Download The

SIDEROPHORE-MEDIATED IRON METABOLISM IN STAPHYLOCOCCUS AUREUS

Marek John Kobylarz

B.Sc., The University of Victoria, 2010

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF

THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES

(Microbiology and Immunology)

THE UNIVERSITY OF BRITISH COLUMBIA

(Vancouver)

February 2016

Staphylococcus aureus requires iron as a nutrient and uses uptake systems to extract iron from the human host. S. aureus produces the iron-chelating siderophore staphyloferrin B (SB) to scavenge for available iron under conditions of low iron stress. Upon iron-siderophore re-entry into the cell, iron is separated from the siderophore complex to initiate assimilation into metabolism.

To gain insight into how SB biosynthesis is integrated into S. aureus central metabolism, the three SB precursor biosynthetic proteins, SbnA, SbnB, and SbnG, were biochemically characterized. SbnG is a citrate synthase analogous to the citrate synthase enzyme present in the

TCA cycle. The crystal structure of SbnG was solved and superpositions with TCA cycle citrate synthases support a model for convergent evolution in the active site architecture and a conserved catalytic mechanism. Since L-Dap is an essential precursor for SB, the biosynthetic pathway for L-Dap was elucidated. A combination of X-ray crystallography, biochemical assays and biophysical techniques were used to delineate the reaction mechanisms for SbnA and SbnB, demonstrating that SbnA performs a -replacement reaction using O-phospho-L-serine (OPS) and L-glutamate to produce N-(1-amino-1-carboxy-2-ethyl)-glutamic acid (ACEGA). Oxidative hydrolysis of ACEGA catalyzed by SbnB produces -ketoglutarate and L-Dap. Detailed analysis of the substrate specificity of SbnA revealed that OPS binding and conversion to the PLP-- aminoacrylate intermediate in SbnA induced a conformational change and formation of a second substrate binding pocket for L-glutamate. Furthermore, L-cysteine was identified as a competitive inhibitor of SbnA activity, revealing a link between iron uptake and the oxidative stress response in S. aureus. IruO was examined for its role in Fe(III)-siderophore reduction.

Utilizing a combination of visible spectroscopy and enzyme kinetics, a mechanism for electron transfer was proposed. IruO was demonstrated to reduce iron bound hydroxamate-type siderophores to release Fe(II) using NADPH as the electron donor. Under anaerobic conditions,

IruO formed a stable FAD semiquinone intermediate that mediates a single electron transfer from the FAD to the Fe(III)-siderophore complex. These studies have shown how SB precursors are synthesized and led to the development of models for SB biosynthesis integration into central metabolism under conditions of low iron stress.

iii

Preface

Most of the work presented in this thesis is published or drawn from manuscripts under preparation. The following contributions were made by fellow scientists and collaborators:

Chapter 3

Kobylarz, M.J., Grigg, J.C., Sheldon, J.R., Heinrichs, D.E., and Murphy, M.E.P. (2014) SbnG, a citrate synthase in Staphylococcus aureus: a new fold on an old enzyme, J. Biol. Chem. 289,

33797-33807.

Chapter 3 was derived from the published manuscript. Dr. D.E. Heinrichs provided the sbnG plasmid construct. Dr. J.C. Grigg developed the protein expression and purification protocols. I optimized crystallization conditions, determined the structure of SbnG and performed bioinformatics analyses. J.R. Sheldon generated SbnG variant complementation vectors (Tables 2-1 and 2-2) and analyzed SbnG activity in S. aureus, which is presented in

Figure 3-5. I wrote the first draft of the manuscript with contributions from J.R. Sheldon on the methods and results. The manuscript was edited by Dr. J.C. Grigg, Dr. D.E. Heinrichs, and Dr.

M.E.P. Murphy.

Chapter 4

Kobylarz, M.J., Grigg, J.C., Takayama, S.J., Rai, D.K., Heinrichs, D.E., and Murphy, M.E.P.

(2014) Synthesis of L-2,3-diaminopropionic acid, a siderophore and antibiotic precursor, Chem.

Biol. 21, 379-388.

Chapter 4 was assembled from the published manuscript. Dr. D.E. Heinrichs provided the sbnA and sbnB plasmid constructs. Dr. J.C. Grigg and D.K. Rai were responsible for solving the structure of SbnB in both the apo- and NAD+ bound form, which are presented in Figure 4-9. Dr.

S.J. Takayama performed the NMR small molecule analysis of ACEGA and the results are presented in Figure 4-3 and Table 4-1. I optimized the protein expression and purification for

SbnA and SbnB, performed all activity assays and visible spectroscopic analysis. I also solved the structure of SbnB with bound ligands ACEGA and NAD+, and -KG and NADH. I wrote the first draft of the manuscript. Dr. S.J. Takayama provided the NMR results and figures. The manuscript was edited by Dr. J.C. Grigg, Dr. D.E. Heinrichs, and Dr. M.E.P. Murphy.

Chapter 5

Kobylarz, M.J., Grigg, J.C., Liu, Y., Lee, M.S.F., Heinrichs, D.E., and Murphy, M.E.P. (2015)

Deciphering the substrate specificity of SbnA, the enzyme catalyzing the first step in staphyloferrin B biosynthesis. Accepted by Biochemistry.

Chapter 5 was derived from the manuscript currently submitted and under review. Dr.

J.C. Grigg and M.S.F. Lee solved the wild type structure of SbnA, which is presented in Figure

5-1. I optimized the purification protocol for SbnA, identified new SbnA crystallization conditions, solved the structure of SbnA bound to its substrate O-phospho-L-serine and designed all kinetic analysis assays. Y. Liu generated the SbnA active site variants and performed both steady-state kinetic and single turnover analysis on the SbnA variants and the results are presented in Figures 5-5, 5-6, 5-9, 5-13 and 5-14 and Tables 5-1, 5-2 and 5-3. Y. Liu also crystallized and solved the structures of SbnA Y152F and SbnA Y152F/S185G (Figure 5-11). I

wrote the first draft of the manuscript while J.C. Grigg, Dr. D.E. Heinrichs and Dr. M.E.P.

Murphy edited the manuscript.

Chapter 6

Kobylarz, M.J., Heieis, G.A., Loutet, S.A., and Murphy, M.E.P. (2015) The siderophore reductase IruO uses an FAD semiquinone intermediate to catalyze iron reduction. Manuscript in preparation.

Chapter 6 is a draft of a manuscript that will be submitted. I did the cloning, protein expression and purification, and structure determination of IruO. I also measured the sulfhydryl content of IruO, performed the visible spectroscopic analysis of IruO and developed all kinetic analysis assays. G.A. Heieis performed all the kinetic analysis assays, fluorescence spectroscopy and anaerobic visible absorption spectroscopy, which is presented in Figures 6-8, 6-9, 6-11 and

6-12 and Table 6-1. Dr. S.A. Loutet purified IsdI, did the heme degradation assays (Figure 6-10) and wrote the corresponding methods and results sections. I wrote the first draft of the manuscript.

Biosafety Approval

This project required Biohazard Approval for the handling of Staphylococcus aureus and

Escherichia coli and was issued by the UBC Biosafety Committee, Certificate number B13-

0096.

Table of Contents

Abstract ...... ii

Preface ...... iv

Table of Contents ...... vii

List of Tables ...... xiii

List of Figures ...... xiv

List of Abbreviations ...... xviii

Acknowledgements ...... xxii

Chapter 1: Introduction ...... 1

1.1 Staphylococcus aureus ...... 1

1.2 Staphylococcus aureus iron uptake systems ...... 3

1.2.1 Heme uptake ...... 4

1.2.2 Siderophore uptake systems ...... 4

1.2.2.1 Hydroxamate-type siderophore uptake ...... 9

1.2.2.2 Catecholate-type siderophore uptake ...... 11

1.2.2.3 Staphyloferrin (SA) and (SB) uptake ...... 12

1.3 Siderophore biosynthesis in Staphylococcus aureus ...... 13

1.3.1 Staphyloferrin A...... 14

1.3.2 Staphyloferrin B ...... 14

1.4 Iron uptake systems in other Staphylococcus species ...... 17

1.5 S. aureus metabolism ...... 18

1.5.1 Central metabolism overview ...... 18 vii

1.5.2 Major nutrient-response regulators ...... 20

1.6 S. aureus metabolic adaptations to the host ...... 22

1.7 Intracellular iron release mechanisms ...... 25

1.7.1 Iron release from heme ...... 25

1.7.2 Iron release from siderophores...... 27

1.8 Objectives ...... 29

Chapter 2: Experimental Procedures ...... 32

2.1 S. aureus growth conditions ...... 32

2.2 Construction of S. aureus mutant strains ...... 34

2.2.1 Site-directed mutagenesis of S. aureus sbnG for in vivo assays ...... 34

2.3 Cloning, expression and purification for biochemical assays and structure

determination ...... 36

2.3.1 Cloning, expression and purification of SbnG and variants ...... 36

2.3.2 Cloning, expression and purification of SbnA and variants ...... 37

2.3.3 Cloning, expression and purification of SbnB ...... 38

2.3.4 Cloning, expression and purification of IruO ...... 39

2.3.5 Expression and purification of SB biosynthetic enzymes ...... 40

2.3.6 Expression and purification of IsdI ...... 41

2.4 Crystallization and structure determination ...... 41

2.4.1 SbnG and E151Q variant structure determination ...... 41

2.4.2 SbnA and variants Y152F and Y152F/S185G structure determination...... 45

2.4.3 SbnB structure determination ...... 48

2.4.4 IruO structure determination ...... 51 viii

2.5 Determination of oligomeric state in solution ...... 54

2.5.1 SbnG oligomerization state determination...... 54

2.5.2 SbnA and SbnB oligomerization state determination ...... 54

2.5.3 IruO oligomerization state determination ...... 55

2.6 Bioinformatic analysis ...... 55

2.6.1 SbnG structure superposition and phylogenetic analysis...... 55

2.6.2 Multiple sequence alignment of SbnA and homologs ...... 56

2.6.3 Multiple sequence alignment of IruO and homologs ...... 56

2.7 UV-Vis spectroscopy analysis of SbnA ...... 57

2.8 UV-Vis spectroscopy analysis of IruO ...... 57

2.9 Fluorescence spectroscopy analysis of IruO ...... 58

2.10 Citrate synthase assay ...... 59

2.11 Phosphate release assay ...... 59

2.12 NADH assay ...... 60

2.13 -KG assay ...... 60

2.14 Ferrozine assay...... 61

2.15 Heme degradation assay ...... 61

2.16 Measurement of IruO sulfhydryls ...... 62

2.17 Agar plate bioassays ...... 63

2.17.1 Determining citrate production from SbnG and variants ...... 63

2.17.2 Determining L-Dap production from SbnA and SbnB ...... 64

2.18 Mass spectrometry analysis ...... 65

2.19 NMR analysis...... 65 ix

2.20 Steady-state kinetic analysis ...... 66

2.20.1 SbnA and variants steady-state kinetic analysis ...... 66

2.20.2 IruO steady-state kinetic analysis ...... 67

2.21 Stopped-flow absorption spectroscopy for single turnover kinetic analysis ...... 67

2.22 Software for kinetic analysis ...... 68

Chapter 3: SbnG, a citrate synthase in Staphylococcus aureus: a new fold on an old enzyme family ...... 69

3.1 Introduction ...... 69

3.2 Results ...... 70

3.2.1 Overall structure of SbnG ...... 70

3.2.2 SbnG active site ...... 72

3.2.3 Crystal structure of SbnG E151Q variant bound to oxaloacetate ...... 74

3.2.4 Structural comparison of SbnG to the phosphoenolpyruvate/pyruvate domain

superfamily ...... 76

3.2.5 SbnG represents a new family of metal independent enzymes within the class II

aldolase superfamily ...... 79

3.2.6 SbnG active site variants show reduced SB production in a citrate synthase-deficient

S. aureus strain ...... 79

3.3 Discussion ...... 81

Chapter 4: Synthesis of L-2,3-diaminopropionic acid, a siderophore and antibiotic precursor ...... 88

4.1 Introduction ...... 88

4.2 Results ...... 89 x

4.2.1 SbnA produces ACEGA from O-phospho-L-serine and L-glutamate ...... 89

4.2.2 SbnB hydrolyzes ACEGA to form L-Dap and -KG ...... 93

4.2.3 L-Dap and -KG generated from SbnA and SbnB are precursor substrates for SB

biosynthesis ...... 95

4.2.4 Structures of SbnB complexes ...... 96

4.3 Discussion ...... 101

Chapter 5: Deciphering the substrate specificity of SbnA, the enzyme catalyzing the first step in staphyloferrin B biosynthesis...... 107

5.1 Introduction ...... 107

5.2 Results ...... 108

5.2.1 Structure of SbnA ...... 108

5.2.2 SbnA forms a PLP--aminoacrylate intermediate with OPS ...... 110

5.2.3 SbnA active site variants ...... 113

5.2.4 Kinetic analysis of SbnA and variants ...... 117

5.2.5 SbnA variant crystal structures ...... 120

5.2.6 L-cysteine inhibits SbnA activity...... 123

5.3 Discussion ...... 126

Chapter 6: The siderophore reductase IruO uses an FAD semiquinone intermediate to catalyze iron reduction ...... 132

6.1 Introduction ...... 132

6.2 Results ...... 133

6.2.1 IruO binds and reduces hydroxamate-type siderophores ...... 133

6.2.2 Structure of IruO ...... 135 xi

6.2.3 An intramolecular disulfide bond can form in IruO ...... 141

6.2.4 IruO activity is diminished by the formation of an intramolecular disulfide bond 143

6.2.5 IruO has high affinity for hydroxamate-type siderophores...... 146

6.2.6 IruO mediates electron transfer from NADPH to Fe(III)-siderophore complexes via

an FAD neutral semiquinone intermediate ...... 147

6.3 Discussion ...... 150

Chapter 7: Overview and future directions...... 157

7.1 SB biosynthesis does not require a functioning TCA cycle ...... 157

7.2 Metabolic pathways and regulatory networks differentiate SA and SB biosynthesis in

S. aureus ...... 161

7.3 A mechanism for Fe(III)-siderophore iron release in S. aureus ...... 164

7.4 Regulatory links between iron uptake and the oxidative stress response in S. aureus 165

7.5 Future work ...... 168

Bibliography ...... 172

xii

List of Tables

Table 2-1 Bacterial strains, plasmids, and oligonucleotides...... 33

Table 2-2 Primers for SbnG variants...... 35

Table 2-3 A list of phosphorylated mutagenic primers used to generate SbnA active site variants.

...... 37

Table 2-4 X-ray diffraction data collection and refinement statistics for SbnG...... 44

Table 2-5 X-ray diffraction data collection and refinement statistics for SbnA and variants...... 47

Table 2-6 X-ray diffraction data collection and refinement statistics for SbnB...... 50

Table 2-7 X-ray diffraction data collection and refinement statistics for IruO ...... 53

Table 3-1 Superposition statistics of wild type SbnG to structurally characterized homologs determined from DaliLite...... 76

Table 3-2 Pairwise sequence alignment statistics of wild type SbnG to homologs from other siderophore biosynthetic operons...... 83

Table 4-1 1H and 13C chemical shifts of ACEGA at pH 8.0 and 25 ºC...... 93

Table 5-1 Single turnover kinetic parameters of wild type SbnA and SbnA variants with various concentrations of OPS...... 117

Table 5-2 Steady-state kinetic parameters of wild type SbnA and SbnA S185G...... 120

Table 5-3 Single turnover kinetic parameters of wild type SbnA and SbnA variants with various concentrations of L-cysteine...... 125

Table 6-1 Steady-state kinetic constants for rdIruO and oxIruO...... 145

xiii

List of Figures

Figure 1-1 Schematic representation of siderophores highlighting the three different types of iron coordinating functional groups...... 6

Figure 1-2 S. aureus endogenously produced siderophores...... 8

Figure 1-3 Schematic of a general siderophore uptake system in Gram-positive bacteria...... 9

Figure 1-4 Schematic of known siderophore uptake systems in S. aureus...... 11

Figure 1-5 SB biosynthesis in S. aureus...... 17

Figure 1-6 S. aureus central metabolism...... 20

Figure 1-7 Schematic of intracellular iron release from heme and Fe(III)-siderophores in S. aureus...... 26

Figure 3-1 Structure of SbnG...... 71

Figure 3-2 SbnG is a hexamer in solution as determined by SEC-MALS...... 73

Figure 3-3 Oxaloacetate bound to SbnG E151Q...... 75

Figure 3-4 Conservation of active site structure in SbnG homologs...... 78

Figure 3-5 SbnG variants are impaired for citrate-dependent production of SB...... 80

Figure 3-6 Multiple sequence alignment of select SbnG homologs from the phosphoenolpyruvate/pyruvate domain superfamily...... 84

Figure 3-7 Superposition of the metal-binding motif from SbnG with MPS...... 84

Figure 3-8 Comparison of the SbnG active site with TCA cycle citrate synthases...... 86

Figure 4-1 SbnA condenses OPS and L-glutamate to generate ACEGA...... 90

Figure 4-2 Proposed biosynthetic pathway for L-Dap...... 91

Figure 4-3 Structure of ACEGA...... 92 xiv

Figure 4-4 SbnB degrades ACEGA to generate L-Dap and -KG...... 94

Figure 4-5 L-Dap and -KG generated from SbnA and SbnB can be used by Sbn siderophore synthetases to create SB in vitro...... 96

Figure 4-6 Structure of SbnB...... 97

Figure 4-7 Molar mass determination of SbnB by SEC-MALS...... 98

Figure 4-8 SbnB active site...... 99

Figure 4-9 Structural overlay of homologs to SbnB...... 103

Figure 4-10 Proposed catalytic mechanism for oxidative hydrolysis of ACEGA to generate L-

Dap and -KG in SbnB...... 104

Figure 4-11 Summary of the overall stoichiometric equation for O-phospho-L-serine and L- glutamate production required for SB biosynthesis...... 106

Figure 5-1 Structure of SbnA...... 109

Figure 5-2 SbnA is a dimer in solution as determined by SEC-MALS...... 110

Figure 5-3 SbnA adopts a closed conformation when bound to the PLP--aminoacrylate...... 112

Figure 5-4 Identification of OPS discriminating residues...... 114

Figure 5-5 UV-visible absorption spectra of SbnA variants...... 116

Figure 5-6 Rate of formation for the PLP--aminoacrylate intermediate...... 117

Figure 5-7 SbnA active site mutations attenuated ACEGA production...... 118

Figure 5-8 Optimization of SbnA activity...... 119

Figure 5-9 Steady-state kinetics of wild type SbnA (A and B) and S185G SbnA (C and D). ... 119

Figure 5-10 Structural overlay of wild type SbnA (brown) with Y152F SbnA (blue)...... 121

Figure 5-11 Structure of the SbnA variant Y152F/S185G...... 122

Figure 5-12 UV-visible absorption spectra of wild type SbnA and SbnA variants with 1 mM L- cysteine...... 124

Figure 5-13 Inhibition kinetics for (A) wild type SbnA and (B) S185G SbnA against varying concentrations of L-cysteine...... 124

Figure 5-14 Rate of formation for the PLP-cysteine intermediate...... 125

Figure 5-15 Proposed flow diagram of staphyloferrin B biosynthesis integrated into central metabolism of S. aureus under nutrient deprivation and oxidative stress...... 130

Figure 6-1 IruO selectively reduces Fe(III)-hydroxamate-type siderophores...... 134

Figure 6-2 IruO selectively reduces Fe(III)-hydroxamate-type siderophores...... 135

Figure 6-3 Structure of rdIruO...... 137

Figure 6-4 IruO forms a dimer in solution...... 138

Figure 6-5 IruO substrate binding sites...... 140

Figure 6-6 OxIruO contains an intramolecular disulfide bond...... 142

Figure 6-7 Multiple sequence alignment of IruO homologs...... 143

Figure 6-8 Steady-state kinetics of IruO with Fe(III)-DFB and Fe(III)-FCA...... 144

Figure 6-9 Steady-state kinetics of IruO with NADPH...... 145

Figure 6-10 Oxidation of IruO lowers the rate of heme degradation by IsdI...... 146

Figure 6-11 IruO binds hydroxamate-type siderophores with high affinity...... 147

Figure 6-12 IruO reduces Fe(III)-DFB using NADPH as the electron donor via an FAD semiquinone intermediate...... 149

Figure 6-13 IruO maintains the FADsq intermediate under molar excess of NADPH unlike the putative thioredoxin reductase (NWMN_0732) from S. aureus...... 150

Figure 6-14 Proposed mechanism for Fe(III)-siderophore reduction via IruO...... 152 xvi

Figure 6-15 FAD isoalloxazine ring binding in IruO homologs...... 154

Figure 6-16 IruO FAD binding and active site formation is divergent from TrxB...... 155

Figure 7-1 Proposed flow diagram of staphyloferrin B biosynthesis integrated into central metabolism through the CodY regulon under conditions of high glucose and low iron...... 159

Figure 7-2 Integrating SA and SB biosynthesis into central metabolism in S. aureus...... 162

Figure 7-3 Regulatory links between iron uptake and the oxidative stress response in S. aureus.

...... 167

xvii

List of Abbreviations

6x-His Poly-histidine affinity purification tag

-KG -ketoglutarate

Å Angstrom

ACEGA N-(1-amino-1-carboxyl-2-ethyl)-glutamic acid

Acetyl-CoA Acetyl coenzyme A

BCAA Branched chain amino acid

CA-MRSA Community acquired methicillin resistant Staphylococcus

aureus

CAS Chrome azurol S

CcpA Catabolite control protein A

CLS Canadian Light Source

DDGA 2-dehydro-3-deoxy-galactarate aldolase

DNPH 2,4-dinitrophenylhydrazine

DTNB 5,5'-dithiobis-(2-nitrobenzoic acid

DTT Dithiothreitol

EDDHA Ethylenediamine-N,N'-bis(2-hydroxyphenylacetic acid)

FAD Flavin adenine dinucleotide

FADhq Hydroquinone flavin adenine dinucleotide

FADox Oxidized flavin adenine dinucleotide

FADsq Semiquinone flavin adenine dinucleotide

Fe(III)-DFB Fe(III)-desferrioxamine B xviii

Fe(III)-FCA Fe(III)-ferrichrome A

FeS Iron-sulfur cluster

FMN Flavin mononucleotide

FNR Ferredoxin-NADP(H) reductase

Fur Ferric uptake regulator

GaDFB Gallium(III)-desferrioxamine B

HA-MRSA Hospital acquired methicillin resistant Staphylococcus

aureus

Hepes 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

HPLC High performance liquid chromatography

IPTG Isopropyl β-D-1-thiogalactopyranoside

IruO Iron utilization oxidoreductase

Isd Iron-responsive surface determinant

LB Lysogeny broth

LC-ESI-MS Liquid chromatography electrospray ionization mass

spectrometry

L-Dap L-2,3-diaminopropionic acid

MESG 2-amino-6-mercapto-7-methylpurine riboside

MPS Macrophomate synthase

MRSA Methicillin resistant Staphylococcus aureus

NADH Nicotinamide adenine dinucleotide

NADPH Nicotinamide adenine dinucleotide phosphate

NIS NRPS-independent siderophore synthesis xix

NRPS Nonribosomal peptide synthesis

OAS O-acetyl-L-serine

OASS O-acetyl-L-serine sulfhydrylase

OPS O-phospho-L-serine

OPSS O-phospho-L-serine sulfhydrylase

OxIruO Oxidized IruO

PEG Poly(ethylene glycol)

PEG MME Poly(ethylene glycol) monomethyl ether

PLP Pyridoxal 5’-phosphate

PNDO Pyridine nucleotide-disulfide oxidoreductase

RdIruO Reduced IruO r.m.s.d. Root mean square deviation

ROS Reactive oxygen species

RPMI-1640 Roswell Park Memorial Institute 1640

SA Staphyloferrin A

SB Staphyloferrin B

SbnA-AA SbnA -aminoacrylate

SbnA-PLP SbnA pyridoxal 5’-phosphate

SDS-PAGE Sodium dodecyl sulphate polyacrylamide gel

electrophoresis

SEC-MALS Size exclusion chromatography multi-angle light scattering

SeMet Selenomethionine

SH Sulfhydryl xx

SSRL Stanford Synchrotron Radiation Lightsource

TCA Tricarboxylic acid

TCEP Tris(2-carboxyethyl)phosphine

TMS Tris minimal succinate

TSB Tryptic Soy Broth

VISA Vancomycin intermediate Staphylococcus aureus

VRSA Vancomycin resistant Staphylococcus aureus

2x YT 2x yeast extract tryptone

xxi

Acknowledgements

I would like to thank Dr. Michael Murphy for the tremendous opportunity to join his laboratory and to complete my Ph.D. thesis as well as for the support and freedom to explore and develop new project ideas.

My committee members, Dr. Bob Hancock, Dr. Ross MacGillivray and Dr. Lawrence

McIntosh were very supportive and I thank them for their guidance and helpful advice.

I also want to give a special thank you to Dr. Tom Beatty who graciously stepped in as an examiner for my comprehensive exam while Dr. Bob Hancock was away on sabbatical.

I would like to acknowledge funding from the John Richard Turner Fellowship award, as well as project funding from the Canadian Institute of Health Research held by Dr. Michael Murphy.

I want to thank all those who contributed to my research on the technical side, including

Dr. Yun Ling and Marco Yeung from the UBC Mass Spectroscopy Centre for helping to analyze my samples and Dr. Lindsay Eltis, Dr. Rahul Singh, Dr. Tony Ruzzini and Eugene Kuatsjah for access to their spectrophotometers and the many hours of training and helpful advice.

I thank our collaborators Dr. David Heinrichs and members of his laboratory, especially

Johnson Cheung and Jessica Sheldon. Without their help and technical assistance, many of my projects would not have progressed as far as they have.

To the past and present members of the Murphy lab, thank you for making my time here enjoyable and memorable, whether it’s helping me to troubleshoot a protocol, skiing up on the

North Shore Mountains, or even playing basketball in the park.

Finally, I thank my family and friends who supported me 100% over the years as I toiled through my degree. xxii

Chapter 1: Introduction

1.1 Staphylococcus aureus

S. aureus is a Gram-positive coccal bacterium that is a member of the phylum Firmicutes.

S. aureus is distinguished from the other ~40 species of Staphylococci based on the gold pigmented colonies formed and positive results of coagulase, mannitol fermentation and deoxyribonuclease tests (1). A natural reservoir for S. aureus is humans, but S. aureus is also known to colonize other mammals. Approximately 20 % of the human population is persistently colonized, while ~30 % of the human population is transiently colonized (2,3). Typically, S. aureus resides as a commensal on the skin or mucosal surfaces such as the nose and throat and commonly transmitted through direct contact with a colonized individual or via fomites (4).

However, S. aureus is also known to be an opportunistic pathogen capable of infecting most tissue types such as the bloodstream, skin and soft tissue, bones and joints and urinary tract (5).

A S. aureus infection is commonly attributed to a breach in the endothelial barrier as a result of injury or surgery, or immuno-suppressive conditions such as HIV or organ transplant, leading to a wide range of diseases like pneumonia, endocarditis, meningitis and bacteremia (4,5).

Ever since the discovery and application of antibiotics as a treatment for bacterial infections, antibiotic resistance has been a serious problem. Initially, penicillin, a -lactam antibiotic, discovered in the 1940’s was initially successful in treating S. aureus infections, but by 1948, most hospital isolates (59%) were reported to be resistant to penicillin (6). A similar situation has developed for the penicillin derivative antibiotic, methicillin, with an ever increasing prevalence of healthcare-associated methicillin-resistant S. aureus (HA-MRSA) (7).

Previously relegated to only healthcare environments, instances of community-associated 1

methicillin-resistant S. aureus (CA-MRSA) have steadily appeared over the last 20 years (8). Of the 20 or so known distinct CA-MRSA lineages, one strain in particular, USA300, has been responsible for the vast majority of outbreaks in both the community and hospitals in the United

States since the 1990’s (9,10). Compared to hospital-associated strains, CA-MRSA appears to have enhanced virulence and colonization capabilities, which increases transmission (10). More recently, the identification of vancomycin intermediate-resistant S. aureus (VISA) and vancomycin-resistant S. aureus (VRSA) are severely limiting the current available treatment options as vancomycin is widely considered to be the gold standard for treating MRSA infections

(11). Recent statistics in the United States revealed that in 2011 an estimated 80,461 invasive

MRSA infections occurred with an estimated 11,285 resulting in death (12). In Canada, MRSA cases per 1000 admissions increased in hospitals over 10-fold between 1995 and 2003 (13). Of the few cost assessments made publicly available, MRSA infections cost the Canadian health system between $54 and $110 million in 2005 (14).

The success of S. aureus as a pathogen is attributed to its large repertoire of virulence factors. The virulence factors can be roughly divided into three categories: secreted proteins and small molecules; cell surface-bound proteins; and cell surface components (15). Upon tissue invasion, S. aureus can protect itself from the host immune attack by coating its surface with host components like fibronectin, fibrinogen and collagen via surface-bound proteins (16) and can resist phagocytosis by producing a thick outer capsule (17). S. aureus also employs various defense mechanisms such as producing the enzymes catalase, superoxide dismutase and flavohaemoglobin to neutralize the host immune system derived hydrogen peroxide, superoxide, and nitric oxide, respectively (18-20), or by producing staphyloxanthin, a carotenoid pigment that acts as an antioxidant against reactive oxygen species (21). In addition to its defense 2

mechanisms, S. aureus can induce tissue destruction by secreting exoproteins such as cytotoxins, proteases and lipases as a means to further infiltrate through host tissues (22). As a consequence, damaged host tissue aids the release of nutrients required for S. aureus growth and sustaining infection. Therefore, S. aureus has evolved numerous import systems to scavenge for nutrients from their immediate environment, including iron (23).

1.2 Staphylococcus aureus iron uptake systems

As is the case for most organisms, including S. aureus, iron is an essential nutrient required for survival. Iron is a particularly versatile element that can exist in two physiologically relevant oxidation states: ferric form or Fe(III), or ferrous form or Fe(II) (24). However, freely available iron is scarce due to the oxygenic atmosphere where iron predominantly exists in the water insoluble Fe(III) form. Due to the abundance of soluble Fe(II) early in evolutionary history on Earth when oxygen was scarce, iron was readily incorporated into proteins as a cofactors or used for electron transfer processes (24). Iron is incorporated into proteins as either free ions, heme, or iron-sulfur clusters (25). For example, in S. aureus, iron plays an essential role in central metabolism (iron-sulfur clusters), in the synthesis of deoxyribonucleotides for DNA synthesis (free ions) and in electron transport reactions via cytochromes (heme) (25).

Since S. aureus is dependent on its host for certain nutrients like iron, S. aureus has developed mechanisms to extract and import iron from its environment. S. aureus has been shown to grow on host iron sources such as Fe(III)/Fe(II) ions, transferrin, heme and hemoglobin

(26,27). Although there is redundancy in the number of iron sources available, S. aureus has been shown to preferentially utilize heme over transferrin iron sources (27).

1.2.1 Heme uptake

In S. aureus, the majority of heme uptake is mediated by the iron-responsive surface determinant (isd) system. The Isd system is composed of nine iron-regulated proteins directly involved in heme uptake (28-30). Three of the Isd proteins (IsdABH) are anchored to the peptidoglycan by sortase A (SrtA) (30). The fourth cell wall anchored protein, IsdC, is attached by a second sortase called SrtB, which is encoded within the isd gene cluster. Both IsdB and

IsdH reside at the cell wall surface and are used to extract heme from hemoproteins such as hemoglobin (26). Heme is relayed from the outer surface into the cell wall via IsdB or IsdH to

IsdA. From IsdA, heme is then transferred via IsdC to the heme-binding lipoprotein IsdE, located at the cell membrane, using a cog-wheel mechanism that continuously cycles heme from holo-

IsdA to apo-IsdE (31,32). Heme binding and transfer is controlled by a conserved protein fold called the NEAT (for NEAr Transporter) domain (33). Coupled with its membrane permease,

IsdF, IsdEF translocates heme into the cytoplasm. A third predicted membrane protein, IsdD, is assumed to play a role in heme uptake, but its function has not been identified. Once internalized, heme can either be transferred directly to other intracellular heme-requiring proteins, or degraded by two heme monooxygenases, IsdG and IsdI to release Fe(II) and staphylobilin (34,35).

1.2.2 Siderophore uptake systems

Siderophores are specialized Fe(III) chelators that are produced by various organisms to scavenge for iron in the external milieu (36). The vast majority of siderophores isolated and identified have been derived from bacteria and fungi. However, some plants such as grasses are able to secrete Fe(III) chelators called phytosiderophores from their root systems (37). 4

Siderophore-mediated iron acquisition is an important factor in maintaining iron homeostasis and in the case of bacterial pathogens; siderophore production can be a critical for pathogenesis. In

Mycobacterium tuberculosis, the loss of mycobactin siderophore production impaired its growth in macrophage-like THP-1 cells, suggesting that siderophore production is an essential virulence factor (38).

Figure 1-1 Schematic representation of siderophores highlighting the three different types of iron coordinating functional groups. (A) Enterobactin produced by Escherichia coli contains catecholate functional groups. (B) Desferrioxamine B produced by Streptomyces pilosus contains hydroxamate function groups. (C) Rhizoferrin produced by Cunninghamella elegans contains -hydroxycarboxylate functional groups. Iron coordinating atoms are colored red.

Siderophores are roughly categorized based on the functional groups that directly coordinate Fe3+ and include the catecholate, hydroxamate and -hydroxycarboxylate functional groups (Figure 1-1) (36). Siderophore affinity for Fe(III) can be influenced not only by the iron- coordinating functional group present, but also by their hexadentate geometry around the Fe(III) and molecular strain due to the siderophore structure (36). Enterobactin, with its three

-49 catecholate groups, has the highest affinity for Fe(III) with a Kd value of ~10 M (39). Lower affinities for Fe(III) are observed for the hydroxamate-type siderophores like desferrioxamine B

-31 -26 (Kd = ~10 M) (39) and -hydroxycarboxylate containing siderophore rhizoferrin (Kd = ~10

M) (40). High affinity for Fe(III) binding is required to extract Fe(III) from its insoluble Fe(III)

-12 oxide-hydroxide form, or from other Fe(III) binding compounds like citrate (Kd = ~10 M) (41),

-20 or proteins like transferrin (Kd = ~10 M) (42). More recently, siderophores classified as ‘mixed type’ have been identified that contain multiple iron coordinating functional groups such as petrobactin (catecholate and -hydroxycarboxylate) or aerobactin (hydroxamate and - hydroxycarboxylate) (36). Many bacterial pathogens typically produce more than one siderophore such as Pseudomonas aeruginosa (43), Klebsiella pneumoniae (44), and Bacillus anthracis (45). S. aureus produces two -hydroxycarboxylate-type siderophores called staphyloferrin A (SA) and staphyloferrin B (SB) (Figure 1-2). In Gram-positive bacteria like S. aureus, Fe(III)-siderophore complexes can diffuse through the cell wall to the cell membrane where they are imported into the cytoplasm by specific siderophore ABC transporters (Figure 1-

3) (23). In addition to the endogenous SA and SB uptake systems, S. aureus also contains other siderophore uptake systems that are selective to the siderophore iron-coordinating functional groups (46,47).

Figure 1-2 S. aureus endogenously produced siderophores. (A) Staphyloferrin A (SA). (B) Staphyloferrin B (SB). Both siderophores coordinate iron using -hydroxycarboxylate functional groups. Iron coordinating atoms are colored red.

Figure 1-3 Schematic of a general siderophore uptake system in Gram-positive bacteria. Siderophores scavenge for Fe(III) in the extracellular matrix forming a Fe(III)-siderophore complex. Fe(III)-siderophore complexes can then diffuse through the cell wall and are recognized by cell membrane-bound receptors. Next, receptor-bound Fe(III)-siderophores are translocated across the cell membrane through a transmembrane permease and releases the Fe(III)-siderophore into the cytoplasm.

1.2.2.1 Hydroxamate-type siderophore uptake

One of the first uptake systems to be discovered in S. aureus was the Fhu uptake system that recognizes and transports a broad range of hydroxamate-type siderophores (48-52).

Currently, four hydroxamate-type/hydroxamate-mixed siderophores have been shown to provide iron to S. aureus via the Fhu uptake system: aerobactin, desferrioxamine B, coprogen and ferrichrome A (48,50). Since S. aureus does not produce hydroxamate-containing siderophores, this system likely utilizes siderophores produced exogenously from other bacteria or fungi found in their immediate environment. The Fhu uptake system (FhuBGC2-D2/D1) is composed of two

paralogous substrate binding proteins, FhuD1 and FhuD2, a heterodimeric permease, FhuB and

FhuG, and an ATPase, FhuC (Figure 1-4) (48-50,52). Insertional mutagenesis of S. aureus with

Tn917-LTV1 into the fhuG coding region, or deleting both fhuD1 and fhuD2 prevented hydroxamate siderophore uptake and subsequent growth on hydroxamate siderophore iron sources (51,52). While FhuD2 facilitates the uptake of many hydroxamate-type siderophores,

FhuD1 appears to transport a smaller subset of hydroxamate-type siderophores and binds its ligands with less affinity compared to FhuD2 (48). Through the use of X-ray crystallography, the

FhuD2 substrate binding mechanism has been elucidated. Co-crystal structures of FhuD2 bound to ferrichrome A and desferrioxamine B revealed two key interacting residues (Arg199 and

Trp197) that hydrogen bond with the structurally conserved hydroxamate groups of the siderophores (53,54). The importance of the Fhu system in S. aureus virulence is still unclear.

Deletion of the FhuBGC components shows a moderate growth impairment of S. aureus in a murine kidney abscess infection model (49). However, other studies focused on the FhuD2 receptor showed that a FhuD2 deletion imparted a more significant S. aureus growth impairment in the blood and kidneys of a murine model of systemic infection (55). Furthermore, it was demonstrated that exogenously supplied Desferal, a mesylate salt of desferrioxamine B used to treat iron overload diseases like thalassaemia, can exacerbate S. aureus infections in mice in a

Fhu-dependent manner (56). This could explain why exogenous treatment of desferrioxamine for patients with thalassaemia showed a positive correlation with an increased rate of S. aureus infection (57).

Figure 1-4 Schematic of known siderophore uptake systems in S. aureus. Hydroxamate siderophores are recognized and internalized by the Fhu system (blue), catecholate siderophores are recognized and internalized by the Sst system (green), SA is recognized and internalized by the Hts system (orange) and SB is recognized and internalized by the Sir system (purple). The substrate receptors are represented by pac-man shapes (FhuD1, FhuD2, SstD, HtsA and SirA) and the heterodimeric permeases are represented by two parallel rectangles (FhuBG, SstAB, HtsBC and SirBC). The ATPase, FhuC (two black circles) is used for the Fhu system and for both Hts and Sir systems, while the Sst system encodes its own ATPase, SstC (two green circles).

1.2.2.2 Catecholate-type siderophore uptake

A second exogenous siderophore uptake system (SstABCD) was recently discovered that recognizes and transports a broad range of catecholate-type siderophores and Fe(III)-complexed catecholamines such as dopamine and epinephrine (58). Previous studies have shown that norepinephrine can reduce and release Fe(III) from transferrin and lactoferrin (59,60). The Sst uptake system is composed of a substrate binding protein, SstD, an ATPase, SstC and a heterodimeric permease, SstA and SstB (Figure 1-4) (61). The substrate receptor SstD showed sub-M affinity for four ferrated catecholamines (norepinephrine, epinephrine, dopamine and L-

DOPA) and various catecholate-type siderophores (bacillibactin, enterobactin and petrobactin).

The presence of catecholamines in human serum or media supplemented with human transferrin 11

enhanced the growth of S. aureus in vitro in a SstABCD-dependent manner (58). Deletion of the

SstABCD operon prevented the growth-promoting effects of catecholamines (58). A murine sepsis model revealed that knocking out the Sst transporter (sst) alone did not result in significant reductions of bacterial counts in the heart, liver or kidneys (58), nor did reducing SstD levels by antisense sstD expression approach effect S. aureus growth in an intraperitoneal rat chamber implant model (61). Furthermore, mice implanted with a pump that delivered a constant supply of epinephrine did not exacerbate the staphylococcal infections when challenged with S. aureus compared to the sst mutant (58). The role of the Sst iron uptake system as a virulence factor in S. aureus remains to be clarified.

1.2.2.3 Staphyloferrin (SA) and (SB) uptake

Iron restricted growth conditions drive S. aureus to produce the siderophores SA and SB.

Both siderophores have been purified and their chemical structures determined (62-66). To utilize Fe(III)-bound SA and SB, S. aureus expresses two iron-regulated siderophore transporters specific for each endogenous siderophore: HtsABC for SA and SirABC for SB (67,68). Unlike the Fhu and Sst systems, HtsABC and SirABC specifically recognize their respective - hydroxycarboxylate-type siderophore with affinities in the sub-nanomolar range (69,70). The SA and SB uptake systems are composed of substrate binding proteins, HtsA and SirA, and heterodimeric permeases, HtsB and HtsC and SirB and SirC, respectively (Figure 1-4). Neither the hts nor sir gene clusters contain a gene that encodes for an ATPase in their respective operons to power the transport of staphyloferrins into the cell. However, studies have shown that

FhuC, the ATPase associated with the Fhu system, is essential for providing energy to facilitate the transport of staphyloferrins (49). 12

Crystal structures for both staphyloferrin receptors have been determined in both the apo- and holo- states in complex with their respective Fe(III)-siderophore (67,69,70). While both SirA and HtsA share low sequence identity (~33%), they maintain a similar tertiary fold composed of two  domains separated by a long -helical bridge. Between the two domains lies the siderophore binding pocket lined with positively charged residues that are required to bind the anionic staphyloferrins. Upon substrate binding, both receptors undergo significant conformational changes that close around the Fe(III)-staphyloferrins.

1.3 Siderophore biosynthesis in Staphylococcus aureus

Siderophore biosynthetic pathways are categorized into two broad classes: nonribosomal peptide synthesis (NRPS) and NRPS-independent siderophore (NIS) synthesis (71). NRPS biosynthetic pathways have been well studied for many siderophores such as enterobactin (72) and for antibiotics like vancomycin (73). This class of siderophore utilizes a peptidic scaffold that is assembled in a stepwise fashion without the need of a ribosome mRNA template (74).

Unlike NRPS siderophore synthesis, NIS siderophore synthesis is considerably less well characterized. NIS biosynthetic pathways do not utilize a peptidic scaffold and the siderophore is instead built from alternating subunits of dicarboxylic acids (citrate, -ketoglutarate) with diamines, amino alcohols and alcohols (75). NIS siderophore synthesis requires ATP-driven synthetases that catalyze siderophore assembly by forming either amide or ester bonds between two siderophore precursors (75). Each assembly step is catalyzed by a NIS synthetase and is categorized into three types (A, B and C) based on their specificity for certain carboxylic acids

(47). Type A synthetases catalyze the formation of an amide or ester bond between either one of the prochiral carboxyl groups in citrate and an amine or alcohol functional group, respectively. 13

Type B synthetases catalyze the formation of an amide, or ester bond between the carboxyl group of-ketoglutarate (-KG) and an amine or alcohol, respectively. Type C synthetases also catalyzes a similar reaction to type A synthetases, but instead uses a monoamide/monoester derivative of citrate as the source of the carboxyl group. For all three types of synthetases, ATP is required to activate the carboxyl group by adenylation. Both S. aureus siderophores, SA and

SB are assembled through the NIS synthesis pathways (47).

1.3.1 Staphyloferrin A

SA production and efflux is encoded by the sfaABCD biosynthetic gene cluster (67,76).

SA is assembled from two citrates and a D-ornithine. A NIS synthetase, SfaD, condenses D- ornithine with citrate to produce a -citryl-D-ornithine intermediate. Next, the intermediate is condensed with a second citrate by SfaB to form the full length SA. D-ornithine is produced from L-ornithine by the predicted pyridoxal 5’-phosphate (PLP)-dependent racemase SfaC.

Finally, SfaA is essential for SA export as a deletion of sfaA resulted in the reduction of SA in the culture supernatant and concurrent accumulation in the cytoplasm (77). SA appears to play an important role in skin colonization and bacterial skin and soft tissue infections as only SA production, and not SB production, was shown to help facilitate skin colonization and abscess formation in mice injected subcutaneously with S. aureus strain Newman (46).

1.3.2 Staphyloferrin B

Compared to SA, the production of SB is more complicated, requiring more enzymatic steps. SB production and efflux is encoded within the sbnA-I biosynthetic gene cluster using substrates citrate, -KG and two molecules of L-2,3-diamionpropionic acid (L-Dap) (Figure 1- 14

5A) (64,78,79). SB assembly is mediated by three NIS synthetases SbnCEF and a PLP- dependent decarboxylase SbnH (Figure 1-5B). First, SbnE condenses citrate and L-Dap to produce the intermediate citryl-L-2,3-diaminopropionic acid. This intermediate is then decarboxylated by SbnH to produce citryl-diaminoethane. Next, SbnF condenses the citryl- diaminoethane with a second molecule of L-Dap to produce the intermediate L-2,3- diaminopropionyl-citryl-diaminoethane. Finally, SbnC adds an -KG molecule to form the full length SB. Encoded within the SB biosynthetic gene cluster is a predicted efflux pump, SbnD.

However, recent studies have shown that while deleting sbnD reduces SB export, it does not completely abrogate SB export suggesting that SB has an alternative route to cross the cell membrane (77). Also encoded in the sbn gene cluster are three other proteins (SbnABG) that are implicated in SB precursor synthesis. SbnA and SbnB are essential for the production of L-Dap as deletion of either gene abrogated SB production (80). However, SB production can be restored if an exogenous source of L-Dap is provided to the culture (80). Citrate is generated by SbnG, which is a novel citrate synthase that uses substrates acetyl-CoA and oxaloacetate (81). Despite our understanding of how SB is produced and imported, integration of the SB biosynthetic pathway with central metabolism with respect to iron deprivation and how SbnA and SbnB function in tandem to produce L-Dap remains poorly understood. To investigate the role of SB in

S. aureus pathogenesis, numerous murine models have been used and have provided mixed results. In one study, S. aureus strain Newman ΔsbnE mutant (loss of SB production) was tested in a murine kidney abscess model, which resulted in a lack of observable abscesses and low bacterial loads compared to the wild type strain (79). However, another study involving a murine sepsis model of S. aureus strain Newman infection revealed that the loss of either SA or SB biosynthesis did not result in any statistically significant reduction in bacterial load in the mouse 15

heart, liver or kidneys (58). Lastly, a recent study involving a murine bacteremia model showed that loss of citrate production, and subsequent staphyloferrin B production, (ΔsbnG/ΔcitZ) from

S. aureus strain Newman resulted in a significant decrease in bacterial loads in the mouse kidneys, liver and heart compared to wild type S. aureus (82).

Figure 1-5 SB biosynthesis in S. aureus. (A) The biosynthetic gene cluster involved in SB production. Genes involved in L-Dap synthesis, siderophore assembly, siderophore efflux and citrate synthesis are colored green, blue, orange, and red, respectively. The function of sbnI is unknown and colored black. (B) The biosynthetic pathway for SB. The final SB schematic is color coordinated to visualize the contribution from each substrate.

1.4 Iron uptake systems in other Staphylococcus species

Compared to S. aureus, few studies have focused on characterizing iron uptake systems in other Staphylococcal species. Based on genomic analyses, most coagulase-negative

Staphylococcal species contain the sfa (SA production) biosynthetic gene cluster and import 17

machinery (htsABC), but lack both the isd (heme uptake) gene set and the sbn (SB production) biosynthetic gene cluster and import machinery (sirABC) (47). A hypothesis is that inclusion of the isd and sbn gene clusters for iron acquisition enhances the pathogenicity of S. aureus (47).

However, exceptions exist to this generalized iron uptake categorization in Staphylococcal species. Staphylococcal species have been identified to contain the sbn biosynthetic gene cluster, which includes the Staphylococcus intermedius group (S. intermedius, Staphylococcus pseudintermedius and Staphylococcus delphini), in addition to containing the machinery for SA production and uptake (83). Recently, Staphylococcus lugdunensis, a coagulase-negative

Staphylococcal species, was shown to utilize Fe(III)-SA, Fe(III)-SB and heme through the

HtsABC, SirABC and Isd uptake systems, respectively (84). S. lugdunensis is unable to produce either SA or SB due to a large deletion in the sfa gene cluster and the absence of the entire sbn gene cluster, respectively (84). Furthermore, S. lugdunensis appears to contain both the Fhu and

Sst iron uptake systems as identified by genomic analyses (84). The functional significance regarding the inability to produce any staphyloferrins is not known.

1.5 S. aureus metabolism

1.5.1 Central metabolism overview

S. aureus metabolism is a complex network of highly regulated enzymes that not only provides energy to the cell, but also provides a full repertoire of metabolic building blocks to permit cell growth and division. The main metabolic engine of S. aureus is central metabolism, which consists of the glycolytic pathway, pentose phosphate pathway and tricarboxylic acid

(TCA) cycle (Figure 1-6) (1,85). Carbohydrates enter central metabolism through the glycolytic 18

pathway or pentose phosphate pathway where they are catabolized to pyruvate. Substrate-level phosphorylation can generate ATP via the glycolytic enzymes phosphoglycerate kinase and pyruvate kinase. Under anaerobic conditions, fermentative pathways are used to reduce pyruvate, primarily to lactic acid by the enzyme lactate dehydrogenase (86). Alternatively, under aerobic conditions, pyruvate is converted to acetyl-CoA by the pyruvate dehydrogenase complex, which can then enter the TCA cycle (87). The TCA cycle is important for generating reducing potential for the cell as well as to provide biosynthetic precursors for anabolic pathways like amino acid or nucleic acid biosynthesis (1). Redox balance is maintained by the reduced dinucleotides directly feeding into the electron transport chain, which drives ATP production via oxidative phosphorylation.

Central metabolic pathways are tightly regulated in S. aureus to optimize metabolic efficiency. The preferential carbohydrate source for S. aureus is glucose (88,89). Under high nutrient conditions containing excess glucose, TCA cycle activity is effectively repressed

(88,90). Therefore, acetyl-CoA generated is shunted away from the TCA cycle and converted to acetate and ATP in a process known as carbon overflow, which occurs until available glucose resources are exhausted (90,91). Carbon overflow only occurs during the exponential growth phase when nutrients are abundant, allowing for rapid growth and cell division. However, once freely available nutrients have been exhausted, S. aureus transitions from the exponential growth phase to the post-exponential growth phase (1). The post-exponential growth phase is primarily driven by the TCA cycle where nonpreferred carbon sources like acetate are used (91,92).

Acetate is converted back to acetyl-CoA and fed into the TCA cycle. In S. aureus, many of these metabolic processes are modulated by numerous metabolic and nutrient-responsive regulators that can respond to fluctuating nutrient levels. 19

Figure 1-6 S. aureus central metabolism. Central metabolism is composed of the glycolytic pathway, pentose phosphate pathway and the tricarboxylic acid (TCA) cycle. Under aerobic conditions, aerobic respiration is responsible for the production of ATP. Under anaerobic conditions, ATP is primarily produced via glycolysis with pyruvate being fermented to lactate.

1.5.2 Major nutrient-response regulators

The ability of S. aureus to adapt its central metabolism processes to fit changing environments ensures survival during both colonization and invasion phases of infection. S. aureus preferentially utilizes glucose as its primary carbon source versus other carbohydrates

(88,89) and glucose utilization is mediated by the catabolite control protein A (CcpA) (93,94).

CcpA regulatory activity is controlled by the phosphorylated corepressor HP, which forms a

tertiary complex together with CcpA and glycolytic intermediates like glucose-6-phosphate or fructose-1,6-bisphosphate (95). In this form, CcpA binds to cre-binding boxes located upstream of CcpA controlled genes and regulates the transcription of various metabolically-linked genes

(94). While the glycolytic pathway is not dramatically up-regulated in S. aureus via CcpA, CcpA instead directly regulates the TCA cycle by repressing transcription when glycolytic activity is high. Most genes involved in the TCA cycle contain putative cre-sites for CcpA binding (96).

Recently, CcpA has been found to also repress proline and arginine biosynthesis, suggesting a regulatory link between carbohydrate and nitrogen utilization (97,98).

Another major global metabolic regulator in S. aureus is the Gram-positive conserved repressor called CodY. CodY responds to multiple intracellular metabolites including branched chain amino acids (BCAA: isoleucine, leucine and valine) and as well as GTP (99). Therefore,

CodY senses carbon and nitrogen availability via BCAAs and responds in part to the stringent response via GTP intracellular concentrations (85). Under conditions of nutrient deprivation, or environmental stress, CodY repression is alleviated by a decrease in intracellular BCAAs and/or

GTP levels. In S. aureus, CodY is responsible for regulating over 100 genes, mostly involved in peptide and amino acid transport and amino acid catabolism (99,100). Specifically, CodY regulates important linkages between major metabolic pathways such as the conversion of pyruvate to oxaloacetate by pyruvate carboxylase and the synthesis of L-glutamate from L- glutamine and -KG by glutamate synthase. CodY is an important factor in controlling many amino acid biosynthetic pathways (e.g. L-serine, L-glutamate and BCAA’s) (99,100). In addition to metabolism, CodY plays an essential role in virulence factor regulation by directly repressing several secreted proteases and indirectly by repressing the agr quorum sensing system, which is used to positively regulate virulence factor gene expression at high cell densities (99-101). 21

As described earlier, S. aureus must acquire iron from its host in order to survive. Iron is an essential nutrient that is used as a cofactor in many enzymes including those utilized in the

TCA cycle and electron transport chain (102). To maintain iron homeostasis, S. aureus tightly regulates the uptake of iron by using a ferric uptake regulator (Fur) to mediate the transcription of the iron uptake systems (103). Under iron-replete conditions, Fur is bound to Fe(II) and binds to a 19 base pair inverted repeat DNA sequence called the Fur-binding box (103). Fur boxes are generally found upstream of the gene under the control of Fur, which typically acts as a transcriptional repressor. Under iron-depletion, Fe(II) bound to Fur is removed and reallocated, which alleviates transcriptional repression. The majority of genes containing Fur-binding boxes are directly related to the import of iron (e.g. isd, fhu, sir, sbn) (104). Fur is also involved in the regulation of virulence gene transcription, either through direct or indirect regulation (e.g. hemolysins, cytotoxins and immunomodulatory proteins) and other global metabolic sensor proteins like CodY (17). The regulatory activity of Fur provides S. aureus protection against killing by neutrophils and is required for full virulence in a murine infection model (105).

1.6 S. aureus metabolic adaptations to the host

S. aureus primarily colonizes the anterior part of the nasal cavity by binding to corneocytes, a type of differentiated nasal epithelial cell, and to proteinaceous material composed of loricrin, involucrin and cytokeratin (106). S. aureus can also bind to the pseudostratified columnar ciliated epithelium located deeper inside the nasal cavity (106). The nose is a dynamic environment where the upper layers of the epithelium are constantly shed in addition to mucus secretions (106). Therefore, S. aureus is constantly removed via shedding and must replicate to sustain persistent colonization (106). S. aureus would also need to overcome nutrient shortages 22

created by competing microbial communities and by the human host’s natural defense mechanisms. Recently, human nasal secretions were analyzed by metabolomics, which revealed many key nutrients were present in low concentrations (107). Yet, available in relatively high concentrations were both urea and glucose; in contrast, several amino acids were not detected at all (L-methionine, L-glutamine, L-tyrosine, L-isoleucine, L-asparagine, and L-aspartate).

Previous studies have shown that the anterior nares of the human nose is severely iron-limited

(108). Under growth conditions that mimic human nasal secretions, S. aureus up-regulated the gene expression of iron uptake systems (e.g. sbn, isd, sir, sst) (107). Additionally, amino acid biosynthetic pathways were also up-regulated, which included pathways for L-glutamate, L- histidine, L-lysine, L-valine, L-leucine, L-isoleucine and L-methionine biosynthesis (107).

Disrupting the production of L-methionine severely impaired S. aureus colonization in the noses of cotton rats (107). Interestingly, no increase in gene expression for secreted virulence factors

(e.g. proteases, hemolysins, toxins) were observed, which was likely due to the fact that the agr quorum sensing is minimal during nasal colonization (108,109). Furthermore, increased induction of the global virulence factor, RNAIII, which responds to the quorum sensing system, reduced nasal colonization of S. aureus in a cotton rat model (110).

Previous studies have shown that iron deprivation in S. aureus induces a phenomenon termed the “iron sparing” response that redirects central metabolism in an effort to conserve iron

(111). Several proteins in the TCA cycle (e.g. aconitase and succinate dehydrogenase) require iron to function, unlike the glycolytic pathway, which does not require iron to convert glucose to pyruvate. In S. aureus, the glycolytic pathway and lactic acid production is up-regulated, while the TCA cycle is down-regulated (111). Also, through the depression of Fur, iron uptake systems are expressed (111). 23

In addition to persistent colonization in the nasal passage, S. aureus can infiltrate the human body by entering the lungs causing pneumonia, or through breaches in the skin, which can lead to a variety of different diseases. Analysis of S. aureus grown in the lungs of mice, a S. aureus pneumonia mouse infection model, resulted in a similar metabolic pattern as exhibited by

S. aureus grown in synthetic human nasal secretions. After a 30 minute incubation of S. aureus in the lungs of mice, gene expression was up-regulated for glycolysis, lactic acid production and amino acid biosynthesis (CodY regulon), while gene expression was down-regulated for gluconeogenesis, TCA cycle and respiration (112). Furthermore, relatively few iron uptake systems were up-regulated (fhuD and htsABC), but not genes involved in heme uptake (isd) or staphyloferrin production, suggesting that some iron is readily available (112).

S. aureus can also survive and proliferate in the circulatory system, which enables S. aureus to be transported throughout the body, allowing it to infect different tissue types. A study of the human serum metabolome revealed a more nutrient rich environment compared to human nasal secretions (113). A more complete set of amino acids were available at higher concentrations as well as ~10-fold higher concentration of glucose (113). Furthermore, higher concentrations of iron are present in human serum (9 mM) (113); however, iron present in the serum is not freely accessible, but is bound to host proteins such as transferrin, lactoferrin and ferritin (114). Transcriptome analysis of S. aureus grown in human serum or blood over two hours revealed similar metabolic changes to growth in synthetic human nasal secretions (115).

Gene expression was highly up-regulated for all iron uptake transport systems (Fur regulon) and for amino acid biosynthetic pathways (CodY regulon). Interestingly, expression of S. aureus toxins was up-regulated in both human blood and serum, including leukotoxin GH, alpha- hemolysin, Panton-Valentine leukotoxin and gamma hemolysin subunits A, B and C (115). 24

Upregulation of these toxins, specifically gamma hemolysin(s), were due to ingestion of S. aureus by polymorphonuclear leukocytes (115). A similar transcriptional profile was also observed for S. aureus grown in an in vivo tissue cage mouse model where both iron uptake systems and secreted toxins were up-regulated (116).

1.7 Intracellular iron release mechanisms

A wealth of knowledge is available on the many iron uptake systems bacteria use to extract iron from their environment (46,114,117). However, less understood are the pathways and reaction mechanisms involved in iron assimilation after iron import into the bacterial cell, particularly for Fe(III)-siderophore complexes. Mechanisms involved in the release of iron from both heme and Fe(III)-siderophore complexes will be discussed below.

1.7.1 Iron release from heme

Heme degradation is required to liberate iron in an effort to repurpose iron for other cellular processes. S. aureus contains two heme degrading enzymes (heme oxygenases (HOs)) termed IsdG and IsdI that were identified based on sequence similarity with the antibiotic biosynthesis monooxygenase family of monooxygenases and their association with the Isd heme uptake system (118,119). Both IsdG and IsdI convert heme to the products 5-oxo--bilirubin and

15-oxo--bilirubin (also called staphylobilins), Fe(II) and formaldehyde (Figure 1-7A) (35,120).

Since the characterization of IsdG and IsdI in S. aureus, other IsdG-like HO homologs have been identified in Gram-positive bacteria that include Bacillus anthracis and Staphylococcus lugdunensis. IsdG-like HOs are homodimers that adopt a ferredoxin-like fold (121). Upon heme binding to IsdG/IsdI, the porphyrin ring is severely distorted and referred to as heme ruffling 25

(34). Unlike canonical HOs that convert heme to biliverdin, the IsdG-like family of HOs likely utilizes a unique heme degradation mechanism to convert heme to staphylobilin.

Figure 1-7 Schematic of intracellular iron release from heme and Fe(III)-siderophores in S. aureus. (A) Imported heme can either be incorporated into proteins, or degraded by heme oxygenases (IsdG/IsdI) to release Fe(II). (B) Imported Fe(III)-siderophores are reduced by either general, or specific siderophore reductases to release Fe(II).

Recently, a pyridine nucleotide-disulfide oxidoreductase (PNDO), termed iron utilization oxidoreductase (IruO), from S. aureus was shown through in vitro studies to provide electrons to

IsdG and IsdI, leading to the degradation of heme to the staphylobilins and Fe(II) (122).

Furthermore, gene deletion studies in S. aureus identified a second iron-regulated reductase,

NtrA, in addition to IruO, contributed to growth on heme as a sole iron source (123).

Interestingly, a single knockout of either iruO or ntrA in S. aureus did not diminish growth with 26

heme as the sole iron source (123). Only a double mutant (iruO ntrA) prevented S. aureus growth on heme as the sole iron source (123). Several homologs of IruO have been identified in Gram- positive bacteria including S. lugdunensis, B. anthracis and Listeria monocytogenes, which suggests that IruO might also contribute to heme degradation in those organisms as well (122).

1.7.2 Iron release from siderophores

One of the main pathways used for iron assimilation from Fe(III)-siderophore complexes is through iron reduction reactions (36). Iron reduction to the Fe(II) form destabilizes the Fe(II)- siderophore complex both kinetically and thermodynamically compared to the Fe(III)- siderophore complex (36). Due to the large variation in siderophore structure and iron coordination, midpoint potentials for Fe(III)-siderophores complexes vary widely across the three major classes of siderophores. Catecholate-type siderophores possess the lowest values

(and highest affinity complexes) at E’0, pH 7.0 of approximately -750 mV, hydroxamate-type siderophores at E’0, pH 7.0 of approximately -300 mV and carboxylate-type siderophores at E’0, pH 7.0 approximately -100 mV (36). Therefore, many Fe(III)-siderophore complexes at the lower midpoint potentials are within the physiological range of cellular reductants like NADH

(E’0, pH 7.0 of -320 mV) (36). Additionally, various general reductases have been identified that could contribute to Fe(III)-siderophore reduction, but are not considered iron reductase exclusive. For example, the NAD(P)H:flavin oxidoreductase (Fre) from E. coli can reduce hydroxamate-type siderophores like Fe(III)-ferrichrome by reducing and releasing free flavins like FAD (124,125). Currently, four examples of siderophore reductases have been identified and characterized, E. coli (FhuF and YqjH), Bacillus halodurans (FchR), and Vibrio cholerae (ViuB)

(126-129). Additionally, two other oxidoreductases, IruO and NtrA, from S. aureus have been 27

shown to play a role in hydroxamate-type siderophore and SA utilization, respectively (123).

Both IruO and NtrA are hypothesized to directly bind and reduce iron from Fe(III)-siderophore complexes (Figure 1-7B) (123). All six reductases have been shown to be Fur-regulated proteins and up-regulated during iron-restricted growth conditions (123,127,130-132). These six siderophore reductases utilize either an iron-sulfur cluster (FeS), or a FAD/FMN cofactor to catalyze iron reduction reactions.

For the FeS containing siderophore reductases FhuF and FchR, the reduced FeS cluster embedded in the protein transfers an electron to the Fe(III)-siderophore complex reducing Fe(III) to Fe(II) (126,127). Both FhuF and FchR are capable of reducing hydroxamate-type siderophores, but not the catecholate-type siderophore enterobactin (126,127). However, the source of electrons required to maintain the reduced FeS state remains elusive. A study by

Miethke, et al. 2011, suggested that one of three ferredoxin-ferredoxin:NADP+ reductase systems most likely acts as the electron donor to FchR (127). Additionally, siderophore interacting proteins, YqjH and ViuB reduce catecholate-type siderophores enterobactin and vibriobactin, respectively (128,129). YqjH and ViuB are homologs and share structural similarity with the ferredoxin reductase-like family (128). Instead of a FeS cofactor, they utilize a FAD cofactor to transfer an electron from NADPH via FAD to their Fe(III)-siderophore substrates

(128). Mechanistically, YqjH has been investigated thoroughly, revealing a two-step single- electron transfer from the fully reduced FADH2 (FADhq; hydroquinone state) to two Fe(III)- enterobactin complexes (128). FADH2 is generated from the initial transfer of electrons from a single NADPH molecule. The midpoint potential of YqjH was measured to be -331 mV

(FADox/FADsq), and -314 mV (FADsq/FADhq), (FADox; oxidized state and FADsq; semiquinone state) (128). Due to the higher midpoint potential of enterobactin, E. coli contains a enterobactin 28

esterase (Fes) that cleaves the backbone of enterobactin to dihydroxybenzoylserine, which has a significantly lower midpoint potential at -350 mV and thereby allowing reduction of the Fe(III)- siderophore complex by YqjH (36,128). No studies have evaluated Fe(III)-siderophore reduction mechanisms for either IruO or NtrA in S. aureus.

1.8 Objectives

Iron is an essential nutrient required for the growth and survival of S. aureus. Therefore, characterizing siderophore assembly and the subsequent iron assimilation reactions upon internalization could lead to novel therapeutic targets for drug development. Antibiotic- conjugated siderophores have shown promise in treating previously antibiotic resistant pathogens such as the delivery of ampicillin, a -lactam antibiotic, to pathogenic E. coli using enterobactin- ampicillin conjugates (133). Presently, studies have primarily focused on siderophore assembly and uptake. However, relatively little is understood on how the iron-deprived metabolic state in bacteria drives siderophore production and iron assimilation. In S. aureus, the rationale for producing two siderophores is also unclear. With the recent identification of two siderophore reductases in S. aureus (123), elucidation of the reductive mechanisms required to release iron from Fe(III)-siderophore complexes also remains to be determined.

The overall objective of this work is to understand how metabolic alterations due to iron deprivation influences siderophore-mediated iron uptake. I hypothesize that SB biosynthesis is linked to central metabolism through the activities of the SB precursor enzymes SbnA, SbnB and

SbnG. Additionally, I hypothesize that S. aureus utilizes an “iron-free” reductive system to assimilate iron derived from Fe(III)-siderophores.

A second citrate synthase, SbnG, that participates in SB synthesis has been identified in

S. aureus (81). Yet, SbnG does not share sequence similarity to any known TCA cycle citrate synthase. Therefore, a combination of X-ray crystallography, structural superpositions and site- directed mutagenesis were used to gain insight into the catalytic mechanism of SbnG. Compared to TCA cycle citrate synthases, SbnG is structurally unique, yet contained a conserved active site architecture suggesting a conserved reaction mechanism. Active site SbnG variants attenuated the production of SB by impeding the production of citrate in S. aureus mutant strains.

SB assembly requires the production of L-Dap by SbnA and SbnB (80). Despite the prevalence of L-Dap in nature, a L-Dap biosynthetic pathway has yet to be experimentally defined. The goals of this study were to identify L-Dap precursors to characterize the biological functions of SbnA and SbnB, and to investigate SbnA substrate specificity. The functions of

SbnA and SbnB were characterized using LC-ESI-MS, NMR and biochemical assays. The substrates were identified as OPS and L-glutamate, which were converted to the products L-Dap and -KG via a novel nonproteinogenic amino acid intermediate. SbnA substrate specificity was investigated by site-directed mutagenesis and enzyme kinetics to identify several active site residues responsible for substrate recognition and enzyme turnover. Additionally, the structures of SbnA and SbnB bound to substrates were solved by X-ray crystallography in an effort to propose a catalytic mechanism.

Lastly, S. aureus contains an oxidoreductase, IruO, that participates in iron assimilation from Fe(III)-hydroxamate-type siderophores (123). The aim of this study was to characterize

IruO as a siderophore reductase that directly binds to hydroxamate-type siderophores and reduces the Fe(III)-siderophore complex. A combination of UV-Vis/fluorescence spectroscopy and enzyme kinetics revealed that IruO binds and reduces Fe(III)-hydroxamate-type siderophore 30

complexes (ferrichrome A and desferrioxamine B) by transferring electrons from NADPH to an

FAD cofactor, which in turn transfers a single electron to a Fe(III)-siderophore complex.

Furthermore, X-ray crystallography was used to identify the structure of IruO in two distinct conformational states and mediated by a single intramolecular disulfide bond.

Chapter 2: Experimental Procedures

2.1 S. aureus growth conditions

All bacterial strains and plasmids employed in this study can be found in Table 2-1.

Unless otherwise indicated, in vivo experiments were performed using S. aureus Newman. For genetic manipulations, E. coli and S. aureus were routinely cultured in Difco Luria-Bertani Broth

(LB; BD), and Difco Tryptic Soy Broth (TSB; BD), respectively. For selection and maintenance of plasmids, media was supplemented with 100 μg ml-1 of ampicillin for E. coli, or 10 μg ml-1 of chloramphenicol for S. aureus. For bacterial growth under iron-restriction, RPMI-1640 media was prepared as directed by the manufacturer (Life Technologies) in sterile polypropylene vessels using water purified with a Milli-Q water filtration system. The non-metabolizable iron chelator ethylenediamine-di(O-hydroxyphenylacetic acid) (EDDHA, LGC Standards GmbH) was added to media, where indicated.

Table 2-1 Bacterial strains, plasmids, and oligonucleotides. Strain or plasmid designation Description Reference Strains Staphylococcus aureus H627 (Newman) Wild type clinical osteomyelitis isolate (134) H187 (RN6390) Wild type prophage-cured laboratory (135) strain H2708 (ΔcitZ sbnG) Newman ΔcitZ ΔsbnG; citrate synthase- (82) deficient strain H2871 (ΔcitZ sbnG/pALC) H2708 containing blank pALC2073 (82) vector (vehicle control) H2856 (ΔcitZ sbnG/psbnG) H2708/psbnG; citZ sbnG mutant (82) complemented with sbnG H2973 (ΔcitZ sbnG/pE46Q) H2708/pE46Q citZ sbnG mutant Chapter 3 complemented with sbnG variant E46Q H2974 (ΔcitZ sbnG/pH47A) H2708/pH47A citZ sbnG mutant Chapter 3 complemented with sbnG variant H47A H2975 (ΔcitZ sbnG/pR72A) H2708/pR72A citZ sbnG mutant Chapter 3 complemented with sbnG variant R72A H2976 (ΔcitZ sbnG/pH96A) H2708/pH96A citZ sbnG mutant Chapter 3 complemented with sbnG variant H96A H2977 (ΔcitZ sbnG/pE151Q) H2708/pE151Q citZ sbnG mutant Chapter 3 complemented with sbnG variant E151Q H2982 (ΔcitZ sbnG/pD177A) H2708/pD177A citZ sbnG mutant Chapter 3 complemented with sbnG variant E46Q Escherichia coli DC10B E. coli cloning vector for cytosine (136) methylation of foreign DNA prior to transformation into S. aureus Plasmids pALC2073 E. coli/Staphylococcus shuttle vector: (137) ApR CmR psbnG pALC2073-sbnG; pALC2073 derivative (82) for expression of sbnG pE46Q pALC2073 derivative expressing sbnG Chapter 3 variant E46Q pH47A pALC2073 derivative expressing sbnG Chapter 3 variant H47A pR72A pALC2073 derivative expressing sbnG Chapter 3 variant R72A pH96A pALC2073 derivative expressing sbnG Chapter 3 variant H96A pE151Q pALC2073 derivative expressing sbnG Chapter 3 variant E151Q pD177A pALC2073 derivative expressing sbnG Chapter 3 variant D177A

2.2 Construction of S. aureus mutant strains

2.2.1 Site-directed mutagenesis of S. aureus sbnG for in vivo assays

To assess the role of residues implicated in SbnG enzymatic function in the production of

SB within the cell, whole plasmid mutagenesis was employed to generate complementation vectors expressing sbnG mutated for these residues. Using the previously constructed psbnG vector (a pALC2073 derivative expressing wild type sbnG) as a template, complementary mutagenesis primers were used to amplify plasmids bearing SbnG variants E46Q, H47A, R72A,

H96A, E151Q, and D177A. All mutagenic primers used are summarized in Table 2-2. Following amplification, methylated template DNA was removed through digestion with DpnI, the enzyme was heat-inactivated at 80 °C for 5 min, and the resulting reaction mixes were used in the transformation of E. coli DC10B. Plasmids bearing the correct mutation were confirmed through sequencing, and introduced into S. aureus Newman ΔcitZ sbnG (H2708) through electroporation.

The sbnG variants were constitutively expressed, without induction, by the leaky Pxyl/tetO promoter of pALC2073 (137).

Table 2-2 Primers for SbnG variants. Primers for SbnG variants used for in vivo experiments

SbnG variant Primer pair Sequence (5’-3’)a E46Q g136c_a138g-F GGTATGACTTTGTTGTGATTGATACACAGCACGTGGCGA TTAATGAT and and g136c_a138g-R ATCATTAATCGCCACGTGCTGTGTATCAATCACAACAAA GTCATACC H47A c139g_a140c-F GGGTATGACTTTGTTGTGATTGATACAGAAGCCGTGGCG ATTAATGATG and and c139g_a140c-R CATCATTAATCGCCACGGCTTCTGTATCAATCACAACAA AGTCATACCC R72A c214g_g215c-F GCAGCGCATATTATACCAATTGTAGCTGTCACTGCAGTG A and and c214g_g215c-R TCACTGCAGTGACAGCTACAATTGGTATAATATGCGCTG C H96A c286g_a287c-F TGCGAGAGGTATTATTGTGCCAGCCGTTAAAGATCGTGA G and and c286g_a287c-R CTCACGATCTTTAACGGCTGGCACAATAATACCTCTCGC A E151Q g451c_a453g-F ATATTATGGTGATTGCCATGATACAGGATGTTGAAGGGG TTATGGC and and g451c_a453g-R GCCATAACCCCTTCAACATCCTGTATCATGGCAATCACC ATAATAT D177A a530c-F GTCGAAGGTGCCGCAGCTTTATCGCAGTCACTT and and a530c-R AAGTGACTGCGATAAAGCTGCGGCACCTTCGAC Primers for SbnG variants used for in vitro assays

SbnG variant Primer Sequence (5’-3’)a E46Q g136c-F GTGATTGATACACAACACGTGGCGATTAATGATGAG H47A c139g_a140c-F GTGATTGATACAGAAGCCGTGGCGATTAATGATGAG R72A c214g_g215c-F GCGCATATTATACCAATTGTAGCTGTCACTGCAGTG H96A c286g_a287c-F AGAGGTATTATTGTGCCAGCCGTTAAAGATCGTGAG E151Q g451c-F ATGGTGATTGCCATGATACAAGATGTTGAAGGGGTTATG D177A a530c-F GACATGATAGTCGAAGGTGCCGCAGCTTTATCGCAG aItalicized sequences represent the codon altered in the indicated SbnG variant, with the underlined sequences denoting the exact nucleotides changed relative to the S. aureus Newman wild type sequence

2.3 Cloning, expression and purification for biochemical assays and structure determination

2.3.1 Cloning, expression and purification of SbnG and variants

The SbnG coding region was cloned into pET28a as previously described (81). SbnG variants (E46Q, H47A, R72A, H96A, E151Q and D177A) were produced using a modified whole plasmid PCR technique as previously described (Table 2-2) (138). SbnG variants were introduced into E. coli BL21(λDE3) and the mutations were confirmed by DNA sequencing.

6x-His-SbnG and variants (E46Q, H47A, R72A, H96A, E151Q and D177A) were expressed from E. coli BL21(λDE3) cells inoculated from 5 mL overnight cultures into 2x YT media supplemented with 25 g/mL kanamycin and grown at 30 °C to an optical density at 600 nm (OD600) of ~0.8. Cultures were then induced with 0.3 mM isopropyl β-D- thiogalactopyranoside (IPTG) and incubated for ~16 hours at 25 °C with shaking at 200 rpm.

Cell pellets were collected by centrifugation at 4400 × g for 10 min, resuspended in 50 mM Tris

(pH 8.0), 100 mM NaCl and lysed at 10,000 psi with an EmulsiFlex-C5 homogenizer (Avestin).

The supernatants were isolated after centrifugation at 39,000 × g for 45 min and 6x-His-SbnG variants were purified using a 5 mL HisTrap HP column (GE Healthcare) with a linear imidazole gradient (0-500 mM). Protein fractions were dialyzed into 50 mM Tris (pH 8.0), 100 mM NaCl at 4 °C. The 6x-His-tag was removed by thrombin digestion at a 1:500 mass ratio (thrombin to protein) and incubated over 16 hours at 4 °C, followed by dialysis into 50 mM Tris (pH 8.0) for

2 hours at 4 °C. SbnG and variants were further purified by anion exchange chromatography using a Source 15Q column (GE Healthcare) equilibrated with 50 mM Tris (pH 8.0) and eluted with a NaCl gradient (0-500 mM). SbnG and variants were dialyzed into 20 mM Tris (pH 8.0), concentrated to 20 mg/mL and stored at -80 °C. Selenomethionine-incorporated (SeMet) SbnG 36

was expressed by methods previously described using a defined media with exogenously added selenomethionine (139) and purified in a similar manner to native SbnG.

2.3.2 Cloning, expression and purification of SbnA and variants

The SbnA coding region was previously cloned from S. aureus strain Newman genomic

DNA into the pET28a plasmid. Site-directed mutagenesis of pET28a-SbnA was performed using a modified multi-site whole plasmid PCR technique (140). Fifty nanograms of plasmid template was incubated with 0.2 mM dNTPs, 3% dimethyl sulfoxide, 0.5 mM NAD+, 0.36 M 5’- phosphorylated forward primer(s), 5 U/l Ampligase (Epicentre), 2 U/l Phusion (New England

BioLabs) in Phusion HF buffer. Mutagenic primers are shown in Table 2-3. The PCR conditions were 98 °C for 30 s, then 30 cycles of 98 °C (15 s), 55 °C (60 s), 72 °C (30 s/kilobase) and a final extension cycle at 72 °C for 10 min. The PCR products were then digested with DpnI and transformed into Escherichia coli BL21(DE3) by electroporation. All mutations were confirmed by DNA sequencing.

Table 2-3 A list of phosphorylated mutagenic primers used to generate SbnA active site variants. Primer Description Primer Sequencea pET28a-sbnA K47A 5’-/5Phospho/cctggaggcagcatggcagatcgacctgcc-3’ pET28a-sbnA R132A 5’-/5Phospho/ggttatttaatgactgctattgcaaaggtgcaagaactg-3’ pET28a-sbnA Y152F 5’-/5Phospho/gcatattggattaatcaatttgcgaatgagttaaattgg-3’ pET28a-sbnA S185G 5’-/5Phospho/gtcgcgccagtcggcacgacaggtagcattatggg-3’ pET28a-sbnA 5’-/5Phospho/gcatattggattaatcaatttgcgaatgagttaaattgg-3’ Y152F/S185G 5’-/5Phospho/gtcgcgccagtcggcacgacaggtagcattatggg-3’ aUnderlined oligonucleotides are mutated from S. aureus strain Newman DNA sequence

SbnA and variants cultures were inoculated into 2x YT media supplemented with 25 mg/mL kanamycin at 30 °C, grown to an optical density of ~0.8, and expression was induced by adding 0.3 mM IPTG and grown overnight at 25 °C. Cells were lysed in 50 mM Tris pH 8.0, 100 mM NaCl and 2 mM Tris(2-carboxyethyl)phosphine (TCEP) (Gold Biotechnology) using an

EmulsiFlex-C5 homogenizer (Avestin) and cell debris was removed by centrifugation. Proteins were purified with a HisTrap HP column (GE Healthcare), eluted with imidazole and dialyzed into 20 mM Tris pH 8.0 and 2 mM TCEP. The 6x-His tag was removed by thrombin digestion for ~3 hours at room temperature. SbnA and variants were further purified using a Source 15Q column equilibrated with 20 mM Tris pH 8.0, 2 mM TCEP and eluted with a linear NaCl gradient to 500 mM. SbnA and variants were dialyzed into 20 mM Tris pH 8.0, 100 mM KCl, 2 mM TCEP and concentrated to ~20 mg/mL. The purification protocol for wild type SbnA-PLP for structure determination was modified by not including TCEP during the purification procedure and SbnA was passed through a second HisTrap HP column instead of a Source 15Q column. All protein preparations were concentrated to ~20 mg/ml and stored at -80 °C.

2.3.3 Cloning, expression and purification of SbnB

sbnB was amplified from S. aureus Newman genomic DNA and cloned into pET28a with an N-terminal 6x-His tag and transformed into Escherichia coli BL21(DE3). Cultures were inoculated into 2x YT media supplemented with 25 mg/mL kanamycin at 30 °C, grown to an optical density of ~0.8, and expression was induced by adding 0.3 mM IPTG and grown overnight at 25 °C. Cells were lysed in 50 mM Tris pH 8.0, 100 mM NaCl and 2 mM TCEP using an EmulsiFlex-C5 homogenizer (Avestin) and cell debris was removed by centrifugation.

Protein was purified with a HisTrap HP column (GE Healthcare), eluted with imidazole and 38

dialyzed into 20 mM Tris pH 8.0 and 2 mM TCEP. The 6x-His tag was removed by thrombin digestion for ~3 hours at room temperature. SbnB was further purified using a Source 15Q column equilibrated with 20 mM Tris pH 8.0, 2 mM TCEP and eluted with a linear NaCl gradient to 500 mM. SbnB was dialyzed into 20 mM Tris pH 8.0, 100 mM KCl, 2 mM TCEP and concentrated to ~20 mg/mL. The purification protocol for SbnB for structure determination was modified by not including TCEP during the purification procedure. SeMet labelled SbnB was expressed using a previously described method using a defined media with exogenously added selenomethionine (139) and purified as described above without the inclusion of TCEP.

2.3.4 Cloning, expression and purification of IruO

NWMN_2274 (IruO) was cloned from S. aureus Newman genomic DNA into pET28a plasmid as previously described in detail (122). pET28a-iruO was introduced into E. coli

BL21(λDE3) Cultures were inoculated into 2 x YT media supplemented with 25 mg/mL kanamycin at 30°C, and grown to an optical density at 600 nm of ~0.8. Cultures were then induced with 0.3 mM IPTG and incubated for ~16 hours at 25°C with shaking at 200 rpm. Cell pellets were centrifuged at 4400 × g for 10 min and resuspended in 50 mM Tris (pH 8.0), 100 mM NaCl, 2 mM TCEP. Cell lysis was performed at 10,000 psi with an EmulsiFlex-C5 homogenizer (Avestin). The supernatant was isolated after centrifugation at 39,000 × g for 45 min and 6x-His-IruO was purified using a 5 mL HisTrap HP column (GE Healthcare) with a linear imidazole gradient (0-500 mM). Protein fractions were dialyzed into 50 mM Tris-HCl (pH

8.0), 100 mM NaCl, and 2 mM TCEP at 4°C. The 6x-His-tag was removed by thrombin

(Haemotologic Technologies) digestion at a 1/500 (w/w) thrombin to protein ratio and incubated for over 24 hours at 4°C. Digested protein was then dialyzed into 50 mM Tris-HCl (pH 8.0), 2 39

mM TCEP for 2 hours at 4°C. IruO was then purified by anion exchange chromatography using a Source 15Q column (GE Healthcare) equilibrated with 50 mM Tris-HCl (pH 8.0) and 2 mM

TCEP. IruO was eluted with a NaCl gradient (0-500 mM). IruO was dialyzed into 50 mM Tris

(pH 8.0), 300 mM KCl, and 2 mM TCEP overnight at 4 °C and concentrated to 15 mg/mL. The oxidized version of IruO was purified in a similar manner with the exceptions that TCEP was removed from all purification steps and the cell lysates were exposed to molar excess potassium ferricyanide for approximately 30 minutes. IruO prepared in buffer containing TCEP is referred to as reduced IruO (rdIruO). SeMet labeled IruO was produced by a previously described method using a defined media with exogenously added selenomethionine (139) and purified as described above. NWMN_0732 was purified as previously described using the same protocol as for IruO purification (122).

2.3.5 Expression and purification of SB biosynthetic enzymes

SB biosynthesis genes sbnC, sbnE, sbnF and sbnH were amplified from S. aureus

Newman genomic DNA and cloned into pET28a with an N-terminal 6x-His tag and transformed into Escherichia coli BL21(DE3) as previously described (78). Cultures were inoculated into 2x

YT media supplemented with 25 mg/mL kanamycin at 30 °C, grown to an optical density of

~0.8, and expression was induced by adding 0.3 mM IPTG and grown overnight at 25 °C. Cells were lysed in 50 mM Tris pH 8.0, 100 mM NaCl using an EmulsiFlex-C5 homogenizer

(Avestin) and cell debris was removed by centrifugation. The proteins were purified using a

HisTrap HP column (GE Healthcare), eluted with imidazole and dialyzed into 20 mM Tris pH

8.0. The 6x-His tag was removed by thrombin digestion for ~3 hours at room temperature. SB biosynthesis proteins were further purified using a Source 15Q column equilibrated with 20 mM 40

Tris pH 8.0 and eluted with a linear NaCl gradient to 500 mM. Proteins were dialyzed into 20 mM Tris pH 8.0, 100 mM NaCl and concentrated to ~20 mg/mL.

2.3.6 Expression and purification of IsdI

Cultures of E. coli BL21(DE3) containing pET15b-IsdI were inoculated into LB supplemented with 100 g/mL ampicillin and grown at 37 °C until an optical density at 600 nm

(OD600) of ~0.75 was attained. The cultures were then induced with 0.25 mM IPTG and grown overnight at 25 °C with shaking at 200 rpm. Cells were harvested by centrifugation at 4800 × g for 15 min and resuspended in 20 mM Tris pH 7.5 and 200 NaCl. Cells were lysed using an

EmulsiFlex-C5 homogenizer (Avestin) at 10,000 psi and then spun down by centrifugation at

12000 rpm for 40 min. 6x-His-IsdI was purified from the supernatant using a 5 mL HisTrap HP column (GE Healthcare) with a linear imidazole gradient up to 500 mM. Protein fractions were dialyzed overnight back into 20 mM Tris pH 7.5 and 200 NaCl. The 6x-His tag was cleaved with

TEV protease at a1:10 mass ratio (TEV to protein) and incubated over 3 days. TEV protease and

6x-His tag were removed from IsdI by running through the HisTrap column and collecting the flow-through, which only contained IsdI. Purified IsdI was then concentrated to ~15 mg/mL and stored at -80 °C.

2.4 Crystallization and structure determination

2.4.1 SbnG and E151Q variant structure determination

SeMet-labeled SbnG crystals were grown by hanging drop vapor diffusion at room temperature in 1:1 mixtures of 20 mg/mL SbnG in 20 mM Tris (pH 8.0) with reservoir solution of 0.2 M calcium acetate, 0.1 M imidazole (pH 9.0) and 4-6% (w/v) polyethylene glycol (PEG)

8000. Crystals were flash frozen in liquid nitrogen following a brief soak in the same buffer 41

supplemented with 20% (v/v) PEG 400 for cryoprotection. Multi-wavelength anomalous diffraction data for SeMet protein crystals were collected at the Canadian Light Source (CLS) on beam line 08B1. The data were processed and scaled using XDS (141,142). Crystals grew in the space group P312 with one molecule in the asymmetric unit. Phase determination and model building were done using AutoSol (initial figure of merit 0.73) and Autobuild (225 of 259 residues built) programs in Phenix (143). Native SbnG was crystallized under similar conditions with the addition of 5 mM citrate to the reservoir and diffraction data were collected at the

Stanford Synchrotron Radiation Laboratory (SSRL) on beamline 7-1.

SbnG E151Q crystals were grown by sitting drop at room temperature in a 1:1 mixture of

20 mg/mL SbnG E151Q in 5 mM oxaloacetate, 5 mM coenzyme A and 20 mM Tris (pH 8.0) with reservoir solution of 5% (v/v) tascimate (pH 7.0), 0.1 M 4-(2-hydroxyethyl)-

1-piperazineethanesulfonic acid (HEPES) (pH 7.0), 10% (w/v) polyethylene glycol monomethyl ether (PEG MME) 5000 and ~40 mM guanidine hydrochloride. SbnG E151Q crystals were soaked for 10 minutes in 5 mM oxaloacetate/CoA mixture in the same buffer and supplemented with 20% (v/v) PEG 400 for cryoprotection before data collection with a Rigaku MicroMax 007-

HF generator, VariMax HR optics and Saturn CCD 944+ detector. The datasets were processed using Mosflm (144) and scaled using Scala (145). The native structure was solved by molecular replacement using the SeMet-labeled coordinates as the search model in Molrep (146) from the

CCP4 program suite (147). Manual building was done using Coot (148) and refinement was performed with Refmac5 (149) using translation libration screw (TLS) (150) parameters with eight TLS groups. SbnG E151Q was solved by molecular replacement using native SbnG coordinates as a search model in Phaser-MR (151) from Phenix (143). Manual building was done

using Coot (148) and refinement was performed with phenix.refine using TLS refinement. Data collection and refinement statistics are shown in Table 2-4.

Table 2-4 X-ray diffraction data collection and refinement statistics for SbnG. SeMet-SbnG SbnG SbnG-E151Q Peak Inflection Remote Data collectiona Wavelength (Å) 0.97912 0.97929 0.97345 1.00000 1.54180 Resolution range (Å) 33.42 – 2.0 33.44 – 2.0 33.41 – 2.0 76.72 – 1.85 27.05 – 2.60 (2.11 – 2.0) (2.11 – 2.0) (2.11 – 2.0) (1.95 – 1.85) (2.70 – 2.60) Unit cell dimensions (Å) a = 74.2, a = 74.2, a = 74.2, a = 74.3, a = 77.2 b = 74.2, b = 74.2, b = 74.2, b = 74.3, b = 77.2 c = 77.0 c = 77.0 c = 77.0 c = 76.7 c = 75.8 Unique reflections 16550 16573 16538 20746 7119 Completeness (%) 99.8 (100) 99.8 (100) 99.8 (100) 99.5 (100) 87.6 (87.2) Mean I/σI 13.2 (3.0) 14.1 (3.1) 13.9 (2.8) 14.3 (4.6) 11.5 (5.1) Redundancy 7.5 (7.5) 7.5 (7.5) 7.5 (7.5) 8.6 (8.7) 4.9 (4.8)

Rmerge 0.056 (0.266) 0.053 (0.254) 0.054 (0.279) 0.089 (0.314) 0.040 (0.119) Wilson B (Å2) 19.9 20.5 20.6 24.6 39.1 Refinement

Rwork (Rfree) ------0.197 (0.211) 0.228 (0.272) B-factors (Å2) ------All atoms ------27.7 56.1 Protein ------28.1 56.6 Ligands ------37.1 59.7 Water ------20.4 36.3 r.m.s.d. bond length (Å) ------0.009 0.004 Ramachandran plot In most-favourable region ------97.0 97.0 In disallowed regions ------0.0 0.46 PDB Accession Codes ------4TV5 4TV6 aValues in parenthesis represent highest resolution shell

2.4.2 SbnA and variants Y152F and Y152F/S185G structure determination

SbnA was dialyzed into 20 mM Tris pH 8.0 and concentrated to ~20 mg/mL for crystallization. Crystals were optimized by hanging drop vapor diffusion in 2 L drops containing 1 L protein solution and 1 L of 0.1 M Tris pH 8.5, 0.2 M MgCl2, 20-25 % PEG

3350. SbnA crystals were soaked for ~1 minute in mother liquor supplemented with 16 % glycerol and flash frozen in liquid nitrogen. SbnA prepared with 2 mM TCEP for substrate soaking experiments was concentrated to ~20 mg/mL for crystallization and optimized by sitting drop vapor diffusion. Each drop contained 2 L of protein solution mixed with 2 L of 0.9-1.1

M sodium citrate pH 6.5. SbnA crystals were further optimized by seeding drops with 0.4 L of serially diluted crushed crystals in the same solution. Crystals were soaked with 10 mM (O- phospho-L-serine) OPS for ~30 minutes followed by a brief soak in mother liquor supplemented with 30 % glycerol and flash frozen in liquid nitrogen. Crystals of SbnA variants Y152F and

Y152F/S185G were grown by sitting drop diffusion with drops containing 1 L of protein solution and 1 L of 0.2 M MgCl2, 0.1 M Tris pH 8.5 and 29-33 % PEG 4000. Crystals were briefly soaked in mother liquor supplemented with 30 % glycerol and flash frozen in liquid nitrogen.

Diffraction data for the SbnA-PLP complex were collected on beamline 08B1-1 at the

CLS. Data were processed using HKL2000 (152). The crystals grew in the space group P21212 with one molecule in the asymmetric unit. The molecular replacement search model was generated in the program Sculptor (153) from the Phenix program suite (143) using the

Thermatoga maritima OASS structure (PDB ID: 1O58, 37 % identity over 295 of 326 amino acids). The search model excluded the first two histidine residues, loop residues 186, 193 and

contained several conservative mutations in surface residues to minimize differences in sidechain conformers. Molecular replacement phases were obtained using Phaser-MR (151) and an initial model of SbnA-PLP was built using Autobuild (154). The structure was manually built with

Coot (148) and refined with Refmac (149) from the CCP4 program suite (147). Diffraction data for SbnA-AA and the SbnA variants Y152F and Y152F/S185G were collected at SSRL on beamline 7-1. SbnA-AA data was integrated and scaled using XDS (141,142) and phased using

Phaser-MR. SbnA Y152F and Y152F/S185G were integrated with Mosflm (155) and scaled with

Scala (156) from the CCP4 program suite (147). SbnA-AA, Y152F, and Y152F/S185G were manually built using Coot (148) and refined using phenix.refine (157) with 6, 4 and 5 translation libration screw (TLS) groups, respectively (150). Data collection and refinement statistics are provided in Table 2-5.

Table 2-5 X-ray diffraction data collection and refinement statistics for SbnA and variants. SbnA-PLP SbnA-AA SbnA Y152F SbnA Y152F/S185G Data Collectiona Wavelength (Å) 0.97952 1.15869 1.12709 1.12709 Resolution Range (Å) 50.00 – 1.45 37.08 – 1.92 35.38 – 1.50 34.92 – 1.50 (1.50 – 1.45) (1.99 – 1.92) (1.55 – 1.50) (1.55 – 1.50) Unit cell dimension (Å) a = 56.08 a = 56.74 a = 56.36 a = 55.98 b = 115.83 b = 117.41 b = 115.99 b = 115.17 c = 45.14 c = 47.82 c = 45.45 c = 44.67 Unique reflections 53071 25069 47648 46567 Completeness (%) 97.3 (94.7) 99.5 (93.2) 98.2 (88.9) 98.9 (93.4) Redundancy 5.9 (5.6) 5.9 (5.5) 6.6 (4.6) 6.3 (4.3) Mean I/I 28.1 (3.3) 13.1 (2.6) 18.6 (5.3) 19.66 (4.5) Rmerge 0.055 (0.413) 0.082 (0.536) 0.062 (0.208) 0.061 (0.242) Wilson B-factor (Å2) 12.4 22.4 12.7 11.2 Refinement Rwork (Rfree) 12.4 (17.5) 17.1 (21.4) 14.8 (18.0) 15.0 (18.1) No. of water molecules 364 175 351 402 r.m.s.d. bond length (Å) 0.012 0.013 0.014 0.009 Average B-values (Å2) Overall 18.8 28.2 18.1 14.8 Protein 16.9 27.6 16.5 13.1 Ligands 12.9 41.3 21.4 11.5 Water 32.7 35.3 29.8 25.9 Ramachandran plot (%): in most-favourable 97.0 97.0 96.0 98.0 in disallowed 0.59 0.0 0.0 0.0 PDB Accession Code 5D84 5D85 5D86 5D87 aData collection statistics in brackets represents highest resolution shell

2.4.3 SbnB structure determination

Four different conditions were used to obtain crystals of SbnB and of complexes with malonate, citrate, -KG and (N-(1-amino-1-carboxyl-2-ethyl)-glutamic acid) ACEGA. SeMet- labelled SbnB crystals were obtained by hanging drop vapour diffusion in drops containing 1 L protein solution and 1 L of reservoir solution (0.1 M Bis-tris pH 6.5, 0.2 M LiSO4 and 25%

PEG 3350). Malonate and citrate-bound SbnB structures were determined from crystals grown from 1 L of protein solution with 1 L of either 2.4 M sodium malonate pH 7.0, or 1.8 M ammonium citrate tribasic pH 7.0, respectively. Crystals were soaked for ~5 minutes in mother liquor supplemented with 30% glycerol and flash frozen in liquid nitrogen. For ACEGA soaking experiments, SbnB crystals were produced by the sitting drop vapor diffusion method in 0.2 M

MgCl2, 0.1 M HEPES pH 7.5 and 29 – 33% PEG 400. Substrates (~5 mM final concentration of

NAD+, NADH or ACEGA) were added directly to crystals and incubated for ~30 minutes and flash frozen directly in liquid nitrogen.

Single wavelength anomalous diffraction data for SeMet SbnB crystals were collected on beamline 7-1 at the SSRL. Data were integrated and scaled using HKL2000. Preliminary phasing and model building was performed using the programs, Solve (158) and Resolve (159,160) at 2.5

Å resolution. Eight of 10 possible selenium sites were identified with a figure of merit of 0.66 following density modification. Phases were extended to 2.1 Å resolution using DM (161) and

290 of 332 residues were built and placed using ArpWarp (162). The structure was manually edited using Coot (148) and refined with Refmac (149) from the CCP4 program suite (147) using

TLS refinement with 6 TLS groups. The final model comprises all residues excluding Met1 and a loop 51-55, which could not be resolved in the electron density map. Data for the SbnB-NAD+- malonate and SbnB-NAD+-citrate complexes were collected using SSRL beamline 7-1. 48

Structures were phased using MolRep and manually edited using Coot (148) and refined with

Refmac5 (149). Data from substrate soaked SbnB crystals were collected using a Saturn 944+

CCD detector on a Rigaku MicroMax-007 HF X-ray generator. Data were integrated and scaled using HKL3000. The datasets were solved using Phaser-MR (151) from Phenix (143) and the structures were refined with phenix.refine using TLS refinement. Data collection and refinement statistics are provided in Table 2-6. The electrostatic potential of SbnB was calculated by the

Adaptive Poisson-Boltzmann Solver software (163).

Table 2-6 X-ray diffraction data collection and refinement statistics for SbnB. SeMet-SbnB SeMet-SbnB SbnB-NAD+- SbnB-NAD+- SbnB-NADH-ACEGA SbnB-NAD+- (Phasing) (Refinement) Malonate Citrate -ketoglutarate Data Collectiona Wavelength (Å) 0.97649 0.97649 1.00000 1.00000 1.54178 1.54178 Resolution Range (Å) 50.00 – 2.50 58.41 – 2.10 58.04 – 1.85 58.20 – 2.10 29.73 – 2.36 27.25 – 2.10 (2.59 – 2.50) (2.22-2.10) (1.95-1.85) (2.21-2.10) (2.40-2.36) (2.14-2.10) Unit cell dimension (Å) a = 62.93 a = 62.89 a = 62.40 a = 62.66 a = 63.12 a = 61.93 b = 62.93 b = 62.89 b = 62.40 b = 62.66 b = 63.12 b = 61.93 c = 157.90 c = 157.63 c = 157.69 c = 157.08 c = 159.41 c = 151.72 Unique reflections 10543 19092 27695 19131 13910 15473 Completeness (%) 95.3 (94.1) 99.7 (98.1) 99.4 (96.6) 99.9 (100.0) 99.6 (99.1) 92.8 (86.0) Redundancy 7.3 (7.6) 14.6 (13.9) 9.5 (7.5) 6.4 (6.7) 6.9 (5.3) 6.9 (6.9) Average I/I 14.5 (7.2) 8.5 (1.9) 13.6 (2.3) 6.9 (2.0) 10.8 (2.7) 11.8 (5.4)

Rmerge 0.099 (0.280) 0.073 (0.384) 0.043 (0.343) 0.085 (0.390) 0.126 (0.590) 0.108 (0.464) Wilson B-factor (Å2) 27.0 25.8 22.8 27.1 32.7 27.1 Refinement

Rwork (Rfree) -- 17.4 (21.4) 18.4 (22.4) 18.7 (23.9) 22.3 (25.8) 22.9 (26.6) No. of water molecules -- 111 141 129 94 75 r.m.s.d. bond length (Å) -- 0.009 0.011 0.010 0.003 0.003 Average B-values (Å2) Overall -- 36.3 30.4 34.7 37.6 34.1 Protein 36.2 30.2 34.7 37.7 34.0 Cofactor -- -- 25.7 30.4 37.8 37.5 Water -- 37.7 35.1 35.5 34.1 34.2 Ramachandran plot (%): in most-favourable -- 98.0 97.0 97.0 97.0 98.0 in disallowed -- 0.0 0.3 0.3 0.0 0.0 PDB Accession Code -- 4MP3 4MP8 4MP6 4M54 4MPD aData collection statistics in brackets represents highest resolution shell

2.4.4 IruO structure determination

Reduced and oxidized IruO preparations were crystallized in two separate conditions.

TCEP containing apo-IruO and SeMet labeled IruO crystals were optimized by sitting drop vapour diffusion at 4 °C. Each drop contained 2 L protein solution and 2 L 0.2 M magnesium acetate and 8-11% PEG 3350. Crystals were further optimized by seeding drops with 0.4 L crushed crystals diluted to 10-4-10-5. Potassium ferricyanide oxidized IruO crystals were also optimized by sitting drop vapour diffusion at 4 °C. Each protein drop contained 2 L protein solution and 2 L 2.0-3.0 M sodium formate pH 7.0. Crystals were further optimized by streak seeding with crushed crystals. Crystals were soaked for <30 seconds in mother liquor supplemented with 30 % glycerol and flash frozen in liquid nitrogen. Reduced Se-Met-labeled

IruO were crystallized identically to native reduced IruO.

Single wavelength anomalous diffraction data for reduced SeMet IruO crystals were collected at a wavelength of 0.979383 Å on beamline 7-1 at the SSRL. Data was integrated and scaled using XDS (141). SeMet IruO crystals grew in the space group C2 with one molecule in the asymmetric unit. Phase determination and model building were performed using AutoSol and

Autobuild programs from Phenix (143). The initial figure of merit from AutoSol was 0.46 and

AutoBuild built 299 of 344 residues. The structure was manually edited using Coot (148) and refinement was performed with Phenix.refine from Phenix (143). Native reduced IruO was collected at a wavelength of 0.976240 Å on beamline 08B1 at the CLS. Data was integrated and scaled using XDS (141). The structure was solved by molecular replacement using the SeMet- labeled coordinates as the search model in Phenix.phaser from Phenix (143). Native oxidized

IruO was collected at a wavelength of 0.975302 Å on beamline 7-1 at the SSRL. The data was integrated and scaled using HKL2000 (152). Native oxidized crystals grew in the space group 51

P41212. The structure was solved by molecular replacement similarly to native reduced IruO with the exception that the search model was separated into three individual ensembles (residues 1-

120, 126-240 and 249-344). The structures were manually edited using Coot (148) and refinement was performed with Phenix.refine from Phenix (143) using TLS refinement with 3

TLS groups. Data collection and refinement statistics are provided in Table 2-7.

Table 2-7 X-ray diffraction data collection and refinement statistics for IruO SeMet-rdIruO rdIruO oxIruO Peak Data Collectiona Wavelength (Å) 0.97938 0.97624 0.97530 Resolution Range (Å) 33.43 – 1.90 41.72 – 1.82 38.91 – 2.30 (2.00 – 1.90) (1.89 – 1.82) (2.34 – 2.30)

Space group C2 C2 P41212 Unit cell dimension (Å) a = 157.60 a = 157.17 a = 90.6 b = 43.38 b = 43.60 b = 90.6 c = 51.10 c = 51.24 c = 138.5 Unique reflections 27210 31257 26012 Completeness (%) 99.3 (96.5) 100 (99.8) 99.9 (98.7) Redundancy 7.3 (6.9) 3.6 (3.5) 14.3 (14.1) Mean I/I 14.9 (3.2) 2.6 (8.2) 28.0 (5.2)

Rmerge 0.118 (0.680) 0.083 (0.401) 0.064 (0.612) Wilson B-factor (Å2) 17.5 18.4 47.4 Refinement

Rwork (Rfree) -- 18.9 (22.7) 18.3 (20.7) No. of water molecules -- 263 53 r.m.s.d. bond length (Å) -- 0.014 0.014 Average B-values (Å2) Overall -- 35.2 60.8 Protein -- 35.3 61.4 Ligands -- 24.6 46.1 Water -- 37.4 50.2 Ramachandran plot (%): in most-favourable -- 98.0 97.0 in disallowed -- 0.58 0.29 PDB Accession Code -- aData collection statistics in brackets represents highest resolution shell

2.5 Determination of oligomeric state in solution

2.5.1 SbnG oligomerization state determination

SbnG and SbnG E151Q oligomerization were determined by size exclusion chromatography multi-angle light scattering (SEC-MALS). Proteins were individually concentrated to 5 mg/mL in 20 mM Tris (pH 8.0) and injected (100 µL) into a HPLC 1260

Infinity LC (Agilent Technologies) attached to a Superdex 200 10/300 column (GE Healthcare) with a flow rate of 0.2 mL/min and column temperature set to room temperature. Data was collected with the miniDAWN TREOS multi-angle static light scattering device and Optilab T- rEX refractive index detector (Wyatt Technologies). Analysis was performed using the ASTRA6 program (Wyatt Technologies). The molar mass of the protein of interest is determined from

Rayleigh scattering, which is the scattering of light by molecules in solution. The Rayleigh ratio is used to quantify the scattered light intensity at a specific angle as detected from MALS detectors. A model combining the data from at least two MALS detectors is then used to approximate the intensity of light scattered at zero angle, which is directly related to the molar mass of the molecule of interest and is described by Eq. 1 where I is the light intensity, M is the molar mass, θ is the observation angle, c is the concentration of the solute and (dn/dc) is the specific refractive index increment (164).

푑푛 2 퐼(휃) ∝ 푀푐 ( ) Eq. (1) 푠푐푎푡푡푒푟푒푑 푑푐

2.5.2 SbnA and SbnB oligomerization state determination

SbnA and SbnB were dialyzed into 50 mM Tris pH 8.0, 200 mM NaCl. SbnA and SbnB solutions were analyzed by dynamic light scattering using a Wyatt DynaPro plate reader (Wyatt

Technologies). Data represent the average of five readings from three samples. SbnA and SbnB were also analyzed by SEC-MALS. SbnA and SbnB were concentrated to 5 mg/mL in 50 mM

Tris pH 8.0, 100 mM KCl and 2 mM TCEP and 100 L was injected into a HPLC 1260 Infinity

LC (Agilent Technologies) attached to a Superdex 200 10/300 column (GE Healthcare) with a flow rate of 0.2 mL/min. Data were collected using a miniDAWN TREOS multi-angle static light scattering device and an Optilab T-rEX refractive index detector (Wyatt Technologies).

Analysis was performed using the ASTRA6 program (Wyatt Technologies).

2.5.3 IruO oligomerization state determination

RdIruO and oxIruO were analyzed by SEC-MALS; they were concentrated to 2 mg/mL in 50 mM Tris pH 8.0, 300 mM KCl and 100 L was injected into a HPLC 1260 Infinity LC

(Agilent Technologies) attached to a Superdex 200 5/150 column (GE Healthcare) with a flow rate of 0.2 mL/min. Data were collected using a miniDAWN TREOS multi-angle static light scattering device and an Optilab T-rEX refractive index detector (Wyatt Technologies). Analysis was performed using the ASTRA6 program (Wyatt Technologies).

2.6 Bioinformatic analysis

2.6.1 SbnG structure superposition and phylogenetic analysis

SbnG homologs were identified by sequence alignment analysis using BLASTp

(165,166) and by tertiary structure superposition using DaliLite (167). A subset of homologs with characterized functions was selected for sequence alignment and phylogenetic analysis.

Homologs of known structure were overlaid with SbnG using SSM Superposition (168) in the program Coot (148). Sequence alignments were produced on the EMBL-EBI website using 55

programs ClustalW2 (169) and Clustal Omega (170). The phylogenetic tree was constructed using PHYML (171) in Seaview (172) using the LG substitution model and 100 bootstrap replicates.

The active sites of SbnG and SbnG E151Q were overlaid with the crystal structures of citrate synthase type I and type II using Superpose in the CCP4 suite (168). A total of three active site residues (15 atoms) from SbnG E151Q (His47, His96 and Asp177) were selected for alignment with the citrate synthase types I and II.

2.6.2 Multiple sequence alignment of SbnA and homologs

A subset of SbnA homologs predicted to be involved L-Dap biosynthesis were chosen from sequence alignments previously performed by Beasley et al. 2011 (80). Another subset of biochemically validated and structurally characterized OASS and OPSS enzymes were identified by a primary sequence search in the Protein Data Bank using the BLAST method with an E- value cutoff of 10.0 and sequence identities over 30 %. Additionally, the OPSS enzyme CysO from Aeropyrum pernix was included even though its sequence identity was 23 %. Also included was the closest S. aureus homolog, the OASS CysK and a third recently biochemically validated

OPSS enzyme, CysK2, from M. tuberculosis (173). Multiple sequence alignments were generated on the EMBL-EBI website using Clustal Omega set to default parameters (170).

2.6.3 Multiple sequence alignment of IruO and homologs

IruO homologs were identified by primary sequence alignments using BLASTp (165).

The multiple sequence alignments were generated by Clustal Omega set to default parameters available on the EMBL-EBI website (170). 56

2.7 UV-Vis spectroscopy analysis of SbnA

Electronic absorbance spectra of SbnA were collected on a Varian Cary 50 UV-Vis spectrophotometer. The spectra of SbnA-PLP were recorded at a concentration of 20 M in 50 mM Tris pH 8.0, 100 mM KCl, 2 mM TCEP at room temperature. Spectra of the SbnA aminoacrylate aldimine complex were recorded immediately after the addition of 30 M OPS.

To demonstrate the return of the resting state, 3 mM L-glutamate was added to the SbnA-

PLP/OPS mixture, incubated for 10 min. and then recorded. To test if SbnA bound O-acetyl-L- serine (OAS) or L-serine, electronic spectra were collected after 5 min. incubation.

Formation of the PLP--aminoacrylate was monitored by UV-visible spectroscopy was also measured for all SbnA variants. Spectra were measured with 20 M wild type SbnA or

SbnA variants in 50 mM Tris pH 8.0 and 2 mM TCEP with a Varian Cary 50 UV-Vis spectrophotometer at room temperature. Spectra were recorded after the addition of 50 M OPS or 1 mM L-cysteine (Sigma) and incubated for 5 minutes.

2.8 UV-Vis spectroscopy analysis of IruO

Room temperature UV-Vis absorption spectra (300-800 nm) were measured with a

Nanodrop 2000c spectrophotometer (Thermo Scientific) in an anaerobic chamber (Belle

Technology). All reagents were transferred into the chamber and allowed to equilibrate to remove all traces of oxygen. A total of 30 M of rdIruO in 50 mM sodium phosphate buffer, pH

7.4 was used for all assays. Upon the addition of 20 or 150 M NADPH (Calbiochem) or 100

M SDT (Sigma), spectra were recorded every 30 s for 5 minutes. SDT was prepared fresh in the

anaerobic chamber and used immediately. To demonstrate specific electron transfer from the

IruO FADsq to siderophores, 10 M of NADPH was added to IruO and incubated for 5 minutes.

Next, 60 M of Fe(III)-desferrioxamine B (Fe(III)-DFB) (Sigma) was added to the mixture and the spectra were recorded as above. Negative controls were NWMN_0732 in place of IruO and

GaDFB in place of Fe(III)-DFB.

To determine the molar ratio of NADPH required to form the FAD semiquinone, 30 M of rdIruO was titrated with 5 M NADPH aliquots from a 5 mM working stock. After each

NADPH addition, the mixture was allowed to incubate for 5 minutes and then recorded at 600 nm, which corresponded to the formation of the FAD semiquinone. The NADPH titration experiment was repeated in triplicate.

2.9 Fluorescence spectroscopy analysis of IruO

Quenching of internal tryptophan fluorescence of IruO was measured using a Cary

Eclipse fluorescence spectrophotometer at 25 °C. 1 μM rdIruO or 0.5 μM oxIruO, in 50 mM sodium phosphate, pH 7.4, were titrated with 0.5 - 20 μM of Fe(III)-DFB or Fe(III)-ferrichrome

A (Fe(III)-FCA) (Sigma). The excitation and emission slits were 10 nm and the detector voltage was 900 V. Samples were excited at 292 nm and emission spectra (310-380 nm) were measured.

Emission at 350 nm was recorded for analysis. Each replicate is the average intensity of three sequential scans and this was performed in triplicate. Dissociation constants (Kd) were determined by fitting data to a one-phase exponential decay model.

2.10 Citrate synthase assay

Citrate synthase activity of wild type SbnG and variants was measured using the reagent

DTNB (5,’5-dithio-bis(2-nitrobenzoic acid)) to quantify the generation of CoA. The concentration of resulting TNB dianion was monitored at 412 nm on a Varian Cary 50 UV-Vis spectrophotometer using an extinction coefficient of 14150 M-1cm-1. Each reaction contained 50 mM HEPES (pH 8.0), 300 mM NaCl, 0.5 mM oxaloacetate, 0.2 mM acetyl-CoA and 10 M protein. Each reaction was incubated for 1 hour at 37 °C followed by the addition of 0.1 mM

DTNB reagent. Individual blanks for each reaction contained all required components minus protein. Oxaloacetate-independent activity was recorded by performing the reactions in the absence of oxaloacetate. All reactions were performed in triplicate.

2.11 Phosphate release assay

Inorganic phosphate release was measured by a continuous spectroscopic assay as previously described (174). The conditions used were 50 mM Tris pH 8.0, 1 mM MgCl2, 5 mM dithiothreitol (DTT), 100 mM KCl, 0.5 U purine nucleoside phosphorylase, 200 M 2-amino-6- mercapto-7-methylpurine riboside (MESG) and 500 nM SbnA. All reactions were performed at room temperature. The pre-substrate mixture was incubated for 5 minutes to remove contaminating inorganic phosphate and blanked. All data was collected on a Varian Cary 50 UV-

Vis spectrophotometer. Continuous measurement recorded at 360 nm was started after the addition of 1 mM OPS. At two minutes, various compounds were added at a concentration of 2 mM and the assay was run for a total of 20 min. A total of three replicates were collected.

2.12 NADH assay

NADH generation was monitored continuously at 340 nm for 20 min. at room temperature on a Varian Cary 50 UV-Vis spectrophotometer. Coupled reactions contained 50 mM Tris pH 8.0, 100 mM KCl, 2 mM TCEP, 10 M SbnA, 10 M SbnB, 1 mM OPS, 1 mM L- glutamate and 100 M NAD+. After 2 min. incubation and blanking, NAD+ was added to each reaction mixture and the absorbance monitored for a further 18 min. An extinction coefficient of

6220 M-1cm-1 was used for NADH. For NADH consumption assays the reaction mixtures were run for 5 min. Reactions contained 50 mM Tris pH 8.0, 100 mM KCl, 2 mM TCEP, 1 M SbnB,

1 mM L-Dap, 1 mM -KG and 200 M NADH. All reactions were done in triplicate.

The NADH assay was also used as a coupled assay with SbnA to measure the activity of

SbnA variants. NADH generation was monitored by absorbance at 340 nm for 10 minutes at room temperature on a Varian Cary 50 UV-Vis spectrophotometer. Coupled reactions contained

50 mM Tris pH 8.0, 100 mM KCl, 2 mM TCEP, 10 M SbnA, 10 M SbnB, 1 mM OPS, 1 mM

L-glutamate (Sigma) and 100 M NAD+. All reactions were performed in triplicate. The extinction coefficient 6220 M-1cm-1 was used to quantify the amount of NADH generated.

2.13 -KG assay

The assays performed were a modification on a previously described 2,4- dinitrophenylhydrazine assay protocol (175). To summarize, 500 L reactions were performed with 10 M SbnB, 1 mM -KG, 1 mM L-Dap, 1.5 mM NADH in 50 mM Tris pH 8.0, 2 mM

TCEP at room temperature. At one minute intervals, 25 L aliquots were removed and quenched with 450 L of 0.1 M HCl containing 0.5 M 2,4-Dinitrophenylhydrazine (DNPH) and 25 L

dH2O. Mixtures were incubated for 10 minutes. To each quenched reaction, 500 L of 2 M

NaOH was added and incubated for a further 5 minutes. All reaction mixtures were measured at

450 nm on a Varian Cary 50 UV-Vis spectrophotometer. An extinction coefficient of 12,500 M-

1cm-1 was used for the 2,4-dinitrophenylhydrazone derivative covalently linked to -KG. All reactions were blanked with a DNPH solution (0.1 M HCl) containing only dH2O. The control blank was also quenched with an equivalent amount of 2 M NaOH prior to use.

2.14 Ferrozine assay

A ferrozine-based assay was used to measure the generation of Fe(II) as previously described (176). 500 L reaction mixtures contained 1 M rdIruO, 1 mM ferrozine, 100 M

NADPH and 100 M Fe(III)-siderophore in 50 mM sodium phosphate buffer (pH 7.4) at room temperature. Spectra of the reaction mixtures from 400 to 800 nm were monitored every 30 seconds for 10 minutes on a Varian Cary 50 UV-Vis spectrophotometer.

2.15 Heme degradation assay

IsdI was reconstituted with heme as previously described (34). Heme degradation was initiated with 1 M oxIruO or rdIruO and 200 M NADPH in the presence of 500 M catalase and 4 U/mL superoxide dismutase. Spectra of the reaction mixtures from 300 to 700 nm were monitored every minute for twenty minutes or the absorbance at the Soret peak (412 nm) was measured every 0.1 second for 60 seconds on a Varian Cary 50 UV-Vis spectrophotometer. To calculate reaction rates the absorbance at the Soret peak was converted to the concentration of

IsdI-heme based on the published extinction coefficient (126 mM-1 cm-1) (119). Linear

regressions were calculated from data taken between 10 and 60 seconds after initiation of the reaction and the slope of the lines were taken as the reaction rates. Reactions were done in triplicate and data presented in the text are the means and standard errors of the mean. Linear regression R2-values were all above 0.97 and all slopes significantly deviated from a value of zero (p < 0.0001).

2.16 Measurement of IruO sulfhydryls

Exposed sulfhydryls on IruO either prepared with TCEP, or potassium ferricyanide were quantified using a modified DTNB method as previously described (177). A total volume of

500 L was used that contained 50 mM potassium phosphate buffer (pH 7.4) and 10 M of protein. Next, 10 L of 10 mM DTNB was added to the reaction mixture, allowed to incubate for

5 minutes at room temperature and the absorbance (A412s) was recorded. Two reagent blanks were also recorded, one with buffer and DTNB (A412r) and the other with buffer and protein

(A412p). The amount of free sulfhydryls was calculated using Eq. (2) with the extinction

-1 -1 coefficient of 412 = 14,150 M cm for the TNB dianion.

퐴 −퐴 −퐴 푚표푙 푆퐻 = 푡표푡푎푙 푣표푙푢푚푒 퐿 × 412푠 412푟 412푝 Eq. (2) 휀412×푝푎푡ℎ푙푒푛𝑔푡ℎ 푐푚

The molar ratios of SH groups per rdIruO or oxIruO were then calculated based on the total amount of protein used.

2.17 Agar plate bioassays

2.17.1 Determining citrate production from SbnG and variants

Concentrated spent supernatants of the citrate-synthase deficient mutant (ΔcitZ sbnG) complemented in trans by plasmids bearing the sbnG variants, as described above, were prepared from triplicate 10 mL cultures of these strains grown in RPMI-1640, a media conducive only to the production of SB (82). Following a 36 hour incubation with shaking at 37 °C, growth was assessed, and the culture densities normalized to approximately 2, by OD600. Bacterial cells were removed by centrifugation, and the resulting supernatants were lyophilized overnight.

Lyophils were resuspended in 1 mL of sterile ddH2O, passed through a 0.2 micron syringe filter, and assessed for their ability to promote growth of wild type S. aureus RN6390, as previously described (67,82). In brief, 10 μL of the reconstituted supernatants were applied to sterile paper disks, which were then placed on Tris Minimal Succinate (TMS) agar plates (48) seeded with ~1 x 104 CFU mL-1 of RN690 as a reporter strain, and containing 10 μM EDDHA. Plates were incubated at 37 °C and growth radius about the disks was measured after 48 h.

To assess the role of residues implicated in SbnG enzymatic function in the citrate- dependent production of SB within the cell, sbnG complementation vectors bearing the mutated residues (pE46Q, pH47A, pR72A, pH96A, pE151Q, and pD177A) were assessed for their ability to promote SB production in a previously constructed citrate-synthase deficient strain (H2708; citZ sbnG) of S. aureus Newman (82). Spent culture supernatants from the complemented citZ sbnG mutant strains grown in RPMI-1640 for 36 hours, and 10X concentrated, were assessed for the presence of SB using agar plate bioassays, as previously described (67,82). In brief, S. aureus wild type RN6390 was seeded into Tris Minimal Succinate (TMS) agar plates (48), and the above supernatants were applied to sterile paper disks placed onto these plates. SB-dependent 63

growth promotion was assessed by measuring the growth radius about the disks after 48 hour incubation at 37 °C.

2.17.2 Determining L-Dap production from SbnA and SbnB

One pot reactions were assembled that incorporated SbnA, SbnB, OPS, L-glutamate and

NAD+ in place of -KG and L-Dap. Enzyme reactions, 1 mL in volume, consisted of SbnA,

SbnB, SbnC, SbnE, SbnF and SbnH (10 M each), 5 mM ATP, 0.5 mM MgCl2, 0.5 mM PLP, 2 mM sodium citrate, 2 mM OPS, 2 mM L-glutamate and 2 mM NAD+ in 50 mM Hepes pH 8.0 containing 2 mM TCEP. A positive control for staphyloferrin B synthesis was included that contained both -KG and L-Dap. All reactions were incubated for 16 hours at room temperature in the dark. Following incubation, reaction mixtures were spun down with a Nanosep 3K Omega spin column (Pall) to remove all Sbn proteins. Filtrates were stored in the dark at 4 °C.

Siderophore detection was performed using the chrome azurol S (CAS) shuttle solution.

50l aliquots of SB reaction mixture were removed and diluted with 450 mL dH2O followed by the addition of 500 mL CAS shuttle solution. The mixture was incubated at room temperature in the dark for 1 hour and quantitated as previously described (67).

Disk diffusion bioassays were performed as previously described (67,81). A total of 5 L

SB reaction mixtures were combined with 5 L of 50 M FeCl3 and spotted onto sterile paper disks that were placed onto Tris-minimal-succinate agar seeded with 1  104 S. aureus per mL.

Bacteria used were S. aureus strain Newman and S. aureus RN6390 isogenic sirA mutant. Plates were incubated at 37 °C for 36 hours and growth was measured by the diameter of the growth halo surrounding each disk.

2.18 Mass spectrometry analysis

SbnA and SbnB reactions were performed in dH2O for 1 hour at room temperature.

Proteins were removed by centrifugation at 14,000  g for 20 min. The filtrates were then injected onto a LC-MS system comprised of an HP 1100 HPLC with a Waters Alantis hydrophilic interaction chromatography column (100  3.0, 3.0 m, 100 Å) coupled to an

Esquire LC-Ion Trap (Bruker). Separation was done at a flow rate of 550 L/min with a gradient that started at 35% B and increased to 50% B over 12 minutes. Solvent A was acetonitrile and

Solvent B was 20 mM ammonium formate. Analysis was performed with electrospray ionization in positive ion mode with a scanning range 50 to 700 m/z.

2.19 NMR analysis

SbnA reactions were performed in 50 mM potassium phosphate pH 8.0 with 10 mM OPS and 10 mM L-glutamate overnight at room temperature. SbnA was removed from the reaction by centrifugation through a Nanosep 3K Omega spin column (Pall) at 14,000  g for 20 min. NMR analysis of reaction product (~10 mM ACEGA in 50 mM phosphate / 10% D2O pH 8.0) was performed at 25 ºC on a Bruker Avance 600 MHz spectrometer with a cryoprobe. 1H spectrum was obtained using water-presaturation. 13C, 1H-13C HSQC, and 1H-13C HMBC spectra were measured using standard methods provided by TopSpin software (Bruker).

2.20 Steady-state kinetic analysis

2.20.1 SbnA and variants steady-state kinetic analysis

SbnA enzymatic activity was also measured by monitoring inorganic phosphate release from OPS using a coupled enzymatic assay, as previously described (174). All data were collected on a Varian Cary 60 UV-Vis spectrophotometer with a Varian Cary Single Cell Peltier temperature block set to 25 °C. The reaction mixture contained 100 nM wild type SbnA or

S185G in 50 mM Tris pH 8.0, 2 mM TCEP, 1 U purine nucleoside phosphorylase and 200 M 2- amino-6-mercapto-7-methylpurine riboside. The concentrations of OPS (0.005 – 10.0 mM) or L- glutamate (0.1 – 100 mM) were varied in the coupled enzymatic assay reaction mixtures. The absorbance increase at 360 nm was recorded every 0.1 s for 120 s, capturing the linear range of activity. The concentration of inorganic phosphate released from OPS was determined using the extinction coefficient of 11,000 M-1cm-1 (174). Data were fit by non-linear regression analysis using a Michaelis-Menten model in Graphpad Prism 6. The inhibition constant (Ki) for L- cysteine was also determined using the inorganic phosphate assay. 10 nM wild type SbnA or 100 nM S185G were used in 50 mM Tris pH 8.0 and 2 mM TCEP. OPS (0.05 – 10 mM) dependence was measured at various L-cysteine concentrations (0 mM – 10 mM). The Ki was determined with Graphpad Prism 6 using a competitive inhibition model as described by Eq.3 and Eq. 4.

[I] 퐾Mobs = 퐾M × (1 + ) (Eq. 3) 퐾i

푣 [S] 푉 = max (Eq. 4) 표 퐾 +[S] Mobs

2.20.2 IruO steady-state kinetic analysis

The kinetic parameters of IruO were also determined using the ferrozine assay. The ferrozine-Fe(II) complex was detected at 562 nm and an extinction co-efficient of 27900 M-1cm-1 was used to for quantification (176). 500 l reactions were prepared with 50 mM sodium phosphate buffer (pH 7.4), 1 - 60 M of NADPH, 1 mM ferrozine, 1 - 250  and Fe(III)-DFB or Fe(III)-FCA. 0.3 M oxIruO was used for both siderophores and 0.1 M and 0.02 M rdIruO was used for Fe(III)-DFB and Fe(III)-FCA, respectively. Data was collected in triplicate on a

Varian Cary 60 UV-Vis spectrophotometer, at a temperature of 25 °C. Data were fit to a non- linear regression and calculated with a least-squares fit to Michaelis-Menten kinetics.

2.21 Stopped-flow absorption spectroscopy for single turnover kinetic analysis

Data were collected from an SX.18MV stopped-flow reaction analyzer (Applied

Photophysics) using a 4.96 nm/mm bandpass monochromator with the slit width set at 0.5 mm.

The flow cell temperature was maintained at 25 ± 1 °C. A photodiode array detector was used to collect single wavelength measurements. 10 M wild type and variant SbnA in 50 mM Tris pH

8.0 and 2 mM TCEP were mixed with varying concentrations of OPS and L-cysteine. PLP- aminoacrylate formation was measured at 467 nm and PLP binding to L-cysteine was measured at 412 nm. Data was collected in triplicate with 1000 data points collected over 2 seconds. Both the formation of the aminoacrylate intermediate and decay of the internal Schiff base due to

L-cysteine binding were fit to single phase exponential functions described by Eq. 5 where At is the absorbance at time t, Ais the absorbance at time , Ai is the amplitude of the ith transient, and ki is the observed first-order rate constant (kobs) for the ith transient.

−푘푖푡 퐴푡 = 퐴∞ ± ∑𝑖 퐴𝑖푒 (Eq. 5)

All observed rates were plotted as a function of either OPS or L-cysteine concentration.

In the case of OPS, a hyperbolic function was fit to the data. Data for L-cysteine were fit by a linear function. Rate constants were calculated using a rapid equilibrium binding model (Eq. 6), as previously described for O-phospho-L-serine sulfhydrylase (OPSS) enzymes (178), where K1

= 1/Kd, [S] is the substrate concentration and k2 is the observed rate of the PLP-aminoacrylate formation. The reverse observed reaction rate (k-2) is assumed to be negligible under the experimental conditions.

푘2퐾1[S] 푘표푏푠 = + 푘−2 (Eq. 6) 퐾1[S]+1

2.22 Software for kinetic analysis

All kinetic calculations and statistical analyses were done with GraphPad Prism 6.00.

Chapter 3: SbnG, a citrate synthase in Staphylococcus aureus: a new fold on

an old enzyme family

3.1 Introduction

Growth of S. aureus under iron-restricted conditions results in a global shift in gene expression that is collectively known as the “iron-sparing response” (111). Iron restriction results in the up-regulation of iron uptake systems, including the production of the citrate-containing

SB. Conversely, many metabolic pathways with iron-containing enzymes are down-regulated in an effort to conserve the iron pool necessary for survival. A notable down-regulation is observed for TCA cycle enzymes including citrate synthase, CitZ (111,116). To alleviate the loss of citrate production, the sbn biosynthetic gene cluster encodes for a second citrate synthase, sbnG, which generates citrate from oxaloacetate and acetyl-CoA under iron-restricted growth conditions.

Therefore, SbnG provides an alternative route for citrate to enter the biosynthetic pathway for

SB, independent of the TCA cycle.

Homologs of SbnG have also been identified in the biosynthetic gene clusters that produce siderophores achromobactin and vibrioferrin, suggesting that a unique group of citrate synthases may have evolved for specifically for siderophore biosynthesis out of the requirement for citrate to be generated during iron restricted growth. SbnG represents a unique citrate synthase, originally annotated as a metal-dependent class II aldolase by sequence analysis; instead, SbnG was reclassified as a metal-independent citrate synthase (81).

In this study, I sought to elucidate the catalytic mechanism of SbnG. Through the use of

X-ray crystallography, the structures of SbnG and an active site variant bound to oxaloacetate were solved revealing a conserved active site architecture with TCA cycle citrate synthases,

supporting a model of convergent evolution for citrate synthesis. The SbnG active site was validated using site-directed mutagenesis to show that all six SbnG variants were defective for in vivo citrate production. Furthermore, phylogenetic analysis reveals that SbnG forms a separate clade with homologs from other siderophore biosynthetic gene clusters and is representative of a metal-independent subgroup in the phosphoenolpyruvate/pyruvate domain superfamily.

3.2 Results

3.2.1 Overall structure of SbnG

The crystal structure of SbnG was solved to a resolution of 1.85 Å with a single subunit in the asymmetric unit. The structure consists of residues 4 to 258, excluding residues 121 to 132 that lack clearly defined electron density and form part of an apparent disordered loop. The overall structure of the SbnG protomer revealed an ()8 barrel fold formed by seven -helices that pack against an interior β-barrel with an eighth -helix on the C-terminal end projecting outward (Figure 3-1A). Dimers related by a crystallographic 2-fold axis resulted in the domain swapping of the C-terminal α-helixes () of each protomer to complete the ()8 fold. The biological hexamer could be reconstructed through 3-fold crystallographic symmetry (Figure 3-

1B). Additionally, SbnG (~28 kDa subunit) exists primarily as a hexamer in solution with a calculated mass of 171 kDa as determined by multi-angle light scattering (Figure 3-2A). The dimer interface includes one π-stacking interaction between equivalent His244 residues from each protomer as well as extensive hydrophobic interactions between residues Ile24, Leu28,

Ile32, Leu234, Ile241, Leu245 and Leu249. The trimeric interface was comprised of a mixture of electrostatic and hydrophobic contacts, including a H-bond between residues His47 and Asp86

and several hydrophobic interactions involving residues Val76, Leu120, Leu123, Leu124,

Pro185 and Trp186.

Figure 3-1 Structure of SbnG. (A) SbnG protomer is shown as a cartoon with helices, loops and -strands colored brown, blue and yellow, respectively. The C-terminal -helix (8) protrudes outwards from the core ()8 barrel. (B) SbnG homo-hexamer is reconstructed through crystallographic symmetry. The

hexamer is constructed from a trimer of dimers mediated by the protruding α8 helix that completes the ()8 barrel fold. (C) SbnG active site residues are shown as sticks with carbon, oxygen and nitrogen in light blue, red and blue, respectively. A single Ca2+ (yellow sphere) is bound by Glu151 and Asp177. Water molecules are shown as small blue spheres and metal- ligand bonds were indicated by dashed lines. (D) A space-filling representation of the SbnG active site. The electrostatic potential is shown on the molecular surface of one protomer with a neighboring protomer shown in cartoon form after the hexamer was reconstructed. The Ca2+ is visible as a yellow sphere in an acidic patch within the proposed active site, which includes the directly coordinated water molecules shown as blue spheres.

3.2.2 SbnG active site

Based on homology to class II aldolases, the SbnG active site was hypothesized to be situated within a groove at the opening of the ()8 barrel near the 3-fold interface shared with an adjacent SbnG protomer (Figure 3-1C). The putative active site was highly polar and is composed in part of two histidine (His47, His96), two glutamate (Glu46, Glu151), aspartate

(Asp177) and arginine (Arg72) residues. In the center of this site, density for a metal ion was identified residing within an acidic patch on the protein surface (Figure 3-1D). Due to the high calcium acetate concentration (0.2 M) from which crystals were formed, the density was modeled as a Ca2+, a known inhibitor of SbnG (81). The Ca2+ is bound by Glu151-O1 (2.4 Å) and Asp177-O2 (2.2 Å) as well as four ordered water molecules (2.3 to 2.5 Å) that complete the octahedral coordination sphere (Figure 3-1C). The coordinated water molecules participate in a

H-bond network involving the side chains of residues Glu46, His47, Arg72 and His96 based on

H-bond distances under 3.0 Å. Due to the hexameric oligomerization of SbnG, an adjacent protomer sits atop the active site groove at the opening of the ()8 barrel. A single serine residue (Ser119) from a neighboring protomer adopts two different conformations and also participates in the Ca2+ hydration sphere H-bond network (Figure 3-1C). Ser119 was situated within a large loop that extends above of the active site, part of which was poorly defined in the

electron density map (Figure 3-1D). Beside the Ca2+-binding site and within the same protomer lies an extended channel lined with intermittent patches of positive (Arg72, Arg189, Arg216,

Arg238 and Arg243) and near neutral electrostatic potential (Figure 3-1D).

Figure 3-2 SbnG is a hexamer in solution as determined by SEC-MALS. Molar mass determination of wild type SbnG and SbnG E151Q. (A) Wild type SbnG with a calculated molar mass of 171 Da. (B) SbnG E151Q with a calculated molar mass of 169 Da. The Rayleigh ratio is shown as a black line and the calculated molar mass is shown as a blue line.

3.2.3 Crystal structure of SbnG E151Q variant bound to oxaloacetate

Previously, citrate synthase activity of SbnG was shown to be inhibited by low mM concentrations of Mg2+ and Ca2+, and not inhibited by the addition of EDTA (81). These observations were in stark contrast to all characterized class II aldolases, which require Mg2+ for activity. Attempts to co-crystallize SbnG in the presence of either acetyl-CoA or oxaloacetate using crystallization conditions lacking calcium acetate proved unsuccessful. Furthermore, soaking experiments of wild type SbnG crystals failed to produce substrate-containing structures leading to the hypothesis that calcium sterically inhibits substrate binding. An SbnG variant

(E151Q) was created to prevent divalent metal ion inhibition and thus favor substrate binding under buffer conditions suitable for crystallization.

SbnG E151Q crystals soaked in the presence of oxaloacetate and CoA diffracted to a resolution of 2.6 Å. Oxaloacetate was modeled in at ~80% occupancy into a patch of difference electron density discovered within the proposed active site (Figure 3-3A). Oxaloacetate was oriented such that the two carboxylate groups formed H-bonds to residues Glu46, His47, Arg72,

His96 and Gln151, while exposing the carbonyl group to the solvent. Ser119, a residue from an adjacent protomer, formed a H-bond to the carbonyl group of oxaloacetate. The E151Q variant also existed as a hexamer in solution with a mass of 169 kDa as determined by size-exclusion chromatography multi-angle static light scattering (SEC-MALS) (Figure 3-2B).

The wild type SbnG and the substrate bound E151Q variant overlay with a root-mean- square-deviation (r.m.s.d.) of 0.86 Å over all atoms. The largest main-chain differences were displacement of two -helices that contained Glu/Gln151 and Asp177. This displacement likely resulted from the loss of Ca2+ coordination and binding of oxaloacetate. Within the active site,

Ca2+ present in the wild type SbnG structure occupied an equivalent position to the oxaloacetate 74

observed in SbnG E151Q (Figure 3-3B). The amino acid side chain conformations in the active site were similar with the exception of Asp177, in which the side chain is rotated ~45º about 2

(Figure 3-3B).

Figure 3-3 Oxaloacetate bound to SbnG E151Q. (A) Stereoview of the active site of SbnG E151Q variant bound to oxaloacetate. The active site SbnG E151Q is represented as sticks with the active site protomer colored grey and the adjacent protomer colored black. Oxaloacetate (OAA) is colored to match its corresponding protomer active site. Omit difference electron density is shown as light grey mesh contoured at 2.0 . (B) An overlay of wild type SbnG (blue) and E151Q variant (grey) structures. The Ca2+ from wild type SbnG is shown as a yellow sphere. OAA from SbnG E151Q is shown as sticks and is colored to match its respective active site residues. 75

3.2.4 Structural comparison of SbnG to the phosphoenolpyruvate/pyruvate domain superfamily

A search for similar structures of SbnG in the Protein Data Bank using the DaliLite server (167) identified several members of the metal-dependent class II aldolase (top seven unique entries with <2.4 Å r.m.s.d. over >240 C atoms). Examples are the well-characterized 2- dehydro-3-deoxy-galactarate aldolase (DDGA) (179) and macrophomate synthase (MPS) (180).

These class II aldolases also form hexamers that are assembled by domain swapped dimers.

More distantly related, but significant matches in the DaliLite search were many structures of the phosphoenolpyruvate/pyruvate domain superfamily as defined in the SCOP database (181). Of these, the most similar were pyruvate kinase and the -subunit of citrate lyase. A list of selected structural superpositions with supporting statistics is presented in Table 3-1.

Table 3-1 Superposition statistics of wild type SbnG to structurally characterized homologs determined from DaliLite. Protein Organism PDB entry Sequence R.M.S.D. value Aligned residues code identity (%) (Å) /259 DDGA E. coli 1DXE 28.9 1.77 232 YfaU E. coli 2VWS 28.2 1.77 227 B8FRX2 D. hafniense 3QZ6 27.1 1.44 240 MPS M. commelinae 1IZC 27.1 1.72 236 HcpH E. coli 2V5J 25.0 1.64 236 CitE M. tuberculosis 1Z6K 20.5 2.07 171 CitE2 Y. pestis 3QLL 17.0 2.21 159 PtsA E. coli 2HWG 15.6 2.51 179 PtsI S. aureus 2WQD 15.4 2.44 175 PykF E. coli 1PKY 15.0 2.14 173 CitE3 C. necator 3QQW 14.4 2.34 181 GlcB E. coli 1D8C 11.8 2.43 186 GlcB M. tuberculosis 1N8W 10.2 2.46 186

To compare active site architectures among SbnG structural homologs, a total of four superpositions were assembled with DDGA (179), MPS (180), citrate lyase -subunit (CitE)

(182) and malate synthase G (GlcB) (183). The four superpositions with SbnG revealed common active site features. Notably, all these structurally characterized homologs contained an Mg2+ in an equivalent position to the Ca2+ present in wild type SbnG (Figures 3-4A and 3-4B). To the best of our knowledge, all other members of the superfamily employ either Mg2+or Mn2+ as a cofactor. Also, residues equivalent to Glu151 and Asp177 directly coordinate to a divalent metal ion and these two residues are conserved in the phosphoenolpyruvate/pyruvate domain superfamily. Glu46 and Arg72 of SbnG are also conserved in the phosphoenolpyruvate/pyruvate domain superfamily and have been suggested to play a role in ordering water molecules in the active site or in substrate binding (179,184-186). Interestingly, amino acid residues equivalent to

SbnG His47 and His96 show variation across the homologs.

Figure 3-4 Conservation of active site structure in SbnG homologs. Superposition of active site residues from SbnG (blue) with those of (A) E. coli DDGA (PDB ID: 1DXE) (light grey), M. commelinae MPS (PDB ID: 1IZC) (green) and (B) M. tuberculosis CitE (PDB ID: 1Z6K) (dark grey) and M. tuberculosis GlcB (PDB ID: 1N8W) (tan). Active site residues are shown as sticks with oxygen and nitrogen colored as red and blue, respectively. Mg2+ and Ca2+ are depicted as spheres and colored to match their respective active site residues. (C) Unrooted phylogenetic tree of SbnG and homologs. Bootstrap values are presented as percentages at key branch points. Each protein is labeled by the genus, species and protein name. The dashed box represents a novel clade with the phosphoenolpyruvate/pyruvate domain superfamily for SbnG-like citrate synthases identified in other siderophore biosynthetic gene clusters.

3.2.5 SbnG represents a new family of metal independent enzymes within the class II aldolase superfamily

To further classify SbnG within the class II aldolase family, a multiple sequence alignment was constructed and used to derive a phylogenetic tree (Figure 3-4C). Included in the analysis were functionally uncharacterized homologs identified in achromobactin and vibrioferrin biosynthesis gene clusters (81), in addition to the structurally established homologs described above. Inspection of the tree revealed that SbnG and homologs from other siderophore biosynthetic pathways group together to form a separate clade (91% bootstrap value) from other class II aldolases. Homologs outside the SbnG clade are all known to require divalent metal for activity.

3.2.6 SbnG active site variants show reduced SB production in a citrate synthase- deficient S. aureus strain

Based on the interaction of Glu46, His47, Arg72, His96, and Asp177 with oxaloacetate in the E151Q crystal structure, variants of these putative active site residues were generated, in addition to E151Q, to evaluate their role in citrate synthase activity. SbnG variants E46Q, H47A,

R72A, H96A, and D177A were all recombinantly expressed in E. coli BL21 and successfully purified. Production of CoA from acetyl-CoA by wild type and variant forms of SbnG was monitored using DTNB in the presence and absence of oxaloacetate. SbnG activity is oxaloacetate dependent; however, the activity by this assay was modest (~5 nmole CoA released per mg of SbnG) as observed previously (81). The activity of the variants did not differ substantially from wild type enzyme, likely due the low wild type activity. I hypothesized that physiologically relevant activity of SbnG requires unidentified factors present in S. aureus but 79

not present in the biochemical assay. Thus, an in vivo assay was developed to assess activity of

SbnG and variants.

To circumvent the above issue of poor enzyme performance in vitro, SbnG activity was assayed by measuring SB production from a strain of S. aureus in which both citZ and sbnG are inactivated (82). Expression of wild type SbnG from a plasmid in this deletion strain restores SB production as quantified using a disk diffusion assay (Figure 3-5). SB production from S. aureus strains expressing each of the six variants was diminished in comparison to wild type SbnG.

Notably, the assay indicates that the D177A variant is unable to produce sufficient citrate for SB production to be detected.

Figure 3-5 SbnG variants are impaired for citrate-dependent production of SB. Agar plate bioassays were performed using supernatant extracts prepared from a citrate synthase- deficient strain of S. aureus Newman (ΔcitZ ΔsbnG) complemented with vectors, as indicated, expressing wild type SbnG, SbnG variants E46Q, H47A, R72A, H96A, E151Q, and D177A, or a blank pALC2073 vector control (vehicle). Strains were grown for 36 h in RPMI-1640, a media conducive to the production of SB (and not SA), and supernatants were assessed for SB through their ability to promote growth of S. aureus RN6390 seeded in iron-restricted TMS agar plates. Growth radius about paper disks, to which the supernatants were applied, was measured after 48 h and reflects the average of three biological replicates for each strain. 80

3.3 Discussion

SbnG is a citrate synthase that functions to provide citrate for SB biosynthesis (81).

Previously, SbnG was hypothesized to be an aldolase based on sequence alignments to characterized homologs DDGA and YfaU. These alignments revealed a conserved pair of amino acids (Glu151 and Asp177 in SbnG) that in the homologs coordinate the active site Mg2+ required for catalysis (179,185). Thus, SbnG was suggested to have metal-dependent aldolase activity and to possibly participate in SB degradation for iron release (81). Furthermore, a homolog found in the biosynthetic operon for the siderophore achromobactin, AcsB, was speculated to play a role in converting an intermediate into pyruvate and an aldehyde, which would eventually feed into siderophore biosynthesis (187). Yet, recent work by Cheung, et al

2012 has revealed SbnG to exhibit citrate synthase activity in the absence of divalent metals like

Ca2+ and Mg2+.

The crystal structure of SbnG contained electron density for a Ca2+ directly coordinated to Glu151 and Asp177 in the active site. Ca2+ was modeled instead of Mg2+ because the observed bond lengths were not consistent with Mg2+ coordination, which are typically ~2.1 Å (188).

Furthermore, 0.2 M calcium acetate was used in the crystallization condition, a sufficiently high enough concentration for full occupancy of bound Ca2+. In the case of DDGA, the closest structurally characterized homolog to SbnG, Mg2+, Co2+ or Mn2+ can be used as cofactors for catalysis (179,189). To the best of our knowledge, no example exists in the literature where Ca2+ serves as a functional cofactor in class II aldolases. Thus, the SbnG-Ca2+ complex is likely an inhibited form of the enzyme and the binding of Ca2+ may stabilize crystal packing. Ca2+ levels in bacteria are tightly regulated and maintained in a range between 100 to 300 nM (190).

Therefore, due to the high concentrations of Ca2+ required to inhibit SbnG activity and the low 81

concentration of intracellular Ca2+ in bacteria (81,190), Ca2+ is predicted to not act as a physiologically relevant inhibitor.

Binding sites for oxaloacetate and acetyl-CoA to SbnG were inferred by analysis of the

E151Q variant and by comparison to the homolog malate synthase G, respectively. Malate synthase G catalyzes a similar Claisen condensation reaction as citrate synthases using substrates glyoxylate and acetyl-CoA to produce malate (191). Crystals of the SbnG E151Q variant produced in the absence of Ca2+ were soaked in a substrate solution revealing density for oxaloacetate. The possibility that the binding mode of oxaloacetate is altered by the amino acid substitution of Glu151 cannot be excluded. Nonetheless, the overlay of the oxaloacetate and Ca2+ binding sites suggests that Ca2+ may be a competitive inhibitor with this substrate. Density for

CoA was not identified in the E151Q variant electron density map. Inspection of the structure does reveal a large positively charged channel that lies next to the oxaloacetate binding site that could serve as a binding site for acetyl-CoA, analogous to that observed in the distantly related homolog malate synthase G (183), although this channel is on the opposite side of the active site.

Additionally, a disordered loop (residues 121 to 132) from an adjacent protomer lies directly above the active site and could play a role in catalysis by acting as the lid for the reaction. Unlike in SbnG, this loop is well ordered in all structurally characterized homologs and faces away from the adjacent active site.

Table 3-2 Pairwise sequence alignment statistics of wild type SbnG to homologs from other siderophore biosynthetic operons. Protein Organism Sequence identity E-Value (%) SbnG S. pseudintermedius 70 3E-128 RRSL 03032 R. solanacearum 44 9E-65 AcsB S. glossinidius 46 2E-86 AcsB P. syringae 47 4E-84 PSPA7 3089 P. aeruginosa 37 8E-56 MhpE X. oryzae 36 2E-37

SbnG shares a similar overall fold with the phosphoenolpyruvate/pyruvate domain superfamily and features of the active site architecture across this family of proteins are remarkably conserved despite the differences in metal binding and catalytic function. In addition to the polypeptide backbone, the amino acids at the equivalent positions to Glu46, Arg72,

Glu151, Asp177, are all conserved across this family. In contrast, His47 and His96 of SbnG show variation among homologs. Interestingly, in MPS these two residues are also histidines and both enzymes have specificity for oxaloacetate as a substrate (180,192). A multiple sequence alignment of SbnG and select class II aldolase homologs (25 to 70% sequence identity, Tables

3-1 and 3-2) revealed two residues in close proximity to the metal binding site that are differently conserved between homologs within the SbnG clade and the class II aldolases (Figure 3-6). The first residue is Met149, which is a glutamine in all representative class II aldolases (Figure 3-6).

In the MPS crystal structure, this glutamine residue is within H-bonding distance to the pyruvate product and the active site arginine residue (Figure 3-7A). The second residue is Ala175, which is a proline residue in all representative class II aldolases (Figure 3-7B). These substitutions may account for differences in metal binding under physiological conditions.

Figure 3-6 Multiple sequence alignment of select SbnG homologs from the phosphoenolpyruvate/pyruvate domain superfamily. Sequences are identified by the genus, species and protein name. Listed above the sequence alignment is the secondary structure for SbnG with the arrows indicating -sheets and zigzags indicating -helices. Grayscale highlighting represents the degree of conservation between sequences. Conserved active site residues are highlighted in red. Additional conserved regions found only in SbnG and homologs from other siderophore biosynthetic gene clusters are highlighted in blue boxes and by a blue arrow.

Figure 3-7 Superposition of the metal-binding motif from SbnG with MPS. Residues preceding (A) and subsequent (B) to Glu151 of SbnG are drawn as sticks (carbon atoms in blue). Two side-chain conformations modeled for Met149 are drawn. The side chains Arg72 and Asp177 and the Ca2+ are drawn for context. The equivalent residues and the Mg2+ of MPS are shown (carbon atoms in green). Nitrogen, oxygen and sulfur atoms are colored blue, red and yellow, respectively. 84

SbnG was shown to catalyze an identical reaction to TCA cycle citrate synthases (81).

Yet, the fold of both proteins are different as SbnG forms an ()8 barrel, while citrate synthase types I and II are -helical. Two histidines and an aspartic acid participate in catalysis by TCA cycle citrate synthase by acting as general acids and general bases during catalysis. The active site of SbnG contains the same configuration of these three residues: His47, His96 and Asp177.

The side chains of His47, His96 and Asp177 (15 atoms total) of the oxaloacetate bound SbnG

E151Q structure were superimposed on the equivalent residues of porcine heart citrate synthase

(type I) and of Acetobacter aceti citrate synthase (type II) (193,194). The overlays clearly demonstrate a conserved spatial distribution of the three catalytic residues with root mean square deviation (r.m.s.d.) values of 2.2 and 2.3 Å with citrate synthases type I and II, respectively

(Figure 3-8). In SbnG E151Q and the structures of either citrate synthase types I and II, the catalytic Asp residues are not observed to H-bond to oxaloacetate. Additionally, Arg72 of SbnG

E151Q roughly overlays with arginine residues of the two citrate synthases that also interact with bound substrates. However, the orientation of Arg72 relative to the TCA cycle citrate synthases are ~90° apart. The conformation of this arginine may determine the orientation of oxaloacetate binding in the active site which is also rotated by ~90° (Figure 3-8). The acetyl CoA analog

(carboxymethyldethia-CoA, CMX) bound to citrate synthase from A. aceti does not overlay with the proposed binding site in SbnG. Rather, a similar ~90° rotation is required to align the acetyl

CoA with the positively charged groove with respect to the oxaloacetate molecule.

Figure 3-8 Comparison of the SbnG active site with TCA cycle citrate synthases. Overlay of SbnG E151Q active site residues (His47, His96 and Asp177) (blue) to catalytic residues in citrate synthase type I from porcine heart (PDB ID: 4CTS) (green) and type II active site residues from A. aceti (PDB ID: 2H12) (brown). Oxygen and nitrogen atoms were shown in red and blue, respectively. Oxaloacetate (OAA), shown in stick form, was bound in all three structures and colored to match their respective active site residues. Carboxymethyldethia coenzyme A (CMX) is derived from A. aceti citrate synthase and shown in stick form.

The structural similarity between the active sites of SbnG and TCA citrate synthases suggests convergent evolution of a clade of the phosphoenolpyruvate/pyruvate domain superfamily. Expression of SbnG active site variants, including the residues His47, His96 and

Asp177, had significant reductions in SB production in a citrate synthase-deficient strain of S. aureus, relative to wild type SbnG. A decrease in citrate synthase activity was observed in citrate synthase type I from porcine heart when the equivalent catalytic residues His274, His320 and

Asp375 were mutated (195). Reduced production of SB by variants with substitutions at residues

Glu46, Arg72, and Glu151 suggests that these residues are also required for full catalytic activity of SbnG, and is supported by their conservation amongst other SbnG homologs in the new clade defined for the phosphoenolpyruvate/pyruvate domain superfamily (Figures 3-4 and 3-6).

Based on the structural similarity of the active site with those of TCA cycle citrate synthases (Figure 3-8) and the mutagenesis data (Figure 3-5), the catalytic mechanism for SbnG is likely to be similar to that of TCA citrate synthases. Analysis of oxaloacetate binding in the

SbnG E151Q structure did not readily identify the specific catalytic role of His47 and His96. The position of oxaloacetate in the active site may differ in the wild type structure or in presence of acetyl-CoA. Nonetheless, Asp177 is proposed to act as a general acid and either His47 or His96 as a general base to afford the enol of acetyl-CoA. A condensation reaction between the enol and oxaloacetate could be assisted by either His47 or His96 acting as a general acid to yield a citryl-

CoA intermediate. Hydrolysis gives the products citrate and CoA.

In summary, I have defined SbnG as a new structural class of citrate synthase from S. aureus. Therefore, I suggest reclassifying SbnG as a type III citrate synthase. Formally part of the metal-dependent class II aldolase family, I have demonstrated that SbnG has lost the requirement for a metal cofactor and together with homologs identified in other siderophore biosynthetic gene clusters, form a new metal-independent category of class II aldolases.

Chapter 4: Synthesis of L-2,3-diaminopropionic acid, a siderophore and

antibiotic precursor

4.1 Introduction

L-Dap is an amino acid produced by numerous plants and bacteria. L-Dap is commonly incorporated into secondary metabolites as exemplified by the synthesis of neurotoxin -N- oxalyl-l-,-diaminopropionic acid in the seeds of the legume Lathyrus sativus (196) and is a key precursor in the synthesis of many antibiotics such as viomycin and capreomycin, two structurally related anti-tuberculosis drugs that belong to the tuberactinomycin family of nonribosomal peptide antibiotics (73,197). Despite the prevalence of L-Dap in nature, an L-Dap biosynthetic pathway has yet to be experimentally elucidated.

SbnA and SbnB are homologous to VioB and VioK, respectively, two enzymes in the viomycin biosynthetic pathway that were originally proposed to be involved in L-Dap synthesis based on bioinformatic analyses (73). VioK is predicted to be an ornithine cyclodeaminase

(OCD) that converts L-ornithine to L-proline and an ammonium ion using NAD+ as a cofactor.

VioB is a putative O-acetyl-L-serine sulfhydrylase (OASS) that is thought to catalyze a - replacement reaction of the L-serine hydroxyl group, using the ammonium ion liberated from the previous VioK reaction to generate L-Dap (73). Previous precursor labeling studies of viomycin and capreomycin have revealed L-serine as a probable precursor for L-Dap (198,199). Based on sequence homology, SbnA and SbnB are annotated as OASS and OCD, respectively. Four possible schemes were proposed for L-Dap biosynthesis from various metabolites including L- ornithine and O-acetyl-L-serine (80).

In this study, the biochemical properties of SbnA and SbnB were characterized to elucidate the L-Dap biosynthetic pathway. Through a combination of biochemical assays, NMR and mass spectrometry, I demonstrate that SbnA performs a -replacement reaction dependent upon O-phospho-L-serine (OPS) and L-glutamate to produce a serine-glutamate conjugate N-(1- amino-1-carboxy-2-ethyl)-glutamic acid (ACEGA). Oxidative hydrolysis of ACEGA catalyzed by SbnB is shown to release -KG and L-Dap, thereby providing a biological source for two of the three precursors required for SB biosynthesis in S. aureus.

4.2 Results

4.2.1 SbnA produces ACEGA from O-phospho-L-serine and L-glutamate

SbnA is homologous to the Class II PLP-dependent family of enzymes and shares approximately 36% amino acid sequence identity with structurally characterized OASS (200).

An absorption maxima at 412 nm was observed in electronic spectra of recombinantly expressed

SbnA (Figure 4-1A), which was attributed to an internal Schiff base formed between the PLP cofactor and an active site lysine residue. Despite homology to OASS, addition of excess O- acetyl-L-serine (OAS) or L-serine did not alter SbnA absorption spectra, even after 5 minutes of incubation, signifying that the expected aminoacrylate external aldimine did not form in SbnA.

Another SbnA homolog, CysM from Mycobacterium tuberculosis (34% sequence identity), is an

O-phospho-L-serine sulfhydrylase (OPSS) that forms part of an alternative cysteine biosynthetic pathway (201). Adding OPS to SbnA caused a rapid change in electronic spectra with the appearance of peaks at 324 and 467 nm, characteristic of the formation of an external aminoacrylate intermediate coupled with phosphate release (Figure 4-1A). This intermediate was susceptible to nucleophilic attack by an amine donor to either form L-Dap directly or via an 89

intermediate (80). To identify the amine donor, SbnA turnover was measured by monitoring phosphate release from OPS in the presence of various amino acids. Only addition of L- glutamate resulted in a dramatic increase in phosphate release (Figure 4-1B). Also, adding excess

L-glutamate to the aminoacrylate aldimine form of SbnA resulted in a return to the resting state as observed by the single peak at 412 nm in electronic spectra (Figure 4-1A).

Figure 4-1 SbnA condenses OPS and L-glutamate to generate ACEGA. (A) UV-visible absorption spectra of SbnA (black line), SbnA in complex with OPS (blue line) and after the addition of excess L-glutamate (red line). The spectra were recorded with 30 M SbnA, 30 M OPS and 3 mM L-glutamate. (B) SbnA catalyzes the increased release of inorganic phosphate from OPS in the presence of L-glutamate. After two minutes, 2 mM of various amino acids were added to the SbnA reaction mixture. All reactions were performed in triplicate. (C) LC-ESI-MS analysis of the SbnA reaction for product formation. A mass ion species of 235 corresponded to the predicted mass of ACEGA was identified. All mass spectra were recorded with retention times between 4.6-5.5 minutes.

Based on the identified optimal substrates OPS and L-glutamate, the resulting product was proposed to be an L-glutamate-L-serine conjugate ACEGA defined by a secondary amine formed between the serine -carbon and L-glutamate amino nitrogen (Figure 4-2). SbnA reaction mixtures were analyzed using liquid chromatography-electrospray ionization (LC-ESI-MS) mass spectroscopy in positive ion mode. Only the full reaction containing SbnA, OPS and L-glutamate yielded a dominant mass ion at [M+H]+ = 235 Da, which was similar to the predicted mass of

234 Da for ACEGA (Figure 4-1C). When either substrate or protein was omitted the 235 Da mass ion peak was not detected (Figure 4-1C). The structure of ACEGA was verified by NMR using a combination of 1H, 13C, HSQC and HMBC analyses (Figure 4-3). All expected cross- peaks for ACEGA were observed in the HMBC spectra with all protons showing correlation to every carbon separated by two and three bonds (Figure 4-3). Spectra for excess L-glutamate from the reaction and glycerol, a possible contaminant from dialysis, were also observed. Chemical shifts for ACEGA obtained from the 1H, 13C spectra are shown in Table 4-1. These data are in agreement with the suggestion that SbnA utilizes OPS and L-glutamate as substrates in a reaction that generates ACEGA.

Figure 4-2 Proposed biosynthetic pathway for L-Dap. 91

Figure 4-3 Structure of ACEGA. Superimposed HSQC (blue) and HMBC (red) NMR spectra of the SbnA reaction mixture. Corresponding 1D spectra are shown along the axes. 1D 13C spectrum and HSQC peaks are labeled with carbon numbers corresponding to the structure of ACEGA shown at the top left. Signals from L-glutamate and glycerol are indicated with red asterisks and pink shading, respectively. 92

Table 4-1 1H and 13C chemical shifts of ACEGA at pH 8.0 and 25 ºC. Carbon number1 1H Shifts (ppm) 13C Shifts (ppm) 1 - 185.1 2 2.26 36.9 3 1.86, 1.90 31.5 4 3.24 66.0 5 - 182.4 6 3.03 50.6 7 3.74 56.6 8 - 176.9

4.2.2 SbnB hydrolyzes ACEGA to form L-Dap and -KG

SbnB is homologous to the -crystallin/ OCD protein family, which includes a subset of alanine dehydrogenases. These enzymes utilize NAD+ as either a cofactor (OCD), or substrate

(alanine dehydrogenase). SB biosynthesis requires both SbnA and SbnB to produce L-Dap (80) and I hypothesized that ACEGA generated from SbnA is a substrate of SbnB, along with NAD+, to produce L-Dap and -KG (Figure 4-2). To test this, SbnB was incubated with SbnA, OPS, L- glutamate and NAD+ and activity was measured by absorption at 340 nm as an indicator for the generation of NADH. Omitting L-glutamate or OPS from the reaction negated the conversion

NAD+ to NADH (Figure 4-4A). Since ACEGA is not commercially available, the reverse SbnB reaction was analyzed using L-Dap, -KG, and NADH as substrates. NADH consumption was measured by the decrease in absorption at 340 nm, which did not occur in the absence of L-Dap,

-KG, or NADH (Figure 4-4B). To demonstrate that -KG was consumed in the reverse reaction, the -KG concentration was measured using 2,4-dinitrophenylhydrazine (Dnph), a colorimetric assay for detecting ketone and aldehyde functional groups in solution. SbnB reactions with L-Dap, -KG, and NADH resulted in the rapid consumption of -KG (Figure

4-4C). The products of the SbnB reverse reaction were analyzed by LC-ESI mass spectroscopy, 93

revealing a prominent peak at [M+H]+ = 235 Da, identical to the mass derived from the SbnA forward reaction and attributed to ACEGA (Figure 4-4D). These results support that in the forward reaction of SbnB, ACEGA and NAD+ are consumed to produce L-Dap, -KG and

NADH.

Figure 4-4 SbnB degrades ACEGA to generate L-Dap and -KG. (A) SbnB activity was monitored via the generation of NADH. SbnB reactions were coupled with SbnA, OPS and L-glutamate. After two minutes, a total of 50 nmoles of NAD+ was added to each reaction mixture. (B) The SbnB reaction was reversible in the presence of NADH, L-Dap and -KG. SbnB activity was measured by monitoring the consumption of NADH. (C) -KG detection by the 2,4-dinitrophenylhydrazine assay. Aliquots of the reaction mixtures were removed every minute and assayed for total ketone levels. Error bars were recorded as the standard deviation for the mean from three replicates. (D) LC-ESI-MS analysis of the full SbnB reverse reaction for product formation and confirmation of the reaction intermediate. A mass ion species of 235 corresponded to the predicted mass of ACEGA was identified. The mass spectrum was recorded with a retention time between 7.3-8.9 minutes. 94

4.2.3 L-Dap and -KG generated from SbnA and SbnB are precursor substrates for SB biosynthesis

SB can be synthesized in vitro using purified synthetases: SbnC, SbnE and SbnF, along with a decarboxylase, SbnH and the precursor substrates: citrate, -KG and L-Dap (70,78). Since

SbnA and SbnB can produce two of the required substrates for SB biosynthesis (L-Dap and -

KG), I reformulated the complete SB synthesis reaction by adding SbnA, SbnB, OPS, L- glutamate and NAD+ to an in vitro reaction mixture in place of L-Dap and -KG. Reactions containing SbnA, SbnB, OPS, L-glutamate and NAD+ produced a CAS-positive result, demonstrating iron-chelating activity. Furthermore, the CAS-positive reaction mixtures were capable of promoting S. aureus growth in disc diffusion assays (Figure 4-5A). To verify whether

SB was indeed synthesized, I used a previously described S. aureus strain with a chromosomal deletion of the lone SB receptor, SirA (68). Both the CAS positive samples from our assay and from a positive control reaction mixture that included both L-Dap and -KG failed to support growth of the sirA deletion S. aureus strain (Figure 4-5B). These results indicate that both L-Dap and -KG generated from SbnA and SbnB can be incorporated to the complete biosynthetic pathway to form a functional SB siderophore.

Figure 4-5 L-Dap and -KG generated from SbnA and SbnB can be used by Sbn siderophore synthetases to create SB in vitro. (A) Disk diffusion growth assays were performed with TMS agar plates seeded with S. aureus. SB enzyme reaction mixtures were spotted onto sterile paper disks and the diameter of growth was measured 48 hours after incubation. The Full Reaction sample contains enzymes SbnA and SbnB and the SB sample is the positive control containing -KG and L-Dap. Error bars were recorded as the standard deviation for the mean from three replicates. The dashed line represents the diameter of the paper disk. (B) Control disk diffusion growth assays of SB enzyme reaction mixtures spotted onto S. aureus sirA deletion strain seeded plates. SirA is the lone receptor for ferric bound SB and loss of SirA receptor prevents growth with ferric bound SB. The dashed line represents the diameter of the paper disk.

4.2.4 Structures of SbnB complexes

The crystal structure of SeMet labeled SbnB was solved by anomalous dispersion methods and refined to 2.1 Å resolution. Subsequently, structures were solved of SbnB-NAD+ in complex with malonate, citrate and -KG, as well as SbnB-NADH in complex with ACEGA to resolutions of 1.85 to 2.36 Å. All of the SbnB crystals were of the space group P41212 with a single molecule in the asymmetric unit.

Figure 4-6 Structure of SbnB. (A) The overall fold of apo-SbnB with the backbone shown as a schematic diagram. Dimerization and NAD binding domains colored green and orange, respectively. (B) The SbnB dimer as obtained through crystallographic symmetry. Individual SbnB monomers are colored blue and green. (C) -KG bound into the active site of SbnB in the presence of NAD+. Omit difference electron density is shown as a gray mesh contoured at 1.0 All residues and ligands are shown as sticks and the dashed line represents the distance between the C2 atom of -KG and C4 atom of NAD+. (D) ACEGA bound into the active site of SbnB in the presence of NADH. Omit difference electron density is shown as a grey mesh contoured at 1.0 . All residues and ligands are shown as sticks and the dashed line represents the distance between the C2 atom of ACEGA and C4 atom of NADH.

The SbnB monomer was composed of an NAD binding domain containing a 7-stranded parallel -sheet surrounded by helices in a classic dinucleotide-binding fold and a dimerization domain that was made almost entirely of a 6-stranded anti-parallel -sheet (Figure 4-6A). A

SbnB dimer could be reconstructed by the application of crystallographic symmetry (Figure 4-

6B). Analysis by the program PISA (202) showed that the interaction interface was at the large antiparallel -sheets burying ~5450 Å2 of surface area. In solution SbnB (~38 kDa monomer) exists primarily as a dimer with mass estimates of 81 kDa and 71 kDa as determined by dynamic and multi-angle light scattering, respectively (Figure 4-7).

Figure 4-7 Molar mass determination of SbnB by SEC-MALS. The elution profile (black line) was expressed as a rayleigh ratio and the calculated molar mass (blue line) was constrained to the single protein elution peak.

Comparison of the apo-SbnB structure with the ligand bound complexes revealed that

NAD+/NADH was bound at full occupancy in a preformed rigid pocket (Figure 4-8A). Each

SbnB monomer contains an NAD+/NADH bound at the base of the interdomain channel between the NAD binding and dimerization domains (Figure 4-8A). A modest difference between the

+ apo-SbnB and SbnB-NAD structures was a subtle shift of the 4/5 loop between the dimerization domains, highlighted by the 1.1 Å shift of Lys89 C. The loop displacement was

likely the result of a hydrogen bond formed between Arg94 N and N at the N-terminal end of

+ + 5 NAD phosphate oxygen atoms. In the SbnB-NAD structure, the planar nicotinamide ring confirms NAD+ was in the oxidized state as compared to the SbnB-NADH structure, in which the ring was distorted from planarity (Figures 4-6C and 4-6D). The entry and exit channels for substrate and products were ~12 Å long and oriented on opposite faces of the dimer (Figures

4-6B and 4-8A). The substrate binding pocket adjacent to the nicotinamide ring re face was observed to have a predominantly positive electrostatic potential with three arginines (Arg60,

Arg94 and Arg122) and one lysine (Lys78) (Figure 4-8B). This electrostatic tuning was well- suited to bind negatively charged substrates and crystals grown in the presence of malonate and citrate had clear density for the respective molecule in the active site. In both structures, a terminal carboxylate was anchored by a salt bridge to Arg94, pointing to its role in stabilizing a terminal substrate carboxylate (Figure 4-8C).

Figure 4-8 SbnB active site. (A) SbnB active site was located near the dimerization domain. Each monomer contained a single NAD+ molecule. Individual monomers were colored green and blue. NAD+ and active site residues were shown as sticks and colored white, and green or blue, respectively. (B) SbnB active site contains a large amount of positively charged residues surrounding the nicotinamide portion of NAD+. SbnB active site was overlaid with electrostatic surface potential as calculated by APBS (163) in Pymol (203). Malonate was omitted from the active site. (C) Structural 99

superposition of SbnB structures bound to -KG, malonate and citrate in the presence of NAD+. All residues and ligands are shown as sticks.

The product of the SbnA reaction, ACEGA, was soaked into SbnB crystals in the presence of NAD+, or NADH. Soaking with NAD+ and ACEGA resulted in observed electron density in the active site pocket that was modeled as an -KG molecule at 80% occupancy and an average B-factor of 42.1 Å2, indicating that SbnB in crystal form remains catalytically active

(Figure 4-6C). Density for a L-Dap molecule was not observed. A structural superposition shows that -KG binds in a similar manner to both malonate and citrate such that their terminal carboxylates interact with Arg94 (Figure 4-8C). NADH was soaked into SbnB crystals in conjunction with ACEGA and electron density for intact ACEGA was clearly visible and was modeled in at 87% occupancy with an average B-factor of 45.0 Å2 (Figure 4-6D). Overall, no large structural differences were observed between SbnB-NADH-ACEGA and SbnB-NAD+--

KG, which were superimposed with an r.m.s.d. of 0.5 Å over all atoms. ACEGA was bound in the active site in an extended conformation parallel to the nicotinamide ring, such that the

L-glutamate derived portion was at the base of the pocket and the OPS derived moiety was directed towards the solvent channel (Figure 4-6D). The side chain of Arg122 forms a bridged interaction with both ACEGA and the amide group in NADH. A similar interaction was also observed in the SbnB-NAD+ structures in complex with citrate and -KG (Figure 4-8C). An additional hydrogen bond interaction between the terminal carboxylate of -KG and Lys45 was observed in the SbnB-NAD+--KG structure. The L-glutamate derived portion of ACEGA, like

-KG and citrate in the other SbnB complexes, formed hydrogen bonds to Lys78, Arg94 and

Arg122 (Figures 4-6C, 4-6D and 4-8C). The spatial distribution of Lys78, Arg94 and Arg122

100

appears to optimally accommodate acidic substrates up to five carbon atoms long. Additionally,

Arg60 was observed to interact with both L-glutamate and OPS derived portions of ACEGA.

ACEGA was aligned such that the C2 atom from the L-glutamate derived portion was 3.8 Å from the NADH C4 atom (Figure 4-6D). In contrast, the -KG C2 and NADH C4 atoms were offset and the distance elongated to 4.0 Å due to the addition of a ketone, which repositions the

-KG terminal carboxylate to form hydrogen bonds with Lys45 and Arg122 (Figure 4-6C). The ketone in -KG was observed to be within hydrogen bonding distance to Lys78 (Figure 4-6C).

In summary, the binding sites for ACEGA, -KG, citrate, and malonate largely overlap and many of the same residues in SbnB (Figures 4-6C, 4-6D and 4-8C) are involved in hydrogen bond interactions with the bound substrates; however, only in ACEGA is the C2 atom aligned for optimal hydride transfer.

4.3 Discussion

In this study, I demonstrated that SbnA and SbnB function in tandem to produce L-Dap and -KG from OPS and L-glutamate (Figure 4-2). Through LC-MS and X-ray crystallographic experiments, I demonstrated that ACEGA is an intermediate in the L-Dap biosynthetic pathway.

Previously, based on sequence similarity SbnA was annotated as an OASS that displaced the acetyl group of OAS with an ammonium ion to produce L-Dap in one step (73). Our results show that OPS and L-glutamate are substrates for SbnA and forms the nonproteinogenic amino acid

ACEGA (Figure 4-2). Indeed, SbnA shares moderate sequence similarity to CysM a characterized OPSS; however, instead of a sulfur nucleophile, SbnA utilizes nitrogen from L- glutamate, distinguishing this enzyme from all known sulfhydrylases. Interestingly, no reaction

101

is observed for SbnA with OAS in contrast to several characterized OPSS (201,204). Thus, although SbnA is clearly homologous to the OAS/OPSS superfamily, SbnA is the archetype for an enzyme family that condenses OPS with potentially any amino acid with the release of inorganic phosphate. Detailed analysis of the substrate specificity and the overall reaction mechanism for SbnA are currently under investigation.

SbnB catalyzes the second step of the L-Dap pathway to generate L-Dap and -KG from the intermediate ACEGA using NAD+ as a substrate (Figure 4-2). SbnB is homologous to the - crystallin/OCD protein family suggesting that it functions similar to either ornithine cyclodeaminases or NAD+-dependent amino acid dehydrogenases. Our analysis revealed that

SbnB uses NAD+ as a substrate, which is a reaction more closely related to that of the NAD+- dependent amino acid dehydrogenases. A well-characterized example is alanine dehydrogenase

(AlaDH) from Archaeoglobus fulgidus (205). Superposition of the active site of SbnB with those from AlaDH and OCD is presented in Figure 4-9. Four residues that form the base of the active site are conserved in all three enzymes: Lys45, Lys78, Arg122 and Asp313 (SbnB numbering).

Two of these residues, Lys78 and Arg122, hydrogen bond to the terminal carboxylate of substrates bound to SbnB and OCD to align the carbon atom of the substrate with NAD+.

SbnB differs from these homologs in that the active site in has been expanded to accommodate a larger substrate than alanine or ornithine. A unique feature of SbnB is the presence of Arg94 forms a salt bridge to the terminal carboxylate of ACEGA.

102

Figure 4-9 Structural overlay of homologs to SbnB. Structural overlays were assembled for SbnB against two homologs (A) alanine dehydrogenase (AlaDH), (tan) from Archaeoglobus fulgidus (PDB ID: 1OMO) and (B) ornithine cyclodeaminase (OCD), (purple) from Pseudomonas putida (PDB ID: 1X7D). SbnB ligands and residues were colored green with the exception of NADH, which was colored white. All active site residues and ligands are shown as sticks.

Based on the structural homology in the active sites of SbnB with AlaDH, I propose a mechanism for SbnB based on analogy to the mechanism proposed for AlaDH. In this mechanism, SbnB catalyzes a hydride transfer from the C2 position of ACEGA to NAD+, followed by a water-mediated nucleophilic attack to generate -KG and L-Dap (Figure 4-10).

The identity of the base has not been defined in AlaDH or SbnB. Inspection of the superposition of the SbnB structure with that of AlaDH suggests two candidates: Lys45 and Lys78 (Figure 4-

9A). Of these residues Lys45 is not conserved in OCD (Figure 4-9B); however, an external base may not be required in the cyclization reaction to form proline. As noted herein, Lys78 participated in a hydrogen bond interaction with the terminal carboxylate of the substrate limiting a dual role. Lys45 is more distant from the substrate and an alternate side-chain

103

conformation is required for proton access. Future site-directed mutagenesis studies may assist in defining a possible catalytic role to Lys45.

Figure 4-10 Proposed catalytic mechanism for oxidative hydrolysis of ACEGA to generate L-Dap and -KG in SbnB. The first step involves a hydride transfer from the C2 atom of the L-glutamate-derived portion of ACEGA to NAD+. A water molecule attacks the resulting imine at the C2 atom, facilitated by the base, B, to form a carbinolamine intermediate. Collapse of the carbinolamine intermediate results in the formation of -KG and L-Dap.

Currently, the L-Dap biosynthetic pathway has been identified in the one siderophore as characterized in this paper; however, L-Dap biosynthetic genes are found in a handful of known antibiotics (80). Two related L-Dap containing tuberactinomycin antibiotics, viomycin and capreomycin, are used to treat drug resistant M. tuberculosis infections (73,197). Notably, the synthesis of viomycin and capreomycin involve -KG dependent oxygenases VioC and CmnC,

104

respectively (206). Thus, the homologs of SbnA and SbnB in these pathways would also supply both precursors.

Identification of OPS and L-glutamate as substrates for L-Dap synthesis enabled us to explore how SB biosynthesis integrates with primary metabolism. S. aureus grown in the presence of human serum or blood undergoes a global metabolic reorganization that includes responding to low iron stress by the ferric uptake repressor (Fur) regulatory network (103). Low iron stress induces an ‘iron sparing response’ in S. aureus that down-regulates many iron- dependent metabolic pathways, including the tricarboxylic acid (TCA) cycle. Simultaneously, the glycolytic pathway is up-regulated leading to the overproduction of pyruvate and lactate

(111,116). Consequently, both citrate and -KG generated from the TCA cycle would be in limited supply for SB biosynthesis. The recent identification of SbnG as a citrate synthase provided a partial explanation of the paradox for SB biosynthesis under iron deprivation (81).

Now that an alternative pathway is known for -KG production, in addition to L-Dap, a complete set of precursors for the SB biosynthetic pathway is available in the absence of a functioning TCA cycle.

The metabolites OPS and L-glutamate are likely to be available in S. aureus in the absence of a TCA cycle under iron restriction. Analysis of the S. aureus transcriptome data under iron deprivation (115) provides an explanation for how OPS and L-glutamate are supplied for SB biosynthesis. The biosynthetic pathway (serAC) for OPS biosynthesis from L-glutamate and the glycolytic intermediate glycerate 3-phosphate is up-regulated. Second, also up-regulated is a pathway (gltBD) for L-glutamate biosynthesis from L-glutamine and -KG. Analysis of these pathways revealed that L-glutamine and glycerate 3-phosphate yield a stoichiometric amount of

L-DAP and -KG (Figure 4-11). Since SB biosynthesis requires two L-DAP and one -KG, the 105

excess -KG produced could be used to generate additional L-glutamate. Tracing back SB precursor pathways reveal glucose and L-glutamine to be the source for production of the basic building blocks required to generate SB. NMR studies of human serum revealed that glucose and

L-glutamine are two of the four most abundant metabolites available (113).

Figure 4-11 Summary of the overall stoichiometric equation for O-phospho-L-serine and L- glutamate production required for SB biosynthesis. Excess -KG is generated as two L-Dap molecules are required for full SB assembly.

The genes serAC and gltBD encode enzymes in the pathways to produce OPS and L- glutamate and form part of the CodY regulon in S. aureus (99,100). CodY, a global regulator for adapting to nutrient limitations and virulence, is primarily found in Gram positive bacteria. In S. aureus, the CodY regulon appears to complement siderophore production through increasing production of metabolic precursors. Also, microarray analyses show that CodY co-regulates iron uptake systems in Bacillus anthracis along with Fur (207,208). S. aureus appears to have adapted to take advantage of available host nutrients as a strategy to produce SB without the need of a functioning TCA cycle to satisfy its iron requirements.

106

Chapter 5: Deciphering the substrate specificity of SbnA, the enzyme

catalyzing the first step in staphyloferrin B biosynthesis

5.1 Introduction

SbnA is a PLP containing enzyme that catalyzes the first step in the L-Dap biosynthetic pathway by utilizing OPS and L-glutamate as substrates to produce ACEGA. SbnA also shares homology to OASS enzymes, more commonly referred to as L-cysteine synthases (80). SbnA homologs are present in other organisms that are predicted to produce SB (80) or structurally similar secondary metabolites, including the antibiotics zwittermicin A (209), viomycin (73), capreomycin (197) and dapdiamide (210). OASS catalyzes the formation of a PLP-- aminoacrylate intermediate from OAS with release of acetate (211-215). Additionally, a subset of previously annotated OASS enzymes has been discovered to utilize OPS as a substrate instead of OAS. The first identified OPSS was from a hyperthermophile archaea Aeropyrum pernix K1, which catalyzes the formation of L-cysteine from OPS and sulfide (216). Unlike currently characterized OPSS, SbnA utilizes a unique second substrate in L-glutamate instead of a sulfur donor. Given the similarities to OASS and OPSS enzymes, it is unclear how SbnA specificity for

OPS and L-glutamate is achieved.

Here, I demonstrate that SbnA forms a PLP--aminoacrylate intermediate when incubated with OPS and this activity is mediated by residues Arg132, Tyr152 and Ser185. The intermediate state of SbnA induced a conformation change that formed a putative L-glutamate binding site occupied by a citrate molecule. Lastly, I identified L-cysteine as a potent competitive

107

inhibitor of SbnA activity. These findings reveal the structural basis for OPS specificity and a potential regulatory link between SB biosynthesis and the oxidative stress response in S. aureus.

5.2 Results

5.2.1 Structure of SbnA

SbnA crystallized with a single molecule in the asymmetric unit, containing two mixed

/ domains. Each domain was made up of a central -sheet surrounded by several -helices with a single PLP bound within a central cavity between the domains (Figure 5-1A). The overall fold places SbnA in the Class II PLP-dependent enzyme family (217). Size exclusion chromatography in line with multi-angle light scattering demonstrated that SbnA (36 kDa monomer) forms a dimer in solution with mass of 77 kDa (Figure 5-2). A similar mass of 82 kDa was determined by dynamic light scattering (data not shown). The biological SbnA dimer could be reconstructed across the 2-fold crystallographic symmetry axis (Figure 5-1B). The dimer interface was contributed by ~50 residues, burying ~1795 Å2 of surface area from solvent (8% of total surface area) in each monomer, as determined using the program PISA (202).

The active site is in a large cleft lined by positively charged residues from one monomer and both active sites are accessible from same face of the dimer (Figure 5-1B). PLP co-purified with the enzyme and was observed covalently bound to Lys47 at the base of the cleft by an internal Schiff base originating from the PLP si face (Figure 5-1C). The structure of the as- isolated enzyme is referred to as SbnA-PLP. The PLP phosphate formed hydrogen bonds to

Ser185, Thr186, Thr187 and Ser189 and three water molecules (Figure 5-1C). The negative charge on the phosphate was stabilized by the positive helix dipole of the -helix spanning residues 188-200. The pyridoxal portion formed hydrogen bonds with Asn77 and Ser272 (Figure 108

5-1C). The re face PLP pyridine ring is exposed to solvent providing access for substrate to displace the lysine Schiff base at C4’ to form an external aldimine.

Figure 5-1 Structure of SbnA. (A) The overall fold of SbnA is shown as a cartoon. The PLP cofactor, colored magenta, is shown in stick form in the active site cleft. Secondary structures composed of loops, -helices

109

and -sheets are colored brown, blue and teal, respectively. (B) The SbnA dimer reconstructed by applying 2-fold crystallography symmetry. (C) The active site of SbnA with PLP interacting residues shown as sticks. Ordered water molecules are shown as blue spheres and the dashed lines represent potential hydrogen bonds. C, N, O, P atoms are colored brown, blue, red and orange, respectively with PLP carbons colored magenta.

Figure 5-2 SbnA is a dimer in solution as determined by SEC-MALS. The Rayleigh ratio was shown as a black line and the calculated molar mass was shown as a blue line.

5.2.2 SbnA forms a PLP--aminoacrylate intermediate with OPS

OPS was previously shown to be a substrate of SbnA (Chapter 4). Soaking OPS into

SbnA crystals resulted in an observable color change from pale to golden yellow, mirroring the absorption changes previously observed in solution, suggesting that a PLP--aminoacrylate intermediate formed in the crystals. The OPS-soaked SbnA crystal structure (SbnA-AA) was solved to 1.92 Å resolution. In the SbnA-AA structure, clear electron density for the PLP-- aminoacrylate was observed and could be modeled and refined at full occupancy with an average

B-factor of 24 Å2 (Figure 5-3A). Electron density corresponding to the OPS phosphate group was not observed. Lys47 is displaced by PLP--aminoacrylate formation and the sidechain 110

amine occupies replace a water molecule to form a hydrogen bond with the PLP phosphate

(Figure 5-1C and Figure 5-3B). Overlaying the SbnA-PLP and SbnA-AA structures (root mean square deviation (r.m.s.d.) of 1.1 Å over all Catoms) revealed that the C-terminal domain has rotated towards the N-terminal domain narrowing the active site cleft (Figure 5-3C). Within the active site, the pyridoxal ring of the PLP--aminoacrylate is rotated approximately 20° with respect to SbnA-PLP (Figure 5-3D). This rotation repositions the -aminoacrylate carboxylic acid to form hydrogen bonds to Thr74, Ser75, Leu78 and Gln151 (Figure 5-3B). In concert, the active site loop (residues 72-76), which contains a highly conserved serine (Ser75), is translated up to ~4 Å bringing it directly above the -aminoacrylate (Figure 5-3C).

111

Figure 5-3 SbnA adopts a closed conformation when bound to the PLP--aminoacrylate. (A) The active site of SbnA-AA contains the PLP--aminoacrylate intermediate and a citrate molecule. Both molecules are shown as sticks and colored white. The dashed line represents the distance between the -aminoacrylate -carbon and citrate. Omit difference (Fo-Fc) electron density is shown as a grey mesh contoured to 3.0 (B) The SbnA active site with the PLP-- aminoacrylate intermediate binding residues shown as sticks. Ordered water molecules are 112

shown as blue spheres and the dashed lines represent hydrogen bonds. (C) The overall structural overlay of SbnA-PLP (brown) and SbnA-AA (purple) as shown in cartoon reveals significant conformational changes in the C-terminal domain. The black arrow highlights directionality of the conformational change upon PLP--aminoacrylate formation. (D) Structural overlay of SbnA-PLP (brown) and SbnA-AA (purple) active sites reveal a significant conformational change in the loops containing residue Ser75. The dashed line indicated the distance between C atoms of Ser75 upon -aminoacrylate formation and the curved black arrows represent the conformational shift for PLP and Ser75. (E) Citrate binds into a newly identified pocket above the active site in the SbnA-AA structure. Citrate and bonding residues are shown as sticks and the dashed lines represent hydrogen bonds.

A second binding pocket is formed between the two domains directly above the PLP-- aminoacrylate. Citrate was present in the crystallization condition and electron density for a citrate molecule was identified in the newly formed pocket. Citrate was modeled and refined at

2 full occupancy with an average B-factor of 34 Å (Figure 5-3A). The citrate molecule was bound by an extensive hydrogen bonding network mediated by the positively charged sidechains from

Lys100, Arg132, and Arg224 as well as the backbone chain amide groups of Tyr128, Leu129, and Ser232 (Figure 5-3E). The -carbon of the -aminoacrylate was approximately 7.0 Å from the citrate (Figure 5-3A).

5.2.3 SbnA active site variants

Despite sharing homology with OASS enzymes, no activity was detected for SbnA using

OAS as a substrate. To identify residues that could provide specificity for OPS, a multiple sequence alignment was constructed to include homologs from other siderophore or antibiotic biosynthetic gene clusters, as well as characterized OPSS and OASS enzymes. The multiple sequence alignment revealed three residues that are predicted to be associated with L-Dap synthesis, Arg132, Tyr152, and Ser185, due to their conservation in SbnA and homologs in gene clusters with a SbnB homolog (Figure 5-4A). In the SbnA structures, these three residues 113

surround the -aminoacrylate and could interact with the OPS phosphoryl group (Figure 5-4B).

In contrast, in characterized OAS dependent homologs, positions 132, 152 and 185 are substituted by conserved alanine, phenylalanine and glycine residues. To determine the role these three residues play as key determinants of OPS versus OAS specificity, four variants were generated: R132A, Y152F, S185G, Y152F/S185G and evaluated for OPS and OAS dependent activity.

Figure 5-4 Identification of OPS discriminating residues. (A) Multiple sequence alignment of SbnA and homologous sequences found in the tryptophan synthase beta superfamily. Sequences are identified by a single letter for genus, the first two letters of the species name and the annotated protein name. The asterisk placed after the protein 114

name(s) indicates biochemically validated OASS or OPSS enzymes. Conserved active site residues between OASS and OPSS enzymes are highlighted in blue boxes. A four amino acid insert identified in SbnA homologs that are responsible for L-Dap synthesis is highlighted in a red box. Numbering along the top corresponds to the sequence position of S. aureus SbnA. S. au, Staphylococcus aureus; S. ps, Staphylococcus pseudintermedius; R. so, Ralstonia solanacearum; A. ol, Acinetobacter oleivorans; B. th, Bacillus thuringiensis; S. vi, Streptomyces vinaceus; M. no, Methylobacterium nodulans; S. de, Shewanella denitrificans; C. me, Cupriavidus metallidurans; P. ag, Pantoea agglomerans; M. tb, Mycobacterium tuberculosis; S. ty, Salmonella typhimurium; E. co, Escherichia coli; T. ma, Thermotoga maritime; M. ma, Mycobacterium marinum; A. pe, Aeropyrum pernix. (B) Stereo view of the SbnA-AA active site highlighting residues implicated in OPS selectivity. PLP-AA and residues are shown as sticks.

The four variants were tested for the ability to form the PLP--aminoacrylate intermediate with either OPS or OAS. Both the Y152F and S185G SbnA retained the ability to form the PLP--aminoacrylate intermediate with OPS (Figures 5-5A and 5-5B). In contrast, intermediate formation with OPS was abolished in Y152F/S185G and R132A SbnA (Figures 5-

5C and 5-5D). None of the SbnA variants formed the intermediate upon the addition of OAS, suggesting all three substitutions are required or other active site features are essential for reaction with OAS (data not shown). The rate of formation of the PLP--aminoacrylate intermediate with OPS was determined for wild type SbnA and the Y152F and S185G variants

(Figure 5-6). The rate of intermediate formation was best fit by a single-exponential function at each concentration of OPS tested and the dependence of OPS concentration on the apparent first- order rate constants were then best described by hyperbolic functions. Single turnover kinetic parameters for wild type, S185G, and Y152F SbnA are summarized in Table 5-1. Relative to wild type SbnA, mutating Ser185 to Gly did not greatly impact -aminoacrylate formation as evidenced by a modest ~2-fold decrease in the second order rate constant. Mutating Tyr152 to

Phe resulted in an over 1000-fold decrease in the second order rate constant. These data suggest

115

that Arg132, Tyr152, and Ser185 play a role in OPS binding and formation of the PLP-- aminoacrylate intermediate, though each contributes to varying degrees.

Figure 5-5 UV-visible absorption spectra of SbnA variants. (A) Wild type SbnA, (B) S185G SbnA, (C) Y152F/S185G SbnA and (D) R132A SbnA. The black line represents the apo protein and the red line represents the protein incubated with OPS. The spectra were recorded at room temperature for 20 M protein in 50 mM Tris pH 8.0 and 2 mM TCEP.

116

Figure 5-6 Rate of formation for the PLP--aminoacrylate intermediate. Plot of the rates of PLP--aminoacrylate formation as a function of OPS concentration, fit to a hyperbolic function for (A) wild type SbnA, (B) S185G SbnA and (C) Y152F SbnA.

Table 5-1 Single turnover kinetic parameters of wild type SbnA and SbnA variants with various concentrations of OPS. -1 -1 -1 kobs (s ) Kd (mM) K2 (M s ) SbnA WT 3.1 × 102 ± 1.2× 101 2.8 ± 0.2 1.1 × 105 ± 1.0 × 104 SbnA S185G 1.4 × 102 ± 6.6× 101 3.4 ± 0.4 4.0 × 104 ± 5.4 × 103 SbnA Y152F 1.5 ± 0.06 15 ± 1.7 1.0 × 102 ± 1.2 × 101

5.2.4 Kinetic analysis of SbnA and variants

Enzyme turnover was monitored in a coupled assay to monitor ACEGA production that contained SbnA, SbnB, OPS, L-glutamate and NAD+. Product formation was decreased for all four SbnA variants compared to wild type SbnA (Figure 5-7). R132A SbnA had no detectable activity, which correlates with absorption spectra that showed this variant did not form the - aminoacrylate intermediate (Figure 5-5). L-Dap, a product of SbnB, appeared to inhibit the 117

activity of SbnB causing interference in the coupled assay. Thus, SbnA turnover was monitored by a phosphate assay. Optimal wild-type enzyme activity as monitored by phosphate release was observed with no KCl (0-500 mM tested) at pH 8.0 (6.5-8.5 tested) in a 50 mM Tris buffer

(Figure 5-8). KCl likely inhibited enzymatic activity in vitro by impeding formation of the quinonoid intermediate, a phenomenon observed in other PLP-dependent enzymes (218,219).

Steady-state kinetics was used to investigate the mechanism of SbnA catalysis. Enzyme rates were measured for wild type SbnA and the S185G variant as a function of OPS and glutamate concentration (Figure 5-9) and Steady-state kinetic parameters are summarized in Table 5-2.

SbnA S185G was the only variant examined because all other variants were insufficiently active to accurately measure enzyme rates (Figure 5-7). SbnA S185G had a modest decrease in catalytic efficiency of ~4-fold compared to the wild type enzyme (Table 5-2).

Figure 5-7 SbnA active site mutations attenuated ACEGA production. SbnA activity was indirectly monitored by measuring SbnB NADH production. SbnB reactions were coupled with SbnA variants, OPS, L-glutamate and NAD+. Each reaction was collected and measured at 10 minutes in triplicate.

118

Figure 5-8 Optimization of SbnA activity. The phosphate assay was used to determine the rate of inorganic phosphate released over time. The calibration was arbitrarily standardized to 100% for the condition that provided the highest activity.

Figure 5-9 Steady-state kinetics of wild type SbnA (A and B) and S185G SbnA (C and D). The phosphate assay was used to measure the release of inorganic phosphate with respect to the concentrations of both OPS (A and C) and L-glutamate (B and D).

119

Table 5-2 Steady-state kinetic parameters of wild type SbnA and SbnA S185G. -1 -1 -1 Km (mM) kcat (s ) kcat/Km (M s ) OPS SbnA WT 0.072 ± 0.008 2.3 ± 0.06 3.2 × 104 ± 3.6 × 103 SbnA S185G 0.22 ± 0.009 1.6 ± 0.02 7.3 × 103 ± 3.0 × 102 L-glutamate SbnA WT 3.2 ± 0.2 3.6 ± 0.06 1.1 × 103 ± 7.8× 101 SbnA S185G 7.1 ± 1.5 2.0 ± 0.1 2.9 × 102 ± 6.4 × 101

5.2.5 SbnA variant crystal structures

To characterize the structural alterations of the active sites in Y152F and Y152F/S185G

SbnA, X-ray crystal structures were solved to 1.5 Å resolution. Electron density for PLP covalently bound to Lys47 was observed in the active site of both variant structures. The crystal structure of Y152F SbnA displayed minimal conformational change when compared to wild type

SbnA (r.m.s.d. of 0.11 Å over all C atoms) (Figure 5-10). Since minimal structural changes were observed and Y152 does not interact directly with the PLP--aminoacrylate, this residue may interact with the phosphate group of OPS, consistent with a decrease in the rate of intermediate formation. Alternatively, Y152 may mediate the second step in the reaction, the bond formation between the -carbon of the -aminoacrylate and the amino group of L- glutamate. In contrast, the Y152F/S185G SbnA structure displayed a greater conformational change when compared to SbnA-PLP (r.m.s.d. of 0.34 Å over all C atoms) (Figure 5-11). A significant conformational change was observed within the active site loop (residues 72-76), which is translated towards the PLP cofactor by up to ~2.5 Å when compared to SbnA-PLP

(Figures 5-11A and 5-11B). The conformational change shares some of the features that occurred upon -aminoacrylate formation in SbnA-AA. Gln151 was rotated ~40º about 2 to mimic the

120

orientation of Gln151 in SbnA-AA that turns in to bind the -aminoacrylate (Figure 5-11C).

Therefore, SbnA Y152/S185G adopted a similar active site conformation to SbnA-AA, even in the absence of the -aminoacrylate. These similarities suggest that the double mutant inhibits

OPS binding by mimicking the closed active site conformation of the protein, potentially explaining its lack of activity.

Figure 5-10 Structural overlay of wild type SbnA (brown) with Y152F SbnA (blue). The PLP cofactor is shown as sticks in the active site cleft.

121

Figure 5-11 Structure of the SbnA variant Y152F/S185G. Structural overlay of SbnA-PLP (brown), SbnA-AA (purple) and SbnA Y152F/S185G (blue) showing the: (A) overall fold, (B) L-glutamate binding pocket, and (C) PLP-AA binding pocket. All ligands and residues are shown as sticks. The dashed line represents the shift between C atoms of Ser75 upon OPS binding. 122

5.2.6 L-cysteine inhibits SbnA activity

In addition to OAS and OPS, some SbnA homologs can bind and utilize L-cysteine as a substrate (178). When L-cysteine was added to wild type SbnA, the characteristic internal aldimine absorption maxima at 412 nm diminished, but no spectral signature at 467 nm for a

PLP--aminoacrylate intermediate formed (Figure 5-12). Instead a subsequent increase at 320 nm was observed suggesting that L-cysteine bound to the PLP cofactor and formed an external aldimine bond (214). In CysM from M. tuberculosis, the reaction with L-cysteine results in the formation of the PLP--aminoacrylate and release of bisulfide (178). The spectra of SbnA implies that a different intermediate is formed that is unable to release bisulfide to form the

-aminoacrylate. The absorption spectra for all four variants demonstrated a similar decrease at

412 nm, though to varying levels, except for R132A, in which the absorption maxima shifted to

467 nm (Figure 5-12). Inhibition by L-cysteine of SbnA and S185G SbnA activity was characterized by steady-state kinetics (Figure 5-13). The Ki for L-cysteine was 91 ± 16 M, while the Ki for L-cysteine against S185G SbnA was 2.1 ± 0.5 mM, a ~23-fold reduction compared to wild type. Additionally, the single turnover kinetics of intermediate formation from

L-cysteine binding to SbnA and the variants were characterized (Figure 5-14 and Table 5-3). In contrast to reaction with OPS, no significant differences the second order rate constants were observed between wild type SbnA and most variants at physiologically relevant L-cysteine concentrations. However, in Y152F/S185G SbnA reacted with L-cysteine at a 10-fold greater rate than wild type SbnA. The data suggests that reaction with L-cysteine is mediated in a different manner compared to OPS binding and appears to be independent of Arg132, Tyr152 and Ser185.

123

Figure 5-12 UV-visible absorption spectra of wild type SbnA and SbnA variants with 1 mM L-cysteine. The spectra were recorded at room temperature for 20 M protein in 50 mM Tris pH 8.0 and 2 mM TCEP.

Figure 5-13 Inhibition kinetics for (A) wild type SbnA and (B) S185G SbnA against varying concentrations of L-cysteine.

124

Figure 5-14 Rate of formation for the PLP-cysteine intermediate. Plot of the rates of PLP-cysteine formation as a function of L-cysteine concentration, fit to a linear function for (A) wild type SbnA, (B) S185G SbnA, (C) Y152F SbnA, (D) Y152F/S185G SbnA and (E) R132A SbnA.

Table 5-3 Single turnover kinetic parameters of wild type SbnA and SbnA variants with various concentrations of L-cysteine. -1 -1 -1 kobs (s ) Kd K2 (M s ) SbnA WT 4.0 × 102 ± 1.2 × 101 SbnA Y152F 2.0 × 102 ± 8.6 SbnA S185G 7.1 × 102 ± 4.6 × 101 SbnA Y152F/S185G 18 ± 3.7 8.9 ± 3.2 2.0 × 103 ± 8.5 × 102 SbnA R132A 4.4 × 102 ± 1.7 × 101

125

5.3 Discussion

I recently elucidated the L-Dap biosynthetic pathway required for SB biosynthesis in S. aureus. In this pathway, SbnA generates the novel serine-glutamate conjugate, ACEGA, from

OPS and L-glutamate. SbnB then oxidatively hydrolyzes ACEGA to generate L-Dap and - ketoglutarate. Sequence homology places SbnA in the tryptophan synthase beta superfamily, which are all PLP-dependent enzymes. Among the biochemically characterized members of this superfamily, SbnA is most similar in sequence to OASS enzymes involved in L-cysteine biosynthesis. SbnA was hypothesized to produce L-Dap from OAS and an ammonium ion liberated from SbnB (80). However, SbnA falls into a small subclass of OASS-like enzymes that instead utilize OPS as a substrate. Unlike most characterized OASS/OPSS enzymes, SbnA uses a nitrogen donor rather than a sulfur donor such as bisulfide.

An ~100-fold difference in Km values of SbnA for OPS and L-glutamate was observed and may reflect the relative abundance of each metabolite in S. aureus under iron deprivation. In

S. aureus, L-glutamate is one of the most abundant amino acids with intracellular concentrations exceeding 100 mM in some growth conditions (220,221), sufficient to saturate SbnA. The intracellular OPS concentration has not been determined in S. aureus. Based on KEGG pathways

(222,223), the main route for L-serine biosynthesis is derived from glycolysis via OPS by the enzymes SerABC. The micromolar Km for OPS may reflect the low intracellular OPS concentrations in S. aureus. How the intracellular concentration of OPS and other amino acids including L-glutamate changes under iron deprivation remains to be determined.

Superposition of SbnA-PLP and SbnA-AA revealed the enzyme undergoes significant conformational changes. Similar collapsing of the active site upon PLP--aminoacrylate intermediate formation is observed in structures determined for homologs CysK1 and CysM 126

from M. tuberculosis (213,224). Loss of the phosphate group upon formation of the stable PLP-

-aminoacrylate intermediate from OPS precluded direct structural identification of residues involved in OPS specificity. However, the active site of SbnA-AA revealed three residues:

Arg132, Tyr152 and Ser185 are in close proximity to the -carbon of the -aminoacrylate intermediate and all three residues were shown to be involved in formation of the intermediate.

Sequence divergence in these active site residues from OASS enzymes likely allows SbnA to bind OPS with high specificity. I hypothesized that mutating the active site of SbnA to mimic active sites of OASS enzymes could shift substrate specificity from OPS to OAS. However, none of the active site variants reacted with OAS. An overlay between Y152F/S185G SbnA and an

OASS homolog, CysM from S. typhimurium (PDB ID: 2JC3), revealed minor structural differences surrounding the PLP cofactor (215). Therefore, substrate specificity appears to require some other factor in addition to the active sites residues Arg132, Tyr152 and Ser185.

A multiple sequence alignment of SbnA homologs revealed that all three residues identified as unique for OPS binding (Arg132, Tyr152 and Ser185) are conserved in enzymes from putative L-Dap biosynthesis pathways. In addition, examples of OPSS enzymes not involved in L-Dap biosynthesis were identified: CysM (178,201) and CysK2 (173) from M. tuberculosis and CysO (216) from Aeropyrum pernix K1. OPS is the preferred substrate of both

CysM and CysO, but unlike SbnA, they are also capable of reacting with OAS to generate the

PLP--aminoacrylate intermediate (178). CysK2 behaves more like SbnA, with specificity for

OPS alone (173). The CysM and CysO crystal structures have comparable active sites to SbnA except that SbnA is missing a loop that lies directly above the PLP cofactor. This loop contains a conserved arginine that is important for OPS specificity over OAS in CysM (201). The analogous arginine is missing from SbnA, but Arg132 is located on the opposite side of the 127

active site pocket and is essential for OPS activity, suggesting it fulfills a similar role.

Additionally, CysO lacks the conserved Tyr152 present in SbnA, CysM and CysK2, which is replaced with a phenylalanine residue (225,226). A CysO variant K127A crystal structure complexed with OPS revealed that the OPS phosphoryl group had limited interactions with active site residues, though selectivity may not be an issue for CysO, since A. pernix is not predicted to synthesize OAS as genome sequence analyses failed to identify a serine acetyltransferase that catalyzes the formation of OAS from L-serine and acetyl-CoA (216).

Furthermore, A. pernix is a hyperthermophile living in temperatures between 90 and 95 °C where

OAS is labile at high temperatures while OPS is heat stable (216). Therefore, the requirement of a tyrosine residue in CysO to improve substrate specificity may not be essential given its unique environment.

In the SbnA-AA, a citrate molecule was identified bound above the PLP-- aminoacrylate. Since citrate is not a substrate for SbnA and is structurally similar to L-glutamate

I hypothesize that L-glutamate binds at this site with the nitrogen poised to attack the -carbon of the -aminoacrylate. Attempts to co-crystallize or soak L-glutamate into SbnA-OPS crystals were unsuccessful. The high concentration of citrate present in the crystallization solution (~1.0

M) likely outcompetes L-glutamate for binding into the second substrate binding pocket.

Inspection of citrate bound to SbnA suggests that L-glutamate binding is likely mediated by three conserved positively charged residues: Lys100, Arg132 and Arg224. Also, the multiple sequence alignment revealed a four amino acid insertion (G126G127Y128L129) preceding Arg132 that are modestly conserved in SbnA and homologs that are implicated in L-Dap biosynthesis, but not present in other OASS or OPSS enzymes (Figure 5-4A). Tyr128 and Leu129 mainchain atoms form hydrogen bonds with citrate in the SbnA-AA (Figure 5-3E). The Tyr128 sidechain formed 128

two CH/ interactions to Pro99 and Pro122 in both the open and closed conformations.

Additionally, the Leu129 sidechain forms part of the citrate-binding pocket in the SbnA-AA. The location and conservation of these residues supports a role of this pocket in determining the specificity of SbnA for L-glutamate.

Formation of the PLP--aminoacrylate intermediate in SbnA is competitively inhibited by L-cysteine. Typically, L-cysteine is generated as the product from OASS and OPSS enzymes

(201,213). Based on the KEGG database, an analysis of amino acid biosynthetic pathways in S. aureus revealed a direct link between L-Dap and L-cysteine (222,223). L-cysteine is ultimately derived from the glycolytic intermediate 3-phosphoglycerate (Figure 5-15). To produce L- cysteine, OPS is dephosphorylated to produce L-serine by phosphoserine phosphatase (SerB).

Next, L-serine is acetylated by a serine acetyltransferase (CysE) and then converted to L-cysteine by cysteine synthase (CysK). Therefore, L-Dap, L-serine and L-cysteine are all derived from

OPS. Thus, inhibition of SbnA by L-cysteine would reduce OPS consumption and favor the production of L-serine and ultimately L-cysteine.

129

Figure 5-15 Proposed flow diagram of staphyloferrin B biosynthesis integrated into central metabolism of S. aureus under nutrient deprivation and oxidative stress. Green arrows represent up-regulated metabolic pathways; red arrows indicate down-regulated metabolic pathways and black arrows represent unchanged metabolic pathways. Orange, blunt ended lines represent inhibition pathways. ROS stands for reactive oxygen species.

Inhibiting SbnA activity with L-cysteine may be another example of coordination between iron homeostasis and oxidative stress response in S. aureus (227,228). In this model, L- cysteine may play a role in regulating iron homeostasis by limiting the production of SB and ultimately iron uptake to decrease the generation of the more damaging reactive oxygen species from H2O2 via the Fenton reaction (Figure 5-15) (229). In S. aureus, sulfur homeostasis is controlled by the repressor CymR (230). Repression is mediated by CymR when in complex with CysK and is alleviated upon OAS binding to CysK (230). In addition to sulfur metabolism,

CymR also plays a role in the oxidative stress response by indirectly regulating genes such as

130

copA, ahpC, soda, ftnA and dps (231). More recently, CymR was also identified to act as an oxidation-sensing regulator where its sole cysteine thiol is oxidized to sulfenic acid upon exposure to H2O2, resulting in the loss of DNA binding (232). The consequence of losing CymR regulation results in an ~68-fold increase in the intracellular L-cysteine concentration in a S. aureus cymR mutant compared to the parental strain (231). Therefore, under oxidative stress, the increased production of L-cysteine concentration, via CymR regulation, would create a more reducing intracellular environment (Figure 5-15). L-cysteine itself is incorporated into thiol redox enzymes (e.g. thioredoxin reductase and thioredoxin) and is a key constituent of bacillithiol, a low-molecular weight thiol that is crucial in fomycin resistance (233). The increased concentration of L-cysteine in S. aureus would limit the production of L-Dap by inhibiting SbnA activity, thus reducing iron uptake by the SB pathway and limiting the production of damaging reactive oxygen species (Figure 5-15).

131

Chapter 6: The siderophore reductase IruO uses an FAD semiquinone

intermediate to catalyze iron reduction

6.1 Introduction

Once internalized, the iron-siderophore complex must be decoupled to permit iron assimilation. One prominent iron release mechanism is the single electron reduction of Fe(III)- siderophore complexes (36). Reduction of Fe(III) to Fe(II) significantly decreases the affinity siderophores have for the iron ion (36). Four siderophore reductases have been characterized and are specific for each type of Fe(III)-siderophore complex. E. coli FhuF and Bacillus halodurans

FchR reduce hydroxamate-type siderophores (126,127), while E. coli YqjH and V. cholerae

ViuB reduce iron from catecholate-type siderophores (128,129). Recently, two putative siderophore reductases have been described in S. aureus. Bacterial mutagenesis studies revealed that IruO and NtrA participate in the utilization of Fe(III)-DFB and Fe(III)-SA, respectively

(123).

In this study, I examined how IruO (NWMN_2274, SAUSA300_2319, or SACOL2369) participates in iron-siderophore reduction and proposed a mechanism for electron transfer. I demonstrated that IruO specifically reduced iron bound hydroxamate-type siderophores, Fe(III)-

DFB and Fe(III)-FCA to release Fe(II) using NADPH as the electron donor. I also solved the X- ray crystal structure of IruO in two distinct conformational states. Finally, I identified that, under anaerobic conditions, IruO formed a stable FAD semiquinone intermediate that mediated a single electron transfer from the FAD to the iron bound siderophore.

132

6.2 Results

6.2.1 IruO binds and reduces hydroxamate-type siderophores

To assess the activity of IruO against hydroxamate-type siderophores, IruO was incubated with 100 M Fe(III)-DFB and 1 mM ferrozine. Upon the addition of 100 M NADPH to 1 M IruO, an increase in the absorption at 562 nm (Fe(II)-ferrozine complex) was observed over 10 minutes, indicating the generation and release of Fe(II) (Figure 6-1A). Omission of either IruO or NADPH prevented the production of Fe(II) (Figures 6-2A and 6-2B). A similar result was observed with 100 M Fe(III)-FCA, a structurally distinct hydroxamate siderophore

(Figures 6-1B, 6-2C, and 6-2D). Unlike Fe(III)-DFB and Fe(III)-FCA, Fe(III)-EDDHA, a synthetic iron chelator that cannot be used as an iron source by S. aureus, could not be reduced by IruO (Figures 6-1C, 6-2E and 6-2F).

133

Figure 6-1 IruO selectively reduces Fe(III)-hydroxamate-type siderophores. IruO catalyzes the reduction and release of Fe(II) from (A) 100 M Fe(III)-DFB, (B) 100 M Fe(III)-FCA and (C) 100 M Fe(III)-EDDHA in 50 mM sodium phosphate buffer (pH 7.4). An increase in absorbance at 562 nm indicates formation of a Fe(II)-ferrozine complex. The spectra were recorded every 30 seconds for 10 minutes. The red and blue lines indicate the initial and final spectra.

134

Figure 6-2 IruO selectively reduces Fe(III)-hydroxamate-type siderophores. Negative controls for the ferrozine assay in Figure 6-1. Fe(III)-DFB reaction mixture with no IruO (A); Fe(III)-DFB reaction mixture with no NADPH (B); Fe(III)-FCA reaction mixture with no IruO (C); Fe(III)-FCA reaction mixture with no NADPH (D); Fe(III)-EDDHA reaction mixture with no IruO (E); Fe(III)-EDDHA reaction mixture with no NADPH (F). The spectra were recorded every 30 seconds for 10 minutes. The red and blue lines indicate the initial and final spectra

6.2.2 Structure of IruO

The crystal structure of TCEP reduced IruO (rdIruO) was solved to 1.82 Å resolution with a single subunit in the asymmetric unit. The IruO subunit is composed of two distinct domains separated by a short, two stranded antiparallel -sheet (Figure 6-3A). Based on a Pfam 135

search using the amino acid sequence of IruO, IruO is classified as a pyridine nucleotide- disulphide oxidoreductase, containing a smaller Rossmann-fold domain that is presumed to bind

NAD(P)H (residues 121 - 244) and a second, larger pyridine nucleotide-disulfide oxidoreductase

(PNDO) domain that binds flavin adenine dinucleotide (FAD). The latter is assembled from the

N-terminal and C-terminal ends of the polypeptide chain (residues 1 - 120, and 245 – 316) (234).

Projecting from the PNDO domain was a 28 residue C-terminal -helix. The PNDO domain also contains a dimerization interface where the rdIruO dimer could be reconstructed by applying crystallographic symmetry (Figure 6-3B). The buried surface at this interface is ~2573 Å2 as quantified using the program PISA (202). To show that dimerization occurred in solution the mass of rdIruO was measured by SEC-MALS to be 77 kDa (Figure 6-4A), similar to the predicted dimeric mass of 76.4 kDa.

136

Figure 6-3 Structure of rdIruO. (A) Overall fold of rdIruO with the backbone shown as a cartoon diagram. rdIruO is composed of two domains (NADPH, PNDO), a hinge region and a C-terminal -helix. (B) The rdIruO dimer reconstructed from crystallographic symmetry. Individual rdIruO subunits are colored gold and teal. (C) The FAD binding site in the PNDO domain. Selected residues interacting with the cofactor are shown as sticks and water molecules as small blue spheres. The oxygen and nitrogen atoms are colored red and blue, respectively. The dashed lines represent hydrogen bonds. (D) The rdIruO dimer is shown as a surface representation and the glycerol (GOL) and FAD molecules are shown as sticks with carbon atoms colored magenta and blue, respectively. Regions of the surface contributed by the NADPH and PNDO domains are labelled as well as residues of the hinge region and C-terminal -helix. 137

Figure 6-4 IruO forms a dimer in solution. The oligomerization state was determined by SEC-MALS. Molar mass was determined for rdIruO (A) and oxIruO (B). The Rayleigh ratio is shown as a black line and the calculated molar mass is shown as a blue line.

Buried within the PNDO domain is a single FAD modeled at full occupancy (Figure 6-

3C). The FAD is oriented such that the re-face of the isoalloxazine ring was exposed to the solvent and the adenosine and ribitol moieties formed multiple hydrogen bonds with water molecules and residues of the PNDO domain (Figure 6-3C). The isoalloxazine ring is stabilized within the PNDO domain by Tyr43 through -stacking at the si-face and Val324 through an intermolecular hydrophobic interaction at the re-face (Figure 6-3C). Val324 is part of the C-

138

terminal tail and interacted with the FAD isoalloxazine ring predominantly bound to the other subunit.

A shallow pocket is located next to the re-face of the isoalloxazine ring and is constructed from residues located in the PNDO domain, NADPH domain, hinge region and C- terminal tail from the adjacent subunit (Figures 6-3D and 6-5A). The pocket is primarily composed of polar and positively charged residues such as Lys39, Thr159, Arg247, His294,

Tyr323 and His327 (Figure 6-5A). In the rdIruO structure a glycerol molecule is bound in the pocket adjacent to the FAD cofactor. The proximity of the pocket to the cofactor and the polar nature of the residues lining the pocket suggest this is a putative siderophore binding site. The

NADPH binding site was located exclusively in the NADPH domain along a shallow groove as determined based on a structural overlay with a ferredoxin-NADP+ oxidoreductase from Bacillus subtilis (PDBID: 3LZW) (Figures 6-5B and 6-5C). NADPH specificity was hypothesized to be controlled by a positively charged patch within the groove, contributed by residues Arg179 and

Arg205.

139

Figure 6-5 IruO substrate binding sites. (A) Putative siderophore binding site. The two subunits of rdIruO are shown as cartoon diagrams and colored gold and teal. Selected amino acid residues, GOL and FAD are shown as sticks and colored gold/teal, magenta and blue, respectively. Oxygen and nitrogen atoms are colored red and blue, respectively. (B) Putative NADPH binding site within the Rossmann-fold. All residues are shown as sticks. (C) Overlay between the rdIruO (gold) and FNR (magenta) from B. subtilis (PDBID: 3LZW) NADPH binding domain. The NADP+ (white) is derived from FNR and shown in stick form.

140

6.2.3 An intramolecular disulfide bond can form in IruO

The oxIruO crystal structure was solved to 2.3 Å resolution. A structural overlay between rdIruO and oxIruO revealed that the NADPH domain of the oxIruO structure is turned ~90° with respect to the PNDO domain (Figures 6-6A and 6-6B). A disulfide bond is observed between

Cys248 and Cys265, along with disruption of the tertiary structure of the hinge region (Figure 6-

6C). Cys248 and Cys265 are located in the hinge region and in a surface loop of the PNDO domain, respectively. The oxIruO structure did not reveal significant conformational change relative to rdIruO in the PNDO and NADPH domains. However, the unstructured hinge region significantly changed the conformation of the NADPH binding pocket along the base of the groove and significantly narrowed the putative siderophore binding site in the vicinity of the

FAD (Figure 6-8D) compared to rdIruO (Figure 6-3D). Similar to rdIruO, oxIruO was dimeric in solution with a mass of 74 kDa as measured by SEC-MALS (Figure 6-4B).

A multiple sequence alignment revealed that IruO homologs from S. aureus species have equivalent L-cysteine residues (Cys248 and Cys265) (Figure 6-7). To examine if IruO oxidation results the loss of reduced L-cysteines on the surface in solution, IruO was subjected to potassium ferricyanide in molar excess to form oxidized IruO (oxIruO). A DNTB assay detected

~1.7 SH groups per rdIruO subunit compared to ~0.2 SH groups per oxIruO subunit showing that the reduced L-cysteines present on the surface of oxIruO is below the detection limit and suggesting that a disulfide bond has formed.

141

Figure 6-6 OxIruO contains an intramolecular disulfide bond. (A) An overlay between rdIruO (gold) and oxIruO (dark grey). Both forms of IruO are shown as a cartoon diagram and the FAD cofactors are shown as sticks. (B) Top-down view of the NADPH domain of oxIruO overlaid with rdIruO. (C) The intramolecular disulfide bond formed between residues Cys248 and Cys265 in oxIruO (dark grey). An overlay between oxIruO and rdIruO (gold) reveals the conformational changes adopted by Cys248 and Cys265. L-Cysteine residues and FAD molecules are shown as sticks. The FAD colored dark blue and light blue corresponds to rdIruO and oxIruO, respectively. The dashed lines highlight the distance between the C atoms of Cys248 and Cys265 in rdIruO and oxIruO. (D) The molecular surface around the FAD cofactor in oxIruO. The oxIruO dimer is shown as a surface representation. OxIruO

142

subunits are colored grey and light blue and shown as a surface representation. The FAD is colored turquoise and shown as sticks. Oxygen, nitrogen and sulfur atoms are colored red, blue and yellow, respectively. Regions of the surface contributed by the NADPH and PNDO domains are labelled as well as residues of the hinge region and C-terminal -helix.

Figure 6-7 Multiple sequence alignment of IruO homologs. Conserved cysteine residues in S. aureus species are highlighted in red. Sequences are identified by a single letter for genus, the first two letters of the species name and the annotated protein name. The numbering along the top corresponds to the sequence position of S. aureus strain Newman IruO. S. au, Staphylococcus aureus; S. ps, Staphylococcus pseudintermedius; S. lu, Staphylococcus lugdenensis; S. ca, Staphylococcus carnosus; S. sa, Staphylococcus saprophyticus; S. ha, Staphylococcus haemolyticus; S. wa, Staphylococcus warneri; B. an, Bacillus anthracis; L. mo, Listeria monocytogenes; B. su, Bacillus subtilis; C. te, Chlorobaculum tepidum; T. th, Thermus thermophiles.

6.2.4 IruO activity is diminished by the formation of an intramolecular disulfide bond

The ferrozine assay was used to determine steady-state kinetic parameters with either rdIruO or oxIruO against varying concentrations of Fe(III)-DFB, Fe(III)-FCA and NADPH

(Table 6-1 and Figures 6-8 and 6-9). The oxidation of IruO decreased enzyme activity with the specificity constant (kcat/Km) for both Fe(III)-DFB and Fe(III)-FCA decreasing ~5-fold.

However, the Km was negligibly impacted by the disulfide bond formation, except for oxIruO

143

with Fe(III)-FCA, which was 5-fold higher. A distinguishable difference in the kcat/Km (~2-fold) was evident between Fe(III)-DFB and Fe(III)-FCA, which showed that structural differences between the siderophores do impact iron reduction. For NADPH, the kcat/Km was comparable, suggesting that the oxidation of IruO does not impact NADPH utilization. IsdI-heme degradation reactions with rdIruO and NADPH (Figure 6-10A) were faster than those with oxIruO and

NADPH (Figure 6-10B) as judged by the decreasing height of the Soret peak at 412 nm. The second peak at 240 nm corresponds to the presence of NADPH in the assay. Heme degradation rates observed under these conditions were calculated to be 0.39 ± 0.02 min-1 and 0.13 ± 0.005 min-1 for rdIruO and oxIruO, respectively. Overall, the introduction of an intramolecular disulfide bond decreased IruO activity.

Figure 6-8 Steady-state kinetics of IruO with Fe(III)-DFB and Fe(III)-FCA. Reactions were performed with 0.1 or 0.02 M rdIruO (A, C) or 0.3M oxIruO (B, D). The reactions were in 500 M NADPH, 50 mM sodium phosphate buffer (pH 7.4) and either 1 – 250 M Fe(III)-DFB or 1 – 250 M Fe(III)-FCA as indicated. The error bars represent the standard error from three independent experiments and Michaelis-Menten kinetic parameters were fit to the data using non-linear regression. 144

Figure 6-9 Steady-state kinetics of IruO with NADPH. Reactions were performed with 0.1 M rdIruO and 200 M Fe(III)-DFB (A). Reactions performed with 0.3M oxIruO and 200 M Fe(III)-DFB (B). Reactions were performed with 0.02 M rdIruO and 200 M Fe(III)-FCA (C). Reactions performed with 0.3 M oxIruO 200 M Fe(III)-FCA (D). The buffer contained 50 mM sodium phosphate (pH 7.4) and NADPH as indicated. The error bars represent the standard error from three independent experiments and Michaelis-Menten kinetic parameters were fit to the data using non-linear regression.

Table 6-1 Steady-state kinetic constants for rdIruO and oxIruO. Fe(III)-DFB and Fe(III)-FCA substrates with NADPH Kinetic Fe(III)-Desferrioxamine B Fe(III)-Ferrichrome A Parameter rdIruO oxIruO rdIruO oxIruO) Km (μM) 25 ± 3 28 ± 4 22 ± 4 108 ± 14 -1 kcat (s ) 0.082 ± 0.003 0.018 ± 0.007 0.15 ± 0.006 0.084 ± 0.005 -1 -1 kcat/Km (M s ) 3300 660 6700 780 NADPH substrate with Fe(III)-DFB or Fe(III)-FCA Kinetic Fe(III)-Desferrioxamine B Fe(III)-Ferrichrome A Parameter rdIruO oxIruO rdIruO oxIruO Km (μM) 44 ± 8 12 ± 1 66 ± 25 61 ± 10 -1 kcat (s ) 0.045 ± 0.005 0.011 ± 0.001 0.11 ± 0.02 0.13 ± 0.01 -1 -1 kcat/Km (M s ) 1000 950 1700 2100

145

Figure 6-10 Oxidation of IruO lowers the rate of heme degradation by IsdI. IsdI-catalyzed heme degradation with rdIruO (A) and oxIruO (B). The spectra were recorded every two minutes over twenty minutes with spectra taken every two minutes. Red and blue lines indicate the initial and final spectra. Panels are representative of three independent experiments.

6.2.5 IruO has high affinity for hydroxamate-type siderophores

To investigate the binding affinity of Fe(III)-DFB and Fe(III)-FCA to IruO, tryptophan fluorescence quenching assays were performed. Fluorescence quenching was recorded at various concentrations of Fe(III)-DFB and Fe(III)-FCA (0.5 – 20 M) with 1 M rdIruO or 0.5 M oxIruO in 50 mM sodium phosphate buffer, pH 7.4. The measured dissociation constant (Kd) of

IruO with either siderophore was in the low M range (1-2 M) (Figure 6-11), but could not be accurately determined because the concentration of IruO required to observe fluorescence 146

change was not at least 5-fold lower than the calculated Kd. Nonetheless, the results suggest that

IruO can bind two structurally distinct hydroxamate-type siderophores with high affinity.

Figure 6-11 IruO binds hydroxamate-type siderophores with high affinity. Tryptophan fluorescence quenching was used to measure the binding affinities of IruO for two hydroxamate-type siderophores. Reaction mixtures were excited at 292 nm and the emission spectra at 350 nm was recorded. The black and green lines represent Fe(III)-FCA binding to oxIruO and rdIruO, respectively. The blue and red lines represent Fe(III)-DFB binding to oxIruO and rdIruO, respectively.

6.2.6 IruO mediates electron transfer from NADPH to Fe(III)-siderophore complexes via an FAD neutral semiquinone intermediate

IruO was transferred to an anaerobic environment in 50 mM sodium phosphate buffer, pH 7.4 to investigate its electron transfer mechanism. UV-Vis absorption spectroscopy revealed that the FAD cofactor in IruO is in the oxidized state (FADox) as characterized by peaks at 376 and 464 nm (Figure 6-12A), which is similar to spectra observed in other FAD containing enzymes such as NADPH-cytochrome P450 reductase (235). Upon the addition of 20 M

NADPH, the FAD cofactor exhibited a spectrum that corresponded to the neutral semiquinone

147

state (FADsq) defined by a decrease in the peaks at 376 and 464 nm and a subsequent increase in a broad peak between 550-650 nm (Figure 6-12B) (235). To ensure that higher concentrations of

NADPH did not result in the reduction of FAD in IruO to the FAD hydroquinone state (FADhq),

5-fold excess of NADPH was added to IruO and its spectrum was recorded after 5 minutes.

Similar to the initial results, the FAD could only be reduced to the neutral FADsq state and not to the FADhq state (Figure 6-13A). Unlike IruO, the closest homolog to IruO in S. aureus, thioredoxin reductase (NWMN_0732), does not stabilize the FADsq state. Instead, the FAD cofactor is fully reduced to the FADhq state in the presence of excess NADPH (Figure 6-13B).

The FADhq state in IruO could be attained, but in the presence of excess sodium dithionite, a non-physiologically relevant chemical reductant (Figure 6-12C).

Oxidation of NADPH to NADP+ is a two electron process, but the FAD cofactor in IruO only accepts one electron. Therefore, I hypothesized that NADPH transfers the second electron through the interacting PNDO domains to the second FAD cofactor in the adjacent subunit.

Titration of NADPH to a fixed concentration of IruO was performed to observe the formation of

FADsq. Formation of the FADsq was complete when the NADPH concentration was roughly half that of the IruO subunit (Figure 6-12D). Therefore, a single NADPH can reduce two molecules of FADox to FADsq.

Next, I hypothesized that IruO transfers a single electron from the FADsq to a Fe(III)- siderophore complex. Under anaerobic conditions, the UV-Vis absorption spectrum was collected for IruO FADsq in the presence of Fe(III)-DFB and gallium(III)-bound DFB (Ga(III)-

DFB), a redox inert analog. When Fe(III)-DFB was added to IruO FADsq, the peak at 600 nm that corresponded to FADsq decreased (Figure 6-12E). When Ga(III)-DFB was added to IruO

148

FADsq, no spectral changes were observed (Figure 6-12F). Overall, the results show that IruO mediates two single electron transfer reactions via FAD to reduce Fe(III)-DFB.

Figure 6-12 IruO reduces Fe(III)-DFB using NADPH as the electron donor via an FAD semiquinone intermediate. (A) UV-vis absorption spectrum of IruO (30 M) containing oxidized FAD (FADox). (B) UV- Vis absorption spectra of 30 M IruO in the presence of 20 M NADPH. Spectra were recorded every 30 seconds for 5 minutes, showing the transition from the FADox state (red line) to the FADsq state (blue line). (C) UV-Vis absorption spectra of 30 M IruO in the presence of sodium dithionite. Spectra were recorded every 30 seconds for 5 minutes showing the transition from the FADox state (red line) to the FADhq state (blue line). The green line represents the FADsq intermediate state that formed immediately after the addition of sodium dithionite. (D) NADPH titration experiment against 30 M IruO FADox recorded at 600 nm. The intersection between the blue and red lines indicates the concentration of NADPH necessary to fully reduce FADox to FADsq. FADox reduction to FADsq was saturated at 0.5:1 molar ratio of NADPH to IruO subunit. (E) UV-Vis absorption spectra of 30 M IruO, pre-incubated with 20 M NADPH to form the FADsq state, in the presence of 60 M Fe(III)-DFB. Spectra were recorded every 30 seconds for 5 minutes, showing the transition from the FADsq state (red line) to the FADox state (blue line). (F) UV-Vis absorption spectra of 30 M IruO, pre-incubated with 20 M NADPH to form the FADsq state, in the presence of 60 M Ga(III)-DFB. Spectra were recorded every 30 seconds for 5 minutes. Red and blue lines indicate the initial and final spectra. The arrows indicate the direction of spectral changes over time. 149

Figure 6-13 IruO maintains the FADsq intermediate under molar excess of NADPH unlike the putative thioredoxin reductase (NWMN_0732) from S. aureus. IruO (A) and NWMN_0732 (B) (black lines) were incubated with 5-fold molar excess of NADPH for 5 minutes (red lines).

6.3 Discussion

Few siderophore reductases have been identified in bacteria. To date, two types of siderophore reductases have been characterized: FeS cofactor containing reductases and

FAD/FMN cofactor containing reductases. E. coli contains both types of reductases, FhuF (FeS)

(126), and YqjH (FAD) (128), which were shown to reduce hydroxamate-type and catecholate- type siderophores, respectively. Another characterized siderophore reductase is the FeS containing reductase (FchR) that was identified in Bacillus halodurans and is responsible for the reduction of Fe(III)-dicitrate and hydroxamate-type siderophores like Fe(III)-schizokinen (127).

Additionally, several other uncharacterized IruO homologs have been identified in Gram- positive bacteria that include Bacillus subtilis, Bacillus anthracis, and Listeria monocytogenes

(122). Interestingly, the L. monocytogenes gene encoding for the IruO homolog appears to be part of a larger gene cluster predicted to encode for hydroxamate-type siderophore import machinery (236).

150

In addition to IruO, S. aureus contains a second flavin-containing reductase, NtrA, that is involved in iron utilization from Fe(III)-SA (123). I hypothesize that using flavins as electron mediators provide a growth advantage to S. aureus under severe iron limited environments.

Under iron restriction, S. aureus alters its metabolism to conserve iron by inducing the “iron sparing” response, which results in the down-regulation of the tricarboxylic acid cycle and up- regulation of glycolysis (111). Coupled with the recent discovery that S. aureus produces staphyloferrin B without the need of the iron-requiring tricarboxylic acid cycle (82), I propose that S. aureus has adapted its metabolism to utilize flavoproteins for iron assimilation as a strategy to further conserve iron.

IruO is categorized as a PNDO family member. This diverse family contains a variety of enzymes that include thioredoxin reductases and glutathione reductases (237). Although structurally different, IruO is functionally convergent with the ferredoxin-NADP(H) reductase- like (FNR) superfamily that includes the catecholate-type siderophore reductase YqjH from E. coli (128,238). I propose that IruO functions by performing electron transfer reactions from

NADPH to FAD to its Fe(III)-siderophore substrate. However, unlike all known examples of siderophore reductases, the FAD in IruO is only reduced to a stable neutral semiquinone

(FADsq), which is the intermediate used for Fe(III)-DFB and Fe(III)-FCA iron reduction (Figure

6-14). Because NADPH must donate two electrons simultaneously and a single subunit of IruO can only accept one electron, I propose that the second electron is directly transferred through the

PNDO domain’s dimeric interface to the second FAD cofactor (Figure 6-14). To the best of my knowledge, this mechanism of electron transfer has not been proposed for any known flavoprotein.

151

Figure 6-14 Proposed mechanism for Fe(III)-siderophore reduction via IruO. FADH* represents FAD in the semiquinone state.

IruO represents the first solved structure of a Fe(III)-siderophore reductase. Three structural homologs (Z-score higher than 30) were identified using the DaliLite Server (167) and included the ferredoxin-NADP+ oxidoreductases (FNR) from B. subtilis (PDBID: 3LZW) (239), and Chlorobaculum tepidum (PDB ID: 3AB1) (240), as well as a thioredoxin reductase-like protein from Thermus thermophiles (PDBID: 2ZBW). A comparison between IruO and all three homologs revealed a core set of conserved residues responsible for binding the FAD isoalloxazine ring. However, the L-leucine residue (residue 324 in IruO) that interacts with the re-face of the FAD isoalloxazine ring is not conserved (ie. L-tryptophan, L-histidine and L- 152

leucine) (Figure 6-15). IruO is structurally similar to thioredoxin reductases (TrxB; r.m.s.d. ~3.5

Å over 231 C atoms), which are responsible for reducing disulfide bonds in thioredoxin (237).

In S. aureus strain Newman, the closest homolog to IruO is TrxB (NWMN_0732) with a sequence identity of 24%. A comparison of the TrxB structure from S. aureus subsp. aureus

Mu50 (PDBID: 4GCM) with rdIruO revealed a conserved overall fold; however, residues that interact with the FAD isoalloxazine ring are different (Figure 6-16). The isoalloxazine ring in

IruO is partially exposed to the solvent whereas the isoalloxazine ring in TrxB is blocked at the re-face by a redox active disulfide bond that is essential for electron transfer (237).

153

Figure 6-15 FAD isoalloxazine ring binding in IruO homologs. (A) FNR from B. subtilis (PDBID: 3LZW). (B) FNR from C. tepidum (PDB ID: 3AB1). (C) Thioredoxin reductase-like protein from T. thermophiles (PDBID: 2ZBW). Selected residues interacting with the FAD isoalloxazine ring and FAD are shown as sticks. The dashed lines represent hydrogen bonds. 154

Figure 6-16 IruO FAD binding and active site formation is divergent from TrxB. (A) RdIruO (B) TrxB from S. aureus (PDBID: 4GCM). The FAD from IruO and TrxB are colored blue and white, respectively. Selected residues interacting with the FAD isoalloxazine ring and FAD are shown as sticks. The dashed lines represent hydrogen bonds.

The crystal structure revealed that upon exposure to an oxidizing agent, IruO formed an intramolecular disulfide bond. In oxIruO, significant structural changes were observed within both the NADPH binding site and the immediate vicinity surrounding the FAD cofactor that narrowed the channel leading to the isoalloxazine ring. In vitro assays revealed that oxIruO activity is lower than rdIruO for both siderophore iron reduction and heme degradation, leading to my hypothesis that the redox sensitive L-cysteines regulate IruO activity under oxidative stress. Previous microarray studies have shown that oxidative and nitrosative stresses result in the dysregulation of iron-homeostasis in S. aureus leading to the up-regulation of most iron uptake systems (18,241). An excess of free iron could increase the generation of reactive oxygen species that are responsible for biomolecule damage. It is possible that a redox sensitive disulfide bond was introduced into IruO to regulate the amount of iron being released under oxidative

155

stress conditions. The siderophore reductase FhuF in E. coli is repressed by the oxidative response regulator, OxyR (242). Future studies need to be performed to determine if IruO activity can be altered by oxidative stress effectors like H2O2.

156

Chapter 7: Overview and future directions

S. aureus undergoes a major metabolic shift when iron availability is low (111). Low iron stress results in S. aureus up-regulating its iron uptake systems to extract trace iron from its immediate environment. Additionally, many iron-requiring metabolic pathways are down- regulated in an effort to conserve iron for essential functions, a phenomenon known as the iron sparing response (111). The work presented in this thesis was aimed at determining how siderophore-mediated iron metabolism is integrated into the central metabolism of S. aureus under the influence of the iron sparing response caused by low iron stress.

7.1 SB biosynthesis does not require a functioning TCA cycle

SB is assembled from three precursor substrates: citrate, -KG and L-Dap (78). All three precursors must be available for SB biosynthesis under conditions of low iron stress. Through an unknown mechanism associated with low iron availability, the expression of TCA cycle enzymes is down-regulated, while the expression of glycolytic enzymes is elevated (111). Considering both citrate and -KG are primarily generated through the TCA cycle, a paradox arose as to how both SB precursors are produced to sustain SB biosynthesis. This paradox was partially explained by the recent discovery that SbnG functions as a citrate synthase, producing citrate from acetyl-CoA and oxaloacetate (81). Furthermore, the functions of SbnA and SbnB were discovered to be essential for L-Dap production, but their substrates and detailed mechanisms were not identified (80). Therefore, how -KG is produced and identifying L-Dap synthesis pathway under conditions of iron deprivation remained to be determined.

157

In Chapter 4, the L-Dap biosynthetic pathway was elucidated, revealing that SbnA and

SbnB produce L-Dap and -KG from substrates OPS and L-glutamate. The unexpected addition of -KG as a product of the L-Dap biosynthetic pathway now demonstrates how all three SB precursors are generated under low iron stress conditions. In total, the precursors required for SB assembly can now be traced back to four metabolites produced or obtained from the environment by S. aureus: acetyl-CoA, oxaloacetate, OPS and L-glutamate (Figure 7-1). Based on the metabolic pathways present in S. aureus as identified in the KEGG database (222,223), both acetyl-CoA and oxaloacetate can be generated directly from glycolysis via pyruvate dehydrogenase (pdhABC)/pyruvate formate lyase (pflAB) and pyruvate carboxylase (pyc), respectively. L-glutamate can be generated by glutamate synthase (gltBC). Lastly, OPS is an intermediate in the L-serine biosynthetic pathway, which is derived from the glycolytic intermediate 3-phosphoglycerate. First, 3-phosphoglycerate is converted to 3- phosphohydroxypyruvate by 3-phosphoglycerate dehydrogenase (serA) and then converted to

OPS by 3-phosphohydroxypyruvate transaminase (serC).

158

Figure 7-1 Proposed flow diagram of staphyloferrin B biosynthesis integrated into central metabolism through the CodY regulon under conditions of high glucose and low iron. SerAC, Pyc and GltBD enzymes (purple circles) are involved in OPS, oxaloacetate and L- glutamate biosynthesis, respectively and are under the control of the CodY regulon. SbnA, SbnB and SbnG (blue circles) are involved in SB precursor biosynthesis and under the control of Fur. Arrows in green are up-regulated metabolic pathways and red arrows are down-regulated metabolic pathways under iron limitation.

While the four SB precursors are theoretically available based on genome mining for metabolic pathways, each SB precursor biosynthetic pathway must be activated under conditions of low iron stress, a condition where SB biosynthesis is up-regulated. Based on multiple transcriptomic datasets for S. aureus grown in human-like media (e.g. human serum, human

159

whole blood, synthetic nasal secretions and mouse lungs), gltBC, serAC and pyc gene expression is up-regulated to increase the production of L-glutamate, OPS and oxaloacetate, respectively.

All three metabolic pathways are part of the CodY regulon. Therefore, I predict that SB biosynthesis, under the regulation of Fur, synergizes with the CodY regulon to ensure SB precursor availability, which feed directly into SB synthesis. A proteomics study revealed that both Fur and iron positively affect CodY protein expression through an uncharacterized activation mechanism, further suggesting a regulatory link between iron homeostasis and amino acid biosynthesis (111). It appears that S. aureus has synergistically linked iron and amino acid nutrient homeostasis together to ensure survival within the human host under nutrient deprived conditions.

Therefore, based on the identified metabolic pathways described above, a functioning

TCA cycle is not required to produce SB in S. aureus as all metabolites can be generated via glycolysis or by the conversion of L-glutamine to L-glutamate. Indeed, Sheldon, J.R. et al. 2014 showed that a S. aureus mutant lacking the TCA cycle citrate synthase (citZ), which effectively shuts down the TCA cycle, was still able to produce SB (82). Interestingly, tracing metabolic pathways back even further from acetyl-CoA, oxaloacetate, OPS and L-glutamate reduces the number of SB precursors to two metabolites: glucose and L-glutamine (Figure 7-1). Remarkably, glucose and L-glutamine are two of the four most abundant metabolites available in human serum (113), indicating that S. aureus has adapted its metabolism to use readily available host nutrients. Overall, S. aureus appears to have devised a strategy to survive low iron stress by regulating metabolic pathways to ensure SB production, but has also adapted its nutritional requirements to reflect the available nutrients in the human host.

160

7.2 Metabolic pathways and regulatory networks differentiate SA and SB biosynthesis in

S. aureus

S. aureus has the capability to produce two siderophores, SA and SB, under low iron stress. However, most other staphylococcal species, including the coagulase-negative

Staphylococci, can only produce SA (47), which raises the question as to the benefit of S. aureus producing a second siderophore. The leading hypothesis is that SB production is a virulence factor for S. aureus that contributes to pathogenesis.

While both SA and SB are categorized as -hydroxycarboxylate-type siderophores, their chemical structures are different. SA is composed of a single L-ornithine flanked by two citrate derived moieties, while SB is assembled from citrate, -KG and two molecules of L-Dap. Based on the differences in SA and SB precursors, each siderophore biosynthetic pathway would be uniquely integrated into S. aureus central metabolism. The summation of work detailed and discussed in Chapters 3, 4, 5 and Section 7.1 reveal how SB precursors (acetyl-CoA, oxaloacetate, OPS and L-glutamate) could integrate into central metabolism via CodY regulated amino acid biosynthetic pathways (Figure 7-2). However, SA biosynthesis integration into central metabolism has not been investigated compared to SB. One study has shown that the

TCA cycle citrate synthase is required to supply citrate for SA production (82). The source of L- ornithine has not been identified, but based on KEGG pathways, L-ornithine could be derived from L-arginine or L-glutamate (222,223).

161

Figure 7-2 Integrating SA and SB biosynthesis into central metabolism in S. aureus. Proposed metabolic flow diagram of staphyloferrin A and staphyloferrin B biosynthesis integrated into central metabolism through the CodY regulon (blue circles) and CcpA regulon (orange circles) under conditions of high glucose and low iron. Arrows in green are up-regulated metabolic pathways and red arrows are down-regulated metabolic pathways. Under conditions of high glucose, CcpA repress both the TCA cycle and genes encoding enzymes in L-arginine and L-proline biosynthesis (orange circles) to reduce production of citrate and L-ornithine. Conversely, CodY repression is alleviated under amino acid deprivation to express enzymes (blue circles) for the production of OPS, L-glutamate and oxaloacetate.

Staphyloferrin production can be differentiated under specific growth conditions. When

S. aureus is grown in media containing high concentrations of glucose with limited iron, only SB is produced (82). Comparatively, both SA and SB are produced when S. aureus is grown in media containing high concentrations of succinate with limited iron (58). In S. aureus, SA abrogation is attributed to catabolite repression elicited by CcpA when glucose, the preferred carbon source, is available. Under glucose rich conditions, the carbon overflow pathway is 162

activated by CcpA, which favors glycolysis and represses TCA cycle activity (Figure 7-2) (96).

CcpA binds upstream to cre-sites in many TCA cycle genes including citrate synthase (citZ), isocitrate dehydrogenase (citC), succinate dehydrogenase (sdhB), succinyl-CoA synthetase

(sucCD) and -ketoglutarate dehydrogenase (odhAB). Therefore, glucose is preferentially oxidized to acetate instead of entering the TCA cycle to generate citrate. Loss of citrate production derived from the TCA cycle prevents SA production in S. aureus (82). Additionally,

CcpA regulates L-arginine and L-proline biosynthetic pathways, which utilize L-ornithine as an intermediate (97,98). Under catabolite repression by CcpA, ornithine aminotransferase (rocD) and arginase (rocF) are repressed, resulting in L-arginine and L-proline auxotrophy (Figure 7-2)

(97,98), which would abrogate the production of L-ornithine. Once S. aureus has shifted to nonpreferred carbon sources upon glucose depletion, CcpA repression is relieved for the TCA cycle and L-arginine and L-proline biosynthetic pathways (96-98), theoretically increasing the availability of L-ornithine for SA biosynthesis.

In the human body, many of the S. aureus environmental niches (blood, serum, nasal secretions and lungs) have glucose readily available. Therefore, SB production is expected to initially dominate iron uptake under such conditions. While glucose is abundant in the host, it is possible that prolonged S. aureus infections, such as in soft tissue abscesses, or persistence in macrophages after phagocytosis could lead to conditions of glucose depletion. S. aureus can respond by switching siderophores to maintain seamless siderophore-mediated iron uptake by producing a siderophore derived from different metabolites that do not depend on glucose availability. The ability of S. aureus to temporally regulate staphyloferrin production as part of its pathogenicity remains to be characterized.

163

7.3 A mechanism for Fe(III)-siderophore iron release in S. aureus

Once Fe(III)-siderophores complexes are internalized, the iron must be released and integrated into the cell. The most common strategy is through the reduction of Fe(III) to Fe(II)

(36). In bacteria, relatively few Fe(III)-siderophore specific reductases have been identified.

Recently, two putative reductases have been identified to contribute to siderophore-mediate iron metabolism in S. aureus, IruO and NtrA (123). IruO contributed to S. aureus growth on hydroxamate-type siderophores as the sole iron source and NtrA with SA as the sole iron source

(123). Work presented in Chapter 6 provided insight into the mechanism of hydroxamate-type siderophore iron reduction by IruO. In IruO, electron transfer from NADPH to Fe(III)-DFB or

Fe(III)-FCA is mediated by the FAD cofactor. Unlike the YqjH FAD cofactor, which accepts two electrons reducing FADox to FADhq, the FAD in IruO is reduced by a single electron to

FADsq. IruO utilizes a unique electron transfer mechanism compared to all other characterized siderophore reductases.

Although NtrA has been identified to play a role in Fe(III)-SA utilization using gene deletions and growth experiments, the function of NtrA in Fe(III)-siderophore iron reduction has not been biochemically validated. The NCBI Conserved Domains database revealed that NtrA is part of the FMN-dependent nitroreductase-like superfamily that uses NAD(P)H as an electron donor (243). NtrA is also predicted to be a homodimer based on sequence identity (31%) to the structure of nitroreductase-like family protein from Bacillus cereus (PDBID: 2I7H) based on a sequence search in the Protein Data Bank (244). It remains to be seen if the putative FMN cofactor could undergo a single (FMNsq) like IruO, or double (FMNhq) electron reduction like

YqiH to facilitate electron transfer to Fe(III)-SA. Future studies remain to be performed to biochemically characterize NtrA as a SA reductase in S. aureus. To date, no other siderophore 164

reductase has been identified for SB or catecholate-type siderophores in S. aureus, but based on the identification of IruO and NtrA, if both SB and catecholate-type siderophore reductases do exist in S. aureus, they are most likely Fur regulated and expressed under low iron stress conditions.

7.4 Regulatory links between iron uptake and the oxidative stress response in S. aureus

To combat S. aureus infections, the host’s immune system generates reactive oxygen

- species (ROS) such as O2 and H2O2 that kill the invading bacterium via oxidative damage. In bacteria, ROS can inactivate iron-containing proteins like aconitase leading to the release of iron

. from FeS, which can then react with H2O2 to produce HO (245). ROS can lead to further damage of oxidation sensitive proteins and DNA (246). In response, S. aureus has developed an intricate oxidative stress response to neutralize ROS and prevent cell death. For example, S. aureus produces catalase to break down H2O2 to water and oxygen, superoxide dismustase to

- convert O2 to O2 or H2O2 and staphyloxanthin, a carotenoid pigment that quenches ROS

(21,247,248). Oxidative stress resistance genes are primarily under the control of the Peroxide responsive regulator (PerR), which senses manganese (Mn(II)) or Fe(II) intracellular concentrations and co-represses gene transcription (227,228).

Furthermore, iron homeostasis, under the regulatory control of Fur, and the oxidative stress response appear to be co-regulated in S. aureus. Fe(II) is highly reactive to ROS and through Fenton chemistry results in the production of hydroxyl radicals. Increasing iron concentration in S. aureus enhances their susceptibility to killing by H2O2 (249). Studies have shown that PerR controls the expression of the iron storage proteins ferritin and Dps, which alleviate oxidative stress by sequestering free intracellular iron (228) and that catalase expression 165

is positively regulated by Fur through some unidentified mechanism (227). Very recently, PerR from S. aureus was shown to sense H2O2 in the Fe(II)-bound form by iron-mediated histidine oxidation (250).

Under conditions of oxidative stress, S. aureus sequesters free iron to prevent Fenton chemistry while attempting to neutralize ROS. Yet, ROS like H2O2 dysregulates iron homeostasis by causing the de-repression of the Fur regulon (241,251). Therefore I hypothesize that other mechanisms must exist to limit iron uptake to prevent Fenton chemistry and killing by

ROS.

In Chapter 5, L-cysteine was shown to competitively inhibit SbnA activity at predicted physiological concentrations (~90 M) by forming a non-functional complex with the PLP cofactor. The amino acid L-cysteine itself plays a significant role in the oxidative stress response by playing a role in protein redox signaling and repair (252) and incorporation into antioxidants thiols like bacillithiol, a low-molecular weight thiol analogous to glutathione (253,254). L- cysteine biosynthesis is regulated by the sulfur metabolism regulator, CymR (230), which also senses oxidative stress through the oxidation of a single cysteine thiol (232). Loss of CymR repression by thiol oxidation increases the L-cysteine intracellular concentration, which is thought increase the buffering capacity of S. aureus to help defend against ROS (231). In E. coli, high intracellular concentrations of L-cysteine in the presence of iron increased killing by H2O2 through Fenton chemistry (229). Therefore, I hypothesize that higher concentrations of L- cysteine post-translationally regulates SbnA by inhibiting its activity to prevent SB production and the subsequent iron uptake into the cell (Figure 7-3).

166

Figure 7-3 Regulatory links between iron uptake and the oxidative stress response in S. aureus. Proposed metabolic flow diagram of staphyloferrin B biosynthesis integrated into primary metabolism of S. aureus under high glucose and low iron. Arrows in green are up-regulated metabolic pathways and red arrows are down-regulated metabolic pathways. Increased intracellular concentrations of L-cysteine due to redox sensing of the sulfur metabolic regulatory, CymR, would block L-Dap biosynthesis by inhibiting SbnA and thereby prevent SB production. Iron release from hydroxamate-type siderophores would decrease as a result of the IruO redox sensitive L-cysteine residues forming an intramolecular disulfide bond when oxidized, possibly by H2O2. Limiting iron uptake decreases the intracellular pool of Fe(II), which would prevent Fenton reactions and the creation of damaging free radicals.

In Chapter 6, IruO was shown to adopt two distinct conformational states based on the formation of an intramolecular disulfide bond by the oxidant potassium ferricyanide. The oxidized form of IruO had ~5-8 fold lower catalytic efficiency reducing Fe(III)-DFB and Fe(III)-

FCA complexes compared to the TCEP reduced form of IruO. Structurally, the reduced IruO displayed a large active site pocket adjacent to the isoalloxazine ring of FAD, unlike oxidized

IruO where the channel leading to the active site is narrowed and appears to sterically block

167

Fe(III)-siderophore complex binding. In the context of oxidative stress by ROS, limiting Fe(III)- siderophore reduction would prevent iron reduction, release from the siderophore and availability for Fenton chemistry to produce more ROS. Instead, Fe(III) would remain bound to its siderophore, resisting Fenton reactions and preventing further oxidative damage to S. aureus.

Therefore, IruO appears to include redox-sensitive L-cysteine residues in an effort to respond to oxidative stress and prevent unwanted Fenton chemistry (Figure 7-3). Another example of a siderophore reductase linked to the oxidative stress response is found in the E. coli Fe(III)- hydroxamate siderophore reductase, FhuF, where fhuF transcription is repressed by the oxidative response regulator, OxyR, in addition to its regulation by Fur (242).

7.5 Future work

Given the wealth of knowledge known about siderophore-mediated iron uptake, much of that research has primarily focused on siderophore biosynthesis and Fe(III)-siderophore uptake.

Underrepresented is our understanding of how siderophore biosynthesis is integrated into the central metabolism of organisms. Furthermore, relatively little is known about the iron release mechanism used on most Fe(III)-siderophore complexes. Much of our knowledge on siderophore-mediated metabolism has been derived from the archetype E. coli siderophore enterobactin (255). However, siderophores as a whole are structurally diverse, and are assembled from a variety of metabolites and can coordinate iron using different functional groups (36). For a great review on the structural diversity of siderophores, see Hider, R.C. and Kong, X., 2010

(36). Even the strategies used to release iron from Fe(III)-siderophore complexes would have to be numerous to accommodate the sheer diversity of functional groups and unique structural elements each siderophore possesses. The work presented in this thesis provides insight into the 168

siderophore-mediated iron metabolism in S. aureus, providing a link between central metabolism and SB biosynthesis and revealing new mechanisms of iron release from Fe(III)-siderophore complexes. However, this work has also generated several new hypotheses that require further investigation.

In this thesis, several metabolic pathways have been proposed to link SB biosynthesis to central metabolism through the control of metabolic regulators CcpA and CodY. Further investigation is necessary to validate these proposed linkages. Applying metabolomics to determine the relative abundancies of central metabolism metabolites in S. aureus under varying growth conditions like low iron stress would provide confirmation that SB precursors are available to feed into the SB biosynthetic pathway. Furthermore, stable isotope tracing can be utilized to validate specific metabolic pathway connections such as the incorporation of stable isotope labeled L-glutamine to validate L-glutamate utilization or glucose under specific conditions where isotopically labeled citrate is not generated to validate OPS utilization (S. aureus citZ sbnG mutant). Additionally, the proposed regulatory synergism between SB biosynthesis and CcpA and CodY can be investigated by measuring the production of SB from mutant S. aureus lacking CcpA (S. aureus ccpA) and/or CodY (S. aureus codY). Similar studies can also be performed for SA biosynthesis using isotopically labeled L-arginine and/or L- proline.

In Chapters 5 and 6, evidence for new links between iron uptake and the oxidative stress response in S. aureus were presented. For SbnA inhibition with L-cysteine, in vivo experimentation could be performed to confirm that an increased concentration of L-cysteine decreases SB production. L-cysteine can be added exogenously to S. aureus cultures grown under low iron conditions. As a complementary experiment, since a cymR knockout has been 169

shown to increase the relative intracellular concentration of L-cysteine (231), S. aureus cymR mutants can be investigated to demonstrate how the dysregulation of sulfur metabolism affects

SB production. In the case of IruO, exposure to a mild oxidant caused the formation of an intramolecular disulfide bond that decreased activity. However, potassium ferricyanide is not a biologically relevant oxidizing reagent. Therefore, IruO disulfide bond formation should be examined with other oxidizing reagents such as H2O2 or nitric oxide. Furthermore, site-directed mutagenesis could be utilized to generate an oxidation-resistant IruO variant by mutating the redox sensitive L-cysteine residues and comparing its activity to wild type S. aureus under various concentrations of different oxidizing compounds. In vivo experiments should also be performed to validate the physiological response of IruO when exposed to compounds like H2O2.

Since IruO is essential for Fe(III)-DFB utilization (123), wild type S. aureus versus S. aureus

IruO oxidation-resistant mutants can be examined by feeding Fe(III)-desferrioxamine B as the sole iron source and subjecting the bacteria to oxidative stress, followed by measuring growth rates and the relative production of free radicals.

Finally, the IruO structures provided clues on how the substrates NADPH and hydroxamate-type siderophores such as Fe(III)-DFB would bind. Additional crystallographic studies should be performed on IruO to determine a structure with ligands bound. Since NADPH does not turnover in the presence of IruO under a nitrogen environment in an anaerobic chamber

(Chapter 6), co-crystallization experiments can be attempted. Even a ternary complex between

IruO, NADPH and Ga(III)-desferrioxamine is theoretically feasible since gallium(III) is redox inert. A ligand bound structure would potentially reveal alternative conformational states that are hypothesized to be critical for catalysis. Furthermore, ligand bound structures would reveal

170

residues required for substrate binding, which could be experimentally validated by site-directed mutagenesis coupled with activity assays.

The work presented in this thesis represents the first step in the understanding of how siderophore-mediated metabolism integrates into S. aureus central metabolism and sets the foundation for future investigation into S. aureus siderophore-mediated iron metabolism.

171

Bibliography

1. Somerville, G. A., and Proctor, R. A. (2009) The Biology of Staphylococci. in Staphylococci in Human Disease, Wiley-Blackwell. pp 1-18 2. Noble, W. C., H.A., V., and Wolters, C. H. (1967) Carriage of Staphylococcus aureus in random samples of a normal population. Journal of Hygiene Cambridge 65, 567-573 3. Casewell, M. W., and Hill, R. L. R. (1986) The carrier state: methicillin-resistant Staphylococcus aureus. Journal of Antimicrobial Chemotherapy 18, 1-12 4. Lowy, F. D. (1998) Staphylococcus aureus infections. The New England Journal of Medicine 339, 520-532 5. Verbrugh, H. A. (2009) Colonization with Staphylococcus aureus and the role of colonization in causing infection. in Staphylococci in Human Disease, Wiley-Blackwell. pp 253-271 6. Barber, M., and Rozwadowska-Dowzenko, M. (1948) Infection by penicillin-resistant staphylococci. The Lancet 252, 641-644 7. Lowy, F. D. (2013) Methicillin-resistant Staphylococcus aureus: where is it coming from and where is it going? JAMA Internal Medicine 173, 1978-1979 8. Chambers, H. F. (2001) The changing epidemiology of Staphylococcus aureus? Emerging Infectious Diseases 7, 178-182 9. Mediavilla, J. R., Chen, L., Mathema, B., and Kreiswirth, B. N. (2012) Global epidemiology of community-associated methicillin resistant Staphylococcus aureus (CA- MRSA). Current Opinion in Microbiology 15, 588-595 10. Uhlemann, A. C., Otto, M., Lowy, F. D., and DeLeo, F. R. (2014) Evolution of community- and healthcare-associated methicillin-resistant Staphylococcus aureus. Infection, Genetics Evolution 21, 563-574 11. Gardete, S., and Tomasz, A. (2014) Mechanisms of vancomycin resistance in Staphylococcus aureus. The Journal of Clinical Investigation 124, 2836-2840 12. Dantes, R., Mu, Y., Belflower, R., Aragon, D., Dumyati, G., Harrison, L. H., Lessa, F. C., Lynfield, R., Nadle, J., Petit, S., Ray, S. M., Schaffner, W., Townes, J., Fridkin, S., and Emerging Infections Program Active Bacterial Core Surveillance, M. S. I. (2013) National burden of invasive methicillin-resistant Staphylococcus aureus infections, United States, 2011. JAMA Internal Medicine 173, 1970-1978 13. Simor, A. E., Ofner-Agostini, M., Gravel, D., Varia, M., Paton, S., McGeer, A., Bryce, E., Loeb, M., Mulvey, M., and Program, C. N. I. S. (2005) Surveillance for methicillin- resistant Staphylococcus aureus in Canadian hospitals--a report update from the Canadian Nosocomial Infection Surveillance Program. Canada Communicable Disease Report 31, 33-40 14. Valiquette, L., Chakra, C. N., and Laupland, K. B. (2014) Financial impact of health care-associated infections: When money talks. The Canadian Journal of Infectious Diseases & Medical Microbiology 25, 71-74 15. Novick, R. (2006) Staphylococcal pathogenesis and pathogenicity factors: genetics and regulation. Gram-Positive Pathogens, 2nd ed. ASM Press, Washington, DC 26, 496-516 16. Foster, T. J., Geoghegan, J. A., Ganesh, V. K., and Hook, M. (2014) Adhesion, invasion and evasion: the many functions of the surface proteins of Staphylococcus aureus. Nature Reviews Microbiology 12, 49-62 172

17. O'Riordan, K., and Lee, J. C. (2004) Staphylococcus aureus capsular polysaccharides. Clinical Microbiology Reviews 17, 218-234 18. Richardson, A. R., Dunman, P. M., and Fang, F. C. (2006) The nitrosative stress response of Staphylococcus aureus is required for resistance to innate immunity. Molecular Microbiology 61, 927-939 19. Mandell, G. L. (1975) Catalase, superoxide dismutase, and virulence of Staphylococcus aureus. In vitro and in vivo studies with emphasis on staphylococcal--leukocyte interaction. The Journal of Clinical Investigation 55, 561-566 20. Clements, M. O., and Foster, S. J. (1999) Stress resistance in Staphylococcus aureus. Trends in Microbiology 7, 458-462 21. Clauditz, A., Resch, A., Wieland, K. P., Peschel, A., and Gotz, F. (2006) Staphyloxanthin plays a role in the fitness of Staphylococcus aureus and its ability to cope with oxidative stress. Infection and Immunity 74, 4950-4953 22. Otto, M. (2014) Staphylococcus aureus toxins. Current Opinion in Microbiology 17, 32- 37 23. Sheldon, J. R., and Heinrichs, D. E. (2012) The iron-regulated staphylococcal lipoproteins. Frontiers in Cellular and Infection Microbiology 2, 41 24. Andrews, S. C., Robinson, A. K., and Rodriguez-Quinones, F. (2003) Bacterial iron homeostasis. FEMS Microbiology Reviews 27, 215-237 25. Loutet, S. A., Chan, A. C., Kobylarz, M. J., Verstraete, M. M., Pfaffen, S., Ye, B., Arrieta, A. L., and Murphy, M. E. (2015) The Fate of Intracellular Metal Ions in Microbes. Trace Metals and Infectious Diseases, 39-56 26. Torres, V. J., Pishchany, G., Humayun, M., Schneewind, O., and Skaar, E. P. (2006) Staphylococcus aureus IsdB is a hemoglobin receptor required for heme iron utilization. Journal of Bacteriology 188, 8421-8429 27. Skaar, E. P., Humayun, M., Bae, T., DeBord, K. L., and Schneewind, O. (2004) Iron- source preference of Staphylococcus aureus infections. Science 305, 1626-1628 28. Skaar, E. P., and Schneewind, O. (2004) Iron-regulated surface determinants (Isd) of Staphylococcus aureus: stealing iron from heme. Microbes and Infection 6, 390-397 29. Mazmanian, S. K., Skaar, E. P., Gaspar, A. H., Humayun, M., Gornicki, P., Jelenska, J., Joachmiak, A., Missiakas, D. M., and Schneewind, O. (2003) Passage of heme-iron across the envelope of Staphylococcus aureus. Science 299, 906-909 30. Mazmanian, S. K., Ton-That, H., Su, K., and Schneewind, O. (2002) An iron-regulated sortase anchors a class of surface protein during Staphylococcus aureus pathogenesis. Proceedings of the National Academy of Sciences of the United States of America 99, 2293-2298 31. Tiedemann, M. T., Heinrichs, D. E., and Stillman, M. J. (2012) Multiprotein heme shuttle pathway in Staphylococcus aureus: iron-regulated surface determinant cog-wheel kinetics. Journal of the American Chemical Society 134, 16578-16585 32. Grigg, J. C., Vermeiren, C. L., Heinrichs, D. E., and Murphy, M. E. (2007) Heme coordination by Staphylococcus aureus IsdE. The Journal of Biological Chemistry 282, 28815-28822 33. Andrade, M. A., Ciccarelli, F. D., Perez-Iratxeta, C., and Bork, P. (2002) NEAT: a domain duplicated in genes near the components of a putative Fe3+ siderophore transporter from Gram-positive pathogenic bacteria. Genome Biology 3, Research0047 173

34. Lee, W. C., Reniere, M. L., Skaar, E. P., and Murphy, M. E. (2008) Ruffling of metalloporphyrins bound to IsdG and IsdI, two heme-degrading enzymes in Staphylococcus aureus. The Journal of Biological Chemistry 283, 30957-30963 35. Reniere, M. L., Ukpabi, G. N., Harry, S. R., Stec, D. F., Krull, R., Wright, D. W., Bachmann, B. O., Murphy, M. E., and Skaar, E. P. (2010) The IsdG-family of haem oxygenases degrades haem to a novel chromophore. Molecular Microbiology 75, 1529- 1538 36. Hider, R. C., and Kong, X. (2010) Chemistry and biology of siderophores. Natural Product Reports 27, 637-657 37. Romheld, V., and Marschner, H. (1986) Evidence for a specific uptake system for iron phytosiderophores in roots of grasses. Plant Physiology 80, 175-180 38. De Voss, J. J., Rutter, K., Schroeder, B. G., Su, H., Zhu, Y., and Barry, C. E., 3rd. (2000) The salicylate-derived mycobactin siderophores of Mycobacterium tuberculosis are essential for growth in macrophages. Proceedings of the National Academy of Sciences of the United States of America 97, 1252-1257 39. Boukhalfa, H., and Crumbliss, A. L. (2002) Chemical aspects of siderophore mediated iron transport. Biometals 15, 325-339 40. Carrano, C. J., Drechsel, H., Kaiser, D., Jung, G., Matzanke, B., Winkelmann, G., Rochel, N., and Albrecht-Gary, A. M. (1996) Coordination Chemistry of the Carboxylate Type Siderophore Rhizoferrin: The Iron(III) Complex and Its Metal Analogs. Inorganic Chemistry 35, 6429-6436 41. Field, T. B., Mccourt, J. L., and Mcbryde, W. A. E. (1974) Composition and Stability of Iron and Copper Citrate Complexes in Aqueous-Solution. Canadian Journal of Chemistry 52, 3119-3124 42. Aisen, P., Leibman, A., and Zweier, J. (1978) Stoichiometric and site characteristics of the binding of iron to human transferrin. The Journal of Biological Chemistry 253, 1930- 1937 43. Poole, K., and McKay, G. A. (2003) Iron acquisition and its control in Pseudomonas aeruginosa: many roads lead to Rome. Frontiers in Bioscience 8, d661-686 44. Russo, T. A., Olson, R., MacDonald, U., Beanan, J., and Davidson, B. A. (2015) Aerobactin, but not yersiniabactin, salmochelin and enterobactin, enables the growth/survival of hypervirulent (hypermucoviscous) Klebsiella pneumoniae ex vivo and in vivo. Infection and Immunity 83, 3325-3333 45. Wilson, M. K., Abergel, R. J., Raymond, K. N., Arceneaux, J. E., and Byers, B. R. (2006) Siderophores of Bacillus anthracis, Bacillus cereus, and Bacillus thuringiensis. Biochemical and Biophysical Research Communications 348, 320-325 46. Sheldon, J. R., and Heinrichs, D. E. (2015) Recent developments in understanding the iron acquisition strategies of gram positive pathogens. FEMS Microbiology Reviews 39, 592-630 47. Beasley, F. C., and Heinrichs, D. E. (2010) Siderophore-mediated iron acquisition in the staphylococci. Journal of Inorganic Biochemistry 104, 282-288 48. Sebulsky, M. T., Speziali, C. D., Shilton, B. H., Edgell, D. R., and Heinrichs, D. E. (2004) FhuD1, a ferric hydroxamate-binding lipoprotein in Staphylococcus aureus: a case of gene duplication and lateral transfer. The Journal of Biological Chemistry 279, 53152- 53159 174

49. Speziali, C. D., Dale, S. E., Henderson, J. A., Vines, E. D., and Heinrichs, D. E. (2006) Requirement of Staphylococcus aureus ATP-binding cassette-ATPase FhuC for iron- restricted growth and evidence that it functions with more than one iron transporter. Journal of Bacteriology 188, 2048-2055 50. Sebulsky, M. T., Shilton, B. H., Speziali, C. D., and Heinrichs, D. E. (2003) The role of FhuD2 in iron(III)-hydroxamate transport in Staphylococcus aureus. Demonstration that FhuD2 binds iron(III)-hydroxamates but with minimal conformational change and implication of mutations on transport. The Journal of Biological Chemistry 278, 49890- 49900 51. Sebulsky, M. T., and Heinrichs, D. E. (2001) Identification and characterization of fhuD1 and fhuD2, two genes involved in iron-hydroxamate uptake in Staphylococcus aureus. Journal of Bacteriology 183, 4994-5000 52. Sebulsky, M. T., Hohnstein, D., Hunter, M. D., and Heinrichs, D. E. (2000) Identification and characterization of a membrane permease involved in iron-hydroxamate transport in Staphylococcus aureus. Journal of Bacteriology 182, 4394-4400 53. Podkowa, K. J., Briere, L. A., Heinrichs, D. E., and Shilton, B. H. (2014) Crystal and solution structure analysis of FhuD2 from Staphylococcus aureus in multiple unliganded conformations and bound to ferrioxamine-B. Biochemistry 53, 2017-2031 54. Mariotti, P., Malito, E., Biancucci, M., Lo Surdo, P., Mishra, R. P., Nardi-Dei, V., Savino, S., Nissum, M., Spraggon, G., Grandi, G., Bagnoli, F., and Bottomley, M. J. (2013) Structural and functional characterization of the Staphylococcus aureus virulence factor and vaccine candidate FhuD2. The Biochemical Journal 449, 683-693 55. Mishra, R. P., Mariotti, P., Fiaschi, L., Nosari, S., Maccari, S., Liberatori, S., Fontana, M. R., Pezzicoli, A., De Falco, M. G., Falugi, F., Altindis, E., Serruto, D., Grandi, G., and Bagnoli, F. (2012) Staphylococcus aureus FhuD2 is involved in the early phase of staphylococcal dissemination and generates protective immunity in mice. The Journal of Infectious Diseases 206, 1041-1049 56. Arifin, A. J., Hannauer, M., Welch, I., and Heinrichs, D. E. (2014) Deferoxamine mesylate enhances virulence of community-associated methicillin resistant Staphylococcus aureus. Microbes and Infection 16, 967-972 57. Rahav, G., Volach, V., Shapiro, M., Rund, D., Rachmilewitz, E. A., and Goldfarb, A. (2006) Severe infections in thalassaemic patients: prevalence and predisposing factors. British Journal of Haematology 133, 667-674 58. Beasley, F. C., Marolda, C. L., Cheung, J., Buac, S., and Heinrichs, D. E. (2011) Staphylococcus aureus Transporters Hts, Sir, and Sst Capture Iron Liberated from Human Transferrin by Staphyloferrin A, Staphyloferrin B, and Catecholamine Stress Hormones, Respectively, and Contribute to Virulence. Infection and Immunity 79, 2345-2355 59. Freestone, P. P., Lyte, M., Neal, C. P., Maggs, A. F., Haigh, R. D., and Williams, P. H. (2000) The mammalian neuroendocrine hormone norepinephrine supplies iron for bacterial growth in the presence of transferrin or lactoferrin. Journal of Bacteriology 182, 6091-6098 60. Sandrini, S. M., Shergill, R., Woodward, J., Muralikuttan, R., Haigh, R. D., Lyte, M., and Freestone, P. P. (2010) Elucidation of the Mechanism by Which Catecholamine Stress Hormones Liberate Iron from the Innate Immune Defense Proteins Transferrin and Lactoferrin. Journal of Bacteriology 192, 587-594 175

61. Morrissey, J. A., Cockayne, A., Hill, P. J., and Williams, P. (2000) Molecular cloning and analysis of a putative siderophore ABC transporter from Staphylococcus aureus. Infection and Immunity 68, 6281-6288 62. Meiwes, J., Fiedler, H. P., Haag, H., Zahner, H., Konetschny-Rapp, S., and Jung, G. (1990) Isolation and characterization of staphyloferrin A, a compound with siderophore activity from Staphylococcus hyicus DSM 20459. FEMS Microbiology Letters 55, 201- 205 63. Haag, H., Fiedler, H. P., Meiwes, J., Drechsel, H., Jung, G., and Zahner, H. (1994) Isolation and biological characterization of staphyloferrin B, a compound with siderophore activity from staphylococci. FEMS Microbiology Letters 115, 125-130 64. Drechsel, H., Freund, S., Nicholson, G., Haag, H., Jung, O., Zahner, H., and Jung, G. (1993) Purification and chemical characterization of staphyloferrin B, a hydrophilic siderophore from staphylococci. Biometals 6, 185-192 65. Konetschny-Rapp, S., Jung, G., Meiwes, J., and Zahner, H. (1990) Staphyloferrin A: a structurally new siderophore from staphylococci. European Journal of Biochemistry 191, 65-74 66. Madsen, J. L. H., Johnstone, T. C., and Nolan, E. M. (2015) Chemical Synthesis of Staphyloferrin B Affords Insight into the Molecular Structure, Iron Chelation, and Biological Activity of a Polycarboxylate Siderophore Deployed by the Human Pathogen Staphylococcus aureus. Journal of the American Chemical Society 137, 9117-9127 67. Beasley, F. C., Vines, E. D., Grigg, J. C., Zheng, Q., Liu, S., Lajoie, G. A., Murphy, M. E., and Heinrichs, D. E. (2009) Characterization of staphyloferrin A biosynthetic and transport mutants in Staphylococcus aureus. Molecular Microbiology 72, 947-963 68. Dale, S. E., Sebulsky, M. T., and Heinrichs, D. E. (2004) Involvement of SirABC in iron- siderophore import in Staphylococcus aureus. Journal of Bacteriology 186, 8356-8362 69. Grigg, J. C., Cooper, J. D., Cheung, J., Heinrichs, D. E., and Murphy, M. E. (2010) The Staphylococcus aureus siderophore receptor HtsA undergoes localized conformational changes to enclose staphyloferrin A in an arginine-rich binding pocket. The Journal of Biological Chemistry 285, 11162-11171 70. Grigg, J. C., Cheung, J., Heinrichs, D. E., and Murphy, M. E. (2010) Specificity of Staphyloferrin B recognition by the SirA receptor from Staphylococcus aureus. The Journal of Biological Chemistry 285, 34579-34588 71. Miethke, M., and Marahiel, M. A. (2007) Siderophore-based iron acquisition and pathogen control. Microbiology and Molecular Biology Reviews 71, 413-451 72. Walsh, C. T., Liu, J., Rusnak, F., and Sakaitani, M. (1990) Molecular Studies on Enzymes in Chorismate Metabolism and the Enterobactin Biosynthetic-Pathway. Chemical Reviews 90, 1105-1129 73. Thomas, M. G., Chan, Y. A., and Ozanick, S. G. (2003) Deciphering tuberactinomycin biosynthesis: isolation, sequencing, and annotation of the viomycin biosynthetic gene cluster. Antimicrobial Agents and Chemotherapy 47, 2823-2830 74. Crosa, J. H., and Walsh, C. T. (2002) Genetics and assembly line enzymology of siderophore biosynthesis in bacteria. Microbiology and Molecular Biology Reviews 66, 223-249

176

75. Challis, G. L. (2005) A widely distributed bacterial pathway for siderophore biosynthesis independent of nonribosomal peptide synthetases. European Journal of Chemical Biology 6, 601-611 76. Cotton, J. L., Tao, J., and Balibar, C. J. (2009) Identification and characterization of the Staphylococcus aureus gene cluster coding for staphyloferrin A. Biochemistry 48, 1025- 1035 77. Hannauer, M., Sheldon, J. R., and Heinrichs, D. E. (2015) Involvement of major facilitator superfamily proteins SfaA and SbnD in staphyloferrin secretion in Staphylococcus aureus. FEBS letters 589, 730-737 78. Cheung, J., Beasley, F. C., Liu, S., Lajoie, G. A., and Heinrichs, D. E. (2009) Molecular characterization of staphyloferrin B biosynthesis in Staphylococcus aureus. Molecular Microbiology 74, 594-608 79. Dale, S. E., Doherty-Kirby, A., Lajoie, G., and Heinrichs, D. E. (2004) Role of siderophore biosynthesis in virulence of Staphylococcus aureus: identification and characterization of genes involved in production of a siderophore. Infection and Immunity 72, 29-37 80. Beasley, F. C., Cheung, J., and Heinrichs, D. E. (2011) Mutation of L-2,3- diaminopropionic acid synthase genes blocks staphyloferrin B synthesis in Staphylococcus aureus. BMC Microbiology 11, 199 81. Cheung, J., Murphy, M. E., and Heinrichs, D. E. (2012) Discovery of an iron-regulated citrate synthase in Staphylococcus aureus. Chemistry & Biology 19, 1568-1578 82. Sheldon, J. R., Marolda, C. L., and Heinrichs, D. E. (2014) TCA cycle activity in Staphylococcus aureus is essential for iron-regulated synthesis of staphyloferrin A, but not staphyloferrin B: the benefit of a second citrate synthase. Molecular Microbiology 92, 824-839 83. Ben Zakour, N. L., Beatson, S. A., van den Broek, A. H., Thoday, K. L., and Fitzgerald, J. R. (2012) Comparative genomics of the Staphylococcus intermedius group of animal pathogens. Front Cell Infect Microbiol 2, 44 84. Brozyna, J. R., Sheldon, J. R., and Heinrichs, D. E. (2014) Growth promotion of the opportunistic human pathogen, Staphylococcus lugdunensis, by heme, hemoglobin, and coculture with Staphylococcus aureus. MicrobiologyOpen 3, 182-195 85. Somerville, G. A., and Proctor, R. A. (2009) At the Crossroads of Bacterial Metabolism and Virulence Factor Synthesis in Staphylococci. Microbiology and Molecular Biology Reviews 73, 233-248 86. Krebs, H. A. (1937) Dismutation of pyruvic acid in Gonococcus and Staphylococcus. The Biochemical Journal 31, 661-671 87. Gardner, J. F., and Lascelles, J. (1962) The requirement for acetate of a streptomycin- resistant strain of Staphylococcus aureus. Journal of General Microbiology 29, 157-164 88. Collins, F. M., and Lascelles, J. (1962) The effect of growth conditions on oxidative and dehydrogenase activity in Staphylococcus aureus. Journal of General Microbiology 29, 531-535 89. Mah, R. A., Fung, D. Y., and Morse, S. A. (1967) Nutritional requirements of Staphylococcus aureus S-6. Applied Microbiology 15, 866-870 90. Strasters, K. C., and Winkler, K. C. (1963) Carbohydrate Metabolism of Staphylococcus Aureus. Journal of General Microbiology 33, 213-229 177

91. Somerville, G. A., Said-Salim, B., Wickman, J. M., Raffel, S. J., Kreiswirth, B. N., and Musser, J. M. (2003) Correlation of acetate catabolism and growth yield in Staphylococcus aureus: implications for host-pathogen interactions. Infection and Immunity 71, 4724-4732 92. Somerville, G. A., Chaussee, M. S., Morgan, C. I., Fitzgerald, J. R., Dorward, D. W., Reitzer, L. J., and Musser, J. M. (2002) Staphylococcus aureus aconitase inactivation unexpectedly inhibits post-exponential-phase growth and enhances stationary-phase survival. Infection and Immunity 70, 6373-6382 93. Seidl, K., Stucki, M., Ruegg, M., Goerke, C., Wolz, C., Harris, L., Berger-Bachi, B., and Bischoff, M. (2006) Staphylococcus aureus CcpA affects virulence determinant production and antibiotic resistance. Antimicrobial Agents and Chemotherapy 50, 1183- 1194 94. Jankovic, I., and Bruckner, R. (2002) Carbon catabolite repression by the catabolite control protein CcpA in Staphylococcus xylosus. Journal of Molecular Microbiology and Biotechnology 4, 309-314 95. Jankovic, I., Egeter, O., and Bruckner, R. (2001) Analysis of catabolite control protein A- dependent repression in Staphylococcus xylosus by a genomic reporter gene system. Journal of Bacteriology 183, 580-586 96. Seidl, K., Muller, S., Francois, P., Kriebitzsch, C., Schrenzel, J., Engelmann, S., Bischoff, M., and Berger-Bachi, B. (2009) Effect of a glucose impulse on the CcpA regulon in Staphylococcus aureus. Bmc Microbiology 9 97. Nuxoll, A. S., Halouska, S. M., Sadykov, M. R., Hanke, M. L., Bayles, K. W., Kielian, T., Powers, R., and Fey, P. D. (2012) CcpA Regulates Arginine Biosynthesis in Staphylococcus aureus through Repression of Proline Catabolism. Plos Pathogens 8 98. Li, C. L., Sun, F., Cho, H., Yelavarthi, V., Sohn, C., He, C. A., Schneewind, O., and Bae, T. (2010) CcpA Mediates Proline Auxotrophy and Is Required for Staphylococcus aureus Pathogenesis. Journal of Bacteriology 192, 3883-3892 99. Majerczyk, C. D., Dunman, P. M., Luong, T. T., Lee, C. Y., Sadykov, M. R., Somerville, G. A., Bodi, K., and Sonenshein, A. L. (2010) Direct Targets of CodY in Staphylococcus aureus. Journal of Bacteriology 192, 2861-2877 100. Pohl, K., Francois, P., Stenz, L., Schlink, F., Geiger, T., Herbert, S., Goerke, C., Schrenzel, J., and Wolz, C. (2009) CodY in Staphylococcus aureus: a Regulatory Link between Metabolism and Virulence Gene Expression. Journal of Bacteriology 191, 2953- 2963 101. Majerczyk, C. D., Sadykov, M. R., Luong, T. T., Lee, C., Somerville, G. A., and Sonenshein, A. L. (2008) Staphylococcus aureus CodY negatively regulates virulence gene expression. Journal of Bacteriology 190, 2257-2265 102. Nriagu, J. O., and Skaar, E. P. (2015) Trace Metals and Infectious Diseases, MIT Press 103. Xiong, A., Singh, V. K., Cabrera, G., and Jayaswal, R. K. (2000) Molecular characterization of the ferric-uptake regulator, fur, from Staphylococcus aureus. Microbiology 146, 659-668 104. Novichkov, P. S., Kazakov, A. E., Ravcheev, D. A., Leyn, S. A., Kovaleva, G. Y., Sutormin, R. A., Kazanov, M. D., Riehl, W., Arkin, A. P., Dubchak, I., and Rodionov, D. A. (2013) RegPrecise 3.0-A resource for genome-scale exploration of transcriptional regulation in bacteria. BMC Genomics 14 178

105. Torres, V. J., Attia, A. S., Mason, W. J., Hood, M. I., Corbin, B. D., Beasley, F. C., Anderson, K. L., Stauff, D. L., McDonald, W. H., Zimmerman, L. J., Friedman, D. B., Heinrichs, D. E., Dunman, P. M., and Skaar, E. P. (2010) Staphylococcus aureus fur regulates the expression of virulence factors that contribute to the pathogenesis of pneumonia. Infection and Immunity 78, 1618-1628 106. Weidenmaier, C., Goerke, C., and Wolz, C. (2012) Staphylococcus aureus determinants for nasal colonization. Trends in Microbiology 20, 243-250 107. Krismer, B., Liebeke, M., Janek, D., Nega, M., Rautenberg, M., Hornig, G., Unger, C., Weidenmaier, C., Lalk, M., and Peschel, A. (2014) Nutrient limitation governs Staphylococcus aureus metabolism and niche adaptation in the human nose. PLoS Pathogens 10, e1003862 108. Burian, M., Wolz, C., and Goerke, C. (2010) Regulatory adaptation of Staphylococcus aureus during nasal colonization of humans. PLoS One 5, e10040 109. Burian, M., Rautenberg, M., Kohler, T., Fritz, M., Krismer, B., Unger, C., Hoffmann, W. H., Peschel, A., Wolz, C., and Goerke, C. (2010) Temporal expression of adhesion factors and activity of global regulators during establishment of Staphylococcus aureus nasal colonization. The Journal of Infectious Diseases 201, 1414-1421 110. Pynnonen, M., Stephenson, R. E., Schwartz, K., Hernandez, M., and Boles, B. R. (2011) Hemoglobin promotes Staphylococcus aureus nasal colonization. PLoS Pathogens 7, e1002104 111. Friedman, D. B., Stauff, D. L., Pishchany, G., Whitwell, C. W., Torres, V. J., and Skaar, E. P. (2006) Staphylococcus aureus redirects central metabolism to increase iron availability. PLoS Pathogens 2, e87 112. Chaffin, D. O., Taylor, D., Skerrett, S. J., and Rubens, C. E. (2012) Changes in the Staphylococcus aureus transcriptome during early adaptation to the lung. PLoS One 7, e41329 113. Psychogios, N., Hau, D. D., Peng, J., Guo, A. C., Mandal, R., Bouatra, S., Sinelnikov, I., Krishnamurthy, R., Eisner, R., Gautam, B., Young, N., Xia, J., Knox, C., Dong, E., Huang, P., Hollander, Z., Pedersen, T. L., Smith, S. R., Bamforth, F., Greiner, R., McManus, B., Newman, J. W., Goodfriend, T., and Wishart, D. S. (2011) The human serum metabolome. PLoS One 6, e16957 114. Wandersman, C., and Delepelaire, P. (2004) Bacterial iron sources: from siderophores to hemophores. Annual Review of Microbiology 58, 611-647 115. Malachowa, N., Whitney, A. R., Kobayashi, S. D., Sturdevant, D. E., Kennedy, A. D., Braughton, K. R., Shabb, D. W., Diep, B. A., Chambers, H. F., Otto, M., and DeLeo, F. R. (2011) Global changes in Staphylococcus aureus gene expression in human blood. PLoS One 6, e18617 116. Allard, M., Moisan, H., Brouillette, E., Gervais, A. L., Jacques, M., Lacasse, P., Diarra, M. S., and Malouin, F. (2006) Transcriptional modulation of some Staphylococcus aureus iron-regulated genes during growth in vitro and in a tissue cage model in vivo. Microbes and Infection 8, 1679-1690 117. Braun, V., and Hantke, K. (2011) Recent insights into iron import by bacteria. Current Opinion in Chemical Biology 15, 328-334 118. Sciara, G., Kendrew, S. G., Miele, A. E., Marsh, N. G., Federici, L., Malatesta, F., Schimperna, G., Savino, C., and Vallone, B. (2003) The structure of ActVA-Orf6, a 179

novel type of monooxygenase involved in actinorhodin biosynthesis. Embo Journal 22, 205-215 119. Skaar, E. P., Gaspar, A. H., and Schneewind, O. (2004) IsdG and IsdI, heme-degrading enzymes in the cytoplasm of Staphylococcus aureus. Journal of Biological Chemistry 279, 436-443 120. Matsui, T., Nambu, S., Ono, Y., Goulding, C. W., Tsurnoto, K., and Ikeda-Saitot, M. (2013) Heme Degradation by Staphylococcus aureus IsdG and IsdI Liberates Formaldehyde Rather Than Carbon Monoxide. Biochemistry 52, 3025-3027 121. Wu, R., Skaar, E. P., Zhang, R., Joachimiak, G., Gornicki, P., Schneewind, O., and Joachimiak, A. (2005) Staphylococcus aureus IsdG and IsdI, heme-degrading enzymes with structural similarity to monooxygenases. The Journal of Biological Chemistry 280, 2840-2846 122. Loutet, S. A., Kobylarz, M. J., Chau, C. H. T., and Murphy, M. E. P. (2013) IruO Is a Reductase for Heme Degradation by IsdI and IsdG Proteins in Staphylococcus aureus. Journal of Biological Chemistry 288, 25749-25759 123. Hannauer, M., Arifin, A. J., and Heinrichs, D. E. (2015) Involvement of reductases IruO and NtrA in iron acquisition by Staphylococcus aureus. Molecular Microbiology 96, 1192-1210 124. Coves, J., and Fontecave, M. (1993) Reduction and mobilization of iron by a NAD(P)H - flavin oxidoreductase from Escherichia coli. European Journal of Biochemistry 211, 635- 641 125. Fontecave, M., Coves, J., and Pierre, J. L. (1994) Ferric reductases or flavin reductases. Biometals 7, 3-8 126. Matzanke, B. F., Anemuller, S., Schunemann, V., Trautwein, A. X., and Hantke, K. (2004) FhuF, part of a siderophore-reductase system. Biochemistry 43, 1386-1392 127. Miethke, M., Pierik, A. J., Peuckert, F., Seubert, A., and Marahiel, M. A. (2011) Identification and characterization of a novel-type ferric siderophore reductase from a gram-positive extremophile. The Journal of Biological Chemistry 286, 2245-2260 128. Miethke, M., Hou, J., and Marahiel, M. A. (2011) The siderophore-interacting protein YqjH acts as a ferric reductase in different iron assimilation pathways of Escherichia coli. Biochemistry 50, 10951-10964 129. Wyckoff, E. E., Allred, B. E., Raymond, K. N., and Payne, S. M. (2015) Catechol siderophore transport by Vibrio cholerae. Journal of Bacteriology 130. Butterton, J. R., and Calderwood, S. B. (1994) Identification, cloning, and sequencing of a gene required for ferric vibriobactin utilization by Vibrio cholerae. Journal of Bacteriology 176, 5631-5638 131. Wang, S. N., Wu, Y., and Outten, F. W. (2011) Fur and the novel regulator YqjI control transcription of the ferric reductase gene yqjH in Escherichia coli. Journal of Bacteriology 193, 563-574 132. Stojiljkovic, I., Baumler, A. J., and Hantke, K. (1994) Fur regulon in Gram-negative bacteria - Identification and characterization of new iron-regulated Escherichia coli genes by a fur titration assay. Journal of Molecular Biology 236, 531-545 133. Zheng, T., and Nolan, E. M. (2014) Enterobactin-mediated delivery of beta-lactam antibiotics enhances antibacterial activity against pathogenic Escherichia coli. Journal of the American Chemical Society 136, 9677-9691 180

134. Duthie, E. S., and Lorenz, L. L. (1952) Staphylococcal coagulase; mode of action and antigenicity. Journal of General Microbiology 6, 95-107 135. Peng, H. L., Novick, R. P., Kreiswirth, B., Kornblum, J., and Schlievert, P. (1988) Cloning, characterization, and sequencing of an accessory gene regulator (agr) in Staphylococcus aureus. Journal of Bacteriology 170, 4365-4372 136. Monk, I. R., Shah, I. M., Xu, M., Tan, M. W., and Foster, T. J. (2012) Transforming the untransformable: application of direct transformation to manipulate genetically Staphylococcus aureus and Staphylococcus epidermidis. MBio 3, e00277 137. Bateman, B. T., Donegan, N. P., Jarry, T. M., Palma, M., and Cheung, A. L. (2001) Evaluation of a tetracycline-inducible promoter in Staphylococcus aureus in vitro and in vivo and its application in demonstrating the role of sigB in microcolony formation. Infection and Immunity 69, 7851-7857 138. MacPherson, I. S., Rosell, F. I., Scofield, M., Mauk, A. G., and Murphy, M. E. (2010) Directed evolution of copper nitrite reductase to a chromogenic reductant. Protein Engineering Design & Selection 23, 137-145 139. Van Duyne, G. D., Standaert, R. F., Karplus, P. A., Schreiber, S. L., and Clardy, J. (1993) Atomic structures of the human immunophilin FKBP-12 complexes with FK506 and rapamycin. Journal of Molecular Biology 229, 105-124 140. Hames, C., Halbedel, S., Schilling, O., and Stulke, J. (2005) Multiple-mutation reaction: a method for simultaneous introduction of multiple mutations into the glpK gene of Mycoplasma pneumoniae. Applied and Environmental Microbiology 71, 4097-4100 141. Kabsch, W. (2010) Xds. Acta Crystallographica Section D 66, 125-132 142. Kabsch, W. (2010) Integration, scaling, space-group assignment and post-refinement. Acta Crystallographica Section D 66, 133-144 143. Adams, P. D., Afonine, P. V., Bunkoczi, G., Chen, V. B., Davis, I. W., Echols, N., Headd, J. J., Hung, L. W., Kapral, G. J., Grosse-Kunstleve, R. W., McCoy, A. J., Moriarty, N. W., Oeffner, R., Read, R. J., Richardson, D. C., Richardson, J. S., Terwilliger, T. C., and Zwart, P. H. (2010) PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallographica Section D 66, 213- 221 144. Leslie, A. G. W. (1992). Joint CCP4 and ESF-EACMB Newsletter on Protein Crystallography, 27-33 145. Evans, P. (1997) Scala. Joint CCP4 and ESF-EACMB Newsletter on Protein Crystallography 33, 22-24 146. Vagin, A., and Teplyakov, A. (2010) Molecular replacement with MOLREP. Acta Crystallographica Section D 66, 22-25 147. Winn, M. D., Ballard, C. C., Cowtan, K. D., Dodson, E. J., Emsley, P., Evans, P. R., Keegan, R. M., Krissinel, E. B., Leslie, A. G. W., McCoy, A., McNicholas, S. J., Murshudov, G. N., Pannu, N. S., Potterton, E. A., Powell, H. R., Read, R. J., Vagin, A., and Wilson, K. S. (2011) Overview of the CCP4 suite and current developments. Acta Crystallographica Section D 67, 235-242 148. Emsley, P., Lohkamp, B., Scott, W. G., and Cowtan, K. (2010) Features and development of Coot. Acta Crystallographica Section D 66, 486-501

181

149. Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997) Refinement of Macromolecular Structures by the Maximum-Likelihood Method. Acta Crystallographica Section D 53, 240-255 150. Painter, J., and Merritt, E. A. (2006) Optimal description of a protein structure in terms of multiple groups undergoing TLS motion. Acta Crystallographica Section D 62, 439-450 151. McCoy, A. J., Grosse-Kunstleve, R. W., Adams, P. D., Winn, M. D., Storoni, L. C., and Read, R. J. (2007) Phaser crystallographic software. Journal of Applied Crystallography 40, 658-674 152. Otwinowski, Z., and Minor, W. (1997) Processing of X-ray diffraction data collected in oscillation mode. Methods in Enzymology 276, 307-326 153. Bunkoczi, G., and Read, R. J. (2011) Improvement of molecular-replacement models with Sculptor. Acta Crystallographica Section D 67, 303-312 154. Terwilliger, T. C., Grosse-Kunstleve, R. W., Afonine, P. V., Moriarty, N. W., Zwart, P. H., Hung, L. W., Read, R. J., and Adams, P. D. (2008) Iterative model building, structure refinement and density modification with the PHENIX AutoBuild wizard. Acta Crystallographica Section D 64, 61-69 155. Leslie, A. G. W., and Powell, H. R. (2007) Processing diffraction data with MOSFLM. Evolving Methods for Macromolecular Crystallography 245, 41-51 156. Evans, P. (2006) Scaling and assessment of data quality. Acta Crystallographica Section D 62, 72-82 157. Afonine, P. V., Grosse-Kunstleve, R. W., Echols, N., Headd, J. J., Moriarty, N. W., Mustyakimov, M., Terwilliger, T. C., Urzhumtsev, A., Zwart, P. H., and Adams, P. D. (2012) Towards automated crystallographic structure refinement with phenix.refine. Acta Crystallographica Section D 68, 352-367 158. Terwilliger, T. C., and Berendzen, J. (1999) Automated MAD and MIR structure solution. Acta Crystallographica Section D 55, 849-861 159. Terwilliger, T. C. (2000) Maximum-likelihood density modification. Acta Crystallographica Section D 56, 965-972 160. Terwilliger, T. C. (2003) Automated main-chain model building by template matching and iterative fragment extension. Acta Crystallographica Section D 59, 38-44 161. Cowtan, K. (1994) 'dm': An automated procedure for phase improvement by density modification. Joint CCP4 and ESF-EACBM Newsletter On Protein Crystallography, 5 162. Perrakis, A., Morris, R., and Lamzin, V. S. (1999) Automated protein model building combined with iterative structure refinement. Nature Structural Biology 6, 458-463 163. Baker, N. A., Sept, D., Joseph, S., Holst, M. J., and McCammon, J. A. (2001) Electrostatics of nanosystems: Application to microtubules and the ribosome. Proceedings of the National Academy of Sciences of the United States of America 98, 10037-10041 164. Folta-Stogniew, E. (2006) Oligomeric states of proteins determined by size-exclusion chromatography coupled with light scattering, absorbance, and refractive index detectors. Methods in Molecular Biology 328, 97-112 165. Altschul, S. F., Gish, W., Miller, W., Myers, E. W., and Lipman, D. J. (1990) Basic local alignment search tool. Journal of Molecular Biology 215, 403-410

182

166. Altschul, S. F., and Lipman, D. J. (1990) Protein database searches for multiple alignments. Proceedings of the National Academy of Sciences of the United States of America 87, 5509-5513 167. Holm, L., and Rosenstrom, P. (2010) Dali server: conservation mapping in 3D. Nucleic Acids Research 38, W545-549 168. Krissinel, E., and Henrick, K. (2004) Secondary-structure matching (SSM), a new tool for fast protein structure alignment in three dimensions. Acta Crystallographica Section D 60, 2256-2268 169. Goujon, M., McWilliam, H., Li, W., Valentin, F., Squizzato, S., Paern, J., and Lopez, R. (2010) A new bioinformatics analysis tools framework at EMBL-EBI. Nucleic Acids Research 38, W695-699 170. Sievers, F., Wilm, A., Dineen, D., Gibson, T. J., Karplus, K., Li, W., Lopez, R., McWilliam, H., Remmert, M., Soding, J., Thompson, J. D., and Higgins, D. G. (2011) Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega. Molecular Systems Biology 7, 539 171. Guindon, S., and Gascuel, O. (2003) A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Systematic Biology 52, 696-704 172. Gouy, M., Guindon, S., and Gascuel, O. (2010) SeaView version 4: A multiplatform graphical user interface for sequence alignment and phylogenetic tree building. Molecular Biology and Evolution 27, 221-224 173. Steiner, E. M., Both, D., Lossl, P., Vilaplana, F., Schnell, R., and Schneider, G. (2014) CysK2 from Mycobacterium tuberculosis is an O-phospho-L-serine-dependent S- sulfocysteine synthase. Journal of Bacteriology 196, 3410-3420 174. Webb, M. R. (1992) A continuous spectrophotometric assay for inorganic phosphate and for measuring phosphate release kinetics in biological systems. Proceedings of the National Academy of Sciences of the United States of America 89, 4884-4887 175. Else, A. J., Barnes, S. J., Danson, M. J., and Weitzman, P. D. (1988) A new spectrophotometric assay for citrate synthase and its use to assess the inhibitory effects of palmitoyl thioesters. Biochemical Journal 251, 803-807 176. Stookey, L. L. (1970) Ferrozine - a New Spectrophotometric Reagent for Iron. Analytical Chemistry 42, 779-781 177. Riener, C. K., Kada, G., and Gruber, H. J. (2002) Quick measurement of protein sulfhydryls with Ellman's reagent and with 4,4'-dithiodipyridine. Analytical and Bioanalytical Chemistry 373, 266-276 178. O'Leary, S. E., Jurgenson, C. T., Ealick, S. E., and Begley, T. P. (2008) O-phospho-L- serine and the thiocarboxylated sulfur carrier protein CysO-COSH are substrates for CysM, a cysteine synthase from Mycobacterium tuberculosis. Biochemistry 47, 11606- 11615 179. Izard, T., and Blackwell, N. C. (2000) Crystal structures of the metal-dependent 2- dehydro-3-deoxy-galactarate aldolase suggest a novel reaction mechanism. The EMBO Journal 19, 3849-3856 180. Ose, T., Watanabe, K., Yao, M., Honma, M., Oikawa, H., and Tanaka, I. (2004) Structure of macrophomate synthase. Acta Crystallographica Section D 60, 1187-1197

183

181. Murzin, A. G., Brenner, S. E., Hubbard, T., and Chothia, C. (1995) SCOP: a structural classification of proteins database for the investigation of sequences and structures. Journal of Molecular Biology 247, 536-540 182. Goulding, C. W., Bowers, P. M., Segelke, B., Lekin, T., Kim, C. Y., Terwilliger, T. C., and Eisenberg, D. (2007) The structure and computational analysis of Mycobacterium tuberculosis protein CitE suggest a novel enzymatic function. Journal of Molecular Biology 365, 275-283 183. Anstrom, D. M., and Remington, S. J. (2006) The product complex of M. tuberculosis malate synthase revisited. Protein Science 15, 2002-2007 184. Howard, B. R., Endrizzi, J. A., and Remington, S. J. (2000) Crystal structure of Escherichia coli malate synthase G complexed with magnesium and glyoxylate at 2.0 A resolution: mechanistic implications. Biochemistry 39, 3156-3168 185. Rea, D., Hovington, R., Rakus, J. F., Gerlt, J. A., Fulop, V., Bugg, T. D., and Roper, D. I. (2008) Crystal structure and functional assignment of YfaU, a metal ion dependent class II aldolase from Escherichia coli K12. Biochemistry 47, 9955-9965 186. Rea, D., Fulop, V., Bugg, T. D., and Roper, D. I. (2007) Structure and mechanism of HpcH: a metal ion dependent class II aldolase from the homoprotocatechuate degradation pathway of Escherichia coli. Journal of Molecular Biology 373, 866-876 187. Franza, T., Mahe, B., and Expert, D. (2005) Erwinia chrysanthemi requires a second iron transport route dependent of the siderophore achromobactin for extracellular growth and plant infection. Molecular Microbiology 55, 261-275 188. Dudev, T., and Lim, C. (2004) Monodentate versus bidentate carboxylate binding in magnesium and calcium proteins: What are the basic principles? Journal of Physical Chemistry B 108, 4546-4557 189. Fish, D. C., and Blumenthal, H. J. (1966) [93] 2-Keto-3-deoxy-d-glucarate aldolase. in Methods in Enzymology (Willis, A. W. ed.), Academic Press. pp 529-534 190. Dominguez, D. C. (2004) Calcium signalling in bacteria. Molecular Microbiology 54, 291-297 191. Smith, C. V., Huang, C. C., Miczak, A., Russell, D. G., Sacchettini, J. C., and Honer zu Bentrup, K. (2003) Biochemical and structural studies of malate synthase from Mycobacterium tuberculosis. The Journal of Biological Chemistry 278, 1735-1743 192. Ose, T., Watanabe, K., Mie, T., Honma, M., Watanabe, H., Yao, M., Oikawa, H., and Tanaka, I. (2003) Insight into a natural Diels-Alder reaction from the structure of macrophomate synthase. Nature 422, 185-189 193. Francois, J. A., Starks, C. M., Sivanuntakorn, S., Jiang, H., Ransome, A. E., Nam, J. W., Constantine, C. Z., and Kappock, T. J. (2006) Structure of a NADH-insensitive hexameric citrate synthase that resists acid inactivation. Biochemistry 45, 13487-13499 194. Remington, S., Wiegand, G., and Huber, R. (1982) Crystallographic refinement and atomic models of two different forms of citrate synthase at 2.7 and 1.7 A resolution. Journal of Molecular Biology 158, 111-152 195. Evans, C. T., Kurz, L. C., Remington, S. J., and Srere, P. A. (1996) Active site mutants of pig citrate synthase: effects of mutations on the enzyme catalytic and structural properties. Biochemistry 35, 10661-10672

184

196. Malathi, K., Padmanab.G, and Sarma, P. S. (1970) Biosynthesis of Beta-N-Oxalyl-L- Alpha,Beta-Diaminopropionic Acid, Lathyrus-Sativus Neurotoxin. Phytochemistry 9, 1603-1610 197. Felnagle, E. A., Rondon, M. R., Berti, A. D., Crosby, H. A., and Thomas, M. G. (2007) Identification of the biosynthetic gene cluster and an additional gene for resistance to the antituberculosis drug capreomycin. Applied and Environmental Microbiology 73, 4162- 4170 198. Carter, J. H., 2nd, Du Bus, R. H., Dyer, J. R., Floyd, J. C., Rice, K. C., and Shaw, P. D. (1974) Biosynthesis of viomycin. I. Origin of alpha, beta-diaminopropionic acid and serine. Biochemistry 13, 1221-1227 199. Wang, M., and Gould, S. J. (1993) Biosynthesis of Capreomycin .2. Incorporation of L- Serine, L-Alanine, and L-2,3-Diaminopropionic Acid. Journal of Organic Chemistry 58, 5176-5180 200. Heine, A., Canaves, J. M., von Delft, F., Brinen, L. S., Dai, X., Deacon, A. M., Elsliger, M. A., Eshaghi, S., Floyd, R., Godzik, A., Grittini, C., Grzechnik, S. K., Guda, C., Jaroszewski, L., Karlak, C., Klock, H. E., Koesema, E., Kovarik, J. S., Kreusch, A., Kuhn, P., Lesley, S. A., McMullan, D., McPhillips, T. M., Miller, M. A., Miller, M. D., Morse, A., Moy, K., Ouyang, J., Page, R., Robb, A., Rodrigues, K., Schwarzenbacher, R., Selby, T. L., Spraggon, G., Stevens, R. C., van den Bedem, H., Velasquez, J., Vincent, J., Wang, X., West, B., Wolf, G., Hodgson, K. O., Wooley, J., and Wilson, I. A. (2004) Crystal structure of O-acetylserine sulfhydrylase (TM0665) from Thermotoga maritima at 1.8 A resolution. Proteins 56, 387-391 201. Agren, D., Schnell, R., Oehlmann, W., Singh, M., and Schneider, G. (2008) Cysteine synthase (CysM) of Mycobacterium tuberculosis is an O-phosphoserine sulfhydrylase: evidence for an alternative cysteine biosynthesis pathway in mycobacteria. The Journal of Biological Chemistry 283, 31567-31574 202. Krissinel, E., and Henrick, K. (2007) Inference of macromolecular assemblies from crystalline state. Journal of Molecular Biology 372, 774-797 203. Schrodinger, L. L. C. (2010) The PyMOL Molecular Graphics System, Version 1.3r1. 204. Ishikawa, K., Mino, K., and Nakamura, T. (2010) New function and application of the cysteine synthase from archaea. Biochemical Engineering Journal 48, 315-322 205. Gallagher, D. T., Monbouquette, H. G., Schroder, I., Robinson, H., Holden, M. J., and Smith, N. N. (2004) Structure of alanine dehydrogenase from Archaeoglobus: Active site analysis and relation to bacterial cyclodeaminases and mammalian mu crystallin. Journal of Molecular Biology 342, 119-130 206. Yin, X. H., and Zabriskie, T. M. (2004) VioC is a non-heme iron, alpha-ketoglutarate- dependent oxygenase that catalyzes the formation of 3S-hydroxy-L-arginine during viomycin biosynthesis. ChemBioChem 5, 1274-1277 207. Chateau, A., van Schaik, W., Six, A., Aucher, W., and Fouet, A. (2011) CodY regulation is required for full virulence and heme iron acquisition in Bacillus anthracis. The FASEB Journal 25, 4445-4456 208. van Schaik, W., Chateau, A., Dillies, M. A., Coppee, J. Y., Sonenshein, A. L., and Fouet, A. (2009) The global regulator CodY regulates toxin gene expression in Bacillus anthracis and is required for full virulence. Infection and Immunity 77, 4437-4445

185

209. Zhao, C., Song, C., Luo, Y., Yu, Z., and Sun, M. (2008) L-2,3-diaminopropionate: one of the building blocks for the biosynthesis of Zwittermicin A in Bacillus thuringiensis subsp. kurstaki strain YBT-1520. FEBS letters 582, 3125-3131 210. Dawlaty, J., Zhang, X., Fischbach, M. A., and Clardy, J. (2010) Dapdiamides, tripeptide antibiotics formed by unconventional amide ligases. Journal of Natural Products 73, 441-446 211. Cook, P. F., and Wedding, R. T. (1976) A reaction mechanism from steady state kinetic studies for O-acetylserine sulfhydrylase from Salmonella typhimurium LT-2. The Journal of Biological Chemistry 251, 2023-2029 212. Kredich, N. M., and Tomkins, G. M. (1966) The enzymic synthesis of L-cysteine in Escherichia coli and Salmonella typhimurium. The Journal of Biological Chemistry 241, 4955-4965 213. Schnell, R., Oehlmann, W., Singh, M., and Schneider, G. (2007) Structural insights into catalysis and inhibition of O-acetylserine sulfhydrylase from Mycobacterium tuberculosis. Crystal structures of the enzyme alpha-aminoacrylate intermediate and an enzyme-inhibitor complex. The Journal of Biological Chemistry 282, 23473-23481 214. Schnackerz, K. D., Tai, C. H., Simmons, J. W., Jacobson, T. M., Rao, G. S. J., and Cook, P. F. (1995) Identification and Spectral Characterization of the External Aldimine of the O-Acetylserine Sulfhydrylase Reaction. Biochemistry 34, 12152-12160 215. Chattopadhyay, A., Meier, M., Ivaninskii, S., Burkhard, P., Speroni, F., Campanini, B., Bettati, S., Mozzarelli, A., Rabeh, W. M., Li, L., and Cook, P. F. (2007) Structure, mechanism, and conformational dynamics of O-acetylserine sulfhydrylase from Salmonella typhimurium: comparison of A and B isozymes. Biochemistry 46, 8315-8330 216. Mino, K., and Ishikawa, K. (2003) A novel O-phospho-L-serine sulfhydrylation reaction catalyzed by O-acetylserine sulfhydrylase from Aeropyrum pernix K1. Febs Letters 551, 133-138 217. Eliot, A. C., and Kirsch, J. F. (2004) Pyridoxal phosphate enzymes: Mechanistic, structural, and evolutionary considerations. Annual Review of Biochemistry 73, 383-415 218. Ro, H. S. (2002) Effects of salts on the conformation and catalytic properties of d-amino acid aminotransferase. Journal of Biochemistry and Molecular Biology 35, 306-312 219. Shibata, K., Shirasuna, K., Motegi, K., Kera, Y., Abe, H., and Yamada, R. (2000) Purification and properties of alanine racemase from crayfish Procambarus clarkii. Comparative Biochemistry and Physiology. Part B, Biochemistry & Molecular Biology 126, 599-608 220. Townsend, D. E., Kaenjak, A., Jayaswal, R. K., and Wilkinson, B. J. (1996) Proline is biosynthesized from arginine in Staphylococcus aureus. Microbiology 142, 1491-1497 221. Liebeke, M., Dorries, K., Zuhlke, D., Bernhardt, J., Fuchs, S., Pane-Farre, J., Engelmann, S., Volker, U., Bode, R., Dandekar, T., Lindequist, U., Hecker, M., and Lalk, M. (2011) A metabolomics and proteomics study of the adaptation of Staphylococcus aureus to glucose starvation. Molecular Biosystems 7, 1241-1253 222. Kanehisa, M., Goto, S., Sato, Y., Kawashima, M., Furumichi, M., and Tanabe, M. (2014) Data, information, knowledge and principle: back to metabolism in KEGG. Nucleic Acids Research 42, D199-205 223. Kanehisa, M., and Goto, S. (2000) KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Research 28, 27-30 186

224. Agren, D., Schnell, R., and Schneider, G. (2009) The C-terminal of CysM from Mycobacterium tuberculosis protects the aminoacrylate intermediate and is involved in sulfur donor selectivity. FEBS Letters 583, 330-336 225. Nakamura, T., Kawai, Y., Kunimoto, K., Iwasaki, Y., Nishii, K., Kataoka, M., and Ishikawa, K. (2012) Structural analysis of the substrate recognition mechanism in O- phosphoserine sulfhydrylase from the hyperthermophilic archaeon Aeropyrum pernix K1. Journal of Molecular Biology 422, 33-44 226. Oda, Y., Mino, K., Ishikawa, K., and Ataka, M. (2005) Three-dimensional structure of a new enzyme, O-phosphoserine sulfhydrylase, involved in l-cysteine biosynthesis by a hyperthermophilic archaeon, Aeropyrum pernix K1, at 2.0A resolution. Journal of Molecular Biology 351, 334-344 227. Horsburgh, M. J., Ingham, E., and Foster, S. J. (2001) In Staphylococcus aureus, Fur is an interactive regulator with PerR, contributes to virulence, and is necessary for oxidative stress resistance through positive regulation of catalase and iron homeostasis. Journal of Bacteriology 183, 468-475 228. Horsburgh, M. J., Clements, M. O., Crossley, H., Ingham, E., and Foster, S. J. (2001) PerR controls oxidative stress resistance and iron storage proteins and is required for virulence in Staphylococcus aureus. Infection and Immunity 69, 3744-3754 229. Park, S., and Imlay, J. A. (2003) High levels of intracellular cysteine promote oxidative DNA damage by driving the Fenton reaction. Journal of Bacteriology 185, 1942-1950 230. Soutourina, O., Poupel, O., Coppee, J. Y., Danchin, A., Msadek, T., and Martin- Verstraete, I. (2009) CymR, the master regulator of cysteine metabolism in Staphylococcus aureus, controls host sulphur source utilization and plays a role in biofilm formation. Molecular Microbiology 73, 194-211 231. Soutourina, O., Dubrac, S., Poupel, O., Msadek, T., and Martin-Verstraete, I. (2010) The pleiotropic CymR regulator of Staphylococcus aureus plays an important role in virulence and stress response. Plos Pathogens 6, e1000894 232. Ji, Q. J., Zhang, L., Sun, F., Deng, X., Liang, H. H., Bae, T. O., and He, C. (2012) Staphylococcus aureus CymR is a new thiol-based oxidation-sensing regulator of stress resistance and oxidative response. Journal of Biological Chemistry 287, 21102-21109 233. Posada, A. C., Kolar, S. L., Dusi, R. G., Francois, P., Roberts, A. A., Hamilton, C. J., Liu, G. Y., and Cheung, A. (2014) Importance of bacillithiol in the oxidative stress response of Staphylococcus aureus. Infection and Immunity 82, 316-332 234. Finn, R. D., Bateman, A., Clements, J., Coggill, P., Eberhardt, R. Y., Eddy, S. R., Heger, A., Hetherington, K., Holm, L., Mistry, J., Sonnhammer, E. L., Tate, J., and Punta, M. (2014) Pfam: the protein families database. Nucleic Acids Res 42, D222-230 235. Oprian, D. D., and Coon, M. J. (1982) Oxidation-reduction states of FMN and FAD in NADPH-cytochrome P-450 reductase during reduction by NADPH. The Journal of Biological Chemistry 257, 8935-8944 236. Glaser, P., Frangeul, L., Buchrieser, C., Rusniok, C., Amend, A., Baquero, F., Berche, P., Bloecker, H., Brandt, P., Chakraborty, T., Charbit, A., Chetouani, F., Couve, E., de Daruvar, A., Dehoux, P., Domann, E., Dominguez-Bernal, G., Duchaud, E., Durant, L., Dussurget, O., Entian, K. D., Fsihi, H., Garcia-del Portillo, F., Garrido, P., Gautier, L., Goebel, W., Gomez-Lopez, N., Hain, T., Hauf, J., Jackson, D., Jones, L. M., Kaerst, U., Kreft, J., Kuhn, M., Kunst, F., Kurapkat, G., Madueno, E., Maitournam, A., Vicente, J. 187

M., Ng, E., Nedjari, H., Nordsiek, G., Novella, S., de Pablos, B., Perez-Diaz, J. C., Purcell, R., Remmel, B., Rose, M., Schlueter, T., Simoes, N., Tierrez, A., Vazquez- Boland, J. A., Voss, H., Wehland, J., and Cossart, P. (2001) Comparative genomics of Listeria species. Science 294, 849-852 237. Williams, C. H., Jr. (1995) Mechanism and structure of thioredoxin reductase from Escherichia coli. FASEB Journal 9, 1267-1276 238. Carrillo, N., and Ceccarelli, E. A. (2003) Open questions in ferredoxin-NADP+ reductase catalytic mechanism. European Journal of Biochemistry 270, 1900-1915 239. Komori, H., Seo, D., Sakurai, T., and Higuchi, Y. (2010) Crystal structure analysis of Bacillus subtilis ferredoxin-NADP(+) oxidoreductase and the structural basis for its substrate selectivity. Protein Science 19, 2279-2290 240. Muraki, N., Seo, D., Shiba, T., Sakurai, T., and Kurisu, G. (2010) Asymmetric dimeric structure of ferredoxin-NAD(P)+ oxidoreductase from the green sulfur bacterium Chlorobaculum tepidum: implications for binding ferredoxin and NADP+. Journal of Molecular Biology 401, 403-414 241. Nobre, L. S., and Saraiva, L. M. (2013) Effect of combined oxidative and nitrosative stresses on Staphylococcus aureus transcriptome. Applied Microbiology and Biotechnology 97, 2563-2573 242. Zheng, M., Wang, X. D., Doan, B., Lewis, K. A., Schneider, T. D., and Storz, G. (2001) Computation-directed identification of OxyR DNA binding sites in Escherichia coli. Journal of Bacteriology 183, 4571-4579 243. Marchler-Bauer, A., Derbyshire, M. K., Gonzales, N. R., Lu, S. N., Chitsaz, F., Geer, L. Y., Geer, R. C., He, J., Gwadz, M., Hurwitz, D. I., Lanczycki, C. J., Lu, F., Marchler, G. H., Song, J. S., Thanki, N., Wang, Z. X., Yamashita, R. A., Zhang, D. C., Zheng, C. J., and Bryant, S. H. (2015) CDD: NCBI's conserved domain database. Nucleic Acids Research 43, D222-D226 244. Berman, H., Henrick, K., and Nakamura, H. (2003) Announcing the worldwide Protein Data Bank. Nature Structural Biology 10, 980 245. Jang, S. J., and Imlay, J. A. (2007) Micromolar intracellular hydrogen peroxide disrupts metabolism by damaging iron-sulfur enzymes. Journal of Biological Chemistry 282, 929- 937 246. Keyer, K., and Imlay, J. A. (1996) Superoxide accelerates DNA damage by elevating free-iron levels. Proceedings of the National Academy of Sciences of the United States of America 93, 13635-13640 247. Clements, M. O., Watson, S. P., and Foster, S. J. (1999) Characterization of the major superoxide dismutase of Staphylococcus aureus and its role in starvation survival, stress resistance, and pathogenicity. Journal of Bacteriology 181, 3898-3903 248. Sanz, R., Marin, I., Ruiz-Santa-Quiteria, J. A., Orden, J. A., Cid, D., Diez, R. M., Silhadi, K. S., Amils, R., and de la Fuente, R. (2000) Catalase deficiency in Staphylococcus aureus subsp. anaerobius is associated with natural loss-of-function mutations within the structural gene. Microbiology 146, 465-475 249. Repine, J. E., Fox, R. B., and Berger, E. M. (1981) Hydrogen-Peroxide Kills Staphylococcus-Aureus by Reacting with Staphylococcal Iron to Form Hydroxyl Radical. Journal of Biological Chemistry 256, 7094-7096

188

250. Ji, C. J., Kim, J. H., Won, Y. B., Lee, Y. E., Choi, T. W., Ju, S. Y., Youn, H., Helmann, J. D., and Lee, J. W. (2015) Staphylococcus aureus PerR Is a Hypersensitive Hydrogen Peroxide Sensor using Iron-mediated Histidine Oxidation. Journal of Biological Chemistry 290, 20374-20386 251. Palazzolo-Ballance, A. M., Reniere, M. L., Braughton, K. R., Sturdevant, D. E., Otto, M., Kreiswirth, B. N., Skaar, E. P., and Deleo, F. R. (2008) Neutrophil microbicides induce a pathogen survival response in community-associated methicillin-resistant Staphylococcus aureus. Journal of Immunology 180, 500-509 252. Gaupp, R., Ledala, N., and Somerville, G. A. (2012) Staphylococcal response to oxidative stress. Frontiers in Cellular and Infection Microbiology 2 253. Newton, G. L., Rawat, M., La Clair, J. J., Jothivasan, V. K., Budiarto, T., Hamilton, C. J., Claiborne, A., Helmann, J. D., and Fahey, R. C. (2009) Bacillithiol is an antioxidant thiol produced in Bacilli. Nature Chemical Biology 5, 625-627 254. Newton, G. L., Fahey, R. C., and Rawat, M. (2012) Detoxification of toxins by bacillithiol in Staphylococcus aureus. Microbiology 158, 1117-1126 255. Raymond, K. N., Dertz, E. A., and Kim, S. S. (2003) Enterobactin: An archetype for microbial iron transport. Proceedings of the National Academy of Sciences of the United States of America 100, 3584-3588

189