CHARACTERIZATION OF ProQ: AN RNA BINDING PROTEIN MODULATING EXPRESSION
OF THE OSMOSENSOR AND OSMOREGULATOR ProP OF ESCHERICHIA COLI
A Thesis
Presented to
The Faculty of Graduate Studies
Of
The University of Guelph
by
MICHELLE N. SMITH-FRIEDAY
In partial fulfillment of requirements
for the degree of
Doctor of Philosophy
April, 2009
© Michelle Smith-Frieday, 2009 Library and Archives Bibliotheque et 1*1 Canada Archives Canada Published Heritage Direction du Branch Patrimoine de I'edition
395 Wellington Street 395, rue Wellington OttawaONK1A0N4 OttawaONK1A0N4 Canada Canada
Your file Votre reference ISBN: 978-0-494-55233-9 Our file Notre reference ISBN: 978-0-494-55233-9
NOTICE: AVIS:
The author has granted a non L'auteur a accorde une licence non exclusive exclusive license allowing Library and permettant a la Bibliotheque et Archives Archives Canada to reproduce, Canada de reproduce, publier, archiver, publish, archive, preserve, conserve, sauvegarder, conserver, transmettre au public communicate to the public by par telecommunication ou par Nnternet, preter, telecommunication or on the Internet, distribuer et vendre des theses partout dans le loan, distribute and sell theses monde, a des fins commerciales ou autres, sur worldwide, for commercial or non support microforme, papier, electronique et/ou commercial purposes, in microform, autres formats. paper, electronic and/or any other formats.
The author retains copyright L'auteur conserve la propriete du droit d'auteur ownership and moral rights in this et des droits moraux qui protege cette these. Ni thesis. Neither the thesis nor la these ni des extraits substantiels de celle-ci substantial extracts from it may be ne doivent etre imprimes ou autrement printed or otherwise reproduced reproduits sans son autorisation. without the author's permission.
In compliance with the Canadian Conformement a la loi canadienne sur la Privacy Act some supporting forms protection de la vie privee, quelques may have been removed from this formulaires secondaires ont ete enleves de thesis. cette these.
While these forms may be included Bien que ces formulaires aient inclus dans in the document page count, their la pagination, il n'y aura aucun contenu removal does not represent any loss manquant. of content from the thesis.
•+• Canada ABSTRACT
CHARACTERIZATION OF ProQ: AN RNA BINDING PROTEIN MODULATING EXPRESSION OF THE OSMOSENSOR AND OSMOREGULATOR ProP OF ESCHERICHIA COLI
Michelle N. Smith-Frieday Advisor: University of Guelph, 2009 Professor Janet M. Wood
The ability to both sense and respond to variations in osmolality is essential for the survival of organisms inhabiting environments of extreme or varying osmolalities. ProP is an
H+-osmoprotectant symporter that transports osmoprotectants into Escherichia coli following osmotic upshifts. In the absence of ProQ, the osmotic activation of ProP is reduced in amplitude as ProQ acts to amplify ProP activity at a post-transcriptional level. ProQ contains two distinct domains. The N-terminal domain of ProQ has a circular dichroism (CD) spectrum characteristic of alp ha-helical structure, and through homology can be modeled on the crystal structure of the alpha-helical mRNA-binding protein FinO. The C-terminal domain of ProQ has a CD spectrum characteristic of beta sheet structure and can be modeled on the structure of the RNA chaperone
Hfq. Functional characterization of these domains shows that the N-terminal domain of ProQ alone is able to partially complement aproQ deletion. Full amplification of ProP activity is not seen, however, unless both the N- and C-terminal domains of ProQ are expressed. Deletion mutations within the proU open reading frame, which encodes a second osmoregulatory transporter in E.coli with a similar substrate specificity to ProP, suppressed the proQ phenotype.
Further analysis of this phenomenon revealed that suppression occurs in the absence of the proU coding sequence, rather than in the absence of ProU transport activity. Observations presented here suggest that ProQ acts to amplify proP expression by interacting with a small non-coding RNA (sRNA) and that this sRNA is either encoded within, or regulated by the expression of the prolJ operon. Acknowledgements
I would like to thank Dr. Janet Wood for allowing me to pursue my graduate work in her laboratory. Dr. Wood's continuous support and guidance throughout my graduate work had been greatly appreciated. I would also like to thank the members of the Wood laboratory both past and present for their friendship and support. I would especially like to thank Doreen Laroque for training me in the laboratory. I am also thankful to Divya
Viswanathan and Tanya Romantsov for their help with the ProQ cellular localization experiments, Patrick Soo for his help with experiments involving the effects of RpoS and
ProQ on ProP activity and expression, Anna Potthoff and the laboratory of Dr. Lynne
Howell at the Hospital for Sick Children in Toronto for all of their help and guidance in my attempts to crystallize the N-terminal domain of ProQ, and Chad Gill for his work on the initial characterization of the suppression of the proQ phenotype. For funding my work, I would like to thank the Natural Science and Engineering Research Council of
Canada. I am also thankful to all of my friends and family members who have supported me throughout my program, especially my parents, my husband Jason and my daughters
Courtney and Madelyn. Table of Contents
Table of Contents i
List of Tables viii
List of Figures ix
Glossary of Abbreviations xi
Chapter 1: Literature Review and Research Proposal 1
1.1 Osmoregulation 1
1.1.1 Osmotic downshifts 2
1.1.2 Osmotic upshifts 6
1.2 Transcriptional regulation of proP andprolJ 12
1.2.1 proP 13
1.2.2 proU 15
1.3 Modulation of ProP activity by medium osmolality 17
1.3.1 Osmosensory mechanisms of ProP 17
1.3.2 Adaptive effects of osmolality on ProP 20
1.4 Modulation of ProP activity by ProQ 21
1.5 ProQ homologue analysis 22
1.6 ProQ structure 24
1.7 Regulation of cellular processes by small non-coding RNA (sRNA) 27
i 1.7.1 Types of sRNA interactions 28
1.7.2 Proteins involved in sRNA interactions 34
1.8 Research proposal 37
Chapter 2: Materials and Methods 42
2.1 General laboratory procedures 42
2.1.1 Laboratory equipment 42
2.1.2 Stock solutions and materials 42
2.1.3 Bacterial strains and plasmids 43
2.1.4 Culture conditions 43
2.2 Molecular biology techniques 54
2.2.1 Chromosomal DNA preparations 54
2.2.2 Plasmid DNA isolation 55
2.2.3 Agarose gel electrophoresis 56
2.2.4 Polymerase chain reaction (Brown and Wood, 1992) 56
2.2.5 Site directed mutagenesis 57
2.2.6 DNA purification following PCR amplification 58
2.2.7 Gel purification of DNA fragments 58
2.2.8 Ligation of DNA fragments 59
2.2.9 Chemical transformation 59
2.2.10 Electroporation 59
ii 2.2.11 DNA sequencing 60
2.2.12 Plasmid construction 60
2.2.13 PI mediated transductions 66
2.2.14 Inactivation of chromosomal genes 66
2.2.15 Isolation and identification of RNA co-purifying with ProQ-His6 68
2.2.16 Cloning cDNA complementary to ProQ-associated RNA fragments of
unknown sequence 70
2.2.17 Denaturing agarose gel electrophoresis 74
2.2.18 Urea acrylamide gel electrophoresis 75
2.3 Protein analysis techniques 76
2.3.1 Protein assays 76
2.3.2 SDSPAGE 76
2.3.3 Western immunoblot analysis 77
2.3.4 Western immunoblotting analysis to determine expression levels 79
2.3.5 Crude ribosome preparation 79
2.3.6 Inside out membrane vesicle preparation 80
2.3.7 Protein solubility screening 81
2.3.8 Overexpression 83
2.3.9 Protein Purification 84
2.3.10 Stokes radius determination 88
iii 2.3.11 Proteoliposome preparation 89
2.3.12 Co-reconstitution of ProP and ProQ 90
2.3.13 Mass Spectrometry 91
2.3.14 CD spectroscopy 91
2.3.15 ProQ-His6 proteolysis 92
2.3.16 Crystallization of the N-terminal domain of ProQ 94
2.3.17 P-galactosidase assays 95
2.4 Physiological techniques 95
2.4.1 Transport assays 95
2.4.2 Fluorescence microscopy 99
Chapter 3: Purification and Characterization of ProQ and ProQ-His6 100
3.1 Abstract 100
3.2 Introduction 101
3.3 Results 102
3.3.1 Creation of plasmids for the high level expression of ProQ and
ProQ-His6 102
3.3.2 Lysis conditions alter the solubility of overexpressed ProQ and
ProQ-His6 105
3.3.3 Purification of ProQ and ProQ-His6 107
3.3.4 Creation of an in-frame deletion ofproQ 109
IV 3.3.5 ProQ-His6 expressed from a plasmid complements a chromosomal
deletion 112
3.3.6 The Stokes radii of ProQ and ProQ-His6 suggest the formation of
homodimers or an extended conformation 112
3.4 Discussion 113
Chapter 4: Domain Structure of ProQ 117
4.1 Abstract 117
4.2 Introduction 118
4.3 Results 120
4.3.1 Limited proteolysis reveals a protease sensitive linker region and
supports the predicted domain structure of ProQ 120
4.3.2 The boundaries of the N- and C-terminal domains of ProQ 124
4.3.3 Creation of plasmids for the high level expression of His6-ProQN, His6-
ProQC and His6-ProQNC 124
4.3.4 Creation of plasmids for the physiological level expression of ProQN,
ProQC, ProQNC, ProQNL, ProQLC and ProQL 126
4.3.5 Overexpression and purification of His6-ProQN, His6-ProQC and His6-
ProQNC 126
4.3.6 Increasing the solubility of QH6, H6N and H6C 133
4.3.7 Analysis of the secondary structures of ProQH6, HeProQN and
H6ProQC 133
v 4.3.8 Functional analysis of the domains of ProQ 137
4.3.9 Crystallization Studies 141
4.4 Discussion 141
Chapter 5: The Cellular Role of ProQ: Macromolecular Interactions 146
5.1 Abstract 146
5.2 Introduction 146
5.3 Results 149
5.3.1 An alternative model for the C-terminal domain of ProQ 149
5.3.2 ProQ does not affect ProP on a posttranslational level via direct protein-
protein interactions 152
5.3.3 ProQ does not alter stationary-phase thermotolerance of cells 157
5.3.4 ProQ does not alter the effects of RpoS on ProP expression and
activity 158
5.3.5 Mutations at the proQ locus affect the cellular levels of ProP 163
5.3.6 ProQ alters ProP levels following plasmid based expression 166
5.3.7 ProQ binds RNA 166
5.3.8 Discussion 172
Chapter 6: Suppression of the proQ Phenotype 181
6.1 Abstract 181
6.2 Introduction 182
vi 6.3 Results 183
6.3.1 Bacterial strains and mutations 183
6.3.2 TheproQ phenotype can be suppressed 184
6.3.3 Mutation A(proP-melAB)212 does not suppress the proQ phenotype ..187
6.3.4 Deletion of loci surrounding proP does not affect ProP activity 189
6.3.5 Suppression of the proQ phenotype is independent ofproQ mutant
class 189
6.3.6 Mutations at the proU locus suppress the proQ phenotype 194
6.3.7 Deletion of proU suppresses the prog phenotype 196
6.3.8 Specific sequences in the proU operon are important for the proQ
phenotype 201
6.3.9 The proQ phenotype is suppressed by proU deletions under conditions of
high osmolality and when proP is expressed frompDC79 201
6.3.10 Mutation proU205 partially suppresses the proQ phenotype 203
6.4 Discussion 207
Chapter 7: General Discussion 216
7.1 Discussion 216
7.2 Future work 224
References 228
Appendix 1: Growth media 255
vn Appendix 2: Maps of plasmids constructed in this study 258
Appendix 3: Alignment of identified RNA sequences 271
Appendix 4: Alignment of the sequences of the insertion sequence present in the proU205 mutation with an IS5 element encoded within the genome of E. coli K-12
273
viii List of Tables
Table 1.1 Occurrence of ProP and ProQ in sequenced bacterial genomes ... 23
Table 2.1 Bacterial strains 44 Table 2.2 Plasmids used in this study 48 Table 2.3 Primers used within this study 51
Table 4.1 Masses and identities of peptides derived from tryptic ProQ fragments by in-gel trypsin and Glu-C digestions 21 Table 4.2 Plasmids and encoded proteins 127 Table 4.3 Molecular weights of purified ProQ fragments 132
Table 5.1 Identity of RNA species co-purifying with ProQ-His6 171
Table 6.1 Genotypes of strains used during identification of the proQ suppressor 187 Table 6.2 Effects of mutations on the proQ phenotype 188 Table 6.3 Functions of the products of genes surrounding the proP locus inE. coli 193
IX List of Figures
Figure 1.1 Osmoregulatory proteins in E. coli and some of their substrates ... 3 Figure 1.2 Promoter regions of the osmotically induced proP and pro U loci.. 14 Figure 1.3 Structure of ProP 18 Figure 1.4 Structural models of the putative N- and C-terminal domains of ProQ 25 Figure 1.5 Mechanism of sRNA action to regulate gene expression 30 Figure 1.6 Regulation of F-pilus formation by FinO 36
Figure 2.1 Synthesis of anRNA linker using DNA oligos 73
Figure 3.1 Western blot analysis of cell fractions to determine the cellular localization of ProQ 103 Figure 3.2 Solubility of ProQ and ProQ-His6 following overexpression 104 Figure 3.3 Solubility of ProQ-His6 is improved at increasing NaCl concentrations 106 Figure 3.4 Overexpression and purification of ProQ and ProQ-His6 108 Figure 3.5 Creation and phenotype of an in-frame proQ deletion 110 Figure 3.6 Complementation of a proQ deletion by plasmid based expression ofProQorProQ-His6 Ill Figure 3.7 Standard curve for the determination of Stokes radium (nm) 114
Figure 4.1 Limited trypsin proteolysis 121 Figure 4.2 Protein sequence alignment of ProQ and its homologues 125 Figure 4.3 Analysis of proteins purified by Ni(NTA) and gel exclusion chromatographies 128 Figure 4.4 Analysis of proteins purified by reverse phase HPLC 130 Figure 4.5 Circular dichroism spectra of ProQ-His6, HiseProQN and His6ProQC 135 Figure 4.6 Functional analysis of ProQ fragments 138 Figure 4.7 Effect of expression level on the ability of ProQN and ProQNL to complement a chromosomal proQ deletion 140
Figure 5.1 Alignment of the secondary structures of ProQC and 2QTX 150 Figure 5.2 Homology model of the C-terminal domain of ProQ 151 Figure 5.3 Effects of ProQ-His6 on ProP-His6 activity in proteoliposomes ... 154 Figure 5.4 Cellular location of ProQ 155 Figure 5.5 Effects of ProQ on the stationary phase thermotolerance of cells... 159 Figure 5.6 Effects ofproQ and rpoS onproP expression and ProP activity... 161 Figure 5.7 Effects of ProQ on ProP expression levels in cells grown at various medium osmolalities 164 Figure 5.8 Effects of ProQ on ProP levels in cells grown in exponential and stationary phases 165 Figure 5.9 ProQ modulates ProP levels following expression from pDC79.... 167
x Figure 5.10 Characteristics ofProQ-His6 preparations 169 Figure 5.11 ProQ sediments with the ribosomal fraction 173 Figure 5.12 Model for the role of ProQ in regulation oiproP expression 181
Figure 6.1 The proQ phenotype can be suppressed 190 Figure 6.2 Chromosomal loci surrounding proP 192 Figure 6.3 Effect of mutation A(proP-melAB)212 on the proQ phenotype... 194 Figure 6.4 Effects of kanamycin replacements of ORFs surrounding proP on ProP activity 195 Figure 6.5 Effect ofproQ mutant class on the proQ phenotype 197 Figure 6.6 Effects ofproU mutations on ProU activity 199 Figure 6.7 Effects of proU mutations on the proQ phenotype 200 Figure 6.8 Alignment of protein sequences of E. coli ProV and MalK based on secondary structure 202 Figure 6.9 Effects of proU mutations an the proQ phenotype following growth in high osmolality medium 206 Figure 6.10 Effects of proU mutations and the proQ phenotype when cells express ProP from plasmid pDC79 207 Figure 6.11 Modulation of ProP levels by ProQ in proU mutants 208 Figure 6.12 Characterization oftheproU205 mutation 210 Figure 6.13 Chromosomal loci surrounding proU 214 Figure 6.14 Model for suppression ofthe proQ phenotype 217
XI Glossary of Abbreviations
Abbreviations conform to the standard forms as defined by IUPAC and IUBMB. Other abbreviations are listed below.
A\|/ Membrane potential
ABC ATP binding cassette
ACN Acetonitrile
Amp Ampicillin
AP Alkaline phosphatase
APS Ammonium persulfate
BCA Bicinchoninic acid
BCCT Betaine Carnitine Choline Transporter
BCIP 5-bromo-4-chloro-3-indolyl-phosphate
C C-terminal domain of ProQ (M+Vl 80-F232)
CCD Charge Coupled Device
CD Circular Dichroism
Cm Chloramphenicol
DDM P-D-dodecylmaltoside
DEPC Diethyl pyrocarbonate
DHP 3,4-dehydro-D,L-proline
DIG Digoxigenin
ESMS Electrospray mass spectrometry
xn Faraday constant
FlAsH Fluorescein Arsenical Helix binder
FlAsH-EDT2 Fluorescein Arsenical Helix binder biarsenical reagent
FPLC Fast protein liquid chromatography
H6C Histidine tagged C-terminal domain of ProQ (MRGSH6GSM + VI80- F232)
H6N Histidine tagged N-terminal domain of ProQ (MRGSH6GS + Ml- E130)
H6NC Histidine tagged ProQ lacking the linker domain (MREGH6GS + Ml- E130 + AW + V180-F232)
HEPES 4-(2-hydroxyethyl)-1 -piperazineethanesulfonic acid
HOM High osmolality medium
HPLC High performance liquid chromatography
IS5 Insertion Sequence 5
Kav Relative elution volume
L Linker of ProQ (M+ E130-V180)
LC Linker and C-terminal domain of ProQ (M+E130-F232)
LOM Low Osmolality Medium
MALDI-TOF Matrix assisted laser desorption/ionization time of flight
MES Morpholinoethane sulfonic acid
MFS Major Facilitator Superfamily
MMLV Monkey Maloney Leukemia Virus
xin MOPS 3-[N-morpholino] propanesulfonic acid mRNA Messenger RNA
Msc Mechanosensitive channels
MWCO Molecular weight cut off
N N-terminal domain of ProQ (M1-E130)
NC ProQ lacking the linker domain (M1-E130+AW+V180-F232)
NL N-terminal domain and linker of ProQ (M1 -V180)
NRE Negative Regulatory Element
PAP Poly A polymerase
PIPES 1,4-Piperazinediethanesulfonic acid
PMF Proton Motive Force
PMSF Phenylmethylsulfonyl fluoride
PTS Phosphotransferase System
QH6 ProQ-His6 (ProQRSHHHHHH)
R Universal gas constant
RFP Red Fluorescent Protein
RP HPLC Reverse phase high performance liquid chromatography
Rs Stokes Radius
TAP Tobacco acid pyrophosphatase
TAPS N-Tris(hydroxymethyl)methyl-3-aminopropanesulfonic Acid
TFA Trifluoroacetic acid
xiv TTC 2,3,5-Triphenyltetrazolium chloride
U Units
VE Elution volume
V0 Void volume
VT Total volume
X Wavelength
xv Chapter 1: Literature Review and Research Proposal
1.1 Osmoregulation
The ability of bacteria to sense and respond to a broad range of stresses allows them to inhabit diverse ecological niches. Osmotic stress results when the osmotic pressure of the external medium becomes different from that of the cytoplasm. Bacteria such as
Escherichia coli must be able to sense and respond to both rapid changes in the osmolality of the environment, as well as to adapt to conditions of prolonged osmotic stress. The consequences of osmotic upshifts and downshifts on bacterial cells, and mechanisms to cope with osmotic stress in bacteria have been studied extensively and detailed reviews are available (Csonka and Hanson, 1991; Csonka and Epstein, 1996;
Wood, 1999; Wood, etal. 2001; Martinac, 2001; Morbach and Kramer, 2002; Poolman, et al. 2002; Poolman, et al. 2004). Since the cytoplasmic membrane is permeable to water, but not to solutes contributing to osmolality, changes in the osmotic pressure of the external environment result in rapid water fluxes across the membrane to balance the osmotic pressure. Under hypoosmotic stress, the external medium becomes dilute with respect to the cytoplasm of the cell, causing water to enter the cell. In order to maintain homeostasis, the cell must be able to rapidly export solutes and water from the cytoplasm so that the cell does not swell to the point of bursting. In hyperosmotic stress, the external medium is more concentrated relative to the cytoplasm, causing water to leave the cell. To prevent dehydration of the cytoplasm, solutes are accumulated to balance the osmotic pressure across the membrane. Under conditions of hyperosmotic stress, cells take up osmoprotectants, or organic compounds which stimulate growth at high, but not low osmolalities when provided exogenously. Some examples of organic
1 osmoprotectants and osmoregulatory systems employed by E. coli are summarized in
Figure 1.1. Microorganisms that permanently inhabit extreme environments with high salinity, such as halobacteria and haloarchaea, can employ a salt-in-cytoplasm mechanism to cope with high external solute concentrations. Salt-in-cytoplasm describes cells that maintain a cytoplasmic salt concentration (typically KC1) that is equivalent to the NaCl concentration of the external environment. In some cases, the concentration of
KC1 in the cytoplasm can accumulate to allow for growth in environments saturated with
NaCl (5.2 M) (Grant, 2004). In these cases the cells have evolved to encode proteins whose structure and function require high concentrations of salt within the cytoplasm
(Grant, 2004).
1.1.1 Osmotic downshifts
Osmotic downshifts occur when cells are transferred into a dilute medium.
Downshifts result in an influx of water into the cell, the extent of which depends on the difference in the osmotic pressure between the cytoplasm and the external medium.
Influx of water following an osmotic downshift results in: dilution of the cytoplasm, increased cell turgor, and increased membrane strain (Wood, 1999). To prevent cell bursting, cells must rapidly eliminate cytoplasmic solutes.
1.1.1.1 Solute efflux via mechanosensitive channels
Rapid elimination of cytoplasmic solutes in response to an osmotic downshift is achieved through the mechanosensitive channels (Msc) located in the cytoplasmic membrane. In response to an osmotic downshift, mechanosensitive channels are able to sense an increase strain in the membrane and open to release cytoplasmic solutes.
2 Figure 1.1: Osmoregulatory proteins in E. coli and some of their substrates A) In response to osmotic downshifts, mechanosensitive channels MscL and MscS sense strain in the plane of the membrane, and open to release water and cytoplasmic solutes. Following an osmotic upshift, potassium is accumulated via the Kdp and Trk systems. KdpFABC is a P-type ATPase whose expression is regulated by the two component system KdpDE while TrkG and TrkH are K+/H+ symporters whose activity requires TrkA, an associated NAD+ binding protein, and TrkE (SapD) the ATP binding cassette of the SapFDBCA peptide transporter. Periplasmic trehalose is converted into 2 glucose molecules by TreA. Glucose it then taken up, through the glucose phosphotransfer system (PTS), and converted to glucose-6-P. OtsA and OtsB then convert 2 molecules of glucose-6-P into trehalose. Uptake of organic osmoprotectants in response to an osmotic upshift occurs via ProP, ProU, BetT and BetU. ProP is an H+/osmoprotectant symporter belonging to the Major Facilitator Superfamily (MFS) of proteins while ProU is an osmoprotectant ABC transporter. Both ProP and ProU act to transport a broad range of organic osmoprotectants. The activity of ProP is stimulated by the cytoplasmic protein ProQ resulting in an amplification of ProP transport activity in the presence of ProQ. BetT and BetU are both members of the Betaine Carnitine Choline Transport (BCCT) family of transporters. BetT is a choline/H+ symporter and BetA and BetB are cytoplasmic enzymes that act to convert choline into glycine betaine, while BetU can take up both glycine betaine and proline betaine.
B) Structures of substrates of osmoregulatory transporters of E. coli. Solutes that are accumulated in response to osmotic stress include amino acids (e.g. proline), amino acid analogues (e.g. taurine) and methyl amines and related compounds (e.g. glycine betaine, carnitine, ectoine) (Adapted, with permission, from MacMillan, et al. 1999).
3 Cytoplasmic solutes Water MscL MscS Prof
Glucose phosphotransf system SapDBCAl Peptide ABC' Transporter H- TrkG+ K* Trkll KdpDE K' KdpFABC
n,c ^N^ XOO H,C N c0 H-^N^ ,SO , ~~ °* H II H,c' Taurine Glycine Betainc N,N-Dimethyl Glycine OH H,CX* N^COO" HjC-N COO' NH2 COO' V H,C HjC CH5 Carnitine Proline Proline Betainc
> JL N a HjCNrl COO' I TH,COO Pinecolate Ectoine 1-Carboxvmethvlpvridinium Compounds including proline, potassium, glutamate, trehalose, ATP and even small proteins have been shown to pass through the mechanosensitive channels in response to osmotic downshifts (Hamill and Martinac, 2001). Two mechanosensitive channels have been identified and characterized in E. coli, the channels with large (MscL) and small
(MscS) conductance (Sukharev, et al. 1994; Levina, et al. 1999). MscL and MscS act as emergency valves to release solutes rapidly and non-specifically in response to increased strain in the plane of the membrane (Levina, et al. 1999; Blount, et al. 1997). Cells lacking both MscL and MscS lyse upon osmotic downshifts (Levina, et al. 1999;
Nakamaru, et al. 1999). The crystal structures for the MscL and MscS channels have been solved, and the structures have given insight into the gating mechanism (Chang, et al. 1998; Bass, et al. 2002). MscL is a homoheptameric ring made up of monomers consisting of 2 transmembrane (TM) domains each. A central pore is formed by 7 TM domains, one TM from each monomer, come to a restriction or gate at the cytoplasmic face of the pore (Chang, et al. 1998). Following an osmotic downshift MscL senses alterations in the strain in the plane of the membrane and its central pore opens from 2 to
30 A, allowing for rapid efflux of cytoplasmic solutes (Cruickshank, et al. 1997; Corry, et al. 2005). MscS is a homoheptameric ring formed by monomers containing 3 TM domains each (Bass, et al. 2002). The third TM from each monomer combines to form a central pore within the heptamer, which opens in response to strain in the membrane plane. The narrowest part of the closed pore has a 3.5 A opening and a hydrophobic ring of leucine residues (Bass, et al. 2002).
5 1.1.2 Osmotic upshifts
Osmotic upshifts occur when cells are transferred into a concentrated medium which results in efflux of water from the cell. As for osmotic downshifts, the extent of water flux over the membrane following osmotic upshifts depends on the magnitude of the difference in the solute concentration across the membrane. Efflux of water from the cell following an osmotic upshift can result in: concentration of all solutes in the cytoplasmic compartment, increased macro molecular crowding, decreased cell volume, decreased cell turgor and altered membrane strain (Wood, 1999). Under extreme conditions, cells undergo a process called plasmolysis, or dehydration of the cytoplasm resulting in the cytoplasmic membrane pulling away from the cell wall. Hyperosmotic stress halts cellular processes, requiring cells to rehydrate their cytoplasm before cellular processes can continue (Meury, 1994). The response of E. coli to an osmotic upshift is more complex than its response to an osmotic downshift and depends on the availability and type of osmoprotectants in the external medium. In response to an osmotic upshift, E. coli are able to utilize extracellular potassium as well as a number of organic osmolytes, which are summarized in figure 1.1. The transport systems involved in this response are summarized in figure 1.1 as well as below.
In response to osmotic upshifts, in both the presence and absence of organic osmoprotectants, potassium available in the external medium is taken up through the Trk and Kdp systems, and cells adjust their ionic and non-ionic components accordingly
(Dinnbier, et al. 1988; Walderhaug, et al. 1992; Greie and Altendorf, 2007). The Trk system is a low affinity potassium uptake system with a KA/ for potassium of 1 mM
(Schlosser, et al. 1995) while the Kdp system has a higher affinity for potassium and is
6 active when external concentrations are low. In response to potassium uptake there is a rapid efflux of putrescine, and potassium takes on the role of the counter ion for DNA and RNA within the cell (Record, et al. 1998). Typically potassium accumulation exceeds the capacity for charge balance offered by nucleic acids and glutamate accumulates, through suppression of its use in metabolism, to counterbalance the charge of potassium (Csonka, et al. 1984; Dinnbier, et al. 1988; Cayley, et al. 1991; Yan, et al.
1996). Accumulation of potassium glutamate within the cytoplasm raises the ionic strength, inhibiting many cellular functions (Meury, 1994; Gralla and Vargas, 2006). In the absence of osmoprotectants, trehalose is synthesized from the metabolites glucose-6- phosphate and UDP-glucose. Trehalose accumulates about 30 min after potassium uptake in response to increased osmolality and replaces the majority of potassium following about two hours (Dinnbier, et al. 1988; Record, et al., 1998).
Cells subjected to osmotic upshifts in the presence of osmoprotectants initially take up potassium and accumulate glutamate, but may also take up organic osmoprotectants such as proline and glycine betaine, which can be accumulated within the cytoplasm to molar concentrations without interfering with cellular processes (Meury, 1988; Csonka, et al. 1994). Organic osmoprotectants are present and available for osmoregulation in various habitats of E. coli such as in organic debris, the intestinal tract, as well as when E. coli is ascending the urinary tract. Cytoplasmic accumulation of potassium is suggested to induce expression of osmotically regulated genes including expression from the proP and proU promoters (Sutherland, et al. 1986; Jovanovich, et al. 1989; Prince and
Villarejo, 1990; Xu and Johnson, 1997b). Osmoregulatory transporters such as ProP and
ProU take up a broad spectrum of organic osmoprotectants from the external medium,
7 raising the solute concentration of the cytoplasm and drawing water back into the cell.
ProP is a broad specificity proton/osmoprotectant symporter whose expression and
activity both increase following an osmotic upshift (Wood, et al. 2001). ProU is a broad
specificity osmoprotectant transporter belonging to the ATP binding cassette (ABC)
transporter family (May, et al. 1989; Cairney, et al. 1985a; Dattananda and
Gowrishankar, 1989; Faatz, et al. 1988; Gowrishankar, et al. 1986). Like ProP, the
expression of proU and activity of ProU increase in response to an osmotic upshift
(Cairney, et al. 1985a). Osmoprotectant uptake is also mediated by the Betaine Carnitine
Choline Transport (BCCT) family members BetT and BetU. BetT is a choline/proton
symporter, whose expression is derepressed following osmotic upshifts in the presence of
choline by the transcriptional regulator Betl. Betl also induces expression of the
cytoplasmic proteins BetA and BetB which act to convert choline into glycine betaine
(Andresen, et al. 1988; Lamark, et al. 1996). BetU acts to take up proline betaine and
glycine betaine in response to increased medium osmolality (Ly, et al. 2004). BetU is not present in all strains of E. coli and is not present in E. coli K-12 used in this study (Ly, et
al. 2004). In the presence or absence of osmoprotectants, accumulation of solutes by the
cells increases the osmotic pressure of the cytoplasm, resulting in influx of water into the
cell, restoring cellular hydration.
1.1.2.1 Potassium uptake by Trk and Kdp
In response to an osmotic upshift there is an immediate uptake of potassium from the
external environment in cells grown in medium supplemented with and without
compatible solutes (Dinnbier, et al. 1988; Record, et al. 1998). Osmotic upshifts
influence the cells' energetics, and the ionic strength, ion composition and osmolality of
8 the cytoplasm (Meury, 1994; Record, et al. 1998; Houssin, et al. 1991). Each of these could act as a signal to activate osmoregulatory mechanisms of the cell. The potassium transporters Trk and Kdp are activated in response to osmotic upshifts.
The Trk system in E. coli K-12 consists of two similar transport proteins TrkH and
TrkG and two cytoplasmic proteins TrkA and TrkE (Dosch, et al. 1991; Schlosser, et al.
1995). TrkG and TrkH each act as high rate, low affinity potassium translocators. TrkG
and TrkH share 41% amino acid identity and are thought to have arisen through gene
duplication (Schlosser, et al. 1995); in fact, many natural E. coli isolates encode only one
Trk transport system (Ly, et al. 2004; Durell, et al. 1999). Both ATP and the proton
motive force (PMF) are required for TrkG and TrkH activity (Rhoads and Epstein, 1977),
ATP is not the energy source for transport, but is thought to activate the system (Stewart,
et al. 1985), while the PMF is thought to drive potassium translocation (Stumpe, et al.
1996). TrkA is a peripheral membrane protein, required for TrkG and TrkH activity
(Dosch, et al. 1991; Stumpe, et al. 1996; Parra-Lopez, et al. 1994; Bossemeyer, et al.
1989). TrkA has been shown to contain 2 NAD+ binding sites. However it is not known
whether binding of NAD+ in vivo has functional significance (Stumpe, et al. 1996;
Schlosser, et al. 1993). TrkE null mutants abolish the transport activity of TrkH, while
TrkG activity diminishes (Dosch, et al. 1991). TrkE (SapD) was mapped to the sapFDBCA operon in which it was found to encode an ATP binding cassette of the peptide ABC transporter (Harms, et al. 2001). The requirement of TrkG and TrkH for
ATP seems to be linked to the binding of ATP, and not its hydrolysis, at the SapD
subunit (Harms, et al. 2001). The biological significance of this has not yet been
determined.
9 The Kdp system is composed of the KdpFABC transporter, a 4 subunit P-type
ATPase, and the KdpDE two component system which acts to sense potassium-limiting conditions and upregulate expression of the kdpFABC operon (Walderhaug, et al. 1992;
Brandon, et al. 2000; Polarek, et al. 1992). P-type ATPases are a family of transporters, present in prokaryotic, eukaryotic and archaeal cells, able to couple cation transport to
ATP hydrolysis (Fagan and Saier, 1994). Potassium translocation across the membrane occurs through KdpA while energy for this process is supplied by ATP hydrolysis carried out by the KdpB subunit (Greie and Altendorf, 2007). The roles of the KdpC and KdpF subunits are not as clear, but it is thought that the KdpC subunit acts as a catalytic chaperone while KdpF acts to maintain structural integrity of the transporter unit (Greie and Altendorf, 2007). Expression of kdpFABC is upregulated under conditions of
Potassium limitation as well as under conditions of osmotic stress (Voelkner, et al. 1993).
Transcription of kdpFABC is activated by the KdpDE two-component system. KdpD is a sensor kinase, which autophosphorylates in response to potassium limitation
(Walderhaug, et al. 1992; Polarek, et al. 1992). It was believed that KdpD also responded to osmotic increases, however, recent experiments suggest that KdpD does not act as an osmosensor (Hamann, et al. 2008). The phosphate on KdpD is then transferred to KdpE which allows it to activate transcription of the kdpFABC operon.
1.1.2.2 Trehalose accumulation by TreA, and OtsAB
Trehalose is a compatible solute which is synthesized by E. coli cells that are exposed to high osmolality medium in the absence of osmoprotectants (Wood, 1999). Trehalose is a disaccharide made up of 2 glucose molecules linked by a 1, 1-glycosidic linkage.
Trehalose synthesis in E. coli involves the actions of two enzymes, OtsA, a trehalose
10 phosphate synthase, and OtsB, a trehalose-phosphate phosphatase (Giaever, et al. 1988;
Kaasen, et al. 1994). Expression ofotsBA is induced by increases in medium osmolality as well as by entry into stationary phase (Hengge-Aronis, et al. 1991). Under conditions of increased osmolality, expression of treA is also upregulated. The treA locus encodes a periplasmic trehalose hydrolase, acting to hydrolyze extracytoplasmic trehalose into two glucose molecules (Boos, et al 1990; Rimmele and Boos, 1994). The glucose produced from this reaction can be taken up by the cell through the Phosphoenoylpyruvate (PEP): glucose phosp ho transfer system (PTS). If available during growth in low osmolality medium, trehalose can be taken up by E. coli as a carbon source. In this case, trehalose is taken up by the trehalose phosphotransferase system (TreB), which phosphorylates it to give trehalose-6-phosphate (Elbein, et al. 2003). Trehalose 6-phosphate is then hydrolyzed in the cytoplasm into glucose and glucose-6-phosphate by a hydrolase (TreC)
(Elbein, et al. 2003). Expression of TreBC is induced by TreR with trehalose-6- phosphate. Under conditions of increased osmolality, the genes for catabolism of trehalose are not expressed as the inducer, trehalose-6-phosphate, is hydrolyzed by OtsB
(Horlacher, et al. 1996).
1.1.2.3 Compatible solute accumulation mediated by ProP and ProU
ProP (Figure 1.1 A) is a 500 amino acid integral membrane protein with 12 membrane spanning helices, linked by hydrophilic loops, and is a member of the Major Facilitator
Superfamily of proteins (MFS), known to be involved in symport, antiport and uniport in diverse organisms (Wood, 1999). ProP is able to sense and respond to increased osmolality not only in whole cells but also as a purified protein reconstituted into proteoliposomes (Grothe, et al. 1986; Racher, et al. 1999). Thus ProP does not require
11 other cellular components, other than the phospholipid membrane, to sense and respond to an osmotic upshift. ProP has broad substrate specificity and acts to transport a range
of structurally related, zwitterionic compounds, including proline, glycine betaine and
ectoine, in response to osmotic upshifts (Figure LIB) (MacMillan, et al. 1999).
ProU is an ATP binding cassette (ABC) transporter composed of three different proteins ProV, ProW and ProX (Figure 1.1 A). ProV is the ATP binding subunit which is peripheral to the cytoplasmic surface of the cytoplasmic membrane. ProX is the periplasmic binding protein acting to recruit substrates to the transporter and ProW is the transmembrane protein which accepts the substrate from ProX and acts to transport the
compound into the cell upon hydrolysis of ATP by the ProV subunit (May, et al. 1989;
Dattananda and Gowrishankar, 1989; May, et al. 1986). The three subunits of ProU are
encoded by theproC/operon (proVWX) whose transcription increases more than 100-fold upon osmotic upshifts (Csonka and Epstein, 1996; Gowrishankar, 1989). ProU and ProP
act to transport a similar range of substrates but each has a different affinity for certain
compounds (MacMillan, et al. 1999; Barron, et al. 1987; Gouesbet, et al. 1994; Haardt, et
al. 1995). For example, ProP has a higher affinity for proline than ProU, while ProU has
a higher affinity for glycine betaine.
1.2 Transcriptional regulation of proP and proU
The initial response to an osmotic upshift is for cells to take up potassium and
accumulate glutamate (Dinnbier, et al. 1988; Cayley, et al. 1991). Upregulation of
osmotically induced genes occurs rapidly following osmotic upshifts. Transcription of various osmotically induced genes has been followed using real time PCR (Balaji, et al.
12 2005). Following an osmotic upshift imposed with 0.3 M NaCl, transcription of proU was induced within 4 min, and transcription of proP within 4-6 min following the upshock, consistent with earlier findings using lacZ fusion analysis (Balaji, et al. 2005;
Jovanovich, et al. 1988). For the osmotically induced expression of proP and proU, potassium glutamate is thought to play a key role (Jovanovich, et al. 1989; Prince and
Villarejo, 1990; Xu and Johnson, 1997b). In response to osmotic upshifts, expression of
ProP is upregulated 17-fold, while expression of ProU is increased more than 100-fold
(Csonka and Epstein, 1996; Jovanovich, et al. 1988).
1.2.1 proP
Transcription of proP in E. coli can occur from two distinct, differentially regulated promoters: PI, located at 182 bp and P2, at 95 bp upstream of the translational start site
(Figure 1.2A). Transcription from PI is c70 dependent and is maintained at a basal level throughout the entire growth cycle (Mellies, et al. 1995). Under conditions of low osmolality, binding of the cyclic AMP (cAMP) receptor protein (CRP) binds to a DNA element 121 bp upstream of the promoter and inhibits expression from the PI promoter
(figure 1.2A). When cells are exposed to an osmotic upshift, transcription from the PI promoter increases 17-fold, due to dissociation of the CRP-cAMP regulator (Xu and
Johnson, 1997b; Landis, et al. 1999).
Expression of prbP from the P2 promoter is regulated by growth phase through the actions of RpoS (a38) and a nucleoid protein, Fis (Mellies, et al. 1995; Xu and Johnson,
1995a; Xu and Johnson, 1997a). Fis binds to two sites within the proP promoter
13 PI P2 -182 -95 +1 a70 a38 proP
CRP Fis Fis -303 176 -136 B
Pi P2 -250 -60
proV
Negative Regulatory Element
Figure 1.2: Promoter regions of the osmotically induced proP and proU loci. A: Schematic diagram of the proP regulatory region which consists of two Fis binding sites, centered at -136 and -176 bp from the translational start site ofproP. The CRP binding site is located -303 bp from the translational start site. Expression from the P2 promoter is dependent on a38, as expression from PI is dependent on a70. CRP-cAMP binds to the upstream region resulting in inhibition of the PI promoter and activation of the P2 promoter (Xu and Johnson 1997; McLeod, et al. 2000). Fis binding in collaboration with a38 strongly activate expression from promoter P2 (Xu and Johnson 1995) (Adapted, with permission, from McLeod et al, 2002).
B: Schematic diagram of the proU regulatory region in E. coli. Expression of proVWX can occur from two promoters, the PI and P2 promoters located at -250 and -60 bp from the translational start site of proV. Expression from PI is dependent on a38 while expression from P2 is dependent on a70. A negative regulatory element is located within the 208 bp following the transcriptional start site from promoter P2 and provides binding sites for the nucleoid associated protein FINS (Heat-stable Nucleoid-Structuring protein) resulting in 25-fold repression of expression of proU under conditions of low osmolality (Adapted, with permission, from Gowrishankar and Manna, 1996).
14 at 176 bp and 136 bp upstream of the translational start site and, in the presence of RpoS,
induces expression from P2 50-fold (Xu and Johnson, 1995b). Cellular levels of Fis
have been shown to vary widely during growth and in response to nutrient availability
(Ball, et al. 1992; Ali Azam, et al. 1999). Upon subculture of E. coli into a rich medium,
cellular Fis levels rise rapidly due to transcriptional derepression, but then decrease as
cells enter exponential growth resulting in dilution of Fis following cell division (Ball, et
al. 1992). This results in very low levels of Fis in cells grown into stationary phase
(Ball, et al. 1992). RpoS expression, on the other hand, rises during the bacterial cells'
entry into stationary phase (Jishage and Ishihama, 1995). Since activation oiproP
expression from promoter P2 is dependent on both Fis and RpoS, a pulse of proP
expression is seen in late exponential phase from promoter P2 (Xu and Johnson, 1995a;
Xu and Johnson, 1997a).
1.2.2 proU
Transcription of proU in E. coli occurs from two promoters, PI and P2, located 250
bp and 60 bp upstream of the transcriptional start site, respectively (Figure 1.2B)
(Dattananda, et al. 1991; Manna and Gowrishankar, 1994). Expression from the pro VWX promoter increases more than 100-fold upon exposure to an osmotic upshift
(May, et al. 1989; Cairney, et al. 1985a; Jovanovich, et al. 1988; Dunlap and Csonka,
1985). Following growth in medium of increased osmolality, a steady state ofproVWX
expression is reached which is 27-fold higher than steady state expression in the absence
of an osmotic upshock (Jovanovich, et al. 1988). Addition of exogenous glycine betaine
to the growth medium during exposure to an osmotic shock decreases the duration and
the steady state level of proVWX expression (Cairney, et al. 1985a; Gowrishankar, et al.
15 1986; Jovanovich, et ah 1988). Each of the promoters of proVWX are osmoregulated
with transcription from PI increasing 6-fold and transcription from P2 increasing 8-fold
in response to an osmotic upshift (Dattananda, et ah 1991).
Transcription from the PI promoter is minor in comparison to that from the P2
promoter (Gowrishankar, 1989; Dattananda, et ah 1991). Transcription from the P1
promoter appears to be RpoS dependent, when the PI promoter region has been deleted
(Manna and Gowrishankar, 1994). In the presence of the P2 promoter, RpoS induction of proVWXexpression from promoter PI is not seen (Manna and Gowrishankar, 1994). In
the closely related organism S. enterica transcription of the proVWX operon only occurs
from one promoter, which is similar to the P2 promoter in E. coli (Zhang, et ah 1996;
Overdier, et ah 1989).
Expression from promoter P2 of pro VWX in E. coli is a70 dependent (Ueguchi and
Mizuno, 1993). In both Salmonella and E. coli, the region of DNA downstream of the
P2 promoter, extending into the first structural gene iproV) is a negative regulatory
element (NRE) located within sequences extending about 300 bp downstream of the P2
promoter (Figure 1.2B) (Dattananda, et ah 1991; Overdier and Csonka, 1992). This NRE
acts to repress proVWX expression from the P2 promoter at low osmolalities (Dattananda,
et ah 1991; Overdier and Csonka, 1992). The NRE is thought to act through interactions
with the DNA binding protein HNS (heat stable nucleoid structuring protein) (May, et ah
1990; Lucht, et ah 1994; Jordi and Higgins, 2000). Interactions of the NRE with HNS
result in tight supercoiling of the DNA within this region, resulting in lowered proVWX
expression (Jordi and Higgins, 2000; Nagarajavel, et ah 2007). When the potassium
glutamate concentration in the cytoplasm increases, the supercoiling within the NRE
16 relaxes due to dissociation of HNS, resulting in increased expression of pro VWX (Jordi and Higgins, 2000; Rajkumari, et al. 1996). The amount of HNS bound to the NRE seems to correlate with promoter activity as when the promoter is more active, fewer
HNS proteins are found bound to this region, while under conditions giving low promoter activity, the NRE becomes more populated by HNS molecules (Nagarajavel, et al. 2007).
In cells containing chromosomal deletions at the hns locus expression ofproVWX remains osmoregulated and thus interaction of HNS with the NRE of proVWX is not the only mechanism regulating the osmotic response ofproU expression (Fletcher and
Csonka, 1995; Lucht, et al. 1994).
1.3 Modulation of ProP activity by medium osmolality
1.3.1 Osmosensory mechanisms of ProP
ProP is an osmoregulatory transporter that is able to both sense and respond to increased medium osmolality by transporting a broad range of structurally related osmoprotectants (Wood, 1999). Both the expression and activity of ProP are osmoregulated, and both the K„ and Vmaxof ProP increase with increasing medium osmolality (Racher, et al. 2001). ProP consists of 12 TM helices with both the N- and C- terminal domains of the protein in the cytoplasm (Culham, et al, 2003). Recently, the structure of ProP was modeled on the crystal structure of GlpT (1PW4) (Figure 1.3A)
(Wood, et al. 2005; Lemieux, et al. 2005). GlpT is also an MFS member, functioning as a glycerol-3-phosphate/phosphate antiporter (Lemieux, et al. 2005). The structural model of ProP reveals the presence of 2 helical bundles within the membrane linked by a long
17 Figure 1.3: Structure of ProP. A: Structural model of ProP (Wood, et al. 2005; Protein Data Bank entry 1Y8S) ProP (left) was modeled on the known crystal structure of GlpT (right) (Lemeiux et al, 2005; 1PW4). Hydrophobic residues are yellow, polar uncharged residues are green, positively charged residues are blue and negatively charged residues are red (From Wood, et al. 2005 with permission). B: Structure of the C-terminal anti-parallel coiled-coil of ProP (residues 468-497) as determined by Nuclear Magnetic Resonance (NMR) spectroscopy (Zoetewey, et al. 2003; Protein Data Bank entry 1R48). Helices contributed by the cytoplasmic C-terminus of each ProP monomer are coloured in green or orange with the N-termini lighter than the C- termini. The coiled-coil structure is maintained in the antiparallel orientation by salt bridges formed between R488 on one strand and D475 and D478 on the other (Culham, et al. 2000; Tsatskis, et al. 2008) (coloured in blue and red, respectively).
18 cytoplasmic loop; the first bundle is composed of helices 1-6 and the second of helices 7-
12 (Wood, et al. 2005). Transport of organic osmoprotectants by ProP is powered by the proton motive force (PMF) (MacMillan, et al. 1999; Racher, et al. 2001) and a putative proton translocation pathway has been identified in the N-terminal helix bundle of ProP
(Wood, et al. 2005).
Mutations resulting in ProP activity that is insensitive to osmotic upshifts have also been identified. These include ProP Y44M and S62C (Wood, et al. 2005; Culham, et al. 2008). In each of these cases, the transport activity is retained; however, the activity is independent of the external osmolality. Based on these observations, a model
for the mechanism of osmosensing by ProP has been suggested. It is proposed that key residues in ProP are overhydrated under conditions of low osmolality leading to low ProP
activity. As the osmolality of the medium increases, water is drawn out of the protein, resulting in optimal hydration of key residues and increased ProP transport activity
(Wood, 2007). The C-terminal domain of ProP consists of approximately 62 amino acids, which extend into the cytoplasm (Liu, et al. 2007; Culham, et al. 2000). Within this C- terminal domain is a region of 29 amino acids that are predicted to form homodimeric,
antiparallel, a-helical coiled-coils (Culham, et al. 2000). A peptide corresponding to residues 468 to 497 of ProP has been shown by NMR to form antiparallel coiled coils
(Figure 1.3B) (Zoetewey, et al. 2003). Mutants lacking these C-terminal amino acid residues or that have amino acid replacements that disturb the coiled-coil structure are expressed by the cell, but have an increased threshold for osmotic activation (Culham, et al. 2000; Tsatskis, et al. 2005).
19 1.3.2 Adaptive effects of osmolality on ProP
ProP is able to adapt its response to growth of E. coli in medium with increased osmolality so that transport of compatible solutes activates at a higher medium osmolality following growth in high osmolality medium and a lower osmolality following growth in low osmolality medium (Tsatskis, et al. 2005). Growth in media with elevated osmolality leads to alteration of the phospholipid content of the membrane. The cytoplasmic membrane of E. coli is composed of phospholipids with three types of head groups, phosphatidylethanolamine (PE) which has no net charge, phosphatidylglycerol
(PG), with a net charge of-1 and cardiolipin or di-phosphatidylglycerol (CL), with a net charge of-2. Typically, the cytoplasmic membrane is composed of 75 mol% PE, 20 mol% PG and 5 mol% CL (Cronan, 2003). When cells are grown at elevated osmolalities, the CL content rises more than 2-fold at the expense of PE due to increased transcription at the cardiolipin synthase (els) locus (Tsatskis, et al. 2005). Although the increase in cardiolipin content seems small, it has been shown that cardiolipin concentrates at the poles in E. coli cells (Mileykovskaya, et al. 2001; Romantsov, et al.
2007). If ProP were to localize at the poles of the cell in the CL rich environment, the increased CL content may act as a signal for ProP adaptation (Tsatskis, et al. 2005).
Using fluorescence microscopy, it was found that ProP localized to the poles of E. coli cells, and that this polar localization was dependent on the CL content (Romantsov, et al.
2007). This is in contrast to the lactose permease LacY, also a member of the MFS, which also localizes at the poles of E. coli, however, unlike that of ProP, the localization of LacY is not dependent on the cardiolipin content of the membrane (Romantsov, et al.
2008). It is proposed that the osmotic adaptation phenomenon of ProP occurs as the C-
20 terminal domain of ProP interacts with increasing concentrations of cardiolipin, resulting in an increased activation threshold of the transporter (Romantsov, et al. 2008).
1.4 Modulation of ProP activity by ProQ
A trans-acting cytoplasmic protein, ProQ, has been found to affect the activity of
ProP. Initial experiments aimed at identifying additional proline uptake systems in E. coli, using selection of mutants with increased resistance to a toxic proline analog, 3,4- dehydro-D,L-proline (DHP), resulted in selection of a mutant showing decreased proline uptake under conditions of low osmolality (Stalmach, et al. 1983). This mutation was presumed to be in the locus encoding ProP (Stalmach, et al. 1983). ProP activity returned, however, when cells were cultivated in high osmolality medium (Milner, et al.
1988; Milner and Wood, 1989). The mutation was mapped to a locus not linked to proP in the E. coli chromosome (Milner and Wood, 1989). The proQ locus was predicted to encode protein, ProQ, a 232 amino acid, cytoplasmic protein with a pi of 9.7 (Kunte, et al. 1999). In the absence of ProQ, ProP activity is not abolished, but it is attenuated at each medium osmolality (Milner and Wood, 1989; Kunte, et al. 1999; Smith, et al. 2004).
Initial experiments suggested that ProQ did not significantly alter the expression of a proPwlacZ fusion (Milner and Wood, 1989). Further experiments found that ProP protein levels were not altered by aproQ mutation, as determined by Western blotting analysis (Kunte, et al. 1999). On this basis it was proposed that ProQ acts to amplify
ProP activity on the post-translational level, possibly through direct protein-protein interactions which would stabilize an active conformation of the transporter (Wood,
1999; Kunte, et al. 1999).
21 1.5 ProQ homologue analysis
Putative orthologues of ProP and ProQ were sought by searching the sequenced microbial genomes included in the NCBI database using BLAST, representative results are shown in Table 1.1. Not all genomes encode putative orthologues of these proteins and, where proQ and proP do co-occur, they are not adjacent. It is difficult to unambiguously identify ProP orthologues given the high background sequence similarity among members of the major facilitator superfamily (MFS), regardless of their function.
For example, transporters ProP of E. coli and Corynebacterium glutamicum are both osmosensors and osmoregulatory betaine transporters although they share only 39% sequence identity (Peter, et al. 1998). On the other hand, KgtP and ShiA of E. coli are also quite similar in sequence to ProP (30% and 27% identity, respectively) yet KgtP is an a-ketoglutarate transporter (Seol and Shatkin, 1993) and ShiA is a shikimate transporter (Whipp, et al, 1998). ProP is certainly present without ProQ in at least one organism (C. glutamicum) (Peter, et al. 1998) (Table 1.1). Putative ProQ orthologues exist in many Gram-negative bacteria, but none has been found in a Gram-positive organism (Table 1.1). ProQ may also be present without ProP, since the putative ProP orthologues identified in some Gram-negative organisms (e.g., those in Pasteurella multocida, Haemophilus influenzae, Shewanella oneidensis, Photorhabdus luminescens,
Vibrio cholerae and Haemophilus ducreyi, Table 1.1) are no more similar in sequence to
ProP than is ShiA. Thus the role of ProQ may not be limited to the regulation of ProP.
22 Table 1.1: Occurrence of ProP and ProQ in sequenced bacterial genomes1 putative orthologue Gram organism ProP ProQ stain accession % identity % overlap accession % identity % overlap number number - Escherichia coli MG1655 NP 418535 100 100 YP 026161 100 100 - Shigella flexneri ABF06105 99 100 YP 688896 99 100 - Salmonella enterica serovar AAL23114 97 100 NP460802 99 100 Typhimurium - Citrobacter koseri ABV14830 95 100 ABV12283 91 99 - Enterobacter cancerogenus ZP 03280665 92 100 YP 001907454 86 98 - Klebsiella pneumoniae YP 002240935 91 100 YP 002237782 83 97 - Serratia proteamaculans YP 001478091 91 100 YP 001478347 71 97 - Erwinia tasmaniensis YP 001909206 84 100 YP001907454 70 99 - Pseudomonas putidia NP 745058 82 99 - Agrobacterium tumefaciens AAL45120 65 98 NP_059699 40 31 - Xanthomonas campestris NP 635478 52 92 - Yersinia pestis YP 00465 32 85 NP 404107 74 100 - Pasteurella multocida NP 246196 30 85 AAK02352 45 99 - Haemophilus influenzae NP 438499 29 83 P44286 43 97 - Shewanella oneidensis NP 719559 26 27 NP 718188 39 100 - Photorhabdus luminescens NP 930365 23 71 NP 929918 63 100 - Vibrio cholerae NP 233058 22 72 AAF94652 53 99 - Haemophilus ducreyi NP 873340 21 86 NP 873499 43 63 + Streptomyces coelicolor NP 626583 51 87 + Bacillus cereus NP 979359 43 91 + Corynebacterium glutamicum NP 602258 39 92 + Mycobacterium avium NP 959067 39 89 + Staphylococcus aureus NP 373784 37 89 ' Putative ProP or ProQ orthologues were found using BLAST searches (blastp) of microbial genomes available in the NCBI database (November 2008). The above list is a selected number of these BLAST hits that represent the trends of the findings. Percent overlap is the number of residues of the query sequence aligned continuously divided by the length of the query sequence. Putative ProP orthologues with sequence identity lower than 37% are listed only if a putative ProQ orthologue was also detected. Among the ProP sequences listed, ProP function has been demonstrated for the proteins from E. coli MG1655 (Grothe, et al. 1986), S. enterica serovar Typhimerium (Caimey, Booth and Higgins 1985b), A. tumefaciens (Tsatskis, et al. 2008), and C. glutamicum (Peter, et al. 1998). (Adapted with permission from Smith, et al. 2004).
23 1.6 ProQ structure
Prior to this work, the exact function and mechanism of the ProQ protein were unknown. Although putative ProQ orthologues exist in many bacteria, none of these have been analyzed on a structural or functional level. In silico analysis of ProQ orthologues predicts a 2-domain structure for ProQ. This structure consists of a mostly a- helical, 130-amino acid N-terminal domain, linked by a 50-amino acid, unstructured region, to a mostly P-sheet, 52-amino acid C-terminal domain (Smith, et al. 2004).
BLAST analysis of full length ProQ revealed 25% sequence identity to FinO, an
RNA-binding protein involved in regulation of F-pilus formation in E. coli (Kunte, et al.
1999). The function of FinO has been extensively studied and the structure of this protein (1DVO) has been solved (Ghetu, et al. 2000). Although ProQ and FinO share limited sequence identity, the predicted secondary structure of the N-terminal 130 amino acids of ProQ aligns almost perfectly with the secondary structure of FinO (Smith, et al.
2004). Based on this alignment, Dr. R.A.B. Keates modeled the structure of the N- terminal domain of ProQ on FinO (Figure 1.4) (Smith, et al. 2004). The function of FinO is summarized below (Introductionl.7.2.1).
In collaboration with our laboratory, Dr. R.A.B. Keates attempted to identify a structure suitable for homology modeling of the C-terminal domain of ProQ. While no one protein gave a strong match to ProQ, a single structural type dominated the top 20 structural hits (3D PSSM; Kelley, et al. 2000), consisting of a five-stranded P-meander either classified as an SH3-like domain (9 out of 20 hits) or a closely related Sm motif (3 out of 20 hits) which is superimposable on an SH3-like domain over 3.5 of the 5 P-
24 N FinO-like ! Linker j SH3-like C
t:.
Figure 1.4: Structural models of the putative N- and C-terminal domains of ProQ. The putative N-terminal domain of ProQ corresponds to residues 1-124 which are predicted to be mostly a-helical in nature. This domain can be modeled on the crystal structure of FinO, an mRNA binding protein involved in F-pilus biogenesis in E. coll The N-terminal domain is linked to the C-terminal domain by a 50 amino acid linker, predicted to be unstructured. The putative C-terminal domain, corresponding to amino acid residues 179-232, is predicted to be mostly P-sheet and its structure can be modeled on the SH3-like domain of the Dictyostelium myosin motor domain (1LVK) (Adapted with permission from Smith, et al. 2004).
25 strands, but differs as the SH3-like domain contains a short 3io helix while the Sm proteins do not (Smith, et al. 2004; Scofield and Lynch, 2008). The structure of the SH3- like domain of the Dictyostelium myosin motor domain (1LVK) (Bauer, et al. 1997) was chosen as template for modeling ProQ C-terminal domain (Figure 1.4) as the predicted secondary structure of ProQ closely matched the target structure with almost gap less alignments (Smith, et al. 2004). Typically consisting of 50 to 60 amino acids, SH3 domains share a common fold of a (3-barrel consisting of 5 to 6 antiparallel P-strands
(Mayer, 2001). SH3 domains are involved in protein-protein interactions and tend to bind proline rich regions on their targets (McPherson, 1999). Sm proteins are RNA binding proteins found in every organism to date (Scofield and Lynch, 2008). Sm proteins form homohexameric or homoheptameric ring structures called toroids (Scofield and Lynch, 2008). The C-terminal domain of ProQ (amino acid residues 180-232) was originally modeled on the SH3-like domain of the Dictyostelium discoideum Myosin
Motor domain (1LVK) (Figure 1.4) (Smith, et al. 2004; Bauer, et al. 1997). The homology model of the C-terminal domain on the SH3-like domain is much more speculative than the model of the N-terminal domain on the structure of FinO. The
1LVK template was chosen in this case as it provided a structural alignment which contained only 1 gap in a loop region of the template. It is possible that the C-terminal domain adopts an Sm fold and further experimentation is needed to determine if the homology model based on the SH3-like domain is accurate.
Modeling work described above suggests that the N- and possibly the C-terminal domains of ProQ show structural similarity to RNA binding proteins; the N-terminal domain to FinO and the C-terminal domain to Sm proteins such as Hfq. Regulation by
26 short, non-coding RNA (sRNA) has been extensively studied in eukaryotic organisms, and more recently, it has become a common theme of regulation in prokaryotes
(Barrandon, et al. 2008; Repoila and Darfeuille, 2009; Argaman, et al. 2001; Vogel and
Papenfort, 2006; Bejerano-Sagie and Xavier, 2007). Interestingly, regulation of gene expression through the actions of sRNAs has been shown to play a major role in regulating genes involved in virulence and adaptive responses, as well as to regulate expression of many membrane proteins (Argaman, et al. 2001; Vogel and Papenfort,
2006; Bejerano-Sagie and Xavier, 2007; Toledo-Arana, et al. 2007; Eddy 2001). It is possible that ProQ acts to modulate ProP activity by mediating RNA-RNA interactions within the cell, linking ProP expression to other cellular processes.
1.7 Regulation of cellular processes by small non-coding RNA (sRNA)
In eukaryotic systems, gene expression can be regulated by RNA molecules called microRNAs which act through antisense base pairing with a target mRNA, resulting in degradation of the RNA hybrid. Recent evidence suggests that many processes in prokaryotes are also regulated by regulatory RNA molecules called small, non-coding
RNA (sRNAs). Searches of the E. coli genome for sRNAs have revealed the presence of more than 70 so far and it is predicted that the E. coli genome may encode as many as
200 sRNA molecules (Altuvia, 2007; Vogel and Wagner, 2007). Although many sRNAs have been identified, the function of only a small subset have been determined (Storz,
2002; Storz, et al. 2005).
27 1.7.1 Types of sRNA interactions
Regulation by sRNAs can occur in 2 ways: by providing alternate binding sites for proteins modulating transcription or translation of target genes (see CsrA example below), or by interacting directly with mRNA resulting in sRNA-mRNA duplexes that up regulate or down regulate translation of the target mRNA (see the RhyB and RpoS examples below). Direct interactions of sRNA with mRNA can fall into two categories, cis-acting sRNA and trans-acting sRNA. Cis-acting sRNAs are complementary to their target mRNA, as they are typically encoded on the DNA strand complementary to their target coding sequence. Trans-acting sRNAs are typically encoded within intergenic regions and show little direct sequence complementarity to their mRNA targets (Vogel and Wagner, 2007). The low sequence similarity between the sRNA and mRNA makes the targets of many identified sRNAs and the sRNAs themselves difficult to predict. In many cases, binding of an sRNA to a target mRNA is mediated by the RNA chaperone
Hfq (Host factor Q) (discussed below) (Valentin-Hansen, et al. 2004).
Binding of trans-acting sRNA to a target can have two effects: down regulation of translation due to mRNA degradation, or upregulation of translation due to changes in the structure of the mRNA around the ribosome binding site (Gottesman, 2004; Gottesman, et al. 2006; Storz, et al. 2004). To down regulate translation, the sRNA can bind to a region of the mRNA that blocks translation, usually sequences in the 5' untranslated region (UTR) involved in ribosome binding, but can also involve sequences within the coding sequence of the mRNA. The resulting sRNA-mRNA duplexes are rapidly degraded (Gottesman, 2004; Storz, et al. 2004). Alternatively, binding of sRNA can increase translation of a target gene. In this case, the structure of the 5'UTR of the
28 mRNA would normally prevent its own translation as RNA stem loops and hairpins block the ribosome binding site (Gottesman, 2004; Gottesman, et al. 2006). Interaction with an sRNA can alter the structure of the 5'UTR, exposing the ribosome binding site and upregulating transcription.
1.7.1.1 Modulation of a transcriptional regulator CsrA by a frans-encoded sRNA,
CsrB
The CsrA protein acts as a global regulator of carbon metabolism in E. coli, its expression increasing following entry into stationary phase (Romeo, et al. 1993; Sabnis, et al. 1995). Target mRNAs modulated by CsrA encode proteins involved in glycogen synthesis, glycolysis, and gluconeogenesis (Romeo, et al. 1993; Sabnis, et al. 1995;
Baker, et al. 2002). The CsrA protein regulates target genes through protein-RNA interactions, modulating the stability of the target mRNA by blocking the ribosome binding site (Figure 1.5 A) (Liu and Romeo, 1997; Dubey, et al. 2003). CsrB was discovered as an RNA co-purifying with CsrA and it was determined to have a negative effect on the ability of CsrA to regulate its target mRNAs (Liu, et al. 1997). Each CsrB sRNA was found to contain 22 potential binding sites for the CsrA protein, although the maximum number found bound to each sRNA was closer to 9 (Figure 1.5A) (Liu, et al.
1997). Transcription of CsrB is under the control of the two-component regulatory system BarA-UvrY (Suzuki, et al. 2002), and transcription of CsrB results in the production of alternate binding sites for the CsrA protein. Binding of CsrA to the CsrB sRNA results in upregulation of expression of the CsrA target mRNA as CsrB competes with the mRNA for binding to CsrA (Figure 1.5A).
29 Figure 1.5: Mechanism of sRNA action to regulate gene expression. A: Mechanism of CsrB. (Left) Protein CsrA acts to block translation of mRNAs encoding proteins involved in carbon storage and utilization during entry into stationary phase. (Right) The CsrB sRNA contains multiple binding sites for the CsrA protein. Expression of CsrB results in derepression of mRNA targeted by CsrA as CsrA is removed from the 5' untranslated region (UTR), exposing the ribosome binding site (Liu et al, 1997; Duby et al, 2003).
B: Regulation of protein synthesis by RhyB during iron limitation. (Left) under conditions of iron sufficiency, the Fur protein binds to and represses transcription of rhyB. (Right) when iron becomes limiting, the sRNA RhyB is expressed and it interacts with mRNAs encoding proteins with an iron containing co-factor, resulting in down regulation of expression and increases cytoplasmic iron availability. Interaction of RhyB with its mRNA targets depends on the RNA chaperone Hfq (Masse and Gottesman 2002; Masse et al, 2005).
C: Regulation of rpoS translation by DsrA, RprA and OxyS sRNAs. (Left) The 5'UTR of rpoS mRNA forms a secondary structure which blocks the ribosome binding site. (Right) Expression of the sRNAs DsrA (at low temperatures) or RprA (under cell surface stress) results in their Hfq-dependent interaction with the 5'UTR of rpoS. These sRNA-mRNA interactions alter the structure of the 5'UTR of rpoS, exposing the ribosome binding site, and leading to increased levels of translation. Expression of OxyS sRNA (in response to oxidative stress) down regulates rpoS expression. It is hypothesized that OxyS competes with DsrA and RprA for available Hfq chaperones, resulting in dissociation of DsrA and RprA from the 5' UTR of rpoS and formation of the inhibitory 5' secondary structure (Majdalani et al. 2002; Storz et al. 2004; Zhang et al, 1998).
30 mRNA
No Translation Translation
B -0- ryhB ryhB Notrancription Transcription
RyhBsRNA I Hfq Target mR Degradation
No Translation Translation Hfq UxySsRNA rpo5nR\A DVACT No Translation R:iiA-.RMA
31 1.7.1.2 RyhB regulation of proteins with iron-containing co-factors
RyhB is a 90-nt sRNA of E. coli that regulates the expression of a number of proteins containing iron co-factors under conditions of iron limitation (Masse and Gottesman,
2002). RyhB is an example of a trans-encoded sRNA with multiple mRNA targets. Iron is an important element for cellular metabolism as it is part of many enzyme co-factors.
When cells are in an iron-rich environment, the Fur (ferric uptake regulator) protein represses transcription of genes encoding proteins involved in iron uptake (Masse and
Gottesman, 2002; Vassinova and Kozyrev, 2000). Under these iron-rich conditions, proteins with an iron-containing co-factor are upregulated; however this regulation is not directly dependent on Fur as Fur cannot bind to the promoter regions of these genes
(Hantke, 2001). RyhB is an sRNA that was identified as being repressed by Fur under iron-rich conditions, with its expression activated under iron-limiting conditions (Masse and Gottesman, 2002). RhyB base pairs with several mRNAs, each encoding proteins with iron components, resulting in degradation of their messages and downregulation of protein expression (Figure 1.5B) (Masse, et al. 2005). Binding of RhyB to target mRNAs depends on the RNA chaperone Hfq (Masse, et al. 2003; Geissmann and Touati, 2004).
At least 18 mRNA targets for RhyB have been identified to date and all encode proteins with iron containing co-factors that are not essential to the cell (Masse, et al. 2005). This regulation makes iron available for essential iron containing proteins during iron-limiting conditions.
32 1.7.1.3 RpoS regulation by the trans-encoded sRNAs DsrA, RprA and OxyS
RpoS is an alternate sigma factor (a38) in E. coli which is expressed during entry into stationary phase, as well as following exposure of the cell to a variety of stresses
(Hengge, 2008). RpoS associates with RNA polymerase to activate transcription of up to
10% ofE. coli genes, most of which are involved in stress responses or are important for stationary phase survival (Hengge-Aronis, 2002a; Klauck, et al. 2007). The complex regulation of rpoS expression and RpoS levels has been reviewed (Hengge, 2008;
Repoila, et al. 2003; Hengge-Aronis, 2002b). Expression of rpoS expression is regulated by at least three different sRNAs, DsrA, RprA and OxyS (Sledjeski, et al. 1996;
Majdalani, et al. 2001; Zhang, et al. 1998). Regulation of RpoS illustrates the regulation of a target gene by multiple sRNAs in response to a variety of growth conditions.
DsrA is an 85-nt sRNA whose expression and stability increase 30-fold at low temperatures (Sledjeski, et al. 1996; Repoila and Gottesman, 2001). RprA is a 105-nt sRNA whose expression is activated by the RcsC/YojN/RcsB phosphorelay system in response to cell surface stress (Majdalani, et al. 2002). DsrA and RprA both increase rpoS expression in ways that depend on the RNA chaperone Hfq (Brescia, et al. 2003).
The 5' UTR of the rpoS mRNA contains a hairpin structure which causes it to base pair with itself, blocking the ribosome binding site and preventing translation (Brown and
Elliott, 1996) (Figure 1.5C). DsrA and RprA base pair with the 5'UTR, resulting in exposure of the ribosome binding site and translation of the message (Figure 1.5C).
OxyS is an sRNA that accumulates within cells following oxidative stress due to transcriptional regulation by OxyR (Altuvia 2007; Zhang, et al. 1997). Upregulation of
33 OxyS sRNA decreases RpoS levels (Zhang, et al. 1998). The mechanism by which OxyS acts on rpoS is unclear, but OxyS may compete with DsrA and RprA for available Hfq chaperones, resulting in down-regulation of RpoS levels (Altuvia, 2007; Zhang, et al.
1998).
1.7.2 Proteins involved in sRNA interactions
In cases where trans-sRNA act on target mRNA molecules, the lack of sequence similarity can limit the affinity of the RNA-RNA duplex. In many cases, an RNA- binding protein is required to stabilize these RNA-RNA interactions. Below, two RNA binding proteins are described. FinO is an mRNA-binding protein that is exclusively involved in regulating F-pilus formation by stabilizing the FinP sRNA in vivo (van
Biesen and Frost, 1994). The second is Hfq, a promiscuous RNA-binding protein that has been identified as essential for sRNA-mRNA interactions in many different processes
(Brown and Elliott, 1996; Aiba, 2007; Valentin-Hansen, et al. 2004a; Moller, et al. 2002;
Muffler, et al. 1996; Zhang, et al. 2002). Interestingly, the N-terminal domain of ProQ can be modeled on the crystal structure of FinO, while the C-terminal domain of ProQ can be modeled on the structure of Hfq. If these models were to suggest a function for
ProQ, then it is possible that ProQ acts as an RNA binding protein with the N-terminal domain binding a specific target and the C-terminal domain acting as a more promiscuous RNA binding protein.
1.7.2.1 FinO
Bacterial conjugation, or the process of transferring DNA directly from one bacterial cell to another, involves formation of an F-pilus, comprised of proteins encoded in the tra
34 operon (Lawley, et al. 2003). Expression of the tra operon is controlled by the transcriptional activator TraJ (Lawley, et al. 2003). TraJ levels are regulated by FinP, a cis encoded sRNA able to base-pair with the 5'UTR of the target traJmRNA
(Koraimann, et al. 1996). FinP stability is modulated by its association with the RNA binding protein FinO. FinO binding to FinP increases its half-life 10-fold (Jerome, et al.
1999). FinO also promotes duplexing between the FinP sRNA and traJmRNA, resulting in degradation of the traJ message and preventing activation of the tra operon and F-pilus formation (Figure 1.6) (van Biesen and Frost, 1994). In the absence of FinO, FinP is unstable, and it is no longer able to form FinP-traJ RNA duplexes, resulting in translation of traJ. TraJ can then activate transcription of the tra operon, leading to F-pilus formation and conjugative transfer of DNA (Figure 1.6).
1.7.2.2 Hfq
Hfq is a member of the Sm family of proteins, which form hexameric ring structures when they interact with RNA molecules (Valentin-Hansen, et al. 2004). Unlike FinO, which has a specific sRNA target, Hfq mediates a wide range of sRNA-mRNA interactions involved in regulation of diverse functions, within bacteria and archea. To date, genes encoding Hfq orthologues have been identified in half of all sequenced bacterial and archeal genomes (Valentin-Hansen, et al. 2004; Sun, et al. 2002). The crystal structures of Hfq orthologues from a variety of bacteria and archea have been solved, each revealing a similar structure (Schumacher, et al. 2002; Nielsen, et al. 2007;
Nikulin, et al. 2005; Sidote, et al. 2004; Numata, et al. 2004; Sauter, et al. 2003). Hfq binds poly(A) regions of RNA and AU rich regions of RNA either preceding or following
35 o Transfer of F plasmid I HS hooo finP tra operon
Degradation OO
B oo
Degradation
Tra Genes are not expressed F plasmid is not transferred tra J Kxx: tra operon
Figure 1.6: Regulation of F-pilus formation by FinO. A: In the absence of FinO, Both traJ and FinP are expressed. traJ encodes a regulatory protein, TraJ, which activates expression of the tra operon, resulting in F-pilus formation and conjugative transfer of DNA. FinP is an sRNA that interacts with traJ mRNA, suppressing its translation. FinP is unstable and, in the absence of FinO, its cellular levels do not rise, so it does not act on traJ. B: In the presence of the RNA binding protein FinO, the FinP sRNA is stabilized. This leads to an increased level of FinP in the cell. FinO also promotes FinP :traJ duplexing, resulting in degradation of the traJ message. This results in very low levels of TraJ within the cell, and prevents induction the tra operon. This results in suppression of F- pilus formation and conjugative transfer of DNA.
36 stem-loop structures (Brescia, et al. 2003; Senear and Steitz, 1976; Mikulecky, et al.
2004; Sun and Wartell, 2006). To date, 22 different sRNAs have been identified as interacting with Hfq (Brennan and Link, 2007). It is thought that Hfq stabilizes important stem-loop structures of sRNAs, allowing them to act on their target mRNA molecules through base pairing interactions (Geissmann and Touati, 2004; Moll, et al. 2003).
1.8 Research proposal
Within this work, I set out to define the osmoregulatory role of ProQ in E. coli. Prior to this work, it was known that disruptions at the proQ locus impaired ProP activity
(Milner and Wood, 1989; Kunte, et al. 1999); however, the mechanism by which ProQ affected ProP was unknown.
My goal in this research was to identify the mechanism by which ProQ was able to amplify the activity of ProP. As mentioned, ProQ homologues exist, but none have a known function other than "ProQ". Proteins that are structurally homologous to ProQ may exist and these proteins may have a known function that may hint at the function of
ProQ. Alignment of ProQ homologues from various bacteria revealed the presence of two conserved regions of ProQ, linked by a region that varied in both length and had a low level of conservation amongst the homologues. This observation lead to the hypothesis that ProQ was composed of a conserved N-terminal domain (amino acid residues 1-130 of E. coli ProQ) and a C-terminal domain (amino acid residues 180-232 of
E. coli ProQ), linked by a less conserved linker region. The N-terminal domain of ProQ was homology modeled on the crystal structure of the mRNA binding protein FinO
(described above) while the C-terminal domain could be homology modeled on the
37 crystal structure of the SH3-like domain of the myosin motor domain or the Sm protein
Hfq (described above). In order to test this hypothesis, I planned on using limited trypsin digestion of the full length protein to isolate protease resistant and protease sensitive regions of ProQ. To do this, it was first necessary to overexpress, and purify full length
ProQ.
Previous attempts to purify ProQ resulted in preparations that were contaminated with the DNA binding proteins HU and a histone-like protein, thus, alternative methods for purification of ProQ had to be created. I set out to create a system for the overexpression
and purification of a histidine tagged variant of ProQ, ProQ-His6 and showed that its in vivo function and in vitro characteristics were similar to that of untagged ProQ.
Following confirmation of the presence of two domains using limited trypsin proteolysis, I will then attempt to test the validity of the models of the N- and C-terminal domains. In order to do this, I will create expression vectors for the overexpression of histidine tagged variants of each ProQ domain individually. I will then develop purification schemes to purify each domain under conditions that will maximize its
solubility. Following purification, I will use circular dichroism (CD) analysis to explore the secondary structure of each domain. I predict that the N-terminal domain will be
composed mostly of a-helical structure, as predicted by the homology model based on
FinO, while the C-terminal domain will show a spectrum that is typical of SH3 domains.
SH3 domains have a peculiar CD spectrum as they contain a short 3io helix that results in
a maximum at 220 nm, while in the absence of this helix, the mostly P-sheet structure
give a minimum at 217 nm.
38 Expression of the various domains of ProQ both individually as well as in combination with each other will allow me to identify regions of the ProQ protein that are important to its osmoregulatory function. To do this, I will create expression systems based on the pBAD24 vector, for expression of the N- and C-terminal domains of ProQ
(N and C) as well as for the Linker domain (residues 130-180) (L), the N-terminal and linker domains (residues 1-180) (NL), the linker and C-terminal domain (residues 130-
232) (LC) as well as the N-terminal domain fused to the C-terminal domain (residues 1 -
130 and 180-232) (NC). I predict that the N-terminal domain of ProQ will contain regions that are important for its function as Agrobacterium tumefaciens encodes a functional ProP homologue, and a putative ProQ homologue, Ydh, that aligns only to the
N-terminal domain of E. coli ProQ.
The ability to model both the N- and C-terminal domains of ProQ on RNA binding proteins involved in regulation of protein expression on the translational level calls for the re-examination of the ability of ProQ to modulate the level of ProP in E. coli.
Previous analysis ofproP expression using proP::lacZ fusions and of ProP levels using
Western blotting suggested that ProQ did not affect ProP on the level of transcription or translation (Milner and Wood, 1989; Kunte, et al. 1999), and thus it was suggested by
Kunte et al. (1999) that ProQ acted on a post translational level, possibly through protein- protein interactions with ProP. In order to test this hypothesis, I will first use fluorescence microscopy to determine the cellular localization of ProQ and determine if it concentrates at the poles of E. coli with ProP. The ability of ProQ to act directly on ProP will be tested using an in vitro proteoliposome system containing ProP co-reconstituted
39 with ProQ. If ProQ acts directly on ProP it is predicted that ProQ will concentrate at the
poles of E. coli and amplify the activity of ProP in proteoliposomes.
Given that the N-terminal domain of ProQ can be modeled on the structure of the
mRNA binding protein FinO, while the C-terminal domain can be modeled on the
general RNA chaperone Hfq, it is possible that ProQ acts as an RNA binding protein,
exerting its effects on ProP levels at the level of translation. To test this hypothesis,
Western immunoblot analysis will be performed on extracts from E. coli cells grown in
medium with various osmolalities, as well as of cells grown in LB medium through
exponential and stationary phase, expanding on the previous work performed by Kunte et
al. (1999). If ProQ modulates expression of ProP, differences should be seen in the
levels of ProP expression. To test the ability of ProQ to bind RNA, ProQ will be purified
under low-salt conditions in the absence of additional nucleases and a wavelength scan of
the preparation performed to detect the presence of nucleic acids. If nucleic acids are
present, protein will be removed from the preparation and the remaining nucleic acid
treated with either DNasel or RNaseA. Absence of nucleic acid following DNasel
treatment indicates the co-purification of cellular RNA, while absence of nucleic acid
following treatment with RNaseA indicates co-purification of RNA with ProQ. An
attempt will be made to identify RNA species co-purifying with ProQ.
In the course of my research, I created aproQ' derivative of a proP' strain in order to
test the effect of proQ on various ProP mutants. Serendipitously, it was found that in this
newly created strain, ProQ was no longer required to amplify the activity of ProP (the proQ phenotype was suppressed). Comparison of the genetic mutations in this strain to
one where the proQ phenotype is clearly seen revealed possible sources of this
40 suppression: the type oiproQ mutation; deletion of an open reading frame(s) flanking proP on the chromosome; deletion oiproVWX (proU); or deletion of an open reading
frame(s) flanking proU. In order to determine the source of suppression, in-frame
kanamycin replacements were transduced from the Keio collection (Baba, et al. 2006)
into our laboratory strains in order to test the effects of proQ, proP and proU mutations
on the proQ phenotype.
Data presented within this thesis indicate that ProQ acts as an RNA binding protein
that modulates ProP levels in response to the level of proU expression. This is a
significant finding as it shows that two seemingly redundant osmoregulatory transporters
in E. coli are able to co-ordinate each others levels and that this co-ordination may occur
via the production of a regulatory RNA and an RNA binding protein ProQ.
41 Chapter 2: Materials and Methods
2.1 General laboratory procedures
2.1.1 Laboratory equipment
Optical density (OD) measurements of cell cultures were taken using the Spectronic
88 (Bausch & Lomb, Rochester, NY). The absorbance of a sample within the UV and
visible spectra were determined by performing a wavelength scan with a Cary 100 Bio
UV Vis Spectrophotometer (Varion, Palo Alto, CA). Solution osmolalities were
measured with a Vapro vapour pressure osmometer (Wescor, Logan UT), according to
the manufacturers instructions. The pH of solutions were measured with a pHM93 pH
meter (Radiometer, Copenhagen, DK) according to the manufacturers instructions.
Microcentrifugation was performed using a bench-top Sarstedt microfuge (Montreal,
QC). Liquid cultures were grown at 37°C in a rotary shaking incubator (New Brunswick
Scientific, Edison, NJ) set at 200 rpm unless otherwise specified.
2.1.2 Stock solutions and materials
All reagents were purchased from Fischer Scientific (Ottawa, ON) or Sigma-Alderich
chemical co (Oakville, ON) unless otherwise specified. Restriction endonucleases were
obtained from New England Biolabs (Pickering, ON) or Invitrogen (Burlington, ON).
Oligonucleotide primers were obtained from Cortec DNA laboratory Services (Kingston,
ON). Solutions were prepared using glass distilled water, or MilliQ-filtered water (water
which has been deionized and filtered through a 0.22um filter to give a resistivity that is more than 18.2 MQ) (Millipore, Billerica, MA). Solutions used in RNA work were prepared in diethyl pyrocarbonate (DEPC)-treated water. DEPC-treated water was
42 prepared by adding 0.1% (v/v) DEPC to distilled water, incubating overnight at 37°C and
then autoclaving.
2.1.3 Bacterial strains and plasmids
The strains of E. coli used in this study are listed in table 2.1. All plasmids used in
this study are listed in Table 2.2. Primers used for plasmid construction and DNA
sequencing are listed in Table 2.3. Glycerol stocks of all strains were prepared by
aseptically mixing 1 mL of overnight culture, grown in LB, with 0.5 mL 80% (v/v)
Glycerol. These stocks were maintained at -40°C. Culture identity and purity were
confirmed by performing strain tests on indicator media (TTC proline plates and Lactose
MacConkey plates), antibiotic supplemented LB medium (ampicillin, kanamycin,
streptomycin, chloramphenicol and tetracycline) and MOPS buffer based minimal
medium plates as described in Appendix 1.
2.1.4 Culture conditions
E. coli cultures were grown in LB broth (Miller 1972) or MOPS based minimal
medium (Neidhardt, Bloch and Smith 1974) (See appendix 1 for media recipes). Where
necessary, ampicillin, kanamycin or chloramphenicol were added to the culture medium
to a final concentration of 100 |a.g/mL, 100 ug/mL, or 50 ng/mL respectively, for plasmid
maintenance.
43 Table 2.1: Bacterial strains
Strain Genotype Source or Reference
BL21 Gold E. coli B F ompT hsdS (nrrik) dcrrC Tet gal endA hte Stratagene, LaJolla, CA BW25113 rrnB3, MacZ4787, hsdR514, A(araBAD)567, Baba, et al. 2006 A(rhaBAD)568, rph-1
DH5a F &80d lacZAM15 A(lacZYA-argF) U169 recAl endAl Bethesda Research hsd R17(rumk )sup E44X thi-1 gyrA relA Laboratories JW2652-1 BW25113 AproV758::kan Baba, et al. 2006 JW2653-1 BW25113 AproW759::kan Baba, et al. 2006 JW2654-1 BW25113 AproX770::kan Baba, et al. 2006 JW4068-1 BW25113 AphnB756::kan Baba, et al 2006 JW4069-1 BW25113 AphnA 756::kan Baba, et al. 2006 JW4070-1 BW25113 AyjdA 756::kan Baba, et al. 2006 JW4071 -1 BW25113 AyjcZ::kan Baba, et al. 2006 JW4072-2 BW25113 AproP:: 771::kan Baba, et al. 2006 JW4073-1 BW25113 AbasS756::kan Baba, et al. 2006 JW4074-1 BW25113 AbasR 756::kan Baba, et al. 2006 JW4075-1 BW25113 AeptA756::kan Baba, et al. 2006 JW5300-1 BW25113 AproQ756::kan Baba, et al. 2006 MC4100 F" araD\39 A(argF-lac)U169 rpsL\50 relA 1 deoCl rbsR Casadaban, 1976 fthD530lfruA25 X MG1655 F Blattner, et al. 1997 RH90 MC4100 rpoS: :TnlO Lange and Hengge- Aronis 1991 RM2 F trp lacZ AputPA 101 rpsL thi Wood and Zadworny 1980
44 Strain Genotype Source or Reference SGI 3009 F his pyrD Alon-100 rpsL thi Qiagen, Mississauga, pREP4 ON WG170 RM2proP219 Stalmach, Grothe and Wood 1983 WG174 BM2proQ220::Tn5 Stalmach, Grothe and Wood, 1983
WG203 WG170 pro U205 Grothe, etal. 1986 WG210 KM2proU205 Grothe, et al. 1986 WG350 WG170 A(proU600) A(p Culham, etal. 1993 WG584 RM2 trpDC700::pt itPAl Culham and Vogele, unpublished data WG914 WG210 AproQ676 Smith, et al. 2004 WG997 WG350 AproQ214 This Study WG10422 VJGllOAphnB 756. :kan This Study WG10432 WG914 AphnB756. :kan This Study WG10462 WG914 AphnA 756.\kan This Study WG10482 WG2\0AyjcZ756:: kan This Study WG10492 WG914 AyjcZ756:: kan This Study WG10512 WG2\0AyjdA756:. :kan This Study WG10522 WG9\4AyjdA 756:.:kan This Study WG10542 WG210 AbasS756:.:kan This Study WG10552 WG914 AbasS756: :kan This Study WG10572 WG210 AbasR 756::kan This Study WG10582 WG914 AbasR756: :kan This Study WG10602 WG2lOAeptA756:. 'kan This Study WG10612 WG9\4AeptA756:. :kan This Study WG10632 WG2\0AproP771: :kan This Study
45 Strain Genotype Source or Reference WG10652 WG914 AproP771::kan This Study WG10672 WG350 AproP771::kan This Study WG10692 WG997 AproP771::kcm This Study WG10702 RM2 AproP771::kan This Study WG10722 RM2 AproQ756::kan This Study WG10734 RM2 AproQ856::FRT This Study WG10743 WG170 proQ220::Tn5 This Study WG10772 RM2 AproV758::kan This Study WG10784 RM2 AproV858::FRT This Study WG10802 WG1078 AproQ756::kan This Study WG10822 RM2 AproW759::kan This Study WG10834 KM2AproW859::FRT This Study WG10852 WG1083 AproQ756::kan This Study WG10872 RM2 AproX770::kan This Study WG10884 RM2 AproX870::FRT This Study WG10902 WG1088 AproQ756::kan This Study WG1119 WG584 AproQ870::FRT This Study WG1120 WG584 rpoS359::TnlO This Study WG1121 WG1119 rpoS359::TnlO This Study WG1197 RM2 A(proV-proX)2098 This study WG11982 WG1197 AproQ756::kan This study WG1201 RM2 proV677 This study WG12022 WG1201 AproQ756::kan This study
Unless otherwise indicated all strains are derived from E. coli K-12. 2 Strains were prepared using PI transduction (2.2.13) to transfer the relevant kanamycin replacement from a Keio collection strain based on BW25113 (Baba, et ah 2006)
46 obtained from the laboratory of Dr. Brown (McMaster University, Hamilton, ON) to the indicated recipient strain. 3Strains were prepared using PI transduction (2.2.13) to transfer the proQ220::Tn5 mutation from strain WG174 to the indicated recipient strain. 4Kanamycin replacements were converted to in-frame chromosomal deletions using the FLP recombinase and techniques described in 2.2.14.
47 Table 2.2: Plasmids used in this study
Plasmid Description Primer Names2 Fragment Source or subcloned3 Reference pBAD24 Expression vector (Guzman, et al. 1995) pCP20 Encodes FLP recombinase and (Datsenko a temperature sensitive origin and Wanner of replication 2000) pDC77 pBAD24 encoding proQ (Kunte, et al. 1999) pDC79 pBAD24 encoding proP (Culham, et al. 2000) pDs-Red Encodes Red Fluorescent Clontech, Protein (RFP) Mountain View, CA. pDVl pBAD24 encoding proQ- ProQ-GFP-1 This Study without its termination codon ProQ-3 pDV2 pBAD24 encoding ProQ-RFP ProQ-RFP-1 This Study ProQ-RFP-2 pK03 Allelic exchange vector (Link, encoding the sacB gene from Phillips and B. subtilis and a temperature Church 1997) sensitive origin of replication pMSl pQE-60 encoding ProQ-His6 pSSl-1 (Smith, et al. pSSl-2 2004; This study) pMSlO Vector pQE-80-L encoding ProQ-Nterm-3 (Smith, et al. H6N ProQ-Ntemi-5 2007; This study) pMSll pBAD24 encoding N Ncol- (Smith, et al. Hindlll of 2007; This pMSlO study) pMS13 pQE-80-L encoding H6C ProQ-Cterm-1 (Smith, et al. proQ-2 2007; This study) pMS14 pBAD24 encoding NC ProQ Fusion-1 Ncol- (Smith, et al. ProQ-2 Hindlll of 2007; This pMS9 study) pMS15 pBAD24 encoding C Ncol- (Smith, et al. Hindlll of 2007; This pMS13 study) pMS16 pQE-80-L encoding H6NC Ncol- (Smith, et al. Hindlll of 2007;This pMS14 study)
48 pMS17 pBAD24 encoding ProQ-His6 EcoRl- This study Hindlll of pMSl pMS18 pBAD24 encoding NL ProQNL2 (Smith, et al. ProQ-Nterm-5 2007;This study) pMS19 pBAD24 encoding LC ProQ LC (Smith, et al. ProQ C-term 2007;This study) pMS2 pQE-60 encoding ProQ proQ-2 (Smith, et al. proQ-3 2007;This study) pMS20 pK03 allelic exchange vector ProQNj/ProQNo This study to introduce AproQ214 ProQCj/ProQCo pMS21 pBAD24 encoding NC EcoRI- (Smith, et al. Hindlll of 2007;This pMS14 study) pMS22 pBAD24 encoding L ProQ NL (Smith, et al. ProQ LC 2007;This study) pMS23 pBAD24 encoding RFP ProQ-RFP-1 This study ProQ-RFP-2 pMS24 pGEM-7Z encoding proP ProP - 2048 This study ProP - JK-2 pMS25 pQE-80L encoding the Keio Km For-Hindlll This study collection Kanamycin cassette Km Rev - EcoRI pMS26 pQE-80L encoding ProV and proV upstream 2 This study ProW ProW5 pMS27 pQE80-L encodingproU219 ProVE190Q-l This study and ProW ProV E190Q-2 pMS28 pQE80L encoding 500 bp proV up out 1 This study upstream of proVPl to proV proV up out 2 PI pMS29 pQE80L encoding the 500 bp proX out down 1 This study following the termination proX out down 2 codon of proX pMS30 pGEM7Z encodingproU::Km Hin&llllBamm This Study ofpMS27 HindllllEcom ofpMS28 EcoRI/BamHl ofpMS29 pMS31 pK03 allelic exchange vector BamHl/Sall This Study to introduce mutationproV677 of pMS27
49 pMS9 Vector pBAD24 encoding ProQ-Nterm-3 (Smith, et al. H6QN no stop ProQ-Nterm-4 2007;This study) pQE-60 Expression vector encoding a Qiagen, C-terminal 6 x His tag Mississauga, ON. pQE-80-L Expression vector encoding an Qiagen, N-terminal 6 x His tag and Mississauga, Laclq in cis ON. pRAC2 pK03 allelic exchange vector proQNi/proQNo (Smith, et al. to introduce mutation proQCi/proQCo 2004) AproQ676 pREP-4 Plasmid encoding Laclq to Qiagen, repress gene expression from Mississauga, pQE-60 in trans ON. pSSl pQE60 encoding proQ213 proQ-3 (Smith, et al. proQ-1 2004) ProQ proteins and domains encoded by plasmids are as follows: ProQ-His6 (ProQ-RSH6), N(ProQMl-E130) C(M+ProQV180-F232) L(M+ProQE130-V180) NL(ProQMl-V180) LC(M+ProQE130-F232) NC (ProQMl-130 + AW + ProQ V180-F232) H6N (MRGSH6N) H6C (MRGSH6C) H6NC (MRGSH6NC)
2Names of primers used to amplify the cloned sequence. Primer sequences are listed in Table 2.3 Restriction endonuclease fragment from the indicated plasmid was subcloned into the indicated vector to create the plasmid of interest
50 Table 2.3: Primers used within this study Primer Primer sequence (5'-3') Use name Primers used for plasmid construction Kan GCG CAA GCT TGT GTA GGC TGG AGC Construction of the pro U operon Cassette TGC TTC GAA GTT C allelic exchange vector Forward - Hindlll Kan GCG CGA ATT CAT TCC GGG GAT CCG Construction of the proU operon Cassette TCG ACC TGC AGT T allelic exchange vector Reverse - EcoRI proQ-1 GGT AGA TCT GAA CAC CAG GTG TTC Cloning ProQ into pQE-60 TG Cloning ProQ no stop into pBAD24 proQ-2 GGA TAA GCT TTC AGA ACA CCA GGT Cloning ProQ into pQE-60 GTT Cloning QC into pQE80L Cloning QNC into pBAD24 proQ-3 GGC TCC ATG GAA AAT CAA CCT AAG Cloning ProQ into pQE-60 TTG ProQ-Ci TGT TTA AGT TTA GTG GAT GGG TCT Construction of proQ allelic TTG ATT GTG CGC GCA G exchange vector ProQ-Co CGC GGA TCC CGA ATT TGA CTT TAC Construction of proQ allelic TGTCC exchange vector ProQ- GCG GAT CCA TGG TTT CTG ACA TTT Cloning QC into pQE80L Cterm-1 CAG CTC Cloning QLC into pBAD24 ProQ- GCG CAA GCT TGG GTT TCT GAC ATT Cloning QNC into pBAD24 Fusion-1 TCAG ProQ- GCG CCT GCA GGA ACA CCA GGT GTT Cloning ProQ without a stop GFP-1 C codon into pBAD24 ProQ-LC GCG CCA TGG AGA AAG AAG ACG Cloning QLC into pBAD24 CAC CGC GCC GC Cloning QL into pBAD24 ProQ-Ni CCC ATC CAC TAA ACT TAA ACA CTT Construction of proQ allelic AGG TTG ATT TTC CAT G exchange vector pMS20 ProQ-NL- GCG AAG CTT TCA AAT GTC AGA AAC Cloning QNL into pBAD24 2 CGG GGT GTG Cloning QL into pBAD24 ProQ-No GCG GAT CCG CTG ATG GCG GGA GAA Construction of proQ allelic AC exchange vector pMS20 ProQ- GCG CGG ATC CAT GGA AAA TCA ACC Cloning QN into pQE80-L Nterm-3 TAA GTT GAA TA ProQ- GCG CAA GCT TTC TCA CCA GCA GTT Cloning QN without a stop Nterm-4 GCG GCA codon into pQE80-L ProQ- GCG CAA GCT TTC ACT CAC CAG CAG Cloning QN into pQE80-L Nterm-5 TTG CGG CAG C Cloning QNL into pBAD24
51 ProQ- CGC CAT GCA TAT GGC CTC CTC C Cloning RFP into pDVl RFP-1 ProQ- GCA AGC TTT TAG GCG CCG GTG G Cloning RFP into pDVl RFP-2 ProV ATA TTA TTA ATG GAC CAA GCC TTC Introduction of mutation E190Q-1 TCT GCG CTC GAT CCA TTA ATT CGC pro FE190Q into pMS26 ProV GCG AAT TAA TGG ATC GAG CGC AGA Introduction of mutation E190Q -2 GAA GGC TTG GTC CAT TAA TAA TAT pro FE190Q into pMS26 ProV out CGC GGG ATC CAT TAT CCG CGA TGA Cloning of the pro V and pro W up 1- AGC AGT CCA CGG ORFS into pQE80L BamHI ProV out CGC GAA GCT TAG TAT CAG TGT AGA Cloning of the pro V and pro W up 2- TCA CCA CAA ATT ORFs into pQE80L Hindlll ProV GCG GGG ATC CGA ATG CCG ATG AAA Construction of A (pro V- upstream ATC proX)2098 2 -BamHI ProW5 CGG AAG CTT TTA CTT AAT GAA TGG Cloning of the pro V and pro W GCG GG ORFS into pQE80L ProX out CAG GGA AAT CTA ATT TTT ATT CGG Construction of the proU operon down 1 - GCG GAT AAG GCG allelic exchange vector EcoRI ProX out GCG CGG ATC CTA CCG AGG ATC ATC Construction of A (proV- down 2- ATC GCC AGC GAC proX)2098 BamHI pSSl-1 GCA AAA CGC AAT TGC GAT CGG CTT Correcting R57Stop mutation in TAC GTC TCT AC pSSl pSSl-2 GTA GAG ACG TAA AGC CGA TCG CAA Correcting R57Stop mutation in TTG CGT TTT GC pSSl Primers used to confirm mutations by PCR analysis basR CGG TAC TTT CAT GCC ATA G To confirm a AbasR756::Kan replacement basS CAC TTA ATC TCT GAC GCG To confirm a AbasS758::Kan replacement eptA GCT GTA ATG ACC TGC CTC To confirm an AeptA 756::Kan replacement Km GAT CTC TCA TCT CAC To confirm Km replacements forward Km GAC AGC CGG AAC ACG GCG To confirm Km replacements reverse phnA GCG TCG CAT CAG GCA TCT G To confirm a AphnA 756::Kan replacement phnB CGG GAA TTC GAT GAC AGT GC To confirm a AphnB756::Kan replacement ProP- GTT GGC TGG GAA AGT GGA C To confirm a AproP771::Kan DC3 replacement
52 ProQ- CCG CAA TCA AGT GCA GCG T To confirm a AproQ756::Km AB6170 replacement yjcZ CAA ACC AGT TGA CGC CTG To confirm a AyjcZ756::Km replacement yjdA CAC GCG AAA GAA GTT CCT G To confirm a AyjdA756::Km replacement Primers used to identify RNA co-purifying with ProQ Hindlll GCT GAA TTA AGC TTG GGG GGG GGA To clone unknown RNA poly(G)9- sequences into vector pGEM7Z A
Hindlll- GCT GAA TTA AGC TTG GGG GGG GGC To clone unknown RNA poly(G)9- sequences into vector pGEM7Z C Hindlll- GCT GAA TTA AGC TTG GGG GGG GGG To clone unknown RNA poly(G)9- sequences into vector pGEM7Z G Hindlll- GCT GAA TTA AGC TTG GGG GGG GGT To clone unknown RNA poly(G)9- sequences into vector pGEM7Z T RNA GGG AGA GGA ATT CAG TGT GTG CGG To clone unknown RNA Oligo sequences into vector pGEM7Z Forward SP6 ATT TAG GTG ACA CTA TAG G To create an RNA linker using promoter SP6 polymerase SP6-RNA CCG CAC ACA CTG AAT TCC TCT CCC To create an RNA linker using OligoTem TAT AGT GTC ACC TAA AT SP6 polymerase plate
53 2.2 Molecular biology techniques
2.2.1 Chromosomal DNA preparations
Cell lysis by boiling
Cells from a 1 mL aliquot of overnight LB culture were harvested by centrifugation for 3 min at 13, 000 rpm. The resulting pellets were washed 2 times in a sterile 0.85%
Saline solution at room temperature. Following the second wash, the pellet was resuspended in 0.5 mL sterile Milli-Q water and boiled for 10 min. The resulting lysate was placed on ice for 3 min and then centrifuged for 5 min as above. The resulting supernatant was maintained at 4°C and used as a source for chromosomal DNA in PCR reactions.
Phenol chloroform DNA extraction (Sambrook and Russell, 2001)
Fifty mL of LB was inoculated with the strain of interest and cells cultured overnight at 37°C. Cells were harvested by centrifugation (12,000 x g, 4°C, 10 min) and washed 2 times in J Buffer (100 mM Tris-HCl (pH 8.0) 100 mM Na-EDTA, 150 raM
NaCl). The final pellet was resuspended in 0.8 mL J Buffer. An aliquot of 0.9 mL of lysozyme (10 mg/mL in 250 mM Tris-HCl (pH 8.0)) was added and the resulting mixture incubated at 37°C for 10 min. 0.1 mL of 10 mg/mL RNase, in water, was added and the mixture incubated for another 10 min at 37°C. The mixture was heated to 70° for 3 min and then 67 uL of 30% sarcosyl (w/v) was added. This mixture was incubated at 37-
45°C for an hour. Eleven microliters of 10 mg/mL proteinase K, in water, was added and the resulting slurry incubated at 37 to 45°C for 2 to 4 h. A second 11 uL aliquot of proteinase K was added and the mixture dialyzed at 37°C against 2 L of 10 mM Tris-HCl
54 (pH 8.0), 10 raM Na-EDTA, 0.15 M NaCl in a 50,000 molecular weight cut off (MWCO) membrane overnight. The sample was transferred to a 5 mL cryovial and the DNA extracted by gentle inversion for 10 min against an equal volume of phenol equilibrated against J Buffer. The mixture was spun in the centrifuge (4000 x g, 10 min, room temperature) and the top layer moved into a fresh cryovial. This layer was then extracted against phenol:chloroform:isoamylalcohol (25:24:1) and then against chloroform:isoamylalcohol (24:1). The final aqueous phase was dialyzed against 2L of water at 4°C overnight in a 50,000 MWCO dialysis membrane.
Isolation using the Qiagen DNeasy kit
DNA was prepared from 3 mL of overnight cell culture using the Qiagen DNeasy
Blood and Tissue Kit as directed by the manufacturer (Qiagen, Mississauga, ON). This procedure typically yields 10 |j,g of chromosomal DNA in a final volume of 400 uL of elution buffer supplied with the kit, giving a final DNA concentration of 25 ng/uL.
2.2.2 Plasmid DNA isolation
The Qiagen QIAprep spin Mini-prep kit (Qiagen Mississauga, ON) was used to prepare plasmid DNA from bacterial cells. For high copy number plasmids, cells pelleted from 3 mL of overnight LB culture yielded 20 ug of plasmid DNA. One hundred uL of sterile water was used to elute plasmid DNA from the spin column in the last step giving a final DNA concentration of approximately 0.2 ug/uL.
For large scale plasmid preparations, or for purification of low copy-number plasmids, the Qiagen Midi prep kit was used (Qiagen, Mississauga, ON). Cells were grown overnight in 50 mL of LB at either 37°C (for most plasmids) or 30°C (for plasmids
55 containing a temperature sensitive origin of replication). Cells were harvested by centrifugation (12,000 x g, 10 min, 4°C). Plasmid DNA was then extracted from the resulting pellet using the MIDI prep kit as per the manufacturer's instructions.
The identity of isolated plasmids was confirmed by comparing the restriction enzyme digest fragments by agarose gel electrophoresis with the predicted fragment sizes for the plasmid.
2.2.3 Agarose gel electrophoresis
Agarose gel electrophoresis was carried out to determine the size and or quantity of
DNA fragments resulting from PCR amplification reactions, or restriction endonuclease digestions. Agarose gels were prepared by adding the appropriate weight percentage of
Agarose in lx TBE buffer (89 mM Tris, 89 mM boric acid, 2.5 mM Na-EDTA) and boiling the resulting mixture to melt the agarose. Melted agarose was poured into casting trays and allowed to set. Gels were run in a Bio-Rad MiniGel apparatus in lxTBE buffer at 80V until bromophenol blue migrated to 3A to the full length of the gel. A 1Kb ladder
(Bio-Rad, Mississauga, ONT) was run as a molecular size standard. Gels were stained in
0.5 \ig ethidium bromide in 0.5 x TBE and then visualized under UV translumination.
2.2.4 Polymerase chain reaction (Brown and Wood, 1992)
Polymerase chain reactions were carried out in a GeneAmp 2400 thermocycler
(Perkin Elmer, Wellesley, MA). PCR mixtures contained 0.2 mM dNTP mixtures, 50 ng
DNA template, 40 pmol of each primer and 1 unit of polymerase. When PCR products were used for analysis, PCR was performed using Taq polymerase (Invitrogen,
Burlington, ON) in either the supplied buffer, or in 50 mM Tris-HCl (pH 8.4), 50 mM
56 KC1, 1.5 mM MgCl2, 0.5% (v/v) tween and 0.001% gelatin. When PCR products were used for subsequent cloning, Pfu Turbo (Stratagene, LaJolla, CA), or Pwo polymerase
(Roche, Mississauga, ON) were used instead of Taq as they possess proofreading abilities. In both cases, buffer supplied by the manufacturer was used according to the manufacturers instructions.
Annealing temperatures for each primer set were determined by subtracting 5°C from the lowest estimated primer melting temperature. Approximate melting temperatures were calculated using the formula Tm= (4(G+C) +2(A+T)). An initial melt was performed at 94°C for 5 min, followed by 26 cycles of the following profile: melting at
94°C for 45 s, annealing at Tm-5°C for 30 s, extending at 72°C for 1 min per kb of DNA amplified. The 26 cycles were followed by final extension step at 72°C for 5 min, followed by a holding temperature of 4°C.
2.2.5 Site directed mutagenesis
Site directed mautagenesis was performed using the Quickchange kit from Stratagene
(Stratagene, LaJolla, CA). The target plasmid was amplified using complementary, mutagenesis primers (Listed in Table 2.3). Each complementary primer was designed to introduce the required mutation as well as to add (or remove) a particular restriction site that would be used to screen for mutants. A 51 uL PCR mixture was prepared containing
5 uL 10 x PWO buffer (+MgS04), 1 uL 25 mM dNTP (Roche, Mississauga, ON), 125 ng of each primer, 20-200 ng plasmid DNA template 1 uL 5LV uL Pwo DNA polymerase in water. Thermal cycle conditions were as follows: denaturation for 20 s at 95°C, 18 cycles of amplification each consisting of 30 s of denaturing at 95°C, annealing for 1 min
57 at 55°C, and extension at 68°C for 14 min. A final extension step was set for 7 min at
68°C. Following the PCR reaction, 1 uL ofDpnl (10 U/ uL) was added directly to the
PCR reaction mixture and the mixture incubated at 37°C for 1 h to digest the template, non mutagenized plasmid DNA.
2.2.6 DNA purification following PCR amplification
If PCR products were to be subjected to restriction digestion, the products were cleaned, to remove PCR buffer and left over dNTPs and oligonucleotides, using a Qiagen
QIAquick PCR purification kit (Qiagen, Mississauga, ON) following the manufacturer's instructions. Alternatively, the PCR product was precipitated using 2% (w/v) potassium acetate in 95% ethanol and resuspended in water.
Prior to ligation reactions, restricted PCR products were purified by agarose gel purification. Restricted PCR products were electrophoresed as above (2.2.3). The band corresponding to the expected size of the restricted PCR product was excised following detection by EtBr staining and UV irridation. The PCR product was extracted from the gel fragment as below (2.2.7).
2.2.7 Gel purification of DNA fragments
Prior to cloning or subcloning a DNA fragment, restriction endonuclease digestions were subjected to agarose gel purification to isolate a fragment of appropriate size and exclude other DNA fragments from the subsequent ligation reaction. Agarose gels were cast and run and visualized as outlined (2.2.3). A scalpel was used to excise the band of interest from the gel. DNA was extracted from this gel slice using the Qiagen QIAquick gel extraction kit (Qiagen, Mississauga, ON) as per the manufacturer's instructions.
58 2.2.8 Ligation of DNA fragments
Ligation of DNA fragments was carried out using T4 DNA ligase (Roche,
Mississauga, ON) following this standard procedure. Vector and insert DNA were mixed in an approximate 1:1 ratio, based on band intensities following agarose gel electrophoresis, in 1 x T4 ligase buffer (66 mM Tris-HCl, 5 mM MgCl2, 1 mM dithioerythritol, 1 mM ATP (pH 7.5)) with 20 units of T4 ligase in a total volume of 20 uL. The ligation mixture was incubated at 15°C for 16 h, and used to transform E. coli
DH5a unless otherwise stated.
2.2.9 Chemical transformation
Chemically competent cells were prepared as described by Hanahan et ah, (1983).
Following transformation the cells were incubated at 37°C (or 30°C for plasmids containing temperature sensitive origins) for 1 to 2 h to allow for the expression of antibiotic resistance genes. Transformants were selected on LB medium containing the appropriate antibiotic at 30°C or 37°C overnight or up to an additional 24 h if necessary.
2.2.10 Electroporation
Electrocompetent cells were prepared as outlined by Dower, et al.(1988). A 5 mL LB culture was inoculated and incubated overnight. This 5 mL culture was used to inoculate a 500 mL LB culture the following day. The culture was grown to an OD600 of 0.6, transferred to centrifuge bottles and chilled on ice for 20 min. Cells were harvested by centrifugation (4000 x g, 15 min, 4°C). The resulting pellet was resuspended in 500 mL of ice cold 10% (v/v) glycerol and the cells harvested by centrifugation as above.
Subsequently the cells were washed in 250 mL, then 30 mL of ice cold 10% glycerol
59 with centrifugation steps as above. The final pellet was resuspended in 1.5 mL ice-cold
10% glycerol. Forty uL aliquots were stored at -80°C.
Transformation of DNA into recipient cells by electroporation was accomplished by thawing a 40 uL aliquot of electrocompetent cells on ice, adding 5 rig DNA to the cells and mixing by vortexing, transferring the mixture to an electroporation cuvette and pulsing the mixture in the electroporator in a 0.1 cm pulse (1.8 kV, 1 pulse , with a variable time constant) (Bio-Rad, Mississauga, ON). Immediately following electroporation, 1 mL of SOC medium was added and the cells allowed to recover at
37°C or 30°C to allow for phenotypic expression of antibiotic resistance. 10 \xL, and 100
|xL aliquots were plated on selective medium. The remainder of the cells were harvested by centrifugation and resuspended in a small volume which was plated on selective medium.
2.2.11 DNA sequencing
DNA sequencing was performed by the Advanced Analysis Center Genomics Facility
(University of Guelph, Guelph, ON) using dye termination reactions (College of
Biological Sciences Advanced Analysis Center, 2009). The sequences of the inserts of all plasmids constructed during this research were determined to verify that the constructs were correct and no mutations had been introduced inadvertently.
2.2.12 Plasmid construction
Plasmids constructed within this study are summarized in Table 2.2. Primers used during plasmid construction are listed in Table 2.3. Plasmid maps are provided in
60 Appendix 2. Plasmids were transformed chemically into DH5a and transformants selected on LB supplemented with 100 \xg/mL Amp.
To construct plasmid pMSl (encoding ProQ-Hise) the proQ ORF was amplified with primers proQl and proQ3 using plasmid pDC77 as template. Primer proQ3 created an
Ncol site at the 5' end of proQ, whereas primer proQl introduced a BgRl site 3' ofproQ and changed the stop codon (TGA) to arginine (AGA), in frame with the six histidine codons and the termination codon provided by the vector. The PCR product and pQE-60 were cleaved with Ncol and BgRl; the desired DNA fragments purified, mixed, ligated, and transformed into E. coli SGI3009 pREP4. Transformants were selected for on LB supplemented with 100 fa.g/mL Amp and 50 [ig/mL Km.
Plasmid pSSl, encoding ProQ R57Stop, had been previously constructed by Sarah
Sanowar in the Wood laboratory. This was constructed by insertion of the amplified proQ ORF into plasmid pQE60 at the BgRl and Hindlll site (ensuring that the vector encoded His tag would not be added to the proQ ORF). Following transformation into
DH5a, sequence analysis revealed that plasmid pSSl encoded a ProQ variant that was truncated by a stop codon at R57, termed allele proQ213. To construct plasmid pMS2
(encoding ProQ), the mutation prog 213 was corrected using the QuickChange site- directed mutagenesis kit (Stratagene, LaJolla, CA) with primers pSSl-1 and pSSl-2, and the resulting plasmid, pMS2, was transformed in E. coli SGI3009 pREP4.
Transformants were selected for on LB supplemented with 100 ng/mL Amp and 50 fig/mL Km.
61 Plasmids pMS8-pMS13, pMS15-19 and pMS21-pMS22 encoding ProQ fragments, outlined in Table 2.4, were constructed by PCR amplifying the required proQ sequence
(plus desired flanking sequences), with plasmid pDC77 as the template, or by excising the required proQ sequence from an existing plasmid, then inserting the amplicon or fragment in an appropriate vector, and recovering the resulting plasmid by transformation into E. coli DH5ot followed by subsequent transformation into the relevant E. coli genetic background.
To construct plasmid pMS14, the fragment encoding residues M1-E130 of ProQ was
PCR amplified, adding a 5' BamBl site and a 3' Hindlll site but no termination codon.
The amplified sequence was ligated into vector pQE80L to create an intermediate plasmid, pMS9. The fragment encoding residues V180-F232 of ProQ was amplified using primers ProQ Fusion and proQ-2. Digestion of the resulting PCR product and plasmid pMS9 with EcoKL allowed for the insertion of the C-terminal coding sequence in-frame with the N-terminal coding sequence. The orientation of the insert was confirmed through restriction analysis and sequencing.
The proQ deletion vector pMS20 was created via a two-step PCR (Link, et al. 1997) using Pwo DNA polymerase (Roche, Mississauga, ON). The promoter proximal (596 bp) and distal (604 bp) DNA fragments flanking the chromosomal proQ ORF in strain RM2 were amplified using primer proQNo with proQNi and primer proQCo with proQCi, respectively. The resulting PCR products were annealed and amplified using primers proQNo and proQCo to generate a proQNo-proQCo fragment encoding AproQ214. This fragment was restricted with BamHl and ligated into 2?awHI-cleaved plasmid pK03
62 (Link, et al. 1997). Chromosomal proQ was deleted from E. coli strain WG350 by allelic exchange using the resulting plasmid, pMS20. Putative proQ deletion mutants were screened by PCR analysis using primer proQNo with primer proQCo. Vector pMS20 encodes proQ214 resulting in replacement of the proQ ORF with one encoding the peptide MENKCLSLVDGSLIVRAEHLVF.
To construct plasmid pMS24 (encoding proP), the proP ORF was excised from plasmid pDC79 (Culham, et al. 2000) at the upstream JECORI and downstream Hindlll sites and cloned into the same sites on the pGEM7Z vector.
Plasmid pMS25 (encoding the proV and pro WORFs in pQE80L) was constructed by
PCR amplification ofproVWfcom chromosomal DNA isolated from BW25113 proQ::Km (Baba, et al. 2006) as a template and primers ProVup 2 and ProW5 resulting in the introduction of BamUl and Hindlll restriction endonuclease sites at the 5' and 3' ends respectively. The amplicon and vector pQE80L were both restricted with BamHl and Hindlll and the resulting fragments purified and ligated.
Plasmid pMS26 was created by site directed mutagenesis with plasmid pMS25 as a template and primers ProV-E190Q-l and ProV-E190Q-2 to introduce mutation
ProVE190Q and to remove a Bgll and Haell restriction endonuclease sites creating allele proV677.
Plasmids pMS27 - pMS29 each encode fragments used to create pMS30. Each was created by amplification of a region of DNA using strain BW25113 proQr.Km (Baba, et al. 2006) as template with a particular set of primers. pMS27 (encoding a kanamycin
63 resistance cassette in pQE80L) was created by PCR amplification of the kanamycin
resistance cassette using primers Km Cassette Forward and Km Cassette Reverse to
introduce Hindlll and EcoRl restriction endonuclease sites to the 5' and 3' ends. pMS28
was created by PCR amplification of the 500 bp upstream of the PI promoter of pro V,
using proV up out 2 to introduce BamHl and Hindlll restriction endonuclease sites to the
5' and 3' ends respectively. pMS30 was created by PCR amplification of the 512
nucleotides downstream of the proX termination codon using primers proX down out 1
and proX out down 2 to introduce EcoKl and BamHl sites to the 5' and 3' ends
respectively. Each amplicon and vector pQE80L were cut with their respective
restriction endonucleases, purified and ligated. The ligation mixtures were transformed
into DH5a and the resulting isolates confirmed by restriction endonuclease digestion.
Plasmid pMS30, encoding a Kanamycin replacement of the entire pro VWX operon
(From promoter PI of proV to the stop codon ofproX), was created by sequential cloning
of DNA fragments isolated from plasmids pMS27, pMS28 and pMS29. To begin, the
506 bp region upstream ofproVs PI promoter was excised from plasmid pMS28 using
the BamHl and Hindlll sites. This fragment was cloned into the BamHl and Hindlll sites
of vector pGEM7Z. The EcoKl/Sphl fragment from plasmid pMS29 encoding 512 bp
downstream of the/?roX stop codon was excised and cloned into the EcoKl/Sphl sites of
the pGEM7Z plasmid containing theproVPl upstream sequence. Into this isolate, the
Hindlll/EcoKl fragment, encoding the Km cassette, from plasmid pMS27 was introduced
to give the final pMS30 plasmid.
64 To create plasmid pMS31, theproV611 -proWcoding sequence from pMS26 was cloned into the BamHl and Sail sites in vector pK03. The resulting plasmid, pMS33, was transformed into DH5a and selected for on LB supplemented with 30 |4.g/mL chloramphenicol at 30°C.
Plasmids pDVl and pDV2 were constructed in co-operation with Divya Viswanathan in the Wood Laboratory. To construct pDVl, theproQ fragment was PCR amplified using plasmid pDC77 as a template and primers ProQ3 and ProQGFPl to introduce an
Ncol site at the promoter proximal end of proQ and to remove the stop codon and add a
Pstl site at the promoter distal end of the proQ ORF. The amplicon was restricted, purified and ligated into vector pBAD24 (Guzman, et al. 1995) which had been previously restricted with Ncol and Pstl.
To construct pDV2, the fragment encoding the Red Fluorescence Protein (RFP) was
PCR amplified using vector pDsRed (Clontech, Mountain View, CA) as template and primers ProQ-RFP-1 and ProQ-RFP-2 to add an Nsil site to the 5' end, and a Hindlll site to the 3' end of the amplicon, respectively. The RFP amplicon was ligated into vector pDVl previously digested with Hindlll and Pstl restriction endonucleases.
Plasmid pMS23, encoding RFP in vector pBAD24, was created as a control for fluorescence microscopy experiments. Primers ProQ-RFP-1 and ProQ-RFP-2 were used to PCR amplify the RFP coding sequence from the pDs-Red (Clontech, Mountain View,
CA) template and add Hindlll and Nsil sites to the 3' and 5' ends respectively. The amplicon and vector pBAD24 were restricted, cleaned and then ligated together. The
65 ligation mixture was transformed into DH5a. The identity of the plasmid was confirmed by restriction endonuclease digestion and sequencing.
2.2.13 PI mediated transductions
PI mediated transductions were performed as previously described (Miller 1972).
Phage PI (cml clr_100) was used as the transducing phage in all experiments. It carries chloramphenicol resistance and is lytic at 42°C.
Transductants were selected on LB medium supplemented with 50 |ug/mL kanamycin. Plates were incubated at 37°C overnight. In some instances, plates were incubated for 24 to 48 h before transductant colonies appeared. Transductants were tested for kanamycin resistance, and screened for loss of chloramphenicol resistance.
Transductants were purified by streaking twice on LB medium containing kanamycin, and verified by PCR analysis (2.2.4).
2.2.14 Inactivation of chromosomal genes
Creating ORF deletions using the allelic exchange vector pKQ3
Plasmids pRAC2, pMS20, and pMS32 (Table 2.2) are based on vector pK03 and were constructed for allelic exchange experiments.
Allelic exchange was performed as previously described (Link, et al. 1997). Plasmid pRAC2, pMS20 or pMS31was transformed into a target strain (E. coli WG210, WG350 and RM2 respectively) which was then grown at the permissive temperature of 30°C on
LB medium supplemented with 30 (xg/mL chloramphenicol. Co-integrants, resulting from single recombination events were isolated when the resulting transformants were
66 grown at the non permissive temperature of 42°C in the presence of chloramphenicol.
Isolates in which a s recombination had affected allelic exchange were selected for by
growing the co-integrant cells on LB medium supplemented with 5% Sucrose.
Following allelic exchange, in-frame chromosomal deletions ofproQ introduced
through the use of pRAC2 and pMS20 were confirmed by PCR with primers AB5338
and AB5628 (Table 2.3). Introduction of the proV677 allele into strain RM2 was
confirmed by PCR with primers proVl and proV3 (Table 2.3) followed by restriction
digestion with Bgll in order to screen for loss of the restriction site. The mutations were
confirmed by DNA sequencing.
Construction of gene a gene deletion using the X-RED recombinase system
The ^-RED recombinase system was used as previously described by (Datsenko and
Wanner 2000). The insert present in plasmid pMS31 was amplified by PCR using primers proV up out 1 and proX out down 2, resulting in a fragment encoding a
kanamycin resistance cassette flanked by 500 bp upstream and downstream of the prolJ to give allele A(proV-proX)2098::kan. The resulting PCR product was restricted with
Bgll and gel purified. Purified product was electroporated into RM2 pKD46 (Datsenko
and Wanner 2000) and transformants selected at 42°C on LB supplemented with 50
Hg/mL kanamycin. The kanamycin replacement was then converted into an in-frame deletion using the methods outlined below.
Converting kanamycin replacements into in-frame chromosomal deletions
Keio collection strains, BW25113 containing kanamycin replacements of genes encoding yjdN, yjdM, yjdA, yjcZ, proP, basS, basR, eptA, proQ, proV, proW or proX
67 (Table 2.1)(Baba, et al. 2006) were obtained from Dr. E. Brown (McMaster University,
Hamilton, ON). Kanamycin replacements of each of these open reading frames (ORFs) were introduced into target strains using PI transduction (2.2.13). Conversion of the kanamycin replacements of each ORF, or of the entire pro f/operon (to create A(proV-
/?ra¥)2098::FRT) was made possible by the presence of FLP recombinase sites flanking the kanamycin cassette, and was performed as previously described (Datsenko and
Wanner 2000). Plasmid pCP20 (containing chloramphenicol resistance, a temperature sensitive origin of replication and the yeast FLP recombinase under the control of a temperature inducible promoter) was transformed into the strain and chloramphenicol resistant transformants were selected for at 30°C. Transformants were grown overnight
in LB medium, containing chloramphenicol, at 30°C and then streaked onto unsupplemented LB agar and incubated at 42°C for 24 to 48 h. The resulting colonies were picked and restreaked on LB agar and grown at 42°C for an additional 24 h.
Isolated colonies were tested for loss of both kanamycin and chloramphenicol resistance.
Isolates which had lost the antibiotic resistance markers were further screened for the loss of the ORF by PCR (2.2.4).
2.2.15 Isolation and identification of RNA co-purifying with ProQ-His6
Strain WG805 (SG13009 pREP4 pMSl) was inoculated into 100 mL of LB
supplemented with Amp and Km for plasmid maintenance, and grown overnight at 37°C.
Ten mL of the overnight culture was inoculated into six, 1L LB cultures supplemented with ampicillin and kanamycin. The cultures were grown for 4 h at 37°C, and then the cells were harvested by centrifugation, and washed in 0.85% saline. The pellet was
stored overnight at -40°C. ProQ-His6 was purified (2.3.9.2 and 2.3.9.5), except that the
68 buffers used at each step contained only 300 mM NaCl, instead of 1M NaCl, and
RNaseA and DNasel were omitted from the lysis buffer. Following purification of the protein, a sample was analyzed by UV absorption spectroscopy to determine if nucleic acid was present. Nucleic acids present were further purified from the protein fraction using a phenol chloroform extraction. Specifically, the purified protein-nucleic acid fraction was mixed in an equal volume of phenol at room temperature for 20 min. The aqueous phase was separated from the organic phase by centrifugation (4000 x g, 10 min, at room temperature). The aqueous phase was removed and mixed with an equal volume of phenol: chloroform: isoamylalcohol (25:24:1 (v:v:v)) and mixed at room temperature for 10 min. The aqueous phase was separated by centrifugation as outlined above. The resulting aqueous phase was mixed with an equal volume of chloroform:isoamylalcohol
(24:1 (v:v)), and mixed for 10 min at room temperature and then this mixture centrifuged as above to separate the aqueous and organic phases. The final aqueous phase contained the nucleic acids of interest. Nucleic acids were concentrated by precipitation with 3 x the volume of 95% Ethanol containing 2% potassium acetate at -40°C, overnight. The precipitate was harvested by centrifugation (12,000 x g, 4°C, 20 min), washed in 95%
Ethanol and then the pellet dried. The final dried pellet was resuspended in 1/10 the volume of the original isolated fraction.
The final purified nucleic acid was quantified by determining the absorbance at 260 nm and the quality of the preparation was determined using formamide agarose gel electrophoresis (2.2.3). The preparation was treated with either no nuclease, RNase Free
DNasel (Ambion, Foster City, CA), or DNase free RNaseH (Sigma-Aldrich, Oakville,
ON) by adding 20 ug of enzyme, or an equal volume of water to the preparation and
69 incubating for 30 min at 37°C. The resulting preparations were analyzed by denaturing urea-acrylamide gel electrophoresis (2.2.18).
2.2.16 Cloning cDNA complementary to ProQ-associated RNA fragments of unknown sequence
RNA fragments co-purifying with ProQ-His6 (2.2.15) were converted to cDNA so that they could be cloned into vector pGEM7Z and sequenced. Two different cloning schemes were used to clone the cDNA into the vector.
In order to create a 3' priming site for cDNA synthesis, a poly(A) tail was added to the isolated RNA using 10 Units of poly(A) polymerase (PAP)/25 uL reaction (Ambion,
Foster City, CA), 2.5 mM MgCh, and ImM ATP in 1 x polyA polymerase buffer to add a polyA tail to 0.4 ng/uL RNA. Reactions were carried out for 2 h at 37°C. Three volumes of cold 2% (w/v) potassium acetate in 95% ethanol were added to the mixture and the RNA precipitated at -40°C overnight. Precipitated RNA was harvested by centrifugation at 4°C, washed in cold 95% ethanol and the resulting pellet air dried. The poly(A) tailed RNA was then resuspended in DEPC treated water.
First and second strand cDNA synthesis was carried out using the RETROscript kit
(Ambion, Foster City, CA) and an oligo(dT) primer. First strand cDNA synthesis was carried out by Monkey Maloney Leukaemia Virus (MMLV) reverse transcriptase (5
U/uL) in 1 x Reverse transcriptase buffer (50 mM Tris-HCl (pH 8.3) 7.5 mM KC1, 3 mM
MgCl2, 5 mM DTT) with 2 mM dNTP mix (0.5 mM of each of dATP, dUTP, dCTP and dGTP), 0.5 U/uL RNase inhibitor and 5 uM of the oligo(dT) primer in a 20 |xL reaction mixture, with 1-2 (xg of template RNA, at 44°C for 1 h. The reverse transcriptase was then inactivated by heating the mixture to 92°C for 10 min. 70 A polyA tail was then added to the 3' end of the single stranded cDNA using
Terminal deoxynucleotidyl transferase (TdT) (New England Biolabs, Pickering, ON) and dATP. Fifteen micro litres of RT-PCR product was treated with 15 units of TdT in 1 x
TdT buffer (20 mM Tris-acetate, 50 mM Potassium acetate, 10 mM Magnesium acetate
0.25 mM CoCl2 (pH 7.9)) with 16 mM dATP (Invitrogen, Carlsbad, CA) for 3 h at 37°C.
TdT was inactivated by incubation at 95°C for 10 min.
The single stranded cDNA was then used as a template in a PCR reaction using Pwo
Polymerase and PolyT-EcoRI primer (Table 2.3) which amplified the cDNA into double stranded DNA and added EcoRI sites to both the 5' and 3' ends. The cDNA was then purified using a PCR clean-up kit (Qiagen, Mississauga, ON) and cut with ZscoRI. The cDNA fragments were ligated into EcoRI digested pGEM7z and transformed into DH5a.
Transformants containing pGEM7Z vectors with insert were distinguished from those without insert using LB+Amp plates supplemented with 0.1 mM X-gal and 1 mM IPTG
(Sambrook and Russell 2001).
Colonies containing plasmids with inserts were sent for sequencing to determine the sequence of the co-purifying RNA. Long PolyA/T tracts within the DNA sequences caused problems during sequencing and so an alternate approach was taken to clone the
RNA sequences. The second method is based on a method outlined by Hinas, et al.
(2007).
RNA fragments co-purifying with ProQ-His6 were treated with PolyA polymerase as above, except GTP was substituted for ATP in the tailing reaction. This substitution results in a poly(G) tail at the 3' end of the RNA. An RNA linker was created by mixing
71 the SP6 promoter DNA oligo (Table 2.3) with the SP6-RNA template oligo (Table 2.3) at a 1:2 (v/v) ratio, heating the mixture to 95°C for 5 min and then cooling the mixture on ice to anneal the two fragments. Four micrograms of the mixed DNA oligonucleotides were mixed with 4 mM NTP mix (1 mM of each NTP) (Invitrogen, Carlsbad, CA), 2
U/uL SP6 polymerase (New England Biolabs, Pickering, ON) in 1 x SP6 polymerase buffer (40 mM Tris-HCl, 6 mM MgCl2, 10 mM DTT, 2mM Spermidine (pH 7.9)). The reaction was carried out at 25°C for 2 h. SP6 RNA polymerase recognizes a 17 bp double stranded DNA sequence as its promoter and will transcribe downstream single or double stranded sequences into RNA. The duplexing of these primers creates a template for SP6 polymerase to create the following RNA oligonucleotide: GGA GAG GAA TTC AGT
GTG TGC GG (Figure 2.1). Isolated RNA was treated with 0.5 U/uL Tobacco Acid pyrophosphatase (TAP) (Epicentre Biotechnologies, Madison WI) in 1 x TAP buffer (50 mM Sodium acetate (pH 6.0), 1 mM Na-EDTA, 0.1% p-mercaptoethanol, 0.01% Triton
X-100) for 2 h at 37°C to remove 2 of the 3 phosphates from the 5' end of the isolated
RNA. TAP treatment allows for ligation of the SP6 synthesized RNA linker to the 5' end of the isolated RNA molecule. The RNA linker was ligated to the 5' end of the isolated
RNA using 80 Units of T4 RNA ligase I (New England Biolabs, Pickering, ON) in a total volume of 80 uL 1 x RNA ligase buffer (50 mM Tris-HCl, 2 mM MgCl2, 1 mM DTT, 40 uM ATP (pH 7.5)).
The polyG tail was used as a priming target for first strand synthesis using the
RETROscript kit (Ambion, Foster City, CA) as specified above with MMLV reverse transcriptase and polyG9 primers. Second strand DNA synthesis and subsequent PCR amplification of the cDNA was performed with Pwo polymerase (Roche, Mississauga,
72 SP6 promoter 5 3'
3' 5' SP6-RNA template SP6 RNA polymerase NTPs
SP6 promoter RNA linker 5' 3'
3 5' SP6-RNA template
Figure 2.1: Synthesis of an RNA linker using DNA oligos. DNA oligonucleotides encoding the SP6 promoter (SP6 promoter) and the SP6 promoter and template for the RNA oligo are melted and annealed to each other. SP6 RNA polymerase synthesizes RNA (RNA linker) with sequence reverse and complementary to the template.
73 ON) and a primer complementary to the RNA linker ligated to the 5' end of the original
RNA molecule with an EcoJU site at the 3' end, and a polyG primer containing either an
A, T, C or G at the 3' end and a Hindlll recognition site at the 5' end (table 2.3). The extra 3' nucleotide ensures that priming occurs within the final 9 nt of the polyG tail and limits the length of the polyGC tract. This resulted in amplification of the target DNA
sequence and introduction of EcoBJ. and Hindlll sites at the 5' and 3' ends of the cDNA, respectively. The resulting cDNA was cut with EcoBl and ligated into EcoJU digested pGEM7Z and transformed into DH5a. Transformants containing pGEM7Z vectors with
insert were distinguished form those without insert using LB+Amp plates supplemented with 20 mg/L X-gal and 0.1 mM IPTG. Isolated plasmids containing inserts were sent
for sequencing.
2.2.17 Denaturing agarose gel electrophoresis
Denaturing agarose gel electrophoresis was performed according to the protocol
outlined by Ambion (RiboPure Bacteria Handbook, Ambion Foster City, CA). An
agarose gel containing 1% (w/v) agarose in 40 mM MOPS (pH 7.0), 10 mM sodium
acetate, 1 mM Na-EDTA, 2.2 M formaldehyde was prepared by dissolving 1% agarose in
72% of the final gel volume. A 10 x MOPS, sodium acetate, Na-EDTA buffer was then added at 10% of the total volume. Once the solution had cooled to about 50°C,
formaldehyde was added at 18% of the final volume in the fume hood. The solution was mixed and then cast in a gel apparatus. The polymerized gel was covered in 1XMOPS buffer (40 mM MOPS (pH 7.0) 10 mM sodium acetate, 1 mM Na-EDTA). RNA samples were prepared by bringing 1 ug of RNA up to a final volume of 11 uL in DEPC treated water. To this sample, 5uL of 10XMOPS buffer (400 mM MOPS (pH 7.0) 100 mM
74 sodium acetate, 10 mM Na-EDTA), 9 uL of 12.3 M formaldehyde and 25 uL of formamide was added. The samples were heated to 55°C for 15 min and then cooled on ice. Ten (J.L of loading dye (1 mM Na-EDTA, 0.25% bromophenol blue, 0.25% xylene cyanol, 50% glycerol) was added and the entire 60 |u.L of sample loaded into the well.
Gels were run at 80 V until the bromophenol blue dye had reached 3A the length of the gel. The gel was stained in 0.5 )ag/mL ethidium bromide solution in 1XMOPS buffer for
20 min and destained with 2, 10 min washes in water. The gel was then visualized under a UV transluminator.
2.2.18 Urea acrylamide gel electrophoresis
To prepare a 6% denaturing urea-acrylamide gel, 1.2 mL of 10XTBE (0.89 M Tris,
0.89 M Boric Acid, 25 mM Na-EDTA), 5.25 g urea, 1.8 mL of 40% acrylamide (T+C) and water to a final volume of 12 mL were mixed together. Six uL of TEMED and 50 uL of 10% APS were added to the mixture before casting the gels. RNA samples were prepared by adding a l(j.g sample of RNA to water to get a final volume of 5 (LIL. To this,
15 uT of formamide buffer (660 uL formamide, 210 uL formaldehyde, 130 uL 10 x
TBE) was added. The samples were heated to 55°C for 15 min and then cooled on ice. 3 uL of RNA loading buffer (1 mM Na-EDTA, 0.25% bromophenol blue, 0.25% xylene cyanol, 50% glycerol) was added. The entire sample was loaded onto the gel and the gel run at 80 V in 1XTBE running buffer that had been pre-heated to 50°C. The gel was stained in 0.5 ug/mL ethidium bromide in 0.5XTBE for 20 min, washed 2x10 min in water and the gel visualized under a UV transluminator.
75 2.3 Protein analysis techniques
2.3.1 Protein assays
BCA assays
Protein concentrations were determined by the BCA (bicinchoninic acid) assay
(Smith, et al. 1985) using a kit obtained from Pierce Biotechnology Inc. (Rockford, IL) with bovine serum albumin (BSA) as the standard.
Bradford assays
Bradford assay (Bradford 1976) was performed with reagents from Sigma-Aldrich
(Oakville, ON) according to manufacturers' instructions. BSA was used as a standard.
Schaffher-Weissman assays
Schaffher-Weissman protein assays were performed as previously described
(Schaffner and Weissmann 1973) with BSA as a standard.
2.3.2 SDS PAGE
SDS-polyacrylamide gel electrophoresis was performed as described by (Laemmli
1970) using a Mini-Protean slab cell (Bio-Rad, Mississauga, ON) with gels comprised of
12% or 10% (w/v) acrylamide with 1.1% or 0.9% (w/v) bis-acrylamide, respectively.
Gels were stained with Gel-Code Blue (Pierce, Rockford, IL) according to the manufacturer's instructions. Tricine SDS-PAGE was performed as described by
(Schagger and von Jagow 1987) with a 16.5% T, 6% C separating gel, a 10% T, 3% C spacer gel, and a 4% T, 3% C stacking gel.
76 2.3.3 Western immunoblot analysis
Western immunoblotting was performed as previously described (Towbin, et al.
1992). Proteins resolved by SDS PAGE were transferred onto nitrocellulose membranes
(Bio-Rad, Mississauga, ON) using a Bio-Rad mini trans-blot cell for 1 hour at 80V using transfer buffer (16 mM Tris, 115 mM glycine, 0.02% (w/v) SDS, 20% (v/v) methanol).
The gel was stained with Gelcode blue to detect proteins that did not transfer. To detect transfer efficiency and to visualize molecular weight marker proteins, the nitrocellulose membrane was stained for 5 min in Ponceau S solution (0.1% (w/v) ponceau S in 5%
(v/v) acetic acid) and destained in water. The remaining Ponceau S solution was removed by washing twice for 10 min with TTBS (25 mM Tris, 14 mM NaCl, 2.7 mM
KC1, 0.1% (v/v) Triton X).
2.3.3.1 Anti-ProQ Western immunoblotting
Anti ProQ Western immunoblotting was performed as described by (Kunte, et al.
1999) using anti-ProQ antibodies purified by adsorption against an affinity coloumn containing E. coli cell extracts from aproQ' strain (Kunte, et al. 1999) and alkaline phosphatase-conjugated goat anti-rabbit immunoglobulin G (Sigma-Aldrich, Oakville,
ON). Blots were visualized with the Enhanced chemiluminescence (ECL) kit (GE
Healthcare, Baie d'Urfe, QC) according to the manufacturer's instructions.
2.3.3.2 Anti-ProP Western immunoblotting
Anti ProP Western immunoblotting was performed as described by Racher, et al.
(2001), with the exception that during development, primary anti-ProP antibody was
77 incubated with the nitrocellulose membrane overnight. ProP was detected using anti-
ProP antibodies purified by repeated adsorption of cleared rabbit serum with an affinity matrix bearing membrane proteins extracted from a AproP E. coli strain (Racher, et al.
2001) and alkaline phosphatase conjugated goat anti-rabbit immunoglobulin G (Sigma-
Alderich, Oakville, ON). Blots were visualized with the ECL kit (GE Healthcare, Baie d'Urfe, QC) according to the manufacturer's instructions.
2.3.3.3 Anti ProX Western immunoblotting
Anti ProX Western immunoblotting was performed as described by Culham, et al.
(1994), using anti-ProX antibodies purified by adsorption of cleared rabbit serum with an affinity matrix bearing periplasmic proteins extracted from a AproXE. coli strain
(Barron, et al. 1987) and alkaline phosphatase conjugated goat anit-rabbit immunoglobulin G (Sigma-Alderich, Oakville, ON). Blots were visualized with the ECL kit (GE Healthcare, Baie d'Urfe, QC) according to the manufacturer's instructions.
2.3.3.4 Anti-His6 Western immunoblotting
Western immunoblotting analysis using anti-His6 horseradish peroxidase (HRP) conjugated antibodies was performed as described in the QIAexpress Detection and
Assay Handbook; protocol 7 (Qiagen, Mississauga, ON). Blots were visualized with the
ECL kit (GE Healthcare, Baie d'Urfe, QC) according to the manufacturer's instructions.
78 2.3.4 Western immunoblotting analysis to determine expression levels
Western immunoblotting was used to adjust the expression levels of ProQ fragments as follows. Full length ProQ (QH6) and the ProQ fragments (H6N, H6C, and H6NC) were purified by Ni (NTA) chromatography for use as standards (2.3.9.2). Extracts of cells cultivated in MOPS buffered minimal medium supplemented with arabinose at various levels (0-13.3 mM) were analyzed by Western immunoblotting in comparison with the purified standards as follows. The quantity of full length ProQ expressed from plasmid pDC77 in E. coli WG914, without arabinose induction, was determined by comparing it with purified QH6 on Western immunoblots. Expression of the ProQ fragments at levels equimolar with full length ProQ, as indicated by standards representing the ProQ fragments, was attained by arabinose supplementation of the relevant cultures.
2.3.5 Crude ribosome preparation
Crude ribosome fractions were prepared as previously described by Spedding, et al.
(2008) Cells were cultured in 1L of LB to an OD60o of 2.0. Cells were harvested by centrifugation (10,000 x g, 10 min, 4°C) and the resulting pellet stored at -40°C for further use. The cell pellet was resuspended in 10 mL ribosome lysis buffer (100 mM
NH4CL, 50 mM MgCH3COOH, 20 mM Tris-HCl (pH 7.5) 1 mM DTT, 0.5 mM Na-
EDTA), and cells lysed by passage through a French pressure cell at an internal cell pressure of 1.6 x 108 Pa. The resulting cell lysate was treated with RNase free DNasel (2
|ig/mL) at 4°C for 20 min. Cell debris was removed from the treated cell lysate by centrifugation (30,000 x g, 30 min, 4°C) 3 times. The cleared cell lysate was layered over a 1.1 M Sucrose cushion in ribosome lysis buffer and the ribosome fraction pelleted
79 by centrifiigation (100,000 x g, 16 h, 4°C). The resulting pellet and supernatant fractions were analyzed for ProQ using western blot analysis and anti-proQ antibodies (2.3.3.1).
2.3.6 Inside out membrane vesicle preparation
Inside out membrane vesicles were prepared as previously described (MacMillan, et al. 1999). Briefly, 20 mL of LB growth medium was inoculated with the strain of interest and the culture grown overnight at 37°C. Five mL of this culture were added to two, 1L volumes of MOPS medium and this culture incubated overnight. Five, 1L
MOPS minimal medium cultures were inoculated with this overnight culture to an OD600 of 0.5 and the cultures grown to an OD600 of 1.0. Cells were harvested by centrifugation
(2500 x g, 20 min, 4°C). Each pellet was washed in 30 mL 0.85% saline (12,000 x g, 10 min, 4°C). Pellets were resuspended in 0.85% saline and combined into one pre-weighed
30 mL centrifuge tube. Following centrifugation, the pellet was weighed and stored at -
40°C. Pellets were resuspended in 20 mL of 0.1 M potassium phosphate buffer (pH 6.6) and inverted vesicles were prepared by passing the cells through a French pressure cell once at a pressure of 550 psi. Cell debris and unlysed cells were removed from the preparation using low speed centrifugation (8000 x g, 20 min, 4°C). Inside out vesicles were harvested by ultracentrifugation (145,000 x g, 2 h, 4°C). The resulting pellet was resuspended using a homogenizer in 1 mL 0.1 M potassium phosphate buffer (pH 6.6).
Inside out vesicles were stored as 100 uL aliquots in liquid nitrogen.
80 2.3.7 Protein solubility screening
2.3.7.1 Comparison of ProQ solubility following overexpression from pBAD24 and pQE-60 based plasmids.
Bacteria were inoculated into LB (10 mL) and incubated overnight at 37°C. Bacteria containing plasmid pDC77 (derived from pBAD24 and encoding ProQ) were subcultured into LB (100 mL) supplemented with arabinose (0.2%, w/v) to attain an ODeoo of 0.4 and incubated, with shaking, at 37°C until the OD600 reached 1. Bacteria harbouring plasmids pREP4 and pMS 1 (derived from pQE-60 and encoding ProQ-Hise) or pMS2 (derived from pQE-60 and encoding ProQ) were subcultured into LB to attain an OD600 of 0.4 and incubated, with shaking, at 37°C until the OD600 reached 0.7. IPTG (isopropyl-P-D- thiogalactopyranoside, 1 mM) was added and incubation continued for 4 h. Cells were harvested by centrifugation (12 000 x g for 10 min at 4°C), and the resulting cell pellets were washed in 0.85% (w/v) NaCl. Cell pellets were resuspended in lysis buffer (75 mM potassium phosphate (pH 7.4) 1 mM dithiothreitol (DTT), 5 mM MgS04, 30 ng/mL
DNase, 30 ug/mL RNase, 1 mg/mL lysozyme) (5 mL per gram of wet weight), incubated on ice for 30 min, lysed by sonication on ice (microtip, Sonicator XL2020 (Heat Systems,
NY)), power set to 4, 4 min total alternating 10 s on, 10 s off and centrifuged (12 000 x g,
20 min, 4°C) to produce soluble and particulate fractions, the latter resuspended in an equal volume of 0.85% NaCl. The distribution of soluble and insoluble protein was assessed by SDS PAGE.
81 2.3.7.2 Optimization of cell lysis conditions
To test the solubility of ProQ and ProQ-His6 expressed from pQE60-based vectors in buffers with a variety of salt concentrations, the relevant strains were cultured, and protein expression was induced as described above. The OD600 was measured, and aliquots were taken that corresponded with 1 mL of culture at an OD600 of 1.0. Cells were harvested from these aliquots by microcentrifugation. The pellets were resuspended in appropriate buffers (100 uL of 50 mM Tris-HCl, 5 mM Na-EDTA, pH 7.4, supplemented with lysozyme (1 mg/mL) and NaCl as specified), lysed by four freeze-thaw cycles
(quick freezing on dry ice for 3 min, thawing at 42°C for 3 min with intermittent vortex mixing), and centrifuged (as above) to produce soluble and particulate fractions. The latter was resuspended in 100 uL of the same buffer. The solubility of ProQ or ProQ-His6 upon lysis in each buffer was determined by analyzing the soluble and insoluble fractions by SDS PAGE.
2.3.7.3 To optimize solubility of purified protein
Protein solubilities were explored as described by Collins, et al. (2004), employing the sodium salts of buffer bases 0.1 M citrate (pH 2.3), potassium phosphate (pH 4.2),
MES (pH 5.6), PIPES (pH 6.5), HEPES (pH 7.5), and TAPS (pH 8.5). These buffer bases were supplemented with the following salts at 0.2 and 0.5 M: MgCh, CaCb,
NH4CI, LiCl, KC1, KSCN, NaN03, or NaCl. Briefly, purified proteins at concentrations of 1 to 2 mg/mL were dialyzed into deionized water overnight. The resulting precipitates were resuspended, aliquots of the suspension were put into microcentrifuge tubes, the protein was recovered by centrifugation, and the supernatant removed. The pellets were
82 resuspended by repeated pipetting in 0.1 mL of the buffers specified above or in water, the suspensions incubated at room temperature for 30 min, and the insoluble protein again pelleted by centrifugation. The amount of protein in solution in the trial buffer was measured with the Bradford Assay (2.3.1). Significant solubilization was deemed to have occurred when the quantity of protein present in the supernatant was at least 3-fold greater than the quantity of protein present in water in the same experiment.
2.3.8 Overexpression
For ProQ or ProQ-His6 purification based on their expression using plasmid vector pQE-60, 100 mL LB was inoculated with E. coli WG806 (SG13009 pREP4 pMS2) or
WG805 (SG13009 pREP4 pMSl), respectively, and grown overnight at 37°C. Two liters of LB were inoculated with the overnight culture to reach an OD600 of 0.4. Cultures were incubated at 37°C until an OD600 of 0.7 was reached, then induced with IPTG (1 mM), and incubated for 4 h at 37°C. The cells were harvested by centrifugation (12,000 x g for 20 min at 4°C) and washed twice with 0.1 M potassium phosphate buffer (pH 7.4) or 0.85% NaCl. The resulting pellet was weighed and stored at -40°C.
HeN, H6C, and H6NC were overexpressed by growing E. coli BL-21 Gold harboring the relevant pQE-80L based plasmid (Table 2.2) overnight in LB medium and subculturing the bacteria to an initial OD60o of 0.4 in the same medium. Cultures were grown to an OD6oo of 0.6, and expression of the recombinant protein was induced by adding IPTG to 1 mM. Incubation continued for 4 h at 37°C, and protein expression stopped when cells were harvested by centrifugation and washed once in ice cold saline
(0.85% (w/v)). The resulting pellets were stored at -40°C.
83 2.3.9 Protein Purification
2.3.9.1 ProQ
ProQ purification was performed as previously described (Smith, et al. 2004; Crane
2004); with the exception that chromatography was performed using an AKTA FPLC component system (Amersham Biosystems, Baie D'Urfe, QC).
2.3.9.2 ProQ-His6
Solutions used during the purification of ProQ-His6 were based on buffer A (50 mM sodium phosphate, 1 M NaCl, pH 8.0). Cells were thawed on ice for at least 15 min before being resuspended in lysis buffer 2 (buffer A supplemented with 0.5 mM PMSF, 5 mM s-amino caproic acid, 5 mM imidazole, 30|ag/mL DNase, 30ug/mL RNase, 0.1 mg/mL lysozyme) (5 mL per gram of wet weight). Cells were disrupted by two passes through a French pressure cell at an internal cell pressure of 1.6 x 108 Pa. The lysate was centrifuged at 12 000 x g for 20 min at 4°C. Ni(NTA) resin (Qiagen, Mississauga, ON)
(1 mL of resin per 5 mL of cell lysate) was loaded into a Bio-Rad Econo-column (Bio-
Rad, Mississauga, ON, 1.5 cm x 30 cm) and washed with 50 mL of water, 50 mL of buffer A containing 0.25 M imidazole, and 70 mL of buffer A containing 10 mM imidazole. The primed resin was mixed with the soluble fraction; the mixture was incubated at 4°C for 1 h before loading into the column and collecting the flow-through.
The column was washed (50 mL of buffer A containing 20 mM imidazole) and ProQ-
His6 eluted with buffer A containing 0.25 M imidazole. Fractions (2 mL) were stored at
4°C.
84 2.3.9.3 His-tagged ProQ Domains
Cell lysis and clearing for purification of His-tagged ProQ fragments was carried out as described for ProQ-His6. Buffers used for purification of H6N, H&C, and HeNC by
Ni(NTA) affinity chromatography were based on buffer A (50 mM sodium phosphate, 1
M NaCl, and 1 mM DTT (pH 8.0)). The lysis buffer was buffer A supplemented with lysozyme (1 mg/mL), DNase I (30 (j.g/mL), RNase A (30 |xg/mL), and Roche EDTA free
Protease Inhibitor (5 mg/mL) (Roche, Mississauga, ON). The column wash buffer and elution buffer were buffer A supplemented with 20 mM and 250 mM imidazole, respectively.
2.3.9.4 ProP-His6
In-side out membrane vesicles were prepared from strain WG 709 (2.3.6) which had been grown in MOPS based minimal medium with 2% (w/v) Arabinose to induce ProP-
His6 expression. One milliliter of the ProP membrane vesicles were solubilized in (0.15
M potassium phosphate (pH 8.0), 33% (v/v) glycerol, 2.5 mM P-mercaptoethanol, 1.72%
(w/v) P-D-dodecylmaltoside) by stirring the solution for 30 min on ice. Debris was removed by centrifugation (13,500 x g, 40 min, 4°C). The resulting supernatant was mixed with Ni-NTA resin (Qiagen, Mississauga, ON) that had been equilibrated with 10 column volumes of buffer 1 (50 mM potassium phosphate (pH 8.0) 300 mM NaCl, 10 mM imidazole, 2 mM p-mercaptoethanol, 10% glycerol (v/v), 0.04% p-D- dodecylmaltoside (w/v)) and mixed at 4°C for 45 min. The incubated resin mixture was put back into a gravity flow column and the flow-through collected. The column was washed with buffer 1 and then with buffer 2 (buffer 1 supplemented with 30 mM
85 imidizole) until a baseline was established at 280 nm. ProP-His6 was eluted from the column into 5-6 mL of buffer 3 (buffer 1 supplemented with 200 mM imidazole). A
Bradford protein assay (2.3.1) was used to determine concentration of ProP-His6 obtained.
2.3.9.5 Size exclusion chromatography
Size exclusion chromatography was performed on full length ProQ and ProQ-His6 as well as on the His-tagged fragments of ProQ (QN, QC, and QNC) using a Superdex 75
HR 10/30 gel exclusion column and an AKTA FPLC component system (Amersham
Biosciences, Baie d'Urfe, QC) with 50 mM sodium phosphate, 0.6 M NaCl, 10 mM imidazole, 1 mM DTT, and 1 mM Na-EDTA (pH 8) as the elution buffer at 4°C with a constant flow rate of 0.5 mL/min. Protein elution profiles were monitored using absorbance at 280 nm and peak fractions were collected and analyzed for the presence of protein. Since the C-terminal domain of ProQ does not contain any aromatic residues, 2 mL fractions were collected over the entire length of the run and each fraction subsequently analyzed for the presence of the protein by SDS PAGE.
2.3.9.6 Reverse phase chromatography
Protein purity was analyzed, and the proteins were further purified by reverse phase
HPLC (RP HPLC). Analytical and preparative scale RP HPLC were carried out on an
Agilent 1100 series HPLC (Agilent Technologies, Santa Clara, CA) with a diode array and a Zorbax C8 column (15 cm x 4.6 mm, 5 ja.m pore size). Varying gradients of eluents
A (0.2% trifluoroacetic acid (TFA) in water) and B (0.2% TFA in acetonitrile (ACN)) were used with a constant flow rate of 1 mL/min. Two milliliter fractions were collected
86 and HPLC purified proteins were further analyzed by analytical RP HPLC, SDS-PAGE, and electrospray mass spectrometry (ESMS).
RP HPLC to analyze protein purity
Analytical RP HPLC was performed with a linear gradient of eluents A (0.2% TFA in water) and B (0.2% TFA in ACN) at a rate of 2% per min over 40 min. ProQ-His6 at 10.6 mg/mL, purified by Ni(NTA) affinity and size exclusion chromatography was dialyzed into 0.1 M sodium citrate (pH 2.3) and diluted to 1 mg/mL in solvent A. A sample (20 uL) was analyzed by analytical RP HPLC (2.3.9.6).
H6N at 30 mg/mL in 0.05 M sodium phosphate, 0.6 M NaCl, and 0.25 M imidazole
(pH 8.0) was diluted 200-fold into buffer A. A 20 uL sample was analyzed by RP HPLC
(2.3.9.6).
HeC, purified only by affinity chromatography and present at 1.6 mg/mL in 0.05 M sodium phosphate, 0.6 M NaCl, and 0.25 M imidazole at pH 8.0, was diluted 50-fold into buffer A. A 20 |uL sample of this mixture was analyzed by RP HPLC (2.3.9.6).
RP HPLC to further purify protein preparations
ProQ-His6 at 10 mg/mL in 0.1 M sodium citrate (pH 2.3) was diluted 4-fold into 5 mL 0.2% TFA in water for further purification using a discontinuous linear gradient of solvents A and B as follows. A 2% B per min linear gradient was used from 1 to 20% B, a 1% B per min linear gradient was used from 21 to 28% B, a 0.1% B per min linear
87 gradient was used from 28 to 48% B, and a 4.2% B per min gradient was used from 48 to
90% B.
H6N at 30 mg/mL in 0.05 M sodium phosphate, 0.6 M NaCl and 0.25 M imidazole
(pH 8.0) was diluted 10-fold into 4 mL of solvent A and purified further using RP HPLC with a discontinuous linear gradient of solvents A and B as follows. A 2% B per min linear gradient was used from 1 to 20% B, a 1% B per min linear gradient was used from
21 to 28% B, a 0.1% B per min linear gradient was used from 28 to 48% B, and a 4.2% B per min gradient was used from 48 to 90% B.
H6C, 7 mg in 5 mL of 0.05 M sodium phosphate, 0.6 M NaCl, and 0.25 M imidazole
(pH 8.0), was mixed with 10 mL of 8 M urea in 0.2% TFA (the final buffer was 0.017 M
Na phosphate, 0.2 M NaCl, 0.083 M imidazole, 5.33 M urea, and 0.013% TFA) and purified using a discontinuous linear gradient as follows: 2% B per min gradient from 0 to 24% B, 1% B per min from 24 to 32% B, 0.1% B per min from 32 to 52% B, and 4.8%
B per min from 52 to 100% B.
2.3.10 Stokes radius determination
Size-exclusion chromatography was performed using the FPLC with a Superdex 75 column (Pharmacia LKB, Sweden) equilibrated with 0.1 M potassium phosphate, 1 mM
DTT, pH 7.4. Bovine serum albumin (67 000, Rs = 3.55 nm), ovalbumin (43 000, Rs =
3.05 nm), chymotrypsinogen A (25 000, Rs = 2.09 nm), ribonuclease A (13 700, Rs = 1.64 nm), and blue dextran (2 000 000) were used as molecular weight standards.
88 2.3.11 Proteoliposome preparation
ProP-His6 was reconstituted into proteoliposomes as previously described (Racher, et al. 1999). Briefly, ProP-His6 was purified using Ni-NTA chromatography (2.3.9.4). 100 mg of E. coli polar lipid extract (Avanti Polar Lipids, Alabaster, AL) in chloroform was moved to a round bottom flask and the chloroform removed using a rotary evaporator at
35 °C for 30 min. The lipids were further desiccated under vacuum for 1 hour at room temperature. Lipids were resuspended in 10 mL dialysis buffer (0.1 M potassium phosphate buffer (pH 7.4) 0.5 raM Na-EDTA 1.5% 2 mM p-mercaptoethanol) supplemented with 1.5% octyl glucoside, by stirring at room temperature for 1 hour. The lipids were transferred to dialysis tubing (SpectraPor 2, 2 mL/cm, 12,000-14,000
MWCO) and dialyzed against 1 L of dialysis buffer for 6 h, 4 times. Prepared lipids were stored in liquid nitrogen for later use.
Lipids were thawed and prepared by extrusion through a 0.4 \im filter 21 times.
Reconstitution was carried out at a lipid: protein ration of 100:1 (w/w). Extruded liposomes were diluted to a final concentration of 5 mg/mL lipid in dialysis buffer with
0.24% (v/v) Triton X-100. Purified ProP-His6 was added to the lipid mixture and incubated at room temperature for 10 min. BioBeads (BioRad 152-3921) were previously prepared by adding 200 mL methanol to 30 g BioBeads and stirring for 15 min. Beads were collected in a sintered glass funnel and washed with an additional 500 mL methanol. The beads were immediately washed in 2L H20 and stored at 4°C until used. Biobeads were added to the protein-lipid mixture to remove detergents used during the preparation of purified ProP-His6 (P-D-dodecylmaltoside (DDM)) and during the reconstitution (triton X-100). For every 1 mL of purified protein added to the
89 reconstitution (protein solution contained 0.04% w/v DDM), 4 mg of BioBeads were added and for every 1 mL of lipid solution used (lipid solution contained 0.24% (w/v)
Triton X-100), 12 mg of BioBeads were added. The mixture was shaken on a rotary platform shaker at room temperature for 1 hour. The same calculated amount of
BioBeads was added and the mixture incubated at room temperature for an additional 1 hour. Twice the calculated amount of BioBeads were added and the mixture incubated at
4°C overnight. The liquid was extracted from the overnight BioBead mixture and put into dialysis tubing (SpectraPor 2, 2 mL/cm, 12,000-14,000 MWCO). The preparation was dialyzed against 2 x 1 L of dialysis buffer over a 48 hour period at 4°C.
Proteoliposomes were sedimented by centrifugation (66,000 x g, 8°C, 45 min). The resulting pellet was resuspended to a final concentration of 60 mg of lipid/mL in dialysis buffer. One-hundred uL samples were aliquoted into cryovials and stored in liquid nitrogen. The protein concentration of the proteoliposomes was determined using the
Schaffner-Weissman Protein Assay (2.3.1).
2.3.12 Co-reconstitution of ProP and ProQ
To load ProP-His6 containing proteoliposomes with ProQ, ProP proteoliposomes were prepared (2.3.11) and stored in liquid nitrogen. Three mL of ProQ-His6 purified as outlined (2.3.9.2 and 2.3.9.5) at 1 mg/mL, was first dialyzed against 0.1 M Potassium phosphate (pH 7.4), 0.5 mM Na-EDTA, 2 raM P-mercaptoethanol and then diluted further to 0.6 mg/mL in the same buffer. Proteoliposomes were then loaded with the purified ProQ-His6 preparation at a 1:1 w:w ratio, or an equal volume of the dialysis
90 buffer without ProQ, by extrusion of the mixture 21 times through a 0.45 um membrane.
The resulting proteoliposomes used for transport assays (2.4.1.2).
2.3.13 Mass Spectrometry
2.3.13.1 Electrospray
Purified ProQ-His6, His6-ProQN and His6-ProQC were analyzed by electrospray mass spectrometry (ESMS), performed using a Mariner Biospectrometer Workstation
(Perspective Biosystems, Forester City, CA). ESMS was performed in the laboratory of
Dr. R.S. Hodges at the University of Colorado at Denver and Health Sciences Center.
2.3.13.2 MALDI-TOF
The masses of ProQ fragments produced by Tryptic digestion or by Glu-C digestion were determined by MALDI-TOF mass spectrometry using Reflex III MALDI TOF
(Bruker Daltonics, Billerica, MA). The identities of the analyzed fragments and the resulting coverage of ProQ (accession number P45577) were determined by comparing their masses with those predicted by ProFound (Zhang and Chait 2000). MALDI TOF
MS was performed at the Biological Mass Spectrometry facility (University of Guelph,
Guelph, ON).
2.3.14 CD spectroscopy
Circular dichroism (CD) spectroscopy was performed on a Jasco-810 spectrometer with constant nitrogen flushing (Jasco, Inc., Easton, MD). Circular optical cells with a path length of 0.05 cm were used to determine the spectra of proteins in 0.1 M phosphate phosphate at pH 7.4 or 0.1 M Na citrate at pH 2.3 in the presence and absence of 0.6 M
91 NaCl over a wavelength range of 195-250 nm in 1 nm increments. The concentrations of the protein preparations were determined by amino acid analysis using a Beckman Model
6300 Amino Acid Analyzer (Beckman Coulter Inc, Fullerton, CA) or by using the BCA protein assay (2.3.1). Mean residue ellipticity was calculated using the following
equation:
e= (eobsxMRW), /(10 x path length (cni)x concentration (zrfr)
where 0 is the mean residue molar ellipticity, 90bs is the observed ellipticity, MRW is the mean residue weight, and concentration is the protein concentration determined by amino
acid analysis. Each spectrum is the average of 8 wavelength scans.
2.3.15 ProQ-His6 proteolysis
2.3.15.1 Limited trypsin proteolysis
ProQ-His6 was purified from E. coli SGI3009 pMSl using a combination of Ni
(NTA) affinity and size exclusion chromatography (2.3.9.2 and 2.3.9.5 respectively).
Purified protein at a concentration of 0.5 mg/mL in 50 mM Na phosphate, 0.6 M NaCl,
10 mM imidazole, and 1 mM DTT (pH 8.0) was diluted to a final concentration of 0.3 mg/mL and combined with trypsin (Promega, Madison, WI), present at 0.1 ug/mL in the
same buffer, at a ratio of 500 ProQ/1 trypsin (w/w). The digestion was carried out at
37°C. Samples were taken at 15 min intervals over the first 90 min and immediately boiled in SDS-PAGE sample buffer for 10 min. Protease resistant fragments were visualized via Tricine SDS-PAGE (2.3.2).
92 2.3.15.2 In-gel trypsin or Glu-C digestion
Gel slices containing protein fragments that were trypsin resistant after 30 min of digestion were excised. Control gel slices were obtained from empty lanes of the same gels. The protein fragments in these gel slices were digested to completion with trypsin or
Glu-C (Promega, Madison, WI) and the resulting peptides were extracted from the gel as
described previously (Zachertowska et al, 2006). Bands were excised from the gel and destained in 25 mM NH4HCO3 in 50% acetonitrile (ACN) for 30 min. The destaining
solution was removed from the sample and the gel fragment dehydrated in ACN for 10 min. The gels were rehydrated in 10 mM DTT and incubated for 30 min at 50°C to reduce the peptides and then the fragments dehydrated again in ACN. The gel was then rehydrated in 55 mM idoacetamide in 50 mM NH4HCO3 for 60 min at room temperature
in the dark in order to alkylate the peptides. The gel fragments were then washed in 50 mM NH4HCO3 for 15 min and then dehydrated in ACN. The resulting gel particles were
dried on a Savant Speedvac (GMI, Ramsey MN) for 20 min before being rehydrated in
20 (xL of Sequencing grade trypsin or Glu-C (Promega, Madison, WI) at 0.01 |xg/ul
Trypsin or GluC in 55 mM HC1, 50 mM NH4HCO3 for 1 h at room temperature. An
additional 50 uL of NH4HCO3 was added and the solution incubated at 37°C for 16-18 h.
To extract the digested peptides from the gel fragments, 50 uL of water was added to
each sample and the samples sonicated for 10 min in a waterbath sonicator at room temperature. The resulting liquid phase was transferred to a tube containing 5 uL of 5%
formic acid in 50% ACN. To the gel pieces, 75 uL of 5% formic acid in 50% ACN was
added. The samples were incubated for 2 min and the liquid phase added to the previous
one. Peptide extraction was repeated once more and all three liquid phases combined.
93 The volume of the extract was reduced to 10 to 15 uL using a Savant SpeedVac (GMI,
Ramsey, MN) and then cleaned using a CI8 ZipTip (Millipore, Billerica, MA). Identities of the ProQ-His6 proteolytic fragments were determined using MALDI TOF mass spectrometry (2.3.13.2).
2.3.16 Crystallization of the N-terminal domain of ProQ
The Histidine tagged N-terminal domain of ProQ was overexpressed and purified using nickel affinity chromatography and size exclusion chromatography (2.3.8, 2.3.9.3 and 2.3.9.5 respectively). Purified protein was dialysed into 0.1 M MES (pH 5.6) in
SpectraPor dialysis membrane (MWCO 12,000-14,000) and then concentrated to 30 mg/mL using a centricon centrifugal filter device with a molecular weight cut off of
30,000 (Millipore, Billerica, MA). Concentrated protein or protein diluted to 14 mg/mL was used for crystal trials. The protein mixture was mixed 1:1 (v:v) with each precipitant
(2.5 uL of each) on a glass microscope slide cover, to give final protein concentrations of either 15 or 7 mg/mL in the crystal screens performed. The protein-precipitant mixture was inverted over a reservoir containing 1 mL of the precipitant. Drops were incubated at room temperature for 3 months and were periodically checked for crystal formation.
Buffers screened in these experiments were parts of screening kits obtained from
Hampton Research (Aliso Viejo. CA) (Hampton Crystal Screen I, Hampton Cryo Screen,
Hampton Cryo Screen II, Hampton Crystal Screen II and Hampton MembFac Screen),
Emerald Biosystems (Bainbridge, WA) (Emerald Wizard I, Emerald Wizard II and
Emerald Cryo I) or Qiagen (Mississauga, ON) (Nextal Classic Suite, Nextal Classics lite,
Nextal AmS04, Nextal pHClear, Nextal PEGs, Nextal Anions, Nextal Cations, Nextal
94 pHClear II). Crystallization screens were carried out in collaboration with the laboratory of Dr. Howell at the Hospital for Sick Children (Toronto, ON)
2.3.17 P-galactosidase assays
P-galactosidase activities were measured as described (Wood, et al. 2005)
2.4 Physiological techniques
2.4.1 Transport assays
2.4.1.1 Whole cell transport assays
Bacteria were cultivated in MOPS minimal medium, as described by (Culham, et al.
2003). Bacteria were cultured in LB medium, then subcultured (1% v/v) into MOPS buffered growth medium adjusted to various osmolalities with NaCl as required and grown overnight with shaking at 37°C. Overnight stationary phase cultures were subcultured into the same medium to an optical density at 600 nm of 0.4 and grown to an optical density at 600 nm of 0.9. When required, the appropriate concentrations of arabinose were added to the culture medium to induce expression of ProQ or its domains encoded on pBAD24-based plasmids. Cells were harvested by centrifugation (10,000 x g, 10 min, at room temperature) and washed 2 times in unsupplemented MOPS based buffer (MOPS buffered growth medium without the addition of ammonium chloride, glycerol, vitamin Bl and tryptophan) and then resuspended in 0.5 mL of unsupplemented
MOPS buffer.
Assay mixtures used for proline uptake measurements were composed of the following: 46.5 uL 10XMOPS (Appendix 1), 4.75 uL 0.132 M K2HP04, 5 uL 20% (w/v)
95 glucose, 0.5 uL 80 mg/mL chloramphenicol, NaCl as needed to attain the required
osmolality, and water to bring the final volume to 455uL.
Twenty five microliters of cell suspension was added to the complete assay buffer
and incubated at 25°C for 3 min. At 3 min, proline uptake was initiated by adding 20 uL
of 14[C]-proline to a final concentration of 200 uM (5 Ci/mol). At 20, 40 and 60 s
following proline addition, 150 uL samples of the assay mixture were taken and cells harvested by filtration through 0.45 um filter membranes (VWR International,
Mississauga, ON). The filters were immediately washed with 5 mL isotonic unsupplemented MOPS buffer. The quantity of proline taken up by each culture sample
at each time point was determined by liquid scintillation counting (Beckman Coulter LS
6500, IL). All assays were performed in triplicate and uptake rates were determined by
linear regression. Serine and glycine betaine uptake were measured in the same way using L-serine at a concentration and specific radioactivity of 20 uM and 25 Ci/mol, respectively or 14[C] glycine betaine at a concentration and specific radioactivity of 100
uM and 5 Ci/mol respectively.
2.4.1.2 Proteoliposome transport assays
Transport assays to determine the proline uptake rates of proteoliposomes reconstituted with purified ProP-His6 or co-reconstituted with both ProP-His6 and ProQ-
His6 were performed as previously described (Racher, et al. 1999). ProP was re
constituted into proteoliposomes as described (2.3.11). The osmolality of the extrusion buffer (0.1 M potassium phosphate (pH 7.4), 0.5 mM Na-EDTA, 2 mM 0- mercaptoethanol) and assay buffers (0.1 M sodium phosphate (pH 7.4), 0.5 mM Na-
96 EDTA, and NaCl to adjust the assay buffer osmolality) were determined using a vapour pressure osmometer (Wescor, South Logan, UT). The membrane potential (A*?) of the proteoliposomes in the presence of valinomycin was calculated using the equation:
+ -RT [K ]in AW = In-——
Af = -24.7 In •* ]in At room temperature
+ + Where [K ]inand [K ]out are the potassium concentrations inside and outside of the proteoliposome respectively; R is the universal gas constant (8.314 J K"1 mol1); T is the temperature in K; and F is the Faraday constant (9.65 x 104 C mol"1) and the final units of
A*¥ are millivolts (mV).
+ The initial [K ]jn in 0.1 M potassium phosphate was determined to be 178 mM
(Racher, et al. 1999). During the proteoliposome assay, 3 |nL of proteoliposomes prepared in 0.1 M potassium phosphate buffer are diluted into 0.1 M sodium phosphate
+ buffer to give a final volume of 630 |j,L. Therefore the concentration of [K ]out is calculated to be:
3 Lxl78mM 85 M [n00, = ( " V63o,L=<'- ™
When proteoliposomes are subjected to an osmotic upshock, water efflux occurs and as a result, the proteoliposomes shrink until the osmolality inside and outside of the
+ vesicles are equal. Since the proteoliposomes shrink, the [K ];n increases resulting in a higher A^. Thus when the assay buffer imposes an osmotic upshock on the
97 proteoliposomes, it is necessary to correct the membrane potential by increasing the
+ [K ]0ut- The amount by which this is corrected is directly proportional to the ratio of the osmolalities of the internal and external buffers:
+ + [K hn = [K Vn (7TT°)
The membrane potential of proteoliposomes in an isotonic buffer is -133 mV and through the addition of KC1 to the assay buffer, this potential can be maintained under each assay condition.
[K+Ut = ( ifhn ) ~ 0.85 mM = ( J**]'n ) - 0.85 mM \e^ /-24.7mV)/ \e^ l-2A.7mV)J
Proline uptake was initiated at time 0 by dilution of 3uL of proteoliposomes into 630 uT 0.1 M sodium phosphate buffer (pH 7.4), 0.5 mM Na-EDTA, with NaCl added to adjust the assay medium to various osmolalities and KC1 added to balance the membrane potential with as outlined above. This buffer also contained 0.2 mM L-[3H] proline (5
Ci/mol) and 0.5 uM valinomycin. Samples were removed from this mixture at 20, 40, 60 and 80 s, diluted with 3 mL ice cold Quench buffer (100 mM potassium phosphate, (pH
6.0), 100 mM LiCl) and the proteoliposomes collected on a 0.22 urn filter and rinsed with a s 3 mL aliquot of quench buffer. The quantity of proline taken up by each culture sample at each time point was determined by liquid scintillation counting (Beckman
Coulter LS 6500, Illinois, USA). All assays were performed in triplicate and uptake rates were determined by linear regression.
98 2.4.2 Fluorescence microscopy
Strains to be analyzed were grown in MOPS minimal medium supplemented with 50 mM NaCl or with 300 mM NaCl. Microscope slides were coated with 40 uL of 2%
(w/v) agarose in water and 4 uL of cell suspension were delivered to the agarose coated slide and covered with a glass cover slip.
RFP localization within the cells was visualized with an Imaging RetigaEX CCD camera mounted on an Axiovert 200M inverted fluorescence microscope (Carl Zeiss
Microimaging) equipped with a Zeiss Plan Neofluor lOOXoil NA1.3 objective. The red fluorescence was excited with a 100 W halogen lamp (N HBO 103) with an excitation filter BP560/40 nm, dichroic mirror FT585 and emission filter BP630/75. Images were obtained and processed using Openlab (Improvision, Waltham, MA).
99 Chapter 3: Purification and Characterization of ProQ and ProQ-His6
3.1 Abstract
Putative E. coli ProQ sequence orthologues are found only in Gram-negative bacteria; although, none have been functionally characterized (Introduction 1.5). In E. coli, ProQ acts to amplify the activity of ProP. The mechanism by which ProQ acts on ProP is unknown at this time. This has prompted the purification of ProQ would facilitate in vitro experiments that may further our understanding of the mechanism of ProQ's function. This chapter describes the overexpression, purification and characterization of a ProQ variant with a C-terminal histidine tag (Hise). It was found that plasmid-encoded
ProQ or ProQ-His6 complemented an in-frame chromosomal deletion (AproQ676), restoring ProP activity. Following overexpression, both ProQ and ProQ-His6 were found to be poorly soluble. Increasing the salt concentration of the lysis buffer improved the solubility of both ProQ and ProQ-His6. Untagged ProQ was found to co-purify with
DNA binding proteins HU and a histone-like protein of similar size and isoelectric point using ion exchange and size exclusion chromatographies, as previously reported by R.
Crane (2004), while ProQ-His6 could be purified to near homogeneity using only nickel chelation affinity chromatography. ProQ appeared to form an extended monomer, as determined by Stokes radius analysis. The Stokes radius of ProQ-His6 was higher than that of ProQ and may reflect altered domain interactions resulting in further extension of the protein conformation. These studies suggest that ProQ-His6 can be used in place of
ProQ in future experiments.
100 3.2 Introduction
As discussed above (Introduction 1.4), Stalmach, et al. (1983) identified a Tn5 insertion mproQ that conferred resistance in E. coli against toxic proline analogues. In allele proQ220: :Tn5, the transposon inserted after nt A314 of the proQ ORF, interrupting the codon forE105 (Kunte et al. 1999). MutationproQ220::Tn5 impaired the osmotic activation of ProP but did not impair transcription or translation ofproP, the latter determined by studying the P-galactosidase activity associated with aproP::lacZ fusion and by Western immunoblotting analysis (Milner and Wood, 1989; Kunte, et al. 1999).
Homology modeling performed by Dr. R.A.B. Keates predicted that the N-terminal domain of ProQ is composed of mostly a-helical structure and can be modeled on the crystal structure of FinO, while the C-terminal domain of ProQ is predicted to contain primarily P-sheet structure and can be homology modeled on either the crystal structure of an SH3-like domain or an Sm motif of Hfq (Introduction 1.5). These structural predictions for ProQ are intriguing, and they must be tested as their functional implications are explored.
To facilitate such studies, I set out to overexpress and purify ProQ. Previous attempts had been made by other members of our group to purify the native ProQ protein using a combination of ion exchange and size exclusion chromatographies (Crane, 2004). Using these techniques, ProQ could be purified to near homogeneity, however, basic DNA binding proteins remained as minor contaminants that were not easily removed (Crane,
2004). This chapter reports the overexpression, purification, and characterization of native and histidine-tagged ProQ (ProQ and ProQ-His6, respectively). This work was done to determine whether ProQ would co-purify with other proteins and whether such
101 preparations would support further structural and functional characterization of ProQ.
Results in this chapter have been previously published and are reproduced, in part with permission from Smith, et ah, 2004 © 2004 American Chemical Society. Contributions made by other authors to this paper have been removed from the results section of this chapter.
3.3 Results
3.3.1 Creation of plasmids for the high level expression of ProQ and ProQ-His6
When ProQ was expressed from its native promoter in E. coli RJVI2, Western immunoblot analysis showed ProQ to be present in the soluble fraction but not the membrane fraction obtained by cell fractionation following disruption using a French pressure cell (Figure 3.1, lane WM). Kunte, et ah demonstrated that ProQ could be expressed via the araBAD promoter under the control of arabinose and AraC in vector pBAD24 (Kunte, et ah 1999). We sought to further elevate the expression of ProQ and
ProQ-His6 by placing proQ downstream of the IPTG-inducible bacteriophage T5 promoter in expression vector pQE-60 (Material and Methods 2.2.12). Control of gene expression from pQE-60 is achieved via incorporation of a plasmid, pREP4, which encodes repressor lacP. Efforts to construct ProQ expression plasmids based on pQE-60 were unsuccessful unless ligation mixtures were transformed into bacteria harboring pREP4. IPTG induction of bacteria harboring the resulting plasmids; pMS2 (encoding
ProQ) and pMSl (encoding ProQ-His6) resulted in high-level expression of proteins with the expected molecular mass (Figure 3.2). However, most of this protein was found in the particulate fraction after cell disruption (Figure 3.2). The solubilities of ProQ and
102 M W WMQ 11
45 31 27
Figure 3.1: Western immunoblot analysis of cell fractions to determine the cellular localization of ProQ. E. coli RM2 was grown in 1L LB medium overnight and cells were harvested and lysed as described for purification of ProQ (2.3.9.1). Cell debris was removed by centrifugation (10,000 x g, 10 min, 4°C) to yield the soluble fraction (S). The membrane fraction (M) was harvested by centrifugation (100,000 x g, 2 h, 4°C). The membrane fraction was resuspended in 75 raM potassium phosphate pH 7.4, 1 mM DTT, 5 mM MgS04 1 mM EDTA and harvested following this wash by centrifugation (100,000 x g, 2h, 4°C). The resulting fractions were the washed membrane fraction (WM) and soluble fraction (W). All of these fractions, as well as standard ProQ (Q) were analyzed by Western immunoblot analysis as described in Materials and Methods (2.3.3.1). Numbers to the left represent the positions of molecular weight markers (kDa). ProQ is present as a band migrating with an apparent molecular mass of 31 kDa as compared to the molecular weight marker.
103 WhoteOM Particulate Sinlnbh* 12 3 1 2 J 12 3
Figure 3.2: Solubility of ProQ and ProQ-His6 following overexpression at 37°C. Cells expressing ProQ from pDC77 (lanes 1) or pMS2 (lanes 2) or ProQ-His6 from pMSl (lanes 3) were grown at 37°C as described in Materials and Methods (2.3.8). Cells were disrupted by sonication, the insoluble fraction separated from the soluble fraction by centrifugation (12,000 x g, 20 min, 4°C) and the amount of ProQ or ProQ-His6 present in each fraction analyzed by SDS PAGE (top) and Western immunoblot analysis (bottom) using anti-ProQ antibodies (2.3.2 and 2.3.3.1). The numbers to the left represent the positions of the molecular weight markers (kDa). ProQ is present as a band with an apparent molecular mass of 31 kDa as compared to the standard molecular weight markers. (From Smith, et al. 2004 with permission)
104 ProQ-His6 were examined after culturing bacteria at varying temperatures (15-37°C) and with IPTG at various concentrations (0.05-1 mM). Under all conditions tested, up to one- half of the overexpressed protein was found in the insoluble fraction (data not shown).
The maximum protein yield was attained in these experiments when IPTG (1 mM) was added at a culture OD600 of 0.7 and growth was continued for 4 h at 37°C.
3.3.2 Lysis conditions alter the solubility of overexpressed ProQ and ProQ-His6
Although ProQ protein expressed from its chromosomal promoter appears to be soluble and cytoplasmic (Figure 3.1; Kunte, et al. 1999), following overexpression from the araBAD promoter in pBAD24 or the bacteriophage T5 promoter in pQE-60, over half of the protein was found in the particulate fraction following cell lysis under diverse concentrations outlined above. The conditions under which the bacterial cells were being lysed were further explored in an attempt to supplement the lysis buffer with components that would lead to an increase in the solubility of the protein. In order to achieve this, a protocol obtained from Jeanne Perry (UCLA, Los Angeles) was used to test the solubility of ProQ and ProQ-His6 in 50 mM Tris-HCl (pH 8.0) 50 mM NaCl, 5mM EDTA buffer supplemented with either no additives or 2 M NaCl or 0.5 M urea or 0.2% (v/v) NP-40 or
10% (v/v) glycerol. This screen showed that both urea and NaCl increased ProQ solubility (Figure 3.3 A). Since the solubility of ProQ in urea was most likely associated with its denaturation, this supplement was not explored further. Varying concentrations
NaCl (0 - 2.5M) were introduced into the buffer and the solubility of the protein following cell lysis explored. This showed that a minimum of 0.6 M NaCl or KC1 was required in the lysis buffer to maintain ProQ or ProQ-His6 in solution (Figure 3.3B). In
105 A Additive None NaCI Glycerol NP Urea SP SPS PSPSP
97— 66— 45— ' 31—" • -*• -m- -ii i^m gp n» m i iggr^* ^ ProQ-His,
Figure 3.3: The solubility of ProQ-His6 is improved at increasing NaCI concentrations. ProQ-His6 was expressed in cells as described in Materials and Methods (2.3.8). One mL of cells were lysed by three repeated freeze-thaw cycles (as outlined in 2.3.7.2) in the buffers outlined below. The soluble fraction (S) was prepared by centrifiigation in a benchtop microfuge for 10 minutes. The resulting pellet was resuspended in an equal volume of buffer to produce the pellet fraction (P). Both fractions were analyzed by SDS PAGE (2.3.2). In (B) only the soluble fractions are shown. Numbers to the left indicate positions of molecular weight markers (kDa). (A) 50 mM Tris-HCl (pH 8.0) 5mM EDTA, 50 mM NaCI pH 7.5 1 mg/mL lysozyme (None) supplemented with either 2M NaCI (NaCI), 10% (v/v) glycerol (Glycerol), 0.2% (v/v) Nonidet P40 (NP) or 0.5 M Urea (Urea). (B) 50 mM Tris-HCl (pH 8.0) supplemented with NaCI at concentrations increasing from 0-1.0 M in 0.1 M increments and then at 1.5 M, 2.0 M and 2.5 M NaCI.
106 Figure 3.3, representative data are shown for ProQ-His6 expressed from pMSl and similar results were obtained for ProQ expressed from pMS2.
3.3.3 Purification of ProQ and ProQ-His6
ProQ was fractionated by cation-exchange chromatography using a MonoS column (Crane, 2004). Upon elution, a single peak was recovered and SDS PAGE of the eluate revealed a major component with an apparent molecular mass of 31 kDa (Figure
3.4). Two low molecular mass proteins previously identified as HU (accession no.
P02342) and a histone-like protein (accession no. PI 1457) were found as contaminants in these preparation also (Crane, 2004). Attempts to further separate ProQ from HU and the histone-like protein by size exclusion chromatography using a Superdex 75 column were not successful. The yield of ProQ was approximately 0.8 mg per gram of wet weight of cells when expressed from the pMS2-based system under these conditions. Plasmid pMSl was created so that ProQ-His6 could be purified by nickel-chelate (Ni(NTA)) affinity chromatography, eliminating other basic proteins. Using Ni(NTA) chromatography to purify the His tagged variant of ProQ made it possible to raise the ionic strength of the lysis buffer to 0.6 M NaCl, leading to an increase in the amount of soluble protein available for purification. The use of Ni(NTA) chromatography successfully yielded ProQ-His6 free of the low molecular mass DNA binding proteins and of other contaminants (Figure 3.4). The yield of ProQ-His6 obtained in this way was 8 mg per gram of wet weight of cells, a ten-fold improvement as compared to results obtained for untagged ProQ.
107 97 > 66 '
9? 45 •""! 66
ProQ- 45 — ,.-.. ;,, 31 iiis
JLf 31— 4£^P>l>riiQ 14 — ^ ^M» jb*-- HI' 2 / •• •"*!-™>r IhssiB.,. flip
Figure 3.4: Overexpression and purification of ProQ and ProQ-His6.
Left panel: Purification of ProQ from the extract ofE. coli SGI 3009 pREP4 pMS2 by ion exchange chromatography. Left lane: soluble fraction obtained by ultracentrifugation of the cleared cell lysate. Right lane: purified ProQ protein eluted from the MonoS column with a KC1 gradient. The major band is ProQ; minor bands of lower molecular weight were determined by R. Crane to be HU and the histone-like protein (HLP) (Crane, 2004).
Right panel: Purification of ProQ-His6 by Ni-(NTA) chromatography following overexpression in E. coli SGI3009 pREP4 pMSl. The left lane shows the soluble fraction obtained from ultracentrifugation of the cleared cell lysate and the right lanes show Pure ProQ-His6 eluted from the Ni-(NTA) resin with buffer containing 0.25 M imidazole. The numbers to the left indicate the positions of the molecular weight markers (kDa) (From Smith, et al. 2004 with permission).
108 3.3.4 Creation of an in-frame deletion of proQ
Two-step PCR and allelic exchange (Link, et al. 1997) were used to create an in- frame chromosomal proQ deletion (AproQ676) and introduce it into E. coli strain
WG210 (RM2 proU205) as described in Materials and Methods (2.2.14). E. coli
\WG210 derivatives with diminished ProP activity were identified using radial streak tests. Isolates showing an increased resistance to the toxic proline analog 3,4-dehydro-
D,L-proline (DHP) were further screened by PCR for loss of the proQ open reading frame
(ORF). An isolate showing the correct PCR product (Figure 3.5A) was confirmed by sequence analysis to encode the peptide MENQPKCLCLGMGL rather than the proQ
ORF. Further analysis using Western immunoblotting confirmed the absence of the
ProQ protein from this isolate (Figure 3.5B). The resulting strain, WG914, was designated as harbouring deletion AproQ676. The proline uptake activity of wildtype and AproQ676 mutant bacteria were determined (Figure 3.5B). Strain WG914 clearly showed the proQ phenotype, as proline uptake activity in this strain was attenuated, with respect to the wild type, at each osmolality tested (Figure 3.5C). Normalization of the proline uptake rates for each strain to the maximal rates obtained when assayed at an osmolality of 400 mmol/kg reveals that deletions at the proQ locus do not alter the shape of the ProP osmotic activation curve (Figure 3.5C). Similar impairment of ProP activity was also observed for bacteria harboring lesionproQ220::Tn5 (Kunte, et al. 1999).
These experiments demonstrated that the proQ phenotype is conferred by mutations at the proQ locus and not due to polar effects of the Tn5 insertion on surrounding loci. This then paved the way for future work requiring a stable proQ null mutation.
109 B II -
C -r 14
• WG210 (ProQ'l O WG914 (Prod)
200 300 400 200 300 400
Assay Osmolality (mmol/kg) Assay Osmolality (mmol/kg)
Figure 3.5: Creation and phenotype of an in-frame proQ deletion. Allelic exchange was used to replace the proQ ORF with one encoding the peptide MENQPKCLCLGMGL. (A) PCR analysis of chromosomal DNA isolated from the proQ+ and proQ strains, WG210 (+) and WG914 (-), with primers AB5338 and AB5628 (refer to table 2.3). The PCR products are expected to be 870 bp in WG210 and 150 bp in WG914. The positions of the molecular size standards are shown to the left of the gel (kbp). (B) Right panel: Western immunoblot analysis of cell lysates of WG210 (+) and WG914 (-) using anti-ProQ antibodies. The SDS PAGE gel is shown in the left panel to show equal loading of the cell lysates. The positions of the molecular size markers are shown at the left the gel (kDa). (C) Left panel: The proline uptake activities of E. coli strains WG914 and WG210 were determined as described in Materials and Methods (2.4.1.1). Strains were cultivated in MOPS based minimal medium supplemented with 50 raM NaCl and transport assay was measured in MOPS media supplemented with NaCl to obtain the indicated osmolality. Right Panel: Proline uptake rates were normalized to the maximal rate attained at an osmolality of 400 mmol/kg.
110 ProQ encoded by: Chromosome Plasmid Q QH« - Q QH« Vector B B B 60 60 60 30 JL ^~~- 25
<1) CD +-> 20 m O (Z Q. d) — j*: 0) 15 - JL 03 O +-> Li. O) D £ 10 i TJ r"n
O c(L U < O __ c o F E • < s_ T T •'•• n r1! ND
Figure 3.6: Complementation of deletion AproQ676 by plasmid-based expression of ProQ or ProQ-His6. Amino acid uptake activity and ProQ or ProQ-His6 expression were measured in bacterial lacking or retaining the chromosomal proQ locus (chromosome - or +; E. coli WG914 and WG210 respectively). They did not (-) or did contain a plasmid encoding ProQ (Q), ProQ-His6 (QHe), or did not contain an insert in the multiple cloning site (-). The plasmids were based on vector pBAD24 (B) or pQE60 (60). l4[C]Proline uptake (top panel) or serine uptake (middle panel) were measured as described in Materials and Methods (2.4.1.1) in assay buffer supplemented with 50 mM NaCl (black bars) or in assay buffer supplemented with 170 mM NaCl (grey bars). Expression of ProQ or ProQ- His6 (bottom panel) was measured by Western immunoblotting using anti-ProQ antibodies (2.3.3.1) (Adapted, with permission from Smith, et al. 2007).
Ill 3.3.5 ProQ-His6 expressed from a plasmid complements a chromosomal deletion
Expression of ProQ from pBAD24-based plasmid pDC77 complemented lesion proQ220::Tn5, restoring ProP activity to the level observed inproQ+ bacteria (Kunte, et al. 1999). I further tested the ability of ProQ and ProQ-His6, expressed from a plasmid, to complement lesion AproQ676. Expression of ProQ or ProQ-His6 from pQE-60 orpBAD24 plasmids restored the ProP activity of host strain WG914 (AproQ676) (Figure
3.6). In this experiment, inducers (IPTG or D-arabinose) were omitted from the growth media to minimize protein expression. Nevertheless, ProQ-His6 was expressed to a higher level than ProQ (particularly when expression was from pQE-60), and ProQ was expressed to higher levels than in cells with a single chromosomal copy of proQ (Figure
3.6). ProQ-His6 expressed from the pQE-60 based vector elevated ProP activity above that of bacteria expressing ProQ from the chromosome or a plasmid, however, when
ProQ-His6 was expressed from vector pBAD24 the elevated ProP activity was no longer observed (Figure 3.6). The elevated activity seen when ProQ-His6 is expressed from the pQE-60 based vector could indicate that the elevated levels of ProQ expression result in further amplification of ProP activity. These results also indicate that ProQ-His6 is functional and can be used in subsequent studies and suggest that further investigation of the relationship between ProQ concentration and ProP activity would be warranted.
3.3.6 The Stokes radii of ProQ and ProQ-His6 suggest the formation of homodimers or an extended conformation
The expected molecular masses of ProQ and ProQ-His6, predicted from their sequences, are 25,876 and 26,872 Da, respectively. Each protein migrated as a single band on an SDS PAGE gel with estimated molecular masses of 28,000 and 33,000,
112 respectively (Figure 3.4). The Stokes radii of ProQ and ProQ-His6 were determined by exclusion chromatography, as outlined in Materials and Methods (2.3.10), to be 3.1 and
3.6 nm, corresponding to apparent molecular mass of 48,000 and 64,000, respectively
(Figure 3.7). These values may indicate that these proteins are globular dimers, but they more likely reflect the presence of N and C-terminal domains of ProQ linked by an extended domain.
3.4 Discussion
ProQ is a soluble protein and its absence from E. coli dramatically attenuates the response of osmoprotectant transporter ProP to increasing extracellular osmolality (the
"proQ phenotype") (Figures 3.6 and 3.7) (Stalmach, et al 1983; Milner and Wood, 1989;
Kunte, et al. 1999). The Wood lab reported that lesionproQ220::Tn5 did not appear to alter proQ transcription (Milner and Wood, 1989) or cellular ProP protein levels (Kunte, et al. 1999), at least under the conditions assayed. This suggested that ProQ would act directly on ProP, via direct protein-protein interactions (Wood, 1999; Saier, 2000). To determine the role of ProQ in the osmotic activation of ProP requires further understanding of the ProQ protein itself. It was necessary produce a sufficiently high yield of purified ProQ or a histidine tagged variant ProQ-His6 to study the structure and function of this protein.
It became apparent that ProQ became insoluble upon overexpression (Figure 3.2).
However, most ProQ remained in solution if the overexpressing bacteria were lysed in buffers containing NaCl at a concentration in excess of 0.6 M (Figure 3.3B). The native
113 0.35 -r
0.30 -
0.25 -
0.20 - >
0.15 -
0.10 -
0.05 -
0.00 -• 1.5 2.0 2.5 3.0 3.5 4.0 Stokes Radius (nm)
Figure 3.7: Standard curve of Stokes radius (nm) (Rs) vs. Kav. ^av is the relative elution volume of the protein (Kav = Fe-Fo/Ft-Fo where: Ve is the elution volume of the protein, Vo is the void volume, as determined by the elution volume of blue dextran and Vt is the total volume of the column) to determine the Stokes radius of ProQ and ProQ-His6. Protein standards with known molecular weights and Stokes radii (bovine serum albumin (67,000, Rs 3.55 nm) ovalbumin (43,000, Rs 3.05), chymotrypsinogen A (25,000, Rs 2.09) and ribonuclease A (13,700, Rs 1.64) were used to develop the standard curve, as outlined (2.3.10). Plotting the Stokes radius vs. the Kav resulted in a linear relationship with the equation y = -0.1274 x + 0.4950. The relationship was used to determine the Stokes radius of ProQ (Kav = 0.1009) and ProQ-His6 (Kav = 0.0364).
114 protein could be purified to near homogeneity in a single step (cation-exchange pDC77 system (Crane, 2004) (Figure 3.3). The major component was previously identified as
ProQ by N-terminal sequencing, amino acid analysis, and MALDI-TOF mass spectroscopy with the minor contaminants being identified as HU and a histone-like protein, both basic, DNA binding proteins (Smith, et al. 2004; Crane, 2004). The Stokes radius of ProQ (3.1 nm) corresponds to an apparent molecular mass of 48,000. ProQ may be a dimer, or this high value may reflect a non-spherical overall shape resulting from the folding of ProQ into two domains linked by an unstructured tether, as predicted by structural modeling.
ProQ was also expressed with a histidine-tag to facilitate its purification by affinity chromatography. Like native ProQ, ProQ-His6 reversed the effects of lesions proQ220::Tn5 and AproQ676 on ProP activity (Figure 3.5). While unlike ProQ, ProQ-
His6 could be purified to near homogeneity by nickel affinity chromatography (Figure
3.3). Its Stokes radius (3.6 nm) was larger than that of ProQ, suggesting a higher apparent molecular mass (64,000). This could have arisen because the C-terminal six- histidine tag interferes with a putative association between distinct N- and C-terminal domains of the protein, it could alter the structure of the C-terminal domain, resulting in a further extension of the overall protein structure, or ProQ-His6 could be a homodimer. If
ProQ-His6 does take on an extended conformation, this conformation might not result in alteration of the in vivo function of ProQ. Overexpression of ProQ-His6 amplified ProP activity beyond the levels attained when ProQ was either expressed at physiological levels or expressed at slightly elevated levels from a pQE-60 based plasmid (Figure 3.6).
When ProQ-His6 was expressed from vector pBAD24 to give a more physiological
115 expression level, the levels of ProP activity returned to normal (Figure 3.6). Thus, the high ProP activity seen following expression from a pQE-60 based vector resulted from the particularly high expression level of ProQ-His6 and not from structural rearrangement due to attachment of the C-terminal tag.
Since they did not co-purify with ProQ-His6, protein HU and the histone-like protein likely co-purified with ProQ because they share its highly basic nature, not due to a direct association. ProQ was deemed unlikely to modulate translation (like its putative structural homologue, FinO) by binding proP mRNA, since proQ lesions had not been observed to affect proP expression or ProP levels under the conditions tested (Milner and
Wood, 1989; Kunte, et al. 1999). However, ProQ could act in this way to alter ProP activity indirectly by modulating the expression of another gene, or could modulate proP expression under conditions that had not yet been explored. The implication of a potential
SH3-like domain at the C-terminus of ProQ suggested that it could be involved in protein-protein interactions that do not survive cellular disruption in high salt (but could account for lack of solubility at low ionic strength). In this study, ProQ and ProQ-His6 were produced at purities and yields sufficient to support further structural and functional analyses. In future studies, ProQ-His6 will be used as a functional equivalent for ProQ.
116 Chapter 4: Domain Structure of ProQ
4.1 Abstract
Homology modeling predicted that ProQ possesses an a-helical N-terminal domain
(residues 1-130) and a P-sheet C-terminal domain (residues 181-232), connected by an unstructured linker (Introduction 1.6). In this chapter, the structural model for ProQ is evaluated, the solubility and folding of full length ProQ and its domains in diverse buffers explored, and the impacts of the putative ProQ domains on ProP activity in vivo are tested. Limited tryptic proteolysis of ProQ revealed protease resistant fragments corresponding to the predicted N- and C-terminal domains. Polypeptides corresponding to the predicted domains of ProQ could be overexpressed and purified to near homogeneity using nickel affinity, size exclusion and reverse-phase chromatographies.
Data from circular dichroism spectroscopy of the purified proteins revealed that the N- terminal domain was predominantly a-helical, and the C-terminal domain was predominantly p-sheet, as predicted. The domains were soluble and folded in neutral buffers containing 0.6 M NaCl. The N-terminal domain was soluble and folded in 0.1 M
MES (2-[M-morpholino]-ethane sulfonic acid) at pH 5.6. Despite having high solubilities in Na-citrate (0.1 M pH 2.3), neither the full length protein nor the N-terminal domain of ProQ were structured. Addition of 0.6 M NaCl to the sodium-citrate buffer resulted in increased structure for the full length protein and the N-terminal domain, but it caused a loss of P-sheet structure for the C-terminal domain. The domains and the linker segment of ProQ were expressed at physiological levels in bacteria lacking the chromosomal proQ locus. Among these constructs, the N-terminal domain, expressed from plasmid pMSlO, could partially complement the proQ deletion. Plasmid-based
117 expression of the foil length protein and a construct lacking only the linker region restored foil activity of the ProP protein. Neither of the domains of ProQ, alone, were able to compete with native ProQ to diminish ProP activity when chromosomal ProQ was present. Experiments aimed at crystallization of the N-terminal domain in order to determine its structure were unsuccessful.
4.2 Introduction
A homology model for residues 1-129 of ProQ was derived from the crystal structure of FinO, a basic mRNA binding protein implicated in the regulation of F-pilus biogenesis in E. coli (Smith, et al. 2004; van Biesen and Frost, 1994). A homology model for residues 178-232 of the protein was derived from the crystal structure of an SH3-like domain of a myosin motor protein in Dictyostelium discoideum (Smith, et al. 2004;
Bauer, et al. 1997). The 50-amino acid long, hydrophilic peptide linking these two regions of the ProQ protein was predicted to be mostly unstructured (see Introduction
1.5).
Experiments involving in vitro systems, such as proteoliposomes co-reconstituted with purified ProP and ProQ, are needed to test the hypothesis that the ProQ protein interacts directly with the ProP protein. Such in vitro work requires that ProQ be purified and stable in solutions compatible with ProP function. Native ProQ and histidine-tagged
ProQ (ProQ-His6) can be kept in solution and purified by increasing the NaCl concentration of the cell lysis buffer to at least 0.6 M and maintaining high salinity during subsequent chromatographic steps (Smith, et al. 2004). However, such conditions may not be appropriate for studies designed to test interactions between ProQ and ProP or
118 other interactions involving ProQ. It is possible that only the N- or C-terminal domain of
ProQ is required to amplify ProP activity and that each domain alone may be more soluble than the full length protein, allowing the domain to serve as a proxy for the full length protein in future in vitro studies.
This chapter describes the overexpression, purification and characterization of ProQ fragments, including the putative domains comprising residues 1-130 and 181-232, in order to test the homology models. Assays performed in vivo were used to show that plasmid-based expression of the N-terminal domain of ProQ alone can partially amplify
ProP activity. Although expression of the C-terminal domain was unable to amplify ProP activity on its own, expression of a protein created by fusing the N- and C-terminal domains of ProQ, without the linker, could fully amplify ProP activity.
The ability of the N-terminal domain of ProQ to partially complement aproQ deletion and its high solubility in folded form in 0.1 M MES pH 5.6 made it a suitable target for crystallization studies. However, experiments performed to determine a suitable buffer for crystallization were unsuccessful. Crystallization trials were carried out in the laboratory of Dr. L. Howell at the Research Institute of the Hospital for Sick
Children (Toronto, ON) with the help of a summer research student Anna Potthoff.
Purification of ProQ and its domains by reverse phase chromatography and characterization of these domains by circular dichroism spectroscopy were performed in the laboratory of Dr. R.S. Hodges at the University of Colorado at Denver Health
Sciences Center with the assistance of Dr. S. Kwok. Results in this chapter have been
119 previously published and are reproduced with permission from Smith et ah, 2007 © 2007
American Chemical Society.
4.3 Results
4.3.1 Limited proteolysis reveals a protease sensitive linker region and supports the predicted domain structure of ProQ
Limited proteolysis was used to compare protease resistant fragments of ProQ with the predicted domains. Tryptic digestion was selected because ProQ contains 36 potential tryptic cleavage sites with 18 in the predicted N-terminal domain (residues 1-130), 13 in the predicted linker domain (residues 131-179), and 5 in the predicted C-terminal domain
(residues 180-232).
These experiments identified four protease resistant fragments after trypsin digestion for 30 min, two of which remained after 90 min of digestion (referred to as Fragment 1 and Fragment 3) (Figure 4.1). The identities of these two fragments were further assessed by independent in-gel digestion with trypsin or Glu-C (trypsin and Glu-C cleave peptide bonds C-terminal to the basic amino acid residues and to glutamic acid, respectively).
Fragment 1 had an apparent molecular mass of approximately 14 kDa (Figure 4.1, left).
In-gel digestion of this fragment with trypsin and analysis with MALDI-TOF MS identified four peptides unique to the digest (absent from control gel slices) (Table 4.1).
If the identified peptides were all derived from sequences near the N-terminus of ProQ, then they would span residues 12-100 (Figure 4.1). Their coverage of the putative N- terminal domain (residues 1-130) was calculated to be 42% by dividing the sum of the numbers of amino acids in the peptides (34 + 20 = 54, Table 4.1) by the number of amino
120 Fragment 1 T-me (min) Trypsjn
Glu-C 27-^ 17-4 130 180 232 6.5 -f Fragment 3 mj UJULL tiin'iriinir iim 3.4-4-. Trypsin
Glu-C
0 130 180 232
Figure 4.1: Limited trypsin proteolysis. ProQ-His6 was purified by Ni(NTA) and size exclusion chromatography. (Left panel) Limited proteolysis and Tricine SDS PAGE were performed as described in Materials and Methods (2.3.15.1 and 2.3.2, respectively). The final lane replicates that shown for 30 min of digestion with Fragments 1 and 3 outlined. The numbers to the left represent the positions of the molecular mass markers (kDa). (Right panel) Fragments 1 and 3 were digested, in-gel, to completion with trypsin or Glu-C and the resulting peptides were extracted from the gel as outlined in material and methods (2.3.15.2), and the peptides analyzed by MALDI-TOF mass spectrometry (Table 4.1). The identities of the detected peptides were determined by searching the NCBI database using ProFound (Zhang, et al. 2000) and are shown as horizontal black bars. Partial oxidation of methionine residues during sample preparation led to the identification of the oxidized peptide species (open bars). The box above the coverage maps represents the distribution of trypsin cleavage sites over ProQ, and the numbers below the coverage maps indicate the residues at the limits of the putative ProQ domains. (Reproduced with permission from Smith et al 2007).
121 Table 4.1: Masses and Identities of Peptides Derived from Tryptic ProQ Fragments by In- Gel Trypsin and Glu-C Digestions j _ Sample Enzyme Measured peptide Peptide identity mass (Da)3 (ProQ residues)4 Fragment 1 Trypsin 1047.203 12-20 1175.207 36-45 1745.2215 21-35 211-226 2289.169 81-100 Fragment 1 Glu-C 1010.693 95-105 1150.731 20-28 1336.823 20-30 2281.5 29-49 Fragment 3 Trypsin 1175.215 36-45 1316.252 215-226 1332.222 215-226 (oxidized) 1745.2386 211-226 21-35 (oxidized) 2122.3566 174-193 36-54 (oxidized) 2446.419 171-193 Fragment 3 Glu-C 1336.648 20-30 2302.530 208-228 ProQ protein fragments 1 and 3 were obtained as described in Materials and Methods (2.3.15.1) and figure 4.1. 2In-gel digestion of each fragment with the indicated enzyme was described as described in Materials and Methods (2.3.15.2). 3The peptide masses were determined by MALDI-TOF MS as described in Materials and Methods (2.3.13.2). 4The identities of the detected peptides were determined by searching the NCBI database using ProFound (Zhang and Chait 2000). 5This mass was found to correspond to ProQ residues 21-35 by Glu-C digestion (see text). 6It was not possible to unambiguously assign these masses to either of the listed ProQ peptides (see text). (Reproduced with permission from Smith, et ah 2007).
122 acids in the putative N-terminal domain (130) and multiplying by 100%. However, the peptide with a mass of 1745 Da could correspond to either N-terminal residues 21-35 or
C-terminal residues 211-226 (Table 4.1). This ambiguity was resolved by in-gel digestion with Glu-C, which produced four unique peptides (Table 4.1), that spanned residues 20-
105 and covered 32% of the putative N-terminal domain (Figure 4.1). Thus, Fragment 1 corresponds to a region within the putative N-terminal domain of ProQ.
Fragment 3 had an apparent molecular weight of approximately 9.3 kDa (Figure 4.1, left). In-gel tryptic digestion of this fragment and analysis by MALDI-TOF MS revealed seven unique peptides (Table 4.1), only three of which could be assigned unambiguously to the predicted C-terminal domain of ProQ (Table 4.1 and Figure 4.1). In-gel digestion of Fragment 3 with Glu-C and MALDI-TOF MS analysis revealed two unique peptides, one derived from a sequence near the N-terminus and another derived from a sequence near the C-terminus (Table 4.1 and Figure 4.1). Fragment 3 has an apparent molecular weight of 9 kDa, and thus, it is not large enough to span amino acids 21-228 of ProQ and more likely contained two fragments with similar molecular weights, which were not fully resolved by Tricine SDS PAGE. One of these fragments would be derived from the predicted C-terminal domain of ProQ, and would include amino acids 171-228. The tryptic peptides would then cover 74% of the putative C-terminal domain (residues 180-
232), and the Glu-C peptide would cover 40% of that domain (Figure 4.1). The other fragment present in fragment 3 would include residues 20-54 in the predicted N-terminal domain of ProQ. The tryptic peptides would cover 26% of that domain, whereas the Glu-
C peptide would cover 8% (Figure 4.1).
123 Thus, limited trypsin digestion of purified ProQ yielded protease resistant fragments corresponding to the predicted N- and C-terminal domains. In contrast, fragments corresponding to predicted tryptic fragments within the linker region (amino acids 130-
173) were not found. These observations are consistent with the predicted domain structure of ProQ.
4.3.2 The boundaries of the N- and C-terminal domains of ProQ
A multiple sequence alignment of 12 ProQ homologues was used to define the boundaries of the putative N- and C-terminal domains (Figure 4.2). Each domain showed sequence and secondary structure conservation, whereas the linker varied in length and amino acid composition. Predominantly a-helical and P-sheet secondary structures were predicted for peptides corresponding to E. coli ProQ residues Ml-El 30 and V180-F232, respectively. This was true even though the sequence identity with E. coli ProQ varied from 99% for the Shigella jlexineri homologue to only 39% for the Shewanella oneidensis homologue. Thus, for further studies, M1-E130 and V180-F232 were chosen as boundaries for the N- and C-terminal domains, respectively (Figure 4.2).
4.3.3 Creation of plasmids for the high level expression of His6-ProQN, His6-
ProQC and His6-ProQNC
Plasmids for the overexpression of ProQ domains were based on the expression vector pQE80L. This vector contains a multiple cloning site downstream of the coding sequence of 6 histidine residues following an initiating methionine, arginine, glycine, serine tag (MRGS). Cloning of a ProQ domain into the HindM site of pQE80L results in expression of an MRGSHHHHHHGS tag at the N-terminus of the protein produced within the cell. Sequences encoding the histidine-tagged N- (HeN) and C-terminal (HeC) 124 E.coli_P45577 -MEN- -QPKLNS--SKEVIAFLAERFPHCFSAEGEARPLKIGIFQDLVDRVAGEMNLS 5 3 Y.pestis_EDR32211 -MEN- -QPKLNS--SKEVIAFLAERFPLCFTAEGEARPLKIGIFQDLVERVQGEENLS 5 3 P.luminescens_NP_929918 -MEN- -QPKLNS--SKEIIAFLAERFPLCFVAEGEARPLKIGIFQDIVERIQDEECLS 53 V.cholerae_YP_001217051 -MEN- -TEKLKN—SKEVIAYIAECFPNCFTLEGEAKPLKIGIFQDLADRLNDDPKVS 53 V.vulnificus_NP_9344 39 MLAGRPLKPGQLMTEKLKN—SKEVIAYIAECFPKCFTLEGEAKPLKIGIFQDLAERLSEDEKVS 63 S.onedensis_NP_71818 8 MES TDKLTD—TNAILAYLYETFPLCFIAEGETKPLKIGLFQKLAERLADDSKVS 5 3 H.somnus_ABI25345 MTEIQKLTN—NKEIIAYLAEKFPLCFSLEGEAKPLKIGLFQDLAEALANDEKVS 5 3 P.multocida_AAK02352 MSRLSFFYGYCFYERKLMYLCEDIIVIFSESELNYAEFFYLKIDTQEKR MTEVQKLTN—NKEVIAYLVEKFPLCFSLEGEAKPLKIGLFQDLAEALQDDERVS102 H.influenzae_EDJ88467 MTDTQLS SQVTDV QTEVQKLTN — AKAIITYLAEKFPLCFVLEGEAKPLKIGLFQDLAEALQDDERVS 66 A.pleuropneumoniae_ABY68 72 4 MSEQQVK IQNG-NKTNPSVKEVITYLAEKFPLCFSVEGEAKPLKVGLFQDLAEALANDEKVS 61 H.ducreyi_NP_8734 9S) MSVQPET MPDSSNKTNPTVKEVIAYLADKFPLCFSIEGEAKPLKVGLSQDLAEALADDEKVS 62 A.turaefaciens ABB59513 MLPFRIGIDTDIEKRLRLDAGLS 2 3
E.coli_P45577 KTQLRSALRLYTSSWRY LYG-VKPGATRVDLDGNPCGELDEQHVEHARKQLEEAKARVQAQRAEQQAKKREAA -ATAGEKEDAPRRERKP-RPTTPRRKEGAE- 153 Y.pestis_EDR32211 KTQLRSALRLYTSSWRY LYG-VKVGAERVDLDGNPCGVLEEQHVEHARKQLEEAKARVQAQRAEQQAKKREAA • IAAGETP-E PRRPRPAGKKPAPRREAGAAPEN 15 6 P.luminescens_NP_929918 KTQLRSALRLYTSSWRY LYG-VKEGAQRVDLDGNSCGELEAEHIEHALQQLTEAKARVQAQRAEQRAKKREAE NVAAGEKNERPTAKKPAPRRRANNTEGEKR 155 v.cholerae_YP_001217 051 KTQLRAALRQYTSSWRY LHG-VKPGATRVDLDGNPCGELEEQHVEHAQAALAESKARVEARRKEQVKKVREEA —KANKPKAKKPQQARRPQN 14 3 V.vulnificus_NP_9344 39 KTQLRAALRQYTSSWRY LHG-VKLGATRVDLDGNECGVLEEEHVEHAKATLAESKAKVQARRKEQAQKARDEE — KS-KPKTKKAPQQRRANKP 153 S.onedensis_NP_71818 8 KTQLRVALRRYTSSWRY LKS-VKAGAQRVDLDGQPCGELEQEHIDHAQAMLKESQEKAKAKRAAQTPKAAPAG KAPAKKAPKKVAVPAR 141 H.somnus_ABI25345 KTQLRQALRQYTSNWRY LHG-CRAGAVRVDLNGEPAGILEQEHVEHAAAKLAEAKAKVAERRAVEKANNPKAN KKRSVYHSGNKSEN-KKSAGK 14 5 P.multocida_AAK023 52 KTQLRQALRAYTSNWRY LHG-CKAGAERVDLQGNVCGILEQEHAEHAAQQLAEAKAKVAAMRAAEKAAKP--E KKRPARRVAAKGQHAKETTTN- 193 H.influenzae_EDJ88467 KTQLRQALRQYTSNWRY YG-CEEGAVRVDLQGNPAGVLDAEHVAHAAQQLAEAKARFAEKRKAEAAAKKAQQ KKHP RKPANK- 148 A.pleuropneumoniae_ABY6872 4 KTLLRQALRTYTMSWRY LAC-CKANVQRVGLQGEEAGIVDEAQAEHAAQTLTVAKEAYAARKAEQRKEQR KEFFKK_ 136 H.ducreyi_NP_8734 9 9 KTLLRQVLRSYTMSWRY LAC-CKANAQR]GLQGENVGIVDEAQAEHAAQSLAVAKEAYAARKAEQRKEQR KDFFKK_ 137 A.tumefaciens ABB59513 --DLRKALRRYTHSAAY LYATAQPEALRHDIVGKPCEPVSEDDAEHAAQSLAVAKEAYAARKAEQRKEQR K 132 N
E.coli_P45577 RKPR-AQKP-VEKAPKT VKAPREEQHTPVSDISALTVGQALKVKAGQNAMDATVLETTKDGVRVQLNSGMSLLVRAEHLVF 232 Y.pestis_EDR32211 RKPRQSPRPQQVRPPRP QVEENQPRPVPVTDISKLQIGQEIKVRAGKSAMDATVLEIAKDGVRVQLSSGLAMIVRAEHLQF 237 P.luminescens_NP_929918 QPPRPQKRPQQARKPVAKPVQAKPIQAAPIQIVDVSSLKIGQEIKVRVGKSSVDASVLEVAKDGVRVQLPSGLAMLVRAEHIQF 23 9 V.cholerae_YP_001217051 -APKVEKKPVETRALAA SELNVGNQVNVNMGKGNMAATIVEVNKEDVRVQL.ANGLQMVVKAEHLRA 20 8 V.vulnificus_NP_93 44 39 QAQKPAKQPVETRALNA DELITGKAVNVNMGKGNMAAT1VE iNKDDVRVQLSNGLQMWKAEH LRA 219 S.onedensis_NP_718188 KTERPAKAAPKVEPVVN LVQAQLTDLAKKQRVNVKLGMTPVAGVITDINKEDIHVQLDSGLTTKVKAEHJLL 213 H.somnus_ABI25345 -KFSKPRQVEQIFVNVD LANLQKGDVVRVKAGDKTTKAE I. LEVVKEGARVELENGL TI.TVSADRLFA 211 P.multocida_AAK02352 -KAKVTRKPKVILNAIE LASLQKGDSVKVKVGESAKKATVLEVVKDSARVQLENGLV] TVTAEHLFA 25 9 H.influenzae_EDJ88467 -NLK--KESKLSLSAVD FSQISVGSVVKVKAGDNAKKATVVEVLKDSARVEi_ENGL.[HNVAADRLFA 212 A.pleuropneumoniae_ABY68724 -KAQEEKAKKNAANQVK KTPRVAKEAS-VKATAESLAALTSKFGKGNK 182 H.ducreyi_NP_8734 9 9 -KAREERNAKTMNKAVK KGS-PKKDTF-AKATAESLAVLTHKFSKGNK 18 2 A.tumefaciens ABB59513 Figure 4.2: Protein sequence alignment of ProQ and its homologues Residues identical in all sequences are denoted with * whereas conserved amino acids are denoted . or : below the sequence alignment. Sequences represented (with Accession Numbers) are from E. coli K-12 (P45577), Yersiniapestis (EDR32211), Photorhabdus luminescens (NP929918), Vibrio cholerae (YP001217051), Vibrio vulnificus (NP934439), Haemophilus influenzae (EDJ88467), Haemophilus somnus (ABI25345), Pasteurella multocida (AAK02352), Shewanella oneidensis (NP718188), Actinobacillus pleuropneumoniae (ABY68724), Haemophilus ducreyi (NP873499), and Agrobacterium tumefaciens (ABB59513). Numbers to the right of the sequence indicate the amino acid positions in the individual proteins. Secondary structure predictions made with Jpred (Cole et al., 2008)are indicated as follows, red - predicted a-helix, green - predicted P-sheet, black - predicted coil. The boundaries of the N- and C-terminal domains are indicated with an N and a C respectively. (Adapted with permission from Smith, et al. 2007)
125 domains of ProQ as well as a variant encoding the first 130 residues with the final 52
amino acids of ProQ linked together by amino acids AW (HeNC) (Table 4.2), were
cloned into this vector. Restriction endonuclease and DNA sequence analysis was used
to confirm the identity of each construct.
4.3.4 Creation of plasmids for the physiological level expression of ProQN, ProQC, ProQNC, ProQNL, ProQLC and ProQL
Plasmids used for complementation and competition experiments were based on the
vector pBAD24. With this vector, expression of the target sequence, under control of the araBAD promoter, can be upregulated through supplementation of the growth medium with arabinose. ProQ sequences corresponding to the N-terminal (N), N-terminal and
linker domain (NL), linker domain alone (L), linker and C-terminal domain (LC) C-
terminal (C), as well as the N-terminal domain fused to the C- terminal domain (NC)
were cloned into the Ncol and HindM sites of pBAD24 following excision of the coding
sequence from the pQE80L based plasmids (N, C and NC) (Materials and Methods
2.2.12), or following PCR amplification and restriction of the amplicon (Table 4.2).
4.3.5 Overexpression and purification of His6-ProQN, His6-ProQC and His6- ProQNC
The putative N- and C-terminal domains of ProQ (HeN and H6C, Table 4.2) were
overexpressed and purified to determine whether each could fold as a protease resistant
domain in vivo. Preliminary studies showed that H^ and HeC were expressed maximally
when strain BL21-Gold containing the relevant plasmid was grown in LB medium to an
OD6oo of 0.6, and expression was induced with 1 mM IPTG for 4 h at 37 °C (Figure 4.3).
126 Table 4.2: Plasmids and Encoded Proteins Plasmid Vector Encoded protein Name Structure Schematic pDC77 pBAD24 ProQ pMSll pBAD24 N ProQ(Ml-E130) pMS15 pBAD24 M-ProQ(S181- F232) pMS18 pBAD24 NL ProQ(M 1-1183) pMS19 pBAD24 LC M-ProQ(E131- F232) pMS21 pBAD24 NC ProQ(Ml-E130)- AW-ProQ(V180- AW F232) pMS22 pBAD24 M-ProQ(E131- VI80) pMSl pQE60 QH6 ProQ-RSH6 pMSlO pQE80L H6N MRGSH6GS- ProQ(Ml-E130) pMS13 pQE80L H6C MRGSH6GSM- ProQ(S181-F232) pMS14 pQE80L H6NC MRGSH6GS- ProQ(Ml-E130)- AW AW-ProQ(V180- F232)
'The details of plasmid construction are provided in Materials and Methods (2.2.12). 2The vectors are pBAD24 (Guzman, et al. 1995), pQE60 (Qiagen, Mississauga, ON), and pQE80L (Qiagen, Mississauga, ON). 3Filled rectangles indicate the segments of ProQ encoded by the listed plasmids. N- terminal open squares indicate the N-terminal tag MRGSHHHHHHGS. The C-terminal grey square for the protein encoded by pMS 1 denotes an RSHHHHHH tag. NC and HeNC contain an AW linker between the N- and C-terminal domains. Reproduced with permission from Smith, et al. 2007.
127 A S A1 G1 G2 97- 66" 45-
QHf 21- •itmm-
B A1 G1 G2 97- 66« 45' 31- 21- 14. H6N
A1 G1 G2
HfiC
Figure 4.3: Analysis of proteins purified by Ni(NTA) and gel exclusion chromatographies. Analysis of fractions derived from the purification of (A) QH6, (B) H6N, and (C) H6C. Proteins were purified, and SDS PAGE (QH6 and H6N) and tricine SDS PAGE (HeC) were performed as described in Materials and Methods (2.3.2). The numbers to the left indicate the sizes of the molecular weight markers. S, soluble fraction from the whole cell lysates; Al, proteins purified by Ni(NTA) affinity chromatography; and Gl and G2, fractions obtained by gel filtration chromatography. Gel filtration chromatography of the ProQ fragments partially resolved them from contaminants and representative fractions are shown (Gl and G2). (Reproduced with permission from Smith, et al. 2007).
128 QH6 was overexpressed as previously described (Chapter 3 results and Materials and
Methods 2.3.8). QH6 and HeN were purified by Ni(NTA) affinity and gel exclusion chromatographies as described in Materials and Methods (2.3.9.2, 2.3.9.3 and 2.3.9.5 respectively) (Figure 4.3), and the purity of the resulting preparations was determined by analytical RP HPLC (Figure 4.4A, left column of chromatograms). This material was further purified by preparative RP HPLC and the purity of the resulting preparations determined by analytical RP HPLC (Figure 4.4A, right column of chromatograms)
(Materials and Methods 2.3.9.6). In each of the resulting RP HPLC preparations, a single protein of the expected molecular mass was evident upon SDS PAGE and dominated the mass spectrum (Figure 4.4B and Table 4.3). Analytical RP HPLC of QH6 purified by
Ni(NTA) affinity and gel exclusion chromatographies (Figure 4.4A, QH6 Left) yielded a
SDS PAGE (Figure 4.4B, Table 4.3). The basis for the resolution of these protein fractions was not determined. Protein from the peak with a retention time of 57.3 min was used during subsequent analyses. The composition of that material is illustrated in
Figure 4.4B. H6N was purified to near homogeneity by Ni(NTA) affinity chromatography (Figure 4.3, lane Al). Size exclusion chromatography removed minor contaminants present in the preparation (Figure 4.3, fraction G2). Analytical RP HPLC detected some minor contaminants (Figure 4.4A, left). HgN was resolved from the minor contaminants by preparative RP HPLC, as shown by subsequent analytical RP HPLC
(Figure 4.4A, right), SDS PAGE (Figure 4.4B), and ESMS (Table 4.3).
HgC could be partially purified, but many contaminants were evident when it was viewed by Tricine SDS PAGE (Figure 4.3, lane Al). Size exclusion chromatography resolved part of the HeC from the contaminants (Figure 4.3, lane G2); however, the
129 Figure 4.4: A: Analysis of proteins purified by reverse phase HPLC. Analytical RP HPLC was performed as described in Materials and Methods (2.3.9.6). Proteins were further purified by preparative RP HPLC as described in Materials and Methods (2.3.9.6). Peaks with retention times below 40 min or above 80 min were not detected for any of the protein preparations. (QH6, Left) analytical RP HPLC of QH6 purified by Ni(NTA) affinity and gel exclusion chromatographies, dialyzed into 0.1 M Na-Citrate buffer pH 2.3. (QH6, Right) Analytical RP HPLC of QH6 following purification by RP HPLC. The fraction analyzed in this chromatogram is from a fraction corresponding to the second peak in the chromatogram of the partially purified protein preparation (*) (QH6, Left). (H6N, Right) analytical RP HPLC of H6N, purified by Ni(NTA) and size exclusion chromatography as described in Materials and Methods (2.3.9.6). (H6N, Right) Analytical HPLC of a fraction obtained from preparative HPLC of the partially purified HeN protein preparation. Analysis was performed on a fraction corresponding to the first major peak in the chromatogram of the partially purified protein (*) (H^N, left). (H6C, Left) Analytical RP HPLC of H6C, purified by Ni(NTA) affinity chromatography only as described in Materials and Methods (2.3.9.6). (HeC, Right) Analytical RP HPLC of a fraction of HeC following purification by preparative RP HPLC. The fraction analyzed corresponds to protein present in the second major peak in the chromatogram of the affinity purified protein (*) (HeC, Left). B: SDS PAGE analysis of the major peaks present in the partially purified and RP HPLC purified preparations. C- Analysis of protein purified by affinity and gel exclusion chromatographies. 1, 2 and 3 - Correspond to fractions obtained from the major peaks present in the RP HPLC chromatogram of the partially purified protein. P - SDS PAGE analysis of RP HPLC purified protein. (Adapted with permission from Smith, et al. 2007).
130 B QH« C 1 2 J P 97.4 66 45 31 g&^ 21 14
Z3 1 UN C I 2 3 P 97.4 E c 66 o 45 CM
CO 31 21 c 14 TO •e o HX C I 2 P 97.4 — * 66 — 45 — 31 — 21 — L> 14 _
¥> *Z. Retention Time (min)
131 Table 4.3: Molecular masses of purified ProQ fragments Protein Mass (Da) Name Sequence Predicted Measured1 Z QH6 ProQ-RSH6 26958.6 26963.6 26962.23 H6N MRGSH6-ProQ(Ml -E130) 15980.9 15999.2 H6C MRGSH6GSM-ProQ(S 181- 7125.2 7127.4 F232) Masses of proteins purified by RP HPLC were determined by ESMS as described in Materials and Methods (2.3.13.1). 2Mass of ProQ-His6 present in peak 2 of the RP HPLC chromatogram (retention time of 57.3 min) (Figure 4.4). 3Mass of ProQ-His6 present in peak 3 of the RP HPLC chromatogram (retention time of 61.9 min) (Figure 4.4).
132 protein recovery was poor. For further experimentation, HeC was only purified by
Ni(NTA) affinity chromatography to reduce protein losses. Analytical RP HPLC of this preparation showed that many contaminants remained (Figure 4.4A, left). Instability in
solution prevented further purification of HeC by RP HPLC using the buffers employed
for the other ProQ fragments. It was, therefore, denatured in 0.05 M sodium-phosphate,
0.6 M NaCl, and 0.25 M imidazole at pH 8.0 containing 6 M urea prior to RP HPLC purification. H6C was resolved from the contaminants by preparative RP HPLC, as
shown by subsequent analytical RP HPLC (Figure 4.4A, right), Tricine SDS PAGE
(Figure 4.4B), and ESMS (Table 4.3). These experiments showed that N- and C-terminal
fragments of ProQ folded in vivo and could be purified, supporting the hypothesis that they represent domains present in full length ProQ.
4.3.6 Increasing the solubility of QH6, H6N and H6C
The solubilities of HgN, HeC, and full length QH6 were explored as described by
Collins et al. (2004). The buffers tested are listed in Materials and Methods (2.3.7.3).
Among the buffers tested, only sodium citrate (0.1 M, pH 2.3) could resolubilize
aggregated HeN, W&C, and full length QH6. Purified QH6 and HeN could be dialyzed into this citrate buffer and concentrated above 25 mg/mL using Millipore YM-10 (MWCO
10,000) centricon devices (Millipore, Mississauga, ON). HeN was also soluble in MES
(0.1 M, pH 5.6), but H6C and QH6 were not.
4.3.7 Analysis of the secondary structures of QH6, H6N and HeC
Circular dichroism (CD) spectroscopy was used to determine whether HeN and HeC would fold in vivo to form domains with the predicted secondary structures and to
133 compare the secondary structures of these proteins in buffers that enhance their solubilities. In CD spectra, a-helical secondary structure is characterized by strong ellipticity minima at 208 and 222 nm. In highly a-helical proteins, the ellipticities at 222 and 208 nm are similar. In proteins with mixed a-helical and random structure, the ellipticity at 222 nm is much less than that at 208 nm because random structure exhibits a
CD spectral minimum below 200 nm (as shown in Figure 4.5 for QHe). Canonical P- sheet structure exhibits a weak minimum at 217 nm.
The CD spectrum of QH6, purified by RP HPLC, lyophilized and resuspended in 0.1
M potassium phosphate buffer at pH 7.4 (Figure 4.5, top) was similar to that of ProQ in
10 mM potassium phosphate at pH 7.4(Smith, et al. 2004). These spectra were taken to represent the native secondary structure of ProQ. The addition of 0.6 M NaCl did not noticeably change the CD spectrum (Figure 4.5, top). Similar spectra were also obtained with the low pH sodium citrate buffer (0.1 M, pH 2.3) in the presence of 0.6 M salt.
However, a different spectrum, which showed considerably more random coil structure, was obtained in the absence of salt at pH 2.3.
Like QH6, H6N was expected to assume its native conformation in 0.1 M phosphate phosphate buffer at pH 7.4. The CD spectrum of HeN indicated that it is mostly a-helical
(Figure 4.5 Middle), in agreement with the model for this domain based on the crystal structure of FinO which predicts 62/130 amino acids to take on a-helical structure and
6/130 amino acids to take on a (3-sheet secondary structure (Smith, et al. 2004; Ghetu, et al. 2000). There was very little change in the CD spectrum when 0.6 M NaCl was added to this preparation (Figure 4.5, middle). As for QH6, the change of pH to 2.3 in the
134 o E TE3 o Wavelength a>
& * 8 •*»»", ^ .9- Qj © H6N
©
• 0 1 M K-PhospftaJe pH 7 4 a: O 0 1MK-PhOSph«!e06MN»ClpH74 • 0 1 M Na-Cdrale pH 2 3 c V 0 1 M Na-CsfraW 0.6 M NSO PH 2 3 (0 0) s Wavelength ,J3 220 0 HfiC
Figure 4.5: Circular dichroism spectra of QH6, HeN, and HeC. RP HPLC purified and lyophilized QH6, f^N, and HeC were dissolved at 1.6 mg/mL in 0.1 M potassium- phosphate at pH 7.4. The solutions were dialyzed into 0.1 M potassium-phosphate at pH 7.4 or 0.1 M sodium-citrate at pH 2.3. The dialysates were diluted with the same buffer or with the same buffer supplemented with NaCl to give 0.04 mg/mL protein in 0.1 M K phosphate buffer at pH 7.4 (circles) or 0.1 M sodium-Citrate buffer at pH 2.3 (triangles), alone (closed symbols) or supplemented with 0.6 M NaCl (open symbols). HeN was also analyzed in 0.1 M MES at pH 5.6 (squares). CD spectra were recorded as described in Materials and Methods (2.1.4.). (Adapted, with permission, from Smith, et al. 2007).
135 presence of 0.6 M NaCl had no effect on the spectrum of H6N. Also, as for QH6, the spectrum for F^N showed a large decrease in a-helical structure in the absence of salt at pH 2.3. CD analysis of H6N in 0.1 M MES showed that it folded similarly in that buffer and in phosphate phosphate (Figure 4.5, middle). Thus, QH6 and F^N behaved similarly, being sensitive to low pH in the absence of salt. (Figure 4.5, middle), in agreement with the model for this domain based on the crystal structure of FinO which predicts 62/130 amino acid residues to take on an a-helical secondary structure.
CD spectroscopy studies of F^C showed that its structure was mostly (3-sheet, in agreement with the predicted secondary structure for this domain of 28/52 residues taking on a P-sheet, and none of the residues taking on an a-helical structure. The spectra were similar at pH 7.4 in the presence and absence of salt and at pH 2.3 in the absence of salt.
However, the addition of 0.6 M NaCl to the pH 2.3 buffer disrupted the structure (Figure
4.5, bottom).
These experiments showed that tagged, full length ProQ (QHe) and the expressed N- and C-terminal fragments of ProQ (F^N and HeC, Table 4.2) folded to form domains with the anticipated secondary structures in 0.1 M potassium phosphate buffer at pH 7.4.
CD spectroscopy offered no evidence for the formation of an SH3-like domain by F^C
(see Discussion). Although these proteins were soluble in sodium citrate buffer at pH 2.3, the absence of salt increased the random coil structure of QH6 and F^N. Interestingly, the presence of salt at pH 2.3 decreased the structure of F^C. F^N was also found to be soluble and structured in 0.1 M MES buffer at pH 5.6, where it can be concentrated to more than 30 mg/mL. Thus, 0.1 M MES is a suitable buffer for further in vitro studies of the N-terminal domain of ProQ.
136 4.3.8 Functional analysis of the domains of ProQ
Untagged ProQ fragments (Table 4.2) were expressed at physiological levels to test
their in vivo functions. To determine if expression of these proteins could complement a
chromosomal proQ deletion or impair the function of full length ProQ in vivo, it was
necessary to ensure that they were expressed at comparable levels. This study employed
the vector pBAD24, which allows for the tight control of protein expression from the
arabinose inducible araBAD promoter. ProQ was present at 2 nmol/mg cell protein when
proQ was expressed from the pBAD24 derived plasmid pDC77 in E. coli WG914
cultivated in MOPS medium (Materials and Methods 2.3.4). The levels of arabinose
supplementation yielding equivalent expression levels for the ProQ fragments were
determined by Western immunoblot analysis, comparing the expressed protein to purified
full length ProQ (QH6), the ProQ domains (HeN and HeC), and a variant lacking the
linker region (HeNC) at known concentrations. The fragments that included a linker (NL,
LC, and L) and that lacking the linker (NC) were degraded in vivo (Figure 4.6B). This
degradation yielded sub fragments of similar size to the N- or the C-terminal domain
identified by limited proteolysis in vitro (Figure 4.6B), suggesting that these two ProQ
domains were also present in vivo.
Previous studies have demonstrated that transposon insertion in or deletion of proQ
decreases ProP activity. Expression of ProQ or ProQH6 from a plasmid complements proQ lesions, indicating that proQ is responsible for the effects on ProP (Kunte, et al.
1999; Smith, et al. 2004). In this study, the impacts of ProQ fragments on ProP activity
were determined by measuring the proline uptake activities of bacteria in which ProP was
the only proline transporter expressed, and proQ was either expressed from the
137 A I) Complementation li) Competition
e I. a.
ffii ! •
s 3f a. I •g a. Chromosome +++++++++ v Q N NL C LC NC L Plasmid - - Q N NLC LC NC L
B Standards Q N CNC "**"•" ^9K
Figure 4.6: Functional analysis of ProQ fragments. Panel A: The proline uptake activities of bacteria expressing ProQ fragments were determined as outlined in Materials and Methods (2.2.1.1). Proline uptake rates were measured in the presence (+) or absence (-) of chromosomally encoded ProQ (Chromosome) and in the absence of plasmid (-), in the presence of the empty pBAD24 vector (-v), or in the presence of plasmid-encoded Q, N, C, L, NL, LC, or NC (Plasmid). (The ProQ fragments are defined in Table 4.2). Cells harbored a proQ deletion (I) or wild type proQ (II) at the position of the chromosomal prog locus. Panel B: Protein expression levels were determined by Western immunoblotting with anti-ProQ antibodies as outlined in Materials and Methods (2.3.3). Samples were loaded at a concentration of 0.08 mg of total cell protein. The central panel in B shows histidine tagged protein standards, purified via Ni(NTA) affinity chromatography. Quantities of these proteins equimolar to the quantity of ProQ in cells with plasmid pDC77 were used. These quantities were 0.1 ug (QH6), 0.082|ug (H6NC), 0.06 jig (H6N), and 0.026 |ug (H6C). (Reproduced with permission from Smith, et al. 2007).
138 chromosome {E. coli WG210) or absent because of an in-frame chromosomal deletion {E. co//WG914).
As previously observed, the proQ deletion decreased ProP activity, and expression of proQ from plasmid pDC77 restored it. Plasmid-basedprog expression yielded higher
ProQ levels than chromosome-based prog expression, yet the resulting ProP activities were similar (Figure 4.6AI). The expression of L, LC, or C did not have an effect on amplifying ProP activity, whereas the expression of N or NL partially restored ProP activity (Figure 4.6AI). Increasing the expression of N or NL by inducing their expression from the pBAD24 based plasmid with arabinose did not increase ProP activity above this intermediate level (Figure 4.7). The expression of a ProQ variant lacking the
50-amino acid linker region (NC) fully restored ProP activity (Figure 4.6AI). As anticipated, N, NL, LC, and NC were all expressed at levels at least equivalent to that of
ProQ (Figure 4.6BI). Neither C nor L could be detected despite full arabinose induction
(Figure 4.6B). However, the failure of LC to complement the proQ deletion suggests that expression of these subfragments would also be ineffective. Thus, the N-tenninal domain of ProQ may contain residues that are important for ProQ function because this domain gives partial complementation. However, both the N- and C-domains of ProQ are important for its complete in vivo function. Because ProQ was predicted to interact with
ProP, the expression of a competing domain alone in the presence of the full length protein might be expected to decrease ProP activity. However, none of the tested ProQ fragments impaired the ability of full length ProQ to amplify ProP activity (Figure 4.6IIA and 4.6IIB).
139 ProQ + -NNNNNN + - NL NL NL NL NL NL %Arabinose 0 0 0 0.006 0.02 0.06 0.08 0.2 0 0 0 .0002.0006.002 .006 .02
B
Figure 4.7: Effect of expression level on the ability of N and NL to complement a chromosomal ProQ deletion. Panel A: The proline uptake activities of ProQ+ (+), ProQ" (-) or ProQ" strains expressing N from plasmid pMSl 1 (N) or NL from plasmid pMS18 following growth in MOPS minimal medium supplemented with arabinose to the indicated percentage (w/v). Transport rates were measured as outlined in Materials and Methods (2.2.1.1). Panel B: Western immunoblot analysis using anti-ProQ antibodies was performed as described in Materials and Methods (2.3.3) in order to determine the amount of N (Left) or NL (Right) present in cells following growth in various concentrations of arabinose.
140 4.3.9 Crystallization Studies
The N-terminal domain of ProQ was purified to near homogeneity using Ni(NTA) and size exclusion chromatography (Figure 4.3B). This preparation could be concentrated to levels above 30 mg/mL in 0.1 MMES (pH 5.6) and results from CD spectroscopy revealed that it retained a folded structure (Figure 4.5B). This protein was therefore selected as a target for crystallization studies.
Crystallization of FL^N was attempted using the hanging drop method as described in
Materials and Methods (2.3.16). This involves mixing the purified protein 1:1 (v:v) with the crystallant buffer on a cover slip. The drop was then inverted over 1 mL of crystallant solution and allowed to equilibrate over time by diffusion of vapour from the protein solution to the crystallant solution. For crystallization trials, the final concentrations of H6N in the hanging drop following dilution in the crystallant were 7 mg/mL and 14 mg/mL for each buffer screened. A variety of buffers were used in the crystallization screen as outlined in Materials and Methods (2.3.16), however none of these conditions yielded crystals of HeN. It is possible that a lower range of protein concentrations used in future crystallization studies as many of the trials resulted in formation of a precipitate following mixing with the crystallant, indicating that either the protein or crystallant was at too high a concentration.
4.4 Discussion
The full length ProQ protein was previously purified and found to have limited solubility. Smith et al. (2004) proposed that ProQ comprises an N-terminal a-helical domain connected to a C-terminal SH3-like domain by an unstructured linker based on
141 comparative modeling studies performed by Dr. R.A.B. Keates (Introduction 1.5;Smith, et al. 2004). The mechanism by which ProQ acts on ProP cannot be predicted via sequence analysis because a BLAST search did not identify homologues with known functions. To test the structural model and design functional tests, we first assessed the domain structure of ProQ by limited proteolysis. As predicted, N-terminal and C-terminal fragments of ProQ were protease resistant and hence apparently folded, whereas the central linker was protease sensitive and hence apparently unstructured (Figure 4.1). The protease resistant domains persisted even after 90 min treatments with trypsin, even though many trypsin cleavage sites are present in each domain. Further, each domain could be purified following expression in E. coli to give a protein that is consistent with the predicted molecular weight (Figure 4.3, Table 4.3) indicating proper folding of these domains and no proteolytic cleavage in vivo.
The identification of functional ProQ domains may allow us to engineer a functional protein that is more stable and soluble in a wider variety of buffer systems than full length ProQ in vitro. The boundaries of the putative N- and C-terminal domains ofE. coli
ProQ were estimated to be M1-E130 and V180-F232, respectively (Figure 4.2). These fragments could be overexpressed and purified using a combination of Ni(NTA) affinity, gel exclusion, and reverse phase chromatographies. Purification of ProQ-His6 using RP
HPLC revealed two protein peaks, both containing ProQ-His6 as determined by ESMS and SDS PAGE (Figure 4.3, 4.4 and Table 4.3). It is possible that ProQ does not fully denature in buffers used for RP HPLC, resulting in the presence of a fully unfolded and a partially unfolded protein. It is possible that these different forms of protein interact
142 differently with the reverse phase column due to differences in surface exposure of hydrophobic residues.
Following purification, each domain contained the predicted secondary structure, as determined by CD spectroscopy, because the N-terminal domain (HeN) was shown to be a-helical, whereas the C-terminal domain (HeC) included P-sheet structure. By performing simple solubility screens, QH6, F^N, and HeC were found to be soluble at high concentrations in a second buffer, 0.1 M sodium citrate at pH 2.3. High concentrations of tySf but not QH6 or HeC could be attained in a third buffer, 0.1 M MES at pH 5.6. The impacts of these buffers on the secondary structures of the purified proteins were examined. Each protein was soluble in 0.1 M sodium citrate at pH 2.3. QRe and HeN lost secondary structure in this buffer, but the structure could be restored in the presence of 0.6 M NaCl. However, HeC maintained its structure in this buffer but lost structure in the presence of salt. Previous studies showed that high salt cell lysis buffers improved the yields of ProQ and QH6 by enhancing their solubilities (Smith, et ah 2004).
The secondary structures of QH6 and HeN were maintained in high salt buffers, but that of HeC was perturbed. These data suggest that the native structure is maintained when the
N-terminal domain of ProQ is expressed separately, but the secondary structure of the C- terminal domain is more labile.
HeC was predicted to be SH3-like and hence to possibly mediate interactions of ProQ with other proteins. Although SH3 domains are predominantly P-sheet in structure, their
CD spectra are atypical in having positive ellipticity with a maximum at 220 nm due to the presence of a short 3 io helix (Maxwell and Davidson 1998). The CD spectrum of
H6C had a minimum at 220 nm in the tested buffers, and thus, this work does not indicate
143 that HeC is SH3-like, but it does support the prediction that the C-terminal domain of
ProQ is composed of mostly P-sheet structure.
To identify functionally important regions of ProQ, fragments N, NL, LC, C, L, and
NC were expressed at physiologically relevant levels using plasmids based on pBAD24
(Guzman, et al. 1995). Some of the expressed ProQ fragments were degraded in vivo to
subfragments corresponding in size to those identified as the N- and C-terminal domains
after limited proteolysis in vitro. This further suggested the presence of two ProQ domains.
Functional analysis revealed that the N-terminal domain is important for the amplification of ProP activity. The expression of fragment N, alone, partially restored
ProP activity to the proQ null mutant while the C-terminal fragment (C) did not.
Interestingly, expression of the ProQ fragment lacking the linker (NC) completely complemented the proQ null mutation. In contrast to the other fragments, neither C nor L could be detected within the cells used for these assays by Western blotting. However,
fragments corresponding in size to L and C were produced by cells expressing fragment
LC, and the expression of LC and these fragments did not affect ProP activity. It is, therefore, unlikely that fragment L or C, alone, could amplify ProP activity.
It was hypothesized that expression of fragments N and NL, those that partially complemented a ProQ deficiency, might also compete with ProQ in vivo and hence decrease ProP activity. None of the fragments, when expressed in the proQ+ background, reduced ProP activity. This may indicate that the N-terminal domain contains residues
144 that are important for functional interactions, but these interactions may be weak and require stabilization by another part of ProQ, possibly the C-terminal domain.
Attempts to crystallize the N-terminal domain of ProQ were unsuccessful. It is possible that the boundaries selected for the N-terminal domain are not suitable for crystallization, even though this fragment was able to partially complement a deletion at the proQ locus, as they may result in the presence of unstructured regions which interfere with protein packing within the crystal.
This study shows that ProQ of E. coli is composed of two domains connected by a protease accessible linker. Residues important for ProQ activity may occur within the N- terminal, 130 amino acid domain of the protein. Full complementation of a proQ deletion requires the presence of the N-terminal 130 and the C-terminal 52 residues but not the 50 amino acid linker. When expressed separately, the N-terminal domain appears to retain its secondary structure under conditions similar to those giving structure to the full length protein. Thus, the N-terminal domain or the fused N- and C-terminal domains of ProQ may serve as alternatives to full length ProQ protein for further in vitro studies of interactions between ProP and ProQ.
145 Chapter 5: The Cellular Role of ProQ: Macromolecular Interactions
5.1 Abstract
Previous analysis ofproP expression and of ProP levels suggested that ProQ did not affect the level of transcription or translation of ProP. Thus, ProQ was predicted to amplify ProP activity post-translationally through protein-protein interactions. Data presented in this chapter, however, suggests that ProQ does not act directly on ProP as
ProQ introduced into the lumen of proteoliposomes containing reconstituted ProP did not amplify its activity. Further, expression of a derivative of ProQ fused to the red fluorescent protein (RFP) (ProQ-RFP), from a pBAD24 based plasmid, did not concentrate with ProP at the poles of E. coli cells. Recognition of structural similarities between the N- and C-terminal domains of ProQ and the mRNA binding proteins FinO and Hfq led to a re-examination of the effects of ProQ on ProP levels. Contrary to previous results, it was found that ProQ was able to modulate the expression of ProP.
This modulation appears to occur on the translational level following growth of cells in minimal medium adjusted to various medium osmolalities, as well as in cells grown in rich medium to various growth phases. Regulation of ProP levels by ProQ may involve a regulatory RNA. ProQ was found to co-purify with RNA identified as 16S and 23S rRNAs and tRNA. Data from cellular fractionation techniques showed that ProQ also co- sedimented with ribosomes.
5.2 Introduction
Initial attempts to produce a homology model for the C-terminal domain of ProQ resulted in a model based on the crystal structure of the myosin motor domain from
146 Dictyostelium disodium (1LVK). Circular dichroism experiments performed in the previous chapter do not support this model as the spectrum is typical of P-sheet proteins, but not of SH3 domains. In this chapter, the C-terminal domain is homology modeled on an alternative structure, Hfq, an RNA binding Sm protein (2QTX) (See Introduction 1.6 and 1.7.2.2).
Previous experiments showed that transcription of proP and cellular levels of ProP were not affected by mutations at the proQ locus (Milner and Wood 1989; Kunte, et al.
1999). In these studies, the (3-galactosidase activities of proQ+ and proQ~ bacteria with a transcriptional chromosomalproPr.lacZ fusion were examined following growth at an osmolality of either 0.22 mol/kg or 0.8 mol/kg (Milner and Wood, 1989). Further, the effects of a proQ mutation on ProP levels were determined using Western blot analysis following growth of cells in medium with an osmolality of approximately 0.8 mol/kg
(Kunte, et al. 1999). Since no differences in proP or ProP expression were seen, these observations led to the hypothesis that ProQ affects ProP on the post-translational level, possibly via direct protein-protein interactions.
Here, I report that co-reconstitution of ProP and ProQ into proteoliposomes does not directly affect ProP activity under the conditions tested. Furthermore, cellular localization studies performed using fluorescence microscopy showed that ProP localized to the poles, as expected, while ProQ was evenly distributed throughout the cytoplasm.
Cellular levels of ProP depend on medium osmolality and growth phase and this regulation occurs at the level of transcription. For a detailed description of this regulation refer to Introduction section 1.2.1. Transcription of proP occurs from two, differently
147 regulated promoters. Transcription from promoter PI increases with growth in medium of increased osmolality (Mellies, et al 1995), while transcription from P2 is activated by the combined actions of Fis and RpoS, resulting in a pulse ofproP expression in late exponential to early stationary phase (Xu and Johnson, 1995a). Given this complex control ofproP expression, I considered the possibility that ProQ exerts its effects by acting as a transcriptional or translational regulator, regulating expression ofproP from
PI and P2 differently. Alternatively, ProQ may act on one of the various regulators of proP transcription (such as RpoS or Fis).
Western blotting analysis was used to determine the effects of ProQ on ProP levels.
When cells were grown in media with osmolalities less than 0.7 mol/kg, the levels of
ProP were noticeably lower in ProQ" than in ProQ+ strains. This evoked a change in thinking in that it was no longer believed that ProQ acted on ProP via direct protein- protein interactions with ProP, but rather that ProQ acts as a translational regulator of proP acting independently of rpoS, a transcriptional regulator ofproP.
ProQ is composed of two domains, an N-terminal domain which can be modeled on the mRNA binding protein FinO, and a C-terminal domain which can be modeled on either and SH3-like domain or on an Sm motif (Introduction 1.6). In Chapter 4 (4.3.7),
CD analysis of purified C-terminal domain gave a spectrum that was typical of p-sheet structure, but not of SH3 domains (Chapter 4). These results support the homology model of the C-terminal domain of ProQ derived from the structure of the Sm protein,
Hfq (Introduction 1.6). The structural homology to RNA binding proteins led to the hypothesis that ProQ is an RNA binding protein, acting at the level of translation, to amplify ProP levels. The ability of ProQ to bind RNA was confirmed and co-purifying
148 RNAs were identified as 23 S and 16S rRNA and the valine tRNA. ProQ was also found in ribosome preparations.
The co-localization studies reported in this chapter were performed in collaboration with Divya Viswanathan, an undergraduate research project student, working under my supervision, and Tanya Romantsov, a postdoctoral fellow in the Wood lab. The effects of an rpoS mutation on the proQ phenotype were performed by Patrick Soo, an undergraduate research project student working under my supervision.
5.3 Results
5.3.1 An alternative model for the C-terminal domain of ProQ
A homology model for the C-terminal domain of ProQ based on the crystal structure of the SH3-like domain the myosin motor domain (1LVK) was developed by Dr. R.A.B.
Keates (See Introduction 1.6). Circular dichroism analysis, performed in the previous chapter, of purified C-terminal domain showed a spectrum that was typical of P-sheet structure, but not typical of SH3-like domains (Chapter 4; Smith, et al. 2007). We focused on finding an alternative target structure for modeling of the C-terminal domain.
Hfq from Methanococcus jannischii (2QTX) (Nielsen, et al. 2007) was detected by
GenThreader as an Sm protein that was a good template for modeling of the C-terminal domain. Alignments of the secondary structure predictions for the C-terminal domain of
ProQ with the known secondary structure of Hfq reveal a low level of amino acid identity, however the secondary structures of the proteins can be aligned (Figure 5.1).
Dr. R.A.B Keates therefore produced the homology model seen in Figure 5.2.
149 2QTX(sstr) HHHHHH •• ".:• " HHH
2QTX (seq) - - PNFE YARRLNGKKVKIFLRNGE VLDAE VTGVSNYEIMVKVG- DRNLLVFKHAIDY IE Y
ProQC(seq) HTPVSDISALTVGQALKVKAGQNA-MDATVLEITKDGVRVQLNSGMSLIVR AEHLVF
ProQC (sspred) EEEEEE EEEEEEE EEEEE EEEEEE EEEE —
Figure 5.1: Alignment of the secondary structure (sstr) of Hfq (2QTX) from Methanococcus jannischii and the predicted secondary structure (sspred) of the C-terminal domain of ProQ (ProQC) as determined by Jpred 3 (Cole et al, 2008). Alpha- helical structure (H), P-sheet structure (E) and random coil structure (-) are indicated for each sequence (Dr. R.A.B. Keates, unpublished data). Amino acid sequences of each polypeptide are given (seq).
150 Figure 5.2: Homology model of The C-terminal domain of ProQ (ProQC) on the crystal structure of Hfq from Methanococcus jannischii (2QTX) (Nielsen, et al. 2007). The top panels show the monomeric forms, coloured by secondary structure type, of Hfq (Left) and ProQC (Right). The bottom panels show the hexameric ring structures coloured by subunit. This model was created by Dr. R.A.B Keates by the manual alignment of the secondary structure of Hfq and the predicted secondary structure of ProQC. This alignment was used to model the ProQC on 2QTX using MODELLER (Eswar, et al. 2008) (Dr. R.A.B. Keates, Unpublished Data).
151 5.3.2 ProQ does not affect ProP on a posttranslational level via direct protein- protein interactions
5.3.2.1 Co-reconstitution experiments
The ability of ProQ to act directly on ProP via protein-protein interactions was tested using a system in which the lumen of proteoliposomes containing purified ProP-His6 protein was loaded with either a buffer solution containing purified ProQ-His6 or the buffer solution alone. If ProQ were able to directly act on ProP to amplify its activity, it would be expected that the activity of ProP in ProQ-loaded proteoliposomes, as measured by proline transport assay analysis, would be higher than the activity of ProP in proteoliposomes loaded with only buffer.
ProP-His6 was purified and reconstituted into E. coli polar lipid extract as outlined in
Materials and Methods (2.3.11). Purified ProQ-His6 was loaded into the proteoliposome
lumen via extrusion as outlined in Materials and Methods (2.3.12). Previous reports have
shown that it is possible to load proteoliposomes with various proteins such as hemoglobin, and BSA using extrusion (Gaber, et al. 1983; Chandaroy, et al.200l).
During extrusion, ProQ-His6 was present at a concentration of 0.3 mg/mL. This
concentration is approximately 10-fold higher than the intracellular level estimated by
Western blot analysis following expression of ProQ from plasmid pDC77 (estimated to be 0.025 mg/mL) (Chapter 4; Smith, et al. 2007). ProP(His)6 was present in the proteoliposomes at a concentration of 0.183 mg/mL. In whole cells, it is estimated that
ProQ is expressed at a level of 1.25 ug/mL cell culture following growth in MOPS based minimal medium supplemented with 50 mM NaCl to an OD6oo of 0.8, while ProP is
estimated to be expressed at 26 ug/mL following growth in MOPS based minimal
152 medium supplemented with 300 mM NaCl (based on ProP activity in membrane vesicles). This gives a ratio of approximately 1 ProQ:20 ProP in cells grown at high osmolality. In the proteoliposome system, the ratio of ProQ:ProP was 16:1. Thus,
ProQ(His)6 loading was at a higher level (with respect to ProP levels) in the proteoliposome system. Previous experiments involved titration of proQ expression from pDC77 in an attempt to determine if ProQ overexpression altered the proQ phenotype. It was found that ProQ overexpression did not have an effect in increasing ProP activity
(Smith, et al. 2004; Crane 2004). Loading of the proteoliposome lumen with ProQ-His6 did not alter ProP activity in proteoliposomes (Figure 5.3).
5.3.2.2 Cellular localization studies
It has been shown that ProP concentrates at the poles of E. coli cells in a manner that is dependent on the concentration of the phospholipid cardiolipin (Romantsov, et al.
2007; Romantsov, et al. 2008) (Introduction 1.3). If ProQ were affecting ProP directly, then ProQ may co-localize with ProP at the cell poles.
Plasmids encoding a ProQ-Red Fluorescent Protein (RFP) or RFP alone were constructed (pDV2 and pMS23, respectively) (Materials and Methods 2.2.12). Sequence analysis and restriction digestion analysis of pDV2 confirmed fusion of the proQ ORF to the rfp ORF with a short RV linker between F232 of ProQ and Ml of RFP (data not shown).
The ability of proQ-rfp to complement aproQ deletion was tested to determine if the
RFP protein interfered with ProQ function (Figure 5.4B). ProQ-RFP complemented the proQ mutation, resulting in increased ProP activity, while the plasmid containing the rfp
153 600
100 200 300 400 500 600 700 Assay Osmolality (mmol/kg)
Figure 5.3: Effects of ProQ-His6 on ProP-His6 activity in proteoliposomes. Proteoliposomes containing purified ProP-His6 were prepared as described in Materials and Methods (2.3.11). Proteoliposomes were loaded with ProQ-His6 (0.6 mg/mL in buffer (closed circles)), or buffer alone (open circles) extrusion (Materials and Methods 2.3.12) to give a final ProQ-His6 luminal concentration of 0.3 mg/mL. The proline uptake activity of each preparation was assayed using transport assays as described in Materials and Methods (2.4.1.2) at the indicated assay medium osmolalities.
154 Figure 5.4: Cellular location of ProQ. A) Images obtained from fluorescence microscopy. Strains analyzed were WG914 pDV2 {proQ praP+ProQ-RFP), WG1065 pDV2 (proQproF ProQ-RFP) and WG914 pMS23 {proQ, proP+ RFP). All were grown in MOPS-based minimal medium supplemented with either 50 mM (low osmolality) or 300 mM (high osmolality) NaCl. Cells were visualized by fluorescence microscopy to determine the cellular localization of ProQ. B) The ability of ProQ-RFP to complement chromosomal deletion of proQ was tested. Strains WG914 (chromosome -) and WG210 (chromosome +) were transformed with plasmid pBAD24 (V), pMS23 (RFP), pDC77 (Q), or pDV2 (Q-RFP) and the proline uptake rates of the bacteria were determined. C) Western blot analysis of ProQ and ProQ-RFP expression levels in strains as denoted in A.
155 ProQ ProP Plasmid
Low Osmolality 0.22 mol/kg
High Osmolality 0.7 mol/kg
•r* 14 "c B '5 | 12 8 D) 10
'c E "5 E c B a:to
"S. a> c "o a. Chromosome Plasmid Q-RFP 66 45
31
156 gene alone did not (Figure 5.4B). Thus, ProQ-RFP could be used in place of ProQ during imaging studies to determine the cellular location of ProQ. ProP+ and ProPdeficient bacteria (strains WG914and WG1065) expressing ProQ-RFP or RFP alone were grown at both low (0.22 mol/kg) and high (0.7 mol/kg) osmolalities. In both the presence and absence of ProP, ProQ-RFP was located in the cytoplasm, and not at the poles of the cell,
following growth at both low and high osmolalities (for representative images, see Figure
5.4A). RFP was also cytoplasmic following growth at high and low osmolalities (see
Figure 5.4A for representative images).
Western blot analysis using anti-ProQ antibodies identified a reactive protein with an apparent molecular mass of 54 kDa (molecular mass predicted for ProQ-RFP is 53.4 kDa) with little proteolytic cleavage of ProQ from the RFP protein (very small amount of reactive protein present with a molecular weight intermediate to ProQ and ProQ-RFP),
indicating that the ProQ-RFP protein product was intact and functional in vivo, and that this protein represented the cellular location of untagged ProQ (Figure 5.4C). Thus ProQ did not concentrate with ProP at the poles of the cell.
5.3.3 ProQ does not alter stationary-phase thermotolerance of cells
It is possible that the effect of ProQ on ProP expression results from an effect of ProQ on a regulator of ProP expression, leading to indirect regulation of ProP levels and activity. Transcription of proP is regulated by RpoS, Fis and cAMP-CRP (Landis, et al.
1999; Xu and Johnson 1995a; McLeod, et al. 1999). RpoS is a sigma-factor that is
induced during entry into stationary phase, and, along with the nucleoid associated protein Fis, acts to induce ProP expression upon entry into stationary phase. It is possible
157 that ProQ modulates the activity of RpoS. Modulation of RpoS activity by ProQ could also explain the previous observation that the effects of ProQ are not limited to ProP
(Milner and Wood 1989; Smith, et al. 2004). The stationary-phase thermotolerance of
ProQ+ and ProQ" cells was examined to test the effect of ProQ on RpoS activity.
Stationary-phase thermotolerance depends on trehalose synthesis which is catalyzed by proteins OtsA and OtsB whose expression is RpoS dependent (Hengge-Aronis, et al.
1991). Bacteria that are thermotolerant following growth into stationary phase can survive exposure to 55°C, while bacteria that are not tolerant do not survive this treatment. If ProQ were to regulate RpoS expression or activity, ProQ" strains may have an altered stationary phase thermotolerance.
ProQ" variants of E. coli stains MC4100 (RpoS+) and RH90 {rpoS359::TnlO) were created by introducingproQ220::Tn5 via PI transduction (see Materials and Methods
2.2.13). MC4100 (rpoS?) and itsproQ220::Tn5 derivative were both thermotolerant, while RH90 (rpoS~) and itsproQ220::Tn5 derivative were similarly thermosensitive
(Figure 5.5). Thus, ProQ does not appear to regulate rpoS expression or RpoS activity under the conditions tested.
5.3.4 ProQ does not alter the effects of RpoS on ProP expression and activity
E. coli strainWG584 (Table 2.1) contains IproP loci, each under the control of the PI and P2 promoters of proP. One is at the usual chromosomal location; the other is present within the trp locus (Mellies, et al. 1995; D. Culham and R.T. Voegele, unpublished data). Expression of the wildtype locus yields ProP protein which mediates proline uptake, while the second locus encodes a translational fusion with lacZ encoded
158 1000
100
CO > V. 3 v^ CO ""^--•v ^ :>. 10 - •w • \ V
J , ! , , , , 4 6 10 12 Time (min)
Figure 5.5: Effects of ProQ on the stationary phase thermotolerance of cells. Strains MC4100 (closed circles), MC4100proQ220::Tn5 (open circles), RH90 (rpoS359:: TnlO- closed inverted triangles) and RH90 proQ220::Tn5 (open inverted triangles) were grown overnight in MOPS based minimal medium supplemented with 50 raM NaCl. Cells were harvested by centrifugation and resuspended in 0.85% (w/v) NaCl to an OD60o of 0.7 and incubated at 55°C. Samples of each culture were removed at 1 minute intervals and dilutions plated on LB medium to determine the viable cell count.
159 following the first 190 residues of ProP. This proPr.LacZ fusion was originally constructed by Mellies et al, (1995) and they found that expression of P-galactosidase from this fusion was similar to that seen with a construct in which lacZ was inserted after the stop codon for proP (1614 bp following the transcriptional start site) (Mellies, et al.
1995). Expression of the proP::lacZ fusion along with wildtype proP, allows monitoring of proP expression, through measurements of p-galactosidase activity, and ProP activity, through transport assay experiments, in the same strain, grown under the same conditions.
The proQ locus was deleted from strain WG584 to yield strain WG1119 and mutation rpoS359::TnlO was introduced into both strains to yield strains WG1120 (proQ+, rpoS359::TnlO) and WG1121 (AproQ, rpoS359::TnlO). Measurements of the p- galactosidase activities of these strains showed that proP expression in each proQ derivative was lower than that of the corresponding proQ+ strain when cells were grown in exponential phase (Figure 5.6 Left, Top and Middle). The level of expression of proP seen in stationary phase for strain WG584 (rpoS+ proQ+) was higher than its rpoS~ derivative, as expected (Figure 5.6, Left, Center). Although rpoS mutations significantly reduced proP expression during stationary phase, the ratios of the P-galactosidase activities of proQ+ to proQ strains were similar for exponential and stationary phase bacteria (Figure 5.6 Left, Bottom), indicating that the rpoS defect did not significantly alter the effect of ProQ on proP expression.
The effects of rpoS and proQ defects on ProP activity were also explored. As for proP expression, the proline uptake activities of these strains were impaired by theproQ mutation in a manner that was largely independent of the rpoS mutation (Figure 5.6B).
160 Figure 5.6: Effects ofproQ and rpoS onproP expression and ProP activity. Strains WG584 (proQ+, rpoSf), WG1119 {proQ, rpoS+), WG1121 (proQ+, rpoS~) and WG1120 {proQ, rpoS~) were grown overnight in LB medium. Cells were subcultured into fresh LB medium to an OD600 of 0.1 and grown to an OD600 of 0.8 (for exponential phase, top) or overnight (for stationary phase, middle). Bottom: the ratio of ProQ+/ProQ" in cells expressing or containing mutations at the rpoS locus (RpoS+ or RpoS" respectively) in each culture condition: exponential growth phase (Exp) or stationary phase (Stat). Left: P-galactosidase activities for the/?raP../acZtranslational fusions in each strain were determined as described in Materials and Methods (2.3.17). Right: Following growth into exponential or stationary phase, cells were harvested by centrifugation, washed and resuspended in unsupplemented MOPS buffered medium with an osmolality of 0.42 mol/kg. Proline uptake rates for these strains were determined as outlined in Materials and Methods (2.4.1.1).
161 Exponential Exponential I E i-i "c "E
+ proO + + + 1 rpoS + Stati Stationary IB DC
IB J Q. : Z> V i _j _c "6 i ! &_ a. i i o in. + ...J :. + rpoS 4- + v ProQ+/ProQ- IB ProQ+/ProQ" J* •B ••-» .„X„ Q.
V "5 a. .1 +•» JB a> J5 + + + 0) RpoS RpoS* RpoS RpoS- RpoS* RpoS- RpoS RpoS- Exp Exp Stat Stat Exp Exp Stat Stat
162 There seemed to be a more significant reduction in ProP activity than in proP expression in the rpoS mutant in stationary phase. It is possible that growth of the cells in
LB medium resulted in uptake of osmoprotectants, resulting in an attenuation of the osmotic signals normally activating ProP. Attenuation of ProP activity in the presence of an osmoprotectant could also occur differently in ProQ+ and ProQ" strains resulting in the
1.5 to 2-fold repression of ProP activity in the absence of ProQ (typically this repression is on the order of 3 to 5-fold following growth in minimal medium devoid of osmoprotectants).
5.3.5 Mutations at theproQ locus affect the cellular levels of ProP
Previously, Kunte, et al. (1999) reported that ProQ deficiency did not alter the cellular levels of ProP. They examined the ProP levels in proQ+ and proQ bacteria grown in MOPS minimal medium supplemented with 350 mM NaCl (with an osmolality of about 0.8 mol/kg). When the same strains are grown at lower osmolalities (MOPS minimal medium supplemented with NaCl to achieve osmolalities in the range of 0.12 mol/kg to 0.7 mol/kg) and analyzed with Western immunoblotting, it was found that
ProQ did, in fact, affect ProP expression levels (Figure 5.7).
Further experimentation showed that ProQ also affected the amount of ProP present in cells grown in LB medium as they entered stationary phase. Strains RM2 (proQ+) and
WG174 (proQ220::Tn5) were grown from an OD600 of 0.04 until the reached stationary phase (OD600 = 4.1). Growth was monitored over the 24-h period and revealed no differences between the proQ+ and proQ' strains (Figure 5.8 A). Samples were taken following 1, 2, 3, 4, 5, 6, 7, and 24 h of growth and analyzed by Western blotting using
163 ProQ +-+ +_ + + - + + -++- + ProP + + - + + - + + _ + + _+ + -
k~ L*> *• few i _ ww <- ProP
0.4mol/kg 0.5mol/kg 0.6 mol/kg 0.7mol/kg 0.8 mol/kg
Figure 5.7: Effects of ProQ on ProP expression levels in cells grown at various medium osmolalities. Stains RM2 (proQ+, proP+), WG174 {proQ, proP+), and WG170 (proQ+, proP) were grown in MOPS minimal medium supplemented with 150, 200, 250, 300 or 350 mM NaCl to give medium osmolalities of 0.4, 0.5, 0.6, 0.7 and 0.8 mol/kg respectively as described in Materials and Methods (2.4.1.1). The levels of ProP determined by Western blot analysis as described in Materials and Methods (2.3.3.2). Following growth in minimal medium with an osmolality up to 0.7 mol/kg, there was a visual difference in the levels of ProP in proQ+ and proQ bacteria.
164 / /-
s to Q o o o
0.01
Time (hours) B ProQ + - + - + - + • + - + +
Oh 1 h 2h 3h 4h 5h 6h 7h 24 h Figure 5.8: Effects of ProQ on ProP levels as cells grown in exponential and stationary phases. A) Growth of E. coli strains RM2 (proQ+ closed symbols) and WG174 (proQ220::Tn5 open circles) in LB medium. B) Western blot analysis of ProP levels. Strains RM2 (ProQ (+)) and WG174 (ProQ (-)) were grown in LB medium from an initial OD60o of 0.04. A 10-mL sample was taken from each culture at the indicated time interval following inoculation. Cells were harvested by centrifugation and resuspended in 400 uL of 0.85% Saline. The protein contents of the suspensions were analyzed at each time point and the amount of total protein was normalized to the ProQ+ strain. Ten micro litres of the resulting suspension was loaded onto an SDS PAGE gel and and analyzed by Western blotting for cellular levels of ProP using anti-ProP antibodies as described in Materials and Methods (2.3.3.2). 165 anti ProP antibodies (Figure 5.8 B). The ProQ deficiency reduced the ProP level during late exponential and in stationary phase (after 3 h of growth in LB medium). 5.3.6 ProQ alters ProP levels following plasmid based expression ProQ modulates ProP levels whenproP is expressed from the chromosome (Figures 5.5 and 5.6). The ability of ProQ to modulate ProP levels when proP was expressed from a heterologous promoter was also explored. In plasmid pDC79, proP is expressed from the araBAD promoter of vector pBAD24 (Culham, et al. 2000; Guzman, et al. 1995). Expression of proP can be induced from pDC79 with arabinose. However, the levels of ProP in the absence of arabinose, are similar to the levels of proP expression from the chromosome induced by osmotic stress (Wood, 1988). The levels of ProP expressed from pDC79 were significantly higher in proQ+ bacteria than in their proQ derivatives (Figure 5.9). The ability of ProQ to modulate ProP expression from both its own chromosomal promoters as well as from a heterologous promoter suggests that ProQ is not modulating proP transcription, but suggests that ProQ may be a translational effector ofproP. 5.3.7 ProQ binds RNA 5.3.7.1 ProQ co-purifies with rRNA and tRNA from E. coli The hypothesis that ProQ is an RNA binding protein that modulates proP translation was tested by determining whether ProQ co-purifies with RNA. Previous purification schemes for ProQ-His6 involved overexpressing ProQ-His6, lysing cells in buffer containing high concentrations of NaCl (at least 0.6 M) and adding nucleases to remove nucleic acids (Chapter 3; Smith, et al. 2004). The NaCl in the purification buffers was 166 proP219 proP+ pDC79 pDC79 Q+ Q" Q+ Q" *M» ProQ Figure 5.9: ProQ modulates ProP levels following expression from a heterologous promoter. Plasmid pDC79 was introduced into strains WG170 (containing a chromosomal proP219 mutation) and a derivative WG174 (proQ220::Tn5) and RM2 (proP+) and a derivative WG1072 {proQ756(del)::Kan). Cells were grown as described for transport assay analysis in Materials and Methods (2.4.1.1) and levels of ProP and ProQ within these cells were analyzed using Western blotting with anti-ProP and anti-ProQ antibodies as described in Materials and Methods (2.3.3.2 and 2.3.3.1). 167 presumed to prevent non-specific ionic interactions between the basic ProQ monomers (pi of 9.7) and other negatively charged cellular structures, preventing aggregation of ProQ following cell lysis (Chapter 3; Smith, et al. 2004). Bacteria expressing ProQ-His6 without IPTG induction were lysed and purified in low salt buffers based on 50 mM sodium-phosphate, 300 mM NaCl (pH 8.0) prepared in DEPC treated water. The absorption spectrum of the resulting preparation was compared with that of ProQ-His6 purified in high salt buffers in the presence of nucleases (Figure 5.10B). Under low salt conditions, ProQ-His6 co-purified with a compound that absorbed strongly at 260 nm while this peak was not seen when ProQ-His6 was purified in high salt buffers (Figure 5.10A). DNA and RNA absorb light at 260 nm. To determine if the co-purifying nucleic acid was RNA or DNA, the nucleic acid was isolated using a phenol chloroform extraction, and then treated with either RNaseA (to degrade RNA within the sample), DNasel (to degrade DNA present within the sample) or the corresponding buffer (to control for any endogenous nucleases present in the sample). The treated samples were analyzed by urea acryamide gel electrophoresis as outlined in Materials and Methods (2.2.18) (Figure 5.10C). No material staining with ethidium bromide was detectable in the sample treated with RNaseA, while treatment with DNAsel had no apparent effect when compared with the control sample (Figure 5.10C). Thus, the co-purifying nucleic acid was determined to be RNA. The molecular sizes of the RNAs co-purifying with ProQ-His6 were small (less than 300 bp) and they did not form bands with discrete molecular sizes (Figure 5.10C). Given that the RNA present within this sample originated from a purified protein preparation, it is possible that nucleases present within the cell acted on the RNA molecules present, resulting in degradation of the molecules. 168 Figure 5.10: Characteristics of ProQ-His6 preparations. A: Spectroscopic analysis of purified ProQ-His6. Dotted line: ProQ-His6 was purified under high salt conditions in the presence of nucleases as described (2.3.9.2) (Refer to figure 3.4 for SDS PAGE analysis of a similarly prepared sample). This spectrum is shown with an expanded absorbance in the inset to show a peak at 280 nm. Solid line: ProQ-His6 purified with low salt conditions and no added nuclease (2.2.15) (Refer to panel B for SDS PAGE analysis of this sample). In each case the ProQ-His6 protein was diluted to 0.1 mg/mL in elution buffer for spectroscopic analysis. B) SDS PAGE analysis of ProQ-His6 purified under low salt conditions (2.2.15). L: Cell lysate, P: Cell debris following cell lysis, S: Soluble fraction following cell lysis, F: Flow-through fraction from Ni(NTA) purification, W: Wash fraction from Ni(NTA) purification, E: Elution fraction showing pure ProQ-His6. C) Nuclease sensitivity of the material co-purifying with ProQ-His6. ProQ-His6 protein was removed from the low salt preparation described above using a phenol chloroform extraction (2.2.15). The isolated nucleic acid was treated with 10 mg/mL RNase-free DNasel (D) to reveal the presence of RNA, or 10 mg/mL DNase-free RNaseA (R) to reveal the presence of DNA, in 10 mM Tris-HCl, 2.5 mM MgCt, 0.5 mM CaCh (pH 7.6), or was incubated without treatment (B) for 30 minutes at 37°C. The resulting samples were analyzed by urea acrylamide gel electrophophoresis (2.2.18) followed by staining with ethidium bromide. The nucleic acids present were visualized under a UV light. Numbers to the right indicate the positions of a 100-bp ladder (bp). 169 B Absorbance -0.2 - 0.0 - 0.4 - 0.6 - 0.8 - 0.2 - 1.0 - 45 14 ^S 21 —' 6 97 31 —* 240 I L PSFWE S-- "*v.0.06• 260 2830 Wavelength (nm) 170 \ 0.00 \ -0.02 Absorbance o o QH fi 8 g 0 24262830 l I Wavelength B DR 320 320 34 340 Table 5.1: Identities of RNA species co-purifying with ProQ-His6 Isolate name Aligns to gene' Start End % Coverage2 A3-8 rrsH 144 183 2.6 A5-2 rrsH 141 179 2.5 A4-3 rrsH 144 179 2.3 A3-6 rrsH 143 179 2.4 A4-11 rrsH 917 940 1.6 A3-11 rrsH 803 821 1.2 A4-10 rrsH 795 816 1.4 A3-13 valU 40 65 34 W24 rrlH 2700 2800 6.9 Alignments of isolated cDNA sequences to the identified genes are shown in Appendix 3 0 Coverage is calculated by (# of nucleotides in the recovered fragment)/(# of nucleotides in RNA) 171 To determine the identities of the co-purifying RNAs, cDNA with 5' and 3' adapters was created using reverse-transcriptase PCR as outlined in Materials and Methods (2.2.16). The resulting cDNA was cloned into vector pGEM7z and sequenced. Nine different isolates were identified (Table 5.1). Seven of the isolates were identified as 16S rRNA, one isolate as 23S rRNA and the final isolate as a valine tRNA (Table 5.1). 5.3.7.2 ProQ sediments with ribosomal RNA A previous report showed ProQ to be ribosome associated (Jiang, et al. 2007), and our results indicate that ribosomal RNA co-purified with ProQ (Table5.1). To determine if ProQ would co-purify with ribosomes following growth under conditions where an effect of ProQ on ProP was seen, the ribosomal fraction was prepared for strain RM2 (proP+, proQ+). ProQ was found to be in the ribosome fraction by Western blotting, suggesting that ProQ binds the ribosome (Figure 5.11, Lane R). Interestingly, ProQ was also present in less dense fractions of the cell lysate (Figure 5.11, Lanes U and L). This suggests that ProQ may be present in multiple forms within the cell. 5.3.8 Discussion Previous experiments suggested that the effect of proQ mutations on ProP activity was not due to altered proP transcription or translation (Milner and Wood, 1989; Kunte, et al. 1999). These observations led to the hypothesis that ProQ acted to regulate ProP on a posttranslational level, possibly through direct protein-protein interactions. Evidence for direct effects of ProQ on ProP activity was sought using an in vitro system. ProP- His6 can be purified and reconstituted into proteoliposomes, where it functions as both an osmosenser and osmoregulator (Racher, et al. 1999). It was further shown that the 172 C U L R 97.4 66 45 31 :«- ProQ 21 Figure 5.11: ProQ sediments with the ribosomal fraction. Strain RM2 was grown in LB medium and cells harvested and lysed as described in Materials and Methods (2.3.5). The 70S ribosome and polysome fractions were separated by centrifugation following treatment of the cell lysate with RNase free DNase and centrifugation to remove cell debris as described in (2.3.5). Western blot analysis using anti-proQ antibodies (Materials and Methods 2.3.3.1) revealed the presence of ProQ in the cell lysate (C), the fraction that did not migrate into the sucrose cushion (U), the fraction migrating into but not through the sucrose cushion (L), and the fraction migrating with the ribosome fraction, through the sucrose cushion, to form a pellet at the bottom of the tube(R). 173 maximal activity attained by re-constituted ProP-His6 was lower than that attained in whole cell systems, and this difference was proposed to be due to the lack of ProQ in the lumen of the proteoliposomes (Culham, et al. 2003). When ProQ was introduced into the lumen of proteoliposomes containing purified ProP-His6, amplification of the transport activity of ProP was not seen and thus, the lower activity of ProP-His6 in proteoliposomes did not result from a lack of ProQ. Within this system, ProQ(His)6 was present in a 16:1 ratio with ProP(His)6. Estimates of the cellular levels of this ratio vary from 1:12 at low osmolalities to 1:20 at high osmolalities. Thus, the levels of ProQ in the proteoliposome lumen were higher than what is seen in whole cells. It is interesting that the levels of ProP within cells increase with increasing medium osmolality, while the levels of ProQ do not. Further, the levels of ProP within the cell are much higher than the levels of ProQ. If ProQ were to interact with ProP, it would be expected that the ProQ:ProP ratio within cells would be closer to 1:1. The lack of an effect of ProQ on ProP activity could be due to the presence of histidine tags on both proteins. This is unlikely, however, as ProQ-His6 has been shown to be functional in vivo (Chapter 3; Smith, et al, 2004) and ProQ was shown by Crane (2004) to amplify the activity of ProP-His6 also in vivo. To further test the hypothesis that ProQ affects ProP activity through direct interactions, the cellular location of ProQ was determined using fluorescence microscopy. Addition of a FlAsH tag (CCPGCC) following the initiating methionine of ProP did not perturb the osmosensing or the osmoregulatory activity of ProP (Romantsov, et al. 2007). Use of this tag to visualize ProP by fluorescence microscopy revealed it to be concentrated at the poles of the cell (Romantsov, et al. 2007; Romantsov, et al. 2008). The subcellular location of ProQ was determined to be cytoplasmic following fusion of 174 the RFP to the C-terminus of ProQ, when compared to the cytoplasmic localization of RFP alone (Larrainzar, et al 2005). Western blot analysis revealed that minimal proteolysis of the ProQ-RFP protein occurred and thus the lack of concentrated fluorescence at the cell poles is not due to proteolysis of RFP from the ProQ-RFP protein. Data obtained from complementation analysis showed that the C-terminal RFP did not interfear with ProQ activity in vivo and thus, the cellular localization of ProQ-RFP is likely to be the same as the cellular localization of untagged ProQ. Based on these results, it is unlikely that ProQ directly affects ProP activity through protein-protein interactions. These observations lead us to reexamine the role of ProQ in osmoregulation of E. coli. The homology models of the N- and C-terminal domains were based on RNA binding proteins involved in translational regulation. This lead to the new hypothesis that ProQ acts as an RNA binding protein, regulating the translation proP. ProQ may directly regulate proP expression, or it may regulate expression of a transcriptional regulator ofproP, or the impact of a transcriptional regulator on proP expression. Expression of proP is modulated by the transcriptional regulators Fis, CRP- cAMP, and RpoS (Landis, et al. 1999; Xu and Johnson, 1995a; Hengge-Aronis, et al. 1993). The ability of RpoS to affect expression of proP in proQ+ and proQ cells was examined using aproP::lacZ transcriptional fusion. If ProQ affects the ability of RpoS to mediate proP expression, an rpoS mutation would alter the proQ phenotype. In both exponential and stationary phase cells, the proQ mutation resulted in a 1.3 to 1.5-fold repression of proP expression and a 1.5 to 2.0-fold repression of ProP activity 175 and this effect was independent of the rpoS mutation (Figure 5.6 Bottom). It should be noted that the effects of ProQ on ProP activity seen in these experiments were smaller than what is typically seen (3 to 5-fold). It is possible that when cells are grown in LB medium, in the presence of osmoprotectants, the osmotic response of ProP is attenuated and this occurs differently in the presence and absence of ProQ. This result contradicts earlier conclusions drawn by Milner and Wood (1989) who found no such correlation between the absence of proQ andproP::lacZ expression levels. In that case, expression of the proP::lacZ fusion inproQ+ strains was about 1.3-fold higher than in proQ strains, and thus these measurements were not significantly different due to a higher level of experimental error (Milner and Wood, 1989). Data from Western immunoblotting analysis indicated that ProQ altered ProP protein levels when transcription occurred from either the chromosome or from a plasmid vector under the control of the araBAD promoter, in the absence of 5 'untranslated proP sequences. The effects of ProQ on ProP levels are similar in magnitude to the effects of proQ mutations on ProP activity (approximately 3 to 5-fold). Figure 5.8 B reveals how growth phase modulates ProP expression. Transcription from promoter P2 of proP has been shown to be upregulated by both RpoS and Fis. Fis is a DNA binding protein whose levels change dramatically with growth phase. Fis levels increase rapidly upon culture of E. coli cells into a rich medium and shut off upon entry into exponential growth (Ball, et al. 1992). The large pulse ofproP expression seen following 1 h of growth is most likely due to an increased level of Fis within the cell at this time, acting with RpoS still present within the cell as it enters exponential phase. The levels of Fis and RpoS decrease during exponential growth, resulting in decreased ProP levels (Figure 176 5.8B 2-3 h following subculture). As cells approach stationary phase (4 h following subculture), the Fis levels have been shown to drop significantly (Ball, et al. 1992), but induction of rpoS expression by these cells results in a pulse of proP expression (Figure 5.8B). ProP levels are then seen to decline in stationary phase cells. This decrease is not due to dilution of the protein by cell growth as these cells are in stationary phase. It is likely that this decrease in ProP levels is due to proteolysis of ProP. It is possible that ProQ exerts its effect on ProP by preventing its degradation. It is unlikely, however, that ProQ acts in this manner as the level of ProP seems to decrease at an equal rate in both ProQ+ and ProQ" strains (Figure 5.8). From these experiments, it can be seen that ProQ affects the levels of ProP, and these differences can be seen when proP is expressed in cells grown in minimal or rich medium, or when proP is expressed from a heterologous promoter on a pBAD24 based plasmid. The results suggest that ProQ exerts its effects on proP expression in a promoter-independent manner, possibly on the translational level. It is possible that ProQ modulates ProP levels at the level of protein-mRNA interactions. Regulation of protein expression via protein-RNA interactions typically involves a small, non-coding RNA (sRNA) that is able to act on a target mRNA through base-pairing. These sRNA-mRNA interactions can result in upregulation or down regulation of expression of the target mRNA (Argaman, et al. 2001; Eddy, 2001; Gottesman, 2004; Gottesman, 2006). Given that both the N- and C-terminal domains of ProQ can be modeled on RNA binding proteins, the ability of ProQ-His6 to co-purify with RNA was examined. 177 Purification of ProQ-His6 in low salt conditions and in the absence of nucleases resulted in co-purification of ProQ-His6 and cellular RNA. The sequences identified by this method, 16S rRNA, 23S rRNA and valine tRNA, correspond to RNA which is abundant and stable in the cell due to intricate secondary structures and associations with other cellular molecules. These RNAs may have been recovered with ProQ because of their stability and abundance, rather than because of specific interactions with ProQ itself. This may have resulted in the stable RNAs diluting out less stable mRNA or sRNA targets of ProQ. It is also interesting that 4 of the identified sequences correspond to the same region of the 16S rRNA. It could be that this region of this RNA interacts with ProQ. However, it is also possible that the long poly(G) tract within these RNA fragments acted as a particularly efficient priming site for RT-PCR with poly(C) primers. Co-purification of ProQ with ribosomal RNA could have biological significance. Hfq, with which ProQ is related structurally, also associates with ribosomal RNA (DuBow, et al. 1977). Further, previous experiments by Jiang et al, (2008) showed that ProQ was present in ribosomal fractions. In order to determine if ProQ is ribosome-associated under conditions where the proQ phenotype is observed, E. coli RM2 was grown in LB medium to an OD600 of 2.0, a condition where there was clearly a difference in ProP expression between ProQ+ and ProQ" cells. Western blot analysis revealed the presence of ProQ in the ribosome fraction of these cells. ProQ was also present in the low density fraction, indicating that it exists in a ribosome associated form as well as at least one non- ribosome associated form. The ability of ProQ to bind RNA and to modulate proP expression in a promoter- independent manner suggests the following model for its mode of action. Since the 178 presence of ProQ results in an increase in the amount of ProP, ProQ must act as a positive regulator of proP expression. For expression to be regulated through RNA-RNA interactions, a short, non-coding RNA (sRNA), typically encoding a sequence that is complementary to the target, associates with the target mRNA (Repoila and Darfeuille, 2009; Repoila, et al. 2003). Formation of this sRNA-mRNA duplex typically results in degradation of the mRNA target, leading to a decrease in the amount of protein encoded by the target (Repoila and Darfeuille, 2009; Gottesman, et al. 2006). RNA binding proteins typically act on sRNA molecules, affecting their affinity for their mRNA targets (Gottesman, et al. 2006; Aiba, 2007). The impact of ProQ on ProP expression may occur as follows. ProQ may bind a small regulatory RNA molecule which encodes a sequence complementary to the proP message. This interaction may prevent duplexing of the regulatory RNA and the proP mRNA, resulting in stabilization and increased expression of proP. In the absence of ProQ, the regulatory RNA would be able to interact with the proP message, resulting in its degradation and decreased proP expression (Figure 5.12). Work presented in this chapter has shown that ProQ does not modulate ProP activity through protein-protein interactions, but rather acts to modulate ProP expression on the translational level. A new hypothesis has emerged predicting that ProQ modulates proP levels through ProQ-RNA interactions. These interactions may involve ProQ binding to a regulatory RNA, preventing it from interacting with and degrading the proP message, leading to increased proP translation and higher ProP protein expression levels within the cell. Attempts to identify the regulatory RNA involved in this model resulted in sequences corresponding to the 16S and 23S rRNA as well as to the valine tRNA. 179 Figure 5.12: Model for the role of ProQ in regulation of proP expression. Top: In ProQ+ cells, ProQ (Q) acts to bind a regulatory RNA (regRNA) and prevent it from interacting with the ProP message (proP). This results in stabilization of proP and an increase in the amount of ProP protein within the cell. Bottom: In ProQ" cells, the regulatory RNA is able to interact with the proP message, resulting in degradation of proP mRNA. Degradation results in a decrease in the amount ofproP translation and a decrease in the amount of ProP protein expressed by the cell. 180 Chapter 6: Suppression of the proQ Phenotype 6.1 Abstract ProQ is proposed to regulate ProP expression at the level of translation by preventing interactions between an unidentified regulatory RNA and proP mRNA (Chapter 5). If ProQ acts in this manner, deletion of the regulatory RNA should result in suppression of the proQ phenotype. A strain was identified which no longer showed the proQ phenotype as ProP activity and expression levels for proQ strains were as high as that of the normal proQ+ strains. It was determined that mutations deleting part or all of the proU operon resulted in partial suppression of the proQ phenotype, with proW and proX deletions suppressing the proQ phenotype under every condition tested. A pro V deletion showed partial suppression of the proQ phenotype following expression of proP from the chromosome, however this mutation suppressed fully when proP was expressed from plasmid pDC79. A point mutation mproV, which inactivated ProU, did not suppress the proQ phenotype under any condition tested, while insertion of IS5 within pro V suppressed the proQ phenotype when ProP was expressed from plasmid pDC79, but not when ProP was expressed from the chromosome. In each strain showing suppression of the proQ phenotype, the level of ProP remained high in the presence or absence of ProQ, indicating that regulation by proU and proQ of ProP expression occur at the same level. It is proposed that the proU operon either encodes or regulates the expression of a regulatory RNA capable of duplexing with proP mRNA, resulting in degradation and 181 decreased levels of ProP expression. In this model, ProQ acts to prevent duplexing, resulting in increased persistence of the proP message, and increased levels of ProP. 6.2 Introduction Data presented in Chapter 5 show that ProQ modulates the expression of ProP, and that this regulation is likely to occur on the translational level, possibly through interactions with a regulatory RNA molecule. Attempts to isolate the putative regulatory RNA through co-purification experiments were unsuccessful, however, I was able to show that ProQ co-purified with RNA (Chapter 5). If ProQ acts to prevent an interaction between a regulatory RNA and the proP mRNA, it should be possible to suppress the proQ phenotype through deletion of this regulatory RNA. This chapter describes the identification of E. coli strains in which the proQ phenotype is suppressed. These strains differed from those showing a proQ phenotype in that they contained large, uncharacterized chromosomal deletions in regions surrounding and including the proP gene and the proll operon. These strains also contained a deletion of the putPA genes (AputPAlOl), but this mutation was also present in strains which showed the proQ phenotype and thus was not a candidate for further investigation. Precise, characterized mutations of the proP and pro U loci were created here in aproQ+ and proQ background and assays for the proQ phenotype performed. It was determined that deletion of the proU locus was responsible for suppression of the proQ phenotype. ProU, like ProP, is an osmoregulatory transporter that both senses increased osmolality and responds by transporting osmoprotectants into the cell (Faatz, et al. 1988). 182 ProP and ProU share a similar set of substrates, but differ in their affinity for these substrates (Wood, 1999). ProU is an ABC transporter composed of three subunits: ProV, the ATP binding subunit; ProW, the integral membrane protein; and ProX, the periplasmic osmoprotectant binding protein (May, et al. 1989; Gowrishankar, 1989; Stirling, et al. 1989). The pro t/operon encodes all three of these polypeptides (proVWX) under the control of two promoters, PI and P2 (Dattananda, et al. 1991; Gowrishankar and Manna, 1996). A discussion of the regulation of proU transcription can be found in section 1.2.2 of the Introduction. Previous reports have indicated that inactivation of ProU stimulates ProP activity as ProU no longer mediates osmoprotectant uptake, preventing attenuation of the osmotic signals that activate ProP (Sutherland, et al. 1986; Cairney, et al. 1985a). This work shows that mutations disrupting the coding sequence of pro C/suppress the proQ phenotype while mutations that inactivate ProU do not. Chad Gill, an undergraduate student, working with me, discovered the suppression of the proQ phenotype through transport assay analysis of WG709 and WG997 and characterized the phenotype in these strains. 6.3 Results 6.3.1 Bacterial strains and mutations Initial characterization of the proQ phenotype and its suppression were carried out in plasmid pDC79 bearing derivatives of strains WG350, WG997, WG170 and WG1074 (Table 6.1). Following this initial characterization, it became apparent that suppression of the proQ phenotype arose from mutations at the proU locus. Characterization of the 183 ability ofproU mutations to affect the proQ phenotype was carried out in E. coli strains RM2 and WG1072 (Table 6.1). A number ofproP, proQ and proU mutations are described in this chapter and their characteristics, as well as their effects on the proQ phenotype, are summarized in Table 6.2. 6.3.2 The proQ phenotype can be suppressed The proQ phenotype can be observed by measuring the ProP activities ofproQ+ and proQ bacteria. AproQ defect decreases the activity of ProP at each osmolality tested (Figure 6.1 A). To facilitate analysis of the impact of ProQ on ProP structural variants, mutation AproQ214 was introduced to strain WG350 (A(putPA)lOl, A(proP-melAB)212, AproU600) to create strain WG997. E. coli strain WG350 is used as a host for plasmid based expression and analysis of proline transporters and their variants. It is devoid of proline uptake activity due to spontaneous deletions of proline transporter-encoding loci putP, proP and pro £/(Culham, et ah 1993). These deletions were selected before allelic exchange technologies become widely available and have uncharacterized end points. ProP is routinely expressed in WG350 from plasmid pDC79, a derivative of vector pBAD24 in which proP expression is controlled by the araBAD promoter (Culham, et ah 2000). When the transport activities of strains WG350_pDC79 and WG997_pDC79 were measured, it was found that the proQ phenotype was suppressed (Figure 6.1, Right). Plasmid based expression of ProP did not cause the suppression as it was evident when the ProP activities of strains WG170 (A(putPA)lOh, proP219, prolf) and WG1074, its proQ220::Tn5 derivative were compared (Figure 6.1, Left). Comparison of the 184 Table 6.1: Genotypes of strains used during identification of the proQ suppressor Strain Name putPA proP2 prolf proQ2 RM2 A(putPA)101 proP prolf proQ+ WG170 A(putPA)101 proP219 prolf proQ+ WG350 A(putPA)101 A(proP- A(proU)600 proQ+ melAB)212 WG997 A(putPA)101 A(proP- A(proU)600 AproQ214 melAB)212 WG1067 A(putPA)101 AproP771::kan A(proU)600 proQ+ WG1069 A(putPA)101 AproP771::kan A(proU)600 AproQ214 WG1070 A(putPA)101 AproP771::kan prolf" proQ+ WG1071 A(putPA)101 AproP771::kan prolf" AproQ856 WG1072 A(putPA)101 proP prolf" AproQ756::kan WG1074 A(putPA)101 proP219 prolf proQ220::Tn5 WG210 A(putPA)101 proP+ proU205 proQ+ WG914 A(putPA)101 proP+ proU205 AproQ676 WG1078 A(putPA)101 proP+ AproV858::FRT proQ+ WG1080 A(putPA)101 proP+ AproV858::FRT AproQ756::kan WG1083 A(putPA)101 proP+ AproW859::FRT proQ+ WG1085 A(putPA)101 proP+ AproW859::FRT AproQ756::kan WG1088 A(putPA)101 proP+ AproX870::FRT proQ+ WG1090 A(putPA)101 proP+ AproX870::FRT AproQ756::kan WG1197 A(putPA)101 proP+ proV677 proQ+ WG1198 A(putPA)101 proP+ proV677 AproQ756::kan WG1201 A(putPA)101 proP A(proV- proQ+ proX)2098::FRT WG1202 A(putPA)101 proP"" A(proV- AproQ756::kan proX)2098::FRT All strains are derived from RM2 Descriptions of alleles are presented in table 6.2 185 Table 6.2:Effects of mutations on the proQ phenotype Presence of the proQ phenotype Mutation Description LQM' HOM2 pDC793 proP219 Point mutation that 1W ND + truncates ProP after A408 A(proP-melAB)212 Deletion of the proP ORF ND ND + as well as an unknown number of upstream and downstream genes AproP771::kan Replacement of the proP ND ND + ORF by a gene conferring kanamycin resistance AproQ756::kan Replacement of the proQ + + + ORF by a gene conferring kanamycin resistance proQ220::Tn5 Tn5 insertion in proQ at the + ND + codonfor El 05 AproQ856::FRT In-frame deletion of proQ + ND + AproQ676 In frame deletion of proQ + ND ND A(proU)600 Uncharacterized deletion of ND ND - the/?ro£/operon proU205 Is5 insertion in pro V + + - A(proV- Deletion of the proU operon - - - proX)2098::FRT from the PI promoter to the termination codon of proX proV677 Encodes ProVE190Q + + + AproV858::FRT In-frame deletion ofproV -/+5 -/+ - AproW859::FRT In-frame deletion oiproW - - - AproX870::FRT In-frame deletion oiproX - - - ProP activity was measured in the presence and absence of chromosomal proQ following growth in MOPS minimal medium supplemented with 50 mM NaCl. ProP activity was measured in the presence and absence of chromosomal proQ following growth in MOPS minimal medium supplemented with 250 mM NaCl. 3ProP activity was measured in the presence and absence of chromosomal proQ following growth in MOPS minimal medium supplemented with 50 mM NaCl and with expression of proP from plasmid pDC79. 4ND indicates that these conditions were not tested for this allele -/+ indicates a partial proQ phenotype 186 genotypes of strains WG350 and WG170 revealed mutations that may be responsible for the suppression of the ProQ phenotype. Specifically, the genetic differences between these strains are: i) at the proP locus, A(proP-melAB)212 mutation in strains WG350 and WG997 vs. theproP2\9 mutation in strains WG170 and WG1074 ii) at theproQ locus, AproQ214 mutation in stain WG997 vs. theproQ220::Tn5 mutation in strain WG1074 and iii) at the proU locus, AproU600 mutation in strains WG350 and WG997 vs. wildtype proU in strains WG170 and WG1074 (Table 6.1). 6.3.3 Mutation A(proP-melAB)212 does not suppress the proQ phenotype Mutation A(proP-melAB)2\2 is a chromosomal deletion that includes proP and an unknown number of flanking ORFs, extending at least to the melAB genes (Figure 6.2 and Table 6.3). To test the hypothesis that deletion of any open reading frame (ORF) flanking proP in strain WG350 results in suppression of theproQ phenotype, ORFs flanking proP were restored in both WG350 and its proQ derivative using PI transduction of the AproP771::kan replacement from the Keio collection (Baba, et al. 2006) as described in Materials and Methods (2.2.13), yielding strains WG1067 and WG1069. As a control, the AproP77l::kan replacement was also introduced into strains RM2_pDC79 and its AproQ856 derivative to yield strains WG1070 and WG1071. If deletion of an ORF flanking proP resulted in suppression of the proQ phenotype, then the proQ phenotype should be observed when it is restored with AproP771::kan. Restoration of ORFs surrounding proP did not alter ProP activity in WG1067_pDC79 (WG350 187 200 300 400 Assay Osmolality (mmol/kgj Assay Osmolality fmmol/kg! Figure 6.1: The proQ phenotype can be suppressed. Left panel: ProP activity in WG170 pDC79 (proP219, A(putPA)101, prolf) (filled circles) and itsproQ220::Tn5 derivative (WG1074 pDC79) (open circles). Right panel: ProP activity in WG350 pDC79 (A(proP-melAB)212, A(putPA)101, AproU600) (filled circles) and WG350 AproQ214 pDC79 (open circles). In both cases, cells were grown in MOPS minimal medium supplemented with 50 mM NaCl, ProP activity was determined at the indicated assay medium osmolalities as described in Materials and Methods (2.4.1.1). 188 AproP771::kan) and WG1069_pDC79 (WG1067 AproQ214) (Figure 6.3). Therefore deletion of an ORF flanking proP did not suppress XheproQ phenotype. Further, mutation of the proP ORF itself did not suppress the proQ phenotype as ProQ amplifies ProP activity in strain WG1070_pDC79 (RM2 AproP771::kan) when compared with WG1071_pDC79 (WG1070 AproQ856::FRT) (Figure 6.3). 6.3.4 Deletion of loci surrounding proP does not affect ProP activity Given the large number of genes with unknown function directly upstream of proP (Figure 6.2 and Table 6.3), the impacts of mutations in these genes on ProP activity and the proQ phenotype were tested individually. PI transduction was used to replace the phnB, phnA, yjdA, yjcZ, proP, basS, basR and eptA ORFs with a kanamycin resistance determinant using Keio collection strains as donors and strains WG914 and WG210 as recipients as described in Materials and Methods (2.2.13). Strains WG210 (RM2 proU205) and WG914 (WG210 AproQ676) are used extensively to characterize the proQ phenotype as they lack ProU. The ProP activity of each strain was determined and it was found that kanamycin replacements of loci surrounding proP neither increased nor decreased ProP activity, nor did they suppress the proQ phenotype (Figure 6.4). A kanamycin replacement of proP abolished proline uptake activity in both proQ+ and proQ strains as expected (Figure 6.4). 6.3.5 Suppression of the proQ phenotype is independent of proQ mutant class Five different pro Q mutations were tested to determine if the type of proQ mutation affected the proQ phenotype; proQ220::Tn5, AproQ676, AproQ214, AproQ756::kan, 189 c J. 4,312,367 L phnCDEFGHIJKLMNOP yjdA ^| yjcZ > [ proP >< tarf j<^7|< ep^ K adiC ^adiy\^ adiA (meiR melA / melB ^ 4,342,813 Figure 6.2: Chromosomal loci surrounding proP in E. coli from 4,312,367 bp to 4,342,813 bp. Function of the protein encoded by each ORF is listed in Table 6.3. 190 Table 6.3: Functions of the products of genes surrounding the proP locus in E. coli Gene(s) Function Map position Reference PhnC-P Phosphonate metabolism 4,312,367-4,323,188 (Metcalf and Wanner 1993) YjdN Conserved protein with 4,323,764-4,323,321 (Riley, et al. 2006) unknown function YjdM Conserved protein with 4,324,757-4,324,422 (Riley, et al. 2006) unknown function YjdA Conserved protein with 4,325,158-4,327,386 (Riley, et al. 2006) unknown function YjcZ Conserved protein with 4,327,383-4,328,261 (Riley, et al. 2006) unknown function ProP Osmoregulatory transporter 4,328,525-4,220,027 (Stalmach, et al. 1983) BasS Sensor kinase of two 4,331,295-4,330,204 (Hagiwara, et al.2004) component regulatory system involved in iron regulation BasR Response regulator of two 4,331,973-4,331,305 (Hagiwara, et al. 2004) component regulatory system involved in iron regulation EptA Predicted metal dependent 4,333,613-4,331,970 (Stenberg, et al. 2005) hydrolase AdiC Arginine:agmatine 4,335,054-4,333,717 (Gong, et al.2003) antiporter AdiY Transcriptional activator of 4,335,952-4,335,191 (Stim-Herndon, et the adi genes al. 1996) AdiA Subunit of arginine 4,338,547-4,336,277 (Stim and Bennett 1993) decarboxylase MelR Melibiose transcriptional 4,339,651-4,338,743 (Caswell, et al. 1992) activator MelA Melibiase (a-galactosidase) 4,339,934-4,341,289 (Burstein and Kepes 1971) MelB Melibiose cation transproter 4,341,404-4,342,813 (Hanatani, et al. 1984) Map position on the E. coli K-12 genome (Keseler, et al. 2005) 191 ^~~ 100 -j 'c £ "i—o Q. — 80 - 8 O) E 'c 60 - E "o E ^ 40 - B "ai_s Figure 6.3: Effect of mutation A(proP-melAB)212 on the proQ phenotype. ProQ+ (+) and ProQ" (-) derivatives of strains WG350 pDC79 (A(proP-melAB)212, AproWOO), WG1067 pDC79 (proP::Km, AproU600) and WG1070 pDC79 {proPr.Km, prolf) were grown in MOPS based minimal medium supplemented with 50 mM NaCl and ProP activity was measured at an osmolality of 0.49 mol/kg, as described in Materials and Methods (2.4.1.1). 192 4- 14 'c X X 2 12 Q. 1 O) 10 cE 8 - E "o E B 6H "TO 0) CD Q. 2 - I 1 C 1> Q. L - 1 r T ^ h-^ i 1 1 T T 1 i i 1 1 1 i Km Replacement None yjdN yjdM yjdA yjcZ proP basS basR eptA ProQ + Figure 6.4: Effects of kanamycin replacements of ORFs surrounding proP on ProP activity. Kanamycin replacements (from the Keio collection) were introduced to strains WG210 (proP+,proQ\proU205) (ProQ+) and WG914 (WG210 AproQ676) (ProQ). The resulting strains contained kanamycin replacements of the yjdN, yjdN, yjdA, yjcZ, proP, bass, basR or eptA loci. These strains were grown in MOPS minimal medium supplemented with 50 mM NaCl. ProP activity was determined using transport assay analysis at an assay medium osmolality of 0.49 mol/kg as described in Materials and Methods (2.4.1.1). 193 and AproQ856 ::FRT (Table 6.2). In order to test the effects of various proQ mutant classes on suppression of the proQ phenotype, mutations proQ220::Tn5, AproQ756::kan and AproQ856 were put into the RM2 background and the activities of ProP in the proQ and proQ+ strains compared (Figure 6.5). The proQ phenotype was found to be independent of the proQ mutant class. 6.3.6 Mutations at the proU locus suppress the proQ phenotype Experiments above suggested that the A(proU)600 mutation suppressed the proQ phenotype. To further characterize this suppression, an in-frame chromosomal deletion of the/?rot/operon was created. This mutation involved deletion of the entire proU operon, as well as its 5'untranslated region (UTR), spanning from the upstream PI promoter of proU to the coding sequence for the termination codon of proX. Initially a kanamycin replacement of the entire proU operon (A(proV-proX)2098::kan) was introduced into strain RM2 using the X-RED recombinase system (Materials and Methods section 2.2.14). This kanamycin replacement was then converted to an in-frame deletion (A(proV-proX)2098::FRT) using the FLP recombinase (Materials and Methods 2.2.14). DNA from the resulting strain (WG1201), was used as PCR template with primers ProVupoutl and ProX_out_down_2 (Table 2.2) to confirm loss of the proU operon. Clones showing loss of the proU operon were confirmed by DNA sequencing (data not shown). The resulting RM2 AproU strain was converted to a AproQ756::kan derivative through PI transduction to yield strain WG1202, as described in Materials and Methods (2.2.13). 194 it - 'c £ -r- o 12 - a. el l o U) 10 - E 'c 'E 8 - "o E c 6 - *—•* £ CD T T -r 0 4 - XL CD -*-Q^. D CD 2 - C "o u. a 0 - I 1 1 ' ' i ' " 1 RM2 RM2 RM2 RM2 RM2 proQ* prvQ756::kan proQ220::Tn5 hpmQ514 &pmQ676 Figure 6.5: Effect oiproQ mutant class on the proQ phenotype. E. coli RM2 derivatives containing intact chromosomal proQ (proQ+), or various chromosomal proQ mutations, as defined in Table 6.2, were grown in MOPS minimal medium supplemented with 50 mM NaCl. Proline uptake by each strain was measured at an assay osmolality of 0.49 mol/kg as described in Materials and Methods (2.4.1.1). 195 The activities of ProP and ProU in strain RJVI2 and the A(proV-proX)2098 derivative were measured, showing that WG1201 (RM2 A(proV-proX)2098) had a much lower glycine betaine uptake activity than RJVI2 (Figure 6.6). Background levels of glycinebetaine uptake retained in this mutant are attributable to chromosomally encoded ProP. Western blot analysis using anti-ProX antibodies did not detect expression of ProX, confirming the deletion of the proU operon (Figure 6.6). The effects of the A(proV-proX)2098 mutation on ProP activity were determined in strains WG1201 (RM2 A(proV-proX)2098) and WG1202 (WG1201 AproQ756::kan). The prog phenotype was absent from these strains, as their ProP activities were similarly high, but the proQ phenotype was present in strains RM2 (prolf proQ+) and WG1072 (KMlprolf AproQ756::kan) (Figure 6.7). 6.3.7 Deletion oiproU suppresses the proQ phenotype ProU and ProP are osmoregulatory transporters acting on similar substrates. It has been suggested that the activity of ProP or ProU will modulate the activity or expression of the other through signal attenuation (Sutherland, et al. 1986; Cairney, et al. 1985a). To determine if the transport activity of ProU or the coding sequence of the proU locus regulates proP, a point mutation was introduced to eliminate ATP hydrolysis by the ProV subunit of ProU, thereby eliminating ProU activity while maintaining an intact pro U locus. Mutation ProV E190Q (proV677) was chosen by comparing the amino acid sequence of ProV with those of the ATP binding cassettes of other ABC transporters. 196 B S<~ProX Figure 6.6: Effects ofproUmutations on ProU activity. E. coli RM2 and derivatives either encoding an intact pro V, pro W or proXORF (+), an in-frame deletion of the indicated ORF (-), or alleles pro V677 (E190Q) or proU205 (IS5) were grown in MOPS based minimal medium supplemented with 300 mM NaCl to induce expression of the proU operon. ProU activity was assayed at an assay osmolality of 0.75 mol/kg using transport assay analysis with 10 uM, 5Ci/mol [14C] glycine betaine as the substrate as described in results. 197 o CL 0 proV + + E190QE190Q IS5 IS5 - - + + + + proW + + + + + + + + - - + + proX + + + + + + + + + + - - proQ + - + - + - + + - + _ + _ Figure 6.7: Effects ofproU mutations on the proQ phenotype. ProQ+ (+) or ProQ" (-) variants of E. coli RM2 and derivatives either encoding an intact proV, proWorproX ORF (+), an in-frame deletion of the indicated ORF (-), or alleles pro V677 (E190Q) or proU205 (IS5), were grown in MOPS based minimal medium supplemented with 50 mM NaCl. ProP activity was determined at an assay medium osmolality of 0.49 mol/kg using proline uptake assays as described in Materials and Methods (2.4.1.1). 198 Previous research has shown that mutation of the Glu residue immediately following the Walker B motif of the ATP binding cassette protein of MJ0769 from a Methanococcus jannaschii ABC transporter abolished transport activity as the modified transporter can no longer bind a water molecule playing a key role in ATP hydrolysis (Moody, et al. 2002; Smith, et al. 2002). The 3D PSSM server was used to align the secondary structure prediction of Pro V with the known secondary structure of MalK, the ATP binding protein of the maltose ABC transporter in E. coli (Figure 6.8). Using this approach, key catalytic residues in ProV could be identified (Figure 6.8). The residue E190 of ProV was identified as a putative catalytic residue involved in ATP hydrolysis. Introduction of an E190Q mutation within ProV should result in disruption of ProU activity while introducing only 2 nucleotide changes. The ProV E190Q (proV677) mutation was introduced to the chromosome of strain RM2 by double homologous recombination resulting in strain WG1197 (RM2 proV677). A point mutation, resulting in replacement of a glutamic acid (E) residue at position 190 with a glutamine (E190Q), was introduced toproVby site-directed mutagenesis using the Quick-change method. Introduction of the ProV E190Q mutation (proV677) into RM2 abolished ProU activity (Figure 6.6). Western blot analysis revealed that the proU operon was still expressed, as the levels of ProX within this stain were similar to those observed in pro if bacteria (Figure 6.6). Allele AproQ756::kan was introduced into strain WG1197 using PI transduction to yield strain WG1198. The activity of ProP was measured in both of these strains, and it was found that the proQ phenotype was present (Figure 199 Walker A sspred EEEEE-EEEEE HHHHHHHHH—EHHHEEEE EEEEE-EEEEE EEEEE HHHHHHHHH ProV MAIKLEIKNLYKIFGEHPQRAFKYIEQGLSKEQILEKTGLSLGVKDASLAIEEGEIFVIMGLSGSGKSTMVRLLNRLIEPTRG MalK MAGVRLVDVWKVFGEVTAVREMSLEVKDGEFMILLGPSGCGKTTTLRMIAGLEEPSRG SStru —EEEEEEEEEEE—EEEE EEE EEEEE HHHHHHHHH Q-loop sspred EEEE E HHHH-HHHH EEEEE-E HHHH HHH HHHHHHHHHHHHHH HHHH HH — ProV QVLIDGVDIAKISDAELREVRRKKIAMVFQSFALMPHMTVLDNTAFGMELAGINAEERREKALDALRQVGLENYAHSYPDELS MalK QIYIGDKLVADPEKGIFVPPKDRDIAMVFQSYALYPHMTVYDNIAFPLKLRKVPRQEIDQRVREVAELLGLTELLNRKPRELS SStru EEEE—EEE HHH—EEEEEE HHHHH HHH HHHHHHHHHHHHHH HHHHH LSGGQ motif Walker B HD motif sspred —HHHHHHHHHHHH EEE--p=-HHHHHHHHHHHHHHHHHHHHHH EEEEEE—HHHHHHH EEEE EEEEE—H ProV GGMRQRVGLARALAINPDILLMlllAFSALDPLIRTEMQDELVKLQAKHQRTIVFISHDLDEAMRIGDRIAIMQNGEVVQVGTP MalK GGQRQRVALGRAIVRKPQVFLMIIiPLSNLDAKLRVRMRAELKKLQRQLGVTTIYVTHDQVEAMTMGDRIAVMNRGVLQQVGSP SStru HHHHHHHHHHHHHH EEEEEF-HHH--HHHHHHHHHHHHHHHHH EEEEEEE-HHHHHHH—EEEEEE—EEEEE—H sspred HHHH HHHHHH EEE EEE EEEE HHHHHHHH EEEEE ProV DEILNNPANDYVRTFFRGV--DISQVFSAKDIARRTPNGLIRKTP GFGPRSALKLLQDEDREYGYVI ERG MalK DEVYDKPANTFVAGFIGSPPMNFLDAIVTEDGFVDFGEFRLKLLPDQFEVLGELGYVGREVIFGIRPEDLYDAMFAQVRVPGE SStru HHHHH HHHHHH EEEEEEEEE—EEEE--EEEEE EEEEEEE—EEEEE sspred —EEEEEEHHHHHHH EEEEEEE--EEEEE HHHHHHHH EEEE EEEEEEEHHHHHHHH ProV NKFVGAVSIDSLKTALTQQQGLDAALIDAPLAVDAQTPLSELLSHVGQAPCAVPVVDEDQQYVGIISKGMLLRALDREGV-NNG MalK NLVRAWEIVENLGSERIVRLRVG GVTFVGS FRSESRVR-EGVEVDVVFDMKKIHIFDKTTGKAIF SStru -EEEEEEEEEEEE--EEEEEEEE- -EEEEEE E EEEEEEEHHH-EEEEHHHH Figure 6.8: Alignment of the protein sequences of E. coli ProV and MalK produced by 3D PSSM (Kelly, et al. 2000). The known secondary structure of MalK (sstru) is aligned with the predicted secondary structure of ProV (sspred). Helical secondary structure is indicated with an H, P-strand secondary structure is indicated with an E and coil structure is indicated with a -. Catalytic residues in MalK and their corresponding residues in ProV are bolded. The catalytic glutamate (E) residue (boxed), thought to act in hydrolysis of ATP to ADP is located immediately after the Walker B motif (Moody, et al. 2002; Smith, et al. 2002). 200 6.7),indicating that the proU locus and not ProU activity is important for the proQ phenotype. 6.3.8 Specific sequences in the proU operon are important for the proQ phenotype To determine whether removal of one of the ORFs within proU would suppress the proQ phenotype, in-frame deletions of pro V, proW and proX were created as described in to give strains WG1078 (RM2 AproV858::FRT), WG1083 (RM2 AproW859::FRT) and WG1088 (RM2 AproX870::FRT). ProU activity could not be detected in these strains, with background glycine betaine uptake activity attributable to chromosomally encoded ProP (Figure 6.6). Western blot analysis revealed that deletion of the pro For pro WOKF, in strains WG1078 and WG1083 respectively, did not affect expression of proX, as the ProX levels seen in these strains are similar to those observed inprolf bacteria (Figure 6.6). Deletion of proX in WG1088 abolished expression of the ProX protein, as expected (Figure 6.6). A AproQ756::kari derivative of each strain was created (Table 6.1). The ProP activities of the resulting strains were measured and it was determined that the proQ phenotype was suppressed when pro W or proX were deleted. The proQ phenotype was partially suppressed when pro V was deleted (Figure 6.7). 6.3.9 The proQ phenotype is suppressed by proU deletions under conditions of high osmolality and when proP is expressed from pDC79 Two other conditions were tested to further explore conditions yielding suppression of the proQ phenotype. These growth conditions involved titrating the levels of proP and proU expression with respect to each other. First, expression of proP and proU from the 201 chromosome were both induced by growing cells in MOPS buffered minimal medium supplemented with 250 mM NaCl, or high osmolality medium (HOM). Second, proP was expressed from plasmid pDC79 during growth in MOPS buffered minimal medium supplemented with 50 mM NaCl, or low osmolality medium (LOM) (with an osmolality of 0.25 mol/kg). Based on ProP activity measurements in strain RM2, grown in either LOM or HOM, and those for RM2 pDC79 grown in LOM, the level of uninduced ProP expression from plasmid pDC79 is approximately 2.5-fold higher than the level of ProP expressed from the chromosome in HOM and approximately 6.6-fold higher than the level expressed from the chromosome in LOM (Figures 6.7, 6.9 and 6.10). Thus, growth in LOM with pDC79 would result in induction of proP expression (6.6-fold higher than what would normally be experienced at that osmolality), while the levels of proU would remain low. E. coli strain RM2 and derivatives with various proU mutations with or without AproQ756::kan (Table 6.2) were grown in high osmolality medium (HOM) and the resulting ProP activity measured. Alleles A(proV-proX)2098::FRT, AproW859::FRT, and AproX870::FRTsuppressed the proQ phenotype, while allelesprolf andproV677 did not (Figure 6.9). Allele AproV858::FRT, did not fully suppress theproQ phenotype, but it did slightly decrease ProP activity in the absence of chromosomal proQ following growth in high osmolality medium (Figure 6.9). E. coli strain RM2 and derivatives, bearing plasmid pDC79, with various pro U mutations with or without AproQ756::kan (Table 6.2) were grown in low osmolality medium (LOM) and the resulting ProP activity measured. Alleles A(proV- 202 proX)2098::FRT, AproW859::FRT, AproX870::FRT, and AproV858::FRT suppressed the proQ phenotype, while alleles pro if andproV677 did not (Figure 6.10). The expression level of ProP in each strain following growth in HOM and following growth in LOM with proP expression from pDC79 was analyzed using Western blotting (Figure 6.11). In cases where the proQ phenotype is seen, the levels of ProP expression in the ProQ+ strain are clearly elevated with respect to their ProQ" derivative. On the other hand, suppression of the proQ phenotype also seems to suppress the requirement for ProQ to elevate ProP levels. 6.3.10 Mutation proU205 partially suppresses thzproQ phenotype Analysis of the proQ phenotype in strain WG210 (proU205) and its AproQ676 derivative WG914, revealed that mutation proU205 suppresses the proQ phenotype only when proP is expressed from plasmid pDC79, but not when proP is expressed from the chromosome during growth at low or high medium osmolality (Figures 6.7, 6.9 and 6.10). The nature of the proU205 mutation was therefore determined. The pro U ORFs in strains WG210 and RM2 were compared by PCR in order to characterize theproU205 mutation. Primers proVI and proX2 were used to amplify a region of the proUORF outlined in Figure 6.12A. The PCR product obtained with DNA from WG210 as a template was about 1 kb larger than the expected 3 kb product seen when DNA from strain RM2 was used as template (Figure 6.12B). The amplicon was purified as described (Materials and Methods 2.2.7) and subjected to restriction endonuclease digestion with Pvul, Ndel and Sail. A 1 kb insert appeared to be present 203 ,4—" 35 -i 1 '5 •*-• 30 - Q2. el l o CD 25 - E 'jU U c 20 - o E c 15 - (1) -*-CD» i_ CD 10 - j* ca »*—Q* . D CD 5 - C "o L_ D_ 0 - proV + + IS5 IS5E190QE190Q - - --++ + + prvW + + ++ ++.. ++. _ + + proX + + ++ ++.. ++++ -- proQ +- + _ +.+. + . + . +. Figure 6.9: Effects of proU mutations on the proQ phenotype following growth in high osmolality medium. ProQ+ (+) or ProQ' (-) variants of E. coli RM2 and derivatives either encoding an intact proV, proW or proX ORF (+), an in-frame deletion of the indicated ORF (-), or allelesproV677 (E190Q) orproU205 (IS5), or mutationproU205 (Tables 6.1 and 6.2), were grown in MOPS based minimal medium supplemented with 250 mM NaCl as described in Materials and Methods (2.4.1.1). ProP activity was determined at an assay medium osmolality of 0.65 mol/kg using proline uptake assays as described in Materials and Methods (2.4.1.1). 204 1 UU - 'c rot e ,,DC, T 80 - J3=- -r p^ Pn T m g cel l p n ' 60 - E 40 - rat e (nmo l r=- r=1 20 - uptak e Prolin e n - u i i i i i i i i i i i i i i proV + + IS5 IS5E190QE190Q - - --++ + + proW + + ++ ++-. ++. . + + proX + + ++ ++-. ++++ -- proQ + - + - + - + - + - + - + - Figure 6.10: Effect of pro U mutations on the proQ phenotype when cells express ProP from plasmid pDC79. ProQ+ (+) or ProQ* (-) variants of E. coli RM2 and derivatives either encoding an intact pro V, proW or proXORF (+), an in-frame deletion of the indicated ORF (-), or a\MesproV677 (E190Q) orproU205 (IS5), were grown in MOPS based minimal medium supplemented with 50 mM NaCl. ProP activity was determined at an assay medium osmolality of 0.49 mol/kg using proline uptake assays as described in Materials and Methods (2.4.1.1). 205 proV + + IS5 IS5 E190QE190Q + proW + + + + + + + proX + + + + + + + + + + - proQ + + + + + - + *•*. Chromosome ^ 4 ' &&ttU&*»A<. .*,,_, -asaasiseif^ -'-*«*< > *'*?v&tr' *«mm».. "&itmm&sW 'IJlMllMMW •- ^ PfOP Chromosome HProP + pDC79 Figure 6.11: Modulation of ProP levels by ProQ inprolJ mutants. Western blot analysis of ProP levels in ProQ+ (+) or ProQ" (-) variants of E. coli RJV12 and derivatives either encoding an intact pro V, proW or proX ORF (+), an in-frame deletion of the indicated ORF (-), or alleles proV677 (E190Q) or pro U205 (IS5), when ProP is expressed: Top panel: from the chromosome following growth of cells in MOPS based minimal medium supplemented with 250 mM NaCl, or Bottom panel: from plasmid pDC79 following growth in MOPS based medium supplemented with 50 mM NaCl. Western blotting was performed as described in Materials and Methods (2.3.3.3) 206 upstream of the Sail site in the genome of strain WG210 (Figure 6.12A). The amplicon was sequenced, revealing an insertion sequence that aligned with insertion sequence 5 (IS5) (Appendix 4). This sequence had inserted between the P347 and L348 codons of proV. MutationproU205 abolishes ProU activity and also results in polarity on the downstreamproWandproXgenes (Figure 6.6) 6.4 Discussion The results presented in chapter 5 suggest that ProQ regulates ProP protein levels within the cell on the level of translation. This regulation is proposed to involve another cellular factor, possibly a regulatory RNA molecule. If this model is correct, then the proQ phenotype should be suppressed by deleting the locus encoding the putative regulatory RNA. An E. coli strain showing suppression of the proQ phenotype was identified and analysis of the various mutations present within this strain showed that deletion of proQ, proP, or an ORF flankingproP, or theproU operon was responsible for suppression of the proQ phenotype. The proP locus is flanked by genes with unknown functions (yjdN, yjdM, yjdA and yjcZ) and genes whose protein products are involved in regulation of other loci (e.g. BasRS forms a two component regulatory system involved in iron regulation) (Hagiwara, et al. 2004) (Figure 6.2 and Table 6.1). It was deemed possible that one of these proteins Flanking proP would regulate proP expression. However, replacement of each ORF with a kanamycin cassette did not alter ProP activity or the proQ phenotype under the conditions tested. 207 Tn5 I proV proW proX 1 > > 2> 356bp I 791bp 1008 bp ' 917 bp proVl proX2 Ndel Sail PVHI B RM 2 1 WG 2 I 5-2— 4.0 Ez9i ** n A. 2.0 H pww 1.6 SI WiSm l.o- 0.5- Figure 6.12: Characterization of the proU205 mutation. A) Primers proVl and proX2 were used to amplify the specified region of the proU operon. Positions of Ndel, Sail and Pvul sites are indicated. The size of the amplicon for the wildtype operon is 3081 bp. B) PCR amplification of the proU operon from strains RM2 {prolf) and WG210 (RM2proU205) using primers proVl and proX2. Template DNA from RM2 yielded the predicted 3012 bp amplicon, while template DNA fromWG210 yielded an amplicon that is approximately 1 Kb larger than that given when RM2 DNA was used as a template. C) The PCR amplicon obtained with DNA from WG210 was purified (Uncut) and then subjected to digestion using Ndel, Sail and Pvul restriction endonucleases. 208 Previous results suggested that the activity of ProP or ProU would attenuate the activity of the other transporter. This attenuation is thought to arise following osmoprotectant uptake by the transporter, resulting in a decreased response to osmotic stress and the signals sensed by each transproter (Sutherland, et al. 1986; Cairney, et al. 1985b). Initially, AproU600 was shown to suppress theproQ phenotype. It was unclear whether this resulted from a lack of ProU activity, possibly through a lack of signal attenuation, or from the absence of a chromosomalproU locus. To determine if suppression of the proQ phenotype was due to absence of ProU transport activity, the effects of ProU activity or absence of the proU coding sequence itself, various mutants were prepared. To disentangle the effects of ProU activity and the proU coding sequence on the proQ phenotype, the ProP activity in proQ+ and proQ' variants of WG1201 (RM2 A(proV-proX)2098::FRT) and WG1197 (RM2proV677) were compared and data showed that both mutations abolished ProU activity but the proQ phenotype was maintained in strains encoding ProVE190Q (WG1197 and WG1198), while the phenotype was suppressed in strains from which the/?ro£/operon was deleted (WG1201 and WG1202). This analysis indicates that suppression of the proQ phenotype is not due to a lack of ProU activity, but instead arises from a loss or alteration of sequences within the proU operon. Theprot/operon is composed of three separate reading frames, proV, proW and proX and it is possible that at least one of these regions is required for the proQ phenotype. In- frame deletions of each ORF were created in E. coli RM2, abolishing ProU transport activity in each of the resulting strains. Further, ProX was fully expressed in WG1078 (RM2 AproV858::FRT) and WG1083 (RM2 AproW859::FRT), but not in WG1088 209 (BM2proX870::FRT), as expected indicating that deletions of these ORFs did not affect expression of the remainder of the operon following cell growth at high osmolalities. The presence of the proQ phenotype was tested in strains carrying a deletion of each ORF of the pro U operon. Strains WG1083/WG1085 (AproW859::FRT) and WG1088/WG1090 (AproX870::FRT) showed suppression of the proQ phenotype, following growth in LOM or HOM, while strains WG1078/WG1080 (AproV858::FRT) retained theproQ phenotype when grown in LOM, and showed a weak proQ phenotype following growth in HOM. The proQ phenotype was suppressed when ProP was expressed from plasmid pDC79 in these strains regardless of the ORF deleted. If a specific sequence within prolJ is important for the proQ phenotype, it is likely located within the pro W or proX ORF, as deletion of these genes suppresses XheproQ phenotype under all of the conditions tested here. It is possible that variable suppression of the proQ phenotype in the AproV mutant resulted from altered expression patterns of downstream genes, due to deletion of the NRE. Regulation ofproU expression is not only achieved through actions at the promoters, but it is also affected by a regulatory sequence located within pro V termed the negative regulatory element (NRE). Deletion of this element derepresses;?ro£/ expression promoters under conditions of low osmolality, but does not affect the promoter activity at high osmolalities (Introductionl.2.2) (Dattananda, et al. 1991; Overdier and Csonka, 1992). When the NRE is present, proU expression at lower osmolalities is approximately 25-fold lower than when the NRE is absent (Dattananda, et al. 1991). In the case of the AproV mutant studied here, we would expect upregulation of the proU expression under low osmolality conditions, and typical levels of expression from the proU promoter 210 following growth in high osmolality medium. If a sequence downstream of pro V encoded a regulatory RNA which was able to interact with the proP mRNA message, upregulating its expression would result in a requirement for ProQ to prevent these interactions, giving the proQ phenotype. When cells are grown in high osmolality medium, the levels of proU expressed are no longer altered by the absence of the NRE. Under these conditions we see a partial requirement for ProQ to amplify ProP activity, as is seen with the AproVmutation. It is also possible that sequences located downstream of the pro £/operon regulate proP expression or contribute to the proQ phenotype. Altered expression of these ORFs could be achieved when part of the proU operon is deleted, bringing the ProU promoter in closer proximity to downstream genes. Previous research revealed that expression of the MprA protein from a plasmid vector blocked the osmotic induction of proU (del Castillo, et al. 1990). Further analysis revealed that deletion of this gene from the chromosome did not alter the osmotic regulation of proU expression (del Castillo, et al. 1991). The mprA locus is located 1.3 kb downstream of the pro [/operon and it regulates production of EmrA and EmrB which form the membrane fusion protein and inner membrane transproter of the EmrAB-TolC multi drug efflux pump (Lomovskaya, et al. 1995). Also present immediately downstream ofproX are ygaXand ygaY, each encoding a protein of unknown function. TheygaXlocus is 192 bp downstream from proX and is predicted to encode an 88-amino acid polypeptide, while the ygaY locus is 488 bp downstream of proX and is predicted to encode a 293-amino acid polypeptide. In-frame deletions within the proU operon, would bring ygaX and ygaY into proximity with the 211 nrdH VJ™!L> 2,798,745 nrdE / nrdF / proV y proW / proX y j / ygaY / ysa^ y rs°H/ mprA/ emrA 1 emrB 1 2,812,176 Figure 6.13: Chromosomal loci surrounding proU (proVWX) in E. coli from 2,798,745 bp to 2,812,176 bp. 212 proU promoters, altering expression of these genes. It is possible that altered expression of these downstream loci could be involved in suppression of the proQ phenotype. In contrast to the effects of most other proU alleles, strains with allele proU205 show a clear proQ phenotype when ProP is expressed from the chromosome following growth in low or high osmolality medium (Figures 6.7 and 6.9 respectively). However, this allele suppresses the proQ phenotype if ProP is expressed from plasmid pDC79 (Figure 6.10). Analysis of proU205 revealed that it arose through insertion of IS5 (Figure 6.12 and Appendix B). This mutation inactivates ProU (Figure 6.6A) and lowers expression of ProX (Figure 6.6B). The reduced ProX expression level indicates that XheproU operon is expressed in strains containing alleleproU205, but the expression or the stability of this transcript is reduced with respect to the wildtype levels. It is possible that the insertion sequence within pro Flowers expression of the downstream pro U genes, as well as the putative regulatory RNA, as it places them further away from the proU promoter. In cases where the proQ phenotype is observed, there is a clear difference in the level of ProP expression in the presence and absence of ProQ (Figure 6.11). Introduction of pro U mutations which result in suppression of the prog phenotype express levels of ProP which are similar to what is seen in wildtype cells, and causes this expression to be independent of proQ mutations (Figure 6.11). This indicates that ProU and ProQ do not independently regulate ProP levels, but that ProQ modulates ProP levels in a manner that is dependent on ProU. 213 A model for the involvement of proU in theproQ phenotype has been developed. The DNA sequence encoding the/?rot/operon also encodes a regulatory RNA molecule or affects production of a regulatory RNA encoded at another locus. This regulatory RNA interacts with proP mRNA, resulting in RNA-RNA duplexes which are degraded, resulting in lowered ProP expression. ProQ binds to the regulatory RNA, preventing it from forming RNA-RNA duplexes with proP mRNA. This increases the lifetime of the proP mRNA and increases expression of ProP (Figure 6.14). It is interesting that deletions ofproV, proW and proX all have an effect on the proQ phenotype. It is possible but unlikely that the entire pro U operon encodes a regulatory RNA which interacts with the proP message. It is more likely that a small RNA exists within the coding regions of proW and proX, since deletion of these ORFs always results in suppression of the proQ phenotype, while sequences present in the pro FORF result in regulation of the expression of the regulatory RNA molecule. Alternatively, it is possible that expression of proVWX is able to alter the expression of the regulatory RNA, responsible for the proQ phenotype, encoded elsewhere on the chromosome. This study demonstrates that the proU operon includes a sequence that is important in the regulation of ProP expression. Analysis of the proU operon for smaller or alternate reading frames, or for potential regulatory RNA sequences may also led to the identification of a regulatory RNA, or another protein involved in regulation of ProP expression via ProQ. Further, the genes immediately downstream of proX should be deleted in order to determine their effects on the proQ phenotype and ProP activity. 214 ProQ+ proV y proWy proXy QT negRNA $ / ProQ- proV y proWy proXy N Figure 6.14: Model for suppression of the proQ phenotype by deletions inproU. The proU operon (proVWX) encodes a regulatory RNA molecule or affects production of a regulatory RNA encoded at another locus (regRNA). In ProQ" cells, this regulatory RNA interacts with proP mRNA (proP), resulting in RNA-RNA duplexes which are degraded, resulting in lowered ProP expression. In ProQ+ cells, ProQ binds to the regulatory RNA, preventing it from forming RNA-RNA duplexes with proP mRNA. This increases the lifetime of the proP mRNA and increases expression of ProP. 215 Chapter 7: General Discussion 7.1 Discussion This thesis describes experiments aimed at further understanding the role of the osmoregulatory protein ProQ in E. coli. Prior to this work, it was known that disruptions at the proQ locus impaired ProP activity (Milner and Wood 1989; Kunte, et al. 1999); however, the mechanism by which ProQ affected ProP activity was unknown Prior to this work, it was not possible to obtain large quantities of pure ProQ due to contamination of ProQ preparations with DNA binding proteins (Crane, 2004; Smith, et al. 2004). Thus, alternative methods for purification of ProQ were sought. I was able to develop a system for the overexpression of a histidine-tagged version of ProQ, ProQ-His6 and showed that it maintained in vivo function by complementing a chromosomal deletion at the proQ locus. Initial characterization of ProQ and ProQ-His6 had been difficult as the proteins were poorly soluble following overexpression. It was believed that inclusion bodies were formed upon overexpression; however, poor solubility was found to arise due to lysis in buffer with low ionic strength. By increasing the NaCl concentration of the lysis buffer to 0.6 M, the solubility of both ProQ and ProQ-His6 were improved dramatically allowing me to further characterize the structure and function of ProQ. Pure preparations of ProQ-His6 were produced and the homology models of the isl and C-terminal domains deduced by Dr. R.A.B. Keates were tested (Introduction 1.6). Using limited trypsin proteolysis experiments, I was able to show the presence of protease-resistant domains within ProQ. Further analysis of these protease-resistant 216 domains, by in-gel digestion, and identification by mass spectrometry, revealed that they corresponded to regions within the predicted N- and C-terminal domains, and not to the intervening linker region, predicted to be mostly unstructured and hence likely to be trypsin sensitive (Chapter 4; Smith, et al. 2007). Plasmids were constructed for the overexpression of N-terminally histidine tagged variants of the N-terminal and C- terminal domains of ProQ (HeN and HeC respectively). I purified both of these domains and characterized their secondary structure using circular dichroism (CD) spectroscopy (Chapter 4). The N-terminal domain was composed of mostly a-helical structure, as the spectrum had negative ellipticities at 208 and 222 nm, supporting the model of the N- terminal domain based on the structure of FinO. The CD spectrum revealed that the C- terminal domain contained mostly of P-sheet structure, as the spectrum showed a minimum at 217 nm, supporting the secondary sequence prediction for this domain. The CD-spectrum of the C-terminal domain was not characteristic of SH3-like domains, which, although they are composed of mostly P-sheet structure, have characteristic CD spectrum with a maximum at 220 nm (Maxwell and Davidson, 1998; Smith, et al. 2007). The negative ellipticity at 217 nm was consistent with homology models of the C- terminal domain of ProQ based on an Sm motif, and thus, Dr. Keates was able to produce a more accurate model of the C-terminal domain of ProQ on the structure of Hfq from Methanococcus jannischii (2QTX) (Nielsen, et al. 2007; R.A.B. Keates, unpublished data). I created a number of plasmid constructs based on the pBAD24 vector, for expression of various domains of ProQ to test their in vivo function. Complementation studies performed with various domains and combinations of domains of ProQ revealed that 217 residues within the N-terminal domain of ProQ are necessary for function, while those in the C-terminal domain are required to enhance the function of the N-terminal to give full activity of ProQ (Chapter 4). Previous analysis ofproP based on proP::lacZ fusions and Western blot analysis revealed that ProQ did not appear to affect ProP at the level of either transcription or translation (Milner and Wood, 1989; Kunte, et al. 1999). ProQ was thus suggested to amplify ProP activity post-translationally, through protein-protein interactions. However, using an in vitro proteoliposome system containing ProP co-reconstituted with ProQ, I was unable to show a direct effect of ProQ on ProP activity (Chapter 5). Previous experiments involving fluorescence microscopy experiments revealed that ProP localizes to the poles of E. coli cells (Romantsov, et al. 2007). A ProQ-RFP fusion was made to examine the cellular localization of ProQ. The presence of RFP was detected uniformly in the cytoplasm and not localized at the poles (Chapter 5). These results indicate that ProQ is unlikely to act directly on ProP. Using Western blot analysis of cells grown in medium of various osmolalities, as well as of cells grown in LB medium through exponential and stationary phase, I showed that ProQ amplified ProP levels when cells were grown in minimal medium with an osmolality that was lower than 0.7 mol/kg, or during late exponential phase when cells were grown in LB medium (Chapter 5). Further, I showed that the presence of proQ resulted in amplified ProP expression under control of the pBAD promoter, indicating that the effect of ProQ on ProP levels was not occurring at the transcriptional level (Chapter 5). Given that the N-terminal domain of ProQ can be modeled on the structure of the mRNA binding protein FinO, while the C-terminal domain can be modeled on the 218 general RNA chaperone Hfq, it was proposed that ProQ acts to modulate interactions between an sRNA and proP mRNA. In this model, the presence of ProQ would prevent interactions between the sRNA and proP mRNA, resulting in increased ProP expression, while in the absence of ProQ, the sRNA is able to interact with proP mRNA, resulting in either its degradation, as is seen for many sRNA-mRNA duplexes, or inhibition of its translation (Barrandon, et al. 2008; Repoila and Darfeuille, 2009; Aiba, 2007). In this work it was shown that ProQ could co-purify with RNA, identified as 16S and 23S rRNA and valine tRNA (Chapter 5). It was also shown that ProQ was present in crude ribosome preparations (Chapter5). ProQ was previously identified as a ribosome- associated protein by Jiang, et al. (2007). In those studies, the ribosome fraction was purified and all proteins present tagged with isobaric tags. These tags facilitated the detection of the proteins by mass spectrometry. Previous reports have also shown Hfq to be ribosome associated (DuBow, et al. 1977), however, the functional significance of this association is not yet known (Valentin-Hansen, et al. 2004; DuBow, et al. 1977). It is possible that ProQ interacts with the ribosome through the C-terminal Hfq-like domain. If this were the case, the in vivo function of ProQ may depend on ribosome interactions, as expression of the N-terminal domain is able to partially complement aproQ deletion, but full complementation requires the presence of the C-terminal domain (Chapter 4; Smith, et al. 2007). Hfq is a general RNA chaperone, stabilizing mRNA-sRNA interactions found in at least half of the sequenced bacterial and archeal genomes to date (Introduction 1.5.2.2) (Valentin-Hansen, et al. 2004; Sun, et al. 2002). The structures of Hfq from multiple species of bacteria have been solved (Schumacher, et al. 2002; Nielsen, et al. 2007; 219 Nikulin, et al. 2005; Sidote, et al. 2004; Numata, et al. 2004; Sauter, et al. 2003) and they reveal a similar fold and hexameric structure. The C-terminal domain of ProQ was modeled on the Hfq protein from M. jannaschii. The overall fold of the archeal Hfq is similar to that found in E. coli and other eubacteria, but differs in the length of the N- terminal helix. Like Hfq from eubacteria, Hfq from M. jannachsii is able to form hexameric ring structures, but the overall diameter of the hexamer is smaller (Nielsen, et al. 2007). Although these structural differences exist, Hfq from M. jannaschii is able to complement the growth defects of a chromosomal hfq deletion in E. coli (Nielsen, et al. 2007). Homologues of ProQ have yet to be identified in Gram-positive bacteria, while Hfq homologues are present in these organisms but have not been as extensively studied. Interestingly, studies performed in Staphylococcus aureus and Bacillus subtilis indicate that Hfq may not play a global regulatory role in these Gram-positive organisms, as Hfq is not required for the action of sRNA on any of the mRNA targets, as it is in the Gram- negative organisms (Bohn, et al. 2007; Gaballa, et al. 2008; Heidrich, et al. 2006). Deletions in hfq from Gram-positive Listeria monocytogenes, on the other hand, have been found to decrease salt and alcohol tolerance, and these mutants show a decreased survival during prolonged amino acid starvation (Christiansen, et al. 2004). Experiments performed to determine the identities of sRNAs co-purifying with Hfq froml. monocytogenes identified only three, LhrA, LhrB and LhrC, the targets of which have not yet been discovered (Christiansen, et al. 2006). When compared to the 22 sRNAs that have been identified as able to interact with Hfq in E. coli using similar techniques, the number of Hfq-interacting sRNAs in L. Monocytogenes is rather limited (MoHer, et al. 220 2002; Moll, et al. 2003). In Gram-negative organisms such as E. coli, Pseudomonas aeruginosa, Legionella pnemophila and Salmonella typhimerium, deletions in hfq showed global effects on the ability of sRNAs to regulate their mRNA targets (Brown and Elliott, 1996; Tsui, et al. 1994; Sonnleitner, et al. 2006; McNealy, et al 2005). It is possible that the requirement of Hfq in Gram-negative bacteria indicates that regulation by sRNAs in these organisms is somehow different than it is in Gram-positives. This may be why ProQ homologues are not found in Gram-positive organisms, even when a ProP orthologue is present. To date, no proteins other than ProQ have been identified with an Hfq-like domain. Functionally important residues for the amplification of ProP activity by ProQ are located in the N-terminal domain of ProQ, which can be modeled on FinO from E. coli, as expression of this domain mproQ strains results in partial restoration of ProP activity. FinO is an mRNA binding protein involved in regulation of F-pilus expression (Introduction, 1.5.2.1) (van Biesen and Frost, 1994). It is possible that the N-terminal domain of ProQ is responsible for interactions with the sRNA and modulation of ProP levels, while the C-terminal domain is responsible for interactions with other cellular structures such as the ribosome. In this work, I was also able to show that ProQ was not always required for full ProP activity (the proQ phenotype could be suppressed). By testing for the proQ phenotype in the presence of a variety of mutations at the proP and proQ loci, as well as within the proU operon, I was able to determine that deletion of sequences within the/?ro£/operon resulted in suppression of the proQ phenotype and that it was the nucleic acid sequence of proU that was important and not the function of the ProU transporter (Chapter 6). 221 WhenproP was expressed from the chromosome, deletions of proW, proX, or the entire jproC/operon resulted in suppression of the proQ phenotype while deletions of proV only partially suppressed the proQ phenotype. This indicates that sequences within the pro W and/or the proX ORF are important for the proQ phenotype. When proP was expressed from plasmid pDC79, all deletions within proU resulted in suppression of the proQ phenotype (Chapter 6). In all cases, a point mutation, ProVE190Q, which disrupted ProU transport activity while maintaining the proU coding sequence, maintained the proQ phenotype (Chapter 6). The differences seen between plasmid based expression of ProP and chromosomally expressed proP on the ability of proU mutations to suppress the proQ phenotype, are proposed to be due to alterations in the stoichiometry between the levels of an sRNA and those of proP mRNA. The ability of proU deletions to suppress the proQ phenotype indicate that the coding sequence for proU includes a sequence for an sRNA, or that expression of intact proU mRNA affects the cellular level of an sRNA able to interact with the proP message. If pro U encoded both an sRNA and mRNA sequence, it would be called a bi-functional RNA. Current literature on regulation of target genes by small RNAs typically implicates short sequences of RNA which do not encode functional polypeptides (Barrandon, et al. 2008; Repoila and Darfeuille, 2009; Vogel and Papenfort, 2006; Gottesman, 2005). Only three examples of bi-functional RNAs have been identified in bacteria to date. Glucose is taken up, via the glucose phosphotransfer system (PTS),and converted to glucose-6-phosphate by the action of IICBglc. SgrS of E. coli encodes a sRNA which down regulates IICB8'0 by base pairing with the encoding mRNA (ptsG) (Vanderpool, 222 2007; Vanderpool and Gottesman, 2004). The sgrS RNA also encodes a small 43 amino acid polypeptide (Wadler and Vanderpool, 2007). This polypeptide has been shown to inhibit the activity of the IICBGlc system through direct protein-protein interactions (Wadler and Vanderpool, 2007). A second example, RNAIII from Staphylococcus aureus, is able to affect expression of several virulence genes through base-pairing mechanisms (Huntzinger, et al. 2005). RNAIII has also been shown to encode a short polypeptide, however, functional significance of this protein has not yet been determined (Huntzinger, et al. 2005; Boisset, et al. 2007). Finally, tmRNA, found in eubacteria, acts as an alanyl tRNA as well as encoding a peptide, SsrA, which acts as a signal to the cell to mark incomplete polypeptides during a process called trans-translation et al. 2008; Akimitsu, 2008; Dulebohn, et al. 2007). It is difficult to predict the targets of known sRNA molecules as base pairing interactions between the two molecules is usually not continuous and depends on the secondary structures taken on by both the regulatory RNA as well as by the mRNA target (Vogel and Wagner 2007). Attempts were made, in silico, to identify regulatory RNA that would bind to the proP message using sRNATARGET (Zhao, et al. 2008). This program attempts to find regions of homology between known and predicted regulatory RNA and the input sequence. Using this program, regulatory RNA within the region of theproUcoding sequence were not identified. Further, attempts were made to predict the secondary structure of both the proP and pro U mRNA using the program RNAfold (Hofaker et al. 1994). In the case of proP both the full length mRNA (including the 5'UTR) as well as the mRNA encoding the start to stop codons of the ProP protein were used as input data. For ProU, each individual ORF, as well as from the pro V and proX 223 ORFs together were used as an input sequence. The output from this experiment was extremely complex and regions of homology between the mRNA molecules could not be identified. Overall, work presented in this thesis contributes to our understanding of osmoregulation in E. coli. Before this work was performed, little was known concerning the mechanism by which ProQ amplified ProP activity. I have shown that ProQ acts to modulate ProP levels within the cell on the level of translation. This work also raises the possibility of regulation of ProP levels via RNA-RNA interactions and that this regulation is dependent onproU, a second osmoregulatory protein in E. coli with similar substrate specificity to ProP. ProQ may act to link ProP levels and activity to those of ProU, acting to co-ordinate the osmoregulatory response. 7.2 Future work This work indicates that ProQ modulates ProP levels on the level of translation. Further work should be performed to show the effects of ProQ on proP mRNA levels. This could be accomplished using real time PCR to quantify the levels of proP mRNA in proQ+ and proQ bacteria. The effects ofproU mutations on the ability of ProQ to modulate proP mRNA levels could also be explored in this manner. It has been determined in this work, as well as by previous work by Jiang, et al. (2007) that ProQ associates with the ribosome. Future work could also be aimed at determining if the association of ProQ with ribosomes requires proP or proU mRNA. Further experimentation could also be performed to determine if the N-terminal or C- terminal domains alone co-purify with the ribosome fraction. The latter experiment will 224 give particularly interesting results, as it has been shown that the N-terminal domain of ProQ is able to partially complement aproQ deletion. If the N-terminal domain is also found to be ribosome associated, then it is likely that this association is linked to the proQ phenotype. Alternatively, the C-terminal domain of ProQ alone was unable to complement aproQ deletion. However, the C-terminal domain expressed with the N- terminal domain was able to give full complementation of aproQ deletion. If only the C- terminal domain were to be ribosome-associated, it would show that the proQ phenotype is not entirely dependent on ProQ's association with the ribosome; however, this association is required for the full effect of ProQ on proP expression. If one domain was found to co-purify with the ribosomal fraction, the other domain could then be used as "bait" to fish for the putative regulatory RNA proposed to interact with the proP transcript, preventing identification of more abundant ribosomal RNA by these screens and identifying less abundant RNA co-purifying with ProQ. These experiments may be more successful than using the entire ProQ protein as "bait" as was performed in this study as this led to the identification of stable and abundant rRNA and tRNA. Alternatively, more sensitive techniques involving microarray analysis of mRNA present in ProQ preparations may lead to the identification of sRNA bound to ProQ. Full length ProQ could also be purified from cell extracts lysed in buffers with varying levels of salt. These experiments may reveal a level of salt that dissociates non specific RNA from species of RNA that are more tightly bound to ProQ allowing for the identification of a regulatory RNA modulating proP expression. 225 Within this work, the C-terminal domain of ProQ was modeled on the structure of Hfq from M. jannischii (2QTX) (Nielsen, et al. 2007; R.A.B. Keates, unpublished data). Hfq has been shown to form hexameric ring structures, and it is possible that ProQ also forms hexamers via its C-terminal domain. Further characterization of this domain alone, as well as with the remainder of the ProQ protein by size exclusion chromatography may reveal that what was previously believed to be aggregated protein present in the excluded volume of the column, is actually composed of physiologically relevant ProQ multimers or ProQ-RNA complexes. Sequences within the pro t/operon seem to play an important role in the proQ phenotype, as deletions within the/?ro[/operon, but not point mutations resulting in an inactive transporter, result in suppression of the proQ phenotype (Chapter 6). Functionally important sequences are likely to be localized within the proWand/or the proXORFs as deletion of these sequences results in suppression of the prog phenotype, while deletion of the pro FORF shows only partial suppression of the proQ phenotype. Future work to further characterize this phenomenon could include cloning of various portions or the entire /?rot/operon onto a plasmid vector in aprolf background and looking for the presence of the proQ phenotype. If a particular sequence within proU could be identified, further experiments could then be performed to characterize its interaction with ProQ and proP mRNA. Within this work, attempts were made to crystallize the N-terminal domain of ProQ. Although I was able to purify the N-terminal domain and concentrate it to 30 mg/mL, I was unable to obtain crystals under any of the conditions tested. It is possible that the C- terminal boundary chosen for the N-terminal domain of ProQ was not suitable for 226 crystallization, even though this domain has been shown to be properly folded in vivo. Future work could focus on identifying a new C-terminal end point for the N-terminal domain which may result in a domain suitable for crystallization studies. Alternatively, other structural techniques such as NMR could be pursued as the N-terminal domain of ProQ is relatively small making it a good candidate. 227 References Aiba, H. (2007) Mechanism of RNA silencing by Hfq-binding small RNAs. Curr. Opin. Microbiol. 10, 134-139. Akimitsu, N. (2008) Messenger RNA surveillance systems monitoring proper translation termination. J. Biochem. 143, 1-8. Ali Azam, T., Iwata, A., Nishimura, A., Ueda, S. and Ishihama, A. (1999) Growth phase-dependent variation in protein composition of the Escherichia coli nucleoid. J. Bacteriol. 181, 6361-6370. Altuvia, S. (2007) Identification of bacterial small non-coding RNAs: experimental approaches. Curr. Opin. Microbiol. 10, 257-261. Andresen, P. A., Kaasen, I., Styrvold, O. B., Boulnois, G. and Strom, A. R. (1988) Molecular cloning, physical mapping and expression of the bet genes governing the osmoregulatory choline-glycine betaine pathway of Escherichia coli. J. Gen. Microbiol. 134, 1737-1746. Argaman, L., Hershberg, R., Vogel, J., Bejerano, G., Wagner, E. G., Margalit, H. and Altuvia, S. (2001) Novel small RNA-encoding genes in the intergenic regions of Escherichia coli. Curr. Biol. 11, 941-950. Baba, T., Ara, T., Hasegawa, M., Takai, Y., Okumura, Y., Baba, M., Datsenko, K. A., Tomita, M., Wanner, B. L. and Mori, H. (2006) Construction of Escherichia coli K-12 in- frame, single-gene knockout mutants: the Keio collection. Mol. Syst. Biol. 2, 2006.0008. Baker, C. S., Morozov, I., Suzuki, K., Romeo, T. and Babitzke, P. (2002) CsrA regulates glycogen biosynthesis by preventing translation of glgC in Escherichia coli. Mol. Microbiol. 44, 1599-1610. Balaji, B., O'Connor, K., Lucas, J. R., Anderson, J. M. and Csonka, L. N. (2005) Timing of induction of osmotically controlled genes in Salmonella enterica Serovar Typhimurium, determined with quantitative real-time reverse transcription-PCR. Appl. Environ. Microbiol. 71, 8273-8283. Ball, C. A., Osuna, R., Ferguson, K. C. and Johnson, R. C. (1992) Dramatic changes in Fis levels upon nutrient upshift in Escherichia coli. J. Bacteriol. 174, 8043-8056. 228 Barrandon, C, Spiluttini, B. and Bensaude, O. (2008) Non-coding RNAs regulating the transcriptional machinery. Biol. Cell. 100, 83-95. Barron, A., Jung, J. U. and Villarejo, M. (1987) Purification and characterization of a glycine betaine binding protein from Escherichia coli. J. Biol. Chem. 262, 11841-11846. Bass, R. B., Strop, P., Barclay, M. and Rees, D. C. (2002) Crystal structure of Escherichia coli MscS, a voltage-modulated and mechanosensitive channel. Science 298, 1582-1587. Bauer, C. B., Kuhlman, P. A., Bagshaw, C. R. and Rayment, I. (1997) X-ray crystal structure and solution fluorescence characterization of Mg.2'(3')-0-(N- methylanthraniloyl) nucleotides bound to the Dictyostelium discoideum myosin motor domain. J. Mol. Biol. 274, 394-407. Bejerano-Sagie, M. and Xavier, K. B. (2007) The role of small RNAs in quorum sensing. Curr. Opin. Microbiol. 10, 189-198. Blattner, F. R., Plunkett, G.,3rd, Bloch, C. A., Perna, N. T., Burland, V., Riley, M., Collado-Vides, J., Glasner, J. D., Rode, C. K., Mayhew, G. F., Gregor, J., Davis, N. W., Kirkpatrick, H. A., Goeden, M. A., Rose, D. J., Mau, B. and Shao, Y. (1997) The complete genome sequence of Escherichia coli K-12. Science 277, 1453-1474. Blount, P., Sukharev, S. I., Moe, P. and Kung, C. (1997) Mechanosensitive channels of E. coli: a genetic and molecular dissection. Biol. Bull. 192, 126-127. Bohn, C., Rigoulay, C. and Bouloc, P. (2007) No detectable effect of RNA-binding protein Hfq absence in Staphylococcus aureus. BMC Microbiol. 7, 10. Boisset, S., Geissmann, T., Huntzinger, E., Fechter, P., Bendridi, N., Possedko, M., Chevalier, C, Heifer, A. C, Benito, Y., Jacquier, A., Gaspin, C, Vandenesch, F. and Romby, P. (2007) Staphylococcus aureus RNAIII coordinately represses the synthesis of virulence factors and the transcription regulator Rot by an antisense mechanism. Genes Dev. 21, 1353-1366. Boos, W., Ehmann, U., Forkl, H., Klein, W., Rimmele, M. and Postma, P. (1990) Trehalose transport and metabolism in Escherichia coli. J. Bacteriol. 172, 3450-3461. Booth, I. R. and Louis, P. (1999) Managing hypoosmotic stress: aquaporins and mechanosensitive channels in Escherichia coli. Curr. Opin. Microbiol. 2, 166-169. 229 Bossemeyer, D., Borchard, A., Dosch, D. C, Helmer, G. C, Epstein, W., Booth, I. R. and Bakker, E. P. (1989) K+-transport protein TrkA of Escherichia coli is a peripheral membrane protein that requires other trk gene products for attachment to the cytoplasmic membrane. J. Biol. Chem. 264, 16403-16410. Bradford, M. M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248-254. Brandon, L., Dorus, S., Epstein, W., Altendorf, K. and Jung, K. (2000) Modulation of KdpD phosphatase implicated in the physiological expression of the kdp ATPase of Escherichia coli. Mol. Microbiol. 38, 1086-1092. Brennan, R. G. and Link, T. M. (2007) Hfq structure, function and ligand binding. Curr. Opin. Microbiol. 10, 125-133. Brescia, C. C, Mikulecky, P. J., Feig, A. L. and Sledjeski, D. D. (2003) Identification of the Hfq-binding site on DsrA RNA: Hfq binds without altering DsrA secondary structure. RNA 9, 33-43. Brown, E. D. and Wood, J. M. (1992) Redesigned purification yields a fully functional PutA protein dimer from Escherichia coli. J. Biol. Chem. 267, 13086-13092. Brown, L. and Elliott, T. (1996) Efficient translation of the RpoS sigma factor in Salmonella typhimurium requires host factor I, an RNA-binding protein encoded by the hfq gene. J. Bacteriol. 178, 3763-3770. Burstein, C. and Kepes, A. (1971) The alpha-galactosidase from Escherichia coli K12. Biochim. Biophys. Acta 230, 52-63. Casadaban, M. J. 1976. Transposition and fusion of the lac genes to selected promoters in Escherichia coli using bacteriophage lambda and Mu. J. Mol. Biol. 104, 541-555. Cairney, J., Booth, I. R. and Higgins, C. F. (1985a) Osmoregulation of gene expression in Salmonella typhimurium: proU encodes an osmotically induced betaine transport system. J. Bacteriol. 164, 1224-1232. 230 Cairney, J., Booth, I. R. and Higgins, C. F. (1985b) Salmonella typhimurium proP gene encodes a transport system for the osmoprotectant betaine. J. Bacteriol. 164, 1218- 1223. Caswell, R., Williams, J., Lyddiatt, A. and Busby, S. (1992) Overexpression, purification and characterization of the Escherichia coli MelR transcription activator protein. Biochem. J. 287 ( Pt 2), 493-499. Cayley, S., Lewis, B. A., Guttman, H. J. and Record, M. T.,Jr. (1991) Characterization of the cytoplasm of Escherichia coli K-12 as a function of external osmolarity. Implications for protein-DNA interactions in vivo. J. Mol. Biol. 222, 281- 300. Chandaroy, P., Sen, A. and Hui, S. W. (2001) Temperature-controlled content release from liposomes encapsulating Pluronic F127. J. Control. Release 76, 27-37. Chang, G., Spencer, R. H., Lee, A. T., Barclay, M. T. and Rees, D. C. (1998) Structure of the MscL homolog from Mycobacterium tuberculosis: a gated mechanosensitive ion channel. Science 282, 2220-2226. Christiansen, J. K., Larsen, M. H., Ingmer, H., Sogaard-Andersen, L. and Kallipolitis, B. H. (2004) The RNA-binding protein Hfq of Listeria monocytogenes: role in stress tolerance and virulence. J. Bacteriol. 186, 3355-3362. Christiansen, J. K., Nielsen, J. S., Ebersbach, T., Valentin-Hansen, P., Sogaard- Andersen, L. and Kallipolitis, B. H. (2006) Identification of small Hfq-binding RNAs in Listeria monocytogenes. RNA 12, 1383-1396. College of Biological Sciences Advanced Analysis Center (2009). CBS DNA sequencing protocols. Retrieved March, 2009 from http://www.uoguelph.ca/~genomics/Genomics%20Facility%20-%20Protocols.htm. Collins, B. K., Tomanicek, S. J., Lyamicheva, N., Kaiser, M. W. and Mueser, T. C. (2004) A preliminary solubility screen used to improve crystallization trials: crystallization and preliminary X-ray structure determination of Aeropyrum pernix flap endonuc lease- X.Acta Crystallogr. D Biol. Crystallogr. 60, 1674-1678. Corry, B., Rigby, P., Liu, Z. W. and Martinac, B. (2005) Conformational changes involved in MscL channel gating measured using FRET spectroscopy. Biophys. J. 89, L49-51. 231 Crane, R. A. (2004) The genetic, biochemical, and functional characterization of ProQ, a regulator of the osmosensor and osmoregulator ProP of Escherichia coli. Ph.D. Thesis. University of Guelph, Guelph ON. Cronan, J. E. (2003) Bacterial membrane lipids: where do we stand? Annu. Rev. Microbiol. 57, 203-224. Cruickshank, C. C, Minchin, R. F., Le Dain, A. C. and Martinac, B. (1997) Estimation of the pore size of the large-conductance mechanosensitive ion channel of Escherichia coli. Biophys. J. 73, 1925-1931. Csonka, L. N. and Epstein, W. (1996) Osmoregulation, p. 1210-1223 In F.C. Neidhardt, R. Curtiss III, J.L. Ingraham, E.C.C. Lin, K.B. Low, B. Magasanik, W.S. Reznikoff, M. Riley, M. Schaechter, and H.E. Umbarger (ed.), Escherichia coli and Salmonella: cellular and molecular biology, 2nd ed. ASM Press, Washington, D.C. Csonka, L. N. and Hanson, A. D. (1991) Prokaryotic osmoregulation: genetics and physiology. Annu. Rev. Microbiol. 45, 569-606. Csonka, L. N., Ikeda, T. P., Fletcher, S. A. and Kustu, S. (1994) The accumulation of glutamate is necessary for optimal growth of Salmonella typhimurium in media of high osmolality but not induction of the pro C/operon. J. Bacteriol. 176, 6324-6333. Culham, D. E., Emmerson, K. S., Lasby, B., Mamelak, D., Steer, B. A., Gyles, C. L., Villarejo, M. and Wood, J. M. (1994) Genes encoding osmoregulatory proline/glycine betaine transporters and the proline catabolic system are present and expressed in diverse clinical Escherichia coli isolates. Can. J. Microbiol. 40, 397-402. Culham, D. E., Henderson, J., Crane, R. A. and Wood, J. M. (2003) Osmosensor ProP of Escherichia coli responds to the concentration, chemistry, and molecular size of osmolytes in the proteoliposome lumen. Biochemistry 42, 410-420. Culham, D. E., Lasby, B., Marangoni, A. G., Milner, J. L., Steer, B. A., van Nues, R. W. and Wood, J. M. (1993) Isolation and sequencing of Escherichia coli gene proP reveals unusual structural features of the osmoregulatory proline/betaine transporter, ProP. J. Mol. Biol. 229, 268-276. Culham, D. E., Tripet, B., Racher, K. I., Voegele, R. T., Hodges, R. S. and Wood, J. M. (2000) The role of the carboxyl terminal alpha-helical coiled-coil domain in osmosensing by transporter ProP of Escherichia coli. J. Mol. Recognit. 13, 309-322. 232 Culham, D. E., Vernikovska, Y., Tschowri, N., Keates, R. A., Wood, J. M. and Boggs, J. M. (2008) Periplasmic loops of osmosensory transporter ProP in Escherichia coli are sensitive to osmolality. Biochemistry 47, 13584-13593. 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. U. S. A. 97, 6640-6645. Dattananda, C. S. and Gowrishankar, J. (1989) Osmoregulation in Escherichia coli: complementation analysis and gene-protein relationships in the proU locus. J. Bacteriol. Ill, 1915-1922. Dattananda, C. S., Rajkumari, K. and Gowrishankar, J. (1991) Multiple mechanisms contribute to osmotic inducibility ofprot/operon expression in Escherichia coli: demonstration of two osmoresponsive promoters and of a negative regulatory element within the first structural gene. J. Bacteriol. 173, 7481-7490. del Castillo, I., Gomez, J. M. and Moreno, F. (1990) mprA, an Escherichia coli gene that reduces growth-phase-dependent synthesis of microcins B17 and C7 and blocks osmoinduction of prof/when cloned on a high-copy-number plasmid. J. Bacteriol. 172, 437-445. del Castillo, I., Gonzalez-Pastor, J. E., San Millan, J. L. and Moreno, F. (1991) Nucleotide sequence of the Escherichia coli regulatory gene mprA and construction and characterization of m/?rv4-deficient mutants. J. Bacteriol. 173, 3924-3929. Dinnbier, U., Limpinsel, E., Schmid, R. and Bakker, E. P. (1988) Transient accumulation of potassium glutamate and its replacement by trehalose during adaptation of growing cells of Escherichia coli K-12 to elevated sodium chloride concentrations. Arch. Microbiol. 150, 348-357. Dosch, D. C, Helmer, G. L., Sutton, S. H., Salvacion, F. F. and Epstein, W. (1991) Genetic analysis of potassium transport loci in Escherichia coli: evidence for three constitutive systems mediating uptake potassium. J. Bacteriol. 173, 687-696. Dower, W. J., Miller, J. F. and Ragsdale, C. W. (1988) High efficiency transformation of E. coli by high voltage electroporation. Nucleic Acids Res. 16, 6127- 6145. 233 Dubey, A. K., Baker, C. S., Suzuki, K., Jones, A. D., Pandit, P., Romeo, T. and Babitzke, P. (2003) CsrA regulates translation of the Escherichia coli carbon starvation gene, est A, by blocking ribosome access to the est A transcript. J. Bacteriol. 185, 4450- 4460. DuBow, M. S., Ryan, T., Young, R. A. and Blumenthal, T. (1977) Host factor for coliphage Q beta RNA replication: presence in procaryotes and association with the 30S ribosomal subunit in Escherichia coli. Mol. Gen. Genet. 153, 39-43. Dulebohn, D., Choy, J., Sundermeier, T., Okan, N. and Karzai, A. W. (2007) Trans- translation: the tmRNA-mediated surveillance mechanism for ribosome rescue, directed protein degradation, and nonstop mRNA decay. Biochemistry 46, 4681-4693. Dunlap, V. J. and Csonka, L. N. (1985) Osmotic regulation of L-proline transport in Salmonella typhimurium. J. Bacteriol. 163, 296-304. Durell, S. R., Hao, Y., Nakamura, T., Bakker, E. P. and Guy, H. R. (1999) Evolutionary relationship between K(+) channels and symporters. Biophys. J. 11, 775- 788. Eddy, S. R. (2001) Non-coding RNA genes and the modern RNA world. Nat. Rev. Genet. 2,919-929. Elbein, A. D., Pan, Y. T., Pastuszak, I. and Carroll, D. (2003) New insights on trehalose: a multifunctional molecule. Glycobiology 13, 17R-27R. Eswar, N., Eramian, D., Webb, B., Shen, M. and A. Sali. (2007). Protein Structure Modeling with MODELLER. In Methods in Molecular Biology, Vol. 426: Structural Proteomics: High-throughput Methods, eds Kobe, B., Guss, M. and T. Huber. Humana Press, Totowa, NJ Faatz, E., Middendorf, A. and Bremer, E. (1988) Cloned structural genes for the osmotically regulated binding-protein-dependent glycine betaine transport system (ProU) of Escherichia coli K-12. Mol. Microbiol. 2, 265-279. Fagan, M. J. and Saier, M. H.,Jr. (1994) P-type ATPases of eukaryotes and bacteria: sequence analyses and construction of phylogenetic trees. J. Mol. Evol. 38, 57-99. Fletcher, S. A. and Csonka, L. N. (1995) Fine-structure deletion analysis of the transcriptional silencer of the pro t/operon of Salmonella typhimurium. J. Bacteriol. Ill, 4508-4513. 234 Gaballa, A., Antelmann, H., Aguilar, C, Khakh, S. K., Song, K. B., Smaldone, G. T. and Helmann, J. D. (2008) The Bacillus subtilis iron-sparing response is mediated by a Fur-regulated small RNA and three small, basic proteins. Proc. Natl. Acad. Sci. U. S. A. 105, 11927-11932. Gaber, B. P., Yager, P., Sheridan, J. P. and Chang, E. L. (1983) Encapsulation of hemoglobin in phospholipid vesicles. FEBSLett. 153, 285-288. Geissmann, T. A. and Touati, D. (2004) Hfq, a new chaperoning role: binding to messenger RNA determines access for small RNA regulator. EMBOJ. 23, 396-405. Ghetu, A. F., Gubbins, M. J., Frost, L. S. and Glover, J. N. (2000) Crystal structure of the bacterial conjugation repressor finO. Nat. Struct. Biol. 7, 565-569. Giaever, H. M., Styrvold, O. B., Kaasen, I. and Strom, A. R. (1988) Biochemical and genetic characterization of osmoregulatory trehalose synthesis in Escherichia coli. J. Bacteriol. 170, 2841-2849. Gong, S., Richard, H. and Foster, J. W. (2003) YjdE (AdiC) is the arginine:agmatine antiporter essential for arginine-dependent acid resistance in Escherichia coli. J. Bacteriol. 185, 4402-4409. Gottesman, S. (2005) Micros for microbes: non-coding regulatory RNAs in bacteria. Trends Genet. 21, 399-404. Gottesman, S. (2004) The small RNA regulators of Escherichia coli: roles and mechanisms. Annu. Rev. Microbiol. 58, 303-328. Gottesman, S., McCullen, C. A., Guillier, M., Vanderpool, C. K., Majdalani, N., Benhammou, J., Thompson, K. M., FitzGerald, P. C, Sowa, N. A. and FitzGerald, D. J. (2006) Small RNA regulators and the bacterial response to stress. Cold Spring Harb. Symp. Quant. Biol. 71, 1-11. Gouesbet, G., Jebbar, M, Talibart, R., Bernard, T. and Blanco, C. (1994) Pipecolic acid is an osmoprotectant for Escherichia coli taken up by the general osmoporters ProU and ProP. Microbiology 140 ( Pt 9), 2415-2422. Gowrishankar, J. (1989) Nucleotide sequence of the osmoregulatory pro t/operon of Escherichia coli. J. Bacteriol. 171, 1923-1931. 235 Gowrishankar, J. (1986) ProP-mediated proline transport also plays a role in Escherichia coli osmoregulation. J. Bacteriol. 166, 331-333. Gowrishankar, J., Jayashree, P. and Rajkumari, K. (1986) Molecular cloning of an osmoregulatory locus in Escherichia coli: increased proU gene dosage results in enhanced osmotolerance. J. Bacteriol. 168, 1197-1204. Gowrishankar, J. and Manna, D. (1996) How is osmotic regulation of transcription of the Escherichia coli proU operon achieved? A review and a model. Genetica 97, 363-378. Gralla, J. D. and Vargas, D. R. (2006) Potassium glutamate as a transcriptional inhibitor during bacterial osmoregulation. EMBO J. 25, 1515-1521. Grant, W. D. (2004) Life at low water activity. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 359, 1249-66; discussion 1266-7. Greie, J. C. and Altendorf, K. (2007) The K+-translocating KdpFABC complex from Escherichia coli: a P-type ATPase with unique features. J. Bioenerg. Biomembr. 39, 397- 402. Grothe, S., Krogsrud, R. L., McClellan, D. J., Milner, J. L. and Wood, J. M. (1986) Proline transport and osmotic stress response in Escherichia coli K-12. J. Bacteriol. 166, 253-259. Guzman, L. M., Belin, D., Carson, M. J. and Beckwith, J. (1995) Tight regulation, modulation, and high-level expression by vectors containing the arabinose PBAD promoter. J. Bacteriol. 177,4121-4130. Haardt, M., Kempf, B., Faatz, E. and Bremer, E. (1995) The osmoprotectant proline betaine is a major substrate for the binding-protein-dependent transport system ProU of Escherichia coli K-12. Mol. Gen. Genet. 246, 783-786. Hagiwara, D., Yamashino, T. and Mizuno, T. (2004) A Genome-wide view of the Escherichia coli BasS-BasR two-component system implicated in iron-responses. Biosci. Biotechnol. Biochem. 68, 1758-1767. Hamann, K., Zimmann, P. and Altendorf, K. (2008) Reduction of turgor is not the stimulus for the sensor kinase KdpD of Escherichia coli. J. Bacteriol. 190, 2360-2367. Hamill, O. P. and Martinac, B. (2001) Molecular basis of mechanotransduction in living cells. Physiol. Rev. 81, 685-740. 236 Hanatani, M., Yazyu, H., Shiota-Niiya, S., Moriyama, Y., Kanazawa, H., Futai, M. and Tsuchiya, T. (1984) Physical and genetic characterization of the melibiose operon and identification of the gene products in Escherichia coli. J. Biol. Chem. 259, 1807- 1812. Hantke, K. (2001) Iron and metal regulation in bacteria. Curr. Opin. Microbiol. 4, 172-177. Harms, C, Domoto, Y., Celik, C, Rahe, E., Stumpe, S., Schmid, R., Nakamura, T. and Bakker, E. P. (2001) Identification of the ABC protein SapD as the subunit that confers ATP dependence to the K+-uptake systems Trk(H) and Trk(G) from Escherichia coli K-12. Microbiology 147, 2991-3003. Heidrich, N., Chinali, A., Gerth, U. and Brant 1, S. (2006) The small untranslated RNA SRI from the Bacillus subtilis genome is involved in the regulation of arginine catabolism. Mol. Microbiol. 62, 520-536. Hengge, R. (2008) The two-component network and the general stress sigma factor RpoS (sigma S) in Escherichia coli. Adv. Exp. Med. Biol. 631, 40-53. Hengge-Aronis, R. (2002a) Recent insights into the general stress response regulatory network in Escherichia coli. J. Mol. Microbiol. Biotechnol. 4, 341-346. Hengge-Aronis, R. (2002b) Signal transduction and regulatory mechanisms involved in control of the sigma(S) (RpoS) subunit of RNA polymerase. Microbiol. Mol. Biol. Rev. 66, 373-95, table of contents. Hengge-Aronis, R., Klein, W., Lange, R., Rimmele, M. and Boos, W. (1991) Trehalose synthesis genes are controlled by the putative sigma factor encoded by rpoS and are involved in stationary-phase thermotolerance in Escherichia coli. J. Bacteriol. 173,7918-7924. Hengge-Aronis, R., Lange, R., Henneberg, N. and Fischer, D. (1993) Osmotic regulation of rpoS'-dependent genes in Escherichia coli. J. Bacteriol. 175, 259-265. Hillar, A., Culham, D. E., Vernikovska, Y. I., Wood, J. M. and Boggs, J. M. (2005) Formation of an antiparallel, intermolecular coiled coil is associated with in vivo dimerization of osmosensor and osmoprotectant transporter ProP in Escherichia coli. Biochemistry 44, 10170-10180. 237 Hinas, A., Reimegard, J., Wagner, E. G., Nellen, W., Ambros, V. R. and Soderbom, F. (2007) The small RNA repertoire of Dictyostelium discoideum and its regulation by components of the RNAi pathway. Nucleic Acids Res. 35, 6714-6726. Hofacker, I.L., Fontana, W., Stadler, P.F., Bonhoeffer, S., Tacker, M. and P. Schuster (1994). Fast Folding and Comparison of RNA Secondary Structures. Monatshefte f. Chemie 125: 167-188 Horlacher, R., Uhland, K., Klein, W., Ehrmann, M. and Boos, W. (1996) Characterization of a cytoplasmic trehalase of Escherichia coli. J. Bacteriol. 178, 6250- 6257. Houssin, C, Eynard, N., Shechter, E. and Ghazi, A. (1991) Effect of osmotic pressure on membrane energy-linked functions in Escherichia coli. Biochim. Biophys. Acta 1056, 76-84. Huntzinger, E., Boisset, S., Saveanu, C, Benito, Y., Geissmann, T., Namane, A., Lina, G., Etienne, J., Ehresmartn, B., Ehresmann, C, Jacquier, A., Vandenesch, F. and Romby, P. (2005) Staphylococcus aureus RNAIII and the endoribonuclease III coordinately regulate spa gene expression. EMBOJ. 24, 824-835. Jerome, L. J., van Biesen, T. and Frost, L. S. (1999) Degradation of FinP antisense RNA from F-like plasmids: the RNA-binding protein, FinO, protects FinP from ribonuclease E. J. Mol. Biol. 285, 1457-1473. Jiang, M., Sullivan, S. M., Walker, A. K., Strahler, J. R., Andrews, P. C. and Maddock, J. R. (2007) Identification of novel Escherichia coli ribosome-associated proteins using isobaric tags and multidimensional protein identification techniques. J. Bacteriol. 189, 3434-3444. Jishage, M. and Ishihama, A. (1995) Regulation of RNA polymerase sigma subunit synthesis in Escherichia coli: intracellular levels of sigma 70 and sigma 38. J. Bacteriol. Ill, 6832-6835. Jordi, B. J. and Higgins, C. F. (2000) The downstream regulatory element of the proU operon of Salmonella typhimurium inhibits open complex formation by RNA polymerase at a distance. J. Biol. Chem. 275, 12123-12128. 238 Jovanovich, S. B., Martinell, M., Record, M. T.,Jr and Burgess, R. R. (1988) Rapid response to osmotic upshift by osmoregulated genes in Escherichia coli and Salmonella typhimurium. J. Bacteriol 170, 534-539. Jovanovich, S. B., Record, M. T.,Jr and Burgess, R. R. (1989) In an Escherichia coli coupled transcription-translation system, expression of the osmoregulated gene proU is stimulated at elevated potassium concentrations and by an extract from cells grown at high osmolality. J. Biol. Chem. 264, 7821-7825. Kaasen, I., McDougall, J. and Strom, A. R. (1994) Analysis of the otsBA operon for osmoregulatory trehalose synthesis in Escherichia coli and homology of the OtsA and OtsB proteins to the yeast trehalose-6-phosphate synthase/phosphatase complex. Gene 145,9-15. Kelley, L. A., MacCallum, R. M. and Sternberg, M. J. (2000) Enhanced genome annotation using structural profiles in the program 3D-PSSM. J. Mol. Biol. 299, 499-520. Keseler, I. M., Collado-Vides, J., Gama-Castro, S., Ingraham, J., Paley, S., Paulsen, I. T., Peralta-Gil, M. and Karp, P. D. (2005) EcoCyc: a comprehensive database resource for Escherichia coli. Nucleic Acids Res. 33, D334-7. Klauck, E., Typas, A. and Hengge, R. (2007) The sigmaS subunit of RNA polymerase as a signal integrator and network master regulator in the general stress response in Escherichia coli. Sci. Prog. 90, 103-127. Koraimann, G., Teferle, K., Markolin, G., Woger, W. and Hogenauer, G. (1996) The FinOP repressor system of plasmid Rl: analysis of the antisense RNA control of traJ expression and conjugative DNA transfer. Mol. Microbiol. 21, 811-821. Kunte, H. J., Crane, R. A., Culham, D. E., Richmond, D. and Wood, J. M. (1999) Protein ProQ influences osmotic activation of compatible solute transporter ProP in Escherichia coli K-12. J. Bacteriol. 181, 1537-1543. Laemmli, U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685. Lamark, T., Rokenes, T. P., McDougall, J. and Strom, A. R. (1996) The complex bet promoters of Escherichia coli: regulation by oxygen (ArcA), choline (Betl), and osmotic stress. J. Bacteriol. 178, 1655-1662. 239 Landis, L., Xu, J. and Johnson, R. C. (1999) The cAMP receptor protein CRP can function as an osmoregulator of transcription in Escherichia coli. Genes Dev. 13, 3081- 3091. Lange, R. and Hengge-Aronis, R. (1991). Identification of a central regulator of stationart phase gene expression in Escherichia coli. Mol. Microbiol. 5, 49-59. Larrainzar, E., O'Gara, F. and J.P. Morrissey (2005). Application of qutofluorescent proteins for in situ studies in microbial ecology. Annu Rev. Microbiol. 59:257-77 Lawley, T. D., Klimke, W. A., Gubbins, M. J. and Frost, L. S. (2003) F factor conjugation is a true type IV secretion system. FEMS Microbiol. Lett. 224, 1-15. Lemieux, M. J., Huang, Y. and Wang da, N. (2005) Crystal structure and mechanism of GlpT, the glycerol-3-phosphate transporter from E. coli. J. Electron. Microsc. (Tokyo) 54 Suppl 1, i43-6. Levina, N., Totemeyer, S., Stokes, N. R., Louis, P., Jones, M. A. and Booth, I. R. (1999) Protection of Escherichia coli cells against extreme turgor by activation of MscS and MscL mechanosensitive channels: identification of genes required for MscS activity. EMBOJ. 18, 1730-1737. Link, A. J., Phillips, D. and Church, G. M. (1997) Methods for generating precise deletions and insertions in the genome of wild-type Escherichia coli: application to open reading frame characterization. J. Bacteriol. 119, 6228-6237. Liu, F., Culham, D. E., Vernikovska, Y. I., Keates, R. A., Boggs, J. M. and Wood, J. M. (2007) Structure and function of transmembrane segment XII in osmosensor and osmoprotectant transporter ProP of Escherichia coli. Biochemistry 46, 5647-5655. Liu, M. Y., Gui, G., Wei, B., Preston, J. F.,3rd, Oakford, L., Yuksel, U., Giedroc, D. P. and Romeo, T. (1997) The RNA molecule CsrB binds to the global regulatory protein CsrA and antagonizes its activity in Escherichia coli. J. Biol. Chem. 272, 17502-17510. Liu, M. Y. and Romeo, T. (1997) The global regulator CsrA of Escherichia coli is a specific mRNA-binding protein. J. Bacteriol. 179,4639-4642. Lomovskaya, O., Lewis, K. and Matin, A. (1995) EmrR is a negative regulator of the Escherichia coli multidrug resistance pump EmrAB. J. Bacteriol. Ill, 2328-2334. 240 Lucht, J. M. and Bremer, E. (1991) Characterization of mutations affecting the osmoregulated/>ro £/promoter of Escherichia coli and identification of 5' sequences required for high-level expression. J. Bacteriol. 173, 801-809. Lucht, J. M., Dersch, P., Kempf, B. and Bremer, E. (1994) Interactions of the nucleoid-associated DNA-binding protein H-NS with the regulatory region of the osmotically controlled pro t/operon of Escherichia coli. J. Biol. Chem. 269, 6578-6578. Ly, A., Henderson, J., Lu, A., Culham, D. E. and Wood, J. M. (2004) Osmoregulatory systems of Escherichia coli: identification of betaine-carnitine-choline transporter family member BetU and distributions of betU and trkG among pathogenic and nonpathogenic isolates. J. Bacteriol. 186,296-306. MacMillan, S. V., Alexander, D. A., Culham, D. E., Kunte, H. J., Marshall, E. V., Rochon, D. and Wood, J. M. (1999) The ion coupling and organic substrate specificities of osmoregulatory transporter ProP in Escherichia coli. Biochim. Biophys. Acta 1420, 30- 44. Majdalani, N., Chen, S., Murrow, J., St John, K. and Gottesman, S. (2001) Regulation of RpoS by a novel small RNA: the characterization of RprA. Mol. Microbiol. 39, 1382- 1394. Majdalani, N., Hernandez, D. and Gottesman, S. (2002) Regulation and mode of action of the second small RNA activator of RpoS translation, RprA. Mol. Microbiol. 46, 813-826. Manna, D. and Gowrishankar, J. (1994) Evidence for involvement of proteins HU and RpoS in transcription of the osmoresponsive proU operon in Escherichia coli. J. Bacteriol. 176, 5378-5384. Martinac, B. (2001) Mechanosensitive channels in prokaryotes. Cell. Physiol. Biochem. 11, 61-76. Masse, E., Escorcia, F. E. and Gottesman, S. (2003) Coupled degradation of a small regulatory RNA and its mRNA targets in Escherichia coli. Genes Dev. 17, 2374-2383. Masse, E. and Gottesman, S. (2002) A small RNA regulates the expression of genes involved in iron metabolism in Escherichia coli. Proc. Natl. Acad. Sci. U. S. A. 99, 4620- 4625. 241 Masse, E., Vanderpool, C. K. and Gottesman, S. (2005) Effect of RyhB small RNA on global iron use in Escherichia coli. J. Bacteriol. 187, 6962-6971. Maxwell, K. L. and Davidson, A. R. (1998) Mutagenesis of a buried polar interaction in an SH3 domain: sequence conservation provides the best prediction of stability effects. Biochemistry 37, 16172-16182. May, G., Dersch, P., Haardt, M., Middendorf, A. and Bremer, E. (1990) The osmZ (bglY) gene encodes the DNA-binding protein H-NS (HI a), a component of the Escherichia coli K12 nucleoid. Mol. Gen. Genet. ll'X, 81-90. May, G., Faatz, E., Lucht, J. M., Haardt, M., Bolliger, M. and Bremer, E. (1989) Characterization of the osmoregulated Escherichia coli proU promoter and identification of Pro V as a membrane-associated protein. Mol. Microbiol. 3, 1521-1531. May, G., Faatz, E., Villarejo, M. and Bremer, E. (1986) Binding protein dependent transport of glycine betaine and its osmotic regulation in Escherichia coli K12. Mol. Gen. Genet. 205, 225-233. Mayer, B. J. (2001) SH3 domains: complexity in moderation. J. Cell. Sci. 114, 1253- 1263. McLeod, S. M., Xu, J., Cramton, S. E., Gaal, T., Gourse, R. L. and Johnson, R. C. (1999) Localization of amino acids required for Fis to function as a class II transcriptional activator at the RpoS-dependent proP P2 promoter. J. Mol. Biol. 294, 333- 346. McNealy, T. L., Forsbach-Birk, V., Shi, C. and Marre, R. (2005) The Hfq homolog in Legionella pneumophila demonstrates regulation by LetA and RpoS and interacts with the global regulator CsrA. J. Bacteriol. 187, 1527-1532. McPherson, P. S. (1999) Regulatory role of SH3 domain-mediated protein-protein interactions in synaptic vesicle endocytosis. Cell. Signal. 11, 229-238. Mellies, J., Wise, A. and Villarejo, M. (1995) Two different Escherichia coliproP promoters respond to osmotic and growth phase signals. J. Bacteriol. Ill, 144-151. Metcalf, W. W. and Wanner, B. L. (1993) Evidence for a fourteen-gene, phnC to phnP locus for phosphonate metabolism in Escherichia coli. Gene 129, 27-32. 242 Meury, J. (1994) Immediate and transient inhibition of the respiration of Escherichia coli under hyperosmotic shock. FEMS Microbiol. Lett. 121, 281-286. Meury, J. (1988) Glycine betaine reverses the effects of osmotic stress on DNA replication and cellular division in Escherichia coli. Arch. Microbiol. 149, 232-239. Mikulecky, P. J., Kaw, M. K., Brescia, C. C, Takach, J. C, Sledjeski, D. D. and Feig, A. L. (2004) Escherichia coli Hfq has distinct interaction surfaces for DsrA, rpoS and poly(A) RNAs. Nat. Struct. Mol. Biol. 11, 1206-1214. Mileykovskaya, E., Dowhan, W., Birke, R. L., Zheng, D., Lutterodt, L. and Haines, T. H. (2001) Cardiolipin binds nonyl acridine orange by aggregating the dye at exposed hydrophobic domains on bilayer surfaces. FEBS Lett. 507, 187-190. Miller, J. H. (1972) Experiments in Molecular Genetics Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Milner, J. L., Grothe, S. and Wood, J. M. (1988) Proline porter II is activated by a hyperosmotic shift in both whole cells and membrane vesicles of Escherichia coli K12. J. Biol. Chem. 263, 14900-14905. Milner, J. L. and Wood, J. M. (1989) InsertionproQ220::Tn5 alters regulation of proline porter II, a transporter of proline and glycine betaine in Escherichia coli. J. Bacteriol. 171,947-951. Moll, I., Leitsch, D., Steinhauser, T. and Blasi, U. (2003) RNA chaperone activity of the Sm-like Hfq protein. EMBO Rep. 4, 284-289. Moller, T., Franch, T., Hojrup, P., Keene, D. R., Bachinger, H. P., Brennan, R. G. and Valentin-Hansen, P. (2002) Hfq: a bacterial Sm-like protein that mediates RNA-RNA interaction. Mol. Cell 9, 23-30. Moody, J. E., Millen, L., Binns, D., Hunt, J. F. and Thomas, P. J. (2002) Cooperative, ATP-dependent association of the nucleotide binding cassettes during the catalytic cycle of ATP-binding cassette transporters. J. Biol. Chem. 277, 21111-21114. Morbach, S. and Kramer, R. (2002) Body shaping under water stress: osmosensing and osmoregulation of solute transport in bacteria. Chembiochem 3, 384-397. 243 Muffler, A., Fischer, D. and Hengge-Aronis, R. (1996) The RNA-binding protein HF- I, known as a host factor for phage Qbeta RNA replication, is essential for rpoS translation in Escherichia coli. Genes Dev. 10, 1143-1151. Nagarajavel, V., Madhusudan, S., Dole, S., Rahmouni, A. R. and Schnetz, K. (2007) Repression by binding of H-NS within the transcription unit. J. Biol. Chem. 282, 23622- 23630. Nakamaru, Y., Takahashi, Y., Unemoto, T. and Nakamura, T. (1999) Mechanosensitive channel functions to alleviate the cell lysis of marine bacterium, Vibrio alginolyticus, by osmotic downshock. FEBS Lett. 444, 170-172. Neidhardt, F. C, Bloch, P. L. and Smith, D. F. (1974) Culture medium for enterobacteria. J. Bacteriol. 119, 736-747. Nielsen, J. S., Boggild, A., Andersen, C. B., Nielsen, G., Boysen, A., Brodersen, D. E. and Valentin-Hansen, P. (2007) An Hfq-like protein in archaea: crystal structure and functional characterization of the Sm protein from Methanococcus jannaschii. RNA 13, 2213-2223. Nikulin, A., Stolboushkina, E., Perederina, A., Vassilieva, I., Blaesi, U., Moll, I., Kachalova, G., Yokoyama, S., Vassylyev, D., Garber, M. and Nikonov, S. (2005) Structure of Pseudomonas aeruginosa Hfq protein. Acta Crystallogr. D Biol. Crystallogr. 61, 141-146. Numata, T., Ishimatsu, I., Kakuta, Y., Tanaka, I. and Kimura, M. (2004) Crystal structure of archaeal ribonuclease P protein Phl771p from Pyrococcus horikoshii OT3: an archaeal homolog of eukaryotic ribonuclease P protein Rpp29. RNA 10,1423-1432. Overdier, D. G. and Csonka, L. N. (1992) A transcriptional silencer downstream of the promoter in the osmotically controlled pro t/operon of Salmonella typhimurium. Proc. Natl. Acad. Sci. U. S A. 89, 3140-3144. Overdier, D. G., Olson, E. R., Erickson, B. D., Ederer, M. M. and Csonka, L. N. (1989) Nucleotide sequence of the transcriptional control region of the osmotically regulated pro t/operon of Salmonella typhimurium and identification of the 5' endpoint of the proU mRNA. J. Bacteriol. 171, 4694-4706. 244 Parra-Lopez, C, Lin, R., Aspedon, A. and Groisman, E. A. (1994) A Salmonella protein that is required for resistance to antimicrobial peptides and transport of potassium. EMBO J. 13, 3964-3972. Perry, J. Solubility Studies: Preparation of soluble/insoluble protein from cells. 2008, Retrieved December 2008, from http://www.embl- heidelberg.de/ExternalInfo/protein_unit/draft_frames_save/frame_extract_clar_ext.html. Peter, H., Weil, B., Burkovski, A., Kramer, R. and Morbach, S. (1998) Corynebacterium glutamicum is equipped with four secondary carriers for compatible solutes: identification, sequencing, and characterization of the proline/ectoine uptake system, ProP, and the ectoine/proline/glycine betaine carrier, EctP. J. Bacteriol. 180, 6005-6012. Polarek, J. W., Williams, G. and Epstein, W. (1992) The products of the kdpDE operon are required for expression of the Kdp ATPase of Escherichia coli. J. Bacteriol. 174,2145-2151. Poolman, B., Blount, P., Folgering, J. H., Friesen, R. H., Moe, P. C. and van der Heide, T. (2002) How do membrane proteins sense water stress? Mol. Microbiol. 44, 889-902. Poolman, B., Spitzer, J. J. and Wood, J. M. (2004) Bacterial osmosensing: roles of membrane structure and electrostatics in lipid-protein and protein-protein interactions. Biochim. Biophys. Acta 1666, 88-104. Prince, W. S. and Villarejo, M. R. (1990) Osmotic control ofprolJtranscription is mediated through direct action of potassium glutamate on the transcription complex. J. Biol. Chem. 265, 17673-17679. Racher, K. I., Culham, D. E. and Wood, J. M. (2001) Requirements for osmosensing and osmotic activation of transporter ProP from Escherichia coli. Biochemistry 40, 7324- 7333. Racher, K. I., Voegele, R. T., Marshall, E. V., Culham, D. E., Wood, J. M., Jung, H., Bacon, M., Cairns, M. T., Ferguson, S. M., Liang, W. J., Henderson, P. J., White, G. and Hallett, F. R. (1999) Purification and reconstitution of an osmosensor: transporter ProP of Escherichia coli senses and responds to osmotic shifts. Biochemistry 38, 1676-1684. 245 Rajkumari, K. and Gowrishankar, J. (2001) In vivo expression from the RpoS- dependent PI promoter of the osmotically regulated proU operon in Escherichia coli and Salmonella enterica serovar Typhimurium: activation by rho and hns mutations and by cold stress. J. Bacteriol. 183, 6543-6550. Rajkumari, K., Kusano, S., Ishihama, A., Mizuno, T. and Gowrishankar, J. (1996) Effects of H-NS and potassium glutamate on sigmaS- and sigma70-directed transcription in vitro from osmotically regulated PI and P2 promoters of proU in Escherichia coli. J. Bacteriol. 178,4176-4181. Record, M. T.,Jr, Courtenay, E. S., Cayley, D. S. and Guttman, H. J. (1998) Responses of E. coli to osmotic stress: large changes in amounts of cytoplasmic solutes and water. Trends Biochem. Sci. 23, 143-148. Repoila, F. and Darfeuille, F. (2009) Small regulatory non-coding RNAs in bacteria: physiology and mechanistic aspects. Biol. Cell. 101, 117-131. Repoila, F. and Gottesman, S. (2001) Signal transduction cascade for regulation of RpoS: temperature regulation of DsrA. J. Bacteriol. 183, 4012-4023. Repoila, F., Majdalani, N. and Gottesman, S. (2003) Small non-coding RNAs, co ordinators of adaptation processes in Escherichia coli: the RpoS paradigm. Mol. Microbiol. 48, 855-861. Rhoads, D. B. and Epstein, W. (1977) Energy coupling to net K+ transport in Escherichia coli K-12. J. Biol. Chem. 252, 1394-1401. Riley, M., Abe, T., Arnaud, M. B., Berlyn, M. K., Blattner, F. R., Chaudhuri, R. R., Glasner, J. D., Horiuchi, T., Keseler, I. M., Kosuge, T., Mori, H., Perna, N. T., Plunkett, G.,3rd, Rudd, K. E., Serres, M. H., Thomas, G. H., Thomson, N. R., Wishart, D. and Wanner, B. L. (2006) Escherichia coli K-12: a cooperatively developed annotation snapshot—2005. Nucleic Acids Res. 34, 1-9. Rimmele, M. and Boos, W. (1994) Trehalose-6-phosphate hydrolase of Escherichia coli. J. Bacteriol. 176, 5654-5664. Romantsov, T., Helbig, S., Culham, D. E., Gill, C, Stalker, L. and Wood, J. M. (2007) Cardiolipin promotes polar localization of osmosensory transporter ProP in Escherichia coli. Mol. Microbiol. 64, 1455-1465. 246 Romantsov, T., Stalker, L., Culham, D. E. and Wood, J. M. (2008) Cardiolipin controls the osmotic stress response and the subcellular location of transporter ProP in Escherichia coli. J. Biol. Chem. 283, 12314-12323. Romeo, T., Gong, M., Liu, M. Y. and Brun-Zinkernagel, A. M. (1993) Identification and molecular characterization of csrA, a pleiotropic gene from Escherichia coli that affects glycogen biosynthesis, gluconeogenesis, cell size, and surface properties. J. Bacteriol. 175, 4744-4755. Sabnis, N. A., Yang, H. and Romeo, T. (1995) Pleiotropic regulation of central carbohydrate metabolism in Escherichia coli via the gene csrA. J. Biol. Chem. 270, 29096-29104. Saier, M. H.,Jr. (2000) Families of transmembrane transporters selective for amino acids and their derivatives. Microbiology 146 ( Pt 8), 1775-1795. Sambrook, J. and Russell, D. W. (2001) Molecular Cloning, a Laboratory Manual, 3rd ed., Cold Spring Harbour, Cold Spring Harbour, N.Y. Sauter, C, Basquin, J. and Suck, D. (2003) Sm-like proteins in Eubacteria: the crystal structure of the Hfq protein from Escherichia coli. Nucleic Acids Res. 31, 4091-4098. Schaffner, W. and Weissmann, C. (1973) A rapid, sensitive, and specific method for the determination of protein in dilute solution. Anal. Biochem. 56, 502-514. Schagger, H. and von Jagow, G. (1987) Tricine-sodium dodecyl sulfate- polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal. Biochem. 166, 368-379. Schlosser, A., Hamann, A., Bossemeyer, D., Schneider, E. and Bakker, E. P. (1993) NAD+ binding to the Escherichia coli K(+)-uptake protein TrkA and sequence similarity between TrkA and domains of a family of dehydrogenases suggest a role for NAD+ in bacterial transport. Mol. Microbiol. 9, 533-543. Schlosser, A., Meldorf, M., Stumpe, S., Bakker, E. P. and Epstein, W. (1995) TrkH and its homo log, TrkG, determine the specificity and kinetics of cation transport by the Trk system of Escherichia coli. J. Bacteriol. Ill, 1908-1910. 247 Schumacher, M. A., Pearson, R. F., Moller, T., Valentin-Hansen, P. and Brennan, R. G. (2002) Structures of the pleiotropic translational regulator Hfq and an Hfq-RNA complex: a bacterial Sm-like protein. EMBOJ. 21, 3546-3556. Scofield, D. G. and Lynch, M. (2008) Evolutionary diversification of the Sm family of RNA-associated proteins. Mol. Biol. Evol. 25, 2255-2267. Senear, A. W. and Steitz, J. A. (1976) Site-specific interaction of Qbeta host factor and ribosomal protein SI with Qbeta and R17 bacteriophage RNAs. J. Biol. Chem. 251, 1902-1912. Seol, W. and Shatkin, A. J. (1993) Membrane topology model of Escherichia coli alpha-ketoglutarate permease byphoA fusion analysis. J. Bacteriol. 175, 565-567. Sidote, D. J., Heideker, J. and Hoffman, D. W. (2004) Crystal structure of archaeal ribonuclease P protein aRpp29 from Archaeoglobus fulgidus. Biochemistry 43, 14128- 14138. Sledjeski, D. D., Gupta, A. and Gottesman, S. (1996) The small RNA, DsrA, is essential for the low temperature expression of RpoS during exponential growth in Escherichia coli. EMBOJ. 15, 3993-4000. Smith, M. N., Crane, R. A., Keates, R. A. and Wood, J. M. (2004) Overexpression, purification, and characterization of ProQ, a posttranslational regulator for osmoregulatory transporter ProP of Escherichia coli. Biochemistry 43, 12979-12989. Smith, M. N., Kwok, S. C, Hodges, R. S. and Wood, J. M. (2007) Structural and functional analysis of ProQ: an osmoregulatory protein of Escherichia coli. Biochemistry 46, 3084-3095. Smith, P. C, Karpowich, N., Millen, L., Moody, J. E., Rosen, J., Thomas, P. J. and Hunt, J. F. (2002) ATP binding to the motor domain from an ABC transporter drives formation of a nucleotide sandwich dimer. Mol. Cell 10, 139-149. Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H., Provenzano, M. D., Fujimoto, E. K., Goeke, N. M., Olson, B. J. and Klenk, D. C. (1985) Measurement of protein using bicinchoninic acid. Anal. Biochem. 150, 76-85. 248 Sonnleitner, E., Schuster, M., Sorger-Domenigg, T., Greenberg, E. P. and Blasi, U. (2006) Hfq-dependent alterations of the transcriptome profile and effects on quorum sensing in Pseudomonas aeruginosa. Mol. Microbiol. 59, 1542-1558. Spedding, G., Strader, M. and Robinette, D. (2008) Protocol to extract ribosomal proteins for LC/MS analysis. Retrieved November 2008 From: http://bioinfo.med.unc.edu/glabwiki/index.php/Protocol_to_Extract_Ribosomal_Proteins _for_LC/MS_analysis. Stalmach, M. E., Grothe, S. and Wood, J. M. (1983) Two proline porters in Escherichia coli K-12. J. Bacteriol. 156, 481-486. Stenberg, F., Chovanec, P., Maslen, S. L., Robinson, C. V., Hag, L. L., von Heijne, G. and Daley, D. O. (2005) Protein complexes of the Escherichia coli cell envelope. J. Biol. Chem. 280, 34409-34419. Stewart, L. M., Bakker, E. P. and Booth, I. R. (1985) Energy coupling to K+ uptake via the Trk system in Escherichia coli: the role of ATP. J. Gen. Microbiol. 131, 77-85. Stim, K. P. and Bennett, G. N. (1993) Nucleotide sequence of the adi gene, which encodes the biodegradative acid-induced arginine decarboxylase of Escherichia coli. J. Bacteriol. 175, 1221-1234. Stim-Herndon, K. P., Flores, T. M. and Bennett, G. N. (1996) Molecular characterization of adiY, a regulatory gene which affects expression of the biodegradative acid-induced arginine decarboxylase gene (adiA) of Escherichia coli. Microbiology 142 ( Pt5), 1311-1320. Stirling, D. A., Hulton, C. S., Waddell, L., Park, S. F., Stewart, G. S., Booth, I. R. and Higgins, C. F. (1989) Molecular characterization of the proU loci of Salmonella typhimurium and Escherichia coli encoding osmoregulated glycine betaine transport systems. Mol. Microbiol. 3, 1025-1038. Storz, G. (2002) An expanding universe of noncoding RNAs. Science 296, 1260- 1263. Storz, G., Altuvia, S. and Wassarman, K. M. (2005) An abundance of RNA regulators. Annu. Rev. Biochem. 74, 199-217. 249 Storz, G., Opdyke, J. A. and Zhang, A. (2004) Controlling mRNA stability and translation with small, noncoding RNAs. Curr. Opin. Microbiol. 7, 140-144. Stumpe, S., Schlosser, A., Schleyer, M. and Bakker, E. P. (1996) K+ circulation across the prokaryotic membrane: K+ uptake systems; in Handbook of Biological Physics Volume 2: Transport Processes in Eukaryotic and Prokaryoptic Organelles, Konings, W. N., Kaback, H. R. and Lo llama, J. S. (eds.), pp. 473-500, Enslaver, New York. Sukharev, S. I., Blount, P., Martinac, B., Blattner, F. R. and Kung, C. (1994) A large- conductance mechanosensitive channel in E. coli encoded by mscL alone. Nature 368, 265-268. Sun, X. and Wartell, R. M. (2006) Escherichia coli Hfq binds A18 and DsrA domain II with similar 2:1 Hfq6/RNA stoichiometry using different surface sites. Biochemistry 45, 4875-4887. Sun, X., Zhulin, I. and Wartell, R. M. (2002) Predicted structure and phyletic distribution of the RNA-binding protein Hfq. Nucleic Acids Res. 30, 3662-3671. Sutherland, L., Cairney, J., Elmore, M. J., Booth, I. R. and Higgins, C. F. (1986) Osmotic regulation of transcription: induction of the pro C/betaine transport gene is dependent on accumulation of intracellular potassium. J. Bacteriol. 168, 805-814. Suzuki, K., Wang, X., Weilbacher, T., Pernestig, A. K., Melefors, O., Georgellis, D., Babitzke, P. and Romeo, T. (2002) Regulatory circuitry of the CsrA/CsrB and BarA/UvrY systems of Escherichia coli. J. Bacteriol. 184, 5130-5140. Thomas, J. R. and Hergenrother, P. J. (2008) Targeting RNA with small molecules. Chem.Rev. 108, 1171-1224. Toledo-Arana, A., Repoila, F. and Cossart, P. (2007) Small noncoding RNAs controlling pathogenesis. Curr. Opin. Microbiol. 10, 182-188. Towbin, H., Staehelin, T. and Gordon, J. (1992) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. 1979. Biotechnology 24,145-149. Tsatskis, Y., Khambati, J., Dobson, M., Bogdanov, M, Dowhan, W. and Wood, J. M. (2005) The osmotic activation of transporter ProP is tuned by both its C-terminal coiled- 250 coil and osmotically induced changes in phospholipid composition. J. Biol. Chem. 280, 41387-41394. Tsatskis, Y., Kwok, S. C, Becker, E., Gill, C, Smith, M. N., Keates, R. A., Hodges, R. S. and Wood, J. M. (2008) Core residue replacements cause coiled-coil orientation switching in vitro and in vivo: structure-function correlations for osmosensory transporter ProP. Biochemistry 47, 60-72. Tsui, H. C, Leung, H. C. and Winkler, M. E. (1994) Characterization of broadly pleiotropic phenotypes caused by an hfq insertion mutation in Escherichia coli K-12. Mol. Microbiol 13, 35-49. Ueguchi, C. and Mizuno, T. (1993) The Escherichia coli nucleoid protein H-NS functions directly as a transcriptional repressor. EMBO J. 12, 1039-1046. Valentin-Hansen, P., Eriksen, M. and Udesen, C. (2004a) The bacterial Sm-like protein Hfq: a key player in RNA transactions. Mol. Microbiol. 51, 1525-1533. Valentin-Hansen, P., Eriksen, M. and Udesen, C. (2004b) The bacterial Sm-like protein Hfq; a key player in RNA transactions. Mol. Microbiol. 51, 1525-33. van Biesen, T. and Frost, L. S. (1994) The FinO protein of IncF plasmids binds FinP antisense RNA and its target, traJ mRNA, and promotes duplex formation. Mol. Microbiol. 14, 427-436. Vanderpool, C. K. (2007) Physiological consequences of small RNA-mediated regulation of glucose-phosphate stress. Curr. Opin. Microbiol. 10, 146-151. Vanderpool, C. K. and Gottesman, S. (2004) Involvement of a novel transcriptional activator and small RNA in post-transcriptional regulation of the glucose phosphoenolpyruvate phosphotransferase system. Mol. Microbiol. 54, 1076-1089. Vassinova, N. and Kozyrev, D. (2000) A method for direct cloning of/wr-regulated genes: identification of seven new/wr-regulated loci in Escherichia coli. Microbiology 146 Pt 12, 3171-3182. Voelkner, P., Puppe, W. and Altendorf, K. (1993) Characterization of the KdpD protein, the sensor kinase of the K(+)-translocating Kdp system of Escherichia coli. Eur. J. Biochem. 217, 1019-1026. 251 Vogel, J. and Papenfort, K. (2006) Small non-coding RNAs and the bacterial outer membrane. Curr. Opin. Microbiol. 9, 605-611. Vogel, J. and Wagner, E. G. (2007) Target identification of small noncoding RNAs in bacteria. Curr. Opin. Microbiol. 10, 262-270. Wadler, C. S. and Vanderpool, C. K. (2007) A dual function for a bacterial small RNA: SgrS performs base pairing-dependent regulation and encodes a functional polypeptide. Proc. Natl. Acad. Sci. U. S. A. 104, 20454-20459. Walderhaug, M. O., Polarek, J. W., Voelkner, P., Daniel, J. M., Hesse, J. E., Altendorf, K. and Epstein, W. (1992) KdpD and KdpE, proteins that control expression of the kdpABC operon, are members of the two-component sensor-effector class of regulators. J. Bacteriol. 174, 2152-2159. Whipp, M. J., Camakaris, H. and Pittard, A. J. (1998) Cloning and analysis of the shiA gene, which encodes the shikimate transport system of Escherichia coli K-12. Gene 209, 185-192. Wood, J. M. (2007) Bacterial osmosensing transporters. Methods Enzymol. 428, 77- 107. Wood, J. M. (1999) Osmosensing by bacteria: signals and membrane-based sensors. Microbiol. Mol. Biol. Rev. 63, 230-262. Wood, J. M. (1988) Proline porters effect the utilization of proline as nutrient or osmoprotectant for bacteria. J. Membr. Biol. 106, 183-202. Wood, J. M., Bremer, E., Csonka, L. N., Kraemer, R., Poolman, B., van der Heide, T. and Smith, L. T. (2001) Osmosensing and osmoregulatory compatible solute accumulation by bacteria. Comp. Biochem. Physiol. A. Mol. Integr. Physiol. 130, 437- 460. Wood, J. M., Culham, D. E., Hillar, A., Vernikovska, Y. I., Liu, F., Boggs, J. M. and Keates, R. A. (2005) A structural model for the osmosensor, transporter, and osmoregulator ProP of Escherichia coli. Biochemistry 44, 5634-5646. Wood, J. M. and Zadworny, D. (1980) Amplification of the put genes and identification of the put gene products in Escherichia coli K12. Can. J. Biochem. 58, 787- 796. 252 Wower, J., Wower, I. K. and Zwieb, C. (2008) Making the jump: new insights into the mechanism of trans-translation. J. Biol. 7, 17. Xu, J. and Johnson, R. C. (1997a) Activation of RpoS-dependent/?raP P2 transcription by the Fis protein in vitro. J. Mol. Biol. 270, 346-359. Xu, J. and Johnson, R. C. (1997b) Cyclic AMP receptor protein functions as a repressor of the osmotically inducible promoter proP PI in Escherichia coli. J. Bacteriol. 179,2410-2417. Xu, J. and Johnson, R. C. (1995a) Fis activates the RpoS-dependent stationary-phase expression of proP in Escherichia coli. J. Bacteriol. Ill, 5222-5231. Xu, J. and Johnson, R. C. (1995b) Identification of genes negatively regulated by Fis: Fis and RpoS co-modulate growth-phase-dependent gene expression in Escherichia coli. J. Bacteriol. Ill, 938-947. Yan, D., Ikeda, T.P., Shauger, A.E. and Kustu, S. (1996) Glutamate is required to maintain the steady-state potassium pool in Salmonella typhimurium. Proc. Natl. Acad. Sci. USA 93, 6527-31. Zhang, A., Altuvia, S. and Storz, G. (1997) The novel OxyS RNA regulates expression of the sigma s subunit of Escherichia coli RNA polymerase. Nucleic Acids Symp. Ser. 36, 27-28. Zhang, A., Altuvia, S., Tiwari, A., Argaman, L., Hengge-Aronis, R. and Storz, G. (1998) The OxyS regulatory RNA represses rpoS translation and binds the Hfq (HF-I) protein. EMBOJ. 17, 6061-6068. Zhang, A., Wassarman, K. M., Ortega, J., Steven, A. C. and Storz, G. (2002) The Sm- like Hfq protein increases OxyS RNA interaction with target mRNAs. Mol. Cell 9, 11-22. Zhang, W. and Chait, B. T. (2000) ProFound: an expert system for protein identification using mass spectrometric peptide mapping information. Anal. Chem. 72, 2482-2489. Zhang, X., Fletcher, S. A. and Csonka, L. N. (1996) Site-directed mutational analysis of the osmotically regulated pro U promoter of Salmonella typhimurium. J. Bacteriol. 178, 3377-3379. 253 Zhao, Y., Li, H., Hou, Y., Cha, L., Cao, Y., Wang, L., Ying, X. and W. Li (2008).Construction of two mathematical models for prediction of bacterial sRNA targets. Biochem Biophys Res Commun. 2008 Jul 25;372(2):346-50. Zoetewey, D. L., Tripet, B. P., Kutateladze, T. G., Overduin, M. J., Wood, J. M. and Hodges, R. S. (2003) Solution structure of the C-terminal antiparallel coiled-coil domain from Escherichia coli osmosensor ProP. J. Mol. Biol. 334, 1063-1076. 254 Appendix 1: Growth media LB medium (Miller. 1972) LB medium contained 10 g peptone from casein (tryptone), 10 g NaCl and 5 g Bacto-yeast extract in 1 L distilled or Milli-Q Water and the medium was sterilized by autoclaving. For LB medium agar, 15 g of agar was added prior to autoclaving. For LB agar containing antibiotics, aliquots of filter sterilized stock solutions of antibiotics were aseptically added, following sterilization, to a final concentration of: 50 (ig/mL of kanamycin, 30 |ag/mL of chloramphenicol, 100 (ig/mL of streptomycin, 100 (ig/mL of ampicillin and 50 (xg/mL of tetracycline. MOPS buffered minimal medium (Neidhardt et al. 1974s) MOPS (3-[N-morpholino] propanesulfonic acid) buffered minimal medium was prepared as described by Neidhardt et al, 1974, except NaCl was not added to the MOPS lOx concentrate base. Concentrations of the constituents of 10 x MOPS medium base are as follows: 40 mM MOPS, 4 mM Tricine, 0.01 mM FeS04-7H20, 0.276 mM K2S04, 5 x 4 4 5 10" mM CaCl2-2H20, 0.528 mM MgCl2-6H20, 4 x 10" mM H3BO3, 3 x 10" mM CoCl2, 5 5 6 5 1 x 10" mM CuS04, 1 x 10" mM ZnS04, 3 x 10" mM (NH4)6(Mo07)24, 8 x 10" mM 6 4 5 MnSO-4H20, 3 x 10" mM (NH4)6(Mo7)24, 4 x 10" mM H3BO3, 3 x 10" mM CoCl2, 1 x 5 5 5 10" mM CuS04, 8 x 10" mM MnS04, 1 x 10" ZnS04, 1.32 mM K2HP04. The MOPS 10 x concentrate was prepared by adding MOPS and Tricine to autoclaved water and adjusting the pH to 7.4 with KOH. The remainder of the constituents were then added and the solution mixed for at least 2 h. The solution was filter sterilized and stored at 4°C. 255 To prepare MOPS buffered minimal medium (1L), 100 mL of MOPS 10 x concentrate was aseptically mixed with 10 mL of 0.132 M K2HPO4, 5 mL 80% (v/v) glycerol, 5 mL 1.9 M NH4C1, 5 mL 0.02% (w/v) Vitamin Bl, 5 mL 1% (w/v) Tryptophan and sodium chloride added to adjust the medium osmolality as required. SOB medium (Hanahan et al. 1983) To prepare 1 L of SOB medium, 20 g tryptone 5 g yeast extract 0.5 g NaCl, and 0.2 g KC1 were mixed in 980 mL of water. The resulting solution is autoclaved and cooled. To this solution, 10 mL 1M MgCl2 and 10 mL 1M MgS04 are added. SOC medium (Hanahan et al. 1983) SOC medium was prepared by adding a sterile glucose stock solution to SOB medium to yield 20 mM glucose. Transformation buffer (Hanahan et al. 1983) To prepare 1 L of transformation buffer, 10 mL of 1 M MES (morpholinoethane solfonic acid) adjusted to a pH of 6.3 with KOH, 7.5 g KC1, 8.9 g MnCl2-4H20, 1.5 g CaCl2'2H20, and 0.8 g hexamine cobalt chloride are combined. The resulting solution was filter sterilized and stored at 4°C. LB top agar To prepare LB top agar, 0.7% agar (w/v) was added to LB medium prior to autoclaving. Top agar was maintained at 50°C until used, or melted in the steamer and maintained at 50°C until use. 256 Lactose-MacConkey plates MacConkey plates (1L) were prepared by mixing 50 g Bacto-MacConkey agar (Fisher) and 1.5 g agar in water. The medium was sterilized by autoclaving and plates stored at room temperature. 2.3.5-Triphenyltetrazolium chloride (TTC) proline plates (Bochner et ah, 1977) Seven grams K2HPO4, 3 g KH2PO4 and 2 g protease peptone were mixed together in 370 mL water to prepare Solution A. Fifteen grams of agar and 1.25 mL of 1 M MgS04 were mixed together with 600 mL of water to prepare solution B. Solutions A and B were then autoclaved in separate flasks, cooled to 50°C and mixed. Ten mM proline, 1.6 mM vitamin Bi, 0.24 mM Tryptophan and 0.0025% (w/v) TTC were added as sterile solutions to give 1 L of TTC agar. MOPS buffer based minimal medium agar plates To prepare minimal medium plates, 15 g of agar was dissolved in 870 mL of water and autoclaved. After autoclaving, the medium was cooled to 50 ° C and a sterile nutrient mix composed of: 100 mL 10 x MOPS, 5 mL 1.9 M NH4C1, 10 mL 0.132 M K2HPO4, 10 mL 20% (w/v) glucose and 5 mL 0.02% (w/v) vitamin B,, was added and the media mixed thoroughly. Tryptophan was added to a final concentration of 0.24 mM when required to satisfy strain auxotrophic requirements. In this case, the total volume of water added was adjusted so that the final volume of the media was 1 L. 257 Appendix 2: Maps of plasmids constructed in this study Plasmid pDVl encodes ProQ without a stop codon in pBAD24 Ncol EcoRI. Plasmid pDV2 encodes pDV2 ProQ-RFP ProQ-RFP in pBAD24 (ProQ 5873 bps KL-RFP) AmpR 258 Plasmid pMS 1 encodes ProQ-His6 (ProQ-RSH6) in pQE-60. Strains containing pMSl also contain plasmid pREP4. Plasmid pMS2 encodes ProQ in pQE-60. Strains containing pMS2 also contain plasmid pREP4 259 EcoRI BamHI Ncol Pvul Plasmid pMS9 encodes The N-terminal domain of ProQ without a stop codon (Ml-El30) in pBAD24 EcoRI BamHI Ncol .Hindlll Nhel Bgll... Plasmid pMSlO encodes a histidine tagged N-terminal pMS10 domain of ProQ 5108 bps (MRGSH6RSProQ Ml-El 30) in pQE80L 260 Plasmid pMS 11 encodes The N-terminal domain of ProQ (Ml-E130)inpBAD24 His6-CtermProQ Bgll... Plasmid pMS13 encodes m AmpR 4000 histidine tagged C-terminal I pMS13 domain of ProQ (MRGSH6GSM 4880 bps ProQV180-F232) in pQE80L 261 Plasmid pMS14 encodes The N-terminal domain of ProQ fused to the C-terminal domain ofProQ(ProQMl-E130-AW- ProQ V180-F232) in pBAD24 Plasmid pMS15 encodes The C-terminal domain of ProQ (M-ProQV180-F232)in pBAD24 262 Hindlll Pvull .EcoRI Hindlll Plasmid pMS16 encodes an N-terminally histidine tagged N- terminal domain of ProQ fused to the C-terminal domain (MRGSH6RS ProQ M1-E130- AW-ProQV180-F232)in Pvull pQE80L Pvull V Plasmid pMS17 encodes pMS17 The ProQ-Hise (ProQ-RSH6) in EcoRI 5222 bps Ncol pBAD24 \ Pvull \\ EcoRI \ Bglll Hindlll 263 Plasmid pMS18 encodes The N-terminal domain and linker of ProQ (Ml-VI80) in pBAD24 Plasmid pMS19 encodes The Linker and the C-terminal domain of ProQ (M-ProQE130- F232) in pBAD24 .Pttlll EcoRI Hindlll 264 Plasmid pMS20 is an allelic exchange plasmid based on pK03. It is used to introduce mutation AproQ214 into the E. coli chromosome by double homologous recombination. Plasmid pMS21 encodes an PMS21 N-terminally histidine tagged N- 5062 bps terminal domain of ProQ fused to the C-terminal domain (MRGSH RS ProQ Ml-El 30- proQC 6 AW-ProQV180-F232)in •. •. Pvull \ EcoRI pQE80L Hindlll Bgll 265 Plasmid pMS23 encodes the linker region of ProQ (M-ProQ E130-V180)inpBAD24 EcoRI Kpnl Smal Plasmid pMS23 encodes Xmal Acc!£?l RFP in pBAD24 266 Muni Aalll. : EcoRI Plasmid pMS25 encodes the f 1 Hindlll kanamycin resistance cassette MCS from the Keio collection in pQE80L Plasmid pMS6 encodes ProV and ProW in pQE80L Hindlll 267 Plasmid pMS27 encodes ProVE190Q and ProW in pQE80L Hindlll .Nhel Plasmid pMS28 encodes the 500 bp upstream of the PI Ncol promoter of the pro U operon in s* pQE80L Hpal Sapl EcoRV Ndel 268 Plasmid pMS29 encodes the 500 bp downstream of proX in pQE80L Hpal Sapl / EcoRV Ndel _ „, .BamHI EcoRI.; BsaAl Bgll BamHI, Sphl.j «ss=»4ar-^ Bgll. / 4^ 'ProXdown Nael. / \ > Plasmid pMS30 encodes the 5000 BsaAl,. / kanamycin resistance cassette form the Keio collection flanked by 500 bp upstream and downstream of the proU operon in pQE80L EcoRV BamHI Nsil 269 Plasmid pMS31 is an allelic exchange plasmid based on pK03. It is used to introduce mutation pro V677 into the E. coli chromosome by double homologous recombination, giving expression of 270 Appendix 3: Alignment of identified RNA sequences. rrsH AAATTGAAGAGTTTGATCATGGCTCAGATTGAACGCTGGCGGCAGGCCTAACACATGCAA 60 rrsH GTCGAACGGTAACAGGAAGAAGCTTGCTTCTTTGCTGACGAGTGGCGGACGGGTGAGTAA 120 rrsH TGTCTGGGAAACTGCCTGATGGAGGGGGATAACTACTGGAAACGGTAGCTAATACCGCAT 180 A3-8 GGGGGATAACTACTGGAAACGGTAGCTAATACCGCAT A5-2 GGAGGGGGATAACTACTGGAAACGGTAGCTAATACCGCA A 4-3 GGGGGATAACTACTGGAAACGGTAGCTAATACCGCA A3-6 AGGGGGATAACTACTGGAAACGGTAGCTAATACCGCA rrsH AACGTCGCAAGACCAAAGAGGGGGACCTTCGGGCCTCTTGCCATCGGATGTGCCCAGATG 24 0 A3-8 AAC rrsH GGATTAGCTAGTAGGTGGGGTAACGGCTCACCTAGGCGACGATCCCTAGCTGGTCTGAGA 300 rrsH GGATGACCAGCCACACTGGAACTGAGACACGGTCCAGACTCCTACGGGAGGCAGCAGTGG 360 rrsH GGAATATTGCACAATGGGCGCAAGCCTGATGCAGCCATGCCGCGTGTATGAAGAAGGCCT 420 rrsH TCGGGTTGTAAAGTACTTTCAGCGGGGAGGAAGGGAGTAAAGTTAATACCTTTGCTCATT 480 rrsH GACGTTACCCGCAGAAGAAGCACCGGCTAACTCCGTGCCAGCAGCCGCGGTAATACGGAG 54 0 rrsH GGTGCAAGCGTTAATCGGAATTACTGGGCGTAAAGCGCACGCAGGCGGTTTGTTAAGTCA 600 rrsH GATGTGAAATCCCCGGGCTCAACCTGGGAACTGCATCTGATACTGGCAAGCTTGAGTCTC 660 rrsH GTAGAGGGGGGTAGAATTCCAGGTGTAGCGGTGAAATGCGTAGAGATCTGGAGGAATACC 72 0 rrsH GGTGGCGAAGGCGGCCCCCTGGACGAAGACTGACGCTCAGGTGCGAAAGCGTGGGGAGCA 780 rrsH AACAGGATTAGATACCCTGGTAGTCCACGCCGTAAACGATGTCGACTTGGAGGTTGTGCC 84 0 A3-11 GTCCACGCCGTAAACGATG A4-10 CCCTGGTAGTCCACGCCGTAAA rrsH CTTGAGGCGTGGCTTCCGGAGCTAACGCGTTAAGTCGACCGCCTGGGGAGTACGGCCGCA 900 rrsH AGGTTAAAACTCAAATGAATTGACGGGGGCCCGCACAAGCGGTGGAGCATGTGGTTTAAT 960 A4-11 GAATTGACGGGGGCCCGCACAAGC ValU GGGTGATTAGCTCAGCTGGGAGAGCACCTCCCTTACAAGGAGGGGGTCGGCGGTTCGATC 60 A3-13 GAGGGG GGTTCGATC ValU CCGTCATCACCCACCA 7 6 A3-13 CCGTC rrlH GGGGGCTGCTCCTAGTACGAGAGGACCGGAGTGGACGCATCACTGGTGTTCGGGTTGTCA 2700 W2 4 A 1 rrlH TGCCAATGGCACTGCCCGGTAGCTAAATGCGGAAGAGATAAGTGCTGAAAGCATCTAAGC 2 7 60 W2 4 TGCCAATGGCACTGCCCGGTAGCTAAATGCGGAAGAGATAAGTGCTGAAAGCATCTAAGC 61 rrlH ACGAAACTTGCCCCGAGATGAGTTCTCCCTGACTCCTTGAGAGTCCTGAAGGAACGTTGA 2820 W2 4 ACGAAACTTGCCCCGAGATGAGTTCTCCCTGACTCCTTGAGAGTCCTGAAGGAACGTTGA 121 271 rrlH AGACGACGACGTTGATAGGCCGGGTGTGTAAGCGCAGCGATGCGTTGAGCTAACCGGTAC 2 880 W2 4 AGACGACGACGTTGATAGGCCGGGTGTGTAAGCGCAGCGATGCGTTGAGCTAACCGGTAC 180 rrlH TAATGAACCGTGAGGCTTAACCTT 2 904 W2 4 TAATGAACCGTGAGGCTTAA rrsH gives the DNA sequence encoding the 16S ribosome in E. coli K-12 valU is the DNA sequence encoding valine tRNA in E. coli K-12 rrlH is the DNA sequence encoding the 23 S ribosome in E. coli K-12 Numbers indicate the nucleotide position within the open reading frame 272 Appendix 4: Alignment of the sequences of the insertion sequence present in the proU205 mutation with an IS5 element encoded within the genome of E. coli K-12 PROU205 TAAGGAAGGTGCGAATAATCGGNNNNNNNCTTCTCGGCTGACTCAGTCATTTCATTTCTT 112 IS5 TAAGGAAGGTGCGAATAAGCGGGGAAATTCTTCTCGGCTGACTCAGTCATTTCATTTCTT 2099829 ****************** *** *************************** * * * PROU205 CATGTTTGAGCCGATTTTTTCTCCCGTAAATGCCTTGAATCAGCCTATTTAGACCGTTTC 172 IS5 CATGTTTGAGCCGATTTTTTCTCCCGTAAATGCCTTGAATCAGCCTATTTAGACCGTTTC 2099889 ********************************************************* PROU205 TTCGCCATTTAAGGCGTTATCCCCAGTTTTTAGTGAGATCTCTCCCACTGACGTATCATT 232 IS5 TTCGCCATTTAAGGCGTTATCCCCAGTTTTTAGTGAGATCTCTCCCACTGACGTATCATT 2099949 ********************************************************* PROU205 TGGTCCGCCCGAAACAGGTTGGCCAGCGTGAATAACATCGCCAGTTGGTTATCGTTTTTC 292 IS5 TGGTCCGCCCGAAACAGGTTGGCCAGCGTGAATAACATCGCCAGTTGGTTATCGTTTTTC 2100009 ************************************************************ PROU205 AGCAACCCCTTGTATCTGGCTTTCACGAAGCCGAACTGTCGCTTGATGANGCGAAATGGG 352 IS5 AGCAACCCCTTGTATCTGGCTTTCACGAAGCCGAACTGTCGCTTGATGATGCGAAATGGG 2100069 ************************************************* ********** PROU205 TGCTCCACCCTGGCCCGGATGCTGGCTTTCATGTATTCGATGTTGATGGCCGTTTTGTTC 412 IS5 TGCTCCACCCTGGCCCGGATGCTGGCTTTCATGTATTCGATGTTGATGGCCGTTTTGTTC 2100129 ********* ****************** ********************* PROU205 TTGCGTGGATGCTGTTTCAAGGTTCTTACCTTGCCGGGGCGCTCGGCGATCAGCCAGTCC 472 IS5 TTGCGTGGATGCTGTTTCAAGGTTCTTACCTTGCCGGGGCGCTCGGCGATCAGCCAGTCC 2100189 ********************************************** *********** PROU205 ACATCCACCTCGGCCAGCTCCTCGCGCTGTGGCGCCCCTTGGTAGCCGGCATCGGCTGAN 532 IS5 ACATCCACCTCGGCCAGCTCCTCGCGCTGTGGCGCCCCTTGGTAGCCGGCATCGGCTGAG 2100249 *********************************************************** PROU2 05 ACAAATTGCTCCTCTCCATGCAGCANATTANCCCAGNTGANTGAGGTCATGCTCGTTGGC 592 IS5 ACAAATTGCTCCTCTCCATGCAGCAGATTA-CCCAGCTGATTGAGGTCATGCTCGTTGGC 210030E ****************************** ***** *** ******************* PROU205 CNCGGNGGTGACCANGCTGTGGGTCAGGNCACTCTTNGCATCNACNNCNATGTGNNNCTT 652 IS5 CGCGGTGGTGACCAGGCTGTGGGTCAGGCCACTCTTGGCATCGACACCAATGTGGGCCTT 2100368 * * * * * * * t* **********V t * * ** * ***** PROU205 CNNGNCNAAGTGCCNCTGNATNGCCTTTCTTGGTCTGATGCATCTCC 699 IS5 CATGCCAAAGTGCCACTG-ATTGCCTTTCTTGGTCTGATGCATCTCC 2100414 * * * ******* *** ** ************************* Sequences of proU205 fcomE. coli strain WG210 are aligned with the sequence for IS5 found within the indicated map positions of the E. coli K-12 chromosome. * indicate exact identity. 273