced all figures in this work, with the exception of Figure 1-4, which was hand drawn by Ms. Kailun Jiang.  Chapter 2: The interaction between SecY and its binding partners Syd and SecA. •  This research was originally published in the Journal of Biological Chemistry K. Dalal, N. Nguyen, M. Alami, J. Tan, T. Moraes, W. Lee, R., Maurus. S. Sligar, G. Brayer and F. Duong. Structure, binding, and activity of Syd, a SecY-interacting protein. J. Biol. Chem. 2009; 12:7897-7902. © the American Society for Biochemistry and Molecular Biology.  iii  In Dalal et al., 2009, I performed all experiment that lead to the following figures shown in this thesis: Figure 2-1, Figure 2-3B, Figure 2-4, Figure 2-6; and Figure 27. In Figure 2-3A, I performed gel filtration analyis of Syd protein, while Trevor F. Moraes performed multi-angle light scattering measurements. The multiple sequence alignment shown in Figure 2-5 was created by Jennifer Tan, who also cloned the Syd cysteine mutants. Protein translocation assays in Figure 2-7, A and B, and nativePAGE analysis in Figure 2-7C was performed by Meriem Alami. The crystallization and structure determination (Table 2-1) of Syd protein was performed by Gary. D. Brayer, Nham Nguyen, Robert Maurus and Woo-Cheol Lee. I performed the electrostatic analysis of the Syd crystal structure as shown in Figure 2-4. The manuscript was jointly written between my supervisor and myself, with contributions from the other authors.  In addition a portion of Chapter 2 is based on a submitted manuscript.  •  Dalal, K., Chan, C.S., Sligar, S.G., and Duong, F. (2011). The SecY channel dimer and acidic lipids are required for SecA-dependent preprotein translocation. Submitted.  In the submitted article, I performed all experiments and jointly wrote the manuscript with my supervisor. Figure 2-2 and Figure 2-3B in this thesis are based on figures from the submitted manscript.  iv  Chapter 3: The SecY channel dimer and acidic lipids are required for SecA-dependent preprotein translocation. Chapter 3 is based on the submitted article that was described for Chapter 2. In addition, parts of the introduction and discussion for Chapter 3 were adapted from a publication in EMBO Journal in 2007  •  Dalal, K., Chan, C.S., Sligar, S.G., and Duong, F. (2011). The SecY channel dimer and acidic lipids are required for SecA-dependent preprotein translocation. Submitted.  •  Alami, M., Dalal, K., Lelj-Garolla, B., Sligar, S.G., and Duong, F. (2007). Nanodiscs unravel the interaction between the SecYEG channel and its cytosolic partner SecA. EMBO J. 8, 1995-2004.  In Chapter 3, I performed all experiments, but must acknowledge Jean-François Montariol for purification of SecYEG complexes and Catherine S. Chan for cloning the pBAD22-HisEYFF-YG and pBAD22-HisEYFF -YEG. Catherine S. Chan also helped with purification of the SecY L106C mutant complex and initial set-up of crosslinking assays as shown in Figure 3-7. This manuscript was written jointly between my supervisor and me. All Appendices related to this chapter are based on my work alone.  v  Chapter 4: The SecY complex forms a channel capable of ionic discrimination. Chapter 4 is based on a composite of publications in EMBO reports, 2009 and Channels (Austin), 2010.  •  Dalal, K., and Duong, F. (2009). The SecY complex forms a channel capable of ionic discrimination. EMBO Rep. 7, 762-768.  •  Dalal, K., Bao, H., and Duong, F. (2010). Modulation of the SecY channel permeability by pore mutations and trivalent cations. Channels (Austin) 2,  In the first article (EMBO reports) I performed all experiments, produced all figures and jointly wrote the manuscript with my supervisor.  The figures appearing in the  thesis that were taken from this publication were Figure 4-1 to Figure 4-3, and Figure 4-5 to Figure 4-8. In Figure 4-3B, Jacques Courtades assisted with Alkaline Phosphatase measurements. All Appendices related to this chapter are based on my work alone.  In the second article (Channels), I performed all experiments and jointly wrote the manuscript with my supervisor. Figures 4-4 and Figure 4-9 were taken from this publication. Huan Bao assisted with the preparation of membrane vesicles in some of the experiments in the Channels article.  vi  Appendix A: Purification of the SecY complex and reconstitution into Nanodiscs Appendix A is based on a methods article published in Methods in Molecular Biology in 2010.  •  Dalal, K., and Duong, F. (2010). Reconstitution of the SecY translocon in nanodiscs. Methods Mol. Biol. 145-156.  I wrote the manuscript jointly with my supervisor and produced all figures in this publication.  vii  Table of Contents Abstract .................................................................................................................................... ii Preface ..................................................................................................................................... iii Table of Contents ................................................................................................................. viii List of Tables ........................................................................................................................ xiii List of Figures ....................................................................................................................... xiv List of Equations .................................................................................................................. xvi List of Symbols, Abbreviations and Terms ...................................................................... xvii Acknowledgements .............................................................................................................. xxi Dedication ............................................................................................................................ xxii Chapter 1: Introduction ........................................................................................................ 1 1.1  Historical perspective................................................................................................ 1  1.2  The Sec channel: coordinating the many components of protein translocation ....... 2  1.3  Structure of the SecY complex ................................................................................. 3  1.3.1 1.4  Mechanism of SecY channel opening................................................................... 5 Involvement of SecA in post-translational translocation across the SecY complex 6  1.4.1  The structure of the SecA ATPase ........................................................................ 7  1.4.2  Interactions between SecA and translocation components ................................... 9  1.4.3  The interaction of SecA with the SecY complex ................................................ 11  1.4.4  The role of lipids in SecA-driven translocation .................................................. 12  1.5  Co-translational translocation: Interactions between SecY and the ribosome........ 14  1.5.1  The SRP targeting pathway ................................................................................ 14  1.5.2  Structure of the ribosome bound Sec complex ................................................... 15  1.6  The role of SecY oligomerization ........................................................................... 18  1.6.1  Dimers and higher SecY oligomers in protein translocation .............................. 18  1.6.2  Orientation of the SecY dimer ............................................................................ 19  1.7  The ion gate-keeping activity of Sec channels ....................................................... 22  1.7.1  Maintaining the integrity of the membrane for small molecules ........................ 22  1.7.2  The role of pore ring and plug in SecY ionic conductance................................. 22  viii  1.7.3  Comparison of the conductance properties of SecY and Sec61 channels .......... 23  1.8  Quality control mechanisms: interaction between SecY and Syd .......................... 24  1.9  The SecY general docking platform ....................................................................... 25  1.10  Overview of objectives ........................................................................................... 27  Chapter 2: The interaction between SecY and its binding partners Syd and SecA ...... 29 2.1  Introduction ............................................................................................................. 29  2.2  Materials and methods ............................................................................................ 31  2.2.1  Plasmids and biological reagents ........................................................................ 31  2.2.2  Purification of Syd .............................................................................................. 31  2.2.3  Purification of SecA ............................................................................................ 32  2.2.4  Crystallization and structure determination of Syd ............................................ 32  2.2.5  Analytical gel filtration ....................................................................................... 36  2.2.6  Preparation of inner membrane vesicles (IMVs) and in vitro protein translocation  experiments ..................................................................................................................... 36 2.2.7  Preparation and iodogenation of preprotein substrates ....................................... 37  2.2.8  Nanodisc reconstitutions and other methods ...................................................... 38  2.3 2.3.1  Results ..................................................................................................................... 39 Reconstitution of the SecY complex in Nanodiscs and binding to Syd or SecA  proteins ............................................................................................................................ 39 2.3.2  The SecY complex binds to Syd or SecA in a 1:1 molar ratio ........................... 43  2.3.3  Crystal structure of Syd ...................................................................................... 45  2.3.4  Interacting surface between SecY and Syd and exclusion of SecA.................... 49  2.3.5  Effect of Syd on translocation efficiency and stability of the SecY complex .... 51  2.4 2.4.1  Discussion ............................................................................................................... 55 Syd provides a new framework for understanding interactions between SecY and  its partners ....................................................................................................................... 55 2.4.2  The stoichiometry between SecY and SecA ....................................................... 56  Chapter 3: The SecY channel dimer and acidic lipids are required for SecA-dependent preprotein translocation ....................................................................................................... 58 3.1  Introduction ............................................................................................................. 58  3.2  Materials and methods ............................................................................................ 60 ix  3.2.1  Biological reagents.............................................................................................. 60  3.2.2  Nanodisc reconstitutions ..................................................................................... 61  3.2.3  Translocation ATPase measurements ................................................................. 62  3.2.4  Other methods ..................................................................................................... 62  3.3  Results ..................................................................................................................... 63  3.3.1  The SecY dimer supports the SecA translocation ATPase ................................. 63  3.3.2  The dimer remains active when one SecY copy is defective for SecA binding . 67  3.3.3  The SecY monomer suffices to bind the signal sequence................................... 70  3.3.4  The back-to-back dimer is the predominant formation in discs ......................... 72  3.3.5  Complementation between two SecY mutants restores protein translocation .... 75  3.4  Discussion ............................................................................................................... 77  3.4.1  The requirement and functional asymmetry of the SecY dimer ......................... 77  3.4.2  Comparison of the translocation reaction and channel dynamics in Nanodiscs  and membranes ............................................................................................................... 78 3.4.3  Role of lipids in protein translocation ................................................................. 80  3.4.4  Possible advantages of the SecY dimer .............................................................. 81  Chapter 4: The SecY complex forms a channel capable of ionic discrimination........... 82 4.1  Introduction ............................................................................................................. 82  4.2  Materials and methods ............................................................................................ 83  4.2.1  Biological and chemical reagents ....................................................................... 83  4.2.2  Measurement of ∆ψ and ∆pH ............................................................................. 84  4.2.3  Protein translocation alkaline phosphatase assays .............................................. 85  4.3  Results ..................................................................................................................... 86  4.3.1  Ionic specificity of the open SecY channel......................................................... 86  4.3.2  Contribution of the pore ring residues to ionic selectivity.................................. 90  4.3.3  Phenylalanine substitution in the pore ring create a chloride permeability ........ 92  4.3.4  Ionic selectivity is a conserved characteristic ..................................................... 94  4.3.5  Proton permeability of the open channel ............................................................ 95  4.3.6  Ionic conductance of the active channel ............................................................. 97  4.3.7  Binding of trivalent cations to the channel creates a chloride permeability ..... 100  4.4  Discussion ............................................................................................................. 102 x  4.4.1  Possible mechanisms for ionic discrimination .................................................. 102  4.4.2  Properties of specific residues in the pore ring ................................................. 103  4.4.3  Possible relationship with ClC chloride transporters ........................................ 104  4.4.4  Implications of SecY conductance for the cell ................................................. 105  Chapter 5: General conclusions and future work........................................................... 106 References ............................................................................................................................ 114 Appendices ........................................................................................................................... 126 Appendix A Purification of the SecY complex and reconstitution into Nanodiscs .......... 126 A.1  SecY purification materials............................................................................... 126  A.2  Membrane scaffold protein (MSP) ................................................................... 127  A.3  Lipids ................................................................................................................ 128  A.4  SecY reconstitution materials ........................................................................... 128  A.5  Size exclusion chromatography materials ........................................................ 129  A.6  SDS-PAGE materials ........................................................................................ 129  A.7  Native-PAGE materials .................................................................................... 130  A.8  Method for SecY purification and reconstitution into Nanodiscs .................... 131  A.9  Protocol for the purification of the SecY complex ........................................... 133  A.10  Preparation of membrane scaffold protein ........................................................ 135  A.11  Preparation of lipids .......................................................................................... 135  A.12  Protocol for Nanodisc reconstitution of the SecY complex.............................. 136  A.13  Size exclusion chromatography of Nanodisc reconstitutions ........................... 137  A.14  SDS-PAGE protocol ......................................................................................... 138  A.15  Native-PAGE protocol ...................................................................................... 138  A.16  Notes ................................................................................................................. 139  Appendix B Cartoon representation of SecY monomer or dimer incorporated into Nanodiscs .......................................................................................................................... 141 Appendix C Cartoon Representation of Nanodisc-SecY-SecA complexes ...................... 142 Appendix D SecA Translocation ATPase kinetics curves with Nanodiscs ...................... 143 Appendix E Reconstitutions with MSP2N2 and sucrose density purification ................. 144 Appendix F Position of cysteine residues introduced into the SecY complex ................. 145 Appendix G Western blot analysis of wild-type and mutant IMVs.................................. 146 xi  Appendix H Oxonol VI fluorescence responds linearly to the membrane potential ........ 147 Appendix I Schematic of SecY conductance assay .......................................................... 148 Appendix J The membrane potential across the SecYPrlA4 membrane collapses in the presence of halide anions .................................................................................................. 149 Appendix K Collapse of the membrane potential at low halide ion concentration .......... 150 Appendix L Structural representation of the SecY pore ring residues ............................. 151 Appendix M Temperature dependence of chloride leakage during translocation ............ 152 Appendix N Acclamation ................................................................................................. 153  xii  List of Tables Table 2-1: Summary of structure determination statistics for Syd ......................................... 34  xiii  List of Figures Figure 1-1: Crystal structures of the SecY complex. ................................................................ 4 Figure 1-2: Crystal structure of bacterial SecA ATPase monomer. ......................................... 8 Figure 1-3: SecA interacts with numerous binding partners. ................................................. 10 Figure 1-4: Post- and co-translational modes of translocation in bacteria.............................. 13 Figure 1-5: Cyro-electron microscopy structure of a translating ribosome bound to the Sec61 channel. ................................................................................................................................... 17 Figure 1-6: Representation of contact points between SecA and the back-to-back SecY dimer. ...................................................................................................................................... 21 Figure 1-7: Electrostatic interactions between SecY and SecA, and SecY and Syd. ............. 26 Figure 2-1: Binding of Syd onto the SecYEG-Nanodisc. ....................................................... 40 Figure 2-2: The SecY monomer and dimer in Nanodiscs binds to SecA. .............................. 42 Figure 2-3: Stoichiometry between the SecY complex and Syd or SecA. ............................. 44 Figure 2-4: Crystal structure and electrostatic surface of Syd. ............................................... 46 Figure 2-5: Database retrieval and multiple sequence alignment of Syd. .............................. 47 Figure 2-6: Contact surface between Syd and SecY and exclusion of SecA.......................... 50 Figure 2-7: Translocation activity and stability of the SecY complex in the presence of Syd. ................................................................................................................................................. 53 Figure 3-1: Two SecY copies are necessary to activation the SecA translocation ATPase. .. 65 Figure 3-2: In vivo and in vitro activity of the SecY mutant channels employed in this study. ................................................................................................................................................. 66 Figure 3-3: The dimer remains active when a SecY copy is defective for SecA binding. ..... 68 Figure 3-4: Pull-down of 125I-labelled SecA with Nanodiscs. ................................................ 69 Figure 3-5: The SecY monomer suffices to bind the signal sequence. ................................... 71 Figure 3-6: Orientation of the SecY copies within the disc. ................................................... 73 Figure 3-7: Crosslinking of SecE-L106C and SecY-A103C in detergent solution. ............... 74 Figure 3-8: Trans-complementation between SecYFF and SecYE mutant channels. .............. 76 Figure 4-1: Ionic conductance of the SecY complex. ............................................................. 88 Figure 4-2: IMVs bearing the SecYPrlA4 complex cannot maintain an electrical potential in the presence of chloride. ............................................................................................................... 89  xiv  Figure 4-3: The SecY pore mutants that are the most active in protein translocation are also the most leaky, but for only monovalent anions. .................................................................... 91 Figure 4-4: Effect of phenylalanine residues introduced in the SecY pore ring on the channel ion conductance and protein translocation activity. ................................................................ 93 Figure 4-5: The ionic specificity of the SecY complex is a conserved characteristic. ........... 94 Figure 4-6: The Cl- conductive SecY mutant channels are impermeable to protons.............. 96 Figure 4-7: Ionic conductance during protein translocation. .................................................. 98 Figure 4-8: Example of traces showing the chloride conductance during translocation. ....... 99 Figure 4-9: The trivalent cation Al3+ increases the Cl- permeability of the SecY channel. .. 101 Figure A-1: SecY purification and reconstitution into Nanodiscs.....…...………....……….132  xv  List of Equations Equation 3-1: One site quadratic binding equation................................................................. 62  xvi  List of Symbols, Abbreviations and Terms ∆ψ: difference in electrical potential across a membrane ∆pH: difference in pH across a membrane Å: Angstrom 10-10 meters ACMA: 9-amino-6-chloro-2-methoxyacridine, a pH sensitive fluorescent dye AMP-PNP: non-hydrolyzable analog of ATP APBS: Adaptive Poisson-Boltzmann Solver ATP: adenosine tri-phosphate ATPγS: non-hydrolyzable analog of ATP B. subtilis: Bacillus subtilis BiP: Binding-immunoglobulin protein BL21: E. coli strain suitable for protein over expression BME: β-mercapto-ethanol, a reducing agent BMOE: bis-maleimido-ethane, a bifunctional cysteine crosslinker with an ethane spacer group C4 loop: the loop region between TMS6 and TMS7 in SecY or Sec61α subunits C5 loop: the loop region between TMS8 and TMS9 in SecY or Sec61α subunits CCCP: carbonyl cyanide m-chlorophenyl hydrazone, a protonophore CJ107: Temperature sensitive strain of E. coli that expresses the thermo-sensitive SecY24 mutant CL: Cardiolipin, an acidic phospholipid made from two covalently attached PG molecules CP3: Copper3(Phenanthroline), an oxidizing agent Cryo-EM: cryo-electron microscopy DDM: n-Dodecyl β-D-maltoside, a non-ionic detergent xvii  DTT: di-thiothreitol, a reducing agent E.coli: Escherichia coli EDTA: Ethylene-di-amine-tetra-acetic acid ER: endoplasmic reticulum FRET: Fluorescence resonance energy transfer FtsY: the bacterial version of SRP FtsH: a protease in the bacterial inner membrane that targets and degrades compromised SecY subunits GTP: guanosine tri-phosphate HSD: Helical Scaffold Domain, a domain in SecA containing the two helix finger IODO-GEN: 1,3,4,6-tetrachloro-3α,6α-diphenylglucoluril IPTG: isopropyl 1-thio-β-D-galactopyranoside IMV: inverted (inside out) inner membrane vesicles kDa: kilodalton = 1000 gram/mole KM9: a strain of E. coli that lack the F0F1 ATP synthase gene MD: molecular dynamics simulations M. jannaschii: Methanococcus jannaschii MSP: membrane scaffold protein mRNA: messenger RNA NADH: Nicotinamide adenine dinucleotide NBD: Nucleotide binding domain, a domain of SecA that interacts with ATP NEM: N-ethylmaleimide, blocks disulphide bond formation NMR: nuclear magnetic resonance  xviii  OmpA: outer membrane protein A lacking a signal sequence on its N terminus Oxonol VI: bis-3-propyl-5-oxoisoxazol-4-yl pentamethine oxonol, a voltage sensitive dye PAGE: poly-acrylamide gel electrophoresis PBD: poly-peptide binding domain (same as PPXD), a domain of SecA that interacts with preprotein substrates P. furiosus: Pyrococcus furiosus PhoA: Alkaline Phosphate A, a preprotein that is transported across the inner membrane by SecYEG. pOA: outer membrane protein with a signal sequence that is transported across the inner membrane by SecYEG PPXD: poly-peptide crosslinking domain of SecA, see PBD. Prl: protein localization, a class of mutations in SecY that bypass the requirement of the signal sequence for protein translocation, and destabilize the close state of the channel. pOmpA: see pOA proOmpA: see pOA. RNC: ribosome-nascent-chain rRNA: ribosomal RNA SDS: sodium dodecyl sulfate SecA: the cytosolic ATPase in bacteria that drive post-translational translocation across the SecY channel. SecB: a cytosolic chaperone SecY: the protein-conducting channel in bacteria.  xix  Sec61: the protein-conducting channel in eukaryotes. SR: signal recognition particle receptor, the receptor for SRP, called FtsY in bacteria SRP: signal recognition particle, called Ffh in bacteria Syd: suppressor of SecY dominance. TCA: tri-chloro acetic acid T. maritima: Thermatoga maritima TMS: trans-membrane segment, a membrane spanning α-helix of a membrane protein  xx  Acknowledgements I would like to thank my supervisor, Dr. Franck Duong, for allowing me to work on this exciting project, and teaching me the basics of membrane protein biochemistry. I would also like to thank several laboratory members who greatly assisted me over the years including: H. Bao, X. Zhang, C.S. Chan, J.F. Montariol, H. Li, S. Macdonald, A. Mills, S. Lalani and M. Alami. My supervisory committee, including Dr. Eric Jan and Dr. Robert Molday, also provided much appreciated direction over the course of my degree. Lastly, I would like to thank my family and friends for supporting me over the years. The work presented in this thesis was generously funded by fellowships and scholarships from the B.C. provincial government, the Natural Sciences and Engineering Council of Canada, and the University of British Columbia.  xxi  Dedication  To my family  xxii  Chapter 1: Introduction 1.1  Historical perspective Membrane proteins are arguably the most difficult macromolecules to handle because  they exist in the complex environment of the lipid bilayer. The intrinsic dynamics of membrane proteins and lipids have impeded the elucidation of structural and mechanistic details of many membrane receptors, channels and transporters. Exemplifying these difficulties is the Sec (secretory pathway) protein-conducting channel, the subject of several controversies with respect to its quaternary structure and channel dynamics. This universally conserved membrane channel, termed SecY in bacteria and Sec61 in eukaryotes, creates a protein-conducting pathway by which newly synthesized polypeptides make their way through the lipid bilayer. Several decades ago, pioneering cell biology and genetic studies provided basic information on the identity and function of the SecY channel, but did not give insight into structure of the transporter or the mechanism of protein translocation. In the following years, biochemical methods were developed to dissect the translocation mechanism, purify the channel and determine its three dimensional structure. Subsequent observations have shown that the channel is not a simple passive pore; it displays remarkable complexity by interacting with numerous soluble partners which drive polypeptide transport or otherwise modulate channel activity. These breakthroughs have motivated us to continue with a biochemistry approach to study protein translocation. Accordingly, in the work presented here, we employ biochemical techniques to control the oligomeric state of the SecY channel, characterize its interactions with binding partners, and investigate the gating of substrates and small molecules across the channel pore.  1  1.2  The Sec channel: coordinating the many components of protein translocation Each minute, hundreds of different polypeptides are targeted, transported across or  integrated into the bacterial plasma or the eukaryotic endoplasmic reticulum (ER) membrane with precision accuracy [Rapoport, 2007]. This accuracy critically depends on targeting signals such as N-terminal cleavable signal peptides and transmembrane segments (TMS) [von Heijne, 1988; Gouridis et al., 2009]. In fast growing cells such as bacteria and yeast, translocation can occur post-translationally, whereas in mammals, transport is mostly cotranslational [Osborne et al., 2005]. Regardless of the mode of translocation, nascent proteins are transferred to a membrane complex or channel termed SecYEG in bacteria, SecYEβ in archaea and Sec61αβγ in eukaryotes. The Sec assembly is comprised of a central subunit, SecY/Sec61α, made of 10 TMS associated with two smaller subunits: SecG/β, which consist of one or two TMS depending on the organism, and SecE/Sec61γ that contains a single TMS, except in some Gram-negative bacteria where SecE possesses two additional Nterminal TMS [Brundage et al., 1990; Gorlich and Rapoport, 1993]. Together, the three subunits associate into a 1:1:1 stoichiometry to form a membrane channel and a binding platform for many partners, such as the SecA ATPase and the ribosome that drives protein transport, or FtsY and Syd that ensures the efficient delivery of substrates and the proper assembly of the complex, respectively. Like many other conserved cellular processes, protein transport has mostly been analyzed in bacteria and yeast, and considerable insight into the working mechanism of translocation has been obtained by crystallographic and cryo-electron microscopy (cryo-EM) analysis. Genetic, biochemical and computational approaches have then linked this structural information to the dynamics of protein translocation, making the Sec channel one of the best  2  understood molecular machines in the cell. This chapter will introduce the bacterial SecY complex and its best characterized partners, and will summarize the major findings that have been obtained following the high-resolution structure of the channel [van den Berg et al., 2004].  1.3  Structure of the SecY complex The structure of the SecY complex, particularly the architecture of the channel pore  and binding site for channel partners, is a central concept in this thesis. The first highresolution crystal structure of the SecY complex was from M. jannaschii and has been fundamental for understanding the working mechanism of the channel [van den Berg et al., 2004]. Three main characteristics contained within the SecY subunit were identified (Figure 1-1): the “pore ring”, made from six hydrophobic residues at the middle point of the channel; the “plug” domain formed by a small α-helix seated on the pore on its extracytosolic side; and the “lateral gate” creating an opening toward the bilayer at the interface between transmembrane segments TMS2 and TMS7. Viewed from the top, the SecY subunit resembled a clamshell and when viewed from the side, it could be described as an hourglass topped with a plug (Figure 1-1, A and B; and for cartoon representation Figure 1-1D). Two large loops protrude from the cytosolic face of the complex, and form part of the binding interface for channel partners (Figure 1-1B; and see Section 1.9 for further discussion). Sequence comparison and additional atomic structures have confirmed that these structural characteristics are strictly conserved across evolution. The long TMS of the SecE/γ subunit is also conserved and binds to the SecY subunit on the side opposed to the lateral gate, also termed the “back” of the complex. The SecG/β subunit, which is the least conserved element  3  of the channel with respect to its sequence and possibly function, is located at the periphery of the channel near SecE/γ.  Figure 1-1: Crystal structures of the SecY complex. The SecYEβ complex from M. jannaschii (PDB: 1RHZ; 3.2 Å resolution; van den Berg et al., 2004) viewed (A) from the cytosol or (B) from the plane of the membrane. The complex is coloured as follows: SecY TMS 1-5 (grey); SecY TMS6-10 (cyan); SecE (yellow); Secβ (orange); SecY pore ring (red); SecY plug domain (blue). The locations of the hinge region and cytosolic loops of SecY are noted by “*” and “**” respectively. The position of the membrane surface is represented by the dark grey lanes. (C) The co-crystal structure of the SecYEG-SecA complex from T. maritima (PDB: 3DIN; 4.5 Å resolution; Zimmer et al., 2008) viewed from the plane of the membrane. A single SecYEG complex (grey, yellow, and orange) is bound to a single SecA molecule (blue). Note that the SecA two-helix finger (red) is inserted into the cytoplasmic funnel made by the SecY subunit. See Section 1.4 for a description of SecA structure. (D) Schematic cross-section of the SecY channel pore. SecG is not shown in this representation and the orientation of the SecY complex is inverted compared to (B) and (C). S120 denotes the position on the SecE subunit with which the plug can form disulphide bonds (see Section 1.3.1). Arrows denote the position on SecY that interact with various binding partners. The presented structures in A-C were modified from Dalal and Duong, 2011. The cartoon in D was modified from Tam et al., 2005.  4  1.3.1  Mechanism of SecY channel opening Solving the SecY complex crystal structure made it possible to formulate specific  hypotheses regarding the mechanism of channel gating. The first prediction was that the insertion of a signal sequence or a transmembrane segment would occur at the lateral gate. This notion was supported by an earlier photocrosslinking analysis showing that a signal sequence can simultaneously contact TMS2 and TMS7 during transport [Plath et al., 1998]. Site-directed cysteine crosslinking analysis have now confirmed the prediction [Osborne and Rapoport, 2007], whereas molecular dynamics (MD) simulations, which allow modeling of the subtle conformational changes that may occur during transport, indicate that the opening of the lateral gate is possible because of a flexible hinge sequence located between TMS5 and TMS6 at the back of the complex [Gumbart and Schulten, 2007]. The structural information also led to the prediction that the pore ring, which is only 5-8 Å in diameter in the resting state, would need to expand in diameter to allow insertion of the polypeptide chain. This widening would follow the opening of the lateral gate because TMS2 and TMS7 contain three residues out of the six amino-acids that create the pore structure (Figure 1-1A). Although not yet experimentally tested, the interior of the channel during polypeptide transport would be expected to approach an inner horizontal diameter of 20 Å x 15 Å following a motion of 15° at the hinge [van den Berg et al., 2004]. This seems to be the case at least in silico, since MD simulations show that the pore ring can expand to allow the passage of beads up to 16 Å in diameter [Tian and Andricioaei, 2006]. The plug domain, which is linked to the channel with unstructured loops, was another element predicted to be involved in the gating of the channel. During transport, this domain (TMS2a) was found located near the carboxyl terminus of the SecE subunit, more than 20Å  5  away from its normal position [Harris and Silhavy, 1999; Tam et al., 2005; see Figure 1-1D for a cross section representation of SecY indicating the position of the plug]. The use of crosslinking agents of different lengths has now confirmed that a plug displacement of at least 13Å is necessary to allow unrestricted polypeptide transport [Lycklama et al., 2010] and as expected, a plug domain locked in its resting position adjacent to the pore ring inactivates the channel [Lycklama et al., 2010; Maillard et al., 2007]. The opening of the channel is largely influenced by the association between the plug and the pore because some mutations in these structures can destabilize the closed state of the channel. MD simulations indicate that perturbations in the hydrogen bonding network between the SecY TMS can eventually be transmitted to the pore and plug to destabilize the closed state of the channel [Bondar et al., 2010]. This destabilization may in turn allow the export of substrates with defective signal sequences [prl mutations, Smith et al., 2005]. In normal conditions, the disruption of the hydrogen bonding network might be facilitated by the docking of SecA or the ribosome onto the channel. These two channel partners drive separate modes of protein translocation in bacteria, and are discussed in detail in the next two sections.  1.4  Involvement of SecA in post-translational translocation across the SecY complex Much of the work presented in this thesis is focused on the mechanistic details of  SecA-driven protein translocation. The SecA ATPase functions as a motor protein to ‘push’ the preprotein across the SecY channel. Before docking with SecY, SecA must interact with different partners to acquire the preprotein substrate. In this section, we describe the current understanding of the structure of SecA and its interactions with the components of translocation.  6  1.4.1  The structure of the SecA ATPase Knowledge of the structure of SecA and its organization is necessary to understand  how preproteins are targeted to the SecY complex for translocation. SecA is a 100 kDa protein that is mostly dimeric in solution with a dissociation constant of ~0.5 µM [Woodbury et al., 2002], far less than its normal concentration in the cell of ~8 µM [Akita et al., 1991]. Accordingly, virtually all crystal structures of SecA display the protein as a dimer [for review see Sardis et al., 2010]. The protein has a complex domain organization (Figure 1-2) that allows it to interact with various partners and also to use the energy from ATP hydrolysis to drive polypeptide movement across the channel. Extensive crystallographic analysis shows that the ATP binding site is formed by two nucleotide binding domains (NBD1 and NBD2), which together form the ‘motor’ of SecA. ATP hydrolysis and subsequent activation of the motor is coupled to movement of two SecA domains that interact with the preprotein substrate, the polypeptide cross-linking domain (PPXD) and C-domain. Observed conformational changes of the PPXD and C-domains upon nucleotide hydrolysis [Economou and Wicker, 1994; Gellis et al., 2007] are thought to move the preprotein into the SecY pore, although this has yet to be experimentally proven. The C-domain constitutes the C-terminal third of the protein and contains a sub-domain termed the helical scaffold domain (HSD). The longest of the three helices in the HSD interacts with all the other SecA domains, providing structural integrity to the protein (Figure 1-2). In addition a hairpin helix structure in the HSD domain, known as the ‘two helix finger’, has been show to contact the preprotein during transport (see Section 1.4.3). Following the HSD domain, the extreme C terminal  7  residues of SecA, termed the ‘C-tail’, are important in signal sequence recognition, lipid binding and recruitment of other translocation components such as SecB (see next section).  Figure 1-2: Crystal structure of bacterial SecA ATPase monomer. The structural representation of SecA from T. maritima (PDB: 3DIN; 4.5 Å resolution; Zimmer et al., 2008) was created with pymol. SecA normally exists as a dimer (see text), but for simplicity, the monomer of SecA is shown here. The indicated domains of SecA are individually coloured. The ‘Motor’ of SecA is comprised of the NBD1 and NBD2 domains. The HSD domain is composed of three helices, the longest acting as a scaffold linking other SecA domains, and the other two forming the two helix ‘Finger’. The flexible C-tail of SecA was not resolved in this structure but would appear after the C terminus (indicated with a ‘C’) directly following the two helix finger. ‘N’ refers the N-terminus of the protein.  8  1.4.2  Interactions between SecA and translocation components SecA is essential for recognizing, targeting and pushing polypeptide substrates across  the SecY channel. These polypeptide substrates generally carry a signal sequence that is moderately hydrophobic [Huber et al., 2005]. SecA must interact with different binding partners to ultimately deliver the substrate to the SecY complex for transport (Figure 1-3). The recognition by SecA may begin as early as the polypeptide emerges from the ribosome because co-sedimentation experiments show that SecA HSD domain has affinity for the protein L23 at the ribosome exit tunnel [Huber et al., 2011]. Certain substrates are also transferred to SecA via the export chaperone SecB, which prevents polypeptides from acquiring their tertiary structure and keeps them in a translocation-competent and unfolded state [Bechtluft et al., 2007, and for review on SecB see Bechtluft et al., 2010]. The SecASecB interaction requires the first 11 N-terminal SecA amino-acid residues [Randall and Henzl, 2010], as well as the C-tail [Breukink et al., 1995; Fekkes et al., 1998]. SecA would then accept the unfolded substrate-SecB complex by recognizing the signal sequence of the preprotein. NMR and FRET analysis confirm the existence of a signal sequence binding site at the surface of SecA formed by the PPXD, HSD and NBD1 domains [Gelis et al., 2007; Auclair et al., 2010]. The C-tail appears to control the accessibility of the signal sequence for its binding site on SecA [Gelis et al., 2007]. A recent crystal structure of SecA further shows that a hydrophobic clamp formed by the adjacent PPXD and HSD domains is capable of interactions with the preprotein mature segment [Zimmer et al., 2009].  9  Figure 1-3: SecA interacts with numerous binding partners. The structure of SecA from B. subtilis (PBD: 1M6N; 2.7 Å resolution; Hunt et al., 2002) was created in pymol. SecA is rotated approximate 180° vertically compared to Figure 1-2. The C-tail of SecA is resolved in this structure and is shown in cyan.  10  1.4.3  The interaction of SecA with the SecY complex The recently solved co-crystal structure of SecA bound to SecYEG has provided new  clues for how these two proteins interact together to catalyze transport (Figure 1-1C) [Zimmer et al., 2008]. In the structure, the SecA two-helix finger was found inserted at the entrance of the SecY cytoplasmic funnel. The insertion of the finger allowed a partial opening of the lateral gate interface, as well as a slight shift of the plug domain toward the periplasmic side of the channel. It is reasonable to assume that the conformational change provoked by the finger domain is crucial for the gating of SecY and therefore, that the finger may act as an allosteric activator capable of “priming” the channel during the initial steps of translocation [Economou, 2008]. This priming step may be related to the SecY-dependent SecA ATPase activity that is observed in the absence of protein substrate; this ATPase is supported by the SecY complex in detergent solution or when embedded in Nanodisc particles (soluble nanoscale lipid bilayers) [Alami et al., 2007; Robson et al., 2009]. During transport, the polypeptide chain captured inside the SecA clamp would be transferred to the two-helix finger, then to the center of the channel. An elegant and systematic cysteine crosslinking analysis designed to map the protein translocation path reveals that the polypeptide chain is most likely threaded in an extended conformation between the SecA clamp and the pore ring [Bauer et al., 2009]. A similar crosslink strategy was employed to show that the SecA finger moves toward the SecY pore upon binding of ATP and probably drags the polypeptide chain with it [Erlandson et al., 2008]. Analysis of the thermodynamic energy released by SecA upon docking of the polypeptide onto SecYEG indicates that the signal sequence contributes to lowering the activation energy barrier for translocation  11  initiation [Gouridis et al., 2009]. With the energy barrier overcome, SecA would then be able to initiate cycles of ATP hydrolysis to push consecutive segments of the protein substrate across the membrane (Figure 1-4A).  1.4.4  The role of lipids in SecA-driven translocation Acidic lipids may also contribute to reducing the activation energy barrier of  translocation [de Vrije et al., 1988]. These lipids, whose deficiency severely impairs protein transport both in vivo and in vitro, cause dissociation of the SecA dimer into monomers [Or et al., 2002; Alami et al., 2007] and stimulates the ATPase activity of SecA when bound to the channel [Lill et al., 1990; Gold et al., 2010]. Recently, it was shown that the binding of SecA to the SecYEG complex is enhanced by cardiolipin (CL) in detergent solution [Gold et al., 2010]. This acidic lipid is tightly bound to the channel as it remains associated even during purification of the Sec complex in detergent solution. In detergent solution, cardiolipin also specifically relieves the Mg2+ dependent allosteric inhibition of SecA, which normally prevents the futile hydrolysis of ATP in the cytosol [Gold et al., 2007]. An atomic model of the dimeric SecY channel (discussed in Section 1.6) indicates that CL could possibly bind at the interface of the two SecY copies and relay long-distance conformational changes to the NBD domains of SecA to modulate its ATPase activity. It will be necessary to understand the contribution of CL and other acidic lipids on the quaternary structure and activity of the SecYEG-SecA complex as it exists in the membrane, and not only in detergent solution. In Chapter 3, we will demonstrate using Nanodiscs that the SecY complex and acidic lipids are essential to stimulate SecA ATPase activity. Nanodisc technology controls  12  the oligomeric state of membrane proteins by incorporating them into water-soluble disc-like particles (see Appendix A and B).  Figure 1-4: Post- and co-translational modes of translocation in bacteria. (A) SecA driven post-translational translocation through the dimeric SecY channel. The SecA two-helix finger is not shown in this cartoon. Left – The SecA nucleotide binding domain (dark green) docks onto the cytosolic loops of a non-translocating copy of the SecY channel (blue), as described in Section 1.6. Middle – Upon binding of ATP, the SecA clamp (light green) transfers the signal sequence (yellow) to the active copy of the SecY channel. Insertion of the signal sequence may induce the movement of the plug (red) away from its central position. Right – SecA binds and hydrolyses ATP, with each cycle pushing a segment of polypeptide chain across the SecY channel. (B) Ribosome driven co-translational translocation. Left - The signal sequence (yellow) that emerges from the ribosome (green) is recognized by SRP (orange) and targeted to the membraneassociated FtsY (pink). The ribosome exit tunnel is shown as a dashed line. The mRNA with 5’ and 3’ ends is shown bound to the small 40S ribosomal subunit. Middle - The ribosome nascent chain (RNC) is transferred to the SecY channel (blue), which is followed by the insertion of the signal sequence. Right – The elongation of the polypeptide chain resumes after the SRP dissociates from the signal sequence. This figure was modified from Dalal and Duong, 2011. 13  1.5  Co-translational translocation: Interactions between SecY and the ribosome The investigation of ribosome-driven translocation has been critical in elucidating the  range of channel dynamics during polypeptide transport, and also in defining the interface between the SecY/Sec61 complexes and their binding partners. Although we did not analyze this mode of translocation in the work presented in the thesis, a brief description of ribosomedriven translocation is warranted, and will show the remarkable ability of the Sec complex to coordinate the many different translocation components. It also gives a historical perspective on attempts to determine the structure of the channel in complex with a binding partner.  1.5.1  The SRP targeting pathway SecA is only found in bacteria, but the signal-recognition particle (SRP) is found in  all organisms. The SRP-dependent pathway delivers protein substrates in a co-translational manner to the membrane (Figure 1-4B) [Halic and Beckmann, 2005; Zimmermann et al., 2011; du Plessis et al., 2011] and in bacteria, this mechanism serves for the integration of most membrane proteins [for review on membrane protein integration via the Sec complex see Rapoport, 2007; duPlessis et al., 2011]. Targeting begins when SRP (termed Ffh in bacteria) binds and arrests the translation of the first TMS (of an integral membrane protein) that emerges from the ribosome. The ribosome-nascent chain complex (RNC), when bound to SRP, is then targeted to the SRP receptor at the membrane (termed FtsY in bacteria). Following GTP hydrolysis, the SRP dissociates and the RNC is transferred to the SecY channel where translation of the polypeptide chain resumes [Shan et al., 2007]. FtsY has a critical role for this sequence of events because the receptor provides the bridge between the  14  soluble SRP-RNC complex, the membrane and the Sec channel. FtsY indeed interacts with the phospholipid bilayer via two lipid-binding helices located at the N and A domains of the protein [Parlitz et al., 2007; Braig et al., 2009], and FtsY displays low but significant affinity for the cytosolic loops of the channel [Kuhn et al., 2011]. It is remarkable that the association of the lipid-binding helices with liposomes is strongly enhanced by phosphatidyl-glycerol (PG), as shown by photocrosslinking and SPR experiments [Braig et al., 2009; Lam et al., 2010]. The presence of acidic lipids in the vicinity of SecY may be functionally important because PG lipids were found increase the GTPase activity of the SRP-FtsY complex [Lam et al., 2010], and therefore may contribute to the unloading of the nascent chain to the channel.  1.5.2  Structure of the ribosome bound Sec complex The understanding of the mechanism of co-translational translocation has been  greatly aided by progress in the field of cryo-electron microscopy. The first snapshots of the RNC bound to the yeast Sec61 channel were obtained at low resolution [Beckmann et al., 1997; Beckmann et al., 2001; Menetret et al., 2005], leading to the conclusion that two to four Sec complexes were attached beneath the ribosome. With better instrumentation and resolving power, cryo-EM analysis now shows that the RNC-Sec61 complex contains of only one SecY/Sec61 heterotrimer [Becker et al., 2009], in agreement with the X-ray structure which indicates that a single complex is sufficient to create a protein-conducting channel. In these high resolution pictures, the Sec61 loops were found in proximity to the ribosomal proteins L23 and L35, as well as specific rRNA helices (Figure 1-5). L23 is now considered a general docking platform for factors that act on nascent chains, such as SecY/Sec61, SRP  15  and SecA [Kramer et al., 2009; Huber et al., 2011]. Image reconstructions also show that the channel pore is aligned directly below the ribosome exit tunnel, thus creating a continuous pathway for the transfer of the polypeptide substrate [Becker et al., 2009; Menetret et al., 2007]. In the latest cryo-EM analysis, the channel was found bound to a ribosome with nascent chain inserted into the translocation pore, causing the middle region of the lateral gate to open toward the lipid bilayer (~5 Å) [Becker et al., 2009]. In a crystal structure of SecYEβ from P. furiosus, the lateral gate was also found to be open, but throughout its entire length [Egea and Stroud, 2010]. In this latter case, a crystal packing artifact in which the Cterminal α-helix of one SecY is inserted into the lateral gate of a second juxtaposed SecY, and may have mimicked the conformational change that normally occurs when a nascent chain transits through the channel. It is remarkable that these channels, seemingly trapped in their active state, were still capped with their plug domain in closed position. This last observation has suggested that the lateral gate and the plug domain may move independently from each other at the early stage of protein insertion.  16  Figure 1-5: Cyro-electron microscopy structure of a translating ribosome bound to the Sec61 channel. (A) The ribosome nascent chain complex is shown bound to the mammalian Sec61 channel. The positions of the 60S and 40S ribosomal subunits, as well as the tRNA are shown. (B) The structure of the channel-ribosome junction (PDB: 2WWB; Becker et al., 2009). The Sec61 cytosolic loops C4 and C5 (green) are shown to be in proximity to the ribosomal proteins L23 (red) and L35 (cyan), as well as to the rRNA helices H50 and H7 (blue). The arrow shows the location of the ribosome exit tunnel. The color code employed for the Sec channel is the same as in Figure 1-1. This figure was modified from Dalal and Duong, 2011.  17  1.6 1.6.1  The role of SecY oligomerization Dimers and higher SecY oligomers in protein translocation The two different translocation modes underlines the dynamic nature of the SecY  complex, and the possibility that SecY oligomers have a role in protein transport (Figure 14). In fact, the SecY complex can form various oligomers in the membrane or in detergent solution, making the analysis of its functional quaternary structure difficult. Dimers and higher order SecY oligomers have been detected by native-PAGE , analytical ultracentrifugation [Collinson et al., 2001; Bessonneau et al., 2002] and electron microscopy [Meyer et al., 1999; Manting et al., 2000; Scheuring et al., 2005]. In two dimensional crystalline membranes, SecY complexes were related to each other by a two-fold symmetry with their interface formed from the long TMS of SecE [Breyton et al., 2002]. This dimeric arrangement was termed the “back-to-back” orientation. Several studies have since focused on understanding the role of SecY oligomerization for protein transport. Earlier electron microscopy analysis concluded that SecA and protein substrates play an important role in modulating the oligomeric status of the complex in the membrane [Meyer et al., 1999; Manting et al., 2000; Scheuring et al., 2005]. Later, it was shown using a genetically fused SecYEG dimer that a cysteine crosslink between a polypeptide substrate and the pore of a mutant channel, otherwise defective for binding of SecA, can occur only if a wild type SecY copy is covalently attached [Osborne et al., 2007]. This structural complementation was explained if SecA binds asymmetrically to the SecY dimer, with one copy actively engaged with the two-helix finger and the other (i.e. passive copy) simply serving as a docking site for the nucleotide binding domains of SecA [Osborne and Rapoport, 2007; Figure 1-4A]. The  18  second and passive SecY copy could therefore act as an anchor to prevent the complete dissociation of SecA when the two-helix finger disengages from the active copy upon ATP hydrolysis. Additional experimental support is needed, but the model is consistent with the SecYEG-SecA atomic resolution structure that show the two-helix finger is inserted into the translocation pore [Zimmer et al., 2008], while the remaining unbound portion of SecA – including the first nucleotide binding domain – is available to bind another SecY copy. This other copy might not solely act as a docking site for SecA however, because residue R357 located on the SecY cytosolic loop C5 seems important to trigger a conformational change that lead to the activation of SecA [Tsukazaki et al., 2008].  1.6.2  Orientation of the SecY dimer The question regarding the oligomeric state of SecY remains unsettled and the  orientation of SecY protomers within the oligomers further complicates the problem. A back-to-back dimer can spontaneously form in the membrane, as judged by the high efficiency of cysteine crosslink between two SecE subunits when the complex is overproduced [Veenendaal et al., 2001; Deville et al., 2011]. Recent single molecule analysis in reconstituted liposomes shows that a crosslinked back-to-back dimer is functional, whereas its monomeric counterpart can interact with preprotein but displays only a fraction of the activity of the dimer [Deville et al., 2011]. The work presented in Chapter 3 will also provide strong support that a dimeric channel assembly in this orientation is functional in translocation. Recent in-silico models of the back-to-back mode of association have been constructed based on the atomic resolution structure of the SecYEG-SecA complex [Gold et al., 2010; Deville et al., 2011]. These models are consistent with various cysteine  19  crosslinking analyses that have identified the contact points between SecY and SecA in the translocation active complex [van der Sluis et al., 2006; Osborne and Rapoport, 2007; Tsukazaki et al., 2008]. In the back-to-back model (Figure 1-6), the passive SecY unit would provide an additional surface for the binding of SecA [Zimmer et al., 2008]. It is interesting that the acidic phospholipid cardiolipin facilitates the dimerization of SecYEG in detergent solution, but also strengthens the binding of SecA to the channel and enhances SecA ATPase activity in the membrane [Gold et al., 2010]. It is possible that the SecY dimer bound to this particular lipid creates an optimal surface for the binding and activation of SecA [Gold et al., 2010]. The back-to-back orientation might not be an exclusive mode of interaction and other conformations may exist -or become stabilized- upon binding of the ribosome and polypeptide substrates. For example, a “front-to-front” orientation can be fitted into the low resolution cryo-EM structure of the SecY-ribosome complex [Mitra et al., 2005], and this mode of association seems supported by a recent in vivo crosslinking analysis which utilized a photo-reactive probe incorporated into SecA [Das and Oliver, 2011].  20  Figure 1-6: Representation of contact points between SecA and the back-to-back SecY dimer. The structures of SecY and SecA are derived from the SecYEG-SecA crystal structure (PDB: 3DIN; Zimmer et al., 2008). Proteins and domains are coloured as follows: SecY (grey); SecE (yellow); SecG (orange); SecA (blue); SecA PPXD domain (purple); SecA NBD1 domain (brown); space filling representation of amino acids (green). All residues are numbered according to E. coli proteins. Active SecY copy - The cytosolic loops C4 and C5 of the active SecY copy interact with the SecA PPXD domain [Zimmer et al., 2008] but for simplicity, only a few interactions are represented here. Residues in the PPXD domain of SecA that are in proximity to the SecY cytosolic loops were selected if within a distance of 4 Å from the indicated SecY residue. For example, residue R255 on the C4 loop of SecY is close to residues S350-D351 of SecA, whereas residue K348 on the C5 loop is close to the SecA residues Y299-S300. The position R357 is not shown because this residue does not interact with SecA in the SecYEG-SecA crystal. Passive SecY copy – The location of the cysteine crosslinks between the passive SecY copy and SecA that were identified in Osborne and Rapoport (2007) are shown. The residue R255 in the C4 loop of the passive SecY copy establishes crosslinks with the positions A48, E55, A163 and K202 in the nucleotide binding domain 1 (NBD1) of SecA. This figure was modified from Dalal and Duong, 2011.  21  1.7  The ion gate-keeping activity of Sec channels The oligomerization of the SecY complex and mechanism of preprotein transport are  not the only outstanding questions regarding translocation. Sec complexes are also channels through which ions and other small molecules may pass, and this process influences the electrochemical gradient across bacterial and eukaryotic cells.  1.7.1  Maintaining the integrity of the membrane for small molecules Recent studies have addressed the problem of membrane permeability during protein  transport, a notion particularly important in bacteria since the membrane needs to maintain a stable electrochemical gradient. Pioneering studies showed that synthetic signal sequence peptides can open large ionic conducting channels, whereas channels jammed with a translocation intermediate during SecA-driven translocation were leaky for chloride ions [Simon and Blobel, 1992; Schiebel and Wickner, 1992]. The latter observation is still puzzling because the SecY channel bearing an arrested nascent chain (emerging from the ribosome exit tunnel) suggested that the pore ring, made of six hydrophobic side chains that project toward the channel interior, creates a tight seal around the polypeptide chain, preventing the permeation of small molecules [Park and Rapoport, 2011]. The lateral sidechains contributing to a membrane seal are indeed within disulphide bridge distance to the polypeptide chain during translocation [Cannon et al., 2005].  1.7.2  The role of pore ring and plug in SecY ionic conductance The role of the pore-plug assembly in the gating of the resting channel is better  characterized than the role of individual side chains in the pore ring. Electrophysiology  22  experiments in planar lipid bilayers have shown that a partial deletion of the plug domain, or locking the plug in its open conformation, is sufficient to render the channel ion-conductive [Saparov et al., 2007]. In vivo, the stabilization of the plug domain in its open state is lethal, suggesting that the pore ring on its own is insufficient to seal the channel [Harris and Silhavy, 1999; Maillard et al., 2007]. Certain amino-acid substitutions in the pore ring produce ion conductive channels, most likely because these mutations destabilize the interactions that keep the plug in its closed position [Saparov et al., 2007; Dalal and Duong, 2009]. In Chapter 4, we show these leaky channels are specific for monovalent anions only [Dalal and Duong, 2009; Dalal et al., 2010], which could explain why they can be produced in the bacterial membrane [Li et al., 2007]. As it will be discussed later, a selective ion conductance may allow the cell to tolerate leaky SecY complexes.  1.7.3  Comparison of the conductance properties of SecY and Sec61 channels The channel characteristics of the eukaryotic Sec61 complex are quite different from  the bacterial counterpart, and illustrate the diversity of membrane permeability in different organisms. Yeast can tolerate Sec61 channels with defective and even missing plug domains [Junne et al., 2006; Junne et al., 2007]. The presence of permanently open channels may be less critical for yeast growth because the solute composition across the ER membrane may be quite similar. In fact, electrophysiology experiments indicate that the resting Sec61 complex conducts a diversity of small molecules, including Ca2+ [Le Gall et al., 2004; Wonderlin, 2009; Erdmann et al., 2009; Erdmann et al., 2010]. A recent study has identified calmodulin as a binding partner of Sec61α [Erdmann et al., 2011]. Molecular modeling suggests that calmodulin may fit into an open cavity that normally exists at the junction between the  23  ribosome and the Sec61 channel. Binding of calmodulin at this interface would regulate the Ca2+ conductance across the channel. On the opposite side of the membrane, the ER chaperone protein BiP (binding immuno-globulin protein) may serve to seal the luminal face of the Sec61 channel both at rest and during the early stages of protein translocation [Hamman et al., 1998]. It remains to be determined if any SecY binding partners can perform a similar function in sealing the bacterial membrane.  1.8  Quality control mechanisms: interaction between SecY and Syd Channels that are constitutively open or jammed with a polypeptide substrate are  lethal in E. coli. The FtsH protease degrades the SecY channels that are non-assembled with SecE [Kihara et al., 1995], or blocked with a translocation intermediate [van Stelten et al., 2009]. In addition to this proteolytic mechanism that eliminates defective and potentially toxic channels, some Gram-negative bacteria contain a small protein termed Syd, which verifies the proper assembly of the SecY complex in the membrane. Syd was originally isolated as a suppressor of a dominant-negative mutation into SecY capable of sequestering SecE into an inactive complex (hence the name suppressor of SecY dominance) [Shimoike et al., 1995]. Accordingly, Syd interferes with protein translocation only when the channel displays abnormal (i.e. weakened) SecY-SecE associations [Matsuo et al., 1998]. These observations have suggested that Syd belongs to a quality control system that proofreads the SecY complex, leading to its degradation by the FtsH protease when the complex is compromised by abnormal heterotrimeric associations. Our work in Chapter 2 will show how the Syd protein forms electrostatic interactions with the SecY cytosolic loops, and also will provide further evidence that Syd targets and disassembles malformed SecY complexes.  24  1.9  The SecY general docking platform The SecY electropositive loops appear to be a general docking platform for all  channel partners, including SecA, Syd and the ribosome. The structures of Syd and the SecYEG-SecA complex serve to highlight an electrostatic mode of binding of these partners to the channel. In particular, the C4 and C5 cytosolic loops on SecY and Sec61α contain several conserved arginine and lysine residues (position K250, K268, R340, K347, K348 and R357; Figure 1-7, left). The SecY mutation R357E, which introduces an acidic residue into the C5 electropositive loop, disrupts the interactions with the negatively charged ribosomal RNA [Menetret et al., 2007] and causes defects in the activation of SecA by SecY [Mori and Ito, 2001]. We will show in Chapter 2 how deletion of R357 also reduces the binding affinity of Syd for the channel. The SecYEG-SecA crystal structure from T. maritima reveals how the cytosolic loops of SecY establish contacts with the PPXD domain of SecA [Zimmer et al., 2008]. Analysis of the electrostatic potential on the surface of SecA reveals discrete areas of negative charge density in the PPXD domain (Figure 1-7, middle). In the SecYEG-SecA crystal structure, the positively charged residues on the C4 and C5 loops of SecY are within salt bridge distance (~4 Å) to the negatively charged regions of SecA. For example, the conserved basic residues at positions 250 and 255 in the C4 loop of SecY interact with SecA near the conserved serine and aspartate at positions 350-351. The conserved arginine at position 357 of SecY does not bind directly to SecA, however other basic residues exist in the C5 loop, such as R348, which are located near the negatively charged region of the PPXD domain containing a tyrosine and serine at positions 299-300. These observations suggest that electrostatic interactions play an important role during the binding of SecA to SecYEG.  25  The crystal structure of Syd presented in Chapter 2 reveals the presence of a concave electronegative groove that could form electrostatic interactions with the cytosolic loops of SecY (Figure 1-7, right). The interaction between the SecY C4 loop and the concave surface of Syd was shown by disulphide crosslinking analysis to occur between positions 255 of SecY and position 115 of Syd [see Chapter 2].  Figure 1-7: Electrostatic interactions between SecY and SecA, and SecY and Syd. The SecY complex, SecA and Syd were analysed using the pymol software (version v0.99) and the electrostatic representation was generated with the APBS plug-in. Blue and red colours represent electropositive and electronegative potential, respectively. The surface potential was set between -4.0 and +4.0 kT/e for the SecYEG complex, and between -2.5 and +2.5 kT/e for SecA and Syd. The solvent-accessible area option of the software was employed in the calculation. Surface representations were constructed from the T. maritima SecYEG-SecA crystal (PDB: 3DIN; Zimmer et al., 2008) and E. coli Syd (PDB: 3FFV; Dalal et al., 2009). All residues are numbered according to E. coli proteins. The residues coloured in yellow where chosen to highlight the possible electrostatic interactions taking place between SecY and SecA, or SecY and Syd (see text). (Left) On the SecY protein, conserved basic residues were highlighted to show the positive charge density on the C4 and C5 cytosolic loops. (Middle) The indicated residues in the SecA PPXD domain lie in regions of negative charge density that are in proximity to the SecY cytosolic loops. (Right) the amino acid at the position 115 on the electronegative and concave surface of Syd interacts with residues located in the cytosolic loop C4 of SecY. This figure was modified from Dalal and Duong, 2011.  26  1.10 Overview of objectives Several outstanding questions remain regarding the SecY complex: the functional oligomeric state of the channel, the binding stoichiometry of SecY with its binding partners, the mechanism of protein translocation, and the importance of ionic conductance across the channel.  These issues stem from the high dynamic nature of the SecY complex in  membranes, as well as the surprising flexibility of the channel to accommodate a diverse range of preprotein substrates, small molecules and ions. To resolve these questions, three main projects in regard to protein translocation through the SecY complex were undertaken in the laboratory:  1) How does the SecY complex interact with its binding partners, Syd and SecA, and with what stoichiometry? 2) What is the functional quaternary organization of the SecY complex, and its implications for the mechanism of SecA-driven protein translocation? What is the role of lipids in this process? 3) How is gating for small molecules achieved by the SecY channel, and how does it compare to the properties of the Sec channel in organisms other than bacteria? What are the implications of SecY channel ionic conductance for the cell?  To address the first question, we analyzed the bacterial SecY complex from E. coli with supported nano-scale lipid bilayers termed Nanodiscs. This method allows controlling the SecY oligomeric state and number of lipid molecules, creating the ideal platform to study 27  the binding and stoichiometry of Syd, SecA and preprotein. Nanodiscs were analyzed with gel electrophoresis and multi-angle light scattering to investigate the binding stoichiometry with the SecY channel partners. The crystal structure of Syd was solved, and the binding interface with SecY was mapped by cysteine crosslinking. We also investigated the physiological role of Syd by observing its effect on the channel structure and activity in vitro. Next, we incorporated both monomers and dimers of the SecY complex into Nanodiscs. Both oligomeric forms of the channel were tested for SecA and preprotein binding, and for the ability to stimulate the preprotein dependent SecA ATPase. Since the immediate environment around the SecY complex can be controlled with Nanodiscs, the contribution of lipids to SecA binding and activation was also studied. A cysteine crosslinking approach was then used to determine the orientation of the two SecY copies within the discs or in membranes. We did not limit our analysis to Nanodiscs, and test the function of SecY dimers in the physiological context. In addition to clarifying several questions about the quaternary assembly of the SecY complex, this study also highlights Nanodiscs as a powerful tool for investigating membrane protein structure, function and oligomerization. Finally, the ionic conductance properties of the SecY channel were tested using inverted inner membrane vesicles (IMVs) and voltage sensitive dyes. By exposing the SecY channel to various salts, we could determine if the channel conducts only certain ions during protein transport. Mutagenesis of SecY pore ring and plug structures also allowed us to identify specific residues that contribute to maintaining a membrane seal. We discuss the implications of the SecY ionic conductance for the bacterial cell.  28  Chapter 2: The interaction between SecY and its binding partners Syd and SecA 2.1  Introduction The interactions between the protein-conducting channel and its binding partners  involve the large electropositive cytosolic loops of the SecY subunit (Figure 1-6 and Figure 1-7). Several residues in these loops have been identified as being indispensible for channel function, or for the recruitment of the ribosome, SecA and Syd (see Section 1.9). Although the interactions between the channel and these components has been somewhat characterized, the stoichiometry and dynamics of their associations are not entirely clear, and must be resolved for a complete understanding of the translocation reaction. As the crystal structure of Syd appeared in 2009 (this study), its interaction with the SecY complex could only be previously inferred by genetic or biochemical evidence [Shimoike et al., 1995; Matsuo et al., 1998].  Syd is a non-essential and hydrophilic protein  of 181 amino-acid residues that exists only in Gram-negative bacteria. Ito and co-workers [Shimoike et al., 1995; Matsuo and Ito, 1998] originally identified its gene as a multi-copy suppressor of the dominant-negative secY_d1 mutation (thus termed Syd for suppressor of SecY dominance) (see Section 1.8). The secY_d1 mutant complex exerts a dominant negative effect by sequestering SecE into inactive complexes in the bacterial membrane, even in the presence of the wild type SecY allele on the chromosome [Shimoike et al., 1995]. In addition, Syd was shown to interfere with protein translocation, but only in cells in which the SecY and SecE subunits interact weakly [Matsuo et al., 1998]. Despite the ability of Syd to influence translocation, its physiological role remains uncertain, especially as the deletion of the gene causes no obvious phenotype [Shimoike et al., 1995].  29  In contrast, the role of SecA in protein translocation is well understood, and the SecA structure has been extensively studied by crystallography [for review see Sardis et al., 2010]. Questions remain, however, regarding the mode of SecA binding with the SecY complex. SecA exists mostly as a dimer in solution [Woodbury et al., 2002], but its oligomeric state after binding to the membrane or to the SecY complex during the translocation cycle is controversial; it is proposed to be monomer or dimer depending on the study [Or et al., 2002; Or et al., 2005; de Keyzer et al., 2005; Osborne and Rapoport, 2007; Kusters et al., 2011]. As implied by models of the SecYEG-SecA association depicting the SecY dimer bound to a single SecA molecule (Figure 1-6), the stoichiometry between these two proteins has deep implications for the mechanism of translocation. In the work presented in this chapter, the association between the channel and its binding partners Syd and SecA are shown to occur in a 1:1 stoichiometry using Nanodiscs, an emerging technology [Bayburt and Sligar, 2003] that allow the preparation of nanometersized soluble particles containing phospholipids and the SecYEG complex. Nanodiscs faithfully recreate a lipid bilayer and thus permit investigating the reactivity and interaction of Syd or SecA with the SecYEG complex in a near native environment, without use of detergent. We also solve the crystal structure of Syd and show that the protein most likely makes electrostatic contacts with the cytosolic loops of the SecY channel. Using Nanodiscs with a larger diameter, it is also possible to capture two SecY complexes within the disc. This SecY dimer interacts with single SecA molecule, supporting the models shown in Figure 1-4A and Figure 1-6. Whereas the protein translocation activity of the SecY dimer bound to SecA is discussed in detail in Chapter 3, the physiological role of Syd as a quality control mechanism for the SecY complex is described here.  30  2.2 2.2.1  Materials and methods Plasmids and biological reagents The Syd open reading frame was PCR-amplified from the E. coli genome and cloned  into the expression vector pET23a (Clontech) using the restriction sites NdeI and XhoI. The plasmid pET23a-hisSecA was previously described [Duong, 2003]. The plasmid pBAD22hisEYG and purification of the SecY complex are described in Appendix A. The deletion ∆251–258 and ∆354–357 into SecY and ∆7–67 into SecE was obtained by PCR amplification using primers introducing a BglII restriction site on each side of the deletion. MSP1 (24.6 kDa) and MSP3 (32.6 kDa) membrane scaffold proteins were previously described [Denisov et al., 2004; Ritchie et al., 2009; see Section 3.2.1 for description].  2.2.2  Purification of Syd Plasmid pET23-Syd was transformed in E. coli strain BL21 (DE3). Overproduction  of Syd was initiated at OD600 nm ~ 0.5 with 1 mM isopropyl 1-thio-β-D-galactopyranoside for 3 h. Cells were collected in TSG buffer (50 mM Tris-Cl, pH 7.5; 50 mM NaCl; 10% glycerol; no DTT) and lysed with a French press (8,000 p.s.i., three passes). After centrifugation (100,000 x g, 1 h at 4 °C), the supernatant was applied onto a Ni2+-chelated Sepharose column (GE Healthcare) equilibrated in TSG buffer. Syd was eluted with 500 mM imidazole and applied onto a 5-mL Q-Sepharose Fast Flow column (GE Healthcare) equilibrated in TS buffer (50 mM Tris-Cl, pH 7.5; 50 mM NaCl; 1 mM DTT). Syd was eluted with 250 mM NaCl and concentrated to 40 mg/mL using an Amicon 5-kDa centrifugation device. For selenomethionine labelling, cells were grown in 9 liters of M9  31  media. At OD600 ~ 0.3, amino-acids were added (L-lysine, L-phenylalanine, L-threonine, Lisoleucine, L-leucine, L-valine, and L-selenomethionine; each 50 mg/liter), and Syd expression was induced with 1.5 mM isopropyl 1-thio-β-D-galactopyranoside (IPTG) during 16 h at 30 °C. 125I labelling was performed using IODO-GEN-coated tubes containing 60 µg of Syd and 25 µCi of Na125I. The reaction was quenched with 5 mM DTT, and the protein was desalted through a G-25 spin column equilibrated in TSG buffer. 125I-Syd (~2 x 105 cpm/µg) was stored at -80 °C.  2.2.3  Purification of SecA His-tagged SecA was expressed in E. coli BL21(DE3) and isolated by Ni2+-chelating  chromatography [Duong, 2003] and further purified on a Superdex 200 HR10/30 column (Amersham Biosciences) equilibrated in TSG buffer (50 mM Tris pH 7.9, 50mM NaCl, 10% glycerol and 1mM DTT).  2.2.4  Crystallization and structure determination of Syd Crystallization and structure determination of Syd was performed by Gary D. Brayer  and coworkers (see Preface). Native E. coli Syd was crystallized using the hanging drop vapour diffusion method in 0.8–1.0 M sodium citrate, 0.2 M sodium chloride, and 0.1 M Tris, pH 7.0. The starting protein concentration was 20 mg/mL. Crystals reached dimensions of up to 0.45 X 0.40 X 0.25 mm. Selenomethionine E. coli Syd crystallized under the same conditions, although the resultant crystals were smaller (up to 0.2 X 0.18 X 0.1 mm). All crystals were cryo-cooled in the presence of 30% sodium malonate, pH 7.0, prior to x-ray data collection at the Stanford Synchrotron Radiation Laboratory (Stanford, CA). Details of  32  data collection, processing, and structural refinement statistics are given in the supplemental material. Each dataset was integrated and scaled using the programs MOSFLM and SCALA, respectively [CCP4, 1994]. The structure of Syd was determined by the single wavelength anomalous diffraction method. Two selenomethionine sites, one each from the two Syd molecules in the asymmetric unit, were located and refined using the program SOLVE, and phases were subsequently improved by density modification using the program RESOLVE [Terwilliger and Berendzen, 1999]. The program ARP/WARP was used to build the initial model [Perrakis et al., 1999]. The primary sequence used in structure refinement was that from GenBankTM (accession number ABE08615). The program REFMAC5 [CCP4, 1994] was used to refine the initial model using the native Syd dataset. This involved iterative cycles of fitting and rebuilding using the COOT program [Emsley et al., 2004]. Further structural refinement was conducted using the native Syd dataset and CNS [Brünger et al., 1998]. Note that the 6-histidine tag attached to each Syd molecule in the asymmetric unit was not observed in electron density maps and therefore was not included in the structural model. Notably, the structure determination clearly indicates the presence of a disulfide bridge between cysteines 147 and 154 in Syd. These two cysteines are ideally positioned with respect to one another within the polypeptide chain fold to form this linkage. However, for one of the molecules of Syd in the asymmetric unit, there is some evidence from electron density maps that this disulfide bridge may be broken in ~20% of the crystallized molecules. The other two cysteines present in Syd are too far removed from one another to form a disulfide bridge. Structures were visualized using Pymol (version 0.99), and electrostatic maps were obtained using the APBS plug-in (version 1.0.0). Atomic charges and radii were  33  generated with the AMBER option at the Protein Data Bank code 2PQR on-line service. Refinement statistics for the structural determination are given in Table 2-1. Table 2-1: Summary of structure determination statistics for Syd  Se-Met anomalous dataset  Native dataset  0.9788  0.9116  R3  R3  Cell dimensions (Å)  a = b = 167.8, c = 40.9  a = b = 167.6, c = 41.6  Resolution range (Å)  83.9 ~ 2.0  ∞ - 2.0  99.6 (99.6)  99.8 (99.9)  Data collection Wavelength (Å) Space group  Completeness within range (%)† Multiplicity  5.6(5.4)  Mean I/σI † Merging R-factor (%)  †  21.2 (5.8)  19.3 (8.8)  7.9 (27.0)  6.8 (15.1)  Phasing Resolution (Å)  50 ~ 2.2  Figure of merit  0.35 (0.71) 1  2  7.7  Z-score  N/A  Structure Refinement Values Number of reflections  29401  Resolution range (Å)  50-2.0  Completeness within  99.8 (99.9)  range (%)† Number of protein atoms  2920  Number of solvent atoms  509  Average thermal factors (Å2) Protein atoms  25.0  Solvent atoms  40.7  34  Final R-factors Final R-free value (%)‡  23.0  Final R-factor (%)  18.2  Final Structure Stereochemistry  (r.m.s. deviations)  bonds (Å)  0.008  angles (°)  1.4  Ramachandran Plot for Syd Highly favored (%)  90.5  Allowed (%)  9.5  Generously allowed (%)  0.0  Disallowed (%)  0.0  1  The value in parentheses is after density modification by RESOLVE. As determined by the program SOLVE. † Values in parentheses are for the highest resolution shell: 2.0 – 2.11 Å. ‡ For the R-free test, 5% of the diffraction data were kept aside. This table was modified from Dalal et al., 2009. 2  35  2.2.5  Analytical gel filtration Analytical gel filtration was performed using a Superdex 200 HR 10/30 column  (Amersham Biosciences) connected in-line to miniDAWN multi-angle light scattering equipment coupled to an interferometric refractometer (Wyatt Technologies). Data analysis was recorded in real time using the ASTRA software (Wyatt Technologies). Molecular masses were calculated using the Debye fit method.  2.2.6  Preparation of inner membrane vesicles (IMVs) and in vitro protein  translocation experiments Inverted inner membrane vesicles were prepared by growing E. coli BL21(DE3) cells harbouring pBAD22 plasmids encoding for wild-type and mutant SecY complexes (OD600 ~ 0.4), and then inducing expression with 0.2% arabinose for 1.5 hours. Cell lysis was performed by passage through a French press (three passes, 8000 p.s.i.) in TSG buffer. Breaking the cell under high pressures causes membrane vesicles to become inverted, exposing the cytosolic face of membrane proteins to the external environment. Cell lysates were spun at low speed in a Beckmann JA.25 rotor (5000 x g, 10 min, 4°C) to remove unbroken cells, followed by high speed in a Beckman Ti60 ultracentrifuge rotor (200,000 x g, 45 min, 4°C) to pellet crude membranes. Outer and inner membranes were separated on a step-wise sucrose gradient of 70, 50 and 20% sucrose solutions (in water) by layering crude membranes on top of the gradient and centrifugation in a Beckmann SW41 swinging bucket rotor (234,000 x g, 4 hours, 4°C). The inner membrane fraction was collected and diluted 3 fold in TSG, and then spun again in a Beckmann Ti60 rotor (200,000 x g, 30 min, 4°C). The  36  pelleted IMVs were resuspended in TSG buffer at a membrane protein concentration of approximately 4 g/L as judged by the Bradford Reagent. In vitro protein translocation experiments were performed as previously described [Dalal and Duong, 2009]. In a typical experiment, 5 µg of IMVs are mixed with 25,000 c.p.m. 125I-proOmpA (or other preprotein substrate), 0.5 µg of SecA, 0.2 g/L BSA in a final reaction volume of 50 µL of TL buffer (50 mM Tris-HCl, pH 7.9, 50 mM KCl, 50 mM NaCl, 5 mM MgCl2 and 1 mM DTT). Translocation is initiated with ATP (1 mM) for 10 min at 37 °C, which transports the radio-labeled preprotein through the SecY complex into the lumen of the vesicles. To degrade any non-transported substrate, ~10 µg proteinase K is added to the reaction mixture for 15 min on ice. The entire mixture is then precipitated with 17% ice cold TCA for 30 minutes on ice, followed by centrifugation for 10 minutes at 13,000 x g at 4°C in a micro-centrifuge. Pellets containing precipitated proteins are washed with acetone then solubilized with SDS-PAGE sample buffer. Samples are analyzed by SDS-PAGE and phosphorimaging. The relative intensity of the protein band after autoradiography indicates the total amount of protein translocation taking place across IMVs. The translocation of proOmpA normally generates two bands: the full length protein (upper band) and a version lacking the N-terminal signal sequence after proteolytic processing of leader peptidase in the inner membrane (lower band).  2.2.7  Preparation and iodogenation of preprotein substrates Preprotein substrates such as proOmpA were expressed in E. coli (pET23a plasmid)  and purified by metal affinity chromatography (6-Histidine tag) from inclusion bodies in 6M urea as described in [Tam et al., 2005; Alami et al., 2007]. Prior to labelling preprotein  37  substrates, 50 µL of 1 g/L IODO-GEN in chloroform was dried inside a 1.5 mL microcentrifuge tube under a gentle N2 stream. 60 µg of proOmpA (or other substrate) in 50 mM Tris-HCl, pH 7.9, 4 M urea was mixed with 25 µCi of Na125I and incubated for 10 minutes on ice. The reaction was quenched by adding 2 mM DTT, and excess 125I was removed by desalting on a home-made 1 mL G-25 column equilibrated in 50 mM Tris-HCl, pH 7.9, 4M urea, 1 mM DTT. Detection of 125I-labeled proteins and densitometry scanning were performed using a phosphorimager scanner.  2.2.8  Nanodisc reconstitutions and other methods The preparation of Nanodisc reconstituted SecYEG and analysis by native-PAGE is  described in Appendix A. Briefly, reconstitution of the detergent purified SecY complex into MSP1 were performed at a molecular ratio SecYEG:MSP1:lipids of 1:5:40. Reconstitutions of the SecYEG with MSP3 were performed at a molecular ratio of 1:3:20. The covalently linked SecYEG dimer, fusing the N terminus of a first SecY molecule with C terminus of a second (termed SecYY, Duong, 2003), was reconstituted into MSP3 at a molecular ratio of 1:5:40. Reconstitutions were mixed in 50 mM Tris-HCl pH 7.9, 300 mM NaCl and 5% glycerol supplemented with 0.08% DDM (n-Dodecyl β-D-maltoside), and detergent removal was performed overnight with 100 µL BioBeads per 300 µL reaction volume. The preparation of urea-stripped IMVs, as well as the conditions for blue-native gel electrophoresis was described previously [Bessonneau et al., 2002; Alami et al., 2007]. Procedures for making and running native gels are also given in Appendix A (and see A.15 for description of running Blue-native gels).  38  2.3 2.3.1  Results Reconstitution of the SecY complex in Nanodiscs and binding to Syd or SecA  proteins Nanodiscs permit investigation of the biochemistry of the SecY complex, without using liposomes or detergent [Alami et al, 2007; Dalal and Duong, 2010]. Each particle, also referred to as Nd-SecYEG or Nd-Y, is made of a single SecY complex embedded in a small patch of lipid bilayer supported by two membrane scaffold proteins (For cartoon representation see Appendix B). Although water-soluble, the Nd-Y particles (~125 kDa) were smeared on native (non-denaturing) PAGE gels (Figure 2-1A), but migrated as a single band when incubated with purified Syd protein, which possibly binds to SecY and modifies the electrophoretic characteristics of the Nanodisc (Figure 2-1A). Syd protein labelled with Iodine-125 migrated at the bottom of the native-PAGE gel but shifts to a higher molecular weight position when incubated with Nd-Y, indicating a complex between the two proteins (Figure 2-1B). When the SecY subunit carries small internal deletion in one or the other of its largest cytosolic loops, C4 and C5, respectively, the binding of 125I-Syd appears reduced (Figure 2-1B, and see Section 1.9 for discussion of the SecY cytosolic loops). The stability of the Nd-SecYEG-Syd complex was further assessed by sucrose density ultracentrifugation. The 125I-Syd protein, which floats at the top of the sucrose gradient, sediments to a higher density in the presence of Nd-Y (Figure 2-1C). The sedimentation is only partial when the SecY subunit carries the internal deletion in either C4 or C5 loops (Figure 2-1C). Interestingly, the absence of salt in the buffer increases the stability of the SecYEG-Syd complex (Figure 2-1C, bottom panel), suggesting that electrostatic forces mediate the interaction.  39  440 232 132 66  Figure 2-1: Binding of Syd onto the SecYEG-Nanodisc. (A) The indicated amount of Syd was incubated with 3 µg of Nd-SecYEG in TSG buffer (5 min at room temperature). The samples were analyzed by native-PAGE followed by Coomassie Blue staining of the gel. The molecular weight markers (MWM) are (top down) ferritin (440 kDa), catalase (232 kDa), and bovine serum albumin (66/132 kDa). The Nd-SecYEG preparation contains some free membrane scaffold protein (MSP) dimers that migrate near the 66-kDa marker. (B) 125I-Syd (∼25,000 cpm; 150 ng) was incubated with NdSecYEG, either wild-type or carrying the indicated deletion in the SecY subunit. Samples were analyzed by native-PAGE and phosphorimaging. (C) 125I-Syd (∼250,000 cpm; 1.25 µg) was analyzed by sucrose density centrifugation with the SecYEG-Nanodiscs (∼35 µg), either wild-type (WT) or carrying the indicated deletion in the SecY subunit. The gradient (6–13%) was prepared in 50 mM Tris, pH 7.9, 5% glycerol, 1 mM DTT containing 100 mM NaCl, or no salt as indicated. The centrifugation was at 197,000 × g for 16 h at 4°C in a Beckman SW41 rotor. Equal fractions were collected and analyzed by SDS-PAGE and phosphorimaging. This figure was modified from Dalal et al., 2009.  40  In addition to forming a stable complex with Syd, the Nd-Y particles also interact with SecA which facilitates the electrophoretic mobility of the discs on native-PAGE (Figure 2-2, lanes 2 and 3; Alami et al., 2007; Dalal and Duong, 2010). Since the SecY channel exists as an oligomer in detergent solution, we attempted to capture the SecY dimer using the longer membrane scaffold protein MSP3, which extends the disc diameter from ~9.7 nm to ~12.1 nm [Ritchie et al., 2009] (see Appendix B for cartoon representation). Following the same reconstitution protocol, two distinct populations of discs were obtained (Figure 2-2, lane 4) and each was able to form a stable complex with Syd or SecA (lanes 5 and 6, see Appendix C for cartoon representation of Nd-Y-SecA complexes). To ascertain that the higher molecular weight species corresponded to discs containing two SecY complexes (termed Nd-Y2), we loaded on the same gel a covalently-linked SecY dimer (also referred to as fused SecY dimer, see Section 2.2.8) reconstituted with MSP3 (termed Nd-YY; Figure 22, lanes 7-9). Collectively the results indicate that Nanodisc embedded SecY monomers and dimers can form a stable interaction with Syd or SecA, possibly through electrostatic interactions with the SecY cytosolic loops.  41  Figure 2-2: The SecY monomer and dimer in Nanodiscs binds to SecA. The SecY monomer (Y), the dimer (Y2) and the fused dimeric version (YY) were reconstituted in Nanodiscs (Nd) with the indicated membrane scaffold protein (MSP). The Nanodiscs (3 µg each) containing one SecY (labelled Nd-Y), two SecY (labelled Nd-Y2) or the fused dimer (labelled Nd-YY), were incubated with SecA (labelled A2) and Syd (2 µg each) followed by native-PAGE and Coomassie blue staining. SecA and Syd bind to the SecY complex with high affinity which greatly improves the electrophoretic mobility of the discs [Dalal and Duong, 2010]. The protein samples were analyzed on the same gel but to facilitate figure labelling and to compare the relative migration, lanes 5 and 6 on left panel are duplicated on right panel.  42  2.3.2  The SecY complex binds to Syd or SecA in a 1:1 molar ratio The oligomeric state of Syd and the stoichiometry between Syd and Nd-SecYEG was  assessed by analytical gel filtration. In the presence of an excess of Syd, most of the NdSecYEG is complexed as judged by the shift and the symmetry of the elution peak during gel filtration chromatography (Figure 2-3A). Multi-angle light scattering analyses indicate that the purified Nd- SecYEG-Syd complex is mono-disperse with a measured molecular mass of 152 kDa ± 3%. On their own, the Nd-SecYEG particles and the Syd protein are also homogeneous but with a measured molecular mass of ~126 ± 5%. and ~23 kDa ± 5%., respectively (Figure 2-3B). Thus, the Syd protein is monomeric in solution and forms a stable 1:1 stoichiometry with the SecYEG complex. We previously reported that the Nd-Y complex (i.e. the SecYEG monomer) could bind to a single molecule of SecA [Alami et al., 2007]. To determine the stoichiometry of the SecY dimer with SecA, SecA was stabilized as a dimer with a pair of intermolecular disulphide bridges [termed SecACP3; Alami et al., 2007]. This linked SecA dimer formed a complex with Nd-Y2 (Figure 2-3C, lane 2) and Nd-YY (Figure 2-3C, lane 5), but compared to native SecA, each assembly had a higher molecular weight (Figure 2-3C compare lanes 13 and lanes 4-6; Appendix C). Thus, one SecY forms a binding site for one SecA, and two SecY still form a binding site for one SecA.  43  Figure 2-3: Stoichiometry between the SecY complex and Syd or SecA. (A) SecYEG-Nanodisc (∼200 µg) incubated with a molar excess of Syd (∼100 µg) and then applied onto a Superdex 200 HR10/10 column equilibrated in TSG buffer. The fractions containing the complex Nd-SecYEGSyd were pooled and concentrated to 1.2 mg/mL for subsequent multi-angle light scattering analysis. (B) Multiangle light scattering of Syd, Nd-SecYEG, and Nd-SecYEG-Syd. In each case, about 100 µg of protein is loaded onto a Superdex 200 HR 10/30 column equilibrated in 10 mM HEPES, 50 mM NaCl, pH 7.4. (C) The non-fused or fused SecY dimer (3 µg of Nd-Y2 and Nd-YY, respectively) was incubated with the disulfidelinked SecA dimer (SecACP3, 4 µg). In lane 3 and 6, the sample was incubated with 1 mM DTT (37°C for 2 min) before loading on the gel. Native SecA is dimeric in solution (~204 kDa) and migrates next to the 201 kDa marker (lanes 7 and 9). On its own, the crosslinked SecA dimer migrates as a smear (lane 8). Expected molecular weights: Nd-Y2/YY + SecA - 318 kDa; Nd-Y2/YY + SecACP3 - 418 kDa. These values do not include the contribution of lipid molecules within the disc. A and B This were modified from Dalal et al., 2009.  44  2.3.3  Crystal structure of Syd Syd was crystallized and its structure solved by single wavelength anomalous  dispersion to 2.0 Å resolution (Figure 2-4A). There are two molecules in the asymmetric unit, and their refined structures include all residues, except for the C-terminal 6-histidine tag used to purify the protein. The Syd structure is relatively compact and globular, being primarily comprised of a six-stranded antiparallel β-sheet and two long α-helices. A spacefilling representation of Syd reveals that part of this β-sheet structure forms a concave cavity (or groove) on the protein surface. Sequence alignments show that the residues forming the concave surface of this groove are highly conserved (Figure 2-5). The periphery of the observed groove is made of two protruding stalks located at the N terminus (residues 30–34; 45–48) and C terminus (residues 138–143) of the protein. Examination of surface electrostatics indicates that Syd has a distinct negatively charged patch localized on the concave groove surface and its surrounding stalks (Figure 2-4B). A similar electrostatic analysis of the SecY complex shows that the two SecY loops involved in the binding of Syd are electropositive, because of several conserved basic residues in those regions (Figure 24C). These observations strongly suggest that the negatively charged groove found in Syd forms the binding site for the SecY electropositive loops.  45  Figure 2-4: Crystal structure and electrostatic surface of Syd. (A) Protein crystallization and structure determination of Syd was performed by Gary. D. Brayer and coworkers (see Preface). Ribbon diagram representation of Syd (β-sheets, yellow; α-helices, magenta; loops, green) and space-filling representation showing the concave structure of the protein (180° rotation compared with right panel). Arrows denote the position of the “stalk” regions that consist of negatively charged protruding loops that delineate the concave region. (B) Electrostatic potential of the concave (left) and convex surface (right, 180° rotation). Blue and red represent electropositive and electronegative potential, respectively. The surface potential was set between -2.5 and +2.5 kT/e using the solvent-accessible area option of the software. (C) Electrostatic map of the archaeal SecY complex (Protein Data Bank code 1RH5) showing the position of the positively charged loops C4 and C5. The surface potential was generated using the parameters described in B. Positive charges are located near the predicted location of the phospholipids head group. Note that in E. coli, the loops C4 and C5 contain additional basic residues compared with Methanococcus jannaschii. In particular, three arginine residues are located at the tip of the loop C4 in the E. coli complex. This figure was modified from Dalal et al., 2009.  46  Figure 2-5: Database retrieval and multiple sequence alignment of Syd. All proteomic data were retrieved from NCBI’s RefSeq database. Syd exists only in Gram-negative bacteria which express SecE proteins with 3 TMS (see Section 2.3.5). The archived SecE proteins were classified based on the existence of 1 or 3 TMS, presence of Syd, organism taxonomy and Gram stain. BLASTP was performed using E. coli Syd as the query sequence against the Microbes sub-division of the entire RefSeq database. The multiple sequence alignment built using the Syd proteins reveals the conserved surface electrostatics located at the concave cavity (red bar) and the surrounding protruding stalks. The Syd alignment was prepared using ClustalW (www.ebi.ac.uk/clustalw/) and visualized with JalView. This figure was modified from Dalal et al., 2009.  47  48  2.3.4  Interacting surface between SecY and Syd and exclusion of SecA Single cysteine residues were introduced at different positions on the concave surface  of Syd (Figure 2-6A) and tested for reactivity with IMVs enriched for the SecY complex carrying the mutation R255C. The position SecY-255C is located at the tip of loop C4, and it is involved in the binding of SecA [Osborne and Rapoport, 2007; Figure 1-6]. None of the purified Syd variants formed cysteine-linked dimer in solution (data not shown) nor crossreacted with IMVs bearing the wild-type SecY complex (Figure 2-6B, top). In contrast, a strong SecY-Syd cysteine cross-link was detected when Syd-115C was incubated with the SecY-255C IMVs (Figure 2-6B). This cross-link is specific and efficient because it appeared without the addition of oxidizing agent, and it engaged most of the SecY proteins in the corresponding titration experiment (Figure 2-6B, bottom). Weaker cross-links were observed when the cysteine residue was located at other positions along the groove (90, 97, and 135; Figure 2-6B, top), suggesting more distant or transient interactions with the SecY-255. Previous Scatchard analyses (to determine affinity constants) indicate that Syd does not interfere with the binding of SecA onto membrane-embedded SecYEG [Matsuo et al., 1998]. In support of the notion that membrane bound SecY preferentially binds SecA, we show that SecA can displace Syd from the SecY complex in IMVs. The cysteine cross-link between Syd- 115C and SecY-255C is almost completely abolished when the IMVs are incubated with an equimolar amount of SecA and Syd (Figure 2-6C). Interestingly, when the SecYEG complex in Nanodiscs is incubated with an equimolar amount of Syd and SecA, most of the resulting complex consists of Syd only (Figure 2-6D). Thus, Syd and SecA seem to possess different affinity for the SecY complex whether it is embedded in the membrane  49  as an oligomer or integrated in Nanodiscs as a monomer. Further discussion will require knowledge of the Kd values, but the results already indicate that the binding of Syd and SecA is mutually exclusive. The conclusion is consistent with the overlap of the interaction site, along with the size consideration of SecA and Syd.  Figure 2-6: Contact surface between Syd and SecY and exclusion of SecA. (A) Location of the unique cysteine residues (90, 97, 115, and 135) introduced at the surface of Syd is indicated in yellow. (B) Purified Syd proteins (each 2 µg) were mixed with IMVs (5 µg) enriched for the SecY complex, either wild-type (WT) or carrying the cysteine mutation at position 255 (SecY-255C) in 100 µL of TSG buffer (without DTT). After 5 min of incubation at room temperature, the disulfide bond formation was stopped with N-ethylmaleimide (NEM; 10 mM). Samples were analyzed by SDS-PAGE followed by immuno-staining with a polyclonal antibody directed against SecY. The presence of N-ethylmaleimide during the incubation prevents the formation of the SecY-Syd cross-link (lower panel, left lane). (C) IMVs SecY-255C (5 µg) were incubated with a fixed amount of Syd-115C (2 µg) and a variable amount of SecA (0.0–5.0 µg) in 100 µL of TSG buffer without DTT. Samples were analyzed by SDS-PAGE followed by immuno-staining with anti-SecY antibodies. (D) Approximately 3 µg of SecYEG-Nanodisc was incubated with a fixed amount of SecA (1 µg) and a variable amount of Syd (0.0–1.0 µg) in 50 µL of TSG buffer. After 5 min of incubation at room temperature, samples were analyzed by native-PAGE and Coomassie Blue staining of the gel. This figure was modified from Dalal et al., 2009. 50  2.3.5  Effect of Syd on translocation efficiency and stability of the SecY complex Search of genome data banks reveals that Syd exists only in Gram-negative bacteria  and solely in those where SecE is comprised of three transmembrane segments (TMS). In general, SecE contains a single TMS except in some Gram-negative bacteria where two additional TMS are located at the N terminus of the protein. Remarkably, all of the Syd containing species of bacteria listed in the sequence alignment (Figure 2-5) belong to the gamma proteobacteria family, which are Gram-negative bacteria that include many important and well characterized human pathogens [Williams et al., 2010]. In E. coli, the two Nterminal TMS are not essential for cell viability or translocation activity but seem important for the stability of the SecY complex [Schatz et al., 1991; Nishiyama et al., 2000] (Figure 27A). The SecE subunit is indeed critical for the biogenesis of SecY, and any unassembled SecY is rapidly eliminated by the FtsH protease [Kihara et al., 1995]. The possible evolutionary relationship between Syd and SecE prompted us to test the effect of Syd on in vitro protein translocation when the two N-terminal TMS of SecE are deleted from the SecY complex (deletion SecE ∆7–67). This in vitro assay uses inverted IMVs bearing wild-type or mutant SecY complexes in the presence of SecA and radio-labelled preprotein substrate; upon addition of ATP, the substrate is transported into the lumen of the vesicle and can be quantified by autoradiography (see Section 2.2.6 for description). As reported previously, in vitro protein translocation into wild-type IMVs is not appreciably affected by Syd [Matsuo et al., 1998] (Figure 2-7B). In contrast, the concentration of Syd required to achieve halfinhibition of translocation appears ~5 times lower for the mutant SecY complex carrying the deletion SecE ∆7–67 (Figure 2-7B). A similar inhibition of translocation is also observed when the SecY complex carries the mutation G240D (or SecY24) [Matsuo et al., 1998]. The 51  mutation G240D is located at the interface between SecY and SecE, and it reduces the stability of the SecYE complex [Taura et al., 1994; van den Berg et al., 2004]. Together, the observations suggest a relationship between the activity of Syd and the strength of the SecYEG heterotrimeric associations. Syd seems to preferentially recognize misassembled SecYEG complexes. In the membrane, the two N-terminal TMS of SecE stabilize the SecY complex and may also prevent the binding of Syd.  52  Figure 2-7: Translocation activity and stability of the SecY complex in the presence of Syd. (A) SecY complex, wild-type (WT) or carrying the mutation secY24 (G240D) or SecE∆7-67, was overproduced in E. coli, and the IMVs were immuno-stained with anti-SecY antibodies. The SecY mutant complex cannot be overproduced to the same level as the wild type, probably due to a higher instability. (B) IMVs were ureastripped, and their concentration was adjusted so that a comparable amount of SecY complex is tested in the corresponding in vitro translocation assays. The IMVs were premixed with the indicated amount of purified Syd before addition of the translocation substrate 125I-proOmpA and SecA (0.25 µg). Translocation was initiated with ATP (1 mM) for 10 min at 37 °C. After proteinase K digestion (10 min on ice), samples were analyzed by SDS-PAGE and phosphorimaging. The results were quantified by densitometry and expressed on the graph curve (right panel). (C) 125I-SecY complex (∼75,000 cpm; 0.75 µg) was incubated with the indicated amount of purified Syd protein. The incubation was for 10 min at 22 °C in TSG buffer containing the indicated amount of detergent. Samples were analyzed by blue-native PAGE and phosphorimaging. (D) Detergent-soluble SecYEG complex (600 µg in 50 mM Tris, pH 7.9, 600 mM NaCl, 5% glycerol, 0.03% DDM) was mixed with a molar excess of Syd (450 µg in TSG buffer) and incubated for 5 min with 0.03% DDM at room temperature or with 0.1% Triton X-100 (TX-100) at 37 °C. The mixtures were applied onto a Superdex 200 HR10/30 gel filtration column equilibrated in 50 mM Tris, pH 7.9, 300 mM NaCl, 5% glycerol containing either 0.03% DDM or 0.1% Triton X-100 as indicated. The eluted fractions were analyzed by SDS-PAGE and Coomassie staining. This figure was modified from Dalal et al., 2009.  53  We further analyzed the effect of Syd on the stability of the SecY complex in detergent solution using Blue-Native PAGE, which uses a blue dye in the running buffer that coats protein complexes with a negative charge. The SecY complex in DDM migrates as a dimer during Blue-Native PAGE analysis; the ratio with monomers depends on the concentration of the detergent [Bessonneau et al., 2002]. However, in the presence of Syd, the equilibrium is shifted, and most of the SecYEG dimers are dissociated in monomers (Figure 2-7C). In other detergents such as Triton X-100, the SecY dimer is relatively labile and easily dissociates in single subunits [Maillard et al., 2007]. In the presence of Syd, the dissociation of the SecY complex is further enhanced (Figure 2-7C). Thus, the SecYEG associations are weakened in detergent micelles, and Syd contributes to their further destabilization. These observations are consistent with a model in which Syd acts against a translocation channel in which the SecY-SecE interactions are malformed, unstable, or compromised [Matsuo et al., 1998]. Although the effects observed are proportional to the concentration of Syd, we failed to detect the presence of this protein in the complexes revealed by Blue-Native PAGE (Figure 2-7C), suggesting weaker interaction in detergent solution or the disruption of this interaction by the charged blue dye during the gel electrophoresis. To differentiate the two possibilities, the SecYEG complex was incubated with Syd in detergent solution and analyzed by gel filtration chromatography. In the DDM micelles, the SecYEG complex and Syd are co-eluted together in a seemingly stoichiometric ratio, whereas excess of Syd is eluted in the later fractions (Figure 2-7D). In the Triton X-100 micelles, SecY and Syd are also co-eluted together, but the SecE/SecG proteins clearly appear in the later fractions  54  (Figure 2-7D). Thus, Syd and the SecY subunit alone have the capacity to form a stable complex, as seen in detergent solution.  2.4 2.4.1  Discussion Syd provides a new framework for understanding interactions between SecY  and its partners Characterizing the interactions between the SecY channel and its partners has been difficult because SecA and the ribosome are highly dynamic proteins that contain multiple and complex functional domains. In contrast, Syd is a monomeric and hydrophilic protein with a conserved electronegative and concave surface, making the interaction with the channel rather straightforward. For all three partners, the interactions with the SecY cytosolic loops may largely involve electrostatics. For Syd, the concave surface formed disulphide bonds with the SecY C4 loop and high salt disrupted the SecY-Syd interaction on sucrose gradients, both suggesting an electrostatic mode of binding. In the case of the ribosome, cryo-electron microscopy studies indicate that the acidic phosphate backbone of the rRNA helices at the ribosome exit tunnel establishes major contact points with the basic residues in the SecY loops [Menetret et al., 2007; Menetret et al., 2008; Figure 1-5]. In the case of SecA, the exact binding mode needs to be clarified, but one of the few electronegative patches at the surface of the protein would be prime candidate for the interaction (see Figure 1-7). The effect of Syd on the function and assembly of the SecY complex was also addressed in this study. The capacity of Syd to dissociate the dimeric and intramolecular SecYEG associations in detergent solution is remarkable. It is consistent with the Syd  55  inhibitory activity against certain SecY mutant complexes presenting abnormal SecY-SecE associations [Shimoike et al., 1995; Matsuo et al., 1998; and this study]. Together the results support a model in which Syd represents a quality control system for the correct assembly of the SecY channel. In the cell membrane, the mal-associated SecY complex would be preferentially recognized and further dissociated upon binding of Syd. The unassembled SecY subunit would then be targeted for degradation by the FtsH protease [Kihara et al., 1995].  2.4.2  The stoichiometry between SecY and SecA In addition to detecting a 1:1 binding stoichiometry between Syd and the SecY  complex, our results demonstrate that one SecA molecule is bound with the SecY monomer [Alami et al., 2007] or SecY dimer (Figure 2-3C). Consistent with our observations, a stoichiometry of a single SecA molecule and two SecY complexes was observed by photocrosslinking in intact cells [Das and Oliver, 2011], but it is not known whether this assembly is engaged in protein translocation or not. Similarly, a single SecA molecule is present in the SecYEG-SecA structure (Figure 1-1C) [Zimmer et al., 2008], but may not represent the translocation active state as the complex was crystallized in detergent solution and without preprotein. Another study provided functional evidence for the activity of SecA monomers, as a SecA mutant (deletion of the C terminal 11 residues) that is predominantly monomeric (>98%) in solution can support in vitro translocation across the SecY complex, and can sustain cellular growth [Or et al., 2005]. The fact that acidic lipids, which are required for protein translocation, cause the monomerization of SecA [Or et al., 2002; Benach et al., 2003] also seems to favour a single SecA molecule as the functional state.  56  In contrast, chemically cross-linked SecA dimers appear to be fully active in translocation [de Keyzer et al., 2005]. Furthermore, an equilibrium of SecA monomers and dimers bound to the SecY complex were detected by various techniques. For example, both SecA oligomeric forms bind to SecYEG in detergent solution when analyzed by nativePAGE or gel filtration [Bessonneau et al., 2002; Duong, 2003; Tziatzios et al., 2004], and also in membranes by recent fluorescence measurements probing SecYEG-SecA interactions at the single molecule level [Kusters et al., 2011]. Clearly, the functional oligomeric state of SecA continues to be controversial, with both monomer and dimers being implicated during the translocation cycle. Unlike the many techniques described above, Nanodiscs allowed direct visualization of the binding between the channel and SecA by native-PAGE, which is not possible with liposomes or detergent solution due to the uncontrollable oligomerization of the SecY complex. As described in the next chapter, a binding stoichiometry of one SecA molecule and two SecY complexes creates a functional apparatus in Nanodiscs, and has important implications for the mechanism of protein translocation.  57  Chapter 3: The SecY channel dimer and acidic lipids are required for SecA-dependent preprotein translocation 3.1  Introduction Although the involvement of SecA in post-translational translocation has been well  characterized (see Section 1.4), the mechanism of the preprotein dependent SecA activation remains unclear. Some details of this mechanism have been elucidated by thermodynamic analysis of SecA-driven translocation, which shows that the docking of the substrate signal sequence with the channel lowers the temperature necessary to stimulate SecA ATPase activity (i.e. the activation energy barrier) and initiate protein transport [Gouridis et al., 2009]. This docking event stimulates the SecA ATPase which, under optimal conditions, leads to the net movement of 20-30 amino acids of the preprotein across the membrane [Scheibel and Wickner, 1991]. This SecA ATPase activity has been defined as the ‘SecA translocation ATPase’ [Cunningham and Wickner, 1989] because it is stimulated ~6 to 9 fold in the presence of a preprotein substrate [Gouridis et al., 2009; Hendrick and Wickner 1991]. The SecA translocation ATPase also depends on acidic lipids in the membrane, such as phosphatidylglycerol (PG) [de Vrije et al., 1998; Lill et al., 1990]. As discussed in Chapter 1, the atomic structure of the SecY complex, alone or associated with SecA, has revealed the location of the protein-conducting channel at the centre of the SecY subunit [van den Berg et al., 2004; Zimmer et al., 2008]. This finding has been confirmed by thiol-crosslinking [Cannon et al., 2005; Bauer et al. 2009] and by cryoelectron microscopy analysis of the eukaryotic Sec complex bound to a translating ribosome [Becker et al., 2009; Frauenfeld et al., 2011; Figure 1-5]. It is thus well established that a 58  single SecY copy is sufficient to form the translocation pathway, yet it remains mysterious that the channel exists as oligomers (see Section 1.6). An earlier study employing a covalently linked SecYEG dimer showed that a preprotein can be transported across a defective channel provided a functional SecY copy is fused to it [Deville et al., 2011]. It was proposed that each copy had a different role, one serving as docking site for SecA and the other as translocation channel (a model termed ‘fraternal twins’; Osborne and Rapoport, 2007; Duong, 2007; Figure 1-4A). In possible support to the model, a photo-crosslinking analysis in intact cells indicated that SecA can simultaneously contact two SecYs [Das and Oliver, 2011]. More recently, a single molecule analysis in proteoliposomes indicated that a single SecY is sufficient to bind the preprotein, but the dimer is also necessary to support significant transport [Deville et al., 2011]. These earlier studies have highlighted the importance of the dimer, but the exact role and the function of each copy needs additional support. Protein translocation taking place at one SecY copy might be facilitated by the second, or translocation may strictly depend on the dimer. It is also possible that artificially fused SecY complexes form higher oligomers in the membrane (e.g. tetramers) that would be responsible for the translocation activity observed in proteoliposomes. The question is further complicated by the dynamic dimeric state of SecA and whether one or two SecA molecules bind to the channel [Akita et al., 1991; Tziatzios et al., 2004; Or et al. 2005; Alami et al., 2007; Kusters et al., 2011; Wowor et al., 2011]. As shown in Chapter 2, the SecY monomer incorporated into Nanodiscs was sufficient to bind SecA. Here, we show that the SecY monomer is sufficient to bind SecA and the signal sequence, yet the activation of SecA occurs only when a second SecY, and acidic lipids, are present in the disc. The two copies are predominantly arranged in a back-  59  to-back manner and create a binding site for one SecA only. Consistent with the fraternal twin model, the SecY dimer can activate the SecA ATPase, provided SecA can bind to the assembly. To confirm the involvement of the dimer in the cell context, we combined a mutant defective for channel opening and a mutant defective for SecA binding. When coproduced together, these otherwise inactive SecY channels created a functional assembly. These results strongly argue that two SecY copies are necessary for preprotein transport.  3.2 3.2.1  Materials and methods Biological reagents Membrane scaffold protein (MSP) variants were prepared as described [Denisov et  al., 2004; Alami et al., 2007]: MSP1D1 (referred to as MSP1; 24.6 kDa), MSP1E3D1 (referred to as MSP3; 32.6 kDa) and MSP2N2 (45.5 kDa). pBAD22-based plasmids encoding for His-tagged SecYEG and covalently-linked SecYEG dimer were previously described [Collinson et al., 2001; Duong, 2003]. The SecYEG complexes were expressed in E. coli BL21 (DE3) and purified by Ni-NTA metal-affinity and cation exchange chromatography [Dalal and Duong, 2010]. To express two different SecY complexes from the same plasmid, a second secY gene together with an identical ribosome binding site were introduced into plasmid pBAD22-hisEYG, in-between secY and secG. Briefly, the restriction site SacI and XbaI were inserted by site-directed mutagenesis at the 3’ end of secY and 5’ end of secG, respectively. Next, a second secY flanked by a ribosome binding sequence was generated by PCR amplification and inserted in-between SacI and XbaI. The resulting plasmids pBAD22-HisEYFF-YG and pBAD22-HisEYFF -YEG were verified by sequencing analysis and restriction mapping. The first 202 amino-acids of the alkaline phosphatase A  60  precursor (PhoA1-202) were expressed from plasmid pET-23 and purified from inclusion bodies using Ni-NTA metal-affinity chromatography under denaturing conditions (50 mM Tris-HCl, pH 7.9, 6 M Urea, 1 mM DTT). 125I-labelling of SecA was performed with iodogen-coated tubes (Pierce) as previously described [Duong, 2003]. DOPG lipids (dioleoyl-phosphatidylglycerol) and E.coli total lipids were purchased from Avanti Polar Lipids. Cu2+-phenanthroline3 (CP3) and N-ethylmaleimide (NEM) were from Sigma.  3.2.2  Nanodisc reconstitutions Reconstitution of the SecY complex into MSP1 was performed as previously  described [Alami et al., 2007; Dalal and Duong, 2010; and Appendix A] at a molecular ratio SecYEG:MSP1:lipids of 1:5:40. Reconstitutions of the SecYEG dimer into MSP3 were performed at a molecular ratio of 1:3:20. The covalently linked SecYEG dimer (termed SecYY, Duong, 2003) was reconstituted into MSP3 or MSP2N2 at a molecular ratio of 1:5:40. Nanodiscs containing only PG lipids or E. coli lipids were prepared at a molecular ratio of MSP3:lipids of 1:20. Purification of the reconstituted discs was by gel filtration chromatography on a Superdex 200 HR10/30 column equilibrated in 50 mM Tris-HCl pH 7.9, 100 mM NaCl, 5% glycerol (TSG buffer). Alternatively, sucrose density purification was performed by layering the Nanodisc preparation on top of a 10 mL sucrose gradient (613% in TSG buffer), followed by centrifugation at 188,000 x g in a Beckmann SW41 rotor (17 hours, 4°C). Fractions (0.5 mL) were collected top-down and analyzed by native-PAGE.  61  3.2.3  Translocation ATPase measurements The purified Nanodisc preparations (0-155 nM) were incubated with SecA (0.2 µM)  and ATP (1 mM) in TL buffer (50 mM Tris-HCl pH 7.9, 50 mM KCl, 50 mM NaCl, 5 mM MgCl2, 1 mM DTT), and 0.8 µM PhoA1-202 or PhoA1-202-L14R. The release of inorganic phosphate was measured by colorimetric method (Malachite Green) as previously described [Alami et al., 2007]. ATPase rates in the presence or absence of preprotein substrate were measured in intervals over a range of 30 minutes at each Nanodisc concentration, followed by fitting the initial rates to a one site quadratic binding equation to determine kcat values as previously described [Robson et al., 2009] and shown in Equation 3-1:  v=  Vmax * L + [ E 0 ] + K d − ([ L] + [ E0 ] + K d ) 2 − 4[ E0 ][ L] 2[ E 0 ]  Equation 3-1: One site quadratic binding equation.  Where v is equal to enzyme activity rate, Vmax is the maximum enzyme activity, [L] is the Nanodisc concentration, [E0] is total SecA concentration, and Kd is the dissociation constant for SecA. The rate of inorganic phosphate release observed in the absence of the substrate was subtracted in each experiment.  3.2.4  Other methods Protein concentration was determined using the Bradford reagent (Bio-Rad).  Colorless native gels (4–13% linear gradient) and electrophoresis conditions were performed as described in Appendix A. The molecular weight markers employed on native-gel are: ferritin, 440 kDa; catalase, 232 kDa; BSA (trimer/dimer/monomer), 201/134/67 kDa. The SecA stabilized dimer (SecACP3) was obtained following oxidation with CP3 and size  62  exclusion chromatography as previously described [Alami et al., 2007]. Dynamic light scattering (DLS) measurements on the Nanodisc particles (0.1 µg/mL) were performed at 25ºC on a Wyatt DynaPro Nanostar equipped with a 661 nm laser beam. Affinity pull-down experiments were performed by binding the Nanodiscs onto Ni-NTA beads (GenScript) via a 6-Histidine N-terminal tag on MSP1 and MSP3, followed by incubation with SecA (see Figure 3-4). In vitro protein translocation experiments were carried out as previously described in Section 2.2.6 using 125I-labelled PhoA1-202 or fluorescently labeled PhoA1-202 (see Section 2.2.7) and IMVs prepared from E. coli strain KM9 (see Section 4.2.1). Dyelabelling of PhoA1-202 (100 µg) was performed by incubation with Alexa Fluor® 680 (40 ng/µL; Invitrogen Molecular Probes) in 50 mM Tris-HCl pH 7.9, 6 M urea, 1 mM EDTA, and 5 µM Tris(2-carboxyethyl)phosphine buffer for 2 hrs at room temperature. The reaction was quenched with 1 mM DTT, and excess dye was removed by gel filtration chromatography in 50 mM Tris-HCl pH 7.9, 5 M urea, and 1 mM DTT. The membrane insertion assay was performed as described in Economou and Wickner (1994). In vivo transcomplementation experiments were performed in E. coli conditional lethal strain CJ107 carrying the secY24 mutation [Shimoike et al., 1995].  3.3 3.3.1  Results The SecY dimer supports the SecA translocation ATPase The SecY complex and PG lipids reconstituted with MSP3 (capturing both SecY  monomers and SecY dimers, see Section 2.3.1), was purified by gel-filtration chromatography and the fractions analyzed by native-PAGE (Figure 3-1, A and B). The fractions were also incubated with SecA, ATP and the preprotein substrate PhoA1-202  63  (alkaline phosphatase residues 1-202; Figure 3-1C). A significant ATPase stimulation occurred with the fractions #5 to #9, which are mostly enriched for the SecY dimer. In contrast, no ATPase activity was detected with the fractions enriched for the SecY monomer (fraction #10 to #12), as previously reported [Alami et al., 2007]. The possibility that the SecYEG Nanodisc preparation was contaminated with proteoliposomes was unlikely because only 20 mole of lipids were added per mole of SecY during the reconstitution. Dynamic light scattering also showed that more than 99.8% of the total mass of fraction #7 consisted of particles with a mean diameter of ~15 nm (Figure 3-1D). These particles were not highly mono-disperse, which might be explain by populations of discs with slightly different amount of lipids, but nevertheless roughly agrees with the diameter of Nanodiscs [Ritchie et al., 2009]. Additional controls showed that the SecA ATPase activity was not supported by a preprotein substrate carrying a defective signal sequence (PhoA1-202-L14R; Figure 3-1E), nor when the SecYEG complex was reconstituted with E. coli total lipids which contain lesser amount of PG (~30% acidic and 70% neutral lipids; Figure 3-1E). In addition, the SecYEG complex carrying the mutation R357E in the large cytosolic loop of SecY (labeled Nd-Y2E ) or the double mutation I82F/I187F in the pore ring (labeled Nd-Y2FF) failed to support the SecA translocation ATPase activity in discs (Figure 3-1F). These mutations, further discussed below, strongly decreased the SecY translocation activity in membranes (Figure 3-2). Together, the results showed without ambiguity that activation of the SecA translocation ATPase depends on the SecY dimer, a functional signal sequence and acidic lipids. The same dependencies define the SecA translocation ATPase in the membrane [Lill et al., 1990].  64  Figure 3-1: Two SecY copies are necessary to activation the SecA translocation ATPase. (A and B) The SecY complex reconstituted in Nanodiscs with MSP3 and PG lipids was purified by gel filtration chromatography. The corresponding fractions were supplemented with Syd (1 µg) to facilitate analysis by native-PAGE. (C) The same fractions were incubated with SecA (0.2 µM), PhoA1-202 (0.8 µM) and ATP (1 mM) for 30 minutes at 37°C. The release of inorganic phosphate was determined by colorimetric assay. Error bars were derived from 3 independent measurements. (D) Dynamic light scattering of fraction #7 in (C). (E) The Nanodiscs were prepared with the indicated lipids (PG: dioleoylphosphatidylglycerol; Ec: E. coli total lipid extract). The monomeric and dimeric populations were separated by gel filtration chromatography as in (A). The SecA translocation ATPase was determined after plotting the initial ATPase activity rates (see Appendix D) obtained in the presence of Nanodiscs and PhoA1-202 or PhoA1-202-L14R (0.8 µM). (F) The SecY complex carrying the mutation R357E (termed YE) or I82F/I187F (termed YFF) was reconstituted into discs. The population of the discs containing the dimer (labeled Nd-Y2E and Nd-Y2FF, respectively) was isolated by size-exclusion chromatography and tested for translocation ATPase as in (C). Nanodiscs containing only PG lipids (labeled Nd-PG) or E. coli total lipids (labeled Nd-Ec) do not support the SecA translocation ATPase activity.  65  Figure 3-2: In vivo and in vitro activity of the SecY mutant channels employed in this study. (A) E. coli CJ107 (secY24) transformed with the indicated plasmids was grown for 6h in LB broth, then serial diluted and spotted on LB-agar plates containing 0.2% arabinose to induce plasmid expression of Syd and the indicated SecY complex. Overproduction of Syd is lethal in CJ107 because Syd destabilizes the already thermo-sensitive SecY24 complex, which becomes rapidly proteolysed [Shimoike et al., 1995; Dalal et al., 2009]. (B) The in vitro protein translocation activity was assayed using 125I-labelled PhoA1-202 as described in Section 2.2.7. The lane pBAD22 refers to IMVs prepared from bacteria expressing endogenous level of SecYEG. Right lane shows 20% of the input material. The % of translocated PhoA1-202 was determined by densitometry using the ImageJ software. The lower band in each lane is the proteolysed version of the substrate which lacks the N-terminal signal sequence (see Section 2.2.6) (C) Western blot analysis of IMVs using a polyclonal antibody directed against SecY.  66  3.3.2  The dimer remains active when one SecY copy is defective for SecA binding To understand the contribution of the SecY copies in the binding and activation of  SecA, we employed a covalently-linked SecY dimer carrying the mutation R357E on either one or both copies (termed YYE, YEY and YEYE respectively). This construct links the C terminus of a SecY molecule to the N terminus of a second [Duong, 2003], and allow mutations to be selectively introduced into either. Thus, a homogeneous population of the covalently linked SecY dimer, carrying the desired mutations, can be expressed and purified. These hybrids were reconstituted into discs and the association with SecA was monitored by native-PAGE (Figure 3-3A) and affinity pull-down (Figure 3-4). When present on both copies, the mutation R357E impaired the binding of SecA (~3 fold; Figure 3-4) and abolished the SecA translocation ATPase (Figure 3-3C) and protein transport in membranes (Figure 3-2B). In contrast, the ATPase activity remained unchanged when the heterodimer contained one wild-type SecY copy (Nd-YY, Nd-YYE or Nd-YEY; Figure 3-3C), which also restored the binding of SecA (Figure 3-3A). As expected, the single SecYE copy alone in the disc had a weak affinity for SecA compared to the wild-type SecY complex (Figure 33B). The ATPase measurements in Figure 3-3C were performed with hybrids reconstituted in to a larger membrane scaffold protein (MSP2N2, see Appendix E), because YY, YYE and YEY complexes stimulated the best translocation ATPase when incorporated into these discs compared to with MSP3. Together, these results confirmed that the SecA translocation ATPase depends on two SecY copies, although only one is needed for the binding of SecA.  67  Figure 3-3: The dimer remains active when a SecY copy is defective for SecA binding. (A) A fixed amount of Nanodisc containing the indicated fused SecY dimer (3 µg made with MSP3 and E. coli lipids) was incubated with an increasing amount of SecA. Alternatively, a fixed amount of SecA (2 µg) was incubated with an increasing amount of discs. The complexes were analysed by native-PAGE and Coomassie blue staining. The Nanodiscs containing the fused SecY dimer with the mutation R357E on either the first or the second copy are labelled Nd-YEY and Nd-YYE, respectively. (B) As in (A) but using the indicated monomeric SecY complex reconstituted with MSP1 and E. coli lipids. (C) Nanodiscs containing the fused SecY dimer were purified by sucrose density ultra-centrifugation (see Appendix E for analysis of sucrose gradient fractions). The kinetics of the measured translocation ATPase are presented in Appendix D.  68  Figure 3-4: Pull-down of 125I-labelled SecA with Nanodiscs. (A) The indicated His6-tagged Nanodiscs (10 µg, same series as in Figure 3-3, A and B) were incubated with the indicated amount of 125I-labelled SecA (non-tagged). The complex was isolated via Ni2+-NTA pull-down assay (500 µL reaction in TSG buffer). The beads were sedimented at 600 x g and washed twice with the same buffer containing 50 mM imidazole. Nanodiscs and bound SecA were eluted in TSG buffer containing 500 mM imidazole, followed by analysis on 12% SDS-PAGE and autoradiography. (B) The amount of bound SecA was quantified by densitometry using ImageJ software. 100% is defined as the density observed after pull-down of 320 ng 125I-SecA using Nd-YEY. Nd-Ec lipids refer to Nanodiscs made with E. coli total lipid extract.  69  3.3.3  The SecY monomer suffices to bind the signal sequence We next tested if the SecY oligomeric state is important for preprotein binding.  Previous crosslinking analysis showed that the signal sequence binds near the amino-acyl position 97, close the SecY channel pore [Osborne and Rapoport, 2007, and Appendix F for model]. To test whether one or two SecY copies were necessary for binding the signal sequence, PhoA1-202 carrying a unique cysteine in its signal sequence (PhoA1-202-5C) was incubated with the SecY97C monomer and dimer purified in Nanodiscs. If the signal sequence can interact with its corresponding binding site on SecY, a disulphide linked SecYsubstrate complex will be detected on SDS-PAGE. Under oxidizing conditions, the signal sequence of PhoA1-202 formed a disulphide linkage with SecY97C irrespective of the oligomeric state of the complex (Figure 3-5A, lanes 5-12). The interaction between the signal sequence and SecY97C in the disc or in detergent solution did not depend on SecA and ATP, unlike the SecY97C complex in proteoliposomes which did depend on these components (Figure 3-5A, lanes 19-21). It is possible that the accessibility of the SecY pore is more restricted when the SecY complex is embedded in the membrane lipid bilayer. We also measured the capacity of the SecY complex monomer and dimer to support SecA ‘membrane insertion’. When bound to the SecY complex in the presence of a non-hydrolysable ATP analog, SecA undergoes a conformational change that causes a domain of 30 kDa to become protease-resistant [Economou and Wickner, 1994]. Formation of the 30 kDa domain was supported in Nanodiscs when at least one SecY complex was able to bind SecA (Figure 35B). Thus, although the dimer is necessary to activate SecA, the SecY monomer was sufficient to bind the signal sequence and to support SecA insertion.  70  Figure 3-5: The SecY monomer suffices to bind the signal sequence. (A) 125I-labelled PhoA1-202 (~25,000 c.p.m., 200 ng) bearing a unique cysteine residue at position 5 of the signal sequence (labelled 5C) was incubated with the indicated Nanodiscs (3µg each) in TL buffer for 5 minutes at 37°C in the presence of SecA (0.1 µM) and ATP (1 mM). The oxidation of the cysteines was started with CP3 (0.2 mM for 5 min at room temperature) and terminated with NEM (10 mM). The crosslink products were analyzed by 12% SDS-PAGE and autoradiography. The same cysteine crosslinking experiment was performed with the purified SecY97CEG complex in detergent solution (3 µg in TL buffer + 0.03% dodecyl maltoside) or reconstituted in liposomes (3 µg at a protein:lipid ratio of 1:2000). The cysteine crosslink between SecY97C and PhoA1-202-5C is labelled SecY97C-5C. A fraction of PhoA1-202-5C forms cysteine linked dimers (labelled 5C2). (B) 125I-labelled SecA (~50,000 c.p.m.; 0.02 µM) was incubated with the indicated Nanodiscs (3 µg each), PhoA1-202 (0.8 µM) and AMP-PNP (1 mM) for 10 minutes at 37°C. Trypsinolysis (0.08 µg/mL trypsin final) was for 15 minutes on ice. Samples were precipitated with TCA (17% final) and analyzed by SDS-PAGE and autoradiography. Nanodiscs bearing the monomeric SecY complexes (Nd-Y, Nd-YE and Nd-YFF) were reconstituted with MSP1 and PG lipid. Nanodiscs bearing the covalently linked SecY dimer were prepared with MSP2N2 as described in Appendix E.  71  3.3.4  The back-to-back dimer is the predominant formation in discs Two different dimeric arrangements have been observed in the lipid bilayer or in  detergent solution: the ‘front-to-front’ where the SecY lateral gates are facing each other and the ‘back-to-back’ with the SecE subunits forming the interface [Kaufmann et al., 1999; Breyton et al., 2002; van der Sluis et al., 2002; Mitra et al., 2005; Deville et al., 2011; Das and Oliver, 2011]. To identify these two conformations (see Appendix F for hypothetical representation), we introduced a cysteine residue at position 106 on SecE (YE106CG) or at position 103 on SecY (Y103CEG). When overproduced in the membrane, both SecE-SecE and SecY-SecY disulfide linkages were detected (Figure 3-6A), suggesting that both dimeric arrangements exist in the membrane, although artifacts related to SecY oligomerization due to protein overproduction cannot be excluded. In detergent, only SecE-SecE crosslinks were detected but this was seen only after dissociation of the SecYEG heterotrimer by temperature or by SDS (Figure 3-7). When the mutant complexes were reconstituted with MSP3 and PG lipids (Figure 3-6B) SecE-SecE crosslinks were readily detected upon oxidation, but SecYSecY crosslinks were absent (Figure 3-6C). Monomers of the SecY complex bearing either cysteine did not form crosslinks either (data not shown). Thus, the SecY dimers are arranged in a back-to-back manner in Nanodiscs and this formation supports the preprotein-dependent SecA translocation ATPase activity.  72  Figure 3-6: Orientation of the SecY copies within the disc. (A) Inner membrane vesicles (2 µg) enriched for the SecY complex carrying the mutation SecE-L106C or SecY-A103C were incubated for 5 minutes at room temperature followed by addition of 0.2 mM CP3 (2 minutes) then NEM (10 mM). Protein samples were analyzed by 12% SDS-PAGE and Western blot using a polyclonal antibody against SecE, or monoclonal antibody against SecY. (B) The indicated SecY complexes were purified and reconstituted in Nanodiscs (using MSP3 and PG lipids). A sample was analysed by nativePAGE and Coomassie blue staining. (C) The same samples were oxidized with 0.2 mM CP3 (2 minutes at 37°C) followed by NEM treatment (10 mM), 12% SDS-PAGE and immuno-staining with anti-SecE or antiSecY antibodies. Cysteine crosslinking is not detected when the samples are treated with NEM before addition of CP3.  73  Figure 3-7: Crosslinking of SecE-L106C and SecY-A103C in detergent solution. The purified SecY complex in detergent solution (2 µg in 0.03% DDM) was incubated for 5 minutes at the indicated temperature or in the presence of 0.2% SDS followed by incubation with 0.2 mM CP3 (2 min, room temperature) and quenching with NEM (10 mM). Immuno-detection using anti-SecE or anti-SecY antibodies were employed to increase the detection limit of the protein crosslinks.  74  3.3.5  Complementation between two SecY mutants restores protein translocation Our results in Section 3.3.2 showed that one copy within the SecY dimer was needed  for SecA binding. Protein translocation might occur at this SecA-bound SecY copy, or it could occur at the other. To understand how the SecY copies may work together, and which are involved in SecA binding and protein transport, we expressed the SecYFF and SecYE mutant copies in the same cell. The mutation SecYFF (I82F/I187F) alters the pore ring structure and most likely compromises the ability of the channel to open (Figure 3-2). This mutant otherwise interacts with SecA as strongly as the wild-type channel (Figure 3-3B). The SecYE complex is defective for SecA binding (Figure 3-3B) but contains an unaltered translocation pore. Neither SecYFF nor SecYE expressed on their own from a plasmid restored the growth of a thermo-sensitive SecY mutant strain (Figure 3-8, A and B). In contrast, the co-production of both mutants restored cell viability (Figure 3-8, A and B). Similarly, the SecYFF and SecYE complexes could support in vitro protein translocation and SecA insertion reactions, but only when both complexes were expressed in the same membrane (Figure 3-8, C and D). Thus, the two inactive SecY complexes must have associated together in order to form a functional channel. In this scenario, the SecY copy bearing wild-type cytosolic loops (SecYFF) would allow SecA to bind whereas the functional pore of the neighboring copy (SecYE) would provide the conduit for the substrate to cross the membrane.  75  Figure 3-8: Trans-complementation between SecYFF and SecYE mutant channels. (A) The cell growth complementation assay was performed as described Figure 3-2A. Cellular growth is observed when the YFF and YE mutant channels (labeled YFF + YE) are co-expressed together in the same strain. (B) The experiment as in (A) was performed in CJ107 cells (without Syd) incubated at 30°C and 42°C. The thermo-sensitive SecY24 complex is compromised at 42°C. (C) In vitro translocation assay using inner membranes vesicles enriched with the indicated SecY complexes. Each assay was performed using 16 ng of fluorescent dye-labeled PhoA1-202 mixed with 100 ng of unlabelled PhoA1-202. Translocation activity was determined at different temperatures (left panel) or at 30°C at different time points (right panel). (D) The SecA membrane insertion assay was performed as in Figure 3-5B, using 3 µg of membrane vesicles containing the indicated SecY complex. (E) The total amount of SecY is comparable in the various membrane preparations used in (C) and (D). Immuno-detection was using a polyclonal antibody directed against SecY. The large cytoplasmic loop of SecY is sensitive to proteolysis. On the right lane, the detergent purified SecYEG complex (1µg) was incubated with trypsin (~300 ng for 15 min at 4°C), followed by precipitation with 17% TCA prior to loading on the gel; this band is absent from IMVs. The bands labelled with an asterisk are cross-reacting proteins with the SecY antibody. 76  3.4 3.4.1  Discussion The requirement and functional asymmetry of the SecY dimer  The difficulty of understanding the oligomerization of the SecY complex has led to uncertainty regarding the functional state of the channel. Here, the nanodisc allowed isolating the SecY dimer which, unlike the monomer, was able to support the SecA translocation ATPase. The importance of SecY dimerization was also observed in membrane vesicles and in vivo because the co-production of two inactive SecY subunits, each for a different reason, recreated a functional unit. Together, these results would indisputably argue that the SecY dimer is crucial for the activation of SecA and subsequent preprotein transport. Yet, a recent analysis concluded that a single SecY channel suffices to support SecA-driven protein translocation [Kedrov et al., 2011]. In that study, the SecYEG complex was incorporated into giant liposomes at extremely dilute protein concentration. Using single molecule fluorescence spectroscopy, it was found that the SecY monomer supported SecA binding and formation of a preprotein translocation intermediate. Remarkably, the preprotein was not detected with the SecY dimer. In contrast, in another single molecule analysis at even lower protein to lipid ratio, the monomer was found insufficient to support protein translocation [Deville et al., 2011]. The opposed conclusions reached in these earlier liposome assays may highlight the difficulty of controlling or measuring the SecY oligomeric state in the fluidic lipid environment. In nanodiscs, the monomer was sufficient to bind SecA and the preprotein signal sequence (Figure 3-3B and Figure 3-5A). These observations were consistent with earlier disulphide crosslinks and confocal microscopy analysis that probed the interaction of the  77  preprotein with the channel [Canon et al., 2005; Deville et al., 2011], yet they render the reason for the presence of a second copy obscure. To help clarify the question, we employed two SecY mutants defective in translocation for different reasons and showed that their coproduction in the same membrane allows for the recreation a functional apparatus. Thus, although the SecY dimer is made of two identical copies, both are required and the function of each is different. This functional asymmetry has also been previously detected by sitedirected cysteine crosslinking experiments where SecA and the preprotein were found separately engaged with two distinct SecY copies during protein translocation [Osborne and Rapoport, 2007].  3.4.2  Comparison of the translocation reaction and channel dynamics in Nanodiscs  and membranes The mechanism of the SecA translocation ATPase itself is still unclear. Arrhenius plots have indicated that the docking of the signal sequence onto the channel lowers the SecA activation energy barrier, a process termed ‘triggering’ [Gouridis et al., 2009]. This step seems to be followed by the irreversible engagement of the substrate within the channel (trapping) and cycles of ATP hydrolysis coupled to protein transport (secretion). In the membrane, this step would be normally followed by the irreversible engagement of the substrate with the channel and by cycles of ATP hydrolysis coupled to protein transport. The later step was apparently not reproduced in the nanodisc, perhaps as a result of the low kcat supported by the system (29.6 min-1, Figure 3-1E) compared to membrane vesicles and proteoliposomes (70 min-1 and 456 min-1, respectively; Gouridis et al., 2009; Robson et al., 2009). The number of lipid captured inside the disc (<40-50 lipids per leaflet given size  78  constraints) may also be insufficient for the signal sequence to interact productively with the channel. This limitation might explain why the preprotein-dependent SecA translocation ATPase was stimulated only 2-3 fold, compared to the 6-9 fold in the membrane [Gouridis et al., 2009]. In addition, the SecY conformation may be affected by the membrane lateral pressure perhaps absent in the disc. This other limitation may explain why the binding of the signal sequence was not dependent on ATP (Figure 3-5A) and why the SecY pore mutant could still facilitate SecA insertion (Figure 3-5B). Nevertheless, the fact that the SecA ATPase activity was dependent on a correct signal sequence is strong evidence that the triggering step of the reaction has been recreated in the disc. The apparent limitation of the Nanodisc to recreate only the triggering step of translocation may explain why a stable complex between SecY and the preprotein was not detected. It is also conceivable that during some point in the translocation reaction, the dissociation of the second SecY copy is required, which would be prevented by the membrane scaffold protein surrounding the Nanodisc. Although we detected a 1:1 binding stoichiometry between the SecY dimer and SecA (Chapter 2), it is also possible that a second SecA molecule is needed for protein transport sometime after translocation initiation. These possibilities cannot be ruled out, and may further explain why ATP dependent signal sequence insertion, and subsequent transport of the substrate’s mature segment, was not observed with the Nanodisc SecY complex. The organization of the SecY copies in the functional dimer has been controversial. Our results show that both front-to-front [van der Sluis et al., 2002; Das and Oliver, 2011] and back-to-back [Kaufmann et al., 1999; Deville et al., 2011] conformations exist in the membrane but most likely as a result of protein overproduction. Furthermore, the SecE  79  subunit self-dimerizes when unbound to SecY (Figure 3-7; Matsuo et al., 2003), which complicates earlier cysteine crosslink analysis performed on membrane vesicles. In nanodiscs, the majority of the SecY complex was arranged in a back-to-back manner. Since the formation of the disc is a self-assembly process, the back-to-back orientation may be an energetically favourable state preferentially selected during the reconstitution. These results are compatible with previous experiments showing that a disulphide stabilized back-to-back dimer is active in liposomes [Deville et al., 2011]. These results do not exclude, however, that other functional arrangements exist. In fact, the exact orientation of the monomers may not be critical as long as the dimeric assembly satisfies signal sequence binding and SecA activation.  3.4.3  Role of lipids in protein translocation It is known that acidic lipids are essential for the activity of SecA [de Vrije et al.,  1988; Cunningham and Wickner, 1991] and a recent report has highlighted a specific role for cardiolipin in the oligomeric state of the channel [Gold et al., 2010]. With Nanodiscs, it was possible to control the lipid composition neighbouring the SecY complex and to show directly that acidic lipids are critical to support the translocation ATPase of SecA bound at the SecY dimer. Recent cryo-EM analysis of the ribosome-bound SecY complex in Nanodiscs reveals exposed lipids at the surface of the disc which could be available to interact with SecA [Frauenfeld et al., 2011]. Discs composed of purely PG lipids (Figure 31F, Nd-PG), or the SecY monomer reconstituted with PG and MSP3 (Figure 3-1E), did not support the translocation ATPase, indicating that lipids only facilitate preprotein-dependent SecA activation in the context of the SecY dimer. How acidic lipids contribute to lower the  80  activation energy of SecA remains to be determined. These lipids seem to enhance the affinity of SecA for the channel [Alami et al., 2007; Gold et al., 2010] and also favour the monomerization of the SecA dimer [Or et al., 2002; Benach et al., 2003]. Together, these effects could directly contribute to the triggering step of the translocation reaction.  3.4.4  Possible advantages of the SecY dimer Considerable progress has been made in understanding the SecY complex, but why  this channel must dimerize and the possible functional advantage, if any, is not entirely understood. It has been proposed the SecY dimer serves to increase the number of sites for recruiting different binding partners [Osborne and Rapoport, 2007], to create a microenvironment of acidic lipids necessary to activate SecA [Gold et al., 2010; Deville et al., 2011], and to facilitate channel opening through some allosteric communications between the SecY subunits [Bostina et al., 2005; Tam et al., 2005]. Our results with Nanodiscs add that one SecY copy in the dimer is the major binding site for SecA (Figure 3-3A), and complementation experiments suggests the other copy engages the preprotein substrate (Figure 3-8). That a single gene suffices to perform two different functions is perhaps a genetic advantage in bacteria. Random mutagenesis indicated that, except for the substitution R357E, SecY is surprisingly tolerant to mutations in the large cytosolic loops that create the SecA binding site [Mori and Ito, 2001]. Since SecA polypeptide-binding and nucleotidebinding domains seem to establish contact with both SecY copies [Osborne and Rapoport, 2007; Zimmer et al., 2008; Das and Oliver, 2011], a single mutation in the SecY cytosolic loops would only modestly affect the overall SecA binding to the channel, thereby increasing the robustness of the translocation process.  81  Chapter 4: The SecY complex forms a channel capable of ionic discrimination 4.1  Introduction The final component of this thesis is the analysis of the ionic conductance properties  of the SecY channel. This may appear unrelated to the previous chapters describing the quaternary assembly of the SecY complex, but as shown below, ionic conductance is a normal consequence of SecA-driven polypeptide movement through the SecY channel. Thus, a complete picture of the translocation reaction must include the mechanism of ionic conductance and selectivity by the SecY channel. During channel opening and movement of the polypeptide chain, the SecY complex must apparently maintain the impermeability of the membrane to small molecules and ions. Surprisingly, in vivo, the toxicity induced by the locked-open channel (i.e. crosslinking the plug in the open state see Section 1.3.1) does not originate from immediate membrane depolarization or cell rupture, as would be expected from a permanently open membrane channel [Harris & Silhavy, 1999]. It is also unexpected that a SecY complex carrying destabilizing prl mutations or truncations in the plug domain, then creating a loosely sealed channel, can be overproduced in the E. coli membrane [Li et al., 2007; Maillard et al., 2007]. In yeast, similar alterations in the Sec61 complex do not produce any obvious growth defect either [Junne et al., 2006; Junne et al. 2007], but the membrane of the endoplasmic reticulum is naturally leaky for small molecules [Le Gall et al., 2004; Lizak et al., 2008; Wonderlin, 2009] (see Section 1.7.3 for discussion). In bacteria, however, tightly regulated ion permeability is crucial to maintaining a correct electro-chemical gradient between the cytosol  82  and the cellular environment. A loosely sealed or open SecY channel might be tolerated if, for example, strong counter-acting pumps compensate for the ion leakage. Alternatively, the tolerance might be explained if the open channel only conducts certain ions. In some circumstances, chloride and proton leakage has been reported during protein transport [Schiebel & Wickner, 1992; Kawasaki et al., 1993], suggesting that the functioning channel might indeed have some particular ionic channel characteristics. Here, we analyzed the ionic specificity of the SecY channel in both resting and active states. We show that the open channel is selective for small monovalent anions, with a strong preference for Cl-; however, K+, Na+, SO42-, PO42- or even H+ cannot easily permeate through it. Mutation of the pore ring residues increases the ionic conductance, but the specificity is maintained. The ionic selectivity is also conserved in the archaeal SecY complex and at the onset of the protein translocation reaction. We propose that the channel forms a dual-stage conductance regulator, with the plug domain acting as a seal and the pore ring contributing to a selective filter. This unique characteristic might explain why imperfectly sealed SecY channels can be tolerated by the cell.  4.2 4.2.1  Materials and methods Biological and chemical reagents Strain KM9 (unc::Tn10, rna10, relA1, spoT1 and metB1) is a derivative of DK8  [Klionsky et al., 1984] and lacks the F1F0 ATP synthase gene. A stable proton gradient can be maintained in membranes lacking the ATP synthase, because this membrane protein complex normally transports protons down its concentration gradient. Plasmid pBAD22 encoding for the wild-type, mutant and archaeal SecY complexes has been described  83  previously [Collinson et al., 2001; van den Berg et al., 2004; Maillard et al., 2007]. Plasmids and methods for preparation of OmpA, proOmpA and LpK (mutation A11K in the leader peptide of proOmpA) were as described in Section 2.2.7. Preparation of SecA was described in Section 2.2.3. IMVs were prepared as described in Section 2.2.6, except that membranes were isolated from E. coli strain KM9 transformed with pBAD22 plasmids, and cell lysis and resuspension of membranes were performed in either: 50 mM Tris-SO4, pH 7.9, 25 mM K2SO4, 10 mM MgSO4 and 0.2 mg/mL BSA (Buffer A), or 50 mM Tris-HCl, pH 7.9, 25 mM KCl, 10 mM MgSO4 and 0.2 mg/mL BSA (Buffer B). The over-production of wild-type and mutant SecY complexes in IMVs was confirmed by Western blot analysis (Appendix G). Nucleotides, ionophores and bismaleimidoethane were purchased from Sigma (www.sigmaaldrich.com). Oxonol VI (bis-3-propyl-5-oxoisoxazol-4-yl pentamethine oxonol) and ACMA (9-amino-6-chloro-2-methoxyacridine) were obtained from Molecular Probes (www.invitrogen.com). Unless stated otherwise, all ions used in ∆ψ dissipation experiments were potassium salts.  4.2.2  Measurement of ∆ψ and ∆pH Measurements were taken using a Varian Cary Eclipse spectro-fluorometer. A typical  experiment that consisted of mixing 40 µg of IMVs and 150 mL of buffer A (50 mM TrisSO4, pH 7.9, 25 mM K2SO4, 10 mM MgSO4 and 0.2 mg/mL BSA) or buffer B (50 mM TrisHCl, pH 7.9, 25 mM KCl, 10 mM MgSO4 and 0.2 mg/mL BSA) in the quartz cuvette was carried out. All experiments were carried out at 25°C and we applied the Henderson– Hasselbalch equation to determine the distribution anions. For example, at pH 8.2, the  84  concentration of divalent K2H(PO4)2- was 4.65 mM, whereas that of monovalent KH2(PO4)was 0.35 mM; by contrast, at pH 7.4, the distribution was 1.93 mM for monovalent KH2(PO4)- and 3.07 mM for divalent K2H(PO4)2-. Unless stated otherwise, the proton gradient was generated using the electron donor NADH, which causes the inner membrane protein NADH oxidase to pump H+ ions into the lumen of the IMV. The membrane potential ∆ψ was followed by fluorescence quenching of oxonol VI (2 µM) at excitation and emission wavelengths of 610 and 640 nm (each of 10 nm slit width), respectively. Calibration of oxonol VI fluorescence with known membrane potentials is described in Appendix H. The proton gradient ∆pH was followed by fluorescence quenching of ACMA (4 µM) at excitation and emission wavelengths of 409 nm (10 nm slit width) and 474 nm (20 nm slit width), respectively. The ∆ψ and ∆pH components were collapsed by the addition of 1 mM of the protonophore carbonyl cyanide m-chlorophenyl hydrazone (CCCP). Fluorescence experiments with oxonol VI were repeated three times and each bar on the graph reports the mean and standard deviation of the values.  4.2.3  Protein translocation alkaline phosphatase assays E. coli strain MPh1061 (phoA61, L14R, a generous gift from Dr J. Beckwith)  expresses PhoA with the mutation L14R in the signal sequence [Derman et al., 1993]. The strain was transformed with plasmid pBAD22 encoding for the SecY complex and mutant forms. Cells were grown to mid-log phase in rich media and induced for 1 h with 0.2% arabinose. Alkaline phosphatase activity assays were carried out as described previously [Derman et al., 1993; Maillard et al., 2007]. One unit of alkaline phosphatase is defined as  85  (OD420 - 1.75 OD550) / (OD600 x volume x time) x 1,000. In vitro protein translocation assays were performed as described in Section 2.2.6.  4.3 4.3.1  Results Ionic specificity of the open SecY channel Conductance experiments were carried out using inner membrane vesicles (IMVs)  prepared from E. coli KM9. An electrochemical membrane potential - inside positive and acidic - was established with NADH and monitored using the voltage-sensitive dye oxonol VI [Clarke & Apell, 1989] (see Appendix I for schematic). The fluorescence emission of this dye is quenched in response to an inside positive charge across membranes. The stability of the electrical potential (∆ψ) was then challenged in the presence of various salts. The IMVs prepared from E. coli wild-type or from E. coli overproducing the SecY complex (termed YWT) were largely impermeable to small or organic ions, as only marginal dissipation of the membrane potential occurred on addition of these salts (Figure 4-1A and Appendix J). Halide anions such as Br- and especially I- have the ability to permeate through phospholipid bilayers and, accordingly, to dissipate the electrical gradient depending on their concentration (Appendix K; Paula et al., 1998). The conductance experiments were then carried out using IMVs enriched with the SecY prlA4 mutant complex (mutation I408N in the pore ring domain). This prl mutation is thought to stabilize the open state of the channel [van den Berg et al., 2004; Smith et al., 2005]. We also used a SecY mutant carrying a deletion in the plug domain (SecY∆33; Maillard et al., 2007). This deletion creates an imperfect plug [Li et al., 2007] and the channel is thus poorly sealed [Saparov et al., 2007]. The conductance measurements were also performed using a mutant carrying a pair of cysteine residues at  86  position 67 of SecY (also known as mutation prlA3) and at position 120 of SecE (Figure 11D and Appendix I). In this mutant (termed SecY67CC), the plug domain is locked away from the centre of the channel on cysteine crosslinking [Tam et al., 2005]. IMVs containing these mutants were all able to generate a stable proton gradient on the addition of NADH (for example, see Figure 4-1B right panel). However, in sharp contrast to the wild-type IMVs, monovalent anions, including chloride, formate, bicarbonate and to a lesser extent acetate, could significantly dissipate the membrane potential across IMVs bearing the mutant complexes (Figure 4-1, B and C). Strikingly, the monovalent phosphate [H2(PO4)-] was permeable, whereas the divalent form [H(PO4)2-] was not. The greatest conductance was for chloride, and titration experiments showed that the ∆ψ could be collapsed with as little as 1 mM of KCl (Appendices J and K). Accordingly, the ∆ψ could not be generated in these membranes if chloride was present in the buffer (Figure 4-2A). By contrast, the membrane potential remained stable in the presence of K+, Na+, Mg2+, SO42- or glutamate (Figure 4-1, B and C). The impermeability to K+ was tested further using potassium acetate in the presence of the ionophore valinomycin, which is able to carry potassium ions across the membrane. A K+-specific electrical potential was generated with a magnitude and stability identical to those of the wild type and mutants (Figure 4-2B). Taken together, these results show that the various mutant SecY channels tested are highly selective for monovalent anions, with a strong preference for chloride.  87  Figure 4-1: Ionic conductance of the SecY complex. (A) Inner membrane vesicles (IMVs) with the wild-type (SecYWT) complex were diluted in buffer A (50 mM Tris–SO4 pH 7.9, 25 mM K2SO4, 10 mM MgSO4, 0.2 mg/mL BSA) in the presence of oxonol VI (2 µM). The membrane potential ∆ψ, inside positive, was generated with NADH (5 mM) which causes fluorescence quenching of oxonol VI. The stability of ∆ψ was then challenged using various potassium salts (each 5 mM) for approximately 1 min and then dissipated with carbonyl cyanide m-chlorophenyl hydrazone (CCCP; 1 µM). The graph (left panel) represents the percentage of ∆ψ that has been dissipated 10 s after the addition of salt. The traces (right panel) are examples of fluorescence recordings. (B, C) IMVs with the indicated SecY mutant complex were analysed as described above. The SecY mutants that were tested are: (B) SecYPrlA4—mutations I408N and F286Y in SecY (recordings shown in the right panel); (C, left panel) SecY∆33—deletion of 33 amino-acid residues in the SecY plug domain; and (C, right panel) SecY67CC + bismaleimidoethane (BMOE) - SecY complex with cysteine residues at position 67 of SecY and position 120 of SecE in a crosslinked state. Error bars indicate the standard deviation between three experiments. This figure was modified from Dalal and Duong., 2009.  88  Figure 4-2: IMVs bearing the SecYPrlA4 complex cannot maintain an electrical potential in the presence of chloride. (A) IMVs containing the SecYWT or SecYPrlA4 complex were incubated in buffer B (50 mM Tris-HCl pH 7.9, 25 mM KCl, 10 mM MgSO4, 0.2 mg/mL BSA) in the presence of oxonol VI. At the indicated time, 5 mM NADH was added in order to generate a membrane potential. (B) IMVs containing the SecYWT or SecYPrlA4 complex were incubated in 50 mM Tris-SO4, pH 7.9. The addition of 200 mM K-acetate and 1 µM valinomycin generates a membrane potential, inside positive, in both membranes. The membrane potential was monitored by oxonol VI fluorescence and partially dissipated with CCCP (1 µM). This figure was modified from Dalal and Duong., 2009.  89  4.3.2  Contribution of the pore ring residues to ionic selectivity The atomic structure of the archaeal SecY channel suggests that the pore is formed by  a ring of six hydrophobic amino-acids, generally isoleucine [van den Berg et al, 2004]. Our results show that the substitution I408N - that is, mutation prlA4 - in E. coli leads to a specific leakage of ions. Thus, we tested the contribution of the other five hydrophobic side chains - that is, Ile82, Ile86, Ile187, Ile191 and Ile278 - after substituting them with asparagines (see Appendix L for structural representation). The IMVs with the mutations I82N and I187N showed ionic conductance as low as that observed for the wild type (Figure 4-3A). By contrast, the other mutants - I86N, I191N and I278N - acquired a strong permeability for chloride, formate, H2(PO4)- and bicarbonate ions (Figure 4-3A). These last three mutants also showed an enhanced in vivo signal-sequence suppression activity and in vitro translocation activity (Figure 4-3, B and C). Thus, the mutations in the pore ring that lead to increased translocation activity also lead to an increased ionic permeability, but the selectivity remains unchanged.  90  Figure 4-3: The SecY pore mutants that are the most active in protein translocation are also the most leaky, but for only monovalent anions. (A) The ionic conductance of inner membrane vesicles with the indicated SecY mutant complex was analysed as described in Figure 4-1A. (B) The in vivo protein translocation activity of the mutants was assayed using the alkaline phosphatase 14R as a reporter. The 14R mutation leads to a defective signal sequence that prevents the preprotein from being recognized by the wild-type SecY complex. 22 refers to cells expressing only endogenous levels of SecY. (C) The in vitro protein translocation activity of the mutants was assayed using 125Ilabelled proOmpA (pOmpA) as described in Section 2.2.6. The quantification was carried out by densitometry scanning using Image J (build 1.40 g). WT, wild type. This figure was modified from Dalal and Duong., 2009. 91  4.3.3  Phenylalanine substitution in the pore ring create a chloride permeability We observed monovalent anion conductance across the SecY channel when  isoleucine residues in the pore ring were replaced by hydrophilic residues (Figure 4-3A). To determine if the membrane seal was dependent solely on the hydrophobicity of these residues, each of the six isoleucines (position 82, 86, 187, 191, 278 and 408; see Appendix L for representation) was mutated into phenylalanine and the mutant SecY complexes were produced in the E. coli inner membrane. After verification that the SecYEG expression levels were similar for all mutants (data not shown) IMVs were prepared and their ionic permeability was measured as described in Figure 4-3. The IMVs containing the wild type complex were rather impermeable to chloride, as indicated by the slow recovery of fluorescence in the presence of 10 mM KCl (Figure 4-4A). In contrast, the presence of a phenylalanine residue at position 278 and 408 caused strong chloride permeability, as judged by the 4 fold increase of the initial rate of fluorescence recovery (Figure 4-4B). The phenylalanine residues placed at the other positions produced chloride leakage but to a lesser extent (Figure 4-4B). As discussed in the previous section, a relationship exists between the chloride permeability of the SecY channel and its translocation activity. The same observation is made here; the phenylalanine mutation that led to the highest rate of chloride leakage also lead to the most active complex, and vice versa (compare Figures 4-4, B and C).  92  Figure 4-4: Effect of phenylalanine residues introduced in the SecY pore ring on the channel ion conductance and protein translocation activity. (A) The ionic conductance of inner membrane vesicles with the indicated SecY mutant complex was analysed as described in Figure 4-1A. (B) The results are expressed as the percentage of the ∆ψ that is dissipated 10 seconds after the addition of 5 mM KCl. The errors bars represent the standard deviation obtained from three independent experiments. (C) The protein translocation activity of the same set of IMVs was assayed using 125Ilabeled proOmpA (pOA) as a preprotein substrate, in the presence of 0.02 g/L of SecA, 0.2 g/L BSA and 1 mM ATP, as previously described in Section 2.2.6. The density of the bands was quantified using Image J software (build 1.40 g). The lane labelled pBAD22 represents the translocation efficiency obtained with IMVs containing endogenous levels of SecY. This figure was modified from Dalal and Duong., 2009.  93  4.3.4  Ionic selectivity is a conserved characteristic The hydrophobicity of the residues that form the pore ring is a conserved  characteristic, but amino-acyl variations exist. The specificity for small molecules was thus tested using IMVs prepared from E. coli KM9 enriched with the evolutionarily distant SecY complex from Methanococcus jannaschii. This archaeal SecY complex does not support in vitro protein translocation (F.D., unpublished data) and membranes enriched for this complex do not significantly conduct any ions (Figure 4-5). However, these membranes become leaky for chloride, formate, H2(PO4)- and bicarbonate when a prlA4-like mutation is introduced into the pore ring of the archaeal SecY complex (mutation L406N) (Figure 4-5). Thus, the ionic selectivity is a conserved characteristic of the SecY channel.  Figure 4-5: The ionic specificity of the SecY complex is a conserved characteristic. A membrane potential ∆ψ was generated and measured as described in Figure 4-1A, but using inner membrane vesicles with the Methanococcus jannaschii SecYE complex, either (A) wild-type (mjEY) or (B) carrying a prlA4-like mutation L406N (mjEY-PrlA4). The graphs represent the percentage of ∆ψ that has been dissipated 10 s after the addition of salt. This figure was modified from Dalal and Duong., 2009.  94  4.3.5  Proton permeability of the open channel As an additional means to probe the specificity of the SecY channel, we used the  ∆pH-dependent dye 9-amino-6-chloro-2- methoxyacridine (ACMA). The magnitude of the proton gradient created in wild-type and mutant membranes was similar and the proton gradient was not dissipated with a solution of KCl (Figure 4-6A). Instead, greater acidification was observed with the mutant membranes, probably because a flow of Cl- ions in the lumen of the vesicles dissipates the membrane potential and allows for further pumping of protons. Finally, we used limiting amounts of NADH to allow the detection of a minor proton outflow (Figure 4-6B). In these conditions, proton leakage could be observed, but it was similar for both the wild-type and all SecY mutant complexes tested, including the archaeal ones (Figure 4-6, B and C). Thus, destabilization of the closed state of the channel (with either a prl mutation or partial deletion of the plug) or the stabilization of its open state (through covalent linkage of the plug) does not render membranes leaky for protons.  95  Figure 4-6: The Cl- conductive SecY mutant channels are impermeable to protons. (A) The inner membrane vesicles (IMVs) with the indicated SecY mutant were incubated in buffer A in the presence of 9-amino-6-chloro-2-methoxyacridine (ACMA; 4 µM). A gradient of protons, inside acidic, was generated with 1 mM NADH and abolished with 1 mM carbonyl cyanide m-chlorophenyl hydrazone (CCCP). The injection of 5 mM KCl causes a Cl- influx in the mutant allowing for further acidification inside the vesicles. (B) IMVs bearing the SecY∆33 complex were prepared as described in (A), but mixed with limiting amounts of NADH (final concentration as indicated). (C) The experiment in (B) was repeated using IMVs with the indicated SecY mutants. For each mutant listed, 100% corresponds to the ∆pH that was generated with 1 mM NADH. This figure was modified from Dalal and Duong., 2009.  96  4.3.6  Ionic conductance of the active channel Schiebel & Wickner (1992) reported that a preprotein jammed across the membrane  creates halide anion permeability whereas Kawasaki et al (1993) reported that a countermovement of protons occurs at the onset of protein translocation. Here, we revisited these pioneering observations using IMVs enriched for the SecYEG or SecYE complex (Figure 47). When IMVs were incubated in Tris-Cl buffer with the SecA ATPase and the preprotein proOmpA, a striking and immediate collapse of the membrane potential was triggered with ATP (Figure 4-7A). Translocation also caused a limited dissipation of ∆ψ in the presence of formate and H2(PO4)- (Figure 4-7B). The large conductance for chloride was, however, not observed on addition of the non-hydrolysable ATP analogue ATPγS, or when the protein substrate had an altered (LpK) or deleted (OmpA) leader peptide (Figure 4-7C, and example of the recordings in Figure 4-8). Thus, SecA, nucleotides and preprotein do not have the capacity on their own to induce chloride permeability and to trigger opening of the channel (this study and Tam et al., 2005). The chloride permeability was partial or null when using membranes with the translocation-inactivating mutation R357E or RPG357EDP introduced into the large cytosolic loop of SecY (Figure 4-7C; Tam et al., 2005). Furthermore, when the translocation was initiated at low temperatures to restrict molecular motions, a corresponding reduction in chloride conductance was recorded (Figure 4-7C and traces in Appendix M). Clearly, the initiation of protein transport, and not only a stalled translocation intermediate [Schiebel & Wickner, 1992], creates chloride permeability across the SecY channel. In the same conditions, we could not detect a counter-movement of protons (Figure 4-7, D and E). Thus, we conclude that protein translocation dissipates the ∆ψ component but not the ∆pH component of the bacterial membrane proton motive force.  97  Figure 4-7: Ionic conductance during protein translocation. (A) 40 µg of inner membrane vesicles (IMVs) with the SecYWT complex was incubated in 150 mL of buffer B (containing 25 mM KCl) in the presence of bovine serum albumin (30 µg), SecA (3 µg) and the protein substrate proOmpA (pOmpA; 3 µg). ∆ψ was generated with 5mM NADH and protein translocation was initiated with 1mM ATP. (B) IMVs with the wild-type SecY complex (40 µg) were tested as described above, but in buffer A in which K2SO4 had been replaced by 25 mM of the indicated salt. The graph represents the percentage of ∆ψ that was dissipated 10 s after the addition of ATP. (C) SecYWT IMVs (40 µg) were tested as described in (A), but the translocation reaction was carried out under the following conditions: 1 mM ATPγS; 3 µg of the translocation-incompetent LpK or OmpA; and 40 µg of IMVs with the indicated SecY mutant, at 4-25 °C. The graph was generated as described in (B). (D, E) IMVs enriched for the SecYEG or SecYE complexes were tested as in (A) but using 4 µM 9-amino-6-chloro-2-methoxyacridine (ACMA) to follow the variations of ∆pH. CCCP, carbonyl cyanide m-chlorophenyl hydrazone; WT, wild type. This figure was modified from Dalal and Duong., 2009. 98  Figure 4-8: Example of traces showing the chloride conductance during translocation. 40 µg of IMVs bearing the SecYWT (panels A, B, C, and D) or SecY mutant complex (panels E and F) were incubated at 25°C in 150 µl of buffer B in the presence of BSA (30 µg), SecA (3 µg) and the protein substrate proOmpA, LpK or OmpA (3 µg each). A ∆ψ was generated with 5 mM NADH and protein translocation was initiated with 1 mM ATP or ATPγS (arrow). The ∆ψ was monitored with oxonol VI fluorescence and quenched with CCCP. This figure was modified from Dalal and Duong., 2009.  99  4.3.7  Binding of trivalent cations to the channel creates a chloride permeability The ionic conductance properties of the SecY channel were next analyzed in the  presence of trivalent cations, recently shown to modify the voltage gating characteristics of the Sec61 channel [Erdman et al., 2009]. IMVs enriched for the wild type SecY complex were diluted in a buffer containing KCl, and the stability of the membrane potential was challenged with aluminum (III) sulfate (Figure 4-9A). Note that the wild type complex is largely impermeable to chloride and the membrane potential remains stable over a large range of KCl concentrations (Figure 4-9B, dark bars). In the presence of Al2(SO4)3 however, there is a rapid dissipation of the membrane potential and with a rate depending on the concentration of the salt (Figure 4-9A, left panel for trace examples). To test whether the collapse of the ∆ψ was caused by the permeation of SO42- through the channel, the experiments were performed in a buffer lacking KCl. Some electric dissipation occurred at high concentration of Al2(SO4)3 (Figure 4-9A, dark bars), but this was much less than the dissipation observed in the presence of KCl (Figure 4-9A, white bars). Another set of measurements using increasing concentrations of KCl in the presence of a fixed amount of Al2(SO4)3 also showed the increased Cl- conductance of the channel in the presence of Al3+ (Figure 4-9B, white bars). Furthermore, the membrane potential across wild type E. coli membranes containing endogenous levels of SecYEG was unaffected by Al2(SO4)3 (data not shown), indicating that the trivalent cation is acting on SecY, not merely on the membrane. Thus, the trivalent cation Al3+ specifically increases the permeability of the channel for chloride anions. However, Al3+ did not cause an obvious effect on the protein translocation activity of the channel in the conditions tested (Figure 4-9C).  100  Figure 4-9: The trivalent cation Al3+ increases the Cl- permeability of the SecY channel. (A) IMVs enriched for the wild type SecY complex were incubated in buffer A containing 10 mM KCl. About 15 seconds after stabilization of the membrane potential, the indicated amount of Al2(SO4)3 was added. The right panel shows the percentage of the ∆ψ that is dissipated 10 seconds after the addition of Al2(SO4)3, in the presence (white bars) or absence (grey bars) of 10 mM KCl. (B) The membrane potential was generated in the presence (white bars) or absence (grey bars) of 2 mM Al2(SO4)3 in buffer A without KCl. The bars represent the percentage of the ∆ψ that is dissipated across the wild type IMVs 10 seconds after the addition of the indicated amount of KCl. (C) The in vitro translocation efficiency of proOmpA was tested with IMVs enriched for the SecY complex in the presence of the indicated concentrations of Al2(SO4)3. This figure was modified from Dalal et al., 2010.  101  4.4 4.4.1  Discussion Possible mechanisms for ionic discrimination This study shows that the SecY channel, despite having a strong ionic conductance  when stabilized in the open state with prl mutations or a locked-open plug [Saparov et al., 2007], is nonetheless able to maintain astonishing ionic specificity. The channel can discriminate between anions and cations, and between the monovalent and divalent forms of the same anion - that is, H2(PO4)- versus H(PO4)2-. Furthermore, assays with formate, acetate and glutamate indicate that the selectivity also operates based on the physical size of the anion. The amino-acyl side chains in and around the pore ring are probably the structures that create the selectivity filter because specificity is maintained when the plug is locked away from the centre of the channel and the permeability of the channel is modulated by the nature of pore-ring residues. It is possible that, as is also proposed for other ion channels [Beckstein & Sansom, 2004], the exclusion of water from the SecY hydrophobic pore environment, together with a discrete distribution of positive and negative charges along the channel walls, would account for the observed selectivity. It is also possible that the opening of the lateral gate would lead to localized re-organization of phospholipids, which, in turn, might enhance the conductance of anions, and especially halide ions, across the lipid bilayer. Site-directed crosslinking studies have indicated that the leader peptide and the lateral groove that forms the leader peptide-binding site are indeed in direct contact with phospholipids [Plath et al., 1998]. We also observed that Al3+ cations increases the chloride conductance of the wild type SecY channel, which is otherwise impermeable to water and small molecules. For the Sec61 complex, it was shown that binding of lanthanum ions (La3+) stabilizes the open  102  conformation of the channel and also changes the selectivity toward negative ions [Erdmann et al., 2009]. In the Erdmann et al. study, molecular modeling suggested that trivalent cations bind to negatively charged amino-acid residues that reside in the SecY lateral gate, implicating them in anion repulsion. Our experiments seem to confirm that the binding of trivalent cations would decrease the overall negative charge density around the pore, and facilitate the conductance of small anions such as chloride. In addition to ionic discrimination by the resting channel, our results show a monovalent anion leakage across the wild-type SecY complex at the onset of SecA driven translocation (Figure 4-7 and Figure 4-8). Thus, the pore ring may only imperfectly seal the SecY channel during preprotein transport. In contrast, a recent work by Park and Rapoport (2011) employed ribosomes bearing a stalled nascent chain (of different lengths) which was inserted into the SecY pore, and this assembly completely prevented chloride ions from leaking across the cell membrane. As the chloride leakage was tested only after the nascent chain was inserted into the channel, is it unknown if the onset of ribosome-driven translocation also creates an anionic permeability. It is possible that the pore ring only seals the channel when the nascent chain is stationary; ions might escape across the channel when ribosome-driven translocation is actually taking place. Nevertheless, the Park and Rapoport study reinforces the idea that the pore ring residues contribute to a gasket-like seal around the polypeptide chain.  4.4.2  Properties of specific residues in the pore ring Our results show that amino-acid substitutions in the SecY pore ring can cause a  monovalent anion conductance. Interestingly, we show that the membrane permeability can  103  be compromised when a phenylalanine residue is introduced into the pore ring, indicating that the hydrophobicity of the pore is not the sole determinant to maintain the membrane seal. The hydrophobic but bulky side chain of phenylalanine may instead cause a steric clash that would distort the shape of the pore and compromise the seal. The location of the mutation within the pore also seems critical since it is mostly the positions 278 and 408, whether substituted by phenylalanine or asparagines (Figure 4-3 and Figure 4-4) which have the most striking effect on channel permeability. The residues at 278 and 408 are located in transmembrane helices 7 and 10 respectively, and the crystal structure of the SecY-SecA complex reveals that these helices are mobile and shifted outward from the center of the channel during opening [Zimmer et al., 2008]. It is possible the substitutions at position 278 and 408 mimic this transition by forcing a similar shift of the helices. In addition, the positions 278 and 408 may contribute the most to the interaction of the pore with the plug domain in order to maintain the channel in the closed state. The current results support the concept that the mutations that destabilize the ionic membrane seal of the channel also decrease the energetic barrier required to activate the translocation channel [Tam et al., 2005; Saparov et al., 2007; and this study].  4.4.3  Possible relationship with ClC chloride transporters In addition to amino-acid side chains in the pore ring and plug, ionic discrimination  might also be conferred by side chains similar to those found in ClC chloride transporters. These hourglass-shaped channels possess a selectivity filter and a constriction located at the mid-point of the membrane, much like the SecY complex. The ClC selectivity filter allows permeation of Cl- because the size and negative charge of this ion interacts favourably with  104  the shared protons from tyrosine and serine hydroxyl groups, as well as by main chain amide groups [Gouaux and MacKinnon, 2005]. Structures of the archaeal and bacterial SecY complexes show that several tyrosine and serine residues exist around the pore or along the channel walls [van den Berg et al., 2004; Zimmer et al., 2008; Tzukasaki et al., 2008], raising the possibility these side chains are involved in ionic discrimination.  4.4.4  Implications of SecY conductance for the cell The ionic specificity of the SecY channel is intriguing and the precise mechanism  needs to be elucidated, especially as, in eukaryotes, the Sec61 homologue seems to be naturally leaky for small molecules including Ca2+ [Le Gall et al., 2004; Wonderlin, 2009; Erdmann et al., 2011]. In bacteria, the observed anion conductance is clearly not required for the translocation efficiency of proteins [Schiebel and Wickner, 1992]. Perhaps, the selectivity of the SecY channel and, most importantly, the exclusion of protons might simply represent a selective advantage in bacteria because most of the metabolic energy is derived from the proton gradient across the inner membrane. This selectivity would allow the cell, for example, to tolerate leaky channels because of a prl or plug mutation leading to an imperfect seal. The limited conductance would also explain why the expression of channels with their plug locked open, although lethal, does not lead to an immediate membrane depolarization and cell rupture [Harris and Silhavy, 1999].  105  Chapter 5: General conclusions and future work Protein translocation is a basic and essential process in all organisms, and understanding the SecY channel interactions and activity is critical for a complete picture of cellular function in bacteria. The findings presented in this thesis provide important information about the quaternary assembly of the SecY complex, and the gating of substrates and small molecules. Because the Sec channel is conserved in all organisms, our results have relevance in a broader context than only bacteria. While the thesis is unified by the investigation of the properties of the SecY complex, it can be broken into three distinct subcategories: First, to understand the nature of the interactions between SecY and its binding partners; second, to understand the overall quaternary assembly of SecY and SecA in post-translational translocation; and third, to understand the mechanism and consequences of ionic conductance through the channel. Much remains to be discovered about the interaction between the SecY complex and Syd and SecA proteins. Previous studies provided genetic evidence to indirectly show that Syd forms an interaction with the SecY complex [Shimoike et al., 2005; Matsuo et al., 1998]. Similarly, the SecY-SecA interaction has only been characterized in liposomes or detergent solution, and only with respect to the binding interface between the two proteins and the apparent binding strength [Zimmer et al., 2008; Gold et al., 2010; Deville et al., 2011]. Since the SecY complex oligomerizes in detergent solution and membranes (see Section 1.6), the stoichiometry between SecY and either SecA or Syd could not be precisely measured in these earlier assays. Our work in Chapters 2 and 3 now directly reveals, through native-PAGE and other analysis of Nanodisc reconstituted SecY complexes, the mode and stoichiometry of the SecY-Syd and SecY-SecA interactions. Light scattering of the Nanodisc embedded SecY  106  monomer in complex with Syd indicated a 1:1 binding stoichiometry (Figure 2-3). Similarly, gel-shift assays of the Nd-Y or Nd-Y2 complexes indicated that only one SecA copy is bound to either the SecY monomer or dimer (Figure 2-3). This last result seems to support the ‘fraternal twin’ model in which the SecY dimer is proposed to form a binding platform for only one SecA molecule (Figure 1-6; Osborne and Rapoport, 2007). In addition to clarifying the stoichiometry, we demonstrate that the SecY complex embedded in Nanodiscs faithfully recreates the binding site for Syd and SecA because altering the SecY cytosolic loops, either by deletion or point mutations, leads to a corresponding reduction in affinity of these two proteins for the channel (Figure 2-1 and Figure 3-4). However, we detected no obvious difference in SecA binding between the SecY monomer and dimer by native-PAGE or affinity pull-down techniques (Figure 2-2 and Figure 3-4). This is surprising since in detergent solution, the SecY dimer has a ~12 fold greater affinity for SecA compared to the SecY monomer [Deville et al., 2011]. To resolve this conflict, quantitative measurements of the interaction between Nanodisc-SecYEG particles and binding partners should be performed with more sensitive techniques such as surface plasmon resonance (SPR) or isothermal titration calorimetry (ITC). Preliminary results show a tight binding between the SecY monomer and Syd or SecA (nano-molar range), indicating that SPR or ITC will be useful for determining the binding affinity of SecA to the SecY dimer [Dalal and Duong; unpublished results]. As shown in Chapter 2, the interaction between the SecY cytosolic loops and channel partners is likely electrostatic in nature. Amongst bacteria, there is considerable sequence conservation within the 4th and 5th SecY cytosolic loops, approximately 60% similarity and 25% identity [van den Berg et al., 2004]. Several of the positively charged  107  residues in this region are conserved even in the eukaryotic versions of the channel (see Figure 1-7), and charge reversal mutations in these regions impair binding to channel partners (see Section 1.9). We showed by crystallography and disulphide crosslinking that Syd protein interacts with the SecY loops via the concave electronegative ‘groove’ (Figure 2-4 and Figure 2-5). Further evidence for an electrostatic interaction was provided by sucrose density centrifugation experiments, where the binding strength of Syd to Nanodisc-SecY particles was reduced in the presence of salt (Figure 2-1). Although these results clearly define the interaction between SecY and Syd, the interaction between SecA and the channel is more complex. The SecYEG-SecA crystal structure shows that the SecY cytosolic loops interact with SecA polypeptide crosslinking domain [Zimmer et al., 2008], but this is by no means the only electronegative surface on SecA (see Figure 1-7). Several other domains of SecA are involved in binding to the SecY cytosolic loops, including the first nucleotide binding domain and helical scaffold domain (see Figure 1-3). Whether these SecA domains also form electrostatic interactions with SecY is currently unknown. In addition, the crystal structure is only one snapshot of many different possible conformations of both the channel and SecA, and should be interpreted with caution. For example, it is puzzling why the R357 residue on SecY, which is critical for channel function [Mori and Ito, 2001], does not contact SecA in the SecYEG-SecA structure [Zimmer et al., 2008 and Figure 1-6]. More work will need to be done to understand how exactly SecA docks onto the SecY complex. In addition to clarifying the molecular basis of the interaction between SecY and its binding partners, we also found a relationship between Syd and bacterial taxonomy. It is remarkable that all Syd containing bacterial species that we identified in the sequence  108  alignment (Figure 2-5) belong to the gammaproteobacteria class that contained SecE with 3 TMS. Gammaproteobacteria species were originally classified by sequence similarity in the 16S ribosomal RNA between these species [Williams et al., 2010]. As discussed in Section 2.3.5, these bacteria are Gram-negative (containing inner and outer membranes) and comprise many human pathogens from the Enterobacteriaceae family, including E. coli and species from Samonella, Vibrio, Shigella, Klebsiella and Yersinia. The gammaproteobacteria class is extremely diverse and the reason these bacteria express Syd, and the relevance for protein translocation, remains unclear. Since protein translocation is an essential for delivery of some virulence factors to host cells, perhaps an additional level of quality control for SecY is beneficial for these pathogens. Furthermore, Syd may have additional binding partners in gammaproteobacteria that have yet to be determined. We are currently using mass spectrometry based approaches to search for new binding partners of Syd in E. coli cells. The uncontrollable oligomerization of the channel in membranes has created uncertainty about the requirement of SecY dimers in protein translocation. In Chapter 3, we show unequivocally with Nanodiscs that a dimeric channel assembly is required to stimulate SecA ATPase activity. The ATPase was strictly dependent on a preprotein containing a correct signal sequence (Figure 3-1E), demonstrating that at least the ‘triggering’ step (see Section 3.4.2) has been recreated with Nanodiscs.  Although we did not show that the  subsequent ‘trapping’ and ‘secretion’ steps were taking place across the SecY dimer in Nanodiscs, the trans-complementation assay (Figure 3-8) demonstrated that both SecY copies are required for activity, and the function of each is different. This work provides the much needed support the fraternal twins model proposed by Osborne & Rapoport (2007) (Figure 1-4A), where SecA and the preprotein are engaged with distinct channels. An  109  interesting follow up study would involve introducing a cysteine into the signal sequence binding site (at position 97) of SecYFF or SecYE. During in vitro translocation across IMVs or liposomes bearing both SecY complexes, a disulphide bond should only form between Pho1-202-5C and the SecYE mutant, but not the SecYFF mutant which has a defective channel pore. The results from Chapter 3 seems to indicate that the SecY dimer is the minimal unit that is required for protein translocation, in direct contrast with a recent report from Kedrov et al. (2011) suggesting that a monomer is the functional unit. As discussed in Section 3.4.1, Kedrov et al. used experimental conditions that highly favour SecYEG monomerization (dilute SecY protein concentration), and transient interactions with a second SecY complex were not ruled out. In addition, the kinetics of protein translocation was not studied and it is possible that SecY dimers would support a much higher rate of transport, as they do in the other recent single molecule analysis of SecY in liposomes [Deville et al., 2010]. The increasingly complex in vitro analysis of SecY in these liposome assays may heavily influence the oligomeric state of SecY, channel activity and experimental outcome. Our results with Nanodiscs also do not exclude that higher oligomers (i.e. tetramers) of SecYEG have some role in protein translocation. Larger discs with longer scaffold protein belts and greater amount of incorporated lipids will be required to test the activity of SecY oligomers. The development of larger discs is currently underway in our laboratory. In addition to clarifying the function of the two SecY copies, we clearly demonstrate the requirement of acidic lipids in protein translocation by controlling the environment surrounding the SecY complex (Figure 3-1). In this respect, the Nanodisc has a significant advantage over detergent solution and liposomes because the oligomeric state of a membrane  110  protein and lipid contributions can be investigated simultaneously. It is unknown how the anionic PG head-group interacts with SecY or SecA, but our results convincing show that acidic lipids only exert their effect in the context of the SecY dimer. It is interesting that cardiolipin, which is formed from two covalently linked PG molecules, increases SecY dimerization in detergent solution (see Section 1.6.2; Gold et al., 2010). In Nanodiscs, the SecY dimer reconstituted in discs in the presence of CL stimulates even more translocation ATPase activity than with PG [Dalal and Duong, unpublished results]. Models of the SecYEG dimer have led to the proposal that cardiolipin might somehow facilitate dimerization by binding at the interface of two SecY subunits [Gold et al., 2010]. Other acidic lipids, such as phosphatidic acid (PA) and phosphatidyl serine (PS), should be incorporated with SecY into discs and tested for their ability to stimulate the SecA translocation ATPase. Finally, our work in Chapter 4 provides a rationale for the observed ionic conductance through the bacterial SecY channel. Compensating for specific monovalent anion leakage would be far easier for the cell than mitigating general membrane permeability. The permeability may simply represent and imperfect membrane seal that exclude protons and other ions, but is insufficient to completely block all anions from crossing the channel. In this way, the proton gradient, which is used for ATP production by the cell, can be preserved. Although our work, and the study from Park and Rapoport (2011), demonstrates that the pore ring residues and plug domain are involved in forming this membrane seal, the contribution of other channel residues in ionic discrimination cannot be excluded. This is illustrated by the effect of binding trivalent cations onto the Sec61/SecY channel [Erdmann et al., 2009; and Figure 4-9], which now implicate several negatively  111  charged lateral gate residues (mostly glutamate and aspartate) in sealing the wild-type channel against anions. It remains a mystery which amino-acid residues contribute to the exclusion of protons and other cations from the channel. In addition to mutagenesis, insight into the mechanism of ionic conductance and translocation might be provided by screening for small molecules that interact with SecY. In collaboration with the Center for Drug Research and Development (UBC), we have developed an assay to screen for small molecules that can block the chloride conductance across the SecY I278N mutant in membrane vesicles. To date, we have screened thousands of drugs and identified several molecules, at micro-molar concentration, that block the chloride conductance, but apparently have no effect on in vitro protein translocation [N. Honson and K. Dalal, 2010, unpublished results]. One such compound, termed amodiaquine (4-[(7-chloroquinolin-4-yl)amino]-2-[(diethylamino)methyl]phenol), shows particular promise in sealing the mutant I278N SecY channel. Finding a compound that affects both SecY conductance and channel activity will help to better define the relationship between these two processes. A similar high-throughput screening approach that instead targets SecA or the channel pore could be an additional means to dissect the translocation mechanism. For example, inhibition of the ATPase enzyme activity of SecA is an attractive target for discovery of small antibacterial molecules [Segers et al., 2011]. In combination with recent literature, the work presented here helps to define the mechanism and consequences of protein translocation in bacteria [Alami et al., 2007; Osborne and Rapoport, 2007; Deville et al; 2011; Park and Rapoport, 2011]. Many of our conclusions regarding the SecY complex will help to guide work on the eukaryotic and archaeal versions of the channel. In addition, Nanodiscs will be useful to investigate the  112  interaction with other binding partners of the SecY complex, including FtsY and the ribosome. Given the inherent difficulty of analyzing protein interactions in the lipidic environment, questions such as the oligomeric state of SecY and Sec61 channels will likely remain an area of important investigations, largely depending on parallel technical advances in membrane biology. For example, recent studies with Nanodiscs have attempted to understand the structure of the SecY complex and other membrane proteins by analysis with cryo-EM [Frauenfeld et al., 2011] or NMR [Gluck et al., 2009]. Using our results as a foundation, the application of these techniques to study the Nanodisc-reconstituted SecY complex, free from the non-compliant environment of the membrane lipid bilayer, opens up many prospects for future research.  113  References Akita, M., Shinkai, A., Matsuyama, S., and Mizushima, S. (1991). SecA, an essential component of the secretory machinery of Escherichia coli, exists as homodimer. Biochem. Biophys. Res. Commun. 1, 211-216. Alami, M., Dalal, K., Lelj-Garolla, B., Sligar, S.G., and Duong, F. (2007). Nanodiscs unravel the interaction between the SecYEG channel and its cytosolic partner SecA. EMBO J. 8, 1995-2004. Auclair, S.M., Moses, J.P., Musial-Siwek, M., Kendall, D.A., Oliver, D.B., and Mukerji, I. (2010). Mapping of the signal peptide-binding domain of Escherichia coli SecA using Forster resonance energy transfer. Biochemistry 4, 782-792. Bauer, B.W., and Rapoport, T.A. (2009). Mapping polypeptide interactions of the SecA ATPase during translocation. Proc. Natl. Acad. Sci. U. S. A. 49, 20800-20805. Bayburt, T.H., and Sligar, S.G. (2003). Self-assembly of single integral membrane proteins into soluble nanoscale phospholipid bilayers. Protein Sci. 11, 2476-2481. Bechtluft, P., Nouwen, N., Tans, S.J., and Driessen, A.J. (2010). SecB--a chaperone dedicated to protein translocation. Mol. Biosyst 4, 620-627. Bechtluft, P., van Leeuwen, R.G., Tyreman, M., Tomkiewicz, D., Nouwen, N., Tepper, H.L., Driessen, A.J., and Tans, S.J. (2007). Direct observation of chaperone-induced changes in a protein folding pathway. Science 5855, 1458-1461. Becker, T., Bhushan, S., Jarasch, A., Armache, J.P., Funes, S., Jossinet, F., Gumbart, J., Mielke, T., Berninghausen, O., Schulten, K. et al. (2009). Structure of monomeric yeast and mammalian Sec61 complexes interacting with the translating ribosome. Science 5958, 1369-1373. Beckmann, R., Bubeck, D., Grassucci, R., Penczek, P., Verschoor, A., Blobel, G., and Frank, J. (1997). Alignment of conduits for the nascent polypeptide chain in the ribosomeSec61 complex. Science 5346, 2123-2126. Beckmann, R., Spahn, C.M., Eswar, N., Helmers, J., Penczek, P.A., Sali, A., Frank, J., and Blobel, G. (2001). Architecture of the protein-conducting channel associated with the translating 80S ribosome. Cell 3, 361-372. Beckstein, O., and Sansom, M.S. (2004). The influence of geometry, surface character, and flexibility on the permeation of ions and water through biological pores. Phys. Biol. 1-2, 42-52. 114  Benach, J., Chou, Y.T., Fak, J.J., Itkin, A., Nicolae, D.D., Smith, P.C., Wittrock, G., Floyd, D.L., Golsaz, C.M., Gierasch, L.M., and Hunt, J.F. (2003). Phospholipid-induced monomerization and signal-peptide-induced oligomerization of SecA. J. Biol. Chem. 6, 3628-3638. Bessonneau, P., Besson, V., Collinson, I., and Duong, F. (2002). The SecYEG preprotein translocation channel is a conformationally dynamic and dimeric structure. EMBO J. 5, 995-1003. Bondar, A.N., del Val, C., Freites, J.A., Tobias, D.J., and White, S.H. (2010). Dynamics of SecY translocons with translocation-defective mutations. Structure 7, 847-857. Bostina, M., Mohsin, B., Kuhlbrandt, W., and Collinson, I. (2005). Atomic model of the E. coli membrane-bound protein translocation complex SecYEG. J. Mol. Biol. 5, 10351043. Braig, D., Bar, C., Thumfart, J.O., and Koch, H.G. (2009). Two cooperating helices constitute the lipid-binding domain of the bacterial SRP receptor. J. Mol. Biol. 3, 401413. Breukink, E., Nouwen, N., van Raalte, A., Mizushima, S., Tommassen, J., and de Kruijff, B. (1995). The C terminus of SecA is involved in both lipid binding and SecB binding. J. Biol. Chem. 14, 7902-7907. Breyton, C., Haase, W., Rapoport, T.A., Kuhlbrandt, W., and Collinson, I. (2002). Threedimensional structure of the bacterial protein-translocation complex SecYEG. Nature 6898, 662-665. Brundage, L., Hendrick, J.P., Schiebel, E., Driessen, A.J., and Wickner, W. (1990). The purified E. coli integral membrane protein SecY/E is sufficient for reconstitution of SecA-dependent precursor protein translocation. Cell 4, 649-657. 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 & NMR system: A new software suite for macromolecular structure determination. Acta Crystallogr. D Biol. Crystallogr. Pt 5, 905-921. Cannon, K.S., Or, E., Clemons, W.M.,Jr, Shibata, Y., and Rapoport, T.A. (2005). Disulfide bridge formation between SecY and a translocating polypeptide localizes the translocation pore to the center of SecY. J. Cell Biol. 2, 219-225. Clarke, R.J., and Apell, H.J. (1989). A stopped-flow kinetic study of the interaction of potential-sensitive oxonol dyes with lipid vesicles. Biophys. Chem. 3, 225-237. Collaborative Computational Project, Number 4. (1994). The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D Biol. Crystallogr. Pt 5, 760-763. 115  Collinson, I., Breyton, C., Duong, F., Tziatzios, C., Schubert, D., Or, E., Rapoport, T., and Kuhlbrandt, W. (2001). Projection structure and oligomeric properties of a bacterial core protein translocase. EMBO J. 10, 2462-2471. Cunningham, K., and Wickner, W. (1989). Specific recognition of the leader region of precursor proteins is required for the activation of translocation ATPase of Escherichia coli. Proc. Natl. Acad. Sci. U. S. A. 22, 8630-8634. Dalal, K., and Duong, F. (2011). The SecY complex: conducting the orchestra of protein translocation. Trends Cell Biol. 9, 506-514. Dalal, K., Bao, H., and Duong, F. (2010). Modulation of the SecY channel permeability by pore mutations and trivalent cations. Channels (Austin) 2, Dalal, K., and Duong, F. (2010). Reconstitution of the SecY translocon in nanodiscs. Methods Mol. Biol. 145-156. Dalal, K., and Duong, F. (2009). The SecY complex forms a channel capable of ionic discrimination. EMBO Rep. 7, 762-768. Dalal, K., Nguyen, N., Alami, M., Tan, J., Moraes, T.F., Lee, W.C., Maurus, R., Sligar, S.S., Brayer, G.D., and Duong, F. (2009). Structure, binding, and activity of Syd, a SecYinteracting protein. J. Biol. Chem. 12, 7897-7902. Das, S., and Oliver, D.B. (2011). Mapping of the SecA{middle dot}SecY and SecA{middle dot}SecG Interfaces by Site-directed in Vivo Photocross-linking. J. Biol. Chem. 14, 12371-12380. de Keyzer, J., van der Sluis, E.O., Spelbrink, R.E., Nijstad, N., de Kruijff, B., Nouwen, N., van der Does, C., and Driessen, A.J. (2005). Covalently dimerized SecA is functional in protein translocation. J. Biol. Chem. 42, 35255-35260. de Vrije, T., de Swart, R.L., Dowhan, W., Tommassen, J., and de Kruijff, B. (1988). Phosphatidylglycerol is involved in protein translocation across Escherichia coli inner membranes. Nature 6178, 173-175. Denisov, I.G., Grinkova, Y.V., Lazarides, A.A., and Sligar, S.G. (2004). Directed selfassembly of monodisperse phospholipid bilayer Nanodiscs with controlled size. J. Am. Chem. Soc. 11, 3477-3487. Derman, A.I., Puziss, J.W., Bassford, P.J.,Jr, and Beckwith, J. (1993). A signal sequence is not required for protein export in prlA mutants of Escherichia coli. EMBO J. 3, 879888.  116  Deville, K., Gold, V.A., Robson, A., Whitehouse, S., Sessions, R.B., Baldwin, S.A., Radford, S.E., and Collinson, I. (2011). The oligomeric state and arrangement of the active bacterial translocon. J. Biol. Chem. 6, 4659-4669. du Plessis, D.J., Nouwen, N., and Driessen, A.J. (2011). The Sec translocase. Biochim. Biophys. Acta 3, 851-865. Duong, F. (2007). Cell biology: fraternal twins. Nature 7137, 741-743. Duong, F. (2003). Binding, activation and dissociation of the dimeric SecA ATPase at the dimeric SecYEG translocase. EMBO J. 17, 4375-4384. Economou, A. (2008) Clamour for a Kiss. Nature 455, 879–880. Economou, A., and Wickner, W. (1994). SecA promotes preprotein translocation by undergoing ATP-driven cycles of membrane insertion and deinsertion. Cell 5, 835843. Egea, P.F., and Stroud, R.M. (2010). Lateral opening of a translocon upon entry of protein suggests the mechanism of insertion into membranes. Proc. Natl. Acad. Sci. U. S. A. Emsley, P., and Cowtan, K. (2004). Coot: model-building tools for molecular graphics. Acta Crystallogr. D Biol. Crystallogr. Pt 12 Pt 1, 2126-2132. Erdmann, F., Jung, M., Eyrisch, S., Lang, S., Helms, V., Wagner, R., and Zimmermann, R. (2009). Lanthanum ions inhibit the mammalian Sec61 complex in its channel dynamics and protein transport activity. FEBS Lett. 14, 2359-2364. Erdmann, F., Jung, M., Maurer, P., Harsman, A., Zimmermann, R., and Wagner, R. (2010). The mammalian and yeast translocon complexes comprise a characteristic Sec61 channel. Biochem. Biophys. Res. Commun. 3, 714-720. Erdmann, F., Schauble, N., Lang, S., Jung, M., Honigmann, A., Ahmad, M., Dudek, J., Benedix, J., Harsman, A., Kopp, A. et al. (2011). Interaction of calmodulin with Sec61alpha limits Ca2+ leakage from the endoplasmic reticulum. EMBO J. 1, 17-31. Erlandson, K.J., Miller, S.B., Nam, Y., Osborne, A.R., Zimmer, J., and Rapoport, T.A. (2008). A role for the two-helix finger of the SecA ATPase in protein translocation. Nature 7215, 984-987. Fekkes, P., de Wit, J.G., Boorsma, A., Friesen, R.H., and Driessen, A.J. (1999). Zinc stabilizes the SecB binding site of SecA. Biochemistry 16, 5111-5116. Frauenfeld, J., Gumbart, J., Sluis, E.O., Funes, S., Gartmann, M., Beatrix, B., Mielke, T., Berninghausen, O., Becker, T., Schulten, K., and Beckmann, R. (2011). Cryo-EM  117  structure of the ribosome-SecYE complex in the membrane environment. Nat. Struct. Mol. Biol. 5, 614-621. Gelis, I., Bonvin, A.M., Keramisanou, D., Koukaki, M., Gouridis, G., Karamanou, S., Economou, A., and Kalodimos, C.G. (2007). Structural basis for signal-sequence recognition by the translocase motor SecA as determined by NMR. Cell 4, 756-769. Gluck, J.M. et al. Integral membrane proteins in nanodiscs can be studied by solution NMR spectroscopy. J. Am. Chem. Soc. 131, 12060-12061 (2009). Gold, V.A., Robson, A., Clarke, A.R., and Collinson, I. (2007). Allosteric regulation of SecA: magnesium-mediated control of conformation and activity. J. Biol. Chem. 24, 17424-17432. Gold, V.A., Robson, A., Bao, H., Romantsov, T., Duong, F., and Collinson, I. (2010). The action of cardiolipin on the bacterial translocon. Proc. Natl. Acad. Sci. U. S. A. 22, 10044-10049. Gorlich, D., and Rapoport, T.A. (1993). Protein translocation into proteoliposomes reconstituted from purified components of the endoplasmic reticulum membrane. Cell 4, 615-630. Gouaux, E., and Mackinnon, R. (2005). Principles of selective ion transport in channels and pumps. Science 5753, 1461-1465. Gouridis, G., Karamanou, S., Gelis, I., Kalodimos, C.G., and Economou, A. (2009). Signal peptides are allosteric activators of the protein translocase. Nature 7271, 363-367. Gumbart, J., and Schulten, K. (2007). Structural determinants of lateral gate opening in the protein translocon. Biochemistry 39, 11147-11157. Halic, M., and Beckmann, R. (2005). The signal recognition particle and its interactions during protein targeting. Curr. Opin. Struct. Biol. 1, 116-125. Hamman, B.D., Hendershot, L.M., and Johnson, A.E. (1998). BiP maintains the permeability barrier of the ER membrane by sealing the lumenal end of the translocon pore before and early in translocation. Cell 6, 747-758. Harris, C.R., and Silhavy, T.J. (1999). Mapping an interface of SecY (PrlA) and SecE (PrlG) by using synthetic phenotypes and in vivo cross-linking. J. Bacteriol. 11, 3438-3444. Hendrick, J.P., and Wickner, W. (1991). SecA protein needs both acidic phospholipids and SecY/E protein for functional high-affinity binding to the Escherichia coli plasma membrane. J. Biol. Chem. 36, 24596-24600.  118  Huber, D., Cha, M.I., Debarbieux, L., Planson, A.G., Cruz, N., Lopez, G., Tasayco, M.L., Chaffotte, A., and Beckwith, J. (2005). A selection for mutants that interfere with folding of Escherichia coli thioredoxin-1 in vivo. Proc. Natl. Acad. Sci. U. S. A. 52, 18872-18877. Huber, D., Rajagopalan, N., Preissler, S., Rocco, M.A., Merz, F., Kramer, G., and Bukau, B. (2011). SecA Interacts with Ribosomes in Order to Facilitate Posttranslational Translocation in Bacteria. Mol. Cell 3, 343-353. Hunt, J.F., Weinkauf, S., Henry, L., Fak, J.J., McNicholas, P., Oliver, D.B., and Deisenhofer, J. (2002). Nucleotide control of interdomain interactions in the conformational reaction cycle of SecA. Science 5589, 2018-2026. Junne, T., Schwede, T., Goder, V., and Spiess, M. (2006). The plug domain of yeast Sec61p is important for efficient protein translocation, but is not essential for cell viability. Mol. Biol. Cell 9, 4063-4068. Junne, T., Schwede, T., Goder, V., and Spiess, M. (2007). Mutations in the Sec61p channel affecting signal sequence recognition and membrane protein topology. J. Biol. Chem. 45, 33201-33209. Kaufmann, A., Manting, E.H., Veenendaal, A.K., Driessen, A.J., and van der Does, C. (1999). Cysteine-directed cross-linking demonstrates that helix 3 of SecE is close to helix 2 of SecY and helix 3 of a neighboring SecE. Biochemistry 28, 9115-9125. Kawasaki, S., Mizushima, S., and Tokuda, H. (1993). Membrane vesicles containing overproduced SecY and SecE exhibit high translocation ATPase activity and countermovement of protons in a SecA- and presecretory protein-dependent manner. J. Biol. Chem. 11, 8193-8198. Kedrov, A., Kusters, I., Krasnikov, V.V., and Driessen, A.J. (2011). A single copy of SecYEG is sufficient for preprotein translocation. EMBO J. 30, 4387-4397 Kihara, A., Akiyama, Y., and Ito, K. (1995). FtsH is required for proteolytic elimination of uncomplexed forms of SecY, an essential protein translocase subunit. Proc. Natl. Acad. Sci. U. S. A. 10, 4532-4536. Klionsky, D.J., Brusilow, W.S., and Simoni, R.D. (1984). In vivo evidence for the role of the epsilon subunit as an inhibitor of the proton-translocating ATPase of Escherichia coli. J. Bacteriol. 3, 1055-1060. Kramer, G., Boehringer, D., Ban, N., and Bukau, B. (2009). The ribosome as a platform for co-translational processing, folding and targeting of newly synthesized proteins. Nat. Struct. Mol. Biol. 6, 589-597.  119  Kuhn, P., Weiche, B., Sturm, L., Sommer, E., Drepper, F., Warscheid, B., Sourjik, V., and Koch, H.G. (2011). The bacterial SRP receptor, SecA and the ribosome use overlapping binding sites on the SecY translocon. Traffic Kusters, I., van den Bogaart, G., Kedrov, A., Krasnikov, V., Fulyani, F., Poolman, B., and Driessen, A.J. (2011). Quaternary structure of SecA in solution and bound to SecYEG probed at the single molecule level. Structure 3, 430-439. Lam, V.Q., Akopian, D., Rome, M., Henningsen, D., and Shan, S.O. (2010). Lipid activation of the signal recognition particle receptor provides spatial coordination of protein targeting. J. Cell Biol. 4, 623-635. Le Gall, S., Neuhof, A., and Rapoport, T. (2004). The endoplasmic reticulum membrane is permeable to small molecules. Mol. Biol. Cell 2, 447-455. Li, W., Schulman, S., Boyd, D., Erlandson, K., Beckwith, J., and Rapoport, T.A. (2007). The plug domain of the SecY protein stabilizes the closed state of the translocation channel and maintains a membrane seal. Mol. Cell 4, 511-521. Lill, R., Dowhan, W., and Wickner, W. (1990). The ATPase activity of SecA is regulated by acidic phospholipids, SecY, and the leader and mature domains of precursor proteins. Cell 2, 271-280. Lizak, B., Csala, M., Benedetti, A., and Banhegyi, G. (2008). The translocon and the nonspecific transport of small molecules in the endoplasmic reticulum (Review). Mol. Membr. Biol. 2, 95-101. Lycklama A Nijeholt, J.A., Bulacu, M., Marrink, S.J., and Driessen, A.J. (2010). Immobilization of the plug domain inside the SecY channel allows unrestricted protein translocation. J. Biol. Chem. 31, 23747-23754. Maillard, A.P., Lalani, S., Silva, F., Belin, D., and Duong, F. (2007). Deregulation of the SecYEG translocation channel upon removal of the plug domain. J. Biol. Chem. 2, 1281-1287. Manting, E.H., van Der Does, C., Remigy, H., Engel, A., and Driessen, A.J. (2000). SecYEG assembles into a tetramer to form the active protein translocation channel. EMBO J. 5, 852-861. Matsuo, E., and Ito, K. (1998). Genetic analysis of an essential cytoplasmic domain of Escherichia coli SecY based on resistance to Syd, a SecY-interacting protein. Mol. Gen. Genet. 3, 240-249. Matsuo, E., Mori, H., and Ito, K. (2003). Interfering mutations provide in vivo evidence that Escherichia coli SecE functions in multimeric states. Mol. Genet. Genomics 6, 808815. 120  Matsuo, E., Mori, H., Shimoike, T., and Ito, K. (1998). Syd, a SecY-interacting protein, excludes SecA from the SecYE complex with an altered SecY24 subunit. J. Biol. Chem. 30, 18835-18840. Menetret, J.F., Hegde, R.S., Aguiar, M., Gygi, S.P., Park, E., Rapoport, T.A., and Akey, C.W. (2008). Single copies of Sec61 and TRAP associate with a nontranslating mammalian ribosome. Structure 7, 1126-1137. Menetret, J.F., Hegde, R.S., Heinrich, S.U., Chandramouli, P., Ludtke, S.J., Rapoport, T.A., and Akey, C.W. (2005). Architecture of the ribosome-channel complex derived from native membranes. J. Mol. Biol. 2, 445-457. Menetret, J.F., Schaletzky, J., Clemons, W.M.,Jr, Osborne, A.R., Skanland, S.S., Denison, C., Gygi, S.P., Kirkpatrick, D.S., Park, E., Ludtke, S.J., Rapoport, T.A., and Akey, C.W. (2007). Ribosome binding of a single copy of the SecY complex: implications for protein translocation. Mol. Cell 6, 1083-1092. Meyer, T.H., Menetret, J.F., Breitling, R., Miller, K.R., Akey, C.W., and Rapoport, T.A. (1999). The bacterial SecY/E translocation complex forms channel-like structures similar to those of the eukaryotic Sec61p complex. J. Mol. Biol. 4, 1789-1800. Mitra, K., Schaffitzel, C., Shaikh, T., Tama, F., Jenni, S., Brooks, C.L.,3rd, Ban, N., and Frank, J. (2005). Structure of the E. coli protein-conducting channel bound to a translating ribosome. Nature 7066, 318-324. Mori, H., and Ito, K. (2001). An essential amino acid residue in the protein translocation channel revealed by targeted random mutagenesis of SecY. Proc. Natl. Acad. Sci. U. S. A. 9, 5128-5133. Nishiyama, K., Suzuki, H., and Tokuda, H. (2000). Role of the non-essential region encompassing the N-terminal two transmembrane stretches of Escherichia coli SecE. Biosci. Biotechnol. Biochem. 10, 2121-2127. Or, E., Boyd, D., Gon, S., Beckwith, J., and Rapoport, T. (2005). The bacterial ATPase SecA functions as a monomer in protein translocation. J. Biol. Chem. 10, 9097-9105. Or, E., Navon, A., and Rapoport, T. (2002). Dissociation of the dimeric SecA ATPase during protein translocation across the bacterial membrane. EMBO J. 17, 4470-4479. Osborne, A.R., and Rapoport, T.A. (2007). Protein translocation is mediated by oligomers of the SecY complex with one SecY copy forming the channel. Cell 1, 97-110. Osborne, A.R., Rapoport, T.A., and van den Berg, B. (2005). Protein translocation by the Sec61/SecY channel. Annu. Rev. Cell Dev. Biol. 529-550.  121  Park, E., and Rapoport, T.A. (2011). Preserving the membrane barrier for small molecules during bacterial protein translocation. Nature 7346, 239-242. Parlitz, R., Eitan, A., Stjepanovic, G., Bahari, L., Bange, G., Bibi, E., and Sinning, I. (2007). Escherichia coli signal recognition particle receptor FtsY contains an essential and autonomous membrane-binding amphipathic helix. J. Biol. Chem. 44, 32176-32184. Paula, S., Volkov, A.G., and Deamer, D.W. (1998). Permeation of halide anions through phospholipid bilayers occurs by the solubility-diffusion mechanism. Biophys. J. 1, 319-327. Perrakis, A., Morris, R., and Lamzin, V.S. (1999). Automated protein model building combined with iterative structure refinement. Nat. Struct. Biol. 5, 458-463. Plath, K., Mothes, W., Wilkinson, B.M., Stirling, C.J., and Rapoport, T.A. (1998). Signal sequence recognition in posttranslational protein transport across the yeast ER membrane. Cell 6, 795-807. Rapoport, T.A. (2007). Protein translocation across the eukaryotic endoplasmic reticulum and bacterial plasma membranes. Nature 7170, 663-669. Randall, L.L., and Henzl, M.T. (2010). Direct identification of the site of binding on the chaperone SecB for the amino terminus of the translocon motor SecA. Protein Sci. 6, 1173-1179. Ritchie, T.K., Grinkova, Y.V., Bayburt, T.H., Denisov, I.G., Zolnerciks, J.K., Atkins, W.M., and Sligar, S.G. (2009). Chapter 11 - Reconstitution of membrane proteins in phospholipid bilayer nanodiscs. Methods Enzymol. 211-231. Robson, A., Gold, V.A., Hodson, S., Clarke, A.R., and Collinson, I. (2009). Energy transduction in protein transport and the ATP hydrolytic cycle of SecA. Proc. Natl. Acad. Sci. U. S. A. 13, 5111-5116. Saparov, S.M., Erlandson, K., Cannon, K., Schaletzky, J., Schulman, S., Rapoport, T.A., and Pohl, P. (2007). Determining the conductance of the SecY protein translocation channel for small molecules. Mol. Cell 4, 501-509. Sardis, M.F., and Economou, A. (2010). SecA: a tale of two protomers. Mol. Microbiol. 5, 1070-1081. Schatz, P.J., Bieker, K.L., Ottemann, K.M., Silhavy, T.J., and Beckwith, J. (1991). One of three transmembrane stretches is sufficient for the functioning of the SecE protein, a membrane component of the E. coli secretion machinery. EMBO J. 7, 1749-1757.  122  Scheuring, J., Braun, N., Nothdurft, L., Stumpf, M., Veenendaal, A.K., Kol, S., van der Does, C., Driessen, A.J., and Weinkauf, S. (2005). The oligomeric distribution of SecYEG is altered by SecA and translocation ligands. J. Mol. Biol. 2, 258-271. Schiebel, E., Driessen, A.J., Hartl, F.U., and Wickner, W. (1991). Delta mu H+ and ATP function at different steps of the catalytic cycle of preprotein translocase. Cell 5, 927939. Schiebel, E., and Wickner, W. (1992). Preprotein translocation creates a halide anion permeability in the Escherichia coli plasma membrane. J. Biol. Chem. 11, 7505-7510. Segers, K., Klaassen, H., Economou, A., Chaltin, P., and Anne, J. (2011). Development of a high-throughput screening assay for the discovery of small-molecule SecA inhibitors. Anal. Biochem. Shan, S.O., Chandrasekar, S., and Walter, P. (2007). Conformational changes in the GTPase modules of the signal reception particle and its receptor drive initiation of protein translocation. J. Cell Biol. 4, 611-620. Shimoike, T., Taura, T., Kihara, A., Yoshihisa, T., Akiyama, Y., Cannon, K., and Ito, K. (1995). Product of a new gene, syd, functionally interacts with SecY when overproduced in Escherichia coli. J. Biol. Chem. 10, 5519-5526. Simon, S.M., and Blobel, G. (1992). Signal peptides open protein-conducting channels in E. coli. Cell 4, 677-684. Smith, M.A., Clemons, W.M.,Jr, DeMars, C.J., and Flower, A.M. (2005). Modeling the effects of prl mutations on the Escherichia coli SecY complex. J. Bacteriol. 18, 64546465. Tam, P.C., Maillard, A.P., Chan, K.K., and Duong, F. (2005). Investigating the SecY plug movement at the SecYEG translocation channel. EMBO J. 19, 3380-3388. Taura, T., Akiyama, Y., and Ito, K. (1994). Genetic analysis of SecY: additional exportdefective mutations and factors affecting their phenotypes. Mol. Gen. Genet. 3, 261269. Terwilliger, T.C., and Berendzen, J. (1999). Automated MAD and MIR structure solution. Acta Crystallogr. D Biol. Crystallogr. Pt 4, 849-861. Tian, P., and Andricioaei, I. (2006). Size, motion, and function of the SecY translocon revealed by molecular dynamics simulations with virtual probes. Biophys. J. 8, 27182730.  123  Tsukazaki, T., Mori, H., Fukai, S., Ishitani, R., Mori, T., Dohmae, N., Perederina, A., Sugita, Y., Vassylyev, D.G., Ito, K., and Nureki, O. (2008). Conformational transition of Sec machinery inferred from bacterial SecYE structures. Nature 7215, 988-991. Tziatzios, C., Schubert, D., Lotz, M., Gundogan, D., Betz, H., Schagger, H., Haase, W., Duong, F., and Collinson, I. (2004). The bacterial protein-translocation complex: SecYEG dimers associate with one or two SecA molecules. J. Mol. Biol. 3, 513-524. Van den Berg, B., Clemons, W.M.,Jr, Collinson, I., Modis, Y., Hartmann, E., Harrison, S.C., and Rapoport, T.A. (2004). X-ray structure of a protein-conducting channel. Nature 6969, 36-44. van der Sluis, E.O., Nouwen, N., and Driessen, A.J. (2002). SecY-SecY and SecY-SecG contacts revealed by site-specific crosslinking. FEBS Lett. 1-3, 159-165. van der Sluis, E.O., Nouwen, N., Koch, J., de Keyzer, J., van der Does, C., Tampe, R., and Driessen, A.J. (2006). Identification of two interaction sites in SecY that are important for the functional interaction with SecA. J. Mol. Biol. 5, 839-849. van Stelten, J., Silva, F., Belin, D., and Silhavy, T.J. (2009). Effects of antibiotics and a proto-oncogene homolog on destruction of protein translocator SecY. Science 5941, 753-756. Veenendaal, A.K., van der Does, C., and Driessen, A.J. (2001). Mapping the sites of interaction between SecY and SecE by cysteine scanning mutagenesis. J. Biol. Chem. 35, 32559-32566. von Heijne, G., and Gavel, Y. (1988). Topogenic signals in integral membrane proteins. Eur. J. Biochem. 4, 671-678. Williams, K.P., Gillespie, J.J., Sobral, B.W., Nordberg, E.K., Snyder, E.E., Shallom, J.M., and Dickerman, A.W. (2010). Phylogeny of gammaproteobacteria. J. Bacteriol. 9, 2305-2314. Wonderlin, W.F. (2009). Constitutive, translation-independent opening of the proteinconducting channel in the endoplasmic reticulum. Pflugers Arch. 4, 917-930. Woodbury, R.L., Hardy, S.J., and Randall, L.L. (2002). Complex behavior in solution of homodimeric SecA. Protein Sci. 4, 875-882. Wowor, A.J., Yu, D., Kendall, D.A., and Cole, J.L. (2011). Energetics of SecA dimerization. J. Mol. Biol. 1, 87-98. Zimmer, J., Nam, Y., and Rapoport, T.A. (2008). Structure of a complex of the ATPase SecA and the protein-translocation channel. Nature 7215, 936-943.  124  Zimmer, J., and Rapoport, T.A. (2009). Conformational flexibility and peptide interaction of the translocation ATPase SecA. J. Mol. Biol. 4, 606-612. Zimmermann, R., Eyrisch, S., Ahmad, M., and Helms, V. (2011). Protein translocation across the ER membrane. Biochim. Biophys. Acta 3, 912-924.  125  Appendices Appendix A Purification of the SecY complex and reconstitution into Nanodiscs Traditional methods for studying the SecY channel involve lipid vesicles, but the oligomeric state of the SecY complex cannot be controlled. Alternatively, it involves detergent micelles, although this may artificially stabilize or destabilize the association of the complex with its cytosolic partners. A novel method termed Nanodiscs has been developed by the laboratory of S. Sligar at the University of Illinois. The technology uses a modified form of the apolipoprotein A-I, with the optimized capacity to wrap around small patch of lipid bilayers and detergent purified membrane proteins [Denisov et al., 2004]. Hence, it becomes possible to study the SecY complex without the need for liposomes or detergents. As many observations in Chapters 2 and 3 were made using Nanodisc technology a detailed protocol for the purification of the SecY complex and reconstitution into Nanodiscs is given here.  A.1  SecY purification materials  1. Strain Escherichia coli BL21 transformed with the plasmid pBAD22 encoding for His6tagged SecE, SecY and SecG under the control of the araBAD promoter. 2. Luria Bertani (LB) broth (Sigma, St. Louis, MO). 3. 150 mM Phenylmethylsulfonylfluoride (PMSF) (Sigma). 4. 20% (w/v) L-arabinose (Alfa Aesar, Ward Hill, MA). 5. Lysis buffer: 50 mM Tris pH 7.9, 300 mM NaCl, 10% glycerol. 6. French Press cell (SLM Aminco) and hydraulic pump (see Section A.16 Note 1). 7. 10% (v/v) Triton® X-100 (Bioshop Canada, Burlington, Ontario) in MilliQ H2O.  126  8. 0.3 M NiSO4 in MilliQ H2O. 9. IMAC (Immobilized metal affinity chromatography) running buffer: 50 mM Tris pH 7.9, 300 mM NaCl, 10% glycerol, 0.03% (w/v) n-dodecyl-β-D-maltoside (DDM; Anatrace, Maumee, OH). 10. IMAC running buffer supplemented with 30 mM imidazole. 11. IMAC running buffer with 50 mM NaCl (instead of 300 mM). 12. IMAC elution buffer: 50 mM Tris pH 7.9, 50 mM NaCl, 10% glycerol, 0.03% (w/v) DDM, 500 mM imidazole. 13. Ion exchange running buffer: same as item 11. 14. Ion exchange gradient buffer: 50 mM Tris pH 7.9, 600 mM NaCl, 10% glycerol, 0.03% (w/v) DDM. 15. FPLC system (ÄKTA purifierTM or equivalent apparatus [GE Healthcare Biosciences AB, Uppsala, Sweden]). 16. IMAC FPLC column (10 mL of packed Chelating SepharoseTM beads in a 10/10 Tricorn column [GE Healthcare Biosciences]). 17. HiTrapTM SP HP 5 mL cation exchange column (GE Healthcare Biosciences). 18. Amicon® Ultra centrifugal filter, MWCO 50 kDa (Millipore, Billerica, MA). 19. Sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) buffers and apparatus for running a 12% SDS-PAGE gel.  A.2  Membrane scaffold protein (MSP)  1. ~100 mg of lyophilized MSP powder (MSP1D1, 24.6 kDa; Denisov et al., 2004). 2. Reconstitution buffer: 20 mM Tris-HCl pH 7.9, 100 mM NaCl, 0.5 mM EDTA.  127  A.3  Lipids  1. Purified E. coli total lipid extract (Avanti polar-lipids, Alabaster, AL). 2. 100% chloroform solution (Fisher Scientific Canada, Ottawa, ON). 3. Standard nitrogen (N2) gas supply. 4. Vacuum dessicator (Corning Life Sciences, Lowell, MA).  A.4  SecY reconstitution materials  1. Concentrated SecY complex at ~3g/L in 0.03% (w/v) DDM detergent micelles (see Section A.9, item 19). 2. Membrane scaffold protein (see Section A.10). 3. 10% (w/v) and 1% (w/v) DDM solutions in MilliQ H2O. 4. 1000 nmole of dried E. coli lipids in a 1.5 mL screwcap microtube. 5. Lipid reconstitution buffer: 50 mM Tris pH 7.9, 50 mM NaCl. 6. Branson 2510 Ultrasonic water bath or equivalent (Branson, Danbury, CT). 7. Nanodisc reconstitution buffer: same as Ion exchange gradient buffer (see Section A.1, item 14). 8. Bio-beads SM-2 Adsorbent (BioRad, Hercules, CA). 10 mL of dry beads are washed successively with 50 mL of ethanol, methanol, and MilliQ H2O. Beads are stored in 25 mL 50 mM Tris pH 7.9, 50 mM NaCl. 9. Rotating shaker apparatus.  128  A.5  Size exclusion chromatography materials  1. FPLC system (ÄKTA purifierTM [GE Healthcare Biosciences]). 2. Superdex-200 prep grade resin packed in a Tricorn 10/10 column (GE Healthcare Biosciences). 3. S200 buffer: 50 mM Tris pH 7.9, 100 mM NaCl, 5% glycerol. 4. Amicon® Ultra centrifugal filter, MWCO 50 kDa (Millipore). 5. Native-PAGE buffers and apparatus for running a 4–12% gradient native gel.  A.6  SDS-PAGE materials  1. 5X SDS-PAGE sample buffer: 250 mM Tris-HCl pH 7.0, 10% (w/v) SDS, 10% (v/v) βmercaptoethanol (BME), 50% glycerol, 0.05% bromophenol blue. Stored at −20°C. 2. 4X SDS-PAGE separating gel buffer: 1.5 M Tris-HCl pH 8.8. 3. 4X SDS-PAGE stacking gel buffer: 0.5 M Tris-HCl pH 6.8. 4. 10% Ammonium persulfate (APS) in MilliQ H2O. 5. 40% mixed Acrylamide/bis solution (37:5:1) (BioRad). 6. N,N,N,N -tetramethyl-ethylenediamine (TEMED) (Bio- Rad). 7. 100% Isopropanol. 8. 10X SDS-PAGE electrophoresis running buffer: 250 mM Tris base, 1.9 M glycine, 0.15 (w/v) SDS. Stored at room temperature. To be diluted to 1X with MilliQ H2O before use. 9. PAGE staining buffer: 40% methanol, 10% acetic acid, 0.025% (w/v) Coomassie Blue R250 (BioRad). 10. PAGE destaining buffer: 20% ethanol, 10% acetic acid.  129  11. BioRad Mini PROTEAN® 3 apparatus, or equivalent (Bio-Rad).  A.7  Native-PAGE materials  1. 10X Native-PAGE loading buffer: 50% glycerol, 0.005% bromophenol blue. Stored at −20°C. 2. 4X Native-PAGE gel buffer: 1.5 M Tris-HCl pH 8.8. 3. 10% APS and TEMED. 4. 40% acrylamide solution and 2% bis-acrylamide solution (BioRad). 5. 1X native PAGE electrophoresis running buffer: 25 mM Tris, 190 mM Glycine, pH 8.8. Stored at 4°C. 6. PAGE staining and destaining buffer (see Section A.6, items 9 and 10). 7. BioRad Mini PROTEAN® 3 apparatus and Multi Casting Chamber, or equivalent (BioRad). 8. 2-well gel mixing chamber. 9. Peristaltic pump (Pump P-1, Pharmacia Biotech).  130  A.8  Method for SecY purification and reconstitution into Nanodiscs The essential component of the Nanodisc reconstitution is an amphipathic helix  termed membrane scaffold protein (MSP). The MSP surrounds the acyl chains of the lipids in which the SecY complex is embedded (termed Nanodisc-SecY complex) (Figure A-1A). The reconstitution requires highly purified, detergent soluble SecY protein complex. The purification of the SecY complex is accomplished by IMAC chromatography (eluted SecY fractions shown in Figure A-1B). A cation ion exchange step is required to further purify the SecY complex to sufficient quality for Nanodisc reconstitution (eluted SecY fractions shown in Figure A-1C). The Nanodisc reconstitution involves mixing lipids, membrane scaffold protein and the soluble SecY complex. Detergent is removed by adsorption onto BioBeads during overnight incubation. The progressive removal of the detergent initiates the self assembly process, resulting into the association of the SecY complex with the MSP. The separation of the Nanodisc-SecY complex from the adducts is obtained by size exclusion chromatography (Figure A-1D). The chromatography has enough resolution to separate aggregates and “empty” discs (not containing the SecY complex). The fractions are analyzed by SDS-PAGE (Figure A-1E) to identify those containing the Nanodisc-SecY complex (i.e., SecY, SecE/G, and MSP). Alternatively, the Nanodisc-SecY complex is detected by nondenaturing native-PAGE (Figure A-1F). The protocol for this procedure is outlined in Sections A.9 – A.16.  131  Figure A-1: SecY purification and reconstitution into Nanodiscs (A) The acyl chains of the phospholipids are surrounded by the amphipathic helical membrane scaffold protein. (B) IMAC chromatography of the his-tagged SecY complex and analysis of the eluted fractions by 12% SDSPAGE and Coomassie staining. (C) Additional purification of the SecY complex using cation exchange chromatography. The SecY complex was eluted from the HiTrapTM HP SP cation exchange column with a linear 50–600 mM NaCl gradient (see Section A.9). (D) Superdex-200 size exclusion chromatogram of the Nanodisc-SecY reconstitution (see Section A.13). Aggregates appear in the column void volume (2.8 mL), and the Nanodisc-SecY complex is eluted at 4.1 mL. (E) SDS-PAGE and (F) Native-PAGE analysis of the Nanodisc-SecY fractions recovered from the size exclusion chromatography. The SecY-Nanodisc complex that was pipetted into each lane of the native gel was first mixed with 1 µg of purified His6-tagged Syd protein (see Section A.16 Note 6). This figure was modified from Dalal and Duong, 2010.  132  A.9  Protocol for the purification of the SecY complex  1. 3L of LB, supplemented with 0.1 g/L ampicillin, are inoculated with 30 mL of an overnight culture of E. coli BL21(DE3) transformed with plasmid pBAD22-hisEYG. 2. The cells are shaken at 37°C until an OD600 nm of 0.6. 3. The cells are induced with 0.2% (w/v) L-arabinose for 3 h. 4. The cells are collected at 5000 g for 10 min at 4°C. The pellet is resuspended with ~50 mL of lysis buffer and 200 µL of 150 mM PMSF. All subsequent steps are carried out at 4°C. 5. The cells are broken by passage through a French press (3X) at 8000 psi. The unbroken cells are removed by centrifugation at 5000 g (10 min). The supernatant is spun again at 125,000 g (60 min) to isolate the crude membranes. 6. The crude membrane pellet is resuspended to 10 g/L in lysis buffer, and then solubilized with 1% Triton ® X-100. The solution is gently rotated overnight at 4°C. 7. The membranes are spun at 125,000 g for 60 min to remove unsolubilized material. The supernatant is collected and kept on ice. 8. The pumps of the FPLC ÄKTA Purifier are washed in MilliQ H2O, and a Tricorn 10/10 IMAC column (see Section A.1, item 16) is mounted. All steps are performed at a flow rate of 2 mL/min. 9. The resin is washed with 2 column volumes of MilliQ H2O, then chelated with 20 mL of 0.3 M NiSO4, and further washed to remove unbound metal. Protein concentrations are monitored at A280 nm (absorbance at 280 nm wavelength). 10. The column is equilibrated with 2 column volumes of IMAC running buffer, followed by injection of the supernatant from step 7.  133  11. The column is washed with 3 column volumes of IMAC running buffer or until the A280 nm value becomes stable. 12. To remove non-specifically bound protein, the column is further washed with 3 volumes of IMAC running buffer supplemented with 30 mM imidazole. An important elution of nonspecifically bound proteins is observed at this point. 13. The column is finally equilibrated with 2 volumes of IMAC running buffer, with the NaCl concentration reduced to 50 mM. This is a preparative step for ion exchange chromatography (steps 16–19) since a low salt concentration is needed to ensure binding of the SecY complex onto the HiTrapTM HP SP column. 14. The elution of the SecY complex is achieved with the IMAC elution buffer, and approximately 30 fractions of 1 mL are collected. About 20 µL of each fraction are mixed with 5 µL of 5X SDS-PAGE sample buffer and analyzed by 12% SDS-PAGE followed by Coomassie blue staining (see Section A.14). 15. The fractions enriched for the SecY complex (shown in Figure A-1B) are pooled and stored at −80°C until ready for cation exchange chromatography. 16. A 5 mL HiTrapTM SP HP cation exchange column is mounted on the chromatography system and washed with 5 column volumes of Ion exchange buffer (containing 50 mM NaCl). All subsequent steps are carried out at 1 mL/min. 17. The pooled SecY complex IMAC fractions from step 15 are loaded onto the column, and eluted with a linear gradient of NaCl (50-600 mM) prepared in Ion exchange buffer. 18. For each 750 µL fraction collected, 20 µL are mixed with 5 µL of 5X SDS-PAGE sample buffer and analyzed by a 12% SDS-PAGE (see Section A.14).  134  19. Fractions containing purified SecY complex (shown in Figure A-1C) are pooled and concentrated to approximately 3 g/L using an Amicon® Ultra centrifugal filter. The concentrated SecY complex is stored at −80°C until needed for the Nanodisc reconstitution (see Section A.12). The purification yield is about 1.5 mg/L of culture.  A.10  Preparation of membrane scaffold protein  1. The MSP are solubilized in reconstitution buffer to a stock concentration of 15 g/L and stored at −80°C. 2. A working solution of MSP is diluted to 5 g/L (with reconstitution buffer) from the stock solution and stored at −80°C until needed for the Nanodisc reconstitution (see Section A.12).  A.11  Preparation of lipids  1. E. coli total phospholipids contained in the manufacturer vial are dissolved in chloroform at a concentration of 20 nmole/µL (see Section A.16 Note 2). 2. Aliquots of 1000 nmole of the chloroform soluble lipids (50 µL of stock) are placed in 1.5 mL screwcap microtubes. 3. The lipids are dried under a gentle stream of nitrogen. A pipetman tip is fixed to the outlet tubing to ease the application of the nitrogen stream. 4. The lipids are further dried in a vacuum dessicator overnight. 5. Lipids are stored at −20°C until needed (see Section A.12).  135  A.12  Protocol for Nanodisc reconstitution of the SecY complex  1. A typical reconstitution consists of mixing together of SecY complex at a SecY:MSP:lipid molar ratio of 1:4:100 (see Section A.16 Note 3). The reaction is carried out in a final volume of 1 mL at 4°C. 2. 75 µL of 1% DDM are added to ensure initial solubilisation of all components. 3. 1000 nmole of dried lipids are resuspended to 5 nmole/µL with 200 µL of lipid reconstitution buffer using alternating vortexing and sonication. The lipids are solubilized with 0.5 % DDM by pipetting 10 µL of 10% DDM into the suspension. Exactly 120 µL (600 nmole) of lipids are added to the reconstitution mixture. 4. Followed by the addition of 24 nmole of MSP (564 µg, or 113 µL of a 5 g/L solution). 5. Followed by the addition of 6 nmole of detergent soluble SecY complex (450 µg protein, or 150 µL of a 3 g/L solution). 6. The Nanodisc reconstitution buffer is finally added to bring the total volume to 1 mL. 7. This is followed by the addition of 300 µL of BioBeads. 8. The mixture is gently rotated overnight at 4°C. 9. The following day, the microtube is placed on ice for 10 min, so that the BioBeads settle at the bottom of the tube. The supernatant is next transferred to a new 1.5 mL ultracentrifuge tube. 10. The sample is then spun at 100,000 g for 20 min at 4°C to remove aggregates. 11. The supernatant is collected into a new 1.5 mL microtube and stored at 4°C until the size exclusion chromatography step (see Section A.16 Note 4).  136  A.13  Size exclusion chromatography of Nanodisc reconstitutions  1. The ÄKTA purifierTM pumps are washed with MilliQ H2O, then with S200 buffer (see Section A.5, item 3). 2. A Tricorn 10/10 column packed with S200 prep grade resin is mounted and equilibrated with 2 column volumes of S200 buffer. The flow rate for all steps is 0.5 mL/min. 3. 300 µL maximal volume of the reconstituted Nanodisc-SecY complex (from Section A.12, item 11) are injected onto the column (see Section A.16 Note 5). The protein concentration is monitored at A280 nm. 4. Fractions of 250 µL are collected as the Nanodisc-SecY complex is eluted. About 20 µL of each fraction are mixed with 5 µL of 5X SDS-PAGE loading buffer and analyzed by 12% SDS-PAGE (Section A.14 and Figure A-1E). In parallel, 20 µL of each fraction is mixed with 1 µg of Syd protein (see Section A.16 Note 6) and allowed to incubate at room temperature for 5 min. Subsequently, a few µL of 10X native-PAGE loading buffer (~2 µL) is added to each fraction and the samples are analyzed by 4–12% gradient native-PAGE (Section A.15 and Figure A-1F). 5. The profile of the size exclusion chromatogram indicates that aggregates are eluting in the void volume of the column (2.8 mL), whereas the Nanodisc-SecY complex is eluting in the latter fractions (Figure A-1D). 6. The remaining crude Nanodisc-SecY complex (from step 3) can be purified by repeating steps 2–4. From beginning to end, the efficiency of the reconstitution is approaching 25%. 7. The fractions containing the Nanodisc-SecY complex are pooled and concentrated to approximately 1.5 g/L using an Amicon® Ultra centrifugal device. The prep is stored at 4°C and is stable for a few days (see Section A.16 Note 7).  137  A.14  SDS-PAGE protocol  1. The reader is referred to the instruction booklet provided online (http://www.biorad.com/LifeScience/pdf/Bulletin_4006193A.pdf) concerning the mounting and running of the SDS-PAGE Mini PROTEAN® 3 system (BioRad).  A.15  Native-PAGE protocol  1. The following steps describe how to run a native-PAGE using the BioRad Mini PROTEAN® 3 system (BioRad). The volumes indicated are for the preparation of six minigels using the PROTEAN® 3 Multi Casting Chamber. 2. A 12% native-PAGE gel is prepared by mixing 14.6 mL of 40% acrylamide, 8 mL of 2% bis-acrylamide, 12.5 mL of 4X native-PAGE gel buffer, 10 mL of 100% glycerol and 4.9 mL of MilliQ H2O (50 mL final volume). 3. A 4% native-PAGE gel is prepared by mixing 4.9 mL 40% acrylamide, 2.7 mL 2% bisacrylamide, 12.5 mL of 4X native-PAGE gel buffer and 29.9 mL of MilliQ H2O (50 mL final volume). 4. 27.5 mL of the 12% and 4% native gel mixes (from steps 2 and 3, respectively) are added to the gel mixing chamber. Tubing is attached to the gel mixing chamber and to the peristaltic pump, which is attached in serial to the multi-caster chamber. 5. Polymerization of the 12% solution is initiated with 58 µL of 10% APS and 5.8 µL of TEMED. Similarly, 145 µL of 10% APS and 14.5 µL of TEMED are added to the 4% solution.  138  6. The separating valve on the gel mixing device is set to the position “mix” and the pump is started at a medium flow rate. Both chambers are stirred with magnetic bars for adequate mixing and to prevent the premature polymerization of acrylamide. Combs for 12 or 15 lanes are placed on the top of the gels after complete filling of the gel caster. The polymerization should occur within 40 min. 7. Each gel is mounted on the BioRad apparatus and sufficient amount of native-PAGE running buffer is added to the inner and outer chambers of the apparatus. The electrophoresis is run at 20 mA at constant voltage (500 V max) until the dye from the sample buffer reaches the bottom of the gel. A Blue native-PAGE can be run by using native-PAGE running buffer supplemented with 0.015% Serva Blue G dye (Helixx Technologies) in the cathode (inner) chamber. 8. The gel is removed and incubated in 15 mL of PAGE staining buffer for 30 min. The gel is rinsed with water and destained with 15 mL of PAGE destaining buffer until the protein bands become clearly visible.  A.16  Notes  1. The large cell of the French press holds a maximum volume of approximately 35 mL. It should be cooled down to 4°C before use, and the piston lubricated with 100% glycerol. 2. Screwcap microtubes possess o-ring lids that will prevent the evaporation of chloroform. 3. Different SecYEG:MSP:lipid ratios are possible. Reconstitution without lipids is also feasible. The purification and the reconstitution will work for certain SecY mutants, but only if the mutation does not destabilize the heterotrimeric SecYEG associations. Different lipid mixtures can be used such as cardiolipin, phophatidylcholine, and phosphatidylgycerol.  139  4. Nanodiscs will partially precipitate after a few days at 4°C. The size exclusion chromatography following the reconstitution should be performed as soon as possible. 5. Warming up the Nanodiscs to 42°C for 5 min before loading onto the gel filtration column reduces aggregation. 6. The Nanodisc-SecY complex migrates as a smeary band on native-PAGE. To obtain sharper bands (as shown in Figure A-1F), the protein Syd is added to the sample. Syd is a small 23 kDa SecY-binding protein (see Chapter 2). Syd forms a complex with SecY and alters the isoelectric point of the Nanodisc-SecY particles, resulting in a better migration on the native gel [Alami et al., 2007]. 7. Concentrating the Nanodisc-SecY complex beyond 1.5 mg/mL may lead to increased aggregation.  140  Appendix B Cartoon representation of SecY monomer or dimer incorporated into Nanodiscs  (Left) The MSP1 Nanodisc has a diameter of ~9.7 nm and is sufficient to capture a single SecY complex. (Right) The MSP3 Nanodisc is slightly larger (12.1 nm) and can capture a second SecY complex. The shown components are coloured as follows: green, MSP belt protein; yellow, lipids; red; SecY complex copy 1; blue SecY complex copy 2.  141  Appendix C Cartoon Representation of Nanodisc-SecY-SecA complexes  Nanodisc-SecY complexes are the same as shown in Figure 2-2 and Figure 2-3C.  142  Appendix D SecA Translocation ATPase kinetics curves with Nanodiscs  Initial rates of SecA translocation ATPase were determined in the presence of the indicated concentrations of purified Nanodiscs, using the malachite green colorimetric assay. Reactions were performed with 0.2 µM SecA, 0.8 µM PhoA1-202 (open squares) or L14R signal sequence mutant (triangles) over a range 30 minutes at 37°C. The results were fitted to a one site quadratic binding equation to determine kcat values (in Figure 3-1, E and F, and Figure 3-3C) as described in Section 3.2.3. Initial rates were determined at least 3 times, which reflect the error bars shown at each point.  143  Appendix E Reconstitutions with MSP2N2 and sucrose density purification  (A) The covalently linked SecY dimer reconstituted with the MSP2N2 scaffold protein and PG lipids was loaded on top of a 6-13% sucrose gradient. After centrifugation (188,000 x g, 17 hours, 4°C), equal fractions were collected and analyzed by native-PAGE in the presence of Syd. Translocation ATPase measurements in Figure 3-3C were performed on the fractions labelled with an asterisk. The fused SecY dimer supported the highest translocation ATPase when reconstituted with the MSP2N2 scaffold protein. (B) Nanodiscs bearing the indicated SecY mutants were purified as in (A). (C) Assessment of the purity of the Nd-YY complex by dynamic light scattering. The trace shows that the Nanodisc complex (~8 nm radius) is 99.8% pure from any liposomes (50-100 nm) or large aggregates (>100 nm).  144  Appendix F Position of cysteine residues introduced into the SecY complex  (A) Top view (from the cytosol) of the SecY complex with position 97 indicated on TMS2 in the signal sequence binding site. (B) Hypothetical ‘front-to-front’ and ‘back-to-back’ representations of the SecY dimer (side view). The SecY complex is coloured as follows: grey SecY; yellow SecE; orange SecG; cyan TMS2; red TMS7; green cysteine residue. The representations are based on SecYEβ structure from Methanococcus jannaschii (PDB: 1RHZ; van den Berg et al., 2004).  145  Appendix G Western blot analysis of wild-type and mutant IMVs  (A) 4 µg of IMVs containing the wild-type or the mutant SecY complexes were analyzed by SDS-PAGE followed by immuno-detection using antibodies directed against the N-terminal domain of SecY. The immunocomplex was revealed using a goat-anti-rabbit antibody linked to a fluorescent dye excited at 680 nm (Licor biosciences). (B) Western blot analysis of membranes bearing the SecY67CC complex (i.e. cysteine residues at position 67 of SecY and position 120 of SecE) using an anti-HA antibody to detect HA-tagged SecE. The IMVs were incubated with BMOE (0.1 mM) or DTT (2 mM) to stabilize or abolish crosslink formation between SecE and the SecY plug. This figure was modified from Dalal and Duong., 2009.  146  Appendix H Oxonol VI fluorescence responds linearly to the membrane potential  40 µg of IMVs (prepared in 20 mM K2SO4) bearing the SecYWT complex were incubated in 150 µL of Tris-SO4 buffer with 20-1000 mM potassium acetate and 2 µM oxonol VI. A K+-specific membrane potential was generated with 1 µM Valinomycin similar to as shown in Figure 4-2B. The percent quenching of oxonol VI fluorescence was calculated according to the expression FQ/FI, where FQ = fluorescence magnitude 360 seconds after addition of valinomycin, and FI=initial fluorescence signal. The fluorescence quenching was plotted against the membrane potential, ∆ψ, derived from the Nernst equation: ∆ψ=(RT/F)*ln([Kout]/[Kin]) where T = 298.15°K, R and F are the classical thermodynamic constants and [Kout]/[Kin] are the concentrations of potassium outside and inside the IMVs. For example, at [Kout]/[Kin] = [1000]/[20], the membrane potential ∆ψ = 0.100504V = 101mV, which will quench ~75% of oxonol VI fluorescence at these IMV and dye concentrations. This figure was modified from Dalal and Duong., 2009.  147  Appendix I Schematic of SecY conductance assay  The addition of NADH to inverted inner membrane vesicles causes NADH oxidase (blue circle) to pump proteins into the vesicle lumen. (Left) Ions are not permeable through the wild-type SecY complex. (Right) When the plug domain is crosslinked in the open state, or when the pore ring residues are mutated, anions enter the vesicle lumen and neutralize the positive charge generated by proton influx. This dissipates the membrane potential ∆ψ. (Bottom) Both IMVs described above can generate a stable ∆ψ upon addition of NADH monitored by fluorescence quenching of oxonol VI. Upon addition of KCl however, IMVs containing the mutant SecY complex allows permeation of chloride ions (red trace), whereas the wild-type IMVs do not (black trace).  148  Appendix J The membrane potential across the SecYPrlA4 membrane collapses in the presence of halide anions  (A) IMVs prepared from E. coli KM9 - bearing the empty pBad22 vector, or enriched for the wild type or prlA4 SecY complexes, were diluted in buffer A (50 mM Tris-SO4 pH 7.9, 25 mM K2SO4, 10 mM MgSO4, 0.2 mg/mL BSA). The membrane potential ∆ψ, inside positive, was generated with 5 mM NADH and monitored by oxonol VI fluorescence. The ∆ψ was then challenged with various potassium salts (each 5mM). (B) The membrane potential ∆ψ, was challenged with increasing amounts of KCl. This figure was modified from Dalal and Duong., 2009.  149  Appendix K Collapse of the membrane potential at low halide ion concentration  IMVs were diluted in buffer A (50 mM Tris-SO4 pH 7.9, 25 mM K2SO4, 10 mM MgSO4, 0.2 mg/mL BSA). The membrane potential ∆ψ, inside positive, was generated with 5 mM NADH and monitored by oxonol VI fluorescence. The ∆ψ was then challenged with low amount of potassium halide ions (0.5 mM each) and abolished with 1 µM CCCP. This figure was modified from Dalal and Duong., 2009.  150  Appendix L Structural representation of the SecY pore ring residues  Top view structure of the channel pore ring with the plug domain removed (modeled in pymol v0.99 using PDB: 1RHZ; van den Berg et al., 2004). The isoleucine residues are represented in red and numbered according to E. coli SecY. Grey: SecY, Yellow: SecE, Green: SecG, Blue: lateral gate. This figure was modified from Dalal et al., 2010.  151  Appendix M Temperature dependence of chloride leakage during translocation  40 µg of IMVs bearing the SecYWT complex were incubated at the indicated temperature in 150 µl of buffer B in the presence of BSA (30 µg), SecA (3 µg) and proOmpA (3 µg). A ∆ψ was generated with 5 mM NADH and protein translocation was initiated with 1 mM ATP (arrow). The ∆ψ was monitored with oxonol VI fluorescence and quenched with CCCP. For comparison, the initial fluorescence of each trace was taken as 100%. This figure was modified from Dalal and Duong., 2009.  152  Appendix N Acclamation •  Science 2007, Editor’s Choice, Vol 316, page 174.  153  •  Nat. Struct. Mol. Biol. 2009, Research Highlights, Vol 16, page 106.  154  

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.24.1-0072537/manifest

Comment

Related Items