EVIDENCE FOR A DIRECT LINK BETWEEN THE TOL-PAL PROTEIN COMPLEX AND GRAM NEGATIVE BACTERIA CELL DIVISION VIA AN INTERACTION BETWEEN TOLQ AND THE DIVISOME PROTEIN FTSN
Mary A. Teleha
A Dissertation
Submitted to the Graduate College of Bowling Green State University in partial fulfillment of the requirements for the degree of
DOCTOR OF PHILOSOPHY
August 2013
Committee:
Dr. Ray A. Larsen, Advisor
Dr. Roudabeh J. Jamasbi Graduate Faculty Representative
Dr. Rex L. Lowe
Dr. Adam C. Miller
Dr. Vipaporn Phuntumart
Dr. Scott O. Rogers
ii
ABSTRACT
Ray Larsen, Advisor
The TolQ protein functions to couple cytoplasmic membrane-derived energy to support outer membrane processes in Gram negative bacteria. With other products of the widely-conserved tol-pal gene cluster, TolQ has been linked to the process of bacterial cell division. When present in excess, TolQ disrupts cell division, leading to filamentous growth of Escherichia coli. The potential role of TolQ in Gram negative cell division was investigated by a number of methods, including growth assays and immunoblot, two-hybrid, and mutational analyses. This filamentation phenotype is specific for TolQ over-expression independent of TolA and TolR levels, with the degree of filamentation directly proportional to TolQ levels. Over-expression of
E. coli TolQ in closely related species indicates that this property of TolQ is not E. coli specific, as excess TolQ leads to a comparable phenotype in other Gram negatives. Bacterial two-hybrid analysis indicates a potential in vivo interaction between TolQ and the divisome protein FtsN, ostensibly one that competitively diverts FtsN from functioning efficiently during late-stage cell division. Filamentation resulting from TolQ over-expression can be suppressed in cells when
FtsN is concurrently expressed in excess. Mutational analysis of the 19 amino acid TolQ N- terminus suggests that specific residues within and/or conformation of the extreme N-terminal region of TolQ are essential for filamentation. Results of this study indicate that the link between the Tol-Pal proteins and cell division is mediated in part through a direct interaction between the periplasmic regions of TolQ and FtsN, with a possible role for the Tol system in stabilization of the divisome during Gram negative cell division. These findings are presented in iii
this dissertation along with a broader consideration of a role for the Tol-Pal complex in cell division and a dynamic nature for the bacterial cell division apparatus itself.
iv
ACKNOWLEDGEMENTS
I would first like to acknowledge Dr. John Crooks of Lorain County Community College and Dr. Stan Smith of Bowling Green State University for laying the foundation for my graduate programs. While Dr. Crooks has moved on to a new position at LCCC and Dr. Smith passed away while I was at BG, I would not have had this opportunity without the efforts of both. A special thank you to all at BG and LCCC who helped facilitated my program to its end. In particular, I would like to thank Dr. Jeff Miner.
To my research advisors, Dr. Ray Larsen and Dr. Adam Miller, I am also appreciative.
Thank you, Dr. Larsen, for introducing me to this work and teaching me the research and lab skills to make it my best work, that which I can be proud of. I very much appreciate your willingness to expand the scope of your lab to include the work in this dissertation and your working so hard with me to see this dissertation to completion. To Dr. Miller, thank you for giving so much of yourself to this project and my program. You have been an invaluable teacher and collaborator to me and your support has always meant so much. Without you both, Ray and
Adam, I would not be writing this today.
I have been so fortunate to have had such an exceptionally supportive and admirable committee. To Dr. Rex Lowe, Dr. Scott Rogers, and Dr. Vipaporn Phuntumart, I thank you each for the kindness, encouragement, and guidance you have given me. I admire and respect so much about each of you, from how you conduct yourselves as fellow scientists to your passions about your work, which each of you have shared so willingly with me. I would also like to offer my sincere appreciation for Dr. Roudabeh Jamasbi. If I were to have chosen for myself, I don’t believe I would have found a more compassionate, supportive, and fair representative for this committee. You are truly a professional with a kind heart. As I leave Bowling Green, it is my v
hope that I can take with me the many gifts of character that each of you have shared with me and serve them justice throughout my own professional career.
To my family, Mark, Michael, and Kristen, I wish to offer my appreciation and love.
Thank you for your support and efforts to encourage me throughout this process. Michael and
Kristen, I am so proud of you both for being the best “you” that is possible. To my sister
Connie, thank you for always supporting me, always loving me, and always encouraging me.
Lastly, to my father, John Simon, thank you for recognizing my hard work and for telling me that you do. As a parent, I think we often feel pride in our children, but sharing that feeling with your child can make a difference when the road becomes bumpy and uncertain, as it did for me.
I would like to acknowledge two students who were interested enough in my project to work with me as undergraduates: Brittany Jacob and Donna Ruth. Donna, who has worked with me on this project long enough to develop her own related project, tells me that I inspire her.
Donna, you are the one who inspires me. You inspire me to be kinder, more forgiving, stronger, and more determined. You are so much more capable than you think you are. Don’t ever stop asking questions. I would also like to thank Ron Jantz at LCCC for photographing some of my research results, which eventually made it into this dissertation.
Lastly, I would like to acknowledge a small group of individuals who have in one way or another become my biggest supporters, my closest friends, and my heroes: Elin LeClaire,
Kathryn French, Dave Karohl, Barb Schmittgen, Lysa Styfurak, Katie Hogan, Suzie Moreno, and
Chad Braley. Additionally, I would like to acknowledge those individuals at LCCC who were first my teachers so long ago and are now my colleagues and friends. It means so much to me that you believed in me back then, and continue to believe in me today.
vi
TABLE OF CONTENTS
Page
CHAPTER I. INTRODUCTION ...... 1
Gram Negative Bacteria ...... 1
The Tol-Pal Complex ...... 5
Gram Negative Cell Division...... 16
Research Aims ...... 20
CHAPTER II. CORRELATION BETWEEN ARABINOSE INDUCTION LEVELS
AND TOLQ PROTEIN EXPRESSION AMOUNTS AND THE CONSERVED
NATURE OF THE TOLQ OVER-EXPRESSION DIVISION PHENOTYPE ...... 23
Introduction ...... 23
Methods ...... 27
Results ...... 33
Discussion ...... 44
CHAPTER III. POTENTIAL INTERACTION BETWEEN TOLQ AND FTSN
AND THE IMPACT OF THEIR DUAL OVER-EXPRESSION ON DIVISION
PHENOTYPE ...... 49
Introduction ...... 49
Methods ...... 58
Results ...... 78
Discussion ...... 85
vii
CHAPTER IV. THE CONTRIBUTION OF THE N-TERMINAL REGION OF
THE TOLQ PROTEIN TO THE OVER-EXPRESSION DIVISION PHENOTYPE
AS ANALYZED THROUGH SEQUENCE MODIFICATION ...... 89
Introduction ...... 89
Methods ...... 95
Results ...... 107
Discussion ...... 127
CHAPTER V. SUMMARY AND GENERAL CONCLUSIONS...... 138
REFERENCES ...... 152
APPENDIX A Over-expression of the Escherichia coli TolQ protein leads to a null-FtsN-like division phenotype ...... 165
APPENDIX B. Approval for Use of Figure 6 ...... 167 viii
LIST OF FIGURES
Figure Page
1 Diagram of the characteristic components of the Gram negative cell envelope ...... 4
2 Predicted membrane topologies of the CM-associated TonB system and Tol-Pal
system proteins and predicted interactions between TolQ, TolR, and TolA ...... 8
3 Components of the TonB and Tol-Pal systems of Escherichia coli ...... 9
4 Over-expression of several Tol system proteins in W3110 E. coli cells grown
for 24 hours under Amp selection +/- induction with 0.1% (w/v) ʟ-arabinose ...... 14
5 W3110 E. coli cells grown for 24 hours under Amp selection at various
concentrations of ʟ-arabinose used to induce expression of TolQ protein ...... 15
6 Known essential divisome proteins and their predicted membrane topology ...... 17
7 Epitope recognized by anti-TolQ antibody ...... 32
8 tolQ, tolR, and tolA deletion mutants grown at standard osmolarity and under
low salt conditions ...... 34
9. Tol deletion mutants RA1027 (ΔtolQ), RA1028 (ΔtolR), and
RA1038 (ΔtolA) carrying tolQ (pRA031) grown for 24 hours with and without
0.1% w/v ʟ-arabinose ...... 37
10. TolQ over-expression in E. coli W3110, C. mutygensii, E. amnigenus, and E. coli
strain BL21...... 40
11. Western blot analysis of TolQ expression levels ...... 43
12. pBT plasmid vector ...... 54
13. pTRG plasmid vector ...... 55
14. Diagram of transcriptional activation of the HIS3 and aadA (Strr) genes ...... 56 ix
15. BacterioMatch® II positive control plasmids ...... 57
16. BacterioMatch® II two-hybrid Strategy ...... 65
17. Vector map and MCS of pBAD24 ...... 74
18. Vector map and MCS of pBAD18-Cm ...... 75
19. Vector map and MCS of pBAD18-Kan ...... 76
20. Map of pPro18 as a representative of the pPro18-Kan vector, and MCS of pPro
Vectors ...... 77
21. Spot plates identifying putative protein-protein interactions via a bacterial two-
hybrid system ...... 79
22. W3110 E. coli cells grown under dual selective conditions of chloramphenicol
and kanamycin for 24 hours (1) ...... 83
23. W3110 E. coli cells grown under dual selective conditions of chloramphenicol
and kanamycin for 24 hours (2) ...... 84
24. Nucleotide alignment of modified TolQ constructs...... 104
25. Full-length amino acid alignment of modified TolQ constructs ...... 105
26. W3110 E. coli cells over-expressing pRA031 (unmodified TolQ), and pRA019
(TolQ with N-terminal T7 tag addition) ...... 111
27. W3110 E. coli cells over-expressing pMT015 (TolQ8-230) and pMT016
(TolQ4-230) ...... 112
28. W3110 E. coli cells over-expressing pMT017 (TolQThr2Ala), pMT018
(TolQAsp3Ala), pMT019 (TolQMet4Arg), and pMT020 (TolQLys12Glu) ...... 113
29. 0.25% (w/v) DOC growth curves for W3110+pBAD24, RA1033+pBAD24,
RA1033+pRA031, and RA1033+pRA019 ...... 117 x
30. 0.25% (w/v) DOC growth curves for W3110+pBAD24, RA1033+pBAD24,
RA1033+pRA031, RA1033+pMT016 and RA1033+pMT015 ...... 118
31. 0.25% (w/v) DOC growth curves for W3110+pBAD24, RA1033+pBAD24,
RA1033+pRA031, RA1033+pMT017, RA1033+pMT018, RA1033+pMT019,
and RA1033+pMT020 ...... 119
32. Colicin spot titer assays for W3110+pBAD24, RA1033+pBAD24, and
RA1033+pRA031 ...... 123
33. Colicin spot titer assays for RA1033+pRA031 and RA1033+pRA019 ...... 124
34. Colicin spot titer assays for RA1033+pRA031, RA1033+pMT016, and
RA1033+pMT015 ...... 125
35. Colicin spot titer assays for RA1033+pRA031, RA1033+pMT017,
RA1033+pMT018, RA1033+pMT019, and RA1033+pMT020 ...... 126
xi
LIST OF TABLES
Table Page
1 Strains and plasmids used in Chapter II ...... 28
2 Strains and plasmids used for two-hybrid screening and dual over-expression
in Chapter III ...... 64
3 Primers used for PCR amplification for two-hybrid analysis and dual over-
expression studies in Chapter III...... 66
4 Summary of two-hybrid spot identifications and results ...... 80
5 E. coli strains and plasmids used in Chapter IV ...... 97
6 Primers used for the generation of constructs used in Chapter IV ...... 98
7 Summary of cell division phenotypes, 0.25% (w/v) DOC growth assays, and
Colicins A and Ia sensitivity assays for TolQ modifications ...... 128 1
CHAPTER I
INTRODUCTION
Gram Negative Bacteria
Escherichia coli has long been the bacterial model system for laboratory studies of the unique qualities associated with Gram negative bacteria. Gram negative bacteria are characterized by the presence of an inner cytoplasmic membrane (CM) and a distinct outer membrane (OM), separated by a semi-aqueous compartment termed the periplasmic space. This space and its contents are referred to as the periplasm. The periplasmic space can range from 13-
25 nm in width and contains a variety of proteins that function in the transport of sugars and amino acids, in chemotaxis, in the synthesis of cell envelope components, and as enzymes
(Silhavy et al., 2010). The high numbers of periplasmic proteins present in the periplasmic space contribute to the viscosity of the periplasm. Osmoregulated periplasmic glucans (OPG’s) protect the cell from hypoosmotic conditions (Schumann, 2006). A thin layer of peptidoglycan that acts as a rigid cytoskeleton resides within the periplasm, providing the cell with shape and structure, as well as with protection against cytoplasmic osmotic pressure (Oliver, 1996). The peptidoglycan layer is comprised of repeating disaccharides that are cross-liked by pentapeptide side chains (Silhavy et al., 2010). The CM is a phospholipid bilayer that serves as a permeability barrier, allowing cells to establish and maintain ion electrochemical gradients (Kadner, 1996).
The CM is primarily composed of the phospholipids phosphatidylethanolamine and phosphotidylglycerol with smaller amounts of cardiolipin and phosphatidylserine (Silhavy et al.,
2010) and is stabilized by hydrogen bonds, hydrophobic interactions, and associated divalent cations such as Mg2+ and Ca2+. A variety of proteins are either embedded within or associated peripherally with the CM that provide for transport, signaling, energy transduction, and other 2
functions. Energy-dependent transport across the CM is driven by transmembrane ion gradients, in particular those created by the flux of protons across the membrane. In respiring organisms, a proton gradient is established and maintained by an electron transport system consisting of one or more sets of integral membrane and peripheral membrane proteins. This establishes an electrochemical gradient with more protons on the exterior of the CM, thus having both a pH and a charge component. The potential energy of this gradient is used to drive many processes including active transport, protein secretion, and flagellar rotation. This potential energy is also converted to chemical energy by the CM resident ATPase complexes that provide the bulk of the cytoplasmic ATP pool (Kadner, 1996).
Unlike the CM, the OM is an asymmetric bilayer with an inner leaflet of phospholipids and an outer leaflet containing large amounts of the glycolipid lipopolysaccharide (LPS)
(Kamino and Nikaido, 1976). LPS is specific to Gram negative bacteria and is composed of three distinct regions: a membrane-proximal lipid A region, a core oligosaccharide region, and a membrane-distal O-antigen. The lipid A region of LPS is hydrophobic, while the O-antigen is hydrophilic. The lipid A region is widely conserved and possesses four unique fatty acid chains that become variably esterified with additional fatty acids to anchor the LPS in the membrane.
This LPS region confers toxicity to many Gram negatives. The core region of LPS is a heavily modified and varied oligosaccharide that serves to link the O-antigen to the lipid A anchor. The
O-antigen of LPS is a chain of repeating oligosaccharide subunits, the specific structure of which varies greatly between strains. The presence or absence of the O-antigen determines whether bacterial strains are “rough” or “smooth,” with strains displaying rough LPS lacking full-length
O-antigen on their surface more permeable to hydrophobic molecules. The composition of full- 3
length LPS renders the surface of the OM charged, hindering the diffusion of many agents soluble in standard lipid bilayers.
The LPS molecules of the OM contain largely saturated acyl chains which allow the LPS molecules to pack tightly together. Additionally, the presence of OM-associated cations increases the rigidity of the OM, rendering this layer as a nonfluid, selective barrier (Silhavy et al., 2010). The OM also contains a number of outer membrane proteins including lipoproteins, the majority being Braun’s lipoprotein (Lpp) and β-barrel proteins which function as either general porins such as OmpC and OmpF or dedicated porins that recognize and interact with a specific substrate. Lpp is embedded by lipid moieties in the inner leaflet of the OM and covalently links to peptidoglycan, anchoring the OM to the cell wall (Braun, 1975). The presence of porin proteins in the OM provides aqueous channels allowing for the free diffusion of hydrophilic molecules smaller than 600 Da across the OM. Thus the OM constitutes a diffusion barrier, allowing entry of hydrophilic nutrients while excluding detergents, dyes, and larger molecules, including large, essential nutrients. A consequence of OM architecture is that it prevents the establishment of ion electrochemical gradients sufficient to energize transport
(Nikaido, 1996). Together with the presence of phosphatases in the periplasmic space (Oliver,
1996), this renders the OM unable to access the standard energy sources used to drive energy- dependent processes (Postle and Kadner, 2003). The general structure of the Gram negative cell envelope is depicted in Fig. 1.
4
Figure 1. Diagram of the characteristic components of the Gram negative cell envelope. The Gram negative bacteria typically possess a cytoplasmic membrane (CM) composed of a phospholipid bilayer and an asymmetric outer membrane (OM) with an inner leaflet of phospholipids and an outer leaflet predominantly composed of lipopolysaccharide (LPS). LPS is comprised of a proximal lipid A component, a core polysaccharide, and a distal O-specific polysaccharide. The periplasmic compartment is located between the CM and OM. The periplasmic compartment contains a thin layer of peptidoglycan and a number of lipoproteins. Peripheral, integral, and transmembrane proteins are either associated with or reside in the CM. OM proteins and porins are located in the OM layer. The OM, with its negative charge, is stabilized by cations (Ca2+, Mg2+) 5
The Tol-Pal Complex
Maintenance of the OM and most OM-related cellular processes, such as the import of
large nutrients, require CM-derived energy. The known energy-dependent processes of the OM
are coupled to and energized by the proton gradient of the cytoplasmic membrane (Cascales et
al., 2007; Noinaj et al., 2010). Transduction of energy from the CM to the OM is mediated by
protein complexes associated with the OM, CM, and periplasmic compartment. One such
system, the Tol system (the Tol-Pal system), was identified through mutational analysis
(reviewed in Lazzaroni et al., 2002). Mutations in the Tol system were first noted to confer
tolerance to filamentous bacteriophage (Sun and Webster, 1987) and group A colicins (reviewed
in Cascales et al., 2007), both of whose entry into cells requires the passage of large
macromolecules across the diffusion barrier of the OM. Tol system mutants were later found to
exhibit a variety of phenotypes that included heightened sensitivity to a number of antibiotics,
detergents, and hydrophobic dyes, shedding of OM, and leakage of periplasmic contents, all
suggesting disruption of the integrity of the OM. Additionally, Tol mutants experience
disruptions in cell division when grown under non-standard laboratory conditions (Lazzaroni et
al., 1989; Meury and Devilliers, 1999; Gerding et al., 2007).
Characterization of the Tol system has revealed it to be the product of potentially seven genes, five of which encode proteins with established roles (Vianney et al., 1996). The tol-pal
gene cluster contains two promoters, P1 and PB. The genes ybgC, tolQ, tolR, tolA, tolB, pal, and
ybgF are transcribed from P1, while the internal promoter PB regulates the synthesis from a
second transcript composed of tolB, pal, and ybgF (Vianney et al., 1996). TolQ and TolR,
anchored in the CM, form a heteromultimeric complex that is presumed to harvest the energy of
the CM proton gradient to energize TolA, the energy transducer protein of the Tol system. The 6
predicted ratio of TolQ:TolR:TolA in the cell is estimated to be 4-6:2:1 (Goemaere et al., 2007).
The TolQ protein possesses three predicted transmembrane domains, with an 18-19 residue N- terminal domain located in the periplasm (Vianney et al., 1994; Goemaere et al., 2007). TolR possesses a single predicted transmembrane domain present near its N-terminus, with the majority of the protein displayed in the periplasmic space (Muller et al., 1993). A similar topology is suggested for TolA, which is anchored by one transmembrane domain near its N- terminus and extends across the periplasm (Levengood et al., 1991) to interact with the OM associated protein TolB and the lipid-anchored OM protein peptidoglycan-associated lipoprotein
(Pal) (Vianney et al., 1996). Pal interacts with peptidoglycan to mediate interactions between the OM and the cell wall (Leduc, et. al., 1992). The Tol system, mediated through Pal, interacts with two non-tol-encoded proteins, OmpA and the major lipoprotein Lpp. Two additional genes,
ybgC and ybgF, encode proteins with no known contribution to Tol system function (Cascales et al., 2002). The ybgC gene product localizes to the cytoplasm, where it functions as a thioesterase (Zhaung et al., 2002). The function of the periplasmically localized YbgF protein is unclear. However, genetic and physical evidence suggests that it does interact with other proteins of the Tol system (Walburger et al., 2002; Krachler, et al., 2010).
TolQ and TolR share sequence similarities with the ExbB and ExbD proteins of the well-
characterized TonB system (Eick-Helmerich and Braun, 1989). Here, the energy of the CM
proton gradient (proton motive force [pmf]) is coupled via TonB protein to OM proteins that
participate in a number of energy-dependent processes including iron siderophore transport via
FepA and cobalamin transport (reviewed in Postle and Larsen, 2007). Anchored in the CM, the
energy-harvesting pair ExbB/ExbD transfers CM-derived energy to TonB, which in turn
associates with a number of ligand-bound OM receptors. This interaction appears to require Lpp 7
and OmpA (Higgs et al., 2002). While a number of models have been proposed for the mechanistic details of how TonB transduces energy to OM receptors (reviewed in Gresock et al.,
2011), these details have yet to be fully described. However, sufficient data exists to support the model in which CM-derived energy is harvested by ExbB/ExbD, transferred to TonB, and subsequently drives ligand transport across the CM (Reynolds et al., 1980; Bradbeer, 1993;
Postle, 1993; Braun, 1995).
The nucleotide sequence identities of exbB with tolQ and exbD with tolR were found to be 59.2% and 49.7%, respectively. Additionally, amino acid sequence comparisons show that
ExbB and TolQ share 26.3% identical amino acids and 79.1% of the non-identical amino acids representing conservative replacements. Similarly, the amino acid sequences between ExbD and
TolR are 25% identical and 70% conserved (Eick-Helmerich and Braun, 1989). Membrane topologies for TolQ and TolR are very similar to those of ExbB and ExbD, respectively, with the highest amino acid sequence homology present in the transmembrane regions (Kampfenkel and
Braun, 1993). The predicted TolA and TonB sequences are similar only in their N-terminal transmembrane domains (Germon et al., 2001). Biochemical and genetic analyses indicate that the components of the Tol system and the TonB system function largely through interactions involving their transmembrane domains (Lazdunski et al., 1995; Lazzaroni et al., 1995; Larsen et al., 2007). Although TolQ/TolR and ExbB/ExbD support distinct processes at the OM, each pair serving to energize separate processes and interact with their protein partners, TolA and TonB, respectively, the two homolog pairs can inefficiently substitute for one another through a process that is referred to as crosstalk (Braun and Herrmann, 1993). However, crosstalk does not occur between TolA and TonB, reflective of their derived functions in the cell (Braun and Herrmann,
1993). Evidence indicates that TolQ/TolR, ExbB/ExbD, as well as the more distantly related 8
MotA/MotB pair may form aqueous ion channels in the CM and function as molecular motors, converting pmf to mechanical energy. Conserved residues located in the transmembrane regions of each pair have been identified as necessary to support cellular functions that require CM- derived energy (Goemaere et al., 2007; Xiang, et al., 2011). From a historical standpoint, it is likely that the Tol system and the TonB system are paralogues, having originated from a common transport system in E. coli (Eick-Helmerich and Braun, 1989). The current understanding of the membrane topology and partitioning of the Tol-Pal and TonB complex components are summarized in Figs. 2 and 3.
Figure 2. Predicted membrane topologies of the CM-associated TonB system and Tol-Pal system proteins and predicted interactions between TolQ, TolR, and TolA. ExbB/TolQ and ExbD/TolR likely represent sequence and structural homologs. While TolA and TonB share limited sequence identity, their C-terminal domains appear to be structurally similar. Functional homology exists between ExbB/ExbD and TolQ/TolR. The Ton and Tol proteins have been shown to interact through their transmembrane (TM) domains. Evidence indicates that TolQ TM1 interacts with both the TolA TM and its own TM3. TolQ TM2 and TM3 and TolR TM are proposed to form an aqueous channel that transmits ions or protons. The TolA C-terminus was shown to interact with TolB and Pal in the periplasm (Derouiche et al., 1995; Lazzaroni et al., 1995; Germon et al., 1998; Journet et al., 1999; Zhai et al. 2003; Xiang, et al., 2011). 9
Figure 3. Components of the TonB and Tol-Pal systems of Escherichia coli. (Left) The topology and partitioning of the TonB system and the representative proteins with which it interacts are depicted at the figure left. TonB associates with the CM through a heteromultimeric complex of ExbB and ExbD and services the TonB-dependent OM transporters, represented here by the ferric enterobactin transporter FepA. TonB also interacts with the OM proteins Lpp and OmpA. (Right) The topology and partitioning of proteins encoded by the tol-pal operon, and proteins with which Tol proteins are known to interact are depicted at the right of the figure. A heteromultimeric complex of TolQ and TolR, along with TolA, associates with the CM as depicted. OM-associated tol gene products are TolB and Pal. Other proteins known to interact with Tol system components include Lpp, OmpA, and the porin OmpF. The tol-associated roles of the two other tol-pal-encoded proteins YbgC and YbgF are unknown.
10
The true function(s) of the Tol system is unknown. However, it is clear that many cellular processes are influenced by this collection of proteins. Because TolQ and TolR can together substitute for ExbB and ExbD, which have themselves been clearly demonstrated to couple the TonB protein to the electrochemical gradient of the CM (Larsen et al., 1999, Larsen and Postle, 2001), it was presumed that TolQ/TolR complexes similarly harvest CM-derived energy. Indeed, much as ExbB/ExbD mediates energy-dependent conformational changes in
TonB (Larsen et al., 1999), TolQ/TolR has been shown to support energy-dependent conformational changes in TolA (Germon et al., 2001; Goemaere et al., 2007). The regulation of
TolA conformation and interaction between TolA and Pal have been shown to require pmf of the
CM (Cascales et al., 2000, 2001; Germon et al., 2001; Lloubes et al., 2001). By using maintenance of the OM as an indicator of Tol system function, it has been possible to study the
Tol system and to develop an understanding of the underlying mechanisms of its energy- dependent roles in the Gram negative cell. Characterization of the Tol system has led to the proposal that TolQ/TolR pair function as a molecular motor, as is also believed do ExbB/ExbD
(Geomaere et al, 2007; Xiang et al., 2011). Homologies between these two complexes, as well as their ability to substitute for one another in mutational analyses, lend further support to the conclusion that the Tol-Pal and TonB systems carry out their distinct roles in a functionally similar manner.
Because it has proven difficult to decipher the true physiological role(s) of the Tol system, most studies have focused on its role in the entry of colicins and filamentous phage, as well as the effect of tol mutations on OM integrity. It is thought that Tol-dependent colicin uptake is not dependent upon energy input, but rather requires a structurally competent Tol-Pal system (Bourdineaud et al., 1990; Lazdunski et al., 1998; Journet et al, 2001). Conversely, 11
evidence indicates that OM maintenance requires energy derived from the CM pmf and that a
Tol-Pal interaction is required for the transport of cell wall components (Lloubes et al., 2001).
For example, the polymerization and expression of the O antigen is potentially influenced by
TolA and Pal (Gaspar et al., 2000; Vines et al., 2005). The TolA-Pal interaction has been shown to be dependent on both energy derived from the CM pmf (Cascales et al., 2000; Lloubes et al.,
2001) and the TolQ/TolR protein pair (Cascales et al., 2001). Additionally, conformational
changes in TolA associated with OM stability are mediated through the pmf and TolQ/TolR
(Germon et al., 2001).
Mutational analyses have provided insight to the specific interactions between the Tol
proteins associated with OM maintenance. The identification of residues conserved between
TolQ/TolR and ExbB/ExbD has provided further evidence that these protein complexes share
both structural and functional homology. Interactions between TolA, TolQ, and TolR have been
shown to occur within transmembrane (TM) α-helix domains. TolQ interacts with the TM of its
energy-harvesting partner TolR through specific residues located in its third TM region
(Lazzaroni et al., 1995). TolQ interaction with TolA occurs via the TolA N-terminal domain
(Derouiche et al., 1995) and the first TM domain of TolQ (Germon et al., 1998), and an
interaction between TolA and TolR occurs through both of their TM domains (Journet et al.,
1999). Intramolecular interactions within TolQ have also been detected and appear to occur
between TM I and TM II as well as between TM I and TM III, with contacts near the
cytoplasmic face (Xiang et al., 2011). Additionally, using C-terminal hemagglutinin (HA) and
8-histidine tags, TolQ dimerization was confirmed to occur independent of TolR (Xiang et al.,
2011). Molecular modeling and mutational analyses have led to a predicted model in which
TolQ/TolR comprise an ion-conducting channel (Zhai et al., 2003; Goemaere et al., 2007; Xiang, 12
et al., 2011). This aqueous channel may be formed by the second and third TM regions of TolQ
and the TM region of TolR, with specific residues in the two C-terminal TolQ TM regions and in
the TolR TM region necessary for proton or ion translocation (Geomaere et al., 2007; Xiang et al, 2011). CM-derived pmf is converted to mechanical energy as TolA undergoes a conformational change to interact with TolB and Pal, and through Pal, with OmpA and Lpp
(Clavel et al., 1998). While details of Tol system function remain largely left to be uncovered,
sequence, structural, and functional homology, the existence of key conserved residues, and
accumulated data that contribute to the understanding of Tol protein interactions and function all
indicate that this protein complex may function as an CM-energized system similar to the more
fully characterized TonB system.
I initially began working with the Tol-Pal system for my Master’s thesis. As a part of this work, the amino terminal sequence of tolQ was modified and expressed in a pBAD expression vector. The unintended consequence of this N-terminal modification was that it altered TolQ-related phenotypes. Therefore, the focus of my Master’s project became to characterize the phenotypes associated with this N-terminally modified TolQ protein. During the study, it was necessary to determine the proper induction levels for the plasmid-encoded protein that would allow expression of TolQ at wild type levels. This system allowed us to express TolQ at wild type levels from the plasmid. Additionally, it also made it possible to examine the phenotypes associated with high or low/absent levels of TolQ expression. E. coli cells expressing N-terminally modified TolQ protein had apparently lost the ability to support energy-
dependent Tol-associated activities within the cell, including maintenance of outer membrane
integrity. While these mutants retained colicin A sensitivity, this finding was not surprising
because group A colicin entry into the cell does not appear to require CM-derived energy 13
(Bourdineaud et al., 1990; Goemaere et al., 2007). During this study, it became relevant to also consider the possibility that the N-terminally modified TolQ was unstable and that the phenotypes associated with this mutant might be caused by premature degradation and/or low protein levels in the cells rather than a functional impairment. To determine if this was the case,
Western blot analysis was used to confirm protein levels using an antibody to the N-terminal addition. Additionally, phenotypic assays were conducted to compare this mutant with plasmid encoded wild-type TolQ over a range of arabinose-induced expression levels. Therefore, cells were analyzed by phenotype under induction levels that ranged from no induction, to wild type, and to high levels of induction. Microscopic analysis showed that expression of the TolQ protein well above wild type levels appeared to alter the division process. Under these conditions cells grew as filaments, a phenotype that was not associated with over-expressed TolR or TolA (Fig.
4). In each case, the degree of filamentation correlated directly with the concentration of inducer, with the longest filaments seen at the highest tolQ induction levels (Teleha, 2009; Fig.
5). A review of the literature available at this time indicated that studies had tentatively linked the Tol-Pal protein system to the process of cell division. It had been reported that E. coli tolA mutants display impaired cell division patterns at high and low osmolarities (Meury and
Devilliers, 1999). This distinct morphological phenotype characterized by chaining cells has also been observed in tol-pal mutants of Vibrio cholerae (Heilpern and Waldor, 2000),
Pseudomonas putida (Llamas et al., 2000) and Erwinia chrysanthemi (Dubuisson et al., 2005).
In E. chrysanthemi, mutations in tolQ, tolA, tolB, and pal exhibited both altered cell morphology and membrane blebbing characteristic of tol mutants as well as incorrect localization of the division septum and filamentation (Dubuisson et al., 2005) indicating a critical role for the Tol- pal complex in cell division in E. chrysanthemi. More recently, a role in cell division had been 14
proposed for the Tol system as a mediator of OM invagination (Gerding et al., 2007). Based on these findings, it became apparent that if I were going to continue to work with the Tol system, whose true physiological role within the cell remains elusive, I would need to explore the possibility of a role for this protein complex and particularly for TolQ in the process of cell division.
Figure 4. Over-expression of several Tol system proteins in W3110 E. coli cells grown for 24 hours under Amp selection +/- induction with 0.1% (w/v) ʟ-arabinose. Cells expressing wild type or induced levels of TolQ are shown in Panels A and B, respectively. Cells expressing wild type or induced levels of TolR are shown in Panels C and D, respectively. Cells expressing wild type or induced levels of TolA are shown in Panels E and F, respectively. These images reveal cell filamentation is only apparent in cells over-expressing the TolQ protein (panel B). The over-expression of neither TolR nor TolA has no obvious impact on cell division (Figure from Teleha, 2009). 15
Figure 5. W3110 E. coli cells grown for 24 hours under Amp selection at various concentrations of ʟ-arabinose used to induce expression of TolQ protein. The concentrations of ʟ-arabinose (w/v) were as follows: A. 0.0%, B. 0.00001%, C. 0.0001 , D. 0.001%, E. 0.01%, and F. 0.1%. Cell elongation first becomes apparent at 0.0001% ʟ-arabinose and the extent increases corresponding with increases in ʟ-arabinose levels and presumably, TolQ protein induction levels (Figure from Teleha, 2009).
16
Gram Negative Cell Division
Bacterial cell division is a complex process that involves the coordinated invagination of the cytoplasmic membrane, remodeling of the peptidoglycan, and what is generally believed to be the passive tethering that brings about invagination of the OM. The divisome is a macromolecular assembly that coordinates the process of cell division in bacteria. Ten essential division proteins have been identified, and the number of known non-essential division proteins that associate with the divisome or contribute to the division process continues to grow
(Bernhardt and de Boer, 2003; Goehring and Beckwith, 2005; Vicente et al., 2006; Alexeeva et al., 2010; de Boer, 2010). A homologue of the eukaryotic GTPase tubulins, FtsZ initiates the formation of the divisome by forming a scaffold called the Z ring upon which the remaining divisome proteins assemble (Adams and Errington, 2009). FtsA and ZipA (or ZapA) anchor
FtsZ to the CM in an assembly referred to as the proto-ring (Vicente and Rico, 2006). FtsK coordinates chromosome segregation with the initiation of septation and FtsQ, FtsB and FtsL join the proto-ring to further stabilize and connect the components of the ring. FtsW and FtsI join the ring at an even later stage to synthesize and remodel septal peptidoglycan (Weiss, 2004).
FtsN is the last essential protein to join the divisome, initiating Z ring constriction and recruiting amidases, including AmiC and EnvC that split the septal wall (Rico et al., 2010). Known components of the divisome and their relative order of recruitment to the division site are shown in Fig. 6 (Lutkenhaus, 2009a). It is likely that the order of ring assembly is not linear. Rather, a number of divisome components have been shown to have the potential to form sub-complexes before joining the septal ring (Georhing and Beckwith, 2005).
17
Figure 6. Known essential divisome proteins and their predicted membrane topology. The predicted order of divisome component recruitment to the cell division site is as follows: FtsZ→(FtsA+ZipA)→FtsK→FtsQ+(FtsB+FtsL)→FtsW+FtsI→FtsN. Diagram from Lutkenhaus, J. 2009a, with permission (Appendix A).
While FtsZ and other essential divisome proteins are widely conserved among bacteria, it
has long been thought that FtsN was much less conserved, with sequence homologues largely
confined to the enteric bacteria. Using more broad and rigorous analyses, Möll and Thanbichler
(2009) have identified structural and functional homologues of FtsN in a number of
proteobacteria, indicating that the FtsN protein is more widely conserved than previously
believed. FtsN mutants show an impaired division phenotype, forming long, smooth filaments in
culture (Dai et al., 1993), similar to those observed in cells that over-express TolQ (Teleha,
2009). Although a late recruit to the divisome, there is evidence that FtsN may serve an 18
important role in the stabilization of the divisome, with its absence causing a destabilization and disassembly of early proto-ring components (Rico et al., 2010).
Two negative regulatory systems are responsible for positioning the divisome at the cell septum. The nucleoid occlusion (NO) regulatory system inhibits Z-ring assembly near the chromosomes and requires the DNA-binding protein SlmA in E. coli (Bernhardt and de Boer,
2005). The Min system, comprised of MinC, D, and E, oscillates from one cell pole to the other to inhibit Z-ring assembly at the cell poles. MinD, a membrane-associated ATPase, recruits
MinC to the membrane (Hu et al., 2003). Activated by MinD, Min C inhibits division by interfering with Z-ring assembly through a direct physical interaction with FtsZ (Hu et al., 1999;
Shen and Lutkenhaus, 2011). The complex of MinC/MinD associated with the CM strongly inhibits Z-ring assembly. MinE regulates the positioning of the MinC/MinD complex through rapid pole-to-pole oscillations that restrict the complex to the cell poles. The rapid binding and hydrolysis of ATP by MinD leads to the repeated cyclic binding and release of the proteins from the CM (Meinhardt and de Boer, 2001). While a number of recent studies have served to further describe the regulation of Z-ring assembly by the Min system, details of Min function, such as the control of Min oscillation, remain to be fully understood (reviewed in de Boer, 2010).
Chimerel (2012) reported that a reduction in the membrane potential was shown to prevent formation of the Z-ring. Therefore, MinC/MinD oscillation may require the electrochemical membrane potential.
Z-ring assembly occurs on the cytoplasmic face of the CM. FtsZ forms filaments in a
GTP-dependent manner and these filaments arrange into a ring-like structure at the division site.
This proto-ring is anchored to the CM by both FtsA and ZipA (at least one is required), which interact with the C-terminus of FtsZ (Adams and Errington, 2009). Before the recruitment of 19
additional divisome components occurs, murein synthesis begins at the division site (de Pedro,
1997). How this occurs remains unknown, but evidence indicates that FtsZ has a role in the process. The remaining divisome components join the septal ring in a relatively predictable order, with the joining of FtsN coinciding with the beginning of cell constriction (Addinall et al.,
1997; Moll and Thanbichler, 2009). Cell constriction involves CM invagination and division, septal murein synthesis and splitting, and OM invagination and fission (de Boer, 2010). It has recently been reported that Z-rings (polymers of FtsZ alone) have the ability to generate a force that can cause constriction of the CM (Osawa et al., 2008) and that condensation of FtsZ filaments might possibly produce enough contractile force to power cell division (Lan et al.,
2009). The rearrangement of FtsZ monomers appears to be dependent upon GTP hydrolysis during cell constriction, but it remains unknown the true origin of the force to drive the constriction process (Cytrynbaum et al., 2012). A number of models have been proposed to describe the mechanism and participants in the generation of force that is necessary to drive cell division. One such model involves the Tol-Pal complex (Gerding et al. 2007).
In 2007, Gerding et al. proposed a role and a mechanism for the Tol-Pal proteins as a sub-complex of the cell division machinery in E. coli. Using fluorescent-tagged Tol proteins, they concluded that all five core Tol proteins (TolA, TolQ, TolR, TolB, and Pal) localize to the septum in dividing cells and that the specific recruitment of TolQ to the septum requires FtsN.
Previous studies had suggested that the Tol-Pal system bridges the IM and OM layers of the cell envelope (Guihard et al, 1994), but this was the first proposal for a specific mechanistic role for the system in cell division. The model proposed by Gerding et al. (2007) asserts that active FtsN at the septal ring (SR) recruits the Tol system components to cell constriction sites as invagination of the CM and peptidoglycan (PG) layer occur. At the SR, pmf-energized TolA 20
reaches through newly synthesized PG to interact with Pal, tethering the OM to the CM and
serving to both spatially and temporally coordinate invagination of the OM, PG, and the CM
(Gerding et al., 2007).
Whether the role of the Tol-Pal complex in cell division is pmf-dependent remains
unclear. However, evidence for either a direct or peripheral role in cell division continues to
grow. Because tol-pal mutants of a number of Gram negative bacteria form chains or filaments
when grown at low osmotic conditions (Meury and Devilliers, 1999; Heilpern and Waldor, 2000;
Llamas et al., 2000; and Dubuisson et al., 2005), this highly conserved complex may tether the
OM to the invaginating CM and peptidoglycan layer during cell constriction. Additionally, in
Caulobacter crescentus, tol-pal mutants fail to complete cell division (Yeh et al., 2010). C.
crescentus divides asymmetrically, producing both swarmer and stalked cells. Interestingly, C.
crescentus lacks Lpp, whose uniform distribution around the periphery of the E. coli cell anchors the OM to peptidoglycan. In C. crescentus, TolA and Pal likely fulfill this role, and in the absence of TolA or Pal, mutants fail to divide.
Research Aims
Based on the preliminary data of my thesis, it appeared to be unlikely that the Tol-Pal system serves an essential role in the E. coli cell division process. Rather, its role may be peripheral or accessory, albeit advantageous in a natural environment. It is more likely that the over-expression of TolQ, based on its location and topology in the CM, interacts with and serves to sequester an essential divisome component, possibly one that recruits the Tol-Pal complex to the divisome upon cell constriction. This study tests the hypothesis that a specific and direct interaction between E.coli TolQ and the divisome protein FtsN relates to cell division and that 21
excess TolQ in E. coli sequesters the essential divisome protein FtsN, inhibiting it from carrying out its role in cell division. To test this hypothesis, four specific aims were developed.
Aim 1: Further characterize the TolQ over-expression division phenotype.
Objectives:
1. Examine Tol system mutant cells under normal and low osmolarity growth
conditions for evidence of division impairment.
2. Over-express TolQ using arabinose induction in various Tol system mutants to
observe the impact on division independent of complete Tol complexes.
3. Evaluate the TolQ over-expression division phenotype as an E. coli-specific
phenomenon by over-expression of E. coli TolQ in other proteobacteria.
4. Confirm the equivalency of arabinose induction levels with protein expression using
Western blot analysis and an antibody specific to the TolQ protein.
Aim 2: Identify potential protein-protein interactions between TolQ and FtsN using a bacterial two-hybrid system.
Objectives:
1. Determine whether TolQ and FtsN can interact.
2. If so, identify specific regions where TolQ and FtsN interact.
22
Aim 3: Examine the division phenotype associated with the dual over-expression of TolQ and
FtsN.
Objectives:
1. Over-express TolQ and FtsN using different pBAD vector constructs and simultaneous
arabinose induction conditions.
2. Over-express TolQ using a pBAD vector and arabinose induction while over-
expressing FtsN using a pPro vector and propionate induction to eliminate the possibility
of an arabinose dilution effect on division by the presence of multiple pBAD constructs.
Aim 4: Modify TolQ to identify key regions associated with the over-expression division phenotype.
Objectives:
1. Generate TolQ construct with N-terminal addition to evaluate the impact on the over-
expression division phenotype.
2. Generate constructs of TolQ with N-terminal truncations to assess the impact on the
over-expression division phenotype.
3. Generate TolQ mutants with various amino acid substitutions to evaluate the impact on
the over-expression division phenotype.
23
CHAPTER II
CORRELATION BETWEEN ARABINOSE INDUCTION LEVELS AND TOLQ PROTEIN
EXPRESSION AMOUNTS AND THE CONSERVED NATURE OF THE TOLQ OVER-
EXPRESSION DIVISION PHENOTYPE
Introduction
Much of what is known about the Tol-Pal system comes from studies in Escherichia coli.
A variety of phenotypes describe tol mutants: loss of OM integrity, increased sensitivity to antibiotics and detergents, decreased or loss of sensitivity to colicins, leakage of periplasmic proteins, and the formation of outer membrane vesicles (Lazdunski, et. al., 1998). Cell division abnormalities have also been reported in E. coli tol-pal mutants (Meury and Devilliers, 1999;
Gerding et al., 2007). Since its identification and characterization in E. coli, the Tol-Pal system has been found to be conserved in a growing number of other Gram negative bacteria species.
Homologues have been identified in Haemophilus influenzae (Deich et al., 1988; Sen et al.,
1996), Pseudomonas aeruginosa (Dennis et al., 1996; Lim et al., 1997) Pseudomonas putida
(Rodriguez-Herva et al., 1996), Brucella abortus (Tibor et al., 1994), Vibrio cholera (Heilpern and Waldor 2000), Salmonella enterica (Bowe et al., 1998), Erwinia chrysanthemi (Dubuisson et al., 2005), as well as in hundreds of other species. Conservation of the Tol-Pal system is both structural as well as functional, as tol-pal mutants in other Gram negatives show phenotypes comparable, although not always identical, to those in E. coli (Rodriquez-Herva et al., 1996;
Heilpern and Waldor 2000; Dubuisso et al., 2005; Yeh et al., 2010). In some cases, the isolation of tol mutants has proven to be difficult, suggesting that for certain bacterial species the Tol-Pal complex performs unique or additional functions beyond those performed in E. coli. This can be seen in the inability to isolate viable Pseudomonas aeruginosa tolQRAB mutants (Dennis et al., 24
1996). In many studies, tol mutants have been observed to not only display impairments in cell
envelope integrity, synthesis, and maintenance, but also to show division abnormalities,
suggesting a common link to the cell division process in species in which the Tol-Pal system is
conserved (Heilpern and Waldor, 2000; Llamas et al., 2000; Dubuisson et al., 2005; Yeh et. al.,
2010).
The structure of the tol-pal operon is conserved across the Gram negative bacteria. It has not been found in the Gram positives, which can be expected based on its role in OM maintenance. Gram negatives lacking the cluster tend to be intracellular parasites (Fraser et al.,
1997; 1998; Anderson et al., 1998; Casjens et al., 2000; Parkhill et al., 2000; Tettelin et al.,
2000). The tolA sequence is the least conserved, even among close relatives, and its
identification proves difficult unless it is analyzed in the context of the entire tol-pal gene cluster.
The tolQ and pal genes are widely conserved, with the tolQ/tolR pair being the most widely
distributed (Sturgis, 2001). The entire gene cluster, from ygbC to pal, is generally well
conserved within the α-,β-, and γ- proteobacteria, while outside these subdivisions, ybgC is
absent and the gene cluster begins with tolQ. Outside of the gene cluster, tolQ and tolR are often
found as a pair, such as in Neiserria meningitidis, which lacks the tol-pal gene cluster. In other
genomes, such as that of Pseudomonas aeruginosa, tolQ and tolR are present as multiple copies.
In fact, the tolQ/tolR pair is found widespread across the eubacteria except in the Gram positives
and the spirochetes. Finally, in Rickettsia prowazekii, the genes encoding the Tol-Pal proteins
are found scattered throughout the genome (Sturgis, 2001). These findings of a widespread
distribution across the eubacteria suggest an ancient origin and conserved function for the tol-pal
gene products. 25
Evidence for both the conserved nature and evolutionary divergence of the tol-pal system
is abundant. That TolQ/TolR, ExbB/ExbD, and MotA/MotB share both structural and functional
homology among their transmembrane regions and are thought to operate as ion potential-driven
motors suggests both that the ancestral protein pair is ancient and that evolutionary
diversification has led to specialized function among these three protein pairs (Cascales et al.,
2001). Between more closely related groups of bacteria, tol function is still relatively
homologous. As with E. coli tol mutants, P. putida, V. cholera, and Caulobacter crescentus tol mutants exhibit characteristic OM disruptions including vesicle formation, increased sensitivity to antibiotics, and leakage of periplasmic proteins. Additionally, these tol mutants also exhibit division impairments (Heilpern and Waldor, 2000; Llamas et al., 2000; Dubuisson et al., 2005;
Yeh et. al., 2010). Furthermore, in C. crescentus, it was reported that the accumulation of TolA
and Pal at the division site is dependent upon FtsZ assembly (Yeh et al., 2010). Likewise, in E.
coli, polymerization of FtsZ at the division site appears to be required for TolA recruitment
(Gerding et al., 2007).
The ability of Tol-Pal system components from one bacterial species to compliment
mutants of another has been addressed. E. coli tolR mutants are complimented by P. pudita tolR
(Llamas et al., 2000). The Yersinia enterocolitica TolA homologue is able to compliment an E.
coli tolA mutant (Weitzel and Larsen, 2008). While E. chrysanthemi tol-pal genes have been
shown to compliment E. coli tol-pal mutants, E. coli tol-pal genes are unable to compliment E.
chrysanthemi tol-pal mutants (Dubuisson et al., 2005). The variable abilities of tol-pal genes to
compliment tol-pal mutants across species raises the question of whether the E. coli tolQ over-
expression phenotype can be induced by over-expression of E. coli tolQ in other proteobacteria
or if this phenotype is specific to E. coli. 26
Aim 1, to further characterize the TolQ over-expression division phenotype, was designed with four objectives in mind. The first objective was to assess more fully the division phenotypes associated with the specific tol deletion mutants used in this study. While these tolA,
tolQ, and tolR mutants have been previously examined in regards to OM integrity and colicin
and phage sensitivity in the Larsen lab, cell division phenotypes at both standard and low
osmotic conditions had not. Additionally, over-expression of TolQ in each of these tol mutants
would provide insight into whether or not an intact, stoichiometrically balanced Tol-Pal complex
is necessary to produce the filamentation phenotype. The third objective was to investigate the
phenotype, if any, associated with the over-expression of TolQ in other proteobacteria. Two
species of Gram negative bacteria of the closely related Enterobacter/Klebsiella lineage (Paradis
et al., 2005) and the E. coli strain BL21 were transformed with a plasmid encoding E. coli TolQ.
Division phenotypes associated with the over-expression of E. coli TolQ in these related species
was then assessed visually. The fourth objective of Aim 1 was to have an antibody to TolQ
synthesized to be used in Western blot analysis to confirm TolQ expression levels above wild
type and to correlate levels of inducer (ʟ-arabinose) with resultant TolQ levels in E. coli cells.
Western blot analysis could then confirm that TolQ levels attained in the cell directly correlated
with the degree of filamentation observed at each inducer concentration. In the preliminary work
to this study, an antibody-detectable tag was added to the N-terminus of TolQ to detect protein
levels. However, this tag altered TolQ energy-dependent phenotypes and this construct was
therefore, not appropriate for this aim (Teleha, 2009).
27
Methods
Media
Bacterial strains were maintained on Luria-Bertani (LB) plates. Cells were grown in LB broth with shaking at 37ºC (Miller, 1972). For filamentation assays cells were grown for 24 hrs in LB broth supplemented with various concentrations of ʟ-arabinose (0.1%, 0.01%, 0.001%,
0.0001%, 0.00001% or 0.0% (w/v)). For colicin sensitivity determination, cells were plated in
T-top on T-plates (Miller, 1972) as previously described (Larsen et al., 2003). Plates and broth
were supplemented with 100μg ml-1 ampicillin, as necessary. All supplements to media were
made on a basis of weight per volume (w/v).
Strains and Plasmids
Bacterial strains and plasmids are summarized in Table 1. The Escherichia coli K-12
strain W3110 (Hill and Harnish, 1981) was used in this study as the wild type. The W3110
derivatives RA1027, RA1028, and RA1038 carried precise, complete deletions of the tolQ, tolR,
and tolA genes, respectively. The construction of RA1038 was previously described (Weitzel
and Larsen, 2008). Deletions of tolQ and tolR were similarly created using the λ red
recombination technique (Datsenko and Wanner, 2000, Brinkman, 2007). As with tolA, this
approach cleanly replaced the predicted open reading frame for each gene with a “scar” region
containing a stop codon and a ribosome-binding site to minimize disruption of downstream
genes. The Enterobacter amnigenus strain ATCC51816 and Cronobacter muytjensii strain
ATCC51329 were purchased from the American Type Culture Collection. The E. coli B strain
BL21 was purchased from New England Biolabs (NEB).
The Tol phenotypes of deletion mutants and their complementation by plasmids were
evaluated by resistance to deoxycholate and sensitivity to group A colicins as previously 28
described (Brinkman and Larsen, 2008; Weitzel and Larsen, 2008; Teleha, 2009). Expression of specific Tol system proteins was achieved using plasmids derived from pBAD24 (Guzman et al.,
1995) as previously described (Brinkman and Larsen, 2008; Teleha, 2009). The relevant strains and plasmids including the characteristics of the pBAD24 vector used in this study are described in Table 1.
Table 1. Strains and plasmids used in Chapter II
Strain Relevant characteristics1 Reference or source W3110 F-IN(rrnD-rrnE)1 Hill and Harnish, 1981 RA1027 W3110-ΔtolQ Brinkman, 2007 RA1028 W3110-ΔtolR Brinkman, 2007 RA1038 W3110-ΔtolA Weitzel and Larsen, 2008 Enterobacter amnigenus Clinical isolate ATCC (ATCC51816) Cronobacter muytjensii Clinical isolate ATCC (ATCC51329) BL21 fhuA2[lon] ompT gal [dcm] ΔhsdS New England BioLabs Plasmid Relevant characteristics Reference or source pBAD24 araBAD promoter, AraC, amp r Guzman et al., 1995 pRA002 pBAD24 encoding E. coli TolR Brinkman and Larsen, 2008 pRA004 pBAD24 encoding E. coli TolA Brinkman and Larsen, 2008 pRA031 pBAD24 encoding E. coli TolQ Teleha, 2009
1The specific nature of the deletions∆) ( are as follows: The ∆tolQ deletion removed the predicted tolQ codons 1-230 and the TAA termination codon; the ∆tolR deletion removed the predicted tolR codons 1-142 and the TAA termination codon; the ∆tolA deletion removed the predicted tolA codons 1-421, leaving the TAA termination codon in place.
29
Transformation
W3110, E. amnigenus, C. muytjensii, and BL21 cells were made competent and transformed using a procedure based on Chung et al. (1989). Specifically, overnight cultures (1.5 ml) were pelleted by centrifugation and resuspended in 100 µl ice cold TSS (LB with 10% (w/v) polyethelyne glycol [PEG 8000], 5% (w/v) dimethyl sulfoxide [DMSO], and 50 mM MgCl2, pH
6.5). A 30-minute incubation period on ice was followed by heat shock for 2 minutes at 37°C.
Five hundred microliters of warmed SOC (2% (w/v) tryptone, 0.5% (w/v) yeast extract, 10 mM
NaCl, 2.5 mM KCl, 10 mM MgCl2, 10 mM MgSO4 and 20 mM glucose) was added and the culture was shaken for 60 minutes at 37°C. Transformation cultures were spread as 100 µl aliquots on warmed LB agar supplemented with 100 μg ml-1 ampicillin. Plates were incubated
overnight at 37°C.
Cell division phenotypes of tolQ, tolR, and tolA deletion mutants under growth conditions of
standard and low osmolarity.
Overnight broth cultures of RA1027, RA1028, and RA1038 were subcultured 1:200 in
LB broth and grown for 24h at 37ºC with shaking. For low osmotic assays, cells were grown in a modified LB broth, made without the normal addition of 1.0% (w/v) NaCl. Cells were then sampled using an inoculating loop, heat fixed to a glass slide, and stained with safranin before being viewed under light microscopy for division phenotypes. Images were obtained using a
Nikon H550S series compound light microscope at 1000x total magnification, and NIS Elements
Documentation Software.
Over-expression of tolQ in tolQ, tolR, and tolA deletion mutants
Overnight cultures of RA1027, RA1028, and RA1038 TSS-transformed with pRA031
grown in LB supplemented with ampicillin (100μg ml-1) were subcultured 1:200 in LB 30
supplemented with 100 μg ml-1 ampicillin with 0.1% (w/v) ʟ-arabinose and grown for 24 h at
37ºC with shaking. Cells were then stained with safranin and viewed for division phenotypes.
Cells were viewed and images were captured as described above.
Over-expression of E. coli TolQ in Enterobacter amnigenus, Cronobacter muytjensii, and BL21
Overnight cultures of W3110, Enterobacter amnigenus, Cronobacter muytjensii, and E.
coli strain BL21 TSS-transformed with pRA031 grown in LB supplemented with ampicillin
(100μg ml-1) were subcultured 1:200 in LB supplemented with 100 μg ml-1 ampicillin with 0.1%
(w/v) ʟ-arabinose and grown for 24 h at 37ºC with shaking. Cells were then stained with
safranin and viewed for division phenotypes. Cells were viewed and images captured as
described above.
Western analysis of TolQ expression levels
An antibody to wild-type, unmodified TolQ was raised by Pacific Immunology for an appropriate cytoplasmic region of TolQ. The predicted linear epitope occurring within the large cytoplasmic loop of TolQ that was chosen as the most immunogenic by Pacific Immunology is
TolQ47-62. The primary sequence containing the epitope of TolQ recognized by this antibody is
shown in Fig. 7. This antibody was used in Western blot analyses to determine the level of TolQ
present in wild type E. coli and to detect protein levels in the cell under ʟ-arabinose induction.
Overnight cultures of W3110 carrying pRA031 were subcultured 1:200 in LB ampicillin (100µg
ml-1) supplemented with ʟ-arabinose at 0.1%, 0.01%, 0.001%, 0.0001%, 0.00001%, and 0.0%
(w/v) and grown at 37°C to an A550 of 0.7 as determined with a Spectronic 20 spectrophotometer
with a path length of 1.5 cm. At A550 of 0.7, cells were sampled with a heat-sterilized nichrome
bacteriological loop and heat fixed to a slide, and then stained with safranin to examine division 31
phenotypes. To immediately precipitate proteins, trichloroacetic acid (TCA) was added to culture samples to a final concentration of 10% (w/v), mixed, and incubated on ice for 15 min.
The samples were then centrifuged at 4ºC for 5 min, aspirated, and the pellets washed with 1 ml of 100 mM Tris-HCl (pH 7.9). The wash was immediately removed by aspiration and the pellet was centrifuged for 5 min at 4ºC. The remaining supernatant was aspirated, the pellet was resuspended in 25 μl 1 M Tris-HCl and 25 μl 2X Laemmli sample buffer, and tubes were boiled for 5 minutes (Abelson et al., 1990). Samples were subjected to electrophoresis on SDS 11% polyacrylamide gels. Resolved proteins were electrotransferred to Immobilon P® (Millipore
Corp.) membrane and Western blot analyses performed using the TolQ antibody described above
(Pacific Immunology), anti-rabbit Immunoglobulin horseradish peroxidase (HRP) conjugate and enhanced chemiluminescence (ECL), as previously described (Larsen et al., 1993, Higgs et al.,
1998).
32
10 20 30 40 50 60 GTG ACT GAC ATG AAT ATC CTT GAT TTG TTC CTG AAG GCT AGC CTT CTG GTT AAA CTT ATC ATG TTG CAC TGA CTG TAC TTA TAG GAA CTA AAC AAG GAC TTC CGA TCG GAA GAC CAA TTT GAA TAG TAC AAC Met Thr Asp Met Asn Ile Leu Asp Leu Phe Leu Lys Ala Ser Leu Leu Val Lys Leu Ile Met Leu>
70 80 90 100 110 120 130 ATT TTG ATT GGT TTT TCA ATC GCA TCT TGG GCC ATT ATT ATC CAG CGG ACC CGT ATT CTT AAC GCA TAA AAC TAA CCA AAA AGT TAG CGT AGA ACC CGG TAA TAA TAG GTC GCC TGG GCA TAA GAA TTG CGT Ile Leu Ile Gly Phe Ser Ile Ala Ser Trp Ala Ile Ile Ile Gln Arg Thr Arg Ile Leu Asn Ala>
140 150 160 170 180 190 GCG GCG CGC GAA GCC GAA GCG TTT GAA GAT AAA TTC TGG TCT GGA ATC GAA CTC TCT CGC CTC TAT CGC CGC GCG CTT CGG CTT CGC AAA CTT CTA TTT AAG ACC AGA CCT TAG CTT GAG AGA GCG GAG ATA Ala Ala Arg Glu Ala Glu Ala Phe Glu Asp Lys Phe Trp Ser Gly Ile Glu Leu Ser Arg Leu Tyr>
200 210 220 230 240 250 260 CAA GAG AGC CAG GGG AAA CGG GAT AAT CTG ACT GGT TCG GAA CAA ATC TTT TAC AGC GGG TTC AAA GTT CTC TCG GTC CCC TTT GCC CTA TTA GAC TGA CCA AGC CTT GTT TAG AAA ATG TCG CCC AAG TTT Gln Glu Ser Gln Gly Lys Arg Asp Asn Leu Thr Gly Ser Glu Gln Ile Phe Tyr Ser Gly Phe Lys>
270 280 290 300 310 320 330 GAG TTT GTG CGC CTG CAT CGT GCC AAT AGC CAT GCG CCG GAA GCC GTA GTG GAA GGG GCG TCG CGT CTC AAA CAC GCG GAC GTA GCA CGG TTA TCG GTA CGC GGC CTT CGG CAT CAC CTT CCC CGC AGC GCA Glu Phe Val Arg Leu His Arg Ala Asn Ser His Ala Pro Glu Ala Val Val Glu Gly Ala Ser Arg>
340 350 360 370 380 390 GCT ATG CGT ATC TCC ATG AAC CGT GAA CTT GAA AAT CTG GAA ACG CAC ATT CCG TTC CTC GGT ACG CGA TAC GCA TAG AGG TAC TTG GCA CTT GAA CTT TTA GAC CTT TGC GTG TAA GGC AAG GAG CCA TGC Ala Met Arg Ile Ser Met Asn Arg Glu Leu Glu Asn Leu Glu Thr His Ile Pro Phe Leu Gly Thr>
400 410 420 430 440 450 460 GTT GGC TCC ATC AGC CCG TAT ATT GGT CTG TTT GGT ACG GTC TGG GGG ATC ATG CAC GCC TTT ATC CAA CCG AGG TAG TCG GGC ATA TAA CCA GAC AAA CCA TGC CAG ACC CCC TAG TAC GTG CGG AAA TAG Val Gly Ser Ile Ser Pro Tyr Ile Gly Leu Phe Gly Thr Val Trp Gly Ile Met His Ala Phe Ile>
470 480 490 500 510 520 GCC CTC GGG GCG GTA AAA CAA GCA ACA CTG CAA ATG GTT GCG CCC GGT ATC GCA GAA GCG TTG ATT CGG GAG CCC CGC CAT TTT GTT CGT TGT GAC GTT TAC CAA CGC GGG CCA TAG CGT CTT CGC AAC TAA Ala Leu Gly Ala Val Lys Gln Ala Thr Leu Gln Met Val Ala Pro Gly Ile Ala Glu Ala Leu Ile>
530 540 550 560 570 580 590 GCG ACT GCA ATT GGT CTG TTT GCC GCT ATC CCG GCA GTT ATG GCC TAC AAC CGC CTC AAC CAG CGC CGC TGA CGT TAA CCA GAC AAA CGG CGA TAG GGC CGT CAA TAC CGG ATG TTG GCG GAG TTG GTC GCG Ala Thr Ala Ile Gly Leu Phe Ala Ala Ile Pro Ala Val Met Ala Tyr Asn Arg Leu Asn Gln Arg>
600 610 620 630 640 650 660 GTA AAC AAA CTG GAA CTG AAT TAC GAC AAC TTT ATG GAA GAG TTT ACC GCG ATT CTG CAC CGC CAG CAT TTG TTT GAC CTT GAC TTA ATG CTG TTG AAA TAC CTT CTC AAA TGG CGC TAA GAC GTG GCG GTC Val Asn Lys Leu Glu Leu Asn Tyr Asp Asn Phe Met Glu Glu Phe Thr Ala Ile Leu His Arg Gln>
670 680 690 GCG TTT ACC GTT AGC GAG AGC AAC AAG GGG TAA CGC AAA TGG CAA TCG CTC TCG TTG TTC CCC ATT Ala Phe Thr Val Ser Glu Ser Asn Lys Gly ***>
Figure 7. Epitope recognized by anti-TolQ antibody. Nucleotide and predicted amino acid sequence of E. coli TolQ. Highlighted in yellow is the region that comprises the large cytoplasmic loop, amino acids 39-126. The peptide used to produce the TolQ antibody comprises amino acids 47-62 and is highlighted in green.
33
Results
Tol system mutants grown at standard and low osmolarity
Division phenotypes of E. coli tol mutants lacking essential components of the energy-
harvesting and transducing Tol system were examined microscopically. To determine if the
absence of either member of the TolQ/TolR pair or the TolA protein halted the cell division
process, mutants were grown for 24 hrs and division phenotypes were determined
microscopically. As Tol system mutants of E. coli and other Gram negatives have been reported
to form chains or filaments when grown under low osmotic conditions (Meury and Devilliers,
1999; Heilpern and Waldor, 2000; Llamas et al., 2000; Dubuisson et al., 2005; Gerding et al.,
2007), tolA, tolQ, and tolR mutants were also grown in standard media without added NaCl.
RA1027 (ΔtolQ), RA1028 (ΔtolR), and RA1038 (ΔtolA), grown in standard LB broth, appeared
to undergo the cell division process to completion. When stained and viewed microscopically,
RA1027, RA1028, and RA1038 cells displayed cell dimensions somewhat smaller than those of
the wild type (W3110) cells (Fig. 8, left panels). The same tol mutants, grown in LB lacking
added NaCl, formed short chains of cells that appeared to be arrested at a late stage of cell
division (Fig. 8, right panels). The chaining phenotype observed with E. coli tol mutants grown
in low salt conditions is similar to that reported by Meury and Devilliers (1999) and Gerding et al. (2007). Notably, the chaining phenotype observed in tolQ, tolR, and tolA deletion mutants grown under low osmotic conditions is distinct from the long, smooth filaments that result from tolQ over-expression as observed in preliminary studies (Teleha, 2009) and this study, which is more similar to that reported for FtsN mutants by Dai et al. (1993).
34
Figure 8. tolQ, tolR, and tolA deletion mutants grown at standard osmolarity and under low-salt conditions. Panels A and B are micrographs of E. coli W3110 grown at standard osmolarity (left) and in LB media with no added NaCl (right). Panels C and D are micrographs of RA1038 (ΔtolA), panels E and F are micrographs of RA1028 (ΔtolR), and panels G and H are micrographs of RA1027 (ΔtolQ). Tol deletion mutants grown for 24 h. in standard LB (left panels) appear to divide normally as do wild type cells. In the right panels, the same tol system deletion mutants form short cell chains when grown under low-salt conditions. 35
Over-expression of tolQ in tolQ, tolR, and tolA deletion mutants
Many Tol system functions that collectively preserve the integrity of the OM have been
shown to require CM-derived energy (Cascales et al., 2007; Noinaj et al, 2010). The TolQ/TolR pair appears to function as a molecular motor that energizes TolA (Geomaere et al., 2007).
TolQ/TolR are hypothesized to comprise an ion-conducting channel (Zhai et al., 2003) that converts CM-derived energy (pmf) into mechanical energy as TolA undergoes a conformational change to interact with TolB and Pal, and through Pal, with OmpA and Lpp (Clavel et al., 1998;
Cascales et al.; 2000, 2001, Germon et al.; 2001, Lloubes et al., 2001). Interactions between
TolQ, TolR, and TolA have been shown to be necessary for Tol functions, including OM maintenance, and to involve transmembrane domains of each protein (Derouiche et al., 1995;
Lazzaroni et al., 1995; Germon et al., 1998; Journet et al., 1999). To determine whether the
TolQ over-expression phenotype is dependent upon Tol-derived energy, division phenotypes of tolA, tolQ, and tolR deletion mutants over-expressing TolQ were analyzed. It was expected that
if filamentation of cells that over-express TolQ required CM-derived energy via the Tol system,
the over-expression of TolQ in Tol deletion mutants would not result in cell filamentation. The
over-expression of TolQ in tolQ (RA1027), tolR (RA1028), and tolA (RA1038) deletion mutants
resulted in cell filamentation that was comparable to that observed in W3110 (wild type) cells
that over-express TolQ. Furthermore, filamentation phenotypes between each mutant strain
over-expressing TolQ were indistinguishable from one another. When grown for 24 h in LB
supplemented with 0.1% (w/v) ʟ-arabinose, RA1027, RA1028, and RA1038 carrying
pRA031grew as filaments, while cells grown under conditions of no arabinose induction did not
(Fig. 9). As can be seen in both Figs. 4 and 9, it appears that the over-expression of TolQ causes
cell filamentation regardless of whether all of the Tol system components are present or TolQ, 36
TolR, or TolA are absent. Because the energy-dependent functions of the Tol-Pal system have been shown to require the five essential tol-pal gene products (TolA, TolQ, TolR, TolB, Pal), the observation of filamentation in tolQ, tolR, and tolA mutants might suggest a role for TolQ in cell division that is independent of an energized Tol-Pal system. These findings corroborate what was observed in preliminary studies. When TolQ and TolR were over-expressed together in
ΔtolQ/ΔexbB/ ΔexbD E. coli mutants, cells filamented (Teleha, 2009). Because cell filamentation occured in the absence of both TolQ/TolR and the ExbB/ExbD pair, which are known to substitute for TolQ/TolR in energy-dependent processes (Braun and Herrmann, 1993), these results strongly indicated that the filamentation phenotype was Tol-derived energy-independent
and might instead reflect a structural role for TolQ or an additional, as yet undefined, role for
TolQ in the bacterial cell division process. On the other hand, if as suggested by Gerding et al.
(2007), an energized Tol-Pal complex does indeed participate in the cell division process, these
results might suggest that its participation is mediated through a direct physical interaction
between TolQ and FtsN.
Over-expression of E. coli TolQ in Enterobacter amnigenus, Cronobacter muytjensii, and E.coli
BL21
The Tol-Pal system is widely conserved across the Gram negative bacteria (Deich et al.,
1988; Tibor et al., 1994; Sen et al., 1996; Dennis et al., 1996; Rodriguez-Herva et al., 1996; Lim
et al., 1997; Bowe et al., 1998; Heilpern and Waldor, 2000; Sturgis, 2001; Dubuisson et al.,
2005). Conservation of the Tol-Pal system is both structural as well as functional, as tol mutants
in other Gram negatives show phenotypes comparable, although not always identical, to those in
E. coli (Rodriquez-Herva et al., 1996; Heilpern and Waldor, 2000; Dubuisson et al., 2005; Yeh
et al., 2010). The ability of E. coli Tol system proteins and Tol proteins from other Gram 37
negatives to compliment mutants of each other is variable (Sun and Webster, 1987; Dubuisson et
al., 2005; Weitzel and Larsen, 2008). This ability is likely influenced by the degree of
relatedness between the bacterial species, and therefore, the degree of homology that exists
between Tol proteins.
Figure 9. Tol deletion mutants RA1027 (ΔtolQ), RA1028 (ΔtolR), and RA1038 (ΔtolA) carrying tolQ (pRA031) grown for 24 hours with and without 0.1% (w/v) ʟ-arabinose. Panels A and B are micrographs of TolQ expressed without or with 0.1% ʟ-arabinose induction in RA1027, respectively. Panels C and D are micrographs of TolQ expressed in RA1028 without or with 0.1% ʟ-arabinose, respectively. Panels E and F are micrographs of TolQ expressed without or with 0.1% ʟ-arabinose induction in RA1038, respectively. These images reveal over- expression of TolQ results in similar filamentation phenotypes in deletion mutants of tolQ, tolR, and tolA. 38
Between more closely related groups of bacteria, Tol-Pal function is still relatively homologous.
As with E. coli tol mutants, P. putida, V. cholera, and C. crescentus tol mutants exhibit characteristic OM disruptions including vesicle formation, increased sensitivity to antibiotics, and leakage of periplasmic proteins. Additionally, these tol mutants also exhibit division impairments (Heilpern and Waldor, 2000; Llamas et al., 2000; Dubuisson et al., 2005; Yeh et. al., 2010). In order to determine if the TolQ over-expression phenotype of filamentation is also conserved or is strictly an E. coli-specific phenomenon, additional Gram negative species were chosen in which to over-express plasmid-encoded TolQ. The species chosen for this investigation were Enterobacter amnigenus, Cronobacter muytjensii, and E. coli strain BL21.
Enterobacter amnigenus, first identified as a distinct species within the Enterobacter genus in
1981 (Izard et al., 1981) belongs to the Enterobacteriaceae class within which E. coli is also classified. E. amnigenus is a close relative of Enterobacter cloacae (Izard et al., 1981).
Currently, very few Enterobacter bacterial genomes are available, as no entire genomes have been sequenced. While individual Enterobacter protein sequences are available, no
Enterobacter TolQ protein sequences are available for comparison (NCBI). Similarly, while the entire genome of Cronobacter muytjensii has yet to be sequenced and a TolQ protein sequence from C. muytjensii is not yet available, individual protein sequences from C. muytjensii have been used to also classify this species within the γ-proteobacteria class. The Cronobacter genus was first proposed in 1980 (Farmer et al., 1980) and C. muytjensii was categorized into this genus in 2008 by Iversen et al. BL21 is a laboratory strain of E. coli that like the K-12 strain, has been widely used in studies of basic biology, medicine, and biotechnology. BL21 is a descendant of the E. coli strain named B by Delbrück and Luria in 1942 and was used for phage studies by Delbrück, Luria, and Hershey (reviewed in Daegelen et al., 2009). Like E. coli K-12, 39
B strains are not encapsulated. The BL21 strain used in this study is deficient in the proteases
Lon and OmpT, rendering the cells very efficient at expression of and producing proteins from cloned genes. As the Tol protein complex is conserved among Gram negative species, E. coli
TolQ should theoretically be translated, processed, and functional in other closely related Gram negative species. E. amnigenus, C. muytjensii, and E. coli strain BL21 were each transformed with the pRA031 plasmid encoding E. coli TolQ and grown under conditions of arabinose- induced over-expression. As predicted, under conditions of 0.1% (w/v) ʟ-arabinose induction, each of the species and strains displayed the characteristic filamentation phenotype observed in
K-12 E. coli when over-expressing its native TolQ protein. Additionally, cells carrying the empty pBAD24 vector did not filament at this arabinose concentration (Fig. 10). These results confirm that the TolQ over-expression effect is not a strain- or species-specific phenomenon, but rather is attributable to a characteristic of the TolQ protein itself.
40
Figure 10. TolQ over-expression in E. coli W3110, C. mutygensii, E. amnigenus, and E. coli strain BL21. Left panels are micrographs of wild tpye E coli W3110 (panels A and B), E. amnigenus (panels C and D), C. muytjensii (Panels E and F), and BL21 (panels G and H), carrying pRA031 encoding tolQ, under no ʟ-arabinose induction. Right panels show cells after 24 h. grown at 0.1% (w/v) ʟ-arabinose. Each species or strain exhibits filamentation when TolQ is over-expressed. 41
Western blot analysis of TolQ expression levels
Without a suitable antibody for the detection of TolQ levels, previous analyses were dependent upon the assumption that protein levels correspond to the concentration of inducer ( ʟ- arabinose) used for assays. While the degree of filamentation appeared to correlate directly with
ʟ-arabinose levels, there had still not been a method for assessing protein levels directly. Once an anti-TolQ antibody was obtained, this was now possible. The goal was to show a direct link between ʟ-arabinse levels, the degree of filamentation at each, and TolQ protein levels in the cell. For Western blot analysis, ʟ-arabinose induction levels were 0.0%, 0.00001%, 0.0001%,
0.001%, and 0.01% and 0.1% (w/v) ʟ-arabinose. These ʟ-arabinose levels were chosen to match those used in previous filamentation assays (Fig. 5). Western blot analysis confirmed the approximate ʟ-arabinose concentration required to obtain wild type expression levels of plasmid- encoded TolQ, which was previously determined experimentally using a number of Tol function- dependent assays to be 0.0001-0.001% (w/v) (Fig. 11, lanes 2 and 5). It also showed a direct correlation between ʟ-arabinose concentration and the TolQ protein levels recovered from cells
(Fig. 11, lanes 3-8). Slides prepared from samples taken at the time of protein collection confirmed the direct correlation between TolQ protein levels and the degree of filamentation
(Fig. 11, panels A-F). In the context of this study, Western blot analysis verified that the direct correlation between ʟ-arabinose induction levels and the degree of filamentation observed at each reflected the amount of TolQ produced. Cell elongation became visually apparent at an ʟ- arabinose concentration of 0.001% (w/v) (Fig. 11, panel D), the same concentration shown by
Western blot analysis at which above-wild type levels of TolQ were first evident (Fig. 11, lane
6). The predicted mass of the TolQ protein is ~26 kDa and TolQ has been shown to migrate to 42
~30 kDa (Xiang et al., 2011). A second, minor band was observed at ~52 kDa, similar to that noted by Xiang et al. (2011), who suggested this might represent TolQ dimers.
43
A
B
Figure 11. Western blot analysis of TolQ expression levels. A. Stained preparations of W3110 bearing a plasmid carrying the tolQ gene under the control of the pBAD promoter grown -1 to A550 = 0.7 at 37˚C with aeration in Miller LB supplemented with 100 µg ml ampicillin and either no (panel A), 0.00001% (panel B), 0.0001% (panel C), 0.001% (panel D), 0.01% (panel E) or 0.1% (w/v) ʟ-arabinose (panel F) are shown. B. Western blot analysis of protein samples from -1 cells grown to A550 = 0.7 in Miller LB supplemented with 100 µg ml ampicillin and either 0.1% ʟ-arabinose for RA1027 (lane1) and W3110 carrying the control plasmid pBAD24 (lane 2) (“∆tolQ” and “chromosomal”, respectively) or as indicated for W3110 carrying a pBAD24 derivative bearing the tolQ gene under ʟ-arabinose control (lanes3-8). Samples were resolved by SDS-PAGE on an 11% polyacrylamide gel, transferred to a PVDF membrane and visualized by enhanced chemiluminescence using a monospecific anti TolQ antiserum as described in Methods. The positions of molecular mass standards are indicated as kDa values at the right side of the developed blot. 44
Discussion
The first aim of this dissertation was to further characterize the TolQ over-expression phenotype that was first observed in previous studies (Teleha, 2009). It had been determined that the over-expression of TolQ, but not the over-expression of plasmid-encoded TolA or TolR, under control of an ʟ-arabinose inducible promoter, led to cell filamentation, a visually observable phenotype (Teleha, 2009). Furthermore, the degree of filamentation directly correlated with the amount of arabinose inducer added to the growth media (Teleha, 2009). One objective of Aim 1 of this study was to identify this filamentation phenotype as distinct from that previously reported for Tol mutants in E. coli (Meury and Devilliers, 1999; Gerding et al., 2007) and in other proteobacteria (Heilpern and Waldor, 2000; Llamas et al., 2000; Dubuisson et al.,
2005; Yeh et. al., 2010). The second objective of this aim was to over-express TolQ in ΔtolQ,
ΔtolR, and ΔtolA mutants to confirm that the filamentation phenotype is not dependent upon a complete Tol protein system, as is outer-membrane maintenance. It was previously observed that TolQ over-expression in tolQ/exbB/exbD mutants still led to the formation of cell filaments
(Teleha, 2009), suggesting that even if the phenotype required a complete Tol protein system, it did not require Tol-derived energy, since in the absence of TolQ/TolR, TolA can be energized by
ExbB/ExbD (Braun and Herrmann, 1993). A third objective designed to further characterize the
TolQ over-expression phenotype was to over-express the E. coli TolQ protein in other proteobacteria. Because TolQ is well-conserved among the Gram negative bacteria and highly conserved among proteobacteria (Deich et al., 1988; Tibor et al., 1994; Dennis et al., 1996;
Rodriguez-Herva et al., 1996; Sen et al., 1996; Lim et al., 1997; Bowe et al., 1998; Heilpern and
Waldor, 2000; Sturgis, 2001; Dubuisson et al., 2005), close relatives of E. coli are likely to have the ability to transcribe and translate the E. coli TolQ protein. What remained unknown was 45
whether or not the over-expression of this translation product would induce a similar phenotype
in relatives of E. coli. The over-expression of E. coli TolQ in other proteobacteria would allow for a determination of whether the TolQ over-expression filamentation phenotype observed in
W3110 E. coli is an E. coli-specific phenomenon or if relatives of E. coli experience comparable disruptions in cell division when over-expressing the E. coli TolQ protein. Lastly, a final
objective of aim one of this dissertation was to directly correlate the TolQ over-expression
phenotype with the amount of TolQ protein present in the cell. In preliminary studies, this had
only been accomplished indirectly. TolQ expression, under control of the pBAD arabinose-
inducible promoter (Guzman et al., 1995), indicated that the higher the concentration of ʟ-
arabinose present in the media, the more pronounced the degree of filamentation observed in
cells (Teleha, 2009, Fig. 5). Without an antibody to TolQ, it had not been possible to use
Western blot analysis to determine TolQ levels in the cell at various ʟ-arabinose levels. An N-
terminal antibody-detectable tag added to the TolQ protein disrupted normal cell growth and led
to the loss of various Tol-dependent functions (Teleha, 2009) and therefore, did not provide a
suitable method for detecting protein levels. This last objective of aim one of this dissertation
was to obtain an antibody to TolQ that could be used to directly asses TolQ levels at previously-
tested ʟ-arabinose concentrations, linking induction levels with the severity of the filamentation
phenotype.
RA1038 (ΔtolA), RA1027 (ΔtolQ), and RA1028 (ΔtolR) deletion mutants, when grown
in LB at standard osmolarity did not form filaments. Likewise, when grown in medium lacking
added NaCl, the Tol mutants did not form filaments. However, the cells did appear to suffer
division impairment. Deletions of tolA, tolQ, and tolR appeared to halt cell division at a late
stage (Fig. 8). The cells formed short chains, as first reported by Maury and Devilliers (1999). 46
These short chains of cells were only apparent in tolA, tolQ, and tolR deletion mutants when cells
were grown at low osmolarity (no added NaCl). Furthermore, the short chains formed by tol mutants were distinct from the long, smooth filaments formed when TolQ was over-expressed.
These results confirm earlier reports of cell division impairment in tol mutants (Meury and
Devilliers, 1999; Heilpern and Waldor, 2000; Llamas et al., 2000; Dubuisson et al., 2005;
Gerding et al. 2007) and also indicate that the TolQ over-expression division impairment is distinct from the division impairment observed with tol mutants grown under select conditions.
It has been suggested that the CM-associated Tol complex consisting of TolA, TolQ, and
TolR are present in the cell at a precise stoichiometry (Guihard et al., 1994). From this it has been inferred that there is required a specific ratio of TolA:TolQ:TolR to carry out Tol- dependent functions. In order to determine if the TolQ over-expression filamentation phenotype is also somehow dependent upon the presence of a complete Tol CM-complex, TolQ was over- expressed in Tol deletion mutants RA1027 (ΔtolQ), RA1028 (ΔtolR), and RA1038 (ΔtolA). E. coli deletion mutants formed filaments comparable to wild type (W3110) when TolQ was over- expressed. This observation strongly sugests that filamentation is attributable to TolQ alone and that this phenotype occurs independent of a complete Tol protein complex, as TolA and TolR are not required for the phenotype (Fig. 9). Filamentation also occurs in cells lacking ExbB/ExbD
(Teleha, 2009). Therefore, it may be that this phenotype is not dependent upon Tol-derived energy or any link between Tol-derived energy and the cell division process is mediated through
TolQ.
The over-expression of E. coli TolQ in E. amnigenus, C. muytjensii and E. coli BL21 also caused cell filamentation (Fig. 10). Unlike W3110, E. coli strain BL21 is not K-12 derived, but like K-12 derivatives, including W3110, it is also a widely-used laboratory strain of E. coli. Both 47
W3110 and BL21 are defective in O-antigen polymerization and capsule formation, although as
a result of different mutations, which makes each a suitable lab strains that cannot colonize
animals (Studier et al., 2009). Like E.coli strains W3110 and BL21, E. amnigenus and C.
muytjensii are γ-proteobacteria. The genome of neither is available, but each was chosen for this
study because they both belong to the Enterobactericeae family into which E. coli is classified,
making them close relatives. That strain Bl21, E. amnigenus, and C. muytjensii each filament
when the E. coli TolQ protein is over-expressed suggests that the phenotype is not W3110 E.
coli specific and that whatever the role of TolQ, if any, in E. coli cell division, this role is
conserved among close relatives of E. coli.
It was previously observed that when E. coli cells carrying plasmid-encoded TolQ
expressed from the araBAD promoter (pRA031) were grown at the ʟ-arabinose concentration of
0.1% (w/v) they formed filaments (Teleha, 2009). Furthermore, when cells were grown at
increasing ʟ-arabinose concentrations, the degree of filamentation that resulted correlated
directly with ʟ-arabinose concentration (Teleha, 2009; Fig. 5). The last objective of aim one of
this dissertation was to have an antibody to TolQ produced commercially that could be used in
Western blot analyses to detect TolQ protein levels in E. coli cells grown over a range of ʟ-
arabinose levels. Western blot analysis indeed provided evidence that both the degree of
filamentation and the amount of TolQ in the cells were dependent upon ʟ-arabinose induction levels. As determined visually, cell elongation becomes readily apparent at an ʟ-arabinose concentration of 0.001% (w/v) (Fig. 11, panel D), the same concentration shown by Western blot analysis at which above-wild type levels of TolQ are first evident, and becomes increasingly pronounced at higher ʟ-arabinose-induced expression levels (Fig. 11, lane 6). 48
The objectives of Chapter II in this dissertation were designed to carry out the specific
aim of further characterizing the TolQ over-expression phenotype of cell filamentation.
Preliminary studies indicated that filamentation results from the over-expression of TolQ, but not
of TolR, or TolA (Teleha, 2009; Fig. 4). As tol mutants grown under conditions of low
osmolarity form short chains of cells, and tol mutants grown at both standard salt concentration
and in media lacking added NaCl fail to form filaments, it can be concluded that filamentation is not a tol mutant phenotype and instead results solely from the over-expression of TolQ (Fig. 8).
It might also be concluded that filamentation that results from TolQ over-expression is not
dependent on the presence of TolR or TolA, and likewise, Tol-derived energy, as filamentation
occurs in ΔtolA, ΔtolQ, and ΔtolR mutants when TolQ is over-expressed (Fig. 9), as well as in
ΔtolQ/ΔExbB/ΔExbD mutants (Teleha, 2009). Filamentation that results from the over- expression of TolQ in close relatives of E. coli (Fig. 10) provides evidence that the role that
TolQ might fulfill in bacterial cell division is conserved, as are the structure and many functions of the Tol-Pal complex. Lastly, Western blot analysis provides a direct link between TolQ
protein levels and the degree of filamentation observed in cells (Fig. 11). Based on the results of
studies carried out thus far, it appears that TolQ, when present at increased levels in the cell,
interrupts normal bacterial cell division. For this to occur, it is likely that excess TolQ interrupts
cell division by interacting with and sequestering a binding partner that likely has a direct role in
the cell division process. Chapter III of this dissertation addresses the identity of this binding partner and the region of TolQ that participates in the interaction. Additionally, the cell division phenotype associated with the manipulation of expression levels of both TolQ and its binding partner was analyzed to determine if the TolQ over-expression phenotype could be alleviated by the concurrent over-expression of the potential binding partner. 49
CHAPTER III
POTENTIAL INTERACTION BETWEEN TOLQ AND FTSN AND THE IMPACT OF
THEIR DUAL OVER-EXPRESSION ON DIVISION PHENOTYPE
Introduction
FtsN was first identified as a multicopy suppressor of a temperature-sensitive ftsA mutation. FtsN was determined to be an essential E. coli cell division protein when mutations in ftsN led to cell filamentation and death (Dai et al., 1993). As a result of ftsN disruption, long, smooth filaments containing few to no constrictions are formed. That nucleoids appear to be regularly spaced indicates that in the absence of FtsN, DNA synthesis and nucleoid segregation is not affected (Dai et al., 1993). Since its identification, studies of FtsN have led to the determination of its topology in the CM (Dai et al., 1996) and its late timing in joining the septal ring (Addinall et al., 1997). More recent studies have identified protein interactions between
FtsN and a number of other divisome components (Goehring et al., 2006) and suggested functional roles for FtsN in the bacterial cell division process (Geissler and Margolin, 2005;
Goehring et al., 2007; Gerding et al., 2009; Lutkenhaus, 2009b; Rico et al., 2010). Gerding et al. (2007) found that Tol mutants experience delayed OM invagination and contain outer membrane blebs at constriction sites and cell poles. In that study, chimeras consisting of each of the Tol proteins translationally fused with green or red fluorescent protein (GFP or RFP) appeared to localize to constriction sites during cell division. This localization did not occur when any of the Tol-GFP fusion proteins were expressed in cells depleted of the divisome protein FtsN. These data led the authors to suggest a model in which Tol system components are recruited by FtsN to cell constriction sites where they then couple the OM to the divisome to coordinate division of the outer membrane with septum formation and cell division (Gerding et 50
al., 2007). In this study, it was observed that only TolA and TolQ localize to constriction sites independent of each other and of other Tol proteins, indicating that each is recruited by an unnamed division protein. For recruitment to the division site, TolA appears to require both FtsZ polymerization and FtsN activity, TolQ appears to require active FtsN, and no Tol proteins accumulate at the division site in the absence of FtsN (Gerding et al., 2007). Because FtsN joins the divisome late in the division process, its recruitment, as well as the recruitment of the Tol-Pal proteins could be dependent upon a number of early divisome proteins. However, given that
TolQ over-expression produces a phenotype that strikingly resembles that of ftsN mutants (Dai et al., 1996), FtsN was identified as a potential binding partner for TolQ in this study. It was hypothesized that a direct protein-protein interaction between TolQ and FtsN leads to cell filamentation when TolQ is present in excess. To test this hypothesis, a protein-fragment complementation assay was used to test potential non-CM localized interactions between TolQ and FtsN. Upon completion of this assay, the phenotype associated with this protein-protein interaction was explored further through dual-over-expression studies.
Yeast two-hybrid screening systems are useful in the identification of protein-protein interactions. These systems are protein-fragment complementation assays that rely upon the activation of a reporter gene by the binding of a transcription activator to an upstream activating sequence. In the two-hybrid system, the transcription factor is divided, with each fragment encoded on a separate plasmid. Plasmids are constructed to encode proteins or protein fragments of interest, fused to the two transcription factor fragments. One of these proteins, referred to as the “bait” protein, is usually a known protein for which one wishes to identify new binding partners. The “target” is either a selected protein that is suspected of interacting with the bait protein or it may be from a cDNA library, representing a collection of proteins from an 51
organism. An interaction between the bait and target proteins brings together the fragments of the transcription factor and the reporter gene is transcribed. Since first developed by Fields and
Song (1989), the yeast two-hybrid screening system has served as a model for the development of other two-hybrid screening systems that are used in a variety of applications. The bacterial two-hybrid system is one such screening method for detecting protein-protein interactions in vivo in E. coli. The Bacteriomatch II ® two-hybrid system (Stratagene) was used in this dissertation to test the hypothesis that TolQ interacts with the essential divisome protein FtsN. The
BacterioMatch II ® two-hybrid system is E. coli-based, making it a suitable system for screening
for in vivo protein-protein interactions in E. coli. Because E. coli grows faster than and is
transformed with a higher efficiency than yeast, and because DNA (plasmid) collection is easier
than DNA collection from yeast, screening can be accomplished relatively quickly and easily. In
this use of the BacterioMatch II ® two-hybrid system, selected regions of the TolQ protein
served as the bait protein, while selected regions of FtsN served as the target protein. In this
study, the bait protein was fused to the full-length bacteriophage λ repressor protein, λcI,
encoded in the pBT plasmid. This protein contains the N-terminal DNA-binding domain and the
C-terminal dimerization domain of λcI. The target protein is fused to the N-terminal domain of
the α-subunit of RNA polymerase, encoded in the pTRG plasmid. The pBT plasmid carries the
cat gene for chloramphenicol resistance, while the pTRG plasmid carries the tetA gene for
tetracycline resistance, with each used for transformant selection and both used for dual-
transformant selection. Relevant features of pBT and pTRG are shown in Figs. 12 and 13,
respectively. Expression of gene sequences in the pBT and pTRG plasmids is regulated by the
lac-UV5 promoter, which is induced by isopropylthio-β-galactoside (IPTG). Therefore, the
genes encoding the fusion bait and target proteins are transcribed upon the addition of IPTG. 52
Bait and target constructs are screened in the BacterioMatch II ® Screening Reporter Competent
strain, which carries the reporter gene cassette (HIS3) on an F´ episome. The reporter strain
carries lacIq, a modified lac promoter that contains a single λ operator at position -62. The
modified lac promoter is not inducible by IPTG, but is instead induced by λcI, ensuring that the
addition of IPTG, while inducing expression of the bait and target fusions, does not induce the
two reporter genes. DNA binding of λcI to the λ operator on DNA tethers the bait protein close
to the operator. Interaction between the bait and the target proteins recruits and stabilizes RNA
polymerase at the promoter, which activates transcription of the HIS3 reporter gene (Fig. 14).
HIS3 encodes a component of the histidine biosynthetic pathway that complements a hisB
mutation engineered into the reporter strain. Low basal levels of HIS3 transcription allow the
reporter strain to grow on minimal media lacking histidine. The compound 3-amino-1,2,4-
triazole (3-AT) is a competitive inhibitor of the HIS3 gene product. The reporter strain is unable
to grow on media lacking histidine if 3-AT is present. When a positive protein-protein
interaction occurs and transcriptional activation increases expression, thus increasing the amount
of the HIS3 gene product, growth of the reporter strain on media lacking histidine and containing
5mM 3-AT occurs. Only when protein-protein interaction is present and transcriptional activation occurs is enough of the HIS3 gene product present to overcome competitive inhibition by 3-AT. In addition to the HIS3 gene, a second reporter gene, aadA, conferring streptomycin resistance, is also present in the reporter cassette to confirm the bait and target interaction.
Positive protein-protein interaction allows the reporter strain to grow on selective screening
media containing 3-AT, tetracycline, and chloramphenicol. Interaction of the bait and target
proteins is validated by growth on dual-selective medium containing 3-AT, tetracycline,
chloramphenicol, and streptomycin. Positive and negative controls are used to test the media, 53
experimental design, and to confirm results. Transformants of two plasmids encoding proteins
known to interact in E. coli are used as a positive control: pBT-LGF2, encoding the dimerization
domain of the GAL4 protein fused to the λcI repressor and pTRG-GAL11P, encoding a domain
of a mutant Gal 11 protein fused to the α-subunit of RNA polymerase (Fig. 15). Transformants of pBT and pTRG vectors into which no bait or target proteins have been cloned serve as the negative control. Using the membrane topologies of TolQ and FtsN to identify possible protein- protein interaction domains, sequences encoding TolQ fragments were cloned into the pBT plasmid. Similarly, sequences encoding regions of the FtsN protein were cloned into the pTRG plasmid. These plasmids were dually-transformed into the BacterioMatch II ® Validation
Reporter Competent cells and plated on non-selective media containing chloramphenicol and
tetracycline, selective media (non-selective media with 3-AT added), and dual selective media
(selective media with streptomycin added). Dual transformants with pBT-LGF2 and pTRG-
Gal11P and with pBT and pTRG with no fusions cloned served as positive and negative controls
for each protein fragment pair tested. The growth of dual-transformants was analyzed as
described in the BacterioMatch II ® handbook (Stratagene, 2003).
54
Figure 12. pBT plasmid vector. Diagram of the pBT plasmid showing relevant features and their locations. The p15A origin of replication is located from position 581-1493. The chloramphenicol resistance gene is located from position 2770-219. lac-UV5 promoter is located from position 1556-1583. λ-cI operator is located from position 1631-2341. The multiple cloning site region (sequence shown in diagram), with restriction sites labeled, is located from position 2342-2394 (Diagram from Stratagene).
55
Figure 13. pTRG plasmid vector. Diagram of the pTRG plasmid, showing features and their locations. The ColE1 origin of replication is located from position 1243-2475. The tetracycline resistance gene is located from position 3120-4310. The lpp promoter is located from position 47-76 and the lac-UV5 promoter is located from position 119-148. The RNA polymerase-α open reading frame is located from position 243-992. The multiple cloning site region (sequence shown in diagram), with restriction sites labeled, is located from position 978-1065 (Diagram from Stratagene).
56
Figure 14. Diagram of transcriptional activation of the HIS3 and aadA (Strr) genes. The bait protein is fused to λ-cI repressor protein gene. The target protein is fused to the N- terminal domain of the α-subunit of RNA polymerase. Interaction between bait and target proteins recruits and stabilizes the binding of RNA polymerase at the promoter, resulting in transcription of the HIS3 reporter gene and growth on selective media lacking histidine and containing 3-AT. Transcriptional activation that results from bait and target interaction also allows the reporter strain to grow on media containing streptomycin, validating the protein-protein interaction. The modified lac promoter, engineered into the reporter strain contains the λ-operator, centered at position -62 and is not inducible by IPTG. Transcription of the reporter genes only occurs through the binding of λ-cI to the λ operator. The reporter strain used for two-hybrid analyses was BacterioMatch ® II Validation Reporter Competent Cells. (Diagram from Stratagene).
57
Figure 15. BacterioMatch® II positive control plasmids. pBT-LGF2 and pTRG-GAL11P encode fusions that have been shown to interact in E. coli. pBT-LGF2 encodes the dimerization domain of the Gal4 transcriptional activator protein fused to the λ-cI repressor. pTRG-GAL11P encodes a 90 amino acid domain of the mutant form of the GalII protein fused to the α-subunit of RNA polymerase. When cotransformed into BacterioMatch ® II two-hybrid system reporter strain competent cells, the interaction of the Gal4 and Gal11 domains allows cell growth on media lacking histidine supplemented with 5mM 3-AT and streptomycin (Diagram from Stratagene).
58
Methods
Media
Bacterial strains were maintained on Luria-Bertani (LB) agar (Miller, 1972). Plasmid-
bearing strains were maintained on LB agar supplemented with 100 μg ml-1 ampicillin, 30 μg
ml-1 kanamycin, and/or 34 μg ml-1 chloramphenicol, as necessary. For all other assays cells were grown in standard LB broth, supplemented with antibiotics as necessary and ʟ-arabinose as indicated. Two hybrid analyses were performed on a variety of M9-based minimal selective media prepared as described in the BacterioMatch Two-Hybrid Instruction Manual (Stratagene),
with no alterations. These included: His-dropout Broth (1 X M9 minimal salts containing 1X
His-dropout supplement amino acids (Clontech) Laboratories, Inc. Mountain View, CA), 0.4%
(w/v) glucose, 200 μM adenine HCl, 1 mM MgSO4, 1 mM Thiamine HCl, 10 µM ZnSO4, 100
μM CaCl2 and 50 μM IPTG); Nonselective agar (His-dropout Broth, 1.5% (w/v) agar,
chloramphenicol 25 μg ml-1 and tetracycline 12.5 μg ml-1); Selective screening agar
(nonselective agar with 5 mM 3-amino-1,2,4 triazole[3-AT]); and Dual selective screening agar
(selective screening agar with 12.5 μg ml-1 streptomycin). As indicated for cloning, SOC
medium (2% (w/v) tryptone, 0.5% yeast extract, 0.05% NaCl, 20 mM glucose, 10 mM MgCl2,
10 mM MgSO4) was used.
Bacterial Strains
Bacterial strains used in this study are summarized in Table 2. The Escherichia coli K-12
strain W3110 (Hill and Harnish, 1981) was used in this study as the wild type. For cloning, 5-
alpha F´ Iq E. coli (NEB) and Turbo E. coli (NEB) competent cells were used. For two-hybrid analysis the BacterioMatch II ® two hybrid system reporter strain (Δ(mcrA)183 Δ(mcrCB- 59
hsdSMR-mrr)173 endA1 hisB supE44 thi-1 recA1gyrA96 relA1 lac [F´ lacIq HIS3 aadA Kmr]) of
E. coli K-12 was purchased (Stratagene).
Plasmids
Plasmids used in this study are summarized in Table 2 and primers used for the
construction of these plasmids are described in Table 3. For two-hybrid analysis, regions
encoding the bait domains of TolQ were amplified by polymerase chain reaction (PCR) and
cloned into the plasmid pBT to generate in-frame fusions with a gene encoding the λcI protein.
Four such fusions were constructed; TolQ codons 1-19 (pMT005), 39-135 (pMT006), 157-174
(pMT007), and 194-230 (pMT008). Similarly, regions encoding target domains of FtsN were amplified and cloned into the plasmid pTRG to generate in-frame fusions with a gene encoding the α-subuint of RNA polymerase. Three such fusions were generated; FtsN codons 1-33
(pMT009), 54-234 (pMT010), and 54-319 (pMT010). For dual over-expression analysis, tolQ was cloned into the vector pBAD18-Cm (Guzman et al., 1995; Fig. 17), while ftsN was cloned into both the vector pBAD18-Kan (Guzman et al., 1995; Fig. 17) and the vector pPro18-Kan
(Lee and Keasling, 2005; Fig. 18). The tolQ and ftsN genes were over-expressed in two separate systems. First, tolQ and ftsN were concurrently over-expressed in W3110 from pBAD18-Cm
and pBAD18-Kan, respectively, at an ʟ-arabinose concentration of 0.1% (w/v). Cells were stained and viewed to observe any reduction in filamentation of cells compared to those that over-expressed tolQ alone. The assay was repeated with the over-expression of tolQ in pBAD18-Cm and ftsN in pPro18-Kan to confirm results using two distinct induction methods.
60
Construction of two hybrid plasmids
To construct pMT005 and pMT008, PCR was carried out using forward primer oMT001
and reverse primer oMT002 to amplify TolQ1-19 and forward primer oMT003 and reverse primer
oMT004 to amplify TolQ194-230 from a W3110 template. PCR products were confirmed on a 1%
(w/v) agarose gel, cleaned using the QIAquick PCR Purification kit (QIAGEN), and restriction
digested with EcoRI (NEB) and NotI (NEB) for 120 min at 37° C. pBT was also digested with
EcoRI and NotI for 120 min at 37°C. Digests were then purified using the QIAquick PCR
Purification kit (QIAGEN) and eluted in 50 µl buffer (10 mM Tris-Cl, pH 8.5). TolQ1-19 and
TolQ194-230 fragments were ligated into pBT using T4 DNA ligase (NEB) in an overnight incubation at room temperature. Ligation mixtures were transformed as 5 µl aliquots into NEB
5-alpha F´ Iq E. coli competent cells according to protocol (NEB). After outgrowth of
transformation cultures for 60 min at 37°C with shaking, 50 µl each transformation was plated
on LB supplemented with ampicillin at 100 μg ml-1 and incubated overnight at 37°C. From each transformations plate, five colonies were grown overnight in LB with ampicillin at 100 μg ml-1
before plasmids were collected using the QIAquick Spin Miniprep kit (QIAGEN). Plasmids
were screened by linearlization with BamHI (NEB). A second screening was conducted using
EcoRI and NotI to restrict the insert back out of the construct, and final screenings with BmtI
(NEB) for pMT005 and with BsgI (NEB) for pMT008 were carried out to confirm the constructs.
Digests were run on 1% (w/v) agarose gels to visualize restriction products. All five
transformants screened for each cloning appeared to contain the correct size plasmid and release
the correct insert. Two plasmids of each were chosen for sequence analysis, which confirmed
correct plasmid construction. One sequence-confirmed construct of TolQ1-19 in pBT was named 61
pMT005 and one sequence-confirmed construct of TolQ194-230 in pBT was named pMT008.
pMT005 and pMT008 were then used in two-hybrid analyses.
pMT006 and pMT007 were similarly generated by PCR amplification of TolQ39-135 and
TolQ157-174 and cloning into pBT. Using W3110 as a template, TolQ39-135 was amplified with
forward primer oMT005 and reverse primer oMT006 and TolQ157-174 was amplified using forward primer oMT007 and reverse primer oMT008. PCR products were confirmed on a 1%
(w/v) agarose gel, cleaned using the QIAquick PCR Purification kit and restriction digested with
EcoRI and NotI for 120 min at 37°C. pBT was also digested with EcoRI and NotI at 37°C for
120 min. Digests were purified using the QIAquick PCR Purification kit and eluted in 50µl
buffer (10 mM Tris-Cl, pH 8.5). Following purification, the TolQ39-135 and TolQ157-174 fragments
were ligated into pBT using T4 DNA ligase in an overnight incubation at room temperature.
Ligation mixtures were transformed as 5 µl aliquots into 5-alpha F´ Iq E. coli competent cells according to protocol (NEB). After outgrowth of transformation cultures for 60 min at 37°C with shaking, 100 µl aliquots each transformation was plated on LB supplemented with ampicillin at 100 μg ml-1 and incubated overnight at 37°C. Seven colonies of each transformation plate were grown overnight in LB with ampicillin at 100 μg ml-1 before plasmids were collected using the QIAquick Spin Miniprep kit. TolQ39-135 in pBT plasmids were screened by restriction digest at 37°C for 90 min for linearlization with XhoI (NEB). TolQ157-174 in pBT
plasmids were screened by restriction digest with HindIII (NEB) and XhoI for 90 min at 37°C.
Digests were run on 1% (w/v) agarose gels to visualize restriction products. Five of the seven
transformants screened for each cloning appeared to contain the correct size plasmid and/or
release the expected insert after digestion. Two plasmids of each were chosen for sequence
analysis, which confirmed correct plasmid construction. One sequence-confirmed construct of 62
TolQ39-135 in pBT was named pMT006 and one sequence-confirmed construct of TolQ157-174 was named pMT007, and both were then used in two-hybrid analyses.
To construct pMT010 and pMT011, PCR was carried out using forward primer oMT009 and reverse primer oMT011 to amplify FtsN54-243 and forward primer oMT009 and reverse primer oMT010 to amplify FtsN54-319 with W3110 as a template. PCR products were confirmed
on 1% (w/v) agarose gel, cleaned using the QIAquick PCR Purification kit and restriction
digested with BamHI and XhoI for 90 min at 37° C, as was pTRG. Digests were then purified
using the QIAquick PCR Purification kit and eluted in 30 µl buffer (10 mM Tris-Cl, pH 8.5).
The FtsN54-243 and FtsN54-319 fragments were ligated into pTRG using T4 DNA ligase in an
overnight incubation at room temperature. Ligation mixtures were transformed as 5 µl aliquots
into NEB Turbo E. coli competent cells according to protocol (NEB). After outgrowth of cultures for 60 min at 37°C with shaking, 50 µl each transformation was plated on LB supplemented with tetracycline at 20 μg ml-1 and incubated overnight at 37°C. The one transformant from the FtsN54-319 into pTRG ligation and five transformants from the FtsN54-243
into pTRG transformation plates were grown overnight in LB with tetracycline at 20 μg ml-1 for screening. Plasmids were collected using the QIAquick Spin Miniprep kit and screened by restriction with XbaI (NEB). The single FtsN54-319 into pTRG and three of five FtsN54-243 into
pTRG constructs appeared to contain the correct size insert. The FtsN54-319 into pTRG construct
and two FtsN54-243 into pTRG constructs were confirmed by sequencing. FtsN54-319 in pTRG was named pMT011, while one FtsN54-243 into pTRG construct was named pMT010. pMT011 and pMT010 were then used for two-hybrid analysis.
FtsN1-33 was amplified by PCR from W3110 to be cloned into pTRG using forward
primer oMT012 and reverse primer oMT013. The PCR product was confirmed on 1% (w/v) 63
agarose gel, cleaned using the QIAquick PCR Purification kit and restriction digested with
BamHI and XhoI for 90 min at 37°C, as was pTRG. The digest was then purified using the
QIAquick PCR Purification kit from QIAGEN and eluted in 30 µl buffer (10 mM Tris-Cl, pH
8.5). The FtsN1-33 fragment was ligated into pTRG using T4 DNA ligase in an overnight
incubation at room temperature. Following incubation, 5 µl ligation mixture was transformed
into NEB Turbo E. coli competent cells according to protocol (NEB). After outgrowth of transformation culture for 60 min at 37°C with shaking, 50 µl was plated on LB supplemented with tetracycline at 20 μg ml-1 and incubated overnight at 37° C. Five transformants from the
FtsN1-33 into pTRG ligation were grown overnight in LB with tetracycline at 20 μg ml-1 for
screening. Plasmids were collected using the QIAquick Spin Miniprep kit and screened by
restriction with BamHI and with XbaI. The five FtsN1-33 into pTRG constructs appeared to contain the correct size insert and two of the five constructs were confirmed by sequencing. One of the FtsN1-33 in pTRG constructs was named pMT009 and then used in two-hybrid analysis.
64
Table 2. Strains and plasmids used for two-hybrid screening and dual over-expression in Chapter III
Strain Relevant Characteristics Source/Reference
W3110 F- IN(rrnD-rrnE)1 Hill & Harnish, 1981 BacterioMatch II hisB lac [F´ lacIq HIS3 aadA Kmr] Stratagene reporter 5-alpha F´ Iq E. F´ proA+B+ lacIq ∆(lacZ)M15 zzf::Tn10 NEB coli (TetR) / fhuA2∆(argF-lacZ)U169 phoA glnV44 Φ80Δ(lacZ)M15 gyrA96 recA1 relA1 endA1 thi-1 hsdR17 Turbo E. coli F´ proA+B+ lacIq ∆ lacZ M15/ fhuA2 ∆(lac- NEB proAB) glnV gal R(zgb-210::Tn10)TetS endA1 thi-1 ∆(hsdS-mcrB)5 Plasmid Relevant Characteristics Source/Reference
pBT lambda cI fusion site, cmr Stratagene pBT-LGF2 lambda cI-LGF2 fusion, cmr Stratagene pTRG RNAP-α fusion site, tetr Stratagene pTRG-Gal 11P RNAPα-Gal 11P fusion, tetr Stratagene pMT005 pBT lambda cI – TolQ (1-19) Present study pMT006 pBT lambda cI – TolQ (39-135) Present study pMT007 pBT lambda cI – TolQ (157-174) Present study pMT008 pBT lambda cI – TolQ (194-230) Present study pMT009 pTRG RNAP-α – FtsN (1-33) Present study pMT010 pTRG RNAP-α – FtsN (54-243) Present study pMT011 pTRG RNAP-α – FtsN (54-319) Present study BacterioMatch II hisB lab [F´ lacIq HIS3 aadA Kmr] Stratagene Validation Reporter pBAD24 araBAD promoter, AraC, ampr Guzman et al. (1995) pBAD18-Cm araBAD promoter, AraC, cmr Guzman et al. (1995) pBAD18-Kan araBAD promoter, AraC, Kmr Guzman et al. (1995) pPRO18-Kan prpBCDE promoter, Kmr Lee & Keasling (2005) pRA031 pBAD24 encoding TolQ Teleha (2009) pMT001 pBAD24 encoding FtsN Present study pMT002 pBAD18-Cm encoding TolQ Present study pMT003 pBAD18-Kan encoding FtsN Present study pMT004 pPRO18-Kan encoding FtsN Present study
Strains, relevant genotypes, and source shown in Table 2. Transformants of pBT fusions were selected in NEB 5-alpha F´ Iq competent E. coli cells. Transformants of pTRG fusions were selected in NEB Turbo competent E. coli cells. NEB: purchased from New England Biolabs.
65
Figure 16. Bacteriomatch II ® Two-hybrid Strategy. A. Diagram of pBT and pTRG vectors and the location of inserts into pBT (TolQ) and pTRG (FtsN) (Diagram from Stratagene). B. Membrane topologies of FtsN and TolQ. FtsN1-33 comprises the amino terminus and is located in the cytoplasm. The transmembrane domain of FtsN34-53 precedes a largely unstructured region, FtsN58-242, containing three partially formed helices. FtsN243-319 comprises the fully structured carboxyl terminus of FtsN. TolQ spans the CM three times, TolQ19-35, TolQ139-156, and TolQ171-186. The amino terminus of TolQ, TolQ1-18 resides in the periplasm, while the carboxyl terminus, TolQ187-231 resides in the cytoplasm. TolQ36-138 forms a large cytoplasmic loop and TolQ157-170 forms a small periplasmic loop (Redrawn from Yang et al., 2004 (FtsN) and Goemaere et al., 2007 (TolQ)). C. BacterioMatch® II Two Hybrid-System screening. Summary of the TolQ and the FtsN domains of interest that have been cloned into bait and target plasmids. TolQ 1-19, TolQ 39-135, TolQ 157-174, and TolQ 194-230 were cloned into pBT, the bait plasmid. FtsN 1-33, FtsN 54-243, and FtsN 54-319 were cloned into pTRG, the target plasmid. Experimental pairs are those that were tested together in dual transformants to identify potential protein-protein interactions. 66
Table 3. Primers used for PCR amplification for two-hybrid analysis and dual over- expression studies in Chapter III.
Protein Restriction Primer Sequence Fragment Vector Site
oMT001 5'-GAGCGGCCGCAGTGACTGACATG-3' TolQ(1-19) pBT NotI Forward oMT002 5'-GATGAATTCAACCAGAAGGCTAGCCTTCAG-3' TolQ(1-19) pBT EcoRI Reverse oMT003 5'-CAGCGGCCGCCTACAACCGCCTC-3' TolQ(194- pBT NotI 230) Forward oMT004 5'-TGGAATTCCCCTTGTTGCTCTCGCT-3' TolQ(194- pBT EcoRI 130) Reverse oMT005 5'-CTGCGGCCGCTATTATCCAGCGGACCCG-3' TolQ(39-135) pBT NotI Forward oMT006 5'-CCGAGGAACGAATTCTGCGTTTCCAG-3' TolQ(39-135) pBT EcoRI Reverse oMT007 5'-GCCCTCGCGGCCGCAAAACAAGCA-3' TolQ(157- pBT NotI 174) Forward oMT008 5'-GCAGTCGAATTCAACGCTTCTGCG-3' TolQ(157- pBT EcoRI 174) Reverse oMT009 5'-CGGTGGTGGATCCTTCATTACGC-3' FtsN(54→) pTRG BamHI Forward oMT010 5'-ACCCCCGGCCTCGAGCGAATGCA-3' FtsN(54-319) pTRG XhoI Reverse oMT011 5'-TTTCTCGAGCGTCGGTTTTGGCGC-3' FtsN(54-243) pTRG XhoI Reverse oMT012 5'-GCGGATCCATAGTGGCACAACGAG-3' FtsN(1-33) pTRG BamHI Forward oMT013 5'-CATCTCGAGAGAAACCGCAGGCAG-3' FtsN(1-33) pTRG XhoI Reverse oMT014 5'-GCACCATGGCACAACGAGA-3' FtsN pBAD24 NcoI Forward oMT015 5-GGGATCTAGAGGGTTTCAA-3' FtsN pBAD24 XbaI Reverse
Listed for each primer are the primer sequence, the gene or gene region cloned and the primer direction, the vector each PCR product was cloned into, and the restriction site generated by each primer. Constructs were confirmed by sequence analysis using pBT and pTRG primers recommended by Stratagene ® or universal pBAD primers (Eurofins MWG Operon).
67
Construction of dual over-expression plasmids
Over-expression of the tolQ and ftsN proteins was achieved using plasmids derived from
ʟ-arabinose-regulated plasmids (Guzman et al., 1995) and from a propionate-regulated plasmid
(Lee and Keasling, 2005). To clone ftsN into pBAD18-Kan (Fig. 19) and pPro18-Kan (Fig. 20),
ftsN was first amplified by PCR (from W3110) using forward primer oMT014 and reverse
primer oMT015. The PCR product was confirmed on a 1% (w/v) agarose gel, cleaned using the
QIAquick PCR Purification kit and restriction digested with NcoI and XbaI for 90 min at 37°C,
as was pBAD24 (Fig. 17). The digests were purified using the QIAquick PCR Purification kit
and eluted in 30 µl buffer (10 mM Tris-Cl, pH 8.5). The ftsN amplimer was ligated into
pBAD24 using T4 DNA ligase in an overnight incubation at room temperature. Following
ligation, 5 µl mixture was transformed into NEB 5-alpha F´ Iq E. coli competent cells according to protocol (NEB). Following outgrowth of transformation culture for 60 min at 37°C with shaking, 50 µl aliquots were plated on LB supplemented with ampicillin at 100 μg ml-1 and
incubated overnight at 37°C. Eight transformants were grown overnight in LB with ampicillin at
100 μg ml-1 for screening. Plasmids were collected using the QIAquick Spin Miniprep kit and
screened by restriction with BsgI. Digests were visualized on a 1% (w/v) agarose gel. The ftsN
into pBAD24 constructs all appeared to contain the correct size insert (960 bp) and one was
chosen to serve as the source of ftsN for ligation into pBAD18-Kan and pPro18-Kan. This
construct was confirmed by sequencing and named pMT001. The ftsN gene was first cloned into
pBAD24 before being cloned into pBAD18-Kan and pPro18-Kan because the optimized
ribosome binding (RBS) site (SD) engineered into pBAD24 is not present in pBAD18-Kan or
pPro18-Kan. To achieve high tolQ and ftsN expression levels, the RBS upstream of ftsN in
pBAD24 was cloned into pBAD18-Kan and pPro18-Kan with ftsN. To clone ftsN into pBAD18- 68
Kan, pMT001 was restriction digested with BamHI and SphI (NEB) for 90 min at 37°C. To
isolate the RBS-ftsN fragment to be cloned, the digest was run on a 1% (w/v) agarose gel and the
band containing the correct size fragment was gel purified using the QIAquick Gel Extraction
Kit. pBAD18-Kan was also digested with BamHI and SphI and was then purified using the
QIAquick PCR Purification Kit. The fragment was ligated into pBAD18-Kan using T4 DNA
ligase in an overnight incubation at room temperature. Following ligation 5 µl mixture was
transformed into 5-alpha F´ Iq E. coli competent cells according to protocol (NEB). After
outgrowth of transformation culture for 60 min at 37°C with shaking, 200 µl was plated on LB
supplemented with kanamycin at 30 μg ml-1 and incubated overnight at 37°C. Eight
transformants were grown overnight in LB kanamycin 30 μg ml-1 for screening. Plasmids were
collected using the QIAquick Spin Miniprep kit and screened by restriction with BsgI. Digests
were visualized on a 1% (w/v) agarose gel. Four of the eight constructs appeared to contain the
correct size fragment. Two of the four ftsN in pBAD18-Kan constructs were confirmed by sequencing and one was named pMT003 to be used in dual over-expression analyses. To clone
FtsN into pPro18-Kan, pMT001 and pPro18-Kan were restriction digested with SphI and NheI
for 90 min at 37°C. The pPro18-Kan digest was purified using the QIAquick PCR Purification
Kit. The RBS-ftsN fragment restricted from pMT001 was purified from a 1% (w/v) agarose gel
using the the QIAquick Gel Extraction Kit. The fragment was ligated into pPro18-Kan using T4
DNA ligase in an overnight incubation at room temperature. An aliquot of 5 µl ligation mixture
was transformed into NEB 5-alpha F´ Iq E. coli competent cells according to protocol (NEB).
After outgrowth of transformation culture for 60 min at 37°C with shaking, 100 µl was plated on
LB supplemented with kanamycin at 30 μg ml-1 and incubated overnight at 37°C. Eight
transformants were grown overnight in LB with kanamycin at 30 μg ml-1 for screening. 69
Plasmids were collected using the QIAquick Spin Miniprep kit and screened by restriction with
SphI and then with NheI and SphI. Digests were visualized on a 1% (w/v) agarose gel. Each of
the 8 constructs appeared to contain the correct size fragment. Two of the four ftsN in pPro18-
Kan constructs were confirmed by sequencing and one was named pMT004 to be used in dual
over-expression analyses. To clone the pBAD24 optomized RBS and tolQ into pBAD18-Cm
(Fig. 18), pRA031 and pBAD18-Cm were first restriction digested with BamHI and SphI for 120
min at 37°C. Digested pRA031 was run on a 1% (w/v) agarose gel and the fragment encoding
RBS-tolQ was gel purified using the QIAquick Gel Extraction Kit. The pBAD18-Cm BamHI
and SphI digest was purified using the QIAquick PCR Purification Kit. The purified fragment
was ligated into pBAD18-Cm using T4 DNA ligase in an overnight incubation at room
temperature. An aliquot of 5 µl ligation mixture was transformed into NEB 5-alpha F´ Iq E. coli competent cells according to protocol (NEB). After outgrowth of transformation culture for 60 min at 37°C with shaking, 75 µl was plated on LB supplemented with chloramphenicol at 34 μg ml-1 and incubated overnight at 37°C. Eight transformants were grown overnight in LB with chloramphenicol at 34 μg ml-1 for screening. Plasmids were collected using the QIAquick Spin
Miniprep kit and screened by restriction with KpnI (NEB). Digests were visualized on a 1%
(w/v) agarose gel. Each of the eight constructs was not digested with KpnI, as expected,
indicating the insert was present. Transformants were then screened by their ability to cause cell
filamentation when cells were grown in LB with chloramphenicol at 34 μg ml-1 supplemented
with 0.1% (w/v) ʟ-arabinose. Two of the eight tolQ in pPAD18-Cm constructs were confirmed
by sequencing and one was named pMT002 to be used in dual over-expression analyses.
70
BacterioMatch II ® Two Hybrid analysis
Before cotransformation and screening of recombinant pBT and pTRG plasmid pairs, the
empty pBT and the pTRG-Gal 11P, a non-interacting pair, were cotransformed into the reporter
strain as the negative control to verify media quality. As a positive control, the known
interactive pair pBT-LGF2 and pTRG-Gal 11P were cotransformed into the reporter strain.
Additionally, all recombinant pBT constructs were cotransformed with the empty pTRG vector
and recombinant pTRG constructs were cotransformed with the empty pBT vector to confirm
bait and target protein fragments do not activate the reporter cassette on their own. This control
tested the suitability of the chosen bait and target inserts for protein-protein interaction analyses
using the BacterioMatch system. Sequence-confirmed constructs were cotransformed into the
BacterioMatch II ® Validation Reporter Competent Cells as described in the BacterioMatch II ®
Instruction Manual (Stratagene). Plasmid pairs were chosen as depicted in Fig. 16, C. For dual transformation, BacterioMatch II ® validation reporter competent cells were thawed on ice and gently mixed. Thawed competent cells (100 µl aliquots) were transferred to chilled 14 ml BD
Falcon® tubes and 1.7 µl β-mercaptoethanol (β-ME) was added to each tube. Following a ten
min incubation on ice, 50 µl each pBT and pTRG plasmid were added to each tube and mixed,
and tubes were further incubated on ice for 30 min. The mixture was then heat-shocked for 35 s
at 42°C and returned to ice for 2 min before 900 µl SOC (prewarmed to 42°C) was added to
tubes, which were then incubated at 37°C with shaking at 225 rpm for 90 min. Cells were
pelleted by a 10 min centrifugation at 2000 x g, and resuspended in 1 ml His-dropout broth after
supernatant was aspirated. Cells were then collected a second time after another 10 min
centrifugation at 2000 x g and resuspended in 1 ml fresh His-dropout broth. Cultures were
incubated with shaking at 225 rpm for 2 h before plating. Cells were plated according to 71
recommendations (Stratagene). The negative control (pBT and pTRG-Gal 11P) and self- activation tests were plated as 200 µl aliquots on selective screening media (5 mM 3-AT).
Aliquots of 20 µl and 200 µl of a 1:100 dilution of cells were plated on non-selective media (no
3-AT). Plates were incubated for 24 h at 37°C and then transferred to the dark at room temperature for an additional 24 h. The number of colonies on each plate were counted and compared to guidelines from the BacterioMatch manual to determine suitability of constructs for protein-protein interaction analyses.
To identify possible protein-protein interactions between bait and target pairs,
BacterioMatch II ® Validation Reporter Competent Cells were dual-transformed with protein- protein interaction pairs according to protocol (Stratagene), as described above. Reporter cells were also cotransformed with positive and negative control pairs for each recombinant plasmid
pair to be tested. For plating of dual-transformants of bait and target fusions, 100 µl each
cotransformation mixture was plated on both nonselective screening medium and selective
screening medium. All pairs were plated in triplicate. Plates were incubated at 37°C for 24 h
and then transferred to room temperature and kept in the dark for an additional 24 h to allow for
growth of cells containing fusions of weak interactors or toxic proteins. Growth of recombinant
pairs on selective screening medium indicated protein-protien interaction, as was observed with
the positive control (pBT-LGF2 and pTRG-Gal11P). Growth on non-selective media confirmed
stable cotransformation. Putative positive colonies were patched on LB plates supplemented
with chloramphenicol at 25 μg ml-1 and tetracycline at 12.5 μg ml-1 to be used in further analyses.
To verify positive protein-protein interactions, cells were patched from selective screening medium (5 mM 3-AT) onto dual selective screening medium (5 mM 3-AT plus streptomycin at
12.5 μg ml-1). Activation of the second reporter gene, aadA, conferring streptomycin resistance 72
was used to confirm putative positive interactions identified by growth on selective screening
medium. Growth of putative positives on dual selective media indicated activation of the second reporter gene and verified the positive interactions. Cotransformants were also verified by spot plating on dual selective screening media. Putative positives were taken from LB plates supplemented with chloramphenicol at 25 μg ml-1 and tetracycline at 12.5 μg ml-1 and grown overnight in minimal media to adapt the cells from growth on rich media to growth in minimal media. Cultures were then diluted 50-fold and plated as 5 µl spots on non-selective, selective, and dual selective screening medium, as well as on plates supplemented with chloramphenicol at
25 μg ml-1 and tetracycline at 12.5 μg ml-1. Plates were incubated for at 37°C for 24 h and then
at room temperature for an additional 24 h. All controls, self-activation tests, and recombinant
pBT and pTRG pairs were plated as 5 µl spots as described above. Images were then captured to
record data.
Dual over-expression of TolQ and FtsN
Once pMT002 (pBAD18-Cm encoding TolQ), pMT003 (pBAD18-Kan encoding FtsN),
and pMT004 (pPro18-Kan encoding FtsN) were constructed and sequence confirmed, W3110
wild type E. coli was dually transformed with plasmid pairs. First, W3110 was transformed with
pMT002 using the one step TSS transformation protocol (Chung et al., 1989), as described in
Chapter II. Transformants of W3110 with pMT002 were grown overnight in LB supplemented
with chloramphenicol at 34 µg ml-1 and subcultured 1:200 in LB supplemented with
chloramphenicol at 34 µg ml-1 with and without 0.1% (w/v) ʟ-arabinose to confirm that the over-
expression of TolQ from pMT002 also resulted in cell filamentation. Indeed, W3110 E. coli carrying pMT002 grown overnight in LB supplemented with chloramphenicol 34 µg ml-1 and
0.1% (w/v) ʟ-arabinose filamented to a degree comparable to wild type (W3110) cells that over- 73
express TolQ from the pRA031 (pBAD24 encoding tolQ) plasmid (data not shown). The
control, W3110 TSS-transformed with pBAD18-Cm, did not filament when grown at 0.1% (w/v)
arabinose. W3110 carrying pMT002 was then secondarily transformed using the one step TSS
transformation protocol (Chung et al., 1989) with 5 µl pMT003 (pBAD18-Kan encoding FtsN)
as described above. Transformation mixture was plated as 200 µl aliquots on LB supplemented
with chloramphenicol at 34 µg ml-1 and kanamycin at 30 µg ml-1 and then incubated overnight at
37°C to select for dual transformants of pMT002 and pMT003. As a control, W3110 was dual
transformed with pMT002 and pBAD18-Kan. Similarly, W3110 carrying pMT002 was dual
transformed with pMT004 (pPro18-Kan encoding FtsN) and, as a control, with pPro18-Kan, as
described above. For filamentation assays, overnight cultures of dual-transformants were
subcultured 1:200 and grown for 24 h before being stained with safranin and viewed
microscopically. Media was supplemented with chloramphenicol at 34 µg ml-1 and kanamycin at 30 µg ml-1 and with either 0.0% or 0.1% (w/v) ʟ-arabinose (W3110+pMT002+pMT003 or
W3110+pMT002+pBAD18-Kan), or 0.0% or 0.1% (w/v) ʟ arabinose and 0 or 50 mM sodium propionate (W3110+pMT003+pMT004 or W3110+pMT002+pPro18-Kan). Images were obtained using a Nikon H550S series compound light microscope at 1000x total magnification, and NIS Elements documentation software.
74
A
B
Figure 17. Vector map and MCS of pBAD24. A. Diagram of pBAD24, 4.542 kb, highlighting locations of notable features. pBAD24 is derived from the pKK223-3 vector backbone. pBAD24 contains an optimized Shine-Dalgarno (SD) sequence and a translation start codon at the NcoI site of the multiple cloning site (MCS). MCS: 1300-1363. Ampicillin resistance gene; 1885-2745. pMB1 (pBR322) origin; 3337-3956. F1 origin; 2804-3103. araC gene; 974-96. PBAD promoter; 1248-1276. The paired BamHI sites were utilized in the cloning strategy and are included in the diagram in red (Diagram modified from Addgene Vector Database). B. MCS of pBAD24, demonstrating restriction sites and SD sequence (Guzman et al., 1995). 75
A
B
Figure 18. Vector map and MCS of pBAD18-Cm. A. Diagram of pBAD18-Cm, 6.026 kb, highlighting locations of notable features. pBAD18-Cm is also derived from the pKK223-3 vector backbone. pBAD18-Cm lacks the optimized SD sequence and translation start codon at the NcoI site of the multiple cloning site (MCS) that is found in pBAD24. MCS: 1300-1363. Chloramphenicol resistance gene (CAT/CamR): 3765-3106. pMB1 (pBR322) origin; 4746- 5365. F1 origin; 4213-4512. araC gene; 974-96. PBAD promoter; 1248-1276. The paired BamHI sites were utilized in the cloning strategy and are included in the diagram in red (Diagram modified from LabLife Vector Database). B. MCS of pBAD18-Cm, demonstrating restriction sites (Guzman et al., 1995).
76
A
B
Figure 19. Vector map and MCS of pBAD18-Kan. A) Diagram of pBAD18-Kan, 5.437 kb, highlighting locations of notable features. pBAD18-Kan is also derived from the pKK223-3 vector backbone. pBAD18-Kan lacks the optimized SD sequence and translation start codon at the NcoI site of the multiple cloning site (MCS) that is found in pBAD24. MCS: 1300-1363. Kanamycin resistance gene (KanR) 2310-3125. pMB1 (pBR322) origin; 4157-4776. F1 origin; 3624-3923. araC gene; 974-96. PBAD promoter; 1248-1276. The paired BamHI sites were utilized in the cloning strategy and are included in the diagram in red (Diagram modified from LabLife Vector Database). B) MCS of pBAD18-Cm, demonstrating restriction sites (Guzman et al., 1995). 77
Figure 20. Map of pPro18 as a representative of the pPro18-Kan vector, and MCS of pPro vectors. A. The pPro vectors are derived from manipulation of the pBAD vectors, with pPro18- Kan a derivative of pBAD18-Kan. pPro18-Kan possesses the pBR322 origin of replication, the prpBCDE promoter (PprpB) controlling expression of propionate catabolic genes (prpBCDE), and the prpR gene encoding the positive regulator of the promoter. PprpB and prpR replace PBAD and araC of the pBAD vectors. Derived from pBAD18-Kan, pPro18-Kan contains the kanamycin resistance gene (KanR), the MCS1 (B), an M13 origin of replication for phage packaging, and the rrnB transcription terminator. All restriction sites above are unique in pBPro18-Kan except for SmaI and XmaI (diagram and description from Lee and Keasling, 2005).
78
Results
Bacteriomatch II ® two-hybrid analysis
Two-hybrid screening showed a positive interaction between pMT005 (pBT-TolQ 1-19)
and pMT011 (pTRG-FtsN54-319) as well as between pMT005 and pMT010 (pTRG-FtsN 54-243)
(Fig. 21). Spots 1-4 and 14-16 represent self-activation tests for each construct, spotted as 5µl
aliquots. Spots 5-8 and 11-12 represent 5µl aliquot spots of protein-protein pairings. Spot 10 is
a negative interaction control, and spot 9 is the positive interaction control. Spot 13 is an
untransformed negative control for antibiotic selection. Plates were incubated for 24 h at 37°C
and then moved to 25°C for an additional 24-48 h to detect less robust interactions before images
were captured. Two-hybrid analysis indicates that the periplasmic N-terminal domain of TolQ
interacts with the periplasmic region of FtsN (Fig. 21, spots 7 and 8). Furthermore, the FtsN
domain required for this interaction includes the CM-proximal α-helices (approx. amino acid 62-
123) of the unstructured proline/glutamine-rich linker region (amino acids 124-243), as indicated
by the growth seen at spot 8. This region has been previously described as necessary to support
cell division is FtsN71-105 (Gerding et al., 2009). Table 4 summarizes results of two-hybrid screening in this study.
Dual over-expression of TolQ and FtsN
Evidence of a physical interaction between TolQ and FtsN as supported by the two- hybrid results provides support to the hypothesis that excess amounts of the TolQ protein can serve to sequester enough endogenous FtsN to significantly reduce the amount of protein available at the divisome, thereby indirectly resulting in a null ftsN-like appearance. If the
filamentation phenotype associated with over-expression of TolQ is truly the result of a
reduction of available FtsN, then concurrent over-expression of FtsN should provide more 79
Figure 21. Spot plates identifying putative protein-protein interactions via a bacterial two- hybrid system. Spot numbers correspond to the plasmid pairings described in Table 2. Transformants were plated on standard LB media (A), nonselective His dropout media containing chloramphenicol 25 µg ml-1 and tetracycline 12.5 µg ml-1 (B), selective screening media comprised of nonselective agar supplemented with 5 mM 3-amino-1,2,4 triazole (C), and dual selective screening media comprised of selective screening agar supplemented with streptomycin 12.5 µg ml-1 (D). Growth corresponding to spots 7 and 8 on selective and dual selective media indicate a positive interaction between the N-terminal periplasmic region of TolQ and the periplasmic C-terminal portion of FtsN. This interaction is stronger with the entire FtsN periplasmic region present (spot 7) as indicated by the greater extent of growth, but is still evident with the truncated portion missing the last 76 residues but still possessing the identified functional domain (spot 8). Spot 9 represents the positive control for robust interaction, and all other plasmid combinations tested showed little to no significant growth on screening medias, indicating a lack of protein-protein interaction.
80
Table 4. Summary of two-hybrid spot identifications and results.
Spot Identification of spot Growth on Growth on Number selective dual selective screening media media
Two-hybrid control spots Spot 9 pBT-LGF2 and pTRG-Gal11P (positive control for interaction) +++ +++ Spot 10 pBT and pTRG-Gal11P (negative control for interaction) - - Spot 13 Bacteriomatch reporter strain Untransformed negative control - - For antibiotic selection
Two-hybrid self-activation test spots Spot 1 pBT and pMT011 - - Spot 2 pBT and pMT010 - - Spot 3 pMT007 and pTRG - - Spot 4 pMT005 and pTRG - - Spot 14 pBT and pMT009 - - Spot 15 pMT008 and pTRG - - Spot 16 pMT006 and pTRG - -
Two-hybrid protein-protein interaction spots Spot 5 pMT007 and pMT011 - - Spot 6 pMT007 and pMT010 - - Spot 7 pMT005 and pMT011 ++ ++ Spot 8 pMT005 and pMT010 ++ + Spot 11 pMT008 and pMT009 - - Spot 12 pMT006 and pMT009 - -
The results for all two-hybrid pairings are summarized. A negative (-) indicates no growth at all or minimal growth below the significant level expected for background. A positive (+++) represents the greatest amount of growth obtained, as seen for the positive control (pBT-LGF2 and pTRG-Gal11P). Positive (++) and (+) represent extent of growth relative to positive controls (+++). Growth corresponding to spots 7 and 8 on selective and dual selective media indicate a positive interaction between the N- terminal periplasmic region of TolQ and the periplasmic C-terminal portion of FtsN.
81
unbound protein free to interact with other divisome components and therefore alleviate the filamentation phenotype observed with over-expression of TolQ alone. Two different coexpression models were employed to investigate this possibility. The first used two distinct pBAD constructs that varied by selective marker. This allowed for the maintenance of both vectors when grown under dual selective conditions. The pBAD18-Cm vector was used to carry the tolQ gene (pMT002), while pBAD18-Kan carried ftsN (pMT003). W3110 wild-type cells were cotransformed either with pMT002 and pMT003, or pMT002 along with the empty pBAD18-Kan vector as a negative control. Cells were then induced with 0.1% (w/v) ʟ-arabinose for 24 hours and stained for microscopic viewing. As expected, uninduced cells did not filament
(Fig. 22, panel A). However, the cells coexpressing both TolQ (pMT002) and FtsN (pMT003) showed a significant reduction in the intensity of the filamentation phenotype (Fig. 22, panel D) compared to cells over-expressing the TolQ protein alone (Fig. 22, panel B). This change was not simply the result of these cells carrying an additional plasmid construct, since cells cotransformed with pMT002 and the empty pBAD18-Kan vector still displayed the characteristic filamentation phenotype (Fig. 22, panel C).
To discount the possibility that somehow the arabinose-induction effect was simply diluted by the inclusion of a second pBAD construct, an additional set of co-expression analyses was performed using a second expression vector with a different induction system. The ftsN gene was cloned into a pPro18-Kan vector that uses sodium propionate as an expression inducer
(pMT004). Cells were either transformed with both pMT002 and pMT004 or with the pMT002 and the pPro18-Kan empty vector, and induced under conditions of 0.1% (w/v) ʟ-arabinose only,
0.1% (w/v) ʟ-arabinose and 50mM sodium propionate, or neither inducer present. The data presented in Fig. 23 indicate that filamented cells were present when cotransformed with either 82
construct combination under ʟ-arabinose induction alone (Fig. 23, panels B and E). However,
the inclusion of the propionate inducer significantly reduced the degree of filamentation in the
cells carrying the pMT004 construct, to the point where slightly elongated cells were prevalent,
but filaments were not observed (Fig. 23, panel F). Importantly, cells that were carrying the
empty pPro18-Kan vector did not show a similar reduction in filamentation with the presence of both inducers (panel C), demonstrating that this is a phenomenon specific to FtsN expression and not the result of the addition of the propionate metabolite alone.
83
Figure 22. W3110 E. coli cells grown under dual selective conditions of chloramphenicol and kanamycin for 24 hours (1). Panels A and B are micrographs of E. coli cells cotransformed with tolQ in pBAD18-Cm (pMT002) and pBAD18-Kan under conditions of no arabinose induction and 0.1% (w/v) ʟ-arabinose induction, respectively. Panels C and D are micrographs of E. coli cells cotransformed with pMT002 and ftsN in pBAD18-Kan (pMT003) under conditions of no arabinose induction and 0.1% (w/v) ʟ-arabinose induction, respectively. Filamentation is pronounced in cells over-expressing TolQ alone (panel B), but this phenotype is markedly reduced with concurrent ʟ-arabinose-induced expression of FtsN (panel D).
84
Figure 23. W3110 E. coli cells grown under dual selective conditions of chloramphenicol and kanamycin for 24 hours (2). Panels A, B and C are micrographs of E. coli cells cotransformed with tolQ in pBAD18-Cm (pMT002) and pPro18-Kan under conditions of no arabinose induction, 0.1% (w/v) ʟ-arabinose induction, and 0.1% (w/v) ʟ-arabinose + 50 mM sodium propionate induction, respectively. Panels D, E and F are micrographs of E. coli cells cotransformed with tolQ in pBAD18-Cm (pMT002) and ftsN in pPro18-Kan (pMT004) under conditions of no arabinose induction, 0.1% (w/v) ʟ-arabinose induction, and 0.1% (w/v) ʟ- arabinose + 50 mM sodium propionate induction, respectively. In both sets of cotransformants arabinose induction of TolQ resulted in numerous filamented cells (panels B and E). While elongated cells are still apparent, there are no observable filaments present under simultaneous sodium propionate induction of FtsN (panel F).
85
Discussion
There were two specific objectives designed to investigate a potential protein-protein interaction between TolQ and FtsN. A bacterial two hybrid system and concurrent over- expression of TolQ and FtsN were the tools used to investigate the two-fold aim of identifying the specific interaction between TolQ and FtsN and linking this interaction to the phenotype of filamentation. For the first goal of this aim, two-hybrid analysis was used to test the ability of specific cell-compartment localized regions of TolQ and FtsN to interact in vivo. In this two- hybrid analysis, periplasmic regions of TolQ were paired with periplasmic regions of FtsN and the potential for interaction between regions of the two proteins was investigated. Similarly, cytoplasmic regions of each protein were analyzed for their ability to interact as well. Previous reports had linked the Tol-Pal protein complex to cell division (Llamas et al., 2000; Heilpern and
Waldor, 2000; Dubuisson et al., 2005; Yeh et. al., 2010), as tol mutants have been shown to form filaments when grown under conditions of non-standard osmolarity. Gerding et al. (2007) found that GFP and RFP fusions of the Tol-Pal proteins accumulate at the division site in E. coli and that TolQ localizes to the division site in the absence of other Tol proteins in a manner that specifically requires FtsN activity. Therefore, to test the hypothesis that a direct physical interaction occurs between TolQ and an essential divisome protein and that when over- expressed, TolQ interaction with this protein leads to division impairment, FtsN was chosen as the “target” for the “bait” TolQ protein in two-hybrid analysis. While investigating a potential interaction between the entire TolQ and FtsN proteins would have provided useful information, the strategy used for this analysis was designed to not only identify a protein-protein interaction between TolQ and FtsN but to also identify the regions of interaction between the two proteins based on membrane topology. 86
Two-hybrid analysis was used to evaluate the potential interaction between TolQ and
FtsN and determine the specific domains involved in such an interaction. This analysis of
protein-protein interactions between TolQ and FtsN suggests an in vivo interaction between the
two proteins. In this study, only two pairings in particular demonstrated a strong positive
interaction acceptable above background growth levels with respect to the various controls in
both selective and dual selective screens. Both screens indicate that this interaction occurs within the periplasmic space and specifically involves the N-terminal domain of TolQ (TolQ1-19) and the proximal periplasmic domain of FtsN (FtsN54-243), the same FtsN domain previously identified as necessary to support proper cell division (Dai et al., 1996; Yang et al., 2004) and to display high sequence conservation among a number of organisms (Yang et al. 2004; Moll and
Thanbichler, 2009). Dai et al. (1996) determined that as long as it is replaced by another localization signal, the MalE cleavable signal sequence, the transmembrane region of FtsN is not required for function and that the N-terminal and cytoplasmic domains of FtsN are similarly not required for FtsN function in the cell division process. It was reported by Yang et al. (2004) from NMR analysis that the C-terminal 77 residues of FtsN form a folded globular domain that shares sequence homology with the peptidoglycan-binding (murein-binding) domain of an autolysin. The remaining periplasmic portion of FtsN is largely unstructured except for 3 partially folded α-helices at the periplasmic membrane proximal region. It was suggested that it
is this helical, periplasmic region of FtsN that interacts with other proteins of the septal ring
during cell division (Yang et al., 2004). Gerding et al. (2009) further characterized the
periplasmic, helical, membrane-proximal region of FtsN, concluding that the essential function
of FtsN is centered at the 35 residues including and surrounding the second partial α-helix (FtsN
71-105). Furthermore, characterization of the C-teriminal murein-binding domain of FtsN 87
indicated that this domain recognizes and binds septal murein that is available only during cell
constriction. Although both the entire periplasmic domain (FtsN54-319) and a truncated form of
the periplasmic domain (FtsN54-243) demonstrated a significant positive interaction with TolQ in two hybrid analysis, the C-terminal truncated FtsN consistently displayed weaker growth on
selective media relative to the full-length construct. In this screening system the stronger the
protein-protein interaction, the more significant the growth on selective media, indicating a
possible requirement for the extreme C-terminal murein-binding domain in proper FtsN
localization or conformation and optimal TolQ interaction. It was not possible to exclude the
potential that further interaction between TolQ and FtsN between the transmembrane domains of
each or between periplasmic domains of one and cytoplasmic domains of the other exist, since
TM domains are not suitable to be analyzed by the BacterioMatch II ® two-hybrid system and
opposite cellular compartment pairings were not examined.
The prediction that the over-expression of TolQ inhibits cell division by saturation or
sequestration of FtsN was also supported by results of this study. Both preliminary studies
(Teleha, 2009) and studies in Chapter II of this dissertation had demonstrated that TolQ over-
expression leads to cell filamentation. Two hybrid analysis demonstrated a positive protein-
protein interaction between TolQ and the cell division protein FtsN. This interaction appears to
occur between the N-terminus of TolQ and the C-terminal region of FtsN. If cell division were
disrupted upon TolQ over-expression because the interaction between TolQ and FtsN is such that
FtsN is sequestered away from either its position or functional role at the divisome, it would be
expected that an increase in the amount of available FtsN would alleviate the division
impairment. Indeed, the concurrent over-expression of FtsN and TolQ using two different
induction methods alleviates the filamentation phenotype resulting from over-expression of TolQ 88
alone, lending further support to an FtsN sequestration effect. When wild-type E. coli over- expressed TolQ under ʟ-arabinose induction, filamentation resulted. The concurrent over- expression of FtsN, under either ʟ-arabinose or propionate induction alleviated the filamentation phenotype resulting from TolQ over-expression alone. The results obtained from bacterial two- hybrid analysis and TolQ and FtsN dual over-expression studies supported the hypothesis that filamentation that results from TolQ over-expression occurs as FtsN becomes titrated by TolQ and unavailable to the cell division process, essentially leading to a null ftsN phenotype.
89
CHAPTER IV
THE CONTRIBUTION OF THE N-TERMINAL REGION OF THE TOLQ PROTEIN TO THE
OVER-EXPRESSION DIVISION PHENOTYPE AS ANALYZED THROUGH SEQUENCE
MODIFICATIONS
Introduction
Earlier objectives of this dissertation provided informative details regarding the phenotype associated with the over-expression of the TolQ protein. It was determined that filamentation occurs due to the over-expression of TolQ but not the over-expression of TolA or
TolR. Western blot analysis confirmed that the degree to which cell division is disrupted correlates directly with protein levels in the cell, with the most pronounced filamentation observed in cells in which TolQ is present at above wild type levels. It was also determined that filamentation results whether or not other Tol system components are present and that the over- expression of E. coli TolQ in the γ-proteobacteria E. amnigenus and C. muytjensii and in the E. coli strain BL21 leads to a comparable disruption of cell division. Using bacterial two-hybrid analysis, it was determined that a protein-protein interaction likely occurs between the N- terminal periplasmic region of TolQ and the C-terminal periplasmic region of the cell division protein FtsN. Furthermore, the over-expression of FtsN in cells over-expressing TolQ alleviates the filamentation phenotype observed when TolQ alone is over-expressed. Together, these results suggest that the interaction between TolQ and FtsN leads to filamentation when TolQ is present in excess as a consequence of TolQ-sequestration of FtsN, which inhibits it from carrying out its role in the cell division process. Therefore, the TolQ over-expression phenotype is essentially a null ftsN phenotype. Two-hybrid analysis identified the N-terminal domain of 90
TolQ as a unique participant in this interaction. This domain of TolQ is comprised of ~19 amino acids, a relatively small number of residues of the entire protein. The last objective of this dissertation was to modify the TolQ protein by the addition, deletion, and substitution of specific amino acids in an attempt to further narrow down the residues of TolQ that contribute to the
TolQ over-expression filamentation phenotype. Additionally, these modified TolQ proteins were assessed for their ability to carry out other Tol-dependent functions in the cell. Specifically, it was determined if modified TolQ proteins support cell filamentation, colicin A sensitivity, cross- talk with the TonB system (as determined using colicin Ia sensitivity assays) and maintenance of the OM (as determined using deoxycholate sensitivity assays).
In this study, the over-expression of TolQ consistently led to filamentous growth of E. coli. In E. coli cells that filament when TolQ is present in excess, the degree of filamentation appears to be variable. A culture generally appears as a mixture of cells ranging in length from those that appear near wild type in length to others that form very long filaments. Furthermore, in many preparations of these cells, septation is variably evident. Two factors contribute to this seemingly inconsistent phenotype. First, while the plasmid expression system used in this study, based on the araBAD promoter, allows for high levels of protein expression when cells are grown at saturating levels of the inducer ʟ-arabinose, subsaturating levels of inducer generate populations of cells that are either fully induced or uninduced (Siegele and Hu, 1997;
Khlebnikov et al, 2000). Gene expression through induction of the araBAD promoter is described as autocatalytic, or “all-or-none” gene expression (Khlebnikov et al., 2000).
Autocatalytic gene expression results when genes encoding the transporter for the inducer are under control of the inducer (Khlebnikov et al., 2000). Therefore, gene expression from this
promoter is subject to “bursts” of expression in cells that accumulate inducer at any given time. 91
Because induction within a culture varies in timing from cell to cell, a culture grown at subsaturating inducer concentrations can appear as a mixed culture at any given time. The proportion of induced cells to uninduced cells in a culture increases as inducer concentrations approach saturating levels. A second factor to consider that might lead to the variable degrees of filamentation and septation observed in cultures that over-express TolQ is that, if filamentation results from a direct interaction between TolQ and FtsN that sequesters FtsN from the divisome in dividing cells, this “tying up” of FtsN is not an “all-or-none” phenomenon but rather is a result of a titration, a competitive type of inhibition. As TolQ expression levels fluctuate in individual cells over time, the proportion of FtsN that interacts with TolQ would also vary, resulting in a population of cells with varying amounts of FtsN present at the divisome to participate in the cell division process. Visually, this culture would appear just as those in this study do, as a mixed culture of cells exhibiting varying degrees of septation and cell elongation.
The first objective of this final dissertation aim was to test the ability of variously modified TolQ proteins to induce filamentation when over-expressed. Because two-hybrid analysis indicated that an interaction between TolQ and FtsN occured via the TolQ N-terminus and dual over-expression studies showed that over-expression of FtsN alleviated the TolQ over- expression phenotype, TolQ modifications were targeted to the N-terminus. W3110 wild type cells over-expressing these modified TolQ proteins were analyzed for their cell division phenotypes. The final aims of this dissertation are discussed below.
Some strains of E. coli produce bacteriocins, protein toxins that are lethal to related strains, presumably in defense of a particular ecological niche. These bacteriocins are called colicins when produced by E. coli. The ability to produce colicins is encoded on extrachromosomal genetic elements, plasmids that are referred to as colicinigenic factors. These 92
plasmids are thought to be maintained naturally by a number of plasmid addiction systems
(reviewed in Kroll et al., 2010). Plasmids are transferred through conjugation and are either
present as small, multicopy plasmids within the cell or as large, monocopy plasmids that also
encode additional genes advantageous for the bacterium (Cascales et al., 2007). Colicin-
producing strains possess an immunity mechanism that protects them from the effects of the
bacteriocin. A small protein called the immunity protein is encoded by the same plasmid that encodes the colicin and protects the colicin-producing strain from its own toxin. In a colicin
operon, the first gene encodes the colicin. The immunity protein is either encoded within the
operon downstream of the colicin gene (nuclease colicins) or on the DNA strand opposite the
colicin gene (pore-forming colicins). The last gene in the colicin operon encodes the lysis
protein, necessary for colicin release into the medium. The colicin operon is largely repressed by
the LexA protein, which also represses the SOS genes that are activated upon cell stress or
damage. Therefore, colicins are produced in large quantities under stress conditions (Cascales et
al., 2007). Colicins are composed of three globular protein domains. The central domain
recognizes and binds to the receptor on a sensitive cell. The N-terminal domain mediates the
translocation of the colicin into the sensitive cell and the C-terminal domain is responsible for
the lethal action of the colicin on the cell (Braun et al., 1994). Colicin action on the sensitive cell
can be in the form of nucleases that degrade RNA or DNA, or the colicin can physically create
pores in the target cell membrane. Secondary effects include nutrient starvation, as colicins often
utilize iron and vitamin B receptors (Cascales et al., 2007).
Colicins are classified into two groups based on the receptors and cellular machinery they
parasitize to enter sensitive cells. Group A colicins bind to the receptors BtuB, Tsx, OmpF,
OmpA, or IutA, specifically, and are dependent upon the Tol-Pal system for import to reach their 93
cellular targets. Group A colicins include colicins A, E1-E9, K, L, N, S4, U, and Y. Most group
A colicins are pore-forming colicins, although some group A colicins display nuclease cytotoxity. Group B colicins utilize the TonB system for translocation after binding to the receptors FepA, Cir, FhuA, or Tsx. Group B colicins include colicins B, D, H, Ia, M, 5, and 10.
Group A colicins are encoded on small, multicopy plasmids and are released into the medium, while group B colicins are encoded on large plasmids and are not secreted (Cascales et al.,
2007). Numerous studies have been undertaken to determine the energy requirements for colicin entry into the cell. It has been concluded from results of these studies that while group B colicins appear to require energy provided by the ExbB/ExbD/TonB complex to cross the cell envelope, group A colicins do not appear to require Tol-derived energy at any time during translocation, although they require the structural presence of various combinations of Tol proteins (Cascales et al., 2007).
Colicin A, a group A colicin, interacts with the TonB-dependent vitamin B12 transporter
BtuB as its receptor. For translocation across the cell envelope, colicin A requires the outer membrane protein OmpF, as well as the TolA, TolB, TolQ, and TolR proteins. Colicin A is a pore-forming colicin that creates small perforations in phospholipid bilayers. These pores reduce the membrane electrochemical potential, effectively killing cells. Colicin Ia is a group B colicin and is also a pore-forming colicin. Colicin Ia utilizes Cir as its receptor, which natively binds to metal chelates and requires the TonB, ExbB, and ExbD proteins for translocation. Because the mode of colicin action, receptor identity and translocation systems of sensitive cells have been well-characterized (reviewed in Cascales et al., 2007), colicins are often used as tools in the laboratory to assess the presence and/or activity of the Tol-Pal and the TonB system proteins.
Particularly in mutational analyses, a loss of sensitivity to colicins can serve as a clear indicator 94
of loss of function in one or all of the components of either system. Not only can colicin
sensitivity serve as an indicator of Tol or TonB complex function, but it can also be used in
cross-reactivity studies of mutants of one system to assess its ability to substitute, even
inefficiently, for the loss of the other. In this study, sensitivity to colicin A serves as a test for
Tol system function. Cells expressing variously altered TolQ proteins were assessed for their
sensitivity to colicin A, which requires TolQ/R/A/B. Additionally, the ability of these altered
TolQ proteins to, with TolR, substitute for ExbB/ExbD in colicin Ia sensitivity assays was tested.
While the precise mechanisms by which the Tol proteins contribute to outer membrane
(OM) integrity are not fully understood, tol mutants exhibit numerous phenotypes associated
with the loss of OM integrity. Mutations in the Tol system not only confer tolerance to
filamentous bacteriophage (Sun and Webster, 1987) and group A colicins (reviewed in Cascales
et al., 2007), but tol mutants also exhibit a variety of phenotypes that include heightened
sensitivity to a number of antibiotics, detergents, and hydrophobic dyes, shedding of OM, and
leakage of periplasmic contents, all apparently resulting from disruption of the OM. One test
that is used to assess Tol-Pal system function is to determine the state of OM integrity by subjecting cells to growth in media containing sodium deoxycholate, the sodium salt of deoxycholic acid, a bile salt that causes bacterial cell lysis in species lacking tolerance mechanisms. Animal pathogens and commensals generally tolerate deoxycholate and other bile salts, which function in the animal to emulsify and solubilize dietary fats. Secondarily, bile acts as an antimicrobial agent in the animal. Bile salts dissolve membrane lipids, disassociate integral membrane proteins, disrupt the stability of macromolecular complexes, induce the formation secondary structures in RNA, damage DNA, and chelate calcium and iron, decreasing their availability (reviewed in Begley et al., 2005). Although the mechanisms employed are poorly 95
understood, Gram negative bacteria are more resistant to bile salts than are Gram positive
bacteria. In particular, E. coli is very resistant to bile and can even colonize the gall bladder and small intestine of animals (Ganzle et al., 1999). In E. coli, LPS contributes to OM stability and
LPS mutants show bile sensitivity (Picken and Beacham, 1977; Thanassi et al., 1997).
Additionally, bile acids can enter the cell through OM porins, as alterations to OmpF diameter increase susceptibility. Mutations in the tol genes (tolQRA) increase bile sensitivity by destabilizing the OM, allowing bile salts to enter the cell (reviewed in Begley et al., 2005). In this study, deoxycholate sensitivity was used as a method to assess Tol-Pal system function, as a loss of function of TolQ leads to OM integrity loss and cell lysis when cultures are treated with deoxycholate.
Methods
Media
Bacterial strains were maintained on Luria-Bertani (LB) plates. For assays cells were grown in LB broth with shaking at 37ºC (Miller, 1972). For filamentation assays, cells were grown for 24 h in LB broth supplemented with either 0.0% or 0.1% (w/v) ʟ-arabinose and ampicillin at 100 μg ml-1. For colicin A and B sensitivity assays, cells were plated in T-top
supplemented with 100 μg ml-1 ampicillin and 0.001% (w/v) ʟ-arabinose on T-plates (Miller
1972) as previously described (Larsen et al., 2003). Deoxycholate sensitivity assays were
performed in LB broth supplemented with ampicillin at 100 μg ml-1, 0.001% (w/v) ʟ-arabinose,
and 0.25% (w/v) deoxycholic acid sodium salt (Fisher Scientific). For transformation of wild
type cells (W3110) and RA1033 (ΔtolQ/ΔexbB/ΔExbD), the one step TSS procedure (Chung et
al., 1989) was employed and plates and broth were supplemented with 100 μg ml-1 ampicillin as necessary. All media supplements were calculated as weight per volume (w/v). 96
Bacterial Strains
Bacterial strains used in this study are summarized in Table 5. The E. coli K-12 strain
W3110 (Hill and Harnish, 1981) was used in this study as the wild type. For cloning, NEB 5-
alpha F´ Iq E. coli competent cells were used. For filamentation assays, plasmids encoding
TolQ modifications were expressed in W3110 E. coli. RA1033 (ΔtolQ/ΔexbB/ΔexbD) was used for colicin A, colicin Ia, and deoxycholate sensitivity assays (Brinkman and Larsen, 2008).
Plasmids
Plasmids used in this study are summarized in Table 5 and primers used for the construction of these plasmids are described in Table 6. For this study, the TolQ protein was modified by an N-terminal addition, two N-terminal truncations, and four N-terminal substitutions. Seven such modified TolQ proteins were utilized, each expressed under control of the arabinose promoter in pBAD24 (Chapter III, Fig. 17). An N-terminal T7 tag was originally added to TolQ in an attempt to determine TolQ levels in the cell under ʟ-arabinose induction.
The fusion with the N-terminal T7 tag added 16 amino acids and was detectable with an anti-T7
antibody for Western blot analysis. However, this addition altered other TolQ phenotypes,
including the TolQ over-expression phenotype (Teleha, 2009). This construct was briefly
examined in my Master’s thesis work, but has been included for a thorough analysis of the TolQ
N-terminus in this study following sequence confirmation. Two N-terminal-truncated TolQ
proteins were also constructed for this study. TolQ truncations of the first three and first seven
amino acids were generated and used for analysis. Lastly, four TolQ constructs were generated
with amino acid substitutions near the N-terminus. In the first substituted construct, the native
threonine residue at position two was replaced with an alanyl residue. For the second 97
substitution, the native aspartic residue at position three was replaced by an alanyl residue. In the third construct, the native methionyl residue at position four was replaced with an arginyl
residue. Lastly, a substitution at amino acid position 12 replaced a lysyl residue with a glutamic
residue. In each case, the residue substitutions significantly changed the chemical property of the residue side group in each position from that of the native residue and were thus non- conservative substitutions.
Table 5. E. coli strains and plasmids used in Chapter IV.
Strain Genotype Source/Reference W3110 F- IN(rrnD-rrnE)1 Hill & Harnish, 1981
RA1033 ΔtolQ/ΔexbB/D Brinkman and Larsen, 2008
5-alpha F´ Iq E. coli F´ proA+B+ lacIq ∆(lacZ)M15 zzf::Tn10 (TetR) / fhuA2∆(argF- NEB lacZ)U169 phoA glnV44 Φ80Δ(lacZ)M15 gyrA96 recA1 relA1 endA1 thi-1 hsdR17 Plasmids Relevant characteristics Source/Reference pBAD24 araBAD promoter, AraC, ampr Guzman et al. (1995)
pRA019 pBAD24 encoding N-terminal T7- Larsen lab, unpublished tagged TolQ (8-230) pMT015 pBAD24 encoding TolQ Present study
(4-230) pMT016 pBAD24 encoding TolQ Present study
pMT017 pBAD24 encoding TolQ(Thr 2 Ala) Present study
pMT018 pBAD24 encoding TolQ(Asp 3 Ala) Present study
pMT019 pBAD24 encoding TolQ(Met 4 Arg) Present study
pMT020 pBAD24 encoding TolQ(Lys 12 Glu) Present study
Strains, relevant genotypes, and source shown in Table 5. Filamentation assays were performed with transformants of wild type E. coli (W3110). Colicin and deoxycholate assays were performed with transformants of RA1033 (ΔtolQ/ΔexbB/D). Screening for cloning was carried out using in 5-alpha F´ Iq E. coli from NEB. 98
Table 6. Primers used for the generation of constructs used in Chapter IV
TolQ Restriction Primer Sequence Modification Site 7 amino acid N- oMT018 5'-GAATACCATGGATTTGTTCCTGAAGGC-3' term truncation NcoI Forward 3 amino acid N- None oMT019 5'-[Phos]- term truncation (Blunt end) AATATCCTTGATTTGTTCCTGAAGGCTAGCC-3' Forward oMT020 5'-CGCGGAGTTTAAGCCATGGCTGACATG-3' Thr 2 Ala NcoI Forward oMT021 5'-[Phos]- Asp 3 Ala None ACTGCCATGAATATCCTTGATTTGTTCCTG-3' Forward (Blunt end) oMT022 5'-[Phos]- Met 4 Arg None ACTGACAGGAATATCCTTGATTTGTTCCTG-3' Forward (Blunt end) oMT023 5'-[Phos]- Lys 12 Glu None ACTGACATGAATATCCTTGATTTGTTCCTGG-3' Forward (Blunt end) oMT024 5'-CACGCGCTCTAGACATGGCTTACC-3' TolQ reverse XbaI
Listed for each primer are the nucleotide sequence, the gene fragment or modified gene amplified, the primer direction, and the restriction site generated by the given primer. Each PCR product was cloned into the pBAD24 vector as described in Methods. Constructs were confirmed by sequence analysis using universal pBAD primers (Eurofins MWG Operon).
Construction of plasmids for amino acid addition, deletion, and substitution analysis
pRA019. pRA019 (pBAD24 encoding N-terminal T7-tagged TolQ) was previously constructed and tested by restriction analysis and mutant complementation (Larsen lab, unpublished). Expression was confirmed by Western blot analysis (Teleha, 2009). The construct was sequence-confirmed for this study.
pMT015. pMT015 (pBAD24 encoding TolQ with N-terminal seven amino acid truncation) was generated using standard PCR amplification of the sequence from W3110 genomic DNA. Forward primer oMT018 and reverse primer oMT024 were used for amplification. The PCR product was visualized on a 1% (w/v) agarose gel and then both the
PCR amplimer and pBAD24 were restriction-digested with NcoI (NEB) and XbaI (NEB) for 90 99
min at 37°C. Each digestion reaction was cleaned using the QIAquick PCR Purification kit and
eluted in 30 µl buffer (10 mM Tris-Cl, pH 8.5). Overnight ligation at room temperature using T4
DNA ligase was followed by transformation into NEB 5-alpha F´Iq chemically competent E. coli
cells according to protocol and shaken at 37°C for 60 min. Transformation mixture was plated
as 100 µl aliquots on LB agar plates supplemented with 100 µg ml -1 ampicillin. Plates were
incubated for 24 h at 37°C. Eight transformants were picked for screening and grown in LB
broth supplemented with 100 µg ml -1 ampicillin overnight before plasmids were collected using the QIAquick Spin Miniprep kit. Plasmids were digested with BsgI (NEB) and with HindIII
(NEB) for 90 min at 37°C and run on a 1% (w/v) agarose gel to screen for the truncated TolQ
insert. Three of the plasmids gave the correct size restriction fragment, and one was chosen for
sequence confirmation. Once confirmed to have the correct sequence, this plasmid was named
pMT015 to be used for further analyses.
pMT016. pMT016 (pBAD24 encoding TolQ with N-terminal three amino acid
truncation) was generated using forward primer oMT019 and reverse primer oMT024. To
prepare the vector, pBAD24 was restriction-digested with NcoI for 60 min at 37°C. The
resultant recessed 3’ termini of the vector were then filled using the Large (Klenow) fragment of
DNA Polymerase I (NEB) to create a blunt end to be ligated to the PCR amplimer. Following
heat inactivation of the Klenow enzyme for 20 min at 70°C, the treated vector was purified with
the QIAquick PCR Purification kit. The NcoI-digested, Klenow filled pBAD24 vector and the
truncated TolQ PCR amplimer were restriction-digested with XbaI for 60 min at 37°C, purified
using the QIAquick PCR Purification kit and eluted in 30 µl buffer (10 mM Tris-Cl, pH 8.5).
The vector was then dephosphorylated for 30 min at 37°C using Antarctic phosphatase (NEB).
The phosphatase enzyme was heat-inactivated at 65°C for 5 min. Following overnight ligation 100
with T4 DNA ligase, 5-alpha F´ Iq chemically competent E. coli cells were transformed with 5
µl ligation mixture according to protocol (NEB) and shaken at 37°C for 60 min. Transformation mixtures were plated as 200 µl aliquots on LB agar plates supplemented with 100 µg ml -1
ampicillin. Plates were incubated for 24 h at 37°C. Eighteen transformants were picked for
screening and grown overnight in LB broth supplemented with 100 µg ml -1 ampicillin before plasmids were collected using the QIAquick Spin Miniprep kit. Plasmids were digested with
BsgI for 60 min at 37°C and run on a 1% (w/v) agarose gel to screen for the truncated TolQ insert. Four of these 18 plasmids were then further screened with SphI (NEB), with three producing the expected digestion fragments. One construct was chosen for sequence confirmation. Once confirmed to have the correct sequence, this plasmid was named pMT016 to be used for further analyses.
pMT017. pMT017 (pBAD24 encoding TolQThr 2 Ala) was generated using standard PCR
with primers designed to substitute an alanyl residue at amino acid position two for the native
threonine residue. This change at the near N-terminus substitutes a nonpolar, hydrophobic
residue for a polar, uncharged residue. Using forward primer oMT020 and reverse primer
oMT024, the TolQThr 2 Ala fragment was amplified and confirmed on a 1% (w/v) agarose gel. The
TolQThr 2 Ala amplimer and pBAD24 were restriction digested with NcoI and XbaI for 90 min at
37°C. Each was purified using the QIAquick PCR Purification kit and eluted in 30 µl buffer (10
mM Tris-Cl, pH 8.5). Overnight ligation at room temperature with T4 DNA ligase was followed
by transformation of 5 µl ligation mixture into 5-alpha F´ Iq chemically competent E. coli according to protocol (NEB) and shaken at 37°C for 60 min. Transformation mixture was plated as 200 µl aliquots on LB agar plates supplemented with 100 µg ml -1 ampicillin. Plates were incubated for 24 h at 37°C. Eight transformants were picked for screening and grown overnight 101
in LB broth supplemented with 100 µg ml -1 ampicillin before plasmids were collected using the
QIAquick Spin Miniprep kit. Plasmids were digested with SphI for 90 min at 37°C and run on a
1% (w/v) agarose gel to screen for insert. Four of these eight constructs were then further screened with digestions using BsgI and then ClaI (NEB), with each producing the expected size digestion products. Two constructs were chosen for sequence confirmation. Following sequence confirmation, the plasmid containing the correct substitution was named pMT017 to be used for further analyses.
pMT018, pMT019, and pMT020. To generate pMT018 (pBAD24 encoding TolQAsp3Ala), pMT019 (pBAD24 encoding TolQMet4Arg), and pMT020 (pBAD24 encoding TolQLys12Glu), pBAD24 was prepared for cloning using the following procedure. pBAD24 was restriction digested with NcoI for 85 min at 37°C. Following digestion, the enzyme was heat inactivated for 20 min at 65°C before the recessed 3’ termini of the vector were filled using Klenow fragment for 30 min at 35°C to create a blunt end to be ligated to the PCR amplimer. The vector was purified with the QIAquick PCR Purification kit and restriction digested with XbaI for 90 min at 37°C. Following XbaI digestion, the prepared vector was purified again using the
QIAquick PCR Purification kit and then dephosphorylated with Antarctic phosphatase for 30 min at 37°C. The phosphatase enzyme was heat-inactivated for 5 min at 65°C. To amplify
TolQAsp3Ala, a full-length TolQ with the replacement of alanine (nonpolar hydrophobic) in place of the native aspartic residue (polar acidic) at amino acid position three, standard PCR was carried out with forward primer oMT021 and reverse primer oMT024 using W3110 genomic
DNA as a template. To amplify TolQMet4Arg, a full-length TolQ with an arginyl residue (polar basic) in place of the native methionyl residue (nonpolar hydrophobic) at amino acid position four, forward primer oMT022 and reverse primer oMT024 were used in standard PCR with 102
W3110 genomic DNA serving as a template. PCR amplification of TolQLys12Glu, full-length
TolQ with a glutamic residue (polar acidic) at amino acid position 12 in place of the native lysyl
(polar basic) was carried out using forward primer oMT023 and reverse primer oMT024, with
W3110 genomic DNA serving as a template. PCR products were confirmed on a 1% (w/v)
agarose gel, purified using the QIAquick PCR Purification kit and eluted in 50 µl buffer (10 mM
Tris-Cl, pH 8.5). Each purified PCR product was restriction digested using XbaI for 90 min at
37°C before being purified again using the QIAquick PCR Purification kit and eluted in 30 µl
buffer (10 mM Tris-Cl, pH 8.5). Overnight ligations of each amplimer with the prepared
pBAD24 vector described above (NcoI-digested, Klenow-filled, XbaI-digested,
dephosphorylated) were carried out at 16°C. Following overnight ligation with T4 DNA ligase,
5-alpha F´ Iq chemically competent E. coli cells were transformed with 5 µl each ligation mixture according to protocol (NEB) and shaken at 37°C for 60 min. Transformation mixtures were plated as 200 µl aliquots on LB agar plates supplemented with 100 µg ml -1 ampicillin.
Plates were incubated for 24 h at 37°C.
Eight transformants of TolQAsp3Ala in pBAD24 were picked for screening and grown
overnight in LB broth supplemented with 100 µg ml -1 ampicillin before plasmids were collected using the QIAquick Spin Miniprep kit. Plasmids were digested with ClaI for 60 min at 37°C and run on a 1% (w/v) agarose gel to screen for size. Five of the eight constructs appeared the correct size and four were sent out for sequence confirmation. One construct was confirmed to be correctly substituted by sequencing. Following sequence confirmation, the plasmid containing the correct substitution was named pMT018 to be used for further analyses. Eight transformants of TolQMet4Arg in pBAD24 were picked for screening and grown overnight in LB
broth supplemented with 100 µg ml -1 ampicillin before plasmids were collected using the 103
QIAquick Spin Miniprep kit. Plasmids were digested with ClaI for 60 min at 37°C and run on a
1% (w/v) agarose gel to screen for size. Six of the eight constructs appeared to be the correct
size. Two were chosen for sequence analysis and one was sequence-confirmed as correctly
substituted. Following sequence confirmation, this construct was named pMT019 to be used for
further analyses. Eight transformants of TolQLys12Glu in pBAD24 were picked for screening and grown overnight in LB broth supplemented with 100 µg ml -1 ampicillin before plasmids were
collected using the QIAquick Spin Miniprep kit. Plasmids were digested with SphI for 90 min at
37°C and run on a 1% (w.v) agarose gel to screen for insert. Two of these eight constructs were
then further screened by digestion with BsgI and then ClaI, with each digestion producing the
expected size digestion products as seen on a 1% (w/v) agarose gel. Both constructs were
confirmed by sequencing. Following sequence confirmation, one of the two plasmids containing
the correct substitution was named pMT020 to be used for further analyses.
Following sequence confirmation of each modified TolQ construct, each was subjected to filamentation assays in wild type (W3110) cells and colcins A and Ia spot titer and DOC growth assays in RA1033 (ΔtolQΔexbBΔexbD) to eliminate any possible contribution to phenotype by crosstalk from the TonB complex. Alignments highlighting nucleotide additions, deletions, and substitutions are depicted in Fig. 24 and protein alignments are depicted in Fig. 25.
104
Figure 24. Nucleotide alignment of modified TolQ constructs. Alignment of the 60 nucleotide N-terminal region of unmodified TolQ (pRA031) compared to the various modifications. The pRA019 line shows the sequence with the 48 nucleotide addition encoding the T7 tag. The pMT017, pMT018, pMT019, and pMT020 lines show the specific nucleotide substitutions introduced, highlighted with non-shaded boxes. The pMT016 and pMT015 lines show the 9 and 18 nucleotide deletions, respectively.
105
Figure 25. Full-length amino acid alignment of modified TolQ constructs. Protein alignment of unmodified TolQ (pRA031) compared to the various modifications. The pRA019 line shows the sequence with the 16 residue addition comprising the T7 tag. The pMT017, pMT018, pMT019, and pMT020 lines show the specific residue substitutions introduced, highlighted with non-shaded boxes (see Table 5 for amino acid substitution identities). The pMT016 and pMT015 lines show the 3 and 7 residue deletions, respectively.
106
Filamentation assays
For filamentation assays, W3110 E. coli was transformed with each sequence-confirmed modified TolQ construct using the TSS transformation protocol (Chung et al., 1989). Overnight cultures of W3110 carrying pRA019, pMT015, pMT016, pMT017, pMT018, pMT019, or pMT020 were subcultured 1:200 in LB broth supplemented with ampicillin at 100 µg ml-1 and
with either no ʟ-arabinose or 0.1% (w/v) ʟ-arabinose and grown for 24 h at 37ºC with shaking.
Triplicate samples of cells were then heat-fixed to a slide, stained with safranin and viewed
under light microscopy for division phenotypes. Images were obtained using a Nikon H550S
series compound light microscope at 1000x total magnification, and NIS Elements
Documentation Software.
Deoxycholate sensitivity assay
Deoxycholate (DOC) sensitivity assays were performed to determine the ability of the
modified TolQ proteins to participate in maintenance of outer membrane integrity. RA1033
(ΔtolQ/ΔexbB/D) was transformed with each sequence-confirmed modified TolQ construct using
the TSS transformation protocol (Chung et al., 1989). Overnight cultures of cells were
subcultured 1:100 in LB with 100 μg ml-1 ampicillin at 0.001% (w/v) ʟ-arabinose and grown
with shaking at 37°C for 60 min before the addition of DOC at 0.25% (w/v). A550 readings were
taken on a Spec20 spectrophotometer with a path length of 1.5 cm at 60 min intervals for 240 min. Outer membrane integrity was determined based on whether cell density (absorbance readings) increased throughout the assay or decreased after the addition of DOC, indicative of cell lysis. Cells with compromised outer membranes are more susceptible to lysis in the presence of DOC. DOC sensitivity assays were performed in triplicate. 107
Colicin A and colicin Ia spot titers
Overnight cultures of RA1033 carrying each modified TolQ construct were subcultured
1:100 in LB supplemented with 100 μg ml-1 ampicillin at 0.001% (w/v) ʟ-arabinose and grown
with shaking at 37°C to an A550 of 0.4, as determined with the Spec20 spectrophotometer. An aliquot of 100 µl cell culture was suspended in 3.5 ml molten T- top agar containing 100 μg ml-1 ampicillin and 0.001% (w/v) ʟ-arabinose and overlaid onto room temperature T-plates. Serial
five-fold dilutions of crude colicin A and colicin Ia preps were applied to plate surfaces as 5μl
aliquots and plates were incubated overnight at 37ºC. Results were recorded as the highest
dilution at which clearing of the bacterial lawn was observed. Colicin A and colicin Ia spot titers
were performed in triplicate.
Results
Cell division phenotypes of variously modified TolQ proteins
Previous work with an N-terminally modified TolQ was carried out using a pBAD24
derived plasmid that encoded TolQ/R, with a T7 antibody-detectable tag fused to the N-terminus
of TolQ (Teleha, 2009). This construct was designed to conserve the stoichiometric balance of
TolQ and TolR, previously suggested as important for Tol system function (Cascales et al.,
2001). Results of current studies indicate that this balance is not necessary for inducing
filamentation in cells that over-express TolQ. Therefore, a second construct of plasmid encoded,
N-terminally T7-tagged TolQ was used for further analysis. This construct, pRA019 (Larsen,
unpublished) is an N-terminally modified TolQ protein with the addition of a 16 amino acid T7
antibody-detectable tag and was used for analysis of cell division phenotypes associated with N-
terminally modified TolQ for this study. The over-expression of TolQ alone has been shown to 108
result in filamentation, indicating that there is no need for the TolQ partner TolR to be
concurrently expressed in this context. Multiple lines of evidence from this study support this
conclusion. Furthermore, if filamentation results from the sequestration of FtsN by excess TolQ
and a protein-protein interaction occurs between the N-terminal region of TolQ and FtsN as indicated from two-hybrid studies, it was expected that modifications to the N-terminal region of
TolQ that eliminate or alter residues necessary for its interaction with FtsN would prevent this
interaction, ultimately alleviating the filamentation phenotype. To test this prediction, N-
terminally modified TolQ over-expression phenotypes were analyzed. Initially, division
phenotypes were investigated with the over-expression of the T7-tagged TolQ protein,
representing an N-terminal addition. Secondly, N-terminal truncated versions of TolQ of three
and seven amino acids were analyzed for their ability to induce filamentation when over-
expressed. Lastly, non-conservative substitutions in TolQ at amino acids two, three, four, and 12
were generated and these N-terminally modified TolQ proteins were over-expressed and cells
were analyzed for cell division phenotypes.
pRA019. W3110 E. coli carrying pRA019 (TolQ with N-terminal T7 tag addition) were
grown for 24 h in LB supplemented with 0.1% (w/v) ʟ-arabinose and then stained and viewed to
determine if the N-terminal addition affected the formation of filaments. These cells were
compared to W3110 carrying the empty pBAD24 vector as a negative control and W3110 over-
expressing unmodified TolQ (pRA031) as the positive control. W3110 carrying pRA031 formed
filaments when grown in LB supplemented with 0.1% (w/v) ʟ-arabinose (Fig. 26, panels C and
D), while W3110 carrying either pBAD24 (Fig. 26, panels A and B) or pRA019 (Fig. 26, panels
E and F) grown under the same conditions for 24 h did not form filaments. These results 109
indicate that the addition of a 16 amino acid T7 tag to the TolQ N-terminus eliminated the filamentation phenotype.
pMT015, and pMT016. W3110 E. coli carrying the plasmids pMT015 (TolQ8-230) or pMT016 (TolQ4-230), when grown for 24 h in 0.1% (w/v) ʟ-arabinose did not form filaments (Fig.
8-230 27, panels B and D). Cell dimensions of W3110 over-expressing TolQ from pMT015 were
comparable to those observed for cells grown in media with no added inducer (Fig. 27, panels A
and C) and for cells carrying the empty pBAD24 vector grown with or without added ʟ-
arabinose (data not shown), although RA1033+pMT016 cells over-expressing TolQ4-230 appeared
to be slightly elongated. W3110 E. coli over-expressing unmodified TolQ (pRA031) formed
filaments when grown at 0.1% (w/v) arabinose, characteristic of those previously observed,
while those grown in media lacking added inducer did not (Fig. 26, panel D). The absence of
filamentation with the over-expression of N-terminal truncations of three and seven amino acids
of TolQ eliminated filamentation caused by native TolQ over-expression.
pMT017, pMT018, pMT019, and pMT020. The over-expression of pMT017 (TolQThr2Ala)
in wild type E. coli resulted in a qualitative reduction, but not the elimination of filamentation
that results when native TolQ is over-expressed. As previously shown, the over-expression of
native TolQ induces the formation of long smooth filaments in wild type E. coli cells (Fig. 26,
panel D). Stationary phase E. coli cells are typically 2.0 µm long and when native TolQ has
been over-expressed in these cells, the length of cells appears to increase to varying degrees.
The over-expression of TolQThr2Ala from pMT017 caused cells to increase in length variably, and
this increase in cell length was not observed in cells grown without added ʟ-arabinose (Fig. 28,
panels A and B). When TolQAsp3Ala was over-expressed from pMT018 in W3110, cells formed long, smooth filaments (Fig. 28, panel D), comparable to those seen when native TolQ is over- 110
expressed (Fig 26, panel D). The over-expression of TolQMet4Arg from pMT019 in wild type E. coli did not lead to the formation of cell filaments at all (Fig. 28, panel F). Cells did appear to be slightly elongated, but filaments did not result. Lastly, when TolQLys12Glu was over-expressed from pMT020 in W3110, cells formed long filaments similar to those over-expressing
TolQAsp3Ala (Fig. 28, panels D and H) and characteristic of those that result from the over- expression of native TolQ (Fig. 26, panel D). These results suggest that non-conservative N- terminal substitutions at residues three and 12 did not affect the ability of TolQ to induce filamentation when over-expressed, while the substitution at residue two reduces the severity of the phenotype, and the substitution at residue four eliminated filamentation altogether.
111
Figure 26. W3110 E. coli cells over-expressing pRA031 (unmodified TolQ) and pRA019 (TolQ with N-terminal T7 tag addition). Panels A and B are micrographs of E. coli W3110 carrying the empty pBAD24 vector grown with no added ʟ-arabinose and at 0.1% (w/v) ʟ- arabinose, respectively. Panels C and D are micrographs of E. coli W3110 expressing unmodified TolQ (pRA031) grown with no added ʟ-arabinose and at 0.1% (w/v) ʟ-arabinose, respectively. Panels E and F are micrographs of W3110 expressing N-terminally T7-tagged TolQ (pRA019) grown with no added ʟ-arabinose and at 0.1% (w/v) ʟ-arabinose, respectively. Slides were prepared from 24 h cultures.
112
Figure 27. W3110 E. coli cells over-expressing pMT015 (TolQ8-230) and pMT016 (TolQ4- 230). Panels A and B are micrographs of E. coli W3110 expressing N-terminal 3 residue- truncated TolQ (pMT016) grown with no added ʟ-arabinose and at 0.1% (w/v) ʟ-arabinose, respectively. Panels C and D are micrographs of E. coli W3110 expressing N-terminal 7 residue-truncated TolQ (pMT015) grown with no added ʟ-arabinose and at 0.1% (w/v) ʟ- arabinose, respectively. Slides were prepared from 24 h cultures.
113
Figure 28. W3110 E. coli cells over-expressing pMT017 (TolQThr2Ala), pMT018 (TolQAsp3Ala), pMT019 (TolQMet4Arg), and pMT020 (TolQLys12Glu). Panels A and B are micrographs of E. coli W3110 expressing TolQThr2Ala (pMT017) grown with no added ʟ- arabinose and at 0.1% (w/v) ʟ-arabinose, respectively. Panels C and D are micrographs of E. coli W3110 expressing TolQAsp3Ala (pMT018) grown with no added ʟ-arabinose and at 0.1% (w/v) ʟ-arabinose, respectively. Panels E and F are micrographs of E. coli W3110 expressing TolQMet4Arg (pMT019) grown with no added ʟ-arabinose and at 0.1% (w/v) ʟ-arabinose, respectively. Panels G and H are micrographs of E. coli W3110 expressing TolQLys12Glu (pMT020) grown with no added ʟ-arabinose and at 0.1% (w/v) ʟ-arabinose, respectively. Slides were prepared from 24 h cultures. 114
Contribution of variously modified TolQ proteins to maintenance of outer membrane integrity as
determined by deoxycholate sensitivity assays
A number of studies have indicated that the synthesis, function, and maintenance of the
OM requires energy of the pmf via the Tol proteins (Cascales et al., 2000, 2001; Germon et al.,
2001; Lloubès et al., 2001; Goemaere et al., 2007). Previous studies had suggested that an N- terminally modified TolQ showed impaired energy-dependent functionality (Teleha, 2009).
Results of the current study indicate that the extreme N-terminus of TolQ is the region of this protein responsible for causing filamentation when TolQ is over-expressed. The N-terminal T7 tag addition to TolQ eliminated filamentation (Fig. 26). Truncated TolQ, lacking the first three or seven residues, also failed to induce filamentation when over-expressed (Fig. 27). Lastly, an
N-terminal non-conservative substitution at residue four eliminated filamentation and another at
residue two decreased the degree of filamentation, while those at residues three and twelve did
not (Fig. 28). Taken together, it is likely that the interaction between TolQ and FtsN can be narrowed down to within the first seven residues of the protein, with a necessity for residue four and a possible contribution by residue two in the generation of filaments upon TolQ over- expression. Results obtained earlier in this dissertation indicated that Tol-derived energy is
unnecessary for generating filaments upon TolQ over-expression, as excess TolQ induced
filamentation in cells lacking TolR and TolA (Fig. 9). Therefore, the variously-modified TolQ
proteins generated in this dissertation were assessed for their ability to contribute to the
maintenance of the OM, an energy-dependent Tol system function. The presence of an
energetically competent TolQ protein in cells expressing TolR and TolA but lacking ExbB and
ExbD would be expected to stabilize the OM, rendering cells resistant to deoxycholate (DOC).
Energetically incompetent TolQ would cause cells to be sensitive to DOC. Each TolQ construct 115
was expressed in RA1033 E. coli (ΔtolQ, ΔexbB, ΔexbD) and cells were treated with 0.25%
(w/v) DOC. The ability of the variously modified TolQ proteins to support OM integrity was
determined by whether cultures grown in LB supplemented with DOC demonstrated an increase
in cell density or whether the culures lysed. Resistance to DOC is indicative of an energetically
competent Tol-Pal system. Importantly, deoxycholate growth assays have the potential to provide confirmation that modified TolQ proteins become inserted correctly in the CM, which is necessary for OM stability via the Tol-Pal complex. Absorbance readings were taken at 60 minute intervals for 240 min following the addition of DOC at 0.25% (w/v). Growth curves are shown as Figs. 28-30.
pRA019. When compared to W3110 wild type E. coli as well as RA1033
(ΔtolQ/Δexb/ΔexbD) expressing plasmid-encoded unmodified TolQ (pRA031) and the ΔtolQ
strain RA1033 carrying the empty pBAD24 vector, the N-terminal modification of TolQ
(RA1033+pRA019) rendered cells susceptible to DOC, comparable to cells with the mutant
phenotype (RA1033+pBAD24 empty vector; Fig. 29). These results indicated that modification
of the TolQ N-terminus by the T7 tag addition not only eliminated the filamentation phenotype
upon TolQ over-expression but also lead to a loss of OM integrity.
pMT015, and pMT016. DOC sensitivity assays showed that RA1033+pMT016 (TolQ
truncation of first three residues) allowed cells to continue to grow after the addition of DOC at
0.25% (w/v) to the medium. However, when compared to RA1033 expressing unmodified TolQ
RA1033+pRA031) and wild type W3110 E. coli, growth was retarded, as measured by lower
absorbance readings for cultures at each time point. RA1033 cells expressing the TolQ
truncation of seven (pMT015) residues did not continue to grow upon the addition of DOC.
RA1033 cells expressing truncated TolQ from pMT015 lysed in a time frame comparable to 116
those completely lacking TolQ (RA1033+pBAD24). These results indicated that the loss of the three N-terminal residues of TolQ partially compromised OM integrity, while the loss of the N-
terminal seven residues abolished the OM barrier function. Results of DOC growth assay are
represented graphically as Fig. 30.
pMT017, pMT018, pMT019, and pMT020. Non-conservative amino acid substitutions at
residues two, three, and twelve did not appear to inhibit the ability of TolQ to function in
maintenance of the OM in RA1033 (ΔtolQ, ΔexbB, ΔexbD) cells. Cultures of RA1033+pMT017
(TolQThr2Ala), RA1033+pMT018 (TolQAsp3Ala), and RA1033+pMT020 (TolQLys12Glu) each continued to grow in the media supplemented with 0.25% (w/v) DOC, comparable to the growth of both wild type (W3110) cells and RA1033 expressing unmodified TolQ from pRA031.
Conversely, RA1033+pMT019 (TolQMet4Arg) cultures lysed upon the addition of DOC, as did
RA1033 carrying the empty pBAD24 vector (Fig. 31). These results indicated that not only was
residue four (Met) of TolQ necessary for the filamentation of cells over-expressing TolQ, but
also in the maintenance of OM integrity.
117
Deoxycholate 0.25% 0.8
0.7
) 0.6 550
0.5
0.4
Cell Density Density Cell (A W3110-pBAD24 0.3 RA1033-pBAD24 RA1033-pRA031 0.2 RA1033-pRA019
0.1
0.0 -60 0 60 120 180 240 Time (min)
Figure 29. 0.25% (w/v) DOC growth curves for W3110+pBAD24, RA1033+pBAD24, RA1033+pRA031, and RA1033+pRA019. Growth curves of ΔtolQ/ΔexbB/ΔexbD (RA1033) E. coli expressing unmodified TolQ (pRA031) and N-terminally T7-tagged TolQ (pRA019) grown in LB at 0.25% deoxycholate (DOC) at 37°C. Cell density is depicted on the y-axis as measured by light absorbance at 550 nm (A550) over time (x-axis). DOC was added at t=0 and absorbance was measured spectrophotometrically at 60 min intervals for 240 min. W3110+pBAD24 (empty vector) represents DOC tolerance of wild type cells with chromosomally-encoded TolQ and RA1033+pRA031 represents ΔtolQ cells expressing plasmid- encoded, unmodified TolQ. Complete lysis was observed in RA1033+pBAD24 (negative control) and RA1033+pRA019 cultures. Assays were performed in triplicate. The data are plotted as the mean, with error bars representing standard error from the mean.
118
Deoxycholate 0.25% 1.0
0.9
0.8
) 0.7 550
0.6
0.5 W3110-pBAD24 0.4 RA1033-pBAD24 Cell Density Density Cell (A RA1033-pRA031 0.3 RA1033-pMT016 RA1033-pMT015 0.2
0.1
0.0 -60 0 60 120 180 240 Time (min)
Figure 30. 0.25% (w/v) DOC growth curves for W3110+pBAD24, RA1033+pBAD24, RA1033+pRA031, RA1033+pMT016 and RA1033+pMT015. Growth curves of ΔtolQ/ΔexbB/ΔexbD (RA1033) E. coli expressing unmodified TolQ (pRA031), TolQ4-230 (pRA016), and TolQ8-230 (pMT015), grown in LB at 0.25% DOC at 37°C. Cell density is depicted on the y-axis as measured by light absorbance at 550 nm (A550) over time (x-axis). DOC was added at t=0 and absorbance was measured spectrophotometrically at 60 min intervals for 240 min. W3110+pBAD24 (empty vector) represents DOC tolerance of wild type cells with chromosomally-encoded TolQ and RA1033+pRA031 represents ΔtolQ cells expressing plasmid- encoded, unmodified TolQ. Complete lysis was observed in RA1033+pBAD24 (negative control) and RA1033+pMT015, while RA1033+pMT016 demonstrated measurable resistance to DOC. Assays were performed in triplicate. The data are plotted as the mean, with error bars representing standard error from the mean 119
Deoxycholate 0.25% 1.2
1.0
) 550 0.8
W3110-pBAD24 0.6 RA1033-pBAD24
Cell Density Density Cell (A RA1033-pRA031 RA1033-pMT017 0.4 RA1033-pMT018 RA1033-pMT019 RA1033-pMT020 0.2
0.0 -60 0 60 120 180 240 Time (min) .
Figure 31. 0.25% (w/v) DOC growth curves for W3110+pBAD24, RA1033+pBAD24, RA1033+pRA031, RA1033+pMT017, RA1033+pMT018, RA1033+pMT019, and RA1033+pMT020. Growth curves of ΔtolQ/ΔexbB/ΔexbD (RA1033) E. coli expressing unmodified TolQ (pRA031), TolQThr2Ala (pMT017), TolQAsp3Ala (pMT018), TolQMet4Arg (pMT019), and TolQLys12Glu (pMT020) grown in LB at 0.25% DOC at 37°C. Cell density is depicted on the y-axis as measured by light absorbance at 550 nm (A550) over time (x-axis). DOC was added at t=0 and absorbance was measured spectrophotometrically at 60 min intervals for 240 min. W3110+pBAD24 (empty vector) represents DOC tolerance of wild type cells with chromosomally-encoded TolQ and RA1033+pRA031 represents ΔtolQ cells expressing plasmid- encoded, unmodified TolQ. Complete lysis was observed in RA1033+pBAD24 (negative control) and RA1033+pMT019 cultures. Assays were performed in triplicate. The data are plotted as the mean, with error bars representing standard error from the mean.
120
The effects of TolQ modifications on sensitivity to colicin A, a group A colicin, and colicin Ia, a
group B colicin
As group A colicins require the Tol proteins (Lazzaroni et al., 2002) to gain entry into E.
coli cells but do not appear to require energy to cross the cell envelope (reviewed in Cascales et
al., 2007), sensitivity to group A colicins cannot provide information about the energy
competence of Tol proteins but can serve as an indicator of a structurally competent
TolQ/TolR/TolA complex present in the CM, which is required for colicin import. Conversely, both pmf-derived energy and the TonB/ExbB/ExbD proteins are required for the entry of group
B colicins into E. coli cells (reviewed in Cascales et al., 2007). Because TolQ has been shown to partially complement ΔexbB mutants (Braun, 1989), testing the sensitivity to group B colicins can be used to assess the ability of a modified TolQ/native TolR pair to restore sensitivity in
ΔexbB/ΔExbD cells. In this dissertation, each variously modified TolQ protein was assessed for its ability to restore sensitivity to a group A colicin, colicin A, and a group B colicin, colicin Ia, in cells lacking ExbB, ExbD, and TolQ. Sensitivity to colicin A provides confirmation that a structurally competent modified TolQ protein is present in the CM, while sensitivity to colicin Ia provides for confirmation that a both structurally and energetically-competent TolQ protein is present in the CM. While the ability of modified TolQ to along with TolR engage in cross-talk with TonB would suggest that the TolQ modification does not impair energy-dependent processes of TolQ, colicin spot titers were primarily carried out to confirm that modified TolQ proteins become correctly inserted in the CM, positioning them to interact with FtsN. While
Western blot analysis might be used to confirm levels of modified TolQ in cells, these data would still not confirm correct localization of these proteins as do colicin sensitivity assays.
Colicin sensitivity assays were carried out as spot titers, with results scored as the highest 5-fold 121
dilution of colicins A and Ia to show clearing of cells, indicating colicin sensitivity. Fig. 32 depicts the sensitivity of W3110 wild type E. coli (panelA) and the ΔtolQ/ΔexbBΔexbD strain
RA1033 (panel B) to colicins A and Ia, each carrying the empty pBAD24 vector. RA1033 expressing unmodified TolQ from the pRA031 plasmid is shown as panel C. Plates at the left of each panel were spotted with 5-fold dilutions of a crude colicin A prep, while plates at the right
were spotted with dilutions of colicin Ia. Numbers 1-8 have been included in panel A to show the placement of spots corresponding to 5-fold dilutions used for all colicin spot titer assays in this study. As seen in Fig. 32, panel A, wild type W3110 E. coli shows sensitivity to colicin A to six dilutions and to seven dilutions for colicin Ia, with a cloudy spot appearing at the 8th dilution.
Shown in panel B, RA1033 (ΔtolQ/ΔexbBΔexbD) were equally resistant to colicins A and Ia, as
no spots produced clearings indicative of colicin sensitivity. Panel C depicts sensitivity to
colicins A and Ia of RA1033 expressing unmodified TolQ from the pRA031 plasmid, induced
with 0.001% (w/v) ʟ-arabinose. RA1033+pRA031 showed sensitivity to colicin A to 5 five-fold
dilutions and colicin Ia to 6 five-fold dilutions.
pRA019. RA1033 expressing pRA019, TolQ with an N-terminal T7 tag addition were
sensitive to colicin A to 5 five-fold dilutions and to colicin Ia to 4 five-fold dilutions (Fig. 33,
panel B). When compared to RA1033 expressing unmodified TolQ (Fig. 33, panel A), spots
were visible but cloudy, likely indicating a reduced sensitivity to both colicins A and Ia due to
the T7 tag addition.
pMT015, and pMT016. RA1033 expressing pMT015 (TolQ8-230), cells showed resistance
to colicin Ia, observed as no clearings for any dilution spotted on plates (Fig. 34, panel C, right).
Weak (turbid) spots to 4 five-fold dilutions can be seen on plates showing sensitivity to colicin A
(Fig. 34, panel C, left). RA1033 expressing pMT016 (TolQ4-230) were sensitive to both colicins 122
A and Ia to 5 five-fold dilutions (Fig. 34, panel B), levels of sensitivity comparable to RA1033
expressing unmodified TolQ (pRA031) for colicin A and nearly wild type sensitivity to colicin Ia
(Fig. 34, panel A). These results suggested that residues more proximal to the CM were
necessary for both colicins A and Ia sensitivity, while those at the extreme N-terminus were less important. Alternatively, a more complete N-terminus may be more structurally and therefore, functionally competent in regards to colicin entry.
pMT017, pMT018, pMT019, and pMT020. RA1033 expressing pMT017 (TolQThr2Ala) were sensitive to colicin A to 5 five-fold dilutions and to colicin Ia to 6 five-fold dilutions (Fig.
35, panel B), sensitivities comparable to those of RA1033 expressing unmodified TolQ (fig. 35, panel A). When RA1033 expressing pMT018 (TolQAsp3Ala) were spotted with dilutions of colicins A and Ia, sensitivity was also comparable to that of cells expressing unmodified TolQ
(Fig. 35, panel A) with clearings apparent to 5 five-fold dilutions (Fig. 35, panel C). Weak sensitivity to both colicins A and Ia resulted from the expression of pMT019 (TolQMet4Arg) by
RA1033 cells, as can be observed by the production of cloudy clearings up to 4 five-fold
dilutions for colicin A and to 3 five-fold dilutions of colicin Ia on spot titer plates (Fig. 35, panel
D). The expression of pMT020 (TolQLys12Glu) by RA1033 appeared to also restore near wild type sensitivity to both colicins A and Ia (Fig 35. Panel A), with spot plates showing clearings to 5 five-fold dilutions of each (Fig. 35, panel E).
123
Figure 32. Colicin spot titer assays for W3110+pBAD24, RA1033+pBAD24, and RA1033+pRA031. Colicin A and colicin Ia spot plate assays on cultures of W3110+pBAD24, RA1033 (ΔtolQ/ΔexbB/ΔexbD)+pBAD24, and RA1033+pRA031 (encoding unmodified TolQ). Five microliter aliquots of five-fold colicin dilutions were spotted in decreasing colicin concentration from right to left, top to bottom. Sensitivity of each strain to colicins was scored as the highest dilution at which a clearing in the lawn could be observed. Wild type sensitivity (chromosomally-encoded TolQ) to each colicin is shown in panel A. Complete resistance to both colicin treatments was observed as no clearing in RA1033+pBAD24 cells (panel B), and sensitivity of mutants expressing unmodified, plasmid-encoded TolQ is shown in panel C. Assays were performed in triplicate. The data shown here are from one replicate but representative of sensitivity levels observed in all three assays, with numerical values summarized in Table 7. 124
Figure 33. Colicin spot titer assays for RA1033+pRA031 and RA1033+pRA019. Colicin A and colicin Ia spot plate assays on cultures of RA1033 (ΔtolQ/ΔexbB/ΔexbD)+pRA031 (unmodified TolQ) and RA1033+pRA019 (N-terminally T7-tagged TolQ). Five microliter aliquots of five-fold colicin dilutions were spotted in decreasing colicin concentration from right to left, top to bottom. Sensitivity of each strain to colicins was scored as the highest dilution at which a clearing in the lawn could be observed. For comparison, sensitivity of cells expressing unmodified TolQ from pRA031 to each colicin is shown in panel A. Sensitivity to both colicins A and Ia treatments of RA1033+pRA019 cells is shown in panel B. Assays were performed in triplicate. The data shown here are from one replicate but representative of sensitivity levels observed in all three assays, with numerical values summarized in Table 7. 125
Figure 34. Colicin spot titer assays for RA1033+pRA031, RA1033+pMT016 and RA1033+pMT015. Colicin A and colicin Ia spot plate assays on cultures of RA1033 (ΔtolQ/ΔexbB/ΔexbD)+pRA031 (unmodified TolQ), RA1033+pMT016 (TolQ4-230) and RA1033+pMT015 (TolQ8-230). Five microliter aliquots of five-fold colicin dilutions were spotted in decreasing colicin concentration from right to left, top to bottom. Sensitivity of each strain to colicins was scored as the highest dilution at which a clearing in the lawn could be observed. For comparison, sensitivity of cells expressing unmodified TolQ from pRA031 to each colicin is shown in panel A. Sensitivity to each colicin treatment of RA1033+pMT016 and RA1033+pMT015 is shown in panels B and C, respectively. Sensitivity to both colicins appears comparable to wild-type in cells expressing pMT016, while sensitivity to colicin A appears to be decreased and to Ia lost with the deletion of the seven N-terminal residues of TolQ. Assays were performed in triplicate. The data shown here are from one replicate but representative of sensitivity levels observed in all three assays, with numerical values summarized in Table 7. 126
Figure 35. Colicin spot titer assays for RA1033+pRA031, RA1033+pMT017, RA1033+pMT018, RA1033+pMT019, and RA1033+pMT020. Colicin A and colicin Ia spot plate assays on cultures of RA1033 (ΔtolQ/ΔexbB/ΔexbD)+pRA031 (unmodified TolQ), RA1033+pMT017 (TolQThr2Ala), RA1033+pMT018 (TolQAsp3Ala), RA1033+pMT019 (TolQMet4Arg), and RA1033+pMT020 (TolQLys12Glu). Five microliter aliquots of five-fold colicin dilutions were spotted in decreasing colicin concentration from right to left, top to bottom. Sensitivity of each strain to colicins was scored as the highest dilution at which a clearing in the lawn could be observed. For comparison, sensitivity of cells expressing unmodified TolQ from pRA031 to each colicin is shown in panel A. Sensitivity to each colicin treatment of RA1033+pMT017, RA1033+pMT018,RA1033+pMT019, and RA1033+pMT020 is shown in panels B, C, D and E, respectively. Sensitivities to both colicins A and Ia appear to be decreased by the substitution of Arg for Met at residue four in TolQ, while non-conservative substitutions at residues two, three, and 12 do not appear to reduce sensitivity to colicins A or Ia to any degree measurable in this assay. Assays were performed in triplicate. The data shown here are from one replicate but representative of sensitivity levels observed in all three assays, with numerical values summarized in Table 7.
127
Discussion
The present aim of this dissertation was to examine the effect of TolQ modifications on
the protein’s ability to induce filamentation when over-expressed. Because the TolQ over-
expression phenotype appears consistent with a null ftsN mutation in E. coli, it was hypothesized
that filamentation caused by both TolQ over-expression and a loss of functional FtsN are one in
the same phenotype. The experimental design of this dissertation would provide insight into
whether or not filamentation that results from TolQ over-expression occurs because TolQ
interacts with FtsN and, when present in excess, sequesters FtsN and inhibits it from
participating in the cell division process. Two-hybrid analysis did indeed indicate a possible in
vivo interaction between TolQ and FtsN. This interaction was shown to occur between the
periplasmic N-terminus of TolQ and the periplasmic C-terminus of FtsN (Fig. 21). Furthermore,
the hypothesis that excess TolQ sequesters FtsN, leading to cell filamentation, can be supported by the data showing that TolQ-induced filamentation can be suppressed by the concurrent over- induction of plasmid-encoded FtsN (Figs. 22 and 23). Based on results of bacterial two-hybrid analysis, the last aim of this study was focused on the periplasmic N-terminus of TolQ,
comprised approximately of residues 1-19. Modifications of the TolQ N-terminus were
constructed and the altered proteins were then tested for their ability to induce filamentation.
Three different types of TolQ modifications were examined for their ability to induce
filamentation when over-expressed (Figs. 26-28): an N-terminal addition, two N-terminal
truncations, and four N-terminal substitutions. Furthermore, outer membrane integrity of cells
expressing modified TolQ proteins was determined using DOC growth assays (Figs. 29- 31).
Finally, groups A and B colicin spot titer assays were carried out in order to ascertain proper
insertion into the CM and the ability of variously modified TolQ proteins to engage in cross-talk 128
with TonB in the energy-dependent translocation of group B colicins, respectively (Figs 32-35).
A summary of these results is presented as Table 7.
Table 7. Summary of cell division phenotypes, 0.25% (w/v) DOC growth assays, and colicins A and Ia sensitivity assays for TolQ modifications.
Relative Colicin A Colicin Ia DOC 0.25% cell Sensitivity Sensitivity Resistant (R) elongation (in RA1033) (in RA1033) scored Sensitive (S) (in W3110) Scored as highest as highest dilution (in RA1033) dilution with clearing with clearing pBAD24 - 0, 0, 0 0, 0, 0 S, S, S pRA031(TolQ in ++++ 5, 5, 5 6, 6, 6 R, R, R pBAD24) TolQ N-terminal Addition pRA019 (T7-TolQ in - 5 *, 4*, 4* 4 *, 4*, 4* S, S, S pBAD24) TolQ N-terminal Truncations pMT015 (TolQ8-230 in - 4 *, 3*, 4* 0, 0, 0 S, S, S pBAD24) pMT016 (TolQ4-230 in + 5, 5, 5 5, 5, 5 R, R, R pBAD24) TolQ N-terminal Substitutions pMT017 (TolQThr2Ala +++ 5, 5, 4 6, 6, 6 R, R, R in pBAD24) pMT018 (TolQAsp3Ala ++++ 5, 5, 5 5, 5, 5 R, R, R in pBAD24) pMT019 (TolQMet4Arg + 4 *, 4*, 4* 3 *, 3*, 4* S, S, S in pBAD24) pMT020 (TolQLys12Glu ++++ 5, 5, 5 5, 5, 5 R, R, R in pBAD24)
Cell division phenotypes were observed in wild type (W3110) E. coli over-expressing modified TolQ proteins. Cell division phenotypes of W3110 over-expressing the various modified TolQ proteins are scored relative to (-), representing no TolQ over-expression (W3110+pBAD24) and (++++), representing W3110 over-expressing full length, unmodified TolQ (from pRA031). Colicins A and Ia spot titers, performed in RA1033 (ΔtolQ/ΔexbB/ΔexbD) are scored as the highest 5-fold dilution to show a zone of clearing. Cloudy zones of clearing are denoted with (*), representing either partial sensitivity or another undetermined effect of diminished overall cell health. To test OM integrity, 0.25% (w/v) DOC growth assays were also carried out using RA1033 cells expressing modified TolQ proteins. Results are scored as either (R) representing cells resistant to DOC, observed as an increase in cell density over time, or as (S) representing cells sensitive to DOC, observed as a decrease in cell density after the addition of DOC to the growth medium. All assays were performed in triplicate. The data shown here for filamentation assays from one replicate represents a consensus of data observed in all three trials. Colicins A and Ia and DOC sensitivity data are shown as absolute scores for each triplicate performed. 129
When modified by an N-terminal T7 tag addition, TolQ over-expression failed to induce
filamentation (Fig. 26). It is likely that the N-terminal amino acid addition interferes with the
interaction between TolQ and FtsN, possibly by inducing a conformational change in the TolQ
N-terminus. Colicin A spot titer assay results suggest that the N-terminal T7-tagged TolQ
protein becomes inserted in the CM, as cells expressing the protein show sensitivity to colicin A.
Colicin Ia spot titer assays suggest that not only does the T7-tagged TolQ become inserted in the
membrane, but also remains at least partially energetically-competent to support cross talk with
the TonB protein. Spot titers of both colicins A and Ia produced only partial clearings, but
nonetheless indicated functionality. In spot titer assays, cloudy zones can result from a number
of causes, from partial functionality to the previously discussed variability in expression from the
araBAD promoter (Siegele and Hu, 1997), to a general decrease in the viability of cells.
However, even the partial zones of clearing indicate a CM presence for the T7-tagged TolQ (Fig.
33). Interestingly, the N-terminally tagged TolQ was unable to support outer membrane
integrity, as seen in DOC growth assays. RA1033 cells (ΔtolQ/ΔexbB/ΔexbD) expressing this modified TolQ lysed upon the addition of DOC (Fig. 29). These results were consistent with a role for the TolQ N-terminus in the over-expression filamentation phenotype observed throughout this study.
In W3110 wild type E. coli, the over-expression of truncated versions of TolQ failed to induce filamentation. The over-expression of TolQ truncations of three and seven residues each produced similar results, with cells over-expressing the three-residue truncated TolQ appearing
only slightly longer than uninduced cells (Fig. 27). These results suggest that the extreme N-
terminus including residues 1-3 are necessary for a TolQ-FtsN interaction to occur. When
expressed in RA1033 cells (ΔtolQ/ΔexbB/ΔexbD) both truncated TolQ proteins conferred 130
variable sensitivities to both colicins A and Ia, as well as to DOC. TolQ lacking residues 1-3
conferred wild-type sensitivities to both colicins A and Ia (Fig. 34) and complemented mutants
with regard to resistance to DOC (Fig. 30). TolQ lacking residues 1-7, while conferring partial
sensitivity to colicin A, did not support sensitivity to colicin Ia (Fig. 34) or resistance to DOC
(Fig. 30), both believed to require energy of the pmf (Cascales et al., 2000; Germon et al., 2001;
Vankemmelbeke et al., 2009). Importantly, colicin A spot titer assays confirmed the presence of both truncated versions of TolQ in the CM.
Non-conservative amino acid substitutions at residues two, three, four, and twelve in
TolQ were examined for their effect, if any, on the ability of TolQ to induce filamentation when
over-expressed (Fig. 28). The substitutions of an alanyl for an aspartic residue at position three
and of a glutamic for lysyl residue at position 12 had no impact on the filamentation phenotype,
while a substitution of an alanyl for a threonyl residue at position two appeared to reduce the
degree of filamentation slightly. Conversely, the substitution of an arginyl for a methionyl
residue at position four eliminated the TolQ over-expression phenotype altogether. Wild type
cells over-expressing pMT019 (TolQMet4Arg) did not form filaments and appeared only slightly longer than uninduced cells (Fig. 28). The substitution at amino acid 2 replaced an uncharged, polar residue (Thr) with a non-polar residue (Ala). Alanyl residues are generally non-reactive and often considered “ambivalent”, found located in both the interior and on the surface of proteins, whereas threonyl residues are often located interiorly due to the increased hydrophobicity conferred by the presence of a methyl group. This substitution appeared to decrease but not to eliminate filamentation upon over-expression, indicating that the native
threonyl residue at position two in TolQ might participate in the interaction that leads to
filamentation when native TolQ is over-expressed or be important to the conformation of the 131
TolQ N-terminus. The replacement of the negatively charged aspartic residue at position 3 with
an alanyl appeared to have no effect on the degree of filamentation resulting from over-
expression of TolQ. While this substitution is non-conservative, both aspartic and alanyl
residues are often found located to helices, with alanyl reidues exhibiting less specificity for
location and aspartic residues often contributing to the solubility of proteins. Alanyl and and
aspartic residues share identical structures except for the replacement of the β-hydrogen in alanyl by a carboxylic acid group in aspartic residues. Likewise, the substitution of a glutamic residue for the lysyl residue at position twelve did not appear to reduce the degree of filamentation when the modified protein was over-expressed. This change from a positive to negatively charged residue conserved solubility at this position in TolQ. The substitution that had the greatest impact on filamentation was the substitution of an arginyl residue for the non-polar methionyl residue at position four. Methionyl residues are important in protein folding and are almost exclusively found in interior protein folds. Conversely, arginyl residues are almost exclusively located on the protein surface, often in the active site of proteins (Kander, 1996; McKee &
McKee, 2003; Berg et al., 2007). It may be that an effect on protein conformation results from this substitution. Interestingly, while it has been assumed throughout this study that filamentation does not require Tol-derived energy, the substitution at position four from methionyl to arginyl residues also rendered cells sensitive to DOC, while substitutions at residues two, three, and twelve did not appear to affect the integrity of the OM (Fig. 31). Similar conclusions can be drawn from results of colicins A and Ia spot titers. Non-conservative substitutions at residues two, three and twelve did not appear to substantially decrease sensitivity to either colicin A or colicin Ia, indicating that these modified proteins become both correctly inserted into the CM and are sufficiently energetically competent (Fig. 35). The substitution at 132
position four of arginyl for methionyl rendered cells sensitive to DOC (Fig. 31) and conferred partial resistance to both colicins A and Ia (Fig. 35). While it is possible that the substitution at position four causes a general decrease in the viability of cells, the combined results of filamentation, DOC, and colicins A and Ia spot titer assays would suggest that this substitution leads to a decrease in TolQ function or TolQ localization in the CM.
At face value, the results obtained from filamentation, DOC, and colicins A and Ia spot titer assays of variously modified TolQ proteins are informative and rather interesting. Because truncations of three and of seven residues at the N-terminus of TolQ lead to a loss of the over- expression filamentation phenotype, it might be that the presence of the residues at the extreme
N-terminus of TolQ are necessary for interaction with FtsN, and therefore in causing filamentation when unmodified TolQ is over-expressed. Furthermore, the N-terminal tag addition also eliminated filamentation, indicating that the interaction with FtsN could involve multiple residues and native conformation at the TolQ N-terminus that is not achieved when residues are either eliminated or added to TolQ. Substitutions at the TolQ N-terminus also suggest that the conformation of the 19 residue N-terminal region is important to TolQ’s potential role in cell division. Non-conservative substitutions at residues three and twelve did not eliminate filamentation when modified TolQ was over-expressed, while a substitution at position two appeared to decrease the degree of filamentation and another at position four eliminated it altogether. The observation that significantly altering individual residues at positions two and three does not appear to inhibit the TolQ interaction with FtsN, while their collective elimination does, further supports the notion that it is an overall conformation that is important for a TolQ-FtsN interaction. If conformation must be achieved for an interaction with 133
FtsN, it would appear that both amino acids two and four are important in achieving this
conformation and to the interaction.
Colicin A sensitivity spot titer assays provided evidence that all modified TolQ proteins used in this study became structurally present in the CM, at least to the degree of conferring colicin sensitivity. Considering an energy-dependent function of TolQ, as tested using colicin Ia sensitivity assays, each modified TolQ except for the seven-residue truncated form appeared to be energetically competent in supporting crosstalk with the TonB system. These modified TolQ proteins, when expressed in RA1033 (ΔtolQΔexbBΔexbD), conferred some degree of sensitivity to this group B colicin. The three-residue truncated TolQ and the TolQ with non-conservative substitutions at residues two, three, and twelve each appeared to confer wild type sensitivity to colicin Ia, while the N-terminal tag addition and the substitution at residue four led to the
production of cloudy clearings on spot titer assays. DOC assay results resembled colicin Ia spot
titer results, with those modified TolQ proteins exhibiting decreased or no sensitivity to colicin
Ia also showing complete or partial sensitivity to DOC; both truncated forms of TolQ as well as
the TolQMet4Arg appeared unable to compliment mutants in maintenance of the OM. Taken together, the results of assays performed in this dissertation would indicate that the extreme N- terminus (at least residues 1-7), in correct conformation, is required for the interaction with FtsN
that leads to filamentation when TolQ is over-expressed. Furthermore, this conformation or the
presence of residues one-seven, with emphasis on conservation of the methionine at residue four,
is required for the energy-dependent sensitivity to colicin Ia and maintenance of the E. coli OM.
In light of the results of experiments performed in this dissertation, the nature of the N-
terminus of TolQ should also be considered from a broader perspective. The relatively short N-
terminal tail of TolQ precedes the first of three membrane-spanning regions found within the 134
protein. Evidence indicates that this first transmembrane region (TM I) serves as a non-
cleavable, Sec-independent insertion sequence for localization of TolQ in the CM (Lewin and
Webster, 1996). TolQ may be one of the many integral membrane proteins targeted to the inner
bacterial membrane by the signal recognition particle (SRP) pathway, recognized and bound at
its hydrophobic TM1 by the SRP cotranslationally as it exits the ribosome (Luirink & Sinning,
2004; Wang & Dalby, 2011; Luirink et al., 2012). It has been reported that translocation of TM I
does not appear to require the pmf membrane potential but proper insertion of TM II and TM III
likely do (Lewin and Webster, 1996). However, from studies of other proteins with N-terminal
tails in the periplasm such as that of TolQ, it has been hypothesized that the membrane potential
facilitates translocation of negatively charged residues while hindering the translocation of
positively charged residues. Charge distribution along polytopic proteins is believed to have an
important role in the proper orientation of these proteins in the CM, with positive and negative
residues acting as translocation start and stop signals (Nilsson and von Heijne, 1990; Rohrer and
Kuhn, 1990; Anderson and von Heijne, 1994; Cao and Dalbey, 1994; von Heijne, 1994, 2006;
Whitley et al., 1994). With two positively charged lysine residues located to the N-terminal tail
(positions 12 and 18), this region of TolQ is still translocated to the periplasm. However, it is
worth considering that modifications in the N-terminal tail of TolQ that add additional positive
residues, such as with TolQMet4Arg, or those that eliminate negatively charged residues, such as the TolQ truncations TolQ3-230 and TolQ8-230, each having lost the aspartic residue at position three, might reduce the efficiency at which TolQ is inserted correctly in the CM. Furthermore, colicin A spot titer assays performed for modified TolQ proteins suggest that each modification tested becomes inserted in the CM and is functional, either comparable to or somewhat less than is observed for unmodified TolQ. In general, colicin sensitivity assays provide data that is 135
thought of as qualitative, as only a small number of molecules are believed to be required for killing sensitive cells. In the interpretation of results of colicin A spot titer assays in this study, it can be difficult to distinguish between qualitative or quantitative causes for what appears to be a decrease in colicin sensitivity. Is there a decrease in the proportion of TolQ proteins correctly inserted in the CM or are properly CM-located modified TolQ proteins less efficient at supporting sensitivity to colicin A than are unmodified proteins? While no modifications to
TM I were made during this study, alterations of the N-terminal tail could theoretically impact the efficiency at which modified TolQ proteins are inserted into the CM.
When data collected for Colicin Ia and DOC sensitivity assays are considered with those of filamentation and colicin A sensitivity assays, it is tempting to speculate that results for these two energy-dependent assays correlate with results of filamentation assays and with the presumably energy-independent assay for colicin A sensitivity. A number of lines of evidence have been used to determine that energization of TolA necessary for OM maintenence requires energy of the pmf and specific interactions between the transmembrane TM domains of TolQ,
TolR, and TolA (Kampfenkel and Braaun, 1993; Muller et al., 1993; Vianney et al., 1994;
Lazzaroni et al., 1995; Lloubes et al., 2001; Goemaere et al., 2006; Xiang et al., 2011).
Likewise, it is assumed that due to the high degree of homology between TM domains of TolQ and ExbB and between TolR and ExbD, interactions between these TM domains are important in crosstalk between the two systems (Braun, 1989; Braun and Hermann, 1993). In this study, this assumption is relevant to the abilities of modified TolQ proteins to confer sensitivity to colicin Ia in ΔexbB/ΔexbD cells. In this study, each modified TolQ protein was able to support sensitivity to colicin Ia through crosstalk except for the TolQ truncation TolQ8-230 (Fig. 34). Interestingly, this same truncated TolQ showed an average 5- to 25-fold decrease in sensitivity to colicin A and 136
an inability to cause filamentation when over-expressed (Figs. 27 and 34). Cells expressing this
seven-residue truncated TolQ were also sensitive to DOC (Fig. 30). TolQMet4Arg, when over- expressed in W3110 likewise failed to induce filamentation (Fig. 28) and when expressed in cells lacking ExbB and ExbD showed decreased sensitivity to colicin A at least five-fold less than that of cells expressing unmodified TolQ (Fig. 35). This TolQ substituted at position four, while also unable to function in OM maintenance, observed as sensitivity to DOC (Fig. 31), conferred sensitivity to colicin Ia at levels 25- to 125-fold below that of unmodified TolQ (Fig. 35).
Finally, the T7 tag addition to TolQ supported sensitivities to colicin A and colicin Ia reduced to levels 5- to 25-fold below that of unmodified TolQ (Fig. 33) but was unable to support the maintence of OM integrity (Fig. 29).
While it is difficult to draw clear conclusions based on the results of this study alone, these data indicate that the conformation of the N-terminal tail of TolQ is essential for the interaction between TolQ and FtsN and the subsequent disruption in cell division when TolQ is over-expressed. Furthermore, conformation might also contribute to colicin A sensitivity, as
TolQ modifications that eliminate the over-expression filamentation phenotype also result in a decreased sensitivity to colicin A. In this study, it was interesting to note that TolQ modifications that eliminated filamentation and reduced colicin A sensitivity also eliminated or reduced sensitivity to colicin Ia and rendered TolQ unable to support processes that maintain the
OM. As discussed above, it is also possible that particular modifications to the N-terminus of
TolQ decrease the efficiency at which TolQ is properly localized to the CM, causing an overall decrease in the performance of TolQ in phenotypic assays. This uncertainty might be clarified by quantification of TolQ protein from membrane fractions purified from cells expressing these modified proteins. However, the experimental purpose of generating these modified proteins 137
was to identify residues or regions of the TolQ N-terminus that are required for the interaction with FtsN that disrupts cell division when TolQ is over-expressed. If these modified TolQ proteins are indeed correctly targeted to the CM, results of this study would confirm earlier data from two-hybrid analysis that suggest an in vivo interaction between the N-terminus of TolQ and the periplasmic region of FtsN as well as the finding that concurrent over-expression of FtsN alleviates filamentation that occurs when TolQ alone is over-expressed. The most plausible conclusion based on cell division phenotypes observed following the over-expression of TolQ modified by an N-terminal addition, two truncations, and substitutions at residues two, three, four and twelve is that the conformation of the extreme N-terminus of TolQ, with specific requirement for a methionyl reside at position four and possible requirement for the threonyl residue at position two, is necessary for optimal interaction with FtsN. That the T7 tag addition to TolQ, which conserves the residues of the native extreme N-terminus, eliminated the ability of
TolQ to induce filamentation when over-expressed further suggests that the conformation of the
N-terminus is essential for filamentation. This result can be expected if the N-terminal tag addition prohibits the modified protein from achieving the native conformation.
138
CHAPTER V
SUMMARY AND GENERAL CONCLUSIONS
While Meury and Devilliers (1999) linked the TolA protein to proper positioning of the division site and Gerding et al. (2007) proposed a direct role for the Tol-Pal proteins as a subcomplex of the division machinery in Gram negative bacteria, the specific details of such a role for this ambiguously-described protein complex remain largely undefined. E. coli Tol mutants display a variety of phenotypes suggestive of disruption of outer membrane integrity.
Tol mutants form mucoid colonies, display increased sensitivity to DOC and high molecular mass antibiotics (Bernstein et al., 1972), leak periplasmic contents (Lazzaroni and Portalier,
1981), and shed outer membrane vesicles (Bernadac et al., 1998). When grown in media of high ionic strength or low osmolarity tol-pal mutants exhibit filamentous or chaining growth patterns
(Meury and Devilliers, 1999; Gerding et al., 2007). Similar phenotypes have been observed in tol-pal mutants of other proteobacterial species as well (Heilpern and Waldor, 2000; Llamas et
al., 2000; and Dubuisson et al., 2005). The model proposed by Gerding et al. (2007) and described in Chapter I provides a plausible mechanism to describe a role for the Tol system in E. coli during cell division. Several observations from our own studies of tol mutants remained consistent with those previously described. In these amorphic Tol mutants, cell division proceeds to completion and no filamentation was observed under standard conditions. For this dissertation, filamentation that occurs in E. coli cells when plasmid-encoded TolQ is expressed at levels above wild type was examined in detail. In order to determine the role of TolQ for this phenotype, the individual contributions of TolA, TolR, and TolQ with respect to this E. coli cell division mutant phenotype were explored. Deletion mutants of tolA, tolQ, and tolR exhibited no observable cell filamentation when grown under typical osmolarity conditions of 1% (w/v) NaCl 139
in E. coli (Chapter II, Fig. 8). Mutants do display a characteristic morphological phenotype observable as cells that are shorter and wider than wild type cells, but no chaining or filamentation is observed. However, mutants grown in the absence of added NaCl exhibit a chaining growth pattern. Individual cells are apparent in these “chains,” indicating that the cell division process is halted at a late stage, likely during septation of the OM (Chapter II, Fig. 8), similar to the observations of Meury and Devilliers (1999) and Gerding et al. (2007). Any observable differences between phenotypes in these studies and ours likely result from the “all- or-none” expression that results from the use of the araBAD promoter (Siegele and Hu, 1997), methods used to obtain images of cells (light microscopy versus confocal and electron transmission microscopy [Meury and Devilliers, 1999] or fluorescence and differential interference contrast (DIC) microscopy [Gerding et al., 2007]) as well as the specific types of mutants tested. While these results provided further evidence for a role for the Tol system in cell division, they also indicated the need for more extensive characterization of phenotypes associated with Tol system mutations and those that resulted from the over-expression of Tol
proteins.
When the TolQ protein was expressed at elevated levels, cell division impairment was
observed in wild type as well as in tolA, tolQ, and tolR mutant E. coli cells (Chapter II, Fig. 9).
This phenotype was not observed when cells over-expressed either the TolR or TolA proteins
(Chapter I, Fig. 4). This observation strongly suggests that it is the TolQ protein alone that
mediates the cell division disruption. Previous studies have revealed the appropriate extent of
arabinose induction necessary to restore wild-type associated functions in Tol system-deficient
mutants (Teleha, 2009). Therefore, in order to evaluate over-expression of these proteins,
arabinose-induction levels that were four magnitudes higher than levels necessary to simply 140
restore functionality associated with maintenance of OM integrity were used. When wild type cells expressing plasmid-encoded TolQ were grown at increasing arabinose-induction levels, the degree of filamentation also increased in severity (Chapter II, Fig. 11), demonstrating a direct correlation between TolQ expression levels and the filamentation phenotype. This assumption was confirmed by Western blot analysis (Chapter II, Fig. 11), which showed that cell elongation becomes visibly evident at the inducer concentration at which above wild type levels of TolQ are first detected in the cell. Increasing arabinose levels in cells carrying the pBAD24 vector alone resulted in no such response, suggesting this is a protein dosage effect that corresponds with the amount of TolQ present in the cell (data not shown).
It has been demonstrated previously that the tol-pal gene cluster is conserved among
Gram negative species and that the TolQ/TolR pair share homologues throughout the eubacteria, indicating a long evolutionary history for these proteins (Sturgis, 2001; Lazzaroni et al., 1999,
2002). The homology of this system extends beyond sequence and protein structure similarities to functional ones, since Tol mutants in numerous other species exhibit common phenotypes tied to cell envelope integrity, OM maintenance (Llamas et al., 2000; Sturgis, 2001; Dubuisson et al.,
2005; Yeh et al., 2010) and cell division (Meury and Devilliers, 1999; Heilpern and Waldor,
2000; Llamas et al., 2000; Dubuisson et al., 2005; Yeh et al., 2010). The homologous nature of this system would imply that similar roles for TolQ would exist in other Gram negative bacteria.
As might be expected, E. coli TolQ, over-expressed in E. amnigenus, C. muytjensii, and the E. coli strain BL21 resulted in a filamentation phenotype comparable to that observed in wild type
E. coli cells (Chapter II, Fig. 10), indicating that the TolQ over-expression phenotype is not an E. coli-specific phenomenon and is likely tied to the endogenous function of this protein. Because the Tol-Pal complex is widely conserved among Gram negative bacteria and TolQ and TolR are 141
the two most widespread proteins of the complex, sharing structural and functional homology
with ExbB/ExbD and MotA/MotB (Eick-Helmerich and Braun, 1989; Cascales et al., 2001), it is
likely that the Tol-Pal complex originated with the appearance of Gram negative bacteria. This
protein complex, while its true role in the cell remains elusive, is clearly required for the
maintenance of OM integrity (Lazzaroni et al., 1989), a role that can be understood as evidence
that the Tol-Pal complex evolved with the Gram negatives and in particular, the Gram negative
outer membrane. A logical extension to this aim would be to identify less closely-related Gram
negative species in which both TolQ and FtsN are conserved and over-express the native TolQ in
these species. An analysis of this sort might provide more information regarding the degree of
conservation for this link between the Tol-Pal protein complex and the cell division machinery.
In this study, two-hybrid analysis was used to evaluate the potential in vivo interaction
between TolQ and FtsN and to also determine the specific domains involved in such an
interaction. The analysis of protein-protein interactions between TolQ and FtsN did indeed
suggest that the potential for an in vivo interaction between the two proteins exists. Two pairings
in particular demonstrated a strong positive interaction acceptable above background growth
levels with respect to the various controls in both selective and dual selective screens. Both
screens indicated that an in vivo interaction would occur within the periplasmic space and
specifically involve the N-terminal domain of TolQ (TolQ1-19) and the proximal periplasmic
domain of FtsN (FtsN54-243), the same FtsN domain previously identified as necessary to support
proper cell division (Dai et al., 1996; Yang et al., 2004; Gerding et al., 2009) and to display high
sequence conservation among a number of bacteria (Yang et al., 2004). Although both the entire
periplasmic domain (FtsN54-319) and a truncated sequence of the periplasmic domain (FtsN54-243) demonstrated a significant positive interaction with the N-terminus of TolQ, the C-terminal 142
truncated FtsN consistently displayed weaker growth on dual-selective media relative to the full-
length construct (Chapter III, Fig. 21). In this screening system, the stronger the protein-protein
interaction, the more significant the growth on selective media, indicating a possible requirement
for the extreme C-terminal murein-binding domain in proper FtsN conformation and optimal
TolQ interaction. Results of two-hybrid analysis might be further confirmed through in vivo
chemical crosslinking analysis, which is often used to identify both transient and weak protein-
protein interactions. Not only could this method be used to confirm the in vivo interaction
between TolQ and FtsN, but it could also possibly show that increasing TolQ levels in the cell
are accompanied by increasing levels of TolQ-FtsN complexes. It is not possible to rule out that
further interaction between TolQ and FtsN exists between their transmembrane domains or
between cytoplasmic domains of one and periplasmic domains of the other, since these were not
analyzed. However, interactions between these regions are less likely given the membrane
topologies of TolQ and FtsN. To rule out interactions between non-topologically similar regions
of TolQ and FtsN, further two-hybrid analyses could be carried out using the bait and target
fusions constructed for this dissertation (Chapter III, Table 2). Additional two-hybrid analyses
might also be used to define more narrowly the region of FtsN that interacts with TolQ. Cloning
shorter FtsN fragments to serve as “target” fusion products, it might be possible determine
whether or not the region of FtsN previously identified by Gerding et al. (2009) as necessary for
cell division (FtsN71-105) is the same region that interacts with TolQ. Additionally, once more narrow regions of interaction are identified, amino acid substitutions within these bait and target regions followed by two-hybrid and/or over-expression analysis might also lead to the identification of residues involved in this interaction. Finally, the prediction that the over- expression of TolQ inhibits cell division by sequestration of FtsN was also supported by results 143
of this study. The concurrent over-expression of FtsN and TolQ using two different induction
methods alleviated the filamentation phenotype resulting from over-expression of TolQ alone, lending further support to an FtsN sequestration effect (Chapter III, Figs. 22 and 23).
In a final aim of this dissertation, TolQ proteins with N-terminal modifications were examined for their effect on filamentation upon their over-expression. Results from these
analyses in Chapter IV indicated that a truncation of as little as only three residues eliminates the
ability of TolQ to cause filamentation (Chapter IV, Fig. 27). A non-conservative substitution at
residue two decreased the degree of filamentation achieved when the modified TolQ was over-
expressed, while another at residue four eliminated filamentation altogether (Chapter IV, Fig.
28). These results could be interpreted as a necessity for the threonyl at position two and the
methionyl at position four in TolQ for filamentation. However, because the addition of a T7 tag
to the otherwise unmodified TolQ N-terminus also eliminated the TolQ over-expression
phenotype, it is more likely that the native extreme N-terminal conformation of TolQ is a
prerequisite for interaction with FtsN (Chapter IV, Fig. 26). Colicin A spot titer assays
confirmed the physical presence of each modified TolQ in the CM, as tol mutants are resistant to
Group A colicins (Nagel de Zwaig and Luria, 1967), and each modified TolQ protein
complimented mutants by restoring colicin A sensitivity (Chapter IV, Figs. 33-35). The T7 tag
addition, the seven-residue truncation, and the non-conservative substitution at position four in
TolQ either eliminated or decreased sensitivity to colicin Ia when expressed in
ΔtolQ/ΔexbBΔ/exbD cells. Expression of the T7-tagged TolQ, the 7 residue truncated TolQ, and
the non-conservative residue 4-substituted TolQ in these mutants indicated that each modified
protein was also unable to restore OM integrity in mutants. That the results of these two energy-
dependent phenotypic assays nearly mirror those obtained for filamentation and colicin A spot 144
titer assays provides a new view of the function of TolQ and the Tol-Pal system as a whole.
Because it has been repeatedly reported that transmembrane interactions between TolQ, TolR, and TolA are required for functional assembly of the Tol complex (Kampfenkel and Braun,
1993; Lazzaroni et al., 1995; Germon et al., 1998; Braun and Herrmann, 2004; Goemaere et al.,
2006) and no modifications were made to any Tol protein transmembrane domains in this study, the effects of the TolQ modifications described above on OM integrity and colicin Ia sensitivity provided interesting results. A possible future direction from this study might be to further characterize these modified TolQ proteins in regard to functional assays. First, it would be necessary to ascertain whether or not the altered proteins become inserted into the CM at decreased efficiency and in turn, decrease TolQ performance in phenotypic assays. Further experimentation, such as Western blot analysis of the membrane fraction of proteins in cells expressing these modified TolQ proteins compared to that from cells expressing unmodified protein, might distinguish if results obtained in Chapter IV of this dissertation are qualitative or quantitative. If the data is qualitative, resulting from a general decrease in TolQ functionality due to the modifications themselves, these constructs could be used to further characterize the particular role of the TolQ N-terminus in cell division, group A and group B colicin sensitivity, and maintenance of the OM. C-terminal truncations, as described below, might also be used for these studies.
Another possible future study of the N-terminus of TolQ would be a more narrow analysis of the protein’s N-terminal region in filamentation assays. Lewin and Webster (1996) demonstrated that a C-terminal truncated version of TolQ, containing the 19 residue N-terminus,
TM I (approx. residues 20-36) and a large portion of the cytoplasmic loop following TM I
(approx. residues 36-128), is inserted in the CM in the proper orientation in a Sec-independent 145
manner. It was concluded that the large cytoplasmic loop likely directs membrane orientation of the N-terminal region of TolQ, as a truncated version containing only residues 1-36 was inserted into the CM in a Sec-dependent manner in reverse orientation (Lewin and Webster, 1996).
Truncated versions of TolQ, specifically those lacking TM II and TM III could be used to further assess the filamentation phenotype in regard to its dependence upon the TolQ N-terminus. These types of analyses would potentially confirm that the role of TolQ in cell division is indeed independent of the TM interactions believed to be required for energy-dependent Tol-Pal functions (Lazzaroni et al., 1995; Goemaere et al., 2006).
Although FtsN joins the septal ring near the end of its assembly, FtsN depletion leads to disassembly of the early divisome components of the proto-ring, indicating a role for FtsN in stabilization of the division ring. Restored FtsN levels back-recruit early divisome complexes
(Rico et al., 2010) and a number of divisome components have been implicated in sharing a role in stabilization of the divisome, with functional overlap occurring among many (Geissler and
Margolin, 2005; Rico et al., 2010). This is evidenced by the ability of altered or highly expressed divisome proteins to compensate for the loss of others (Geissler and Margolon, 2005;
Bernard et al., 2007). These findings would suggest that divisome assembly and function is more dynamic and cooperative than originally described. Furthermore, interactions among divisome components and between divisome proteins and a growing number of non-essential proteins point to a more complex and cooperative nature for the cell division process itself. In a review of the information available regarding the assembly of the bacterial division apparatus,
Vicente et al. (2006) suggested just that. Based on widespread comparisons between bacterial genomes, the distribution pattern of division genes among different genomes is variable, with only six essential divisome proteins found to be widely conserved (Vicente et al., 2006). Cell 146
division proteins designated as essential in E. coli are found variably distributed outside the γ- proteobacteria (Möll and Thanbichler, 2009). These findings led the authors to suggest that the cell division machinery is flexible and has evolved to operate efficiently within different bacterial groups in accordance with each group’s specific cell envelope, shape, and life cycle needs (Vicente et al., 2006).
If the cell division machinery is indeed ancient and appeared in a precursor to modern bacteria, it should come as no surprise that more thorough investigation has identified functional homologs of FtsN outside the γ-proteobacteria, the class of bacteria to which FtsN sequence homology appears to be limited. Möll and Thanbichler (2009) have reported FtsN-like homologs in α-, β-, and δ-proteobacteria, in Caulobacter crescentus, Burkholderia thailandensis and
Myxococcus xanthus, respectively. Moreover, in B. thailandensis and M. xanthus, these FtsN homologs were shown to localize to the division site. In C. crescentus the FtsN homolog is essential for cell division and viability (Möll and Thanbichler, 2009). Interestingly, C. crescentus lacks Lpp, which in E. coli serves as an important bridge between the IM and OM to maintain cell envelope integrity (Weigand et al., 1976). In C. crescentus, the Tol-Pal complex is essential for OM integrity as well as invagination of the OM during late cell division (Yeh et al.,
2010). It may be that in C. crescentus, the Tol-Pal complex is essential because it serves as a primary bridge between the CM and OM in both dividing and non-dividing cells, a role that is fulfilled by Lpp in E. coli and other enterics (Yeh et al., 2010). It is apparent from these relatively recent studies of bacterial cell division proteins across a wider range of lineages that the cell division machinery, including the Tol-Pal proteins, is most likely as varied and evolved as the diverged lineages within which they are found. 147
In a 2004 analysis of FtsN, Ursinus et al. examined the functional properties of FtsN domains. It was previously determined that the cytoplasmic N-terminal domain (residues 1-33) and the TM helix (residues 34-53) are not essential for cell division (Yang et al., 2004) but likely serve roles in membrane localization and stabilization of the protein and can be replaced by similar CM anchors (Dai et al., 1996; Yang et al., 2004). Targeting of FtsN to the divisome requires the periplasmic domain of the protein (Addinall et al., 1997). A largely unstructured periplasmic region following the TM helix (residues 53-137) appears to be essential for cell division (Ursinus et al., 2004). The region of FtsN essential to cell division was further defined by Gerding et al. (2009) to at minimum include residues 71-105. The FtsN C-terminal 77 residues form a folded globular domain that has been identified as a murein binding domain that binds cell wall material during cell division. This murein binding domain is required for localization to the constriction sites in dividing cells in a self-enhancing manner to interact with the murein transpeptidase penicillin-binding protein 3 (PBP3), also known as FtsI (Gerding et al., 2009). Accumulation of FtsN at the division site then recruits the amidase AmiC (Barnhardt and de Boer, 2003) that remodels the newly synthesized cell wall material. However, it has been shown that the peptidoglycan binding ability of FtsN is not essential for cell division. It may be that a secondary role for FtsN in cell division is to bind to new septal murein, stabilizing the division process, while an essential but yet unknown role is mediated by the CM-proximal periplasmic region (Ursinus et al., 2004). That FtsN is present in the cell at high copy numbers relative to all other division proteins except for FtsZ, which is present at approximately 20,000 copies per cell, suggests that FtsN might have a more general role in peptidoglycan-binding that is unrelated to cell division, since in non-dividing cells, the proteins are distributed around the periphery of the cell and only accumulate at the division septum just before the onset of cell 148
constriction. FtsN is present at 3,000-6,000 copies per cell, while other divisome proteins are carried at copy numbers between 20 and 40 per cell (Bi and Lutkenhaus, 1991). Considering these findings and the functional overlap observed between many divisome components, it is reasonable to consider the possibility that, while FtsN is required for cell division in a number of bacterial species including E. coli, the divergent nature of the cell division machinery might suggest that FtsN and FtsN-like proteins have evolved to fulfill a number of roles in the cell related to OM integrity as well. In particular, FtsN may contribute to maintence of the OM by bridging the OM and CM in both dividing and non-dividing cells. As FtsN sequence conservation is very poor outside the γ-proteobacteria (Möll and Thanbichler, 2009), FtsN-like functional “homologs” might very well have arisen in an analogous manner to support the evolution of the OM over time. Including FtsN, there are four proteins in E. coli that each possess a murein-binding domain and transiently localize to the divisome in dividing cells which appear to contribute to the constriction process, but whose functions remain largely unknown. It is also predicted that over 1,600 proteins from over 500 bacterial species possess these murein- binding domains (Finn et al., 2008). It is plausible that proteins containing murein binding domains, such as FtsN, serve among a growing number of cell envelope-associated proteins that might tether the OM to underlying peptidoglycan and the CM in Gram negative bacteria during non-dividing stages of the bacterial life cycle and localize to the constriction site in a cooperative manner to facilitate the cell division process.
It is also reasonable to consider a dual role for the Tol-Pal complex in the cell. A role for the Tol system in OM maintenance has clearly been established in a number of Gram negative species. Much less is known about the role of the Tol system in cell division. The phenotype observed in cells over-expressing TolQ is distinct from that of Tol mutants grown under low salt 149
conditions. Over-expressed TolQ leads to the formation of long smooth filaments, a characteristic phenotype of ftsN mutants. The chaining phenotype observed in Tol mutants grown at low salt concentrations observed in this study and as reported by Meury and Devilliers
(1999) and Gerding et al. (2007) may be the result of a response to osmotic stress that is exacerbated by the absence of a functional Tol system. Cell filamentation has long been known to be a stress response to unfavorable growth conditions such as nutrient limitation (Young,
2006). Gerding et al (2007) provided evidence that this phenotype is rescued by plasmid- expressed Tol proteins. Under standard conditions, tol mutants continue to divide, ruling out an essential role for the Tol-Pal complex as a core component of the divisome. However, an accessory role for the Tol system in divisome stability is a very plausible physiological role for this system. The Tol-Pal system links the IM and OM through a TolA-Pal interaction (Cascales et al., 2000), during both non-dividing and dividing stages of the bacterial life cycle (Rodriquez-
Herva et al. 1996; Heilpern and Waldor, 2000; Dubuisson et al., 2005; Yeh et al., 2010). During both stages, this link would have been advantageous throughout the evolution of the Gram negatives. The ability of the Tol-Pal complex to interact with divisome components, in particular TolQ to interact with FtsN or FtsN-like proteins, would also have been advantageous, as such an interaction might provide additional stability to the divisome while physically linking the OM to the underlying peptidoglycan and CM during cell division, facilitating a more efficient cell division process through cooperative binding. That TolQ and TolR are the most conserved Tol proteins, found throughout the eubacteria, as well as in an archean (Sturgis, 2001), provides evidence for the ancient origin of these genes. TolA homologs are much more difficult to identify because sequence conservation is very poor between genomes. Either the TolA gene is highly divergent or TolA homologs are not homologs at all but rather are analogs that have 150
evolved separately in various lineages to interact with the more ancient TolQ/TolR pair to
support the Gram negative OM. Because the TolQ/TolR pair is so widely conserved suggests a
possible alternative function for these proteins that led to their widespread distribution across the
Gram negative phylogeny (Sturgis, 2001). Additionally, because ybgC and ybgF genes are evolutionary latecomers to the tol-pal operon, appearing with the emergence of the α, β- and γ-
proteobacteria, the early beginnings of Tol-Pal complex can be traced to at minimum a protein
pair ancestral to TolQ/TolR, ExbB/ExbD, and MotA/MotB that provided some benefit to an
early double-membrane bacteria and may have evolved to give rise to three functionally distinct
“molecular motors” whose common ancestry is still evident through sequence, structural, and
functional comparisons (Eick-Helmerick and Braun, 1989; Cascales et al., 2001). It is logical to
infer that as new Gram negative lineages emerged, these TolQ/TolR ancestors took on new
partners and assumed additional roles within each lineage in support of the evolving OM.
Evidence indicates the Tol proteins interact through their transmembrane domains to collectively
perform energy-dependent functions in the cell (Germon et al., 2001) and that a necessary
stoichiometric balance of Tol system components preserves this functionality (Guihard et al.,
1993). While the present study does not specifically address whether or not the energy-
transducing functions of the Tol system contribute to the filamentation phenotype observed in
cells over-expressing TolQ, this can be inferred from analysis of the data. Even in cells
completely lacking one or more Tol system proteins and ExbB/ExbD, cell division still occurs.
The details of Tol complex action at the division site remain to be determined, including the
potential role of Tol-Pal-derived energy in cell division. However, the results of this study
provide physical support for previous assertions that the Tol-Pal complex, specifically TolQ,
does indeed directly interact with the cell division machinery in E. coli and its relatives. Two- 151
hybrid analyses have identified the domains of interaction between TolQ and FtsN and provided evidence that this interaction is substantial enough to interfere with the cell division process when TolQ is present in excess. Through dual over-expression experiments, it was also shown that this disruption in cell division can be alleviated with simultaneous over-expression of FtsN.
The mechanism by which the over-expression of TolQ and its interaction with FtsN disrupt the cell division process remain to be clarified, perhaps as the physiological function(s) of FtsN become better understood and the specific role(s) for the TolQ and the Tol-Pal complex during cell division are clarified.
152
REFERENCES Abelson, J.N., Simon, M.I., Deutscher, M.P. 1990. Guide to protein purification. Methods Enzymol. 182:78-79. Adams, D.A., Errington, J., 2009. Bacterial cell division: assembly, maintenance and disassembly of the Z ring. Nature Reviews: Microbiol. 7:642-653. Addinall, S.G., Cao, C., Lutkenhaus, J. 1997. FtsN, a late recruit to the septum in Escherichia coli. Mol. Microbiol. 25:303-309. Anderson, H., von Heijne, G. 1994. Membrane protein topology: effects of ΔµH+ on the translocation of charged residues explain the "Positive Inside" rule. EMBO J. 13:2267- 2272. Andersson, S.G., Zomorodipour, A., Andersson, J.O., Sicheritz-Ponten, T., Alsmark, U.C., Podowski, R.M.M., Naslund, A.K., Eriksson, A.S., Winkler, H.H., Kurland, C.G. 1998. The genome sequence of Rickettsia prowazekii and the origin of mitochondria. Nature. 396:133-140. Begley, M., Gahan, C.G.M., Hill, C. 2005. The interaction between bacteria and bile. FEMS Microbiol. Rev. 29:625-651. Berg, J.M., Tymoczko, J.L., Stryer, L 2007. Protein Composition and Structure. In: Biochemistry (6th Edition) Berg, J.M., Tymoczko, J.L., Stryer, L. (Eds) pp 25-60. W.H. Freeman and Company, New York. Bernadac, A.M., Gavioli, J.C., Lazzaroni, J.C., Raina, S., Lloubes, R. 1998. Escherichia coli tol- pal mutants form outer membrane vesicles. J. Bacteriol. 180:4872-4878. Bernhardt, T.G., de Boer, P. A. J. 2003. The Escherichia coli amidase AmiC is a periplasmic septal ring component exported via the twin-arginine transport pathway. Mol. Microbiol. 48:1171-1182. Bernhardt, T.G., de Boer, P.A. 2005. SlmA, a nucleoid-associated, FtsZ binding protein required for blocking septal ring assembly over chromosomes in E. coli. Mol. Cell 18:555-564. Bernstein, A., Rolfe, B., Onodera, K. 1972. Pleiotropic properties and genetic organization of the tolA, B locus of Escherichia coli K-12. J. Bacteriol. 112:74-83. Bi, E., Lutkenhaus, J. 1991. FtsZ ring structure associated with division in Escherichia coli. Nature. 354:161-164. Bourdineaud, J.P., Boulanger, P., Lazdunski, C., Letellier, L. 1990. In vivo properties of colicin A: Channel activity is voltage dependent but translocation may be voltage independent. Proc. Natl. Acad. Sci. USA. 87:1037-1041. Bowe, F., Lipps, C. J., Tsolis, R. M., Groisman, E., Heffron, F., Kusters, J. G. 1998. At least four percent of the Salmonella typhimurium genome is required for fatal infection of mice. Infect Immun. 66:3372–3377. 153
Bradbeer, C. 1993. The proton motive force drives the outer membrane transport of cobalamin in Escherichia coli. J. Bacteriol. 175:3146-3150. Braun, V. 1975. Covalent lipoprotein from the outer membrane of Escherichia coli. Biochim. Biophys. Acta. 415:335-377. Braun, V. 1989. The structurally related exbB and tolQ genes are interchangeable in conferring tonB-dependent colicin, bacteriophage, and albomycin sensitivity. J. Bacteriol. 171:6387-6390. Braun, V. 1995. Energy-coupled transport and signal transduction through the gram-negative outer membrane via TonB-ExbB-ExbD-dependent receptors. FEMS Microbiol. Rev. 16:295-307. Braun, V., Herrmann, C. 1993. Evolutionary relationship of uptake systems for biopolymers in Escherichia coli: cross-complementation between the TonB-ExbB-ExbD and the TolA- TolQ-TolR proteins. Mol. Micriobiol. 8:261-268. Braun, V., Herrman, C. 2004. Point mutations in transmembrane helices 2 and 3 of ExbB and TolQ affect their activities in Escherichia coli K-12. J. Bacteriol. 186:4402-4406. Braun, V., Pilsl, H., Groß, P. 1994. Colicins: structures, modes of actions, transfer through membranes, and evolution. Arch. Microbiol. 161:199-206. Brinkman, K.K. 2007. Using chimeras to evaluate cross-talk, energy transfer, and protein- protein interactions in the TonB and TolA systems. Doctoral Dissertation, Bowling Green St. University. Brinkman, K.K., Larsen, R.A. 2008. Interactions of the energy transducer TonB with noncognate energy-harvesting complexes. J. Bacteriol. 190:421-427. Cao, G., Dalbey, R. 1994. Translocation of amino-terminal tails across the plasma membrane. EMBO J. 13:4662-4669. Cascales, E., Bernadac, A., Gavioli, M., Lazzaroni, J.C., Lloubes, R. 2002. Pal lipoprotein of Escherichia coli plays a major role in outer membrane integrity. J. Bacteriol. 184:754- 759. Cascales, E., Buchanan, S.K., Duché, D., Kleanthous, C., Lloubès, R., Postle, K., Riley, M., Slatin, S., Cavard, D. 2007. Colicin biology. Microbiol. Mol. Biol. Rev. 71:158-229.
Cascales, E., Gavioli, M., Sturgis, J.N., Loubes, R. 2000. Proton motive force drives the interaction of the inner membrane TolA and outer membrane Pal proteins in Escherichia coli. Mol. Microbiol. 38:904-915. Cascales, E., Lloubes, R., Sturgis, J.N. 2001. The TolQ-TolR proteins energize TolA and share homologies with the flagellar motor proteins MotA-MotB. Mol. Microbiol. 42:795-807.
154
Casjens, S., Palmer, N., van Vugt, R., Huang, W.M., Stevenson, B., Rosa, P., Lathigra, R., Sutton, G., Peterson, J., Dodson, R.J., Haft, D., Hickey, E., Gwinn, M., White, O., Fraser, C.M. 2000. A bacterial genome in flux: the twelve linear and nine circular extrachromosomal DNAs in an infectious isolate of the Lyme disease spirochete Borrelia burgdorferi. Mol. Microbiol. 35:490-516. Chimerel, C., Field, C.M., Pinero-Fernandez, S., Keyser, U., Summers, D. 2012. Indole prevents Escherichia coli cell division by modulating membrane potential. Biochimica et Biophysica Acta. 1818:1590-1594. Chung, C.T., Niemela, S.L., Miller, R.H. 1989. One-step preparation of competent Escherichia coli; transformation and storage of bacterial cells in the same solution. Proc. Natl. Acad. Sci. USA. 86:2172-2175. Clavel, T., Germon, P., Vianney, A., Portalier, R., Lazzaroni, J.C. 1998. TolB protein of Escherichia coli K-12 interacts with the outer membrane peptidoglycan-associated proteins Pal, Lpp, and OmpA. Mol. Microbiol. 29:359-367. Clontech Laboratories, Inc. Mountain View, CA, 94053, USA. Cytrynbaum, E.N., Li,.D., Allard, J.F., Mehrabian, H. 2012. Estimating the bending modulus of a FtsZ bacterial-division protein filament. Phys. Rev. 85:011902. Daegelen, P., Studier, F., Lenski, R., Cure, S., Kim, J. 2009. Tracing ancestors and relatives of Escherichia coli B, and the derivation of B strains REL606 and BL21 (DE3). J. Mol. Biol. 394:634-643. Dai, K., Xu, Y., Lutkenhaus, J. 1993. Cloning and characterization of ftsN, an essential cell division gene in Escherichia coli isolated as a multicopy suppressor of ftsA12(Ts). J. Bacteriol. 175:3790-3797. Dai, K., Xu, Y., Lutkenhaus, J. 1996 Topological characterization of the essential Escherichia coli cell division protein FtsN. J. Bacteriol. 178:1328-1334. Datensko, K.A., Wanner, B.L. 2000. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA. 97:6640-6645. de Boer, P.A.J. 2010. Advances in understanding E. coli cell fission. Curr. Opin. Microbiol. 13:730-737. de Pedro, M.A., Quintela, J.C., Holtje, J-V., Schwartz, H. 1997. Murein segregation in Escherichia coli. J. Bacteriol. 179:2823-2834. Deich, R.A., Metcalf, B.J., Finn, C.W., Farley, J.E., Green, B.A. 1998. Cloning of genes encoding a 15,000-dalton peptidoglycan-associated outer membrane lipoprotein and an antigenically related 15,000-dalton protein from Haemophilus influenzae. J. Bacteriol. 170:489-498. Dennis, J.J., Lafontaine, E.R., Sokol, P.A. 1996. Identification and characterization of the tolQRA genes of Pseudomonas aeruginosa. J. Bacteriol. 178:7059-7068. 155
Derouiche, R., Bénédetti, H., Lazzaroni, J., Lazdunski, C., Lloubes, R. 1995. Protein complex within Escherichia coli inner membrane. J. Biol. Chem. 270:11078-11084. Dubuisson, J.F., Vianney, A., Hugouvieux-Cotte-Pattat, N., Lazzaroni, J.C. 2005. Tol-Pal proteins are critical cell envelop components of Erwinia chrysanthemi affecting cell morphology and virulence. Microbiology. 151:3337-3347. Eick-Helmerick, K., Braun, V. 1989. Import of biopolymers into Escherichia coli: Nucleotide sequences of the exbB and exbD genes are homologous to those of the tolQ and tolR genes, respectively. J. Bacteriol. 171:5117-5126. Eurofins MWG Operon. Huntsville, AL, 35805, USA. Farmer, J.J., III, Ashbury, M.A., Hickman, F.W., Brenner, D.J. The Enterobacteriaceae Study Group (USA). 1980. Enterobacter sakazakii: a new species of “Enterobacteriaceae” isolated from clinical speciemens. Int. J. Syst. Bacteriol. 30:569-584. Fields S, Song O. 1989. A novel genetic system to detect protein-protein interactions. Nature 340:245–246. Finn, R.D., Tate, J., Mistry, J., Coggoll, P.C., Sammut, S.J., Hotz, H.R., Ceric, G., Forslund, K., Eddy, S.R., Sonnhammer, E.L., Bateman, A. 2008. The Pfam Protein Families Database. Nucleic Acids Res. 36:D281-D288. Fisher Scientific Company L.L.C. Pittsburg, PA, 15275, USA. Fraser, C.M., Casjens, S., Huang, W.M., Sutton, G.G., Clayton, R., Lathigra, R., White, O., Ketchum, K.A., Dodson, R., Hickey, E.K., Gwinn, M., Dougherty, B., Tomb, J.F., Fleischmann, R.D., Richardson, D., Peterson, J., Kerlavage, A.R., Quackenbush, J., Salzberg, S., Hanson, M., van Vugt, R., Palmer, N., Adams, M.D., Gocayne, J., Weidman, J., Utterback, T., Watthey, L., McDonald, L., Artiach, P., Bowman, C., Garland, S., Fujii, C., Cotton, M.D., Horst, K., Roberts, K., Hatch, B., Smith, H.O., Venter, J.C. 1997. Genomic sequence of a Lyme disease spirochaete, Borrelia burgdorferi. Nature. 390:580-586. Fraser, C.M., Norris, S.J., Weinsock, G.M., White, O., Sutton, G.G., Dodson, R., Gwinn, M., Hickey, E.K., Clayton, R., Ketchum, K.A., Sodergren, E., Hardham, J.M., McLeod, M.P., Salzberg, S., Peterson, J., Khalak, H., Richardson, D., Howell, J.K., Chidambaram, M., Utterback, T., McDonald, L., Artiach, P., Bowman, C., Cotton, M.D., Fijii, C., Garland, S., Hatch, B., Horst, K., Roberts, K., Sandusky, M., Weidman, J., Smith, H.O., Venter, J.C. 1998. Complete genome sequence of Treponema pallidum, the syphilis spirochete. Science. 281:375-388. Ganzle, M.G., Hertel, C., van der Vossen, J.M.B.M., Hammes, W.P. 1999. Effect of bacteriocin- producing lactobacilli on the survival of Escherichia coli and Listeria in a dynamic model of the stomach and the small intestine. Int. J. Food Microbiol. 48:21-35. 156
Gaspar, J.A., Thomas, J.A., Marolda, C.L., Valvano, M.A. 2000. Surface expression of O- specific lipopolysaccharides in Escherichia coli requires the function of the TolA protein. Mol. Microbiol. 38:262-275. Geissler, B., Margolin, W. 2005. Evidence for functional overlap among multiple bacterial division proteins: compensating for the loss of FtsK. Mol. Mocrobiol. 58:596-612. Gerding, M.A., Ogata, H., Pecora, N.D., Niki, H., de Boer, P.A.J. 2007. The trans-envelope Tol- Pal complex is part of the cell division machinery and required for proper outer- membrane invagination during cell constriction in E. coli. Mol. Microbiol. 63:1008-1025. Gerding, M.A., Liu, B., Bendezu, F.O., Hale, C.A., Bernhardt, T.G., de Boer, P.A.J. 2009. Self- enhanced accumulation of FtsN at division sites and roles for other proteins with a SPOR domain (DamX, DedD, and RlpA) in Escherichia coli cell constriction. J. Bacteriol. 191:7383-7401. Germon, P., Clavel, T., Vianney, A., Portalier, R., Lazzaroni, J.C. 1998. Mutational analysis of the Escherichia coli K-12 TolA N-terminal region and characterization of its TolQ- interacting domain by genetic suppression. J. Bacteriol. 180:6433-6439.
Germon, P., Ray, M.C., Vianney, A., Lazzaroni, J.C. 2001. Energy-dependent conformational change in the TolA protein of Escherichia coli involves its N-terminal domain, TolQ, and TolR. J. Bacteriol. 183:4110-4114. Goehring, N.W., Beckwith, J. 2005. Diverse paths to midcell: assembly of the bacterial cell division machinery. Curr. Biol. 15:514-526. Goehring, N.W., Gonzalez, M.D., Beckwith, J. 2006. Premature targeting of cell division proteins to midcell reveals hierarchies of protein interactions involved in divisome assembly. Mol. Microbiol. 61:33-45. Goehring, N.W., Robichon, C., Beckwith, J. 2007. Role for the nonessential N terminus of FtsN in divisome assembly. J. Bacteriol. 189:646-649. Goemaere, E.L., Cascales, E., Lloubes, R. 2007. Mutational analyses define helix organization and key residues of a bacterial membrane energy-transducing complex. J. Mol. Biol. 366:1424-1436. Gresock, M.G., Savenkova, M.I., Larsen, R.A., Ollis, R.A., Postle, K. 2011. Death of the TonB shuttle hypothesis. Front. Microbio. 2:206. Doi:10.3389/fmicb.2011.00206. Guihard, G., Boulanger, P., Bénédetti, H., Lloubès, R., Besnard, M., Letelliers, L. 1994. Colicin A and the Tol proteins involved in its translocation are preferentially located in the contact sites between the inner and outer membranes of Escherichia coli cell. J. Biol Chem. 269:5874-5880. Guzman, L.M., Belin, D., Carson, M.J., Beckwith, J. 1995. Tight regulation, modulation, and high-level expression by vectors containing the arabinose P-BAD promoter. J. Bacteriol. 177:1739-1747. 157
Heilpern, A.J., Waldor, M.K. 2000. CYXphi infection of Vibrio cholerae requires the tolQRA gene products. J Bacteriol. 182:1739-1747. Higgs, P.I., Myers, P.S., Postle, K. 1998. Interactions in the TonB-dependent energy transduction complex: ExbB and ExbD form homomultimers. J. Bacteriol. 180:6031- 6038.
Higgs, P.L., Letain, T.F., Merriam, K.K., Burke, N.S., Park, H., Kang, C., Postle, K. 2002. TonB interacts with nonreceptor proteins in the outer membrane of Escherichia coli. J. Bacteriol. 184:1640-1648. Hill, C.W., Harnish, B.W. 1981. Inversions between ribosomal RNA genes of Escherichia coli. Proc. Natl. Acad. Sci. USA. 78:7069-7072. Hu, Z., Mukherje, A., Pichoff, S., Lutkenhaus, J. 1999. The MinC component of the division site selection system in Escherichia coli interacts with FtsZ to prevent polymerization. Proc. Natl. Acad. Sci. USA. 96:14819-14824. Hu, Z., Saez, C., Lutkenhaus, J. 2003. Recruitment of MinC, an inhibitor of Z-ring formation, to the membrane in Escherichia coli: role of MinD and MinE. J. Bacteriol. 185:196-203. Iversen, C., Mullane, N., McCardell, B., Tall, B.D., Lehner, A., Fanning, S., Stephan, R., Joosten, H., Iversen, C. 2008. Cronobacter gen. nov., a new genus to accommodate the biogroups of Enterobacter sakazakii, and proposal of Cronobacter sakazakii gen. nov. comb. nov., C. malonaticus sp. nov., C. turicensis sp. nov., C. muytjensii sp. nov., C. dublinensis sp. nov., Cronobacter genomospecies 1, and of three subspecies, C. dublinensis sp. nov. subsp. dublinensis subsp. nov., C. dublinensis sp. nov. subsp. lausannensis subsp. nov., and C. dublinensis sp. nov. subsp. lactaridi subsp. nov.. IJSEM 58:1442–1447. Izard D., Gavini, F., Trinel, P.A., LeClerc, H. 1981. Deoxyribunucleic acid relatedness between Enterobacter cloacae and Enterobacter amnigenus sp. nov. Int. J. Syst. Bacteriol. 31:35- 42. Journet, L., Rigal, A., Lazdunski, C., Bénédetti, H. 1999. Role of TolR N-terminal, central, and C-terminal domains in dimerization and interaction with TolA and TolQ. J. Bacteriol. 181:4476-4484.
Kadner, R.J. 1996. Cytoplasmic membrane. In: Escherichia coli and Salmonella, Cellular and Molecular Biology (2nd Edition). Neidhart, F.C. Curtiss III, R., Ingraham, J.L., Lin, E.C.C., Low, K.B., Magasanik, B., and Umbarger, H.E. (Eds.) pp 58-87. ASM Press, Washington, DC.
Kamio, Y., Nikaido, H. 1976. Outer membrane of Salmonella typhimurium: accessibility of phospholipid head groups to phospholipase c and cyanogen bromide activated dextran in the external medium. Biochemistry 15:2561–2570.
158
Kampfenkel, K., Braun, V. 1993. Membrane topologies of the TolQ and TolR proteins of Escherichia coli: Inactivation by a missense mutation in the proposed first transmembrane segment. J. Bacteriol. 175:4485-4491.
Khlebnikov, A., Risa, O., Skaug, T., Carrier, T.A., Keasling, J.D. 2000. Regulatable arabinose- inducible gene expression system with consistent control in all cells of a culture. J. Bacteriol. 182:7029-7034. Kleanthous, C. 2010. Swimming against the tide: progress and challenges in our understanding of colicin translocation. Nat. Rev. Microbiol. 8:843-848.
Krachler, A., Sharma, A., Cauldwell, A., Papadakos, G., and Kleanthous, C. 2010. TolA modulates the oligomeric status of YbgF in the bacterial periplasm. J. Mol. Biol. 403:270-285.
Lan, G., Daniels, B.R., Dobrowsky, T.M., Wirtz, D., Sun, S.X. 2009. Condensation of FtsZ filaments can drive bacterial cell division. Proc. Natl. Acad. Sci. USA. 106:121-126.
Larsen, R.A., Chen, G.J., Postle, K. 2003. Performance of standard phenotypic assays for TonB activity as evaluated by varying the level of functional, wild-type TonB. J. Bacteriol. 185:4699-4706.
Larsen, R.A., Deckert, G.E., Kastead, K.A., Devanathan, S., Keller, K.L., Postle, K. 2007. His20 provides the sole functionally significant side chain in the essential TonB transmembrane domain. J. Bacteriol. 189:2825-2833.
Larsen, R.A., Postle, K. 2001. Conserved residues Ser16 and His20 and their relative positioning are essential for TonB activity, cross-linking of TonB with ExbB, and the ability of TonB to respond to proton motive force. J. Biol. Chem. 276:8111-8117.
Larsen, R.A., Thomas, M.G., Postle, K. 1999. Protonmotive force, ExbB and ligand-bound FepA drive conformational changes in TonB. Mol. Microbiol. 31:1809-1824.
Larsen, R.A., Wood, G.E., Postle, K. 1993. The conserved proline-rich motif is not essential for energy transduction by Escherichia coli TonB protein. Mol. Microbiol. 10:943-953.
Lazdunski, C.J., Bouveret, E., Rigal, A., Journet, L., Lloubes, R., Bénédetti, H. 1998. Colicin import into Escherichia coli cells. J. Bacteriol. 180:4993-5002.
Lazzaroni, J.C., Dubuisson, J.F., Vianney, A. 2002. The Tol proteins of Escherichia coli and their involvement in the translocation of group A colicins. Biochimie. 84:391-397.
Lazzaroni, J.C., Fognini-Lefebvre, N., Portalier, R. 1989. Cloning of excC and excD genes involved in the release of periplasmic proteins by E. coli K-12. Mol. Gen. Genet. 246:460-464.
159
Lazzaroni, J.C., Portalier, R.C. 1981. Genetic and biochemical characterization of periplasmic- leaky mutants of Escherichia coli K-12. J. Bacteriol. 145:1351-1358.
Lazzaroni, J.C., Vianney, A., Popot, J.L., Bénédetti, H., Samatey, F., Lazdunski, C., Portalier, R., Geli, V. 1995. Transmembrane α-helix interactions are required for the functional assembly of the Escherichia coli Tol complex. J. Mol. Bio. 246:1-7.
Leduc, M., Ishidate, K., Shakibai, N., Rothfield, L. 1992. Interactions of E. coli membrane lipoproteins with the murein sacculus. J. Bacteriol. 174:7982-7988. Lee, S.K., Keasling, J.D. 2005. A proprionate-inducible expression system for enteric bacteria. Appl. Environ. Microbiol. 7:6856-6862.
Levengood, S., Beyer Jr., W., Webster, R. 1991. TolA: A membrane protein involved in colicin uptake contains an extended helical region. Proc. Natl. Acad. Sci. USA. 88:5939-5943. Lewin, T., and Webster, R. 1996. Membrane insertion characteristics of the various transmembrane domains of the Escherichia coli TolQ protein. J. Biol. Chem. 271:14143- 14149. Lim, A.,Jr., De Vos, D., Brauns, M., Mossialos, D., Gaballa, A., Qing, D., Cornelis, P. 1997. Molecular and immunological characterization of OprL, the 18 kDa outer-membrane peptidoglycan-associated lipoprotein (PAL) of Pseudomonas aeruginosa. Microbiology. 143:1709-1716.
Llamas, M.A., Ramos, J.L., Rodriguez-Herva, J.J. 2000. Mutations in each of the tol genes of Pseudomonas putida reveal that they are critical for the maintenance of outer membrane stability. J. Bacteriol. 182:4764-4772.
Lloubès, R., Cascales, E., Walburger, A., Bouveret, A., Lazdunski, C., Bernadac, A., Journet, L. 2001. The Tol-Pal proteins of the Escherichia coli cell envelope; an energized system required for the outer membrane integrity? Res. Microbiol. 152:523-529. Luirink, J., Sinning, I., 2004. SRP-mediated protein targeting: structure and function revisited. Biochimica et Biophysica Acta. 1694:17-35. Luirink, J., Yu, Z., Wagner, S., de Gier, J-W. 2012. Biogenesis of inner membrane proteins in Escherichia coli. Biochimica et Biophysica Acta. 1817:965-976. Lutkenhaus, J. 2009. FtsN-trigger for septatation. J. Bacteriol. 191:7381-7382. McKee, J.R., McKee, T. 2003. Amino Acids, Peptides, and Proteins. In: Biochemistry: The Molecular Basis of Life (3rd Edition) McKee, J.R., McKee, T. (Eds) pp 110-115. McGraw Hill Companies, New York. Meinhardt, H., de Boer, P.A.J. 2001. Pattern formation in Escherichia coli: a model for the pole- to-pole oscillations on Min proteins and the localization of the division site. Proc. Natl. Acad. Sci. USA. 98:14202-14207. 160
Meury, J., Devilliers, G. 1999. Impairment of cell division in tolA mutants of Escherichia coli at low and high medium osmolarities. Biol. of the Cell. 91:67-75 Miller, J.H. 1972. Experiments in Molecular Genetics. Cold Spring Harbor Laboratory Press. Cold Spring Harbor, NY. Millipore Corp. Bedford, MA, 01730, USA. Möll, A., Thanbichler, M. 2009. FtsN-like proteins are conserved components of the cell division machinery in proteobacteria. Mol. Microbiol. 72:1037-1053. Muller, M., Vianney, A., Lazzaroni, J.C., Webster, R., and Portalier, R. 1993. Membrane topology of the Escherichia coli TolR protein required for cell envelope integrity. J. Bacteriol. 175:6059-6061. Nagel de Zwaig, R., Luria, S.E. 1967. Genetics and physiology of colicin-tolerant mutants of Escherichia coli. J. Bacteriol. 94:1112-1123.
New England BioLabs (NEB). Ipswich, MA, 01938, USA.
Nikaido, H. 1996. Periplasm. In: Escherichia coli and Salmonella, Cellular and Molecular Biology (2nd Edition). Neidhart, F.C., Curtiss III, R., Ingraham, J.L., Lin, E.C.C., Low, K.B., Magasanik, B., and Umbarger, H.E. (Eds.) pp 29-47. ASM Press, Washington, DC.
Nilsson, I., von Heijne, G. 1990. Fine-tuning the topology of a polytopic membrane protein: role of positively and negatively charged amino acids. Cell. 62:1135-1141.
Noinaj N., Guillier M., Barnard T.J., Buchanan S.K. 2010. TonB-dependent transporters: regulation, structure, and function. Annu Rev Microbiol. 64:43-60.
Oliver, D.B. 1996. Periplasm. In: Escherichia coli and Salmonella, Cellular and Molecular Biology (2nd Edition). Neidhart, F.C., Curtiss III, R., Ingraham, J.L., Lin, E.C.C., Low, K.B., Magasanik, B., and Umbarger, H.E. (Eds.) pp 88-103. ASM Press, Washington, DC.
Osawa, M., Anderson, D.E., Erickson, H.P. 2008. Reconstruction of contractile FtsZ rings in liposomes. Science. 320:792-794.
Pacific Immunology. Ramona, CA, 92065, USA.
Parkhill, J., Achtman, M., James, K.D., Bentley, S.D., Churcher, C., Klee, S.R., Morelli, G., Basham, D., Brown, D., Chillingworth, T., Davies, R.M., Davis, P., Devlin, K., Feltwell, T., Hamlin, N., Holroyd, S., Jagels, K., Leather, S., Moule, S., Mungall, K., Quail, M.A., Rajandream, M.A., Rutherford, K.M., Simmonds, M., Skelton, J., Whitehead, S., 2000. Complete DNA sequence of a serogroup A strain of Neisseria meningitidis Z2491. Nature. 404:502-506. 161
Picken, R.N., Beacham, I.R. 1977. Bacteriophage-resistant mutants of Escherichia coli K-12. Location of receptors within the lipopolysaccharide. J. Gen. Microbiol. 102:305-318.
Postle, K. 1993. TonB protein and energy transduction between membranes. J. Bioenerget. Biomemb. 25:591-601. Postle, K. 1999. Active transport by customized β-barrels. Nature Struct. Biol. 6:3-6. Postle, K., Kadner, R.J. 2003. Touch and Go: tying TonB transport. Mol. Microbiol. 49:869-882.
Postle K., Larsen, R.A. 2007. TonB-dependent energy transduction between outer and cytoplasmic membranes. BioMetals. 20:453-465.
QIAGEN Inc. Valencia, CA, 91355, USA.
Reynolds, P.R., Mottur, G.P., and Bradbeer, C.J. 1980. Transport of vitamin B-12 in Escherichia coli. Some observations on the roles of the gene products of BtuC and TonB. J. Biol. Chem. 255:4313-4319. Rico, A.I., Garcia-Ovalle, M., Palacios, P., Casanova, M., Vicente, M. 2010. Role of Escherichia coli FtsN in the assembly and stability of the cell division ring. Mol. Microbiol. 76:760- 771. Rodriguez-Herva, J.J., Ramos, J.L. 1996. Characterization of an OprL null mutant of Pseudomonas putida. J. Bacteriol. 178:5836-5840. Rodriguez-Herva, J.J., Ramos-Gonzales, M.I., Ramos, J.L. 1996. The Pseudomonas putida peptidoglycan-associated outer membrane lipoprotein is involved in the maintenance of the integrity of the cell envelope. J. Bacteriol. 178:1699-1706. Rohrer, J., Kuhn, A. 1990. The function of a leader peptide in translocating charged amino acyl residues across a membrane. Science. 250:1418-1421. Schumann, W. 2006. Dynamics of the Bacterial Chromosme. pp. 1-27. Wiley-VCH Verlag GmbH & Co. KGaA. Bayreuth, Germany.
Sen, K., Sikkema, D.J., Murphy, T.F. 1996. Isolation and characterization of the Haemophilus influenzae tolQ, tolR, tolA., tolB genes. Gene. 178:75-81.
Shen, B., Lutkenhaus, J. 2011. Differences in MinC/MinD sensitivity between polar and internal Z rings in Escherichia coli. J. Bacteriol. 193:367-376.
Seigele, D.A., Hu, J.C. 1997. Gene expression from plasmids containing the araBAD promoter at subsaturating inducer concentrations represents mixed populations. Proc. Natl. Acad. Sci. USA. 94:8168-8172.
Silhavy, T. J., Kahne, D., Walker, S. 2010. The Bacterial Cell Envelope. Cold Spring Harb. Perspect. Biol. 2101;2:a000414.
162
Stratagene. 2003. BacterioMatch II Two-Hybrid System Vector Kit. Instruction Manual. Stratagene: Agilent Technologies Inc., La Jolla, CA.
Studier, F.W., Daegelen, P., Lenski, R.E., Maslov, S., Kim, J.F. 2009. Understanding the difference between genome sequences of Escherichia coli B strains REL606 and BL21(DE3) and comparison of the E. coli B and K-12 genomes. J. Mol. Biol. 394:653- 680.
Sturgis, J.N. 2001. Organisation and evolution of the tol-pal gene cluster. J. Mol.Micriobiol. Biotechnol. 3:113-122.
Sun, T.P., Webster, R.E. 1987. Nucleotide sequence of a gene cluster involved in entry of E colicins and single-stranded DNA of infecting filamentous bacteriophages into Escherichia coli. J. Bacteriol. 169:2667-2674.
Teleha, M. 2009. The Significance of the N-terminal Region of TolQ in Maintaining Tol- associated Energy-dependent Functions and Cell Division in Escherichia coli. Master’s Thesis, Bowling Green State University.
Thanassi, D.G., Cheng, L.W., Nikaido, H. 1997. Active efflux of bile salts by Escherichia coli. J. Bacteriol. 179:2512-2518.
Tettelin, H., Saunders, N.J., Heidelberg, J., Jeffries, A.C., Nelson, K.E., Eisen, J.A., Ketchum, K.A., Hood, D.W., Peden, J.F., Dodson, R.J., Nelson, W.C., Gwinn, M.L., DeBoy, R., Peterson, J.D., Hickey, E.K., Haft, D.H., Salzberg, S.L., White, O., Fleischmann, R.D., Dougherty, B.A., Mason, T., Ciecko, A., Parksey, D.S., Blair, E., Cittone, H., Clark, E.B., Cotton, M.D., Utterback, T.R., Khouri, H., Qin, H., Vamathevan, J., Gill, J., Scarlato, V., Masignani, V., Pizza, M., Grandi, G., Sun, L., Smith, H.O., Fraser, C.M., Moxon, E.R., Rappuoli, R., Venter, J.C. 2000. Complete genome sequence of Neisseria meningitidis serogroup B strain MC58. Science. 287:1809-1815.
Tibor, A., Weynants, V., Denoel, P., Lichtfouse, B., De Bolle, X., Saman, E., Limet, J.N., Letesson, J.J. 1994. Molecular cloning, nucleotide sequence, and occurrence of a 16.5- kilodalton outer membrane protein of Brucella abortus with similarity to PAL lipoproteins. Infect. Immun. 62:3633-3639.
Ursinus, A., van den Ent, F., Brechtel, S., de Pedro, M., Höltje, J-V., Löwe, J., Vollmer, W. 2004. Murein (Peptidoglycan) binding property of the essential cell division protein FtsN from Escherichia coli. J. Bacteriol. 186:6728-6737.
Vankemmelbeke, M., Zhang, Y., Moore, G.R., Kleanthous, C., Perfold, C.N, James, R. 2009. Energy-dependent immunity protein release during Tol-dependent nuclease colicin translocation. J. Biol. Chem. 284:18932-18941.
163
Vianney, A., Lewin, T., Beyer Jr, W., Lazzaroni, J.C., Portalier, R., Webster, R. 1994. Membrane topology and mutational analysis of the TolQ protein of Escherichia coli required for the uptake of macromolecules and cell envelope integrity. J. Bacteriol. 176:822-829.
Vianney, A., Muller, M.M., Clavel, T., Lazzaroni, J.C., Portalier, R., Webster, R. 1996. Characterization of the tol-pal region of Escherichia coli K-12: Translational control of tolR expression by TolQ and identification of a new open reading frame downstream of pal encoding a periplasmic protein. J. Bacteriol. 178:4031-4038.
Vicente, M., Rico, A. I. 2006. The order of the ring: assembly of Escherichia coli cell division components. Mol. Microbiol. 61:5-8.
Vicente, M., Rico, A.I., Martinez-Arteaga, R., Mingorance, J. 2006. Septum enlightenment: assembly of bacterial division proteins. J. Bacteriol. 188:19-27.
Vines, E., Marolda, C.L., Balachandran, A., Valvano, M.A. 2005. Defective O antigen polymerization in TolA and Pal mutants of Escherichia coli in response to cytoplasmic stress. J. Bacteriol. 187:3359-3368.
von Heijne, G. 2006. Membrane-protein topology. Nat. Rev. Mol. Cell Biol. 7:909-918.
von Heijne, G. 1994. Sec-independent protein insertion into the inner E. coli membrane. A phenomenon in search of an explanation. FEBS Lett. 346:69-72.
Walburger, A., Lazdunski, C., and Corda, Y. 2002. The Tol/Pal system function requires an interaction between the Carboxyl-terminal domain of TolA and the Amino-terminal domain of TolB. Mol. Microbiol. 44:695-708.
Wang, P., Dalbey, R.E. 2011. Inserting membrane proteins: The YidC/Oxa1/Alb3 machinery in bacteria, mitochondria, and chloroplasts. Biochimica et Biophysica Acta. 1808:866-875.
Weigand, R.A., Vinci, K. D., Rothfield, L.I. 1976. Morphogenesis of the bacterial division septum; a new class of septation-defective mutants. Proc. Natl. Acad. Sci. USA. 73:1882- 1896.
Weiss, D. 2004. Bacterial cell division and the septal ring. Mol. Microbiol. 54:577-597.
Weitzel, A.C., Larsen, R.A. 2008. Differential complementation of ΔtolA Escherichia coli by a Yersinia enterocolitica TolA homologue. FEMS Micriobiol. Lett. 282:81-88.
Whitley, P., Zander, T., Ehrmann, M., Haardt, M., Bremer, E., von Heijne, G. 1994. Sec- independent translocation of a 100-residues long periplasmic N-terminal tail in the E. coli inner membrane protein ProW. EMBO J. 13:4653-4661.
164
Xiang, Y., Zhang, -Z., Goemaere, E., Seddiki, N., Celia, H., Gaviolo, M., Cascales, E. 2011. Mapping the interactions between Escherichia coli TolQ transmembrane segments. J. Bio. Chem. 286:11756-11764.
Yang, J-C., van den Ent, F., Neuhaus, D., Brevier, J., Lowe, J. 2004. Solution structure and domain architecture of the divisome protein FtsN. Mol. Microbiol. 52:651-660.
Yao, X., Walter, J., Burke, S., Stewart, S., Jericho, M.H., Pink, D., Hunter, R., Beveridge, T.J. 2002. Atomic force microscopy and theoretical considerations of surface properties and turgor pressures of bacteria. Colloids and Surfaces B: Biointerfaces. 23:213-230.
Yeh, Y., Comolli, L., Downing, K., Shapiro, L., McAdams, H. 2010. The Caulobacter Tol-Pal complex is essential for outer membrane integrity and the positioning of a polar localization factor. J. Bacteriol. 192:4847-4858.
Young, K.D. 2006. The selective value of bacterial shape. Microbiol. Mol. Biol. Rev. 70:660- 703.
Zhai, Y.F., Heijne, W., Saier, M.H. 2003. Molecular modeling of the bacterial outer membrane receptor energizer, ExbBD/TonB, based on homology with the flagellar motor, MotAB. Biochim. Biophys. Acta. 1614:201-210.
Zhuang, Z., Song, F., Martin, B.M., Dunaway-Mariano, D. 2002. The YbgC protein encoded by the ybgC gene of the tol-pal gene cluster of Haemophilus influenzae catalyzes acyl- coenzyme A thioester hydrolysis. FEBS Lett. 516:161-163.
165
APPENDIX A
Over-expression of the Escherichia coli TolQ protein leads to a null-FtsN-like division
phenotype
This appendix is a manuscript submitted to MicrobiologyOpen in its final submission format. The manuscript contains work carried out for this dissertation, preliminary work to this dissertation (Teleha, 2009), and supporting data not included in this dissertation. The manuscript was accepted for publication June 3, 2013.
166
Over-expression of the Escherichia coli TolQ protein leads to a null-
FtsN-like division phenotype
Mary A. Teleha1,2 , Adam C. Miller2 and Ray A. Larsen1#
RUNNING TITLE:
TolQ over-expression phenotype
KEY WORDS:
Escherichia coli, TolQ, FtsN, Cell division
1Department of Biological Sciences, Bowling Green State University, Bowling Green, OH
43403.
2Division of Science and Math, Lorain County Community College, Elyria, OH 44035.
#Correspondence: [email protected] 167
Summary
Mutations involving the Tol-Pal complex of Escherichia coli result in a subtle phenotype in
which cells chain when grown under low-salt conditions. Here, the non-polar deletion of individual genes encoding the cytoplasmic membrane associated components of the complex
(TolQ, TolR, TolA) produced a similar phenotype. Surprisingly, the over-expression of one of these proteins, TolQ, resulted in a much more overt phenotype in which cells occurred as elongated rods coupled in long chains when grown under normal salt conditions. Neither TolR nor TolA over-expression produced a phenotype, nor was the presence of either protein required for the TolQ-dependent phenotype. Consistent with their native membrane topology, the amino-
terminal domain of TolQ specifically associated in vivo with the periplasmic domain of FtsN in a
cytoplasm-based two-hybrid analysis. Further, the concomitant over-expression of FtsN rescued
the TolQ-dependent phenotype, suggesting a model wherein the over-expression of TolQ
sequesters FtsN, depleting this essential protein from the divisome during Gram negative cell
division. The role of the Tol-Pal system in division is discussed. 168
Introduction
The Tol system is a set of proteins that ostensibly function to couple cytoplasmic membrane
(CM)-derived energy to outer membrane (OM) processes in the Gram-negative envelope
(Cascales et al., 2000; Germon et al., 2001). These proteins are encoded from the tol-pal gene
cluster, which is conserved among many Gram-negative bacteria with products sharing
similarities in both sequence (Sturgis, 2001) and, where examined, function (Webster, 1991;
Dennis et al., 1996; Heilpern & Waldor, 2000; Prouty et al., 2002; Llamas et al., 2003). While the specific physiological role of the Tol system remains unclear, numerous phenotypes have been observed in Tol system mutants. Initially identified as conferring tolerance to certain colicins (Nagel de Zwaig & Luria, 1967) and later filamentous phage (Sun and Webster, 1986), tol mutants display a variety of traits consistent with disruption of outer membrane integrity.
These include formation of mucoid colonies, hypersensitivity to deoxycholate and high molecular mass antibiotics (Bernstein et al., 1972), leakage of periplasmic contents (Lazzaroni &
Portalier, 1981), and shedding of outer membrane-derived vesicles (Bernadac et al., 1998).
Other studies have suggested a potential role in the expression of O-specific lipopolysaccharides known to participate in OM integrity (Gaspar et al., 2000; Vines et al., 2005).
The tol-pal gene cluster consists of seven open reading frames, five of which encode proteins
with established roles in the Tol system (Vianney et al., 1996). Two of these proteins, TolQ and
TolR, are components of a heteromultimeric CM protein complex that appears to couple a third
protein, TolA, to the electrochemical gradient of the CM (Cascales et al., 2000, 2001; Germon et
al., 2001). TolA protein spans the periplasmic space to interact with a number of outer 169
membrane-associated proteins, including two tol-pal gene cluster products, TolB and Pal (Isnard et al., 1994; Cascales et al., 2002). The Tol protein complex, mediated through Pal, interacts with two non-tol-encoded proteins, OmpA and the major lipoprotein Lpp. Two additional genes, ybgC and ybgF, encode proteins whose contributions to the Tol system remain unclear (Cascales et al., 2002). The ybgC gene product localizes to the cytoplasm, where it functions as a thioesterase (Zhuang et al., 2002). The function of the periplasmically localized YbgF protein is unclear. However, genetic and physical evidence suggests that it does interact with other proteins of the Tol system (Walburger et al., 2002; Krachler et al., 2010).
While maintenance of OM integrity is the most often cited role for the Tol system, indirect evidence has implicated this protein complex in the cell division process. Meury and Devilliers
(1999) observed impaired cell division patterns in E. coli tolA mutants when grown under conditions of either low osmolarity or high ionic strength. This distinct morphological phenotype, consisting of filamenting or chaining cells, was subsequently observed in tol mutants of Vibrio cholerae (Heilpern and Waldor, 2000), Pseudomonas putida (Llamas et al., 2000) and
Erwinia chrysanthemi (Dubuisson et al., 2005). Gerding et al. (2007) found that Tol mutants experience delayed OM invagination and contain outer membrane blebs at constriction sites and cell poles. In that study, chimeras consisting of each of the five Tol proteins translationally fused with fluorescent proteins (GFP and cRFP) appeared to localize to constriction sites during cell division. This localization did not occur when any of the Tol-GFP fusion proteins were expressed in cells depleted of the divisome protein FtsN. These data led the authors to suggest a model where Tol system components are recruited by FtsN to cell constriction sites where they then couple the OM to the divisome to coordinate division of the outer membrane with septum 170
formation and cell division (Gerding et al., 2007). The subsequent observation that in the
absence of FtsN, an FtsA suppressor mutation could support the localization of TolA-GFP to the
divisome, indicating that the role of FtsN in this process is probably indirect (Bernard et al.,
2007).
In previous studies we had noted that even moderate over-expression of one particular Tol
system component; TolQ hindered the growth of E. coli strains in both fluid (Brinkman, 2007)
and solid phase cultures (Brinkman and Larsen, 2008). Subsequently we observed that TolQ
over-expression resulted in cells with a smooth, highly elongated appearance, a phenotype
similar to that previously observed in FtsN-depleted cells (Dai et al., 1993). In this study, we
document this TolQ-specific over-expression phenotype and provide genetic and biochemical
evidence for direct interactions between TolQ and the divisome protein FtsN.
171
Results
The morphology of ∆tol strains is altered when grown in low salt medium. Distinct
phenotypes for ∆tol derivatives of W3110 were not evident by light microscopy for cells grown
to stationary phase in standard LB at 37˚C (Fig. 1). Conversely, when grown in low salt LB,
stationary phase ∆tol strains occurred as short cocco-bacillary cells in chains, whereas the
W3110 parent cells had a normal appearance as short, single rods (Fig. 1). Thus the chaining
phenotype previously noted specifically for tolA mutants grown in low salt (Meury and
Devilliers, 1999; Gerding et al., 2007) also occurred for tolQ and tolR mutants.
The over-expression of TolQ uniquely results in cell filamentation. The absence of
individual Tol proteins resulted in a phenotype that suggested an impact on cell division when
grown under non-standard osmolarity conditions. Interestingly, the over-expression of these
proteins resulted in distinctly different phenotypes. For TolA and TolR, over-expression had no evident impact on cell division when examined in stationary phase (Fig. 2). Surprisingly, TolQ over-expression resulted in a novel filamentation phenotype (Fig. 2, lower right panel). The majority of cells appeared as elongated rods growing in long chains, with regions suggestive of septum formation. This filamentation phenotype was not unique to stationary phase cells, as a similar morphology was evident in exponential phase cells in the first hours of growth following induction of tolQ expression (data not shown).
The extent of cell filamentation corresponds to the level of TolQ expression. If the filamentation phenotype occurs as a result of TolQ over-expression, the degree of filamentation 172
should vary with level to which TolQ is over-expressed. Examination of cells carrying plasmids bearing the pBAD-regulated tolQ gene and grown to stationary phase in LB supplemented with increasing 10-fold concentrations of ʟ-arabinose suggested this to be the case (Fig. 3a). In these
experiments, cells grown with ʟ-arabinose at levels of 0.0001% (w/v) or less were
indistinguishable from those grown without ʟ-arabinose supplementation,whereas filamentation
is first evident in cells grown with 0.001% (w/v) ʟ-arabinose and becomes increasingly apparent
for cells grown in 0.01 and 0.1% (w/v) ʟ-arabinose. Immunoblot analysis using a polyclonal
monospecific TolQ antiserum allowed for the comparison of TolQ levels obtained by ʟ-arabinose
induction relative to the level of TolQ protein resulting from normal expression of the
chromosomal tolQ gene (Fig. 3b). The relative levels of TolQ expression in these cells, assayed in late exponential growth, correlated directly with the filamentation phenotypes evident in stationary cells grown in the same concentrations of ʟ-arabinose. Specifically, cells grown with
ʟ-arabinose concentrations that resulted in a TolQ level similar to chromosomally-encoded levels did not show evidence of filamentation, with the filamentation phenotype first evident at an ʟ-
arabinose level that induces a moderate over-expression of TolQ (Fig. 3a, 0.001%) and becoming
more extensive with ʟ-arabinose levels that induce greater levels of TolQ protein.
TolQ-induced cell filamentation occurs independent of TolR and TolA. The above results
suggested that neither the normal role of the CM Tol complex, nor a complete complex itself was
involved in the filamentation phenotype. To test this possibility, TolQ protein was over-
expressed in each of the ∆tol strains, such that only in the case of the ∆tolQ strain would all three
components of the CM Tol complex be present (Fig. 4). The over-expression of TolQ resulted in 173
the filamentation phenotype in all three ∆tol strains, indicating that neither TolA nor TolR made
essential contributions to the TolQ-dependent filamentation phenotype.
Because TolQ normally occurs in as part of a cytoplasmic membrane protein complex with
TolR, the possibility that filamentation was a response to a stoichiometric imbalance between
TolQ and Tol R was examined. Overexpression of TolQ and TolR from an arabinose-regulated cloned operon similarly resulted in cell filamentation, with levels of TolQ achieved from the plasmid-borne tolQR operon slightly lower than those achieved from a plasmid encoded tolQ alone (Fig S1)
TolQ and TolR are paralogues of ExbB and ExbD of the TonB system, with enough shared function that molecular crosstalk occurs between the systems (Braun, 1989). Despite its structural and functional similarity to TolQ, over-expression of ExbB (with ExbD) to levels
similar to those that result in extensive filamentation for TolQ did not alter the division
phenotype of the cell (Fig S1). This suggests filamentation is a very specific response to the
over-expression of TolQ, involving protein interactions unique to TolQ.
The TolQ amino-terminal domain interacts with the FtsN periplasmic domain. The TolQ-
dependent filamentation phenotype resembled the morphological phenotype previously reported
for temperature-sensitive ftsN mutants grown under restrictive conditions (Dai et al., 1993). This
suggested the possibility that the over-expression of TolQ somehow interfered with an FtsN-
dependent process in cell division. The previous observation that FtsN appeared to play a role in
recruiting GFP-Tol protein fusions to the divisome (Gerding et al., 2007) suggested a potential 174
physical interaction between TolQ and FtsN. To address this possibility, “bait” domains of TolQ fused with the lambda cI protein and “target” domains of FtsN fused the α-subunit of RNA polymerase were paired for evaluation by bacterial two-hybrid analysis (Fig. 5). All strains and pairings grew on normal LB (Fig. 5 row A), and all but the “no plasmid control” grew on the chloramphenicol- and tetracycline-supplemented non-selective His dropout medium (Fig. 5, row
B). The ability to grow on under either selective (row C) or dual selective (row D) conditions was the criteria for putative interactions between paired domains of FtsN and TolQ, with the pBT-LGF2/pTRGal11Ppairing serving as a positive control for interaction, and the cI/RNα pairing providing a negative control. For the cytoplasmic domains, none of the pairings between
FtsN and TolQ domains were productive (TC1/FC, TC2/FC), nor were the self-activation controls (cI/Fc, TC1/RNα, TC2RNα). For the periplasmic domains, a productive pairing occurred between the first periplasmic domain of TolQ and the periplasmic domain of FtsN, as indicated by growth on both selective and dual selective media (TP1/FP). Similarly, pairing of the entire periplasmic region of FtsN (residues 54-319) with the first periplasmic domain of TolQ also resulted in growth on both selective and dual selective media (data not shown). Productive interactions were not evident when the other TolQ periplasmic domain was paired with the FtsN periplasmic domain (TP2/FP), nor were the self-activation controls productive (cI/FP, TP1RNα,
TP2/RNα).
Over-expression of FtsN suppresses TolQ-induced cell filamentation. The data from the two-hybrid analysis strongly support the likelihood of an in vivo interaction between TolQ and
FtsN. In this setting, the over-expression of TolQ might result in a filamentation phenotype simply by binding to FtsN and preventing it from performing its normal function at the divisome. 175
If so, then over-production of FtsN should compensate for the loss of available FtsN due to TolQ sequestration and should alleviate the phenotype. To test this possibility, TolQ and FtsN were simultaneously over-expressed, using two ʟ-arabinose-inducible pBAD expression vectors (Fig.
6a). As per previous experiments, the filamentation phenotype was observed when expression of the tolQ gene was induced with 0.1% (w/v) ʟ-arabinose (Fig. 6a, ptolQ + pBAD24).
Nevertheless, when paired with a plasmid bearing an ʟ-arabinose-inducible ftsN gene, the filamentation phenotype was not observed in cultures grown with 0.1% (w/v) ʟ-arabinose (Fig.
6a, ptolQ + pftsN). This did not appear to result from an arabinose dilution effect as the control pairing of TolQ with TonB, a CM protein with a topology similar to that of FtsN still produced filamentous cells (Fig. 6a, ptolQ + ptonB). Despite the difference in filamentation phenotype, co-over-expression of TolQ with either FtsN or TonB resulted in similar levels of TolQ, indistinguishable from that of TolQ over-expressed alone (Fig. 6b)
Over-expression of E. coli TolQ induces filamentation in other Gram-negative species. The over-expression of TolQ protein appeared to interfere with some aspect of cell division in E. coli
K12 strains. To determine if the TolQ-dependent filamentation phenotype was unique to this laboratory-adapted lineage or involved a general aspect of cell division in Gram-negative enterics, the ʟ-arabinose-regulated E. coli tolQ gene was transformed into Enterobacter amnigenus and Cronobacter muytjensii. Under 0.1% (w/v) ʟ-arabinose induction, both species, like E. coli, displayed a filamentation phenotype similar to that seen in E. coli when carrying the
ʟ-arabinose-regulated E. coli tolQ gene, but not when carrying the pBAD24 vector alone (Fig. 7).
176
Discussion
First noted as a set of genes required for resistance to specific colicins (Nagel de Zwaig & Luria,
1967), the tol genes and their products have been the subjects of much scrutiny. However, identification of their actual physiologic function has remained elusive. Evidence to date suggests that the Tol system couples the ion electrochemical gradient of the CM to the OM, with a variety of tol phenotypes reflecting energy dependence (Cascales et al., 2000; Germon et al.,
2001; Vankemmelbeke et al., 2009). The pleiotropic phenotype displayed by tol mutants is suggestive of a role in the maintenance of outer membrane integrity (Lloubés et al., 2001).
Interestingly, certain tol system mutations impact the microscopic morphology of gram-negative organisms, with cells under conditions of low osmolarity or high ionic strength displaying filamentous or chaining growth patterns (Meury and Devilliers, 1999; Heilpern and Waldor,
2000; Llamas et al., 2000; Dubuisson et al., 2005). The observation that Tol proteins fused to fluorescent proteins localized to division septa and that this trafficking was dependent upon the divisome protein FtsN led Gerding et al. (2007) to propose that the Tol proteins comprise a sub- complex of the division machinery, tethering the OM to the peptidoglycan corset during constriction of the divisome.
In the present study, we noted that E. coli strains individually deleted for tolQ, tolR, or tolA occurred as single cells when grown in standard LB medium, but, similar to previous observations with tolA strains (Meury and Devilliers, 1999; Gerding et al., 2007) formed multi- septate cell chains when grown in LB lacking NaCl (Fig. 1). These observations indicate that the
CM-associated components of the Tol system are not essential for normal resolution of division when cells are grown under standard osmotic conditions. However, when grown under low 177
osmotic conditions, cells lacking any single component of the CM Tol complex displayed a
chaining phenotype. In previous studies, the accumulation of either GFP-tagged TolA or TolQ at the nascent division septum occurred independent of the presence of other Tol proteins, whereas the trafficking of TolR to the divisome required the presence of TolQ (Gerding et al., 2007).
Together these observations suggest the participation of the entire CM Tol protein complex in the division process, with central roles played by TolA and TolQ.
The division phenotype of strains lacking a component of the CM Tol complex was subtle, evident only when cells are grown under abnormal osmotic conditions. In contrast, the over- expression of TolQ resulted in a more obvious division phenotype in which cells appeared as elongated rods coupled in long chains (Fig. 2). Sites of apparent septal formation separated
“individual” elongated rods. The spacing between these septal sites varied, likely reflecting the all-or-none nature of induction of the arabinose-inducible promoter used in this study (Siegele
and Hu, 1997). Consistent with this interpretation, the number of filaments relative to the
number of “normal” cells increased in a dose-dependent manner (Fig. 3). Unlike the chaining
phenotype of deletion mutants under low osmotic conditions, this filamentation phenotype was
associated only with TolQ, with the over-expression of either TolA or TolR resulting in no
evident division phenotype (Fig. 2). This suggested that the mechanism by which the over-
expression of TolQ disrupted division was distinct from the mechanism by which the CM Tol
complex contributes to cell division. Indeed, the filamentation phenotype occurred when TolQ
was over-expressed regardless of the presence or absence of other components of the CM Tol
complex (Fig. 4) indicating that this phenotype was not the result of an excess input of the
normal contribution of the Tol system to cell division. 178
The phenotype of cells over-expressing TolQ is similar to that seen in FtsN-depleted cells, and it
had been noted that FtsN was involved in recruiting Tol proteins to the divisome (Gerding et al.,
2007); thus we hypothesized that high levels of TolQ might disrupt FtsN function. One possible
mechanism for such a disruption would involve physical interactions between over-expressed
TolQ protein and FtsN, limiting the availability of the latter to the divisome. To test this possibility we used a two-hybrid analysis strategy, first to evaluate the potential in vivo
interaction between TolQ and FtsN and second to identify the specific domains involved in such
an interaction (Fig. 5). These experiments suggested that TolQ and FtsN could interact in vivo;
specifically the amino-terminal domain of TolQ (TolQ (1-19)) and the proximal periplasmic domain of FtsN (FtsN (54-243)) were identified as interacting regions. Significantly, this is the
same FtsN domain previously identified as necessary to support proper cell division (Dai et al.,
1996; Yang et al., 2004). While these results are consistent with the topological partitioning of
these membrane proteins, it was not possible to rule out additional interactions between the
transmembrane domains of TolQ and FtsN by this approach.
If the TolQ-dependent filamentation phenotype did result from the sequestration of FtsN, the
concomitant over-expression of FtsN should alleviate the phenotype. This was indeed the case
(Fig. 6). Previously, it was shown that over-expressed TolQ localized to the membrane in a
physiologically relevant conformation (Lewin and Webster, 1996). Presumably, over-expressed
FtsN is also trafficked to the CM appropriately. It should be noted that neither these, nor the
two-hybrid experiments exclude the possibility that interactions might occur between the
cytoplasmic domain of TolQ and the periplasmic domain of nascent FtsN prior to its Sec- 179
mediated insertion into the CM. However, such a mechanism would not be temporally
consistent with the apparent recruitment of TolQ to the divisome.
The tol-pal gene cluster is conserved among Gram-negative species and the TolQ-TolR pair
share homologues throughout the Eubacteria, indicating a long evolutionary history (Sturgis,
2001; Lazzaroni et al., 1999, 2002). The homology of this system extends beyond sequence and
protein structure similarities to functional ones, with Tol mutants in other species exhibiting
common phenotypes tied to cell envelope integrity, OM maintenance (Sturgis, 2001) and cell
division (Meury and Devilliers, 1999; Heilpern and Waldor, 2000; Llamas et al., 2000; and
Dubuisson et al., 2005). The homologous nature of this system would imply that similar roles
for TolQ would be present in other Gram-negative bacteria. As expected, E. coli TolQ over- expressed in E. amnigenus and C. muytjensii resulted in a filamentation phenotype comparable to that observed in wild type E. coli cells (Fig. 7), indicating that the TolQ over-expression phenotype is not an E. coli-specific phenomenon and is likely tied to the endogenous function of this protein. Because the Tol-Pal complex is widely conserved among Gram-negative bacteria and TolQ and TolR are the two most widespread proteins of the complex, sharing structural and functional homology with ExbB/ExbD and MotA/MotB (Cascales et al., 2001; Eick-Helmerich and Braun, 1989), it is likely that the Tol-Pal complex originated with the appearance of Gram- negative bacteria. This protein complex, while its true role in the cell remains elusive, is clearly required for the maintenance of OM integrity (Lazzaroni et al., 1989), a role that can be understood as evidence that the Tol-Pal complex evolved alongside Gram-negatives and in particular, the Gram-negative OM.
180
Taken together, these data present an intriguing picture of a potential role for TolQ and by
extension the Tol system in the process of cell division. The phenotype observed in cells over-
expressing TolQ is distinct from that of Tol mutants grown in low salt conditions. Over- expressed TolQ leads to the formation of long smooth filaments, a characteristic phenotype of ftsN mutants. The chaining phenotype observed in Tol mutants grown under low salt conditions observed here and elsewhere (Meury and Devilliers, 1999; Gerding et al., 2007) may reflect a response to osmotic stress by the absence of a functional Tol system. Cell filamentation has long been recognized as a stress response to unfavorable growth conditions such as nutrient limitation
(Young, 2006). Gerding et al. (2007) provided evidence that plasmid-expressed Tol proteins rescue this phenotype, while we found that excess TolQ results in a distinctly different division phenotype. TolQ-sequestered FtsN would potentially be unavailable in the cell to interact with another downstream essential cell division component to carry out its role in the process of cell division.
Although FtsN joins the septal ring near the end of its assembly, FtsN depletion leads to disassembly of the early divisome components of the proto-ring, indicating a specific role for
FtsN in stabilization of the division ring. When restored to normal levels, FtsN back-recruits early divisome complexes (Rico et al., 2010) and a number of divisome components have been
implicated in sharing a role in stabilizing the divisome, with functional overlap occurring among
many (Geissler and Margolin 2005; Rico, et al., 2010). This is evidenced by the ability of
altered or highly expressed divisome components to compensate for the loss of others (Geissler
and Margolin 2005; Bernard et al., 2007). Such findings support a model of divisome assembly
and function that is more dynamic and cooperative than originally described. Further, 181
interactions between divisome proteins and a growing number of non-essential proteins point to a more complex nature for the cell division process itself. It is reasonable to consider a dual role for the Tol-Pal complex in the cell. A role for the Tol system in outer membrane maintenance has clearly been established in a number of Gram-negative species (Sturgis, 2001; Meury and
Devilliers, 1999; Heilpern and Waldor, 2000; Llamas et al., 2000; and Dubuisson et al., 2005).
Much less is known about the role of the Tol system in cell division. Under standard conditions,
Tol mutants continue to divide, ruling out an essential role for the Tol-Pal complex as a core component of the divisome. However, an accessory role for the Tol system in divisome stability is a very plausible physiological role for this system. The Tol-Pal system links the IM and OM, during both non-dividing and dividing stages of the bacterial life cycle. During both stages, this link would have been advantageous throughout the evolution of the Gram-negatives. The ability of the Tol-Pal complex to interact with divisome components, in particular TolQ to interact with
FtsN or FtsN-like proteins, would also have been advantageous, as such an interaction might provide additional stability to the divisome while physically linking the OM to the CM during cell division, facilitating a more efficient cell division process through cooperative binding.
Evidence indicates the Tol proteins interact through their transmembrane domains to collectively perform energy-dependent functions in the cell (Germon et al., 2001) and that a necessary stoichiometric balance of Tol system components preserves this functionality (Guihard et al.,
1994). Although the present data suggest an energized Tol system is unnecessary for generating the filamentation phenotype observed in cells over-expressing TolQ, it does not provide insight into the potential contribution of Tol system-derived energy in the division process as proposed by Gerding et al. (2007). While that study suggests that FtsN is normally involved in the recruitment of Tol proteins to the divisome, its role in this process appears to be indirect, as a 182
suppressor mutation in FtsA can provide for TolA recruitment in an FtsN depleted strain
(Bernard et al., 2007). This suggests that the physical interaction between the periplasmic
domains of TolQ and FtsN could occur following the recruitment of the Tol components to the
divisome, with the over-expressed TolQ protein then diverting FtsN from its contributions to maintaining the divisome.
The details of Tol complex action at the division site remain to be determined. However, the results of this study support previous findings that the Tol-Pal complex, specifically TolQ, does indeed interact with the cell division machinery in E. coli and its relatives. We have identified
the domains of interaction between TolQ and FtsN and provided evidence that this interaction is
substantial enough to interfere with the cell division process when TolQ is present in excess. We
have also shown that this disruption in cell division can be alleviated with simultaneous over-
expression of FtsN. The mechanism by which the over-expression of TolQ and its interaction
with FtsN disrupts the cell division process remain to be clarified, perhaps as the function of
FtsN at the divisome becomes better understood and the specific role for the Tol-Pal complex
during cell division is clarified.
Methods
Media:
Bacterial strains were maintained on Luria-Bertani (LB) agar (Miller, 1972). Plasmid-bearing
strains were maintained on LB agar supplemented with 100 μg ml-1 ampicillin, and/or 34 μg ml-1
chloramphenicol as necessary. For low osmotic assays, cells were grown in a modified LB 183
broth, made without the normal addition of 1.0% (w/v) NaCl. For all other assays cells were
grown in standard LB broth, supplemented with antibiotics as necessary, and ʟ-arabinose as
indicated. Colicin sensitivity assays were performed using cells suspended in T-top on T-plates
(Miller 1972) as previously described (Larsen et al., 2003), with both the T-top and T-plates
supplemented with antibiotics as necessary and ʟ-arabinose as indicated. Two-hybrid analyses
were performed on a variety of M9-based minimal selective media made as described in the
BacterioMatch Two-Hybrid Instruction Manual (Stratagene: Agilent Technologies Inc., La Jolla,
CA). These included: His-dropout Broth (1 X M9 minimal salts containing 1X His-dropout
supplement amino acids (Clontech Laboratories, Inc. Mountain View, CA), 0.4% (w/v) glucose,
200 μM adenine HCl, 1 mM MgSO4, 1 mM Thiamine HCl, 100 μM CaCl2, 50 μM IPTG and 10
-1 µM ZnSO4); nonselective agar (His-dropout broth, 1.5% (w/v) agar, chloramphenicol 25 μg ml and tetracycline 12.5 μg ml-1); selective screening agar (nonselective agar with 5 mM 3-amino-
1,2,4 triazole); and dual selective screening agar (selective screening agar with 12.5 μg ml-1 streptomycin).
Strains:
Bacterial strains and plasmids are summarized in Table 1. The Escherichia coli K12 strain
W3110 (Hill and Harnish, 1981) was used in this study as the wild type. The W3110 derivatives
RA1027, RA1028, and RA1038 carried precise, complete deletions of the tolQ, tolR, and tolA genes, respectively. The construction of RA1038 was previously described (Weitzel and Larsen,
2008). In the present study, deletions of tolQ and tolR were similarly created using the λ red recombination technique (Datsenko and Wanner, 2000) to replace the predicted open reading frame for each gene with a “scar” region containing a stop codon and a ribosome-binding site to 184
minimize polar effects on downstream genes. Tol phenotypes of deletion mutants and their complementation by plasmids (see below) were confirmed by sensitivities to deoxycholate and group A colicins (data not shown) as previously described (Brinkman and Larsen, 2008; Weitzel and Larsen, 2008). For two-hybrid analysis the “BacterioMatch II two hybrid system reporter” strain (Δ(mcrA)183 Δ(mcrCB-hsdSMR-mrr)173 endA1 hisB supE44 thi-1 recA1 gyrA96 relA1 lac [F´ lacIq HIS3 aadA Kmr]) of E. coli K12 was purchased (Stratagene; Agilent
Technologies Inc., La Jolla, CA). The Enterobacter amnigenus strain ATCC51816 and
Cronobacter muytjensii strain ATCC51329 were obtained from the American Type Culture
Collection.
Plasmids:
Expression of specific Tol system proteins was achieved using plasmids derived from ʟ- arabinose-regulated plasmids (Guzman et al., 1995) as previously described (Brinkman and
Larsen, 2008). For this study, the plasmid carrying only tolQ (pRA031) was constructed from pRA003 (carrying both tolQ and tolR) by removal of tolR with specific restriction, followed by ligation. This tolQ construct was then excised by restriction and ligated into the identically restricted pBAD18-Cm to create pMT002.
The ftsN gene was amplified by the polymerase chain reaction (PCR) from W3110 genomic
DNA and inserted into pBAD24 at the NcoI and XbaI sites to create pMT001.
For two-hybrid analysis regions encoding the “bait” domains of TolQ were amplified by PCR and cloned into the plasmid pBT to generate in-frame fusions with a gene encoding the lambda 185
cI protein. Four such fusions were constructed, using tolQ codons 1-19 (pMT005), 39-135
(pMT006), 157-174 (pMT007), and 194-230 (pMT008). Similarly, regions encoding “target” domains of FtsN were amplified and cloned into the plasmid pTRG to generate in-frame fusions with a gene encoding the α-subunit of RNA polymerase. Three such fusions were generated, using ftsN codons 1-33 (pMT009), 54-234 (pMT010), and 54-319 (pMT011).
The identities of all constructs were confirmed by DNA sequence analysis. Specific primers used for PCR are available on request.
Microscopy
Fresh overnight cultures (grown at 37˚C with agitation in LB broth supplemented with antibiotics as necessary) were diluted 1:200 into 5 ml aliquots of fresh LB broth, supplemented with antibiotics as necessary and ʟ-arabinose as indicated. In preliminary studies cultures were grown by shaking at 37˚C, with cells harvested as indicated with a bacteriological loop and heat- fixed onto clean glass slides. Fixed cells were stained with 0.6% (w/v) Safranin O in 20% (v/v) ethanol and then examined by light microscopy. Images of representative fields were digitally recorded under oil immersion on a Nikon H550S series compound light microscope using NIS
Elements Documentation Software (Nikon Instruments Inc., Melville, NY).
Immunoblot analysis
The relative levels of TolQ protein present in wild type E. coli and in plasmid-bearing cells under ʟ-arabinose induction was determined by immunoblot analysis. For the determination of relative induction levels (Fig. 3b), overnight cultures of W3110 carrying either pBAD24 or 186
pRA031 were sub-cultured 1:200 in LB ampicillin 100µg ml-1 supplemented with ʟ-arabinose at
0.1%, 0.01%, 0.001%, 0.0001%, 0.00001% or 0.0% (w/v) and incubated at 37˚C with agitation.
Cells were harvested in late exponential phase (A550 = 0.7, as determined with a Spectronic 20 spectophotometer with a path length of 1.5 cm) and precipitated at 4 ˚C with 10% trichloroacetic acid (TCA) to limit proteolysis (Abelson et al., 1990), rinsed with 1ml 100mM Tris-HCl (pH
7.9), then suspended in 25μl 1M Tris-HCl, mixed with 25μl 2X Laemmli sample buffer, incubated at 98˚C for 5 minutes and stored at -20˚C. Samples for dual overexpression studies were similarly processed, with cells cultivated in LB containing 34 µg ml-1 chloramphenicol and100µg ml-1 ampicillin and supplemented with ʟ-arabinose at 0.0 or 0.1%, and harvested in stationary phase coincident with the harvesting of cell for microscopy (Fig 6b). Samples were then resolved on a sodium dodecyl sulfate (SDS) 11% polyacrylamide gel (Laemmli, 1970).
Resolved proteins were transferred to a polyvinylidene fluoride membrane (Millipore Corp.
Bedford, MA) and probed with a monospecific polyclonal rabbit antiserum raised against a synthetic peptide corresponding to TolQ residues 47-62 (generated by Pacific Immunology, San
Diego, CA), and visualized using an anti-rabbit immunoglobulin HRP conjugate and enhanced chemiluminescence (ECL), as previously described (Larsen et al., 1993, Higgs et al., 1998).
Bacterial two-hybrid analysis
Bait and target constructs were co-transformed into BacterioMatch ® II Screening Reporter
Competent Cells according to the BacterioMatch ® II Two-Hybrid System protocol (Stratagene
®, CA 92037, USA). Cotransformation pairs consisted of pBT-TolQ (1-19) + pTRG-FtsN (54-319), pBT-TolQ (1-19) + pTRG-FtsN (54-243), pBT-TolQ (157-174) + pTRG-FtsN (54-319), pBT-TolQ (157-174) + pTRG-FtsN (54-243), pBT-TolQ (39-135) + pTRG-FtsN (1-33), and pBT-TolQ (194-230) + pTRG-FtsN (1- 187
33). Cotransformants were plated on non-selective, selective, and dual-selective media, with dual transformants maintained on LB agar supplemented with tetracycline and chloramphenicol as described in the BacterioMatch Two-Hybrid Instruction Manual (Stratagene).
188
Acknowledgements:
Strains RA1027 and RA1028 were constructed by Kerry Brinkman. The plamid pKP660 and the rabbit α-ExbB antisera were provided by Kathleen Postle. We thank Ron Jantz for
photographing the plates for the images used for Figure 5, and Donna Ruth and Brittany Jacob
for assistance with dual transformant screenings. This research was supported by awards from
the Department of Biological Sciences at Bowling Green State University and the Division of
Science and Math at Lorain County Community College.
189
Table 1: Bacteria and Plasmids
Strains: Relevant characteristics1 Source Escherichia coli: W3110 F- IN(rrnD-rrnE)1 Hill & Harnish (1981) RA1027 W3110-ΔtolQ Present study RA1028 W3110-ΔtolR Present study RA1038 W3110-ΔtolA Weitzel & Larsen (2008) BacterioMatch II reporter hisB lac [F´ lacIq HIS3 aadA Kmr] Stratagene2 Enterobacter amnigenus (ATCC51816) Clinical isolate ATCC3 Cronobacter muytjensii (ATCC51329) Clinical isolate ATCC3 Plasmids: pBAD18-Cm araBAD promoter, AraC, cmr Guzman et al. (1995) pBAD24 araBAD promoter, AraC, ampr Guzman et al. (1995) pKP315 pBAD24 encoding TonB Larsen et al. (1999) pKP660 pBAD24 encoding ExbBD Ollis & Postle (2012) pBT lambda cI fusion site, cmr Stratagene pBT-LGF2 lambda cI-LGF2 fusion, cmr Stratagene pTRG RNAPα fusion site, tetr Stratagene pTRG-Gal11P RNAPα-Gal11P fusion, tetr Stratagene pRA002 pBAD24 encoding TolR Brinkman & Larsen (2008) pRA003 pBAD24 encoding tolQ/R Brinkman & Larsen (2008) pRA004 pBAD24 encoding TolA Weitzel & Larsen (2008) pRA031 pBAD24 encoding TolQ Present study pMT001 pBAD24 encoding FtsN Present study pMT002 pBAD18-Cm encoding TolQ Present study pMT005 pBT lambda cI – TolQ(1-19) Present study pMT006 pBT lambda cI – TolQ(39-135) Present study pMT007 pBT lambda cI – TolQ(157-174) Present study pMT008 pBT lambda cI – TolQ(194-230) Present study pMT009 pTRG RNAPα – FtsN(1-33) Present study pMT010 pTRG RNAPα – FtsN(54-243) Present study pMT011 pTRG RNAPα – FtsN(54-319) Present study
1The specific nature of the deletions (∆) is as follows: The ∆tolQ deletion removed the predicted tolQ codons 1-230 and the TAA termination codon; the ∆tolR deletion removed the predicted tolR codons 1- 142 and the TAA termination codon; the ∆tolA deletion removed the predicted tolA codons 1-421, leaving the TAA termination codon in place. 2Purchased from Stratagene 3Purchased from the American Type Culture Collection (ATCC) 190
Figure legends:
Figure 1: Th e morphology of ∆tol strains is altered when grown in low salt medium.
Stained preparations of W3110 (”Wild type”) and the ∆tol derivatives RA1038 (ΔtolA), RA1028
(ΔtolR) and RA1027 (ΔtolQ); grown for 24 hrs at 37˚C with aeration in either normal Miller LB or modified Miller LB lacking added sodium chloride are shown. All panels are displayed at the same relative magnification, with a bar representing 10 µm provided in each panel for scale.
Figure 2: The over-expression of TolQ uniquely results in cell filamentation. Stained preparations of W3110 cells bearing plasmids carrying either the tolA (ptolA), tolR (ptolR) or tolQ (ptolQ) gene under the control of the pBAD promoter grown for 24 hours at 37˚C with aeration in Miller LB supplemented with 100 µg ml-1 ampicillin and either no ʟ-arabinose or
0.1% (w/v) ʟ-arabinose are shown. All panels are displayed at the same relative magnification, with a bar representing 10 µm provided in each panel for scale. The arrows in panel F highlight examples of possible septal formation.
Figure 3: The extent of cell filamentation corresponds to TolQ expression levels. Figure
3a: Stained preparations of W3110 bearing a plasmid carrying the tolQ gene under the control of the pBAD promoter grown for 24 hours at 37˚C with aeration in Miller LB supplemented with
100 µg ml-1 ampicillin and either 0.0% or 10-fold increments of ʟ-arabinose are shown. All panels are displayed at the same relative magnification, with a bar representing 10 µm provided in each panel for scale. Figure 3b: Immunoblot analysis of samples from cells grown to exponential phase in Miller LB supplemented with 100 µg ml-1 ampicillin and either 0.1% (w/v) 191
ʟ-arabinose for W3110 carrying the control plasmid pBAD24 (“chromosomal” and “∆tolQ” respectively) or as indicated for W3110 carrying a pBAD24 derivative bearing the tolQ gene under ʟ-arabinose control. Samples were resolved by SDS-PAGE on an 11% polyacrylamide gel, transferred to a PVDF membrane and visualized by enhanced chemiluminescence using a monospecific anti-TolQ antiserum as described in Methods. The positions of molecular mass standards are indicated as kDa values at the right side of the developed blot. At higher levels of induction an additional, minor species of greater apparent molecular mass is detected by the anti-
TolQ antiserum (0.1% lane).
Figure 4: TolQ-induced cell filamentation occurs independent of TolR and TolA. Stained preparations of the W3110 ∆tol derivatives RA1027 (ΔtolQ), RA1028 (ΔtolR), and RA1038
(ΔtolA) bearing a plasmid carrying the tolQ gene under the control of the pBAD promoter grown for 24 hours at 37˚C with aeration in Miller LB supplemented with 100 µg ml-1 ampicillin and either no ʟ-arabinose or 0.1% (w/v) ʟ-arabinose are shown. All panels are displayed at the same relative magnification, with a bar representing 10 µm provided in each panel for scale.
Figure 5: The TolQ amino-terminal domain interacts with the FtsN periplasmic domain.
The potential for various regions of the TolQ and FtsN proteins to interact were tested by a two- hybrid analysis, using plasmids encoding the FtsN domains fused to RNαP2 and the TolQ domains fused to lambda cI. The individual regions considered for each protein are depicted by the diagram at the top, and identified by protein and topology relative to the cytoplasmic membrane. Thus, the cytoplasmically exposed FtsN residues 1-33 are termed “FC”, while the periplasmically localized FtsN residues 54-243 are termed “FP”. Similarly, the cytoplasmically 192
exposed TolQ residues 39-135 and 194-230 are termed “TC1” and “TC2” respectively, whilst the periplasmically localized TolQ residues 1-19 and 157-174 are termed “TP1” and TP2”, respectively. Indicator cells co-expressing various fusions and/or control products were plated on 4 different media and scored for growth at 24 hrs: “A”, LB medium; “B”, nonselective His dropout medium containing chloramphenicol 25µg ml-1 and tetracycline 12.5 µg ml-1;“C”, selective screening medium comprised of nonselective agar supplemented with 5mM 3-amino-
1,2,4 triazole; and “D”, dual selective screening medium comprised of selective screening agar supplemented with streptomycin 12.5 µg ml-1. Results are displayed as three sets: “Controls”, consisting of indicator cells with no plasmid, a positive control of two fusion proteins known to interact (LGF2/Gall11P), and a negative control pairing the lambda cI and RNαP2 encoding plasmids lacking fusion domains (cI/RNα); “Cytoplasmic pairings” consisting of indicator cells bearing combinations of plasmids encoding fusions of the cytoplasmic domains of FtsN and
TolQ paired with each other or with cognate controls and; “Periplasmic pairings, consisting of indicator cells bearing combinations of plasmids encoding fusions of the periplasmic domains of
FtsN and TolQ paired with each other or with cognate controls.
Figure 6: Over-expression of FtsN suppresses TolQ-induced cell filamentation. Figure 6a:
Stained preparations of W3110 cells bearing a chloramphenicol-selected plasmid carrying the tolQ gene under the control of the pBAD promoter and paired with either a ampicillin-selected control plasmid (ptolQ + pBAD24), with a pBAD24 derivative carrying ftsN under the control of the pBAD promoter (ptolQ + pftsN) or with a pBAD24 derivative carrying tonB under the control of the pBAD promoter (ptolQ + ptonB) grown for 24 hours at 37˚C with aeration in
Miller LB supplemented with 100 μg ml-1 ampicillin and 34 μg ml-1 chloramphenicol are shown. 193
Cultures were either not supplemented with ʟ-arabinose or were supplemented with 0.1% (w/v)
ʟ-arabinose.
Figure 6b: Immunoblot analysis of samples taken of the above cells at 24 hrs culture. A serial
2-fold dilution series was made for samples from cells co-expressing TolQ and either FtsN or
TonB to allow for comparison of relative levels of TolQ protein, with “1.0” representing the
undiluted sample and the other values representing the relative fraction thereof loaded onto the
gel. Samples expressing TolQ alone and samples from cultures not supplemented with ʟ-
arabinose were loaded onto the gel without dilution. Samples were resolved by SDS-PAGE on
an 11% polyacrylamide gel, transferred to a PVDF membrane and visualized by enhanced
chemiluminescence using a monospecific anti-TolQ antiserum as described in Methods. The
positions of molecular mass standards are indicated as kDa values at the right side of the
developed blot.
Figure 7: Over-expression of E. coli TolQ induces filamentation in other Gram-negative species. Stained preparations of E. coli, Cronobacter muytjensii , and Enterobacter amnigenus strains bearing a plasmid carrying the tolQ gene under the control of the pBAD promoter grown for 24 hours at 37˚C with aeration in Miller LB supplemented with 100 µg ml-1 ampicillin and
either no ʟ-arabinose or 0.1% (w/v) ʟ-arabinose are shown. All panels are displayed at the same
relative magnification, with a bar representing 10 µm provided in each panel for scale.
194
References:
Abelson, J.N., M.I. Simon, and Deutscher, M.P. (1990) Guide to protein purification. Methods
Enzymol. 182: 78-79.
Bernadac, A., Gavioli, M., Lazzaroni, J.C., Raina, S., and Lloubés, R. (1998) Escherichia coli tol-pal mutants form outer membrane vesicles. J. Bacteriol. 180: 4872-4878.
Bernstein, A., Rolfe, B., and Onodera, K. (1972) Pleiotropic properties and genetic organization of the tolA,B locus of Escherichia coli K-12. J. Bacteriol. 112: 74-83.
Bernard, C.S., Sadasivam, M., Shiomi, D., and Margolin, W. (2007) An altered FtsA can compensate for the loss of essential cell division protein FtsN in Escherichia coli. Mol.
Microbiol. 64: 1289-1305.
Bernhardt, T.G., and de Boer, P. A. J. (2005) The Escherichia coli amidase AmiC is a periplasmic septal ring component exported via the twin-arginine transport pathway. Mol.
Microbiol. 48: 1171-1182.
Braun, V. (1989) The structurally related exbB and tolQ genes are interchangeable in conferring tonB-dependent colicin, bacteriophage, and albomycin sensitivity. J. Bacteriol 171:6387-6390
195
Brinkman, K.K. (2007) Using chimeras to evaluate cross-talk, energy transfer, and protein-
protein interactions in the TonB and TolA systems. Doctoral Dissertation, Bowling Green State
University.
Brinkman, K.K., and Larsen, R.A. (2008) Interactions of the energy transducer TonB with
noncognate energy-harvesting complexes. J. Bacteriol. 190: 421-427.
Cascales, E. Lloubès, R., and Sturgis, J.N. (2001) The TolQ-TolR proteins energize TolA and
share homologies with the flagellar motor proteins MotA-MotB. Mol. Microbiol. 42: 795-807.
Cascales, E., Buchanan, S.K., Duché, D., Kleanthous, C., Lloubès, R., Postle, K., Riley, M.,
Slatin, S., and Cavard, D. (2007) Colicin biology. Microbiol. Mol. Biol. Rev. 71: 158-229.
Cascales, E., Bernadac, A., Gavioli, M., Lazzaroni, J.C., and Lloubés, R. (2002) Pal Lipoprotein
of Escherichia coli plays a major role in outer membrane integrity. J. Bacteriol. 184: 754-759.
Cascales, E., M. Gavioli, J.M. Sturgis, and Lloubés R. (2000) Proton motive force drives the
interaction of the inner membrane TolA and outer membrane Pal proteins in Escherichia coli.
Mol. Microbiol. 38: 904-915.
Dai, K., Xu, Y., and Lutkenhaus, J. (1993) Cloning and characterization of ftsN, an essential cell
division gene in Escherichia coli isolated as a multicopy suppressor of ftsA12(Ts). J. Bacteriol.
175: 3790-3797. 196
Dai, K., Xu, Y., and Lutkenhaus, J. (1996) Topological characterization of the essential
Escherichia coli cell division protein FtsN. J. Bacteriol. 178: 1328-1334.
Datsenko, K.A., and Wanner, B.L. (2000) One-step inactivation of chromosomal genes in
Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA. 97: 6640-6645.
Dennis, J.J., Lafontaine E.R., and Sokol P.A. (1996) Identification and characterization of the tolQRA genes of Pseudomonas aeruginosa. J. Bacteriol. 178: 7059-7068.
Dubuisson, J.F., Vianney, A., Hugouvieux-Cotte-Pattat, N., and Lazzaroni, J.C. (2005) Tol-Pal proteins are critical cell envelope components of Erwinia chrysanthemi affecting cell morphology and virulence. Microbiology. 151: 3337-3347.
Eick-Helmerick, K., and Braun, V. (1989) Import of biopolymers into Escherichia coli:
Nucleotide sequences of the exbB and exbD genes are homologous to those of the tolQ and tolR genes, respectively. J. Bacteriol. 171: 5117-5126.
Gaspar, J.A., Thomas J.A., Marolda C.L., and Valvano M.A. (2000) Surface expression of O- specific lipopolysaccharide in Escherichia coli requires the function of the TolA protein. Mol.
Microbiol. 38: 262-275.
Geissler, B., and Margolin, W. (2005) Evidence for functional overlap among multiple bacterial cell division proteins: compensating for the loss of FtsK. Mol. Microbiol. 58: 596-612. 197
Gerding, M. A., Ogata, Y., Pecora, N.D., Niki H., and de Boer, P.A.J. (2007) The trans-envelope
Tol-Pal complex is part of the cell division machinery and required for proper outer-membrane invagination during cell constriction in E. coli. Mol. Microbiol. 63: 1008-1025.
Gerding, M. A., Liu, B., Bendezú, F.O., Hale, C.A., Bernhardt, T.G., and de Boer, P.A.J. (2009)
Self-Enhanced Accumulation of FtsN at Division Sites and Roles for Other Proteins with a
SPOR Domain (CamX, DedD, and RlpA) in Escherichia coli Cell Constriction. J. Bacteriol.
191: 7383-7401.
Germon, P., Ray, M.C., Vianney, A., and Lazzaroni, J.C. (2001) Energy-dependent conformational change in the TolA protein of Escherichia coli involves its Amino-terminal domain, TolQ, and TolR. J. Bacteriol. 183: 4110-4114.
Goemaere, E.L., Cascales, E., and Lloubés, R. (2007) Mutational analyses define helix organization and key residues of a bacterial membrane energy-transducing complex. J. Mol. Biol.
366: 1424-1436.
Guihard, G., Boulanger, P., Benedetti, H., Lloubés, R., Besnard, M., and Letellier, L. (1994)
Colicin A and the Tol proteins involved in its translocation are preferentially located in the contact sites between the inner and outer membranes of Escherichia coli cells. J. Biol. Chem.
269: 5874-5880.
198
Guzman, L.M., Belin, D., Carson, M.J., and Beckwith, J. (1995) Tight regulation, modulation,
and high-level expression by vectors containing the arabinose P-BAD promoter. J. Bacteriol.
177: 4121-4130.
Heilpern, A.J., and Waldor, M.K. (2000) CTXφ infection of Vibrio cholerae requires the tolQRA gene products. J. Bacteriol. 182: 1739-1747.
Higgs, P.I., Myers, P.S., and Postle, K. (1998) Interactions in the TonB-dependent energy transduction complex: ExbB and ExbD form homomultimers. J. Bacteriol. 180: 6031-
6038.
Hill, C.W., and Harnish, B.W. (1981) Inversions between ribosomal RNA genes of Escherichia coli. Proc. Natl. Acad. Sci. USA 78: 7069-7072.
Isnard, M., Rigal, A., Lazzaroni, J-C., Lazdunski, C., and Lloubés, R. (1994) Maturation and localization of the TolB protein required for colicin import. J. Bacteriol. 176: 6392-6396.
Kampfenkel, K., and Braun, V. (1993) Membrane topologies of the TolQ and TolR proteins of
Escherichia coli: Inactivation by a missense mutation in the proposed first transmembrane segment. J. Bacteriol. 175: 4485-4491.
Krachler, A., Sharma, A., Cauldwell, A., Papadakos, G., and Kleanthous, C. (2010) TolA modulates the oligomeric status of YbgF in the bacterial periplasm. J. Mol. Biol. 403: 270-285. 199
Laemmli, U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 277: 680-685.
Larsen, R.A., Chen, G.J., and Postle, K. (2003) Performance of standard phenotypic assays for
TonB activity as evaluated by varying the level of functional, wild-type TonB. J. Bacteriol. 185:
4699-4706.
Larsen, R.A., Thomas, M.G., and Postle, K. (1999) Protonmotive force, ExbB and ligand-bound
FepA drive conformational changes in TonB. Mol. Microbiol. 31:1809-1824.
Larsen, R.A., Wood, G.E., and Postle, K. (1993) The conserved proline-rich motif is not essential for energy transduction by Escherichia coli TonB protein. Mol. Microbiol. 10: 943-953.
Lazzaroni, J.C., Dubuisson, J.F., and Vianney, A. (2002) The Tol proteins of Escherichia coli and their involvement in the translocation of group A colicins. Biochimie. 84: 391-397.
Lazzaroni, J.C., and. Portalier, R.C. (1981) Genetic and biochemical characterization of periplasmic-leaky mutants of Escherichia coli K-12. J. Bacteriol. 145: 1351-1358.
Lazzaroni, J.C., Fognini-Lefebvre, N., and Portalier, R.C. (1989) Cloning of excC and excD genes involved in the release of periplasmic proteins by E. coli K-12. Mol. Gen. Genet. 246:
460-464.
200
Lazzaroni, J.C., Germon, P., Ray, M.C., and Vianney, A. (1999) The Tol proteins of Escherichia coli and their involvement in the uptake of biomolecules and outer membrane stability. FEMS
Microbiol. Lett. 177: 191-197.
Lazzaroni, J.C., Vianney, A., Popot, J.L., Bénédetti, H., Samatey, F., Lazdunski, C., Portalier, R., and
Geli, V. (1995) Transmembrane α-helix interactions are required for the functional assembly of the
Escherichia coli Tol complex. J. Mol. Biol. 246: 1-7.
Levengood, S., Beyer Jr., W., and Webster, R. (1991) TolA: A membrane protein involved in colicin uptake contains an extended helical region. Proc. Natl. Acad. Sci. USA. 88: 5939-5943.
Lewin, T.M., and Webster, R.E. (1996) Membrane insertion characteristics of the various transmembrane domains of the Escherichia coli TolQ protein. J. Biol. Chem. 271: 14143-14149.
Llamas, M.A., Ramos, J.L., and Rodriguez-Herva, J.J. (2000) Mutations in each of the tol genes of Pseudomonas putida reveal that they are critical for the maintenance of outer membrane stability. J. Bacteriol. 182: 4764-4772.
Llamas, M.A., Rodriquez-Herva, J.J., Hancock, R.E.W., Bitter, W., Tommassen, J., and Ramos,
J.L. (2003) Role of Pseudomonas putida tol-oprL gene products in uptake of solutes through the cytoplasmic membrane. J. Bacteriol. 185: 4707-4716.
201
Lloubés, R., Cascales, E., Walburger, A., Bouveret, E., Lazdunski, C., Bernadac, A., and Journet,
L. (2001) The Tol-Pal proteins of the Escherichia coli cell envelope: an energized system required for outer membrane integrity? Res. Microbiol. 152: 523-529.
Meury, J., and Devilliers, G. (1999) Impairment of cell division in tolA mutants of Escherichia coli at low and high medium osmolarities. Biol. Cell. 91: 67-75.
Miller, J.H. (1972) Experiments in Molecular Genetics. Cold Spring Harbor Laboratory Press.
Cold Spring Harbor, NY.
Möll, A., and Thanbichler, M. (2009) FtsN-like proteins are conserved components of the cell division machinery in proteobacteria. Mol. Microbiol. 72: 1037-1053.
Muller, M., Vianney, A., Lazzaroni, J.C., Webster, R., and Portalier, R. (1993) Membrane topology of the Escherichia coli TolR protein required for cell envelope integrity. J. Bacteriol.
175: 6059-6061.
Nagel de Zwaig, R., and Luria, S.E. (1967) Genetics and physiology of colicin-tolerant mutants of Escherichia coli. J. Bacteriol. 94: 1112-1123.
Ollis, A.A. and Postle, K. (2012) ExbD mutants define initial stages in TonB energization. J.
Mol. Biol. 415:237-247.
202
Prouty, A.M., Van Velkinburgh, J.C., and Gunn, J.S. (2002) Salmonella enterica serovar
Typhimurium resistance to bile: identification and characterization of the tolQRA cluster. J.
Bacteriol. 184: 1270-1276.
Rico, A.I., Garcia-Ovalle, M., Palacios, P., Casanova, M., and Vicente, M. (2010) Role of
Escherichia coli FtsN protein in the assembly and stability of the cell division ring. Mol.
Microbiol. 76: 760-771.
Siegele, D.A., and Hu, J.C. (1997) Gene expression from plasmids containing the araBAD
promoter at subsaturating inducer concentrations represents mixed populations. Proc. Natl.
Acad. Sci. USA 94: 8168-8172.
Sun, T-P., and Webster, R.E. (1986) fii, a bacterial locus required for filamentous phage
infection and its relation to colicin-tolerant tolA and tolB. J. Bacteriol. 165: 107-115.
Sturgis, J.N. (2001) Organisation and evolution of the tol-pal gene cluster. J. Mol. Microbiol.
Biotechnol. 3: 113-122.
Vankemmelbeke, M., Zhang, Y., Moore, G.R., Kleanthous, C., Perfold, C.N, and James, R.
(2009) Energy-dependent immunity protein release during Tol-dependent nuclease colicin
translocation. J. Biol. Chem. 284: 18932-18941.
203
Vianney, A., Lewin, T., Beyer Jr, W., Lazzaroni, J.C., Portalier, R., and Webster, R. (1994)
Membrane topology and mutational analysis of the TolQ protein of Escherichia coli required for the uptake of macromolecules and cell envelope integrity. J. Bacteriol. 176: 822-829.
Vianney, A., Muller, M.M., Clavel, T., Lazzaroni, J.C., Portalier, R., and Webster, R. (1996)
Characterization of the tol-pal region of Escherichia coli K-12: Translational control of tolR
expression by TolQ and identification of a new open reading frame downstream of pal encoding
a periplasmic protein. J. Bacteriol. 178: 4031-4038.
Vines, E.D., Maroida. C.L., Balachandran. A., and Valvano, M.A. (2005) Defective O-antigen
polymerization in tolA and pal mutants of Escherichia coli in response to extracytoplasmic
stress. J. Bacteriol. 187: 3359-3368.
Walburger, A., Lazdunski, C., and Corda, Y. (2002) The Tol/Pal system function requires an
interaction between the Carboxyl-terminal domain of TolA and the Amino-terminal domain of
TolB. Mol. Microbiol. 44: 695-708.
Webster, R.E. (1991) The tol gene products and the import of macromolecules into Escherichia
coli. Mol. Microbiol. 5: 1005-1011.
Weitzel, A.C., and Larsen, R.A. (2008) Differential complementation of ΔtolA Escherichia coli
by a Yersinia enterocolitica TolA homologue. FEMS Microbiol. Lett. 282: 81-88.
204
Yang, J-C., van den Ent, F., Neuhaus, D., Brevier, J., and Löwe, J. (2004) Solution structure and
domain architecture of the divisome protein FtsN. Mol. Microbiol. 52: 651-660.
Young, K.D. (2006) The selective value of bacterial shape. Microbiol. Mol. Biol. Rev. 70: 660-
703.
Zhuang, Z., Song, F., Martin, B.M., Dunaway-Mariano , D. (2002) The YbgC protein encoded
by the ybgC gene of the tol-pal gene cluster of Haemophilus influenzae catalyzes acyl-coenzyme
A thioester hydrolysis. FEBS Lett. 516:161-163.
205
Appendix
Figure S1: TolQ-dependent cell filamentation is not mitigated by concurrent expression of
TolR; whereas the TolQ paralogue ExbB does not induce filamentation when over- expressed.
Figure S1a: Stained preparations of W3110 cells bearing plasmids carrying either the ʟ-
arabinose-regulated plasmid pBAD24 (pBAD24) or pBAD24 derivatives encoding tolQ (ptolQ),
tolQR (ptolR) or exbBD (pexbBD) genes under the control of the pBAD promoter grown for 24
hours at 37˚C with aeration in Miller LB supplemented with 100 µg ml-1 ampicillin and either no
ʟ-arabinose or 0.1% (w/v) ʟ-arabinose are shown. All panels are displayed at the same relative magnification, with a bar representing 10 µm provided in each panel for scale. Figure S1b:
Immunoblot analysis of samples from cells used for stained preparations in Fig S1a (above).
Samples were prepared by TCA precipitation as described in Methods, then resolved by SDS-
PAGE on 11% polyacrylamide gels, transferred to a PVDF membrane and visualized by enhanced chemiluminescence using the a monospecific anti-TolQ antiserum and a polyspecific anti-ExbB antiserum as described in Methods. The positions of molecular mass standards are
indicated as kDa values between the developed blot. Two non-specific bands are evident in the
samples probed with anti-ExbB (indicated by “*”), with the ExbB band evident at 25 kDa.
206
207
208
209
210
211
212
213
214
APPENDIX B Approval for Use of Figure 6 Dear Mary Teleha,
Thank you for your request.
Permission is granted for you to use the material requested for your thesis/dissertation subject to the usual acknowledgements and on the understanding that you will reapply for permission if you wish to distribute or publish your thesis/dissertation commercially.
Permission is granted solely for use in conjunction with the thesis, and the material may not be posted online separately.
Any third party material is expressly excluded from this permission. If any material appears within the article with credit to another source, authorisation from that source must be obtained.
Kind Regards
Emma Willcox Permissions Assistant Wiley
-----Original Message----- From: [email protected] on www.wiley.com [mailto:[email protected]] Sent: Thursday, April 04, 2013 10:18 AM To: Permissions - US Subject: Republication/Electronic Request Form
A01_First_Name: Mary A02_Last_Name: Teleha A03_Company_Name: Bowling Green State University A04_Address: Life Sciences Building A05_City: Bowling Green A06_State: OH A07_Zip: 43403 A08_Country: U.S.A. A09_Contact_Phone_Number: 440 865 2946 A10_Fax: A11_Emails: [email protected] A12_Reference: A13_Book_Title: eLS A40_Book_or_Journal: Journal A14_Book_Author: A15_Book_ISBN: A16_Journal_Month: December 215
A17_Journal_Year: 2009 A18_Journal_Volume: N/A A19_Journal_Issue_Number: N/A A20_Copy_Pages: Figure 2 A21_Maximum_Copies: unknown A22_Your_Publisher: Bowling Green A23_Your_Title: Interaction Between TolQ and FtsN in E. coli and Implications for the Role of the Tol System in Gram Negative Bacterial Cell Division A24_Publication_Date: August 2013 A25_Format: Other A41_Ebook_Reader_Type: A26_If_WWW_URL: A27_If_WWW_From_Adopted_Book: A28_If_WWW_Password_Access: A45_WWW_Users: A29_If_WWW_Material_Posted_From: A30_If_WWW_Material_Posted_To: A42_If_Intranet_URL: A32_If_Intranet_From_Adopted_Book: A33_If_Intranet_Password_Access: A48_Intranet_Users: A34_If_Intranet_Material_Posted_From: A35_If_Intranet_Material_Posted_To: A50_If_Software_Print_Type: A60_If_Other_Type: dissertation A37_Comments_For_Request: This source has no identifiable volume or issue number. The reference is as follows: Lutkenhaus, Joe(Dec 2009) Bacterial Cell Division. In: eLS. John Wiley & Sons Ltd, Chichester. http://www.els.net [doi: 10.1002/9780470015902.a0000294.pub2]