BIOSYNTHETIC AND FUNCTIONAL STUDIES OF BACILLITHIOL IN GRAM- POSITIVE BACTERIA

BY

ZHONG FANG A Dissertation Submitted to the Graduate Faculty of

WAKE FOREST UNIVERSITY GRADUATE SCHOOL OF ARTS AND SCIENCES

in Partial Fulfillment of the Requirements

for the Degree of

DOCTOR OF PHILOSOPHY

Chemistry

May 2015

Winston-Salem, North Carolina

Approved By:

Patricia C. Dos Santos, Ph.D., Advisor

H. Alexander Claiborne Jr, Chair

Rebecca W. Alexander, Ph.D.

Stephen Bruce King, Ph.D.

Mark E. Welker, Ph.D.

TABLE OF CONTENTS LIST OF ILLUSTRATIONS AND TABLES ...... iv

LIST OF ABBREVIATIONS ...... viii

ABSTRACT ...... ix

CHAPTER 1 Introduction ...... 1

1.1 Biothiol ...... 1

1.2 Cysteine ...... 2

1.3 Glutathione ...... 6

1.4 Mycothiol ...... 10

1.5 Bacillithiol ...... 17

CHAPTER 2 Cross-functionalities of Bacillus deacetylases involved in bacillithiol biosynthesis and bacillithiol-S-conjugate detoxification pathways ...... 25

Abstract ...... 26

Introduction ...... 27

Method and materials ...... 30

Results ...... 37

Discussion ...... 51

CHAPTER 3 Protective role of bacillithiol in superoxide stress and metal homeostasis in Bacillus subtilis ...... 56

Abstract ...... 57

Introduction ...... 58

Method and materials ...... 62

Results ...... 69

Discussion ...... 86

ii

CHAPTER 4 Characterization of NifS cysteine desulfurase and its role on NAD biosynthesis in Bacillus subtilis ...... 90

Introduction ...... 91

Method and materials ...... 99

Results ...... 104

Discussion ...... 115

CHAPTER 5 Functional Studies of Nfu in Bacillus subtilis ...... 119

Introduction ...... 120

Method and materials ...... 124

Results ...... 128

Discussion ...... 132

CHAPTER 6 CONCLUDING REMARKS ...... 135

REFERENCES ...... 139

SCHOLASTIC VITA ...... 155

iii

LIST OF ILUSTRATIONS AND TABLES FIGURES

CHAPTER 1

Figure 1.1 Common low-molecule-weight thiols found in living organisms. 1

Figure 1.2 Mechanism of oxidoreductase-catalyzed formation of disulfide bonds. 3

Figure 1.3 The common cysteine oxidation states in biological systems. 4

Figure 1.4 Common metal-thiol cofactors found in biological enzymes. 6

Figure 1.5 The catalytic cycle of glutathione reductase. 9

Figure 1.6 Biosynthetic pathway of mycothiol. 12

Figure 1.7 Detoxification pathways of fosfomycin through FosA, FosB and FosX. 21

CHAPTER 2

Figure 2.1 The proposed biosynthetic pathway and detoxification mechanism of 27 bacillithiol.

Figure 2.2 Substrate saturation curves of BA1557, BC1534, BA3888 and BC3461. 39

Figure 2.3 Sequence identity (grey boxes) and similarity (white boxes) between the N- deacetylase enzymes 40

Figure 2.4 Drawgram of deacetylase enzymes 40

Figure 2.5 Far-UV circular dichroism spectra of deacetylases 41

Figure 2.6 Active site region of B. cereus BC1534 (PDB Accession 2IXD). The Zn2+ is 41 coordinated with His12, Asp15 and His113.

Figure 2.7 Amino acid sequence alignment of deacetylase enzymes. 42

Figure 2.8 Deacetylation activity of apo-BA1557 reconstituted with different 43 stoichiometric ratio of Zn2+.

Figure 2.9 Activity pH profiles deacetylase enzymes. 45

Figure 2.10 Activity pH profiles deacetylase activity of BA1557. 46

iv

Figure 2.11 Proposed mechanism for BshB deacetylation reaction 46

Figure 2.12 Model of the active site of BC1534 (PDB code 2IXD) in complex with 49

GlcNAc-Mal.

CHAPTER 3

Figure 3.1 Growth profiles of B. subtilis wild type, ∆bshA, and ∆bshC AmyE:Pxyl-bshC strains in LB and Spizizen’s MM plus leu, Isoleu, Gln and Glu. 70

Figure 3.2 Growth profile of B. subtilis strain lacking BSH in MM. 71

Figure 3.3 Activities of the Fe-S proteins glutamate synthase (GOGAT), aconitase (ACN), isopropyl malate isomerase (LeuCD) and the non-Fe-S protein malate dehydrogenase

(MDH) in cell lysates of B. subtilis wild type and ΔbshA strain. 73

Figure 3.4 Activities of mononuclear iron enzyme threonine dehydrogenase and quercetin 2,3-dioxygenase. 75

Figure 3.5 Growth curves of wild-type B. subtilis and ∆bshA strain in MM with 0.05% casamino acid in presence of various stress challenges. 77

Figure 3.6 Cellular BSH redox status upon metal and superoxide stress. 78

Figure 3.7 Specific activities of ACN (A) and GOGAT (B) under stress conditions. 80

Figure 3.8 Total iron and manganese concentration analysis in wild-type B. subtilis and

∆bshA strain after exposure to 100 µM of paraquat stress. 82

Figure 3.9 MnSOD activity in the cell lysates of CU1065 (wt) and HB11002 (bshA) before and 30 min after the treatment of 100 µM paraquat. 83

Figure 3.10 Western blot analysis of Fe-S biosynthetic enzymes 83

CHAPTER 4

Figure 4.1 De novo and salvage NAD biosynthetic pathways. 92

Figure 4.2 Biosynthetic conversion steps in the formation of quinolinic acid. 94

Figure 4.3 The nicotinic acid de novo gene region. 95

Figure 4.4 In-vitro NadA activation. 96

v

Figure 4.5 NifS cysteine desulfurase reaction. 97

Figure 4.6 Cysteine saturation curves of NifS in terms of alanine and sulfide produced.

105

Figure 4.7 A. pH-activity profile of cysteine desulfurase NifS. 106

Figure 4.8 Cysteine desulfurase activities of NifS under various conditions. 108

Figure 4.9 Far-UV circular dichroism spectra of NadA+NadB, NifS+NadA, NifS+NadB, and NifS+NadA+NadB. 109

Figure 4.10 Analytical gel filtration chromatograms of NifS, NadA, NadB and different combinations of the three enzymes. 111

Figure 4.11 UV-visible absorption spectra of as isolated B. subtilis NadA, after anaerobic treatment with cyanide and EDTA and after reconstitution. 112

Figure 4.12 Reconstitution of apo-NadA in presence of various sulfur sources under reducing and non-reducing condition. 113

Figure 4.13 Reconstitution of apo-NadA in presence of equivalent NadB under reducing

(A) and non-reducing (B) conditions. 114

Figure 4.14 Reconstitution of apo-NadA with Fe and sulfide in presence of 2 mM DTT,

BSH or none. 115

CHAPTER 5

Figure 5.1 Biosynthetic scheme of Fe-S cluster. 120

Figure 5.2 Schematic representations of domain structures of NifU, NfuA, IscA, Nfu, and

IscU proteins from bacteria. Azotabacter vinelandii (Av), E. coli (Ec), B. subtilis (Bs) and

S. aureus (Sa). 122

Figure 5.3 Sequence alignments of B. subtilis Nfu with E.coli NfuA U domain and

A.vinelandii NfuA U domain. 123

Figure 5.4 A. Complementation analysis of B.subtilis Nfu in wild type (MG) and ΔnfuA E. coli strains under oxidative stress and iron starvation conditions. 129

vi

Figure 5.5 Enzymatic activities of GOGAT, ACN, and MDH in MG1655 and ∆NfuA strains containing different plasmids. 130

Figure 5.6 UV-visible absorption spectra of as isolated Nfu, apo-Nfu and reconstituted.

131

Figure 5.7 A. apo-AcnA was incubated with reconstituted Nfu at room temperature under anaerobic conditions, and aliquots from different time points were taken for activity assays as described in Method and Materials. B. apo-AcnA activation as a function of the concentration of reconstituted Nfu. 132

TABLES

Table 2.1 Substrate specificity of B. anthracis and B. cereus N-deacetylases. 38

Table 2.2 Kinetic parameters of B. anthracis and B. cereus N-deacetylases 38

Table 2.3 Activity of BA1557 upon reconstitution with different metals 44

Table 2.4 pH dependence of different enzymes/mutants for N-deacetylation of GlcNac-

Mal 46

Table 5.1 Oligonucleotides used for plasmids construction. 125

vii

LIST OF ABBREVIATIONS

ACN - aconitase

Ala – alanine

BSH - bacillithiol

BSA – bovine serum albumin

CD - circular dichroism

Cys – cysteine

DTT – dithiothreitol

Fe/S – iron-sulfur

GOGAT – glutamate synthase

GSH - glutathione

ICP-OES - Inductively coupled plasma atomic emission spectrometry

ICP-OES - Inductively coupled plasma optical emission spectrometry

LMW – low-molecular-weight

MDH – malate dehydrogenase

MM – minimal medium

MSH - mycothiol

NDA - naphthalene-2, 3-dicarboxaldehyde

PLP – pyridoxal-5’-phosphate

S2-- sulfide

viii

ABSTRACT

Low-molecular-weight (LMW) thiols play vital roles on a variety of cell metabolic pathways including oxidative stress management, xenobiotic detoxification, post- translational modification, and metal homeostasis. Previous studies have identified glutathione (GSH) as the dominant LMW thiol in eukaryotic cells as well as Gram-negative bacteria and mycothiol (MSH) as the dominant LMW thiol in

Actinobacteria. Functional studies on those thiols have revealed their importance in cell metabolism, also their potential as target for antimicrobial development.

Interestingly, neither GSH nor MSH was found in low G+C Gram-positive bacteria, including some notable pathogens, such as B. anthracis, B. cereus and

S. aureus, etc. In 2009, a novel LMW thiol, bacillithiol (BSH), was identified in this class of bacteria and suggested to play analogous function as GSH and MSH.

This dissertation aims at understanding the biosynthesis and function of BSH.

Kinetic assays with the non-commercial natural substrate have been performed on the second enzyme, BshB, in the biosynthetic pathway. The results revealed a general acid-base deacetylation mechanism for BshB enzymes and provided insights on the substrate binding mechanism. The growth experiments on B. subtilis wild type and BSH null strains have identified a slow growth phenotype linked to the decreased activities of Fe-S enzymes. Deficiency of BSH also resulted in decreased levels of intracellular Fe accompanied by increased levels of Mn and altered expression levels of Fe-S cluster biosynthetic Suf components.

The initial results linking BSH to Fe-S metabolism were further explored by studying the involvement of BSH on Fe-S biosynthesis by using cysteine

ix

desulfurase NifS and the 4Fe-4S containing quinolinate synthase NadA as a model. We characterized an Fe-S carrier or/and scaffold Nfu in B. subtilis, and investigated the cooperative relation between Nfu and BSH. Altogether, this dissertation provides the characterization of BSH deacetylase enzymes involved in the biosynthesis and detoxification pathways and describes phenotypic and biochemical analysis of the involvement of BSH in oxidative stress response and

Fe-S metabolism.

x

CHAPTER 1 INTRODUCTION

1.1 Biothiols Sulfur is an essential element for all living organisms. In humans, it is the eighth most abundant element by weight (Parcell, 2002). Sulfur belongs to the same elemental group as oxygen, in which each atom contains six valence electrons.

This property gives sulfur a variety of valence charge from -2 to +6. Thiols in -2 oxidation state contain the most sulfur in cells, which are involved in numerous cellular functions. Biothiols can be divided into large-molecular-weight thiols which are also named protein thiols and low-molecular-weight (LMW) thiols. The common LMW thiols are shown in Figure 1.1.

Figure 1.1 Common low-molecule-weight thiols found in living organisms.

1

1.2 Cysteine

Among the LMW thiols, cysteine (Cys) has been studied most extensively considering its essentiality for cell metabolism. In humans, Cys is synthesized through a five-step pathway using L-methionine and L-serine as starting materials. The biosynthesis starts with the conversion from L-methionine to S- adenosyl-L-methionine which is further demethylated to form S-adenosyl-L- homocysteine. The intermediate S-adenosyl-L-homocysteine then undergoes degradation to produce L-homocysteine which transfers sulfur to L-serine to form

L-Cys (Meier et al., 2001). The concentrations of free Cys in mammalian cells are maintained at µM range to meet the needs of protein synthesis and the production of other essential molecules that include glutathione, coenzyme A, taurine, and inorganic sulfur (Brigham et al., 1960; Jones et al., 2002; Stipanuk et al., 1992). The interplay between Cys and other metabolites makes the regulation of cysteine metabolism extremely complicated (Feldman-Salit et al.,

2012; Stipanuk et al., 2006).

In plants and bacteria, the pathway of Cys biosynthesis is accomplished through a two-step conversion (Bonner et al., 2005; Wirtz and Hell, 2003). The first step is catalyzed by serine acetyltransferase to acetylate the hydroxyl group on L- serine to obtain O-acetyl-L-serine (Olsen et al., 2004). The second step involves the reaction between hydrogen sulfide and O-acetyl-L-serine to make final product Cys and release acetate as side product (Rabeh and Cook, 2004). The biosynthetic pathway is strictly regulated through a feedback inhibition by Cys on

2

step one (Olsen et al., 2004) as well as a regulatory effect at the genetic level on enzymes in step-two reaction (Hulanicka et al., 1979).

The thiol group on Cys exhibits a pKa of 8.5 (Clement and Hartz, 1971), which indicates most Cys molecules are in protonated state under physiological conditions (Vincent E. Bower et al., 1961). Upon oxidation, a reversible reaction occurs with two Cys molecules to form one cystine. The reversible conversion from Cys to cystine is a formal two-electron reaction. Biologically, molecular oxygen, nicotinamide, flavin cofactors and other disulfides participate as electron donors or acceptors in Cys conversion. In normal human plasma, the redox ratio of Cys/cystine is around 0.1-0.25 while bacteria have a significantly higher ratio fluctuating in the range of 1-10 (Crawhall and Segal, 1967; Parsonage et al.,

2010). The formation of disulfide bonds in proteins can be an enzyme catalyzed

(Fass, 2012). Proteins that are capable of catalyzing protein disulfide bond formation can be divided into two classes: oxidoreductases and nonthioredoxin- like enzymes (Sevier and Kaiser, 2002). Oxidoreductases belong to the thioredoxin superfamily, these enzymes contain a signature CXXC motif

(Kouwen et al., 2008; Pedone et al., 2004). In vivo, oxidoreductases rapidly oxidize proteins by donating their disulfides to pairs of cysteine in substrate proteins through a thiol-disulfide exchange reaction as shown in Figure 1.2.

Figure 1.2 Mechanism of oxidoreductase-catalyzed formation of disulfide bonds.

3

In an exchange reaction, a thiolate anion from substrate Cys thiol attacks one sulfur of the disulfide bond in oxidoreductase to form a bridge disulfide bond between the two proteins. The remaining thiolate in substrate protein then attacks the bridge disulfide bond to form an intracellular disulfide bond while releasing oxidoreductase as free thiol (Araki and Inaba, 2012; Kofoed et al.,

2011) . Protein Cys disulfide formation is found in diverse bioprocesses such as catalysis, protection against oxidative stress, the regulation of biological activity and the stabilization of proteins (Fass, 2012; Hogg, 2003).

Reactive oxygen species (ROS) are natural products of the normal cell metabolism (Ray et al., 2012). However, excess ROS can overwhelm the cellular antioxidant defense system and cause damage to nucleic acids, proteins, and lipids (Cabiscol et al., 2000; Cooke et al., 2003; Halliwell, 1991). In recent years, a new role for ROS has emerged as its redox-mediated modifications on Cys residues as shown in Figure 1.3.

Figure 1.3 The common cysteine oxidation states in biological systems.

These modifications along with the changes in redox status of Cys residues regulate the initiation of signal transduction pathways and the induction of gene

4

expression (Green and Paget, 2004; Ray et al., 2012). In the presence of oxidative stress, Cys residues undergo single oxidation into sulfenic acid (-SOH) which can form disulfide bonds with nearby Cys residues or undergo further oxidation to sulfinic (-SO2H) or sulfonic (-SO3H) acid (Green and Paget, 2004; Lo

Conte and Carroll, 2013). When the Cys residues are in a lower oxidation state than sulfinic acid, these redox modifications are reversible by reducing systems such as thioredoxin and peroxiredoxin.

An additional property of Cys is its unique metal binding affinity. Metal thiolate bonds are present in the active sites of many metalloproteins and are essential for their biological functions (Dance, 1986; Giles et al., 2003). For example, iron- sulfur enzymes utilize multiple Cys residues as the ligands to stabilize iron-sulfur center for structure or catalysis purpose (Johnson et al., 2005a). The common metal thiolate coordination in metalloprotein sites as revealed by X-ray diffraction are listed in Figure 1.4. The flexibility of Cys redox state also plays an important role on the regulation of metal-thiol binding affinity which further alters the function of metalloproteins. Metallothionine (MT) is a major zinc binding protein which is involved in the zinc storage and transfer. The mammalian MT consists of about 60 amino acids, of which 20 are Cys residues. A refined crystal structure revealed that these residues tightly bind seven zinc atoms in a Zn3Cys9 and a

Zn4Cys11 cluster (Robbins et al., 1991). At physiological pH, the dissociated constant of Zn in MT was estimated to be as high as 5 × 1012 M, which indicates that the metal transfer from MT to apo-proteins is rather inefficient. However, due to the extreme abundance of Cys residues, the Zn/sulfur clusters are redox

5

sensitive, and ligand oxidation results in Zn release. A number of oxidants such as cystine, cystamine, selenocystamine, and glutathione disulfide were suggested to participate in this process (Maret and Vallee, 1998).

Figure 1.4 Common metal-thiol cofactors found in biological enzymes.

1.2 Glutathione

In the late 19th century, another thiol compound named “philothion” was found widely distributed in yeast and mammalian cells (Rey-Pailhade, 1888a, b). With the pioneering work from many laboratories over nearly 50 years, the structure of this sulfur compound was determined to be a tripeptide consisting of L-Cys,

6

glycine and L-glutamate. Since then this thiol compound was named as glutathione (GSH) based on its components. GSH is present in almost all eukaryotes with the exception of those that do not have mitochondria or chloroplasts, but its production among prokaryotes is limited to Gram-negative species as well as a few strains of Gram-positive bacteria (Fahey et al., 1978).

Despite the differences of GSH concentration in different organisms, the biosynthesis of GSH shares a unique two-step pathway (Lu, 2013). In the first step, L-Cys and L-Glu are ligated together through a conjugation reaction under the catalysis of ATP-dependent enzyme γ-glutamylcysteine synthetase (GCS) to form γ-glutamylcysteine. GCS is composed of a large catalytic subunit and a small modifier subunit, which are encoded by a single gene in bacteria and yeast.

The catalytic subunit is the only domain involved in the ligation reaction, which is under the feedback inhibition of GSH at mM concentration (Seelig et al., 1984).

Even though the catalytic subunit itself can catalyze the reaction, the KM of L- glutamate is significantly higher than the physiological concentration (Huang et al., 1993a). Subsequent studies revealed that the small modifier subunit could lower the KM of GCS and raise the Ki (inhibition constant) of GSH (Huang et al.,

1993b). The second step of GSH biosynthesis involves the addition of glycine onto the carboxyl group of cysteine, resulting in the formation of GSH

(Oppenheimer et al., 1979). The addition reaction is catalyzed by GSH synthase, which is highly specific for γ-glutamylcysteine. Under normal growth conditions,

GSH synthase does not show significant involvement on the regulation of GSH synthesis (Oppenheimer et al., 1979). However, abnormal GSH synthase activity

7

and expression have been observed in the presence of stressful or pathological conditions (Lu, 2000). Besides the biosynthetic pathway, GSH can also be acquired through GSH transporters (Bachhawat et al., 2013).

The physiological concentration of GSH ranges from 0.1 to about 10 mM in bacteria, while it is present in mammalian tissues at 1–10 mM concentration

(Fahey et al., 1978). GSH mainly exists in the reduced form and its disulfide

(GSSG) form, and the cellular ratio of GSH/GSSG is normally around 100-400

(Hwang et al., 1992). The GSH/GSSG ratio changed in response to nutritional status, drugs, and especially oxidants. In the presence of oxidative stress, the

GSH/GSSG ratio may fall into 1-10 range (Chai et al., 1994), and the enormous level of GSSG will be converted back to GSH by GSSG reductase (GR) to maintain normal reducing thiol pool. GR is an oxidoreductase-family enzyme

(Deponte, 2013), which is conserved in all living organisms. The reduction of

GSSG by GR is achieved through a four-step reaction as shown in Figure 1.5

(Pai and Schulz, 1983).

The first step involves the reduction of FAD by NADPH to form a transient FADH anion. The FADH anion then reduces the conserved Cys disulfide bond in GR to release one reduced Cys and forms a complex with the other Cys. In the third step, GSSG enters into the active site and undergoes thiol-disulfide exchange with the reduced Cys to form a Cys-SG bond while releasing one reduced GSH.

Finally, the Cys-SG bond is attacked by the Cys to release another reduced glutathione and reform the redox active Cys disulfide (Deponte, 2013; Pai and

Schulz, 1983).

8

FADH- Figure 1.5 The catalytic cycle of glutathione reductase.

Besides the convertible oxidation into GSSG, GSH also serves as the substrate of biochemical reactions to protect cells from oxidative stress. Dehydroascorbate reductase is a key component of the glutathione-ascorbate cycle, which utilizes

GSH as substrate to produce ascorbate to detoxify H2O2 (Xu et al., 1996). In addition, GSH cooperates with GSH peroxidases to reduce lipid hydroperoxides to their corresponding alcohols and to reduce free hydrogen peroxide to water

(Carmagnol et al., 1983). GSH S-transferases (GSTs) play a vital role against oxidative stress by conjugating GSH with a range of harmful electrophiles, including 2-alkenals acrolein, crotonaldehyde, cholesterol-5,6-oxide, and epoxyeicosatrienoic acid. (Allocati et al., 2009; Vuilleumier, 1997). The conjugated compounds are then processed through the mercapuric acid pathway to excrete electrophiles from the cells (Hinchman et al., 1991). In the past fifty years, GSTs have been extensively studied and a wide range of substrates for

GSTs has been found including antibiotics, industrial intermediates, (Townsend

9

and Tew, 2003) and environmental chemicals (Habig et al., 1974; Hinchman et al., 1991). Although best known for their ability to conjugate xenobiotics to GSH and thereby detoxify cellular environments, GSTs are also capable of binding nonsubstrate ligands to regulate cell signaling (Ketley et al., 1975).

Other than the important roles of GSH on detoxifying oxidants and xenobiotics,

GSH is also involved in chemical modification of protein cysteine residues

(Cooper et al., 2011; Dalle-Donne et al., 2007; Gallogly and Mieyal, 2007).

Protein S-glutathionylation (PSSG) can occur either by thiol–disulfide exchange between oxidized glutathione (GSSG) and a protein Cys residue or by reduction of oxidized protein Cys residue utilizing reduced GSH (Grek et al., 2013).

Although several mechanisms can account for the formation of PSSG, the regulations as well as the specificity of PSSG are still far from being understood.

Through proteomic studies, a list of proteins shown to be glutathionylated has been identified under different conditions. The cellular responses of those modifications have not been investigated thoroughly, but it has been suggested the involvement of S-glutathionylation on redox-regulation, energy metabolism, cellular signaling, calcium homeostasis, protein folding and stability, etc (Rouhier et al., 2008; Zaffagnini et al., 2012).

1.3 Mycothiol

The importance of GSH on redox regulation and other cell metabolism has extended the studies of this LMW thiol in many organisms. However, GSH is absent in most Gram-positive bacteria lacking this thiol, for example,

10

actinobacteria whose members distribute in a wide range of ecosystems, from soil and seawater to the skin, lungs, and gastrointestinal tract of humans. It was hypothesized that other LMW thiols may present in those species and play analogous function as GSH. The monobromobimane (mBBr) -based thiol analysis in Streptomycetes identified a novel thiol in the high-performance liquid chromatography profile (Newton et al., 1996). The following studies determined that the structure of this unknown thiol involves a cysteine residue in which the amino group is acetylated and the carboxyl group is linked to 1-myo-inositol-D- glucosamine through an amide bond (MissetSmits et al., 1997). The common name was proposed as mycothiol (MSH) since it was present in all mycobacteria to date. The intracellular concentration of MSH varies from 0.015 mM to 10 mM

(Newton et al., 1996), which is very close to that of GSH. Like GSH, MSH mainly exists in its reduced form and its disulfide state (MSSM), and the ratio of

MSH/MSSM is usually around 100:1 (Ung and Av-Gay, 2006). The highly reduced status of MSH is maintained through mycothione reductase (MR) which also belongs to the oxidoreductase family (Patel and Blanchard, 1999). The reduction mechanism of MR is similar to GR, in which NADPH is utilized as electron donor and an intracellular disulfide bond is used for the sequential reduction of MSSM (Patel and Blanchard, 1999, 2001). The remarkable structural difference between MSH and GSH implies that MSH may be synthesized through a distinct pathway. With the pioneering work from several laboratories, the biosynthesis pathway of MSH was determined to be a five-step

11

process including four enzymes, MshA, MshB, MshC and MshD as illustrated in

Figure 1.6 (Newton et al., 2008; Newton et al., 2006a).

Figure 1.6 Biosynthetic pathway of mycothiol.

The first step of MSH biosynthesis is catalyzed by MshA which ligates the 1-L- myo-inositol-1-phosphate moiety to the carbon 1 position of UDP–N- acetylglucosamine to generate 3-phospho-1-D-myo-inosityl-2-acetamido-2- deoxy-a-D-glucopyranoside (GlcNAc-Ins-P) and release UDP molecule (Newton et al., 2006b). MshA belongs to the GT-4 family of glycosyltransferase enzymes, in which most are directly involved with cell wall biosynthesis (Vetting et al.,

2008). The deletion of mshA completely depleted the MSH level in M. smegmatis, which further confirms the involvement of MshA in the biosynthesis of MSH (Xu et al., 2011). Kinetic studies suggested that MshA utilizes an ordered sequential mechanism with UDP–GlcNAc binding first, followed by 1L-Ins-1-P (Vetting et al.,

2008). Before the reaction of MshB, GlcNAc-Ins-P needs to be dephosphorylated to form GlcNAc-Ins. In M. tuberculosis, multiple copies of inositol monophosphatases have shown activities using several phosphate containing

12

substrates (Nigou et al., 2002). Since no single inositol monophosphatase mutant has been shown to be MSH-depleted, it was concluded that more than one enzyme can catalyze the dephosphorylation (Vetting et al., 2008). The first MshB was identified in M. tuberculosis and shown to be a homolog of the MSH S- conjugate amidase (Mca) (Newton et al., 2000). The enzymatic assays with

GlcNAc-Ins as substrate have been performed with purified M.tuberculosis MshB

-1 -1 which showed an efficient catalytic capacity with a kcat/KM of 1440+360 M s

(Newton et al., 2006a). However, the deletion of mshB did not deplete the level of

MSH, for instance, M. smegmatis ∆mshB strain showed a 10-fold decrease of

MSH with an enormous accumulation of GlcNAc-Ins (Rawat et al., 2003). This observation suggested other deacetylase(s) may participate in the biosynthesis of MSH. Mycothiol amidse (Mca) is the most likely alternative MshB enzyme since it shows high sequence similarity to MshB on amino acid sequence. The

Mca from M. tuberculosis showed a 30,000-fold lower catalytic activity than MshB on the deacetylation of GlcNAc-Ins (Steffek et al., 2003). A M. smegmatis

∆mshB/∆mca strain was constructed and showed no detectable MSH (Xu et al.,

2011). These observations suggested the involvement of Mca as a backup MshB enzyme. The structure of MshB from M. tuberculosis was solved with a α/β fold

C-term domain and a lactase dehydrogenase-like N-terminal domain (Maynes et al., 2003). Structural studies also revealed a Zn2+ ion coordinated with 2 conserved His and 1 conserved Asp (Maynes et al., 2003). MshB was completely inactivated by chelator 1,10-phenathroline (Newton et al., 2006a), and the apo-

MshB could be reactivated after incubating with other metals following an overall

13

trend Fe2+>Co2+>Zn2+>Mn2+>Ni2+>Fe3+(Huang et al., 2011). Among all the metal cofactors, Zn showed the highest binding affinity, however, MshB may prefer

Fe2+ under anaerobic conditions (Huang et al., 2011). A general acid-base mechanism was proposed for MshB based on its similar active site to other metal-dependent hydrolases. The catalysis of MshB on the commercial substrate

GlcNAc exhibits a bell-shaped dependence on pH, indicating that the occurrence of two ionization events in the deacetylation reaction (Huang et al., 2011b).

These observations are consistent with MshB proceeding through general acid- base mechanism, in which the conserved Asp15 on the top of the Zn center was suggested as a general base for water activation corresponding to the first ionization. Mutation of Asp15 to Ala resulted in the loss of the first pKa, confirming the role of Asp15 as general base. The initial candidates of general acid were suggested to be the adjacent protonated Asp15 or protonated His144, however, neither mutation on these residues eliminated the second ionization.

Surprisingly, mutation of Tyr142 caused the loss of the second pKa, indicating that Tyr142 acts as the general acid for MshB catalysis (Huang and Hernick,

2012).

MshC catalyzes the third step of MSH biosynthesis by forming an amide bond between L-Cys and GlcN-Ins. MshC is an ATP-dependent ligase which shares significant sequence and structural similarities with the cysteinyl-tRNA synthetase (Newberry et al., 2002). The kinetic studies were performed with

-1 purified M.smegmatis MshC which exhibits a kcat value of 3.52 s and KM values of 1.84 + 0.06 mM, 0.10 + 0.01 mM and 0.16 + 0.05 mM for ATP, Cys and GlcN-

14

Ins, respectively (Fan et al., 2007). Initial attempts to construct a M. tuberculosis mshC knockout failed, suggesting that mshC is essential under laboratory conditions (Sareen et al., 2003). The deletion of bshC was successfully achieved in Mycobacterium smegmatis mc(2)155, and the mutant strain did not show a a detectable amount of MSH. The last step of MSH biosynthesis involves the transfer of an acetyl group from acetyl-CoA to Cys-Gln-Ins to form the final product MSH. MshD belongs to GCN5-related N-acetyltransferases family, in which most enzymes occur as dimer (Dyda et al., 2000). Both GCN5-related N- acetyltransferase domains in MshD have shown acetyl-CoA binding affinity; however, only C-terminal domain is biologically active while the N-terminal domain was suggested to be structural (Vetting et al., 2003; Vetting et al., 2006).

The regulation of MSH biosynthesis is still not clear. A variety of growth conditions, such as oxidative stress, low oxygen and growth in macrophage have been investigated in M. tuberculosis (Schnappinger et al., 2003; Voskuil et al.,

2003; Voskuil et al., 2004), however, no significant change of mRNA levels were observed for any MSH biosynthetic genes. This observation suggested that the conditions tested may have not been strong enough to trigger the response or the regulation of MSH biosynthesis undergoes a novel stress-response mechanism. Even though the regulation of MSH in mycobacteria is still not understood, the studies in Streptomeyces coelicolor have suggested a sigma factor linked to up-regulation of mshA, mshC and mshD genes (Lee et al., 2005).

The activity of sigma factor is under the control of RsrA which responses to the thiol-disulfide balance. A disulfide in RsrA undergoes reduction by intracellular

15

free thiol to generate a reduced RsrA that further interacts with a zinc ion to form an RsrA-zinc complex (Park and Roe, 2008). The complex binds with sigma factor σR preventing transcription activation of this associated target genes including msh genes. The deletion of sigma factor in S. coelicolor only caused a modest lower expression of msh genes, which suggests the presence of another regulatory mechanism.

A series of MSH mutants have been constructed to study its in-vivo function and a wide array of phenotypes has been identified. In M. smegamatis, mshC mutant strains were successfully produced by both chemical mutagenesis and transposon mutagenesis. The transposon mutated mshC strains displayed an increased sensitivity to alkylating reagents including mBBr, iodoacetamide, chlorodinitrobenzene, and diamide (Newton et al., 1999). Similar effects were also observed under oxidative stress, especially under hydrogen peroxide. The depletion of MSH blocked the MSH-dependent detoxification pathway, which resulted in an increased sensitivity to a list of antibiotics including erythromycin, azithromycin, vancomycin, and penicillin G (Rawat et al., 2002; Rawat et al.,

2007). Examination of drugs used in the treatment of mycobacteria infection also revealed that transposon mutants are more sensitive to rifamycin and rifampin

(Rawat et al., 2007). However, no difference on resistance was observed on ethambutol or pyrazinamide (Rawat et al., 2002). The specificity of MSH on antibiotics and drug detoxification pathways as well as the associated mechanisms of these reactions still need further investigation.

Protein S-glutathionylation serves as one of the major post translation

16

modifications in GSH-containing species. However, limited research has been conducted on studying protein S-mycothiolation, and it seems that mycothiolation is not as common as glutationlation overall. Recently, protein S-mycothiolation was studied in Corynebacterium glutamicum through transcriptomic studies under NaOCl challenge. The results identified 25 mycothiol-modified proteins including methionine synthase (MetE), enzymes for the biosynthesis of nucleotides (GuaB1, GuaB2, PurL, NadC) and thiamine (ThiD), translation proteins, and antioxidant enzymes (Tpx, Gpx, MsrA) (Chi et al., 2014).

1.4 Bacillithiol

In some low-(G+C)-content Gram-positive bacteria, such as Bacillus species, neither GSH nor MSH was detected which led to the idea that another LMW thiol(s) could serve in this capacity. In Bacillus, cysteine concentration is highcompared with other bacteria, leading to the notion that cysteine could act as the role LMW thiol in this organism. It has been suggested that cysteine could protect protein thiols by forming reversible disulfide bond under oxidative stress and it could also function for repairing damaged Fe-S enzymes. However, considering the reduction potential of cysteine (-223 mV) along with low cellular cysteine/cystine ratio, it is unlikely for cysteine to work as the major redox buffer.

In addition to cysteine, Coenzyme A (CoA) is also present in high concentrations in Firmicutes species, especially in B. anthracis and Staphylococcus aureus

(Newton et al., 1996). Early research in B. megaterium showed that CoA is able to thiolate protein thiols. However, CoA is also an unlikely candidate since it

17

cannot serve as a reservoir for regeneration of cysteine, and in cellular environments is mostly conjugated with acetate as acetyl-CoA.

In 2009, three groups have simultaneously identified a 398-Da thiol compound, and subsequent thiol analysis through monobromobimane labeling in

Deinococcus radiodurans allowed its isolation and characterization (Newton et al.,

2009). Similar molecules were also reported in B. anthracis and B. subtilis with the same molecular weight. The hydrolysis of this thiol molecule released L- malate, L-Cys and D-glucosamine. The structure of this thiol includes 1D-L-

Malate-α-D-glucopyranoside, in which the hydroxyl group at the 2-position of the glucose moiety is replaced by an (L-cysteinyl)amido group (Figure 1.1). Since the initial analysis of all seven Bacillus strains indicated significant amounts of this LMW thiol, bacillithiol was proposed as a common name with an abbreviation.

The biosynthesis of BSH was proposed to be a three-step pathway involving a glycosyltransferase (BshA), a deacetylase (BshB) and a cysteine ligase (BshC).

Initial bioinformatic analysis in B. subtilis revealed five candidates with a significant similarity to glycosyltransferases (Gaballa et al., 2010). One noticeable candidate is ypjH next to ypjG encoding an MshB homolog. The deletion of ypjH completely depleted the intracellular BSH pool (Gaballa et al., 2010), and the kinetic studies with purified YpjH (BshA) enzyme showed KM and Vmax values of

410 µM and 11.2 µmol·min-1·mg-1 for L-malate and 370 µM and 8.7 µmol·min-

1·mg-1 for UDP-GlcNAc (Upton et al., 2012). The crystal structure of BshA from B. anthracis was solved and an overlay with the MshA crystal structure demonstrates a conserved active site (Parsonage et al., 2010). Besides the

18

natural substrate L-malate and UDP-GlcNAc, a variety of structurally similar compounds have been tested on BshA and no significant activity was observed

(Upton et al., 2012). Moreover, BshA was shown to be feedback inhibited by

BSH with an IC50 of 0.7 mM, which is within the biological level of BSH (Upton et al., 2012).

The second enzyme in the BSH biosynthetic pathway was proposed to be the adjacent gene of bshA. In B. subtilis, YpjG shares a 62% identical amino acid sequence as a known deacetylase BcZBP in B. cereus (Gaballa et al., 2010).

The orthologous enzyme of YpjG in B. anthracis (named BA1557) was purified

-1 and showed a KM of 0.16 mM and kcat of 42 s using GlcNAc-malate as substrate

(Parsonage et al., 2010). However, the deletion of ypjG only showed a 60% decrease of BSH level, similar to the deletion of mshB in M. smegmatis (Gaballa et al., 2010; Rawat et al., 2003). Further investigation revealed another protein

YojG as the BshB2 enzyme and a double mutation of ypjG and yojG resulted in no detectable BSH (Gaballa et al., 2010).

The last step was hypothesized to be the ligation of cysteine on the free amine group based on the structure difference between GlcN-malate and BSH. MshC belongs to a class of tRNA-cysteinyl synthetase (CysS) (Fan et al., 2007), suggesting BshC may also be a homolog of CysS. However, BLAST search against B. subtilis failed to identify any CysS homolog, and an S. aureus CysS enzyme did not show any significant activity on BshC reaction (Gaballa et al.,

2010) . These observations implied that the BshC activity may be catalyzed by an uncharacterized enzyme and informatic analysis was able to identify a co-

19

occurrence of bshA and bshB genes to a gene named yllA in B. subtilis. Deletion of yllA eliminated BSH while causing an accumulation of GlcN-Malate (Gaballa et al., 2010). The structure of YllA was reported recently and showed a distinct structure fold compared to MshC or other CysS enzymes (VanDuinen et al.,

2015). Despite strong evidence pointing to the role of YllA as the BshC enzyme, no reports to date have shown biochemical evidence of this reaction.

The intracellular ratio of BSH/BSSB was measured to be around 50-400 in different species (Parsonage et al., 2010; Rajkarnikar et al., 2013), which is similar to GSH/GSSG and MSH/MSSM. It is anticipated that BSH undergoes oxidation to form BSSB in presence of oxidative stress. This was confirmed by the challenge of B. subtilis culture with paraquat stress, showing an increase of

BSSB by five-fold (described in Chapter 3). To recycle the BSSB back into BSH as redox buffer, an enzyme similar to GSH reductase or MSH reductase may be necessary. Several enzymes have been proposed as BSH reductase in B. subtilis (Gaballa et al., 2010), but the identity as well as reaction mechanism are still under investigation. The thiol-disulfide redox potential of BSH was determined through a coupled thiol/disulfide equilibrium reaction with GSH

(Sharma et al., 2013). As estimated, the redox potential of BSH was -221 mV and higher than those reported for Cys (-223 mV) and CoA (-234 mV). The pKa of the thiol group on BSH was determined to be around 7.97, significantly lower than those of Cys and CoA (Sharma et al., 2013). Considering the low pKa as well as dominant concentration of BSH, it is anticipated that BSH is the preferred

LMW thiol as the reducing agent.

20

BSH deletion strains have been constructed in many species to study the function of this thiol under normal as well as stress conditions. In B. subtilis, the zone inhibition assays on wild type and ∆bshA strain showed undistinguishable growth under a variety of conditions including H2O2, diamide, mBBr, etc (Gaballa et al., 2010). A dramatic increase in fosfomycin sensitivity was observed in

∆bshA strain (Gaballa et al., 2010). Fosfomycin inhibits bacterial cell wall biogenesis by inactivating the enzyme UDP-N-acetylglucosamine-3- enolpyruvyltransferase, also known as MurA (Kahan et al., 1974). Several enzymes including FosA, FosX in Gram (-) and FosB in Gram (+) bacteria have been associated with detoxification of fosfomycin through nucleophilic attack on carbon 1 of fosfomycin, which opens the epoxide ring and renders the drug ineffective (Figure 1.7) (Rigsby et al., 2005). Purified FosB from S.aureus and B. subtilis exhibited a preference for BSH as the thiol substrate to react with fosfomycin, suggesting the involvement of BSH on detoxification of fosfomycin

(Lamers et al., 2012; Roberts et al., 2013).

Figure 1.7 Detoxification pathways of fosfomycin through FosA, FosB and FosX.

21

The deletion of BSH in S. sureus displayed an increased sensitivity to alkylating stress, oxidative stress, and metal stress (Rajkarnikar et al., 2013). Moreover, the addition of BsmB in wild type and ∆bshB S. aureus revealed a BshB (Bca)- catalyzed mercapturic acid detoxification pathway (Rajkarnikar et al., 2013).

Even though the deletion of BSH did not cause any growth defect in LB medium

(Gaballa et al., 2010), a slow growth phenotype of the BSH null strain in minimal media was observed, which could be recovered upon the addition of selected amino acids (Leu/Ile or Glu/Gln), supplementation of iron, or chemical complementation with BSH disulfide (BSSB) to the growth medium (described in

Chapter 3). Interestingly, the Fe-S cluster containing enzymes isopropylmalate isomerase (LeuCD) and glutamate synthase (GOGAT) showed decreased activities in the BSH null strain.

Protein glutathionylation is a common post translation modification in GSH- containing species. In the species containing BSH as the unique LMW thiol, it is likely that protein thiols undergo thiolation with BSH to form a disulfide modification. Transcriptomic and redox proteomic analysis have revealed S- bacillithiolation of the OhrR repressor, two methionine synthases MetE and YxjG, the inorganic pyrophosphatase PpaC, and the 3-D-phosphoglycerate dehydrogenase SerA as major protection mechanisms against hypochlorite stress in B. subtilis (Chi et al., 2013a). Further studies have identified two bacillithioredoxin candidates able to reduce the disulfide in the bacillithiolation sites on MetE and OhrR (Gaballa et al., 2014). However, the extents of bacillithiolation as well as the functions are still far from being understood.

22

Upon the determination of BSH structure, the malate group of BSH was suggested as a potential intracellular metal chelator (Newton et al., 2009). A recent report has indicated that BSH serves as a major buffer of the labile zinc pool (Ma et al., 2014). Cells lacking BSH are more sensitive to zinc stress, and they induce zinc efflux at lower external zinc concentrations. Deficiency of BSH also resulted in decreased levels of intracellular Fe2+ accompanied by increased levels of Mn2+ under paraquat stress (described in Chapter 3).

Since the discovery of BSH as the LMW thiol in low G+C Gram-positive bacteria, it has been proposed as a novel target for the treatment of notable pathogens, such as B. anthracis, B. cereus and S. aureus. Initial studies including the results in this dissertation have revealed its involvement on oxidative stress, detoxification, metal homeostasis, and post-translational modification.

This dissertation aims to explore the biosynthesis and function of BSH in Gram- positive species. This study describes the biochemical characterization of the

BshB enzymes as well as the Bca enzyme participating in the mercapturic acid detoxification pathway (described in Chapter 2). Mechanistic studies on those enzymes also revealed a general acid-base mechanism for the deacetylation reaction. The studies led to the investigation of the roles of BSH which are described in Chapter 3. Depletion of BSH caused a slow grow phenotype in minimal medium, which has been linked to the decreased Fe-S enzymatic activities involved in branched-chain amino acid biosynthesis. The initial results linking BSH to Fe-S metabolism stimulates an extensive research on the involvement of BSH on Fe-S biosynthesis by using cysteine desulfurase NifS and

23

the 4Fe-4S containing quinolinate synthase NadA as a model in Chapter 4. The cysteine desulfurase and Fe-S cluster assembly assays were performed with organic synthesized BSH. Lastly, we characterized an Fe-S carrier (or/and scaffold) Nfu in B. subtilis, and investigated the cooperative relation between Nfu and BSH described in Chapter 5. These results are in agreement with a recent report describing the involvement of Nfu as in S. aureus (Mashruwala et al.,

2015). Altogether, this study provides the characterization of BSH deacetylase enzymes involved in the biosynthesis and detoxification pathways and describes phenotypic and biochemical analysis of the involvement of BSH in oxidative stress response and Fe-S metabolism.

24

CHAPTER 2

Cross-functionalities of Bacillus deacetylases involved in bacillithiol

biosynthesis and bacillithiol-S-conjugate detoxification pathways

Zhong Fang*, Alexandra A. Roberts†, Karissa Weidman*, Sunil V. Sharma†, Al

Claiborne‡, Christopher J. Hamilton† and Patricia C. Dos Santos*

The work contained in this chapter was published on Biochemical Journal (2013) 454 (239–

247). Zhong Fang, Alexandra Roberts, Karissa Weidman and Sunil Sharma performed experiments. Sunil Sharma has synthesized the substrates used in the kinetic analysis,

Alexandra Roberts performed kinetic assays with BSH and BSSB, Karissa Weidman assisted with purification of B. cereus deacetylases, and Zhong Fang performed all the expereiments described in this chapter. Al Claiborne and Patricia Dos Santos conceived the idea, and provided overall direction. Zhong Fang, Christopher Hamilton and Patricia Dos

Santos planned experiments, analysed data and wrote the paper.

25

Abstract

BshB, a key enzyme in bacillithiol biosynthesis, hydrolyses the acetyl group from

N-acetyl-glucosamine malate to generate glucosamine malate. In Bacillus anthracis, BA1557 has been identified as the N-acetyl-glucosamine malate deacetylase (BshB); however, a high content of bacillithiol (~70%) was still observed in the B. anthracis ∆BA1557 strain. Genomic analysis led to the proposal that another deacetylase could exhibit cross-functionality in bacillithiol biosynthesis. In the present study, BA1557, its paralogue BA3888 and orthologous Bacillus cereus enzymes BC1534 and BC3461 have been characterized for their deacetylase activity towards N-acetyl-glucosamine malate, thus providing biochemical evidence for this proposal. In addition, the involvement of deacetylase enzymes is also expected in bacillithiol-detoxifying pathways through formation of S-mercapturic adducts. The kinetic analysis of bacillithiol-S-bimane conjugate favours the involvement of BA3888 as the B. anthracis bacillithiol-S-conjugate amidase (Bca). The high degree of specificity of this group of enzymes for its physiological substrate, along with their similar pH- activity profile and Zn2+-dependent catalytic acid-base reaction provides further evidence for their cross-functionalities.

26

Introduction

The similarities on the modular composition of BSH and MSH suggested that their biosynthetic schemes would involve a sequence of similar chemical reactions catalyzed by enzymes displaying conservation in their amino acid sequences and structures. Supporting this proposal, recent reports demonstrated that the first step on BSH biosynthesis uses a glycosyltransferase, BshA, to promote the transfer of N-acetyl-glucosamine (GlcNAc) from UDP-GlcNAc to the

2-OH position of L-malate. The catalytic mechanism is similar to that for MshA to ligate inositol-1-phosphate to GlcNAc. The product of this first reaction, GlcNAc- malate, is the substrate of the second biosynthetic step, catalyzed by the deacetylase BshB. This reaction leads to activation of the amino group in the glucosamine moiety enabling subsequent ligation to the carboxyl-moiety of cysteine that requires a not-yet-characterized protein BshC (Figure 2.1).

Figure 2.1 The proposed biosynthetic pathway and detoxification mechanism of bacillithiol.

27

In many Bacilli (e.g. B. subtilis, B. anthracis, B. cereus) bshA and bshB are adjacent genes located in the same operon. Whilst BshA and BshC are essential for BSH biosynthesis, in many cases bshB knockout mutants display only a partial reduction in BSH levels (Gaballa et al., 2010; Parsonage et al., 2010).

This has been attributed to the presence of one, or more, additional bshB-like gene products, which can also N-deacetylate GlcNAc-Mal. This has already been demonstrated for B. subtilis where a ΔbshB1 (ypjG) mutant still retains almost half of the wild type levels of BSH, whereas the ΔbshB1 + ΔbshB2 (yojG) double knockout strain showed no detectable levels of BSH. A similar phenotype is observed for the analogous N-deacetylase in MSH biosynthesis (MshB) where 5-

10% of wild type MSH levels are still produced in ΔmshB knockout mutants due to the residing background N-deacetylase activity of Mca (Rawat et al., 2003;

Steffek et al., 2003). Interestingly, in S. aureus, which only contains a single bshB-like gene, preliminary studies have indicated the presence of a bacillithiol-

S-conjugate amidase (Bca) pathway which is capable of detoxifying the electrophilic xenobiotic monobromobimane. Considering all these in vivo observations, it seems plausible that one or more of these BshB-like enzymes could also present Bca activity in xenobiotic detoxification. In S. aureus, the N- deacetylase activity with GlcNAcMal and the amidase activity with BSmB are most likely performed by the same enzyme, although no kinetic studies of this enzyme have yet been reported. As yet, the relative N-deacetylase (BshB) and amidase (Bca) capabilities of the redundant BshB enzymes within other Bacilli

28

have also not been explored.

Prior to the discovery of BSH, the crystal structure of a B. cereus zinc-binding protein (BC1534) (herein shown to be a BshB) was reported (Fadouloglou et al.,

2007), which was characterized as a potential chitin-deacetylase based on its

-1 observed N-deacetylase activity with GlcNAc (KM = 3 μM, kcat = 2 s ) and

-1 chitobiose (KM = 3 μM, kcat = 98 s ) (Deli et al., 2010). Since then, initial kinetic studies of the homologous (97% sequence identity) B. anthracis enzyme

(BA1557) have demonstrated its BshB activity with GlcNAc-Mal, but no notable

GlcNAc deacetylase activity was observed (Gaballa et al., 2010; Parsonage et al.,

2010).

So far, BSH-deficient mutants have been shown to display impaired sporulation, sensitivity to acid and salt, increased sensitivity to fosfomycin (Gaballa et al.,

2010; Pothera et al., 2013) and reduced viability in mouse macrophage cell lines.

A detailed understanding of the enzymes that mediate BSH biosynthesis and

BSH-dependent drug resistance is central to identifying how/if such biological targets could be exploited for the design of new antibiotic chemotherapies.

Herein, we have investigated the BshB and Bca properties of the B. anthracis N- deacetylases BA1557, BA3888, and BA3524 as well as the orthologous B. cereus enzymes BC1534 and BC3461. The results indicate that both BA1557 and BA3888 (and the B. cereus homologs) are competent in catalyzing the N- deacetylation of GlcNAc-Mal. No kinetically significant Bca activity was observed for any of these enzymes. Assays with GlcNAc-Mal substrate analogs

29

demonstrate that the malate portion of the molecule plays a role in controlling substrate specificity. In addition, a site-directed mutagenesis approach provides insight into the mechanistic details of the Zn-dependent catalytic acid-base reaction. The apparent discrepancy between the reported GlcNAc N-deacetylase activities of the 97% identical BA1557 and BC1534 enzymes is also addressed.

MATERIALS AND METHODS

Materials

BSmB (Newton et al., 2012), GlcNAc-Mal (Roberts et al., 2013), GlcNAc-OBn

(Yeager and Finney, 2005) and GlcNAc-OMe (Ma et al., 2004) were chemically synthesised as previously described. BS-fosfomycin was enzymatically synthesized via FosB-catalysed S-conjugation of BSH with fosfomycin (Roberts et al., 2013). GlcNAc was purchased from Acros, and naphthalene-2,3- dialdehyde (NDA) was from Anaspec. All other reagents were purchased from

Fisher Scientific or Sigma.

Expression and purification of B. anthracis BA1557, BA3524, BA3888 and B. cereus BC1534, BC3461

The cloning of BA1557 (BshB) was described by Parsonage et al. (Parsonage et al., 2010). The codon-optimized genes for BA3524 and BA3888 were synthesized by GenScript and sub-cloned into pET28a (+) (Novagen). BC1534 in pET26b (+) (Deli et al., 2010) and BC3461 in pET24a (+) were generously provided by Dr. Vassilis Bouriotis. All variants of BA1557 were prepared using

30

the QuickChange site-directed mutagenesis kit (Stratagene), and confirmed by

DNA sequencing. All proteins described in this study contained a C-terminal hexahistidine tag.

BA1557, BC1534 and all BA1557 variants were purified by the following general protocol. The plasmid was transformed into chemically competent E. coli BL21

(DE3) cells and plated on LB plates with 40 µg/mL kanamycin in a 37 ºC incubator overnight. For protein expression, 3 L of LB supplemented with 40

µg/mL kanamycin were inoculated with freshly transformed cells and incubated at

37 ºC/300 rpm. Expression was induced upon addition of 5.8 mM lactose when the OD600 reached 0.5, and the culture was further incubated at 15 ºC/300 rpm overnight before being harvested by centrifugation at 5000 x g for 10 min at 4 ºC.

The cells were stored at -20°C until further use. Cells were resuspended with 25 mM Tris, 150 mM NaCl, and 10% glycerol, pH 8.0 (buffer A), and then lysed using an EmulsiFlex-C5 high pressure homogenizer. The cell debris was separated by centrifugation at 12,000 g for 30 min, and the supernatant was cleared upon treatment with 1% w/v streptomycin sulfate followed by centrifugation (12,000 g for 20 min). After the treatment, the clear supernatant was loaded onto pre-equilibrated (buffer A) zinc- or cobalt-IMAC columns (GE

Healthcare). The column was washed with buffer A until the Abs280 reached baseline. Proteins associated with the resin were eluted in a step gradient of 5%,

15%, and 50% of buffer B (25 mM Tris, 150 mM NaCl, 300 mM imidazole, and 10% glycerol, pH 8.0). The fractions eluted from 50% of buffer B containing pure proteins were diluted five times with buffer C (25 mM Tris, 10% glycerol, and pH

31

8.0), and loaded onto a 5 mL Hitrap Q FF (GE Healthcare) column pre- equilibrated with buffer C. The desired protein was eluted with 0.5 M NaCl in the same buffer. Protein concentration was determined by Bradford Assay using

BSA as the standard (Bradford, 1976). Protein aliquots were frozen with liquid nitrogen and stored at -80 °C.

Due to the poor solubility experienced with BA3888 pET vector expression, the gene was subcloned into a pBAD vector. The purification of pBAD-BA3888 and pET24a-BC3461 shared the same purification procedure. The plasmids were transformed into E. coli BL21 (DE3) cells, and cells were plated on LB agar with

100 µg/mL ampicillin or 40 µg/mL kanamycin. The inoculum was prepared with several colonies taken from a plate and transferred into 100 mL LB broth with appropriate antibiotics. After one hour incubation at 37 oC/300 rpm, the inoculum was added to 4L of LB broth with antibiotics and incubated at 37 oC/300 rpm.

Protein expression was induced at OD600 of 0.4 by the addition of 5.8 mM lactose or 20 mM arabinose final concentration. The culture was shaken at 37 °C for another 4 h before being harvested by centrifugation. The cells were lysed as described above, and the soluble proteins were loaded onto a nickel-IMAC column pre-equilibrated with buffer A. The column was washed with buffer A, and proteins associated to the column were eluted with a step gradient (5%, 15%,

30%, 50% and 100%) of buffer B. Fractions containing deacetylase activity were pooled, diluted six-fold with 25 mM HEPES, 10% glycerol, pH 8.0 (buffer D), and loaded onto a MonoQ column (GE Healthcare) equilibrated with buffer D.

Proteins associated to the column were eluted in a 20-mL linear gradient from 0

32

to 0.75 M NaCl in buffer D. Pure protein fractions (by SDS-PAGE) containing activity were combined and aliquots were then frozen into liquid nitrogen for storage at -80 oC. Within ten days from purification date, freshly thawed aliquots were used for kinetic experiments.

Circular dichroism (CD) spectroscopy

CD spectra were performed using an Aviv CD spectrometer (Model 215; AVIV

Biomedical) with a 1-mm quartz cuvette (Hellma Analytics) and a bandwidth of 1 nm. Protein samples were analyzed at 5 µM in 10 mM phosphate buffer (pH 7.4) and scanned from 250 nm to 190 nm with 0.5 nm increments. The final spectrum of each sample is the average of 10 scans.

Enzyme assays for N-deacetylation of GlcNAc-Mal, GlcNAc-OBn, GlcNAc-

OMe and GlcNAc

The deacetylation activity was assayed by quantitation of the primary amine product. In general, a 200 µL assay mixture containing 50 mM MOPS, pH 7.4, and 0.1-20 µg enzyme was pre-equilibrated at 37 °C for 5 min. Reactions were initiated by the addition of substrate. Sample aliquots (20 µL) were taken at different time points and quenched with 20 µL acetonitrile. The primary amine products (GlcN, GlcN-Mal, etc.) were then derivatized upon addition of 200 µL of freshly prepared NDA-mix (0.1 M borate pH 9, 2.5 mM KCN and 0.5 mM naphthalene-2, 3-dialdehyde). The solution mix was allowed to react in the dark for 30 min before being read in a flat-bottom microplate (Costar) using a Synergy

H1 plate reader (Biotek) with Ex390/Em480. Initial velocity was determined from

33

the slope of a plot of fluorescence units (FU) versus time using 4 different time points over a period of up to 3 hours. A glucosamine standard curve was used to convert the FU/min slopes into [product]/min; the fluorescence intensity of the

GlcN-NDA product was identical to that of the GlcN-Mal-NDA. Initial deacetylation rates of GlcNAc-Mal were measured over a concentration range of

0-3.5 mM. For determination of the steady-state parameters, results representing the mean of triplicate values were fitted to the Michaelis-Menten equation using

SigmaPlot 11.0.

Enzyme assays for amidase activity with BSmB, BSH and BS-Fos

Amidase activity of BSH was assayed by quantitation of the monobromobimane

(mBBr) derivative of Cys (CySmB) produced during hydrolysis of BSmB. A representative assay contained 1.5-20 µg of enzyme in 200 µL 50 mM MOPS pH

7.4 pre-warmed at 37 ºC for 5 min. The reaction was initiated by addition of

BSmB. After incubation for various time intervals, reaction aliquots (10 µL) were quenched by the addition of 20 µL acetonitrile followed by centrifugation (5 min).

The supernatant (20 µL) was diluted 50-fold with 5 mM HCl before analysis by

HPLC using previously described procedures (Newton et al., 2009).

The amidase activity towards BSH was assayed by quantification of Cys with fluorescent reagent mBBr. The assay was performed by first equilibrating 40 µM of the enzyme in 50 mM HEPES buffer (pH 7.4) for 10 min at 37 °C. The reactions were initiated upon addition of substrate (0.2 to 10 mM BSH) followed by incubation for 30 min at 37 °C. Ten-microliter time points were collected and

34

reaction was quenched by heating at 80 °C for 10 min. Reaction aliquots (4 µl) were then reacted with 6 µl of 100 mM mBBr for 15 min at room temperature to label Cys product and quenched upon addition of 0.25 µl 5 M HMeSO3 and centrifuged for 5 min. Finally, the supernatant was diluted with 100 µl of 10 mM

HMeSO3 and 50 µl of sample was analyzed by HPLC as described in (Newton et al., 2009). The activities were calculated from a calibration curve using CysmB as standard. The mean of three duplicates were fitted into Michaelis-Menten equation to determine the steady-state parameters.

The activity towards Bs-fosfomycin (Bs-Fos) was assayed by quantifying Cys- fosfomycin and GlcN-Mal using AccQ Fluor reagent kit (Waters) to detect the primary amine products. A solution containing 40 µM of enzyme in 50 mM

HEPES (pH 7.4) was prewarmed at 37 °C for 5 min and reaction was started upon addition of 0.5 mM Bs-Fos. Thirty-minute reactions were quenched by heating the samples at 85 °C for 10 min. The derivation of amine products with

AccQ Fluor reagent was performed as recommended by the manufacture. The samples were assayed with HPLC method as described in (Parsonage et al.,

2010).

Preparation of metal-free deacetylase and metal reconstitution

To prepare metal-free BA1557, Co2+-IMAC purified protein (~75 µM) was incubated with 25 mM Tris, 25 mM diethylene triamine pentaacetic acid, 10% glycerol, pH 7.5, on ice for 30 min. The protein solution was then dialyzed over 3

× 2 L of 25 mM Tris, 10% glycerol (pH 7.5) buffer at 4 ºC. The concentration of

35

residual metal ion was determined to be less than 5% by ICP-AES (Tededyne

Leeman Labs) analysis.

2+ For the reconstitution with Zn , apo-BA1557 (~10 µM) was incubated with different stoichiometric ratios of ZnSO4 on ice for 30 min. The solution was then dialyzed against 1 L of 25 mM Tris (8.0), 10% glycerol at 4 ºC to remove unbound Zn2+, and the protein metal content was measured by ICP-AES. pH-dependent activity profiles

For the pH-dependence experiments, the following buffers were used: 50 mM

MES, pH 6.0-7.0; 50 mM MOPS pH 7.0-8.0; 50 mM bicine pH 8.0-9.0; 50 mM borate pH 9.0-10.0. The standard assay protocol was conducted by equilibrating

200 µL buffer containing ~0.4 µg deacetylase at 37 ºC. The deacetylation reaction was initiated by the addition of 0.25 mM GlcNAc-malate. Samples (20 µL) were taken at intervals, and the reaction was terminated by addition of an equal volume of acetonitrile. Steady-state kinetic parameters KM, kcat, and kcat/KM for deacetylase activity were determined by fitting initial velocities to the Michaelis-

Menten equation. Equation 1 was fit to the pH rate profile, wherein V is the observed rate of the reaction, K is pH-independent rate constant for GlcNAc or

GlcNAc-Mal substrates, and Ka and Kb are the ionization constants of the acid and base species, respectively (McClerren et al., 2005).

36

RESULTS

Previous work described the ability of BA1557 to catalyse the hydrolysis of

GlcNAc-Mal using an assay that involved HPLC analysis of GlcN-Mal product formation after derivatization with the amine-specific fluorophore (AccQ tag)

(Parsonage et al., 2010). In the present study, BA1557 and other BshB-like enzymes have been extensively characterized through the application of a faster assay procedure that bypasses the need to quantify derivatized reaction products by HPLC separation. This was achieved by developing a fluorescence microplate assay to detect GlcN-Mal formation following derivatization of the primary amine with NDA (Selbach et al., 2010). This method is more cost-effective and faster than the commercial AccQ tag kit (Parsonage et al., 2010), and is ∼50-fold more sensitive than methods involving a direct fluorescamine-based procedure (Huang and Hernick, 2011). Despite the aforementioned benefits of this method, NDA- derivatization of primary amines is not specific to the deacetylation product and background signal associated with other primary amines present in the reaction mixture is a recurring limitation of such amine-specific fluorescence-based assays (Hernick, 2011). Nevertheless, the high sensitivity of this method allowed us to optimize the use of synthetic substrates (Table 2.1), many of which were only available in limited quantities. The activity of BA1557 with GlcNAc-Mal (19 +

1.1 μmol/min per mg), when tested using this method, was comparable with that obtained with the AccQ tag (21.5+0.2 μmol/min per mg) or fluorescamine procedures (19.8 + 1.0 μmol/min per mg). BA3888 also displayed comparable substrate kinetics with those of GlcNAc-Mal, thereby demonstrating the functional redundancy of BA1557 in the second step of BSH biosynthesis (Table 2.2 and

37

Figure 2.2).

Table 2.1 Substrate specificity of B. anthracis and B.

cereus N-deacetylases.

The kinetic analysis of both enzymes fits well with the in vivo data, which demonstrated a 70% reduction in BSH levels in the B. anthracis ∆bshB strain

(Parsonage et al., 2010). As expected, the orthologous BC1534 (97% sequence identity) displays very similar kinetic behaviour in the BshB reaction (Table 2.2 and Figure 2.2), supporting the proposed role for this enzyme in the biosynthesis of BSH.

Table 2.2 Kinetic parameters of B. anthracis and B. cereus N-deacetylases

38

The catalytic efficiencies (kcat/KM) of BA1557 and BC1534 with GlcNAc were four orders of magnitude lower than that determined against GlcNAc-Mal, whereas

BA3888 showed no detectable activity (Table 2.2 and Figure 2.2). The calculated kcat/KM values for BA1557 and BC1534 with GlcNAc were substantially

(six orders of magnitude) lower than the values reported previously for the B. cereus enzyme (Deli et al., 2010), but similar to the values reported for B. anthracis (Parsonage et al., 2010).

Figure 2.2 Substrate saturation curves of BA1557 (▲), BC1534 (■), BA3888 (●) and BC3461 (▼).

Additional N-deacetylase candidates BA3524 and BC3461 were investigated in the BshB reaction. The two enzymes are 95% identical in sequence and are likely to perform similar intracellular functions (Figure 2.3). Phylogenetic analysis shows that they constitute a separate branch of N-acetyl hydrolases, different from those including BA1557 or BA3888 (Figure 2.4). In our hands, BC3461 showed limited activity with GlcNAc-Mal (Table 2.2), whereas BA3524 displayed no detectable activity.

39

Similarity (%)

BA1557 BA3888 BA3524 BC1534 BC3461

BA1557 - 26 24 97 24 (%)

BA3888 45 - 47 26 47 BA3524 40 63 - 24 95 BC1534 99 45 42 - 26 Identity BC3461 42 64 98 42 - Figure 2.3 Sequence identity (grey boxes) and similarity (white boxes) between the N-deacetylase enzymes analysed in the present study from B. anthracis str. Ames and B. cereus ATCC-14579.

Far-UV CD spectra and ICP-AES analyses of these proteins indicated comparable secondary structure, and Zn2+ content with those of BA1557,

BA3888 and BC1534 (Figure 2.5). This observation indicates that the lack of activity seems not to be attributed to protein misfolding or lack of metal cofactor resulting from heterologous E. coli expression.

Figure 2.4 Drawgram of deacetylase enzymes included in Figure 2.10. PHYLIP rooted phylogenetic tree phenogram (Drawgram) constructed in Biology Workbench depicts the phylogenetic distribution of deacetylases. Shaded boxes indicate proposed functional assignments for deacetylase candidate sequences.

40

Figure 2.5 Far-UV circular dichroism spectra of deacetylases BA1557, BC1534, BA3888, BC3461 and BA3524 (Panel A) and BA1557 variants H110A, D14A, R53A, R53K and R109K (Panel B).

Zn2+-dependent deacetylase activity

The Zn2+-binding site identified in the structure of BC1534 (BcZBP) displayed a similar arrangement to that found in the active sites of other Zn2+-dependent deacetylases (Fadouloglou et al., 2007). Like the MshB structure, the BC1534 active site displays two histidine residues and one aspartate residue (His12,

His113 and Asp15) providing a facial triad Zn2+ coordination (Fadouloglou et al.,

2007; Maynes et al., 2003) (Figure 2.6). All three metal-ion-coordinating residues are strictly conserved in the MshB and Mca sequences, in addition to the BshB and Bca candidate enzymes from B. anthracis and other BSH-producing species

(Figure 2 .7).

Figure 2.6 Active site region of B. cereus BC1534 (PDB Accession 2IXD). The Zn2+is coordinated with His12, Asp15 and His113.

41

Figure 2.7 Amino acid sequence alignment of deacetylase enzymes. ClustalW alignment including deacetylase sequences from Bacillus anthracis str. Ames (BA1557, BA3888, BA524), Bacillus cereus ATCC 14579 (BC1534, BC3461), Bacillus subtilis subsp. subtilis str. 168 (BSU22470, BSU19460), Bacillus pumilus SAFR-032 (BPUM1978, BPUM_1869), Bacillus megaterium DSM 319 (BMD_1367, BMD_1967), Staphylococcus aureus 04-02981 (SA2981_0544), Staphylococcus saprophyticus subsp. saprophyticus ATCC 15305 (SSP2151), Mycobacterium smegmatis str. MC2 155 (MSMEG5261, MSMEG5129), Mycobacterium tuberculosis H37Rv (Rv1082, Rv1170), Corynebacterium efficiens YS-314 (CE1052, CE1158), Streptomyces coelicolor A3(2) (SCO4967, SCO5126).

42

Hexahistidine-tagged recombinant BA1557, purified by Zn2+-affinity chromatography contains 2.55 + 0.04 Zn2+ cations per monomer. Treatment with

20 mM EDTA or EGTA had a minor effect on either metal content or enzyme activity (less than 20% decrease upon 1 h of incubation), whereas treatment with the stronger metal-chelating reagent DETAPAC removes nearly 98% of the bound Zn2+ and almost completely inactivates the enzyme.

This indicates that Zn2+ is a tight-binding metal ion cofactor in this enzyme. The apo form of the enzyme displays an identical far-UV CD spectrum, indicating that neither DETAPAC treatment nor Zn2+ displacement causes major changes to secondary structure (Figure 2.8, inset). Zn2+ titration of the apo form completely reconstitutes the activity of the enzyme, correlating with a stoichiometry of one

Zn2+/mol of enzyme (Figure 2.8).

Figure 2.8 Deacetylation activity of apo-BA1557 reconstituted with different stoichiometric ratio of Zn2+. The assays were conducted with 0.2 µg enzyme (0.035 µM) in the presence of 0.25 mM GlcNAc-Mal. The inset panel shows the Far-UV circular dichroism spectrum of holo-BA1557 (solid line) and apo-BA1557 (dashed line).

43

Further addition of >1 molar equivalent of Zn2+ causes inhibition. This Zn2+ activation/inhibition profile is very similar to that observed for UDP-3-O-(R-3- hydroxymyristoyl)-GlcNAc deacetylase (LpxC). It has been proposed that LpxC has two Zn2+-binding sites, one necessary for catalytic activity and a second site with allosteric inhibitory properties (Jackman et al., 1999)

In addition to Zn2+, BA1557 can also be activated upon stoichiometric addition of other metal cations such as Ni2+,Co2+ and Fe3+ (Table 2.3 ). These results indicate that BA1557 shares similar profiles for non-specific metal-dependent deacetylation, comparable with the characterized cambialistic behaviours of

MshB (Huang et al., 2011), LpxC (Gattis et al., 2010) and the histone deacetylases (Dowling et al., 2010).

Table 2.3 Activity of BA1557 upon reconstitution with different metals

General acid–base catalysis

The pH–activity profile for the BshB reaction with BA1557 presents a bell-shaped curve indicative of general acid–base catalysis (McClerren et al., 2005). The maximum catalytic activity at subsaturating concentrations of GlcNAc-Mal was

44

reached at pH 7.8. The pH–activity curve fit is consistent with two ionization states with associated pKa values of 6.5 and 8.5. Interestingly, BA3888 and

BC1534 displayed nearly identical pH–activity bellshaped curves and similar ionization constants (Figure 2.9 and Table 2.4). The pH profile of Zn2+-or Co2+- reconstituted BA1557 displayed similar ionization events (Figure 2.10). Likewise, the pH-dependence on the GlcNAc deacetylation reaction displayed pKa values of 6.7 and 9.3 (Figure 2.9). These results discard the participation of the malate carboxyl group(s) of the substrate or the metal cofactor on either ionization events. The similar pH–activity profile of BshB enzymes suggests the participation of conserved residues in the general acid–base catalytic mechanism of the BshB reaction.

1 1 A B

0.1

0.001 0.1 0.01

0.0001 0.001

10-5 5.5 6.5 7.5 8.5 9.5 0.01 0.0001 5.5 6 6.5 7 7.5 8 8.5 9 9.5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 pH pH Figure 2.9 Activity pH profiles deacetylase enzymes. Panel A shows the relative activity (V/K) of BA1557 (), BC1534 (▼), BA3888 (), and BC3461 (, inset). Panel B shows relative activity (V/K) of BA1557 wild type (), H110A (), and D14A (▲). Assays were measured with 0.25 mM GlcNAc-Mal under pH range from 5.97 to 9.56. All the pH curves exhibit bell shape except D14A which loses the basic limb. The pKa values were determined by fitting the curve into the equation 1 shown in the “Method and Materials”.

45

Table 2.4 pH dependence of different enzymes/mutants for N-deacetylation of GlcNac-Mal

Close inspection of the protein environment surrounding the Zn2+ cation at the active site of BC1534 (Figure 2.6) initially suggested His110 and Asp14 as possible candidates for the general acid and general base respectively. Both residues are strictly conserved in all members of this family of enzymes including the BshB, MshB and Mca sequences (Figure 2.7).

1 1 A B

0.1 0.1

0.01 0.01 5.5 6 6.5 7 7.5 8 8.5 9 9.5 6 6.5 7 7.5 8 8.5 9 9.5 pH pH Figure 2.10 Activity pH profiles deacetylase activity of BA1557. Panel A shows the relative activity (V/K) of Co-BA1557 () and Zn-BA1557 (▲) using GlcNAc-Mal as substrate. Panel B shows relative activity (V/K) of Zn-BA1557 with GlcNAc-Mal (▲) and GlcNAc () as substrates. Assays were measured with 0.25 mM GlcNAc-Mal or 50 mM GlcNAc under pH range from 5.5 to 9.5. The line is the best fit of equation 1 as described in materials and methods, the pKas and pKbs were shown in the insert table.

ND1 of His110 and OD2 of Asp14 are located on opposite sides of the acetate

46

group that interacts with Zn2+, approximately 4 Å (1 Å=0.1 nm) from the metal centre. Individual alanine substitutions of His110 and Asp14 in BA1557 resulted in decreased activity levels of 4-fold and 3000-fold respectively (Figure 2.10 and

Table 2.5). The pH–activity profile of H110A BA1557 retained a bell-shaped curve, ruling out the possible participation of this residue as the acid or base in this mechanism. Interestingly, in the structure of BC1534, OD2 of Asp112 is located within hydrogen-bonding distance from both NE2 of His110 and NE2 of

His12, which is a known ligand for the Zn2+ (Figure 2.6).

Table 2.5

Steady-state kinetic parameters of BA1557 variants for N-deacetylation of GlcNAc-Mal

GlcNAc-Mal

Km kcat kcat/Km Relative Enzyme -1 -1 -1 (mM) (s ) (s .M ) Efficiency WT 0.19 ± 0.03 8.24 ± 0.46 43,300 ± 7263 1

D14A 0.15 ± 0.03 (2.8 ± 0.14) x10-4 1.8 ± 0.37 4.1x10-5 H110A 0.69 ± 0.08 2.19 ± 0.22 3,170 ± 486 0.073 R53A ---a ---a ---a 0 R53K 0.31 ± 0.05 (1.4 ± 0.1) x10-2 45 ± 7.9 1.0x10-3 R109K 0.24 ± 0.03 1.29 ± 0.16 5,380 ± 947 0.124

GlcNAc

Km kcat kcat/Km Relative Enzyme -1 -1 -1 (mM) (s ) (s .M ) Efficiency WT 57.1 ± 8.5 (3.3 ± 0.14) x10-3 0.067 ± 0.010 1 R53K ---b ---b ---b 0 R109K 18.6 ± 1.8 (2.9 ± 0. 1 ) x10-3 0.16 ± 0.016 2.32 H110A ---c (0.9 ± 0.1) x10-3 c ---c a Assays were performed using 20 µg enzyme and 0 – 5 mM GlcNAc-Mal in 50 mM MOPS (pH 7.4) at 37 °C. b Assays were performed in presence of 20 µg enzyme and 0 -120 mM GlcNAc in 50 mM MOPS (pH 7.4) at 37°C. c No substrate saturation was observed up to 180 mM GlcNAc, calculated Kcat was based on the velocity at 120 mM.

In BC1534, the D112A substitution completely eliminates the GlcNAc deacetylase activity of this enzyme (Deli et al., 2010). Unexpectedly, the D14A

47

BA1557 variant retained the ionization event associated with pKa1, but not that for pKa2, suggesting its role as a general acid in the reaction. Although the identity of the general base remains unknown, the pH–activity profile of the D14A variant eliminates the possibility of any dual involvement of this residue, in both ionization events of this reaction mechanism.

Figure 2.11 Proposed mechanism for BshB deacetylation reaction.

Gatekeepers of substrate binding and hydrolysis

The very low catalytic efficiency of the enzyme with GlcNAc compared with

GlcNAc-Mal prompted the investigation of substrate analogues where the malate aglycone was replaced by an uncharged methyl or benzyl motif. Table 2.1 shows that all substrates lacking malate at this position were at least three orders of magnitude less effective as substrates than GlcNAc-Mal. An in silico auto- docking approach was used to explore potential binding modes of GlcNAc-Mal with BC1534 (Trott and Olson, 2010). The results from these docking experiments pointed to two arginine residues, Arg53 and Arg109, as providing potential electrostatic interactions with the substrate (Figure 2.11). These residues are located on opposite sides of the active site, ∼10 Å from each other and∼5 Å from the C1-position of bound acetate, whereas the distances between the Zn2+ to the ω-N of Arg53 and Arg109 were 5.72 Å and 8.03 Å respectively.

48

Interestingly, structural and functional analyses of malate dehydrogenases have shown that two arginine residues (∼12 Å apart) are involved in substrate binding providing electrostatic interactions with both carboxyl groups of the L-malate substrate (Bell et al., 2001). Compared with WT BA1557, an R53A mutation completely eliminated the BshB activity, whereas a more conservative R53K substitution displayed 103-fold lower activity for GlcNAcMal and no detectable activity with GlcNAc (Table 2.5).

Figure 2.12 Model of the active site of BC1534 (PDB code 2IXD) in complex with GlcNAc-Mal. The GlcNAc-Mal molecule was generated by Gaussian09 to achieve the most stable conformation. The model is the result of a simulated annealing docking of GlcNAc-Mal to BC1534 using Autodock Vina (Trott and Olson, 2010). Arg53 and Arg109 on each site of the catalytic Zn2+ with a distance 6.17 Å and 2.39 Å respectively are shown.

Both substitutions did not impair the secondary structure of this enzyme; the far-

UV CD spectrum was nearly identical with that of the WT (Figure 2.5). Sequence alignment of several deacetylases including MshB and Mca indicate that Arg53 is strictly conserved. It has been suggested that Arg68 and His144 of MshB

(equivalent to Arg53 and His110 of BC1534 and BA1557) participate in

49

electrostatic interactions with a sugar hydroxy group of the substrate. Our results with the Arg53 mutants provide support for the involvement of this residue in substrate binding. Arg109, however, is only conserved among a group of deacetylases limited to the BshB1 candidates; including BA1557, BC1534 and

YpjG (Figure 2.7). The BA1557 R109K variant showed a 6-fold decrease in kcat for GlcNAc-Mal (Table 2.5). Although this substitution impaired the catalytic efficiency of BA1557 with GlcNAc-Mal, the R109K variant showed a modest improvement in reactivity against GlcNAc. This variant enzyme had a lower KM for GlcNAc (18.6 mM) and similar kcat (2.9×10−3 s−1) when compared with the

WT constants of 57 mM and 3.3×10−3 s−1 respectively (Table 2.5).

Deacetylase enzymes participating in detoxification pathways

BSH is already known to be implicated in some aspect of drug detoxification as exemplified by BST (bacillithiol-S-transferase) (FosB). A second class of BSTs has also recently been identified, but their target electrophilic substrates are not yet known. On the basis of what has been observed previously in MSH metabolism, enzymes catalysing the BshB reaction (the hydrolysis of GlcNAc-

Mal to yield GlcN-Mal) could also potentially be capable of catalysing the hydrolysis of bacillithiol S-conjugates produced in BSH-dependent pathways. The proposed Bca enzyme is thought to be involved in the subsequent step, thus utilizing the products of the BST reaction as substrates. Because of the similarities in mechanism between the BshB and Bca reactions, we have determined the reactivity of BA1557 and BA3888 against BSmB, BSH and BS–

Fos. The activity of BA3888 in the hydrolysis of the CysmB side chain of BSmB

50

was 200-fold higher than that of BA1557; however, at 5 mM BSmB the turnover rate of BA3888 was only 1.84 min−1 (Table 2.2). Kinetic analysis of BA1557 and

BA3888 with BSH also showed low rates of BSH degradation by formation of cysteine and GlcN, indicating that these enzymes do not contribute to the pool of reduced cysteine levels. Interestingly, the KM of BA1557 for BSH (0.48 mM) is in the same range as cellular BSH concentrations, suggesting that BSH could act as a feedback inhibitor of this enzyme (Table 2.2). On the other hand, the

BA3888 Km for BSH was not determined since no substrate saturation was reached up to 5 mM BSH. In our hands, neither BA1557 nor BA3888 showed any detectable activity against BS–Fos. The remarkable selectivity of this group of enzymes for a restricted subset of substrates highlights the importance for in vivo screening of candidate compounds in strains lacking these enzymes.

DISCUSSION

In the present study, we performed comparative kinetic analysis of the B. anthracis and B. cereus deacetylases in performing BshB and Bca reactions to provide biochemical evidence for the cross functionality of these enzymes in the biosynthesis of BSH. This conclusion supports in vivo demonstrations that inactivation of bshB in B. anthracis and bshB1(ypjG) in B. subtilis decreased, but did not eliminate, the levels of BSH. Kinetic analyses of the BshB reactions with

BA1557 and BA3888 show that they display equivalent catalytic efficiencies towards GlcNAcMal. However, BA3888 showed a nearly 200-fold higher activity with BSmB compared with BA1557 (Table 2.2). The reactivity of this enzyme towards larger bacillithiol conjugates provides experimental evidence supporting

51

the role of BA3888 as a catalyst for the Bca reaction. However, it is worth noting that the substrate kinetics for BSmB with BA3888 are still poor and it remains to be seen what the physiologically relevant bacillithiol conjugate substrate(s) is

(are) for this enzyme. With BA1557, the very weak amidase activity observed with BSH and a KM value comparable with its cellular concentration indicates that

BSH functions as a feedback inhibitor of this enzyme. However, such feedback inhibition is redundant in the presence of a functional BA3888 enzyme, which can substitute for the GlcNAc-Mal activity of BA1557.

Interestingly, in S. aureus, only one BshB-like enzyme has been identified.

Among the enzymes described in the present paper, BA3888 is the closest orthologue of S. aureus BshB (70/49% identity/similarity) supporting the notion that a single enzyme in S. aureus could fulfill both BshB and Bca functions, although this remains to be proved. Amino acid sequence alignment indicates the presence of a ten-residue insertion (GDPFFANRET in BA3888) in Bca/BshB2 sequences, including BA3888 and S. aureus BshB (Figure 2.7). Although the structural location of sequence is not known, adjacent residues in the BC1534 structure constitute a short loop surrounding the active site. It is possible that this decameric insertion sequence could expand the flexibility of the active site to accommodate potentially larger substrates as in the case of Bca. Nevertheless, the low reactivity of these enzymes against BSmB/BS–Fos suggests the possible occurrence of alternative pathways for detoxification of such adducts and/or the involvement of this group of enzymes in serving as amidases to a specific group of bacillithiol adducts.

52

Compared with GlcNAc-Mal, the low activity values against GlcNAc, GlcNAc-

OMe and GlcNAc-OBn reveal the high degree of specificity for the malate aglycone. Our results from site directed mutagenesis pointed to BA1557 Arg109 as one potential gatekeeper in controlling this specificity or stabilizing a reaction intermediate. Although this model is favoured for BshB enzymes containing arginine residues at positions equivalent to Arg109, this scheme may not apply to all deacetylase enzymes capable of catalysing the BshB reaction (e.g. BA3888 and BC3461 contain valine at the equivalent position). Nonetheless, all four enzymes tested in the present study showed a high degree of selectivity for

GlcNAc-Mal. Controlled substrate binding and hydrolysis may serve as a molecular strategy used by this group of enzymes to restrict the repertoire of physiological substrates.

Functional assignment of residues surrounding Zn2+ at the active site confirmed the involvement of Asp14 with the second ionization event in the deacetylation mechanism. The abnormally high pKa of 8.5 is 4 units above the pKa of free aspartate. Large shifts in aspartate pKa (>4 units) have been reported for the active-site aspartate residues of human thioredoxin and bacteriorhodopsin. In both cases, the active-site environment surrounding the aspartate side chain favours the protonated acid form during turnover. Studies on the MshB pH– activity profile showed that the rate of deacetylation was also dependent on two ionization events (pKa1=7.4 and pKa2=10.5). The first ionization event was attributed to Asp15 which was proposed to be the general base on this catalytic mechanism. Whereas there is an agreement for the role of Asp15 as a general

53

base in the MshB mechanism, the identity of the general acid remains controversial. Solvent isotope effect and site-directed mutagenesis experiments suggest the participation of His144 as the general acid; however, pH- dependence studies on H144A MshB and recent structural analysis of this enzyme do not support this model. A previous proposal has evoked the dual role of Asp15 as a single general acid and base. In this later model, it would be expected that the pKa associated with the deprotonation of Asp15 (pKa2) would be higher than 7.4 (pKa1). In the structure of BA1534, Asp14 and His110 occupy positions equivalent to Asp15 and His144 of MshB. Interestingly, in the present study, we identified the pKa associated with Asp14 of BA1557 (pKa2 of 8.6,

Table 2.4) to be only 1.2 pH units higher than the pKa1 of the MshB reaction.

Collectively, these results support a mechanism for the BshB reaction involving

Asp14 as a general acid during the protonation of the nascent amino group on

GlcN-Mal product.

The identity of the residue associated with the first ionization event awaits further investigation. Results of the present study rule out the possible participation of

His110 and Asp14 in performing this function. The pH–activity curve of BA1557 using GlcNAc as substrate showed an identical profile eliminating the potential involvement of the malate carboxy groups of the substrate in the first ionization event (Figure 2.10). Alternatively, the activation of a water molecule by the active-site Zn2+,as observed with carbonic anhydrase, or the participation of the

Zn2+-coordinating Asp15 in abstracting the proton from this water, provide possible candidates for the general base in this mechanism. In any event, the

54

conservation of amino acids within the active site imposes a controlled substrate specificity (along with conserved kinetic behaviour) of this group of deacetylases and opens the door for further investigation into their roles in BSH biosynthesis and BSH-mediated xenobiotic detoxification.

55

CHAPTER 3

Protective role of bacillithiol in superoxide stress and metal homeostasis in

Bacillus subtilis

Zhong Fang and Patricia C. Dos Santos

The work contained in this chapter was accepted to be published in them journal MicrobiologyOpen in 2015. Patricia C. Dos Santos and Zhong Fang designed the experiment, and Zhong Fang performed the experiments. The manuscript was drafted by Zhong Fang and Patricia C. Dos Santos.

56

Abstract

Glutathione serves as the prime thiol in most organisms as its depletion increases antibiotic and metal toxicity, impairs oxidative stress responses, and affects Fe and Fe-S cluster metabolism. Many Gram-positive bacteria lack glutathione, but instead produce other structurally-unrelated yet functionally-equivalent thiols. Among those, bacillithiol

(BSH) has been recently identified in several low G+C Gram-positive bacteria. In this work, we have explored the link between BSH and Fe-S metabolism in B. subtilis. We have identified that B. subtilis lacking BSH is more sensitive to oxidative stress

(paraquat), and metal toxicity (Cu(I) and Cd(II)), but not H2O2. Furthermore, a slow growth phenotype of BSH null strain in minimal medium was observed, which could be recovered upon the addition of selected amino acids (Leu/Ile and Glu/Gln), supplementation of iron, or chemical complementation with BSH disulfide (BSSB) to the growth medium. Interestingly, Fe-S cluster containing isopropylmalate isomerase

(LeuCD) and glutamate synthase (GOGAT) showed decreased activities in BSH null strain. Deficiency of BSH also resulted in decreased levels of intracellular Fe accompanied by increased levels of Mn and altered expression levels of Fe-S cluster biosynthetic SUF components. Together, this study is the first to establish a link between

BSH and Fe-S metabolism in B. subtilis.

57

INTRODUCTION

Biothiols are involved in numerous cellular functions, including but not limited to a variety of biosynthetic pathways, detoxification by conjugation, metal metabolism, and cell division (Hider and Kong, 2011; Jacob et al., 2006; Winterbourn and

Metodiewa, 1999). They can be classified as large-molecular-weight thiols, also called protein thiols, and low-molecular-weight (LMW) free thiols (Sen and

Packer, 2000). Cysteine is the most common LMW biothiol and its sulfhydryl group provides unique chemical functionality distinct from other standard amino acids. Its neutral pKa, nucleophilicity, redox properties, metal ion affinity, and bonding characteristics make Cys a versatile site for many chemical processes

(Bulaj et al., 1998; Jacob et al., 2006; Roos et al., 2013).

In the late 19th century, glutathione (GSH) was identified as another abundant

LMW thiol widely distributed in most cells. GSH mainly exists in its reduced form and its disulfide form (GSSG) with the cellular ratio of GSH:GSSG estimated to be around 100:1 as opposed to cysteine, whose thiol:disulfide ratio lies around

25:1 (Newton et al., 2009; Sharma et al., 2013). Early studies have shown that

GSH plays a key role in a wide array of biochemical functions. First, GSH can directly serve as an antioxidant by donating electrons to reactive radicals through autoxidation reactions to form GSSG (Chesney et al., 1996; Meister, 1988;

Meister and Anderson, 1983). Thus, under oxidative stress conditions, the

GSH/GSSG ratio may fall into the 1-10 range (Reed, 1990). Second, GSH serves as the substrate of biochemical reactions. For example, dehydroascorbate reductase which is a key component of the GSH-ascorbate cycle uses GSH as a

58

substrate to produce ascorbate to detoxify H2O2 (Crook, 1941). Besides the various roles of GSH on detoxifying oxidants and toxins, GSH is also involved in chemical modification of protein cysteine residues (Dalle-Donne et al., 2009). S- glutathionylation is considered a specific post-translational modification on protein cysteine residues in response to oxidative stress or nitrosative stress

(Dalle-Donne et al., 2009; Giustarini et al., 2005; Martinez-Ruiz and Lamas,

2007). Several findings have also reported the vital role for S-glutathionylation in redox-regulation, energy metabolism, cellular signaling, calcium homeostasis, protein folding, and stability (Cotgreave et al., 2002; Dalle-Donne et al., 2003;

Dalle-Donne et al., 2009; Demasi et al., 2008; Pan and Berk, 2007; Reed, 1990;

Rodriguez-Pascual et al., 2008).

Interestingly, the effect of GSH on the superoxide-sensitive Fe-S cluster- containing aconitase (ACN) of Escherichia coli has also been demonstrated

(Gardner and Fridovich, 1993a). The results showed a ~20% decrease of ACN activity in a GSH-deficient E. coli strain, displaying a phenotype which could be recovered upon addition of GSH to the growth medium (Gardner and Fridovich,

- 1993a). Earlier research indicated that the reactivation of O2 -inactivated ACN in cell extract was blocked by adding EDTA or 2, 2’-dipyridyl (Gardner and

Fridovich, 1992; Kennedy et al., 1983). This observation led to the proposal that

- the reactivation of O2 -inactivated ACN would involve the restoration of [3Fe-4S] clusters by free Fe2+. Overall, in this model, GSH would act as either a reductant or Fe2+ donor to facilitate repair of damaged Fe-S clusters. Recently, it has been reported that GSH also acts as a key component of the cytoplasmic labile iron

59

pool providing further support for the important role of GSH in iron homeostasis

(Hider and Kong, 2011). Moreover, changes in the redox ratio of GSH/GSSG has been correlated to induction of Mn-dependent superoxide dismutase expression

(Gardner and Fridovich, 1993b). Despite compelling evidence for GSH’s involvement in oxidative stress, thiol homeostasis, and Fe-S cluster metabolism, the complete inventory of biological processes involving GSH as well as the exact mechanism of GSH-dependent reactions is still not fully understood.

Although GSH is recognized as the dominant LMW thiol in eukaryotes and Gram- negative prokaryotes, GSH is not detected in many Gram-positive species. This observation suggested the existence of other LMW thiols replacing the role of

GSH in those species. In fact, in Actinobacteria, mycothiol was identified as a novel cysteine derivative (Newton et al., 1996). Mycobacterium smegmatis strains lacking mycothiol (MSH) are more sensitive to reactive oxygen and nitrogen species, such as hydrogen peroxide, menadione, plumbagin, and t-butyl hydrogen peroxide (Newton et al., 2008; Newton et al., 1999; Rawat et al., 2002).

A recent study on WhiD, a member of the WhiB-like family of Fe-S proteins, indicated that methyl mycothiol exerted a small but significant protective effect against WhiD cluster loss under oxidative stress (Crack et al., 2009).

In some low-(G+C)-content Gram-positive bacteria, such as Bacillus species, bacillithiol (BSH) is a major LMW thiol recently discovered in bacteria lacking

GSH and MSH (Newton et al., 2009). Because of the crucial roles of GSH and

MSH in thiol-redox homeostasis, protein post-translational modification, and xenobiotic detoxification, it is anticipated that some of these functions might also

60

be performed by BSH (Helmann, 2011). The BSH redox potential was characterized to be -221 mV through in-vitro coupled assay with GSH/GSSG, which is higher than Cys (-223 mV) and CoASH (-234 mV) (Sharma et al., 2013).

This implies that BSH does not have better capacity to buffer oxidative stress through autoxidation reaction. However, considering the low thiol pKa value of

- BSH (pKaSH/S = 7.97) and the relatively high concentration of BSH (0.5-5 mM)

(Sharma et al 2013), it is still possible that BSH is preferable to Cys and CoA as a major redox buffer in B. subtilis. In addition, BSH has been shown to modulate the intracellular labile zinc pool by forming a tetrahedral BSH2:zinc complex with two sulfur and two oxygen ligands (Ma et al., 2014). Although BSH is not an essential metabolite under standard laboratory conditions, initial studies have identified that bacterial strains unable to synthesize this thiol displayed impaired sporulation, sensitivity to acid and salt, depleted NADPH level, and increased sensitivity to fosfomycin, hypochlorite, diamide, and methylglyoxal (Chandrangsu et al., 2014; Gaballa et al., 2010; Lamers et al., 2012; Posada et al., 2014;

Roberts et al., 2013). While defects in metabolism resulting from BSH- phenotypes have been identified, biochemical pathways involving BSH have not been fully explored.

The goal of this work was to investigate the role of BSH in Fe-S metabolism under normal and stress conditions. We show that lack of BSH caused a slow growth phenotype in minimal medium, which may be due to the lower activity of

Fe-S enzymes involved in branched chain amino acid biosynthesis. Furthermore, growth assays under various stress conditions revealed a protective role of BSH

61

against copper stress, superoxide stress, cadmium stress, and iron starvation.

Along with the parallel Fe-S enzyme assays, thiol analysis, and intracellular metal analysis suggest the involvement of BSH in superoxide stress and metal homeostasis. Moreover, western blot analysis was performed on Fe-S cluster biosynthetic proteins in wild-type and mutant strains. Metal analysis of a BSH deletion strain under paraquat stress also revealed a link between BSH and manganese (Mn) homeostasis.

EXPERIMENTAL PROCEDURES

Reagents

3-Isopropylmalic acid and monobromobimane were purchased from Sigma-

Aldrich. N-acetyl-glucosamine-malate and BSSB were synthesized as described in (Sharma et al., 2011) and (Lamers et al., 2012) respectively. Primary antibodies were generated by Thermo Scientific Pierce using laboratory purified protein antigen (SufC, SufD, or MnmA) for rabbit immunization, while Goat anti-

Rabbit IgG (H+L) with AP conjugate as secondary antibody and Lumi-Phos WB chemiluminescent substrate were both from Thermo Scientific. All the other reagents were obtained from Fisher Scientific and Sigma Aldrich.

Bacterial growth

B. subtilis CU1065 (wt), HB11002 (∆bshA), HB110079 (∆bshC) and HB110091

(∆bshC amyE::PxylAbshC) strains were kindly provided by Prof. John Helmann

(Cornell University) (Gaballa et al., 2010). Strains were outgrown either in Luria broth (LB) medium or Spizizen’s minimal medium plus 10 µg/ml tryptophan (MM).

62

For the growth experiments, supplementary components were added where appropriate to the following concentrations: casamino acid, 0.05%; L-glutamine,

120 µg/ml; L-glutamate, 120 µg/ml; L-leucine, 46 µg/ml; L-isoleucine, 34 µg/ml.

When antibiotics were needed, medium was supplemented with 12.5 µg/ml erythromycin and 12.5 µg/ml lincomycin or 40 µg/ml kanamycin. In growth assays, strains cultured up to OD600 of 1.0 (measured using a 1 cm cuvette) in MM were diluted to OD600 of 0.05 into 200 µl of fresh medium and allowed to grow at 37 °C on Falcon 96-well plate. The cell growth rate was monitored at Abs600 using

Synergy H1 plate reader from BioTek.

Enzymatic assays

Activities of Fe-S cluster containing proteins, mononuclear iron enzymes and the non Fe-S cluster containing malate dehydrogenase were determined in clear cell free lysate. B. subtilis wild type and ∆bshA were grown in MM to OD600 of 0.8-1.0.

Cells were harvested by centrifugation at 5000 g for 10 min and frozen in -80 °C until use. Cell pellets from 500 ml cultures were transferred into anaerobic glove box and washed by 5 ml of 25 mM Tris-HCl, 150 mM NaCl, 10% glycerol, pH 8.0 degassed buffer. The resulting pellets were then resuspended in 2 ml buffer and disrupted by sonication anaerobically. Clear cell extracts were collected by centrifugation at 13200 g for 10 min and stored in sealed vials for assays. Protein concentrations were determined by Bradford protein assay using BSA as standard.

63

Glutamate synthase (GOGAT) assays were conducted using the protocol as described in reference (Outten et al., 2004) with minor modifications. To ensure consistent reaction conditions, 200 mM Tris-HCl (pH 8.0), 0.1 M buffered α- ketoglutarate, and 8 mM NADPH were mixed in a ratio of 32:2:1 to make a stock reaction mix. In each assay, 875 µl of fresh stock reaction mix and 75 µl of clear cell lysate were injected into a sealed cuvette filled with argon gas. The decreasing slope of absorbance at 340 nm (NADPH oxidation) was recorded as

S1, and then 50 µl of 0.2 M L-glutamine was added and second slope was recorded as S2. The specific activity of GOGAT was calculated by the following equation:

S2 × 1.05 − S1 (μ · · ) = 0.00622 × mg total protein

Isopropylmalate isomerase (LeuCD) was assayed by using 3-isopropylmalic acid as substrate and by following the formation of the intermediate dimethylcitraconate at 235 nm (Fultz and Kemper, 1981; Manikandan et al.,

2011). In brief, 800 µl of 200 mM phosphate buffer (pH 7.0) and 100 µl of cell extract were mixed first, followed by addition of 100 µl of 10 mM 3-isopropylmalic acid to start the reaction. Blanks were carried out in the absence of substrate, and extinction coefficient of 4.668 mM-1.cm-1 was used to calculate the activity.

Aconitase (ACN) catalyzes a reversible reaction between citrate and iso-citrate through an intermediate cis-aconitate which has absorption at 240 nm (Alen and

Sonenshein, 1999). Each assay reaction contained 700 µl of 100 mM Tris-HCl, pH 8.0, 50 µl of cell extract and 100 µl of 200 mM Tris-Na-Citrate. The increase in absorbance at 240 nm was measured and a millimolar extinction coefficient of

64

3.4 mM-1.cm-1 was used to calculate the activities. Control reactions were performed without substrate.

The activity of malate dehydrogenase (MDH) was measured by following the consumption of NADH at 340 nm (Siegel, 1969). The reaction mix contained 50

µmol of Tris-HCl (pH 7.4), 0.38 µmol of oxaloacetate, 0.15 µmol NADH and 50 µl of cell extract in a total volume of 1 ml. The decrease in absorption at 340 nm was measured over time.

Quercetin 2,3-dioxygenase (QD) catalyzes the cleavage of the O-heteroaromatic ring of quercetin to produce 2-protocatechuoylphloroglucinol carboxylic acid and carbon monoxide. Quercetin 2,3-dioxygenase activity was measured based on the decomposition of quercetin as indicated by the decrease of absorption at 367 nm (Bowater et al., 2004). Cell lysate was prepared anaerobically as described above or by adding 5 µM quercetin in lysis buffer to stabilize metal cofactor. A typical assay reaction contained 1 ml of 200 mM Tris-HCl (pH, 8.0), 100 µl of crude extract and 50 µl of 1.2 mM quercetin in DMSO. The absorption at 367 nm was monitored over time for 2 min and an extinction coefficient of 20 mM-1cm-1 was used to calculate the activity.

Assays of threonine dehydrogenase (Tdh) were performed by following the previous reported protocol (Gu and Imlay, 2013). The cell lysate was prepared anaerobically using Tris-HCl buffer plus 5 mM L-threonine. In each assay, 600 µl of 50 mM Bicine (pH, 8.5), 100 µl of cell extract and 20 µl of 50 mM NAD were mixed in a sealed cuvette, and then 300 µl of buffered 90 mM L-threonine was

65

added to initiate the reaction. The activity of Tdh was measured by monitoring the production of NADH at Abs340.

ICP-OES analysis for cellular metal concentration

B. subtilis wild-type and ∆bshA strains were grown in MM until OD600 reached 1.0.

Culture aliquots (300 ml) were removed before the addition of 100 µM paraquat, and 20 min, 50 min after stress exposure. Cells were harvested by centrifugation and washed twice with 10 ml of chilled 25 mM phosphate buffer (pH.7.4) containing 1 mM EDTA. The resulting cell pellets were immediately washed twice with the same buffer without EDTA and resuspended with 10 ml of phosphate buffer. Aliquots of resuspended cells (2ml) were transferred to an eppendorf tube, in which 20 µl of 20 mg/ml lysozyme was added. The sample was incubated at

37 °C for 30 min, followed by adding 2 ml of 5% nitric acid with 0.1% (v/w) Triton

X-100. The sample was digested at 95 °C for 30 min and supernatant was collected by centrifugation. Each supernatant was diluted with an equal volume of deionized water for the measurement of total metal content.

The concentrations of labile metal (defined as metals not associated with a higher molecular weight fraction, <3000 Da) were quantified as follow. Eight milliliters of cell suspension from above was disrupted by an Avestin EmulsiFlex-

C5 homogenizer anaerobically, and centrifuged at 13,200 g for 20 min.

Supernatant was transferred and adjusted to 8 ml with degassed phosphate buffer. Cell free extract (2 ml) was mixed with 2 ml of 5% nitric acid containing

0.1% (v/w) Triton X-100, followed by incubation and dilution as described above.

66

The resulting sample contained both labile metal and protein-associated metal.

The remaining 6 ml of supernatant was transferred to an Amicon Ultra-15 centrifugal tube with 3000 MW cut-off, and spun at 5,000 g. Two milliliter of flow- through was applied to prepare the sample containing labile metal by following the same procedure as described above.

All the samples were analyzed using inductively coupled plasma optical emission spectrometry (ICP-OES) from Teledyne Leeman Labs. The wavelengths for the detection of each metal were as follows (nm): Fe, 238.204, 239.563; Mn, 257.610,

259.372; Sample concentrations were calculated using metal standard solutions and normalized to the dry weights of cell pellets.

Quantification of thiol levels under stress conditions

The contents of reduced thiol BSH, oxidized thiol BSSB and protein thiol

(BSSProt) were prepared as described by Gerald L. Newton (Chi et al., 2013b;

Newton et al., 2009). In brief, 25 ml of cultures in MM were challenged with stress at OD600 of 1.0 and harvested at 0, 20 and 50 min after exposure.

Controls were prepared by adding 5 mM of N-ethylmaleimide (NEM) to block free thiol, and followed by derivatization with 2.5 mM of monobromobimane (mBBr).

The preparations of disulfide samples were similar to controls, except being reduced with dithiothreitol (DTT) prior derivatization. Reduced thiol was prepared by direct incubation with 2 mM mBBr. Total thiol content was measured by incubating cell pellet with 500 µl of 25 mM HEPES, 2 mM DTT, pH 8.0 on ice for

1 h, followed by adding 500 µl of 8 mM mBBr in acetonitrile. The mixture was

67

incubated at 65 °C for 20 min and quenched by the addition of 10 µl of 2 N HCl.

All samples were diluted 5 times with 5 mM HCl and injected 50 µl for HPLC analysis. The HPLC method was conducted with a Waters Symmetry C8 column

(3.9 x 150 mm, 5 µm) using the gradient as described in (Newton et al., 2009).

The intracellular concentrations of BSH derivatives were calculated based on

BSH standard and normalized to cell dry weight (DW).

Western blot analysis

Cells were grown in MM to OD600 of 0.8-1.0 and then challenged with different stress for 30 min. Cultures were harvested by centrifugation and stored in -20 °C.

Cell pellets were resuspended and disrupted by sonication, the supernatant containing soluble protein extract was then quantified using the Bradford assay

(Biorad). For SDS-PAGE, 50 µg of total protein was loaded into each well and the gel was subjected to Western blot analysis. Custom primary anti-bodies were generated by Thermo scientific through rabbit immunization with heterologously- expressed, purified B. subtilis SufC, SufB and MnmA. The western blot membrane was incubated with each primary antibody, secondary antibody followed by Lumi-Phos WB chemiluminescent substrate, and then exposed to

Carestream® Kodak film from Fisher to fix the protein bands. The films were then been scanned and images were processed by ImageJ software, each band area was quantified (Gassmann et al., 2009). The amount of protein in terms of µg was calculated based on a standard curve from pure proteins in the range of

0.02-0.1 µg.

68

RESULTS

BSH null strains show a slow growth phenotype in minimal medium

The biosynthesis of BSH was proposed to be a three-step reaction, utilizing BshA,

BshB and BshC enzymes. BshA catalyzes the ligation of L-malate onto the carbon 1 position of UDP-N-acetyl-glucosamine to release UDP molecule

(Parsonage et al., 2010). BshB is a metal-dependent deacetylase which cleaves the amide bond of N-acetyl-glucosamine-malate to form N-glucosamine-malate

(Fang et al., 2013; Gaballa et al., 2010). The last step of BSH biosynthesis was suggested to be the condensation reaction of L-Cys and N-glucosamine-malate

(Helmann, 2011). Inactivation of bshA or bshC completely depleted intracellular

BSH while deletion of bshB only caused a modest decrease of BSH levels. Initial studies showed that inactivation of BSH biosynthetic genes caused no effects on

B. subtilis growth rates in LB rich medium (Gaballa et al., 2010). This observation suggested that the depletion of BSH did not cause major defects in metabolism.

However, it was also possible that the components in LB medium replaced the role of BSH and concealed phenotypes in mutant strains. To rule out this possibility, B. subtilis wild type and mutant strains were grown in Spizizen’s minimal medium (MM). Interestingly, unlike the similar growth profiles in LB

(Figure 3.1), the ∆bshA strain displayed a modest yet-reproducible slower growth rate in MM as shown in Figure 3.2A.

69

Figure 3.1 Growth profiles of B. subtilis wild type (square), ∆bshA (circle), and ∆bshC AmyE:Pxyl-bshC (triangle) strains in LB (black) and Spizizen’s MM plus leu, Isoleu, Gln and Glu (empty). Since bshA is located in an operon along with ten other coding sequences including bshB, the slow growth phenotype could have been attributed to a polar effect on downstream genes. We have addressed this potential concern in three experiments. First, the addition of enzymatically produced N-acetyl-glucosamine- malate (GlcNAc-malate) (Fang et al., 2013), the product of the BshA reaction, to cultures of ∆bshA strain recovered the phenotype in MM (Figure 3.2A). Second, the growth rate of a strain containing a deletion within another biosynthetic gene,

∆bshC strain, showed a similar growth profile in MM (Figure 3.2B). In B. subtilis, bshC is a monocystronic gene located remotely from other bshA and bshB biosynthetic genes. Third, addition of BSSB, the oxidized form of chemically synthesized BSH, to either ∆bshA or ∆bshC cultures was able to repair these growth defects (Figure 3.2A and 3.2B). Interestingly, addition of cystine or

GSSG did not rescue the phenotype associated with lack of BSH. Collectively,

70

these results confirmed that phenotypes associated with ∆bshA strain were attributed to the absence of BSH.

Figure 3.2 Growth profile of B. subtilis strain lacking BSH in MM. A. Growth curves of wild-type B. subtilis (square) and ∆bshA strain (circle) in Spizizen’s MM (empty), MM

plus 200 µM of N-acetyl-glucosamine-malate (gray) or MM plus 200 µM of BSSB (black). B. Growth curves of wild type (square), ∆bshC amyE:P bshC (triangle) in xylA MM (empty), MM plus 2% xylose (black) and MM plus 200 µM of BSSB (gray) C. Growth curves of wild-type (square), ∆bshA (circle), and ∆bshC amyE:P bshC xylA (triangle) strains in MM (empty) and in MM plus 0.5% casamino acid (black). D. Growth curves in the presence (black) and absence (empty) of 50 µM of Fe2+ for wild

type (square) and ∆bshA strain (circle). The curves shown are representatives of at least three independent experiments.

71

Further studies revealed that the growth phenotype was fully restored with the supplement of 0.05% casamino acid (Figure 3.2C). Via the screening of different combinations of amino acids, it was concluded that L-glutamine, L-glutamate and branched-chain amino acids L-leucine and L-isoleucine were able to correct the growth defect of the BSH-deletion strain (Figure 3.1). Interestingly, the addition of 50 µM Fe2+ also restored growth defects associated with lack of BSH in MM

(Figure 3.2D), while other divalent metals (Mn, Mg, Ca, and Zn) showed no drastic restoring effect (data not shown), confirming the specificity of Fe2+.

The depletion of BSH decreases the activities of Fe-S enzymes

Initial growth experiments suggested the involvement of BSH in the synthesis of selected amino acids (leucine, isoleucine, glutamine and glutamate). The biosynthetic pathways of these amino acids include Fe-S enzymes. Glutamine oxoglutarate aminotransferase (GOGAT) containing two 4Fe-4S clusters and one

3Fe-4S cluster catalyzes the reversible conversion from L-glutamine and α- ketoglutarate to L-glutamate. LeuCD, a dihydroxy-acid dehydratase in the biosynthesis of L-isoleucine, requires [4Fe-4S] cluster for its function. Likewise, the 4Fe-4S enzyme aconitase in TCA cycle also indirectly contributes to synthesis of amino acids. The E. coli GSH null strain displayed a decrease in aconitase activity (80% of WT levels) under standard growth conditions (Gardner and Fridovich, 1993a). Therefore, we hypothesized that the growth defects of the

∆bshA strain were caused by decreased Fe-S enzyme activities. To test this hypothesis, the specific activities of GOGAT, LeuCD, and ACN were determined in cell lysates from cultures harvested at mid-log phase (OD600 of 0.8-1.0 when

72

using 1 cm cuvette). The activity levels of these enzymes in the wild-type cell lysates were comparable to the previously reported activities (Alen and

Sonenshein, 1999; Bohannon and Sonenshein, 1989), and modestly reduced in the strain lacking BSH (Figure 3.3A).

Figure 3.3 A. Activities of the Fe-S proteins glutamate synthase (GOGAT), aconitase (ACN), isopropyl malate isomerase (LeuCD) and the non-Fe-S protein malate

dehydrogenase (MDH) in cell lysates of B. subtilis wild type (black) and ΔbshA strain (gray) cultured in MM (filled bar) or MM containing an additional 50 µM of Fe2+ (diagonal

striped bar) B. B. subtilis wild type strain (black) and ΔbshA strain (gray) cultured MM and the activity of Fe-S enzymes was quantified before (filled bar) and after the addition

of 50 µM of Fe2+ to cell lysates (upright striped bar). The activities of GOGAT, ACN, LeuCD and MDH for wild type B. subtilis were measured to be 40, 131, 17 and 1051

(nmol/min/mg total protein in crude extract) respectively. The activities of the ΔbshA strain were normalized to the activities of wild type in MM. All assays were performed in

triplicates. The statistical analysis was performed using unpaired t test, p values were obtained by comparing the ΔbshA without iron to wild type w/o Fe and to ΔbshA with Fe.

Comparisons of wild-type activities with and without Fe in both panels were not statistical significant. (NS, not significant, p*<0.05, p**<0.01, p***<0.001, p****<0.0001).

73

As a control, the activity of malate dehydrogenase, a non-Fe/S enzyme, was not significantly different in either strain. Supplementation of growth cultures with 50

µM Fe2+, a condition identified to suppress BSH- phenotype in MM, led to an increase of GOGAT, ACN, and LeuCD activities in ∆bshA strain by 28%, 17%, and 16% respectively while causing no effect to the levels detected for the wild- type cell lysates (Figure 3.3A). Similar results were also reported in S. cerevisiae

∆gsh1 strain, in which the growth defect from deletion of GSH was also repaired upon addition of Fe3+ in the medium, and the LeuCD activity was also enhanced after 1-hour exposure to Fe3+ (Kumar et al., 2011). In B. subtilis cultures, however, the addition of 50 µM of Fe2+ in the cell lysates after disruption did not improved the activity levels of GOGAT, ACN, and LeuCD in both the wild type and ∆bshA strains (Fig 3.3B).

In addition to Fe-S enzymes, mononuclear iron enzymes are also known to be susceptible to changes in the cellular Fe status and found to respond to reactive oxygen species (Anjem and Imlay, 2012). In this study, we determined the activity of two of these metalloenzymes, quercetin 2,3-dioxygenase (QD) and threonine dehydrogenase (Tdh). The lack of BSH caused a slight decrease in their activities (Figure 3.4). The addition of exogenous iron into the cell lysates did not enhance the activities, which confirmed the preparation of cell lysate was strictly anoxic.

74

Figure 3.4 Activities of mononuclear iron enzyme threonine dehydrogenase (A) and quercetin 2,3-dioxygenase (B). Activities were determine with wild type B. subtilis (black) and ∆bshA strain (gray) cell lysates from MM (upright diagonal bar) and MM plus 50 µM of exogenous Fe2+ (upright diagonal bar). All assays were performed in triplicates. The statistical analysis was performed using unpaired t test, p values were obtained by comparing each sample with the ∆bshA w/o Fe (NS, not significant, p*<0.05, p**<0.01, p***<0.001, p****<0.0001).

BSH participates in metal stress response

The effects of oxidative stress on the B. subtilis ∆bshA strain have been previously investigated using zone inhibition assays demonstrating that both wild

- type and BSH strain had the same sensitivity to H2O2 (Gaballa et al., 2010).

Herein, the role of BSH against a wide range of stresses was investigated in MM supplemented with casamino acids. This growth medium not only ensured similar growth rates for both strains under no stress condition (Figure 3.2C and Figure

3.5), but also eliminated the interference from potential protective biomolecules present in rich medium. Initial screens involved growth assays in presence of 0-

75

300 µM H2O2 (data not shown), which showed no differences between wild type and bshA strains, in agreement with a previous report (Gaballa et al., 2010).

Given the involvement of GSH in metal homeostasis, Fe-S biogenesis, and bacterial response to environmental stressors, we investigated the growth profiles of both B. subtilis wild type and bshA deletion strains under conditions known to destabilize metal homeostasis and Fe-S metabolism such as copper

(Cu(I)), cadmium (Cd(II)), as well Fe starvation conditions.

Cu(I) can strongly trigger Fenton-like reactions to produce reactive hydroxyl radicals, which further cause a global oxidative stress response, such as decreased activities for Fe-S enzymes (Macomber and Imlay, 2009). As shown in

Fig. 3.5A, the addition of Cu(I) did not inhibit the initial growth rate of the wild- type strain, but decreased the OD600 in stationary phase from 0.4 to 0.265, displaying similar growth inhibition as previously reported for B. subtilis

(Chillappagari et al., 2010). The BSH mutant, however, showed a concentration- dependent inhibition of the initial growth rate, which indicated the protective role of BSH against Cu(I) stress. The addition of Cu(I) slightly decreased the concentration of BSH (from 0.68 to 0.44 µmol/g DW) and BSSB (from 0.052 to

0.035 µmol/g DW), but it significantly stimulated protein S-bacillithiolation by 6- fold (from 0.02 to 0.12 µmol/g DW) (Figure 3.6A). Wild-type cells when cultured in MM and challenged with Cu(I) for 30 min showed reduction of 50% and 87% of

ACN and GOGAT activities (Figure 3.7).

76

Figure 3.5 Growth curves of wild-type B. subtilis (left panel) and ∆bshA strain (right panel) in MM with 0.05% casamino acid in presence of various stress challenges. A. Cultures contained 0 µM (square), 100 µM (circle) and 300 µM (triangle) of CuCl; B. Cultures contained 0 µM (square), 20 µM (circle) and 50 µM (triangle) of CdCl2. C. Cultures contained 0 µM (square), 100 µM (circle) and 200 µM (triangle) of 2,2’-dypridyl D. Cultures contained 0 µM (square), 100 µM (circle) and 250 µM (triangle) of paraquat. The curves shown are representatives of at least three independent experiments.

77

Figure 3.6 Cellular BSH redox status upon metal and superoxide stress. Intracellular levels of BSH (square), BSSB (circle) and BSSProt (triangle) were determined in B. subtilis wild-type cells cultured in MM before (time 0) and after challenge with 100 µM of Cu (I) (A), 50 µM of Cd (II) (B) or 100 µM of paraquat (C). Oxidized BSH was determined when associated to intracellular small molecular weight thiol (BSSB) or large-molecule thiol (BSSProt) as described in the materials and methods. The statistical analysis was performed using unpaired t test, p values compare time 0 sample with other samples (NS, not significant, p*<0.05, p**<0.01, p***<0.001, p****<0.0001).

This supported previous proposals that Fe-S enzymes were the targets for Cu(I) stress in E. coli (Macomber and Imlay, 2009) and B. subtilis (Chillappagari et al.,

2010; Macomber and Imlay, 2009). However, in the mutant strain the activities of these enzymes were not further affected with Cu(I) challenge (Figure 3.7).

78

Cadmium ion (Cd(II)) has a high binding affinity for sulfur (Nies, 2007). Therefore, the toxicity of Cd(II) has been suggested to be the result of the binding of Cd(II) to free thiols, protein thiols, Fe-S centers and other sulfur-rich compounds

(Helbig et al., 2008). Cd(II) challenge affected the growth profile of both wild-type and mutant strains while displaying a more dramatic effect on the growth rate of the strain lacking BSH (Figure 3.5B). Thiol analysis showed that despite the levels of BSH remaining at ~0.8 µmol/g DW within 50-min treatment, the BSSB content increased 7-fold, from 0.07 µmol/g DW to 0.35 µmol/g DW (Figure 3.6B).

The activities of Fe-S enzymes were also determined in cell lysates of cultures subjected to 50 µM Cd(II) challenge for 30 min. While the addition of Cd(II) to wild-type cultures decreased ACN by only 20% (Figure 3.7A), GOGAT was inactivated to nearly half of its initial activity (Figure 3.7B). Similar to the effect observed for Cu(I) stress, the addition of Cd(II) did not cause further damage to the activity of these Fe-S enzymes in BSH null strain. Interestingly, deletion of

BSH caused a 20% decrease of intracellular zinc which is a substrate for the

CadA resistance system (Ma et al., 2014; Tsai et al., 1992). As a result, the

CadA efflux ATPase was repressed and more Cd(II) was accumulated inside cells, which may explain the observed elevated sensitivity of BSH null strain to

Cd(II).

Since exogenous addition of iron to cultures of B. subtilis bshA recovered growth defects in MM and partially corrected biochemical defects associated with the loss of BSH, we sought to determine the effects of iron starvation in both strains. The growth profile of wild type and mutant were examined under iron

79

starvation conditions in the presence of different concentrations of iron chelator

2,2’-dipyridyl (DP) in MM supplemented with casamino acids. Only a minor inhibitory effect was observed in the mutant strain with 200 µM DP (Figure 3.6C).

Nonetheless, DP challenge elicited response similar to what has been observed for Cu(I) and Cd(II) stresses (Figure 3.7); i.e. the addition of DP to growth medium resulted in 80% and 40% decrease of ACN and GOGAT activities in wild-type extracts, while causing modest to no inhibition in cells lacking BSH.

Figure 3.7 Specific activities of ACN (A) and GOGAT (B) under stress conditions. Cells were grown in MM to late log phase (OD600 of 0.8-1.0 measured with a 1 cm cuvette), at 37

°C, and then 100 µM of 2,2’-dypridyl (DP), 100 µM of paraquat (PQ), 100 µM of H2O2 (HP),

50 µM of CdCl2 (Cd) or 100 µM of CuCl (Cu) was added to challenge the cultures for 30 min.

Cell lysates were prepared as described in Materials and Methods. The wild type strain is represented in black bars while ∆bshA strain is shown in gray bars. All assays were repeated in triplicate. The statistical analysis was performed using unpaired t test, p values compare ∆bshA sample with wild type sample (NS, not significant, p*<0.05, p**<0.01, p***<0.001, p****<0.0001).

B. subtilis strains lacking BSH are sensitive to superoxide stress

80

Lastly, these strains were challenged with paraquat (PQ), which can induce superoxide stress in vivo. When treated with 250 µM of PQ, the mutant strain showed a remarkable growth reduction while wild type remained unaffected

(Figure 3.5D). Thiol analysis also indicated that the content of BSSB increased by near five-fold (from 0.07 to 0.34 µmol/g DW) and protein associated BSH increased by 2-fold (from 0.02 to 0.04 µmol/g DW) (Figure 3.6C). PQ challenge has also been associated with iron-starvation and NADPH depletion in B. anthracis (Pohl et al., 2011). Therefore, the total Fe content of cells cultured in the presence of PQ was determined. As shown in Figure 3.8A, the mutant strain displayed only a small decreased in levels of total Fe when compared to the wild type. In the presence of 100 µM of PQ, the activity of GOGAT was more affected in the wild type than in the bshA strain (Figure 3.7B). The ACN activity, on the other hand, remained unchanged in the wild-type strain while displaying a nearly

50% decrease in cells extracts lacking BSH (Figure 3.7A). Collectively, the PQ sensitivity associated with lack of BSH along with the thiol analysis of wild type cultures exposed to PQ provides evidence for a protective role of BSH against superoxide stress.

Lack of BSH leads to Mn uptake under superoxide stress

Oxidative stress has also been shown to trigger the import of Mn and induction of

Mn-dependent superoxide dismutase. In E. coli, the expression of MnSOD is positively correlated to the redox ratio of GSH/GSSG (Gardner and Fridovich,

1993b). The partial cross-regulation between thiol-redox homeostasis and Fe/Mn levels(Helmann, 2014), led us to determine whether the lack of BSH would affect

81

cellular accumulation of these metals. Intracellular content of Fe and Mn levels were determined in wild-type and bshA cells cultured in MM (Figure 3.8A and

3.8B). While the lack of BSH did not cause drastic changes in the levels of Fe under stress conditions, paraquat challenge elicited a time-dependent increase of intracellular Mn for bshA cells. As shown in Figure 3.8B, the Mn level in the wild-type remained unaltered under paraquat stress while the concentration of

Mn in the mutant increased by more than 2-fold over a 50-min treatment. This observation followed a similar response observed under oxidative challenge in E. coli, in which deletion of catalase and peroxidase introduced oxidative stress that led to further increase in intracellular Mn concentration (Anjem et al., 2009).

Figure 3.8 Total iron (A) and manganese (B) concentration analysis in wild-type B.

subtilis (black bar) and ∆bshA strain (gray bar) after exposure to 100 µM of paraquat stress. All measurements were performed in triplicate from independent cultures. The

statistical analysis was performed using unpaired t test, p values compare ∆bshA sample with wild type sample (NS, not significant, p*<0.05, p**<0.01, p***<0.001, p****<0.0001). C. Manganese distribution in ∆bshA strain under paraquat stress as shown labile (black bar) and protein-associated manganese (gray bar).

82

Based on phenotypic and biochemical analysis in E. coli, it has been proposed that Mn could protect peroxide-stressed cells by replacing Fe in mononuclear enzymes and by decreasing levels of Fe-triggered Fenton reaction (Anjem et al.,

2009). In agreement with this model, paraquat challenge led to increased levels of Mn associated with proteins, supporting a protection mechanism of Mn in B. subtilis (Figure 3.8C). In addition, the activity of manganese superoxide dismutase (MnSOD) was determined in cell lysates of wild-type and ∆bshA strains before and 30 min after 100 µM paraquat challenge. As shown in Figure

3.9, the MnSOD activity in ∆bshA of 15.2 + 1.3 unit/min/mg was higher than activity detected for the wild-type strain (11.8 + 1.0 units/min/mg).

Figure 3.9 MnSOD activity in the cell lysates of CU1065 (wt) and HB11002 (bshA) before (black) and 30 min after (gray) the treatment of 100 µM paraquat. The cultures were

grown to OD600 of 1.0 in MM, and samples were taken before and 30 min after the treatment of 100 µM paraquat. Cell pellets were resuspended, lysed and assayed according to the manufactory protocol from Cayman Chemical. One unit is defined as the

amount of enzyme needed to exhibit 50% dismutation of the superoxide radical.

83

This suggests that the deletion of BSH caused a modest increase of MnSOD even under standard conditions (i.e. when cells were cultured in MM). Paraquat treatment cause no significant change to MnSOD activity levels in wild-type, while it increased by 1.5-fold (from 15.2 + 1.3 to 23.3 + 1.0 units/min/mg) in the strain lacking BSH. Activation of MnSOD elicited by paraquat stress in the ∆bshA strain provided additional evidence for the increase levels of protein-associated

Mn determined through ICP analysis. Besides the protective role of Mn mentioned above, early studies in B. subtilis suggested that free Mn was also capable of eliminating the superoxide stress through an unknown mechanism

(Inaoka et al., 1999). Therefore, ICP-OES analysis was also carried out for free

Mn or that associated to low molecular weight (LMW) fraction Mn (Figure 3.8C).

Levels of free Mn showed a similar increasing trend as the protein-associated Mn in response to paraquat stress. These observations suggested a combined protection mechanism of Mn through metallating enzymes and increasing LMW

Mn pool under paraquat stress.

BSH and Fe-S biosynthesis

In B. subtilis, the SUF system encoded by the sufCDSUB operon, is proposed to be the only Fe-S cluster biosynthesis machinery (Albrecht et al., 2010; Boyd et al., 2014; Selbach et al., 2010). The lower levels of Fe-S clusters, judged by the lower activity of Fe-S enzymes, could also be attributed to expression or inefficiency of biosynthetic apparatus in assembling Fe-S clusters in the absence of BSH and/or conditions detrimental to Fe-S metabolism. Therefore the levels of two biosynthetic components, SufC and SufB, were determined by Western blot

84

analysis of lysates of cells cultured in MM. Expression levels of both SufC and

SufB in wild-type were higher than the mutant strain, while the internal control

MnmA protein, proposed to participate in 2-thiouridine formation (Black and Dos

Santos, 2014), showed no significant change (Figure 3.10A). In E. coli, the SUF system is expressed only under oxidative stress or iron starvation to promote Fe-

S biogenesis under these conditions (Outten et al., 2004). To verify the changes in expression levels of the B. subtilis SUF components under stress conditions,

Western blot analyses of SufC were conducted in cell lysates of MM cultures challenged with 2, 2'-dipyridyl and paraquat for 30 min. Overall, SufC protein levels were similar across various conditions in the wild-type strain (Figure

3.10B). When the bshA strain was challenged to various stresses, the levels of

SufC increased to comparable levels to those observed in wild-type cultures

(Figure 3.10B). This result suggested that expression of sufC although constitutive can respond to iron starvation and oxidative stress, but only when

BSH is absent.

A B

Figure 3.10 A. Western blot analysis of Fe-S biosynthetic enzyme SufB, SufC and MnmA in wild type (black) and ∆bshA (gray) strains. B. Western blot analysis of SufC after challenge with dipyridyl (DP) and paraquat (PQ).

85

DISCUSSION

Since the discovery of BSH in Bacillus species in 2009, this biothiol has been proposed to be analogous to GSH and MSH in B. subtilis based on its similar structure and dominant concentration (Newton et al., 2009). Subsequent studies identified the involvement of BSH in protecting cells against hypochlorite, diamide, and fosfomycin challenges (Chi et al., 2011; Chi et al., 2013b; Gaballa et al.,

2010; Posada et al., 2014). However, the stress-related protective mechanisms of BSH as well as other in-vivo roles are still being unveiled. Initial growth assays with wild-type B. subtilis and ∆bshA mutant strains in rich medium, under various conditions, did not show any significant difference (Gaballa et al., 2010).

However, B. subtilis bshA showed a slow growth phenotype in Spizizen’s MM, which was restored upon addition of GlcNAc-malate and BSSB, supporting the idea that growth defects observed in MM were attributed to lack of BSH. Growth improvement was also observed upon supplementation with selected amino acids (Leu, Ile, Glu, and Gln) or iron. This initial result led us to explore a potential link between BSH and iron metabolism which includes Fe-S cluster biosynthesis. The results showed that deletion of BSH caused severe growth defect under both Cu(I) stress and paraquat stress. As early as 1987, it has been reported that Fe-S cluster containing alpha, beta-dihydroxyisovalerate dehydratase in the isoleucine biosynthetic pathway was an target of superoxide

(Kuo et al., 1987). Subsequent studies also suggested that Fe-S dehydratases were primary targets of superoxide stress. Fe-S dehydratases can be oxidized at an extremely high rate (3 ×106 M−1s−1) (Flint et al., 1993), leading to intracellular

86

superoxide overload and consequently a branched-chain amino acid auxotrophy.

The investigation of B. subtilis described in this study showed that similar oxidative damage appeared to be exacerbated in the absence of BSH, as evidenced by the paraquat sensitivity of the mutant strain. In addition to growth inhibition of the bshA strain, paraquat challenge also resulted in increased levels of intracellular Mn, disulfide pool and BSH-associated to proteins. In combination, these observations highlighted the involvement of BSH in protecting cells against superoxide stress.

Pathways containing Fe-S enzymes are susceptible targets of oxidative stress.

Bacterial responses to oxidative stress includes (i) expression of hydrogen peroxide and superoxide scavenging enzymes, (ii) repression of Fe- and [Fe-S]- containing enzymes with concomitant expression of enzymes utilizing alternate cofactors, (iii) mismetallation of enzyme’s Fe binding site for Mn or Zn (Helmann,

2014; Imlay, 2006), and (iv) up-regulation of Fe-S biosynthetic pathways. In E. coli, the two Fe-S cluster biosynthetic machineries ISC and SUF use distinct regulatory strategies to respond to these environmental challenges (Outten et al.,

2004). The housekeeping ISC system supplies Fe-S clusters via its Fe-S cluster scaffold IscU. However under conditions of oxidative stress, transient clusters associated to IscU are also prone to damage thus lowering the efficiency of this system (Jang and Imlay, 2010). Instead, under oxidative stress and Fe starvation conditions, Fe-S cluster biogenesis is accomplished by the SUF system. In fact E. coli strains lacking SUF components are unable to grow in the presence of 2,2’-

87

dipyridyl and are sensitive to paraquat and hydrogen peroxide (Outten et al.,

2004).

B. subtilis and other Gram-positive bacteria lack the ISC pathway yet possess a modified SUF pathway which is thought to be the sole pathway for synthesis of

Fe-S clusters in this group of bacteria. In fact, genes encoding for the SUF components in B. subtilis are essential for survival (Albrecht et al., 2010;

Kobayashi et al., 2003). Therefore, we have investigated if expression of the housekeeping SUF components would also be subjected to regulation during iron starvation or oxidative stress. Unlike the expression profile observed for E. coli,

SufC protein levels did not change significantly with oxidative challenge, suggesting that the B. subtilis SUF pathway was being constantly expressed. Our results are in agreement with a recent transcriptomic study in B. subtilis cultured under 144 different conditions, which identified the suf genes among a small subset of coding sequences (<3% of total genome) that were always expressed at levels higher than the chromosome median (Nicolas et al., 2012). While this study indicated high levels of suf genes expression, it did not reveal conditions under which expression was further augmented. Together, these observations provide additional support for the involvement of the B. subtilis SUF system in the essential housekeeping formation of Fe-S clusters, a role that is different from the specialized function of the E. coli SUF pathway that is only expressed under stress conditions. Furthermore, B. subtilis lacks the Fe-S transcriptional regulator

IscR (Giel et al., 2013; Mettert and Kiley, 2014) and preliminary results described here suggest that this pathway may be subjected to distinct yet-unidentified

88

mechanism(s). Interestingly, depletion of BSH caused a 2-fold reduction of SufC accumulation; however expression levels increased in the presence of 2,2’- dipyridyl.

GSH has been proposed to be the major cytoplasmic iron pool based on a recent metal affinity study (Hider and Kong, 2011). In Salmonella enterica, GSH has been suggested to be a physiological chelator of the labile iron pool (Thorgersen and Downs, 2008), which could provide Fe for Fe-S cluster synthesis, export iron, present ferritin with excess iron, and supply iron directly to cytoplasmic enzymes

(Andrews et al., 2003). The cysteinyl and malyl moieties of BSH are suitable metal ligands (Newton et al., 2009). Although, it is hypothesized that BSH plays an analogous role to GSH in B. subtilis, it is plausible to consider that BSH may also participate in direct metal binding and mediate transport of Fe in bacterial species producing this biothiol.

ACKNOWLEDGEMENTS

The authors thank Dr. John Helmann and Dr. Ahmed Gaballa for kind gifts of B. subtilis wild-type, bshA, and bshC strains and for providing with insightful comments to the manuscript. This research was funded by the National Science

Foundation (MCB-1054623). PDS and ZF have no conflict of interest to declare regarding commercial or other relationships.

89

CHAPTER 4

Characterization of NifS cysteine desulfurase and its role on NAD

biosynthesis in B. subtilis

Zhong Fang, Kelsey Worthy, Lisa Hong and Patricia C. Dos Santos

Patricia C. Dos Santos conceived the idea, and provided overall direction. Lisa Hong and

Zhong Fang constructed the plasmids. Kelsey Worthy, Lisa Hong and Zhong Fang performed experiments, analyzed the data, made the graphs and wrote the paper.

90

Introduction

Nicotinamide adenine dinucleotide (NAD+) is an essential cofactor in all living organisms. It plays an important role in oxidation-reduction reactions such as glycolysis and citric acid cycle of cellular respiration (Deal, 1969). Additionally,

NAD plays key roles as substrate for certain reactions such as those in cellular signaling and in transcription regulation. For example, it acts as a precursor of cyclic ADP-ribose in the second messenger system for cellular signaling (Guse,

2004a). The produced cyclic ADP-ribose binds to ryanodine receptors and releases calcium by opening calcium channels in the membranes of organelles

(Guse, 2004b, 2005). Another example is its role as substrate for NAD- dependent deacetylases such as sirtuins that are involved in regulating transcription by deacetylating histones and modifying nucleosome structure.

Interestingly, recent studies have also shown that NAD also involved in aging and extracellular signaling pathway (North and Verdin, 2004).

NAD is made up with two nucleotides connected by a pair of phosphate groups.

Two primary schemes, salvage and de novo, are utilized by most organisms for the biosynthesis of NAD (Figure 4.1). While bacteria are capable of utilizing the salvage pathway in addition to the de novo route of NAD biosynthesis, other cells, such as those of mammals, rely entirely on the salvage pathway for NAD production. In order for a cell to utilize the salvage pathway, it must have access to the pyridine ring containing-NAD breakdown products, which are necessary to reassemble an intact NAD (Gazzaniga et al., 2009). These products, known collectively as niacin, which includes nicotinic acid, nicotinamide, nicotinic acid

91

riboside, and nicotinamide riboside. They are either imported from the extracellular environment, or recycled from pre-existing products within the cell

(Gazzaniga et al., 2009).

Figure 4.1 De novo and salvage NAD biosynthetic pathways.

When the cell is unable to proceed through the salvage pathway, the de novo pathway provides a means of the biosynthetic production of the necessary precursors to NAD. The two primary de novo pathways have been identified to date and they are distinguished from one another by their unique amino acids precursors. Most eukaryotic cells utilize the Trp-dependent biosynthetic pathway

92

by oxidizing L-tryptophan in a five-step reaction, forming a variety of transition molecules (Magni et al., 2008) (Figure 4.1). Plants and most eubacteria, including some pathogens, begin the de novo pathway with an L-aspartate precursor that proceeds through a more consistent anaerobic two-step reaction

(Saunders and Booker, 2008). Both reaction paths produce quinolinic acid, which is the first common precursor in de novo pathways for all organisms studied to date (Marinoni et al., 2008). Quinolinic acid is then converted into nicotinic acid, at which point the salvage and de novo pathways converge to produce NAD through a simple two-step reaction conserved in all cells (Marinoni et al., 2008).

Given the distinct pathways in NAD production between eukaryotic cells and bacterial cells, the L-aspartatic acid to quinolinic acid conversion steps have garnered increased interest as a potential target for future antibiotic applications

(Rizzi and Schindelin, 2002).

Species within the Bacillus genus utilize L-aspartate to proceed through the de novo pathway, and contain reaction enzymes with homologous equivalents that are found in similar gene regions of other organisms, such as Escherichia coli and Mycobacterium tuberculosis (Saunders et al., 2008). Among these species, the first step in the production of quinolinic acid involves L-aspartic oxidase

(NadB) reacting with flavin adenine dinucleotide (FAD) to oxidize L-aspartate into iminoaspartate. Iminoaspartate is an instable compound which can be spontaneously converted into oxaloacetate and ammonia with a half time of 2.5 min at 25 oC and pH of 8.0. Quinolinic synthase (NadA) then reacts with dihydroxyacetone phosphate (DHAP) to condense the iminoaspartate into

93

quinolinic acid (Figure 4.2) (Sun and Setlow, 1993). The instability of iminoaspartate under physiological condition suggested that NadA and NadB may form a complex. However, experimental validation of this proposal has not yet been reported. B. subtilis is ideal for the study and examination of this process, as it is a thoroughly studied model for low GC-content, Gram-positive bacteria, and a model organism for many pathogenic species (Marinoni et al.,

2008).

Figure 4.2 Biosynthetic conversion steps in the formation of quinolinic acid. NadB

oxidizes L- aspartate into iminoaspartate, and NadA subsequently catalyzes the condensation DHAP from iminoaspartate, resulting in quinolinic acid.

The NAD biosynthetic gene region encodes the enzymes involved in the de novo pathway in B. subtilis (Figure 4.3). The mapping of this gene region has revealed the promoter region controlling expression of nadB, nadC, and nadA genes overlaps to the promoter of nifS and nadR. Interestingly, nicotinic acid significantly reduces the transcription of both nadB and nifS, while the inactivation of nifS has a minimal effect on the transcription of nadB (Sun and

Setlow, 1993).

94

The B. subtilis L-aspartate oxidase NadB, like its E. coli homologue, is involved in the first step of de novo biosynthesis. NadB is a 60-kDa flavoprotein consisting of three domains: a FAD-binding domain, a C-terminal three helical bundle domain, and an α/β capping domain (Rizzi and Schindelin, 2002). NadB binds non- covalently to one molecule of FAD per polypeptide, which is oxidized during the production of iminoaspartate. This causes a structural change on NadB, which effectively forms a shield around its highly soluble and labile product (Rizzi and

Schindelin, 2002).

Figure 4.3 The nicotinic acid de novo gene region. The nadBCA gene region encoding for NAD synthesis proteins shares an overlapping promoter region with nadR and nifS, a transcription repressor and cysteine desulfurase, respectively.

The nadC gene following nadB encodes for the quinolinic phosphoribosyl transferase enzyme known as NadC. The NadC enzyme structure contains two domains consisting of a mixed α/β N-terminal domain and an α/β barrel domain

95

containing seven beta sheets. NadC is able to form a complex with the substrate quinolinic acid, aiding in the formation of nicotinamide adenine mononucleotide.

The quinolinate synthase (NadA), encoded by the nadA gene, contains three conserved cysteine residues that coordinate a [4Fe-4S] cluster cofactor necessary for its activity (Figure 4.4) (Marinoni et al.). Since the cluster is particularly sensitive to oxygen, NadA was identified as the site of oxygen poisoning of NAD biosynthesis for anaerobic organisms.

Figure 4.4 In vitro NadA activation. The quinolinate synthase NadA remains inactive

until it is able to bind to the [4Fe-4S] cluster cofactor. Upon activation, NadA is able to convert iminoaspartate and dihydroxyactecyl phosphate into quinolinic acid, water

and a phosphate group.

NadR, initially named as YrxA, was classified as a transcriptional regulator. This protein has no homologues in Gram-negative bacteria. However, genes for similar proteins are present in all the sequenced genomes of bacilli and other pathogenic Gram-positive bacteria. In-vitro studies revealed that NadR cooperates with nicotinic acid on repressing the transcription of nadBCA and nifS

(Figure 4.3). Primer extension analysis showed that the -10 to -35 promoter

96

regions for nadB and nifS were in fact both encompassed in a dyad symmetrical overlap, serving as a regulatory binding site for NadR.

The nifS gene encodes a hypothetical cysteine desulfurase and the deletion of nifS in B. subtilis caused a nicotinic acid auxotrophic phenotype. NifS from B. subtilis displays significant sequence similarity to the nitrogen-fixing A. vinelandii cysteine desulfurase NifS. In A. vinelandii, NifS catalyzes the conversion of cysteine to alanine and transfers the sulfur for the biosynthesis of Fe-S clusters for nitrogen fixation proteins (Figure 4.5) (Zheng et al., 1993). In A. vinelandii, sulfur mobilization for the biosynthesis of Fe-S cluster not asscociated to nitrogen fixation involves the essential cysteine desulfurase IscS. Disruption of iscS in

E.coli resulted in phenotypic growth defects including nicotinic auxotrophic phenotype (Schwartz et al., 2000). These observations suggest a connection between cysteine desulfurase and nicotinic biosynthesis pathway, which remains to be determined.

Figure 4.5 NifS cysteine desulfurase reaction.

97

Studies have shown that the formation and insertion of complementary metalloclusters is essential to the activity of nif-specific gene products in other organisms. It is possible that the nifS gene product may in fact play a key role by providing sulfur for the Fe-S cluster assembly on NadA for its activation.

Interestingly, the NadA homolog in Arabidopsis thaliana was reported to have an extra C-term SufE-like domain which is believed to participate in the sulfur mobilization for Fe-S cluster assembly. The cysteine desulfurase activity of

CpNifS in A. thaliana was shown to be enhanced by 70-fold in presence of NadA, indicating an efficient sulfur transfer from CpNifS to NadA. Moreover, the disruption of SufE domain resulted in a nicotinic acid auxotrophic phenotype similar to that of the deletion of NadA (Murthy et al., 2009). Therefore, it was hypothesized that the sulfur transfer step may be essential for the functionality of

NadA in the biosynthesis of nicotinic acid (Murthy et al., 2009).

This study proposes that the cysteine desulfurase NifS of B. subtilis plays an essential role in the formation of the [4Fe-4S] cluster on quinolinate synthase

NadA. In order to test this hypothesis, the function of NifS as a cysteine desulfurase was first characterized. Second, the ability of this enzyme ability to reconstitute the [4Fe-4S] cluster on NadA was then compared to the ability of other sulfur sources to reconstitute the NadA cofactor. To account for differences in reaction conditions between the in vitro and the in vivo conditions, the possible formation of multienzyme complexes among the NAD-related enzymes of

Bacillus subtilis was investigated.

98

Method and Materials

Chemicals

All reagents and chemicals were purchased from Fisher Scientific and Sigma-

Aldrich unless otherwise specified. Naphthalene-2,3-dicarboxaldehyde (NDA) was purchased from Anaspec.

Protein expression and purification

Bacillus subtilis NifS was cloned into the pAra13 arabinose induced expression vector. The pAra-NifS plasmid construct was transformed into E. coli CL100

(ΔiscS) strain and spread on a Luria Broth agar plate containing 100 µg/mL ampicillin (LBA). After overnight incubation at 37 °C, a colony was resuspended in 100 mL LBA medium and outgrown for 2 hours at 37 °C. The cell suspension was split into 6 x 500 mL fresh LBA medium and grown at 37 °C to OD600 of 0.2.

The culture was induced with 2 mg/mL arabinose, grown at 15 °C overnight, harvested by centrifugation at 5000 x g for 8 min and stored at -20 °C. For the purification of NifS, pelleted cells were resuspended in 25 mM Tris-HCl pH 8.0,

10% glycerol (buffer A), and disrupted using Emulsiflex-C5 high-pressure homogenizer (Avestin) by following the manufacturer’s protocol. The cell lysis solution was centrifuged at 13,200 g for 20 min and the cleared lysate supernatant was loaded onto a 45 mL Q-sepharose anion exchange column pre- equilibrated with buffer A and washed with excess buffer A until the Abs280 stabilized at a baseline. Bound proteins were eluted with a 150 mL linear gradient from 0-1 M NaCl, and samples were collected in 3 mL fractions. Eluted fractions

99

containing protein (visually inspected by the yellow color characteristic of the enzyme PLP cofactor) were collected and analyzed using sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) to determine the fractions containing a 42 kDa band, corresponding to NifS protein size. Fractions containing NifS with a purity higher than 90% were pooled, frozen in liquid nitrogen, and stored at -80 °C.

An expression plasmid containing an N-term His-tag NadA in pET16b was transformed into E. coli BL21 (DE3). An inoculum of ampicillin resistance colonies was passed into 100 mL LBA broth medium and shaken at 37 °C for 5 hours. The resulting inoculum was grown in 3 L total volume (6 flasks of 0.5 L each). At OD600 of 0.5, 2.0 g/L lactose was added to induce expression for 3 hours. The culture was harvested by centrifugation and the cell pellet was resuspended in degassed 25 mM Tris-HCl pH 8.0, 300 mM NaCl, 10% glycerol

(buffer B) in an anaerobic chamber (Coy). The resuspension was disrupted using argon gas in a high-pressure (12,000 psi) homogenizer. Cellular debris was separated by centrifugation at 13,200 g for 20 min and the cleared lysate supernatant was collected. The sample was loaded into a nickel-IMAC column

(GE healthcare) that was pre-equilibrated with buffer B in an anaerobic chamber.

Purification of the bound sample was eluted by a stepwise gradient of 5%, 15% and 50% buffer C (buffer B plus 300 mM imidazole). The elution profiles were monitored at 280 nm and fractions were analyzed using SDS-PAGE. Fractions from 50% buffer C containing pure NadA protein (with a dark brown color) were

100

pooled and sealed, before freezing in liquid nitrogen. Purified enzyme was stored at -80 °C.

To prepare apo-NadA, pure holo-NadA was incubated with 50 equivalent of

EDTA and 20 equivalent of K3Fe(CN)6. Fe-S cluster chelation was assessed by color change from brown to shiny yellow. The solution was dialyzed at 4 °C using

3 × 1 L degassed buffer B. After dialysis, the colorless protein solution was pelleted in liquid nitrogen and stored at -80 ºC until experiments involving UV-Vis absorption spectroscopy and reconstitution assays.

NadB plasmid was kindly provided by Dr. Gabriella Tedeschi. The purification procedure is described previously in Marinoni et al (Marinoni et al., 2008). Protein concentrations were determined using the Bradford method using a Bio-Rad protein assay kit and bovine serum albumin (BSA) as a standard (Bradford,

1976).

Cysteine desulfurase assay

The desulfurase activity was assayed through the quantification of sulfide end products. An assay mixture (2000 μL) was prepared containing 0.05 mg NifS enzyme, 2 mM DTT in 50 mM 3-(N-morpholino)propanesulfonic acid (MOPS) with a pH range 6.61, 7.24, 7.42, 7.51, 7.72, 7.79, 7.91, and 8.10. The reaction solution was equilibrated at room temperature for 5 min before the addition of 0.5 mM L-cysteine. Sample aliquots of 200 μL were taken at 2 min intervals and analyzed through the methylene blue assay (Selbach et al., 2010). The slopes of standardized sulfide samples were used to convert the data from (ΔA･min-1) to

101

(nmol Sulfide･min-1･mg-1 NifS) and data was fitted to Henderson-Hasselbach curve to determine the pKa value.

Desulfurase activity at varying cysteine concentrations was assayed through the quantification of both alanine and sulfide end products. Sulfide assays were completed as described above except the pH remained at 8.0, the cysteine concentration ranged between 0.02 mM to 0.25 mM, and the final data was fit to a Michaelis-Menten equation. Quantification of alanine was completed using 3 mL assay mixtures containing 0.05 mg NifS enzyme in 2 mM DTT, 50 mM MOPS buffer pH 8.0 and cysteine, unless indicated. Reactants were equilibrated at room temperature for 5 minutes before adding substrate, 200 μL reactant aliquots were quenched with 40 μL 10% trichloroacetic acid (TCA) at various time points as indicated. Collection samples were centrifuged at 13,200 rpm for 5 minutes, and 200 μL aliquots of the supernatant were transferred to 2.0 mL glass vials with 1.0 mL of derivatization solution consisting of 80 mM borate, 4.8 mM potassium cyanide, and 0.64 mM naphthalene-2,3-dicarboxaldehyde (NDA) was added to each sample. Samples were loaded and analyzed using high performance liquid chromatography (HPLC). The slope of standardized alanine samples was converted to (nmol Alanine･min-1･mg-1 NifS) based on an alanine standard curve.

Effect of the reducing agent DTT on cysteine desulfurase on activity was assayed through the quantification of the alanine end product. Assays proceeded as described by the alanine assay above except that reactions took place in 0.25

102

mM cysteine, and 2 mM DTT was included in only some reactions. The standardized alanine samples were used to establish a slope that was used to convert results from (ΔFU･min-1) to (nmol alanine･min-1･mg-1 NifS) and data was fitted to a linear equation.

Circular dichroism (CD) spectroscopy

CD spectra were performed using an Aviv CD spectrometer (Model 215; AVIV

Biomedical) with a 1-mm quartz cuvette (Hellma Analytics) and a bandwidth of 1 nm. Protein samples were analyzed at 5 µM in 10 mM phosphate buffer (pH 7.4) and scanned from 250 nm to 190 nm with 0.5 nm increments. The final spectrum of each sample is the average of 10 scans.

Analytical gel filtration

The oligomerization states of NifS and NadAB were determined using an FPLC system equipped with a 24 mL Superose 10/300 gel filtration column pre- equilibrated with buffer containing 25 mM Tris, 150 mM NaCl and 10% glycerol

(pH 8.0). Molecular markers employed included blue dextran (Mr = 2,000 kDa), conalbumin (Mr = 75 kDa), ovalbumin (Mr = 43 kDa), carbonic anhydrase (Mr =

29 kDa), ribonuclease A (Mr = 13.0 kDa), and aprotinin (Mr = 6.5 kDa).

In vitro reconstitution and activity assay

A reconstitution assay was conducted under anaerobic conditions in a glove box with 9.35 nmol apo-NadA, 60 nmol Fe2+ and different sulfur source in 100 mM

MOPS in the presence or absence of 2 mM DTT. At each reconstitution time

103

point, a 50 μL of aliquot was mixed with 750 μL of 200 mM MOPS, pH 8.0, 100

μL of 1 M ammonium chloride, and 50 μL of 100 mM dihydroxyacetone phosphate and 50 μL of 100 mM oxalate. Each reaction mixture (200 μL) was removed and quenched with 40 μL 10% TCA at 5 min, 10 min, 20 min and 30 min. The activity of NadA was detected by the presence of quinolinic acid through an HPLC-UV based method as described below (Table 1). The samples were injected into a Waters 2489 system equipped with a Zorbax C-18 (4.6 × 150 mm) from Agilent. The separation was initiated by flowing at 0.75 mL･min-1 with

100% solvent A (0.1% TFA in water) for 12 min, followed by increasing to 100% solvent B (acetonitrile) with a linear gradient in 5 min, then re-equilibrating the column through a gradient to 100% A in 5 min. There was a 5 min interval before the next injection. The quinolinic acid peak was eluted with a retention time of 5.6 min at Abs268. The concentration of quinolinic acid was determined using a calibration curve of known concentrations.

Results

Characterization of NifS cysteine desulfurase activity

NifS protein expressed in E. coli CL100 containing pDS23 plasmid was aerobically purified by ion-exchange chromatography. Elution fractions from Q sepharose column containing protein were collected and pooled. NifS purified samples resulted in a PLP occupancy as a high of 93.0%.

The kinetic behavior of NifS for the production of sulfide and alanine in reactions with varying cysteine substrate concentrations were determined (Figure 4.5).

104

The activity data from both alanine (Figure 4.6A) and sulfide (Figure 4.6B) were plotted against the reaction cysteine concentration, and both graphs displayed hyperbolic trends. The values were fit to the Michaelis-Menten equation and the resulting curves provided two sets of kinetic constants. The activity in terms of

-1 -1 alanine detection resulted in Vmax of 120.3 nmol･min ･mg , and KM of 17.1 μM.

-1 -1 The activity in terms of sulfide resulted in Vmax of 136.3 nmol･min ･mg , and KM of 23.0 µM. These data confirmed the cysteine desulfurase reaction catalyzed by

NifS proceeded by releasing equivalent amount of sulfide and alanine without producing any byproduct.

A B

Figure 4.6 Cysteine saturation curves of NifS in terms of Alanine (A) and sulfide (B) produced. Samples of NifS enzyme, DTT-reducing agent, and varying amount of cysteine were assayed in buffer and measured using HPLC coupled to a fluorescence detector and UV absorption method, respectively. The displayed lines of fit were best fit to the Michaelis-Menten equation.

The profile of the pH-activity curve of NifS was determined by observing the rate of sulfide produced in presence of 0.5 mM cysteine (Figure 4.7A). The activity of each sample was recorded in relation to the maximum rate of sulfide formation

(120 nmol･min-1･mg-1) determined previously.

105

These relative-activity values were plotted against pH, and showed that the reaction is dependent on at least one ionization event. The pH-activity profile when fitted to a Henderson-Hasselbach equation shows an associated pKa of

7.36, and indicates that NifS activity is optimum at pH of 8.0 or higher.

Effect of reducing agents on NifS activity was determined by observing the rate of alanine production at optimum pH and substrate conditions under reducing and non- reducing conditions. The activity data were plotted against time and showed a linear trend (Figure 4.7B). The reactions containing DTT resulted in a slope of 1.95 nmol Ala･min-1･mg-1, and those without DTT resulted in a slope of

0.397 nmol Ala･min-1･mg-1, consistent to values reported for other cysteine desulfurases under non-reducing conditions (Rajakovich et al., 2012; Zheng et al.,

1993).

A B

Figure 4.7 A. pH-activity profile of cysteine desulfurase NifS. The relative activity of sulfide production was calculated through normalizing to maximum activity (120 nmol S2-･min-1･mg-1). The values were fitted to a Henderson Hasselbach equation and resulted in a pKa of 7.36 ± 0.21. B. NifS production of alanine under reducing (circle) and non-reducing (diamond) conditions. Reactions were conducted in 0.5 mM cysteine, 50 mM MOPS buffer (pH 8) and 0.25 mg NifS, and when used, reducing agent 2 mM DTT was added. 106

NadA and NadB inhibit the cysteine desulfurase activity of NifS

The kinetic profile of cysteine desulfurase can be modulated in the presence of their physiological sulfur acceptors. For example, the B. subtilis Thil accepts sulfur from cysteine desulfurase NifZ to promote the biosynthesis of 4-thiouridine

(Rajakovich et al., 2012), and the NifZ cysteine desulfurase activity was enhanced by Thil only under non-reducing condition. SufU stimulates the activity of SufS under both non-reducing and reducing conditions, which promotes the sulfur mobilization for housekeeping Fe-S assembly (Selbach et al., 2010).

Interestingly, the activity of YrvO was only enhanced by its partner MnmA in presence of ATP (Black and Dos Santos, 2015a). Since apo-NadA has been hypothesized as the physiological sulfur acceptor of NifS, the catalytic rate of

NifS was determined in the presence of apo-NadA under both reducing and non- reducing conditions. Under reducing condition, the presence of apo-NadA did not affect the cysteine desulfurase activity of NifS. This observation may be because

DTT can efficiently reduce the enzyme persulfide intermediate.

Since DTT is absent in physiological conditions, the same assays were performed without DTT. Surprisingly, the addition of apo-NadA decreased the activity of NifS from 28.3 + 1.0 nmol Ala･min-1･mg-1 to 14 + 0.8 nmol Ala･min-1･ mg-1. To confirm this effect was not caused by the macromolecular crowding effect in the system, control experiments were performed with 20 equivalent of

BSA, which gave an activity of 30 + 1.2 nmol Ala･min-1･mg-1. The cysteine

107

desulfurase assay was also performed with another cysteine desulfurase NifZ in presence of apo-NadA, which did not show the inhibition effect as observed from

NifS (data not shown). Therefore, we concluded that the inhibition of apo-NadA

Figure 4.8 Cysteine desulfurase activities of NifS under various conditions. 0.05 mg (1.2 µmol) of NifS was assayed with 0.5 mM cysteine in presence of 20 equivalent of NadA, NadB, NadAB, BSA and/or 2 mM DTT. The reaction rate was measured through quantitation of produced alanine over time. on NifS is specific.

Earlier studies suggested that NadA and NadB may form a complex under physiological conditions (Griffith et al., 1975; Nasu et al., 1982). Therefore, we tested the possibility of NadB as a sulfur transporter between NifS and NadA. To test this hypothesis, 20 equivalent of NadB was added to NifS cysteine desulfurase assay in presence and absence of 2 mM DTT. The results were very similar to that of NadA, in which unchanged and decreased activity were observed under reducing and non-reducing conditions respectively. Moreover, when 20 equivalents of apo-NadA and NadB were added together, the NifS cysteine desulfurase activity was almost eliminated under non-reducing condition.

108

Protein-protein interaction studies on NifS and NAD proteins

The results from cysteine desulfurase assays revealed an unexpected inhibition effect of NadA and NadB on NifS activity. Therefore, we decided to explore the interaction between NifS and NadAB through circular dichroism (CD) and analytical gel filtration. First, we attempted to explore the possibility of detecting complex formation between NadA and NadB by using circular dichroism. As shown in Figure 4.9A, the experimental CD spectrum of apo-NadA plus NadB was indistinguishable from the theoretical spectrum, indicating that no significant change of secondary structure occurred when both proteins are mixed.

A B

C D

Figure 4.9 Far-UV circular dichroism spectra of NadA+NadB (A) NifS+NadA (B) NifS+NadB (C) and NifS+NadA+NadB (D).

109

Similar experiments were also performed for NifS with NadA, NadB or NadAB by incubating at a 1:1 ratio. These results suggest if protein-protein interaction occurred it did not result in large conformational change.

To further verify the results from circular dichroism, analytical gel filtration was used to study the oligomerization state of those enzymes as well as to investigate possible complex formation. The calculated molecular weight of NifS was 84.2 kDa, confirming that NifS is a dimer. This finding was consistent with the quaternary structures of all studied cysteine desulfurases, such as IscS, SufS,

NifS, etc. The molecular weight of NadA and NadB were estimated to be 52.7 kDa and 49.9 KDa respectively, indicating that both proteins are monomers. Gel filtration analysis of holo-NadA was performed in the glove box and displayed the same retention time as the monomer apo-NadA. Our results differ from earlier studies from Marinoni et al. showing that B. subtilis NadA formed a trimer under both aerobic and anaerobic conditions, while B. subtilis NadB formed a monomer at a NaCl concentration of 150 mM.

Incubation of NifS and NadA showed no evidence of complex formation (Figure

4.10). The NadA-NadB interaction was studied through analytical gel filtration as well as affinity column binding by Marinoni et al (Marinoni et al., 2008). However, different results were obtained from these two independent approaches. The binding experiment of NadB to His-tagged NadA or to GST-tagged NadA showed specific binding, while no complex formation was observed in gel filtration

(Marinoni et al., 2008). In our hands, we accidently found that NadB binds to the immobilized metal ion affinity chromatography column up to 90 mM imidazole.

110

This may explain the binding of NadB to His-tagged NadA as reported earlier

(Marinoni et al., 2008). In summary, neither the circular dichroism nor analytical gel filtration provides any evidence for the stable complex formation between

NifS with NadAB. One possible reason may be the misfolding of NadA and NadB from expression and purification. To rule out this possibility, enzymatic assays for

NadA and NadB were performed and both enzymes displayed comparable activity as reported data.

Figure 4.10 Analytical gel filtration chromatograms of NifS, NadA, NadB and different combinations of the three enzymes. In vitro NadA reconstitution and activity assay

As isolated NadA contained partial occupancy of Fe-S cluster as indicated by the absorption peak at 410 nm (Figure 4.11), the activity was measured to be 120 nmol QA･min-1･mg-1. Apo-NadA generated from cyanide and EDTA treatment showed no detectable activity and was subsequently used for reconstitution assay.

111

Figure 4.11 UV-visible absorption spectra of as isolated B. subtilis NadA (red line), after anaerobic treatment with cyanide and EDTA (green line) and after reconstitution (blue line).

Reactions containing Fe2+ and the sulfur source were allowed to reconstitute for varying times and then the respective NadA activities were determined by observing the rate of quinolinic acid (QA) produced. The assembly of the Fe-S cluster on NadA was observed in the presence of the cysteine desulfurases NifS,

NifZ or SufS+5 SufU with cysteine, and Na2S without cysteine. The reconstitution of NadA with free sulfide as sulfur source under reducing condition completed in

15 min with the highest activity of 120 nmol QA･min-1･mg-1 (Figure 4.12A). The reconstitution reactions with different cysteine desulfurases plus cysteine as sulfur source showed similar reconstitution profiles as sulfide with a highest rate of 180 nmol QA･min-1･mg-1. Moreover, the activation curve of NifS was undistinguishable to other cysteine desulfurases.

112

The NadA reconstitutions under non-reducing condition were more diverse compared to reducing condition. With the free sulfide as sulfur source, the reconstitution remained very low up to 45 min with the activity of NadA around 2 nmol QA･min-1･mg-1. The removal of DTT also diminished the reconstitution rates of NifZ and SufSU to 14 nmol QA･min-1･mg-1 and NifS to 30 nmol QA･min-1

･mg-1. As suggested from Figure 4.12A, the activity of NadA could reach as high as 180 nmol QA･min-1･mg-1, however, all reconstitutions stopped at a much lower activity under non-reducing condition. Considering the excess amount of free sulfide or continuous production of sulfur, it is unlikely that sulfur mobilization caused the early termination. The possible reason may be the limited reducing ability of cysteine which provided electrons for Fe-S cluster assembly. Overall,

NifS showed a 2-fold reconstitution level compared to NifZ and SufSU under non- reducing condition.

A B

Figure 4.12 Reconstitution of apo-NadA in presence of various sulfur sources under reducing (A) and non-reducing (B) condition. Reactions included 9.35 nmol apo- NadA, 60 nmol Fe2+ and different sulfur sources including 0.62 nmol NifS (square), NifZ (circle), SufS ( plus 3.1 nmol SufU) (triangle) with 2.5 mM cysteine or 225 nmol

sulfide (diamond) . Reconstitution under reducing condition included 2 mM DTT. The activities were measured as described in Method and Materials.

113

The effect of NadB on NadA reconstitution was also investigated by adding equivalent amount of NadB into the reaction using NifS and cysteine as sulfur source. As shown in Figure 4.13, the addition of NadB did not cause any significant change on the reconstitution profile under both reducing and non- reducing conditions.

A B

Figure 4.13 Reconstitution of apo-NadA in presence of equivalent NadB under

reducing (A) and non-reducing (B) conditions. Reconstitution assays were performed with 9.35 nmol apo-NadA, 9.35 nmol NadB, 60 nmol Fe2+, 0.62 nmol NifS and 2.5 mM

cysteine in presence or absence of 2 mM DTT. The activities were measured as described in Method and Materials.

Reconstitution of apo-NadA with BSH as the reducing reagent

In B. subtilis, bacillithiol was identified as the major LMW thiol, which may play a role as intracellular redox buffer (Helmann, 2011). Initial studies on BSH also revealed its key role on Fe-S metabolism and biosynthesis. Herein, the reconstitution of apo-NadA was carried out in presence of 2 mM BSH. As shown in Figure 4.14, apo-NadA was reconstituted over time and the highest activation was achieved after 10 min with a specific activity of 60 nmol QA･min-1･mg-1. This

114

confirms the ability of BSH as a reducing reagent for the assembly of Fe-S cluster. However, the activation of BSH is only half efficient compared to physiology-unrelated DTT.

Figure 4.14 Reconstitution of apo-NadA with Fe and sulfide in presence of 2 mM DTT

(square), BSH (triangle) or none (circle). Reconstitution assays were performed with 2+ 9.35 nmol apo-NadA, 60 nmol Fe , 225 nmol sulfide and 2 mM DTT or BSH. Aliquots of reconstitution were applied for NadA activity assay coupled with HPLC for the determination of quinolinic acid.

Discussion

The NifS enzymes were initially proposed to be dedicated to mobilize sulfur for the biosynthesis of metal-thiol cofactors for enzymes involved in the nitrogen- fixation pathway (Johnson et al., 2005b; Zheng et al., 1993). However, NifS homologs were also found in species lacking the unique nitrogen metabolism

(Lacourciere et al., 2000; Sun and Setlow, 1993), suggesting an alternative role for this class of cysteine desulfurases. In B. subtilis, nifS locates next to the genes involved in nicotinic acid biosynthesis and is under the repression of nicotinic acid. The deletion of nifs caused a nicotinic acid auxotrophic phenotype

115

which also showed up in the E.coli iscs deletion strain (Lauhon and Kambampati,

2000; Sun and Setlow, 1993). Based on the function of NifS and the enzyme components in nicotinic acid biosynthesis pathway, we hypothesized that NifS is specifically involved in the Fe-S cluster assembly on NadA. The cysteine desulfurase activity of NifS was measured through the detection of alanine and sulfide under reducing conditions. Both the activity of 120.3 nmol･min-1･mg-1 and

KM of 17.1 μM from alanine assay were comparable to other reported cysteine desulfurases. The addition of NadA did not enhance the NifS cysteine desulfurase activity, instead decreased the activity by 50%. The inhibition of cysteine desulfurases by partner proteins is not common, one example is the inhibition of E.coli SufS activity by SufE under reducing condition when cysteine concentration is lower than 300 µM (Dai and Outten, 2012). This inhibition has been suggested to be a regulatory mechanism for SufS activity under low cysteine concentration (Dai and Outten, 2012). However, it is not likely the same case for NifS-NadA based on the following two reasons, first, NifS showed no interaction with NadA or/and NadB as indicated by circular dichroism and analytical gel filtration, while SufS and SufE form a complex. Second, it is not reasonable to have an oxygen-sensitive regulator to control a pathway independent of oxidative stress. In mammalian systems, the single turnover assays on NifS1-Isd11 showed an inhibition upon incubation of IscU, however, the further addition of frataxin stimulated the single turnover dramatically (Parent et al., 2015). Initial studies on frataxins suggested roles for iron storage, iron delivery and Fe-S cluster assembly (Adinolfi et al., 2009; Bou-Abdallah et al.,

116

2004). Since the assembly of Fe-S cluster requires both iron and sulfur, the involvement of iron donor may facilitate the turnover of cysteine desulfurase. In B. subtilis, a frataxin homolog ydhG has been studied and reported its roles on intracellular iron channeling and Fe-S cluster biosynthesis (Albrecht et al., 2011).

Future experiments are aimed to include ydhG in the cysteine desulfurase assay as well as protein-protein interaction studies.

The reconstitution of NadA with a different sulfur source under reducing condition showed a similar time-dependent activation. A highest activity of 180 nmol QA･ min-1･mg-1 was observed after 15 min. This activity is the highest value ever reported NadA in all species (Saunders et al., 2008). The cysteine desulfurase activities for NifS, NifZ and SufSU at saturated cysteine concentrations in the presence of DTT from previous studies were reported to be 120, 240 and 500 nmol Ala･min-1･mg-1, respectively (Rajakovich et al., 2012; Selbach et al., 2010).

In each reaction, 0.026 mg of cysteine desulfurase was added, which resulted in a rate of sulfide production of 3.12, 6.24 and 13 nmol･min-1 for NifS, NifZ and

SufSU respectively. Therefore, it would require approximately 12, 6 and 3 min respectively for NifS, NifZ and SufSU to produce 36 nmol of sulfur needed to fully activate 9 nmol of NadA. As shown in Figure 4.12A, the complete activation for all cysteine desulfurases was about 12 min, therefore sulfur mobilization step is not the rate-limiting step for Fe-S assembly on NadA. In E.coli, CsdA-CsdE system has been identified as a preferred sulfur source on NadA reconstitution under reducing condition (5 mM DTT) compared to SufS-SufE and IscS (Loiseau et al., 2005). However, in vivo studies suggested that only iscS deletion caused a

117

nicotinic acid auxotrophic phenotype (Lauhon and Kambampati, 2000). This inconsistency may due to the following facts on the reconstitution conditions: 1. physiological unrelated DTT was used in the assay; 2. none of scaffold enzymes were used, such as IscU, NfuA, SufBCD, etc.; 3. the intracellular iron donor was missing (Loiseau et al., 2005). The reconstitution experiments under non- reducing condition indicated that NifS was a preferred sulfur source with a two- fold reconstitution level compared to NifZ and SufSU. We also attempted to reconstitute NadA with the physiological related BSH, which gave a 50% reconstitution compared with DTT. This is the first report showing the capacity of

BSH as a reducing reagent in in vitro Fe-S assembly. Up to now, the major Fe-S scaffold protein has not been identified in B. subtilis, and SufBCD was suggested to be the candidate (Selbach et al., 2014). Besides SufBCD in Suf biosynthetic pathway, Nfu was also able to efficiently assemble and transfer Fe-S cluster in vitro. Future reconstitution will be performed with BSH as reducing reagent in presence of scaffold proteins.

118

CHAPTER 5

Functional Studies of Nfu in Bacillus subtilis

Zhong Fang, Hana Choi and Patricia C. Dos Santos

The work contained in this chapter including figures and schemes, was drafted by

Zhong Fang, and edited by Patricia C. Dos Santos. Zhong Fang performed plasmid construction, proteins purifications, cluster transfer assays. Hana Choi conducted the complementation experiments.

119

Introduction

Iron-sulfur (Fe/S) proteins are essential to all living cells (Johnson et al., 2005a;

Rouault, 2012). Owing to the structural plasticity and versatile chemical features of Fe-S clusters, Fe-S proteins are involved in a variety of biochemical pathways including enzymatic catalysis, electron transport, and regulation of gene expression (Bandyopadhyay et al., 2008a; Johnson et al., 2005a).

The biosynthesis of Fe-S cluster usually contains three major steps: mobilization of sulfur from cysteine, assembly of Fe-S clusters on a scaffold enzyme and transfer Fe-S clusters to target proteins (Ayala-Castro et al., 2008; Roche et al.,

2013; Rouault, 2012) (Figure 5.1).

Desulfurase

Figure 5.1 Biosynthetic scheme of Fe-S cluster. The cysteine desulfurase converts cysteine to alanine and release one sulfur which interacts with an iron donor to form

an Fe/S cluster on a scaffold protein. The cluster is transferred using transfer proteins to the apo protein (inactive) to its holo from (active).

The first step is catalyzed by cysteine desulfurase which has a characteristic pyridoxal-5’ phosphate (PLP) cofactor in the enzyme active site (Mihara and

Esaki, 2002). The reaction occurs through the formation of a cysteinyl persulfide from nucleophilic attack by the active-site cysteine thiolate anion on the sulfur of

120

the cysteine-PLP adduct, releasing alanine in the process. This step leads to the formation of a cysteinyl persulfide which can then be transferred to sulfur acceptor proteins by another nucleophilic attack involving a reaction similar to thiol-disulfide exchange (Black and Dos Santos, 2015b). In the biosynthesis of

Fe-S clusters, the sulfur is transferred to scaffold protein. A scaffold is defined as a protein that binds transiently Fe-S clusters and transfers them to apotargets

(Chandramouli et al., 2007). Up to now, three types of scaffold proteins, SufB- type, A-type and U-type, have been found in most Fe-S biosynthetic pathways

(Py and Barras, 2010). The first Fe-S cluster scaffold identified was the

Azotobacter vinelandii NifU involved in the initial assembly of Fe-S clusters for proteins involved in nitrogen fixation. NifU is organized with an N-terminal domain, a central domain and a C-terminal domain (Yuvaniyama et al., 2000). The N- terminal domain, subsequently referred to as an IscU-type domain, contains one

Fe-S cluster that is transiently formed and can be transferred to apoproteins

(Agar et al., 2000; Nishio and Nakai, 2000). The central domain contains a Fe/S cluster that is stable, disqualifying this region to act as a scaffold. The C-terminal domain, subsequently referred to as the Nfu domain, is able to assembly a Fe/S cluster that is transferrable to apoproteins (Bandyopadhyay et al., 2008b; Nishio and Nakai, 2000). The A-type scaffold proteins are found associated with the ISC,

SUF, and NIF machineries in Fe-S cluster biosynthesis pathways. They contain three highly conserved Cys residues (C-X42-44-D-X20-C-G-C) in their C-terminal regions (Ollagnier-de-Choudens et al., 2004). The exact role of the A-type proteins in Fe-S cluster assembly is not clear. Two hypotheses concerning the

121

biochemical function of the A-type proteins have recently emerged. The first suggests that the A-type proteins are alternative Fe-S scaffolds. The second hypothesis is that the A-type proteins are actually iron chaperones that donate iron for Fe-S cluster assembly on U-type scaffolds. Another protein designated as NfuA with an N-terminal domain similar to A-type scaffold and a C-terminal domain similar to C-terminal domain of NifU was found among bacteria, archaea, and eukarya (Figure 5.2) (Angelini et al., 2008; Bandyopadhyay et al., 2008b).

Figure 5.2 Schematic representations of domain structures of NifU, NfuA, IscA, Nfu, and IscU proteins from bacteria. Azotabacter vinelandii (Av), E. coli (Ec), B. subtilis (Bs) and S. aureus (Sa).

In the Gram-negative Escherichia coli, the N-terminal domain of NfuA shares 24–

28% identity with the A-type Fe-S proteins of E. coli (i.e. IscA, SufA, and ErpA).

However, it does not contain three conserved cysteine residues as Fe-S cluster ligands. The C-terminal domain exhibits 21% identity with the Nfu domain of NifU from A. vinelandii and contains the conserved CXXC motif. The deletion of nfuA in E. coli did not cause any growth defect under normal growth conditions, but did

122

cause an severe slow growth in presence of paraquat or 2,2-dipyridyl stress

(Angelini et al., 2008). A similar phenotype was observed in A. vinelandii ∆nfuA, in which the mutant was more sensitive to oxygen stress (Bandyopadhyay et al.,

2008b). Gene inactivation also showed that both domains of NfuA are essential for its in vivo function. Complementary studies with different cysteine variants of

NfuA in E. coli ∆nfuA strain implied that the cysteine residues of the N-terminal domain are dispensable and two cysteines on the C-terminal Nfu domain are essential for this protein function under oxidative stress and iron starvation conditions (Angelini et al., 2008). Further in vitro analysis has revealed NfuA’s ability to bind a 4Fe-4S cluster and to serve as a carrier to transfer clusters to apoproteins (Angelini et al., 2008; Bandyopadhyay et al., 2008b).

In Gram-positive bacteria, the NfuA orthologous enzymes contain only the Nfu domain which shares a 20-30% similarity along with two conserved cysteine residues (Figure 5.3).

Figure 5.3 Sequence alignments of B. subtilis Nfu with E.coli NfuA U domain and A.vinelandii NfuA U domain.

123

Initial studies of Nfu in S. aureus have revealed its involvement in iron metabolism, ROS management and virulence. In-vitro experiments also suggested its ability to bind an Fe-S cluster and transfer the cluster to apo- aconitase in a very slow rate. Herein, we investigated the role of Nfu through complementation studies of B. subtilis nfu in an E. coli ∆nfuA strain under superoxide stress and iron starvation. Furthermore, Nfu was purified and characterized for its ability to bind and transfer Fe-S clusters.

Method and Materials

Plasmid construction and protein purification

The B. subtilis nfu gene was amplified by PCR using P1 and P2 primers, and ligated into TOPO-TA cloning vector, yielding pDS51 plasmid. The EcoR1 fragment from pDS51 containing nfu was ligated into the EcoR1 site of pUC7 plasmid to construct PUCnfu plasmid designed as pDS52. Single site- mutagenesis using primer set P3/P4 and P5/P6 on pDS52 yield C82A and C79A mutant in pUC7 vector and labeled as pDS239 and pDS240, respectively. The nfu gene amplified with P7 and P8 was cloned into TOPO-TA vector to generate pDS237, followed by digesting with Nde1 and BamH1 to ligate into Nde1/BamH1 sites in pET16b, yielding pET16nfu-Nt plasmid labeled as pDS238. All the primers used in this study are listed in Table 5.1.

Complementation study of B. subtilis nfu in E.coli ∆nfuA strain

The E. coli MG1655 wild type strain and E. coli ∆nfuA strain (SA002) were kindly provided by Dr. Frederic Barras (Angelini et al., 2008). The plasmids containing nfu and nfu variants were transformed into MG1655 and SA002 strains. Single

124

colonies from transformation were grown overnight in LBA medium at 37 oC and

4 µL of overnight culture with 10-fold serial dilutions starting from 10-1 were spotted on LBA plate containing 0.1 mM IPTG as well as different stress triggers.

The plates were incubated at 37 oC overnight before analysis.

Table 5.1 Oligonucleotides used for plasmids construction.

Enzymatic assays of aconitase, glutamate synthase and malate dehydrogenase

MG1655 and SA002 containing nfu genes were grown in Spizizen’s MM plus 0.1 mM Isopropyl β-D-1-thiogalactopyranoside (IPTG) until OD600 of 0.8-1.0, then challenged with 100 µM paraquat for 30 min with shaking before harvest by centrifugation. Cell pellets were washed and resuspended with 25 mM Tris, 150 mM NaCl, pH 8.0, and then disrupted by sonication under anaerobic condition.

The clear cell lysates were collected by centrifugation and protein concentrations were measured using Bradford method with BSA as standards.

Glutamate synthase (GOGAT) assays were conducted using the protocol as described in reference (Outten et al., 2004) with minor modifications. To ensure consistent reaction conditions, 200 mM Tris-HCl (pH 8.0), 100 mM buffered α-

125

ketoglutarate, and 8 mM NADPH were mixed in a ratio of 32:2:1 to make a stock reaction mix. In each assay, 875 µL of fresh stock reaction mix and 75 µL of clear cell lysate were injected into a sealed cuvette filled with argon gas. The decreasing slope of absorbance at 340 nm (NADPH oxidation) was recorded as

S1, and then 50 µL of 0.2 M L-glutamine was added and another decreasing slope was recorded as S2. The specific activity of GOGAT was calculated by the following equation:

S2 × 1.05 − S1 (μ · · ) = 0.00622 × mg total protein

Aconitase (ACN) catalyzes a reversible reaction between citrate and isocitrate through an intermediate cis-aconitate which has absorption at 240 nm (Alen and

Sonenshein, 1999). Each assay reaction contained 700 µL of 100 mM Tris-HCl, pH 8.0, 50 µL of cell extract and 100 µL of 200 mM Tris-Na-Citrate. The increase in absorbance at 240 nm was measured and a millimolar extinction coefficient of

3.4 mM-1.cm-1 was used to calculate the activities. Control reactions were performed without substrate.

The activity of malate dehydrogenase (MDH) was measured by following the consumption of NADH at 340 nm (Siegel, 1969). The reaction mix contained 50

µmol of Tris-HCl (pH 7.4), 0.38 µmol of oxaloacetate, 0.15 µmol NADH and 50 µL of cell extract in a total volume of 1 mL. The decrease in absorption at 340 nm was measured over time.

Purification of Nfu

126

E. coli arctic (DE3) cells transformed with plasmid pDS238 were grown in LB medium with 0.1 mg/mL ampicillin at 30 ºC until they reached OD600 of 0.4-0.6, at which point, 1 g of lactose was added and cultures were further grown at 15 oC for 20 h before harvest by centrifugation. Cells were resuspended with 25 mM

Tris, 150 mM NaCl, and 10% glycerol, pH 8.0 (buffer A), and then lysed using an

EmulsiFlex-C5 high pressure homogenizer. The cell debris collected after homogenization was incubated with buffer A plus 6 M urea (buffer D) for 1 h at

RT with continuous stirring. The resulting sample was then centrifuged at 12,000 g for 20 min, and clear supernatant was loaded onto a nickel-IMAC column pre- equilibrated with buffer D. Bound proteins were eluted with a gradient of 30%, 50% and 100% of buffer B (buffer A containing 300 mM imidazole and 6 M urea). Pure protein fractions were combined and dialyzed in a series of buffers containing 4

M urea, 2 M urea, 0 M urea subsequently to remove the urea. Finally pure proteins were frozen as aliquot and kept as refolded Nfu in -80 oC. Protein concentration was determined by Bradford Assay using BSA as the standard.

Nfu reconstitution and Fe-S cluster transfer to apo-aconitase

Nfu was reconstituted in a Coy anaerobic chamber at RT, and a typical reaction contains 0.33 mM Nfu, 2.64 mM Fe2+, 2.7 mM cysteine, 1.3 µM NifS and 2 mM

DTT. After 1h incubation, the sample was passed through a Zeba spin desalting column to remove the excess iron and sulfur.

Apo-aconitase was prepared by incubating recombinantly expressed and purified

A. vinelandii aconitase A (AcnA) with EDTA and potassium ferricyanide as

127

described previously (Bandyopadhyay et al., 2008b). The time-dependent cluster transfer from holo-Nfu to apo-AcnA was performed by mixing 40 µM holo-Nfu and

4 µM of AcnA in a sealed vial pre-purged with argon gas. At each time point, 100

µL of aliquot was taken and assayed with 800 µL of 25 mM sodium citrate in 200 mM Tris, pH 8.0. The formation of intermediate cis-aconitate as indicated by the increase of absorption at 240 nm was recorded and transformed into activity by using a molar absorption coefficient of 3400 mM-1cm-1. To investigate the ratio of holo-Nfu/apo-AcnA on activation, mixtures contained 4 µM of apo-AcnA and between 0 and 24 µM of holo-Nfu were incubated at RT for 15min, 100 µL of sample was taken to perform enzyme assay as described above.

Results

B. subtilis nfu cannot complement the phenotype of E. coli ∆nfuA

The MG1655 and ∆nfuA strains displayed similar growth after overnight incubation under non-stress condition (Figure 5.4A). However, a growth defect was observed with ∆nfuA strain upon addition of 200 µM paraquat or 400 µM

2,2’-dipyridyl. These results were consistent with the previous report, suggesting that NfuA was required for E. coli to grow under oxidative stress and iron starvation conditions. The overexpression of B.subtilis Nfu in ∆nfuA strain did not restore the growth defect (Figure 5.4A). Interestingly, when B. subtilis Nfu was overexpressed in MG1655, it caused an inhibition effect on the growth under oxidative stress and iron starvation, but not normal growth condition (Figure

5.4B). The mutation on either cysteine residues on Nfu eliminated the growth

128

inhibition, indicating that the essentiality of two cysteine residues are essential for toxic phenotype.

A MG×puc7 MG×PDS52

∆nfuA×puc7 ∆nfuA×PDS52

B MG×PDS52

MG×PDS52-C79A

MG×PDS52-C81A

Figure 5.4 A. Complementation analysis of B.subtilis Nfu in wild type (MG) and ΔnfuA E. coli strains under oxidative stress and iron starvation conditions. The plasmid puc7 (empty vector) or PDS52 (nfu in puc7) was transformed into ΔnfuA and wild type (MG) E. coli strains. Cultures were spotted on LB medium plates containing ampicillin, 0.1 mM of IPTG and paraquat or 2,2’-dipyridyl. Growth was analyzed after overnight incubation at 37 °C.

Previous reports on E. coli and A. vinelandii NfuA suggested its role as a carrier protein to transfer cluster to apoproteins. Therefore, we hypothesized that the overexpression of B. subtilis Nfu may cause a defect on Fe-S metabolism by disrupting the cluster transferring of NfuA. To test this hypothesis, we performed enzymatic assays for Fe-S containing enzymes ACN and GOGAT and non-Fe-S

MDH. As shown in Figure 5.5, the overexpression of B. subtilis Nfu in MG1655 strain led to a 60% decrease of GOGAT activity and a 20% decrease of ACN activity. This observation was consistent with the inhibition effect as shown in

Figure 5.4B. However, the overexpression of Nfu C79A and C82A variants in

MG1655 also displayed decreased activity for GOGAT or/and ACN. Since no

129

growth defects were when C79A or C82A was overexpressed (Figure 5.4B) is unlikely caused by disruption of Fe-S metabolism.

Figure 5.5 Enzymatic activities of GOGAT (black), ACN (green), and MDH (red) in MG1655 and ∆NfuA strains containing different plasmids. Cells were grown in

Spizizen’s MM plus 0.1 mM IPTG until OD600 reached 1.0. Then 100 µM of paraquat was added and challenged for 30 min before harvested by centrifugation. Cell pellets were washed and disrupted anaerobically. Clear cell lysate was used for assays as described in Method and Materials.

Nfu binds 4Fe-4S cluster

When Nfu was exogenously expressed in E. coli cells, the majority of recombinant proteins appeared to be insoluble and formed inclusion bodies.

Therefore, we attempted to use urea to unfold and refold the insoluble Nfu. As analyzed by the SDS-PAGE, 6 M urea was the optimal concentration to achieve the most solublization (data not shown). The as isolated enzyme from purification displayed a light brown color without obvious UV absorption feature

(Figure 5.6). After serial dialysis, the brown color disappeared and the proteins became colorless. The absorption spectrum of chemically reconstituted holo-Nfu

130

(Figure 5.6) showed broad absorbance bands between 300 and 600 nm that are typical of sulfur → iron charge transfer bands of Fe-S proteins. The presence of maxima in the 320- and 420-nm regions suggested that holo-Nfu harbored 4Fe-

4S clusters.

Figure 5.6 UV-visible absorption spectra of as isolated Nfu, apo-Nfu and reconstituted.

The stability of the Fe-S clusters in holo-Nfu was examined by exposing holo-

NfuA to oxygen. Exposure of holo-NfuA to air levels of oxygen led to a 10% decrease in absorption between 320 and 420 nm within 2 h (data not shown).

This result indicated that the Fe-S cluster(s) in B. subtilis holo-Nfu was relatively resistant to oxygen compared to NfuA from Synechococcus sp. PCC 7002 and

Arabidopsis thaliana, which showed a 50% decrease within 2 h exposure to oxygen.

Nfu was able to efficiently transfer cluster to apo-AcnA

Next, we examined whether holo-Nfu can perform similar function as E.coli NfuA

131

to transfer cluster to apo-proteins. The incubation of holo-Nfu and apo-AcnA at a

10:1 ratio resulted in a full activation of AcnA after 6 min (Figure 5.7A), which was similar to reported reconstitution profiles of E.coli NfuA and A. vinelandii

NfuA on apo-AcnA.

Figure 5.7 A. apo-AcnA was incubated with reconstituted Nfu at room temperature under anaerobic conditions, and aliquots from different time points were taken for activity assays as described in Method and Materials. B. apo-AcnA activation as a function of the concentration of reconstituted Nfu.

Interestingly, the C-terminal Nfu domain in E. coli NfuA was able to bind a 4Fe-

4S cluster and transfer it to apo-AcnA at a similar rate as NfuA. This observation implied that only Nfu domain was required for the Fe-S binding and transfer. As shown in Figure 5.7B, the reconstitution of apo-AcnA after 15 min of highest reconstitution was achieved at a molar ratio around 4:1. Considering the iron and sulfur contents in holo-Nfu, it was concluded that the [4Fe-4S] cluster transfer from Nfu to apo-AcnA occurred with a 1:1 stoichiometry.

Discussion

132

The Fe-S biosynthetic pathways in different species have a similar array of reactions, including mobilization of sulfur by cysteine desulfurase, assembly of clusters on scaffold proteins and transferring clusters to apo-proteins. In B. subtilis and other Gram-positive bacteria, the SUF system consists of SufSUBCD enzymes is the only complete Fe-S biosynthetic pathway. Initial studies on the

SUF machinery from B. subtilis revealed SufS as a cysteine desulfurase and

SufU as a zinc-dependent sulfur-transferase which mediates sulfur transfer from

SufS to scaffold protein. The SufBCD was suggested to form a complex and function as the scaffold protein, but further studies are still required to test this hypothesis. The last step of Fe-S assembly involves the transfer of cluster to apoproteins through carrier proteins. In B. subtilis, the A-type protein SufA and

Nfu-type protein Nfu are potential candidates for this unique function. Inactivation of sufA in B. subtilis showed on growth phenotype or defects on Fe-S enzyme activities. In this study, we successfully purified Nfu and performed reconstitution and characterization on this enzyme. It was concluded that Nfu was able to bind

4Fe-4S cluster as well as transfer to apo-AcnA with a comparable rate to NfuA or other carrier proteins.

Heterologous trans-complementation experiments showed that B. subtilis Nfu could not rescue the growth defect of deletion of NfuA under oxidative stress or iron starvation. This observation emphasized the importance of A-domain for the in-vivo function of NfuA. Recently, the C-term Nfu domain of NfuA domain was separately purified and characterized. The results indicated that C-term Nfu domain was capable of transfer cluster, but the N-term A-domain could enhance

133

the transfer. Moreover, the N-term A domain was important for the interaction between NfuA and apoproteins. This may explain why B. subtilis Nfu could not complement the defects caused by deletion of NfuA. Interestingly, a slow growth defect was observed under paraquat or iron starvation condition when Nfu was overexpressed in wild-type E.coli strain. We hypothesized that apo-Nfu may over compete apo-NfuA on Fe-S binding, which results nontransferable Fe-S on Nfu to apoproteins due to lacking of A domain. However, the enzymatic assays on

GOGAT and ACN disqualified this hypothesis since decreased activities were observed for both wild type Nfu and Nfu variants.

Besides our studies of Nfu in B. subtilis, the function of Nfu has also been recently investigated in S. aureus. The studies found that the nfu mutant accumulated both increased intracellular non-incorporated Fe and endogenous reactive oxygen species resulting in DNA damage. In addition, a strain lacking

Nfu was necessary for s. pathogenesis. It is anticipated that B. subtilis Nfu functions in similar role as the S. aureus Nfu. Future research will be targeted to determine the interaction of Nfu with the Suf components as well as potential Fe-

S cluster acceptor proteins.

134

CONCLUDING REMARKS

The biosynthesis of BSH was proposed to occur through a three-step pathway catalyzed by BshA, BshB and BshC, respectively. The kinetic studies on the

BshB enzyme(s) in B. anthracis have provided biochemical evidence to support that multiple enzymes are able to catalyze the BshB deacetylation reaction.

These results confirmed the in vivo data reported previously, moreover, suggested the existence of BSH-dependent mercapturic detoxification pathway catalyzed by Bca enzyme. Activity assays on the two deacetylases, BC1534 and

BC3461, from B. cereus have shown their capacity as BshB enzyme for the biosynthesis of BSH, which is the first report on the BSH biosynthesis in B. cereus.

The function of BSH was studied extensively by utilizing a BSH null B. subtilis strain. The growth experiments in MM revealed that deletion of BSH caused a slow growth phenotype which can be recovered by addition of Gln/Glu/Leu/Isoleu or iron. This result facilitated the following discovery of an overall decreased Fe-S enzyme activities in BSH deletion strain. The involvement of BSH in Fe-S biosynthesis was investigated by westernblot on the two components, SufC and

SufB, in SUF Fe-S machinery. The results indicated a reduced expression level of both enzymes in BSH deletion strain. However, the mechanism of BSH’s involvement on Fe-S biosynthesis is still unknown. To data, the regulation of SUF system in B. subtilis is still not clear, and B. subtilis lacks the regulatory systems presented in other species, such as E. coli, M. tuberculosis, T. potens, etc.

135

Besides the roles of BSH on Fe-S metabolism, our studies also suggested a protective role of BSH against metal toxicity and oxidative stress. Along the thiol analysis with Fe-S assays under stress conditions, it was concluded that BSH could protect cells from metal toxicity by altering the redox status as well as modifying protein thiols. In the presence of paraquat stress, an elevated content of intracellular manganese was observed in BSH-deletion strain, but not in the wild type strain. This elevated manganese has been associated with the increased activity of MnSOD, which is a common response under superoxide stress. Our results demonstrated that the induction of MnSOD under paraquat stress is required only when BSH is absent.

Cysteine desulfurases catalyze the sulfur mobilization on cysteine and transfer sulfur for the biosynthesis of biological thiol-containing cofactors. Initial studies have identified a nicotinic acid auxotrophic phenotype when a cysteine desulfurase-like enzyme, NifS, was deleted in B. subtilis. Our hypothesis is that

NifS directly mobilizes sulfur for the biosynthesis of Fe-S cluster on quinolinate synthase (NadA) involved in nicotinic acid biosynthetic pathway. Circular dichroism and analytical gel filtration analyses did not provide evidence for a strong interaction between NifS and NadA. Interestingly, a specific inhibition effect of apo-NadA on NifS cysteine desulfurase activity was observed under non-reducing condition. However, the basis for this inhibition is still unknown. The

Fe-S reconstitution on NadA has been performed with a variety of sulfur source under both reducing and non-reducing conditions. The results showed non- distinguishable reconstitution rate for all sulfur sources under reducing condition,

136

while NifS was a preferred sulfur source under non-reducing condition. Future reconstitution experiments will include iron source, physiological reducing reagent and other components in Fe-S cluster assembly to mimic the in vivo

NadA reconstitution. Besides the biochemical assay, the overexpression of his- tag NadA through a xylose-inducible vector in ∆nifS strain will provide an in vivo validation for this hypothesis.

The biosynthesis of Fe-S cluster has been extensively studied in Gram-negative bacteria, but not in Gram-positive bacteria. An Nfu-type enzyme in B. subtilis has been purified and characterized as a Fe-S cluster carrier enzyme. Initial studies demonstrated that Nfu-type enzymes are important only under oxidative stress and iron starvation. Having considered the function of BSH under oxidative stress and iron starvation, it is worth to explore the links between Nfu and BSH on Fe-S cluster metabolism and metal homeostasis.

In summary, BSH is the unique LMW thiol in low GC Gram-positive bacteria. Our studies have demonstrated its importance on oxidative stress management, metal homeostasis, xenobiotic detoxification and Fe-S metabolism. However, mechanistic studies are still needed to understand how BSH participates in those pathways. Also, it will be of great interest to investigate the potential of BSH as the reducing agent or iron donor in Fe-S cluster biosynthesis or repair.

137

Strains and plasmids used for this study

Strains/plasmids Relevant genotype/features Reference CU1065 trpC2 attSPβ Gaballa et al., 2010 HB10002 CU1065 bshA::mls Gaballa et al., 2010 HB110079 CU1065 bshC::kan Gaballa et al., 2010 HB110091 HB11079 amyE::Pxyl-bshC Gaballa et al., 2010 MG1655 F- lambda- ilvG- rfb-50 rph-1 Angelini et al., 2008 SA001 MG1655 ∆nfuA::kan, KmR Angelini et al., 2008 BA1557 (his) BA1557 from B. anthracis into pET28A Parsonage et al., 2010 BA1557-D14A BA1557D14A in pET28A This study BA1557-H110A BA1557H110A in pET28A This study BA1557-R109K BA1557R109K in pET28A This study BA1557-R53K BA1557R53K in pET28A This study BA3888 (his) BA3888 from B. anthracis into pET28A This study BC1534 (his) BC1534 from B. cereus into pET26 Deli et al., 2010 BC3461 BC3461 from B. cereus into pRSETA Deli et al., 2010 pDS23 NifS from B. subtilis into pAra13 This study pDS95 (his) NadA from B. subtilis into pET16b This study NadB NadB from B. subtilis into pET-Duet Marinoni et al., 2008 pDS52 Nfu from B.subtilis into pUC7 This study

pDS240 NfuC79A in pUC7 This study

pDS239 NfuC81A in pUC7 This study pDS238 (his) Nfu from B.subtilis into pET16b This study

\

138

REFERENCES

Adinolfi, S., Iannuzzi, C., Prischi, F., Pastore, C., Iametti, S., Martin, S.R., Bonomi, F., and Pastore, A. (2009). Bacterial frataxin CyaY is the gatekeeper of iron-sulfur cluster formation catalyzed by IscS. Nat Struct Mol Biol 16, 390-396. Agar, J.N., Krebs, C., Frazzon, J., Huynh, B.H., Dean, D.R., and Johnson, M.K. (2000). IscU as a scaffold for iron-sulfur cluster biosynthesis: Sequential assembly of [2Fe-2S] and [4Fe-4S] clusters in IscU. Biochemistry-Us 39, 7856- 7862. Albrecht, A.G., Landmann, H., Nette, D., Burghaus, O., Peuckert, F., Seubert, A., Miethke, M., and Marahiel, M.A. (2011). The frataxin homologue Fra plays a key role in intracellular iron channeling in Bacillus subtilis. Chembiochem 12, 2052- 2061. Albrecht, A.G., Netz, D.J.A., Miethke, M., Pierik, A.J., Burghaus, O., Peuckert, F., Lill, R., and Marahiel, M.A. (2010). SufU Is an Essential Iron-Sulfur Cluster Scaffold Protein in Bacillus subtilis. J Bacteriol 192, 1643-1651. Alen, C., and Sonenshein, A.L. (1999). Bacillus subtilis aconitase is an RNA- binding protein. P Natl Acad Sci USA 96, 10412-10417. Allocati, N., Federici, L., Masulli, M., and Di Ilio, C. (2009). Glutathione transferases in bacteria. Febs J 276, 58-75. Andrews, S.C., Robinson, A.K., and Rodriguez-Quinones, F. (2003). Bacterial iron homeostasis. FEMS microbiology reviews 27, 215-237. Angelini, S., Gerez, C., Ollagnier-de Choudens, S., Sanakis, Y., Fontecave, M., Barras, F., and Py, B. (2008). NfuA, a new factor required for maturing Fe/S proteins in Escherichia coli under oxidative stress and iron starvation conditions. J Biol Chem 283, 14084-14091. Anjem, A., and Imlay, J.A. (2012). Mononuclear Iron Enzymes Are Primary Targets of Hydrogen Peroxide Stress. J Biol Chem 287, 15544-15556. Anjem, A., Varghese, S., and Imlay, J.A. (2009). Manganese import is a key element of the OxyR response to hydrogen peroxide in Escherichia coli. Mol Microbiol 72, 844-858. Araki, K., and Inaba, K. (2012). Structure, Mechanism, and Evolution of Ero1 Family Enzymes. Antioxid Redox Sign 16, 790-799. Ayala-Castro, C., Saini, A., and Outten, F.W. (2008). Fe-s cluster assembly pathways in bacteria. Microbiol Mol Biol R 72, 110-+. Bachhawat, A.K., Thakur, A., Kaur, J., and Zulkifli, M. (2013). Glutathione transporters. Bba-Gen Subjects 1830, 3154-3164. Bandyopadhyay, S., Chandramouli, K., and Johnson, M.K. (2008a). Iron-sulfur cluster biosynthesis. Biochem Soc T 36, 1112-1119. Bandyopadhyay, S., Naik, S.G., O'Carroll, I.P., Huynh, B.H., Dean, D.R., Johnson, M.K., and Dos Santos, P.C. (2008b). A proposed role for the Azotobacter vinelandii NfuA protein as an intermediate iron-sulfur cluster carrier. J Biol Chem 283, 14092-14099. Bell, J.K., Yennawar, H.P., Wright, S.K., Thompson, J.R., Viola, R.E., and Banaszak, L.J. (2001). Structural analyses of a malate dehydrogenase with a variable active site. J Biol Chem 276, 31156-31162.

139

Black, K.A., and Dos Santos, P.C. (2014). Shared-intermediates in the biosynthesis of thio-cofactors: Mechanism and functions of cysteine desulfurases and sulfur acceptors. Biochimica et biophysica acta. Black, K.A., and Dos Santos, P.C. (2015a). An Abreviated Pathway for the Biosynthesis of 2-Thiouridine in Bacillus Subtilis. J Bacteriol. Black, K.A., and Dos Santos, P.C. (2015b). Shared-intermediates in the biosynthesis of thio-cofactors: Mechanism and functions of cysteine desulfurases and sulfur acceptors. Biochimica et biophysica acta 1853, 1470-1480. Bohannon, D.E., and Sonenshein, A.L. (1989). Positive Regulation of Glutamate Biosynthesis in Bacillus-Subtilis. J Bacteriol 171, 4718-4727. Bonner, E.R., Cahoon, R.E., Knapke, S.M., and Jez, J.M. (2005). Molecular basis of cysteine biosynthesis in plants - Structural and functional analysis of O- acetylserine sulfhydrylase from Arabidopsis thaliana. J Biol Chem 280, 38803- 38813. Bou-Abdallah, F., Adinolfi, S., Pastore, A., Laue, T.M., and Dennis Chasteen, N. (2004). Iron binding and oxidation kinetics in frataxin CyaY of Escherichia coli. J Mol Biol 341, 605-615. Bowater, L., Fairhurst, S.A., Just, V.J., and Bornemann, S. (2004). Bacillus subtilis YxaG is a novel Fe-containing quercetin 2,3-dioxygenase. Febs Lett 557, 45-48. Boyd, E.S., Thomas, K.M., Dai, Y.Y., Boyd, J.M., and Outten, F.W. (2014). Interplay between Oxygen and Fe-S Cluster Biogenesis: Insights from the Suf Pathway. Biochemistry 53, 5834-5847. Bradford, M.M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72, 248-254. Brigham, M.P., Stein, W.H., and Moore, S. (1960). The Concentrations of Cysteine and Cystine in Human Blood Plasma. The Journal of clinical investigation 39, 1633-1638. Bulaj, G., Kortemme, T., and Goldenberg, D.P. (1998). Ionization-reactivity relationships for cysteine thiols in polypeptides. Biochemistry 37, 8965-8972. Cabiscol, E., Tamarit, J., and Ros, J. (2000). Oxidative stress in bacteria and protein damage by reactive oxygen species. International microbiology : the official journal of the Spanish Society for Microbiology 3, 3-8. Carmagnol, F., Sinet, P.M., and Jerome, H. (1983). Selenium-Dependent and Non-Selenium-Dependent Glutathione Peroxidases in Human-Tissue Extracts. Biochim Biophys Acta 759, 49-57. Chai, Y.C., Ashraf, S.S., Rokutan, K., Johnston, R.B., and Thomas, J.A. (1994). S-Thiolation of Individual Human Neutrophil Proteins Including Actin by Stimulation of the Respiratory Burst - Evidence against a Role for Glutathione Disulfide. Arch Biochem Biophys 310, 273-281. Chandramouli, K., Unciuleac, M.C., Naik, S., Dean, D.R., Huynh, B.H., and Johnson, M.K. (2007). Formation and properties of [4Fe-4S] clusters on the IscU scaffold protein. Biochemistry-Us 46, 6804-6811.

140

Chandrangsu, P., Dusi, R., Hamilton, C.J., and Helmann, J.D. (2014). Methylglyoxal resistance in Bacillus subtilis: contributions of bacillithiol- dependent and independent pathways. Mol Microbiol 91, 706-715. Chesney, J.A., Eaton, J.W., and Mahoney, J.R. (1996). Bacterial glutathione: a sacrificial defense against chlorine compounds. J Bacteriol 178, 2131-2135. Chi, B.K., Busche, T., Van Laer, K., Basell, K., Becher, D., Clermont, L., Seibold, G.M., Persicke, M., Kalinowski, J., Messens, J., et al. (2014). Protein S- Mycothiolation Functions as Redox-Switch and Thiol Protection Mechanism in Corynebacterium glutamicum Under Hypochlorite Stress. Antioxid Redox Sign 20, 589-605. Chi, B.K., Gronau, K., Mader, U., Hessling, B., Becher, D., and Antelmann, H. (2011). S-Bacillithiolation Protects Against Hypochlorite Stress in Bacillus subtilis as Revealed by Transcriptomics and Redox Proteomics. Mol Cell Proteomics 10. Chi, B.K., Roberts, A.A., Huyen, T.T.T., Basell, K., Becher, D., Albrecht, D., Hamilton, C.J., and Antelmann, H. (2013a). S-Bacillithiolation Protects Conserved and Essential Proteins Against Hypochlorite Stress in Firmicutes Bacteria. Antioxid Redox Sign 18, 1273-1295. Chi, B.K., Roberts, A.A., Huyen, T.T.T., Basell, K., Becher, D., Albrecht, D., Hamilton, C.J., and Antelmann, H. (2013b). S-Bacillithiolation Protects Conserved and Essential Proteins Against Hypochlorite Stress in Firmicutes Bacteria. Antioxidants & redox signaling 18, 1273-1295. Chillappagari, S., Seubert, A., Trip, H., Kuipers, O.P., Marahiel, M.A., and Miethke, M. (2010). Copper Stress Affects Iron Homeostasis by Destabilizing Iron-Sulfur Cluster Formation in Bacillus subtilis. J Bacteriol 192, 2512-2524. Clement, G.E., and Hartz, T.P. (1971). Determination of the microscopic ionization constants of cysteine. Journal of chemical education 48, 395-397. Cooke, M.S., Evans, M.D., Dizdaroglu, M., and Lunec, J. (2003). Oxidative DNA damage: mechanisms, mutation, and disease. Faseb J 17, 1195-1214. Cooper, A.J., Pinto, J.T., and Callery, P.S. (2011). Reversible and irreversible protein glutathionylation: biological and clinical aspects. Expert opinion on drug metabolism & toxicology 7, 891-910. Cotgreave, I.A., Gerdes, R., Schuppe-Koistinen, I., and Lind, C. (2002). S- glutathionylation of glyceraldehyde-3-phosphate dehydrogenase: role of thiol oxidation and catalysis by glutaredoxin. Methods in enzymology 348, 175-182. Crack, J.C., den Hengst, C.D., Jakimowicz, P., Subramanian, S., Johnson, M.K., Buttner, M.J., Thomson, A.J., and Le Brun, N.E. (2009). Characterization of [4Fe- 4S]-containing and cluster-free forms of Streptomyces WhiD. Biochemistry 48, 12252-12264. Crawhall, J.C., and Segal, S. (1967). The intracellular ratio of cysteine and cystine in various tissues. The Biochemical journal 105, 891-896. Crook, E.M. (1941). The system dehydroascorbic acid-glutathione. The Biochemical journal 35, 226-236. Dai, Y.Y., and Outten, F.W. (2012). The E. coli SufS-SufE sulfur transfer system is more resistant to oxidative stress than IscS-IscU. Febs Lett 586, 4016-4022.

141

Dalle-Donne, I., Giustarini, D., Rossi, R., Colombo, R., and Milzani, A. (2003). Reversible S-glutathionylation of Cys 374 regulates actin filament formation by inducing structural changes in the actin molecule. Free Radic Biol Med 34, 23-32. Dalle-Donne, I., Rossi, R., Colombo, G., Giustarini, D., and Milzani, A. (2009). Protein S-glutathionylation: a regulatory device from bacteria to humans. Trends in biochemical sciences 34, 85-96. Dalle-Donne, I., Rossi, R., Giustarini, D., Colombo, R., and Milzani, A. (2007). S- glutathionylation in protein redox regulation. Free radical biology & medicine 43, 883-898. Dance, I.G. (1986). The Structural Chemistry of Metal Thiolate Complexes. Polyhedron 5, 1037-1104. Deal, W.C., Jr. (1969). Metabolic control and structure of glycolytic enzymes. IV. Nicotinamide-adenine dinucleotide dependent in vitro reversal of dissociation and possible in vivo control of yeast glyceraldehyde 3-phosphate dehydrogenase synthesis. Biochemistry-Us 8, 2795-2805. Deli, A., Koutsioulis, D., Fadouloglou, V.E., Spiliotopoulou, P., Balomenou, S., Arnaouteli, S., Tzanodaskalaki, M., Mavromatis, K., Kokkinidis, M., and Bouriotis, V. (2010). LmbE proteins from Bacillus cereus are de-N-acetylases with broad substrate specificity and are highly similar to proteins in Bacillus anthracis. Febs J 277, 2740-2753. Demasi, M., Piassa Filho, G.M., Castro, L.M., Ferreira, J.C., Rioli, V., and Ferro, E.S. (2008). Oligomerization of the cysteinyl-rich oligopeptidase EP24.15 is triggered by S-glutathionylation. Free radical biology & medicine 44, 1180-1190. Deponte, M. (2013). Glutathione catalysis and the reaction mechanisms of glutathione-dependent enzymes. Bba-Gen Subjects 1830, 3217-3266. Dowling, D.P., Gattis, S.G., Fierke, C.A., and Christianson, D.W. (2010). Structures of metal-substituted human histone deacetylase 8 provide mechanistic inferences on biological function. Biochemistry-Us 49, 5048-5056. Dyda, F., Klein, D.C., and Hickman, A.B. (2000). GCN5-related N- acetyltransferases: A structural overview. Annu Rev Bioph Biom 29, 81-103. Fadouloglou, V.E., Deli, A., Glykos, N.M., Psylinakis, E., Bouriotis, V., and Kokkinidis, M. (2007). Crystal structure of the BcZBP, a zinc-binding protein from Bacillus cereus - Functional insights from structural data. Febs J 274, 3044-3054. Fahey, R.C., Brown, W.C., Adams, W.B., and Worsham, M.B. (1978). Occurrence of Glutathione in Bacteria. J Bacteriol 133, 1126-1129. Fan, F., Luxenburger, A., Painter, G.F., and Blanchard, J.S. (2007). Steady-state and pre-steady-state kinetic analysis of Mycobacterium smegmatis cysteine ligase (MshC). Biochemistry-Us 46, 11421-11429. Fang, Z., Roberts, A.A., Weidman, K., Sharma, S.V., Claiborne, A., Hamilton, C.J., and Dos Santos, P.C. (2013). Cross-functionalities of Bacillus deacetylases involved in bacillithiol biosynthesis and bacillithiol-S-conjugate detoxification pathways. Biochemical Journal 454, 239-247. Fass, D. (2012). Disulfide Bonding in Protein Biophysics. Annu Rev Biophys 41, 63-79. Feldman-Salit, A., Wirtz, M., Lenherr, E.D., Throm, C., Hothorn, M., Scheffzek, K., Hell, R., and Wade, R.C. (2012). Allosterically gated enzyme dynamics in the

142

cysteine synthase complex regulate cysteine biosynthesis in Arabidopsis thaliana. Structure 20, 292-302. Flint, D.H., Tuminello, J.F., and Emptage, M.H. (1993). The Inactivation of Fe-S Cluster Containing Hydro-Lyases by Superoxide. J Biol Chem 268, 22369-22376. Fultz, P.N., and Kemper, J. (1981). Wild-type isopropylmalate isomerase in Salmonella typhimurium is composed of two different subunits. J Bacteriol 148, 210-219. Gaballa, A., Chi, B.K., Roberts, A.A., Becher, D., Hamilton, C.J., Antelmann, H., and Helmann, J.D. (2014). Redox regulation in Bacillus subtilis: The bacilliredoxins BrxA(YphP) and BrxB(YqiW) function in de-bacillithiolation of S- bacillithiolated OhrR and MetE. Antioxid Redox Signal 21, 357-367. Gaballa, A., Newton, G.L., Antelmann, H., Parsonage, D., Upton, H., Rawat, M., Claiborne, A., Fahey, R.C., and Helmann, J.D. (2010). Biosynthesis and functions of bacillithiol, a major low-molecular-weight thiol in Bacilli. P Natl Acad Sci USA 107, 6482-6486. Gallogly, M.M., and Mieyal, J.J. (2007). Mechanisms of reversible protein glutathionylation in redox signaling and oxidative stress. Current opinion in pharmacology 7, 381-391. Gardner, P.R., and Fridovich, I. (1992). Inactivation-reactivation of aconitase in Escherichia coli. A sensitive measure of superoxide radical. J Biol Chem 267, 8757-8763. Gardner, P.R., and Fridovich, I. (1993a). Effect of glutathione on aconitase in Escherichia coli. Archives of biochemistry and biophysics 301, 98-102. Gardner, P.R., and Fridovich, I. (1993b). Nadph Inhibits Transcription of the Escherichia-Coli Manganese Superoxide-Dismutase Gene (Soda) Invitro. J Biol Chem 268, 12958-12963. Gassmann, M., Grenacher, B., Rohde, B., and Vogel, J. (2009). Quantifying Western blots: Pitfalls of densitometry. Electrophoresis 30, 1845-1855. Gattis, S.G., Hernick, M., and Fierke, C.A. (2010). Active site metal ion in UDP-3- O-((R)-3-hydroxymyristoyl)-N-acetylglucosamine deacetylase (LpxC) switches between Fe(II) and Zn(II) depending on cellular conditions. J Biol Chem 285, 33788-33796. Gazzaniga, F., Stebbins, R., Chang, S.Z., McPeek, M.A., and Brenner, C. (2009). Microbial NAD metabolism: lessons from comparative genomics. Microbiology and molecular biology reviews : MMBR 73, 529-541, Table of Contents. Giel, J.L., Nesbit, A.D., Mettert, E.L., Fleischhacker, A.S., Wanta, B.T., and Kiley, P.J. (2013). Regulation of iron-sulphur cluster homeostasis through transcriptional control of the Isc pathway by [2Fe-2S]-IscR in Escherichia coli. Mol Microbiol 87, 478-492. Giles, N.M., Watts, A.B., Giles, G.I., Fry, F.H., Littlechild, J.A., and Jacob, C. (2003). Metal and redox modulation of cysteine protein function. Chem Biol 10, 677-693. Giustarini, D., Milzani, A., Aldini, G., Carini, M., Rossi, R., and Dalle-Donne, I. (2005). S-nitrosation versus S-glutathionylation of protein sulfhydryl groups by S- nitrosoglutathione. Antioxidants & redox signaling 7, 930-939.

143

Green, J., and Paget, M.S. (2004). Bacterial redox sensors. Nat Rev Microbiol 2, 954-966. Grek, C.L., Zhang, J., Manevich, Y., Townsend, D.M., and Tew, K.D. (2013). Causes and consequences of cysteine S-glutathionylation. The Journal of biological chemistry 288, 26497-26504. Griffith, G.R., Chandler, J.L., and Gholson, R.K. (1975). Studies on the de novo biosynthesis of NAD in Escherichia coli. The separation of the nadB gene product from the nadA gene product and its purification. Eur J Biochem 54, 239- 245. Gu, M.Z., and Imlay, J.A. (2013). Superoxide poisons mononuclear iron enzymes by causing mismetallation. Mol Microbiol 89, 123-134. Guse, A.H. (2004a). Biochemistry, biology, and pharmacology of cyclic adenosine diphosphoribose (cADPR). Current medicinal chemistry 11, 847-855. Guse, A.H. (2004b). Regulation of calcium signaling by the second messenger cyclic adenosine diphosphoribose (cADPR). Current molecular medicine 4, 239- 248. Guse, A.H. (2005). Second messenger function and the structure-activity relationship of cyclic adenosine diphosphoribose (cADPR). Febs J 272, 4590- 4597. Habig, W.H., Pabst, M.J., and Jakoby, W.B. (1974). Glutathione S-Transferases - First Enzymatic Step in Mercapturic Acid Formation. Journal of Biological Chemistry 249, 7130-7139. Halliwell, B. (1991). Reactive Oxygen Species in Living Systems - Source, Biochemistry, and Role in Human-Disease. Am J Med 91, S14-S22. Helbig, K., Grosse, C., and Nies, D.H. (2008). Cadmium toxicity in glutathione mutants of Escherichia coli. J Bacteriol 190, 5439-5454. Helmann, J.D. (2011). Bacillithiol, a New Player in Bacterial Redox Homeostasis. Antioxid Redox Sign 15, 123-133. Helmann, J.D. (2014). Specificity of Metal Sensing: Iron and Manganese Homeostasis in Bacillus subtilis. J Biol Chem 289, 28112-28120. Hernick, M. (2011). Fluorescence-based methods to assay inhibitors of lipopolysaccharide synthesis. Methods Mol Biol 739, 123-133. Hider, R.C., and Kong, X.L. (2011). Glutathione: a key component of the cytoplasmic labile iron pool. Biometals : an international journal on the role of metal ions in biology, biochemistry, and medicine 24, 1179-1187. Hinchman, C.A., Matsumoto, H., Simmons, T.W., and Ballatori, N. (1991). Intrahepatic Conversion of a Glutathione Conjugate to Its Mercapturic Acid - Metabolism of 1-Chloro-2,4-Dinitrobenzene in Isolated Perfused Rat and Guinea- Pig Livers. Journal of Biological Chemistry 266, 22179-22185. Hogg, P.J. (2003). Disulfide bonds as switches for protein function. Trends Biochem Sci 28, 210-214. Huang, C.S., Anderson, M.E., and Meister, A. (1993a). Amino-Acid-Sequence and Function of the Light Subunit of Rat-Kidney Gamma-Glutamylcysteine Synthetase. J Biol Chem 268, 20578-20583.

144

Huang, C.S., Chang, L.S., Anderson, M.E., and Meister, A. (1993b). Catalytic and regulatory properties of the heavy subunit of rat kidney gamma- glutamylcysteine synthetase. J Biol Chem 268, 19675-19680. Huang, X., and Hernick, M. (2011). A fluorescence-based assay for measuring N-acetyl-1-D-myo-inosityl-2-amino-2-deoxy-alpha-D-glucopyranoside deacetylase activity. Anal Biochem 414, 278-281. Huang, X., Kocabas, E., and Hernick, M. (2011). The activity and cofactor preferences of N-acetyl-1-D-myo-inosityl-2-amino-2-deoxy-alpha-D- glucopyranoside deacetylase (MshB) change depending on environmental conditions. J Biol Chem 286, 20275-20282. Huang, X.Y., and Hernick, M. (2012). Examination of Mechanism of N-Acetyl-1- D-myo-inosityl-2-amino-2-deoxy-alpha-D-glucopyranoside Deacetylase (MshB) Reveals Unexpected Role for Dynamic Tyrosine. Journal of Biological Chemistry 287, 10424-10434. Hulanicka, M.D., Hallquist, S.G., Kredich, N.M., and Mojica, T. (1979). Regulation of O-Acetylserine Sulfhydrylase-B by L-Cysteine in Salmonella-Typhimurium. J Bacteriol 140, 141-146. Hwang, C., Sinskey, A.J., and Lodish, H.F. (1992). Oxidized Redox State of Glutathione in the Endoplasmic-Reticulum. Science 257, 1496-1502. Imlay, J.A. (2006). Iron-sulphur clusters and the problem with oxygen. Mol Microbiol 59, 1073-1082. Inaoka, T., Matsumura, Y., and Tsuchido, T. (1999). SodA and manganese are essential for resistance to oxidative stress in growing and sporulating cells of Bacillus subtilis. J Bacteriol 181, 1939-1943. Jackman, J.E., Raetz, C.R., and Fierke, C.A. (1999). UDP-3-O-(R-3- hydroxymyristoyl)-N-acetylglucosamine deacetylase of Escherichia coli is a zinc metalloenzyme. Biochemistry-Us 38, 1902-1911. Jacob, C., Knight, I., and Winyard, P.G. (2006). Aspects of the biological redox chemistry of cysteine: from simple redox responses to sophisticated signalling pathways. Biological chemistry 387, 1385-1397. Jang, S.J., and Imlay, J.A. (2010). Hydrogen peroxide inactivates the Escherichia coli Isc iron-sulphur assembly system, and OxyR induces the Suf system to compensate. Mol Microbiol 78, 1448-1467. Johnson, D.C., Dean, D.R., Smith, A.D., and Johnson, M.K. (2005a). Structure, function, and formation of biological iron-sulfur clusters. Annu Rev Biochem 74, 247-281. Johnson, D.C., Dos Santos, P.C., and Dean, D.R. (2005b). NifU and NifS are required for the maturation of nitrogenase and cannot replace the function of isc- gene products in Azotobacter vinelandii. Biochem Soc T 33, 90-93. Jones, D.P., Mody, V.C., Carlson, J.L., Lynn, M.J., and Sternberg, P. (2002). Redox analysis of human plasma allows separation of pro-oxidant events of aging from decline in antioxidant defenses. Free Radical Bio Med 33, 1290-1300. Kahan, F.M., Kahan, J.S., Cassidy, P.J., and Kropp, H. (1974). The mechanism of action of fosfomycin (phosphonomycin). Annals of the New York Academy of Sciences 235, 364-386.

145

Kennedy, M.C., Emptage, M.H., Dreyer, J.L., and Beinert, H. (1983). The role of iron in the activation-inactivation of aconitase. J Biol Chem 258, 11098-11105. Ketley, J.N., Habig, W.H., and Jakoby, W.B. (1975). Binding of nonsubstrate ligands to the glutathione S-transferases. The Journal of biological chemistry 250, 8670-8673. Kobayashi, K., Ehrlich, S.D., Albertini, A., Amati, G., Andersen, K.K., Arnaud, M., Asai, K., Ashikaga, S., Aymerich, S., Bessieres, P., et al. (2003). Essential Bacillus subtilis genes. P Natl Acad Sci USA 100, 4678-4683. Kofoed, M.A., Wampler, D.A., Pandey, A.S., Peters, J.W., and Ensign, S.A. (2011). Roles of the Redox-Active Disulfide and Histidine Residues Forming a Catalytic Dyad in Reactions Catalyzed by 2-Ketopropyl Coenzyme M Oxidoreductase/Carboxylase. J Bacteriol 193, 4904-4913. Kouwen, T.R.H.M., Dubois, J.Y.F., Freudl, R., Quax, W.J., and van Dijl, J.M. (2008). Modulation of Thiol-Disulfide Oxidoreductases for Increased Production of Disulfide-Bond-Containing Proteins in Bacillus subtilis. Appl Environ Microb 74, 7536-7545. Kumar, C., Igbaria, A., D'autreaux, B., Planson, A.G., Junot, C., Godat, E., Bachhawat, A.K., Delaunay-Moisan, A., and Toledano, M.B. (2011). Glutathione revisited: a vital function in iron metabolism and ancillary role in thiol-redox control. Embo J 30, 2044-2056. Kuo, C.F., Mashino, T., and Fridovich, I. (1987). Alpha,Beta-Dihydroxyisovalerate Dehydratase - a Superoxide-Sensitive Enzyme. J Biol Chem 262, 4724-4727. Lacourciere, G.M., Mihara, H., Kurihara, T., Esaki, N., and Stadtman, T.C. (2000). Escherichia coli NifS-like proteins provide selenium in the pathway for the biosynthesis of selenophosphate. J Biol Chem 275, 23769-23773. Lamers, A.P., Keithly, M.E., Kim, K., Cook, P.D., Stec, D.F., Hines, K.M., Sulikowski, G.A., and Armstrong, R.N. (2012). Synthesis of Bacillithiol and the Catalytic Selectivity of FosB-Type Fosfomycin Resistance Proteins. Org Lett 14, 5207-5209. Lauhon, C.T., and Kambampati, R. (2000). The iscS gene in Escherichia coli is required for the biosynthesis of 4-thiouridine, thiamin, and NAD. J Biol Chem 275, 20096-20103. Lee, E.J., Karoonuthaisiri, N., Kim, H.S., Park, J.H., Cha, C.J., Kao, C.M., and Roe, J.H. (2005). A master regulator sigmaB governs osmotic and oxidative response as well as differentiation via a network of sigma factors in Streptomyces coelicolor. Molecular microbiology 57, 1252-1264. Lo Conte, M., and Carroll, K.S. (2013). The Redox Biochemistry of Protein Sulfenylation and Sulfinylation. J Biol Chem 288, 26480-26488. Loiseau, L., Ollagnier-de Choudens, S., Lascoux, D., Forest, E., Fontecave, M., and Barras, F. (2005). Analysis of the heteromeric CsdA-CsdE cysteine desulfurase, assisting Fe-S cluster biogenesis in Escherichia coli. J Biol Chem 280, 26760-26769. Lu, S.C. (2000). Regulation of glutathione synthesis. Curr Top Cell Regul 36, 95- 116. Lu, S.C. (2013). Glutathione synthesis. Bba-Gen Subjects 1830, 3143-3153.

146

Ma, Z., Chandrangsu, P., Helmann, T.C., Romsang, A., Gaballa, A., and Helmann, J.D. (2014). Bacillithiol is a major buffer of the labile zinc pool in Bacillus subtilis. Mol Microbiol 94, 756-770. Ma, Z.C., Zhang, H.J., and Kong, F.Z. (2004). Synthesis of two oligosaccharides, the GPI anchor glycans from S. cereveside and A. fumigatus. Carbohyd Res 339, 29-35. Macomber, L., and Imlay, J.A. (2009). The iron-sulfur clusters of dehydratases are primary intracellular targets of copper toxicity. Proc Natl Acad Sci U S A 106, 8344-8349. Magni, G., Orsomando, G., Raffelli, N., and Ruggieri, S. (2008). Enzymology of mammalian NAD metabolism in health and disease. Front Biosci-Landmrk 13, 6135-6154. Manikandan, K., Geerlof, A., Zozulya, A.V., Svergun, D.I., and Weiss, M.S. (2011). Structural studies on the enzyme complex isopropylmalate isomerase (LeuCD) from Mycobacterium tuberculosis. Proteins 79, 35-49. Maret, W., and Vallee, B.L. (1998). Thiolate ligands in metallothionein confer redox activity on zinc clusters. P Natl Acad Sci USA 95, 3478-3482. Marinoni, I., Nonnis, S., Monteferrante, C., Heathcote, P., Hartig, E., Bottger, L.H., Trautwein, A.X., Negri, A., Albertini, A.M., and Tedeschi, G. (2008). Characterization of L-aspartate oxidase and quinolinate synthase from Bacillus subtilis. Febs J 275, 5090-5107. Martinez-Ruiz, A., and Lamas, S. (2007). Signalling by NO-induced protein S- nitrosylation and S-glutathionylation: convergences and divergences. Cardiovascular research 75, 220-228. Mashruwala, A.A., Pang, Y.Y., Rosario-Cruz, Z., Chahal, H.K., Benson, M.A., Mike, L.A., Skaar, E.P., Torres, V.J., Nauseef, W.M., and Boyd, J.M. (2015). Nfu facilitates the maturation of iron-sulfur proteins and participates in virulence in Staphylococcus aureus. Molecular microbiology 95, 383-409. Maynes, J.T., Garen, C., Cherney, M.M., Newton, G., Arad, D., Av-Gay, Y., Fahey, R.C., and James, M.N. (2003). The crystal structure of 1-D-myo-inosityl 2-acetamido-2-deoxy-alpha-D-glucopyranoside deacetylase (MshB) from Mycobacterium tuberculosis reveals a zinc hydrolase with a lactate dehydrogenase fold. J Biol Chem 278, 47166-47170. McClerren, A.L., Zhou, P., Guan, Z.Q., Raetz, C.R.H., and Rudolph, J. (2005). Kinetic analysis of the zinc-dependent deacetylase in the lipid a biosynthetic pathway. Biochemistry-Us 44, 1106-1113. Meier, M., Janosik, M., Kery, V., Kraus, J.P., and Burkhard, P. (2001). Structure of human cystathionine beta-synthase: a unique pyridoxal 5 '-phosphate- dependent heme protein. Embo J 20, 3910-3916. Meister, A. (1988). Glutathione metabolism and its selective modification. J Biol Chem 263, 17205-17208. Meister, A., and Anderson, M.E. (1983). Glutathione. Annu Rev Biochem 52, 711-760. Mettert, E.L., and Kiley, P.J. (2014). Coordinate Regulation of the Suf and Isc Fe- S Cluster Biogenesis Pathways by IscR Is Essential for Viability of Escherichia coli. J Bacteriol 196, 4315-4323.

147

Mihara, H., and Esaki, N. (2002). Bacterial cysteine desulfurases: their function and mechanisms. Appl Microbiol Biot 60, 12-23. MissetSmits, M., vanOphem, P.W., Sakuda, S., and Duine, J.A. (1997). Mycothiol, 1-O-(2'-[N-acetyl-L-cysteinyl]amido-2'-deoxy-alpha-D-glucopyranosyl)- D-myo-inositol, is the factor of NAD/factor-dependent formaldehyde dehydrogenase. Febs Lett 409, 221-222. Murthy, U.M.N., Ollagnier-de-Choudens, S., Sanakis, Y., Abdel-Ghany, S.E., Rousset, C., Ye, H., Fontecave, M., Pilon-Smits, E.A.H., and Pilon, M. (2009). Characterization of Arabidopsis thaliana SufE2 and SufE3. Functions in chloroplast iron-sulfur cluster assembly and NAD synthesis (vol 282, pg 18254, 2007). J Biol Chem 284, 27020-27020. Nasu, S., Wicks, F.D., and Gholson, R.K. (1982). L-Aspartate Oxidase, a Newly Discovered Enzyme of Escherichia-Coli, Is the B Protein of Quinolinate Synthetase. J Biol Chem 257, 626-632. Newberry, K.J., Hou, Y.M., and Perona, J.J. (2002). Structural origins of amino acid selection without editing by cysteinyl-tRNA synthetase. Embo J 21, 2778- 2787. Newton, G.L., Arnold, K., Price, M.S., Sherrill, C., Delcardayre, S.B., Aharonowitz, Y., Cohen, G., Davies, J., Fahey, R.C., and Davis, C. (1996). Distribution of thiols in microorganisms: mycothiol is a major thiol in most actinomycetes. J Bacteriol 178, 1990-1995. Newton, G.L., Av-Gay, Y., and Fahey, R.C. (2000). N-acetyl-1-D-myo-inosityl-2- amino-2-deoxy-alpha-D-glucopyranoside deacetylase (MshB) is a key enzyme in mycothiol biosynthesis. J Bacteriol 182, 6958-6963. Newton, G.L., Buchmeier, N., and Fahey, R.C. (2008). Biosynthesis and functions of mycothiol, the unique protective thiol of Actinobacteria. Microbiology and molecular biology reviews : MMBR 72, 471-494. Newton, G.L., Fahey, R.C., and Rawat, M. (2012). Detoxification of toxins by bacillithiol in Staphylococcus aureus. Microbiol-Sgm 158, 1117-1126. Newton, G.L., Ko, M., Ta, P.L., Av-Gay, Y., and Fahey, R.C. (2006a). Purification and characterization of Mycobacterium tuberculosis 1D-myo-inosityl-2- acetamido-2-deoxy-alpha-D-glucopyranoside deacetylase, MshB, a mycothiol biosynthetic enzyme. Protein Expres Purif 47, 542-550. 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. Nat Chem Biol 5, 625-627. Newton, G.L., Ta, P., Bzymek, K.P., and Fahey, R.C. (2006b). Biochemistry of the initial steps of mycothiol biosynthesis. Journal of Biological Chemistry 281, 33910-33920. Newton, G.L., Unson, M.D., Anderberg, S.J., Aguilera, J.A., Oh, N.N., delCardayre, S.B., Av-Gay, Y., and Fahey, R.C. (1999). Characterization of Mycobacterium smegmatis mutants defective in 1-d-myo-inosityl-2-amino-2- deoxy-alpha-d-glucopyranoside and mycothiol biosynthesis. Biochemical and biophysical research communications 255, 239-244. Nicolas, P., Mader, U., Dervyn, E., Rochat, T., Leduc, A., Pigeonneau, N., Bidnenko, E., Marchadier, E., Hoebeke, M., Aymerich, S., et al. (2012).

148

Condition-Dependent Transcriptome Reveals High-Level Regulatory Architecture in Bacillus subtilis. Science 335, 1103-1106. Nies, D.H. (2007). Bacterial transition metal homeostasis. In Molecular microbiology of heavy metals, D.H.N.a. S.Silver, ed. (Springer-Verlag ), pp. 118- 142. Nigou, J., Dover, L.G., and Besra, G.S. (2002). Purification and biochemical characterization of Mycobacterium tuberculosis SuhB, an inositol monophosphatase involved in inositol biosynthesis. Biochemistry-Us 41, 4392- 4398. Nishio, K., and Nakai, M. (2000). Transfer of iron-sulfur cluster from NifU to apoferredoxin. J Biol Chem 275, 22615-22618. North, B.J., and Verdin, E. (2004). Sirtuins: Sir2-related NAD-dependent protein deacetylases. Genome biology 5, 224. Ollagnier-de-Choudens, S., Sanakis, Y., and Fontecave, M. (2004). SufA/IscA: reactivity studies of a class of scaffold proteins involved in [Fe-S] cluster assembly. J Biol Inorg Chem 9, 828-838. Olsen, L.R., Huang, B., Vetting, M.W., and Roderick, S.L. (2004). Structure of serine acetyltransferase in complexes with CoA and its cysteine feedback inhibitor. Biochemistry-Us 43, 6013-6019. Oppenheimer, L., Wellner, V.P., Griffith, O.W., and Meister, A. (1979). Glutathione synthetase. Purification from rat kidney and mapping of the substrate binding sites. The Journal of biological chemistry 254, 5184-5190. Outten, F.W., Djaman, O., and Storz, G. (2004). A suf operon requirement for Fe- S cluster assembly during iron starvation in Escherichia coli. Mol Microbiol 52, 861-872. Pai, E.F., and Schulz, G.E. (1983). The Catalytic Mechanism of Glutathione- Reductase as Derived from X-Ray-Diffraction Analyses of Reaction Intermediates. Journal of Biological Chemistry 258, 1752-1757. Pan, S., and Berk, B.C. (2007). Glutathiolation regulates tumor necrosis factor- alpha-induced caspase-3 cleavage and apoptosis: key role for glutaredoxin in the death pathway. Circulation research 100, 213-219. Parcell, S. (2002). Sulfur in human nutrition and applications in medicine. Alternative medicine review : a journal of clinical therapeutic 7, 22-44. Parent, A., Elduque, X., Cornu, D., Belot, L., Le Caer, J.P., Grandas, A., Toledano, M.B., and D'Autreaux, B. (2015). Mammalian frataxin directly enhances sulfur transfer of NFS1 persulfide to both ISCU and free thiols. Nature communications 6, 5686. Park, J.H., and Roe, J.H. (2008). Mycothiol regulates and is regulated by a thiol- specific antisigma factor RsrA and sigma(R) in Streptomyces coelicolor. Molecular microbiology 68, 861-870. Parsonage, D., Newton, G.L., Holder, R.C., Wallace, B.D., Paige, C., Hamilton, C.J., Dos Santos, P.C., Redinbo, M.R., Reid, S.D., and Claiborne, A. (2010). Characterization of the N-Acetyl-alpha-D-glucosaminyl L-Malate Synthase and Deacetylase Functions for Bacillithiol Biosynthesis in Bacillus anthracis. Biochemistry-Us 49, 8398-8414.

149

Patel, M.P., and Blanchard, J.S. (1999). Expression, purification, and characterization of Mycobacterium tuberculosis mycothione reductase. Biochemistry-Us 38, 11827-11833. Patel, M.P., and Blanchard, J.S. (2001). Mycobacterium tuberculosis mycothione reductase: pH dependence of the kinetic parameters and kinetic isotope effects. Biochemistry-Us 40, 5119-5126. Pedone, E., Ren, B., Ladenstein, R., Rossi, M., and Bartolucci, S. (2004). Functional properties of the protein disulfide oxidoreductase from the archaeon Pyrococcus furiosus - A member of a novel protein family related to protein disulfide-isomerase. Eur J Biochem 271, 3437-3448. Pohl, S., Tu, W.Y., Aldridge, P.D., Gillespie, C., Hahne, H., Mader, U., Read, T.D., and Harwood, C.R. (2011). Combined proteomic and transcriptomic analysis of the response of Bacillus anthracis to oxidative stress. Proteomics 11, 3036-3055. 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. Pothera, D.C., Gierok, P., Harms, M., Mostertz, J., Hochgafe, F., Antelmann, H., Hamilton, C.J., Borovok, I., Lalk, M., Aharonowitz, Y., et al. (2013). Distribution and infection-related functions of bacillithiol in Staphylococcus aureus. Int J Med Microbiol 303, 114-123. Py, B., and Barras, F. (2010). Building Fe-S proteins: bacterial strategies. Nat Rev Microbiol 8, 436-446. Rabeh, W.M., and Cook, P.F. (2004). Structure and mechanism of O- acetylserine sulfhydrylase. J Biol Chem 279, 26803-26806. Rajakovich, L.J., Tomlinson, J., and Dos Santos, P.C. (2012). Functional Analysis of Bacillus subtilis Genes Involved in the Biosynthesis of 4-Thiouridine in tRNA. J Bacteriol 194, 4933-4940. Rajkarnikar, A., Strankman, A., Duran, S., Vargas, D., Roberts, A.A., Barretto, K., Upton, H., Hamilton, C.J., and Rawat, M. (2013). Analysis of mutants disrupted in bacillithiol metabolism in Staphylococcus aureus. Biochem Bioph Res Co 436, 128-133. Rawat, M., Kovacevic, S., Billman-Jacobe, H., and Av-Gay, L.Y.K. (2003). Inactivation of mshB, a key gene in the mycothiol biosynthesis pathway in Mycobacterium smegmatis. Microbiol-Sgm 149, 1341-1349. Rawat, M., Newton, G.L., Ko, M., Martinez, G.J., Fahey, R.C., and Av-Gay, Y. (2002). Mycothiol-deficient Mycobacterium smegmatis mutants are hypersensitive to alkylating agents, free radicals, and antibiotics. Antimicrobial agents and chemotherapy 46, 3348-3355. Ray, P.D., Huang, B.W., and Tsuji, Y. (2012). Reactive oxygen species (ROS) homeostasis and redox regulation in cellular signaling. Cell Signal 24, 981-990. Reed, D.J. (1990). Review of the current status of calcium and thiols in cellular injury. Chem Res Toxicol 3, 495-502. Rey-Pailhade, J.D. (1888a). Nouvelle recherches physiologique sur la substance organique hydrogénant le soufre à froid. Comptes Rendus de l'Acaémie de Sciences 107, 43.

150

Rey-Pailhade, J.D. (1888b). Sur un corps d’origine organique hydrogénant le soufre 1 à froid. Comptes Rendus Hebdomadaire Séances de l’Académie de Sciences 16, 1683. Rigsby, R.E., Fillgrove, K.L., Beihoffer, L.A., and Armstrong, R.N. (2005). Fosfomycin resistance proteins: a nexus of glutathione transferases and epoxide hydrolases in a metalloenzyme superfamily. Methods Enzymol 401, 367-379. Rizzi, M., and Schindelin, H. (2002). Structural biology of enzymes involved in NAD and molybdenum cofactor biosynthesis. Current opinion in structural biology 12, 709-720. Robbins, A.H., Mcree, D.E., Williamson, M., Collett, S.A., Xuong, N.H., Furey, W.F., Wang, B.C., and Stout, C.D. (1991). Refined Crystal-Structure of Cd, Zn Metallothionein at 2.0 a Resolution. J Mol Biol 221, 1269-1293. Roberts, A.A., Sharma, S.V., Strankman, A.W., Duran, S.R., Rawat, M., and Hamilton, C.J. (2013). Mechanistic studies of FosB: a divalent-metal-dependent bacillithiol-S-transferase that mediates fosfomycin resistance in Staphylococcus aureus. Biochemical Journal 451, 69-79. Roche, B., Aussel, L., Ezraty, B., Mandin, P., Py, B., and Barras, F. (2013). Iron/sulfur proteins biogenesis in prokaryotes: Formation, regulation and diversity. Bba-Bioenergetics 1827, 455-469. Rodriguez-Pascual, F., Redondo-Horcajo, M., Magan-Marchal, N., Lagares, D., Martinez-Ruiz, A., Kleinert, H., and Lamas, S. (2008). Glyceraldehyde-3- phosphate dehydrogenase regulates endothelin-1 expression by a novel, redox- sensitive mechanism involving mRNA stability. Molecular and cellular biology 28, 7139-7155. Roos, G., Foloppe, N., and Messens, J. (2013). Understanding the pK(a) of redox cysteines: the key role of hydrogen bonding. Antioxidants & redox signaling 18, 94-127. Rouault, T.A. (2012). Biogenesis of iron-sulfur clusters in mammalian cells: new insights and relevance to human disease. Dis Model Mech 5, 155-164. Rouhier, N., Lemaire, S.D., and Jacquot, J.P. (2008). The role of glutathione in photosynthetic organisms: emerging functions for glutaredoxins and glutathionylation. Annual review of plant biology 59, 143-166. Sareen, D., Newton, G.L., Fahey, R.C., and Buchmeier, N.A. (2003). Mycothiol is essential for growth of Mycobacterium tuberculosis erdman. J Bacteriol 185, 6736-6740. Saunders, A.H., and Booker, S.J. (2008). Regulation of the activity of Escherichia coli quinolinate synthase by reversible disulfide-bond formation. Biochemistry-Us 47, 8467-8469. Saunders, A.H., Griffiths, A.E., Lee, K.H., Cicchillo, R.M., Tu, L., Stromberg, J.A., Krebs, C., and Booker, S.J. (2008). Characterization of quinolinate synthases from Escherichia coli, Mycobacterium tuberculosis, and Pyrococcus horikoshii indicates that [4Fe-4S] clusters are common cofactors throughout this class of enzymes. Biochemistry-Us 47, 10999-11012. Schnappinger, D., Ehrt, S., Voskuil, M.I., Liu, Y., Mangan, J.A., Monahan, I.M., Dolganov, G., Efron, B., Butcher, P.D., Nathan, C., et al. (2003). Transcriptional

151

adaptation of Mycobacterium tuberculosis within macrophages: Insights into the phagosomal environment. J Exp Med 198, 693-704. Schwartz, C.J., Djaman, O., Imlay, J.A., and Kiley, P.J. (2000). The cysteine desulfurase, IscS, has a major role in in vivo Fe-S cluster formation in Escherichia coli. P Natl Acad Sci USA 97, 9009-9014. Seelig, G.F., Simondsen, R.P., and Meister, A. (1984). Reversible dissociation of gamma-glutamylcysteine synthetase into two subunits. J Biol Chem 259, 9345- 9347. Selbach, B., Earles, E., and Dos Santos, P.C. (2010). Kinetic Analysis of the Bisubstrate Cysteine Desulfurase SufS from Bacillus subtilis. Biochemistry-Us 49, 8794-8802. Selbach, B.P., Chung, A.H., Scott, A.D., George, S.J., Cramer, S.P., and Dos Santos, P.C. (2014). Fe-S Cluster Biogenesis in Gram-Positive Bacteria: SufU Is a Zinc-Dependent Sulfur Transfer Protein. Biochemistry-Us 53, 152-160. Sen, C.K., and Packer, L. (2000). Thiol homeostasis and supplements in physical exercise. The American journal of clinical nutrition 72, 653S-669S. Sevier, C.S., and Kaiser, C.A. (2002). Formation and transfer of disulphide bonds in living cells. Nat Rev Mol Cell Bio 3, 836-847. Sharma, S.V., Arbach, M., Roberts, A.A., Macdonald, C.J., Groom, M., and Hamilton, C.J. (2013). Biophysical Features of Bacillithiol, the Glutathione Surrogate of Bacillus subtilis and other Firmicutes. Chembiochem 14, 2160-2168. Sharma, S.V., Jothivasan, V.K., Newton, G.L., Upton, H., Wakabayashi, J.I., Kane, M.G., Roberts, A.A., Rawat, M., La Clair, J.J., and Hamilton, C.J. (2011). Chemical and Chemoenzymatic syntheses of bacillithiol: a unique low-molecular- weight thiol amongst low G + C Gram-positive bacteria. Angew Chem Int Ed Engl 50, 7101-7104. Siegel, S.E.a.L. (1969). Mitochondrial L-Malate Dehydrogenase of Beef Heart Mothods in enzymology 18, 99-106. Steffek, M., Newton, G.L., Av-Gay, Y., and Fahey, R.C. (2003). Characterization of Mycobacterium tuberculosis mycothiol S-conjugate amidase. Biochemistry-Us 42, 12067-12076. Stipanuk, M.H., Coloso, R.M., Garcia, R.A.G., and Banks, M.F. (1992). Cysteine Concentration Regulates Cysteine Metabolism to Glutathione, Sulfate and Taurine in Rat Hepatocytes. J Nutr 122, 420-427. Stipanuk, M.H., Dominy, J.E., Lee, J.I., and Coloso, R.M. (2006). Mammalian cysteine metabolism: New insights into regulation of cysteine metabolism. J Nutr 136, 1652s-1659s. Sun, D., and Setlow, P. (1993). Cloning, nucleotide sequence, and regulation of the Bacillus subtilis nadB gene and a nifS-like gene, both of which are essential for NAD biosynthesis. J Bacteriol 175, 1423-1432. Thorgersen, M.P., and Downs, D.M. (2008). Analysis of yggX and gshA Mutants Provides Insights into the Labile Iron Pool in Salmonella enterica. J Bacteriol 190, 7608-7613. Townsend, D.M., and Tew, K.D. (2003). The role of glutathione-S-transferase in anti-cancer drug resistance. Oncogene 22, 7369-7375.

152

Trott, O., and Olson, A.J. (2010). AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. Journal of computational chemistry 31, 455-461. Tsai, K.J., Yoon, K.P., and Lynn, A.R. (1992). Atp-Dependent Cadmium Transport by the Cada Cadmium Resistance Determinant in Everted Membrane- Vesicles of Bacillus-Subtilis. J Bacteriol 174, 116-121. Ung, K.S.E., and Av-Gay, Y. (2006). Mycothiol-dependent mycobacterial response to oxidative stress. Febs Lett 580, 2712-2716. Upton, H., Newton, G.L., Gushiken, M., Lo, K., Holden, D., Fahey, R.C., and Rawat, M. (2012). Characterization of BshA, bacillithiol glycosyltransferase from Staphylococcus aureus and Bacillus subtilis. Febs Lett 586, 1004-1008. VanDuinen, A.J., Winchell, K.R., Keithly, M.E., and Cook, P.D. (2015). X-ray crystallographic structure of BshC, a unique enzyme involved in bacillithiol biosynthesis. Biochemistry-Us 54, 100-103. Vetting, M.W., Frantom, P.A., and Blanchard, J.S. (2008). Structural and enzymatic analysis of MshA from Corynebacterium glutamicum - Substrate- assisted catalysis. J Biol Chem 283, 15834-15844. Vetting, M.W., Roderick, S.L., Yu, M., and Blanchard, J.S. (2003). Crystal structure of mycothiol synthase (Rv0819) from Mycobacterium tuberculosis shows structural homology to the GNAT, family of N-acetyltransferases. Protein Sci 12, 1954-1959. Vetting, M.W., Yu, M., Rendle, P.M., and Blanchard, J.S. (2006). The substrate- induced conformational change of Mycobacterium tuberculosis mycothiol synthase. J Biol Chem 281, 2795-2802. Vincent E. Bower, Maya Paabo, and Bates, R.G. (1961). A Standard for the Measurement of the pH of Blood and Other Physiological Media. JOURNAL OF RESEARCH of the National Bureau of Standards-A. Physics and Chemistry 65A, 267-274. Voskuil, M.I., Schnappinger, D., Visconti, K.C., Harrell, M.I., Dolganov, G.M., Sherman, D.R., and Schoolnik, G.K. (2003). Inhibition of respiration by nitric oxide induces a Mycobacterium tuberculosis dormancy program. J Exp Med 198, 705-713. Voskuil, M.I., Visconti, K.C., and Schoolnik, G.K. (2004). Mycobacterium tuberculosis gene expression during adaptation to stationary phase and low- oxygen dormancy. Tuberculosis 84, 218-227. Vuilleumier, S. (1997). Bacterial glutathione S-transferases: What are they good for? J Bacteriol 179, 1431-1441. Winterbourn, C.C., and Metodiewa, D. (1999). Reactivity of biologically important thiol compounds with superoxide and hydrogen peroxide. Free radical biology & medicine 27, 322-328. Wirtz, M., and Hell, R. (2003). Production of cysteine for bacterial and plant biotechnology: Application of cysteine feedback-insensitive isoforms of serine acetyltransferase. Amino Acids 24, 195-203. Xu, D.P., Washburn, M.P., Sun, G.P., and Wells, W.W. (1996). Purification and characterization of a glutathione dependent dehydroascorbate reductase from human erythrocytes. Biochem Bioph Res Co 221, 117-121.

153

Xu, X., Vilcheze, C., Av-Gay, Y., Gomez-Velasco, A., and Jacobs, W.R. (2011). Precise Null Deletion Mutations of the Mycothiol Synthesis Genes Reveal Their Role in Isoniazid and Ethionamide Resistance in Mycobacterium smegmatis. Antimicrob Agents Ch 55, 3133-3139. Yeager, A.R., and Finney, N.S. (2005). Synthesis of fluorescently labeled UDP- GlcNAc analogues and their evaluation as chitin synthase substrates. J Org Chem 70, 1269-1275. Yuvaniyama, P., Agar, J.N., Cash, V.L., Johnson, M.K., and Dean, D.R. (2000). NifS-directed assembly of a transient [2Fe-2S] cluster within the NifU protein. P Natl Acad Sci USA 97, 599-604. Zaffagnini, M., Bedhomme, M., Groni, H., Marchand, C.H., Puppo, C., Gontero, B., Cassier-Chauvat, C., Decottignies, P., and Lemaire, S.D. (2012). Glutathionylation in the photosynthetic model organism Chlamydomonas reinhardtii: a proteomic survey. Molecular & cellular proteomics : MCP 11, M111 014142. Zheng, L.M., White, R.H., Cash, V.L., Jack, R.F., and Dean, D.R. (1993). Cysteine Desulfurase Activity Indicates a Role for Nifs in Metallocluster Biosynthesis. P Natl Acad Sci USA 90, 2754-2758.

154

SCHOLASTIC VITA

Zhong Fang

Education

2010-2015 PhD candidate, Department of Chemistry, Wake Forest

University, Winston Salem, NC, USA

2006-2010 B.S degree, Department of Chemistry, Northwest University,

Xi’an, PRC

Professional Positions and Experience

01/2014- 12/2014 Research assistant under grant from National Science

Foundation to investigate the interaction between bacillithiol

and Fe-S metabolism;

09/2011-09/2013 Research assistant under Multidisciplinary Research Grant

from the North Carolina Biotechnology Center to investigate

the biosynthesis and detoxification pathway of bacillithiol in B.

anthracis;

Fellowships and Awards

2014 Outstanding Senior Graduate Student Award, Department of

Chemistry, Wake Forest University

Scientific Conferences Attended

2014 The Gordon Research Conference on Iron-Sulfur Enzymes

2013 7th International Conference on Iron-Sulfur Cluster Biogenesis

and Regulation, Columbia, USC

2013 Southeast Regional Fe-S Symposium, University of Georgia

155

2012 Southeast Regional Fe-S Symposium, Virginia Tech

2012 The southeast regional meeting of the ACS

2011 Southeast Regional Fe-S Symposium, University of

South Carolina

Teaching Experience

12/14-05/15 Teaching assistant for biochemistry Lab,

Department of Chemistry, Wake Forest University

09/13-01/14 Teaching assistant for biochemistry Lab,

Department of Chemistry, Wake Forest University

09/10-09/11 Teaching assistant for general chemistry Lab,

Department of Chemistry, Wake Forest University

09/11-05/15 Undergraduate mentoring Co-mentored nine

undergraduates for an average of two years on research (six

from Wake Forest University, one from Salem College, one

from Guilford College and one from Winston Salem State

Univ.), two students graduated with honor thesis

Publications

1. Dos Santos P., Fang Z., Mason S., Setubal J., Dixon R. (2012). Distribution of nitrogen fixation and nitrogenase-like sequences amongst microbial genomes.

BMC Genomics 13:162 10.1186/1471-2164-13-162 (recognized as highly accessed)

2. Fang, Z.; Roberts, A.A.; Weidman, K.; Sharma, S.V.; Claiborne, A.; Hamilton,

C.J.; dos Santos, P.C. Cross-functionalities of Bacillus deacetylases involved in

156

bacillithiol biosynthesis and bacillithiol-S-conjugate detoxification pathways.

Biochem. J. 2013, 454, 239–247. (Selected as podcast article)

3. Fang, Z. and Dos Santos P. The protective role of bacillithiol against stress conditions in B. subtilis. Accepted on MicrobiologyOpen

4. Young C.A.; Gordon L.D., Fang. Z.; Holder R.C.; Reid S.D.; Copper tolerance and the characterization of a copper-responsive operon, copYAZ, in an M1T1 clinical strain of streptococcus pyogenes. Submitted and revised for resubmission to Journal of Bacteriology.

157